Pediatric Cancer v2 Teratoid Rhabdoid (UnitedVRG)

Pediatric Cancer
Pediatric Cancer, Volume 2

Diagnosis, Therapy, and Prognosis
For further volumes:

http://www.springer.com/series/10167
Pediatric Cancer
Volume 2
Diagnosis, Therapy, and Prognosis
Pediatric Cancer
Teratoid/Rhabdoid,
Brain Tumors, and Glioma
Edited by
M.A. Hayat
Distinguished Professor
Department of Biological Sciences,
Kean University, Union, NJ, USA
Editor
M.A. Hayat
Department of Biological Sciences
Kean University
Room 213, Library building
Morris Avenue 1000
Union, NJ 07083
USA
ISSN 2211-7997
ISSN 2211-8004 (electronic)
ISBN 978-94-007-2956-8
ISBN 978-94-007-2957-5 (eBook)
DOI 10.1007/978-94-007-2957-5
Springer Dordrecht Heidelberg New York London
Library of Congress Control Number: 2011939493
Springer Science+Business Media Dordrecht 2012
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed. Exempted from this
legal reservation are brief excerpts in connection with reviews or scholarly analysis or material
supplied specifically for the purpose of being entered and executed on a computer system, for
exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is
permitted only under the provisions of the Copyright Law of the Publishers location, in its
current version, and permission for use must always be obtained from Springer. Permissions for
use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable
to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility
for any errors or omissions that may be made. The publisher makes no warranty, express or
implied, with respect to the material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)
Although touched by technology, surgical pathology always

has been, and remains, an art. Surgical pathologists, like all
artists, depict in their artwork (surgical pathology reports)
their interactions with nature: emotions, observations, and
knowledge are all integrated. The resulting artwork is a poor
record of complex phenomena.
Richard J. Reed, MD
One Point of View
All small tumors do not keep growing, especially small breast tumors,
testicular tumors, and prostate tumors. Some small tumors may even disappear
without a treatment. Indeed, because prostate tumor grows slowly, it is not
unusual that a patient may die, at an advanced age, of some other causes, but
prostate tumor is discovered in an autopsy study. In some cases of prostate
tumors, the patient should be offered the option of active surveillance followed by PSA test or biopsies. Similarly, every small kidney tumor may not
change or may even regress. Another example of cancer or precancer reversal
is cervical cancer. Precancerous cervical cells found with Pap test, may revert
to normal cells. Tumor shrinkage, regression, reversal, or stabilization is not
impossible. The pertinent question is: Is it always necessary to practice tumor
surgery, chemotherapy, or radiotherapy? Although the conventional belief is
that cancer represents an arrow that advances unidirectionally, it is becoming clear that for cancer to progress, they require cooperative microenvironment (niche), including immune system and hormone levels. However, it is
emphasized that advanced (malignant) cancers do not show regression, and
require therapy. In the light of the inadequacy of standard treatments of
malignancy, clinical applications of the stem cell technology need to be
expedited.
Eric M.A. Hayat
vii
Preface
This is the second volume in the series, Pediatric Cancer. Specifically, it

discusses diagnosis, therapy, and prognosis of Teratoid/Rhabdoid Tumors
and Brain Tumors including Gliomas. Volume 1 contains similar information
on Neuroblastoma. It is recognized that scientific journals and books not only
provide current information but also facilitate the exchange of information,
resulting in rapid progress in the medical field. In this endeavor, the main role
of scientific books is to present current information in more detail after careful additional evaluation of the investigational results, especially those of new
or relatively new therapeutic methods and their potential toxic side-effects.
Although subjects of diagnosis, cancer recurrence including brain tumors,
resistance to chemotherapy, assessment of treatment effectiveness, including
cell therapy and side-effects of a treatment are scattered in a vast number of
journals and books, there is need of combining these subjects in single
volumes. An attempt is made to accomplish this goal in the projected
multi-volume series of Handbooks.
In the era of cost-effectiveness, my opinion may be minority perspective,
but it needs to be recognized that the potential for false-positive or falsenegative interpretation on the basis of a single laboratory test in clinical
pathology does exist. Interobserver or intraobserver variability in the interpretation of results in pathology is not uncommon. Interpretative differences
often are related to the relative importance of the criteria being used.
Generally, no test always performs perfectly. Although there is no perfect
remedy to this problem, standardized classifications with written definitions
and guidelines will help. Standardization of methods to achieve objectivity is
imperative in this effort. The validity of a test should be based on the careful,
objective interpretation of the tomographic images, photo-micrographs, and
other tests. The interpretation of the results should be explicit rather than
implicit. To achieve accurate diagnosis and correct prognosis, the use of
molecular criteria and targeted medicine is important. Equally important are
the translation of molecular genetics into clinical practice and evidence-based
therapy. Translation of medicine from the laboratory to clinical application
needs to be carefully expedited. Indeed, molecular medicine has arrived.
Brain Tumors are the most common solid tumors of childhood, and remain
the leading cause of cancer-related mortality in children. A general introduction
to the principles of diagnosis and treatment of children with brain tumors is
presented in this volume. Molecular characterization of solid tumors is important for providing novel biomarkers of disease and identifying molecular
ix
pathways, which may provide putative targets for new therapies. 1H high
resolution magic spinning NMR spectroscopy is being used to determine
metabolic profiles from small pieces of intact tissue and whole cells in culture, which is ideal for molecular characterization of childhood brain tumor
tissue and cells grown in culture. This technique is explained in this volume.
There are many differences between adult and pediatric brain tumors
beyond simple nomenclature; for example, pediatric tumors are often more
sensitive to adjuvant irradiation and chemotherapy. Some pediatric tumors
may only need complete resection to achieve a cure. It is pointed out that an
experienced neurosurgeon should be aware of the difference between the
adult tumors and pediatric tumors. It is emphasized that pediatric low-grade
gliomas need lower doses of anticancer drugs such as cisplatin/etoposide.
Refinements in clinical and molecular stratification for many types of childhood brain tumors to achieve risk-adapted treatment planning are discussed.
Many tumor suppressor genes and oncogenes directly participate in or
regulate signal transduction pathways. Alterations of these and other cell
cycle regulators play a crucial role in the development and progression of
human malignancies, including those in children. Both p53 and retinoblastoma protein pathways are discussed. The role of the Wnt pathway in pediatric CNS primitive neuroectodermal tumor and its mutational analysis and
immunohistochemistry are explained. Treatment of pediatric supratentorial
neuroectodermal tumor with chemotherapy and radiation is explained.
Platelet-derived growth factor receptor (PDGFR) plays a critical role in the
cellular proliferation, differentiation, and angiogenesis and survival of tumors
in both adults and children. The critical role played by PDGFR signaling in
oncogenic growth and survival promoting pathways in pediatric patients is
emphasized. Because the family of HER receptors modulates neurogenesis
and is connected to the biology of neuroblastic tumors in infancy, the expression and characteristics of EGFR and HER2-4 are explained.
Neurofibromatosis type-1(NF1) is a common genetic disorder with a high
prevalence of CNS abnormalities including tumors. The clinical utility of
cerebral 18F-flurodeoxyglucose positron emission tomography in children
with NF1 in determining optic pathway tumors is discussed in this volume.
The role of various immunostains in differentiating non-neoplastic brain
tumors from glioma and in subtyping glial and nonglial tumors is presented.
Individuals with NF1 are prone to the development of both benign and
malignant tumors of the CNS. The most common tumors in children with
NF1 is an optic pathway glioma, the treatment of which is detailed in this
volume.
Although pediatric optic-hypothalamic gliomas have a favorable prognosis
with regard to the long-term survival, such children may suffer from neurological deficits. Pediatric patients with high-grade gliomas have a very poor
prognosis despite a variety of aggressive therapies. An overview, including
epidemiology, etiology, treatment, and prognostic factors, of these high-grade
gliomas is presented.
Complications caused by treatments in children with leukemia are not
uncommon. Various neuroradiological imaging modalities in children with
leukemia are detailed, including their inherent strengths and weaknesses.
Preface
Preface
xi
Endoscopic neurosurgical techniques for treating pediatric intraventricular

brain tumors are explained. The efficacy of using scheduled non-narcotic
analgesic regimens following cranial and spine neurosurgery is explained.
The use of scheduled doses of alternating acetaminophen and ibuprofen following craniotomy for tumor biopsy or resection is recommended.
Neurosurigcal management of pediatric low-grade glioma, high-grade
glioma, ependymoma, and medulloblastoma is explained in detail. Also
included are results of clinical trials in pediatric brain tumors, such as medulloblastoma, ependymoma, craniopharyngioma, low-grade glioma, high-grade
glioma, brainstem glioma, and germ-cell tumors, using radiotherapy.
Paragangliomas are adrenal tumors of neural crest origin, which are genetically driven tumors. Their linkage to mutation in the subunits of the succinate dehydrogenase complex is explained, underscoring the importance of
genetic testing in patients and their family members. With increased survival
of children diagnosed with CNS tumors, there is a need of better understanding of cancer-related late adverse effects; one of these effects is cardiovascular risk, which is discussed in this volume. An overview of neurocognitive
deficits, including attention, concentration, and verbal memory, common in
pediatric cancer survivors is also presented.
Epilepsy is one of the initial presenting features of a primary brain tumor
in children. Considering clinical importance of this symptom, diagnosis and
treatments (surgery and antiepileptic drugs) of epilepsy in children are discussed in this volume.
Pediatric atypical teratoid/rhabdoid tumors are among the most common
malignant neoplasms in children; the treatments (surgery, chemotherapy, and
radiotherapy) and diagnostic imaging of these tumors are detailed. Several
preoperative imaging techniques and immunohistochemistry used to diagnose this malignancy are included. Differential diagnosis of this tumor is also
included. Present and future therapy for children with this tumor is presented.
The application of various imaging for diagnosing these tumors is included.
By bringing together a large number of experts (oncologists, neurosurgeons, physicians, research scientists, and pathologists) in various aspects
of this medical field, it is my hope that substantial progress will be made
against this terrible disease. It would be difficult for a single author to discuss
effectively the complexity of diagnosis, therapy, and prognosis of any type of
tumor in one volume. Another advantage of involving more than one author
is to present different points of view on a specific controversial aspect of the
pediatric cancer. I hope these goals will be fulfilled in this and other volumes
of this series. This volume was written by 68 contributors representing
10 countries. I am grateful to them to for their promptness in accepting
my suggestions. Their practical experience highlights their writings, which
should build and further the endeavors of the reader in this important area of
disease. I respect and appreciate the hard work and exceptional insight into
the nature of brain tumors provided by these contributors. The contents
of the volume are divided into 3 subheadings: Rhabdoid/Teratoid, Brain
Tumors (General), and Gliomas for the convenience of the readers.
It is my hope that the current volume will join the future volumes of the
series for assisting in the more complete understanding and cure of globally
Preface
xii
relevant brain malignancy. There exists a tremendous, urgent demand by the

public and the scientific community to address to cancer, diagnosis, treatment, cure, and hopefully prevention. In the light of existing cancer calamity,
government funding must give priority to eradicating this deadly childrens
malignancy over military superiority.
I am thankful to Dr. Dawood Farahi and Dr. Kristie Reilly for recognizing
the importance of medical research and publishing at an institution of higher
education
Eric M.A. Hayat
Contents
Part I Teratoid/Rhabdoid
1
Pediatric Atypical Teratoid/Rhabdoid Tumors

(An Overview) .................................................................................
Krishan Kumar Bansal and Deepak Goel
Pediatric Atypical Teratoid/Rhabdoid Tumor.............................

Korgun Koral and Daniel C. Bowers
Pediatric Rhabdomyosarcoma: Role of Cell Cycle

Regulators Alteration .....................................................................
Kenichi Kohashi, Yukiko Takahashi, Tomoaki Taguchi,
and Yoshinao Oda
Pediatric Atypical Teratoid/Rhabdoid Tumors:

Imaging with CT and MRI ............................................................
Monika Warmuth-Metz and Michael Frhwald
Pediatric Atypical Teratoid/Rhabdoid Tumor:
Diagnosis Using Imaging Techniques
and Histopathology.........................................................................
Fabrice Bing
Pediatric Atypical Teratoid/Rhabdoid Tumors:
Differential Diagnosis .....................................................................
Justin A. Bishop and Syed Z. Ali
3
13
23
31
39
53
Part II Brain Tumors (General)

7
Pediatric Brain Tumors (An Overview) .......................................

Eugene I. Hwang and Roger J. Packer
Pediatric CNS Primitive Neuroectodermal Tumor:

Role of the WNT Pathway .............................................................
Hazel A. Rogers and Richard G. Grundy
75
Neuroblastic Tumors Status and Role of HER

Family Receptors ............................................................................
Ewa Izycka-Swieszewska and Agnieszka Wozniak
89
61
xiii
Contents
xiv
10
Children with Neurofibromatosis Type 1:

Positron Emission Tomography ....................................................
Kevin London, Mahendra Moharir, Kathryn North,
and Robert Howman-Giles
99
11
Metabolite Profile Differences in Childhood Brain

Tumors: 1H Magic Angle Spinning NMR Spectroscopy............ 107
Martin Wilson and Andrew Peet
12
Central Nervous System Imaging in Childhood Leukemia ....... 117

Luciana Porto and Heinrich Lanfermann
13
Immunohistochemistry in the Differential Diagnosis

of Adult and Pediatric Brain Tumors ........................................... 129
Aditya Raghunathan
14
Children with Brain Tumors: Role of the Neurosurgeon ........... 143

Peter F. Morgenstern and Mark M. Souweidane
15
Pediatric Intraventricular Brain Tumors: Endoscopic

Neurosurgical Techniques.............................................................. 155
David I. Sandberg and Faiz Ahmad
16
Neurosurgical Management of Pediatric Brain Tumors ............ 165

Mehdi Shahideh, George M. Ibrahim, and James T. Rutka
17
Pediatric Brain Tumor Biopsy or Resection:

Use of Postoperative Nonnarcotic Analgesic
Medication ....................................................................................... 179
R. Shane Tubbs, Martin M. Mortazavi,
and Aaron A. Cohen-Gadol
18
Clinical Trials in Pediatric Brain Tumors; Radiotherapy.......... 183

Anna Skowronska-Gardas, Marzanna Chojnacka,
and Katarzyna Pedziwiatr
19
Epileptic Seizures and Supratentorial Brain

Tumors in Children ........................................................................ 199
Roberto Gaggero, Alessandro Consales, Francesca Fazzini,
Maria Luisa Garr, and Pasquale Striano
20
Postoperative Pain in Children: Advantage of Using

Nonnarcotic Analgesic Regimen ................................................... 207
21
Pediatric Brain Tumors: Application of Stratification

Criteria to Refine Patient Management ....................................... 211
Ian F. Pollack
22
Pediatric Supratentorial Primitive Neuroectodermal Tumor:

Treatment with Chemotherapy and Radiation ........................... 223
Donna L. Johnston and Daniel L. Keene
Contents
xv
23
Pediatric Cancer Survivors: Neurocognitive Late Effects ......... 229

Sarah Hile, Erica Montague, Bonnie Carlson-Green,
Paul Colte, Leanne Embry, and Robert D. Annett
24
Adult Survivors of Pediatric Cancer:

Risk of Cardiovascular Disease ..................................................... 247
Eric Chow and Lillian Meacham
Part III Gliomas

25
Pediatric Glioma: Role of Platelet-Derived Growth

Factor Receptor .............................................................................. 259
Tobey J. MacDonald
26
An Overview of Pediatric High-Grade Gliomas

and Diffuse Intrinsic Pontine Gliomas ......................................... 269
Rishi R. Lulla and Jason Fangusaro
27
Pediatric Low-Grade Glioma: The Role

of Neurofibromatosis-1 in Guiding Therapy ............................... 285
Robert Listernick and David H. Gutmann
28
Treatment of Pediatric Optic-Hypothalamic Gliomas:

Prognosis.......................................................................................... 295
Luca Massimi
29
Pediatric Low-Grade Gliomas: Advantage of Using

Lower Doses of Cisplatin/Etoposide ............................................. 309
Maura Massimino, Veronica Biassoni, and Elisabetta Schiavello
30
Pediatric Paragangliomas: Role of Germline

Mutation in Succinate Dehydrogenase ......................................... 321
Pinki K. Prasad and Elizabeth Yang
Index......................................................................................................... 333
Contents of Volume 1
Introduction
Pediatric Cns Neuroblastoma: Magnetic Resonance Imaging
and Spectroscopy
Pediatric Neuroblastoma-Associated Opsoclonus-Myoclonus-Ataxia
Syndrome: Early Diagnosis
Neuroblastoma Mouse Model
Orbital Metastasis in Neuroblastoma Patients
Pediatric Neuroblastoma: Molecular Detection of Minimal
Residual Disease
A Comprehensive Tissue Microarray-Based Fish Screen
of Alk Gene in Neuroblastomas
Neuroblastoma: Triptolide Therapy
Neuroblastoma: Ornithine Decarboxylase and Polyamines
are Novel Targets for Therapeutic Intervention
Neuroblastoma: Antibody-Based Immunotherapy
Targeting Multidrug Resistance in Neuroblastoma
Neuroblastoma: Perspectives for the Use of Il-21 in Immunotherapy
Neuroblastoma: Role of Hypoxia and Hypoxia Inducible Factors
in Tumor Progression
Neuroblastoma: Role of Gata Transcription Factors
Neuroblastoma: Role of Mycn/Bmil Pathway in Neuroblastoma
Neuroblastoma: Role of Clusterin as a Tumor Suppressor Gene
Refractory NeuRoblastoma Cells: Statins Target Atp Binding
Cassette-Transporters
Neuroblastoma: Dosimetry for Mibg Therapies
Advanced Neuroblastoma: Role of Alk Mutations
xvii
xviii
Pediatric Neuroblastoma: Treatment With Oral Irinotecan

and Temozolomide
Genomic Profiling of Neuroblastoma Tumors- Prognostic
Impact of Genomic Aberrations
Neuroblastoma Patients: Plasma Growth Factor Mildkine
as a Prognostic Growth Factor
Pediatric Neuroblastoma: Role of Tgfbi (Keratoepithelin)
Role of Bone Marrow InfiltratioN Detected by Sensitive Methods
in Patients with Localized Neuroblastoma
Contents of Volume 1
Contributors
Faiz Ahmad Department of Neurological Surgery, University of Miami

Miller School of Medicine and Miami Childrens Hospital, Miami, FL, USA
Syed Z. Ali Department of Pathology and Radiology, The Johns Hopkins
Hospital, Baltimore, MD, USA
Robert D. Annett Department of Pediatrics and Psychology, University of
New Mexico Health Sciences Center, Albuquerque, NM, USA
Krishan Kumar Bansal Department of Neurosurgery, Himalayan Institute
of Medical Sciences, Dehradun, Uttarakhand, India
Veronica Biassoni Fondazione IRCCS Istituto Nazionale Tumori, Milan,
Italy
Fabrice Bing Neuroradiology unit, University Hospital of Grenoble,
Grenoble Cedex 09, France
Justin A. Bishop Department of Pathology and Radiology, The Johns
Hopkins Hospital, Baltimore, MD, USA
Daniel C. Bowers Department of Pediatric, University of Texas
Southwestern Medical Center at Dallas and Childrens Medical Center,
Dallas, TX, USA
Bonnie Carlson-Green Pediatric Neuropsychologist, Childrens Hospitals
and Clinics of Minnesota, Psychology Services, St. Paul, MN, USA
Marzanna Chojnacka Department of Radiotherapy, M. Sklodowska-Curie
Memorial Cancer Centre, Warsaw, Poland
Eric Chow Pediatric Hematology-Oncology, Seattle Childrens Hospital,
Fred Hutchinson Cancer Research Center, Seattle, WA, USA
Aaron A. Cohen-Gadol Pediatric Neurosurgery, Children Hospital,
Birmingham, AL, USA
Paul Colte Division of Hematology/Oncology/BMT, Primary Childrens
Medical Center, Salt Lake City, UT, USA
Alessandro Consales Department of Neurosciences, Gaslini Childrens
Hospital, Genoa, Genoa, Italy
xix
xx
Leanne Embry Pediatric Hematology/Oncology, University of Texas Health

Science Center at San Antonio, San Antonio, TX, USA
Jason Fangusaro Hematology/Oncology/Stem Cell Transplantation, Soares
Lab, Cancer Biology and Epigenomics, Childrens Memorial Hospital,
Chicago, IL, USA
Francesca Fazzini Department of Neurosciences, Gaslini Childrens
Michael Frhwald Abteilung fur Neuroradiologie der Universitat Wurzburg,
Wurzburg, Germany
Roberto Gaggero Department of Neurosciences, Gaslini Childrens
Maria Luisa Garr Department of Neurosciences, Gaslini Childrens
Deepak Goel Department of Neurosurgery, Himalayan Institute of Medical
Sciences, Dehradun, Uttarakhand, India
Richard G. Grundy Department of Brain Tumour Research Centre, Queens
Medical Centre, University of Nottingham, Nottingham, UK
David H. Gutmann Department of Neurology, Washington University
School of Medicine, St. Louis, MO, USA
Sarah Hile Psychology Department, University of New Mexico,
Albuquerque, NM, USA
Robert Howman-Giles Department of Nuclear Medicine, Childrens
Hospital at Westmead, Sydney, NSW, Australia
Eugene I. Hwang Center for Cancer and Blood Disorders, Childrens
National Medical Center, Washington, DC, USA
George M. Ibrahim Division of Neurosurgery, The Hospital for Sick
Children, Toronto, ON, Canada
Ewa Izycka-Swieszewska Department of Nursing Management and
Pathomorphology, Medical University of Gdansk, Gdansk, Poland
Donna L. Johnston Division of Hematology/Oncology, Childrens Hospital
of Eastern Ontario, Ottawa, ON, Canada
Daniel L. Keene Division of Hematology/Oncology, Childrens Hospital of
Eastern Ontario, Ottawa, ON, Canada
Kenichi Kohashi Department of Anatomic Pathology, Pathological Sciences,
Graduate School of Medical Sciences, Kyushu University, Higashi-ku,
Fukuoka, Japan
Korgun Koral Department of Radiology, University of Texas Southwestern
Medical Center at Dallas and Childrens Medical Center, Dallas, TX, USA
Contributors
Contributors
xxi
Heinrich Lanfermann Institut fr Neuroradiologie, Klinikum der Johann

Wolfgang Goethe-Universitt, Frankfurt am Main, Germany
Robert Listernick Department of Neurology, Washington University School
of Medicine, St. Louis, MO, USA
Kevin London Department of Nuclear Medicine, Childrens Hospital at
Westmead, Sydney, NSW, Australia
Rishi R. Lulla Hematology/Oncology/Stem Cell Transplantation, Childrens
Memorial Hospital, Chicago, IL, USA
Tobey J. MacDonald Childrens Healthcare of Atlanta, Aflac Cancer Center
and Blood Disorders Service, Emory Childrens Center, Emory University,
Atlanta, GA, USA
Luca Massimi Institute of Neurosurgery-Pediatric Neurosurgery, A. Gemelli
Hospital, Rome, Italy
Maura Massimino Fondazione IRCCS Istituto Nazionale Tumori,
Milan, Italy
Lillian Meacham Hematology/Oncology and Endocrinology, Emory
University, Atlanta, GA, USA
Mahendra Moharir Department of Nuclear Medicine, Childrens Hospital
at Westmead, Sydney, NSW, Australia
Erica Montague Psychology Department, University of New Mexico,
Albuquerque, NM, USA
Peter F. Morgenstern Departments of Neurological Surgery and Pediatrics,
Wil Cornell Medical College and Memorial Sloan-Kettering Cancer Center,
New York, NY, USA
Martin M. Mortazavi Pediatric
Birmingham, AL, USA
Neurosurgery,
Children
Hospital,
Kathryn North Department of Nuclear Medicine, Childrens Hospital at

Westmead, Sydney, NSW, Australia
Yoshinao Oda Department of Anatomic Pathology, Pathological Sciences,
Graduate School of Medical Sciences, Kyushu University, Higashi-ku,
Fukuoka, Japan
Roger J. Packer Center for Cancer and Blood Disorders, Childrens National
Medical Center, Washington, DC, USA
Katarzyna Pedziwiatr Department of Radiotherapy, M. Sklodowska-Curie
Andrew Peet Cancer Sciences, University of Birmingham, Birmingham, UK
Ian F. Pollack Department of Neurosurgery, Childrens Hospital of Pittsburgh,
Pittsburgh, PA, USA
xxii
Luciana Porto Institut fu r Neuroradiologie, Klinikum der Johann Wolfgang

Goethe-Universita t, Frankfurt am Main, Germany
Pinki K. Prasad Department of Pediatric, Vanderbilt University School of
Medicine, Nashville, TN, USA
Aditya Raghunathan The Methodist Hospital & University of Texas M.D.
Anderson Cancer Center, Houston, TX, USA
Hazel A. Rogers Department of Brain Tumour Research Centre, Queens
Medical Centre, University of Nottingham, Nottingham, UK
James T. Rutka Division of Neurosurgery, The Hospital for Sick Children,
Toronto, ON, Canada
David I. Sandberg Department of Neurological Surgery, University of
Miami Miller School of Medicine and Miami Childrens Hospital, Miami,
FL, USA
Elisabetta Schiavello Fondazione IRCCS Istituto Nazionale Tumori, Milan,
Italy
Mehdi Shahideh Division of Neurosurgery, The Hospital for Sick Children,
Toronto, ON, Canada
Anna Skowronska-Gardas Department of Radiotherapy, M. SklodowskaCurie Memorial Cancer Centre, Warsaw, Poland
Mark M. Souweidane Departments of Neurological Surgery and Pediatrics,
Wil Cornell Medical College and Memorial Sloan-Kettering Cancer Center,
New York, NY, USA
Pasquale Striano Department of Neurosciences, Gaslini Childrens Hospital,
Genoa, Genoa, Italy
Tomoaki Taguchi Department of Pediatric Surgery, Graduate School of
Medical Sciences, Kyushu University, Higashi-ku, Fukuoka, Japan
Yukiko Takahashi Department of Pediatric Surgery, Graduate School
of Medical Sciences, Kyushu University, Higashi-ku, Fukuoka, Japan
R. Shane Tubbs Pediatric Neurosurgery, Children Hospital, Birmingham,
AL, USA
Monika Warmuth-Metz Abteilung fur Neuroradiologie der Universitat
Wrzburg, Wrzburg, Germany
Martin Wilson Cancer Sciences, University of Birmingham, Birmingham,
UK
Agnieszka Wozniak Laboratory of Experimental Oncology and Department
of General Medical Oncology, KU Leuven and University Hospitals,
Herestraat, Leuven, Belgium
Elizabeth Yang Department of Pediatric, Vanderbilt University School of
Medicine, Nashville, TN, USA
Contributors
Part I
Teratoid/Rhabdoid
Pediatric Atypical Teratoid/

Rhabdoid Tumors (An Overview)
Krishan Kumar Bansal and Deepak Goel
Contents
Abstract
Introduction ............................................................
Epidemiology ..........................................................
Clinical Presentation of CNS AT/RT ....................
Imaging of AT/RT ..................................................
Gross and Microscopic Pathology

of AT/RT..................................................................
Molecular Pathology of AT/RT .............................
Treatment of AT/RT: Surgery ...............................
Treatment of AT/RT: Chemotherapy ...................
Treatment of AT/RT: Radiotherapy .....................
References ...............................................................
Pediatric Atypical teratoid/rhabdoid tumors

(AT/RT) of the central nervous system (CNS)
are among the most malignant neoplasms and
most often diagnosed in children smaller than
3 years of age and incidence is 12% of all
brain tumors in children. Sixty-three percent
of the AT/RT of the CNS is seen in infra-tentorial compartment, there are no precise imaging features that differentiate AT/RT from the
other posterior fossa tumor. The rhabdoid
cells are characteristic on cytopathology. It
has been established now that CNS, AT/RT
often shows deletion of the long arm of chromosome 22q11.2. The initial treatment for
most children with AT/RT is surgical with and
without cerebrospinal fluid diversionary procedure. Children with less than 3 years of age
offered chemotherapy but in older children
radiotherapy is given in addition.
Introduction
K.K. Bansal (*) D. Goel

Department of Neurosurgery,
Himalayan Institute of Medical Sciences,
Dehradun, Uttarakhand, India
e-mail: kbansalk@yahoo.com
Pediatric Atypical teratoid/rhabdoid tumors

(AT/RT) of central nervous system (CNS) are
rare and very aggressive malignant lesion of early
childhood. Because of both the infrequency and
rapid course of disease, consensus has not been
made for the standard treatment so far (Rorke
et al. 1996; Strother 2005; Hilden et al. 1998;
Biegel et al. 1990; Gandhi et al. 2004).
M.A. Hayat (ed.), Pediatric Cancer, Volume 2,

DOI 10.1007/978-94-007-2957-5_1, Springer Science+Business Media Dordrecht 2012
Beckwith and Palmer In 1978 first described a

histological variant of Wilms tumor that occurred
primarily in infants and was correlated with
extremely poor prognosis. It was subsequently
called malignant rhabdoid tumor meant for the
reason that the tumor looked like a rhabdomyosarcoma, but the cells did not demonstrate usual
morphological or immuno-histo-chemical features of muscle (Haas et al. 1981). A CNS tumor
composed of rhabdoid cells was first reported in
1985 by Briner et al. but the unique clinical and
pathological features were not well defined until
19951996 (Rorke et al. 1996).
Since approximately 70% of these tumors
contain
fields
indistinguishable
from
Medulloblastoma or primitive neuroectodermal
tumor, pathologists by and large gave one or the
other diagnosis. The histological diagnosis is not
easy, as there may be considerable microscopic
overlap with these embryonal tumors (Biegel
et al. 2000, 2002). However, study of these tumors
with high index of suspicion even in routine H
and E stains disclosed fields of rhabdoid cells
with or without areas of primitive neuroepithelial
cells and in a quarter to a third of samples, mesenchymal and/or epithelial elements were seen as
well. Thus, even though such a combination of
divergent tissue types suggested that these tumors
were teratomas, although they lacked the standard features essential for such a diagnosis.
The diagnostic term that seemed most suitable
was AT/RT and so it was coined (Rorke et al. 1995,
1996; Bhattacharjee et al. 1997). The histogenesis
of this curious and highly malignant tumor of the
early childhood has remained unidentified (Packer
et al. 2002; Burger et al. 1998; Fisher et al. 1996).
Regarding effective therapy of these tumors till
date no standard protocol has been setup and overall survival even after multidisciplinary efforts like
surgery, radiation and chemotherapy has not
improved significantly (Rorke et al. 1996; Tekautz
et al. 2005; Hilden et al. 2004).
Epidemiology
AT/RT of the CNS most frequently occurs in
infants or neonates, the majority of patients
diagnosed being younger than 3 years of age
K.K. Bansal and D. Goel
(Rorke et al. 1996; Gyure 2005) although it is not

often seen in older childrens as well, 70% vs.
30% (Rorke et al. 1996; Tekautz et al. 2005). The
mean age at diagnosis ranges from 17 to
29 months (Rorke et al. 1996; Burger et al. 1998;
Gandhi et al. 2004; Gyure 2005). These tumors
are somewhat more frequent in boys [34: 12,
male: female ratio] in younger than 3 years age
group but in childrens older than 3 year the ratio
is not consistent (Tekautz et al. 2005; Hilden
et al. 2004). Their incidence is 12% of all brain
tumors in children, while some investigators
report that 6.7% of CNS tumors in infants
02 years were AT/RT (Roberts et al. 2000; Wong
2005; Rickert and Paulus 2001).
Most common location is infratentorial
[6070%] in the cerebellum or CP angle and rest
in supratentorial, spinal or multifocal. ATRT of
CNS has recently shown some age specific site
preference, posterior fossa is the most common
site in younger than 3 but in older than 3 preferred location for the development of this tumor
is supratentorial (Rorke et al. 1995, 1996;
Bhattacharjee et al. 1997; Tekautz et al. 2005;
Gyure 2005). AT/RT has been reported in an inutero infant, a pregnant female and also in a
patient of neurofibromatosis-1; in both of the
later tumor was supratentorial (Erickson et al.
2005; Kababri et al. 2006).
Prognosis in patients younger than 3 years is
very grim if compared with older then 3-year
children. Moreover, younger patients are more
likely to present with metastatic disease at diagnosis and tend to develop progression with higher
frequency and earlier in the course of treatment
than older childrens (Rorke et al. 1995; Tekautz
et al. 2005). The results of Hilden et al. suggested
that older children diagnosed with AT/RT have
better survival (Hilden et al. 2004).
Clinical Presentation of CNS AT/RT

Since rhabdoid tumors were originally found in
the kidney, such tumors have been described in
many different organs and soft tissues, as well as
in the CNS. Sixty-three percent of the AT/RT of
the CNS is seen in infra-tentorial compartment,
rest arises in supratentorial [27%] or 8% may be
Pediatric Atypical Teratoid/Rhabdoid Tumors (An Overview)
multifocal (Rorke et al. 1996; Bhattacharjee et al.

1997; Burger et al. 1998; Wong 2005). While the
cerebellum and cerebral hemispheres are the
most common locations, these tumors have a predilection for the cerebellopontine angle (Rorke
et al. 1996). They may also arise in the spinal
cord, pineal gland and supra-sellar region (Burger
et al. 1998; Wong 2005). Posterior fossa is a
favorite location in children younger than 3 years,
as opposed to older children; the few examples in
adults are almost exclusively in the cerebrum
(Rorke et al. 1996; Wong 2005). A small group of
children have both renal and CNS rhabdoid
tumors which most likely represent metachronous tumors, possibly due to a germ-line mutation in the hSNF5 gene.
Infants whose cranial sutures have not yet
fused tend to present with non-specific symptoms
such as macro-cephalic, lethargy, vomiting and/
or failure to thrive (Gyure 2005). Older children
present with head tilts [IV nerve palsy], diplopia
[VI nerve palsy], facial weakness [VII nerve
palsy], headache and/or hemiplegia. The majority of children with posterior fossa AT/RT have
hydrocephalus at presentation due to obstruction
of the cerebrospinal fluid (CSF) flow at the fourth
ventricle. The set of clinical signs and symptoms
in children with AT/RT is similar to those children with PNET/medulloblastoma [PNET/MB]
or any other tumor in posterior fossa.
One-third of children with AT/RT present with
Leptomeningeal spread of tumor at diagnosis, a
rate similar to that seen in children with PNET/
MB (Gyure 2005). There was no noticeable difference in age of the patient for metastatic disease and those who do not, but recent studies
showing that dissemination of the disease is early
and with higher frequency in patients younger
than 3-year (Tekautz et al. 2005). Examination of
the CSF at the time of diagnosis revealed malignant cells in one third of the patients and CSF
may be positive even when cranio-spinal imaging
is negative (Burger et al. 1998).
Fig. 1.1 Sagittal T1 precontrast. Heterogenous mass

appearing to arise from fourth ventricle. High signal likely
represents hemorrhage (methemoglobin) or calcification
Imaging of AT/RT
The imaging procedure of choice in children with
AT/RT is a cranio-spinal MRI with and without
gadolinium (Figs. 1.1 and 1.2). The tumor shows
Fig. 1.2 Sagittal T1 postcontrast. Partially wellcircumscribed mass extending from the foramen magnum
to the superior vermis, contrast enhancement superiorly
and anteriorly. Mild tonsilar herniation Hydrocephalus
persists
low signal intensity on T1 weighted images and

isointense or decreased signal on T2 weighted
images (Rorke et al. 1996). Cysts and hemorrhages are commonly seen. These tumors can be
of heterogeneous intensity with heterogeneous
enhancement with peripheral cystic components
(Cheng et al. 2005). Obstructive hydrocephalus
and peri-ventricular lucency may be seen, especially with tumors located in the posterior fossa
that block the fourth ventricle or its outlet foramina. The main tumor mass enhances in-homogeneously after administration of gadolinium.
Leptomeningeal spread appears as diffuse
enhancement of the meninges and/or enhancing
clumps along the spinal cord and cauda equina as
drop metastasis. All of these features are similar
to those seen in PNET/MB and, in fact, there are
no precise imaging features that differentiate AT/
RT from the PNET/MB. Some children may
undergo CT scanning as part of their diagnostic
workup. As with PNET/MB, the AT/RT appears
as a hyperdense lesion on an unenhanced CT
scan, presumably due to the high cellularity of
the tumor (Rorke et al. 1996).
Gross and Microscopic Pathology

of AT/RT
Macroscopic features of these tumors differ in
no way from those of PNET/MB. They are soft,
pinkish-red, necrotic and/or hemorrhagic. Those
with a prominent mesenchymal component may
be firm and contain tan-white foci. Tumors primarily in the cerebello-pontine angle may
incorporate cranial nerve roots in the vicinity.
Leptomeningeal deposits display no specific
distinguishing features and are basically similar
to PNET/MB. On section, these tumors tend to
infiltrate and have poorly-demarcated margins.
Microscopic characteristics of AT/RT are variable, although it is self-evident that they must
contain rhabdoid cells (Fig. 1.3). Some tumors
consist of only this cell type, whereas more commonly there is a mixture of rhabdoid fields and
areas indistinguishable from classical PNET/MB
(Rorke et al. 1996; Bhattacharjee et al. 1997; Oka
and Scheithauer 1999). Although this portion
Fig. 1.3 Microscopic histopathologic (H & E) photograph showing rhabdoid cells
may rarely contain Flexner-Wintersteiner or

Homer Wright rosettes, neither desmoplastic nor
the nodularneuroblastic histological types have
been observed. Basically, the PNET/MB portions
simply consist of small primitive appearing neuroepithelial cells (Parwani et al. 2005).
The typical rhabdoid cell is of medium to large
size and consists of an eccentric nucleus adjacent
to which is eosinophilic cytoplasm equal to or
larger than the size of the nucleus (Gyure 2005).
This tends to be round or slightly bulbous and
may have a faint pink rim accentuating a denser
pink core. Many nuclei contain a prominent
nucleolus. Mitotic figures are frequent. The rhabdoid cells may range from small to large size,
pale cells may sometimes-containing two nuclei,
in a jumbled architectural arrangement (Burger
et al. 1998). The small cell component resembled
Medulloblastoma and rarely had cords of cells in
a mucinous background, imitating chordoma.
The cytoplasm of the larger cells is prominent
with a somewhat rhabdoid appearance,
although rhabdoid features were not-always
prominent (Rorke et al. 1995).
Rhabdoid and PNET fields tend to remain
separate, as do the epithelial and mesenchymal
components, although there are no sharply delineated margins. A recognizable epithelial component, which may be adenomatous or squamous,
occurs in about a quarter of the tumors, although
a much higher number of tumor cells express epithelial antigens. A small group of these tumors
may mimic Choroid plexus carcinoma; hence
this possibility should be kept in mind. In addition,

about one-third of tumors contain neoplastic
mesenchymal elements, which, in the extreme,
mimic sarcomas (Rorke et al. 1996). These
tumors often exhibit large areas of necrosis, mitoses and hemorrhage, but intrinsic vasculature
generally manifests no distinctive features
(Bhattacharjee et al. 1997).
Certainty in making a specific histological
diagnosis of AT/RT may be improved by studying the tumor with a panel of monoclonal antibodies. Most helpful are the following: epithelial
membrane antigen (EMA), vimentin (V), smooth
muscle actin (SMA), keratin (K), glial fibrillary
acidic protein (GFAP) and neurofilament protein
(NFP) (Bhattacharjee et al. 1997; Burger et al.
1998; Gyure 2005). Desmin is rarely expressed
by the neuroepithelial cells, but not by rhabdoid
cells. Markers for germ cells are consistently
negative (Rorke et al. 1996).
The pattern of expression of these antigens is
complex; hence attention must be paid to which
specific cellular component is expressing the
antigen. The rhabdoid cells typically express
vimentin and EMA, but SMA less frequently
(Gyure 2005). They may also express K, GFAP
and/or NFP. The neuroepithelial cells express
only GFAP and/or NFP, whereas the epithelial
component expresses K plus or minus EMA; the
mesenchymal cells typically express vimentin
and SMA (Rorke et al. 1996).
Ultra structural findings vary, depending upon
sampling. The classical, but not pathognomonic
finding in the rhabdoid cell consists of large bundles of intermediate filaments in the cytoplasm.
For cytological study of these tumors materials can be obtained from smear scraping, squash
preparation or fine needle aspiration (Parwani
et al. 2005). Cyto-morphological study of the
tumor shows hyper-cellularity, primarily large
tissue fragments with tumor cells adjacent to
growing capillaries illustrating a papillary-like
appearance and characteristic rhabdoid cells
i.e., intermediate-sized cells with granular or
fibrillary, brilliantly eosinophilic cytoplasm
with or without globoid inclusions; large,
eccentrically located, single nucleoli, small,
round, primitive neuronal-appearing cells
with a high nuclear to cytoplasmic ratio,

speckled chromatin and atypical, occasionally
multinucleated giant cells. In addition, few
apoptotic bodies, mitoses, considerable necrosis
and prominent dystrophic calcification may be
found (Parwani et al. 2005).
Molecular Pathology of AT/RT

It has been established now that CNS, AT/RT
often show deletion of the long arm of chromosome 22q11.2, further molecular studies have led
to the identification of a rhabdoid suppressor
gene [INI1/hSNF5] at said location (Hilden et al.
2004; Versteege et al. 1998; Biegel 1997).
Somatic mutations in this gene predispose children to develop AT/RT. Earlier, this was often the
only karyotypic change seen in this tumor and it
was thought that a tumor suppressor gene was
contained to that region. Furthermore, it was suggested that loss of one copy of chromosome 22q
could distinguish an AT/RT from a PNET/MB,
which is frequently associated with loss of 17p/
isochromosome 17q. Later, Versteege et al.
(1998) identified deletions and truncating mutations of the hSNF5 gene on chromosome 22q11
in a series of cell lines derived from renal rhabdoid tumors.
The hSNF5 protein is highly conserved and
is not greatly changed between flies, mice and
humans. The hSNF5 protein is the smallest
member of a family of proteins that form a
complex, which regulates the DNA through
changes in the nucleosome. By winding and
unwinding DNA, this complex changes the
configuration of genomic DNA, thus allowing
or denying transcription factors access to the
DNA and changing gene expression patterns.
Biegel et al. (1999) subsequently identified
somatic mutations of hSNF5 in a series of CNS
AT/RT.
Some children with AT/RT are born with
heterozygous germ-line mutations of the hSNF5
gene, suggesting that these children were predisposed to develop AT/RT. In most cases, these
germ line mutations are de novo (a new mutation,
not inherited from the parents), but in some
instances, they may be inherited from phenotypically normal parents (Biegel et al. 1999; Taylor
et al. 2000). Individuals and families with
germ-line mutations of hSNF5 are also at
increased risk to develop carcinoma of the choroid plexus. However, it remains to be determined
whether these are true choroid plexus tumors or
AT/RT which may sometimes be misdiagnosed
as a choroid plexus carcinoma.
Heterozygous mSNF5 knockout mice
develop tumors resembling AT/RT, supporting the
role of hSNF5 as a tumor suppressor gene.
Although most AT/RTs show evidence of some
genetic derangement at the hSNF5 locus, mutational analysis of the hSNF5 gene in a series of
PNET/MBs discovered mutations in only 4/52
tumors (Biegel et al. 2000). Of those 4, 2/4 was
re-classified, as AT/RT on re-examination of the
pathology, but there was insufficient clinical material to establish an accurate diagnosis in the other
two cases. This suggests that tumors diagnosed as
PNET/MB with hSNF5 are most likely AT/RT.
Such confusion is not surprising, given the large
number of AT/RTs, which contain fields indistinguishable from PNET/MB. While mutation/deletion of hSNF5 is not currently sufficient for a
diagnosis of AT/RT, it appears to be related to the
clinical outcome and hence, searching for it is
becoming part of the diagnostic work-up. Overexpression of osteopontin gene has been reported
as a potential diagnostic marker for atypical teratoid/rhabdoid tumor (Kao et al. 2005). In one
study, Alpha-internexin expression is seen in the
atypical teratoid/rhabdoid tumors, indicate that
these primitive tumors usually exhibit neuronal
differentiation (Kao et al. 2003).
Treatment of AT/RT: Surgery

The initial treatment for most children with AT/
RT is surgical. Children presenting in extremis
with severe hydrocephalus require a cerebrospinal fluid (CSF) diversionary procedure, either a
ventriculostomy, a ventriculo-peritoneal shunt or,
more recently, an endoscopic third ventriculostomy (Sainte-Rose et al. 2001).Most children
undergo a craniotomy, with maximal safe resec-
tion of tumor. The interface of the AT/RT and

cerebellum may be abrupt or infiltrative and ill
defined (Burger et al. 1998). Total or near total,
resection of the tumor is feasible in about 50% of
patients. While surgery is excellent for reducing
the mass effect, children who receive surgery
alone with no adjuvant therapy typically die
within 1 month after surgery (Rorke et al. 1996).
There is no high quality, prospective data on
the value of surgical resection in the management
of AT/RT, but in patients with PNET/MB,
progression-free survival in children without disseminated disease at diagnosis is 20% better if
the amount of residual tumor post-operatively is
less than 1.5 cm3 compared to children where the
amount of residual tumor was greater than 1.5 cm3
(Albright et al. 1996). Gross total resection is feasible in 64% of patients with younger than 3 year
while its possible in 78% children more than
3-years of age. Eighty-one percent of patients
younger than 3-year at diagnosis develop recurrent disease within 3 months after surgery and
recurrence is mainly local in more than 70% of
patients (Tekautz et al. 2005).
Treatment of AT/RT: Chemotherapy

Most children with AT/RT receive chemotherapy
at some point during their clinical course, especially those less than 2 years of age, in view to
delay radiation therapy. Several different chemotherapeutic regimens have been tried, including
baby Pediatric Oncology Group (POG) protocols, eight drugs in 1 day, single agent cyclophosphamide and single agent ifosphamide (Rorke
et al. 1996). Most of these regimens were chosen
based on their efficacy in treating PNET/MB.
However, patients with AT/RT respond poorly to
chemotherapy and only 6/33 children who
received chemotherapy alone after surgery or
chemotherapy prior to radiation had a response
as defined by greater than 50% reduction in tumor
mass. In addition, most responses were shortlived, the longest being 10 months (Rorke et al.
1996). Some AT/RTs have been documented to
progress during the course of chemotherapy
(Burger et al. 1998).
In contrast to the older children, recurrent or

progressive AT/RT in children 3 years or younger
appears refractory to chemotherapy (Tekautz
et al. 2005). Two children treated with high-dose
chemotherapy, followed by autologous bone
marrow transplant had a good response, with
one child surviving 19 months and another alive
and well at 46 months of follow-up (Hilden et al.
1998). There is one report where chemotherapy
previously described for use in patients with
parameningeal rhabdomyosarcoma, was administered to three patients with AT/RT. Therapy
included surgery, radiotherapy, chemotherapy
and triple intra-thecal chemotherapy. All three
patients were reported to be alive and well, with
no evidence of disease at 5 years, 2 years and
9 months, respectively (Olson et al. 1995). This
exciting result awaits confirmation in larger,
prospective trials. In one retrospective study,
median survival with chemotherapy in younger
than 3-year is 0.3 year and with Radiation in
older than 3 was 0.4 year and median survival is
0.6 year in those who received both chemotherapy and radiation (Tekautz et al. 2005). Hilden
et al. (1998) reported event free survival of
16 months in children older than 3 in comparison
to only 7.7 months in younger than 3 year,
their patients treated with surgery and chemotherapy only.
Treatment of AT/RT: Radiotherapy

Most children diagnosed with AT/RT are usually
less than 2 years of age, so, because of toxicity of
radiation to young brains, radiotherapy is initially
not offered. Currently, the goal is to continue
with chemotherapy until the child is at least 2 or
3 years of age, at which time, radiation effects are
less severe. As children with AT/RT commonly
present with Leptomeningeal spread or else
develop it at the time of relapse, it is desirable to
administer cranio-spinal radiotherapy, in addition
to treatment of the primary tumor. Some authors
have advocated a boost of radiation to the primary tumor by conventional means or by stereotactic radiosurgery at the time cranio-spinal
therapy are administered. However, radiotherapy
does not seem to alter the progression of disease

in children with AT/RT; indeed, an objective
response to radiotherapy was obtained in only
2/10 patients (Burger et al. 1998).
Recently gamma knife has also been used; but
there are only two reports of the use of gamma
knife performed in patients with AT/RT that
resulted in local control of the post-operative
lesion (Bambakidis et al. 2002; Hirth et al. 2003).
In recent studies radiotherapy has shown promising results with prolonged survival of older children and adults with ATRT and it appears most
effective if administered early in the course of
treatment, though the unacceptable sequelae of
cranial radiation in infants and youngs preclude
its use (Packer et al. 2002). In spite of these
inconsistent statistics, children who are between
2 and 3 years of age and older will receive radiotherapy at some point in the course of their
disease.
In conclusion the prognosis for children with
AT/RT is bleak. The median time to progression
is 4.5 months and the median reported survivals
range from 6 to 11 months (Burger et al. 1998;
Tekautz et al. 2005). Currently, the longest
reported surviving patient was a 3-year-old girl
who survived 5.5 years after presenting with a
thalamic tumor that was treated with craniospinal
irradiation. At relapse, the disease may be local
(31%), in the leptomeninges (11%) or both
(58%). Rorke et al. (1996) had found at postmortem examination, that 10/11 children with
AT/RT had widespread Leptomeningeal metastatic disease. Obviously, current treatments for
this tumor in the form of surgery, chemotherapy
and/or radiation are not sufficient. Identification
and characterization of the rhabdoid tumor
predisposition gene on chromosome 22q may
allow development of more focused, effective
therapeutic agents which may increase survival time.
References
Albright AL, Wisoff JH, Zeltzer PM, Boyett JM, Rorke
LB, Stanley P (1996) Effects of medulloblastoma
resections on outcome in children: a report from the
Childrens Cancer Group. Neurosurgery 38:265271
10
Bambakidis NC, Robinson S, Cohen M, Cohen AR (2002)
Atypical teratoid/rhabdoid tumors of the central nervous system: clinical, radiographic and pathologic features. Pediatr Neurosurg 37:6470
Beckwith JB, Palmer NF (1978) Histopathology and
prognosis of Wilms tumors: results from the First
National Wilms Tumor Study. Cancer 41:19371948
Bhattacharjee M, Hicks J, Langford L, Dauser R, Strother
D, Chintagumpala M (1997) Central nervous system
atypical teratoid/rhabdoid tumors of infancy and childhood. Ultrastruct Pathol 21:369378
Biegel JA (1997) Genetics of pediatric central nervous
system tumors. J Pediatr Hematol Oncol 19:492501
Biegel JA, Rorke LB, Packer RJ, Emanuel BS (1990)
Monosomy 22 in rhabdoid or atypical tumors of the
brain. J Neurosurg 73:710714
Biegel JA, Zhou JY, Rorke LB, Stenstrom C, Wainwright
LM, Fogelgren B (1999) Germ-line and acquired
mutations of INI1 in atypical teratoid and rhabdoid
tumors. Cancer Res 59:7479
Biegel JA, Fogelgren B, Zhou JY, James CD, Janss AJ,
Allen JC, Zagzag D, Raffel C, Rorke LB (2000)
Mutations of the INI1 rhabdoid tumor suppressor gene
in medulloblastomas and primitive neuroectodermal
tumors of the central nervous system. Clin Cancer Res
6:27592763
Biegel JA, Tan L, Zhang F, Wainwright L, Russo P, Rorke
LB (2002) Alterations of the hSNF5/INI1 gene in central nervous system atypical teratoid/rhabdoid tumors
and renal and extra-renal rhabdoid tumors. Clin Cancer
Res 8:34613467
Briner J, Bannwart F, Kleihues P (1985) Malignant small
cell tumor of the brain with intermediate filaments a
case of primary cerebral rhabdoid tumor. Pediatr
Pathol 3:117118
Burger PC, Yu IT, Tihan T, Friedman HS, Strother DR,
Kepner JL, Duffner PK, Kun LE, Perlman EJ (1998)
Atypical teratoid/rhabdoid tumor of the central nervous
system: a highly malignant tumor of infancy and childhood frequently mistaken for medulloblastoma: a Pediatric
Oncology Group study. Am J Surg Pathol 22:10831092
Cheng YC, Lirng JF, Chang FC, Guo WY, Teng MM,
Chang CY, Wong TT, Ho DM (2005) Neuroradiological
findings in atypical teratoid/rhabdoid tumor of the
central nervous system. Acta Radiol 46:8996
Erickson ML, Johnson R, Bannykh SI, de Lotbiniere A,
Kim JH (2005) Malignant rhabdoid tumor in a pregnant adult female: literature review of central nervous
system rhabdoid tumors. J Neurooncol 74:311319
Fisher BJ, Siddiqui J, Macdonald D, Cairney AE, Ramsey
D, Munoz D, Del Maestro R (1996) Malignant rhabdoid tumor of the brain: an aggressive clinical entity.
Can J Neurol Sci 23:257263
Gandhi CD, Krieger MD, McComb JG (2004) Atypical
teratoid/rhabdoid tumor: an unusual presentation.
Neuroradiology 46:834837
Gyure KA (2005) Newly defined central nervous system
neoplasms. Am J Clin Pathol 123:S3S12

Haas JE, Palmer NF, Weinberg AG, Beckwith JB (1981)
Ultrastructure of malignant rhabdoid tumor of the
kidney. A distinctive renal tumor of children. Hum
Pathol 12:646657
Hilden JM, Watterson J, Longee DC, Moertel CL, Dunn
ME, Kurtzberg J, Scheithauer BW (1998) Central
nervous system atypical teratoid tumor/rhabdoid
tumor: response to intensive therapy and review of the
literature. J Neurooncol 40:265275
Hilden JM, Meerbaum S, Burger P, Finlay J, Janss A,
Scheithauer BW, Walter AW, Rorke LB, Biegel JA
(2004) Central nervous system atypical teratoid/
rhabdoid tumor: results of therapy in children enrolled
in a registry. J Clin Oncol 22:28772884
Hirth A, Pedersen PH, Baardsen R, Larsen JL, Krossnes
BK, Helgestad J (2003) Gamma-knife radiosurgery in
paediatric cerebral and skull base tumors. Med Pediatr
Oncol 40:99103
Kababri ME, Andre N, Carole C, Lena G, FigarellaBranger D, Gentet JC (2006) Atypical teratoid rhabdoid tumor in a child with neurofibromatosis-1. Pediatr
Blood Cancer 46:267268
Kao CL, Chiou SH, Chen YJ, Singh S, Lin HT, Liu RS,
Lo CW, Yang CC, Chi CW, Lee CH, Kaya B, Mena H,
Miettinen M, Rushing EJ (2003) Alpha-internexin
expression in medulloblastoma and atypical teratoidrhabdoid tumors. Clin Neuropathol 22:215221
Kao CL, Chiou SH, Chen YJ, Singh S, Lin HT, Liu RS,
Lo CW, Yang CC, Chi CW, Lee CH, Wong TT (2005)
Increased expression of osteopontin gene in atypical
teratoid/rhabdoid tumor of the central nervous system.
Mod Pathol 18:769778
Oka H, Scheithauer BW (1999) Clinicopathological characteristics of atypical teratoid/rhabdoid tumor. Neurol
Med Chir (Tokyo) 39:510518
Olson TA, Bayar E, Kosnik E, Hamoudi AB, Klopfenstein
KJ, Pieters RS, Ruymann FB (1995) Successful
treatment of disseminated central nervous system
malignant rhabdoid tumor. J Pediatr Hematol Oncol
17:7175
Packer RJ, Biegel JA, Blaney S, Finlay J, Geyer JR,
Heideman R, Hilden J, Janss AJ, Kun L, Vezina G,
Rorke LB, Smith M (2002) Atypical teratoid/rhabdoid
tumor of the central nervous system: report on workshop. J Pediatr Hematol Oncol 24:337342
Parwani AV, Stelow EB, Pambuccian SE, Burger PC, Ali
SZ (2005) Atypical teratoid/rhabdoid tumor of the
brain: cytopathologic characteristics and differential
diagnosis. Cancer 105:6570
Rickert CH, Paulus W (2001) Epidemiology of central
nervous system tumors in childhood and adolescence
based on new WHO classification. Childs Nerv Syst
17:503511
Roberts CW, Galusha SA, McMenabin ME, Fletcher CD,
Orkin SH (2000) Haploinsufficiency of Snf5 (integrase interactor 1) predisposes to malignant rhabdoid
tumors in mice. Proc Natl Acad Sci USA
97:1379613800
Rorke LB, Packer R, Biegel J (1995) Central nervous

system atypical teratoid/rhabdoid tumors of infancy
and childhood. J Neurooncol 24:2128
Rorke LB, Packer RJ, Biegel JA (1996) Central nervous
and childhood: definition of an entity. J Neurosurg
85:5665
Sainte-Rose C, Cinalli G, Roux FE, Maixner R, Chumas
PD, Mansour M, Carpentier A, Bourgeois M, Zerah
M, Pierre-Kahn A, Renier D (2001) Management of
hydrocephalus in pediatric patients with posterior
fossa tumors: the role of endoscopic third ventriculostomy. J Neurosurg 95:791797
Strother D (2005) Atypical teratoid rhabdoid tumors of
childhood: diagnosis, treatment and challenges. Expert
Rev Anticancer Ther 5:907915
Taylor MD, Gokgoz N, Andrulis IL, Mainprize TG,
Drake JM, Rutka JT (2000) Familial posterior fossa
11
brain tumors of infancy secondary to germ-line

mutation of the hSNF5 gene. Am J Hum Genet
66:14031406
Tekautz TM, Fuller CE, Blaney S, Fouladi M, Bronicer A,
Merchant TE, Krasin M, Dalton J, Hale G, Kun LE,
Wallace D, Gilbertson RJ, Gujjar A (2005) Atypical
teratoid/rhabdoid tumors (AT/RT): improved survival
in children 3 years of age and older with radiation
therapy and high dose alkylator based chemotherapy.
J Clin Oncol 23:14911499
Versteege I, Sevenet N, Lange J, Rousseau-Merck
MF, Ambros P, Handgretinger R, Aurias A,
Delattre O (1998) Truncating mutations of hSNF5/
INI1 in aggressive paediatric cancer. Nature
394:203206
Wong TT (2005) Increased expression of osteopontin
gene in atypical teratoid/rhabdoid tumor of the central
nervous system. Mod Pathol 18:769778

Rhabdoid Tumor
Korgun Koral and Daniel C. Bowers
Contents
Abstract
Identification of Atypical Teratoid/

Rhabdoid Tumors and Early
Observational Studies ............................................
13
Molecular Biology of Atypical

Teratoid/Rhabdoid Tumors ...................................
15
Imaging of Atypical Teratoid/

Rhabdoid Tumors ..................................................
Preoperative Imaging ...............................................
Postoperative Imaging..............................................
15
15
17
Treatment of Atypical Teratoid/

Rhabdoid Tumors ..................................................
17
Present and Future Therapy

for Children with Atypical
Teratoid/Rhabdoid Tumors ...................................
18
Conclusions .............................................................
19
References ...............................................................
19
Atypical teratoid/rhabdoid tumor of the

central nervous system is a relatively recently
described malignant neoplasm afflicting predominantly infants and young children. This
aggressive neoplasm shares many common
characteristics on routine histopathology and
imaging with the primitive neuroectodermal
tumor/medulloblastoma which is the most
common malignant central nervous system
neoplasm seen in children. Differentiation is
made by immunohistochemical analysis. The
outcome in atypical teratoid/rhabdoid tumor
remains worse than medulloblastoma, despite
recent improvements. In this manuscript clinical and radiological features of atypical teratoid/rhabdoid tumor were reviewed. Patient
outcomes and current treatment options were
discussed.
Identication of Atypical Teratoid/

Rhabdoid Tumors and Early
Observational Studies
K. Koral (*)
Department of Radiology, University of Texas
Southwestern Medical Center at Dallas and Childrens
Medical Center, Dallas, TX, USA
e-mail: korgun.koral@utsouthwestern.edu
D.C. Bowers
Department of Pediatrics, University of Texas
Southwestern Medical Center at Dallas and Childrens
Medical Center, Dallas, TX, USA
Rhabdoid tumors are rare tumors that usually

occur among children during the first 2 years
of life. When rhabdoid tumors are located in
the central nervous system, they are known as
Atypical Teratoid/Rhabdoid Tumors (AT/RTs),
but they can also arise within the kidney (where
they are known as Rhabdoid Tumor of the

13
14
Kidney or RTK) and soft tissues (extra-renal

rhabdoid tumor). On light microscopy, AT/RTs
contain classic rhabdoid cells, including large
nuclei containing a single prominent nucleolus
and cytoplasm with distinct pale eosinophilic
inclusions. In addition, they also often contain
areas of primitive neuroepithelial, mesenchymal, and epithelial components. Prior to two
decades ago, these tumors were often misdiagnosed as medulloblastoma/primitive neuroectodermal tumors (MB/PNETs), choroid plexus
carcinomas or germ cell tumors. But because
of increasing awareness of AT/RTs and more
accurate methods of diagnosis, these tumors
are increasingly being recognized and diagnosed. In contrast to MB/PNETs, they are more
likely to be in the supratentorial compartment
than the infratentorial compartment. Also,
there is a higher frequency of disseminated disease at diagnosis than MB/PNETs (Hilden
et al. 2004).
A report from Bonnin et al. (1984) described
7 children aged 1 day to 17 months who had
embryonal tumors arising in both the brain and
kidney. Six of 7 kidney tumors were recognized
as rhabdoid tumors of the kidney (RTKs); the
brain tumors included 3 cerebellar medulloblastomas, 1 pineoblastoma, 1 primitive neuroepithelial tumor, 1 malignant subependymal
giant cell astrocytoma, and 1 cerebellar
medulloepithelioma with divergent glial and
neuronal differentiation. Six of 7 children died
less than 3 months after diagnosis. The remaining 12-month-old child was treated with chemotherapy (CCNU and vincristine) and
craniospinal radiation therapy in addition to
surgical resection of his tumors and was
reported to be alive 3 years after diagnosis.
Increased intracranial pressure was the first presenting sign in 3 patients and an abdominal mass
was the presenting feature of the other 4 children. The authors stated that it was unlikely that
the second tumor was a metastasis; nevertheless, they also noted that the relationship
between these dissimilar, embryologically unrelated tumors remains enigmatic. In retrospect,
it is most plausible that these children all had
synchronous AT/RTs and RTKs.
K. Koral and D.C. Bowers
Biggs et al. (1987) described autopsy findings of a 3-month-old infant who was found to
have a widely disseminated malignant rhabdoid
tumor. The authors recognized that malignant
rhabdoid tumors had not yet been described as a
primary tumor of the central nervous system.
Lefkowitz et al. (1987) identified and further
defined AT/RTs as unique and distinct from MB/
PNETs. A manuscript by Rorke et al. (1996)
described 52 children with AT/RTs from
Childrens Hospital of Philadelphia and
Childrens Cancer Group clinical trials. They
recognized that presenting symptoms and signs,
imaging studies and pathology of AT/RTs were
often very similar to MB/PNETs, but that the
immunohistochemical profile was unique: epithelial membrane antigen, vimentin, and smoothmuscle actin were positive in the majority of AT/
RTs and markers for germ-cell tumors were consistently negative. Also, abnormalities of chromosome 22, including monosomy 22, were often
detected in AT/RTs (Biegel et al. 1990). Finally,
although treatment regimens were variable, the
survival after diagnosis for these children was
very limited. Another study from Burger et al.
(1998) from the Pediatric Oncology Group
(POG) described 55 children with AT/RTs who
were enrolled on POG clinical trials. This report
further described the immunohistochemical
findings of AT/RTs, confirmed that mutations in
chromosome 22 were often present, and reported
a mean survival of 11 months after diagnosis.
The Childrens Cancer Group protocol #9921
utilized up-front chemotherapy and delayed
radiation therapy for infants less than 3 years of
age with malignant brain tumors (Geyer et al.
2005). The study included 28 children with AT/
RTs. The 5 year event free survival (EFS) and
overall survival (OS) rates for children with AT/
RTs were 14% 7% and 29% 9%, respectively.
No factor was significantly associated with
prognosis in this small subgroup of patients. By
the mid-1990s, AT/RTs were recognized as a
new, distinct form of CNS embryonal tumor
which had a substantially worse outcome than
MB/PNETs. Also, it was recognized that the
standard infant brain tumor protocols, which
utilized relatively modest chemotherapy and
Pediatric Atypical Teratoid/Rhabdoid Tumor
delayed radiation therapy were insufficient to

cause prolonged tumor responses in the majority
of children with AT/RTs.
15
Imaging of Atypical Teratoid/

Rhabdoid Tumors
Preoperative Imaging
Molecular Biology of Atypical

Teratoid/Rhabdoid Tumors
Versteege et al. (1998) examined malignant
rhabdoid tumors and identified mutations in the
hSNF5/INI1 gene, located on chromosome 22,
which encodes a component of the SWI/SNF
multiprotein chromatin-remodeling complex.
These investigators proposed that the hSNF5/
INI1 gene functioned as a tumor suppressor
gene. Likewise, Biegel et al. (1999) confirmed
hSNF5/INI1 gene mutations in germline and
tumor DNA from children with rhabdoid tumors
and proposed that hSNF5/INI1 is a tumor suppressor gene involved in the pathogenesis of AT/
RTs, RTKs and extra-renal rhabdoid tumors. As
a tumor suppressor gene, the identification of
germline hSNF5/INI1 mutations was recognized
as a possible mechanism to explain synchronous
tumors in both the brain and kidney (Jackson
et al. 2007). A subsequent report from Biegel
et al. (2002) identified specific mutations in
the hSNF5/INI1 gene that were associated
with tumors located in the brain, kidney and
elsewhere.
The hSNF5/INI1 gene appears to function as a
classic tumor suppressor gene, where germ line
deletions and mutations predispose the child to
the development of these malignancies, and
somatic loss or mutation of the other allele
constitutes the second hit. Inactivation of both
copies of the gene leads to loss of protein expression in the nucleus, which can be detected by
immunohistochemistry. Subsequently, a primary
mouse monoclonal antibody, BAF47, against the
hSNF5/INI1 gene product has been identified as
highly sensitive and specific in the diagnosis of
AT/RTs and other rhabdoid tumors (Judkins et al.
2004). The genetic nomenclature committee
bestowed the name SMARCB1 (SWI/SNF-related,
Matrix-associated, Actin-dependent Regulator of
Chromatin, subfamily B, member 1) as the new
moniker for the hSNF5/INI1 gene.
The first radiologic descriptions of AT/RTs

appeared in the literature in the early 1990s
(Hanna et al. 1993; Munoz et al. 1995). In several
case reports and case series computerized tomography (CT) and magnetic resonance imaging
(MRI) findings of AT/RTs were described
(Zuccoli et al. 1999; Arslanoglu et al. 2004; Lee
et al. 2004). Although AT/RTs are generally
tumors of infants and young children, they can be
encountered in older children and adults (Takei
et al. 2009). AT/RTs commonly involve the cerebellum and cerebrum, however, no specific central
nervous system location appears immune to primary involvement by this tumor. AT/RTs of the
optic nerves, pineal region, spinal cord (Verma and
Morriss 2008; Niwa et al. 2009; Takei et al. 2009)
have been described. A report of primary leptomeningeal involvement by AT/RT has been published (El-Nabbout et al. 2009).
Computerized Tomography
The role of computerized tomography (CT) is
limited to the detection of the tumor and diagnosis of hydrocephalus in AT/RTs. For adequate
characterization of the tumor and assessment of
distant leptomeningeal spread and drop metastases magnetic resonance imaging (MRI) is necessary. On CT, AT/RTs are generally hyperdense
owing to the increased nuclear/cytoplasmic ratio
of the tightly packed tumor cells and paucity of
extracellular matrix (Parmar et al. 2006). The
lesions may be hemorrhagic or cystic. Punctate
calcifications may be present (Packer et al. 2002;
Parmar et al. 2006; Koral et al. 2008). In a report
of 4 AT/RT patients, all tumors showed small foci
of calcifications on CT (Arslanoglu et al. 2004).
Cheng et al. (2005) reported their experience
in 20 patients with AT/RT. Preoperative CT was
available in 7 patients. Calcifications were present in 2 tumors. In a report by Parmar et al. (2006)
all of the 11 patients with AT/RT had preoperative CT available for review. Six tumors were
16
in the posterior fossa and 5 were in the supratentorial compartment. On CT, approximately 36%
(4/11) of the tumors showed calcifications. Eightytwo percent (9/11) of tumors were either completely or partially hyperdense, due to
hypercellularity of the tumor. In the 7 patients
described by Koral et al. (2008) all tumors were
either completely or partially hyperdense. Two of
the 7 (29%) lesions had punctate calcifications.
Magnetic Resonance Imaging

The magnetic resonance imaging (MRI) publications (Howlett et al. 1997; Zuccoli et al. 1999;
Evans et al. 2001) emphasized heterogenous
appearance of AT/RTs. Arslanoglu et al. (2004)
reported preoperative MRI findings of AT/RTs in
4 patients. Three of the patients had tumor in the
supratentorial compartment. All tumors were
heterogenous and contained cystic components.
The solid components showed enhancement with
intravenous gadolinium. In the report by Cheng
et al. (2005) 7 patients had preoperative MRI of
the brain. Four patients had their tumor in the
posterior fossa, 1 in the temporal lobe, 1 in a lateral ventricle and 1 in the spinal cord. The tumors
were described as heterogenous and enhancing
with intravenous gadolinium. In the report by
Parmar et al. (2006) 7 patients had preoperative
MRI available. The solid components showed
marked enhancement in 6 and moderate
enhancement in 1 tumor. Koral et al. (2008)
reported preoperative imaging findings of 19 AT/
RTs and compared with 36 medulloblastomas.
Medulloblastoma is the primary differential diagnosis of AT/RT in the posterior fossa. One observation they made was the earlier presentation of
patients with AT/RT in comparison with the
medulloblastoma patients: 1.32 years versus
6.52 years. The outcomes for AT/RT were far
worse than the outcome for medulloblastoma.
Survival at 1 year was 27.9% for AT/RT whereas
it was 86.9% for medulloblastoma. Consistent
with other reports (Cheng et al. 2005; Parmar
et al. 2006), approximately half of the AT/RTs
were infratentorial (58%, 11/19) (Fig. 2.1a).
There were 3 patients whose tumors were both
infratentorial and supratentorial. The authors
found that hydrocephalus was less common at
Fig. 2.1 (a) Axial T2 weighted image through the posterior fossa shows a large heterogenous mass occupying the
right cerebellar hemisphere (arrows). cb cerebellum. (b)
Apparent diffusion coefficient (ADC) map shows the
solid component of the tumor (arrows) is more hypointense than normal cerebellum (cb), indicating hypercellularity compared to normal brain parenchyma
presentation for infratentorial AT/RTs compared

to medulloblastomas. This was thought to be due
to the proclivity of AT/RTs to involve the cerebellopontine angle cistern in the posterior fossa
(73%, 8/11) compared to medulloblastomas
which are overwhelmingly midline tumors.
Intratumoral hemorrhage, described as foci of T1
shortening on precontrast images was present in
approximately half of all AT/RTs and was more
common than it was in medulloblastomas.
Koral et al. (2008) described the apparent diffusion coefficient (ADC) values in 6 AT/RTs. The
mean ADC value of AT/RTs was 0.55 0.06 103
(Fig. 2.1b). ADC is a measure of random motion
of water molecules. ADC is reduced in situations
where there are more boundaries impeding the
random movement of water molecules in a given
volume. AT/RTs consist of tightly packed cells
with little extracellular matrix compared to normal
brain parenchyma and other common pediatric
brain tumors, such pilocytic astrocytoma and
ependymoma. Rumboldt et al. (2006) provided
ADC values of 2 AT/RTs: 0.55 and 0.60 103 in
their publication on utility of ADC values in distinguishing pediatric cerebellar tumors. Koral et al.
(2008) reported the mean ADC value of 14
medulloblastomas as 0.47 0.16 103 and they
concluded that ADC values of medulloblastomas
and AT/RTs were not significantly different. The
ADC values of AT/RT and medulloblastoma were
markedly lower than normal brain parenchyma
and other common pediatric cerebellar tumors,
namely pilocytic astrocytomas and ependymomas
(Rumboldt et al. 2006; Koral et al. 2008).
Due to the friable nature of cells in AT/RTs
similar to medulloblastomas, there is a tendency
for intracranial leptomeningeal tumor spread and
drop metastasis (Koeller and Rushing 2003). In
the report of Parmar et al. (2006) of the 11 patients
reported, 7 had information with regards to
intracranial and spinal metastases. Three patients
had tumor dissemination both intracranially and
into the thecal sac. Two patients had spinal drop
metastases only and 2 patients did not have spinal
or brain metastases. Of the 11 patients in whom
imaging findings were reported, Cheng et al.
(2005) reported that 9 of them had leptomeningeal
spread. In this publication it was not reported
17
whether leptomeningeal spread was intracranial

or spinal. Koral et al. (2008) reported the frequency of intracranial leptomeningeal metastases
for AT/RT as 10.5% at presentation. This was not
significantly different than that of medulloblastomas. Spinal drop metastases in AT/RT were seen
in 26.7% of patients at presentation, this figure
was also not significantly different than that of
medulloblastomas (Koral et al. 2008). Screening
of spine with MRI prior to surgery is mandatory
when a cerebellar mass is identified, because following surgery there is always some subdural and
sometimes subarachnoidal fluid and/or blood in
the spinal canal making determination of drop
metastases very difficult, if not impossible. The
spinal postoperative fluid collections usually
resolve after 2 weeks, and if not performed earlier, spine imaging may be performed thereafter.
Postoperative Imaging
Following resection of an AT/RT an early postoperative MRI is obtained within 24 h. This is done
not only to assess the extent of resection, but also
to evaluate operative complications. If a gross
total resection was achieved, follow-up MR
imaging is performed every 3 months. In addition
to conventional precontrast and postcontrast
sequences, we find postcontrast FLAIR (fluid
attenuated inversion recovery) sequence very
useful in identification of primary or subsequent
leptomeningeal tumor spread. FLAIR sequence
nulls the extracellular fluid signal and normal
cortical vascular enhancement is minimal with
this technique. Therefore, when there is abnormal leptomeningeal enhancement, the enhancing
tumor is much easier to detect against a hypointense background.
Treatment of Atypical Teratoid/

Rhabdoid Tumors
The early reports of children with AT/RTs
described a dismal prognosis with often encountered early death due to tumor and very few
long-term survivors. A case series by Olson et al.
18
(1995) described 3 children with newly diagnosed

AT/RT who had prolonged progression-free survival following treatment with surgery, radiation
therapy, chemotherapy and intrathecal chemotherapy. The treatment regimen was based upon a
clinical trial for children with rhabdomyosarcoma
with parameningeal extension [Intergroup
Rhabdomyosarcoma Study-III (IRS-III), Regimen
36]. Two of those 3 patients were long-term survivors. Despite being a retrospective report and
including small numbers of patients, the results
were encouraging.
In the early 2000s, 2 reports suggested that
intensive therapy, including radiation therapy,
might have activity against AT/RTs. In 2004,
Hilden et al. (2004) described 42 children who
were enrolled on an AT/RT registry at the
Cleveland Clinic. The median age of patients
at diagnosis was 24 months. Primary therapy
included chemotherapy in all patients,
radiotherapy in 13 patients (31%), high-dose
chemotherapy with stem cell transplant (HSCT)
in 13 patients (31%), and intrathecal chemotherapy in 16 patients (38%). The median survival was 16.75 months and the median
event-free survival was 10 months. Twentyseven patients (64%) died of tumor from 3 to
62 months from diagnosis and 1 patient died
due to toxicity. Fourteen patients (33%) showed
no evidence of disease (9.596 months from
diagnosis). The authors concluded that aggressive therapy may prolong the survival of some
children with AT/RTs. A study by Tekautz
et al. (2005) described a retrospective review
of 31 patients with AT/RTs from St. Jude
Childrens Research Hospital. Children aged
3 years or older were more likely to have
received craniospinal radiation. The event-free
and overall survival of children aged 3 years or
older was 78% 14% and 89% 11%, respectively. This was significantly better than
younger patients, who had an event-free survival and overall survival of 11% 6% and
17% 8%, respectively. They did not identify
any other clinical characteristics that were predictive of survival. These studies suggest that
the survival of children with ATRTs is improved
with aggressive therapy, including radiation

therapy.
There are two published prospective studies
for children with newly diagnosed AT/RTs. In
2008, Chi et al. (2009) reported the results of the
first prospective trial for children with newly
diagnosed AT/RTs, which included chemotherapy, intrathecal chemotherapy and radiation
therapy. There were 20 patients in the study. The
objective response rate to the first 12 weeks of
chemotherapy (pre-radiation therapy) was 58%.
The objective response rate observed after radiation therapy was 38%. The 2-year progressionfree and overall survival rates were 53% 13%
and 70% 10%, respectively, and the median
overall survival had not been reached. The
authors concluded that this intensive treatment
regimen resulted in a significant improvement in
time to progression and overall survival for
patients with this previously poor-prognosis
tumor. A prospective report from Gardner et al.
(2008) described 3 survivors among 13 children
with AT/RTs treated with high-dose chemotherapy with stem cell rescue per the Head Start I
and II regimens. The authors noted that the 3
survivors were from among the 7 children treated
on the Head Start II regimen, which added high
dose methotrexate to the treatment regimen.
Both of these regimens are quite intensive with
considerable short term and long term morbidities and risks of treatment related mortality.
Nevertheless, intensive therapy is believed to
be justified given the poor prognosis for children with AT/RTs treated on conventional infant
brain tumor protocols.
Present and Future Therapy

for Children with Atypical Teratoid/
Rhabdoid Tumors
At present, the Childrens Oncology Group is
studying the effectiveness of 3 sequential courses
of high-dose chemotherapy and stem cell rescue
for children with AT/RTs. In addition, a consortium of childhood cancer centers lead by St. Jude
Childrens Research Hospital is exploring a
regimen of risk-adapted chemotherapy and radiation therapy. Finally, the Head Start consortium
is further examining the role of high-dose chemotherapy with stem cell rescue for children with
AT/RTs.
Given the relatively modest survival rates with
considerable treatment-related morbidity for children with AT/RTs and increasingly better understanding of the molecular biology of these tumors, it
is hypothesized that future therapies may be developed that improve survival with less toxicity. For
example, Alarcon-Vargas et al. (2006) identified that
survival of rhabdoid tumor cells is dependent on the
presence of cyclin D1, a downstream target of
SMARCB1. They demonstrated that N-(4hydroxyphenyl) retinamide downmodulates cyclin
D1 and induces G1 arrest and apoptosis in rhabdoid
tumor cell lines. Also, N-(4-hydroxyphenyl) retinamide has synergistic activity with 4-hydroxytamoxifen against rhabdoid tumors in in vivo and
in vitro rhabdoid tumor models. In another preclinical study, Wu et al. (2008) hypothesized that
the loss of SMARCB1 activity in rhabdoid tumors
impairs the innate antiviral response of the tumor,
and that these tumors might be susceptible to treatment with oncolytic viruses. Therefore, they
infected mice with rhabdoid tumors implanted in
the brain and flank with a myxoma virus and attenuated vesicular stomatitis virus and they demonstrated anti-tumor activity by the viruses. Preclinical
studies have examined the use of dendritic cell
based therapies against rhabdoid tumors. For
example, Katsumi et al. (2008) have reported activity of the combination of trastuzumab, a humanized monoclonal antibody against human epidermal
growth factor receptor-2, autologous or allogeneic
peripheral blood mononuclear cells augmented by
interleukin-2 against malignant rhabdoid tumor
cells. A report by Ardon et al. (2010) described
2 long-term survivors among 3 children with
relapsed AT/RT who were treated with autologous,
monocyte-derived dendritic cells loaded with
tumor lysate, which was used as source of tumorassociated antigens. At present, such potential
therapies are interesting but will require considerably
more evaluation before they will be implemented
into frontline therapy for children with AT/RTs.
19
Conclusions
AT/RTs are aggressive tumors that occur most
often among very young children and are often
associated with either a germ line or somatic
mutation in the SMARCB1 gene. Whereas they
have historically had a dismal prognosis, more
recent reports have demonstrated improved survival for children with AT/RTs, albeit with high
short-term and long-term morbidities. Further
therapies, based upon the molecular biology of
these tumors, may further improve the prognosis for children with AT/RTs with less toxicity.
References
Alarcon-Vargas D, Zhang Z, Agarwal B, Challagulla K,
Mani S, Kalpana GV (2006) Targeting cyclin D1, a
downstream effector of INI1/hSNF5, in rhabdoid
tumors. Oncogene 25(5):722734
Ardon H, De Vleeschouwer S, Van Calenbergh F, Claes
L, Kramm CM, Rutkowski S, Wollf JE, Van Gool SW
(2010) Adjuvant dendritic cell-based tumour
vaccination for children with malignant brain tumours.
Pediatr Blood Cancer 54(4):519525
Arslanoglu A, Aygun N, Tekhtani D, Aronson L, Cohen
K, Burger PC, Yousem DM (2004) Imaging findings
of CNS atypical teratoid/rhabdoid tumors. AJNR Am
J Neuroradiol 25(3):476480
Biegel JA, Rorke LB, Packer RJ, Emanuel BS (1990)
Monosomy 22 in rhabdoid or atypical tumors of the
brain. J Neurosurg 73(5):710714
tumors. Cancer Res 59(1):7479
Biegel JA, Tan L, Zhang F, Wainwright L, Russo P, Rorke
LB (2002) Alterations of the hSNF5/INI1 gene in central nervous system atypical teratoid/rhabdoid tumors
and renal and extrarenal rhabdoid tumors. Clin Cancer
Res 8(11):34613467
Biggs PJ, Garen PD, Powers JM, Garvin AJ (1987)
Malignant rhabdoid tumor of the central nervous
system. Hum Pathol 18(4):332337
Bonnin JM, Rubinstein LJ, Palmer NF, Beckwith JB
(1984) The association of embryonal tumors originating
in the kidney and in the brain a report of 7 cases.
Cancer 54(10):21372146
Atypical teratoid/rhabdoid tumor of the central nervous
system: a highly malignant tumor of infancy and
20
childhood frequently mistaken for medulloblastoma: a
Pediatric Oncology Group study. Am J Surg Pathol
22(9):10831092
Chang CY, Wong TT, Ho DM (2005)
Neuroradiological findings in atypical teratoid/rhabdoid tumor of the central nervous system. Acta
Radiol 46(1):8996
Chi SN, Zimmerman MA, Yao X, Cohen KJ, Burger P,
Biegel JA, Rorke-Adams LB, Fisher MJ, Janss A,
Mazewski C, Goldman S, Manley PE, Bowers DC,
Bendel A, Rubin J, Turner CD, Marcus KJ,
Goumnerova L, Ullrich NJ, Kieran MW (2009)
Intensive multimodality treatment for children with
newly diagnosed CNS atypical teratoid rhabdoid
tumor. J Clin Oncol 27(3):385389
El-Nabbout B, Shbarou R, Glasier CM, Saad AG (2009)
Primary diffuse cerebral leptomeningeal atypical
teratoid rhabdoid tumor: report of the first case.
J Neurooncol. doi:10.1007/s11060-009-0094-z
Evans A, Ganatra R, Morris SJ (2001) Imaging features of
primary malignant rhabdoid tumour of the brain.
Pediatr Radiol 31(9):631633
Gardner SL, Asgharzadeh S, Green A, Horn B, McCowage
G, Finlay J (2008) Intensive induction chemotherapy
followed by high dose chemotherapy with autologous
hematopoietic progenitor cell rescue in young children
newly diagnosed with central nervous system atypical
teratoid rhabdoid tumors. Pediatr Blood Cancer
51(2):235240
Geyer JR, Sposto R, Jennings M, Boyett JM, Axtell RA,
Breiger D, Broxson E, Donahue B, Finlay JL,
Goldwein JW, Heier LA, Johnson D, Mazewski C,
Miller DC, Packer R, Puccetti D, Radcliffe J, Tao ML,
Shiminski-Maher T (2005) Multiagent chemotherapy
and deferred radiotherapy in infants with malignant
brain tumors: a report from the Childrens Cancer
Group. J Clin Oncol 23(30):76217631
Hanna SL, Langston JW, Parham DM, Douglass EC
(1993) Primary malignant rhabdoid tumor of the brain:
clinical, imaging, and pathologic findings. AJNR Am
in a registry. J Clin Oncol 22(14):28772884
Howlett DC, King AP, Jarosz JM, Stewart RA, al-Sarraj
ST, Bingham JB, Cox TC (1997) Imaging and pathological features of primary malignant rhabdoid
tumours of the brain and spine. Neuroradiology
39(10):719723
Jackson EM, Shaikh TH, Gururangan S, Jones MC,
Malkin D, Nikkel SM, Zuppan CW, Wainwright LM,
Zhang F, Biegel JA (2007) High-density single nucleotide polymorphism array analysis in patients with
germline deletions of 22q11.2 and malignant rhabdoid
tumor. Hum Genet 122(2):117127
Judkins AR, Mauger J, Ht A, Rorke LB, Biegel JA (2004)
Immunohistochemical analysis of hSNF5/INI1 in

pediatric CNS neoplasms. Am J Surg Pathol
28(5):644650
Katsumi Y, Kuwahara Y, Tamura S, Kikuchi K, Otabe O,
Tsuchiya K, Iehara T, Kuroda H, Hosoi H, Sugimoto T
(2008) Trastuzumab activates allogeneic or autologous
antibody-dependent cellular cytotoxicity against
malignant rhabdoid tumor cells and interleukin-2 augments the cytotoxicity. Clin Cancer Res
14(4):11921199
Koeller KK, Rushing EJ (2003) From the archives of the
AFIP: medulloblastoma: a comprehensive review with
radiologic-pathologic correlation. Radiographics
23(6):16131637
Koral K, Gargan L, Bowers DC, Gimi B, Timmons CF,
Weprin B, Rollins NK (2008) Imaging characteristics
of atypical teratoid-rhabdoid tumor in children
compared with medulloblastoma. AJR Am J
Roentgenol 190(3):809814
Lee YK, Choi CG, Lee JH (2004) Atypical teratoid/
rhabdoid tumor of the cerebellum: report of
two infantile cases. AJNR Am J Neuroradiol
25(3):481483
Lefkowitz I, Rorke L, Packer R, Sutton L, Siegel K,
Katnick R (1987) Atypical teratoid tumor of infancy
definition of an entity. Ann Neurol 22(3):448449
Munoz A, Carrasco A, Munoz MJ, Esparza J (1995)
Cranial rhabdoid tumor with marginal tumor cystic
component and extraaxial extension. AJNR Am
Niwa T, Aida N, Tanaka M, Okubo J, Sasano M, Shishikura
A, Fujita K, Ito S, Tanaka Y, Kigasawa H (2009)
Diffusion-weighted imaging of an atypical teratoid/
rhabdoid tumor of the cervical spine. Magn Reson
Med Sci 8(3):135138
Olson TA, Bayar E, Kosnik E, Hamoudi AB, Klopfenstein
KJ, Pieters RS, Ruymann FB (1995) Successful treatment of disseminated central nervous system malignant rhabdoid tumor. J Pediatr Hematol Oncol
17(1):7175
Packer RJ, Biegel JA, Blaney S, Finlay J, Geyer JR,
Heideman R, Hilden J, Janss AJ, Kun L, Vezina G,
Rorke LB, Smith M (2002) Atypical teratoid/rhabdoid
tumor of the central nervous system: report on workshop. J Pediatr Hematol Oncol 24(5):337342
Parmar H, Hawkins C, Bouffet E, Rutka J, Shroff M
(2006) Imaging findings in primary intracranial atypical teratoid/rhabdoid tumors. Pediatr Radiol
36(2):126132
85(1):5665
Rumboldt Z, Camacho DL, Lake D, Welsh CT, Castillo M
(2006) Apparent diffusion coefficients for differentiation of cerebellar tumors in children. AJNR Am
Takei H, Adesina AM, Mehta V, Powell SZ, Langford LA
(2009) Atypical teratoid/rhabdoid tumor of the pineal
region in an adult. J Neurosurg. doi:10.3171/2009.10.
JNS09964
Tekautz TM, Fuller CE, Blaney S, Fouladi M, Broniscer

A, Merchant TE, Krasin M, Dalton J, Hale G, Kun LE,
Wallace D, Gilbertson RJ, Gajjar A (2005) Atypical
teratoid/rhabdoid tumors (ATRT): improved survival
therapy and high-dose alkylator-based chemotherapy.
J Clin Oncol 23(7):14911499
Verma A, Morriss C (2008) Atypical teratoid/rhabdoid tumor
of the optic nerve. Pediatr Radiol 38(10):11171121
Versteege I, Sevenet N, Lange J, Rousseau-Merck MF,
Ambros P, Handgretinger R, Aurias A, Delattre O
(1998) Truncating mutations of hSNF5/INI1
21
in
aggressive
paediatric
cancer.
Nature
394(6689):203206
Wu Y, Lun X, Zhou H, Wang L, Sun B, Bell JC, Barrett
JW, McFadden G, Biegel JA, Senger DL, Forsyth PA
(2008) Oncolytic efficacy of recombinant vesicular
stomatitis virus and myxoma virus in experimental
models of rhabdoid tumors. Clin Cancer Res
14(4):12181227
Zuccoli G, Izzi G, Bacchini E, Tondelli MT, Ferrozzi
F, Bellomi M (1999) Central nervous system atypical
teratoid/rhabdoid tumour of infancy. CT and mr
findings. Clin Imaging 23(6):356360
Pediatric Rhabdomyosarcoma: Role

of Cell Cycle Regulators Alteration
Kenichi Kohashi, Yukiko Takahashi,
Tomoaki Taguchi, and Yoshinao Oda
Contents
Abstract
Introduction ............................................................
24
Rhabdomyosarcoma ..............................................
24
Alteration of Cell-Cycle Regulators

in Rhabdomyosarcoma ..........................................
RB Pathway..............................................................
P53 Pathway.............................................................
The Other Regulators ...............................................
24
24
26
27
Conclusion ..............................................................
27
References ...............................................................
29
K. Kohashi Y. Oda (*)

Department of Anatomic Pathology, Pathological
Sciences, Graduate School of Medical Sciences,
Kyushu University, Maidashi 3-1-1, Higashi-ku,
Fukuoka 812-8582, Japan
e-mail: oda@surgpath.med.kyushu-u.ac.jp
Y. Takahashi T. Taguchi
Department of Pediatric Surgery, Graduate School of
Medical Sciences, Kyushu University, Maidashi 3-1-1,
Higashi-ku, Fukuoka 812-8582, Japan
In the present paper, rhabdomyosarcoma of

the soft tissue and central nervous system are
reviewed, and role of cell cycle regulators
alterations in rhabdomyosarcoma are discussed. Rhabdomyosarcoma, which is the
most common pediatric soft tissue sarcoma,
can be divided into two major types: embryonal type and the more aggressive alveolar type.
Genetic backgrounds are also different:
embryonal type demonstrates complex karyotypes, whereas the alveolar type discloses specific fusion gene transcripts of PAX3/7-FKHR.
Primary rhabdomyosarcoma of the central
nervous system is an extremely rare entity and
nearly all of the embryonal subtype. In cell
cycle regulators, the retinoblastoma protein
(RB) and p53 are well known tumor suppressors, and some reports have discussed the
altered expression and molecular abnormalities of various cell cycle regulatory proteins
including RB and p53 pathway. However, cell
cycle regulators alterations reported as significant in the outcome of the patients or histological subtypes are confined to a small
number including RB, MYCN and Akt related
proteins. Especially, as it has been reported
that RB plays a role in myogenic differentiation, RB may be one of the most important
factors in tumorigenesis and the progression
of rhabdomyosarcoma. For the future, not
only further analyses of major cell-cycle regulators of alteration but also elucidations of the

23
K. Kohashi et al.
24
interactions between major cell-cycle pathway

related genes and fusion genes, such as PAX3,
PAX7, FKHR, Akt and MYCN, may be more
important.
Introduction
Many tumor-suppressor genes and oncogenes
directly participate in or regulate signal transduction pathways. They function in the complex signaling pathways involved in the control of cellular
differentiation and the cell cycle. Therefore,
alterations in cell-cycle regulation may underlie
the development and/or progression of human
malignancies (Takahashi et al. 2004).
Alteration of cell-cycle regulators can disrupt
normal growth control in response to environmental cues, or it can dismantle cell-cycle checkpoints that otherwise limit cell division or that
induce cell suicide in response to DNA damage
or oncogene activation. RB and p53 pathways
play main roles in these processes, and these
pathways are closely related. To clarify the individual specific tumorigenesis, analyses of the
regulation of cell-cycle regulators and discussions of related pathways are necessary.
Rhabdomyosarcoma
Rhabdomyosarcoma is the most commonly
occurring soft tissue sarcoma in children, accounting for 58% of all malignancies. Two major
histologic subtypes of rhabdomyosarcoma can be
identified: embryonal and alveolar. Embryonal
rhabdomyosarcoma is composed of primitive
mesenchymal cells in various stages of myogenesis (Parham and Barr 2002b). Meanwhile,
alveolar rhabdomyosarcoma exhibits round-cell
cytological features reminiscent of lymphomas
but with primitive myoblastic differentiation
(Parham and Barr 2002a). Therefore, the rate of
nuclear immunoreactivity for myogenin varies
from subtype to subtype and is potentially a
useful tool for differential diagnosis (Hostein
et al. 2004). The clinical behavior differs depending on the subtype, as does the genetic back-
ground: the embryonal type demonstrates

complex karyotypes, whereas the alveolar type
discloses specific fusion gene transcripts of
PAX3/7-FKHR (Kohashi et al. 2008).
Primary rhabdomyosarcoma of the central
nervous system is classified as tumors of the
meninges, mesenchymal tumors according to
the WHO classification of tumors of the central
nervous system. It is an extremely rare entity,
with only approximately 40 cases reported
previously (Celli et al. 1998; Paulus et al. 2007;
Guilcher et al. 2008). These cases ranged in age
from 1 year to 68 years, the majority of which
were children (Celli et al. 1998; Guilcher et al.
2008). It is striking that nearly all rhabdomyosarcomas of the central nervous system are of
the embryonal subtype, and no cases of alveolar rhabdomyosarcoma have been reported
(Burger and Scheithauer 2007; Paulus et al.
2007). Therefore, rhabdomyosarcoma of the
central nervous system should be definitively
differentiated from other brain tumors with
rhabdomyosarcomatous differentiation, such
as germ cell tumor, medullomyosarcoma, and
gliosarcoma (Burger and Scheithauer 2007;
Paulus et al. 2007).
Alteration of Cell-Cycle Regulators

in Rhabdomyosarcoma
RB Pathway
The RB gene, located on the long arm of chromosome 13, is a tumor suppressor gene that was first
identified as a retinoblastoma. The RB, which is
phosphorylated or is bound to E2F, plays an
essential role in regulating the cell cycle (from
G1 to S phases) and, consequently, cell proliferation. Hypophosphorylated RB is bound to E2F
and prevents the activation of the E2F target
genes, but hyperphosphorylated RB does not
bind E2F, which remain free to active the target
genes. The target genes of E2F are cyclins, CDKs,
checkpoints regulators, and DNA repair and replication proteins.
RB phosphorylation is controlled by upstream
regulators such as the p16INK4a family, p21CIP1
Pediatric Rhabdomyosarcoma: Role of Cell Cycle Regulators Alteration
family, CDKs and cyclins. Phosphorylation is

triggered in the early G1 phase by the cyclin
D-CDK4-CDK6 complexes. The activities of
CDK4 and CDK6 are inhibited by p16INK4a and
p21CIP1. At the end of the G1 phase, phosphorylation
is completed by cyclin E-CDK2 complexes.
CDK2 activity is inhibited by p21CIP1. These components of the regulatory machinery that controls
the G1-S phase transition behave as tumor
suppressors or proto-oncogenes, and they are
frequently altered in tumor cells (Derenzini
et al. 2007).
RB
Concerning RB protein expression, it was previously reported that immunohistochemical RB
labeling indexes in 31 embryonal rhabdomyosarcomas (median value, 31%) were significantly
reduced in comparison with those observed in
26 alveolar rhabdomyosarcomas (median value,
85%) (P < 0.0001) (Kohashi et al. 2008).
Several analyses of the RB gene have been
performed as part of the molecular background
of RB protein expression. In chromosomal analysis, gains of chromosome 13 are frequently associated with embryonal rhabdomyosarcoma rather
than alveolar rhabdomyosarcoma. An allelic
imbalance at 13q1214 including RB was more
frequently detected in embryonal rhabdomyosarcoma (13/27) than in alveolar rhabdomyosarcoma
(3/20) (P = 0.04). However, no close relationship
between RB expression and allelic imbalance at
13q1214 was found (P = 0.54) (Kohashi et al.
2008). Moreover, although various sarcomas
acquired RB mutations, such alterations were not
found in sporadic rhabdomyosarcoma. Although
RB mutations were not detected in rhabdomyosarcoma, genetic changes were found in genes
encoding proteins that regulate RB function (Xia
et al. 2002).
CDKs
Increased expression of CDK2 mRNA was more
frequently observed in the embryonal rhabdomyosarcoma (6/13: 46%) than in the alveolar rhabdomyosarcoma (2/10: 20%). However, CDK4
mRNA expression was increased in both embryonal rhabdomyosarcoma (11/13: 85%) and
25
alveolar rhabdomyosarcoma (8/10: 80%) (Moretti

et al. 2002).
In another study of 22 rhabdomyosarcomas
(9 alveolar rhabdomyosarcoma, 7 anaplastic
embryonal rhabdomyosarcoma and 6 classic
embryonal rhabdomyosarcoma), amplification
of the CDK4 gene (more than 10 copies) was
found in 2 cases (9%), including 1 alveolar
rhabdomyosarcoma (11%) and 1 anaplastic
embryonal
rhabdomyosarcoma
(14%).
Overrepresentation (from 3 to 10 copies) of this
gene was found in 5 cases (23%), including 2
alveolar rhabdomyosarcomas (22%) and 3 anaplastic embryonal rhabdomyosarcomas (43%).
None of the classic embryonal rhabdomyosarcomas showed evidence of CDK4 amplification or
overrepresentation. Immunohistochemically, 19
positive expression cases (8 alveolar rhabdomyosarcomas, 6 anaplastic embryonal rhabdomyosarcomasand5classicembryonalrhabdomyosarcomas)
for CDK4 proteins, including 10 high-expression
cases (3 alveolar rhabdomyosarcomas, 5 anaplastic
embryonal rhabdomyosarcomas and 2 classic
embryonal rhabdomyosarcomas), were recognized
(Ragazzini et al. 2004).
Cyclins
Increased expression of Cyclin Ds mRNA was
found in 11/13 (Cyclin D1), 10/13 (Cyclin D2)
and 4/13 (Cyclin D3) embryonal rhabdomyosarcoma cases, and 9/10 (Cyclin D1), 6/10 (Cyclin
D2) and 5/10 (Cyclin D3) alveolar rhabdomyosarcoma cases. No difference was found between
the embryonal and alveolar types with regard to
the expression rates of those Cyclin Ds mRNA.
However, the results suggest that mRNA expression rates of Cyclin D1 and D2 were higher than
that of Cyclin D3 (Moretti et al. 2002).
CIP/KIP Family
Increased expression of p21CIP1 mRNA was
detected more frequently in embryonal rhabdomyosarcoma (13/13) than in alveolar rhabdomyosarcoma (4/10) (P < 0.05) (Moretti et al.
2002). In 4 of the 8 embryonal and 1 of the 5
alveolar rhabdomyosarcoma cases, there was
detectable expression of the p27KIP1 gene product.
Although the small number of samples did not
26
K. Kohashi et al.
allow a statistical evaluation, these findings

strongly suggested there are differences in p27KIP1
expression between embryonal and alveolar rhabdomyosarcoma (Moretti et al. 2002). Another
study showed that PAX3FKHR reduced the
expression of p27kip1 protein via elevated Skp2
and 26S proteasome-dependent degradation
(Zhang and Wang 2003).
Upregulation of p21CIP1 and 14-3-3s by p53

imposes G1 and G2 arrest, respectively. MDM2
interacts with p53 and downregulates p53 protein
expression, thereby functioning as an oncogene.
P53 in turn upregulates MDM2 expression, which
functions as an autoregulatory negative feedback
loop. MDM2 activity is inhibited by p14ARF
(Osada and Takahashi 2002).
INK4a/ARF Family
Two of the 6 clinical cases and all 3 cell lines for
embryonal rhabdomyosarcoma, as well as 1 of
the 6 clinical cases and both of the cell lines for
alveolar rhabdomyosarcoma, had homozygous
deletion of the p16INK4a/P15INK4b gene. In cases
without homozygous deletion, no mutations of
this gene were detected (Iolascon et al. 1996).
In another study, 2 of the 4 alveolar rhabdomyosarcoma cell lines were each found to have both a
nonsense and a missense mutation of the p16INK4a
gene. However, no mutation of the p16INK4a/p14ARF
gene was found in any of the remaining cases,
including 3 cell lines and 21 clinical cases of
embryonal rhabdomyosarcoma as well as 11
clinical cases of alveolar rhabdomyosarcoma.
Reduced or absent expression of p16INK4a mRNA
was observed in 11 of 24 samples, including 2
embryonal and 1 alveolar rhabdomyosarcoma
cell lines and 9 clinical cases (the subtypes of the
clinical cases were unknown). Ten of 24 samples,
including 1 alveolar rhabdomyosarcoma cell line
and 9 clinical cases (the subtypes of the clinical
cases were unknown), demonstrated reduced or
absent expression of p14ARF mRNA (Chen et al.
2007). In the embryonal rhabdomyosarcoma cell
line, those authors suggested that deletion of the
p16INK4a gene may not only facilitate growth but
also inhibit myogenic differentiation (Urashima
et al. 1999).
P53
The nuclear accumulation, defined as the staining
of more than 10% of the nuclei, of p53 was
detected in 14 of the 36 (39%) embryonal and 7
of the 34 (21%) alveolar rhabdomyosarcoma
cases. Moreover, 10 (5 embryonal, 4 alveolar and
1 pleomorphic) of the 45 rhabdomyosarcoma
cases (22%) exhibited p53 alterations (exons
59). However, there was no significant correlation between p53 immunoreactivity and p53
mutation status (Takahashi et al. 2004). Other
studies showed the point mutation of p53 in 4 of
36 (11%) embryonal and 1 of 17 (6%) alveolar
rhabdomyosarcoma cases (Xia et al. 2002).
P53 Pathway
P53 is a tumor suppressor gene that plays a central role in cell-cycle checkpoints, apoptosis,
gene stability and inhibition of angiogenesis. The
activated p53 transactivates its target genes,
including p21CIP1, 14-3-3s and MDM2.
MDM2
Overexpression, defined as the staining of more
than 10% of the nuclei, of MDM2 was detected in
7 of the 36 (19%) embryonal and 2 of the 34 (6%)
alveolar rhabdomyosarcoma cases. Moreover, 3
of the 18 cases with MDM2 gene amplification
(subtype unknown) were observed. However,
there was no significant correlation between
MDM2 overexpression and its gene amplification. In addition, MDM2 showed no association
with patient prognosis or with any other clinicopathological parameters, including histological
subtype and age (Takahashi et al. 2004).
In another study of 22 rhabdomyosarcomas
(9 alveolar rhabdomyosarcomas, 7 anaplastic
embryonal rhabdomyosarcomas and 6 classic
embryonal rhabdomyosarcomas), the amplification (more than 10 copies) of the MDM2 gene
was found in 2 cases (9%), including 1
alveolar rhabdomyosarcoma (11%) and 1
anaplastic
embryonal
rhabdomyosarcoma
(14%). Overrepresentation (from 3 to 10 copies)
of this gene was found in 3 cases (14%), including 2 alveolar rhabdomyosarcoma (22%) and 1
anaplastic
embryonal
rhabdomyosarcoma
(14%). None of the classic embryonal rhabdomyosarcomas
showed
evidence
of
MDM2 amplification or overrepresentation.
Immunohistochemically, 12 cases with positive
expression were recognized (5 alveolar rhabdomyosarcomas, 4 anaplastic embryonal rhabdomyosarcomas and 3 classic embryonal
rhabdomyosarcomas) for MDM2 proteins,
including 4 high-expression cases (2 alveolar
rhabdomyosarcomas, 1 anaplastic embryonal
rhabdomyosarcoma and 1 classic embryonal
rhabdomyosarcoma) (Ragazzini et al. 2004).
The Other Regulators

MYCN
MYCN, a member of the MYC family of transcription factors, is thought to drive cell proliferation (that is, it upregulates cyclins and
downregulates p21) and to play a very important
role in regulating cell growth (it upregulates ribosomal RNA and proteins), apoptosis (it downregulates Bcl-2) and differentiation. MYCN is a
proto-oncogene and is upregulated in various
kinds of tumors.
In rhabdomyosarcoma, the FISH method
detected MYCN amplification in 9 of 15 alveolar
cases but in none of the 14 embryonal cases.
Moreover, among the 15 alveolar rhabdomyosarcoma cases, the survival rate of patients with
amplified MYCN oncogene (1 of 9 cases: 11%)
was significantly worse than that of cases with
non-amplified MYCN oncogene (4 of 6 cases:
66%) (P < 0.05) (Hachitanda et al. 1998).
Akt
The PI3K-Akt pathway is a growth-promoting
pathway that regulates cell proliferation. Akt is
known to play a role in the cell cycle, and MDM2
is positively regulated through its phosphorylation by Akt. Activation of Akt is shown to
overcome cell-cycle arrest in the G1 and G2
phases. In Akt Ser473, 4EBP1 Thr37/46, eIF4G
Ser1108 and p70S6 Thr389, high phosphorylation
levels were associated with poor overall and disease-free survival of rhabdomyosarcoma cases.
27
Moreover, the interrelationship between insulin

receptor substrate (IRS-1) activity and the Akt/
mTOR pathway proteins may be altered in the
tumors of patients who subsequently showed
poor survival after chemotherapy compared with
the tumors of patients who were long-term survivors (Petricoin et al. 2007).
Ki-67
Ki-67 protein, a cellular marker for proliferation,
is found throughout the active phases of the cell
cycle (G1, S, G2 and M phases), and it is absent
in resting (G0) cells.
Immunohistochemically, the mean values of
the Ki-67 labeling index in embryonal rhabdomyosarcoma and alveolar rhabdomyosarcoma were 13.3% and 10.4%, respectively. As
for the proliferative activities determined by the
Ki-67 labeling index, there was no significant
difference between the two subtypes (Takahashi
et al. 2004).
Conclusion
As stated above, there have been several studies
of cell-cycle regulators of alterations in rhabdomyosarcoma (Fig. 3.1). However, there were
no significant differences in the frequency of
these alterations in the p53 pathway between
embryonal and alveolar rhabdomyosarcoma.
According to a study by Taubert et al. (1998), the
p53 mutational mean rate in soft tissue sarcoma
was 16.3% (range, 058.5%). There was also no
significant difference in the p53 mutation rate
between embryonal and alveolar types. Moreover,
it has been documented that, in soft tissue sarcomas with specific translocation such as synovial
sarcoma,
myxoid/round-cell
liposarcoma,
Ewings sarcoma/primitive neuroectodermal
tumor and alveolar rhabdomyosarcoma, p53
pathway alterations were a rather rare event (Oda
et al. 2005). Therefore, cell-cycle regulators of
alterations in the p53 pathway may not play
an important role in the tumorigenesis of
rhabdomyosarcoma.
In the RB pathway, the RB labeling index was
significantly higher in alveolar rhabdomyosarcoma
28
K. Kohashi et al.
Fig. 3.1 Frequency of cell-cycle regulators of alterations in rhabdomyosarcoma subtypes (HD homozygous deletion,
Amp gene amplification, HE high expression, IE increased expression, LI labeling index)
Fig. 3.2 Immunohistochemical staining of retinoblastoma

protein (RB) expression. (a) The numbers of tumor cells
with a positive nuclear reaction are reduced; the labeling
index is 12% (0 years old, embryonal rhabdomyosarcoma,
neck). (b) Most tumor cells show the immunoreaction in

the nuclei; the labeling index is 93% (24 years old, alveolar
rhabdomyosarcoma, forearm)
than in embryonal rhabdomyosarcoma (Fig. 3.2).

RB expression varies according to variations in
myogenic differentiation, and plays a role in the
switch from myogenic proliferation to differenti-
ation (Huh et al. 2004). As the activity of

myogenic regulatory factors is tightly coupled to
cell-cycle control, RB may be one of the most
important factors in tumorigenesis and the
progression of rhabdomyosarcoma. Moreover,

the RB labeling index has the potential to
become an ancillary parameter in the differential diagnosis of rhabdomyosarcoma subtypes
(Kohashi et al. 2008).
Previous studies in malignant tumors including soft tissue sarcoma have suggested that
increased Ki-67 labeling index and p53 mutation
are associated with adverse clinical outcomes
(Levine 1999; Taubert et al. 1998). In addition, a
few authors have reported homozygous deletion
or hypermethylation of the p16INK4a gene and their
correlation with the loss of the p16INK4a protein
and poor prognosis in bone and soft-tissue
sarcoma (Oda et al. 2005). However, prognostic
factors of rhabdomyosarcoma in major cell-cycle
regulators such as p53 and p16INK4a are not evident. At the present time, the most important
prognostic factor is histological subtype; patients
with the alveolar subtype survived for a significantly shorter time than those with the embryonal
subtype. Moreover, in the alveolar subtype, PAX7FKHR-positive tumors behave in a more dormant
fashion than PAX3-FKHR-positive ones
(Davicioni et al. 2009; Oda and Tsuneyoshi 2009).
Therefore, the expression of proteins and the
alteration of genes in signaling pathways related
with PAX3, PAX7 and FKHR are intriguing. In
fact, FKHR is directly phosphorylated and regulated by Akt, which is reported to be a prognostic
factor of rhabdomyosarcoma (Biggs et al. 1999).
For the future, not only further analyses of
major cell-cycle regulators of alteration but also
elucidations of the signaling pathway associated
with major cell-cycle pathway-related genes and
fusion genes, such as PAX3, PAX7, FKHR, Akt
and MYCN, may be more important. Therefore,
the elucidation of tumorigenesis and biological
behavior and their corresponding molecular
target therapy should be conducted in
rhabdomyosarcoma.
References
Biggs WH 3rd, Meisenhelder J, Hunter T, Cavenee WK,
Arden KC (1999) Protein kinase B/Akt-mediated
phosphorylation promotes nuclear exclusion of the
winged helix transcription factor FKHR1. Proc Natl
Acad Sci USA 96:74217426
29
Burger PC, Scheithauer BW (eds) (2007) AFIP atlas

of tumor pathology series 4, tumors of the central
nervous system, 1st edn. American Registry of
Pathology, Washington, DC
Celli P, Cervoni L, Maraglino C (1998) Primary rhabdomyosarcoma of the brain: observations on a case
with clinical and radiological evidence of cure.
J Neurooncol 36:259267
Chen Y, Takita J, Mizuguchi M, Tanaka K, Ida K, Koh K,
Igarashi T, Hanada R, Tanaka Y, Park MJ, Hayashi Y
(2007) Mutation and expression analyses of the MET
and CDKN2A genes in rhabdomyosarcoma with
emphasis
on
MET
overexpression.
Genes
Chromosomes Cancer 46:348358
Davicioni E, Anderson MJ, Finckenstein FG, Lynch JC,
Qualman SJ, Shimada H, Schofield DE, Buckley JD,
Meyer WH, Sorensen PH, Triche TJ (2009) Molecular
classification of rhabdomyosarcomagenotypic and
phenotypic determinants of diagnosis: a report from
the Childrens Oncology Group. Am J Pathol
174:550564
Derenzini M, Montanaro L, Vici M, Barbieri S, Ceccarelli
C, Santini D, Taffurelli M, Martinelli GN, Trer D
(2007) Relationship between the RB1 mRNA level
and the expression of phosphorylated RB protein in
human breast cancers: their relevance in cell proliferation activity and patient clinical outcome. Histol
Histopathol 22:505513
Guilcher GM, Hendson G, Goddard K, Steinbok P,
Bond M (2008) Successful treatment of a child with
a primary intracranial rhabdomyosarcoma with chemotherapy and radiation therapy. J Neurooncol
86:7982
Hachitanda Y, Toyoshima S, Akazawa K, Tsuneyoshi M
(1998) N-myc gene amplification in rhabdomyosarcoma detected by fluorescence in situ hybridization:
its correlation with histologic features. Mod Pathol
11:12221227
Hostein I, Andraud-Fregeville M, Guillou L, TerrierLacombe MJ, Deminire C, Ranchre D, Lussan C,
Longavenne E, Bui NB, Delattre O, Coindre JM
(2004) Rhabdomyosarcoma: value of myogenin
expression analysis and molecular testing in diagnosing the alveolar subtype: an analysis of 109 paraffinembedded specimens. Cancer 101:28172824
Huh MS, Parker MH, Scim A, Parks R, Rudnicki MA
(2004) Rb is required for progression through myogenic differentiation but not maintenance of terminal
differentiation. J Cell Biol 166:865876
Iolascon A, Faienza MF, Coppola B, Rosolen A, Basso G,
Della Ragione F, Schettini F (1996) Analysis of cyclindependent kinase inhibitor genes (CDKN2A, CDKN2B,
and CDKN2C) in childhood rhabdomyosarcoma. Genes
Chromosomes Cancer 15:217222
Kohashi K, Oda Y, Yamamoto H, Tamiya S, Takahira T,
Takahashi Y, Tajiri T, Taguchi T, Suita S, Tsuneyoshi
M (2008) Alterations of RB1 gene in embryonal and
alveolar rhabdomyosarcoma: special reference to utility of pRB immunoreactivity in differential diagnosis
of rhabdomyosarcoma subtype. J Cancer Res Clin
Oncol 134:10971103
30
Levine EA (1999) Prognostic factors in soft tissue sarcoma.
Semin Surg Oncol 17:2332
Moretti A, Borriello A, Monno F, Criscuolo M, Rosolen
A, Esposito G, Dello Iacovo R, Della Ragione F,
Iolascon A (2002) Cell division cycle control in
embryonal and alveolar rhabdomyosarcomas. Eur J
Cancer 38:22902299
Oda Y, Tsuneyoshi M (2009) Recent advances in the
molecular pathology of soft tissue sarcoma: implications for diagnosis, patient prognosis, and molecular target therapy in the future. Cancer Sci
100:200208
Oda Y, Yamamoto H, Takahira T, Kobayashi C, Kawaguchi
K, Tateishi N, Nozuka Y, Tamiya S, Tanaka K, Matsuda
S, Yokoyama R, Iwamoto Y, Tsuneyoshi M (2005)
Frequent alteration of p16(INK4a)/p14(ARF) and p53
pathways reduced p14(ARF) expression both correlate
with poor prognosis. J Pathol 207:410421
Osada H, Takahashi T (2002) Genetic alterations of multiple tumor suppressors and oncogenes in the carcinogenesis and progression of lung cancer. Oncogene
21:74217434
Parham DM, Barr FG (2002a) Alveolar rhabdomyosarcoma. In: Fletcher CDM, Unni KK, Mertens F (eds)
WHO classification of tumours, pathology and genetics of tumours of soft tissue and bone. IARG Press,
Lyon, pp 150152
Parham DM, Barr FG (2002b) Embryonal rhabdomyosarcoma. In: Fletcher CDM, Unni KK, Mertens F (eds)
WHO classification of tumours, pathology and genetics of tumours of soft tissue and bone. IARG Press,
Lyon, pp 146149
Paulus W, Scheithauer BW, Perry A (2007) Mesenchymal,
non-meningothelial tumours. In: Louis DN, Ohgaki H,
K. Kohashi et al.
Wiestler OD, Cavenee WK (eds) WHO classification
of tumours of the central nervous system. IARG Press,
Lyon, pp 173177
Petricoin EF 3rd, Espina V, Araujo RP, Midura B, Yeung
C, Wan X, Eichler GS, Johann DJ Jr, Qualman S,
Tsokos M, Krishnan K, Helman LJ, Liotta LA (2007)
Phosphoprotein pathway mapping: Akt/mammalian
target of rapamycin activation is negatively associated
with childhood rhabdomyosarcoma survival. Cancer
Res 67:34313440
Ragazzini P, Gamberi G, Pazzaglia L, Serra M, Magagnoli
G, Ponticelli F, Ferrari C, Ghinelli C, Alberghini M,
Bertoni F, Picci P, Benassi MS (2004) Amplification
of CDK4, MDM2, SAS and GLI genes in leiomyosarcoma, alveolar and embryonal rhabdomyosarcoma.
Histol Histopathol 19:401411
Takahashi Y, Oda Y, Kawaguchi K, Tamiya S, Yamamoto
H, Suita S, Tsuneyoshi M (2004) Altered expression
and molecular abnormalities of cell-cycle-regulatory
proteins in rhabdomyosarcoma. Mod Pathol
17:660669
Taubert H, Meye A, Wrl P (1998) Soft tissue sarcomas
and p53 mutations. Mol Med 4:365372
Urashima M, Teoh G, Akiyama M, Yuza Y, Anderson KC,
Maekawa K (1999) Restoration of p16INK4A protein
induces myogenic differentiation in RD rhabdomyosarcoma cells. Br J Cancer 79:10321036
Xia SJ, Pressey JG, Barr FG (2002) Molecular
pathogenesis of rhabdomyosarcoma. Cancer Biol
Ther 1:97104
Zhang L, Wang C (2003) PAX3-FKHR transformation
increases 26 S proteasome-dependent degradation of
p27Kip1, a potential role for elevated Skp2 expression.
J Biol Chem 278:2736

Rhabdoid Tumors: Imaging
with CT and MRI
Monika Warmuth-Metz and Michael Frhwald
Contents
Abstract
Introduction ............................................................
31
Demographic and Clinical

Characteristics, Localization
and Meningeal Dissemination ...............................
32
Computed Tomography (CT)................................
33
Magnetic Resonance Imaging (MRI) ...................
34
References ...............................................................
36
Atypical Teratoid/Rhabdoid tumors (AT/RT)

of the CNS are highly malignant (WHO grade
IV) heterogeneous embryonal tumors (ICDCode 9508/3 WHO classification 2007), which
are diagnosed with an increasing frequency.
Their incidence in large series is estimated to
be 12% of pediatric brain tumors but the incidence is about tenfold in children below the
age of 3 years, when the diagnosis is as common as medulloblastoma. Although the imaging features have always been described as
non-specific there seem to be certain morphological markers that are more frequently
observed in AT/RT than in other brain tumors
of childhood. Along with the clinical data of a
patient i.e. young age, metastatic lesion,
aggressive neoplasm suggestive features on
Magnetic Resonance Imaging (MRI) and
Computed Tomography (CT) are presented
and make the diagnosis possible in many
cases.
Introduction
M. Warmuth-Metz
Abteilung fr Neuroradiologie der
Universitat Wrzburg, Josef-Schneider-Str. 11,
97080 Wrzburg, Germany
e-mail: warmuth@neuroradiologie.uni-wuerzburg.de
M. Frhwald (*)
Klinik fr Kinder und Jugendliche Klinikum Augsburg,
Stenglinstr. 2, 86156 Augsburg, Germany
The histopathological features of AT/RT are

related to the ones initially described in malignant rhaboid tumors of the kidneys and soft tissue
in infants. The first report of a rhaboid tumor in
the CNS dates to 1985 (Montgomery et al. 1985).
Since the first description of the entity termed AT/
RT by Rorke et al. (1996) the importance of a

31
32
correct diagnosis has been emphasized because

AT/RT have often been misdiagnosed as primitive
neuroectodermal tumors (PNET)/medulloblastomas
(MB). AT/RT contain cell elements that are histopathologically indistinguishable from PNET/MB.
Other differential diagnoses are choroid plexus
carcinomas, germ cell tumors or malignant gliomas
(Judkins et al. 2007).
The histopathological features of AT/RT are
prominent rhabdoid cells and immunopositivity
for EMA (epithelial membrane antigen), vimentin, and keratin. Most importantly an absence of
the nuclear staining for the protein INI1 (also
known as hSNF5/SMARCB1/BAF47) has been
consistently demonstrated in up to 90% of AT/
RT. The genetic hallmark of the tumor is a loss of
the INI1-protein due to base pair mutations or a
loss of the INI1-locus due to deletions at 22q11.2.
In rare instances the diagnosis may be made
in tumors without changes of the INI1 gene
(Schneppenheim et al. 2010).
Germline mutations occur in about one third
of patients, true familial cases are however rare
(Wesseling et al. 2007). Children with metachronous or synchronous multiple malignant rhabdoid tumors, e.g. bilateral renal or additional soft
tissue tumors or affected relatives are almost
always afflicted by the rhabdoid tumor predisposition syndrome (RTPS). In these cases rhabdoid
tumors are highly aggressive and generally lethal
(Janson et al. 2006).
The prognosis of an AT/RT, usually found in
very young children below the age of 23 years is
much worse than in case of a PNET. They usually
do not respond well to chemotherapy strategies
designed for MB/PNET (Geyer et al. 2005). The
proliferation rates of AT/RT in children can be
higher than 50%, locally up to 100% (Ho et al.
2000; Biswas et al. 2009). In adults in whom this
tumor type is only exceptionally found the labeling indices can be considerably lower (Zarovnaya
et al. 2007; Lutterbach et al. 2001).
Rorke et al. and also a number of further investigators (Dang et al. 2003; Meyers et al. 2006;
Parmar et al. 2006) described the imaging features
of AT/RT as variable and essentially indistinguishable from PNETs or Medulloblastomas of
the posterior fossa.
M. Warmuth-Metz and M. Frhwald
MRI is the imaging procedure of choice

especially in children with brain tumors and
therefore many children are examined by MRI
only. We describe typical MRI features helping
in the diagnosis of an AT/RT.
Demographic and Clinical

Characteristics, Localization
and Meningeal Dissemination
Most AT/RT arise in children below the age of
3 years. The mean age is about 2 years. In our own
series of currently 91 patients the oldest was 12 years
of age. As our patients were collected in the context
of a pediatric study a certain bias is apparent. Boys
are more frequently affected (1.6:12:1) (Hilden
et al. 2004; Tekautz et al. 2005; Warmuth-Metz
et al. 2008). In our series the ratio was 49 boys to 42
girls. Nevertheless, adult patients have been
diagnosed with this tumor as well. The oldest
patients were diagnosed in their fifth decade.
Signs of elevated intracranial pressure such as
increase in head circumference, lethargy and vomiting are the non specific presenting symptoms and
are a consequence of the usually large tumor in the
supratentorial compartment or of an obstruction of
CSF pathways in infratentorial tumors.
Limited case reports, and articles focusing on
the imaging aspect found the majority of tumors
infratentorially. However, larger collaborative
studies report as the most frequent localization
the supratentorial compartment (Hilden et al.
2004; Warmuth-Metz et al. 2008). Supratentorially
the frontal lobe is preferred just as in gliomas.
Infratentorial tumors often grow outside the
midline and quite frequently into the supratentorial compartment. The extramidline position
of an infratentorial AT/RT (see Fig. 4.3b) is a
discriminating feature from the main differential diagnosis medulloblastoma (Koral et al.
2008). The infra- or supratentorial position of
an AT/RT appears to be related to the age of the
patient. In young children the infratentorial
compartment is more frequently affected while
in adult patients the supratentorial brain is preferably found to be affected (Tekautz et al. 2005;
Makuria et al. 2008).
Pediatric Atypical Teratoid/Rhabdoid Tumors: Imaging with CT and MRI
Table 4.1 Frequency of meningeal dissemination and

respective age of the patients according to the literature
Author
Hilden et al.
Koral et al.
Meyers et al.
Tekautz et al.
Frequency of
meningeal
disease (%)
21
10.5
24
27
Pamar et al.
Warmuth-Metz et al.
46
15
Age at
presentation
n.a.
n.a.
n.a.
All younger than
3 years
n.a.
Median age
4.5 months
33
cortical destruction can be seen on this MRI as a

loss of hypointensity of the bony margin. These are
clear signs of a bone invasion.
Bifocal tumors without other signs of dissemination have been reported (Warmuth-Metz et al.
2008) and involvement of cranial nerves
(Warmuth-Metz et al. 2008; Parmar et al. 2006;
Wykoff et al. 2008)
Chako et al. (2007) described one case of an
AT/RT coexisting with a pleomorphic xanthoastrocytoma within one single tumor.
Computed Tomography (CT)

Tumors in the spinal cord are rare (Zarovnaya
et al. 2007) and usually cannot be distinguished
by CT or MRI from other primary tumors of the
cord mainly astrocytomas in children. We have
observed one patient with a diffuse meningeal
dissemination without a visible primary tumor.
However, this pattern is not unique in AT/RT but
also encountered in patients with low to high
grade gliomas and PNET/MB.
Meningeal dissemination of an AT/RT is
regarded as prognostically dismal. The incidence
of meningeal dissemination varies but is reported
to be high (see Table 4.1) however the percentage
also depends on the intensity of staging examinations (Parmar et al. 2006). An example of an
infratentorial AT/RT with meningeal dissemination
is given on Fig. 4.3b. Note also the extramidline
position of the tumor, the typical band like
enhancement and the peripheral cyst or necrosis
dorsally. The meningeal disease is covering the
brainstem and can be seen best as linear enhancement on the anterior surface of the normal
brainstem.
Bone involvement is surprisingly frequent (4 of
91 own patients) and was seen in four patients in
recent reports (Arslanoglu et al. 2004; Evans et al.
2001; Heuer et al. 2010; Kazan et al. 2007; WarmuthMetz et al. 2008). As bone invasion is very rare in
other primary brain tumors except gliosarcomas
this seems to be a peculiarity of AT/RT. The frontal
AT/RT given on Fig. 4.2a, b shows a destruction of
the inner tabula of the frontal bone on CT and an
atypically high signal in the calvarial bone overlying the tumor on the coronal T2-weighted MRI. A
As most children are primarily examined by MRI,

descriptions of the typical features of AT/RT on
CT are infrequent. Comparable to other highly
cellular tumors of the CNS AT/RT are characterized
by a high density of their solid parts on unenhanced
CT (Koral et al. 2008). A typical example of the
high density of the solid portions surrounding a
central necrosis of an AT/RT is given von
Fig. 4.1a. A tiny spot of calcification is also visible.
All except three tumors in our database of 20 CT
examinations had increased density compared to
cortex. The remainder was isodens (n = 2) and
hypodense (n = 1). Due to the rapid proliferation
and therefore lack of sufficient energy the tumors
are prone to necrotic changes. The internal
structure especially in larger tumors is more often
inhomogeneous and areas of bleeding, cysts or
necroses are not rare. Calcifications may be visible and are diagnostically not useful. In our
patients we found 7 out of 20 with mainly small
(4 vs. 3) calcifications.
As on MRI contrast enhancement is variable
and usually more often intense than slight. The
enhancement pattern is predominantly inhomogeneous. Most tumors are surrounded by a
perifocal edema.
Bone infiltration and destruction seems to be
more frequent than in other primary tumors of the
CNS but calvarial involvement is usually obvious
on MRI because of gross destruction or even growth
outside the cranial cavity. Subtle changes that could
be detected easier by a bone reconstruction on CT
are unusual (case 1 of Arslanoglu et al. 2004).
34
Fig. 4.1 Right frontal AT/RT with high CT density

reflecting a high cellular density (a). On T2-weighted
MRI (b) high cell density is demonstrated by low signal.
The T1-weighted MRI (c) without contrast shows a small
area of met-hemoglobin after local bleeding dorsally. The

typical pattern of enhancement is seen in the sagittal
T1-weighted image after contrast (d)
Magnetic Resonance Imaging (MRI)
However, the visual evaluation of signal intensities

on MRI is rather subjective and variation is thus
explainable.
On T1-weighted images the signal is usually
mixed with hypointense cysts or necroses and
hyperintense areas of methemoglobin following recent hemorrhage. See Fig. 4.1c with a
linear deposit of methemoglobin dorsally as a
sign of recent bleeding. The amount of
posthemorrhagic signal changes is increased
compared to other highly malignant tumors
(Koral et al. 2008; Warmuth-Metz et al. 2008;
Signalintensities: As a correlate to the high CT

density reflecting a high cellularity of AT/RT the
T2-signal is usually inhomogeneous and either
slightly increased, decreased or isointense to normal cortex. A low signal is shown on Fig. 4.2b in
a frontal AT/RT (same patient as Fig. 4.1a, c, d).
Foci of very low signal intensity possibly representing hemosiderin were found in AT/RT but
also in medulloblastomas and did not show a significantly different frequency (Koral et al. 2008).
35
among the ones with cPNET (65% vs. 36%).

Parmar et al. (2006) found significant perifocal
edema in all of his 11 patients, while Arslanoglu
et al. (2004) reported little if any edema in 4
patients.
Tumor cysts: The incidence of cystic cavitations
within the tumor seems to be higher than in
cPNET and medulloblastomas. While Arslanoglu
et al. (2004) described a higher number of cysts
with enhancement of the cyst walls, Parmar et al.
(2006) reported in only 2 of 11 patients cysts but
in 5 of them necroses. The localization of cysts
between the tumor and the normal brain seems to
be quite characteristic (Arslanoglu et al. 2004;
Warmuth-Metz et al. 2008). The different information in reports may be based on the definition
of a real cyst and its delineation from necrotic
cavities. Koral et al. (2008) who differentiated
between cystic and solid tumors found
a statistically significant difference between
medulloblastomas and AT/RT with more cystic
AT/RT. See Fig. 4.3b showing a peripheral cyst
or area of necrosis delineated by linear enhancing
wall between the normal brain and the tumor.
Fig. 4.2 Bony destruction is obvious on the coronal CT

(a) by a focal loss of the lamina interna of the calvarial
bone. On the coronal T2-weighted MRI (b) the signal
change in the cavarial bone is a sign of the tumor infiltration of the bone marrow and also the cortical bone destruction is visible
Parmar et al. 2006). We evaluated the incidence

of hemorrhage in 23 supratentorial AT/RT and
36 cPNET and found more than twice as many
patients with hemorrhage among patients with
AT/RT (61% vs. 25%).
Edema and tumor borders: The amount and the
presence of a perifocal edema seems to been
more frequent and larger than in cPNET or
medulloblastoma. In half of our own cases the
tumor was surrounded by an edema of a mean
of 1.7 cm. In an unpublished compilation of
supratentorial AT/RT and cPNET the number of
patients with perifocal edema among patients
with AT/RT was twice as high as the number
Tumor size: AT/RT are usually large tumors at

the time of diagnosis. The infratentorial ones are
smaller than the infra- and supratentorially
located or the purely supratentorial tumors. The
mean volume for all tumors was 66 mL with a
mean volume of 35 mL if they were infratentorially located and 82 mL as well for the tumors in
both compartments and the supratentorial ones
(Warmuth-Metz et al. 2008).
Contrast enhancement: Most AT/RT show more
or less intense enhancement after iv-application
of Gadolinium contrast material. The majority of
enhancing tumors demonstrated enhancement in a
high proportion of the tumor volume. But completely missing enhancement does not exclude an
AT/RT (Meyers et al. 2006; Warmuth-Metz et al.
2008). A wavy band of enhancement completely
or partially surrounding a central necrotic area
and frequently composed of multiple small bubbly
lesions was described in a considerable proportion
of AT/RT (12 of 33) irrespective of localization
36
of enhancement (own data). The typical bandlike

enhancement pattern can be seen on Fig. 4.1d in a
right frontal AT/RT and 3b in an extramidline
infratentorial AT/RT.
Diffusion weighted MRI: Signal intensity on diffusion weighted MRI and ADC (apparent diffusion
coefficient) maps are reported to show a restricted
diffusion (Warmuth-Metz et al. 2008) and low ADC
values. See Fig. 4.3a showing the restricted diffusion
on an ADC-image. They are similar to medulloblastomas and do not allow a discrimination (Koral et al.
2008; Rumboldt et al. 2006).
MR-Spectroscopy (MRS): Proton MRS is able
to indentify highly cellular tumors by a high choline component and choline/N-acetylaspartate
(NAA) ratio in the resonance spectrum (Meyers
et al. 2006). Free lipids may be present as a sign
of tumor necrosis (Severino et al. 2010).
Fig. 4.3 ADC-image (a) with very low signal representing high cellular density. The axial T1-weighted MRI in
another patient with an infratentorial AT/RT (b) shows as
well the characteristic enhancement pattern as a peripheral cyst and a meningeal dissemination along the pontine
surface. As well the lateral position of the tumor is in
favor of an AT/RT
Combination of characteristic MR features: If

a more frequently large tumor with features of a
high cellularity irrespective of infra- or supratentorial localization in a preferably young child
shows signs of bleeding, peripheral cysts, an
inhomogeneous internal structure as well on T1
and on T2-weighted MRI and inhomogeneous
enhancement this is highly suspicious of an AT/
RT. Even more suspicious is if a peculiar pattern
of wavy and bandlike enhancement is visible and
if bone destruction is present. In the supratentorial region no specific localization is characteristic but infratentorially the extramidline position
of a tumor or the extension into the supratentorial
compartment is differentiating an AT/RT from
the main differential diagnosis of a medulloblastoma. As well in supra- and infratentorial compartment the presence of a perifocal edema and a
more intense and more complete enhancement is
in favor of an AT/RT and less likely in a PNET/
medulloblastoma.
References
(Warmuth-Metz et al. 2008). This aspect does not
predominate in cPNET or medulloblastomas. In a
comparison of 22 AT/RT and 35 cPNET only one
cPNET but 10 AT/RT showed this peculiar pattern

K, Burger P, Yousem DM (2004) Imaging findings
of CNS atypical teratoid/rhabdoid tumors. AJNR
Am J Neuroradiol 25:476480
Biswas A, Goyal S, Puri T, Das P, Sarkar C, Julka PK,

Bakshi S, Rath GK (2009) Atypical teratoid rhabdoid
tumor of the brain: case series and review of literature.
Childs Nerv Syst 25:14951500
Chako G, Chako AG, Dunham CP, Judkins AR, Biegel
JA, Perry A (2007) Atypical teratoid/rhabdoid tumor
arising in the setting of a pleomorphic xanthoastrocytoma. J Neurooncol 84:217222
Dang T, Vassilyadi M, Michaud J, Jimenez C, Ventureyra
ECG (2003) Atypical teratoid/rhabdoid tumors. Childs
Nerv Syst 19:244248
Evans A, Ganatra R, Morris SJ (2001) Imaging features of
primary malignant rhabdoid tumour of the brain.
Pediatr Radiol 31:631633
Breiger D, Broxon E, Donahue B, Finlay JL,
Miller DC, Packer R, Pucetti D, Radcliff J, Tao ML,
Shiminski-Maher T, Childrenss Cancer Group
(2005) Multiagent chemotherapy and deferred radiotherapy in infants with malignant brain tumors: a
report from the Childrens Cancer Group. J Clin
Oncol 23:76217631
Heuer GG, Kiefer H, Judkins AR, Blasco J, Biegel JA,
Jackson EM, Cohen M, OMalley BW, Strom PB
(2010) Surgical treatment of a clival-C2 atypical teratoid/rhabdoid tumor. J Neurosurg Pediatr 5:7597
(2004) Central nervous system atypical teratoid/rhabdoid tumor: results of therapy in children enrolled in a
registry. J Clin Oncol 22:28772884
Ho DM, Hsu CY, Wong TT, Ling LT, Chiang H (2000)
Atypical teratoid/rhabdoid tumor of the central nervous system: a comparative study with primitive
neuroectodermal
tumor/medulloblastoma.
Acta
Neuropathol 99:482488
Janson K, Nedzi LA, David O, Schorin M, Walsh JW,
Bhattarcharjee M, Pridjian G, Tan Lu, Jukins AR,
Biegel JA (2006) Predisposition to atypical teratoid/
rhabdoid tumor due to an inherited INI1 mutation.
Pediatr Blood Cancer 47:279284
Judkins AR, Eberhart CG, Wesseling P (2007) Atypical
teratoid/rhabdoid tumour. In: Louis DN, Ohgaki H,
Wiestler OD, Cavenee WK (eds) WHO classification
of tumours of the central nervous system. International
Agency for Research on Cancer, Lyon, pp 147149
Kazan S, Goksu E, Mihci E, Gokhan G, Keser I, Gurer I (2007)
Primary atypical teratoid/rhabdoid tumor of the clival
region. Case report. J Neurosurg 106(4 Suppl):308311
Koral K, Gargan L, Bowers DC, Gimi B, Timmins CF,
Weprin B, Rollins NK (2008) Imaging characteristics
of atypical teratoid-rhabdoid tumors in children
compared with medulloblastomas. AJR Am J
Roentgenol 190:809814
Lutterbach J, Liegibel J, Koch D, Madlinger A, Frommhold
H, Pagenstecher A (2001) Atypical teratoid/rhabdoid
tumors in adult patients: case report and review of the
literature. J Neurooncol 52:4956
37
Makuria AT, Rushing EJ, McGrail KM, Hartmann D-P,

Azumi N, Ozdemirli M (2008) Atypical teratoid rhabdoid tumor (AT/RT) in adults: review of four cases.
Meyers SP, Khademian ZP, Biegel JA, Chuang SH, Krones
DN, Zimmerman RA (2006) Primary intracranial
atypical teratoid/rhabdoid tumors of infancy and
childhood: MRI features and patient outcomes. AJNR
Montgomery P, Kuhn JP, Berger PE (1985) Rhabdoid
tumor of the kidney. Urol Radiol 7:4244
Parmar H, Hawkins C, Bouffet E, Rutka J, Shroff M (2006)
Imaging findings in primary intracranial atypical teratoid/rhabdoid tumors. Pediatr Radiol 36:126132
85:5665
(2006) Apparent diffusion coefficients for differentiation of cerebral tumors in children. AJNR Am J
Neuroradiol 27:13621369
Schneppenheim R, Frhwald MC, Gesk S, Hasselblatt M,
Jeibmann A, Kordes U, Kreuz M, Leuschner I, Martin
Subero JL, Obser T, Oyen F, Vater I, Siebert R (2010)
Germline nonsense mutation and somatic inactivation of
SMARCA4/BRG1 in a family with rhabdoid tumor predisposition syndrome. Am J Hum Genet 86:279284
Severino M, Schwartz ES, Thurnher MM, Rydland J,
Nikas I, Rossi A (2010) Congenital tumors of the central nervous system. Neuroradiology 52:531548
Tekautz TM, Fuller CE, Blaney S, Fouladi M, Bronciser
therapy and high-dose alkylator based chemotherapy.
Warmuth-Metz M, Bison B, Dannemann-Stern E, Kortmann
R, Rutkowski S, Pietsch T (2008) CT and MR imaging
in atypical teratoid/rhabdoid tumors of the central nervous system. Neuroradiology 50:447452
Wesseling P, Biegel JA, Eberhart CG, Judkins AR (2007)
Rhabdoid tumor predisposition syndrome. In: Louis
DN, Ohgaki H, Wiestler OD, Cavenee WK (eds) WHO
classification of tumours of the central nervous system. International Agency for Research on Cancer,
Lyon, pp 234235
Wykoff CC, Byron L, Brathwaite CD, Biegel JA,
McKeown CA, Rosenblum MK, Allewelt HB,
Sandberg DI (2008) Atypical teratoid/rhabdoid tumor
arising from the third cranial nerve. J Neuroophthalmol
28:207211
Zarovnaya EL, Pallatroni HF, Hub EB, Ball PA,
Cromwell LD, Pipas JM, Fadul CE, Meyer LP, Park
JP, Biegel JA, Perry A, Rhodes CH (2007) Atypical
teratoid/rhabdoid tumor of the spine in an adult: case
report and review of the literature. J Neurooncol
84:4955

Rhabdoid Tumor: Diagnosis
Using Imaging Techniques
and Histopathology
Fabrice Bing
Contents
Abstract
Introduction ............................................................
40
Clinical Features ....................................................
40
Preoperative Imaging Techniques ........................

CT Scanner...............................................................
Conventional MRI Sequences ..................................
MRI Diffusion ..........................................................
MRI Perfusion..........................................................
MRI Spectroscopy ...................................................
42
42
43
45
45
45
Postoperative Imaging ...........................................
46
Pathologic Findings ................................................

Histopathology .........................................................
Immunohistochemistry
and Cytogenetic Study .............................................
46
47
Discussion and Differential Diagnosis ..................
48
References ...............................................................
50
F. Bing (*)
Neuroradiology unit, University Hospital of Grenoble,
38700 Grenoble Cedex 09, France
e-mail: fabricebing@yahoo.fr
47
Atypical teratoid/rhabdoid tumor (AT/RT) of

the central nervous system is an aggressive
infantile embryonal neoplasm, usually presented as an intra-parenchymatous lesion. On
Computed Tomodensitometry (CT) and
Magnetic Resonance Imaging (MRI), AT/RT
appears as a bulky, heterogeneous lesion,
enhanced after contrast material injection.
Both CT and MRI findings are similar to
those seen in patients with Primitive
Neuroectodermal Tumor/Medulloblastoma
(PNET/medulloblastoma), consisting in the
main differential diagnosis. Very few cases
with diffusion, perfusion and spectroscopy
data have been reported, but it seems that
AT/RTs and PNET/medulloblastomas are not
more distinguishable with more recent MRI
techniques. One major histopathologic
characteristic of AT/RT is the presence of
rhabdoid cells, rarely predominant and not
specific given that AT/RT contains variable
components with primitive neuroectodermal, mesenchymal and epithelial features.
Immunohistochemistry and genetic studies
allow the differentiation between PNET/
medulloblastoma and composite rhabdoid
tumors. A precise diagnosis of AT/RT is
important as this tumor has to be treated with
intensive therapy (surgery, chemotherapy and
radiation therapy). In spite of this treatment,
the pronostic of AT/RT remains very poor.

39
F. Bing
40
Introduction
Malignant rhabdoid tumor was first described in
1978 as an aggressive renal pediatric neoplasm
(Beckwith and Palmer 1978). In 1996, an original
type of central nervous system (CNS) tumor
consisting of rhabdoid cells and resembling a
rhabdoid tumor of the kidney was reported (Rorke
et al. 1996). This new entity also contained areas
of primitive neuroectodermal cells with mesenchymal and epithelial components. It was called
atypical teratoid/rhabdoid tumor (AT/RT). In fact
one major pre-operative imaging differential
diagnosis is the primitive neuroectodermal tumor/
medulloblastoma
(PNET/medulloblastoma),
which represents the most common malignant
CNS tumor in the first decade of life.
Reviewing the literature and according to the
data of our patients, it does not seem possible to
distinguish these two entities with the Computed
Tomodensitometry (CT) scanner and classical
Magnetic Resonance Imaging (MRI) sequences.
Very few AT/RTs have been described with MRI
diffusion, perfusion and spectroscopy, but it
seems that these more recent techniques do not
provide us with specific information.
Even if imaging features of intracranial AT/
RT are non-specific, some of them (signal, localisation and type of enhancement) may be in
favour of this rare diagnosis. The radiologist has
to know this tumoral entity which has often been
misdiagnosed during this last decade probably
because differentiation between AT/RT and
PNET/medulloblastoma was also a challenge for
the anatomopathologist. In fact, AT/RT presents
a distinct histopathological, immunohistochemical and molecular biological pattern that allows
the differential diagnosis. The distinction is
important as AT/RT presents a very poor
prognosis and has to be treated with aggressive
therapies (surgery, radiotherapy and high-dose
chemotherapy). In pre-operative imaging, the
other differential diagnoses are glial tumors such
as malignant astrocytoma, pilocytic astrocytoma,
specially in the cerebellar pontine angle (CPA),
ependymoma and other CNS lesions that may

have rhabdoid cells.
This chapter attempts to present imaging and
pathologic features of intracranial AT/RTs. Based
on our own experience of a few cases (Table 5.1)
and those in literature, we will also discuss the
interest of the different MRI sequences in preoperative imaging and present the main differential
diagnoses of intracranial AT/RT.
Clinical Features
Atypical teratoid and rhabdoid tumor is a tumor
of infancy. In a study of 52 children, (Rorke et al.
1996) reported a mean age of 29 months at diagnosis. Atypical teratoid and rhabdoid tumor has
rarely been described in utero or in adults (Horn
et al. 1992; Rorke et al. 1996). On the contrary,
the age at presentation of PNET/medulloblastoma tumors is about 57 years of age. Even if
PNET/medulloblastoma is more frequent, AT/RT
must be systematically evoked in front of an
aggressive intracranial tumor, sustentorial or
located in the posterior fossa, in a less than
3-year-old child.
One major challenge for the consultant is to
reduce the delay between the primary symptoms
and the diagnosis, as pediatric brain tumors very
frequently present with non specific signs. For
example, one of our patients, a 8-month-old boy,
presented only with digestive signs and a gastro
esophageal transit was performed in first intention. Clinical presentation depends on the age of
onset and the location of the tumor. It remains
difficult to know when cerebral imaging is appropriate. Children less than 3 years of age present
with non specific symptoms as vomiting, lethargy
and failure to thrive. In the first year of life, tens
fontanelle and enlarged head circumference are
described. Hydrocephalus is often present when
the lesion is located in the posterior fossa. Focal
neurologic deficit and seizures can reveal a sustentorial lesion. Older patients may suffer from
headache. Cranial nerve palsies are common
(Rorke et al. 1996).
Headache
Vomiting
Lethargy
Vomiting
4 (F, 10 years)
5 (M, 8 months)
Hyper
Cerebellum
Hyper
ParietoNA
occipital and
intraventricular
Vermis
NA
Cerebellar
pedicle
Pont
Vermis
NA
Dural-based
mass
Vermis
Location
Sustentorial
CT
NE
Hyper
NA not available, NE no enhanced, CE contrast enhanced
Lethargy
Lethargy
Seizures
Fig. 5.2
3 (F, 6 years)
Fig. 5.3
6 (M, 8 months)
Vomiting
2 (M, 3 months)
Fig. 5.1
Patient no. (sex,

Clinical
age in months)
signs
1 (M, 22 months) Seizures
Table 5.1 Presentation of our own cases
NA
NA
NA
NA
CE
++
Hyper
T2-W
Iso
++
Gd-T1
++
Hyper
Hyper
Iso and Iso

hyper
Hypo
Hypo
++
++
Iso and Iso and +

hyper
hyper
Hypo
MRI
T1-W
Hyper
No
No
No
No
Yes
Calcification
No
Yes
Yes
Yes
Yes
Yes
Cysts
Yes
No
No
No
No
No
Edema
Yes
No
No
No
No
Yes (6 months)
Meningeal dissemination
(at initial diagnosis or
follow-up in months)
No
Recurrence
(5 months)
Died (6 months)
Died
(3.5 months)
No recurrence
(96 months)
Follow-up MR
(in months after
initial diagnosis)
Recurrence
(12 months)
Died
(14 months)
Recurrence
(8 months)
Died (9 months)
No recurrence
(72 months)
5
Pediatric Atypical Teratoid/Rhabdoid Tumor: Diagnosis Using Imaging Techniques and Histopathology
41
42
F. Bing
Fig. 5.1 (Patient 6) Preoperative axial non-enhanced CT

scanner (a), contrast-enhanced coronal T1-weighted image
(b), axial diffusion weighted image (c), axial perfusion
MRI (d). Photograph showing diffuse loss of nuclear INI1
expression in large neoplastic cells (e). This extra-axial
lesion presents as a bulky mass, spontaneously hyperdense.

On MRI, there is a heterogeneous enhancement. The cystic
portion is seen in periphery of the tumor (b). There is also
a restriction of the diffusion (c). On perfusion, there is an
elevated cerebral volume (rCBVmax = 4.33)
Preoperative Imaging Techniques
Primitive neuroepithelial tumors are also hyperdense for the same reason. Some authors reported
hemorrhage or calcifications, which can contribute to the spontaneous hyperdensity and heterogeneity (Arslanoglu et al. 2004; Rorke et al. 1996).
Aggressive features are one major characteristic
of AT/RTs: hydocephalus, apparent invasion of
the adjacent brain or dura and marked mass effect
are often reported. These signs are also described
with other aggressive brain tumors, as PNET or
multiform glioblastoma.
The lesion can be solid or can present with
cysts. When cysts are described, they are usually
CT Scanner
Non enhanced CT scan (NECT) and contrast
enhanced by CT scan (CECT) features of AT/RT
have been largely reported in previous studies. No
specific signs have been noticed. Atypical teratoid
and rhabdoid tumor presents as a heterogenous
mass, hyperdense on NECT, which is due to the
high cellular density of the tumor. In our experience, NECT performed for three of our patients
showed a spontaneous hyperdense mass (Fig. 5.1).
43
Fig. 5.2 (Patient 2) Preoperative sagittal T1 SE MRI

without (a) and with (b) gadolinium injection axial T2
SE (c) and spectroscopy (d). This vermian lesion is
heterogeneous on T2 and presents a heterogeneous
enhancement. Hydrocephalus is present. On spectroscopy, there is an elevated peak of lipids. 8 months after
surgery, the patient presented pineal and leptomeningeal metastases (e)
eccentrics, which is rather rare with PNET/

medulloblastoma (Cheng et al. 2005). On CECT,
the lesion presents heterogeneous enhancement
and the periphery of the cysts is classically
enhanced. Despite the aggressive characteristics
of this bulky mass, there may be no oedema
around it. Oedema was present in 52% of the
cases in one study (Warmuth-Metz et al. 2008).
ing radiation. Commonly used MRI sequences

for the exploration of an intracranial tumor are
T1 spin-echo (SE), T2 SE, T1 SE with gadolinium, T2 echo-planar (EP) and fluid-attenuated
inversion recovery (FLAIR). Aggressive patterns
are almost always present: AT/RT appears as a
bulky mass, with heterogeneous signal on T1 SE
and T2 SE (Figs. 5.1, 5.2, and 5.3). In the posterior fossa, the lesion is more often located in the
cerebellum (57%) (Rorke et al. 1996) whereas
medulloblastoma has a tendency to occur in midline. However, in our cases, the three lesions in
the posterior fossa were located in the vermis.
Extraaxial lesions concern the CPA and can
Conventional MRI Sequences

In comparison to CT, MRI provides superior
delineation of the extent of tumor without ioniz-
44
F. Bing
Fig. 5.3 (Patient 3) Preoperative sagittal T1 SE MRI without (a) and with (b) gadolinium injection, axial T2 SE (c).
This parieto-occipital and intraventricular lesion presents a
spontaneous hyperintensity on T1 SE image, corresponding
to hemorrhage. The lesion appears heterogeneous on T2 SE

image and presents a heterogeneous enhancement. Six years
after the initial diagnosis, no recurrence is notified, which is
quite exceptional with AT/RT (axial T2 SE (d))
invade the internal auditory canal (Parmar et al.

2006). In a retrospective study of 55 patients
comparing AT/RT and medulloblastoma, CPA
involvement and intratumoral hemorrhage were
more common in AT/RT (Koral et al. 2008).
When sustentorial, the lesion concerns the hemispheres, the pineal region and rarely the suprasellar region. One of our patients presented with
a unique extraaxial mass with a pseudomeningiomatous aspect (Fig. 5.1) (Bing et al. 2009).
The lesion can also concern the lateral ventricle
(Fig. 5.3). One major differential diagnosis in
that location is the choroid plexus carcinoma,
which can also present rhabdoid features.
In AT/RT cystic components are frequently
noticed. Hyperintense signal on T1 SE corresponds
to hemorrhage, which is confirmed on T2 EG,
showing hypointense signals. Hyperintensities
on T1 SE rarely correspond to microcalcifications.
The FLAIR sequence evaluates the oedema
around the tumor. Brain tumoral infiltration can

also be confounded with this mixed oedema,
vasogenic and cytotoxic. Only one of our cases
presented clear oedema.
Contrast enhancement is common but can be
missing (Warmuth-Metz et al. 2008). When present, the enhancement is heterogeneous and concerns the limits of the cysts, as in our cases.
(Warmuth-Metz et al. 2008) noticed an enhancement pattern consisting of a band-like rim of
strong enhancement. This pattern has not been
described in other studies. It is important to look
for a leptomeningeal dissemination, sometimes
to be found at the moment of the diagnosis. A
disseminated tumor in the leptomeninges at initial
staging was reported in 24% of patients (Meyers
et al. 2006). Leptomeningeal spread is also a
well-known characteristic of medulloblastoma.
One of our patients presented a recurrence with
leptomeningeal and pineal metastases 6 months
after the first MRI (Fig. 5.2e). Primary diffuse

cerebral leptomeningeal AT/RT is exceptional
(El-Nabbout et al. 2010).
These non specific signs are common in other
aggressive primitive brain tumors. Primitive neuroepithelial tumor and multiform glioblastoma
can give extracranial metastases, which is very
rare with AT/RT. In fact, the evolution of AT/RT
is so fast that we can assume that there is not
enough time for the tumor to present an extracranial extension.
MRI Diffusion
Diffusion MR imaging evaluates the microscopic
water diffusion within tissues. Apparent diffusion coefficient (ADC) maps are calculated and
represent an absolute measure of average
diffusion for each voxel. Diffusion weightedimage (DWI) and ADC maps are commonly used
in stroke lesions and to differentiate brain lesions.
In tumoral lesions, a restriction of the diffusion
has been reported when the tumor presents a high
cellular density and an important nuclear area.
In AT/RT, few studies reported a restricted diffusion (Gauvain et al. 2001; Koral et al. 2008;
Meyers et al. 2006; Rumboldt et al. 2006). This
technique has been used to discriminate the different cerebellar tumors in children. Pilocytic
astrocytomas present the highest values after
ependymomas and medullolastomas, and it seems
that there is a link between the grade of the tumor
and the intensity of the signal on DWI. In fact,
medulloblastoma and AT/RT are the two pediatric posterior fossa tumors showing restricted diffusion and their ADC values similar (Koral et al.
2008). Both of our patients explored with DWI
showed a restriction of diffusion (Fig. 5.1c).
MRI Perfusion
MRI perfusion is a dynamic, susceptibilityweighted, contrast-enhanced sequence (DSC)
that provides in vivo assessment of the microvasculature in intracranial brain tumors. This
sequence gives important physiological information concerning the vascularity of the tumor:
45
elevated relative cerebral blood volume (rCBV)

corresponds to areas of microvascular density
and vascular endothelial hyperplasia (Cha et al.
2003) and rCBV values correlate with tumor
grade. Blood brain-barrier disruption can be the
consequence of neoangiogenesis and/or the
presence of permeability factors. DSC MR
imaging has been studied to evaluate vasculature anomalies in glial tumors and it has been
shown that rCBV values correlate with tumor
grade (Aronen et al. 1994).
There is not enough data concerning DSC MR
imaging on AT/RT in the literature. Concerning
PNET/medulloblastoma, rCBV values are the
same as in high grade glioma. It has been shown
that there was no difference between high grade
gliomas and PNET rCBVs (Law et al. 2004). The
perfusion study of the extraaxial sustentorial
dural based AT/RT could have corresponded to
an aggressive meningioma or a PNET (Fig. 5.1d).
In that case, DSC MR imaging revealed a in the
tumor maximal relative cerebral blood volume of
4.33, which is different from the higher maximal
rCBV of meningioma (9.1 4.4 S.D.) (Yang
et al. 2003). However, malignant meningioma
may present with a lower maximal rCBV (5.89
3.86 S.D.) (Zhang et al. 2008) and PNET rCBV
values can be similar.
MRI Spectroscopy
In vivo proton magnetic resonance spectroscopy
(MRS) brings biochemical informations on
tumoral tissues. These additional data may be
useful for preoperative differentiation of brain
tumors (Arle et al. 1997; Kugel et al. 1992; Wang
et al. 1995). If spectral characteristics of the most
common brain tumors are well-known, rare
tumors as AT/RT and even PNET have been
poorly explored in MRS. Concerning pediatric
tumors in the posterior fossa, it is possible to distinguish between medulloblastoma, ependymoma
and astrocytoma with MRS (Arle et al. 1997).
Primitive neuroepithelial tumors MRS spectrum
is characterised by elevated relative choline values, that may correlate with the high cellularity
of the tumor (Becker 1999). In an adult study
(Majos et al. 2002), no lipids were found in this
F. Bing
46
tumor, contrary to other aggressive tumors: it can

be explained by their high cellularity and a relative low amount of intratumoral necrosis, but lipids have also been reported (Jouanneau et al.
2006) and significant necrosis has been noticed
in a cytopathologic study (Parwani et al. 2005). A
pic of myoinositol can be described and PNET is
characterised by a Tau peak.
Concerning AT/RT, very few MRS studies
have been reported. Atypical tratoid and rhabdoid tumor may present the same metabolic
spectrum as PNET: elevated levels of choline
and decreased N-acetylaspartate (NAA). Our
patient with the intraaxial lesion presented this
spectrum, and also an elevated lipid peak
(Fig. 5.2d). When the lesion is extraaxial, one
differential diagnosis may be the meningioma,
even if this entity is rarely described in infants.
In meningioma, proton MR spectrum shows
elevated choline peak, low level of creatine,
glutamate-glutamine signal and often alanine
resonance. Concerning our dural-based lesion,
we did not find peaks of myoinositol and alanine
described with primitive neuroectodermal tumor
and meningioma, respectively (Bing et al. 2009).
There was also a peak of lipids.
perform the same MRI sequences for each

imaging follow-up. T1 SE sequences are done
without and with gadolinium injection, as hyperintensities can correspond to postoperative hemorrhage, presenting the same aspects as residual
tumoral enhancement. Leptomeningeal dissemination has to be systematically searched (Fig.
5.2). In a retrospective study (Meyers et al.
2006), it occurred in 35% of patients presenting
with initial postoperative negative findings. The
median survival rate of patients with initial or
follow-up disseminated tumor in the leptomeninges was dramatically reduced. On the other
hand, postoperative meningeal enhancement,
corresponding to transient changes, is commonly
observed and has to be controlled, considering
the agressivity of AT/RT.
Brains immaturity limits the use of radiotherapy and is only in case of progressive
lesions. Chemotherapy, alone or with radiotherapy, may initially enlarge some tumors thats
why modalities of treatments have to be known
for the interpretation and comparison of imaging (Burger et al. 1998).
Pathologic Findings
Postoperative Imaging
Intensive therapy associating surgery, radiotherapy and chemotherapy is most often proposed.
However, the best treatment for this aggressive
lesion remains unknown and regular follow-up
imaging controls are mandatory. Poor pronostic
factors are young age at diagnosis, presence of
leptomeningeal disssemination and absence of
gross total tumoral resection (Meyers et al. 2006).
The mean postoperative survival of patients is
only 11 months (Burger et al. 1998).
Hence, the aim of the first postoperative MRI
is to confirm the total resection of the tumor.
Whatever the pathologic findings are, the first
postoperative imaging has to be perfomed 4872
hours after surgery. If performed too late, cicatricial tissue may appear as nodular perilesional
enhancement, which may be confounded with a
residual lesion. Moreover, it is a necessity to
The histogenesis of AT/RTs is uncertain.

Rhabdoid cells may derive from pluripotent fetal
cells. As AT/RT and choroid plexus carcinoma
present phenotypic and genotypic overlaps, these
tumors could both derive from a common progenitor (Gessi et al. 2003). Atypical teratoid and
rhabdoid tumor can exceptionally occur in the
setting of other tumors, as it has been reported
with a ganglioglioma (Allen et al. 2006) and a
pleomorphic xanthastrocytoma (Chacko et al.
2007). These cases suggest that rhabdoid cells
may arise from a primitive cell with the capacity
to diverge along multiple differentiation pathways (Allen et al. 2006).
Neither clinical signs nor imaging features can
distinguish AT/RT from PNET. The definitive
diagnosis is always given by the anatomopathologist thanks to histopathologic, immunohistopathologic and molecular genetic studies.
Histopathology
Atypical teratoid and rhabdoid tumor is a hypercellular tumor, presenting a wide range of different histopathologic patterns: it contains neets or
sheets of rhabdoid cells, varying percentage of
PNET cells (causing erroneous diagnosis of AT/
RT as PNET in the past) (Biegel et al. 2000;
Rorke et al. 1996), mesenchymal spindle-shaped
and epithelial-type tumor cells (Burger et al.
1998). However, germ cells and tissue differentiation associated with malignant teratomas are not
described with AT/RTs. Mitotic figures, necrotic
foci, hemorrhage and ill-defined margins with
adjacent brain or dura are commonly seen.
Tumoral cells present a high proliferative activity
with Ki-67/MIB-1 labeling indices often more
than 50%.
One major pathologic characteristic of malignant rhabdoid tumors (renal and extrarenal forms)
and AT/RT is the presence of rhabdoid cells. It is
now clear that these rhabdoid cells can also
constitute the secondary phenotype of various neoplasms (carcinoma, sarcoma, melanoma, meningioma, neuroblastoma, glioma and desmoplastic
small round cell tumor), called composite rhabdoid
tumors (Perry et al. 2005). Even if rhabdoid cells
forming a distinct entity or a secondary phenotype
may not share the same molecular mechanism
involved in their formation, all of them present with
eccentrically positioned, large and vesicular nuclei,
prominent nucleoli and densely eosinophilic cytoplasm (Edgar and Rosenblum 2008; Rorke et al.
1996). Globular and fibrillar paranuclear inclusions
are also described and correspond ultrastructurally
to whorled bundles of intermediate filaments.
Rhabdoid cells can present with variable morphological features: embracing cells may be more
common, corresponding to sickle-shaped cells having dark nuclei and scanty cytoplasm. Other modified rhabdoid cells are large, pale polygonal cells
having clear and round nuclei with prominent
nucleoli and pale, granular eosinophilic cytoplasm.
The rhabdoid component is usually separate from
the other elements. Teratoid components correspond to poorly differentiated neuroepithelial elements of small cell type, commonly seen (Lee et al.
2002): these cells present basophilic nuclei with
47
dispersed chromatin, small nucleoli and sparse

cytoplasm. One difficulty is that only 1325% of
AT/RT are composed solely of rhabdoid cells
(Burger et al. 1998; Rorke et al. 1996) and the neuroepithelial cells are very similar to the cells of
PNET/medulloblastoma and pineoblastoma.
Mesenchymal cells are characterised by a highly
cellular growth of tumor cells resembling a fibrosarcoma. These cells can be predominant (Rorke
et al. 1996). Epithelial cells may be poorly differentiated, showing papillary, adenomatous or
squamous patterns. When the lesion is situated next
to the ventricles and presents with extensive epithelial differentiation, the major differential diagnosis
is the choroid plexus carcinoma.
In the 52 AT/RTs reported by Rorke, 67% of
the tumors contain areas also described in PNET.
Hence we need for the final diagnosis the use of
immunohistochemical techniques, and more
recently, genetic assessment.
and Cytogenetic Study
Atypical teratoid and rhabdoid tumor shows striking polyphenotypic immunoreactivity. In the
rhabdoid cells, epithelial membrane antigen
(EMA) and vimentin (VMA) are always expressed.
Smooth-muscle actin (SMA) is expressed in
8397% of them (Meyers et al. 2006; Rorke et al.
1996). The two primary neural antibodies, glial
fibrillary acid protein (GFAP) and neurofilament
protein (NFP) and keratin may also be positive in
rhabdoid cells. Variable immunoreactivities for
synaptophysin, S-100 protein and desmin are
present. Primitive neuroepithelial tumors elements can also express GFAP and NFP, as well as
desmin. Mesenchymal fields can express vimentin and desmin. Epithelial cells express keratin
and less frequently EMA. No tumoral cells in AT/
RT express germ cell markers.
The genetic hallmark of AT/RT is the mutation or loss of hSNF5/INI1/SMARCB1/BAF47, a
tumor suppressor gene on chromosome 22q11.2.
The alteration of this gene, abbreviated INI1, can
occur sporadically or in a setting of heritable
genetic
predisposition
(rhabdoid
tumor
48
predisposition syndrome), characterised by

potentially multifocal, neural and extra-neural
tumors presenting in the first year of life. All the
tumoral cells in AT/RT present a loss of nuclear
expression of INI1 (Fig. 5.1e). It results from partial chromosal deletion or monosomy 22. A nonsense or frameshift mutation resulting in a stop
codon affect the remaining gene.
It has been demonstrated by cytogenetic and
molecular analyses that most rhabdoid tumors
present an inactivation of INI1, whereas it is
exceptional in composite rhabdoid tumors (Perry
et al. 2005). Primitive neuroepithelial tumors,
medulloblastomas and choroid plexus carcinomas
may exhibit INI1 mutations and some authors
proposed that in that cases, these entities should
be considered as rhabdoid tumors (Biegel et al.
2002). In a molecular analysis of 126 meningiomas, 4 (3%) carried a mutation in the INI1gene,
suggesting that INI1 may also be involved in the
pathogenesis of meningiomas (Schmitz et al.
2001). On the other hand, 1520% of tumors
exhibiting typical morphologic and immunophenotypic features of AT/RTs do not present INI1
deletion or mutation (Biegel 2006). Moreover, the
gene INI1 can be intact with a decreased expression. In fact, immunostaining for INI1, using a
monoclonal antibody to the INI1 protein (BAF47),
is a more sensitive adjunct to the diagnosis of AT/
RT and has to be performed when lesions present
with overlapping histologic and immunohistochemical phenotypes. This technique has been
used to rectify initial diagnosis of choroid plexus
carcinoma in AT/RT (Judkins et al. 2005).
Discussion and Differential Diagnosis

Primary AT/RTs are rare aggressive embryonal
neoplasms of childhood in the CNS. They represent 1.3% of primary CNS tumors in the pediatric
population and 6.7% of central nervous system
neoplasms in children less than 3 years of age
(Rickert and Paulus 2001). Tumor localization is
more often reported as infratentorial and intraaxial (Meyers et al. 2006). Extra-axial lesions are
usually situated in the cerebellopontine angle.
Supratentorial tumors exhibit the same localiza-
F. Bing
tions as primitive neuroectodermal tumors.

Primary spinal AT/RT is extremely rare. In that
location, MRI findings are also similar to PNET
(Kodama et al. 2007). In the clival region, AT/RT
does present as a lytic lesion, mimicking other
aggressive tumors as chordoma, Ewing sarcoma,
neuroblastoma, Langerhans cell histiocytosis or
lymphoma (Kazan et al. 2007).
CT scanner and MRI findings are nonspecific
and show a bulky and heterogeneous
contrast-enhanced mass with cystic and necrotic
areas. The lesion appears hyperdense according
to CT scanner, corresponding to high cellular
density. In our patients, neither calcification nor
hemorrhage were evident upon histopathologic
study, except for one of them. An unusual pattern
of contrast enhancement, corresponding to wavy
band-like enhancing zones surrounding central
cystic or necrotic areas, has been reported
(Warmuth-Metz et al. 2008). This pattern was not
present in our patients, but all the lesions presented cysts and solid tumoral tissue, a regular
finding in AT/RTs. Atypical teratoid and rhabdoid tumors present hyperintense signal intensity
on diffusion-weighted imaging, with hypointense
signal intensity on ADC images, indicating
restricted diffusion (Meyers et al. 2006). This
pattern can be explained by the high cellular
density of the tumor. The two lesions studied in
diffusion presented a hypointense signal on ADC
images. In spectrum single-voxel proton magnetic resonance spectroscopy, intra-axial lesions
reveal elevated levels of choline and decreased
levels of N-acetyl-aspartate. We also found elevated levels of choline and lipids, indicating
aggressive lesion, in both an extraaxial and an
intraaxial AT/RT.
In children, the first main differential diagnosis consists of PNET/medulloblastoma.
Distinguishing PNET/medulloblastoma from
AT/RT is of clinical significance. In fact,
patients with AT/RT are younger and present a
worse prognosis than those with PNETs/
medulloblastomas. These tumors can present
the same CT and MRI findings (Meyers et al.
2006). It does not seem possible to differentiate
AT/RTs from PNETs/medulloblastomas with
diffusion-weighted imaging (Gauvain et al.
2001), perfusion magnetic resonance imaging

(Law et al. 2004), or spectroscopy (Majos et al.
2002). The hyperintensity in diffusion-weighted
imaging is explained by the hypercellularity of
both AT/RTs and PNETs/medulloblastomas.
Perfusion data in the literature concerning
PNETs/medulloblastomas are rare. The rCBV
in MRI perfusion and the spectroscopic spectrum of PNETs/medulloblastomas may be the
same as for AT/RTs, revealing aggressive signs
with an increased rCBV in perfusion and an
elevated peak of choline with decreased creatine
and a peak of lipid in SRM. The histopathologic
differential diagnosis of an AT/RT also focuses
on PNET/medulloblastoma and may be
difficult, especially in a biopsy specimen.
Indeed, 70% of AT/RTs contain histologic fields
indistinguishable from PNETs/medulloblastomas, which can cause confusion in diagnosis
(Rorke et al. 1996). Moreover, PNET/medulloblastoma without rhabdoid features but with a
loss of INI1 expression may be similar to AT/
RTs in biopsy material (Biegel et al. 2000;
Edgar and Rosenblum 2008). We emphasize
that the age of the patient, under 3 year of age
favored the diagnosis of AT/RT rather than MRI
features.
The second main differential diagnosis is the
ependymoma, a WHO grade II glial tumor, constituting almost one-third of all brain tumors in
patients younger than 3 years. Sixty percent of
ependymomas are located in the posterior
fossa and concern most often children in
that location. Infratentorial and supratentorial
ependymomas have the same radiologic characteristics: they present as calcified, hemorrhaged
and cystic lesions with avid enhancement. High
cellularity explains the low apparent diffusion
coefficient. Ependymomas, PNETs/medulloblastomas and also AT/RTs present the same characteristics in MRI perfusion and MRS. Infratentorial
ependymoma is characterised by its propension
to infiltrate the foramen of Lushka and extend in
the cerebellopontine angle. Medulloblastomas
tend to be more aggressive, destroying neural
structures rather than insinuate around vessels
and cranial nerves as do ependymomas (Yuh
et al. 2009). Histologically, ependymomas pres-
49
ent rare mitotic cells, organized in perivascular

pseudorosettes and ependymal rosettes.
The third differential diagnosis in children
is the pilocytic astrocytoma, specially when
the lesion is situated in the cerebellopontine
angle. In fact this benign glial tumor is atypical
as it is a hypervascular tumor with low biologic
activity. The classical presentation is a cystic
lesion with an intensely enhancing mural
nodule but pilocytic astrocytoma can present
either as a necrotic or a solid mass with minimal cystic component, which can look like an
aggressive lesion. Histologically, pilocytic
astrocytomas are characterised by a biphasic
pattern composed of loose glial tissue and
piloid tissue. Calcifications and hemorrhage
are uncommon. The important enhancement is
explained by the presence of hyalinized and
glomeruloid vessels.
The two differential diagnoses evoked in CT
and conventional MRI of the dural-based lesion
(Fig. 5.1) are PNET and meningioma, even if the
latter is extremely uncommon in infancy. In proton MRS, we did not find peaks of myoinositol
and alanine described with primitive neuroectodermal tumor and meningioma, respectively.
Rhabdoid cells are not specific of AT/RTs: they
can be identified in rhabdoid meningioma, rhabdoid glioblastoma, medullomyoblastoma, and
rhabdomyosarcoma. These tumors are called
rhabdoid composite tumors. However, in our
extraaxial dural based tumor, the absence of meningioma features and the polyphenotypic immunoprofile with a loss of INI1 expression excluded
the diagnosis of rhabdoid meningioma. Indeed,
classic rhabdoid meningiomas do not lose INI1
expression (Perry et al. 2005).
In conclusion, a metabolic study using diffusion weighted imaging, perfusion magnetic resonance imaging and proton-magnetic resonance
spectroscopy contributed to a better preoperative
diagnosis of a pediatric intracranial tumor. In the
presence of an aggressive intracranial tumor, a
diagnosis of atypical teratoid/rhabdoid tumor
must always be evoked in children under 3 year
of age, even if the localization is unusual. The
main differential diagnosis on preoperative
imaging remains primitive neuroectodermal
50
tumors/medulloblastoma and distinguishing

these two tumors may also be a challenge for the
neuropathologist. However, immunohistochemical and genetic studies allow a precise diagnosis
when the lesions present phenotypic histologic
overlap.
Acknowledgments I am grateful to Doctor Caroline
Salon, neuropathologist, Grenoble University Hospital,
for her valuable suggestions and help for this write-up.
References
Allen JC, Judkins AR, Rosenblum MK, Biegel JA (2006)
Atypical teratoid/rhabdoid tumor evolving from an
optic pathway ganglioglioma: case study. Neuro Oncol
8:7982
Arle JE, Morriss C, Wang ZJ, Zimmerman RA, Phillips PG,
Sutton LN (1997) Prediction of posterior fossa tumor
type in children by means of magnetic resonance
image properties, spectroscopy, and neural networks.
J Neurosurg 86:755761
Aronen HJ, Gazit IE, Louis DN, Buchbinder BR,
Pardo FS, Weisskoff RM, Harsh GR, Cosgrove GR,
Halpern EF, Hochberg FH et al (1994) Cerebral blood
volume maps of gliomas: comparison with tumor
grade and histologic findings. Radiology 191:4151
K, Burger PC, Yousem DM (2004) Imaging findings
of CNS atypical teratoid/rhabdoid tumors. AJNR Am
J Neuroradiol 25:476480
Becker LE (1999) Pathology of pediatric brain tumors.
Neuroimaging Clin N Am 9:671690
Beckwith JB, Palmer NF (1978) Histopathology and
prognosis of Wilms tumors: results from the
First National Wilms Tumor Study. Cancer 41:
19371948
Biegel JA (2006) Molecular genetics of atypical teratoid/
rhabdoid tumor. Neurosurg Focus 20:E11
6:27592763
Biegel JA, Kalpana G, Knudsen ES, Packer RJ, Roberts
CW, Thiele CJ, Weissman B, Smith M (2002) The role
of INI1 and the SWI/SNF complex in the development
of rhabdoid tumors: meeting summary from the workshop on childhood atypical teratoid/rhabdoid tumors.
Cancer Res 62:323328
Bing F, Nugues F, Grand S, Bessou P, Salon C (2009)
Primary intracranial extra-axial and supratentorial
atypical rhabdoid tumor. Pediatr Neurol 41:453456
F. Bing
Atypical teratoid/rhabdoid tumor of the central nervous system: a highly malignant tumor of infancy and
Pediatric Oncology Group study. Am J Surg Pathol
22:10831092
Cha S, Johnson G, Wadghiri YZ, Jin O, Babb J, Zagzag D,
Turnbull DH (2003) Dynamic, contrast-enhanced perfusion MRI in mouse gliomas: correlation with histopathology. Magn Reson Med 49:848855
Chacko G, Chacko AG, Dunham CP, Judkins AR, Biegel
JA, Perry A (2007) Atypical teratoid/rhabdoid tumor
arising in the setting of a pleomorphic xanthoastrocytoma. J Neurooncol 84:217222
Chang CY, Wong TT, Ho DM (2005) Neuroradiological
findings in atypical teratoid/rhabdoid tumor of the
central nervous system. Acta Radiol 46:8996
Edgar MA, Rosenblum MK (2008) The differential diagnosis of central nervous system tumors: a critical
examination of some recent immunohistochemical
applications. Arch Pathol Lab Med 132:500509
El-Nabbout B, Shbarou R, Glasier CM, Saad AG (2010)
Primary diffuse cerebral leptomeningeal atypical teratoid rhabdoid tumor: report of the first case. J
Neurooncol 98:431434
Gauvain KM, McKinstry RC, Mukherjee P, Perry A, Neil
JJ, Kaufman BA, Hayashi RJ (2001) Evaluating pediatric brain tumor cellularity with diffusion-tensor
imaging. AJR Am J Roentgenol 177:449454
Gessi M, Giangaspero F, Pietsch T (2003) Atypical teratoid/
rhabdoid tumors and choroid plexus tumors: when
genetics surprise pathology. Brain Pathol 13:409414
Horn M, Schlote W, Lerch KD, Steudel WI, Harms D,
Thomas E (1992) Malignant rhabdoid tumor: primary
intracranial manifestation in an adult. Acta Neuropathol
83:445448
Jouanneau E, Guzman Tovar RA, Desuzinges C, Frappaz
D, Louis-Tisserand G, Sunyach MP, Jouvet A, Sindou
M (2006) Very late frontal relapse of medulloblastoma
mimicking a meningioma in an adult: usefulness of 1H
magnetic resonance spectroscopy and diffusionperfusion magnetic resonance imaging for preoperative diagnosis: case report. Neurosurgery 58:E789,
discussion E789
Judkins AR, Burger PC, Hamilton RL, KleinschmidtDeMasters B, Perry A, Pomeroy SL, Rosenblum MK,
Yachnis AT, Zhou H, Rorke LB, Biegel JA (2005)
INI1 protein expression distinguishes atypical teratoid/rhabdoid tumor from choroid plexus carcinoma.
J Neuropathol Exp Neurol 64:391397
Kazan S, Goksu E, Mihci E, Gokhan G, Keser I, Gurer I
(2007) Primary atypical teratoid/rhabdoid tumor of the
clival region. Case report. J Neurosurg 106:308311
Kodama H, Maeda M, Imai H, Matsubara T, Taki W,
Takeda K (2007) MRI of primary spinal atypical teratoid/rhabdoid tumor: a case report and literature
review. J Neurooncol 84:213216
Koral K, Gargan L, Bowers DC, Gimi B, Timmons CF,
Weprin B, Rollins NK (2008) Imaging characteristics of
atypical teratoid-rhabdoid tumor in children compared
with medulloblastoma. AJR Am J Roentgenol 190:

809814
Kugel H, Heindel W, Ernestus RI, Bunke J, du Mesnil R,
Friedmann G (1992) Human brain tumors: spectral
patterns detected with localized H-1 MR spectroscopy.
Radiology 183:701709
Law M, Kazmi K, Wetzel S, Wang E, Iacob C, Zagzag D,
Golfinos JG, Johnson G (2004) Dynamic susceptibility contrast-enhanced perfusion and conventional MR
imaging findings for adult patients with cerebral primitive neuroectodermal tumors. AJNR Am J Neuroradiol
25:9971005
Lee MC, Park SK, Lim JS, Jung S, Kim JH, Woo YJ,
Lee JS, Kim HI, Jeong MJ, Choi HY (2002) Atypical
teratoid/rhabdoid tumor of the central nervous
system: clinico-pathological study. Neuropathology
22:252260
Majos C, Alonso J, Aguilera C, Serrallonga M, Acebes JJ,
Arus C, Gili J (2002) Adult primitive neuroectodermal
tumor: proton MR spectroscopic findings with possible application for differential diagnosis. Radiology
225:556566
Meyers SP, Khademian ZP, Biegel JA, Chuang SH,
Korones DN, Zimmerman RA (2006) Primary intracranial atypical teratoid/rhabdoid tumors of infancy and
childhood: MRI features and patient outcomes. AJNR
Parmar H, Hawkins C, Bouffet E, Rutka J, Shroff M
(2006) Imaging findings in primary intracranial atypical teratoid/rhabdoid tumors. Pediatr Radiol
36:126132
diagnosis. Cancer 105:6570
Perry A, Fuller CE, Judkins AR, Dehner LP, Biegel JA
(2005) INI1 expression is retained in composite rhab-
51
doid tumors, including rhabdoid meningiomas. Mod

Pathol 18:951958
Rickert CH, Paulus W (2001) Epidemiology of central
nervous system tumors in childhood and adolescence
based on the new WHO classification. Childs Nerv
Syst 17:503511
Rorke LB, Packer RJ, Biegel JA (1996) Central nervous system atypical teratoid/rhabdoid tumors of infancy and
childhood: definition of an entity. J Neurosurg 85:5665
(2006) Apparent diffusion coefficients for differentiation of cerebellar tumors in children. AJNR Am J
Schmitz U, Mueller W, Weber M, Sevenet N, Delattre O,
von Deimling A (2001) INI1 mutations in meningiomas at a potential hotspot in exon 9. Br J Cancer
84:199201
Wang Z, Sutton LN, Cnaan A, Haselgrove JC, Rorke LB,
Zhao H, Bilaniuk LT, Zimmerman RA (1995) Proton
MR spectroscopy of pediatric cerebellar tumors.
AJNR Am J Neuroradiol 16:18211833
Warmuth-Metz M, Bison B, Dannemann-Stern E,
Kortmann R, Rutkowski S, Pietsch T (2008) CT and
MR imaging in atypical teratoid/rhabdoid tumors of
the central nervous system. Neuroradiology
50:447452
Yang S, Law M, Zagzag D, Wu HH, Cha S, Golfinos JG,
Knopp EA, Johnson G (2003) Dynamic contrastenhanced perfusion MR imaging measurements of
endothelial permeability: differentiation between
atypical and typical meningiomas. AJNR Am J
Yuh EL, Barkovich AJ, Gupta N (2009) Imaging of ependymomas: MRI and CT. Childs Nerv Syst 25:12031213
Zhang H, Rodiger LA, Shen T, Miao J, Oudkerk M (2008)
Perfusion MR imaging for differentiation of benign and
malignant meningiomas. Neuroradiology 50:525530

Rhabdoid Tumors: Differential
Diagnosis
Justin A. Bishop and Syed Z. Ali
Contents
Abstract
Introduction ............................................................
53
Microscopic Findings .............................................
54
Differential Diagnosis ............................................

Medulloblastoma......................................................
Choroid Plexus Carcinoma ......................................
Glioblastoma ............................................................
Germ Cell Tumors....................................................
Other Tumors ...........................................................
56
56
56
56
57
57
References ...............................................................
57
Atypical teratoid/rhabdoid tumor (AT/RT) is a

rare malignancy of the central nervous system.
It occurs most often in the posterior cranial
fossa of children younger than 2 years old.
The microscopic appearance of AT/RT is quite
variable. Though rhabdoid cells are characteristic of AT/RT, they are often dominated by
other cell populations. Owing to its varying
morphologies, many entities should be considered in the differential diagnosis of AT/RT.
These tumors include medulloblastoma, choroid plexus carcinoma, glioblastoma, and germ
cell tumors. Deletions or mutations on the
long arm of chromosome 22 resulting in
decreased expression of INI1 are seen in
nearly all cases of AT/RT. The recent introduction of an immunohistochemical stain for
INI1 has greatly aided pathologists in distinguishing AT/RT from other morphological
mimics.
Introduction
J.A. Bishop
Department of Pathology, The Johns Hopkins Hospital,
600 North Wolfe Street, Rm PATH 406, Baltimore,
MD 21287, USA
S.Z. Ali ()
Department of Pathology and Radiology,
The Johns Hopkins Hospital, 600 North Wolfe Street,
Rm PATH 406, Baltimore, MD 21287, USA
e-mail: Sali@jhmi.edu
Atypical teratoid/rhabdoid tumor (AT/RT) is a

rare, highly malignant central nervous system
tumor of uncertain histogenesis, first described
by Lefkowitz and colleagues in 1987 (Lefkowitz
et al. 1987; Louis et al. 2007). It is World Health
Organization Grade IV by definition (Louis et al.
2007). It almost always affects children
younger than 2 years of age, and there is a male

53
54
predominance (Burger et al. 1998; Rorke et al.

1996; Hilden et al. 2004). Although AT/RT can
arise anywhere in the CNS, these tumors typically arise in the posterior fossa (Hanna et al.
1993; Burger and Scheithauer 2007). Atypical
teratoid/rhabdoid tumors usually grow rapidly as
expanding masses with necrosis, and are associated with a poor prognosis (Burger and
Scheithauer 2007). Atypical teratoid/rhabdoid
tumor has a characteristic molecular alteration:
mutation of the hSNF5/INI1/SMARCB1 (abbreviated INI1) locus on chromosome band 22q11.2
(Biegel et al. 1999, 2000). INI1 encodes a protein
involved in chromatin remodeling (Edgar and
Rosenblum 2008). This mutation is shared by
other tumors (mostly renal, but also extra renal)
that AT/RT resembles morphologically (Dang
et al. 2003; Biegel et al. 2002). In fact, it is not
uncommon for patients with renal rhabdoid
tumors also to have AT/RT, a condition sometimes referred to as rhabdoid tumor predisposition syndrome (Weeks et al. 1989; Louis et al.
2007).
J.A. Bishop and S.Z. Ali
Fig. 6.1 AT/RT. A sagittal view MRI scan depicts a large

posterior fossa brain mass in a child
Microscopic Findings (Figs. 6.1, 6.2,

6.3, 6.4, 6.5, and 6.6)
Atypical teratoid/rhabdoid tumors is characteristically a very cellular neoplasm. Histologically, AT/
RT displays sheets of cells between fibrovascular
septae and frequent areas of necrosis that frequently calcify (Burger and Scheithauer 2007).
The designation rhabdoid refers to a characteristic population of cells that have large vesicular
nuclei with prominent nucleoli and eccentric cytoplasm with eosinophilic cytoplasmic inclusions
(Burger and Scheithauer 2007; Parwani et al.
2005; Edgar and Rosenblum 2008). Only about
25% of cases, though, are composed entirely of
rhabdoid cells (Edgar and Rosenblum 2008). The
teratoid portion of the tumors name refers to
the multitude of components the tumor displays,
including poorly differentiated neuroepithelial
elements, spindled mesenchymal elements, and
rarely overt epithelial differentiation in the form
of rudimentary glands, rosettes, or squamoid epithelium (Burger and Scheithauer 2007; Edgar and
Fig. 6.2 AT/RT. Histologic section with sheets of

rhabdoid cells displaying eccentric nuclei and macronucleoli (H&E stain, 200)
Rosenblum 2008). Mitoses and apoptotic bodies

are numerous. The tumor cells of AT/RT frequently show artifactual cytoplasmic vacuolization (Louis et al. 2007). Cytologically (e.g. on
touch preparations, fine needle aspiration, or cerebrospinal fluid cytology) AT/RTs yield highly cellular smears with cellular aggregation around
vessels yielding a papillary-like architecture
Pediatric Atypical Teratoid/Rhabdoid Tumors: Differential Diagnosis
Fig. 6.3 AT/RT. Histologic section with large malignant

cells showing the classic rhabdoid phenotype (H&E
stain, 200)
Fig. 6.4 AT/RT. Histologic section with malignant cells

containing abundant eosinophilic cytoplasm, eccentric
nucleus and macronucleous (H&E stain, 400)
Fig. 6.5 AT/RT. Cytologic smear displaying large

pleomorphic malignant cells (H&E stain, 400)
55
Fig. 6.6 Medulloblastoma. A hypercellular cytologic

smear displaying with primitive small round blue cells
(Diff Quik stain, 200)
(Parwani et al. 2005). As on histology, numerous

cellular populations can be seen cytologically. In
addition to the rhabdoid cells, smears show
varying proportions of large, round, pleomorphic
cells with amphophilic cytoplasm and no nucleoli
and small round, neuronal-appearing cells with
speckled chromatin and small nucleoli (Parwani
et al. 2005).
Atypical teratoid/rhabdoid tumor has an
unusual immunohistochemical phenotype. It is
usually immunoreactive for cytokeratins and
epithelial membrane antigen (EMA) as well as
vimentin and glial fibrillary acidic protein
(GFAP) (Burger and Scheithauer 2007). It can
be positive for neuronal markers synaptophysin
and neurofilament (Burger and Scheithauer
2007). The rhabdoid elements are positive for
smooth muscle actin, but unlike tumors of true
skeletal muscle differentiation (e.g., rhabdomyosarcoma), AT/RT does not express desmin or
myogenin. Interestingly, the individual tumor
cells themselves are usually immunoreactive for
only one antigen (Bouffard et al. 2004). The key
diagnostic marker, however, is INI1 protein;
nuclear expression is characteristically lost
completely in AT/RT cells. Positive staining in
normal cells such as endothelial cells or lymphocytes serves as an internal control.
Immunohistochemistry for INI1 is actually more
sensitive than genetic studies for AT/RT, since
up to 30% of cases will not show INI1 deletion
or mutation with molecular modalities (Edgar
and Rosenblum 2008).
56
Largely due to the polymorphous nature of

AT/RT, it can be a difficult diagnosis to make.
This chapter discusses the main considerations in
the differential diagnosis of AT/RT.
AT/RTs dominated by medulloblastoma-like

elements (Edgar and Rosenblum 2008).
Choroid Plexus Carcinoma
Differential Diagnosis
Medulloblastoma
Atypical teratoid/rhabdoid tumors are commonly
misdiagnosed as medulloblastoma. This is not
surprising when one considers that medulloblastomas are much more common than AT/RT, and
when one recalls that AT/RT can possess areas of
smaller, primitive-appearing cells (which can
predominate). Also complicating the issue is the
fact that some medulloblastomas (e.g. of the large
cell or anaplastic type) can display rhabdoid
cytological features (Edgar and Rosenblum
2008). Despite these difficulties, it is important to
distinguish the two entities because AT/RT is not
responsive to the standard chemotherapy regimens used for medulloblastoma (Edgar and
Rosenblum 2008). It is therefore important to at
least consider the diagnosis of AT/RT, especially
in the posterior fossa of a patient younger than 2
years. The finding of truly rhabdoid cells argues
against the diagnosis of medulloblastoma. A very
specific histologic feature of medulloblastoma is
the appearance of pale islands or nodules in
some conventional types (Edgar and Rosenblum
2008). Anaplastic and large cell medulloblastoma/PNET generally shows even a higher degree
of mitotic activity/apoptosis and nuclear hyperchromasia than AT/RT. Nuclear molding is more
common in medulloblastoma than AT/RT.
Immunohistochemistry can help, because
medulloblastoma does not express epithelial
markers or actin. Ultimately, the diagnosis can be
confirmed by immunohistochemistry for INI1;
expression is retained in medulloblastoma but
lost in AT/RT. Rare medulloblastomas have
shown loss of INI1 expression. However, these
tumors have been in very young children,
expressed epithelial antigens, and responded
poorly to chemotherapy (Haberler et al. 2006).
In short, these tumors very likely represented
Choroid plexus carcinoma is a very rare epithelial tumor that arises most often in the ventricles
of children (Burger and Scheitauer 2007; Judkins
et al. 2005). Some choroid plexus carcinomas
have only focal epithelial elements (Judkins et al.
2005). The papillary-like architecture, epithelial
differentiation, and occasional intraventricular
growth, along with high grade features of AT/RT
may lead to diagnostic confusion with choroid
plexus carcinoma. In addition, choroid plexus
carcinomas can show rhabdoid cytologic features. It can be difficult or impossible to differentiate the two types of tumors on H&E histology
alone (Judkins et al. 2005). In these cases, INI1
immunostaining patterns differentiate the two
entities, with choroid plexus carcinomas retaining nuclear expression and AT/RTs losing it
entirely (Judkins et al. 2005). Although there are
reports of choroid plexus carcinomas with loss of
INI1 expression, it is most likely that these tumors
truly represent AT/RTs (Edgar and Rosenblum
2008; Judkins et al. 2005).
Glioblastoma
Glioblastoma is the most common primary brain
neoplasm. Although the diagnosis of glioblastoma is usually straightforward, it can have many
appearances as suggested by its mutiforme
designation. Rare variants of glioblastoma with
epithelioid and/or rhabdoid features may be confused with AT/RT. Both tumors show high grade
features such as necrosis and numerous mitoses.
Immunohistochemical findings of glioblastoma
overlap somewhat with AT/RT, with GFAP and
vimentin positivity; even cytokeratins, EMA, and
actin can be positive in rhabdoid glioblastoma
(Fung et al. 2004). Glioblastomas usually present
in cerebral hemispheres of older adults, in contrast
to AT/RT. Supportive of the diagnosis of glioblastoma is the presence of fibrillary cytoplasmic
Pediatric Atypical Teratoid/Rhabdoid Tumors: Differential Diagnosis
processes or emergence of the tumor from a more

conventional-appearing glioma (Nagai et al.
2009). Once again, INI1 immunohistochemistry
is useful as gliomas consistently retain expression of this protein (Kleinschmidt-DeMasters
et al. 2010).
Germ Cell Tumors (Fig. 6.7)

Atypical teratoid/rhabdoid tumors may be confused with germ cell tumors, particularly when
arising in the pineal or sellar/suprasellar region,
locations far more typical for germ cell tumors.
Like teratomas, AT/RT can show heterologous
elements, but well-developed glioneuronal or
respiratory/gastrointestinal tissues point to
teratoma. While not truly rhabdoid, the tumor
cells of germinoma share some cytologic features
like large nuclei with prominent nucleoli.
A fried-egg appearance, prominent lymphocytic infiltration and granulomatous inflammation,
characteristic of germinoma, are not seen in AT/
RT. Embryonal carcinoma also has high grade
cytologic features, numerous mitoses, necrosis,
and abortive papillae. Fortunately, AT/RT seems
to be uniformly negative for placental alkaline
phosphatase (PLAP), c-kit (CD117), human chorionic gonadotropin (hCG), and OCT-4, some of
the immunohistochemical markers for germ cell
57
tumors. In addition, germ cell tumors have not

been found to have loss of INI1 expression
(Sigauke et al. 2006; Haberler et al. 2006).
Other Tumors
Other less common entities of the posterior fossa
to consider in the diagnosis of AT/RT include
anaplastic or rhabdoid forms of meningioma and
ependymoma. Atypical teratoid/rhabdoid tumor
may occasionally have myxoid areas with cords
of cells mimicking chordoma (Burger and
Scheithauer 2007). Also, if the patient has a prior
history of malignancy, metastatic melanoma,
rhabdomyosarcoma, or carcinoma should be considerations. In addition to imaging studies, clinical history, and immunophenotype, in most cases,
INI1 immunolabeling will be conclusive (Perry
et al. 2005). It is prudent to remember that
although loss of INI1 is very specific for AT/RT
in the central nervous system, other tumors such
as medullary carcinoma of the kidney and epithelioid sarcoma of soft tissue have been shown to
lose INI1 expression (Judkins 2007). In a patient
with the rhabdoid tumor predisposition syndrome, the distinction of metastatic disease from
AT/RT is impossible to make by histology and
immunohistochemistry alone (Louis et al. 2007).
In conclusion, atypical teratoid/rhabdoid tumor
is a rare pediatric tumor of the central nervous system. An appreciation of its variable morphologic
appearances and liberal use of the INI1 immunostain are necessary to avoid misdiagnosis.
References
Fig. 6.7 Germ cell tumor. Cytologic smear of germinoma. Note the large pleomorphic epithelioid cells with
macronucleoli admixed with few lymphocytes
(Papanicolaou stain, 400)
Biegel JA, Zhou JY, Rorke LB, Stenstrom C, Wainwright LM,

Fogelgren B (1999) Germ-line and acquired mutations
of INI1 in atypical teratoid and rhabdoid tumors.
Cancer Res 59:7479
6:27592763
Biegel JA, Tan L, Zhang F, Wainwright L, Russo P,
Rorke LB (2002) Alterations of the hSNF5/INI1 gene
58
in central nervous system atypical teratoid/rhabdoid
tumors and renal and extra renal rhabdoid tumors. Clin
Bouffard JP, Sandberg GD, Golden JA, Rorke LB (2004)
Double immunolabeling of central nervous system
atypical teratoid/rhabdoid tumors. Mod Pathol
17:679683
Burger PC, Scheithauer BW (2007) Tumors of the central
nervous system (AFIP atlas of tumor pathology).
American Registry of Pathology Press, Washington, DC
Atypical teratoid/rhabdoid tumor of the central nervous system: a highly malignant tumor of infancy and
pediatric oncology study group study. Am J Surg
Pathol 22:10831092
Dang T, Vassilyadi M, Michaud J, Jimenez C, Ventureyra EC
(2003) Atypical teratoid/rhabdoid tumors. Childs Nerv
Syst 19:244248
Fung KM, Perry A, Payner TD, Shan Y (2004) Rhabdoid
glioblastoma in an adult. Pathology 36:585587
Haberler C, Laggner U, Slavc I, Czech T, Ambros IM,
Ambros PF, Budka H, Hainfellner JA (2006)
Immunohistochemical analysis of INI1 protein in
malignant pediatric CNS tumors: lack of INI1 in
atypical teratoid/rhabdoid tumors and in a fraction of
primitive neuroectodermal tumors without rhabdoid
phenotype. Am J Surg Pathol 30:14621468
Hanna SL, Langston JW, Parham DM, Douglass EC
(1993) Primary malignant rhabdoid tumor of the brain:
clinical, imaging, and pathologic findings. AJNR Am
J Neuroradiol 14:107115
Hilden J, Meerbaum S, Burger P, Finlay J, Janss A,
Scheithauer BW, Wwalter AW, Rorke LB, Biegel JA
in a registry. J Clin Oncol 22:28772884
Judkins AR (2007) Immunohistochemistry of INI1
expression: a new tool for old challenges in CNS and
soft tissue pathology. Adv Anat Pathol 14:335339

INI1 protein expression distinguishes atypical teratoid/rhabdoid tumor from choroid plexus carcinoma.
Kleinschmidt-DeMasters BK, Alassiri AH, Birks DK,
Newell KL, Moore W, Lillehei KO (2010) Epithelioid
versus rhabdoid glioblastomas are distinguished by
monosomy 22 and immunohistochemical expression
of INI-1 but not claudin 6. Am J Surg Pathol
34:341354
Lefkowitz IB, Rorke LB, Packer RJ (1987) Atypical
teratoid tumor of infancy: definition of an entity. Ann
Neurol 22:448449, (abstract)
Louis DN, Ohgaki H, Wiestler OD, Cavenee WK (2007)
WHO classification of tumours of the central nervous
system. IARC Press, Lyon, pp 147149; 234235
Nagai S, Kurimoto M, Ishizawa S, Hayashi N, Hamada
H, Kamiyama H, Endo S (2009) A rare astrocytic
tumor with rhabdoid features. Brain Tumor Pathol
26:1924
diagnosis. Cancer Cytopathol 105:6570
Perry A, Fuller CE, Judkins AR, Dehner LP, Biegel JA
(2005) INI1 expression is retained in composite rhabdoid tumors, including rhabdoid meningiomas. Mod
Pathol 18:951958
85:5665
Sigauke E, Rakheja D, Maddox DL, Hladik CL,
White CL, Timmons CF, Raisanen J (2006) Absence
of expression of SMARCB1/INI1 in malignant
rhabdoid tumors of the central nervous system,
kidneys, and soft tissue: an immunohistochemical
study with implications for diagnosis. Mod Pathol
19:717725
Weeks DA, Beckwith JB, Mierau GW, Luckey DW (1989)
Rhabdoid tumor of kidney. A report of 111 cases from
the National Wilms Tumor Study Pathology Center.
Am J Surg Pathol 13:439458
Part II
Brain Tumors (General)
Pediatric Brain Tumors

(An Overview)
Eugene I. Hwang and Roger J. Packer
Contents
Abstract
Epidemiology of Pediatric Brain Tumors ............
61
Diagnostic Principles Imaging ...........................

Posterior Fossa .........................................................
Brainstem .................................................................
Optic Pathway/Suprasellar Tumors..........................
Other Tumors of Childhood .....................................
Other Imaging Modalities ........................................
62
63
63
64
64
64
Diagnostic Principles Pathology ........................

Molecular Studies ....................................................
65
66
Therapeutic Principles...........................................
67
Therapeutic Principles Neurosurgery ...............
67
Therapeutic Principles Radiotherapy ...............
68
Therapeutic Principles Chemotherapy .............
69
Therapeutic Principles Other Strategies ..........
71
Conclusion ..............................................................
71
References ...............................................................
72
Brain tumors are the most common solid tumor

of childhood and remain the leading cause of
cancer-related mortality in children. Advances
in diagnostic techniques, neurosurgical procedures, radiotherapy, and chemotherapeutic
regimens have resulted in modest improvements in survival; however, outcomes remain
suboptimal and morbidity frequent. Increased
understanding of the underlying molecular
characteristics of brain tumors in children
has led to the design and early testing of
targeted agents, although continued investigation is needed. This chapter will serve as a
general introduction to the principles of
diagnosis and treatment in children with
brain tumors, and will apply those principles
in a brief overview of the most common
tumors.
Epidemiology of Pediatric Brain

Tumors
E.I. Hwang (*) R.J. Packer

Center for Cancer and Blood Disorders, Childrens
National Medical Center, 111 Michigan Ave, NW,
Washington, DC 20010, USA
e-mail: ehwang@cnmc.org
Brain tumors represent the most common form of

solid tumor in children younger than 19 years of
age (Altekruse et al. 2010), constituting greater
than 20% of all childhood malignancies (CBTRUS
2011). Brain tumors are also the leading cause of
cancer-related mortality in children (Altekruse
et al. 2010); in fact, the proportion of deaths in children due to brain tumors has increased from 17.8%
in 1975 to 25.7% in 2006 (Smith et al. 2010). This

61
62
is primarily attributable to significant advances in

treatment for non-central nervous system (CNS)
cancers when compared to more modest improvements in the treatment of brain tumors.
Nearly 4,000 new CNS tumors per year are
diagnosed in children (Smith et al. 2010), and even
slow-growing tumors can cause significant morbidity and even death. Both tumor location and
therapeutic approaches frequently leave many survivors with a constellation of cognitive, neurologic,
endocrinologic, and neuropsychologic sequelae.
The most common primary brain tumors in
children less than 14 years old are gliomas (55%),
with pilocytic astrocytomas comprising the
largest tumor subtype within this group (18.5%).
The most common malignant histologies are
medulloblastoma (13%), and, in the 2007
CBTRUS report, malignant glioma NOS (14.2%).
Primary spinal tumors are rare in children, constituting approximately 2% of all childhood CNS
tumors. In adults, metastasis to the brain exceeds
the presence of primary CNS tumors; however, in
children metastatic disease in the brain accounts
for less than 5% of brain tumors and are usually
derived from leukemia and sarcoma. Although
most pediatric CNS tumors are sporadic in nature,
several tumor types can be associated with
specific genetic syndromes, including medulloblastoma (Gorlin and Turcot Syndromes, Li
Fraumeni), low-grade gliomas (neurofibromatosis types 1 and 2), choroid plexus tumors, and
high-grade gliomas (Li Fraumeni), for example.
Brain tumors in children are more likely to be
malignant when compared to adult brain tumors
(65.2% versus 33.7% (Kohler et al. 2011)),
although outcomes for children with brain tumors
are generally better than that for adults (5 year
OS 69.9% versus 35.5%, respectively (Smith
et al. 2010)). This has often been attributed to the
relative resilience of children and their ability to
tolerate aggressive therapy, but recent studies
have also demonstrated significant molecular differences between otherwise histologically similar
adult and pediatric brain tumors. These underlying differences may be exploited to improve
treatment regimens and better predict tumor
response and outcomes.
In children, brain tumor incidence, histology,
and location tend to be specific to age. For
E.I. Hwang and R.J. Packer
instance, overall location in children (as in adults)

tends to be supratentorial; however, this tendency
is reversed in the first decade of life, with approximately 60% of tumors in young children infratentorial in location. Brain tumor incidence in
children is highest in the under-5 year old group
and lowest in 1014 year olds (Altekruse et al.
2010). However, among children with brain
tumors, 75% are less than 10 years old, and 35%
less than 5 years old, reflecting the relatively
young age of most children with brain tumors;
this has implications both for diagnosis and therapy that will be discussed further. Malignant
tumors, such as medulloblastoma or malignant
glioma, are most common in children less than
5 years of age and become progressively less
common with age; conversely, tumors such as
pilocytic astrocytoma tend to maintain fairly stable rates throughout age groups. Taken together,
the specificity of tumor histology to age and location
is likely relevant to understanding tumorigenesis
in children, but is not yet fully explainable.
Compared to adult disease, diagnosis of brain
tumors in children is often delayed by several
factors, with a majority of patients diagnosed
only after several physician visits (Mehta et al.
2002). Communication with children regarding
the onset, severity, and extent of their symptoms
is often more difficult due to their younger age.
Moreover, symptoms such as headache, early
morning vomiting, and fatigue secondary to CNS
neoplasms can be nonspecific and common in
other pediatric conditions (Mehta et al. 2002);
this requires a high degree of suspicion in order
to accurately and quickly make a diagnosis. Other
common presenting complaints that may initially
be subtle include partial or complex seizures,
visual disturbance, ataxia, endocrinopathies,
altered personality or mental status, school difficulties or declining school performance, and
poorly described sensory abnormalities.
Diagnostic Principles Imaging

Many advances in diagnostics, such as the advent
and widespread dissemination of CT and MRI,
have enabled earlier and more accurate classification of brain tumors in children. Accurate
Pediatric Brain Tumors (An Overview)
detection of disease is essential to determining an

adequate treatment plan; for example, in a recent
Phase 3 trial, inaccurate or inadequate imaging
was found to worsen the outcome for patients
with medulloblastoma (Packer et al. 2006).
Often, the first imaging study of a presenting
patient is a CT scan due to its wide availability,
rapid turnaround, and the lack of need for sedation in young patients. However, if a CNS tumor
is detected, an MRI will be rapidly obtained. In
assessing brain tumors, MRI has traditionally
provided a preliminary differential using location, cellularity, edema (FLAIR), and disruption
of the blood brain barrier (contrast enhancement).
This differential can inform caregivers as to the
necessity of aggressive resection, obtaining tumor
markers or CSF samples, determine spinal or
other imaging needs, and ultimately required
therapeutic regimens. Contrast enhancement can
also be useful to determine the more biologically
active portions of the tumor, as well as to monitor
for post-surgical residual disease (Keating et al.
2001). This is particularly important, as extent of
resection in children is one of the most critical
common prognostic factors between disparate
histologies such as glioblastoma multiforme,
ependymoma, and medulloblastoma. Because
MRI may be unable to distinguish between postoperative changes and residual tumor, it is important to obtain post-operative scans within 72 h of
surgery (Merchant et al. 2009a).
Although the basic tenets of neuro-radiology
are common between pediatric and adult neurooncology, there are several specific locations and
tumors where neuro-imaging has particular import
in the treatment of children with brain tumors.
Posterior Fossa
Unlike in adults, posterior fossa tumors in children are common, and typically consist of pilocytic astrocytomas, medulloblastomas, and
ependymomas. Other less common posterior
fossa tumors include other primitive neuro-ectodermal tumors (such as atypical teratoid rhabdoid
tumor (ATRT), pineoblastoma), and high-grade
gliomas. Medulloblastomas often arise from the
roof of the fourth ventricle and tend to brightly
63
enhance with contrast, but are hypointense on T1

or T2 (Panigrahy and Bluml 2009); in older children, medulloblastomas tend to arise from the
cerebellar hemispheres. Medulloblastoma may
be disseminated on presentation, thus mandating
spinal imaging and CSF cytologic examination.
ATRT is most common in very young children
less than 2 years old and on imaging resembles
medulloblastoma, but has a much worse outcome
due to underlying biologic differences.
Ependymomas, the other common malignant
posterior fossa tumor in children, are also often
hypointense on T1 but demonstrate high T2 signal while enhancement is more heterogeneous;
these usually arise from the floor of the fourth
ventricle. Ependymomas are more common in
children less than 5 years old, and while metastasis on presentation is rare, neuraxis dissemination
can occur. Low-grade gliomas have T1 and T2
characteristics similar to ependymomas, but have
more enhancement and are typically heterogeneous and cystic.
Brainstem
In the brainstem, imaging findings become even
more essential, as tumors such as diffuse intrinsic
pontine glioma (DIPG) may be diagnosed solely
by a characteristic appearance on MRI (pontine
involvement with diffuse brainstem expansion, T2
hyperintensity with minimal contrast enhancement). Hydrocephalus is uncommon despite narrowing of the 4th ventricle, and mineralization is
likewise uncommon. Given the risk associated with
biopsy of the brainstem, few children with these
findings undergo surgery, and most will undergo
radiation therapy; unfortunately, median survival in
these children is approximately 1012 months.
However, not all primary brain stem tumors are
diffuse intrinsic pontine gliomas. Lower-grade
gliomas and primitive neuroectodermal tumors
may also present in the brainstem, which possess
different prognoses and mandate different treatment approaches. Thus, it is essential that with
atypical MRI findings (such as significant exophytic growth, aberrant enhancement patterns, or
focality in the midbrain, cervicomedullary junction,
or tectum) biopsy is strongly considered.
64
Optic Pathway/Suprasellar Tumors
Other Tumors of Childhood
Another location in children that relies heavily on

radiographic findings is the hypothalamus/optic
chiasm. Tumors in this area are most commonly
low-grade gliomas, but may also be craniopharyngiomas or germ-cell tumors (GCT). Low-grade
gliomas of this region have imaging characteristics as described above, and tend to arise from the
chiasm or other parts of the optic pathway.
Children with optic pathway gliomas either present with or are at risk for significant visual impairment, sometimes progressing to blindness. Rapid
and accurate diagnosis is essential without surgical intervention that may, in itself, cause visual
harm. Thus, with a characteristic MRI, either
chemotherapy or radiation therapy may be
initiated without tissue diagnosis.
Craniopharyngiomas are most common in the
second decade of life. These tumors are often diffusely cystic and enhance with gadolinium
administration. Calcification frequently occurs,
and is evident on both CT and MRI.
Germ cell tumors may primarily present in either
the suprasellar region or the pineal region, and will
on occasion present simultaneously in both as
bifocal disease. They may also be more diffuse or
affect the thalamus, where this is often misdiagnosed. On MRI, GCTs are well-circumscribed
lesions with low T1 and only mildly elevated T2
signal and may have calcifications; this contrasts
with gliomas which will often be hyperintense on
T2 scans. Thickening of the pituitary stalk is
common. In addition, in conjunction with this
radiographic appearance, serum or cerebrospinal
fluid (CSF) bhCG and a-fetoprotein (AFP) levels
can confirm diagnosis of GCT without tissue. CSF
markers are more sensitive than serum markers,
and together are sometimes able to subclassify
germ cell tumors into germinoma, teratoma, or
non-germinomatous GCT (including choriocarcinoma, mixed germ cell tumors, and yolk sac
tumors). Of this group, teratomas have the most
individual imaging findings, including calcifications, cystic components, enhancing solid portions
and fat-containing elements. Teratomas are the
most common brain tumor in the early infant
period.
The same tumors that populate the posterior fossa

can also occur supratentorially in children, including low- and high-grade gliomas, ependymomas,
primitive neuroectodermal tumors (PNETs), and
ATRTs, with imaging characteristics similar to
those described above. Supratentorial astrocytomas can occur throughout the cerebrum in
children and comprise almost one-third of
supratentorial tumors in children (Altekruse et al.
2010). High-grade gliomas are similar on imaging
to adult disease, namely, high T2 signal, low T1
signal with significant peritumoral edema and
contrast enhancement.
Choroid plexus carcinomas and papillomas
comprise 1020% of pediatric brain tumors in
the first year of life, and almost all present before
5 years of age. MRI findings can distinguish
between the more aggressive carcinomas and the
more benign papillomas using size, invasiveness,
and peritumoral edema as criteria. When arising
in children, they are typically in the lateral
ventricles, contrasted against the typical adult
location of the fourth ventricle.
Other less common CNS tumors of childhood include dysembryoplastic neuroepithelial
tumors (DNETs), which are associated with
intractable seizure disorders and appear as a
well-circumscribed mass in the cortex with
cystic components and heterogeneous enhancement. Schwannomas are benign tumors most
often in the internal auditory canal and associated with Neurofibromatosis type 2 (NF2).
Meningiomas, which are more common in
adults, are also associated with NF2 and are intimately associated with the meninges and may
also be secondary tumors after radiation therapy.
Other Imaging Modalities

Neuro-imaging has progressed beyond showing
only gross anatomic findings to providing detailed
information about underlying biologic tissue
function. Recently, MRIs have incorporated spectroscopy (MRS), which can identify markers of
cellular proliferation. Controversy continues to
exist regarding the utility of MRS in differentiating between types of brain tumors in children due
to considerable overlap of findings; however, in
general more aggressive tumors have higher choline and reduction or absence of N-acetylaspartate
(NAA) (Vezina 2005). MRS can sometimes
differentiate between necrosis and active tumor,
as necrotic tumors tend to have decreased NAA
and choline markers with concomitant lactate
elevation. However, exceptions such as germ cell
tumors (with unexpectedly lower choline levels)
and pilocytic astrocytoma (with relatively high
choline levels) continue to highlight the need for
optimization of this particular imaging technique.
Modern scanners have also added diffusionweighted imaging (DWI), used to identify
tumor cellularity by measuring the ability of
water to diffuse through cellular structures.
Diffusion tensor imaging (DTI) also utilizes
water motion and can be used to map white
matter tracts, while perfusion scanning is able
to determine the blood volume and regional
blood flow to define the perfusion state of tissue. Other imaging advances include single
photon emission commuted tomography
(SPECT), which is useful for determining
tumor metabolism and is widely available, but
has poor resolution compared to standard MRI.
Positron emission tomography (PET) scans utilizing 18FDG take advantage of increased tumor
uptake in higher grade malignancies. Attempts
to utilize PET imaging to distinguish tumor
necrosis from progression continue, but thus
far PET cannot be reliably used to differentiate
between the two. All of these imaging modalities continue to undergo evaluation for use in
pediatric neuro-imaging.
Lastly, consistent measurement of tumor size
is essential to determine tumor progression or
response, but can sometimes be difficult in pediatric neuro-oncology. Many brain tumors in children are not amenable to easy cross-sectional
measurements given their complex, heterogeneous appearance. This can make assessment of
tumor response to therapy or radiographic progression difficult, as some portions of tumor may
improve while others worsen. Volumetric
measurements are being explored as a solution to
65
this problem, and have been shown to be more

accurate in determining response and progression
in tumors outside the CNS (Kostis et al. 2004);
however, debate continues as to the validity of
such measurements in brain tumors.
Diagnostic Principles Pathology

As with other solid neoplasms, histopathology
has been the predominant method of identifying
and classifying pediatric CNS tumors. However,
classification by microscopic examination can be
imprecise due to the complexity of pediatric brain
tumors, which often are heterogeneous tumors
with foci that are discordant from the overall
tumor. In addition, biopsies in the brain are often
limited to very small portions of tissue that may
not accurately reflect the majority of the tumor,
further increasing the difficulty of accurate and
complete diagnosis; indeed, biopsy and definitive
tumor resection are potentially discordant in 17%
of cases (Aker et al. 2005).
Besides assessment of the gross cellular
appearance, specific stains can differentiate and
grade pediatric CNS tumors. These stains traditionally include H&E and proliferation markers
with Ki67/MIB1, but may include multiple subsequent immunostains, such as glial markers
(glial fibrillary acidic protein (GFAP), S100);
ependymal markers (EMA/CD99), markers of
medulloblastoma (synaptophysin), germinoma
(OCT3/4, PLAP, ckit), or choriocarcinoma
(bhCG, EMA), among others. Tumors are also
graded based on level of necrosis, mitotic activity, vascularity, and invasiveness. Based on these
findings, histopathology has traditionally been
critical for determining the risk category for pediatric brain tumors. Some of the elements of neuropathology specific to the care of pediatric brain
tumors are discussed below.
In ependymomas, anaplasia seems to indicate
worse outcomes and the need for more aggressive therapy. However, considerable variation can
be found in the assessment of anaplasia by neuropathologists. General agreement may be found
for a tumor that is diffusely anaplastic, but the
implications of focal areas of anaplasia, degree of
66
proliferation indices, vascular proliferation, and

necrosis is unclear. Thus, likely due in part to
these differences, there is some variation in correlation of anaplasia with outcomes.
Several histologic variants of medulloblastoma
exist, including classical, anaplastic (enlarged
nuclei and increased nuclear-to-cytoplasm ratio),
desmoplastic (areas of nodularity and differentiation), and large-cell (mitoses, necrosis). These
differences have been clearly associated with variable outcomes and treatment responses in children, demonstrating the utility of histopathology.
For instance, anaplastic features have conferred a
poorer prognosis in most studies. The desmoplastic variant, which is associated with improved
outcome and sonic hedgehog pathway activation,
is most commonly seen in adults or infants, but
less present in other children. Medulloblastoma
with extensive nodularity similarly confers an
improved prognosis; the improved outcome for
these patients may warrant therapy reductions.
Supratentorial primitive neuroectodermal tumors
(sPNETs) and atypical teratoid/rhabdoid tumors
(ATRTs) are histologically difficult to distinguish
from medulloblastoma, although ATRTs may be
differentiated by INI1 staining, which is typically
absent in ATRTs. However, these tumors are both
characterized by a poor outcome, reflecting underlying differences not accounted for by histopathologic appearance.
Although gliomas are the most common brain
tumor in children, tumor behavior and thus
therapy- is guided in part by histology. Grade I
gliomas, such as juvenile pilocytic astrocytomas,
are common in children and represented by characteristic Rosenthal fibers and low cellularity
with strong GFAP staining, although may exhibit
rare mitotic figures and microvascularization. As
will be discussed later, these tumors in general
are rarely life threatening and may be cured by
complete resection, although can also be recurrent or progressive. Grade II gliomas, such as
pilomyxoid gliomas, are similar but more highly
cellular with more infiltration but relatively rare
mitoses, vascular proliferation, or necrosis. Highgrade gliomas histologically closely resemble
their adult counterparts, with similarly dismal
outcomes. Both Grade III (anaplastic astrocy-
toma) and IV (glioblastoma multiforme) are

characterized by extensive necrosis, multiple
mitotic figures, and infiltration.
Because of the complexity of histopathological diagnosis coupled with the high level of heterogeneity of pediatric brain tumors, second or
even third neuropathological opinions are often
sought prior to initiation of therapy. Still, the
widely variant response to therapy and outcomes
for tumors that are histologically similar demonstrates the need for improved methods of tumor
characterization.
Molecular Studies
Recent investigation has raised the possibility
that tumors that appear histologically similar may
have dissimilar molecular characteristics that
more accurately predict outcome and sensitivity
to therapy.
In ependymomas, prediction of tumor
behavior based solely on histologic and clinical characteristics remains inadequate. Genetic
heterogeneity has been well described, and distinct differences from adult tumors are evident
(Kilday et al. 2009). Various pathways and
proteins have been implicated such as p53,
Cyclin D1, topoisomerase, tenascin, VEGF,
and EGFR, among others. Most recently, one
group has described two groups of posterior
fossa ependymomas, separated into a poorer
outcome group A (characterized by genetic
variation in multiple genes and chromosome
1q gain) and a better-outcome group B. This
group was able to distinguish between the two
categories by staining for LAMA2 (group A)
and NELL2 (group B) (Witt et al. 2011).
Studies in medulloblastoma (Cho et al. 2011)
have revealed several molecular subgroups that
seem to better predict response to therapy and
outcome when compared with traditional measures of disease risk. Medulloblastomas have
varying degrees of genetic variation, including
amplification of MYCN and activation of several
pathways including Sonic hedgehog, Notch and
Wnt, among others. Unsupervised molecular
analysis revealed tumor signatures that were
replicable and provided potential additional targets for future therapeutic options.
DIPGs, long assumed to be similar to cortical
high-grade glioma due to a similar histopathologic appearance, nevertheless have frequent
copy number gains in PDGFR and PARP
(Hawkins et al. 2002), which are not replicated in
cortical pediatric high-grade gliomas (HGGs) or
adult glioblastoma.
Increasing knowledge in juvenile pilocytic
astrocytomas (JPA) has stemmed from study of
children with NF-1, a heritable condition with
neurofibromin deficiency, who are predisposed
to development of JPA. Neurofibromin typically functions by suppressing the RAS/MEK
pathway; deficiency can also result in activation
of the PI3K/PTEN/AKT/mTOR pathways.
These two pathways have also been implicated
in tumorigenesis in sporadic low-grade gliomas
(Jones et al. 2008); for instance, almost half of
pediatric LGGs were reported to harbor a BRAF
activating mutation: V600E (Dougherty et al.
2010). Less is understood about the critical
molecular characteristics of HGGs in children.
They do not seem to share some aspects of adult
HGGs, such as activation of the epidermal
growth factor receptor (EGFR) pathway or
common expression of the variant III EGFR
mutant (Bredel et al. 1999), although they do
have overexpression of that pathway. Similarly,
pediatric HGGs have been found to more commonly have aberrant activation of plateletderived growth factor receptor (PDGFR) A and
B, which is present but less important in adult
HGGs, and more frequently display P53 mutations than adults.
Therapeutic Principles
Treatment of children with CNS tumors involves
challenges in determining effective therapy, but
also in preserving the relatively fragile developing
brain in pediatric patients. Multidisciplinary teams
have been developed that are able to cohesively
present a multifaceted treatment plan for patients
with brain tumors, involving neurosurgeons, neurologist, neuro-oncologists, endocrinologists, and
67
neuropsychologists; together, this team can provide comprehensive care of both direct tumor and
treatment-related adverse effects. In fact, each of
the available treatment modalities namely, resection, radiotherapy, and chemotherapy- has unique
problems related to the young age and location of
brain tumors in children.
Therapeutic
Principles Neurosurgery
As with most solid tumors, surgical resection is
often one of the most important therapeutic
options; however, the location of many brain
tumors in children often precludes complete excision. Thus, neurosurgeons must consider the
delicate balance between aggressive resection
and avoidance of devastating neurologic deficits
in patients. In some tumors such as DIPG or optic
pathway glioma, even a biopsy is often avoided
in the context of characteristic MRI findings
where the risk for life-altering damage is higher
and the gain from tissue diagnosis is lower.
Despite the potential complications, complete
resection is one of a few reliable prognostic indicators in several pediatric brain tumors. In lowgrade gliomas (LGGs), complete resection is
often sufficient for cure with greater than 90%
5-year progression free survival (Wisoff et al.
2011); however, less than one-third of pediatric
LGGs can be completely excised without significant morbidity. Indeed, given the slow growth
patterns of this tumor and excellent long-term
outcome, overly aggressive resection is avoided
unless easily achievable. In HGG, however, the
only two consistent prognostic indicators have
been underlying histology and extent of resection; thus, even with some increased likelihood
of surgically-related deficit, complete resection
is pursued. Similarly, in ependymoma, the extent
of resection was the single most important prognostic factor with a greater than 20% improvement
in outcome associated with gross-total resection
(GTR) (Sanford et al. 2009). The importance of
complete resection in ependymoma is further
highlighted by the current Childrens Oncology
Group (COG) protocol, which advocates for a
68
second look surgery after chemotherapy in an

attempt to achieve GTR. Achieving extensive
resection in medulloblastomas also confers a
highly beneficial effect, with a significantly worse
outcome if resection leaves greater than 1.5 cm2
residual tumor (Zeltzer et al. 1999).
Over the past decade, newer techniques have
improved neurosurgical procedures. Some, such
as brain mapping, can inform the surgeon
regarding eloquent areas and can improve resection totality while reducing morbidity. Additional
neurosurgical advances including microsurgical
and stereotactic techniques, improvements in
anesthesia, and enhancements in functional
radiology (magnetoencephalography (MEG),
intraoperative MRI, and diffusion tensor imaging), have enhanced the ability to surgically
intervene in children with brain tumors.
Nonetheless, post-surgical phenomena such as
posterior fossa syndrome and multiple neurologic deficits remain a risk.
Therapeutic Principles Radiotherapy

Radiation therapy has been a cornerstone of
brain tumor treatment since the description by
Harvey Cushing of prolongation of life with
radiotherapy in patients with medulloblastoma.
Radiation causes damage to DNA by generating
hydroxyl radicals; thus, cells have increased
radiosensitivity if well oxygenated and proliferating in the cell cycle. Conversely, tumor cells
may be intrinsically radioresistant, utilizing
DNA repair mechanisms and supported by the
tumor microenvironment. Dose, interval timing
such as fractionation, and radiation fields are key
considerations to overcoming any resistance
while maintaining overall tolerability (Keating
et al. 2001).
Therapeutic radiation to brain tumors carries
significant toxicity which is especially problematic in young children. Radiation to the brain
carries a high risk for cognitive impairment,
endocrinopathies including growth deficiency,
vasculopathies, secondary tumors, and hearing
deficits, among others (Merchant et al. 2009a).
Risk for toxicity increases with younger age,
more extensive field of radiation, and higher dose

of radiation; even with just a few of these risk
factors, many children have some level of intellectual deficit following therapeutic radiation.
In medulloblastoma, early clinical trials
greatly increased survival by employing craniospinal irradiation (CSI) to 36 Gy with focal
boosts of 54 Gy to the tumor bed, which treated
gross disease as well as prevented metastatic
recurrence. However, the significant long-term
adverse effects subsequently led to serial clinical
trials attempting to reduce the dose of irradiation, ultimately, through the addition of chemotherapy, successfully reducing the CSI dose to
23.4 Gy while preserving efficacy. Further investigation is ongoing into potentially further dose
reduction to 18 Gy in select patient groups by
intensification of adjuvant therapy. In children
with high-risk disease, however, the radiation
dose and field have not been decreased due to
corresponding lower survival rates.
In LGG, focal radiotherapy has long been
accepted as an effective treatment, with greater
than 80% 5 year EFS (Merchant et al. 2009b).
Again because of radiation-associated adverse
effects, the overall strategy of treatment in younger
children has become postponement of radiation
via the use of chemotherapy. Conversely, in HGGs
(including DIPG), radiation is the only non-surgical
treatment modality that has been shown to prolong
survival in children, and XRT can also often provide transient symptomatic relief. Thus, focal XRT
is standard therapy for these tumors. Finally, trials
into optimization of radiosensitization in HGG
continue despite previously disappointing studies
involving temozolomide, capecitabine, gadolinium and texaphrine, among others.
Radiation modalities have striven to improve
delivery to tumor cells while sparing surrounding
normal tissue. Advances in targeting and delivery
have included the advent of conformal therapy
and intensity modulated radiotherapy (IMRT),
which allow more precise delivery of photons
and sparing of surrounding tissue. Proton therapy
is a newer radiation modality utilizing protons,
which are larger particles that reach a peak dose
in the terminal portion of their path. Together, the
larger particle size and focality of dose translate
into significantly less radiation to off-target

points. Although proton beam radiation is available only in a few major centers, it has particular
applications in the radiation of developing brains
(MacDonald et al. 2008). Indeed, for most pediatric tumors, proton therapy may mitigate some
radiation effects, and is becoming recommended
more frequently.
Therapeutic
Principles Chemotherapy
Initial treatment for brain tumors in adults has
traditionally been focused on resection and radiotherapy; however, the considerable surgical and
radiation toxicity in children coupled with suboptimal outcomes have highlighted the need for
additional therapy. Chemotherapy in general
exploits the sensitivity of rapidly dividing neoplastic cells to non-specific damage of DNA or
impairment of vital cellular functions. Children
have many age-specific characteristics which can
affect the clearance and toxicity profile of chemotherapy, including smaller volume of distribution, altered clearance parameters in young
children, and variable ability to absorb orally
administered agents.
Chemotherapy has improved outcomes in
some tumors such as medulloblastoma, where
institution of systemic chemotherapy helped
improve overall survival in the past 30 years
from less than 50% to 73% (Smith et al. 2010). It
has been used successfully in stabilizing tumors
and therefore delaying or even obviating the
need for radiation therapy, which is especially
valuable in very young children (Grundy et al.
2007). In those cases where radiation is necessary, chemotherapy has been effective at reducing the field or dose of radiation necessary
(Packer et al. 2006). However, in other CNS
tumors such as glioblastoma multiforme, craniopharyngioma, and meningioma, among others,
addition of systemic chemotherapy has been
disappointing, with minimal improvements in
overall or event-free survival.
Chemotherapy given prior to definitive local
control is termed neoadjuvant therapy, and
69
seeks to control micrometastatic disease, shrink

the primary tumor and in turn reduce morbidity,
improve success of local control, and provide
information regarding chemosensitivity of the
individual tumor. Some chemotherapy may be
given concomitantly with local control, such as
that used through convection enhanced delivery,
which involves placement of catheters into a
tumor for direct instillation of antineoplastic
agents, thereby bypassing the blood-brain barrier
and limiting systemic exposure. Chemotherapy
during radiation can also theoretically provide
several advantages, including sensitization of
tumor cells to radiation. One such method is by
inhibiting DNA repair mechanisms through
PARP or mTOR inhibition, two pathways thought
to contribute to DNA repair and therefore radiation resistance. Finally, adjuvant therapy is
administered after local control is completed.
Because of the unique complications specific to
CNS tumors, immediate local intervention is
often required, which leaves adjuvant timing as
the most convenient and safest regimen. This
method of administration has one major goal:
elimination of any residual gross or microscopic
disease.
Specific agents, dosing schedule, and route of
administration vary widely depending on histology, grade, and stage of a particular brain tumor.
Combination chemotherapy can be more effective than single agents; however, overlapping
toxicities can worsen in an unexpected manner,
thus necessitating careful consideration of
optimal combinations.
More intense therapy may potentially overcome several barriers for chemotherapy, including penetration of the blood-brain barrier and
overcoming both intrinsic and extrinsic tumor
resistance. High-dose chemotherapy with autologous stem-cell rescue is a strategy recently
employed in various malignant tumors, where the
primary role of the stem-cell infusion is reconstitution of bone marrow, allowing safer administration of higher doses of chemotherapy. This
strategy has been helpful in some scenarios, such
as in infants with brain tumors (Dhall et al. 2008)
or high-risk medulloblastoma, although has been
disappointing in others, such as some recurrent
70
tumors. In addition, the considerable toxicity

associated with high-dose chemotherapy precludes its frequent use unless obvious benefit or
in the context of otherwise poor survival.
Conversely, some chemotherapy plans have
utilized a metronomic dosing schedule. When
compared with conventional chemotherapy regimens, metronomic dosing utilizes lower-dose
and more frequent administration. It is thought
that this strategy can target angiogenesis in
tumors (Kieran 2005) as well as potentially attack
cells in various stages of cell division. One clear
advantage of metronomic schedules is increased
tolerability.
The practical use of chemotherapy in children
with brain tumors varies between tumor types
and stages. The role of chemotherapy in the management of some of the most common tumors
will be discussed here.
It is clear that children with ependymoma can
be responsive to chemotherapy, with some groups
avoiding radiotherapy in a subset of patients by
using chemotherapy alone (Grundy et al. 2007),
and others reporting clinical responses prior to
radiotherapy. However, no clinical trials have
definitively demonstrated a role for chemotherapy in the upfront treatment of ependymoma.
Multiple studies have reported no improved
survival despite prolonged maintenance chemotherapy (Robertson et al. 1998). In young children, however, neoadjuvant chemotherapy was
successful in delaying radiation therapy in a subset of patients, with a response rate of 48%
(Duffner et al. 1999). Investigation continues into
the utility of chemotherapy to provide an
improved resection, superior outcomes, or
avoidance/delay of radiotherapy.
The treatment of medulloblastoma has, however, benefited from the addition of chemotherapy, especially by improved outcomes for
high-risk medulloblastoma and for allowing
reduction of radiotherapy without decreasing survival. Attempts to reduce radiotherapy doses
were initially met with worsening outcomes
(Thomas et al. 2000); however, with the addition
of chemotherapy, lowered radiation doses were
able to maintain overall survival (Packer et al.
1999). In high risk patients, multiple attempts to
improve outcome with chemotherapy have been

attempted, including some reports of benefit utilizing high-dose chemotherapy with stem cell
rescue (Gajjar et al. 2006) and chemotherapybased radiosensitization, with resultant survival
improving to more closely match that of standard
risk patients. Finally, use of agents targeting the
underlying molecular drivers of medulloblastoma
has emerged. For example, inhibition of the sonic
hedgehog pathway, which supports some
medulloblastoma growth, has resulted in evidence of encouraging tumor response (Rudin
et al. 2009).
As previously stated, chemotherapy can be
effective in the control of low-grade gliomas
which are not amenable to complete resection
or radiotherapy. Carboplatin and vincristine
have been shown to achieve this aim in 8090%
of patients with unresectable LGG; however a
significant percentage of these patients will
progress (Packer et al. 1997). In patients with
progressive disease, a multitude of regimens
have been tested, including TPCV, vinblastine,
avastin/irinotecan, and cisplatin/etoposide,
among others. Although a small proportion of
these patients will have permanent disease stability using these agents, currently the primary
role of chemotherapy is to delay radiotherapy
administration.
In high-grade gliomas, chemotherapy has
been less effective. Despite evidence of benefit
from temozolomide administration in adults
with HGG, analogous pediatric studies have not
demonstrated a similar benefit. Other studies
have examined the use of chemoradiation and
intensive chemotherapy regimens, including
high-dose chemotherapy with stem-cell rescue,
again without definitive improvement in survival.
Thus, chemotherapy in the context of a child
with HGG remains controversial. Despite
emerging evidence of the molecular differences
between cortical and brainstem HGG, the overall conclusions for chemotherapy are the same
in DIPG, i.e., there has been no study demonstrating significant prolongation of survival with
any chemotherapy agent.
Although the role of chemotherapy in infants
has not been completely defined, several studies
have attempted to improve upon a historically

poor prognosis with varying chemotherapeutic
regimens. These strategies revolve around utilization of high-dose chemotherapy and stem cell
rescue either upfront or on relapse, administration of prolonged post-operative chemotherapy,
and the use of intrathecal treatment. Some
improvement in outcome has resulted from these
approaches (Dhall et al. 2008), encouraging further investigation into this tactic. Nonetheless,
the prognosis for infants with malignant
brain tumors remains poor despite multimodal
therapy.
Therapeutic Principles Other

Strategies
Due to the challenges associated with effective
chemotherapy use, and in conjunction with an
increasing knowledge of the molecular foundation of brain tumors, molecularly-targeted therapies have recently come under investigation.
Analysis of brain tumor biology has revealed
activation of pathways such as epidermal growth
factor (high-grade glioma, ependymoma), platelet-derived growth factor (DIPG, meningioma),
Sonic hedgehog (medulloblastoma), Notch
(medulloblastoma, glioblastoma multiforme),
Ras/Raf/MAPK (glioma, various), and PI3K/
AKT/mTOR. Inhibiting tumor angiogenesis also
remains a promising avenue for further investigation, as does inhibition of other biologic targets, such as histone deacetylase (HDAC) and
poly (adenosine-diphosphate-ribose) polymerase
(PARP).
However, despite promising pre-clinical data,
successful utilization of molecularly targeted
therapy has not yet been fully realized. This is
due to many factors. Identification and significance of specific pathways critical in brain
tumors is ongoing but incomplete. Delivery of
these molecules faces the same challenges that
chemotherapy does, including crossing the
blood-brain barrier. Further, inhibition of pathways vital to tumorigenesis may also have unexpected systemic toxicities, especially in
developing children. Finally, only an exceed-
71
ingly small number of pediatric patients are

available with specific targets eligible for new
treatment protocols; this problem is worsened by
the growing understanding that simple extrapolation from adult analogs is inadequate. Still,
although many challenges remain in assessing
efficacy and best administration of molecularly
targeted agents, this area continues to hold the
promise of future advances.
Finally, the potential for priming the immune
system for anticancer effect has been investigated recently in several pediatric trials utilizing
vaccines. Vaccine trials directed toward tumor
have generally targeted tumor antigens available in the context of specific HLA types, antigens typically expressed on specific tumor
types, or generation of tumor-specific vaccines.
Multiple clinical investigations continue to
investigate the possible efficacy and optimal
design of vaccine use, with some promising
early reports.
Conclusion
Treatment of children with brain tumors has
improved with enhanced delivery of radiation
therapy, techniques of surgical resection, and
advances in combinations and delivery of chemotherapy. However, outcomes for many pediatric brain tumors remain suboptimal, and
children with CNS tumors require continued
treatment advances, both to maximize survival
and to minimize adverse effects. Given the modest improvement to date, significant progress
will likely require innovation in both methods of
delivery as well as agents employed. New biologic targeting is promising, but optimal partnerships with conventionally employed
modalities of treatment remain elusive and systematic study of new agents is associated with
considerable hurdles. Regardless, improved
comprehension of tumor molecular biology has
led to the generation of new targeted therapies,
and the next era will likely emphasize continued
research into how to best incorporate new agents
into the management of children with brain
tumors.
72
References
Aker FV, Hakan T, Karadereier S, Erkan M (2005)
Accuracy and diagnostic yield of stereotactic biopsy
in the diagnosis of brain masses: comparison of results
of biopsy and resected surgical specimens.
Neuropathology 25(3):207213
Altekruse SF, Kosary CL, Krapcho M, Neyman N,
Aminou R, Waldron W, Ruhl J, Howlader N, Tatalovich
Z, Cho H, Mariotto A, Eisner MP, Lewis DR, Cronin
K, Chen HS, Feuer EJ, Stinchcomb DG, Edwards BK
(eds) (2010) SEER Cancer Statistics Review:
19752007
Bredel M, Pollack IF, Hamilton RL, James CD (1999)
Epidermal growth factor receptor expression and gene
amplification in high-grade non-brainstem gliomas of
childhood. Clin Cancer Res 5(7):17861792
Central Brain Tumor Registry of the US (CBTRUS)
(2011) Statistical report: primary brain tumors in the
United States, 20042007
Cho YJ, Tsherniak A, Tamayo P, Santagat S, Ligon A,
Greulich H, Erhoukim R, Amani V, Goumnerova L,
Eberhart CG, Laug CC, Olson JM, Gilbertson RJ,
Gajjar A, Delattre O, Kool M, Ligon K, Meyerson M,
Mesirov JP, Pomeroy SL (2011) Integrative genomic
analysis of medulloblastoma identifies a molecular
subgroup that drives poor clinical outcome. J Clin
Oncol 29(11):14241430
Dhall G, Grodman H, Ji L, Sands S, Garnder S, Dunkel IJ,
McCowage GB, Diez, Allen JC, Gopalan A, Corenlius
AS, Termuhlen A, Abromowitch M, Sposto R, Finlay
JL (2008) Outcome of children less than three years
old at diagnosis with non-metastatic medulloblastoma
treated with chemotherapy on the Head Start I and II
protocols. Pediatr Blood Cancer 50(6):11691175
Dougherty MJ, Santi M, Brose MS, Ma C, Resnick AC,
Sievert AJ, Storm PB, Biegel JA (2010) Activating
mutations in BRAF characterize a spectrum of pediatric
low-grade gliomas. Neuro-oncology 12(7):621630
Duffner PK, Horowitz ME, Kirscher JP, Friedman HS,
Burger PC, Cohen ME, Sanford RA, Mulhern RK,
James HE, Freeman CR, Kun LE (1999) The treatment of malignant brain tumors in infants and very
young children: an update of the Pediatric Oncology
Group experience. Neuro Oncol 1:152161
Gajjar A, Chintagumpala M, Ashley D, Kellie S, Kun
LE, Merchant TE, Woo S, Wheeler G, Ahern V,
Krasin MJ, Fouladi M, Broniscer A, Krance R,
Hale GA, Stewart CF, Dauser R, Sanford RA, Fuller
C, Lauc C, Boyett JM, Wallace D, Gilbertson RJ
(2006) Risk-adapted craniospinal radiotherapy followed by high-dose chemotherapy and stem-cell
rescue in children with newly diagnosed medulloblastoma (St Jude Medulloblastoma-96): long-term
results from a prospective, multicentre trial. Lancet
Oncol 7(10):813820
Grundy RG, Wilne SA, Weston CL, Robinson K, Lasford
LS, Ironside J, Cox T, Chong WK, Campbell RH,

Bailey CC, Gattamaneni R, Picton S, Thorpe N,
Mallucci C, English MW, Punt JA, Walker DA,
Wllison DW, Machin D (2007) Primary postoperative
chemotherapy without radiotherapy for intracranial
ependymoma in children: the UKCCSG/SIOP prospective study. Lancet Oncol 8(8):696705
Hawkins TL, Detter JC, Richardson PM (2002) Whole
genome amplification applications and advances.
Curr Opin Biotechnol 13(1):6567
Jones DT, Kocialkowski S, Liu L, Pearson DM, Backlund
LM, Ichimura K, Collins VP (2008) Tandem duplication producing a novel oncogenic BRAF fusion gene
defines the majority of pilocytic astrocytomas. Cancer
Res 68(21):86738677
Keating RF, Goodrich JT, Packer RJ (2001) Tumors of the
pediatric central nervous system. Thieme, New York/
Stuttgart
Kieran MW (2005) Anti-angiogenic therapy in pediatric
neuro-oncology. J Neurooncol 75(3):327334
Kilday JP, Rahman R, Dyer S, Ridley L, Lowe J, Coyle B,
Grundy R (2009) Pediatric ependymoma: biological
perspectives. Mol Cancer Res 7(6):765786
Kohler BA, Ward E, McCarthy BJ, Schymura MJ, Ries
LA, Eheman C, Jemal A, Anderson RN, Ajani UA,
Edwards BK (2011) Annual report to the nation on the
status of cancer, 19752007, featuring tumors of the
brain and other nervous system. J Natl Cancer Inst
103(9):714736
Kostis WJ, Yankelevitz DF, Reeves AP, Fluture SC,
Henschke CI (2004) Small pulmonary nodules: reproducibility of three-dimensional volumetric measurement and estimation of time to follow-up CT.
Radiology 231(2):446452
MacDonald SM, Safai S, Trofimov A, Wolfgang J,
Fullerton B, Yeap BY, Bortfeld T, Tarbell NJ, Yock T
(2008) Proton radiotherapy for childhood ependymoma:
initial clinical outcomes and dose comparisons. Int J
Radiat Oncol Biol Phys 71(4):979986
Mehta V, Chapman A, McNeely PD, Walling S, Howes
WJ (2002) Latency between symptom onset and diagnosis of pediatric brain tumors: an Eastern Canadian
geographic study. Neurosurgery 51(2):365372; discussion 372363
Merchant TE, Conklin HM, Wu S, Lustig RH, Xiong X
(2009a) Late effects of conformal radiation therapy for
pediatric patients with low-grade glioma: prospective
evaluation of cognitive, endocrine, and hearing deficits. J Clin Oncol 27(22):36913697
Merchant TE, Kun LE, Wu S, Xiong X, Sanford RA,
Boop FA (2009b) Phase II trial of conformal radiation
therapy for pediatric low-grade glioma. J Clin Oncol
27(22):35983604
Packer RJ, Ater J, Allen J, Phillips P, Geyer R,
Nicholson HS, Jakacki R, Kurczynski E, Needle M,
Finlay J, Reaman G, Boyett JM (1997) Carboplatin
and vincristine chemotherapy for children with newly
diagnosed progressive low-grade gliomas. J Neurosurg
86:747754
Packer RJ, Goldwein J, Nicholson HS, Vezina LG, ALlen
JC, Ris MD, Muraszko K, Rorke LB, Wara WM,
Cohen BH, Boyett JM (1999) Treatment of children

with medulloblastomas with reduced-dose craniospinal radiation therapy and adjuvant chemotherapy: a
Childrens Cancer Group Study. J Clin Oncol
17(7):21272136
Packer RJ, Gajjar A, Vezina G, Rorke-Adams L, Burger
PC, Robertson PL, Bayer L, LaFond D, Donahue BR,
Marymont MH, Muraszko K, Langston J, Sposto R
(2006) Phase III study of craniospinal radiation therapy followed by adjuvant chemotherapy for newly
diagnosed average-risk medulloblastoma. J Clin Oncol
24(25):42024208
Panigrahy A, Bluml S (2009) Neuroimaging of pediatric
brain tumors: from basic to advanced magnetic
resonance imaging (MRI). J Child Neurol
24(11):13431365
Robertson PL, Zeltzer PM, Boyett JM, Rorke LB, Allen
JC, Geyer JR, Stanley P, Li H, Albright AL, McGuireCullen P, Finlay JL, Stevens KR Jr, Milstein JM,
Packer RJ, Wisoff J (1998) Survival and prognostic
factors following radiation therapy for ependymomas
in children: a report of the Childrens Cancer Group.
J Neurosurg 88(4):695703
Rudin CM, Hann CL, Laterra J, Yauch RL, Callahan CA,
Fu L, Holcomb T, Stinson J, Gould SE, COleman B,
LoRusso PM, Von Hoff DD, de Sauvage FJ, Low JA
(2009) Treatment of medulloblastoma with hedgehog
pathway inhibitor GDC-0449. N Engl J Med
361(12):11731178
Sanford RA, Merchant TE, Zwienenberg-Lee M, Kun LE,
Boop FA (2009) Advances in surgical techniques for
resection of childhood cerebellopontine angle
ependymomas are key to survival. Childs Nerv Syst
25(10):12291240
Smith MA, Seibel NL, Altekruse SF, Ries LA, Melbert
DL, OLeary M, Smith FO, Reaman GH (2010)
73
Outcomes for children and adolescents with cancer:
challenges for the twenty-first century. J Clin Oncol
28(15):26252634
Thomas PR, Deutsch M, Kepner JL, Boyett JM, Krischer
J, Aronin P, Albright L, Allen JC, Packer RJ, Linggood
R, Mulhern R, Stehbens JA, Langston J, Stanley P,
Duffner P, Rorke L, Cherlow J, Friedman HS, Finlay JL,
Vietti TJ, Kun LE (2000) Low-stage medulloblastoma:
final analysis of trial comparing standard-dose with
reduced-dose neuraxis irradiation. J Clin Oncol
18(16):30043011
Vezina LG (2005) Neuroradiology of childhood brain
tumors: new challenges. J Neurooncol 75(3):
243252
Wisoff JH, Sanford RA, Heier LA, Sposto R, Burger PC,
Yates AJ, Homes EJ, Kun LE (2011) Primary neurosurgery for pediatric low-grade gliomas: a prospective
multi-institutional study from the Childrens Oncology
Group. Neurosurgery 68(6):15481555
Witt H, Mack SC, Ryzhova M, Bender S, Sill M, Isserlin
R, Benner A, Hielscher T, Milde T, Remke M, Jones
DT, Northcott PA, Garzia L, Bertrand KC, Wittmann
A, Yao Y, Roberts SS, Massimi L, Van Meter T, Weiss
WA, Gupta N, Grajkowska W, Lach B, Cho YJ, von
Deimling A, Kulozik AE, Witt O (2011) Delineation
of two clinically and molecularly distinct subgroups of
posterior fossa ependymoma. Cancer Cell
20(2):143157
Zeltzer PM, Boyett JM, Finlay JL, ALbright AL, Rorke
LB, Milstein JM, ALlen JC, Stevens KR, Stanley P, Li
H, Wisoff JH, Geyer JR, McGuire-Cullen P, Stehbens
JA, Shurin SB, Packer RJ (1999) Metastasis stage,
adjuvant treatment, and residual tumor are prognostic
factors for medulloblastoma in children: conclusions
from the Childrens Cancer Group 921 randomized
phase III study. J Clin Oncol 17(3):832845
Pediatric CNS Primitive

Neuroectodermal Tumor:
Role of the WNT Pathway
Hazel A. Rogers and Richard G. Grundy
Contents
Abstract
Introduction ............................................................
76
Methodology ...........................................................
Sample Cohort .........................................................
Immunohistochemistry ............................................
Mutational Analysis .................................................
Statistical Analysis ...................................................
77
77
77
78
78
WNT/B-Catenin Pathway Status

in CNS PNET..........................................................
Immunohistochemistry ............................................
Mutational Analysis .................................................
Statistical Analysis ...................................................
78
78
81
81
Discussion................................................................
83
References ...............................................................
85
H.A. Rogers R.G. Grundy (*)

Department of Brain Tumour Research Centre,
Queens Medical Centre, University of Nottingham,
D Floor Medical School (D32),
Nottingham NG7 2UH, UK
e-mail: Richard.grundy@nottingham.ac.uk
Central nervous system (CNS) primitive

neuroectodermal tumors (CNS PNET) are
high grade, predominantly pediatric, brain
tumors. Previously they have been grouped
with medulloblastomas, due to their histological similarities. The WNT/b-catenin pathway
has been implicated in many tumor types
including medulloblastoma. Upon pathway
activation, b-catenin (CTNNB1) translocates
to the nucleus where it induces transcription
of target genes. It is commonly up-regulated
in tumors by mutations in the key pathway
components APC and CTNNB1. WNT/bcatenin pathway status was investigated by
immunohistochemical analysis of CTNNB1
and the pathway target cyclin D1 (CCND1) in
49 CNS PNETs and 46 medulloblastomas.
The mutational status of APC and CTNNB1
(b-catenin) was investigated in 33 CNS PNETs
and 22 medulloblastomas. b-catenin nuclear
localization was seen in 36% of CNS PNETs
and 27% of medulloblastomas. A significant
correlation was found between CTNNB1
nuclear localization and CCND1 levels.
Mutations in CTNNB1 were identified in 4%
of CNS PNETs and 20% of medulloblastomas. No mutations were identified in APC. A
potential link between the level of nuclear
staining and a better prognosis was identified
in the CNS PNETs, suggesting that the extent
of pathway activation is linked to outcome.
The results suggest that the WNT/b-catenin

75
H.A. Rogers and R.G. Grundy
76
pathway plays an important role in the pathogenesis of CNS PNETs. However, activation
is not caused by mutations in CTNNB1 or APC
in the majority of CNS PNET cases.
Introduction
The most common solid tumors during childhood
are those of the CNS. CNS primitive neuroectodermal tumors (CNS PNET) are high grade
embryonal tumors that occur at any extracerebellar site in the central nervous system and are
composed of undifferentiated or poorly differentiated neuroepithelial cells (Louis et al. 2007).
Current outcome for children with CNS PNET is
poor with a relatively low overall 5-year survival
rate (Geyer et al. 2005; Reddy et al. 2000).
Relatively little research has been undertaken to
elucidate the molecular basis of CNS PNETs.
Previously they have often been grouped with the
histologically similar tumor medulloblastoma;
both being composed of poorly differentiated
round blue cells with scant cytoplasm (Louis
et al. 2007). An increased understanding of CNS
PNET biology will allow a more targeted
approach to therapy.
Many studies have demonstrated deregulation
of developmental signaling pathways involved in
normal brain development in medulloblastoma.
Similar pathways are likely to be involved in
CNS PNETs. The WNT/b-catenin signaling
pathway plays a key role in many cellular functions related to tumorigenesis including cell proliferation, differentiation, and migration. It was
originally linked to medulloblastoma through
studies of Turcot syndrome where germline
mutations in the APC gene have been identified
(Hamilton et al. 1995).
B-catenin (CTNNB1) is the key downstream
effecter of the pathway. When the pathway is
inactive CTNNB1 is bound in the cytoplasm to a
complex containing the proteins adenomatous
polyposis coli (APC), axin1 and glycogen synthase kinase-3b (GSK-3b). GSK-3b phosphorylates CTNNB1 at specific serine and threonine
residues allowing the protein to be targeted for
degradation through the ubiquitin-proteosome
system (Morin 1999). Upon pathway activation

the protein complex is destabilized, preventing
phosphorylation and enabling CTNNB1 to translocate to the nucleus where it acts as a co-activator
of TCF and LEF transcription factors and leads to
the up-regulation of target genes, including MYC
and cyclin D1(CCND1) (He et al. 1998; Tetsu
and McCormick 1999).
Activating mutations in CTNNB1 have been
identified in many different cancers including
Colon and Hepatocellular (Koch et al. 1999;
Morin et al. 1997). Single base substitutions have
been identified at codons in exon 3 of the gene,
encoding serine and threonine residues targeted
by GSK-3b, or at adjacent residues. These mutations are predicted to prevent phosphorylation
and subsequent degradation of CTNNB1.
Pathway activation through the stabilization and
nuclear accumulation of CTNNB1 has been demonstrated in sporadic medulloblastomas (Clifford
et al. 2006; Eberhart et al. 2000; Ellison et al.
2005; Koch et al. 2001; Thompson et al. 2006).
In the majority of cases this was caused by activating mutations in CTNNB1. A small study
identified a single mutation in CTNNB1 in one
out of four CNS PNETs (Koch et al. 2001). No
further research has been undertaken in CNS
PNET to date.
APC is also commonly mutated in many
tumor types including colon with the majority of
mutations occurring in the mutation cluster
region (Miyoshi et al. 1992). Mutations in APC
are commonly truncated, resulting in proteins
that are not able to form the cytoplasmic complex to target CTNNB1 for degradation. APC
mutations are rare in sporadic medulloblastomas
(Clifford et al. 2006; Ellison et al. 2005; Huang
et al. 2000; Koch et al. 2001; Thompson et al.
2006). To date only one study investigating APC
mutational status in 4 CNS PNET tumors has
been reported where no mutations were found
(Koch et al. 2001).
A significant association between CTNNB1
nuclear immunoreactivity and survival has previously been demonstrated in medulloblastoma,
with nuclear accumulation being associated with
a favorable outcome (Ellison et al. 2005). This is
in contrast to other tumor types, such as colon,
Pediatric CNS Primitive Neuroectodermal Tumor: Role of the WNT Pathway
and breast carcinomas, where nuclear immunoreactivity has been associated with disease
progression and a poorer prognosis (Bondi et al.
2004; Lin et al. 2000).
We investigated the WNT/b-catenin pathway
in a set of CNS PNETs, using immunohistochemistry (IHC) to determine the cellular location of
CTNNB1. This serves as a marker for pathway
status, where nuclear staining represents the
active state and cytoplasmic inactive (Eberhart
et al. 2000). The pathway target CCND1 was also
investigated by IHC and results correlated with
CTNNB1 localization. MKI67 (antigen identified by monoclonal antibody to Ki67) protein
levels were investigated to measure cell proliferation rates and compared to CTNNB1 and CCND1
data. The mutational status of exon 3 of CTNNB1
and the mutation cluster region of APC were
investigated by sequencing and correlated with
the IHC results. The pathway status was also
investigated in a set of medulloblastomas for
comparison. Results were correlated with clinical
information.
Methodology
Sample Cohort
Tumor samples were obtained from the Childrens
Cancer and Leukaemia Group (CCLG) and the
Cooperative Human Tissue Network (CHTN). A
total of 25 snap frozen CNS PNETs, all located in
the cerebral hemispheres, and 22 medulloblastomas were obtained. Five CNS PNETs were recurrences, four with the paired primary. Two
medulloblastomas were recurrences, one paired.
Eight pineoblastomas were also obtained, 6 primary and 2 recurrences (unpaired). Of the primary
medulloblastomas, 85% were classical, 10% desmoplastic and 5% anaplastic. The recurrent
tumors included one classical and one desmoplastic tumor. Medulloblastoma subtypes were
assigned according to the WHO criteria (Louis
et al. 2007). Two CNS PNETs and 4 pineoblastomas were obtained from CHTN. All other tumors
were obtained from CCLG. When cutting a piece
of frozen tissue for analysis a small piece was
77
taken and smeared along a slide which was subsequently stained with haematoxylin and eosin
(H&E) to determine whether the tissue contained
tumor cells.
Forty-two CNS PNETs (all cerebral) 46
medulloblastoma and 7 pineoblastoma samples
were fixed in 4% phosphate buffered formaldehyde and embedded in paraffin. Seven CNS
PNETs were recurrences, 5 with the paired primary. Three medulloblastomas were recurrent
tumors, one paired and two not. Of the primary
medulloblastomas, 44% were classical, 33% desmoplastic, 14% anaplastic, 7% large cell plus one
medullomyoblastoma. The recurrent tumors
included one classical, one anaplastic and one
medullomyoblastoma. Blood samples were
received for 5 CNS PNETs, 3 medulloblastomas
and 2 pineoblastomas. All paraffin tumor samples
were obtained from CCLG. Pineoblastomas were
included in the study due to their histopathological
similarities to other CNS PNETs (Louis et al.
2007). In the UK, pineoblastomas are also treated
with similar protocols to CNS PNETs (Pizer et al.
2006). For analysis they were included in the
CNS PNET cohort.
Clinical information including gender, age at
diagnosis, time to recurrence, date of death or last
follow up if still alive and metastatic status (using
the Chang staging system (Chang et al. 1969)),
was obtained from CCLG and CHTN. Multiple
Centre Research Ethics Committee (MREC)
approval was obtained for the study. Consent for
use of tumor samples was taken in accordance
with national tumor banking procedures and the
human tissue act.
Formalin fixed paraffin-embedded (FFPE) samples were analyzed on a tissue microarray
(TMA). Following review by a pathologist, representative areas of tumor tissue were selected.
Three cores from each tumor, taken from different locations in the section, were included on the
array. For IHC, slides were incubated at 37C
overnight, deparafinized in xylene and hydrated
though decreasing concentrations of ethanol.
78
Antigen retrieval was performed in a pressure

cooker for 1 min at full pressure in sodium citrate buffer (pH 6.0). An endogenous peroxide
block was performed (Dako, Cambs, UK). Slides
were incubated with either CTNNB1 (1:500, cell
signalling technology, Herts UK), CCND1
(1:100, abcam Cambridge UK) or MKI67
(MIB-1 clone, 1:50, Dako, Cambs UK) overnight at 4C. Antigen detection was carried out
using the Dako Chemate Envision Detection Kit
(Dako, Cambs, UK) with DAB chromagen used
for visualization, according to the manufacturers instructions.
Results for CTNNB1 were scored by location either as nuclear (pathway active), cytoplasmic or negative (pathway inactive). Samples
displaying nuclear staining were divided into
two groups depending on the percentage of positive nuclei. Those with <10% of nuclei positive
were labeled as low and those with >10% as
high. CCND1 and MKI67 were scored by calculating the percentage of positive cells.
Hundred cells were counted in 5 randomly chosen fields of view. CCND1 was considered to be
over-expressed if greater than 10% of cells were
positive. Lost cores or those where the majority
of tissue was necrotic were removed from the
analysis.
Mutational Analysis
DNA was extracted from 25 snap frozen CNS
PNETs, 22 medulloblastomas and 8 pineoblastomas. The pineoblastoma samples were included
in the CNS PNET cohort for analysis.
Constitutional DNA from 5 blood samples from
CNS PNET patients, 3 from medulloblastoma
and 2 from pineoblastoma patients was also
extracted. Five to ten milligram of tissue was
lysed in lysis buffer (50 mM Tris pH 8, 100 mM
EDTA pH 8, 100 mM NaCl, 1% SDS) and proteinase K (20 mg/mL) at 37C overnight. DNA was
obtained by phenol: chloroform extraction followed by isopropanol precipitation. Standard
PCR reactions were carried out using previously
published primers designed to amplify exon 3 of
CTNNB1 (Genbank accession number X89579)
(Koch et al. 1999). A combination of published

(Huang et al. 2000) and newly designed
primers (5 primer sequence TGCCACTTGC
AAAGTTTCTTC, 3 primer sequence CATTCC
ACTGCATGGTTCAC, annealing temperature
60C) were used to amplify the mutation cluster
region of APC (Genbank accession number
NM000038). PCR products were purified by
incubation with 0.3 U shrimp alkaline phosphatase (Promega, Hants UK) and 1.5 U exonuclease I (NEB, Herts UK) at 37C for 8 min
followed by 15 min at 72C. Sequencing reactions were performed on 1 mL purified PCR product using Big Dye V1.1 (Applied Biosystems,
Warrington UK), following the manufacturers
protocol.
Statistical Analysis
Association between clinical factors and immunohistochemical status was investigated using the
Fishers Exact Test. Overall, and progression
free, survival were investigated using the Kaplan
Meier method. The differences were estimated
using the log-rank (Mantel-Cox) test. Overall
survival was defined as the time between date of
original diagnosis and date of death. Progressionfree survival was defined as the time between
date of original diagnosis and date of first event
(recurrence or death). Patients still alive at the
end of the study were censored at the date of last
follow up. Median survival was estimated using
Kaplan Meier.
WNT/b-Catenin Pathway Status

in CNS PNET
The cellular location of CTNNB1 was investigated. Results were obtained for 28 primary
tumors in the CNS PNET cohort which included
5 pineoblastomas. Ten displayed CTNNB1
nuclear staining (36%) which included one
pineal tumor. Two patterns of nuclear staining
were noted. In the first only a small number of
nuclei were positive for CTNNB1. In the second

a large number of nuclei were positive for
CTNNB1 across the majority of the tissue analyzed. For scoring the groups were defined by
the percentage of CTNNB1 positive nuclei with
the low nuclear group containing <10% of positive nuclei and the high group containing >10%.
Six samples displayed high CTNNB1 nuclear
staining with four having >30% of nuclei positive. In the other four (including the pineoblastoma), low CTNNB1 nuclear positivity was seen
(Fig. 8.1a, c). Cytoplasmic staining was seen in
most tumors with only one tumor negative.
Results were also obtained for six recurrent
tumors. Three displayed high CTNNB1 nuclear
staining and three cytoplasmic. For two of the
recurrences the primary from the same patient
was analyzed. Both displayed the same staining
patterns. Concordance of results across all cores
was seen for all samples except one, where low
CTNNB1 nuclear staining was seen in only one
core. The sample was scored as low nuclear
CTNNB1.
CTNNB1 cellular location was determined
in 37 primary medulloblastomas with 10 displaying CTNNB1 nuclear staining (27%). The
same pattern of high and low nuclear staining
as the CNS PNET cohort was observed. High
CTNNB1 nuclear staining was seen for 3
tumors, plus 2 with focal high CTNNB1 nuclear
staining. All high CTNNB1 nuclear tumors displayed <30% of nuclei positive. Five tumors
displayed low CTNNB1 nuclear positivity
(Fig. 8.1b, d). Four tumors were negative, the
rest displayed cytoplasmic staining. A result for
only one recurrent sample was obtained which
was negative for CTNNB1. The primary sample
was not analyzed. Concordance across all cores
was seen for all medulloblastomas except 4.
Two displayed low CTNNB1 nuclear staining
in two cores and only cytoplasmic in the other
and were placed in the low CTNNB1 nuclear
group. One sample displayed focal high
CTNNB1 nuclear staining in two out of the
three cores. In the other focal high CTNNB1
staining was seen in one out of the three cores.
The latter two samples were placed in the high
CTNNB1 nuclear group.
79
CCND1 results were obtained for 27 primary

tumors in the CNS PNET cohort. Where positive, the protein was localized in the nucleus.
CCND1 was over-expressed in 12 samples
(44%, no pineoblastomas) with percentages of
positive cells ranging from 11% to 56%
(Fig. 8.1e). Of the 10 tumors with nuclear
CTNNB1, 8 (80%) displayed CCND1 over
expression, which included all tumors with high
CTNNB1 nuclear positivity. Four out of 17
(24%) tumor samples displaying cytoplasmic
or negative CTNNB1 staining also overexpressed CCND1. Six recurrent tumors were
analyzed, 3 with high CTNNB1 nuclear staining and 3 with CTNNB1 cytoplasmic staining.
All 6 displayed CCND1 over expression.
A significant correlation was found between
CTNNB1 nuclear positivity and CCND1 over
expression in the primary CNS PNETs (Fishers
Exact Test p = 0.007).
In medulloblastoma CCND1 over-expression
(nuclear) was seen in 4 out of 37 (11%) primary
tumors with percentages of positive cells ranging
from 14% to 22%. This included 3 out of 10
(30%) tumors with CTNNB1 nuclear staining,
one with low and two with a high level of staining. CCND1 was also over-expressed in one out
of 27 (4%) tumors with cytoplasmic or negative
CTNNB1 staining. An association between
CTNNB1 nuclear staining and CCND1 over
expression in the primary medulloblastomas just
below significance was identified (Fishers Exact
Test p = 0.052). However, 70% of tumors with
nuclear localization of CTNNB1 did not display
CCND1 over-expression.
MKI67, a proliferation marker, was investigated to see if CCND1 over expression was
affecting cell proliferation. MKI67 results were
obtained for 18 primary and 6 recurrent tumors in
the CNS PNET cohort. The number of positive
cells ranged from 0% to 42% (Fig. 8.1f). No correlation was found between MKI67 results and
CCND1, or CTNNB1 localization. MKI67 results
were obtained for 34 primary medulloblastomas
and one recurrence. The number of positive cells
ranged from 0% to 40%. No correlation was
found between MKI67 results and CCND1, or
CTNNB1 localization.
Fig. 8.1 Immunohistochemical analysis of CTNNB1,

CCND1 and MKI67 in the CNS PNET and medulloblastoma
cohorts. Two patterns of CTNNB1 nuclear staining were
seen; low where less than 10% of nuclei were positive (a CNS
PNET, b medulloblastoma) and high were greater than 10%
of nuclei were positive (c CNS PNET, d medulloblastoma).
Over expression of CCND1 was also seen in a subset of

tumors (e, CNS PNET). MKI67 levels were measured in both
cohorts (f, CNS PNET). Levels did not correlate with CCND1.
An additional CNS PNET sample containing a mutation in
CTNNB1 exon 3 was analyzed in a separate experiment and
displayed high CTNNB1 nuclear staining (g)
81
Fig. 8.2 Schematic representation of mutation locations in exon 3 of CTNNB1. Amino acid substitutions are indicated
above the sequence, grey changes represent mutations from medulloblastoma and black from CNS PNET
Mutational Analysis
In the CNS PNET cohort, only one of 26
primary tumors sequenced contained a mutation in exon 3 of CTNNB1 (4%) (Fig. 8.2). No
mutations were found in 6 recurrent samples.
The mutation was a missense point mutation at
codon 34 (GGA > CGA) converting glycine to
arginine. No blood samples contained mutations. The matching blood sample for the tumor
containing a mutation was not available for
sequencing. An IHC result for the CNS PNET
sample for which a mutation in CTNNB1 was
found was not obtained from the TMA due to
core drop out. However, high CTNNB1 nuclear
staining was seen in a separate experiment
(Fig. 8.1g). Four other primary tumors and one
recurrent tumor that displayed CTNNB1 nuclear
staining were sequenced with none containing
mutations.
Four of 20 primary medulloblastomas contained CTNNB1 mutations (20%) (Fig. 8.2). Four
recurrent samples were sequenced with none
containing mutations. All mutations were missense point mutations; one at codon 32
(GAC > TAC) converting aspartic acid to tyrosine;
two at codon 33 (TCT > TGT) converting serine
to cystine; and one at codon 34 (GGA > GAA)
converting glycine to glutamic acid. One sample
with a mutation at codon 33 also contained a missense point mutation at codon 40 (ACT > AGT)
converting threonine to serine. No blood samples
contained mutations. No blood samples from
patients with mutations in their tumors were
sequenced. There was only a small overlap in the
cohorts of medulloblastoma samples analyzed by
IHC and sequencing. Therefore, none of the samples displaying CTNNB1 nuclear staining was
sequenced and no IHC result was obtained for
any of the tumors containing mutations. No mutations were found in the mutation cluster region of
APC in 20 CNS PNET and 19 medulloblastoma
primary tumors sequenced. None of the blood
samples, from both tumor types, contained APC
mutations.
In the CNS PNET cohort, CTNNB1 nuclear cases
contained a higher proportion of males (male:
female ratio 4:1 compared to 0.6:1 in nonnuclear), and displayed a higher 5 year survival
rate (30% compared to 13%) than the non-nuclear
cases. However, no significant association was
seen for any clinical factor tested (Fishers Exact
Test). Analysis could be limited by the small
sample size (n = 28).
Comparison of all CNS PNET CTNNB1
nuclear cases to non-nuclear cases did not reveal
a significant difference in overall or progression
free survival (p = 0.852 and 0.536, respectively)
(Fig. 8.3a). However comparison of high
CTNNB1 nuclear cases to all other tumors (low
CTNNB1 nuclear plus cytoplasmic and negative
cases), although not significant (overall survival,
p = 0.113), suggested a trend towards the association of high CTNNB1 nuclear staining with a
more favorable outcome (Fig. 8.3b). Comparison
of cases with high CTNNB1 nuclear staining to
just those with a low level of nuclear staining did
reveal a significant difference in overall survival
(p = 0.007) (Fig. 8.3c). However, only limited
conclusions can be drawn due to the small number of samples analyzed (n = 10). The results
were supported by the 5 year overall survival
rates. Patients with a high level of CTNNB1
nuclear staining had a 5 year overall survival rate
82
Fig. 8.3 Kaplan Meier curves for analysis of CTNNB1

IHC. Comparison of CNS PNETs displaying nuclear
CTNNB1 to non-nuclear staining did not reveal a
significant difference in overall survival (p = 0.852) (a).
Comparison of high CTNNB1 nuclear tumors (high
nuclear) to the rest of the cohort (rest; tumors displaying
low nuclear, cytoplasmic or negative staining), although
not statistically significant (overall survival, p = 0.113),
suggested a trend towards a better prognosis for the high
CTNNB1 nuclear group (b).A significant difference in
overall survival was seen between CNS PNETs displaying

high and low CTNNB1 nuclear staining (overall survival,
p = 0.007) (c). In the medulloblastoma cohort comparison
of nuclear to non-nuclear CTNNB1 tumors was not significant (overall survival, p = 0.590) but suggested a trend
towards better survival for the nuclear group (d).
Comparison of CTNNB1 high nuclear cases (high nuclear)
to all other tumors in the medulloblastoma cohort (rest)
also suggested the same trend (overall survival, p = 0.310)
(e). All survival times are in months
of 50%, compared to 11% for the rest of the

cohort.
In the medulloblastoma cohort, association
between CTNNB1 nuclear immunoreactivity
and percentage of cases that had relapsed almost
reached significance (Fishers exact test,
p value = 0.056) with a lower percentage of
relapses seen in the nuclear cases. Although not
significant, there was a male bias in the
CTNNB1 nuclear cases (male: female ratio 9:1
compared to 2.9:1). Sixty percent of CTNNB1
nuclear cases were desmoplastic compared to
31% of non-nuclear tumors. CTNNB1 nuclear
immunoreactivity was not significantly linked
to overall or progression free survival (p = 0.590
and 0.517, respectively). However, the Kaplan
Meier curves suggested a difference (Fig. 8.3d).
This was also reflected in the overall survival
rates. At 5 years 56% of patients with CTNNB1

nuclear staining and 46% of patients with only
cytoplasmic or negative staining were still alive.
At 10 years the difference between survival
rates was greater with 56% of CTNNB1 nuclear
patients still alive and 24% of those with only
cytoplasmic or negative staining. It is possible
that significance was not reached due to the
relatively small sample size in this study
(n = 37). Comparison of cases displaying a high
level of CTNNB1 nuclear positivity to the rest
of the cohort was not significant (overall survival, p = 0.310), but suggested a better survival
for the high nuclear group (Fig. 8.3e). This was
supported by the difference in 5 year overall
survival rates of 80% for patients with high
CTNNB1 nuclear staining compared to 44% for
the rest of the cohort.
Discussion
We have extensively investigated the status of the
WNT/b-catenin pathway in CNS PNETs and
have demonstrated pathway activation in a high
proportion of tumors (36%), as well as suggested
a link between pathway activation and a more
favorable outcome. The high percentage of
tumors displaying activation suggests that the
pathway plays an important role in the pathogenesis of CNS PNETs, and is a potential target for
future therapies. Further investigation is needed
to validate findings and understand the biological
role the pathway is playing in tumorigenesis. An
equivalent rate of pathway activation was seen in
the medulloblastomas investigated in this study
(27%), in agreement with previous research
(Eberhart et al. 2000; Ellison et al. 2005).
Although a different CTNNB1 antibody was used
in these studies (BD Transduction Laboratories),
the agreement in the results suggests the two
alternative antibodies are comparable.
Nuclear localization of CTNNB1 was used to
determine pathway activation. The results were
supported by the significant correlation with
CCND1 over expression in both cohorts. CCND1
has previously been shown to be a target of the
WNT/b-catenin pathway (Tetsu and McCormick
1999). The evidence, although significant, was
not as strong in the medulloblastoma cohort, with
70% of tumors displaying nuclear localization of
CTNNB1 showing no CCND1 over-expression.
This included 3 out of 5 tumors with high nuclear
CTNNB1 expression. Correlation of CCND1
over-expression and CTNNB1 nuclear localization was not absolute in either cohort, with some
tumors displaying only cytoplasmic or negative
CTNNB1 staining over-expressing CCND1. This
could suggest an alternative factor is influencing
CCND1 over-expression. CCND1 expression has
been increased in other tumor types by gene
amplification or translocation, or control by alternative cell signaling pathways such as the sonic
hedgehog pathway (Marino 2005). However, the
significant correlation with CTNNB1 nuclear
localization is found, particularly in the CNS
PNET cohort, strongly suggesting that the
83
WNT/b-catenin pathway is increasing CCND1

expression in the tumors with pathway activation
in this study.
The correlation between CTNNB1 nuclear
localization and CCND1 over-expression suggests that WNT/b-catenin pathway activation is
affecting cell proliferation. However, no correlation was found between CCND1 and MKI67 in
either cohort. It may be that WNT/b-catenin
pathway activation is not affecting cell proliferation or that alternative mechanisms are affecting
the proliferation rate in tumors without pathway
activation and, thus masking any correlation that
could be found.
Unlike medulloblastomas, pathway activation in CNS PNETs does not seem to be caused
by mutations in exon 3 of CTNNB1 with only
one CNS PNET in this study containing a mutation. This tumor did display high nuclear staining of CTNNB1, suggesting that mutation could
be the cause of pathway activation in this
sample. The overlap in the cohorts used for IHC
and sequencing was relatively low. Results for
both methods were only obtained for 12 primary and 2 recurrent samples. This included 4
primary and one recurrent tumor with CTNNB1
nuclear staining. Six additional primary and 2
recurrent tumors with CTNNB1 nuclear staining were not sequenced. Therefore, no definite
conclusions can be drawn regarding whether
there is a correlation between CTNNB1 mutation and nuclear staining. However, only one
mutation was found in 32 tumors sequenced,
which included 17 tumors with no IHC result
and, therefore, no known WNT/b-catenin pathway status, suggesting CTNNB1 exon 3 mutation to be rare.
Only four medulloblastomas had both IHC
and sequencing results, all displaying cytoplasmic CTNNB1 staining and containing no mutations in CTNNB1 exon 3. Therefore, it cannot be
concluded whether there is a correlation between
nuclear staining and mutation of CTNNB1.
However, the overall mutation rate identified
(20%) was similar to that found previously where
a correlation was reported, suggesting this is
likely to be the case in this study (Clifford et al.
84
2006; Ellison et al. 2005; Thompson et al. 2006).

Matching blood samples were not available for
any of the CNS PNET or medulloblastoma
tumors containing mutations so it is not known
whether these are somatic or constitutional.
The mutations detected in the medulloblastomas are consistent with those described in earlier studies (Clifford et al. 2006; Eberhart et al.
2000; Ellison et al. 2005; Koch et al. 2001;
Thompson et al. 2006). The mutation in the single CNS PNET identified here is in the same
codon as one of the medulloblastomas and a
CNS PNET in a previous study, where a missense point mutation caused a glycine to valine
substitution (Koch et al. 2001). The substitution
of glycine to arginine, found in the CNS PNET
in this study, has not been seen in CNS PNETs
before but has been found in medulloblastoma
and pancreatic tumors (Abraham et al. 2002;
Haberler et al. 2008). All the mutations altered
residues that are phosphorylation sites for GSK3b (codon 33), or adjacent residues (codons 32,
34 and 40). These are predicted to prevent
phosphorylation of CTNNB1 by GSK-3b and
therefore prevent its degradation.
In colon cancer, the WNT pathway is commonly activated by mutations in the mutation
cluster region of APC. However, no mutations
were found in this region in the CNS PNET or
medulloblastoma cohorts. Further investigation
to determine the molecular basis of pathway activation is needed to help understand if it is playing
a role in disease development. It is possible that
mutations are present in other regions of the
CTNNB1 or APC genes not investigated in this
study. Alternative factors that could cause pathway activation in CNS PNETs include inactivating mutations in the pathway inhibitors AXIN1
and AXIN2. Mutations have been identified in
both genes in different tumor types including
medulloblastoma and colon carcinoma (Baeza
et al. 2003; Koch et al. 2007; Liu et al. 2000).
WNT ligands or their receptors could also be
over-expressed. Previous studies have identified
increased expression of WNT and frizzled receptor genes in different tumor types (Janssens et al.
2004; Merle et al. 2004). Epigenetic alterations
have also been identified, including inactivation
of secreted frizzled related protein (sFRP) genes

in medulloblastoma and colorectal cancer
(Kongkham et al. 2010; Suzuki et al. 2004).
Interaction of WNT/b-catenin signaling with
other signaling pathways has been demonstrated
to affect the levels of signaling as well (Saldanha
et al. 2004).
Although not statistically significant, survival
analysis in the CNS PNET cohort suggested a
trend towards a better prognosis for patients
whose tumors displayed high CTNNB1 nuclear
staining. Significance was achieved when high
CTNNB1 nuclear tumors were compared to low
CTNNB1 nuclear tumors but was limited by the
very small number of samples included in the
analysis. Together with the differences in 5 year
overall survival rates the data suggested a higher
level of pathway activation was linked to a better
outcome and is in agreement with the association
of better prognosis with WNT/b-catenin pathway
activation previously found in medulloblastoma
(Ellison et al. 2005). The number of samples
included in this analysis was relatively low therefore further investigation is needed for confirmation of results.
No significant association was found for
medulloblastomas in this study between pathway
activation and survival. However the Kaplan
Meier curves and overall survival rates suggest a
trend towards better survival for patients with
the pathway active in their tumors. Other clinical
factors also suggest this, including the higher
percentage of relapses in non-nuclear cases.
Significance would need to be examined in a
larger cohort.
Association of pathway activation with a
favorable prognosis is somewhat surprising, considering its link to a poorer outcome and disease
progression in other tumor types, including colon
and breast carcinoma (Bondi et al. 2004; Lin
et al. 2000). However an association with a better prognosis has been seen before in medulloblastoma and other tumor types such as non-small
cell lung carcinoma and ovarian cancer (Catasus
et al. 2004; Ellison et al. 2005; Hommura et al.
2002). There could be a number of reasons for
this link in CNS PNET and medulloblastoma.
Pathway activation could represent a subset of
tumors with a less aggressive phenotype than

other subtypes. Activation of the WNT/b-catenin
pathway can have many different effects on a
cell including influencing proliferation, apoptosis and differentiation. Therefore, pathway activation could be causing a deleterious effect such
as promoting apoptosis. Alternatively, pathway
activation could affect sensitivity to treatment. A
recent study in medulloblastoma cell lines demonstrated activation of the pathway following
irradiation (Salaroli et al. 2008). The high proportion of tumors, displaying pathway activation
in both CNS PNET and medulloblastoma suggest it could be an important treatment target.
However, the reason for the association with
favorable prognosis needs to be understood
before strategies for targeting the pathway are
developed.
A high proportion of the medulloblastomas
displaying nuclear CTNNB1 staining were of the
desmoplastic subtype which differs from previous results (Thompson et al. 2006). Desmoplastic
tumors have been associated with a better survival
than the classic subtype which might suggest this
is the cause of better prognosis in the nuclear
positive cases in this study (Sure et al. 1995). No
statistically significant association between desmoplastic cases and survival was found in this
study. However median survival for desmoplastic
cases was greater than the rest of the cohort
(12 years vs. 2 years) suggesting the lack of statistical significance may be due to the small sample size (n = 37). Median survival for desmoplastic
cases with nuclear CTNNB1 staining was greater
than desmoplastic cases with only cytoplasmic or
negative staining (7 years vs. 3.5 years). Although
not significant this data suggests tumor subtype
was not the cause of better prognosis for the
CTNNB1 nuclear tumors. The most comprehensive analysis of WNT status and survival in
medulloblastomas did not include any desmoplastic tumors (Ellison et al. 2005). The results
from this study suggest future cohorts investigated should include this tumor subtype.
Interestingly all the medulloblastomas that contained mutations were of the classical subtype.
However only 10% of the samples sequenced
were desmoplastic.
85
Association between WNT/b-catenin pathway

activation and chromosome 6 loss has previously
been found in medulloblastoma (Clifford et al.
2006; Thompson et al. 2006). Copy number data
generated from AffymetrixTM SNP chip analysis
for 12 of the CNS PNETs used in this study
(6 nuclear and 6 cytoplasmic CTNNB1) suggested this is not the case for this tumor type.
Only one tumor displaying cytoplasmic CTNNB1
staining had a loss of one copy of chromosome 6
(S Miller, un-published data).
In summary, we found WNT/b-catenin pathway activation in more than one third of CNS
PNETs, suggesting it plays an important role in
the pathogenesis of this tumor type. The percentage of samples displaying pathway activation
was similar to results seen in medulloblastoma.
However, the method of activation appears to differ from mutation of exon 3 of CTNNB1. The
data have also revealed a potential link between
survival and the extent of pathway activation in
CNS PNETs. The mechanism of activation as
well as the role the pathway is playing in the
pathogenesis of these tumors need to be
determined to better understand their biology as
well as help to decide how the pathway could be
targeted as part of future treatment strategies.
References
Abraham SC, Klimstra DS, Wilentz RE, Yeo CJ, Conlon K,
Brennan M, Cameron JL, Wu TT, Hruban RH (2002)
Solid-pseudopapillary tumors of the pancreas are
genetically distinct from pancreatic ductal adenocarcinomas and almost always harbor beta-catenin mutations. Am J Pathol 160:13611369
Baeza N, Masuoka J, Kleihues P, Ohgaki H (2003) AXIN1
mutations but not deletions in cerebellar medulloblastomas. Oncogene 22:632636
Bondi J, Bukholm G, Nesland JM, Bukholm IR (2004)
Expression of non-membranous beta-catenin and
gamma-catenin, c-Myc and cyclin D1 in relation to
patient outcome in human colon adenocarcinomas.
APMIS 112:4956
Catasus L, Bussaglia E, Rodrguez I, Gallardo A, Pons C,
Irving JA, Prat J (2004) Molecular genetic alterations
in endometrioid carcinomas of the ovary: similar frequency of beta-catenin abnormalities but lower rate of
microsatellite instability and PTEN alterations than in
uterine endometrioid carcinomas. Hum Pathol
35:13601368
86
Chang CH, Housepian EM, Herbert C Jr (1969) An operative staging system and a megavoltage radiotherapeutic technic for cerebellar medulloblastomas. Radiology
93:13511359
Clifford SC, Lusher ME, Lindsey JC, Langdon JA,
Gilbertson RJ, Straughton D, Ellison DW (2006) Wnt/
Wingless pathway activation and chromosome 6 loss
characterize a distinct molecular sub-group of
medulloblastomas associated with a favorable prognosis. Cell Cycle 5:26662670
Eberhart CG, Tihan T, Burger PC (2000) Nuclear localization and mutation of beta-catenin in medulloblastomas. J Neuropathol Exp Neurol 59:333337
Ellison DW, Onilude OE, Lindsey JC, Lusher ME, Weston
CL, Taylor RE, Pearson AD, Clifford SC (2005) Betacatenin status predicts a favorable outcome in childhood medulloblastoma: the United Kingdom
Childrens Cancer Study Group Brain Tumour
Committee. J Clin Oncol 23:79517957
Group. J Clin Oncol 23:76217631
Haberler C, Varlet P, Legoix P, Fattet S, Janoueix-Lerosey
I, Lellouch-Tubiana A, Grill J, Doz F, Sainte-Rose C,
Delattre O (2008) Widespread nuclear b-catenin
expression in medulloblastoma correlates with mutation status of CTNNB1 gene encoding b-catenin
[abstract]. Neuro-oncology 10:470
Hamilton SR, Liu B, Parsons RE, Papadopoulos N, Jen J,
Powell SM, Krush AJ, Berk T, Cohen Z, Tetu B,
Berger PC, Wood PA, Taqi F, Booker SV, Petersen
GM, Offerhaus JA, Tersmette AC, Giardiello FM,
Vogelstein B, Kinzler KW (1995) The molecular basis
of Turcots syndrome. N Engl J Med 332:839847
He TC, Sparks AB, Rago C, Hermeking H, Zawel L, da
Costa LT, Morin PJ, Vogelstein B, Kinzler KW (1998)
Identification of c-MYC as a target of the APC pathway. Science 281:15091512
Hommura F, Furuuchi K, Yamazaki K, Ogura S, Kinoshita
I, Shimizu M, Moriuchi T, Katoh H, Nishimura M,
Dosaka-Akita H (2002) Increased expression of betacatenin predicts better prognosis in nonsmall cell lung
carcinomas. Cancer 94:752758
Huang H, Mahler-Araujo BM, Sankila A, Chimelli L,
Yonekawa Y, Kleihues P, Ohgaki H (2000) APC mutations in sporadic medulloblastomas. Am J Pathol
156:433437
Janssens N, Andries L, Janicot M, Perera T, Bakker A
(2004) Alteration of frizzled expression in renal cell
carcinoma. Tumour Biol 25:161171
Koch A, Denkhaus D, Albrecht S, Leuschner I, von
Schweinitz D, Pietsch T (1999) Childhood hepatoblastomas frequently carry a mutated degradation targeting box of the beta-catenin gene. Cancer Res
59:269273

Koch A, Waha A, Tonn JC, Sorensen N, Berthold F,
Wolter M, Reifenberger J, Hartmann W, Friedl W,
Reifenberger G, Wiestler OD, Pietsch T (2001)
Somatic mutations of WNT/wingless signaling pathway components in primitive neuroectodermal tumors.
Int J Cancer 93:445449
Koch A, Hrychyk A, Hartmann W, Waha A, Mikeska T,
Waha A, Schuller U, Sorensen N, Berthold F, Goodyer
CG, Wiestler OD, Birchmeier W, Behrens J, Pietsch T
(2007) Mutations of the Wnt antagonist AXIN2
(Conductin) result in TCF-dependent transcription in
medulloblastomas. Int J Cancer 121:284291
Kongkham PN, Northcott PA, Croul SE, Smith CA, Taylor
MD, Rutka JT (2010) The SFRP family of WNT
inhibitors function as novel tumor suppressor genes
epigenetically silenced in medulloblastoma. Oncogene
29:3017. doi:10.1038/onc.2010.32
Lin SY, Xia W, Wang JC, Kwong KY, Spohn B, Wen Y,
Pestell RG, Hung MC (2000) Beta-catenin, a novel
prognostic marker for breast cancer: its roles in cyclin
D1 expression and cancer progression. Proc Natl Acad
Sci USA 97:42624266
Liu W, Dong X, Mai M, Seelan RS, Taniguchi K,
Krishnadath KK, Halling KC, Cunningham JM,
Boardman LA, Qian C, Christensen E, Schmidt SS,
Roche PC, Smith DI, Thibodeau SN (2000) Mutations
in AXIN2 cause colorectal cancer with defective mismatch repair by activating beta-catenin/TCF signalling. Nat Genet 26:146147
Louis DN, Ohgaki H, Wiestler OD, Cavenee WK,
Burger PC, Jouvet A, Scheithauer BW, Kleihues P
(2007) The 2007 WHO classification of tumours of
the central nervous system. Acta Neuropathol (Berl)
114:97109
Marino S (2005) Medulloblastoma: developmental mechanisms out of control. Trends Mol Med 11:1722
Merle P, de la Monte S, Kim M, Herrmann M, Tanaka S,
Von Dem Bussche A, Kew MC, Trepo C, Wands JR
(2004) Functional consequences of frizzled-7 receptor
overexpression in human hepatocellular carcinoma.
Gastroenterology 127:11101122
Miyoshi Y, Nagase H, Ando H, Horii A, Ichii S, Nakatsuru
S, Aoki T, Miki Y, Mori T, Nakamura Y (1992)
Somatic mutations of the APC gene in colorectal
tumors: mutation cluster region in the APC gene. Hum
Mol Genet 1:229233
Morin PJ (1999) Beta-catenin signaling and cancer.
Bioessays 21:10211030
Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H,
Vogelstein B, Kinzler KW (1997) Activation of betacatenin-Tcf signaling in colon cancer by mutations in
beta-catenin or APC. Science 275:17871790
Pizer BL, Weston CL, Robinson KJ, Ellison DW, Ironside
J, Saran F, Lashford LS, Tait D, Lucraft H, Walker DA,
Bailey CC, Taylor RE (2006) Analysis of patients with
supratentorial primitive neuro-ectodermal tumours
entered into the SIOP/UKCCSG PNET 3 study. Eur J
Cancer 42:11201128
Reddy AT, Janss AJ, Phillips PC, Weiss HL, Packer RJ
(2000) Outcome for children with supratentorial
primitive neuroectodermal tumors treated with surgery,

radiation, and chemotherapy. Cancer 88:21892193
Salaroli R, Di Tomaso T, Ronchi A, Ceccarelli C,
Cammelli S, Cappellini A, Martinelli GN, Barbieri E,
Giangaspero F, Cenacchi G (2008) Radiobiologic
response of medulloblastoma cell lines: involvement
of beta-catenin? J Neurooncol 90:243251
Saldanha G, Ghura V, Potter L, Fletcher A (2004) Nuclear
beta-catenin in basal cell carcinoma correlates with
increased proliferation. Br J Dermatol 151:157164
Sure U, Berghorn WJ, Bertalanffy H, Wakabayashi T,
Yoshida J, Sugita K, Seeger W (1995) Staging, scoring
and grading of medulloblastoma. A postoperative
prognosis predicting system based on the cases of a
single institute. Acta Neurochir (Wien) 132:5965
87
Suzuki H, Watkins DN, Jair KW, Schuebel KE, Markowitz

SD, Chen WD, Pretlow TP, Yang B, Akiyama Y, Van
Engeland M, Toyota M, Tokino T, Hinoda Y, Imai K,
Herman JG, Baylin SB (2004) Epigenetic inactivation
of SFRP genes allows constitutive WNT signaling in
colorectal cancer. Nat Genet 36:417422
Tetsu O, McCormick F (1999) Beta-catenin regulates
expression of cyclin D1 in colon carcinoma cells.
Nature 398:422426
Thompson MC, Fuller C, Hogg TL, Dalton J, Finkelstein D,
Lau CC, Chintagumpala M, Adesina A, Ashley DM,
Kellie SJ, Taylor MD, Curran T, Gajjar A, Gilbertson
RJ (2006) Genomics identifies medulloblastoma subgroups that are enriched for specific genetic alterations. J Clin Oncol 24:19241931
Neuroblastic Tumors Status

and Role of HER Family Receptors
Ewa Izycka-Swieszewska and Agnieszka Wozniak
Contents
Abstract
Introduction ............................................................
89
Neuroblastic Tumors..............................................
90
HER Receptor Characteristics .............................
90
HERs in Development of Peripheral

Nervous System ......................................................
91
HERs in Cancer......................................................
91
HER Family in Neuroblastic Tumors ...................

EGFR .......................................................................
HER2........................................................................
HER3........................................................................
HER4........................................................................
Coexpression Profiling............................................ .
92
92
94
95
95
97
References ...............................................................
97
E. Izycka-Swieszewska ()
Department of Nursing Management and
Pathomorphology, Medical University of Gdansk,
Debinki 7 Street, 80-211, Gdansk, Poland
e-mail: eczis@wp.pl
A. Wozniak
Laboratory of Experimental Oncology and Department
of General Medical Oncology, KU Leuven and
University Hospitals, Herestraat 49 post 815,
Leuven, Belgium
e-mail: Agnieszka.Wozniak@med.kuleuven.be
Neuroblastic tumors (NBs) show diverse

clinical presentation ranging from a regression, through delayed or arrested maturation,
to the metastatic spread. The different biological behavior of NBs seems to be determined
by the molecular mechanisms associated with
neural crest development. HER receptors family which modulates neurogenesis, is involved
in genesis and progression of several types
of cancer. HER receptors are commonly
expressed in NB tumors and it seems that
HER family members play an interrelated and
complex role in their biology. It is suggested
that HERs are related to the neuroblastic cell
differentiation and Schwannian stroma development. Moreover, HER expression is considered a potential prognostic factor. Because
new treatment options are still warranted in
high risk NBs, HER receptors become an
evolving target for novel therapeutic
approaches such as tyrosine kinase inhibitors
or specific antibodies.
Introduction
Neuroblastic tumors (NBs) are common pediatric embryonal neoplasms which originate from
the peripheral neural crest, developing within
the adrenal medulla and in the sympathetic ganglia (Maris et al. 2007). These tumors show
diverse clinical presentation ranging from a

89
90
spontaneous regression, through delayed or

arrested maturation, to metastatic dissemination. The development of new treatment modalities that would target the pathways responsible
for malignant transformation and progression is
necessary especially for high risk NBs (Fong
and Park 2009). The different biological behavior of NBs seems to be determined by the
molecular mechanisms associated with neural
crest development, such as TRK family receptors and NMYC (Nakagawara 2004). HER
receptors family which modulates neurogenesis
is involved in genesis and progression of several types of cancers. In addition, HER receptors constitute modern therapeutic target. The
increasing data show that these receptors are
connected to the biology of NBs.
Neuroblastic Tumors
NBs account for up to 10% of all childhood
cancers, and are the most common type of
malignancy diagnosed during the infancy.
Approximately, 40% of patients present with
localized disease and about half of children have
metastatic disease at the time of diagnosis, with
secondary foci in bone, liver, and distal lymph
nodes. Finally, 5% of children- infants have stage
4s neuroblastoma with small localized tumors,
disseminated into the liver, skin, or bone marrow,
that have a tendency towards a spontaneous
regression (Maris et al. 2007).
NBs are remarkable for their diverse pathological picture concerning the level of neoplastic
cell differentiation and Schwannian stroma
amount. Neuroblastoma cells may follow a
variety of neural lineages: adrenal chromaffin,
extraadrenal chromaffin, and sympathetic
ganglionic line. The basic histological NB
categories include: neuroblastoma (Schwannian
stroma poor), intermixed ganglioneuroblastoma
(Schwannian stroma rich), ganglioneuroma
(Schwannian stroma predominant), and ganglioneuroblastoma nodular (Schwannian stroma
poor/rich/predominant) (Shimada et al. 1999).
The Shimada classification system, based on
tumor histology, mitosis/karryorhexis index
E. Izycka-Swieszewska and A. Wozniak
(MKI) and patients age, is widely accepted for

histological prognostic assessment in NBs.
Current standard for NBs treatment ranges
from the wait-and-see approach to the
complex, multimodal therapy (surgery, chemotherapy, radiotherapy, bone marrow transplantation) which lasts for several months. The newest
adjuvant regimens include immunotherapy with
anti-GD2 antibodies, retinoids, and antiangiogenic drugs (Fong and Park 2009). The therapy
planning is directed by the clinicopathological
and molecular risk factors. In general, the longterm prognosis of children older than 18 months
with metastatic disease is dismal. Moreover,
NMYC amplification found in approximately
20% of NBs strongly correlates with advanced
stage of disease and poor patients prognosis. The
clinicopathological classification system established by International Neuroblastoma Risk
Group (INRG) includes many prognostic factors
such as patients age, tumor stage, histology,
ploidy, NMYC, and 11q status (Cohn et al. 2009).
In the near future the application of a multigene
expression signatures and epigenetic markers,
may serve as reliable biomarkers of individual
prognosis (Vermuelen et al. 2010). Hence, the
understanding of NB biology and identification
of new tumor biomarkers may generate novel
therapeutic strategies.
HER Receptor Characteristics

HER proteins family is composed of four transmembrane growth factor receptors: EGFR (also
known as ERBB1, HER1), HER2 (ERBB2),
HER3 (ERBB3), and HER4 (ERBB4). These
receptors contain an extracellular ligand-binding
domain, a hydrophobic transmembrane component, and an intracellular part with tyrosine kinase
activity. All HERs share structural homology, but
their functional properties are different (Citri and
Yarden 2006). EGFR and HER4 have active
tyrosine kinase domain and known ligands. On
the other hand, HER3 lacks the internal tyrosine
kinase activity while for HER2 no direct ligand is
identified. The family of ligands for HER can be
divided into three groups: the first consists of
Neuroblastic Tumors Status and Role of HER Family Receptors
specific EGFR ligands such as EGF, transforming growth factor alpha and amphiregullin. The
second includes betacellulin, heparin binding
EGF, and epiregulin, which show dual EGFR and
HER4 specificity. Finally, the third group is composed of the neuregulins (Casalini et al. 2004).
The activation of HERs by ligand binding,
causes receptor homo- or heterodimerization,
leading to the phosphorylation of the receptors
internal domain. Such activation triggers important intracellular signaling pathways, mainly
PI3K/AKT and RAF/MEK/MAPK (Moasser
2007). Integrated cellular signaling through HER
receptors regulates a wide range of cellular functions, such as cell proliferation, migration, angiogenesis, differentiation, and survival (Rogers
et al. 2009). Both EGFR and HER2 can activate
PI3K through interaction with adaptor proteins,
albeit it can do it by cooperation with HER3 or
HER4. (Citri and Yarden 2006; Baselga and
Swain 2009). HER2 has the hierarchical catalytic
importance within the EGFR family network
because HER2 exists in a constitutively active
state (Moasser 2007). Effects of HER- mediated
signaling depend on cross-talk between receptors, modulation by other tyrosine kinases and
integrins, and horizontal interactions between
second messengers. Moreover, compartmentalization is a central mechanism that controls output from the ERBB network. For example
internalized receptors might couple effectors in
predegradative intracellular compartments, and
activate pathways distinct from those triggered at
the cell surface (Citri and Yarden 2006).
HERs in Development of Peripheral

Nervous System
HER family is involved in the embryonic development of many organs, including nervous system and heart (Casalini et al. 2004). These
receptors are temporo-spatial regulators of peripheral nervous system including sympathetic chain
and adrenal medulla. The acquisition of mitotic
responsiveness to ligands is associated with
appearance of subpopulation of cells with high
membranous EGFR expression. HER2 plays a
91
pivotal role in embryogenesis dictating the establishment of several cell lineages through mesenchyme- epithelial- neuroectodermal inductive
processes. HER2 and HER3 are essential for
NRG1 signaling, being basically important in
driving migration of neural crest- derived sympathoadrenal progenitors and sensitizing them to
differentiation signals (Morris et al. 1999). HER2null mutant mice have improper differentiation of
neural crest derived neurons, reduction of
Schwann cell number, and maldevelopment of the
neuro-muscular junction. HER3-null mice lack
sympathetic ganglia and present neural migration
defects, depletion of Schwann cells and adrenal
chromaffin cells. Aberrant structure of peripheral
nervous system, cause of lethality, is reported in
HER4 mice knock-outs (Casalini et al. 2004).
HERs in Cancer
HER 14 are implicated in the development and
progression of various malignant tumors, by
influencing cell cycle, sustaining survival and
enhancing invasion of cancer cells. The significance and clinicopathological associations of
HERs expression and gene status are receptordependent and different in different tumor types
(Rogers et al. 2009; Baselga and Swain 2009). In
some tumors HER receptors expression has some
associations with its gene status; in addition, their
location and sorting are regulated by adaptor and
scaffolding proteins.
EGFR was the first tyrosine kinase receptor
which was directly linked to tumorigenesis in
humans. However, the EGFR gene and protein
status associations with patients prognosis and
therapeutic response are still not fully identified.
Up-regulated EGFR signaling through EGFR
gene amplification, leading to the receptor overexpression is found in lung cancer, gliomas, and
head and neck cancer (Rogers et al. 2009; Hirsch
et al. 2009). In some neoplasms EGFRvIII variant
is detected (Moscatello et al. 1995) while EGFR
somatic point mutations are identified in others
including nonsmall cell lung cancer. Moreover,
the presence of the latter predisposes to an antiEGFR treatment response (Lynch et al. 2004).
92
Amplification of HER2, leading to protein

overexpression, detected at first in a subset of
breast cancer was also identified in other human
neoplasms such as ovarian and gastric tumours.
The somatic point mutations of HER2 were
reported in a small subset of several tumors
(Moasser 2007). Membranous protein overexpression and/or amplification of HER2 gene
found in several types of cancer is usually associated with a poor prognosis. Conversely, adverse
significance of HER2- negativity in cancer has
also been shown. Interestingly, HER2 is not
evenly distributed on cellular surface, but may be
localized in subcellular compartments such as in
perisynaptic region. Moreover, another form of
HER2 is its truncated version (p95 or ERBB2
C-terminal fragments)- a potent oncoprotein, which
constitutively homodimerizes and regulates the
expression of many genes (Baselga and Swain
2009; Moasser 2007).
HER3 has an important role in driving
oncogenic cellular proliferation and survival in
several human tumors. HER3-depending signaling cannot occur in the absence of a kinasefunctional dimerization partner, preferentially
HER2. High levels of HER3 expression might be
linked to more aggressive disease course and significantly reduced survival. The recent findings
suggest that HER3 signaling can be up-regulated
to compensate for the inhibition of other HER
family members (Baselga and Swain 2009).
The role of HER4 in tumorigenesis is complex, showing associations with cell differentiation, but also proliferation. The differences in
the activation of signaling cascades are reflected
by the cellular responses stimulated via cytoplasmic HER4 isoforms (Junttila et al. 2000). In
a subset of cancers, HER4 amplification with
overexpression is encountered while in the other
tumors immunonegativity is reported. Also a
few studies show prognostic significance of
HER4 (Gullick 2003).
In adult tumors, HER receptor family members, especially EGFR and HER2 represent validated targets for biological therapy. Pre-clinical
studies proved that treatment with ERBB-targeted
tyrosine kinase inhibitors (TKI) or antibodies
rapidly down-regulated PI3K/AKT, MAPK, SRC
and STAT signaling pathways, leading to the

inhibition of cells proliferation.
Several TKIs such as gefitinib, erlotinib or
monoclonal antibodies (e.g., trastuzumab, cetuximab) have been already approved for cancer
treatment and several others have entered clinical
trials (Lynch et al. 2004; Hirsch et al. 2009).
EGFR and HER2 inhibitors are also known to
potentiate, by counteracting the prosurvival
signaling, the effects of cytotoxic drugs, and
radiotherapy in cancer (Michaelis et al. 2008).
There are also novel therapeutic anti-HER
strategies that include heat shock protein 90
(HSP90) inhibitors, HER dimerization inhibitors, and antibody-chemotherapy conjugates
(Baselga and Swain 2009).
In contrast to adult cancer, the knowledge of
HER family role in pediatric tumours is not so
very broad. In several solid tumor subtypes,
different HERs were detected both at mRNA and
protein levels (osteosarcoma, rhabdomyosarcoma, Wilms tumor). In pediatric malignancies,
HER expression is frequently cytoplasmic, unlike
to membranous staining typically found in adult
tumors. However, the overexpression was rarely
due to genetic amplification, and the receptor
presence does not always directly relate to its
functionality (Gilbertson 2005). Anti-HER targeted therapy in pediatric oncology is a promising new tool for patients survival improvement,
but it is still being investigated and is a subject of
on-going clinical trials (Jakacki et al. 2008).
HER Family in Neuroblastic Tumors

EGFR
In neural cell cultures, EGFR promotes initial
proliferation, but at later phases induces neuronal
differentiation and stimulates the survival and
neurite process outgrowth (Evangelopoulos et al.
2009). Interestingly, some NB cell lines differentiate after serum deprivation from the culture
medium by activating EGFR and MAPK signaling. Moreover, increased EGFR expression occurs
in chemoresistant NB cells or can be induced by
chemotherapy (Michaelis et al. 2008).
93
Fig. 9.1 Membranous EGFR expression in poorly differentiated neuroblastoma and high EGFR staining within
ganglion cells and Schwannian stroma in ganglioneuroblastoma (EGFR, 400)
The data on genetic changes within EGFR

gene in NB are scant. So far EGFR somatic mutations were not detected in any NB cell lines
(Richards et al. 2010; Rossler et al. 2009) or
tumors samples analyzed (Izycka-Swieszewska
et al. 2010a). Available mutation analyses concerned the most commonly mutated codons
1821. In the only one study performed on NBs
clinical samples, three EGFR polymorphisms
were detected, with the frequency similar to
healthy Caucasian population, and there were no
differences between groups with different clinicopathological characteristics. FISH study
showed no EGFR amplification, but in 26% of
cases the polysomy was found, which was not
associated with any clinicopathological features
of NB and no relations between gene copy status
and protein expression, were revealed (IzyckaSwieszewska et al. 2010a).
The expression of EGFR in NB tissue was
the subject of a few studies (Layfield et al.
1995; Ho et al. 2005; Tamura et al. 2007;
Richards et al. 2010). EGFR expression at
mRNA level detected in NB cell lines and in
short series of tumor tissues was up to the level
of fetal brain. Frequent EGFR expression in NB
cell lines was also proved by Western immunoblotting and flow cytometry. The absence of
EGFR protein in some cases detected by cytometric analysis was explained through receptor
recirculation (Richards et al. 2010). In one
study the post- transcriptional down-regulation
of EGFR protein expression was suggested
because mRNA was detected in six NB cell
lines, but protein expression was found only in
one of them (Rossler et al. 2009). In the clinical
tumor samples, membranous and cytoplasmic
immunohistochemical labeling pattern with
frequent high expression was observed. The
stage of neuroblastic cell differentiation
reflected membranous EGFR labeling in poorly
differentiated NB, whereas membrano-cytoplasmic and axonal staining in differentiating
and maturing tumors (Fig. 9.1). Low staining of
Schwannian stroma was also observed in some
cases (Izycka-Swieszewska et al. 2011).
Richards et al. (2010) also suggested the influence of EGFR on neuroblastic cells- stromal
interactions and angiogenesis. The other clinicopathological markers which correlated with
EGFR expression included MKI, histological
risk, and proliferation index. High cytoplasmic
expression was more often connected with
favorable Shimada category, low MKI, and
lower proliferation index (Izycka-Swieszewska
et al. 2011). Prognostic significance of EGFR
94
Fig. 9.2 HER2 labeling in scattered neuroblastic cells in poorly differentiated neuroblastoma (left- HER2, 200) and
spectrum of HER2 expression in ganglioneuroblastoma (right- HER2, 400)
expression, or correlation with risk group,

stage, and NMYC status have not been found in
any results published so far.
Some in vitro and in vivo studies show that
EGFR mediates proliferation of NB cells, which
can be selectively inhibited by gefitinib and erlotinib in a dose- dependent manner (Ho et al. 2005;
Tamura et al. 2007). Nevertheless, the pan ERBB
inhibitor CI-1033 showed more potent in vivo
inhibition effect than single target inhibitors
(Richards et al. 2010). Experimental studies
suggest that EGFR- targeted cytotoxic reagents
represent promising therapeutic method, especially in combination with cisplatin (Tamura
et al. 2007; Michaelis et al. 2008). Jakacki et al.
(2008) reported a beneficial effect of erlotinib as
a single agent or in combination with temozolomide in phase I study concerning children with
refractory solid tumors, including few cases of
neuroblastoma.
HER2
HER2 was initially isolated from an ethylnitrosurea-induced rat neuroblastoma. In NBs, HER2
expression was detected immunohistochemically
in 1380% of cases (Layfield et al. 1995; Goji

et al. 1995) as membranous and cytoplasmic
staining. The expression was further confirmed
by PCR, Western immunoblotting, and mRNA
studies (Ho et al. 2005; Richards et al. 2010).
HER2 labeling pattern reflects the phenotype of
developing peripheral neurons and may be
involved in proliferation and/or differentiation of
neuroblastic cells by autocrine or paracrine fashion (Goji et al. 1995; Gambini et al. 2003). In
NBs, HER2 signaling seems to be related with
differentiation and Schwannian stroma development. It is similar to the embryogenesis of the
peripheral nervous system, where early Schwann
cells precursors rely on axonal neuregulins signals for maintenance and proliferation (Morris
et al. 1999). In our studies, HER2 expression
differed significantly between NBs histological
categories and subtypes. HER2- positivity correlated with favorable Shimada category. The
undifferentiated neuroblastic cells were immunonegative while the most intense membranocytoplasmic labeling concerned maturing
neuroblasts/ganglioid cells (Fig. 9.2). Mature
ganglion cells as well as developing Schwannian
stroma had mainly low membranous staining.
Furthermore, HER2 immunonegativity was more
frequently found in NMYC-amplified and in metastatic tumors. Also, when comparing HER2negative and positive tumors, a strong reverse
correlation between proliferation index Ki-67
and HER2 expression was revealed (IzyckaSwieszewska et al. 2010b).
The prognostic significance of HER2- positivity in NB was reported as poor (Layfield et al.
1995), none (Gambini et al. 2003; Ho et al. 2005),
or as a good marker (Izycka-Swieszewska et al.
2010b). Our studies showed the differences in
overall survival between patients with HER2negative and positive tumors, proving that immunonegativity was an unfavorable predictor of a
long-term survival. Moreover, Cox proportional
hazard model revealed HER2 expression as an
independent prognostic factor. It can be connected
with the modulating and integrating role of this
protein in HER family signaling, as an important
positive regulator and preferred secondary receptor (Moasser 2007).
HER2 gene is localized in 17q12, whilst 17q
gain is seen in 80% of NB cases. Trisomy of
chromosome 17 is typically seen in near-triploid
tumors with favorable prognosis, while selective
gain of 17q is primarily found in an advanced disease (Maris et al. 2007; Vermuelen et al. 2010). It
is possible that 17q gains characterize a larger
population and more heterogenous in terms of
risk. By FISH no amplification of HER2 was
found, but HER2 polysomy concerned 44% of
cases, that did not correlate with HER2 expression or tumor ploidy. HER2 polysomic NBs
were more frequently observed in younger children, in Schwannian stroma- poor tumors, and
characterized tumours with better prognosis
(Izycka-Swieszewska et al. 2011).
HER3
HER3 is an effective signal transducer that has
the multiple binding sites for PI3K p85 regulatory subunit. HER3 plays an important role in
embryogenesis, as postnatally down-regulator
in most derivatives of neural crest cells with
the exception of Schwann cells that require it
for normal function (Baselga and Swain 2009).
95
HER3 is a very rare subject of investigation in

NBs. However, Ho et al. (2005) were the first
to report HER3 expression in NB at the mRNA
level. They observed HER3 level similar to that
of EGFR. Later, variable HER3 expression levels in cell lines and 20 tumor samples with low
cell surface presentation have been described
(Richards et al. 2010). In our study, high,
mainly cytoplasmic HER3 expression in neuroblastic cells in approximately 40% of cases
and low to high Schwannian stroma positivity
was reported (Fig. 9.3). HER3 labeling occurs
in neuroblastic cells in various stages of differentiation, but with increasing tendency parallel to maturation (Izycka-Swieszewska et al.
2011). Stromal positivity may suggest the
HER3 involvement in NB stroma formation in
auto- and paracrine manner. HER3 expression
was significantly correlated with HER2 and
HER4 expression level. Neither significant
clinicopathological association of HER3
expression nor prognostic significance has
been described.
HER4
HER4 is widely expressed in many adult and
fetal tissues. In PC12 neuronal cells, HER4 activation causes neurite outgrowth and protects cells
from low serum- induced apoptosis. Activated
HER4 undergoes proteolytic cleavage at the cell
surface to release a multifunctional intracellular
domain 4ICD. This domain may be located in the
cytosol to mitochondria (induces apoptosis) and
nucleus where it functions as transcriptional
co-activator (Citri and Yarden 2006).
In NBs uptill now, this receptor has been
examined in three reports. Ho et al. (2005)
revealed low HER4 mRNA expression in NB cell
lines and tumor samples. Richards et al. (2010)
found HER4 cell surface expression, by the flow
cytometry, in most cases of NB cell lines. They
reported higher expression level of cleavable
HER4 isoform mRNA than non-cleavable ones.
In addition, they detected HER4 in all lines and
tumor samples by Western immunoblotting.
These authors postulated a relevant role of HER4
96
Fig. 9.3 High HER3 expression in neuroblastoma Schwannian stroma-poor (left- HER3, 100) and in both components of ganglioneuroma (right- HER3, 400)
Fig. 9.4 HER4 cytoplasmic staining in neuroblastic cells with early differentiation (left- HER4, 400) and high HER4
expression of ganglion cells in ganglioneuroma (right- HER4, 200)
in NBs. Above reports did not show any correlation with clinicopathological data. In our study,
half of the cases expressed HER4, with
cytoplasmic, membranous and rarely nuclear
localization. Level of expression was diverse in
histological categories and subgroups, showing
increasing level parallel to differentiation
(Fig. 9.4). These findings support general opinion

that HER4 signaling promotes cellular differentiation. However, high HER4 expression was
more often found in tumors with known adverse
risk factors (stage IV, age over 18 months and
almost diploidy) and in high clinical risk group.
Survival analysis showed a tendency for a high
HER4 expression to be an adverse prognostic

factor (Izycka-Swieszewska et al. 2011).
Coexpression Proling
Dimerization greatly broadens the signaling diversity of HER receptors. The consequence of any
HER activation may to a large extent depend on
the context of the other receptors. Various dimeric
pairs depend on the concentration of receptors,
ligands and the receptors affinity towards each
other (Citri and Yarden 2006; Rogers et al. 2009).
Because HERs function by homo and- heterodimerization, we also analyzed coexpression of
receptors. In our series of NBs, high expression of
all four receptors, three receptors, and at least two
family members characterized around 25% of
tumors each. Therefore, the analysis of paired
profiling associations with clinicopathological
data was performed. Interestingly, tumors
HER3-high + HER4-negative/low and HER2positive +HER4-negative/low were rarely found
in tumors in metastatic stage while all patients
with HER3-high + HER4-negative/low tumors
were NBs survivors (non-censored observations).
Survival analysis showed the difference between
patients with some HER expression patterns: both
HER2/HER3 negative, showed decreased survival; HER2-positive + HER4- negative/low and
EGFR-high + HER2-positive were associated with
a better outcome, which was also found when
analyzed metastatic tumors only. Profile HER2negative HER3-negative was more frequent in
NMYC-amplified and in poorly differentiated NBs
(Izycka-Swieszewska et al. 2011). These findings
suggest the significance of coexpression of HER
receptors in their co-operative signaling, which
needs further exploration.
HERs may represent a new therapeutic target in
NB, encompassing selective or pan-ERBB inhibitors, complementary to conventional methods.
However, their feasibility needs further investigation, due to the interrelated and complex role
of HER family members in NB biology. Better
understanding of the function of these receptors
may enable the establishment of new prognostic
and predictive biomarkers in NBs.
97
References
Baselga J, Swain S (2009) Novel anticancer targets: revisiting ERBB2 and discovering ERBB3. Nat Rev Cancer
9:463475
Casalini P, Iorio M, Galmozzi E, Menard S (2004) Role of
HER receptors family in development and differentiation. J Cell Physiol 200:343350
Citri A, Yarden Y (2006) EGF-ERBB signalling: towards
the systems level. Nat Rev Mol Cell Biol 7:505516
Cohn S, Pearson A, London W, Monclair T, Ambros P,
Brodeur G, Faldum A, Hero B, Iehara T, Machin D,
Mosseri V, Simon T, Garaventa A, Castel V, Matthay
K (2009) The International Neuroblastoma Risk Group
(INRG) classification system: an INRG task force
report. J Clin Oncol 27:289297
Evangelopoulos M, Weis J, Kruttgen A (2009) Mevastatininduced neurite outgrowth of neuroblastoma cells via
activation of EGFR. J Neurosci Res 87:21382144
Fong A, Park J (2009) High-risk neuroblastoma: a therapy
in evolution. Pediatr Hematol Oncol 26:539548
Gambini C, Sementa AR, Boni L, Marino CE, Corce M,
Negri F, Pistoia V, Ferrini S, Corrias MV (2003)
Expression of HER2/neu is uncommon in human neuroblastic tumors and is unrelated to tumor progression.
Cancer Immunol Immunother 52:116120
Gilbertson RJ (2005) ERBB2 in pediatric cancer: innocent
until proven guilty. Oncologist 10:508517
Goji J, Nakamura H, Ito H, Mabuchi O, Hashimoto K,
Sano K (1995) Expression of c-ErbB2 in human neuroblastoma tissues, adrenal medulla adjacent to tumor,
and developing mouse neural crest cells. Am J Pathol
146:660672
Gullick WJ (2003) c-erbB-4/HER4: friend or foe? J Pathol
200:279281
Hirsch F, Varella-Garcia M, Cappuzzo F (2009)
Predictive value of EGFR and HER2 overexpression
in advanced non-small-cell lung cancer. Oncogene
28:S32S37
Ho R, Minturn J, Hishiki T, Zhao H, Wang Q, Cnaan A,
Maris J, Evans A, Brodeur G (2005) Proliferation of
human neuroblastomas mediated by the epidermal
growth factor receptor. Cancer Res 65:98689875
Izycka-Swieszewska E, Brzeskwiniewicz M, Wozniak A,
Drozynska E, Grajkowska W, Perek D, Balcerska A,
Klepacka T, Limon J (2010a) EGFR, PI3KCA and
PTEN genes status and their protein product expression in neuroblastic tumors. Folia Neuropathol
48:238245
Izycka-Swieszewska E, Wozniak A, Kot J, Grajkowska
W, Balcerska A, Perek D, Dembowska-Baginska B,
Klepacka T, Drozynska E (2010b) Prognostic significance of HER2 expression in neuroblastic tumors.
Mod Pathol 23:12611268
Izycka-Swieszewska E, Wozniak A, Drozynska E, Kot J,
Grajkowska W, Klepacka T, Perek D, Koltan S, Bien
E, Limon J (2011) Expression and prognostic significance of HER family receptors in neuroblastic tumors.
Clin Exp Metastasis 28:271282
98
Jakacki R, Hamilton M, Gilbertson R, Blaney S, Terask J,
Krailo M, Ingle A, Voss S, Dancey J, Adamson P
(2008) Pediatric phase I and pharmacokinetic study of
Erlotinib followed by the combination of Erlotinib and
Temozolomid: a childrens oncology group phase I
consortium study. J Clin Oncol 26:49214927
Junttila T, Sundvall M, Maatta J, Elenius K (2000) ErbB4
and its isoforms. Selective regulation of growth factor
responses by naturally occurring receptor variants.
Trends Cardiovasc Med 10:304310
Layfield L, Thompson K, Dodge R, Kerns B (1995)
Prognostic indicators for neuroblastoma: stage, grade,
DNA ploidy, MIB-1-proliferation index, p53, HER-2/
neu and EGFR- a survival study. J Surg Oncol 59:2127
Lynch T, Bell D, Sordella R, Gurubhagavatula S, Okimoto
R, Brannigan B, Harris P, Haserlat S, Supko J, Haluska
F, Louis D, Christiani D, Settleman J, Haber D (2004)
Activating mutations in the epidermal growth factor
receptor underlying responsiveness of non-small-cell
lung cancer to gefitinib. N Engl J Med 350:21292139
Maris J, Hogarty M, Bagatell R, Cohn S (2007)
Neuroblastoma. Lancet 369:21062120
Michaelis M, Bliss J, Arnold S, Hinsch N, Rothweiler F,
Deubzer H, Witt O, Langer K, Doerr H, Wels W,
Cinatl J (2008) Cisplatin-resistant neuroblastoma cells
express enhanced levels of EGFR and are sensitive to
treatment with EGFR-specific toxins. Clin Cancer Res
14:65316537
Moasser MM (2007) The oncogene HER2: its signaling
and transforming functions and its role in human cancer pathogenesis. Oncogene 26:64696487
Morris J, Lin W, Hauser C, Matchuk Y, Getman D, Lee K
(1999) Rescue of the cardiac defect in ErbB2 mutant
mice reveals essential roles of ErbB2 in peripheral
nervous system development. Neuron 23:273283

Moscatello D, Holgado-Madruga M, Godwin A, Ramirez
G, Gunn G, Zoltick P, Biegel J, Hayes R, Wong A
(1995) Frequent expression of a mutant epidermal
growth factor receptor in multiple human tumors.
Nakagawara A (2004) Neural crest development and neuroblastoma: the genetic and biological link. Prog Brain
Res 146:233242
Richards KN, Zweidler-McKay PA, Van Roy N, Speleman F,
Trevino J, Zage P, Hughes D (2010) Signaling of ERBB
receptor tyrosine kinases promotes neuroblastoma growth
in vitro and in vivo. Cancer 116:32333243
Rogers S, Box C, Chambers P, Barbachano Y, Nutting C,
Rhys-Evans P, Workman P, Harrington K, Eccles S
(2009) Determinants of response to epidermal growth
factor receptor tyrosine kinase inhibition in squamous
cell carcinoma of the head and neck. J Pathol 218:
122130
Rossler J, Odenthal E, Geoerger B, Gerstenmeyer A,
Lagodny J, Niemeyer C, Vassal G (2009) EGFR inhibition using Gefitinib is not active in neuroblastoma
cell lines. Anticancer Res 29:13271334
Shimada H, Ambros I, Dehner L, Hata J, Joshi V, Roald B,
Stram D, Gerbing R, Lukens J, Matthay K, Castleberry R
(1999) The international neuroblastoma pathology classification (the Shimada system). Cancer 86:364372
Tamura S, Hosoi H, Kuwahara Y, Kikuchi K, Otabe O,
Izumi M, Tsuchiya K, Iehara T, Gotoh T, Sugimoto T
(2007) Induction of apoptosis by an inhibitor of EGFR
in neuroblastoma cells. Biochem Biophys Res
Commun 358:226232
Vermuelen J, De Preter K, Mestdagh P, Laureys G,
Speleman F, Vandensompele J (2010) Predicting outcomes for children with neuroblastoma. Discov Med
10:2936
Children with Neurobromatosis

Type 1: Positron Emission
Tomography
10
Kevin London, Mahendra Moharir, Kathryn North,

and Robert Howman-Giles
Contents
Abstract
Introduction ............................................................
99
Cerebral Glucose Utilisation

in Children with NF1 .............................................
100
Optic Pathway Gliomas .........................................
100
Gliomas at Other Locations ..................................
103
T2 Hyperintense Lesions on MRI.........................
103
Conclusion ..............................................................
103
References ...............................................................
104
K. London (*)
Discipline of Paediatrics and Child Health,
Faculty of Medicine, University of Sydney and
Department of Nuclear Medicine, The Childrens
Hospital at Westmead, Sydney, NSW 2145, Australia
e-mail: kevin.london@health.nsw.gov.au
M. Moharir
Faculty of Medicine, University of Toronto and Division
of Neurology, The Hospital for Sick Children, Toronto,
Ontario M5G 1X8, Canada
K. North
Discipline of Paediatrics and Child Health,
Faculty of Medicine, University of Sydney and Head,
Institute for Neuroscience and Muscle Research,
The Childrens Hospital at Westmead, Sydney,
NSW 2145, Australia
R. Howman-Giles
Discipline of Imaging, Faculty of Medicine,
University of Sydney and Head, Department of Nuclear
Medicine, The Childrens Hospital at Westmead, Sydney,
NSW 2145, Australia
Neurofibromatosis Type 1 (NF1) is a common

genetic disorder with a high prevalence of
CNS abnormalities including tumors, mainly
gliomas involving the optic pathway. Current
methods of surveillance using clinical and
MRI imaging protocols are not effective in
selecting the patients who have tumors that are
likely to cause progressive symptoms or
undergo high grade transformation, both of
which would benefit from early therapy. This
chapter highlights the clinical utility of cerebral 18F-flurodeoxyglucose positron emission
tomography in children with NF1 in determining optic pathway tumors that require treatment, identifying high grade transformation
of gliomas and in differentiating MRI T2
weighted hyperintense lesions from true
gliomas in this patient population with often
challenging neuroimaging findings.
Introduction
Neurofibromatosis Type 1 is a common autosomal dominant neurocutaneous syndrome with
a prevalence rate of 1:3,500. The diagnosis is
usually based on well-established clinical criteria (Stumpf et al. 1988). The clinical phenotype
of NF1 can involve multiple organ systems and
includes multiple caf-au-lait spots, axillary and
inguinal freckling, hamartomatous lesions in the

99
K. London et al.
100
iris called Lish nodules, optic pathway gliomas

(OPGs), benign and malignant peripheral nerve
sheath tumors (MPNST), and skeletal dysplasia.
Neurocognitive deficits occur in up to 5081%
of patients with NF1 (depending on the definition) (Hyman et al. 2005). The underlying
genetic abnormality is a heterozygous mutation
in the NF1 gene that codes for the protein neurofibromin which is expressed in many tissues and
acts as a tumor suppressor gene. This can occur
as a new dominant mutation (sporadic) or be
inherited as an autosomal dominant trait from an
affected parent.
As well as gliomas of the optic pathway, previous observational studies have reported
increased risk of other tumours of the CNS in
children with NF1 including malignant astrocytomas, ependymomas, meningiomas, medulloblastomas, and MPNST arising from the cranial
nerves (Cohen and Rothner 1989; Korf 2000;
Rosser and Packer 2002). These tumors appear to
have a similar natural history to those which
occur in children without NF1 and the management options are the same (Cohen et al. 1990).
Positron emission tomography (PET) imaging
in the evaluation of CNS tumors involves
the use of radiotracers which map different
biological processes. The glucose analogue
18
F-flurodeoxyglucose (FDG) maps cerebral
glucose metabolism and is by far the most
commonly used PET tracer in oncology. It has
limitations in the detection of CNS tumors due to
high physiologic uptake in normal grey matter.
Other PET radiopharmaceuticals of use in
evaluating CNS tumors include labelled amino
acid and amino acid analogues such as methionine (Galldiks et al. 2010) and fluorophenylalanine
(Tripathi et al. 2009) which map protein synthesis, and the thymidine analogue fluorothymidine
which maps cellular proliferation (Chen et al.
2005; Tripathi et al. 2009). The availability of
these tracers is likely to extend beyond the
research setting if they are shown to provide
additional clinically relevant information above
that of FDG and conventional imaging modalities
in children with CNS tumors. The scientific experience with PET in children with CNS tumors but
without NF (Pirotte et al. 2007; Warren 2009) is
directly applicable to children with NF1.
Areas of hyperintense signal on T2 weighted

cerebral MRI scans, known as T2 hyperintensities (T2H), are seen on average in up to 90% of
patients with NF1 (Gill et al. 2006; Hyman et al.
2003); the reported frequency of these lesions has
increased as MRI imaging has improved, especially with the use of FLAIR and volumetric
sequences. MRI T2 hyperintensities are thought
to represent focal areas of astrocytic gliosis or
dysmyelination (DiPaolo et al. 1995), however
their neuropathological substrate is far from clear.
A significant challenge in children with NF1 is
differentiating T2H from gliomas, especially in
the setting of a lesion which does not show contrast enhancement or a significant mass effect.
There is a small but measurable risk of
inducing a malignancy associated with the use of
imaging modalities which expose the patient to
ionising radiation such as PET (Brenner and Hall
2007; Hall and Brenner 2008). NF1 is a cancer
predisposition syndrome and it is therefore of
heightened importance to exercise the judicious
use of PET in children with this condition.
Cerebral Glucose Utilisation

in Children with NF1
There is emerging evidence that children with NF1
have a different pattern of cerebral metabolism
compared to children without this condition. The
few small studies in this area have suggested
reduced glucose utilisation in the thalamus and less
consistently, in various cortical regions (Balestri
et al. 1994; Kaplan et al. 1997; Buchert et al. 2008).
Well designed comparative studies involving larger
groups of patients are required to add to the current
understanding of cerebral glucose utilisation in
children with NF1. Relevant alterations in cerebral
metabolism can then be determined and allow for
more accurate characterisation of CNS lesions seen
on FDG PET imaging (Fig. 10.1).
Optic Pathway Gliomas

Optic pathway gliomas are the most common
CNS tumor in children with NF1 with a
reported prevalence of up to 19% of patients
10
Children with Neurofibromatosis Type 1: Positron Emission Tomography
101
Fig. 10.1 Cerebral FDG PET scans in children with NF1

(top row) compared to aged matched children without
NF1 (bottom row). The top row demonstrates axial cerebral FDG PET scans at the level of the basal ganglia in
three children with NF1. Immediately below each image
is an aged matched child without NF1 for comparison.
The control patients underwent FDG PET scanning for

staging of non-CNS tumors. The 3 NF1 patients show
reduced tracer uptake in the thalami compared to their
aged matched controls (arrow head). The patients with
NF1 also show variably reduced cortical tracer uptake
(arrows) (London et al., unpublished data, 2010)
(Listernick et al. 1989, 1994). The majority

arise in the optic nerve and can be bilateral.
The majority of OPG present during early
childhood; ~50% become symptomatic (with
visual loss or endocrine dysfunction due to
involvement of the hypothalamus) and the
remainder are asymptomatic with a self-limiting indolent course. Nevertheless more recent
studies have shown that a minority of OPG can
present later in childhood (Thiagalingam et al.
2004; Sylvester et al. 2006; Segal et al. 2010)
and follow an unpredictable clinical course.
Even if diagnosed at an early age clinical deterioration may not occur until much later
(Thiagalingam et al. 2004) requiring continued
surveillance for visual symptoms and precocious puberty throughout childhood.
Although the most appropriate means of
imaging follow up for these patients remains
controversial, most centres caring for children
with NF1 utilise MRI imaging as part of their
surveillance protocol. It should be noted that
MRI has not been shown to be accurate in
predicting OPG progression and selecting

appropriate patients for therapy. Options for
treatment for symptomatic OPG are chemotherapy, and more rarely surgery when there is significant proptosis associated with complete
visual loss in a unilateral intraorbital tumour.
Radiotherapy tends to be avoided due to the risk
of inducing cerebrovascular dysplasia (such as
moyamoya) or secondary malignancy. There is
limited evidence from small case series suggesting that early chemotherapy may reduce the progression to symptomatic OPGs (Blazo et al.
2004) but the challenge is early detection and
ability to predict which tumours will progress.
FDG PET/CT is an imaging modality showing
promise in differentiating aggressive OPGs from
those which will follow an indolent course.
The histopathology of OPGs is usually that of
low grade, pilocytic astrocytoma (WHO grade 1)
(Louis et al. 2007) or rarely fibrillary astrocytomas (Leonard et al. 2006). Pilocytic astrocytomas have been clearly documented to
demonstrate increased FDG accumulation,
102
K. London et al.
Fig. 10.2 The FDG PET appearance of symptomatic

OPG before and after therapy. A 20 month old girl with
NF1 was assessed for right sided proptosis and visual
impairment. The upper row is imaging performed prior to
therapy and shows a bulky contrast enhancing right OPG
extending to the optic chiasm on T1 weighted MRI
(arrow head, left image) with increased FDG accumulation on PET scan (arrow head, right image). The co-registered and fused FDG PET/MRI image accurately
localised the abnormal tracer uptake to the right OPG
(arrow head, centre image). The lower row is imaging

performed after chemotherapy using vincristine and carboplatin. The T1 weighted MRI scan shows mild reduction in size of the right OPG and retained contrast
enhancement (arrow, left image). The FDG PET scan
(right image) and co-registered FDG PET/MRI image
(centre image) shows near complete resolution of the
FDG PET findings with now only minimal tracer activity
in the right OPG post therapy. Normal physiological
FDG uptake is seen in the orbital muscles
similar to that more typically seen in higher grade

lesions (Fulham et al. 1993).
High grade malignant transformation has not
been reported in OPG in NF1 patients. A recent
pilot study by our group suggests that FDG PET/
CT may be useful in predicting which OPGs will
become symptomatic (Moharir et al. 2010). If
this is substantiated in larger prospective studies,
then FDG PET/CT could be used to identify

children for therapy prior to the significant loss of
vision, as well as serving as an imaging biomarker to monitor the effectiveness of treatment.
The utility of FDG PET/CT would be particularly
relevant in younger children with NF1 where the
reliable assessment of visual acuity is difficult
(Fig. 10.2).
10
103
Fig. 10.3 FDG PET in the detection of gliomas in children with NF1. 15 year old boy with NF1 and T2H
lesions in the posterior corpus callosum and brainstem.
Coronal flair MRI (left image) shows the T2H lesion in
the posterior corpus callosum extending into the centrum semiovale bilaterally (solid arrow) and a more heterogeneous T2H lesion in the brainstem extending into
the left cerebellar white matter (open arrows). The FDG
PET scan (right image) and co-registered FDG PET/
MRI image (centre image) shows a focal area of

increased tracer uptake localising to the right side of the
corpus callosum lesion (arrowhead). The remainder of
the corpus callosum lesion and the T2H lesion in the
brainstem do not demonstrate significant tracer accumulation. A stereotactic MRI guided biopsy of the FDG
avid portion of the corpus callosum lesion revealed a
low grade glioma (WHO grade 2 fibrillary
astrocytoma)
Gliomas at Other Locations
white matter. Lesions in different locations tend

to follow a different course and therefore may
have a different pathological basis. T2H in the
basal ganglia, cerebellum and brainstem tend to
resolve or become less intense over time (Gill
et al. 2006). Lesions in the cerebral hemispheres
do not tend to resolve and may actually increase
in number during childhood (Hyman et al. 2003).
Studies have suggested that there may be an association between discrete T2H in the thalami and
cognitive impairment in children with NF1
(Hyman et al. 2007).
T2H have been shown to have reduced FDG
uptake compared to nearby normal brain areas
(Kaplan et al. 1997). This is an important discriminating characteristic enabling these to be
differentiated from non-optic pathway gliomas
especially in the absence of contrast enhancement or mass effect (Fig. 10.3).
The brainstem is the next most common site for

CNS tumors after the optic pathway and tumors in
this region are also usually low grade gliomas
(Guillamo et al. 2003). A previous series in children with NF1 has shown that tumours outside the
optic pathway, rapidly growing or arising in older
children may be more likely to be higher grade
lesions (Leonard et al. 2006). FDG PET is accurate in predicting the tumor grade of CNS lesions
(Borgwardt et al. 2005; Chen 2007) and in children with NF1 is useful in assessing tumors outside of the optic pathway which may be more
susceptible for anaplastic transformation into high
grade malignancies. One confounding factor is
that there are anecdotal reports that brainstem
lesions and optic pathway tumours can spontaneously regress without treatment in individuals with
NF1 (Kim et al. 1998); however the FDG PET
characteristics of these tumours is not known.
Conclusion
T2 Hyperintense Lesions on MRI
Hyperintense lesions (T2H) seen on T2 weighted
MRI sequences occur in the cortical and subcortical grey matter, basal ganglia and in the deep
Cerebral FDG PET is an imaging technique that

holds promise in the assessment of CNS tumors
in children with NF1, and is an important adjunct
to MR imaging. Potential roles exist in identifying children with OPG that would benefit from
104
early instigation of therapy, assessing gliomas for

transformation into higher grade lesions and differentiating T2H from true gliomas particularly
in areas inaccessible to biopsy. There is a requirement for larger prospective studies to be undertaken addressing the expected benefit at a patient
level before clear PET imaging guidelines can be
recommended in this patient group.
References
Balestri P, Lucignani G, Fois A, Magliani L, Calistri L,
Grana C, Di Bartolo RM, Perani D, Fazio F (1994)
Cerebral glucose metabolism in neurofibromatosis
type 1 assessed with [18F]-2-fluoro-2-deoxy-Dglucose and PET. J Neurol Neurosurg Psychiatry
57:14791483
Blazo MA, Lewis RA, Chintagumpala MM, Frazier M,
McCluggage C, Plon SE (2004) Outcomes of systematic screening for optic pathway tumors in children
with neurofibromatosis type 1. Am J Med Genet A
127A:224229
Borgwardt L, Hojgaard L, Carstensen H, Laursen H,
Nowak M, Thomsen C, Schmiegelow K (2005)
Increased fluorine-18 2-fluoro-2-deoxy-D-glucose
(FDG) uptake in childhood CNS tumors is correlated
with malignancy grade: a study with FDG positron
emission tomography/magnetic resonance imaging
coregistration and image fusion. J Clin Oncol
23:30303037
Brenner DJ, Hall EJ (2007) Computed tomography an
increasing source of radiation exposure. N Engl J Med
357:22772284
Buchert R, von Borczyskowski D, Wilke F, Gronowsky M,
Friedrich RE, Brenner W, Mester J, Clausen M,
Mautner VF, Buchert R, von Borczyskowski D, Wilke
F, Gronowsky M, Friedrich RE, Brenner W, Mester J,
Clausen M, Mautner VF (2008) Reduced thalamic
18F-flurodeoxyglucose retention in adults with neurofibromatosis type 1. Nucl Med Commun 29:1726
Chen W (2007) Clinical applications of PET in brain
tumors. J Nucl Med 48:14681481
Chen W, Cloughesy T, Kamdar N, Satyamurthy N,
Bergsneider M, Liau L, Mischel P, Czernin J, Phelps
ME, Silverman DH (2005) Imaging proliferation in
brain tumors with 18F-FLT PET: comparison with
18F-FDG. J Nucl Med 46:945952
Cohen BH, Rothner AD (1989) Incidence, types, and
management of cancer in patients with neurofibromatosis. Oncology 3:2330, discussion 34
Cohen BH, Kaplan AM, Packer RJ (1990) Management
of intracranial neoplasms in children with neurofibromatosis type 1 and 2. The Childrens Cancer Study
Group. Pediatr Neurosurg 16:6672
DiPaolo DP, Zimmerman RA, Rorke LB, Zackai EH,
Bilaniuk LT, Yachnis AT (1995) Neurofibromatosis
K. London et al.
type 1: pathologic substrate of high-signal-intensity
foci in the brain. Radiology 195:721724
Fulham MJ, Melisi JW, Nishimiya J, Dwyer AJ, Di Chiro
G (1993) Neuroimaging of juvenile pilocytic astrocytomas: an enigma. Radiology 189:221225
Galldiks N, Kracht LW, Berthold F, Miletic H, Klein JC,
Herholz K, Jacobs AH, Heiss WD (2010) [11C]-Lmethionine positron emission tomography in the management of children and young adults with brain
tumors. J Neurooncol 96:231239
Gill DS, Hyman SL, Steinberg A, North KN (2006) Agerelated findings on MRI in neurofibromatosis type 1.
Pediatr Radiol 36:10481056
Guillamo JS, Creange A, Kalifa C, Grill J, Rodriguez D,
Doz F, Barbarot S, Zerah M, Sanson M, Bastuji-Garin
S, Wolkenstein P (2003) Prognostic factors of CNS
tumours in Neurofibromatosis 1 (NF1): a retrospective
study of 104 patients. Brain 126:152160
Hall EJ, Brenner DJ (2008) Cancer risks from diagnostic
radiology. Br J Radiol 81:362378
Hyman SL, Gill DS, Shores EA, Steinberg A, Joy P,
Gibikote SV, North KN (2003) Natural history of cognitive deficits and their relationship to MRI
T2-hyperintensities
in
NF1.
Neurology
60:11391145
Hyman SL, Shores A, North KN (2005) The nature and
frequency of cognitive deficits in children with neurofibromatosis type 1. Neurology 65:10371044
Hyman SL, Gill DS, Shores EA, Steinberg A, North KN
(2007) T2 hyperintensities in children with neurofibromatosis type 1 and their relationship to cognitive
functioning. J Neurol Neurosurg Psychiatry
78:10881091
Kaplan AM, Chen K, Lawson MA, Wodrich DL, Bonstelle
CT, Reiman EM (1997) Positron emission tomography in children with neurofibromatosis-1. J Child
Neurol 12:499506
Kim G, Mehta M, Kucharczyk W, Blaser S (1998)
Spontaneous regression of a tectal mass in neurofibromatosis 1. AJNR Am J Neuroradiol 19:11371139
Korf BR (2000) Malignancy in neurofibromatosis type 1.
Oncologist 5:477485
Leonard JR, Perry A, Rubin JB, King AA, Chicoine MR,
Gutmann DH (2006) The role of surgical biopsy in the
diagnosis of glioma in individuals with neurofibromatosis-1. Neurology 67:15091512
Listernick R, Charrow J, Greenwald MJ, Esterly NB
(1989) Optic gliomas in children with neurofibromatosis type 1. J Pediatr 114:788792
Listernick R, Charrow J, Greenwald M, Mets M (1994)
Natural history of optic pathway tumors in children
with neurofibromatosis type 1: a longitudinal study. J
Pediatr 125:6366
Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger
PC, Jouvet A, Scheithauer BW, Kleihues P (2007) The
2007 WHO classification of tumours of the central
nervous system. Acta Neuropathol 114:97109
Moharir M, London K, Howman-Giles R, North K
(2010) Utility of positron emission tomography for
tumour surveillance in children with neurofibroma-
10
tosis type 1. Eur J Nucl Med Mol Imaging

37:13091317
Pirotte B, Acerbi F, Lubansu A, Goldman S, Brotchi J,
Levivier M, Pirotte B, Acerbi F, Lubansu A, Goldman
S, Brotchi J, Levivier M (2007) PET imaging in the
surgical management of pediatric brain tumors. Childs
Nerv Syst 23:739751
Rosser T, Packer RJ (2002) Intracranial neoplasms in children with neurofibromatosis 1. J Child Neurol 17:630
637, discussion 646651
Segal L, Darvish-Zargar M, Dilenge ME, Ortenberg J,
Polomeno RC (2010) Optic pathway gliomas in
patients with neurofibromatosis type 1: follow-up of
44 patients. J AAPOS 14:155158
Stumpf DA, Alksne JF, Annegers JF, Brown SS, Conneally
PM, Housman D, Leppert MF, Miller JP, Moss ML,
Pileggi AJ, Rapin I, Strohman RC, Swanson LW,
105
Zimmerman A (1988) Neurofibromatosis. Conference

statement. National Institutes of Health Consensus
Development Conference. Arch Neurol 45:575578
Sylvester CL, Drohan LA, Sergott RC (2006) Optic-nerve
gliomas, chiasmal gliomas and neurofibromatosis type
1. Curr Opin Ophthalmol 17:711
Thiagalingam S, Flaherty M, Billson F, North K (2004)
Neurofibromatosis type 1 and optic pathway gliomas:
follow-up of 54 patients. Ophthalmology 111:568577
Tripathi M, Sharma R, DSouza M, Jaimini A, Panwar P,
Varshney R, Datta A, Kumar N, Garg G, Singh D,
Grover RK, Mishra AK, Mondal A (2009) Comparative
evaluation of F-18 FDOPA, F-18 FDG, and F-18 FLTPET/CT for metabolic imaging of low grade gliomas.
Clin Nucl Med 34:878883
Warren KE (2009) Noninvasive assessment of pediatric
brain tumors. Cancer Biol Ther 8:18811888
Metabolite Prole Differences

in Childhood Brain Tumors:
1H Magic Angle Spinning NMR
Spectroscopy
11
Martin Wilson and Andrew Peet
Contents
Abstract
Introduction ............................................................
108
Methods...................................................................
Tissue Samples .........................................................
Cell Line Samples ....................................................
HR-MAS ..................................................................
Fitting and Multivariate Analysis.............................
109
109
110
110
110
Results .....................................................................
111
Discussion................................................................
111
References ...............................................................
114
M. Wilson (*) A. Peet

Cancer Sciences, University of Birmingham,
Birmingham, UK
e-mail: martin@pipegrep.co.uk
Background: Brain and nervous system tumours

are the most common solid cancers in children.
Molecular characterisation of these tumours is
important for providing novel biomarkers of
disease and identifying molecular pathways
which may provide putative targets for new
therapies. 1H High Resolution Magic Angle
Spinning NMR spectroscopy (HR-MAS) is an
emerging technique for determining metabolite
profiles from small pieces of intact tissue and
whole cells grown in culture, making it ideal
for molecular characterisation of disease.
Method: 22 tissue samples from children
with brain tumours and 7 cell lines originating
from medulloblastoma (2), neuroblastoma (2),
supertentorial PNET (1) and retinoblastoma
(2) were analysed using HR-MAS. Spectra
were fitted to a library of individual metabolite spectra to provide metabolite values,
Principle Component Analysis (PCA) was
used to investigate the metabolic relationship
between the tumour types.
Results: Primitive neuroectodermal and
glial based tumours were linearly separable
using PCA, demonstrating that histological
features of the tissue were closely related to
their metabolite profiles. Each cultured cell
line was found to have a distinct metabolite
profile. The desmoplastic and classic variants
of medulloblastoma were particularly distinct,
with large differences in the distribution of
choline containing metabolites.

107
108
Conclusion: HR-MAS identified key

differences in the metabolite profiles of childhood brain tumour tissue and cells grown in
culture, improving the molecular characterisation of these tumours and showing the promise
of HR-MAS as a rapid diagnostic aid. Further
investigation of the underlying molecular
pathways is required to assess their potential
as targets for new agents.
Introduction
Childhood brain and nervous system tumours are
the most common solid cancers of childhood.
They comprise a diverse set of diseases from the
highly malignant to low grade indolent tumours
with a corresponding variety of treatments, prognoses and outcomes. Improvements in outcome
have not matched those in other forms of childhood cancer and new methods are required to
understand the biology of these tumours and
develop novel approaches to therapy.
Currently the treatment of these tumours is
largely determined through categorization of the
cases by histopathology, location, stage and
patient age. The most common high grade
tumours can be categorised as primitive neuroectodermal tumours (PNETs) based on their histopathological appearance (Pizzo and Poplack
2010). PNETs are embryonal tumours and have
subgroups which occur in various locations of
the brain, the sympathetic nervous system and
the eye. Neuroblastoma arises from the sympathetic nervous system and often presents with
metastases at diagnosis and is particularly challenging to treat. Intracranial PNETs are all
WHO grade IV tumours which have metastatic potential and follow an aggressive clinical course. Medulloblastomas occur in the
cerebellum, pineoblastomas in the pineal gland
and supratentorial PNETs in other supratentorial
regions. They are all poorly differentiated
tumours with closely related histopathology.
Despite their many similarities, treatment is tailored to the specific sub-type of tumour and
improved characterization is an important
objective.
M. Wilson and A. Peet
Other childhood brain tumours are diverse in

terms of histopathology, grade and clinical behaviour. In addition to PNETs, brain tumours can
belong to another common histopathological
category known as glial tumours. These tumours
are thought to arise from the supportive tissue of
the brain, glia. Astrocytomas, many of which are
WHO grade I, are the most common example of
these tumours in the brain during childhood.
Ependymomas are locally aggressive tumours
which are predominantly WHO grade II and III.
Although histopathology is an important
method of characterizing and diagnosing tumours,
it is not always possible to accurately distinguish
between different tumour types using this method
and the development of new techniques may
improve characterisation and diagnosis in difficult cases. Furthermore, histopathology is often
a poor predictor of tumour behaviour and response
to treatment. The improved characterization of
these tumour types through the discovery of novel
biomarkers is an important step in optimizing
treatment for individual patients.
Tumour genetics is emerging as an important
adjunct to histopathological diagnosis and clinical indicators in determining prognosis and stratifying treatment. Amplification of the MYCN
oncogene is already used clinically as a prognostic marker to stratify treatment in neuroblastoma
and cMyc has been linked to a more aggressive
phenotype in medulloblastoma (Badiali et al.
1991; Maris et al. 2007). Furthermore, gene
expression profiling has been highly successful
in subcategorizing the different subtypes of
PNETs and has led to the discovery of prognostic
markers (Pomeroy et al. 2002). Through this process, specific molecular pathways are being identified for specific tumours leading to the discovery
of potential targets for new therapeutic agents.
With the identification of specific patterns of
gene expression, there is increasing interest in
probing the downstream molecular pathways
related to these changes. One strategy which has
emerged as being of particular interest is the
broad sampling of metabolite levels as a measure
of tumour metabolism, a strategy commonly
termed metabolomics. 1H nuclear magnetic resonance (NMR) spectroscopy can measure the
11
Metabolite Profile Differences in Childhood Brain Tumors: 1H Magic Angle Spinning NMR
concentration of a range of metabolites and is a

particularly powerful tool for measuring metabolite profiles. Several studies have used NMR to
measure metabolite profiles from chemical
extracts of excised brain tumour tissue and found
that specific metabolites differ between brain
tumour and healthy brain (Peeling and Sutherland
1992), low grade and high grade astrocytic
tumours (Kinoshita et al. 1993; Usenius et al.
1994), and glioblastoma and metastatic tumours.
It has been shown more recently that total choline
correlates with tumour progression (Lehnhardt
et al. 2005). The majority of extract studies on
brain tumour tissue have focused on adult astrocytomas, however a study of pediatric posterior
fossa tumours (Sutton et al. 1994) showed that
medulloblastomas could be distinguished from
astrocytomas by their metabolite profile.
A variant of the NMR technique known as 1H
high resolution magic angle spinning NMR
(HR-MAS) allows metabolite profiling of intact
tissue. The technique provides high resolution
data on small (530 mg) inhomogeneous samples
making it ideal for the study of tissue (Griffin and
Shockcor 2004). The technique has had success
in characterising a range of tissues including diseased brain (Cheng et al. 1997), breast tumour
(Sitter et al. 2002, 2006) cervical (Lyng et al. 2007),
liver tumour (Martnez-Granados et al. 2006; Yang
et al. 2007), primary (Tzika et al. 2007; MartnezBisbal et al. 2004; Barton et al. 1999; Wright
et al. 2010) and metastatic (Sjbakk et al. 2008)
adult brain tumours and paediatric brain tumours
(Tzika et al. 2002; Tugnoli et al. 2005). HR-MAS
results also show a good correlation with in vivo
metabolite profiles measured by magnetic resonance spectroscopy in patients (Tzika et al. 2002;
Wilson et al. 2009).
Whilst analysis of tumour tissue is desirable,
it can often be difficult to obtain a large number
of samples for relatively rare diseases such as
childhood brain tumours. The study of cell lines
presents a convenient alternative to tissue, and
has the advantage of producing homogeneous
and reproducible samples which can be genetically manipulated or exposed to drugs. HR-MAS
has been shown to be a useful tool for investigating the metabolic response of transfection (Griffin
109
et al. 2001; Peet et al. 2007), cell type (Griffin

et al. 2002) and therapeutic agents (Borel et al.
2007; Morvan et al. 2003) on cells grown in
culture, making it a suitable method for in-vitro
studies of cancer cells.
Recently, semi-automated methods have been
developed for accurately quantitating metabolite
concentrations from HR-MAS spectra (Rabeson
et al. 2008; Reynolds et al. 2006). Multivariate
techniques such as principle component analysis
(PCA) and linear discriminant analysis (LDA)
may be used to analyse these metabolite profiles
with the goal of improving tumour characterisation and classification (Preul et al. 1996; Davies
et al. 2008). The combination of minimal sample
preparation, speed of collecting HR-MAS data
and automated analysis/classification, gives the
potential for this strategy to provide a rapid diagnostic aid. Current methods used for rapid diagnosis such as frozen section analysis have a low
accuracy and HR-MAS provides a potential
method to improve this. However, more importantly, the analysis of metabolite profiles gives
the opportunity to identify key molecular pathways to improve our understanding of tumour
biology and provide new targets for novel therapeutic agents.
In this study we use the technique of HR-MAS
to investigate the metabolic characteristics of
childhood brain tumour tissue and cells. The initial aim is to ascertain whether the tumour tissue
samples form distinct groups according to their
histopathology, and if so, to determine their characteristic metabolite profiles. HR-MAS is also
performed on a panel of primitive neuroectodermal cell lines to investigate the reproducibility of
the technique and its ability to uniquely characterise cell lines from tumours with similar histological appearances.
Methods
Tissue Samples
Biopsy tissue was snap frozen in liquid nitrogen
shortly after resection and stored at 80C. Just
prior to HR-MAS, tissue was thawed at room
110
temperature and cut to approx 15 mg where

appropriate. The mean tissue mass analysed was
10 mg. The tissue was placed into a 40 ml widemouthed zirconium sample tube (Varian NMR
Inc, Palo Alto, CA, USA) and weighed. Four
microlitre of 3-(trimethylsilyl)proponic-2,2,3,3d4 acid sodium salt (TSP) was dissolved in D2O
at a concentration of 10 mM and was added to the
rotor. The remaining volume of the rotor was
filled with D2O.
Cell Line Samples

Seven cell lines from four primitive neuroectodermal tumour types were studied, two medulloblastomas (D283 derived from a classic
medulloblastoma tumour, DAOY derived from a
desmoplastic medulloblastoma tumour), two
neuroblastomas (KELLY MYCN amplified,
SHEP1 MYCN non-amplified), two retinoblastomas (WERI-RB-1, Y79) and one ST-PNET
(PFSK) cell line. Cells were grown in DMEM/F12
supplemented with 4 mM L-glutamine, 15% fetal
calf serum and MEM non-essential amino acid
solution (Sigma Aldrich, Gillingham, Dorset;
catalogue number M7145). They were incubated
at 37C in 5% CO2 and harvested 24 h after a
final medium change at ~90% confluence.
Single 75-cm2 flasks were harvested by removing the medium and washing the cells three times
while still adherent to the flask with 3 ml ice-cold
phosphate buffer solution. The cells were then
removed from the flask using a manual scraper
and centrifuged at 250 g for 3 min to form a pellet, which was snap-frozen in liquid nitrogen and
stored at 80C. Just before HR-MAS, the cells
were defrosted, and 36 ml was pipetted into a
wide-mouthed zirconium sample tube, 4 ml of
10 mM TSP dissolved in D2O was added as a
chemical shift standard.
HR-MAS
HR-MAS was performed on a Varian 600 MHz
vertical bore spectrometer using a 4 mm gHX
nanoprobe (Varian NMR Inc, Palo Alto, CA,

USA) with a three channel INOVA console running VNMRj software. The probe temperature
was set to 0.1C to minimize sample degradation,
and the sample was spun at 2,500 Hz. These conditions equated to a sample temperature of 6.7C
determined by methanol calibration. A standard
pulse and acquire sequence was used which consisted of a single 90 pulse preceded by 1 s of
water presaturation. This was followed by the
acquisition of 16 K complex points at a sampling
frequency of 7,200 Hz. 256 or 512 scans were
acquired depending on the sample size with a
repetition time of 3.3 s giving a total acquisition
time of 14 or 28 min. Tuning and matching, 90
pulse width and the pre-saturation pulse frequency
were optimised for each sample.
Fitting and Multivariate Analysis

Raw data was Fourier transformed to 16 K points,
phased and referenced to the creatine peak at
3.03 ppm using in-house software. The phased
data was then transformed back to the timedomain and the TARQUIN algorithm (Reynolds
et al. 2006) was applied to determine relative
metabolite concentrations. This algorithm was
chosen as it has been shown to be robust to the
shifting of metabolite peaks caused by pH variation, which is of particular importance in the
analysis of HR-MAS data. The relative concentrations of the following metabolites were determined: acetate (Ace), alanine (Ala), aspartate
(Asp), choline (Cho), creatine (Cr), glutamate
(Glu), glutamine (Gln), glycerophosphocholine
(GPC), glycine (Gly), lactate (Lac), myo-inositol
(m-Ins), N-acetylaspartate (NAA), phosphocholine (PC), phosphorylethanolamine (PEth),
scyllo-inositol (s-Ins), succinate (Suc) and taurine
(Tau).
Metabolite concentrations were imported into
the R statistics software package (R Development
Core Team 2009) and each metabolite profile
vector was normalised by the sum of metabolite
concentrations. All data was whitened prior to
Principal Component Analysis (PCA).
11
Figure 11.2a shows that the metabolite profile

is consistently reproduced for each PNET cell
line with no overlap between the different cell
lines. The second principal component separates
the medulloblastoma from the neuroblastoma
tumours, whereas the first principal component
does not show any obvious splitting between the
tumour types. However, in the first principal
component, the medulloblastoma cell lines
DAOY and D238 have similar scores to the neuroblastoma cell lines SHEP and KELLY respectively. From the corresponding loadings plot
(Fig. 11.2b) choline-containing metabolites
contribute significantly to the first principal
component. The spectral regions associated with
the choline-containing metabolites are shown in
Fig. 11.3, clearly illustrating the similarities
and differences between the neuroblastoma and
medulloblastoma cell lines.
Results
Spectral resolution for both tissue and whole cell
samples was comparable to liquid state NMR and
most metabolites analysed were clearly separable
from spectral noise. Phosphorylethanolamine
was clearly visible in the medulloblastoma samples but was not detected in any other tumour
type. The residual water signal was small enough
to prevent interference with metabolite analysis
and metabolite profiles were found to be stable
within the acquisition duration.
Figure 11.1a shows that the glial and PNET
brain tumour tissue samples studied are linearly
separable using a combination of the first two
principal components. The two main tumour
types, medulloblastoma and grade I astrocytoma,
appear to show a wide range of metabolic heterogeneity whereas the three ependymoma tumours
form a small group separately from the other
tumour types. From Fig. 11.2b PNET tumours
can be characterised as having greater levels of
phosphocholine, glycine and taurine whereas
glial tumours can be characterised as having
greater levels of glutamine, lactate and
glycerophosphocholine.
Discussion
The discrimination of PNETs from glial tumours
is important for two reasons. Firstly, it is clinically useful to be able to differentiate between
these types for tumours in several locations, in
b
4
4
111
0.4
PC
sIns
Cho
Cr
PEth
mIns
Asp
0.0
PC2
0.2
Tau
PC2
Gly
Glu
Ace
PC1
Fig. 11.1 Principal component scores (a) and loadings

(b) of the metabolite profiles from paediatric brain tumour
tissue. The solid black line plotted with the equation y = x
0.2
2
Lac
GPC
Epp
G1Astro
MB
STPNET
0.4
Gln
Ala
Suc
NAA
0.4
0.2
0.0
0.2
0.4
PC1
separates the glial and PNET tumours. Epp ependymoma,

G1Astro grade I astrocytoma, MB medulloblastoma,
ST-PNET supratentorial primative neroectodermal tumour
112
4
Lac
Gln
Asp
Gly
0.0
PEth
mIns
2
Tau
sIns
NAA
Suc
PC
0.2
D283
DAOY
WERI
Y79
PFSK
KELLY
SHEP1
Ala
Cho Ace
GPC
Glu
0.1
PC2
PC2
1
Cr
0.1
b
0.2
0
PC1
0.2
0.1
0.0
PC1
0.1
0.2
Fig. 11.2 Principal component scores (a) and loadings (b) of the metabolite profiles from a panel of primitive neuroectodermal tumour cell lines
Fig. 11.3 A comparison between the choline regions of

the medulloblastoma DAOY and D283 cell lines. The
MYCN amplified cell line KELLY, and MYCN nonamplified cell line SHEP1 are also included to illustrate
the similarities. Choline, phosphocholine and glycerophosphocholine have chemical shifts of 3.205, 3.222 and
3.231 respectively
11
particular the cerebellum and cerebral hemispheres.

Since HR-MAS can be performed on tissue samples as small as 5 mg with minimal processing,
the technique is potentially useful for the rapid
analysis of inter-operative biopsy samples.
Secondly, the identification of key metabolic
markers combined with the analysis of metabolic
networks (Easton et al. 2010) and other molecular data (Pomeroy et al. 2002) may reveal pathways which are key to the disease process and
can be targeted by novel drugs.
Biochemical changes have been noted previously in childhood brain and nervous system
tumours (Davies et al. 2008; Tzika et al. 2002)
but the current understanding of their significance
is limited. High taurine in medulloblastomas has
been reported previously in-vivo (Moreno-Torres
et al. 2004) and in small studies using HR-MAS
(Tugnoli et al. 2005; Davies et al. 2008). Taurine
is known to play an important role in neurodevelopment (Wharton et al. 2004) and has been shown
to correlate with apoptosis in adult gliomas
(Opstad et al. 2009), however, its precise role in
tumourigenesis remains unclear.
Choline metabolism has been related to
tumour growth in numerous studies reviewed by
Podo (Podo 1999), and high phosphocholine/
glycerophosphocholine ratio is seen in rapidly
growing aggressive tumours (Usenius et al.
1994). The high phosphocholine and phosphocholine/glycerophosphocholine ratio seen in
medulloblastomas is therefore consistent with
their rapid growth and high grade and confirms
data obtained on these tumours in-vivo and exvivo (Davies et al. 2008). The increase of glycine
in medulloblastoma has been confirmed in-vivo
(Davies et al. 2010) and a recent study has
reported that glycine may also be a useful biomarker in adult brain tumours (Righi et al. 2010).
Currently, the link between malignancy and glycine
metabolism is poorly understood making it an
interesting topic for further research.
The cell lines studied showed a high level of
reproducibility, to the extent where no overlap
was present between the individual cell lines on a
principal component scores plot. This demonstrates that HR-MAS metabolic profiling is capable of detecting differences between tumour cells
113
which are very similar in terms of their morphology

and are thought to have a common cell of origin.
Despite each cell line having a unique profile, the
main tumour groups (medulloblastoma, neuroblastoma and retinoblastoma) did not cluster
together in the first two principal components
indicating that their histological classification
was not responsible for the greatest level of metabolic variance present across the samples.
Choline metabolism was shown to be important for distinguishing between cell lines as well
as tissue. In our previous work on neuroblastoma
cell lines (Peet et al. 2007) it was shown that cell
lines with an increased expression of the MYCN
oncogene had an associated increase in the phosphocholine/glycerphosphocholine ratio. The two
neuroblastoma cell lines included in this study
showed this trend with the MYCN amplified cell
line KELLY having a greater phosphocholine/
glycerphosphocholine ratio than the MYCN nonamplified cell line SHEP1, as illustrated in
Fig. 11.3. Like the neuroblastoma cell lines, the
medulloblastoma cell lines DAOY and D283 also
show a large variation in their phosphocholine/
glycerphosphocholine ratios with D283 and
DAOY having a ratios similar to that of a MYCN
amplified and non-amplified cell line respectively. Interestingly, it has been shown that the
level of c-myc has a causal relationship to a
more aggressive anaplastic phenotype in
medulloblastoma cell lines, and noted that D283
has a greater expression of c-myc than DAOY
(Stearns et al. 2006).
These data suggest that c-myc may have an
influence on the metabolite profile of medulloblastoma cell lines similar to that of MYCN in
neuroblastoma. This is especially relevant as a
recent study has shown a link between the
retinoic acid treatment of medulloblastoma
cells and a reduction in c-myc (Chang et al.
2007). Retinoic acid treatment has also been
shown to significantly improve outcome for
patients with high risk neuroblastoma
(Reynolds et al. 2003). An interesting extension to this study would be to confirm whether
the level of c-myc expression is directly related
to the concentrations of choline-containing
metabolites.
114
Cell lines present a useful model for investigating

the complex molecular interactions which are
important to the formation and spread of tumour
cells. Once established in culture, cells can provide a consistent supply of biological replicates
making it an attractive model for diseases where
tissue samples are rare such as paediatric brain
tumours. A further advantage of cell lines is that
they can be modified to knock-down or increase
the production of particular proteins via transfection allowing particular molecular pathways to be
investigated in greater detail. An investigation
into the effect of the MYCN oncogene on the
metabolites detectable using HR-MAS (Peet
et al. 2007) is one example of how HR-MAS and
manipulated cells can be used to reveal the downstream effects of altered gene expression.
However, care must be taken in interpreting
results since tissue micro-environment is also
known to influence tumour biology. Where possible, molecular findings should be confirmed
with tissue samples or an appropriate animal
model.
In conclusion, we have shown that the combination of HR-MAS, automated spectral analysis
and chemometric methods to be a useful approach
for establishing the metabolite profiles of paediatric brain tumour tissue and cell lines. Differences
in a number of metabolites have been found
between glial and PNET tumour tissue, and choline metabolism has been identified as a key differentiator between in these tumour types. We
have also shown that the metabolite profiles of
cultured PNET cells can be measured with a high
level of reproducibility, making the technique
suited to the study of in-vitro model systems
which may evaluate the efficacy of drugs or the
downstream effects of a particular gene. We have
also established that the major variability in the
metabolite profiles between these cell lines is not
the histopathology of their tumour of origin and
is more likely to be the genetic defects which
drive their malignant process. By comparing the
metabolite profiles of tumour tissue and manipulated tumour cells grown in culture together with
information from molecular genetics, gene and
protein expression, key molecular pathways
may be identified and validated, improving our
understanding of these tumours and identifying

targets for new drugs.
References
Badiali M, Pession A, Basso G, Andreini L, Rigobello L,
Galassi E, Giangaspero F (1991) N-myc and c-myc
oncogenes amplification in medulloblastomas.
Evidence of particularly aggressive behavior of a tumor
with c-myc amplification. Tumori 77(2):118121
Barton SJ, Howe FA, Tomlins AM, Cudlip SA, Nicholson
JK, Bell BA, Griffiths JR (1999) Comparison of
in vivo 1H MRS of human brain tumours with 1H
HR-MAS spectroscopy of intact biopsy samples
in vitro. MAGMA 8(2):121128
Borel M, Degoul F, Communal Y, Mounetou E, Bouchon
B, C-Gaudreault R, Madelmont JC, Miot-Noirault E
(2007) N-(4-iodophenyl)-N-(2-chloroethyl)urea as a
microtubule disrupter: in vitro and in vivo profiling of
antitumoral activity on CT-26 murine colon carcinoma
cell line cultured and grafted to mice. Br J Cancer
96(11):16841691
Chang Q, Chen Z, You J, McNutt MA, Zhang T, Han Z,
Zhang X, Gong E, Gu J (2007) All-trans-retinoic acid
induces cell growth arrest in a human medulloblastoma cell line. J Neurooncol 84(3):263267
Cheng LL, Ma MJ, Becerra L, Ptak T, Tracey I, Lackner
A, Gonzlez RG (1997) Quantitative neuropathology
by high resolution magic angle spinning proton magnetic resonance spectroscopy. Proc Natl Acad Sci
USA 94(12):64086413
Davies NP, Wilson M, Harris LM, Natarajan K, Lateef S,
MacPherson L, Sgouros S, Grundy RG, Arvanitis TN,
Peet AC (2008) Identification and characterisation of
childhood cerebellar tumours by in vivo proton MRS.
NMR Biomed 21(8):908918
Davies NP, Wilson M, Natarajan K, Sun Y, MacPherson
L, Brundler M, Arvanitis TN, Grundy RG, Peet AC
(2010) Non-invasive detection of glycine as a biomarker of malignancy in childhood brain tumours
using in-vivo 1H MRS at 1.5 Tesla confirmed by
ex-vivo high-resolution magic-angle spinning NMR.
NMR Biomed 23(1):8087
Easton JM, Harris LM, Viant MR, Peet AC, Arvanitis TN
(2010) Linked metabolites: a tool for the construction
of directed metabolic graphs. Comput Biol Med
40(3):340349
Griffin JL, Shockcor JP (2004) Metabolic profiles of
cancer cells. Nat Rev Cancer 4(7):551561
Griffin JL, Mann CJ, Scott J, Shoulders CC, Nicholson JK
(2001) Choline containing metabolites during cell transfection: an insight into magnetic resonance spectroscopy
detectable changes. FEBS Lett 509(2):263266
Griffin JL, Bollard M, Nicholson JK, Bhakoo K (2002)
Spectral profiles of cultured neuronal and glial cells
derived from HRMAS (1)H NMR spectroscopy. NMR
Biomed 15(6):375384
11
Kinoshita Y, Kajiwara H, Yokota A, Koga Y (1993) Proton

magnetic resonance spectroscopy of astrocytic tumors:
an in vitro study. Neurol Med Chir (Tokyo)
33(6):350359
Lehnhardt F, Bock C, Rhn G, Ernestus R, Hoehn M
(2005) Metabolic differences between primary and
recurrent human brain tumors: a 1H NMR spectroscopic investigation. NMR Biomed 18(6):371382
Lyng H, Sitter B, Bathen TF, Jensen LR, Sundfr K,
Kristensen GB, Gribbestad IS (2007) Metabolic mapping by use of high-resolution magic angle spinning
1H MR spectroscopy for assessment of apoptosis in
cervical carcinomas. BMC Cancer 7:11
Maris JM, Hogarty MD, Bagatell R, Cohn SL (2007)
Neuroblastoma. Lancet 369(9579):21062120
Martnez-Bisbal MC, Mart-Bonmat L, Piquer J, Revert
A, Ferrer P, Llcer JL, Piotto M, Assemat O, Celda B
(2004) 1H and 13C HR-MAS spectroscopy of intact
biopsy samples ex vivo and in vivo 1H MRS study of
human high grade gliomas. NMR Biomed 17(4):
191205
Martnez-Granados B, Monlen D, Martnez-Bisbal MC,
Rodrigo JM, del Olmo J, Lluch P, Ferrndez A, MartBonmat L, Celda B (2006) Metabolite identification
in human liver needle biopsies by high-resolution
magic angle spinning 1H NMR spectroscopy. NMR
Biomed 19(1):90100
Moreno-Torres A, et al. (2004) Taurine detection by
proton magnetic resonance spectroscopy in medulloblastoma: contribution to noninvasive differential diagnosis with cerebellar astrocytoma. Neurosurgery. 55(4):
824829, discussion 829
Morvan D, Demidem A, Papon J, Madelmont JC (2003)
Quantitative HRMAS proton total correlation spectroscopy applied to cultured melanoma cells treated by
chloroethyl nitrosourea: demonstration of phospholipid metabolism alterations. Magn Reson Med 49(2):
241248
Opstad KS, Bell BA, Griffiths JR, Howe FA (2009)
Taurine: a potential marker of apoptosis in gliomas. Br
J Cancer 100(5):789794
Peeling J, Sutherland G (1992) High-resolution 1H NMR
spectroscopy studies of extracts of human cerebral
neoplasms. Magn Reson Med 24(1):123136
Peet AC, McConville C, Wilson M, Levine BA, Reed M,
Dyer SA, Edwards EC, Strachan MC, McMullan DJ,
Wilkes TM, Grundy RG (2007) 1H MRS identifies
specific metabolite profiles associated with MYCNamplified and non-amplified tumour subtypes of neuroblastoma cell lines. NMR Biomed 20(7):692700
Pizzo PA, Poplack DG (2010) Principles and practice of
pediatric oncology, 6th edn. Lippincott Williams and
Wilkins, Philadelphia
Podo F (1999) Tumour phospholipid metabolism. NMR
Biomed 12(7):413439
Pomeroy SL, Tamayo P, Gaasenbeek M, Sturla LM,
Angelo M, McLaughlin ME, Kim JYH, Goumnerova
LC, Black PM, Lau C, Allen JC, Zagzag D, Olson
JM, Curran T, Wetmore C, Biegel JA, Poggio T,
Mukherjee S, Rifkin R, Califano A, Stolovitzky G,
115
Louis DN, Mesirov JP, Lander ES, Golub TR (2002)

Prediction of central nervous system embryonal
tumour outcome based on gene expression. Nature
415(6870):436442
Preul MC, Caramanos Z, Collins DL, Villemure J, Leblanc
R, Olivier A, Pokrupa R, Arnold DL (1996) Accurate,
noninvasive diagnosis of human brain tumors by using
proton magnetic resonance spectroscopy. Nat Med
2(3):323325
Rabeson H, Fauvelle F, Testylier G, Foquin A, Carpentier
P, Dorandeu F, van Ormondt D, Graveron-Demilly D
(2008) Quantitation with QUEST of brain HRMASNMR signals: application to metabolic disorders in
experimental epileptic seizures. Magn Reson Med
59(6):12661273
Reynolds CP, Matthay KK, Villablanca JG, Maurer BJ
(2003) Retinoid therapy of high-risk neuroblastoma.
Cancer Lett 197(12):185192
Reynolds G, Wilson M, Peet A, Arvanitis TN (2006) An
algorithm for the automated quantitation of metabolites
in in vitro NMR signals. Magn Reson Med
56(6):12111219
R Development Core Team (2008). R: A language and
environment for statistical computing. R Foundation
for Statistical Computing, Vienna, Austria. ISBN
3-900051-07-0, http://www.R-project.org
Righi V, Andronesi OC, Mintzopoulos D, Black PM,
Tzika AA (2010) High-resolution magic angle spinning magnetic resonance spectroscopy detects glycine
as a biomarker in brain tumors. Int J Oncol
36(2):301306
Sitter B, Sonnewald U, Spraul M, Fjsne HE, Gribbestad
IS (2002) High-resolution magic angle spinning MRS
of breast cancer tissue. NMR Biomed 15(5):327337
Sitter B, Lundgren S, Bathen TF, Halgunset J, Fjosne HE,
Gribbestad IS (2006) Comparison of HR MAS MR
spectroscopic profiles of breast cancer tissue with
clinical parameters. NMR Biomed 19(1):3040
Sjbakk TE, Johansen R, Bathen TF, Sonnewald U, Juul
R, Torp SH, Lundgren S, Gribbestad IS (2008)
Characterization of brain metastases using highresolution magic angle spinning MRS. NMR Biomed
21(2):175185
Stearns D, Chaudhry A, Abel TW, Burger PC, Dang CV,
Eberhart CG (2006) c-myc overexpression causes anaplasia in medulloblastoma. Cancer Res 66(2):673681
Sutton LN, Wehrli SL, Gennarelli L, Wang Z, Zimmerman
R, Bonner K, Rorke LB (1994) High-resolution
1H-magnetic resonance spectroscopy of pediatric
posterior fossa tumors in vitro. J Neurosurg
81(3):443448
Tugnoli V, Schenetti L, Mucci A, Nocetti L, Toraci C,
Mavilla L, Basso G, Rovati R, Tavani F, Zunarelli E,
Righi V, Tosi MR (2005) A comparison between
in vivo and ex vivo HR-MAS 1H MR spectra of a
pediatric posterior fossa lesion. Int J Mol Med
16(2):301307
Tzika AA, Cheng LL, Goumnerova L, Madsen JR,
Zurakowski D, Astrakas LG, Zarifi MK, Scott RM,
Anthony DC, Gonzalez RG, Black PM (2002)
116
Biochemical characterization of pediatric brain tumors
by using in vivo and ex vivo magnetic resonance spectroscopy. J Neurosurg 96(6):10231031
Tzika AA, Astrakas L, Cao H, Mintzopoulos D, Andronesi
OC, Mindrinos M, Zhang J, Rahme LG, Blekas KD,
Likas AC, Galatsanos NP, Carroll RS, Black PM
(2007) Combination of high-resolution magic angle
spinning proton magnetic resonance spectroscopy and
microscale genomics to type brain tumor biopsies. Int
J Mol Med 20(2):199208
Usenius JP, Vainio P, Hernesniemi J, Kauppinen RA
(1994) Choline-containing compounds in human
astrocytomas studied by 1H NMR spectroscopy
in vivo and in vitro. J Neurochem 63(4):15381543
Wharton BA, Morley R, Isaacs EB, Cole TJ, Lucas A
(2004) Low plasma taurine and later neurodevelopment.

Arch Dis Child Fetal Neonatal Ed 89(6):
F497F498
Wilson M, Davies NP, Grundy RG, Peet AC (2009) A
quantitative comparison of metabolite signals as detected
by in vivo MRS with ex vivo 1H HR-MAS for childhood
brain tumours. NMR Biomed 22(2):213219
Wright A, Fellows G, Griffiths J, Wilson M, Bell B, Howe
F (2010) Ex vivo HRMAS of adult brain tumours:
metabolite quantification and assignment of tumour
biomarkers. Mol Cancer 9(1)
Yang Y, Li C, Nie X, Feng X, Chen W, Yue Y, Tang H,
Deng F (2007) Metabonomic studies of human hepatocellular carcinoma using high-resolution magicangle spinning 1H NMR spectroscopy in conjunction
with multivariate data analysis. J Proteome Res
6(7):26052614
Central Nervous System Imaging

in Childhood Leukemia
12
Luciana Porto and Heinrich Lanfermann
Abstract
Contents
Introduction ............................................................
118
Central Nervous System Manifestations

of Leukemia ............................................................
118
Side Effects of the Therapeutic Procedures .........

Cerebrovascular Disorders .......................................
White Matter Changes: Leukoencephalopathy
Versus Transient Abnormalities ...............................
Posterior Reversible Encephalopathy
Syndrome (PRES) ....................................................
Infections..................................................................
Graft-Versus-Host Disease (GHVD)........................
Secondary Brain Tumors..........................................
120
120
123
123
124
125
Neurocognitive Outcome .......................................
125
Conclusion ..............................................................
126
References ...............................................................
126
122
The imaging of CNS has been increasingly

recognized in pediatric oncology as a complex area. Imaging has increased our ability
to diagnose CNS manifestations and treatment complications in childhood leukemia.
Compared to conventional imaging, advanced
MR modalities provide deeper insight into
pathophysiology and pathobiochemistry of
the CNS involvement in pediatric oncologic
diseases. This is important because survivors
of childhood leukemia are at risk of developing therapy-induced secondary neurological
impairments and/or secondary tumors, even
after many years. For a systematic approach
we have divided the CNS imaging into manifestations of leukemia and side effects of the
therapeutic procedures. Under CNS manifestations of the primary disease we understand
the leptomeningeal or parenchymal manifestations, as well as the involvement of intracranial vessels. The side effects of the therapeutic
procedures may include cerebrovascular disorders, white matter lesions, infections, atrophy and the most threatening of all, secondary
tumors. Complications, such as cerebrovascular disorders, can be due either to disease itself
or can be secondary to its treatment.
L. Porto (*) H. Lanfermann

Institut fr Neuroradiologie, Klinikum der Johann
Wolfgang Goethe-Universitt, Schleusenweg 2-16,
D-60528, Frankfurt am Main, Germany
e-mail: luciana.porto@kgu.de
117
118
Introduction
Childhood leukemia represents nearly one third
of all oncologic diseases in children. The
increased survival in this group of pediatric
oncologic patients is partly due to the successful control of central nervous system (CNS)
manifestations. Unfortunately, the prophylaxis
and treatment are associated with treatment
complications, including those affecting the
brain, with CNS therapy, and in particular the
use of radiation therapy, having long-term consequences. This knowledge has triggered the
use of other measures of CNS prophylaxis in
the treatment of childhood leukemia.
The imaging of CNS has therefore been
increasingly recognized in pediatric oncology as
a complex area, with the need for special expertise and multidisciplinary interaction to improve
the care of pediatric patients with leukemia
and CNS complications. The field of pediatric
magnetic resonance imaging (MRI), especially
in the field of neuro-oncology, is now expanding
rapidly. Imaging has increased our ability to
diagnose diseases and treatment complications
in pediatric neuro-oncology. Compared to conventional imaging, advanced MR modalities provide deeper insight into pathophysiology and
pathobiochemistry of the CNS involvement in
pediatric oncologic diseases. Advanced MRI
modalities include diffusion weighted imaging
(DWI) and diffusion tensor imaging (DTI), MR
perfusion, MR spectroscopy and voxel-based
morphometry. These techniques are proving
helpful, not only in comprehending the disease
process itself, but also in evaluating primary
therapeutic effects, as well as secondary deleterious side effects of therapy. This is important
because survivors of childhood leukemia are at
risk of developing therapy-induced secondary
neurological impairments and/or secondary
tumors, even after many years.
For a systematic approach we have divided the
CNS imaging into manifestations of leukemia
and side effects of the therapeutic procedures.
Under CNS manifestations of the primary disease
we understand the leptomeningeal or parenchy-
L. Porto and H. Lanfermann
mal manifestations, as well as the involvement of

intracranial vessels. The side effects of the therapeutic procedures may include cerebrovascular
disorders, white matter lesions, infections, atrophy and the most threatening of all, secondary
tumors. Complications, such as cerebrovascular
disorders, can be due either to disease itself or
can be secondary to its treatment. In our experience (Porto et al. 2004a), most of the CNS complications occurring during therapy or within
3 months of its completion are cerebrovascular
disorders, infections, meningeal leukemia, and
treatment-related neurotoxicity. Cerebrovascular
complications are the most common CNS
abnormality.
Central Nervous System

Manifestations of Leukemia
Leukemia in children usually presents with the
involvement of the hematopoietic organs. Yet the
CNS, considered a sanctuary site, can be affected,
primarily or later as the site of origin of relapse
after therapy. Although there has been a dramatic
drop in the incidence of CNS relapse since the
introduction of methotrexate prophylaxis, it still
occurs. Leukemic cells can involve the calvarial
bone marrow, dura, leptomeninges or all three
(Fig. 12.1). Price and Johnson (1973), who
studied the pathophysiology of CNS leukemia,
showed that leukemic cells initially infiltrate the
wall of superficial arachnoidal veins and surrounding adventitia. Often, with evolution of the
process, the arachnoid trabeculae are destroyed
with consequent leukemic contamination of the
cerebrospinal channels. The process can advance
with expansion into deep arachnoid surrounding
blood vessels in gray and white matter and disruption of the pial membrane, resulting in parenchymal infiltrates.
One of the most difficult challenges in pediatric
patients with leukemia is how to evaluate meningeal enhancement on MRI. In our daily practice,
we see not only patients with minimal enhancement
and positive cytology, but also children with clear
enhancement and negative cytology on a single
lumbar puncture. An easy approach is to divide
12
Central Nervous System Imaging in Childhood Leukemia
119
Fig. 12.2 Three-year-old girl with granulocytic sarcoma

at the right lacrimal fossa. T1-weighted MR image with
FAT SAT after contrast shows an enhancing mass at the
right lacrimal fossa (arrow) with intra- and extra-conal
infiltration and involvement of the M. rectus lateralis
Fig. 12.1 Six-year-old girl with right occipital meningeal leukemia. Axial T1-weighted MR image after contrast shows an enhancing right occipital epidural mass
(thick white arrows) with permeation of skull (black
arrow) and extracerebral soft-tissue component (curved
white arrows)
the meningeal enhancement into leptomeningeal

and pachymeningeal (lineal and thick enhancement
without extension into the gyri). Patients with
reactive meningitis secondary to radiation or
chemical meningitis due to intrathecal chemotherapy usually present with pachymeningeal enhancement. Leptomeningeal enhancement in pediatric
leukemia may result from CNS manifestation or
infection. Rarely can these two complications
present together. As a result, cytologic examination
of cerebrospinal fluid is essential for diagnosis.
Chemical meningitis is a quite relatively frequent
and benign complication, which presents with
transient abnormality of CSF during CNS prophylaxis without CNS involvement. It has been
reported, that CSF abnormality with or without
CNS symptoms may develop in 1060% of
patients who have received intrathecal injections
(Fukushima et al. 1999).
Ophthalmic manifestations in patients with

leukemia (Fig. 12.2) are intraocular or extraocular
with muscle or optic nerve involvement. The
presence of enhancing optic nerve enlargement in
children with a history of leukemia should suggest the diagnosis of leukemic infiltration even in
the absence of previous CNS involvement.
According to Madani et al. (2000), this area might
represent another sanctuary for leukemic cells
due to suboptimal penetration of chemotherapy
in the retrobulbar optic nerve. A rare clinical
manifestation of acute myeloid leukemia is central diabetes insipidus, caused by leukemic infiltration to the hypothalamo-neurohypophyseal
system. Characteristically, there is an absence of
a physiological bright spot of the neurohyphophysis on T1 images before contrast application. Granulocytic sarcoma, so called chloroma
is a rare manifestation of myelogenous leukemias
(38%) in which immature myeloid cells of granulocytic lineage infiltrate soft tissue and bone.
Due to the lesions typical greenish appearance, it
is named chloroma. It has been described in
almost every location, but is usually located in
the skull, orbita, and sinuses (Porto et al. 2004b).
It is most common in the pediatric population and
may present at any time during the course of the
disease, i.e., at presentation or during a remission
or relapse. It may be difficult to diagnose when it
120
precedes the clinical onset of acute myeloid

leukemia. Fat suppression techniques, in combination with gadolinium enhancement, are invaluable in the screening for soft tissue and bone
infiltrations (Porto et al. 2004b). The skull, especially the skull base and the spine are usually
infiltrated in leukemia, either during presentation
or relapse, as bone marrow manifestation. MR
imaging of the bone marrow in affected children
demonstrates low signal intensity leukemic infiltrates on T1-weighted images, particularly apparent in areas where red marrow is converting to
yellow marrow. It should be noted that bone marrow infiltration of skull and spine is common and
is usually not associated with CNS infiltration.
Contrast enhancement of the cauda equina in
children with leukemia is a serious complication
with a wide range of differential diagnosis.
Besides leukemic CNS manifestation, the following should be considered in the differential diagnosis: CNS arachnoiditis (after intrathecal
administration of chemotherapeutic agents such
as methotrexate or cytarabine), post-surgical
arachnoiditis, root compression associated with
inflammation, cytomegalovirus polyradiculopathy and inflammatory demyelinating polyradiculoneuropathy, i.e., Guillain-Barr syndrome.
Anterior lumbosacral radiculopathy can be due to
intrathecal
methotrexate
administration.
According to Vsquez et al. (2002), the clinical
picture consists of progressive, flaccid weakness
of the lower extremities without sensory deficit.
MRI after gadolinium shows enhancement of the
anterior lumbosacral nerve roots.
Side Effects of the Therapeutic

Procedures
Cerebrovascular Disorders
Cerebrovascular thrombosis or intracranial haemorrhage (ICH) can be due either to disease itself
or can be secondary to its treatment. Pediatric
patients with leukemia can show bleeding diathesis secondary to changes in the coagulation factors, such as thrombocytopenia and probable
associated intravascular coagulation. Fulminant
Fig. 12.3 Nine-year-old boy with cerebral hematoma.

CT scan without contrast enhancement shows hyperdense
area in the right frontal lobe with a peripheral rim (curved
white arrow)
leukocytosis, leukostasis or blast cell thrombi

within small arterioles can also generate vascular
injury and haemorrhage. In addition, thrombosis
or ICH (Fig. 12.3) can occur in the course of the
antileukemic treatment as a result of leukocytosis,
thrombocytopenia, sepsis, or coagulopathy. In
our experience (Porto et al. 2004b), about half of
children with leukemia have cerebrovascular
accidents that are not related to prednisone-vincristine-asparaginase treatment. Initial computed
tomography (CT) following intracranial bleeding
usually shows an elliptical or round parenchymal
mass with a density of 5080 Hounsfield units.
As the clot retracts, extravasated serum may produce a peripheral rim (Fig. 12.3). In patients with
leukemia and ICH, further MRI with gradient
echo (GRE-T2*) should be performed in search
for microbleeds. The MR appearance of the ICH
is variable. It not only depends on the age of the
clot, but also on such factors as oxygenation,
hemoglobin status, red blood cell membrane status,
dilution effects, field strength, and pulse sequence
parameters.
12
Fig. 12.4 Seven-year-old boy with sinus thrombosis of

the right transverse and sigmoid sinus. MR venogram
shows lack of flow related enhancement in the right transverse (thick white arrow) and sigmoid sinus (curved white
arrow)
Cerebral infarction in children with leukemia

is mainly related to sinusvenous thrombosis
(SVT), (Fig. 12.4), leukostasis with hypercoagulability, CNS infiltration, as well as the use of
certain chemotherapeutic agents, particularly
L-asparaginase, but also vincristine contribute to
venous occlusion. In our experience, ca. 50% of
the cerebrovascular accidents in patients with
leukemia are related to asparaginase (Porto et al.
2004a). According to Nowak-Gottl et al. (1999),
most patients with thrombotic complications
have one or more hereditary prothrombotic
defects. Treatment with asparaginase leads to the
depletion of plasma proteins involved in both
coagulation and fibrinolysis, and has been linked
to cerebrovascular complications, including not
only cortical infarcts and SVT, but also haemorrhage and haemorrhagic infarcts. Both bleeding
and thrombotic complications occur, at the earliest, 8 days after therapy with asparaginase has
been started (Kieslich et al. 2003). An occlusion
of a large sinus or an extensive smaller vein
obstruction can critically impair venous drainage.
Consequently, the cortex and localized white
matter become congested and swollen, leading to
venous infarction. A cerebral infarction has a
characteristic location and appearance. Typically
121
it has a non-arterial supply and is more frequently

haemorrhagic; it affects primarily the white matter
rather than the cortex. On CT, venous infarcts
usually present as poorly defined hypodense or
mixed attenuation areas with a slight mass effect.
The CT venogram shows thrombus as filling
defects in dural sinus or vein obstruction. On
MR, early venous infarcts may be identified by
reduced diffusion in the described characteristic
regions. Venous infarcts are often haemorrhagic,
with their imaging appearance varying from large
subcortical hematomas to petechial hemorrhages
with edema of the brain parenchyma. Hemosiderin
can also be seen in gradient echo images (T2*) in
case of haemorrhage. Time-of-flight and phase
contrast sequences are studies sensitive to slow
flow and are often used for screening the cerebral
venous flow (Fig. 12.4). Another cause of ischemic
accidents is radiation induced vascular damage.
According to Bowers et al. (2006), the rate of
late-occurring stroke for leukemia survivors is
57.9 per 100,000 person-years. Cranial radiation
is associated with an increased risk for cerebrovascular disease and its effect is dose-dependent
(Bowers et al. 2006). Unfortunately, radiation
can damage the normal brain tissue, probably due
to a disruption of the endothelial layer of the
larger vessels mediated by chronic inflammatory
response. Eventually, intimal fibrosis develops
with secondary narrowing of the vessels, i.e.,
signs of atherosclerosis (Morris et al. 2009).
Radiation-induced cerebrovascular changes in
childhood leukemia survivors include intracranial occlusive disease (including moyamoya),
cavernomas, telangiectasia and mineralizing
microangiopathy (Morris et al. 2009). Nowadays,
an annual neurologic examination is recommended by the Childrens oncology group in
patients who underwent 18 or more gray (Gy)
cranial radiation. If clinically indicated, brain
MRI with DWI and MR angiography should follow (Morris et al. 2009).
Children with leukemia who are treated with
cranial radiation can develop haemorrhagic lesions
within the white matter, similar to cavernous
angiomas on MR images. These lesions are probably capillary telangiectasia secondary to the
damaged venular endothelium, with consequent
122
venous occlusive disease (Vsquez et al. 2002). It

may be isolated or multiple, and can show calcification. Gradient-echo (T2*) and susceptibilityweighted imaging (SWI) sequences are particularly
useful for their identification because of their
greater magnetic susceptibility. It is of note that
children have more frequent radiation-induced
vascular malformations than adults (Morris et al.
2009). Telangiectasias are described in 20% of
survivors after radiation, independent of the dose
(Koike et al. 2004). In contrast to telangiectasias
and cavernous angiomas, which have a restricted
danger of bleeding, radiation-induced aneurysms
can have a high risk of haemorrhage (Morris et al.
2009). Patients can also present after radiation
with recurrent headaches with reversible neurologic deficits; so-called stroke-like migraine.
Characteristically, there is a reversible cortical
ribbon-like enhancement after contrast application on MRI (Partap et al. 2006).
White Matter Changes:

Leukoencephalopathy Versus
Transient Abnormalities
Not only the classical combination of treatment
with cranial radiation therapy and methotrexate
(MTX) but also MTX alone can cause white matter damage. Pathologically, high-dose MTX is
associated with demyelination, white matter
necrosis, loss of oligodendroglia, axonal swelling, microcystic encephalomalacia and atrophy,
which is selective for the deep cerebral white
matter (Roolins et al. 2004). Fortunately, most
white matter changes appear to be transient and
are not associated with neurologic deficits
(Reddick et al. 2005, 2007, 2009). Chu et al.
(2003) have demonstrated with MR spectroscopy
a transient decrease in N-Acetyl Aspartate/choline (NAA/Cho) ratio after high-dose MTX therapy in most patients with acute lymphoblastic
leukemia (ALL), either with or without white
matter changes. The metabolite ratio then returns
to a physiologic increase with brain maturation.
Low levels of NAA, which is considered to be
primarily of neuronal origin, can be interpreted
as a consequence of demyelination. In addition, a
significant reduction of the mean NAA/Cr

(creatine) ratio is reported in children who underwent cranial radiation (Ficek et al. 2010). In corroboration, Reddick et al. (2009) described, using
DTI, a temporary decreased fractional anisotropy
(FA), with a pattern consistent with transient
demyelination. However, while on one hand, the
clinically acute neurotoxicity associated with
intrathecal MTX is not automatically associated
with irreversible cell death, on the other hand
progressive and persistent white matter changes
may be seen in the absence of symptomatic neurotoxicity. It remains an open question whether
the transient abnormalities in the cerebral white
matter may contribute to the development of leukoencephalopathy. Typically, leukoencephalopathy has been associated with demyelination of
the periventricular white matter that can be seen
as early as 9 months after treatment with cranial
radiation and MTX (Porto et al. 2004a).
Nevertheless, it is described that MTX alone can
cause leukoencephalopathy that may be progressive (Reddick et al. 2005). In MRI, leukoencephalopathy is characterized by high signal intensity
on T2-weighted images. It affects the deep white
matter, but may also extend to more peripheral
white matter. Some of the leukoencephalopathy
changes are transient with significant reduction
in the prevalence of leukoencephalopathy
approximately 1.5 years after the completion of
IV MTX therapy (Reddick et al. 2005).
Pathologically, late delayed injury (months to
years after treatment) include white matter necrosis, demyelination, astrocytosis, and vasculopathy (Edwards-Brown and Jakacki 1999). It is
known that delayed toxic effects associated with
radiation therapy can be severe and irreversible,
resulting from white matter vasculopathy. Porto
et al. (2008), using morphometry associated with
FA and mean diffusilibility (MD) measurements
to assess microstructural changes of white matter
integrity, showed a persistent reduced white matter volume in long-term survivors of ALL.
Survivors of childhood ALL who underwent
treatment (with or without radiation) have smaller
white matter volumes compared to controls, even
in the absence of clear lesions on conventional
images. In addition, chemotherapy alone had a
12
123
smaller impact and damage on the white matter

compared to radiation. Thus, transient white matter
changes observed during therapy may lead to
lasting, subtle white matter damage that is usually not recognized by late conventional MRI. In
addition, the grey matter seems to be affected
selectively by treatment. Lower grey matter volume was found in the caudate nuclei heads structures in the irradiated group. Because the caudate
head is involved in cognitive functions, its lower
volumes could contribute to the cognitive deficit
in these patients.
Posterior Reversible Encephalopathy

Syndrome (PRES)
Pediatric patients under treatment for myeloproliferative disorders can present with acute
neurologic complications occurring with chemotherapeutic and immunosuppression medication,
mainly cyclosporine A (CsA) but also tacrolimus
(FK-506), cisplatin, interferon-alpha and erythropoietin. Neurologic manifestations include
headache, nausea, vomiting, seizures, visual
changes, confusion, and coma. To date, no clear
relationship between CsA levels and neurotoxicity has been established (No et al. 2010).
Clinically, PRES was first described as acute
neurologic changes in the background of arterial
hypertension that overcomes the autoregulatory
capacity of the cerebral vasculature. However,
CsA neurotoxicity also occurs in normotensive
individuals, suggesting that additional causative
factors are present. It is generally established
that acute hypertension or drug toxicity damages
the vascular endothelium. Cyclosporine A, for
example, is known to have profound effects on
vascular endothelium and to cause the release of
potent vasoconstrictors such as endothelin, prostacyclin, and thromboxane A2 (Reece et al.
1991). The relative scarcity of sympathetic innervation in the posterior circulation may be related
to the preponderance of posterior cerebral
changes (Cooney et al. 2000). Accordingly, arteriography in patients with posterior encephalopathy shows diffuse vascular narrowing, with
slight preference for the posterior circulation
Fig. 12.5 Four-year-old boy with CsA neurotoxicity.

DWI image shows left parietal corticosubcortical high
signal lesions (white thick arrow)
(Weidauer et al. 2003). The vascular damage

leads to a breakthrough of autoregulation, which
causes blood brain barrier disruption and
vasogenic edema. Characteristically, MR imaging
shows subcortical white matter edema, predominantly in the posterior temporal, parietal, and
occipital areas, albeit also affecting the anterior
frontal lobes, basal ganglia, cerebellar hemispheres and brainstem in more severe cases.
Junctions of vascular watershed zones are usually
involved. It is usually bilateral and often to some
extent asymmetric (Fig. 12.5). The affected areas
usually have increased diffusion (Fig. 12.5) with
elevated apparent diffusion coefficient (ADC),
i.e., vasogenic edema. Reduced ADC may
indicate irreversible infarction. There is usually
a variable patchy enhancement after contrast.
Full neurologic recovery with resolution of neuroimaging abnormalities is often the end result
after removal of the immunosuppressive drug
and/or control of blood pressure. Yet, if untreated,
it can lead to progressive neurologic decline with
infarction, hemorrhage and potential irreversible
neurologic deficit. Therefore, it is disputable if
the R (for reversible) in the denomination
PRES should be disregarded.
124
Fig. 12.6 One-year-old girl with presumed Aspergillus

infection. T1-weighted MR axial image after i.v. contrast
shows disseminated enhancing nodules in the brain. The
associated with alveolitis on lung CT scans prompted
open biopsy of the cerebral and lung lesions. Microscopic
examination of the brain specimens revealed no fungus or
bacteria. The patient was then treated empirically (no isolation of Aspergillus spp. in cultures or histological finding of septate hyphae in affected biopsy) for fungus with
intravenous amphotericin B in high doses, and the symptoms disappeared. Follow-up MRI showed complete resolution of the cerebral lesions
Infections
A series of deficiencies may occur in the immune
systems of patients with leukemia. In addition,
therapy leads to secondary immune deficiency
(Lehrnbecher et al. 1997). Therefore, children
with leukemia and an immuno-compromised status
are susceptible to opportunistic infections. In
addition, mucositis, veinpucture, the use of catheters and bone marrow punctions proportionate
the ideal route for organisms to invade the
host. Neutropenia increases the risk for bacterial
infections; notably, prolonged neutropenia
(>10 days) is associated with invasive fungal
disease (Lehrnbecher et al. 1997; Pui 1999).
These infections are usually caused by Candida
and Aspergillus spp. (Fig. 12.6). Unfortunately,
this last opportunistic mycosis is an important

cause of morbidity and mortality in immunocompromised pediatric patient populations,
including, but not limited to those undergoing
hematopoietic
stem
cell
transplantation
(Lehrnbecher and Groll 2011). There is haematogenic dissemination to the CNS from lung. An
alternative route is the cribriform plate; fungal
sinusitis, usually aspergillosis or mucormycosis,
may progress to rhinocerebral syndrome with
invasion of the CNS. Once in the brain, aspergillus causes infectious vasculopathy leading to
acute infarction or hemorrhage, or extends into
surrounding tissue as infectious cerebritis or
abscess. Multiple haemorrhagic parenchymal
lesions are characteristic in MRI. Ashdown et al.
(1994) describe three imaging patterns in neutropenic patients with cerebral invasive aspergillosis:
(1) cortical-subcortical hyperintense areas on
T2-weighted images, (2) multiple ring-enhancing
lesions, (3) dural enhancement adjacent to sinonasal disease. Infarction or haemorrhage can be an
early radiological presentation, due to the angioinvasive nature of the infection (DeLone et al.
1999). Multiple intracranial bleeding or infarction
in an uncommon distribution, and infectious
lesions with unusual enhancement characteristics,
may result from opportunistic mycosis infection
with vasculopathy. Any of these findings in the
immuno-compromised patient should suggest a
diagnosis of aspergillosis.
Graft-Versus-Host Disease (GHVD)

All patients who receive allogenic marrow are at
risk of developing GVHD. Characteristically, this
disorder does not occur before the recovery of
lymphocytes, which play a central role in the
pathophysiology of GvHD. To date, there has
only been limited information about CNS involvement in GVHD (Padovan et al. 1999; Provenzale
and Graham 1996). Recently, a long-term followup of four allogeneic bone marrow recipients
who developed angiitis-like disease 218 years
after transplantation (following reduction of
immuno-suppressive therapy), most likely due to
GVHD, were reported (Sostak et al. 2010).
12
Unfortunately, MRI was unspecific and variable,

showing generalized atrophy, ischemic lesions or
leukoencephalopathy. The diagnosis of cerebral
angiitis was confirmed by histopathology and
response to immuno-suppressive therapy.
Therefore, the presence of non-infectious angiitislike disease of the CNS in long-term survivors after
bone marrow transplantation (BMT) should suggest the diagnosis of GVHD (Sostak et al. 2010).
Secondary Brain Tumors

As the cure rates in children with leukemia
increase, more children are exposed to the risk of
long-term sequelae, including the development
of a second malignant neoplasm. The cumulative
risk of developing a secondary neoplasm ranges
from 1.2% to 3.3% after 1015 years of follow-up
(Hijiya et al. 2007). Unfortunately, in a long term
follow-up time up to 41.3 years, the cumulative
incidence of secondary neoplasm at 30 years after
ALL is 10.85%. Instead of reaching a plateau at
1520 years, the cumulative incidence of secondary neoplasms continues to increase over 30 years
(Hijiya et al. 2007). In our experience (Porto et al.
2004a), secondary brain tumors develop from 8
up to 23 years after initial treatment. According
to Walter (2004) approximately 1% of patients
who receive cranial radiation will develop brain
tumors; the latency period ranges from 9 years
for high-grade gliomas to 19 years for meningiomas. Fortunately, most of the late intracranial
secondary tumors are meningiomas, but more
aggressive tumors such as gliomas and sarcomas
also occur (Hijiya et al. 2007). Besides, secondary intracranial meningiomas after high-dose cranial radiation are likely to have an atypical
behavior and to recur (Strojan et al. 2000). Other
secondary tumors are ependymoma and lymphoma. While cranial radiation is clearly responsible for the development of secondary brain
tumors, survivors of childhood leukemia without
a previous history of radiation but with secondary
malignant brain tumors have been reported
(Gilman and Miller 1981). Majhail et al. (2011)
also described secondary brain tumors in survivors
of leukemia who underwent hematopoietic cell
125
transplant (HCT). The rates of development of a

new solid cancer in HCT recipients are twice that
expected for the general population. Important
risk factors are age at transplantation, radiation
(conditioning regimen), and chronic graft-versushost-disease (Majhail et al. 2011). In addition,
other factors, such as loss of immune surveillance
and genetic factors have been proposed.
Specifically risk factors for children are cranial
radiation at 5 years of age or younger, genetic
predisposition to tumors and survivors of BMT.
Unfortunately, treatment-induced tumors tend to
be more aggressive and refractory to therapy
(Choi and Seex 2000). Because of the unequivocal risk of cranial and spinal radiation, this treatment is no longer given prophylactically.
Although the pathogenesis of these cancers is
multifactorial, there is increasing evidence that
genetic characteristics, i.e., high genetic susceptibility to secondary tumors, will be an important
issue in the future, and this group of children will
deserve special treatment (Hijiya et al. 2007;
Walter et al. 1998; Relling et al. 1999).
Neurocognitive Outcome
It is widely accepted that survivors of childhood
leukemia are at risk for cognitive impairments
(Kesler et al. 2010). Different studies have shown
white matter changes (Porto et al. 2008; Reddick
et al. 2009) with reduced volume and lower fractional anisotropy after leukemic treatment associated with cognitive deficits (Aukema et al. 2009;
Reddick et al. 2006). A series of factors have a
negative impact on cognitive outcome in survivors of childhood leukemia; the most important
ones being CNS involvement, the dose of radiotherapy and chemotherapeutic agents. As a general rule, children treated for leukemia who
received cranial radiation have a high risk of significant dysfunction. In comparison, survivors
treated without radiation tend to preserve the
cognitive function (Kamps et al. 2010; Tabone
and Leverger 2009). According to Kesler et al.
(2010) the cognitive reserve, as indicated by the
level of maternal education, was inversely associated with the white matter volume in survivals of
126
childhood leukemia without radiation. The

authors suggested that in these children a greater
white matter involvement is required before the
presentation of cognitive defects. Furthermore, it
seems that survivors are able to re-route cognitive
function to uninjured regions. This could explain
the discrepancy between the degree of atrophy/
white matter lesion and the surprising preservation of cognitive function in patients treated only
with chemotherapy (Kesler et al. 2010).
Conclusion
Because of increased survival, there is a growing
aging population of survivors of childhood leukemia. This population requires monitoring by a
multidisciplinary team, such as oncologists,
neurologists and neuroradiologists, who have a
special expertise and are aware of morbidities
that commonly affect the CNS in this population.
CNS radiation, which was performed two decades
ago in all children, is given today only to a minority of children. That is also one of the big challenges of MRI: to better stratify who needs more
intensive therapy and who does not. Consequently,
MRI plays a major role in the diagnosis and monitoring of children with leukemia: first in helping
to establish the disease extent, which may determine therapeutic strategies; second, in ruling out
complications, related either to treatment or
relapse; and last, in evaluating long-term side
effects. More studies with new techniques, such
as morphometry, diffusion, perfusion and spectroscopy, are necessary to determine to which
degree the childhood brain is able to respond and
reorganize itself after therapy. In this way, the
ideal therapy can be selected, minimizing the
neurotoxic effects.
References
Ashdown BC, Tien RD, Felsberg GJ (1994) Aspergillosis
of the brain and paranasal sinuses in immunocompromised patients: CT and MR imaging findings. AJR
Am J Roentgenol 162:155159
Aukema EJ, Caan MW, Oudhuis N, Majoie CB, Vos FM,
Reneman L, Last BF, Grootenhuis MA, Schouten-van

Meeteren AY (2009) White matter fractional
anisotropy correlates with speed of processing and
motor speed in young childhood cancer survivors. Int
J Radiat Oncol Biol Phys 74:837843
Bowers DC, Liu Y, Leisenring W, McNeil E, Stovall M,
Gurney JG, Robison LL, Packer RJ, Oeffinger KC
(2006) Late-occurring stroke among long-term survivors of childhood leukemia and brain tumors: a report
from the Childhood Cancer Survivor Study. J Clin
Oncol 24(33):52775282
Choi D, Seex K (2000) Intracranial meningioma following childhood radiation for leukaemia. Br J Hematol
108:665
Chu WCW, Chik K, Chan Y, Yeung DKW, Roebuck DJ,
Howard RG, Li C, Metreweli C (2003) White matter
and cerebral metabolite changes in children undergoing treatment for acute lymphoblastic leukemia: longitudinal study with MR imaging and 1H MR
spectroscopy. Radiology 229:659669
Cooney MJ, Bradley WG, Symko SC, Patel ST, Groncy
PK (2000) Hypertensive encephalopathy in children
treated for myeloproliferative disorders: report of three
cases. Radiology 214:711716
DeLone DR, Goldstein RA, Petermann G, Salamat MS,
Miles JM, Knechtle SJ, Brown WD (1999)
Disseminated aspergillosis involving the brain: distribution and imaging characteristics. AJNR Am J
Edwards-Brown MK, Jakacki RI (1999) Imaging the central nervous system effects of radiation and chemotherapy of pediatric tumors. Neuroimaging Clin N Am
9:177193
Ficek K, Blamek S, Sygua D, Miszczyk L, So taJakimczyk D, Tarnawski R (2010) Evaluation of the
late effects of CNS prophylactic treatment in childhood acute lymphoblastic leukemia (ALL) using magnetic resonance spectroscopy. Acta Neurochir Suppl
106:195197
Fukushima T, Sumazaki R, Koike K, Tsuchida M, Okada
Y, Maki T, Hamano K (1999) A magnetic resonance
abnormality correlating with permeability of the
blood-brain barrier in a child with chemical meningitis
during central nervous system prophylaxis for acute
leukemia. Ann Hematol 78:564567
Gilman PA, Miller RW (1981) Cancer after acute lymphoblastic leukemia. Am J Dis Child 135:311312
Hijiya N, Hudson MM, Lensing S, Zacher M, Onciu M,
Behm FG, Razzouk BI, Ribeiro RC, Rubnitz JE,
Sandlund JT, Rivera GK, Evans WE, Relling MV, Pui
CH (2007) Cumulative incidence of secondary neoplasms as a first event after childhood acute lymphoblastic leukemia. JAMA 297:12071215
Kamps WA, van der Pal-de Bruin KM, Veerman AJ,
Fiocco M, Bierings M, Pieters R (2010) Long-term
results of Dutch Childhood Oncology Group studies
for children with acute lymphoblastic leukemia from
1984 to 2004. Leukemia 24:309319
Kesler SR, Tanaka H, Koovakkattu D (2010) Cognitive
reserve and brain volumes in pediatric acute lymphoblastic leukemia. Brain Imaging Behav 4:256269
12
Kieslich M, Porto L, Lanfermann H, Jacobi G, Schwabe

D, Bhles H (2003) Cerebrovascular complications of
L-asparaginase in the therapy of acute lymphoblastic
leukemia. J Pediatr Hematol Oncol 25(6):484487
Koike S, Aida N, Hata M, Fujita K, Ozawa Y, Inoue T
(2004) Asymptomatic radiation-induced telangiectasia in children after cranial radiation: frequency,
latency, and dose relation. Radiology 230(1):9399
Lehrnbecher T, Groll AH (2011) Invasive fungal infections in the pediatric population. Expert Rev Anti
Infect Ther 9:275278
Lehrnbecher T, Foster C, Vzquez N, Mackall CL,
Chanock SJ (1997) Therapy-induced alterations in
host defense in children receiving therapy for cancer.
J Pediatr Hematol Oncol 19:399417
Madani A, Christophe C, Ferster A, Dan B (2000)
Perioptic nerve infiltration during leukaemic relapse:
MRI diagnosis. Pediatr Radiol 30:3032
Majhail NS, Brazauskas R, Rizzo JD, Sobecks RM, Wang
Z, Horowitz MM, Bolwell B, Wingard JR, Socie G
(2011) Secondary solid cancers after allogeneic
hematopoietic cell transplantation using busulfancyclophosphamide conditioning. Blood 117:316322
Morris B, Partap S, Yeom K, Gibbs IC, Fisher PG, King
AA (2009) Cerebrovascular disease in childhood cancer survivors: a childrens oncology group report.
Neurology 73(22):19061913
No A, Cappelli B, Biffi A, Chiesa R, Frugnoli I, Biral E,
Finizio V, Baldoli C, Vezzulli P, Minicucci F, Fanelli
G, Fiori R, Ciceri F, Roncarolo MG, Marktel S (2010)
High incidence of severe cyclosporine neurotoxicity in
children affected by haemoglobinopaties undergoing
myeloablative haematopoietic stem cell transplantation: early diagnosis and prompt intervention ameliorates neurological outcome. Ital J Pediatr 36:14
Nowak-Gottl U, Wermes C, Junker R, Koch HG, Schobess
R, Fleischhack G, Schwabe D, Ehrenforth S (1999)
Prospective evaluation of the thrombotic risk in children with ALL carrying the MTHFR TT 677 genotype, the prothrombin G20210A variant and further
prothrombotic risk factors. Blood 1(93):15951599
Padovan C, Bise K, Hahn J, Sostak P, Holler E, Kolb HJ,
Straube A (1999) Angiitis of the central nervous system after allogeneic bone marrow transplantation?
Stroke 30:16511656; Neuroradiology 46:374377
Partap S, Walker M, Longstreth WT Jr, Spence AM (2006)
Prolonged but reversible migraine-like episodes long
after cranial radiation. Neurology 66(7):11051107
Porto L, Kieslich M, Schwabe D, Zanella FE, Lanfermann
H (2004a) Central nervous system imaging in childhood leukaemia. Eur J Cancer 40:20822090
Porto L, Kieslich M, Schwabe D, Zanella FE, Lanfermann
H (2004b) Granulocytic sarcoma in children.
Neuroradiology 46:374377
Porto L, Preibisch C, Hattingen E, Bartels M,
Lehrnbecher T, Dewitz R, Zanella F, Good C,
Lanfermann H, Dumesnil R, Kieslich M (2008)
Voxel-based morphometry and diffusion-tensor MR
imaging of the brain in long-term survivors of childhood leukemia. Eur Radiol 18:26912700
127
Price RA, Johnson WW (1973) The central nervous

system in childhood leukemia: I. The arachnoid.
Cancer 31:520533
Provenzale J, Graham M (1996) Reversible leukoencephalopathy associated with graft-versus-host disease:
MR findings. AJNR Am J Neuroradiol 17:12901294
Pui CH (1999) Childhood leukemias. Cambridge
University Press, Cambridge
Reddick WE, Glass JO, Helton KJ, Langston JW, Xiong
X, Wu S, Pui CH (2005) Prevalence of leukoencephalopathy in children treated for acute lymphoblastic leukaemia with high-dose methotrexate. AJNR Am J
Reddick WE, Shan ZY, Glass JO, Helton S, Xiong X,
Wu S, Bonner MJ, Howard SC, Christensen R, Khan
RB, Pui CH, Mulhern RK (2006) Smaller whitematter volumes are associated with larger deficits in
attention and learning among long-term survivors of
acute lymphoblastic leukemia. Cancer 106:
941949
Reddick WE, Laningham FH, Glass JO, Pui CH (2007)
Quantitative morphologic evaluation of magnetic resonance imaging during and after treatment of childhood leukemia. Neuroradiology 49:889904
Reddick WE, Glass JO, Johnson DP, Laningham FH, Pui
CH (2009) Voxel-based analysis of T2 hyperintensities in white matter during treatment of childhood leukemia. AJNR Am J Neuroradiol 30:19471954
Reece DE, Frei-Lahr DA, Shephard JD, Dorovini-Zis K,
Gascoyne RD, Graeb DA, Spinelli JJ, Barnett MJ,
Klingemann HG, Herzig GP (1991) Neurologic complications in allogenic bone marrow transplant patients
receiving cyclosporine. Bone Marrow Transplant
8:393401
Relling MV, Rubnitz JE, Rivera GK, Boyett JM, Hancock
ML, Felix CA, Kun LE, Walter AW, Evans WE, Pui
CH (1999) Central nervous system imaging in childhood leukemia, high incidence of secondary brain
tumors after radiotherapy and antimetabolites. Lancet
354(9172):3439
Roolins N, Winick N, Bash R, Booth T (2004) Acute
methotrexate neurotoxicity: findings on diffusionweighted imaging and correlation with clinical outcome. AJNR Am J Neuroradiol 25:16881695
Sostak P, Padovan CS, Eigenbrod S, Roeber S, Segerer S,
Schankin C, Siegert S, Saam T, Theil D, Kolb HJ,
Kretzschmar H, Straube A (2010) Cerebral angiitis in
four patients with chronic GVHD. Bone Marrow
Transplant 45:11811188
Strojan P, Popovi M, Jereb B (2000) Secondary intracranial meningiomas after high-dose cranial radiation:
report of five cases and review of the literature. Int J
Radiat Oncol Biol Phys 48:6573
Tabone MD, Leverger G (2009) Outcome of children
cured of acute lymphoblastic leukemia. Bull Acad
Natl Med 193:15191528
Vsquez E, Lucaya J, Castellote A, Piqueras J, Sainz P,
Oliv T, Snchez-Toledo J, Ortega JJ (2002)
Neuroimaging in pediatric leukemia and lymphoma:
differential diagnosis. Radiographics 22:14111428
128
Walter AW (2004) Brain tumors in children. Curr Oncol
Rep 6:438444
Walter AW, Hancock ML, Pui CH, Hudson MM, Ochs JS,
Rivera GK, Pratt CB, Boyett JM, Kun LE (1998)
Secondary brain tumors in children treated for ALL at

St. Jude Childrens Research Hospital. J Clin Oncol
16:37613767
Weidauer S, Gaa J, Sitzer M, Hefner R, Lanfermann H, Zanella
FE (2003) Posterior encephalopathy with vasospasm:
MRI and angiography. Neuroradiology 45:869876
in the Differential Diagnosis
of Adult and Pediatric Brain Tumors
13
Aditya Raghunathan
Contents
Introduction ............................................................
139
References ...............................................................
140
129
Differentiating Non-neoplastic Tissue

from Diffuse Glioma ..............................................
130
Immunohistochemistry in Gliomas ......................
131
Immunohistochemistry in Glioneuronal
Neoplasms ...............................................................
133
Differentiation of Atypical Teratoid/Rhabdoid

Tumors and Primitive Neuroectodermal
Tumors ...................................................................
134
INI1 in Other Tumor Types: Choroid

Plexus Carcinoma ..................................................
135
INI1 in Other Tumor Types: Multiple

Schwannoma Syndromes.......................................
135
Identifying Medulloblastoma Subtypes

by Immunohistochemistry.....................................
135
Immunohistochemistry in the Diagnosis

of Intracranial Germ Cell Tumors .......................
136
Differentiating Meningiomas,
Hemangiopericytomas
and Solitary Fibrous Tumors ................................
137
Diagnosis of Primary CNS Lymphomas ..............
137
Differentiating Hemangioblastoma
from Metastatic Renal Clear Cell Carcinoma .....
138
Brachyury and the Diagnosis of Chordoma ........
138
A. Raghunathan (*)
The Methodist Hospital & University of Texas M.D.
Anderson Cancer Center, Houston, TX, USA
e-mail: araghunathan@tmhs.org
Immunohistochemistry in Carcinoma,
Melanoma and Sarcoma ........................................
Abstract
Immunohistochemical stains are of increasing

importance in evaluating adult and pediatric
intracranial neoplasms. Surgical neuropathologists frequently utilize these in the subclassification and characterization of CNS
neoplasms, assessment of prognosis, identification of therapeutic targets, and possibly
helping predict response to therapy. Here,
we review existing and novel antibodies that
are of immediate utility in the differential
diagnosis of adult and pediatric CNS neoplasms. The discussion focuses on the role
of various immunostains in differentiating
non-neoplastic brain from glioma, in subtyping glial and non-glial tumors, and in providing clinically relevant prognostic and
predictive information.
Introduction
The spectrum of central nervous system (CNS)
neoplasia in adult and children is broad. The
major lineages of primary CNS neoplasia include
astrocytic, oligodendroglial, ependymal, neuronal,
choroid plexus, meningothelial, hematolymphoid,
germ cell, melanocytic, mesenchymal (including

129
130
chondroid, osseous, vascular), as well as neoplasia of uncertain histogenesis. Various epithelial

and mesenchymal neoplasms may also present as
intracranial metastases. The first step in the evaluation of intracranial, and indeed any, neoplasms
is identifying morphologically abnormal cells.
Following this, immunohistochemical stains are
useful in identifying the lineage of the atypical
cells, highlighting specific morphological features, providing an index for estimating aggressiveness, helping assess prognosis, identify
targets for directed therapy, and possibly predict
response. An awareness of the strengths and limitations of various immunostains is also essential
for the pathologist to be able to provide the most
clinically relevant information for patient management. In this chapter, we review key immunohistochemical stains that help clarify the
differential diagnosis of various adult and pediatric
central nervous system neoplasms.
Differentiating Non-neoplastic Tissue

from Diffuse Glioma
The first step in evaluating CNS neoplasia is
determining whether the tissue being examined is
normal for the site of origin. Increased cellularity
greater than typical for a particular region of the
CNS raises the possibility of reactive gliosis,
inflammation, or of involvement by diffuse
glioma. Gliosis is the reaction of astrocytes to
any form of insult, and can lead to an appearance
of increased cellularity. Glial Fibrillary Acid
Protein (GFAP) is a cytoplasmic intermediate
filament found in normal and neoplastic glial
cells. The immunohistochemical stain for GFAP
highlights the cytoplasm and processes of glial
cells, more frequently in astrocytoma than in
oligodendroglioma. Studies by Cosgrove et al.
(1989) and Oh and Prayson (1999), demonstrated
GFAP to have 100% sensitivity in astrocytoma
and glioblastoma. In gliosis, GFAP immunostain
highlights astrocytes that are evenly distributed
in the tissue, have abundant cytoplasmic processes and, consequently, appear to have low
nuclear: cytoplasmic ratios. Gliomas, in contrast,
tend to have unevenly distributed and closely
A. Raghunathan
spaced cells, with expression of GFAP depending

on the amount of cytoplasm. Morphologically,
finding hypercellular areas, with unevenly distributed glial cells that have atypical nuclei should
raise suspicion for subtle involvement of the submitted tissue by a diffuse glioma.
Inflammatory cells, especially macrophages,
may cause tissue hypercellularity, may be morphologically identical to the atypical nuclei seen
in gliosis or gliomas, and so enter the differential
diagnosis of gliosis and low grade gliomas. This
is readily resolved by appropriate use of the histiocytic markers. The CD68 glycoprotein is a
lysosomal marker that is expressed by macrophages, and in benign and malignant histiocytic lesions. However, microglia in the normal
brain may also show CD68 expression. The lack
of GFAP immunostaining combined with positive cytoplasmic CD68 helps to differentiate
macrophages/microglia from glial cells.
The TP53 gene is the most frequently mutated
gene in human malignancies. This tumor suppressor gene is involved cell cycle arrest in G1
phase and initiation of apoptosis. Mutations in
TP53 are found in a high percentage (4060%)
of diffuse astrocytomas, particularly gemistocytic astrocytomas (up to 80%), and are less frequently present in oligodendrogliomas. The
wild-type p53 protein has a half-life of 530 min,
being readily degraded and not normally detectable by immunohistochemical stains. Mutations
of TP53 result in a non-functional p53 protein
that is not degraded as readily. However, the p53
immunoreactivity is not limited to gliomas
alone. Other means of retarded degradation of
wild-type p53 include binding by other oncoproteins such as the murine double minute-2,
and other epigenetic changes that may occur
during conditions of cellular stress. Whatever
the mechanism, the resultant p53 is more stable
and so is detectable by the p53 immunostain.
Kurtkaya-Yapicier et al. (2002) demonstrated
nuclear p53 immunoreactivity in astrocytes,
macrophages, and also in rare neurons, oligodendroglia, endothelial cells and lymphocytes, in
various non-neoplastic lesions that were in the
differential diagnosis of glioma. There were no
TP53 mutations accompanying the positive p53
13
Immunohistochemistry in the Differential Diagnosis of Adult and Pediatric Brain Tumors
immunostaining in any of the non-neoplastic

conditions, unlike diffuse astrocytoma that
frequently exhibits p53 immunoreactivity and
TP53 mutation. This would urge caution in
diagnosing a glioma on the basis of increased
p53 nuclear expression alone.
The labile, non-histone nuclear protein Ki-67
is a marker of cell proliferation. It is expressed in
the G1 thru M phases of the cell cycle and is
rapidly degraded at the end of M phase. It is not
detectable in the resting G0 phase and early in the
G1 phase of the cell cycle. The monoclonal IgG
immunoglobulin MIB-1 is directed against the
Ki-67 protein and is used in formalin-fixed, paraffin embedded tissue. Proliferating cells that are
in the cell cycle show strong nuclear staining for
MIB-1 and widespread nuclear MIB-1 expression is indicative of a rapidly proliferating neoplasm. Although helpful, the MIB-1 labeling
index does not form the basis of grading gliomas.
Used in conjunction with p53, a lymphocytic
marker and a macrophage marker (CD68), MIB-1
may help identify inflammatory cells as the basis
of increased cellularity.
Isocitrate dehydrogenase 1 (IDH1) on 2q33.3
encodes a nicotinamide adenine dinucleotide
phosphate (NADP+) dependent enzyme (IDH1)
in the citric acid cycle that catalyzes the cytosolic
oxidative decarboxylation of isocitrate to alphaketoglutarate, producing reduced NADP+
(NADPH). Mutations of IDH1 predominantly
involve arginine at position 132 in the amino acid
sequence of IDH1 protein. A majority of the
mutations involve exchange of guanine for adenine
(G395A), resulting in replacement of arginine by
histidine at position 132 (R132H). Initially, the
IDH1 status could only be assessed by DNA
sequencing or by polymerase chain reaction
(PCR) utilizing restriction endonuclease-based
detection of mutations in codon 132 of IDH1, as
described by Meyer et al. (2010). Capper et al.
(2009) have demonstrated high specificity and
sensitivity of immunohistochemistry utilizing a
mutation-specific mouse monoclonal antibody
for detecting the mutant IDH1R132H protein in
formalin-fixed paraffin-embedded (FFPE) tissue
sections. Capper et al. (2010a) also reported p53
immunoreactivity in 63% of 120 specimens with
131
reactive gliosis due to non-neoplastic etiologies.

In contrast, none (0%) of the cases were immunoreactive for the mutant-specific IDH1R132H
immunostain. Further, among 19 cases of posttherapy WHO grade II and grade III gliomas
with radiation change and extensive reactive
changes, immunostain for the mutant-specific
IDH1R132H identified positive cells in 13 specimens. In five of these cases, tumor cells were
missed by conventional staining. The mutantspecific IDH1R132H immunostain, therefore,
appears to be tumor-specific, and holds promise
in helping to differentiate reactive glial cells
from grade II and grade III diffuse gliomas, as
well as identify residual diffuse glioma cells
in specimens with prominent therapy-related
changes.
Immunohistochemistry in Gliomas
The glial fibrillary acidic protein (GFAP) is an
intermediate filament expressed by normal glial
cells and by glial neoplasms that can be identified
reliably using immunohistochemistry. While
positive GFAP staining of tumor cells in CNS
neoplasms supports the diagnosis of glioma, it
has sometimes been regarded as more specific for
astrocytic, rather than oligodendroglial, differentiation. However, neoplastic oligodendroglioma
cells, especially minigemistocytes and
gliofibrillary oligodendrocytes, show GFAP
staining as well. Herpers and Budka (1984)
examined GFAP expression by immunohistochemistry in 50 oligodendrogliomas and in 16
mixed oligoastrocytomas, and reported GFAP
expression in 50% of oligodendrogliomas, most
evident in gliofibrillary oligodendroglioma
cells. Kros et al. (1990) analyzed GFAP staining
in 111 oligodendrogliomas and correlated the
findings to tumor cell morphology and to patient
survival. They reported GFAP staining in 68% of
newly diagnosed and in 86% of recurrent
oligodendrogliomas. There was no significant
correlation noted between tumor grade and GFAP
staining. Thus, GFAP is not a reliable marker
for distinguishing oligodendrogliomas from
astrocytomas.
132
The morphologic differential diagnosis of

oligodendroglioma includes neurocytoma, dysembryoplastic neuroepithelial tumor (DNT), clear
cell ependymoma, clear cell meningioma, pilocytic astrocytoma with area showing oligodendroglial-like differentiation, and glioblastoma
with oligodendroglial component. The oligodendrocyte lineage-specific Olig1 and Olig2 genes
that encode basic helix-loop-helix transcription
factors appeared to be restricted to the oligodendrocyte lineage during development as well as in
adults. However, these have now been identified
in other gliomas. Ohnishi et al. (2003) demonstrated the presence of high levels of Olig1
mRNA in oligoastrocytoma, while high levels
of Olig2 mRNAs were identified in anaplastic
astrocytoma. On immunostaining, however, a
difference in the protein expression was identified. Olig2 was the strongest and most diffuse in
anaplastic oligodendroglioma. Astrocytomas,
both low grade and high grade, had a lesser
proportion of cells with positive staining. Takei
et al. (2008) also found Olig1 immunoreactivity
in a majority of pilocytic astrocytomas with oligodendroglia-like cells, while Olig2 immunoreactivity was lesser in extent and intensity. Most
recently, Capper et al. (2010b) demonstrated
positive immunostaining for mutant IDH1R132H
protein in over 92% of oligodendrogliomas
(n = 143) and 84% of oligoastrocytomas among
25 cases. In contrast, none of the 106 other neoplasms that enter the morphological differential
diagnoses of oligodendrogliomas showed any
staining for mutant-specific IDH1R132H, including pediatric oligodendrogliomas, central and
extraventricular neurocytomas, DNTs, clear cell
ependymomas, clear cell meningiomas, primary
glioblastomas with oligodendroglial component
and pilocytic astrocytomas with oligodendrogliallike differentiation. Based on these findings, positive
expression of mutation-specific IDH1R132H
immunostain appears to help identify oligodendrogliomas or oligoastrocytomas, while a negative immunostain should prompt the consideration
of other clear cell neoplasms.
The immunohistochemical stains may also
help distinguish other neoplasms with oligodendroglioma-like features. The epithelial membrane
antigen (EMA) positivity within tumor cells and
A. Raghunathan
along luminal surfaces, combined with GFAP positivity in perivascular cytoplasmic processes helps
identify clear cell ependymoma. Neurocytomas
are negative for GFAP and have diffuse positivity
for neuronal markers, including synaptophysin,
NeuN, neurofilament and microtubule associated
protein - 2 (MAP-2). In contrast, while oligodendrogliomas are generally negative for neuronal
markers, they may focally have synaptophysin
expression in addition to the staining of background neuropil. Clear cell meningiomas are
extra-axial tumors that show immunoreactivity
for EMA, and are negative for GFAP and neuronal
markers.
Fetal ependymal cells are strongly positive for
GFAP, while mature and neoplastic ependymal
cells may show only weak, focal or absent GFAP
expression. In ependymomas, cytoplasmic processes that radiate towards blood vessels show the
strongest and most consistent expression of GFAP.
Ependymomas may also be immunoreactive for
S100, while they are negative for other neuronal
markers. The EMA immunostain may show dotlike staining within the cytoplasm, corresponding
to intracytoplasmic microlumina, as well as ringlike or linear staining of the luminal surfaces. It is
important to emphasize, however, that the
EMA staining may be weak or even very focal,
necessitating a detailed examination of the
immunostained section. The use of additional
testing, including electron microscopy, may be
required to help establish a definitive diagnosis of
ependymoma.
Gliosarcoma is a glioblastoma variant that
shows a biphasic tissue pattern, with alternating
areas displaying glial and mesenchymal differentiation, and is designated World Health
Organization (WHO) grade IV. By definition, the
mesenchymal component shows histologic features of malignancy, including cytologic anaplasia,
brisk and atypical mitoses, and necrosis. A biopsy
that predominantly contains the mesenchymal
component would raise the differential diagnosis
of metastatic sarcomas. The glial and non-glial
components are distinguished by immunohistochemical staining for GFAP and vimentin. The
glial component is highlighted by GFAP, while
the sarcomatous component is negative for GFAP
and strongly positive for vimentin. Histochemical
13
staining for reticulin or immunostaining for collagen type IV help highlight the deposition of
collagen fibers around individual malignant
spindle cells in the sarcomatous areas. The
demonstration of a malignant glial component
that is GFAP positive is essential in excluding
sarcoma involving the CNS.
Pleomorphic xanthoastrocytoma (PXA) is a
WHO grade I glioma that shows abundant pleomorphic astrocytes and focal to abundant lipidized cells. The marked pleomorphism seen in
PXA may raise the differential diagnosis of a
giant cell glioblastoma, a WHO grade IV tumor,
particularly in limited biopsy samples where
necrosis and microvascular proliferation may not
have been sampled. The Ki-67/MIB-1 labeling
index in PXA is generally lower than 1%. These
tumors have no known association with TP53
mutations, and are negative for p53 immunostaining.
Giant cell glioblastoma is a histological variant of
glioblastoma that has a predominance of bizarre,
multinucleated giant cells. In addition to necrosis
and microvascular proliferation, these glioblastoma variants show frequent mitoses, Ki-67/MIB-1
labeling indices of about 1520% on average have
a high frequency of TP53 mutations and high
percentage of p53 immunostain positive nuclei.
PXA may show expression of neuronal markers
including synaptophysin, neurofilament, class III
-tubulin and MAP2, while giant cell glioblastoma
is usually negative for these markers. In addition,
CD34, the human hematopoietic progenitor cell
antigen, is frequently expressed in PXA.
Spontaneous mutations of IDH1 have been
detected with high frequency in WHO grades II
and III astrocytoma, oligoastrocytoma and oligodendroglioma, in cases of gliomatosis cerebri
that also exhibit a solid tumor mass, and in
secondary glioblastoma. In contrast, these mutations are rare in primary glioblastoma and in
other discrete tumors, and are absent in ependymomas and in classical gliomatosis cerebri that
lacks a solid tumor component. Hence, the immunostain for mutant IDH1R132H may help in distinguish diffuse glioma from solid/discrete
glioma types. As a note of caution, Antonelli
et al. (2010) reported pediatric high-grade
gliomas to lack IDH1 mutations, suggesting differences in the biology of pediatric and adult
133
gliomas, despite morphologic similarity.

Therefore, immunostain for mutant IDH1R132H
may be of limited value in assessing pediatric
gliomas.
Apart from its role in glioma biology, there is
some indication that IDH1 mutation and expression status may prove to be important in determining prognosis. Diffuse glioma patients with
IDH1 mutations have been reported to show longer overall survival compared to those with wild
type IDH1. In a study of 382 patients with anaplastic astrocytoma and glioblastoma, Hartmann
et al. (2010) reported glioblastomas with IDH1
mutation to have better outcomes than anaplastic
astrocytoma without IDH1 mutation, and that the
prognostic significance of patient age was possibly due to the IDH1 mutations occurring in
younger patients. The immunostain for mutant
IDH1R132H protein yielded similar results. On
the other hand, Kim et al. (2010) found that while
IDH1 mutations occur at a very early stage of
tumor genesis and with a high frequency among
low-grade diffuse gliomas, there was a significant
association with older age. Further, they found
the IDH1 status was not prognostic for improved
survival among these patients by itself. At present, the effect of IDH1 mutations, and mutant
IDH1 immunostain status, on the prognosis of
diffuse gliomas appears to require further clarification before their role as prognostic indicators
can be firmly established.
in Glioneuronal Neoplasms
Glioneuronal tumors show an admixture of glial
and neuronal components, with both cell types
being considered part of the same neoplastic process. The glial component in glioneuronal tumors
may resemble pilocytic astrocytoma or an infiltrating glioma with astrocytic or oligodendroglial
features, while the neuronal component may be
composed of mature ganglion cells, small but
mature neurons, or dysmature/immature neurons.
Glioneuronal tumors, which behave as low-grade
lesions and are potentially curable by surgery,
must be distinguished from entrapment of preexisting neurons by infiltrating gliomas that do not
134
have a neuronal component. In a majority of

glioneuronal tumors, the glial component is the
main determinant of clinical outcome.
Immunohistochemical stain for GFAP may help
better evaluate the glial component of these
tumors, and nuclear labeling for the proliferating
cell nuclear antigen Ki-67 is generally observed
exclusively in this astrocytic component.
Immunohistochemical staining for neuronal
markers, including synaptophysin, neurofilament
and MAP-2, may help demonstrate the infiltrating glial cells by staining pre-existing axons,
particularly in white matter, and the pre-existing
synapses in the gray matter, in addition to demonstrating the neuronal component.
At the other end of the spectrum, glioneuronal
neoplasms need to be separated from malformations
such as cortical dysplasia that may consist of both
neuronal and glial components. CD34 is a stem
cell marker that is transiently expressed during
early neuronal development. In an immunohistochemical evaluation of CD34 in 47 glioneuronal lesions obtained from patients with
intractable epilepsy, Deb et al. (2006) reported
CD34 staining in dysplastic and atypical neuronal
precursors that were not identified on routine
H&E staining. Gangliogliomas were diffusely
positive for CD34, while focal cortical dysplasia
demonstrated only single cells or small groups
that were immunopositive, and cases of DNTs
were largely negative. However, subsequent studies have demonstrated CD34 expression within a
subpopulation of balloon cells seen in severe cortical dysplasia. Hence, differentiation of severe
cortical dysplasia from a low grade glioneuronal
tumor remains a diagnostic challenge that is not
readily resolved by immunohistochemistry.
Differentiation of Atypical Teratoid/

Rhabdoid Tumors and Primitive
Neuroectodermal Tumors
The atypical teratoid/rhabdoid tumor (AT/RT) is
an important morphological diagnostic consideration for poorly differentiated CNS neuroepithelial
tumors, and may be morphologically indistinguishable from medulloblastoma, pineoblastoma,
A. Raghunathan
and supratentorial primitive neuroectodermal

tumor. Only about 25% of AT/RTs show a
predominance of rhabdoid cells, and these are
obscured by vastly predominant primitive neuroectodermal tumor (PNET)like population in a
significant subset of cases. The diagnosis of AT/
RT also has therapeutic implications, since these
tumors are unresponsive to standard therapeutic
regimens that are effective against PNETs, and
some cases appear to be better controlled by
intensified adjuvant treatment strategies.
The identification of inactivating abnormalities of hSNF5/INI1/SMARCB1/BAF47, a tumor
suppressor gene on chromosome 22q11.2 in AT/
RTs by Biegel et al. (1999) has helped in separating these from the other morphologic differential
diagnoses. Commonly abbreviated to INI1 or
hSNF5/INI1, this gene encodes a ubiquitous
protein that participates in the SWItch/Sucrose
Non-Fermentable ATP-dependent chromatin
remodeling complex. This inactivation follows a
two-hit mechanism for disabling tumor suppressing genes, wherein partial deletion leads to the
loss of one copy of the gene and a nonsense or
frame-shift mutation of the second copy produces
a novel stop codon. Germ line mutations of this
gene underlie a familial rhabdoid tumor syndrome that is characterized by potentially multifocal neural and extraneural primary rhabdoid
tumors presenting in the first year of life.
Many cases that appear to have intact INI1 on
gene sequencing show decreased gene expression
on reverse transcriptasePCR analysis, or undetectable INI1 protein by Western blot. Therefore,
an antibody to the INI1 protein is a more sensitive adjunct to diagnose AT/RT than genetic
assays. Complete absence of INI1 immunostaining in tumor cell nuclei has been consistently
observed in bona fide AT/RTs, as well as renal
and extrarenal rhabdoid tumors. It must be
emphasized that complete absence of INI1
nuclear immunoreactivity is required for confirming the diagnosis of AT/RT. Regional labeling loss may be encountered in non-AT/RT
tumors and may occasionally be very extensive.
The INI1 target protein is also localized in the
nuclei of normal cell types, such as neurons, glia,
endothelial cells and lymphoid cells. In the
13
absence of appropriate reactivity of normal

endothelial and lymphoid nuclei in the tumor
tissue, the results of the immunostain must be
considered non-contributory.
INI1 in Other Tumor Types:

Choroid Plexus Carcinoma
Choroid plexus tumors, especially choroid plexus
carcinoma, have been described as occasionally
exhibiting INI1 mutation. However, AT/RTs may
present as intra-ventricular tumors and may
exhibit differentiation along epithelial lines.
Further, some choroid plexus carcinomas may
demonstrate anaplastic solid areas, including
rhabdoid-appearing cells. Some cases diagnosed
as choroid plexus carcinomas have also been
shown to exhibit an AT/RT-like immunophenotype and may manifest chromosome 22 abnormalities. Analyzing 28 tumors diagnosed as
choroid plexus carcinoma, Judkins et al. (2005)
found INI1 immunoexpression in 20 cases that
were, nevertheless, morphologically diagnosed
as choroid plexus carcinomas on independent
review by expert neuropathologists. Seven of
these cases were subjected to mutation analysis,
and all showed normal INI1 gene profiles. The
diagnosis of six cases that were previously diagnosed as choroid plexus carcinomas and showed
complete loss of INI1 nuclear immunostaining
were revised to AT/RT on expert review, emphasizing the overlap of AT/RTs with choroid plexus
carcinomas that have AT/RT-like genetic features.
INI1 in Other Tumor Types: Multiple

Schwannoma Syndromes
In recent reports, tumors in patients with both
familial and sporadic schwannomatosis were
found to harbor mutations in the SMARCB1 gene.
In a study of 45 schwannomas from patients with
multiple schwannoma syndromes and on 38 solitary, sporadic schwannomas from non-syndromic
patients, Patil et al. (2008) reported a mosaic
pattern of INI1 immunostaining with a mixture
of positive and negative nuclei in 93% of tumors
135
from familial schwannomatosis patients, 55% of

tumors from sporadic schwannomatosis, 83% of
NF2-associated tumors, while cases of solitary,
sporadic schwannomas predominantly showed
diffuse positive nuclear staining for INI1 immunostain. These results suggest a role for the INI1
gene in multiple schwannoma syndromes, and
that a different pathway of tumorigenesis may be
present in solitary, sporadic schwannoma.
Identifying Medulloblastoma
Subtypes by Immunohistochemistry
Medulloblastoma is a common malignant pediatric brain tumor. Although overall survival rates
have improved in recent years, the mortality rate
remains significant. Completely resected tumors
from patients older than 3 years of age with no
leptomeningeal dissemination at diagnosis are
classified as standard risk, whereas all others are
considered high risk. In an analysis of gene
expression profiles and DNA copy number aberrations for 103 primary medulloblastomas,
Northcott et al. (2011) identified the following
four distinct, non-overlapping molecular variants
using multiple unsupervised bioinformaticsbased analyses of transcriptional profiles:
medulloblastomas with alterations in the Sonic
Hedgehog (SHH) pathway, tumors with alterations in the WNT pathway, a Group C, and a
Group D. Analysis of these four subgroups
revealed distinct demographics, clinical presentation, histology, transcriptional profiles, genetic
abnormalities, and clinical outcomes.
Tumors with alterations in the SHH pathway
occur primarily in infants and adults, while Group
C tumors were largely confined to childhood.
Tumors with alterations in the WNT pathway
were almost three times more common in females
than males (17% vs. 6%). Metastases were significantly over-represented in Group C (46.5%)
followed by Group D (29.7%). MYC amplification analysis by interphase FISH was positive in
11 of 98 Group C tumors, but was absent in all
SHH, WNT and group D tumors. Large cell/
anaplastic histology were found much more
commonly among Group C tumors (23%), while
136
WNT tumors were most commonly classic

medulloblastomas.
Although
desmoplastic
tumors were most commonly seen in the SHH
group, they were also found in Group C and
Group D. On genetic testing, while MYC was
highly expressed in Group C and WNT tumors,
and MYCN was highly expressed in SHH tumors,
neither MYC nor MYCN were highly expressed
in Group D tumors.
Immunohistochemical analysis was performed
for subgroup-specific signature genes, that were
DKK1 for the WNT group, SFRP1 for the SHH
group, NPR3 for Group C and KCNA1 for Group
D, was performed on 294 medulloblastomas on
two independent tissue microarrays. The immunohistochemical analysis demonstrated robust
staining for a single subgroup-specific marker
in 288 (approximately 98%) of the 294
medulloblastomas. This four-antibody immunohistochemistry approach to subgroup medulloblastoma appears to offer reproducible
identification of the four demographically, clinically, and genetically distinct medulloblastoma
variants.
in the Diagnosis of Intracranial
Germ Cell Tumors
The panel of immunohistochemical stains for
germ cell neoplasia includes placental alkaline
phosphatase, alpha-fetoprotein, human chorionic
gonadotropin, OCT 3/4, CD30 and CD117 (c-kit).
Placental alkaline phosphatase (PLAP) is a
membrane-bound isoenzyme that is produced by
placental syncytiotrophoblasts and many neoplasms.
Among germ cell tumors, PLAP is expressed in
nearly all germinomas and embryonal carcinomas, most yolk sac tumors, and variably in choriocarcinomas. PLAP is also expressed in
non-germ cell carcinomas including serous carcinomas, and is not specific for germ cell tumors.
Alpha- Fetoprotein is an oncofetal glycoprotein
that is positive in a majority of yolk sac tumors,
although staining may be patchy. Pure germinomas
are negative for alpha-fetoprotein, and its positivity
A. Raghunathan
in a mixed germ cell tumor identifies areas with

yolk sac differentiation. Human chorionic
gonadotropin (HCG) is a glycoprotein composed
of a- and b- subunits. The b- subunit is produced
by syncytiotrophoblasts. HCG is, therefore, a
marker for choriocarcinoma and for identifying
syncytiotrophoblastic cells in germinomas, embryonal carcinomas and yolk sac tumors.
OCT4 is relatively recent addition to the
immunohistochemical battery to evaluate germ
cell neoplasms. Also known as OCT3, OCT3/4,
OTF3 and POU5F1, this transcription factor is
normally expressed in embryonic stem cells and
primordial germ cells. It helps in maintaining
pleuripotency and is required for survival of primordial germ cells. The OCT4 immunohistochemical stain has been shown to have
consistent nuclear expression in classic gonadal
seminomas, dysgerminomas and embryonal carcinomas, as well as the germ cell component of
gonadoblastomas. In contrast, gonadal cases of
spermatocytic seminomas, yolk sac tumors, choriocarcinomas, and teratomas are negative for
OCT4 immunostain.
Hattab et al. (2005) compared OCT4 and
PLAP immunohistochemical antibodies in a series
of 25 intracranial germinomas. Immunostaining
for PLAP was present in 23 of 25 germinomas. In
some instances, however, nonspecific background
PLAP staining interfered with identification of
the membranous and cytoplasmic staining in
tumor cells. All 25 germinomas showed diffuse,
moderate to strong nuclear immunostaining for
OCT4, supporting it to be a sensitive and specific
marker for germinomas. Another differential
diagnosis of CNS germ cell neoplasms includes
AT/RT. In a review of OCT4 immunostaining in
intracranial germ cells, Edgar and Rosenblum
(2008) mentioned assessing several examples of
AT/RT for OCT4 expression, and that none demonstrated nuclear immunostaining. Embryonal
carcinomas have also been shown to express
OCT4, and can be distinguished from germinomas by the cytoplasmic staining for CD30 and
the intense and diffuse expression of various
cytokeratins, that is not found in germinomas.
Further, germinomas also express CD117 (c-kit)
13
in a membranous and, often, golgi pattern that is

not seen in embryonal carcinoma.
Differentiating Meningiomas,
Hemangiopericytomas and Solitary
Fibrous Tumors
Meningiomas account for about 30% of primary
intracranial tumours in the USA. They may
occur at any age, although they are most common among the middle-aged and elderly, while
childhood examples tend to be aggressive. A
majority of meningiomas show immunohistochemical staining for EMA, although immunoreactivity is less consistent in atypical and
anaplastic meningiomas. Hemangiopericytoma
(HPC) and solitary fibrous tumor (SFT) are dural
based mesenchymal tumors that are in the differential diagnosis of meningioma. Meningeal hemangiopericytomas constitute 2.5% of all meningeal
tumors and 1% of all intracranial tumors. They
tend to recur even after macroscopic total resection, with local recurrence rates as high as 91%.
These have been shown to arise from Zimmermann
pericytes, which are contractile spindle cells surrounding capillaries and postcapillary venules
throughout the human body. Solitary fibrous
tumors were first described in the pleura and have
subsequently been observed in a number of extrapleural sites, including soft tissues and intracranial meninges. Typical histologic features of HPC
and SFT include spindle cells arranged in a patternless architecture, the so-called cytologic turbulence, and branching vascular channels with
thin walls. Solitary fibrous tumors show prominent collagen bands that are not a feature of HPC,
and are benign tumors with good prognosis after
surgical resection, with rare recurrence or metastasis. In contrast, HPCs have a greater predilection for local recurrence and metastasis.
Soft tissue SFT and HPC are considered as
part of the spectrum of one entity, although
intracranial cases continue to be diagnosed as
separate tumors. Immunohistochemical stains
have been applied in an attempt to differentiate
intracranial SFT and HPC. In a comparison of 31
137
meningeal HPCs and 12 soft tissue HPC/SFTs,

Ambrosini-Spaltro and Eusebi (2010) found
intracranial tumors to be more cellular than HPC/
SFT of soft tissues, and had fewer collagen
bands. Meningeal HPC in addition had more
mitoses, higher Ki67 index, stained less intensely
for CD34 and BCL2 than HPCSFT of soft tissues. The CD34 immunostain showed patchy to
diffuse positivity in 24 specimens and BCL2
positivity in 13 specimens of the 31 HPCs. On
the other hand, CD34 showed patchy to diffuse
staining in soft tissue SFT/HPCs, while BCL2
was positive in 10 of 12 cases. In other words,
CD34 and BCL2 cannot be used to reliably differentiate HPC from SFT, and higher cellularity
and a lower extent of pericellular reticulin in
HPC may be more helpful in distinguishing these
two entities.
Diagnosis of Primary
CNS Lymphomas
The diagnosis of CNS B- and T- cell lymphomas
is supported by the expression of CD45 (the leukocyte common antigen) and with markers of Bor T-lymphocytic differentiation. Putative
evidence of involvement of the CNS by lymphoma is based on the demonstration of an
abnormal predominance of B-lymphocytes, or
by either loss or aberrant expression of
T-lymphocyte markers. With the use of the antiCD20 drug Rituximab in lymphoma chemotherapy, there may be decreased expression or loss of
surface CD20 expression in post-treatment samples. With this in mind, a second pan- B- lymphocyte marker, such as CD79a, would help
identify the B- lymphocyte population in B- cell
lymphomas that are negative for CD20. The
presence of CNS lymphoma is confirmed by
flow cytometric evidence of an abnormal clonal
lymphocyte population, and by gene rearrangement studies that demonstrate expansion of a
single clone with restriction in the normal diversity of immunoglobulin heavy chain (for B-cell
lymphomas) or the T-cell receptor (in T-cell
lymphomas).
138
Differentiating Hemangioblastoma
from Metastatic Renal Clear Cell
Carcinoma
Hemangioblastoma is a WHO grade I, capillaryrich neoplasm that occurs either sporadically or
in the setting of von Hippel-Lindau (VHL) syndrome, an autosomal dominant disorder caused
by germline mutations of the VHL tumor suppressor gene. The histologic differential diagnosis of hemangioblastoma includes metastatic
clear cell carcinoma (RCC) and microcystic/
angiomatous meningiomas. This differentiation
is particularly important given the drastically different prognostic and therapeutic significance of
these tumors. This distinction is challenging due
to the morphologic similarities of their clear
cell components. Metastatic RCC may also
occur synchronously or metachronously with
hemangioblastoma in patients with VHL disease,
in whom hemangioblastoma and RCC are among
the most common tumors.
Immunohistochemistry is a crucial adjunct in
helping distinguish these two tumors. In general, clear cell renal cell carcinoma is variably
immunoreactive for CD10 and for epithelial
markers such as EMA and cytokeratins, while
hemangioblastoma is generally negative.
However, EMA has been reported in up to 36%
of hemangioblastoma cases, as membranous
staining of stromal cells. Microcystic/angiomatous meningiomas also express EMA, and cannot be distinguished from hemangioblastomas
on the basis of this alone.
Inhibin is a dimeric 32-kd peptide hormone
composed of two subunits, A and B, linked by
disulfide bridges, and is produced by ovarian
granulosa cells and testicular sertoli cells.
Inhibin-A subunit (inhibin A) immunoreactivity was first described in hemangioblastoma by
Hoang and Amirkhan (2003), who reported its
expression in the stromal cells of all 25 cases of
hemangioblastoma, while all cases of renal cell
carcinoma studied were negative. Subsequent
studies have reported lower sensitivity and
specificity of inhibin A for differentiating these
two tumors.
A. Raghunathan
Aquaporin is a family of integral cell membrane

proteins that function as water transport channels. In the normal brain, the aquaporin 1 subtype
has been selectively demonstrated on the surface
of choroid plexus epithelial cells, suggesting a
possible role in cerebrospinal fluid secretion.
Aquaporin 1 was found to be expressed on the
cytoplasmic membranes of stromal cells in
hemangioblastoma, with a higher level of expression
in cystic than in solid hemangioblastoma, suggesting its association with hemangioblastoma
cyst formation. In an analysis of 67 hemangioblastomas and 34 metastatic clear cell RCC,
Weinbreck et al. (2008) found aquaporin 1 to
have a higher sensitivity and specificity than
inhibin-A (97% vs. 88% sensitivity, and 83% vs.
79% specificity) for hemangioblastoma. Clear
cell RCC showed positivity with cytokeratin
AE1/AE3 (100% specificity, 88% sensitivity)
and CD10 (100% specificity, 79% sensitivity). In
their analysis, the combination of aquaporin1 and
AE1/AE3 yielded the highest degree of sensitivity and specificity for differentiating hemangioblastoma and metastatic clear cell RCC. Nearly
all hemangioblastomas (65 of 66) were positive
for aquaporin1 and negative for cytokeratin AE1/
AE3, while all tumors with the opposite profile
(25 of 25), aquaporin1 negative and cytokeratin
AE1/AE3 positive, corresponded to metastatic
clear cell RCC.
Brachyury and the Diagnosis

of Chordoma
Chordoma is a rare, slowly growing, locally
aggressive malignant tumor that is believed to be
derived from vestigial notochordal remnants, and
can occur anywhere along the central axial skeleton. The histologic differential diagnosis includes
myxoid chondrosarcoma, chordoid meningioma,
liposarcoma, metastatic renal cell carcinoma, metastatic mucinous adenocarcinoma, metastatic salivary gland carcinoma and metastatic malignant
melanoma. On immunohistochemistry, chordomas express cytokeratin and EMA and a significant proportion also demonstrate S-100 protein
reactivity. The coexpression of immunostains for
13
epithelial markers and S-100 protein is very helpful

in diagnosing chordoma, when present. However,
cases with sarcomatous transformation usually
lack immunoreactivity for epithelial markers in
the sarcomatoid component.
Brachyury is a transcription factor protein that
is important for mesodermal differentiation and
notochord development in early embryogenesis.
Brachyury immunoreactivity has been reported
in spermatogonia and testicular germ cell tumors,
as well as in hemangioblastoma. In a study by
Vujovic et al. (2006), the brachyury immunostain
had 100% sensitivity and 100% specificity in
identifying all 53 cases of chordomas analyzed,
labeling the neuclei of neoplastic cells in the
chordoid and chondroid areas. Nuclear immunoreactivity was reported in virtually all chordoma
cells in decalcified and non-decalcified material.
In cases of dedifferentiated chordoma, while
brachyury was negative in the undifferentiated
spindle cell component, there was diffuse and
strong immunostaining in areas of more conventional chordoma. In contrast, none of the nonchondroma cases studied was positive for the
brachyury immunostain. In a study of 103 chondroid neoplasms from the head and neck region
and the skull base, including 79 chordomas and
24 chondrosarcomas, Oakley et al. (2008)
reported 89.5% sensitivity and 100% specificity
for brachyury in identifying skull-based chordomas. The combination of cytokeratin and
brachyury improved the sensitivity to 98% and
the specificity to 100% for the detection of
chordoma.
Immunohistochemistry in Carcinoma,
Melanoma and Sarcoma
Cytokeratin (CK) is an intermediate filament
found in epithelial tissues. Immunostains that are
useful in confirming epithelial differentiation
include low-molecular-weight cytokeratins recognized by CAM 5.2 and broad spectrum cytokeratins recognized by AE1/AE3 antibody. Being
markers of epithelial differentiation, their expression would support the diagnosis of metastatic
carcinoma rather than glioblastoma in the setting
139
of a poorly differentiated CNS malignancy.

However, gliomas may also show immunoreactivity to cytokeratins, especially AE1/3 which is a
commonly used antibody preparation that recognizes numerous cytokeratin types. Cosgrove et al.
(1989) identified AE1/3 immunostain in 24 of 30
(30%) of low to high grade astrocytomas. Oh and
Prayson (1999) found 95.7% of glioblastoma
samples showed keratin expression using AE1/3,
while only 4.3% showed expression of CAM5.2,
CK7 and CK20. On the other hand, 3 of 22 metastatic carcinomas showed focal GFAP staining.
Therefore, epithelial differentiation in metastatic
carcinoma is best documented using cytokeratin
antibodies that are not positive in astrocytomas,
such a CAM 5.2, rather than the more widely
reactive AE1/3. The reader is referred to several
excellent reviews, including by Krishna (2010),
that discuss the immunostaining patterns of cytokeratin subtypes in various organs to help determine the origin of metastatic carcinoma.
S-100 is a polyclonal antibody raised against a
relatively non-specific neuroendocrine marker
(S-100 protein), and is the most sensitive marker
of melanocytes. Both nuclear and cytoplasmic
S100 expression should be present for the staining to be interpreted as positive. However, nearly
all malignant gliomas also express S-100 protein.
In fact, the S100 protein was first isolated from
the CNS, making this a poor marker to distinguish gliomas from melanocytic neoplasms.
Further, S100 can be seen in many mesenchymal
and epithelial tumors as well. To confirm the
presence of melanoma, S100 should be combined
with other markers that have higher specificity,
such as HMB-45, Melan-A/MART-1, tyrosinase
or microphthalmia transcription factor protein.
HMB45 is a monoclonal antibody against the
group 100 protein of premelanosomes, and is a
more specific marker than S100 though it lacks
sensitivity. The melan-A is melanocytic differentiation antigen that is recognized by cytotoxic
T- cells. Tyrosinase is an enzyme involved in the
initial stages of melanin biosynthesis in melanocytes and melanoma cells. Clarkson et al.
(2001) compared the expression of S100, HMB45, Melan-A and tyrosinase immunostains in
50 benign and malignant melanocytic lesions.
140
They reported S100 to be the most sensitive

melanocytic marker, while HMB-45 was the least
sensitive. Desmoplastic/spindle cell melanomas
showed patchy to negative staining with all four
of these stains.
The microphthalmia transcription factor
(MITF) gene encodes a basic helix-loop-helixleucine zipper transcription factor that is involved
in the development of the melanocyte lineage,
being involved in the embryonic development
and postnatal viability of melanocytes, and helping in the transcriptional regulation of tyrosinase
and tyrosinase-related proteins 1 and 2. Xu et al.
(2002) demonstrated MITF to be of use in helping
identify HMB-45 negative spindle cell melanomas. The use of MITF may, therefore, be considered in spindle cell neoplasms whose lineage
cannot be identified despite proven tissue viability
for immunostaining.
There is no reliable positive screening marker
for confirming mesenchymal differentiation. The
best approach to sarcomas is to use a short panel
of immunohistochemical stains to first define
mesenchymal lineage while excluding other lineages, such as GFAP, vimentin, CAM 5.2, S100
and CD45. While vimentin has historically been
used indicate mesenchymal lineage, it is coexpressed with keratins in some carcinomas and
is also expressed in melanomas. Reactivity with
vimentin also helps assess the immunoreactivity
of the tissue, and place negative immunostain
results in context. Therefore, it is important to
identify the lineage of origin and exclude others
by using an immunohistochemical panel that
includes with vimentin. Subsequent subtyping of
the tumor may be performed by immunohistochemistry and, if required, cytogenetic, molecular and electron microscopic analysis.
References
Ambrosini-Spaltro A, Eusebi V (2010) Meningeal hemangiopericytomas and hemangiopericytoma/solitary
fibrous tumors of extracranial soft tissues: a comparison. Virchows Arch 456:343354
Antonelli M, Buttarelli FR, Arcella A, Nobusawa S,
Donofrio V, Oghaki H, Giangaspero F (2010)
Prognostic significance of histological grading, p53
A. Raghunathan
status, YKL-40 expression, and IDH1 mutations in
pediatric high-grade gliomas. J Neurooncol
99:209215
tumors. Cancer Res 59:7479
Capper D, Zentgraf H, Balss J, Hartmann C, von Deimling
A (2009) Monoclonal antibody specific for IDH1
R132H mutation. Acta Neuropathol 118:599601
Capper D, Sahm F, Hartmann C, Meyermann R, von
Deimling A, Schittenhelm J (2010a) Application of
mutant IDH1 antibody to differentiate diffuse glioma
from nonneoplastic central nervous system lesions and
therapy-induced changes. Am J Surg Pathol
34:11991204
Capper D, Reuss D, Schittenhelm J, Hartmann C, Bremer
J, Sahm F, Harter PN, Jeibmann A, von Deimling A
(2010b) Mutation-specific IDH1 antibody differentiates oligodendrogliomas and oligoastrocytomas from
other brain tumors with oligodendroglioma-like
morphology. Acta Neuropathol 121:241252
Clarkson KS, Sturdgess IC, Molyneux AJ (2001) The usefulness of tyrosinase in the immunohistochemical
assessment of melanocytic lesions: a comparison of
the novel T311 antibody (anti-tyrosinase) with S-100,
HMB45, and A103 (anti-melan-A). J Clin Pathol
54:196200
Cosgrove M, Fitzgibbons PL, Sherrod A, Chandrasoma
PT, Martin SE (1989) Intermediate filament expression in astrocytic neoplasms. Am J Surg Pathol 13:
141145
Deb P, Sharma MC, Tripathi M, Sarat CP, Gupta A, Sarkar
C (2006) Expression of CD34 as a novel marker
for glioneuronal lesions associated with chronic
intractable epilepsy. Neuropathol Appl Neurobiol
32:461468
Hartmann C, Hentschel B, Wick W, Capper D, Felsberg J,
Simon M, Westphal M, Schackert G, Meyermann R,
Pietsch T, Reifenberger G, Weller M, Loeffler M, von
Deimling A (2010) Patients with IDH1 wild type anaplastic astrocytomas exhibit worse prognosis than
IDH1-mutated glioblastomas, and IDH1 mutation
status accounts for the unfavorable prognostic effect
of higher age: implications for classification of
gliomas. Acta Neuropathol 120:707718
Hattab EM, Tu PH, Wilson JD, Cheng L (2005) OCT4
immunohistochemistry is superior to placental alkaline phosphatase (PLAP) in the diagnosis of central
nervous system germinoma. Am J Surg Pathol
29:368371
Herpers MJ, Budka H (1984) Glial fibrillary acidic protein (GFAP) in oligodendroglial tumors: gliofibrillary
oligodendroglioma and transitional oligoastrocytoma
as subtypes of oligodendroglioma. Acta Neuropathol
(Berl) 64:265272
13
Hoang MP, Amirkhan RH (2003) Inhibin alpha distinguishes hemangioblastoma from clear cell renal cell
carcinoma. Am J Surg Pathol 27:11521156
INI1 protein expression distinguishes atypical teratoid/
rhabdoid tumor from choroid plexus carcinoma.
Kim YH, Nobusawa S, Mittelbronn M, Paulus W,
Brokinkel B, Keyvani K, Sure U, Wrede K, Nakazato
Y, Tanaka Y, Vital A, Mariani L, Stawski R, Watanabe
T, De Girolami U, Kleihues P, Ohgaki H (2010)
Molecular classification of low-grade diffuse gliomas.
Am J Pathol 177:27082714
Krishna M (2010) Diagnosis of metastatic neoplasms: an
immunohistochemical approach. Arch Pathol Lab
Med 134:207215
Kros JM, Van Eden CG, Stefanko SZ, Waayer-Van BM,
van der Kwast TH (1990) Prognostic implications of
glial fibrillary acidic protein containing cell types in
oligodendrogliomas. Cancer 66:12041212
Kurtkaya-Yapcer O, Scheithauer BW, Hebrink D, James
CD (2002) P53 in nonneoplastic central nervous system lesions: an immunohistochemical and genetic
sequencing study. Neurosurgery 51:12461255
Meyer J, Pusch S, Balss J, Capper D, Mueller W, Christians
A, Hartmann C, von Deimling A (2010) PCR- and
restriction endonuclease-based detection of IDH1
mutations. Brain Pathol 20:298300
Northcott PA, Korshunov A, Witt H, Hielscher T, Eberhart
CG, Mack S, Bouffet E, Clifford SC, Hawkins CE,
French P, Rutka JT, Pfister S, Taylor MD (2011)
Medulloblastoma comprises four distinct molecular
variants. J Clin Oncol 29(11):14081414
Oakley GJ, Fuhrer K, Seethala RR (2008) Brachyury,
SOX-9, and podoplanin, new markers in the skull base
chordoma versus chondrosarcoma differential: a tissue
141
microarray-based comparative analysis. Mod Pathol

21:14611469
Oh D, Prayson RA (1999) Evaluation of epithelial and keratin markers in glioblastoma multiforme: an immunohistochemical study. Arch Pathol Lab Med 123:917920
Ohnishi A, Sawa H, Tsuda M, Sawamura Y, Itoh T,
Iwasaki Y, Nagashima K (2003) Expression of the oligodendroglial lineage-associated markers Olig1 and
Olig2 in different types of human gliomas. J
Neuropathol Exp Neurol 62:10521059
Patil S, Perry A, Maccollin M, Dong S, Betensky RA, Yeh
TH, Gutmann DH, Stemmer-Rachamimov AO (2008)
Immunohistochemical analysis supports a role for
INI1/SMARCB1 in hereditary forms of schwannomas, but not in solitary, sporadic schwannomas. Brain
Pathol 18:517519
Takei H, Yogeswaren ST, Wong KK, Mehta V,
Chintagumpala M, Dauser RC, Lau CC, Adesina AM
(2008) Expression of oligodendroglial differentiation
markers in pilocytic astrocytomas identifies two
clinical subsets and shows a significant correlation
with proliferation index and progression free survival.
Vujovic S, Henderson S, Presneau N, Odell E, Jacques TS,
Tirabosco R, Boshoff C, Flanagan AM (2006) Brachyury,
a crucial regulator of notochordal development, is a
novel biomarker for chordomas. J Pathol 209:157165
Weinbreck N, Marie B, Bressenot A, Montagne K, Joud
A, Baumann C, Klein O, Vignaud JM (2008)
Immunohistochemical markers to distinguish between
hemangioblastoma and metastatic clear cell renal cell
carcinoma in the brain: utility of aquaporin 1 combined with cytokeratin AE1/AE3 immunostaining.
Am J Surg Pathol 32:10511059
Xu X, Chu AY, Pasha TL, Elder DE, Zhang PJ (2002)
Immunoprofile of MITF, tyrosinase, melan-A, and
MAGE-1 in HMB45-negative melanomas. Am J Surg
Pathol 26:8287
Children with Brain Tumors:

Role of the Neurosurgeon
14
Peter F. Morgenstern and Mark M. Souweidane
Contents
Abstract
Introduction ............................................................
143
The Pediatric Neurosurgeon .................................
144
Hydrocephalus........................................................
External Ventricular Drain .......................................
Ventriculoperitoneal Shunt.......................................
Endoscopic Third Ventriculostomy..........................
144
145
145
145
Age of the Patient ...................................................
146
Cytoreductive Surgery ...........................................
146
Operative Techniques ............................................
146
Biopsy ......................................................................
146
Staging.....................................................................
147
Second Look Surgery .............................................
147
Tissue Procurement for Investigation ..................
147
Tumor-Specific Considerations .............................

Gliomas ...................................................................
Low-Grade Gliomas.................................................
High-Grade Gliomas ................................................
Medulloblastoma......................................................
Ependymoma ...........................................................
Craniopharyngioma..................................................
Pineal Region Tumors ..............................................
Brain Stem Tumors ..................................................
148
148
148
148
148
150
150
151
152
Future Directions ...................................................
152
References ...............................................................
153
P.F. Morgenstern (*) M.M. Souweidane

Departments of Neurological Surgery and Pediatrics,
Wil Cornell Medical College and Memorial SloanKettering Cancer Center, 525 East 68th Street,
New York, NY 10021, USA
e-mail: mmsouwei@med.cornell.edu
The surgical management of children with

brain tumors has evolved considerably over
the years and continues to be the first line of
therapy for most tumors of the central nervous
system. Outcomes are profoundly influenced
by the level of specialized care and expertise
available. Important elements of therapy
include management of hydrocephalus, diagnostic sampling and tumor removal, each of
which has been influenced by the rapid
advances in the fields of pediatric neurosurgery and neurooncology. Tumor type and
other factors dictate the appropriate course of
action for each patient. The role of the pediatric neurosurgeon continues to expand to
involve investigation of tumor development
and therapies, consideration of disease staging,
and drug delivery.
Introduction
Tumors of the central nervous system (CNS) are
the second most common malignancy and account
for 24% of cancer-related deaths in children
(Heuer et al. 2007). The Central Brain Tumor
Registry of the United States reports an annual
incidence of 4.5 cases per 100,000 person-years,
or about 3,750 cases per year. The complexity of
pediatric neurooncology today is clear when one
considers the many different types of brain tumors
afflicting children and the unique aspects of each.

143
144
However, one commonality is that outcomes are

optimized through the care of coordinated teams
of neurosurgeons, oncologists and other specialists with expertise in brain tumor therapy.
There are major differences between adult and
pediatric brain tumors beyond simple nomenclature. Perhaps most important is the potential for
cure in children. Pediatric tumors are often more
sensitive to adjuvant irradiation and chemotherapy (Merchant et al. 2010) and some tumors may
only need complete resection to achieve a cure. It is
for these reasons that initial management by an
experienced pediatric neurosurgeon can make a
tremendous difference in the patients clinical
outcome.
The Pediatric Neurosurgeon

Sub-specialization has become more common in
neurosurgery, but the number of pediatric neurosurgeons remains small. There are about 0.25
neurosurgeons who specialize in caring for children for every 100,000 children in the United
States. Thus, most children presenting in emergency departments around the country will initially
be treated by neurosurgeons who are not pediatric specialists. Evidence suggests that children
with common brain tumors have better surgical
outcomes in the hands of specialized neurosurgeons (Albright et al. 2000).
One indirect measure of the quality of care
provided at a specific center for a given disease is
the number of patients treated there, the logic
being that centers with more experience can provide better and more up-to-date care. It is estimated that only 37.4% of children with brain
tumors are seen at high volume centers, and there
is ample evidence to suggest serious racial and
ethnic disparities in access to high quality neurosurgical care for these patients. Hispanic children
in the United States, for example, are 32% less
likely to be admitted to high-volume hospitals.
Furthermore, patients from areas with lower
median home values have been shown to have less
access to high-volume centers and there is a direct
link between geographic neurosurgeon density
and the likelihood that a pediatric patient will
P.F. Morgenstern and M.M. Souweidane
receive neurooncological care at a high-volume

center (Mukherjee et al. 2009). As a result, the
majority of children will be initially evaluated and
treated by neurosurgeons not versed in the clinical
and technical aspects of these unique tumors, a
fact that could translate into less than optimal outcome and quality of life.
The pediatrician or emergency room physician is often the first provider to care for children with brain tumors. Thus, education of
pediatricians is an essential element in assuring
that high quality care is available from the start.
Neurosurgical education for pediatricians after
residency is limited, and thus the responsibility
for providing information falls largely on the
pediatric neurosurgeon. This deficit leads to
slower referrals and delayed diagnoses,
adversely affecting outcomes. Improved education on establishing a diagnosis and recognizing
warning signs could improve the initial care of
children with brain tumors (Aldana and
Steinbok 2009).
Hydrocephalus
A common initial presenting concern in children
with brain tumors is hydrocephalus. Because of
the flexibility of the developing skull, hydrocephalus in young children is often better tolerated
than in adults, with many patients showing only
divergent macrocephaly with no other symptoms.
Divergence of the head circumference growth is
more important than any single measurement in
infancy. Presenting symptoms in infancy can also
include vomiting, lethargy, failure to thrive,
delayed development, irritability, downward
deviation of the eyes and seizures. Older children
may complain of headaches, nausea, blurred
vision and other symptoms. Papilledema seen on
a fundoscopic exam is common.
Management of hydrocephalus in children has
been a subject of controversy, with some advocating a permanent diversion or transitory shunting prior to resection to decrease associated
morbidity and mortality. The most effective
approach to hydrocephalus is to deal with its
underlying cause. In patients with tumors, this
14 Children with Brain Tumors: Role of the Neurosurgeon
means urgently removing the mass, thus restoring the normal flow of cerebrospinal fluid (CSF).
Intravenous corticosteroids can be used as a
temporizing measure before surgery. A minority
of patients (1035%) will require CSF diversion
following posterior fossa tumor removal. Thus it
is better, in most cases, to avoid a pre-operative
diversion and the associated morbidity
(Souweidane 2009). In all cases, patients with
signs of uncontrolled intracranial pressure (ICP)
somnolence, hemodynamic instability require
immediate action.
External Ventricular Drain

The simplest approach to hydrocephalus, an
external ventricular drain (EVD) is a catheter
inserted percutaneously into the ventricular system and connected to a strain gauge transducer.
This allows rapid relief and measurement of elevated ICP. In children they are most often used in
the cases of acute hydrocephalus, VP shunt failure and traumatic brain injury.
One significant advantage of an EVD is that it
can be placed at the bedside in emergency situations. While conventional wisdom suggests this
is not preferred because of potential for infectious contamination, it is an available option in an
acutely unstable patient. Potential complications
in children include infection, misplacement,
hemorrhage, malfunction and obstruction.
External ventricular drains are effective for temporary relief of hydrocephalus, and have the
advantages of lacking permanent hardware and
not permanently altering the patients anatomy. If
a ventricular drain is placed, care must be taken
to avoid over draining CSF given the potential for
upward herniation syndromes. Thus, once
placed it is advisable to use a relatively high
threshold (20 cm H2O) for drainage with incremental reductions as needed.
Ventriculoperitoneal Shunt
The ventriculoperitoneal (VP) shunt has been a
mainstay of CSF diversionary procedures for
145
many years and has many variations in valves

and other technology. But the concept is simple:
a catheter reroutes CSF flow proximal to an
obstruction to another part of the body where it
can be reabsorbed. The distal end is most often
placed in the peritoneum. Shunt failure and
infection are the two most common complications of this particular type of diversion, and the
rates are high. Given the morbidities and disruptions of the patients life that come with repeated
shunt revisions, resection of the tumor or other
measures for ICP control are preferred in the
short term. It remains a useful tool for palliation
of elevated ICP in patients with unresectable
tumors.
Endoscopic Third Ventriculostomy

Endoscopic third ventriculostomy (ETV) was
developed as an alternative to the VP shunt in
cases of noncommunicating hydrocephalus. The
surgeon enters a lateral ventricle with a fiberoptic
or rigid lens scope and navigates into the third
ventricle. A small perforation can be made in the
floor of the third ventricle, allowing CSF flow to
bypass any obstruction distal to the third
ventricle.
This procedure is particularly useful for
patients in whom tumor removal is not possible
or immediately necessary, including those who
require biopsies and other pre-resection interventions (i.e. tectal/mesencephalic gliomas, pineal
region germ cell tumors). In children with tumors
that will not be treated surgically, diffuse brain
stem glioma for example, ETV can also be used
as a palliative measure. Complication rates vary
and are likely correlated to experience with the
technique. Regardless, it is considered a safe and
effective treatment for hydrocephalus secondary
to tumors. Given that the majority of children
with posterior fossa tumors will have a reduced
prepontine CSF cistern, the technique of ETV
can be modified to ensure safety and efficacy.
Thus, it is a strong alternative to shunting and
EVD in most situations and its scope continues to
grow as technology and individual expertise
improve.
146
Age of the Patient

The age of the child may not have a considerable
impact on the initial decision to proceed with surgery, but has a profound impact on the ensuing
course of therapy. Brain tumors in children tend
to be more aggressive than in adults, and the
developing brain is highly sensitive to radiation
therapy. As a result, adjuvant therapy for young
children has focused on chemotherapy in order to
delay the need for potentially damaging radiation. This is especially true in children under the
age of 3, for whom radiation has the most deleterious effects. In some patients the use of radiation
is being reconsidered because of improved targeting technology (Merchant et al. 2009). This
allows a focused beam of radiation to be delivered to the tumor with minimal exposure of normal brain. Further, risk stratification has been
used for some tumor types medulloblastoma in
particular to identify patients in whom radiation offers more potential benefit than risk
(Rutkowski et al. 2005).
Cytoreductive Surgery
For most tumors, the degree of resection is a significant predictor of the future disease course.
Total excision should therefore be the goal whenever possible with attention paid to the morbidity
of a more aggressive approach. The balance
between total resection of a tumor and the morbidity associated with potentially injuring normal
brain must always be considered. Tumors that are
responsive to radiation and/or chemotherapy may
accommodate a less aggressive approach as
residual tumor can be addressed with adjuvant
therapy. In addition to this exception, certain
highly aggressive tumors may only be treated
with sub-total de-bulking to reduce symptoms
associated with the tumor.
A variety of strategies have been developed to
enable the surgeon to achieve a maximal resection. High quality pre-operative imaging is a key
component of surgical planning and achieving
accurate stereotactic navigation. In addition to
the standard magnetic resonance imaging (MRI),

magnetic resonance tractography can be used to
elucidate vital pathways. Intra-operative imaging, including ultrasound, computed tomography
(CT) and MRI is becoming more widespread as a
way to limit re-operations by identifying tumor
tissue before closing. Other technologies, such as
fluorescence-guided surgery with 5-aminolevulinic acid, have developed over the last several
years to make maximal resection possible
(Stummer et al. 2006).
Operative Techniques
Traditional operative techniques for a variety of
tumors have been reevaluated over the years as
surgeons have searched for ways to reduce morbidity. One frequently used example for many
pediatric brain tumors is the cerebellar vermian
dissection for posterior fossa tumors, which has
been implicated in the posterior fossa or cerebellar mutism syndrome. This realization has led to
revision of the approach to these tumors in favor
of the less morbid telovelar approach when it is
possible.
Advances in endoscopy have paved the way
for minimally invasive resection of numerous
tumor types, particularly within the ventricular
system. Size is a major limitation, as large tumors
do not lend themselves well to endoscopic resection. But small tumors, typically less than 2 cm,
entirely within the ventricles are well suited for
endoscopic removal (Souweidane and Luther
2006). Endoscopy has also been used in transnasal
approaches to sellar and suprasellar tumors. But
with the risk of CSF leaks and concerns about
post-operative hypothalamic function, it remains
to be seen whether this approach will improve
functional outcomes in pediatric patients.
Biopsy
Maximal resection is the goal for most children
with brain tumors, and therefore biopsy is infrequently indicated. But there are certain scenarios
see Tumor Specific Considerations that
necessitate a tissue diagnosis before proceeding. This is true of tumors that require neoadjuvant therapy prior to surgical treatment, notably
germinomas and nongerminomatous germ cell
tumors. While the latter may be diagnosed
using serum or CSF markers, a biopsy is frequently useful to confirm the diagnosis of pure
germinoma or differentiate between tumor
types when marker status is equivocal (Luther
et al. 2006). Furthermore, endoscopic biopsy
can be paired with an endoscopic third ventriculostomy when relief of hydrocephalus is
needed. Endoscopic biopsy can be performed
with minimal morbidity in patients with and
without ventriculomegaly, and has a high diagnostic yield (Souweidane 2009). A stereotactic
needle biopsy can also be performed if
necessary.
Staging
An essential component of neurosurgical care of
children with brain tumors is staging. The initial
evaluation and staging influences the ensuing
treatment plan. Recent advances in intra-operative staging by way of CSF and arachnoid sampling or through endoscopic observation, although
not assessed on a large prospective scale, have
highlighted the potential role of the neurosurgeon
in disease staging (Souweidane et al. 2009).
Furthermore, the pediatric neurosurgeon should
advocate for the appropriate studies and time
frame for therapy to achieve ideal staging for the
patient.
Second Look Surgery

Maximal resection is among the best prognostic
factors for most pediatric central nervous system
tumors. Thus, second surgeries are increasingly
being used to facilitate this goal. This is particularly true for ependymoma once one has considered the risk of an aggressive second surgery
when balanced against the potential benefit. In
fact, it has been suggested that incompletely
resected ependymoma can be more effectively
147
removed in a second surgery following adjuvant

therapy (Foreman et al. 1997).
Non-germinomatous germ cell tumors are also
being considered as an emerging case in which
second-look surgery may be beneficial. These
tumors are typically treated with chemotherapy.
Following induction chemotherapy, surgery can
be used to remove any remaining tumor and to
determine whether malignant cells are present
(Blakeley and Grossman 2006). This approach
can direct future therapy for the patient.
A third set of cases for which second-look surgery may be valuable is large tumors in infancy.
Atypical teratoid/rhabdoid tumors, pineoblastomas,
choroid plexus carcinomas, and other tumors
that can grow to a significant volume in infancy
present serious risks of complications at the time
of surgery, mostly relating to the potential for
blood loss. As a result, these tumors may be
more safely and effectively treated with induction chemotherapy followed by resection, as it
has been shown that chemotherapy may reduce
tumor volume and vascularity prior to surgery
(Razzaq and Cohen 1997).
Tissue Procurement
for Investigation
As with other cancers, molecular analysis of
pediatric brain tumors is a growing field that
could yield extraordinary advances in care. In
some cases, such as atypical teratoid/rhabdoid
tumors, genetic testing of tumor tissue is needed
to confirm a diagnosis (Biegel et al. 2000).
Through molecular characterization childhood
glioblastoma multiforme (GBM) has been shown
to have distinct alterations compared with adult
tumors (Pollack et al. 2006). Alternatively, there
are tumors about which we know relatively little
regarding their molecular or genetic profiles,
such as diffuse pontine glioma. The best example
highlighting the importance of tissue procurement is the evolving role of defining the molecular biology of medulloblastoma and the impact of
this characterization on prognosis (Pfister et al.
2009). In fact, ongoing cooperative group trials
are testing the importance of these findings
148
relative to conventional prognostic indicators.

Only through the involvement of pediatric neurosurgeons will these valuable issues be clarified.
Tumor-Specic Considerations
Aside from the goals and specific anatomical
considerations that govern the surgical approach,
each tumor type presents a different set of issues
for the patient and surgeon. The concepts above
need to be carefully applied with attention to how
the tumor type influences the treatment plan.
Gliomas
These tumors comprise a large category and are
the most common tumors in children, accounting for more than half of all cases (Nejat et al.
2008). Some are seen more often in children
than in adults. Gliomas are graded using the
World Health Organization (WHO) grading
scale, where WHO grade I and II tumors are
referred to as low-grade gliomas, while WHO
grade III and IV tumors are also known as highgrade gliomas.
Low-Grade Gliomas
Juvenile pilocytic astrocytoma (JPA) is by far the
most common, representing 23.5% of all pediatric brain tumors. It is seen in the posterior fossa
and supratentorially and is often well circumscribed. It is usually seen in younger children
(median age 4) and is best treated with maximal
surgical resection. Re-resection is preferred over
adjuvant radiation or chemotherapy when necessary (Heuer et al. 2007). Because JPAs are well
circumscribed they can even be safely removed
when located in more eloquent brain regions,
thanks to the advent of effective brain mapping
technology, intraoperative monitoring and
microneurosurgical techniques. Other low grade
gliomas, such as diffusely infiltrating WHO grade
II tumors, occur more often in older children
(median age 10). They are treated similarly to
JPAs, though they may progress to higher-grade

tumors over time (Nejat et al. 2008).
There are a few exceptions to the rule of maximal resection, including low-grade gliomas involving the optic nerves and chiasm, hypothalamus or
parts of the brainstem. These should be treated
principally with chemotherapy or radiation, as primary or adjuvant therapies (Nejat et al. 2008). In
rare situations when the tumor mass causes dysfunction of surrounding structures or obstruction
of CSF pathways, partial surgical removal may be
beneficial. The prognosis for patients with lowgrade gliomas is influenced by patient age, tumor
grade and location and, most importantly, extent of
resection. Children with greater than 95% resection have 5-year and 10-year survival rates ranging
from 75% to 100% (Nejat et al. 2008).
High-Grade Gliomas
High-grade gliomas are either anaplastic
astrocytoma (WHO III) or glioblastoma multiforme (WHO IV). In children these tumors are
extremely aggressive, and tend to exhibit high
frequencies of chromosomal aberrations and
other abnormalities (Rickert et al. 2001). As a
result, symptoms generally progress more rapidly, with elevated ICP and focal neurologic deficits. Non-brain stem tumors are treated with
maximal resection with adjuvant chemotherapy
and radiation. Multiple studies have shown a correlation between 5-year survival and extent of
resection in these patients, suggesting that an
aggressive approach is warranted whenever possible (Nejat et al. 2008).
Medulloblastoma
Primitive neuroectodermal tumors are divided
into supratentorial and infratentorial posterior
fossa types, the latter being more commonly
referred to as medulloblastoma. These tumors
consist of poorly differentiated small, round blue
cells with a variety of molecular profiles, suggesting that they are actually a heterogeneous
group of undifferentiated tumor types (Mueller
and Chang 2009). Medulloblastoma is the second

most common pediatric brain tumor, accounting
for 1525% of cases and occurring most often in
children under the age of 10 (Nejat et al. 2008).
These tumors are malignant and invasive. They
typically arise from the cerebellar vermis and
frequently fill the fourth ventricle, causing hydrocephalus. The symptom profile is reflective of
elevated intracranial pressure, accompanied by
ataxia caused by cerebellar involvement.
As with most other pediatric tumors, maximal
resection is the first step in treatment (Nejat et al.
2008). But an additional important component of
the early management of these patients is risk
stratification. Patients with medulloblastoma are
currently divided into two groups: high risk and
standard risk. High risk patients are those with
residual disease (>1.5 cm2) after surgery, metastatic disease at diagnosis, or who are less than
3 years of age (Karajannis et al. 2008). Lumbar
puncture and MRI of the brain and spinal cord
are necessary to look for metastases. Imaging is
ideally completed before surgery because blood
or cellular debris in the spinal canal can complicate the interpretation of the MRI (Dhall 2009).
In contrast, lumbar puncture is rarely possible
preoperatively due to the frequency of non-communicating hydrocephalus and the risk of cerebellar herniation with CSF removal.
Another important pre-surgical consideration
in patients with medulloblastoma is the need for
CSF diversion. Approximately 40% of these
patients will need permanent shunt placement
after surgery with some variation depending on
other elements; factors increasing the likelihood
of this need are young age, larger tumors and profound hydrocephalus preoperatively (Mueller
and Chang 2009). In light of these indicators and
other considerations of the risk of post-operative
hydrocephalus, some patients may benefit from
CSF diversion prior to or at the time of surgery to
obviate the need to do so later.
Surgery in the posterior fossa carries risk, and
it is important to recognize that one must not be
overly aggressive with tumors infiltrating the
brainstem. Though morbidity varies significantly
by tumor type, a number of patients continue to
have limitation of function following posterior
149
fossa tumor removal. Postoperative cerebellar

mutism, which has been reported in as much as
24% of patients (Mueller and Chang 2009), is the
most common complication of surgery, and it can
be associated with ataxia, dysmetria, hypotonia,
dysphagia, hemiparesis and mood lability (Nejat
et al. 2008). While extent of resection has repeatedly been shown to be a positive prognostic indicator, surgical interruption of the deep cerebellar
nuclei and outflow tracts of the cerebellum have
been implicated in contributing to this syndrome.
While maximal resection is the first step for
most children with medulloblastoma, there are
differences in the therapeutic strategy depending
on the age of the patient and assignment of standard
or high risk. For standard-risk patients older than
3 years of age, there have been many studies evaluating the risks and benefits of different radiation
and chemotherapy protocols. While full craniospinal irradiation (36 Gy) has traditionally been
most effective at increasing the progression-free
survival (PFS) in these patients, the goal of reducing radiation in children has led to lower dose
craniospinal treatment (23.4 Gy) with a localized
boost to the posterior fossa (36 Gy). As each successive trial of reduced radiation has shown some
promise, the Childrens Oncology Group is currently carrying out a trial to evaluate a protocol of
lower dose craniospinal irradiation (18 Gy compared with 23.4 Gy) with a boost only to the
tumor bed, with all patients receiving adjuvant
chemotherapy (Mueller and Chang 2009).
Like standard risk patients, high-risk patients
older than age 3 should receive radiation therapy
following surgery. But it was recognized as early
as 1990 that these patients obtain a clearer benefit
from adjuvant chemotherapy than standard risk
patients, with a 5-year PFS of 48% in the treatment group and 0% in the radiation only group
(Evans et al. 1990). The current Childrens
Oncology Group study is evaluating the impact
of adding daily carboplatin as a radiosensitizer
during the course of radiation therapy (Mueller
and Chang 2009).
Children younger than 3 years of age are perhaps the most controversial treatment group.
Adjuvant radiation is harmful to the rapidly
developing nervous system, and thus attempts to
150
find alternatives are ongoing. Chemotherapy

alone, both intravenous and intraventricular, has
been considered and shown to be effective in
young children, potentially avoiding the morbidity of irradiation (Rutkowski et al. 2005).
Innovative treatments including intrathecal chemotherapy and focused involved-field irradiation
have recently been advocated with admirable
results (Merchant et al. 2008; Beutler et al. 2005).
Studies are hindered, however, by the limited
number of very young patients with medulloblastoma. Large collaborations are required to generate sufficient data (Mueller and Chang 2009).
The horizon is densely populated with ongoing
studies of postoperative care for medulloblastoma patients. Survival rates are increasing, and
their results are highly anticipated.
Ependymoma
The third most common brain tumor in children,
ependymoma occurs more often in younger children, with a mean age of diagnosis between 3 and
5 years. It is classified into three WHO grades,
though there is little proven correlation between
grade and eventual outcome (Mueller and Chang
2009). Extent of surgical resection has been
linked positively to prognosis in patients with all
grades of ependymoma, such that second-look
surgery is recommended in patients with residual
disease. Recurrences most often occur at the site
of the primary tumor (Mueller and Chang 2009).
Following gross total resection, 5-year survival
can be 5070%, while survival falls to 20% in
patients with residual disease. Other poor prognostic factors in children with ependymomas
include age less than 3 years, fourth ventricular
location, metastatic disease and anaplasia seen on
pathology (Heuer et al. 2007). Posterior fossa
location is likely on this list because it interferes
with complete resection. Further support of this
notion lies in the relatively worse outcome and
low incidence of total removal in the lateral variant of fourth ventricular ependymoma (FigarellaBranger et al. 2000). Only 3050% of posterior
fossa ependymomas can be completely removed
safely (Nejat et al. 2008).
In children older than 3 years of age, involved

field radiation therapy is frequently used to limit
local recurrences, but chemotherapy has failed to
show a clear and lasting benefit (Heuer et al.
2007). More recently, postoperative high dose
conformal radiation therapy has been shown to
benefit children with ependymoma; and it has
been suggested that future trials focus on treatment stratification (Merchant et al. 2009).
Craniopharyngioma
Craniopharyngioma is a histologically benign
tumor located in the suprasellar region that,
despite its benign pathology, can be damaging
because of its proximity to essential structures.
The adamantinomatous type occurs principally in
children, with a peak incidence between 5 and
10 years of age. Symptoms are usually related to
hydrocephalus, pituitary dysfunction and visual
field deficits (Heuer et al. 2007; Nejat et al. 2008).
Treatment options are varied and need to be
individualized for each patient depending on a
number of variables. These include pre-existing
neurologic or endocrinologic deficits, presence
of hydrocephalus, hypothalamic involvement,
patient age, and degree of cystic components.
Complete surgical resection of the tumor is the
therapy most likely to result in cure. However,
total excision is not always possible and frequently is associated with damage to nearby
structures. When only partial resection is possible
or desired for the sake of hypothalamic preservation, adjuvant radiation therapy is employed to
address the remaining tumor (Heuer et al. 2007).
This multimodality approach has the distinct
benefit of avoiding many of the neurobehavioral
and cognitive complications of treatment while
still maintaining good long-term tumor control
(Nejat et al. 2008). Some recommend considering patient age and other factors when developing the treatment plan. For example, young
patients with small tumors may benefit most from
radical surgical resection while the same patients
with larger tumors may be best served with partial resection followed by deferred radiation
treatment at the time of disease progression
(Puget et al. 2007). This treatment stratification

could protect the developing brain from harmful
radiation and avoid the morbidity of radical resection. Older children with large tumors may also
benefit from a less aggressive surgical approach
accompanied by irradiation (Kalapurakal 2005).
However, some still advocate radical surgical
resection in most cases, pointing out that it offers
a high rate of disease control and possible cure,
while at the same time avoiding the obvious morbidity associated with radiation delivered to the
developing brain. In expert hands, gross total
resection is possible in most cases. Elliott et al.
(2010) showed that children with gross total
resection are less likely to experience significant
learning setbacks and memory deficits. But they
are more likely to be emotionally labile with
limited impulse control, possibly due to the subfrontal approach used in this particular study.
Recurrent craniopharyngiomas are even more
problematic given the secondary changes in the
tumor interface and surrounding vasculature.
Purely cystic recurrences have been successfully
controlled with intracavitary therapy with bleomycin and radioisotopes.
There are several possible surgical approaches
to this tumor and others in the sellar region.
Transphenoidal resection is a common method
used in adults that is beginning to gain acceptance in the pediatric population. Other approaches
are still more common in children, including subfrontal, pterional, transcallosal and sub-temporal
(Nejat et al. 2008). Stereotactic radiosurgery is
increasingly being used for pediatric craniopharyngioma, and has improved the rate of tumor
control. It is most safe and effective for solid
tumors less than 2.5 cm in diameter and at least
3 mm away from the optic chiasm (Kalapurakal
2005). However, it has been recently suggested
that radiation may play a role in the rare events of
malignant transformation of this tumor type
(Aquilina et al. 2010).
Pineal Region Tumors

Tumors in the pineal region are ten times more
common in children than in adults, representing
151
311% of all childhood brain tumors. Symptoms

most commonly include headache, nausea, vomiting and dizziness secondary to hydrocephalus
and elevated ICP. Patients may also experience
extraocular movement and vision abnormalities
(Blakeley and Grossman 2006).
Pineal region tumors as a group represent a
special case in which a biopsy can be a valuable
first step in management owing to the variability
of tumor types that occur in the pineal region.
Germ cell tumors, both germinomatous and nongerminomatous, primitive neuroectodermal
tumors (pineoblastoma), pineal parenchymal
tumors, gliomas, meningiomas and ependymomas occur in this location with some regularity.
Importantly, germinomas can often be cured with
radiotherapy alone and it has been shown that
their resection provides no benefit to survival.
Thus it behooves the surgeon to identify patients
with this diagnosis and avoid a potentially harmful and unnecessary total resection. Prior to any
consideration of surgical intervention for tumor
sampling, serum and, if possible, CSF tumor
markers (b-Human Chorionic Gonadotropin and
a-Fetoprotein) should be assayed. In the presence of positive tumor markers, neoadjuvant chemotherapy should always be the front line therapy
for non-germinomatous germ cell tumors.
There is some variability in the method by
which tissue is obtained, ranging from a
microsurgical approach to stereotactic sampling
to endoscopic biopsy. Stereotactic and endoscopic
biopsy are less invasive then open surgery, but
carry the risk of an indeterminate or even misleading diagnosis given the high frequency of
mixed tumor types (Luther et al. 2006; Blakeley
and Grossman 2006). One important benefit of
the use of endoscopic surgery is the potential for
performing simultaneous endoscopic third
ventriculostomy for treatment of hydrocephalus.
For other pineal lesions, total resection may cure
patients with low-grade tumors and improve the
outcome of patients with more malignant lesions.
For high-grade gliomas in this region, some
advocate observation with hydrocephalus
management alone (Daglioglu et al. 2003), while
others believe that resection and radiation
are needed (Blakeley and Grossman 2006).
152
This issue remains unresolved. Stereotactic

radiosurgery is being investigated as an option
for some pineal tumors. But because very few
patients have been treated using this therapy,
efficacy cannot be fully assessed at this point
(Blakeley and Grossman 2006).
Brain Stem Tumors

The classification of brain stem tumors has undergone many changes over the years. As a group,
they make up 1020% of all pediatric central
nervous system neoplasms (Recinos et al. 2007).
They tend to occur in children age 79, and more
than half of them are diffuse gliomas (Albright
and Pollack 2004). Advances in the treatment of
brain stem tumors over the last 50 years are
largely secondary to improvements in imaging
technology. Until these lesions could be clearly
delineated with imaging, they were considered
inoperable and uniformly fatal. Today, they are
classified into four groups (Type IIV) based on
their radiologic appearances: diffuse, intrinsic
focal, exophytic focal and cervicomedullary
(Recinos et al. 2007). Identification of tumor type
delineates the appropriate course of action for the
patient and neurosurgeon. Biopsy is rarely indicated for patients with brainstem tumors although
it can be performed safely. It is recommended
that biopsy be reserved for indeterminate
lesions of the brainstem when imaging is not
sufficient to guide the next step in treatment
(Leach et al. 2008).
For children whose clinical presentation and
imaging are consistent with diffuse gliomas of
the brainstem, surgery has demonstrated no benefit in survival. Corticosteroids or irradiation have
been shown to temporarily stabilize these lesions
and symptomatic shunting has been used (Recinos
et al. 2007). However, most children die within
18 months (Heuer et al. 2007). As with adult
high-grade gliomas, these tumors in children are
subjects of intense investigation in the hopes of
finding chemotherapeutic agents that can effectively limit their growth.
Focal lesions intrinsic, exophytic or cervicomedullary are handled very differently: surgical
resection is the most effective first step.

Neurophysiological mapping is used to avoid
damage to essential structures of the brainstem.
Surgical morbidity remains higher than for surgery in most other brain regions, and therefore
an over-aggressive surgical approach should be
avoided. While radical resection is preferred and
offers the best long-term prognosis, it has been
shown that greater than 50% resection of intrinsic medullary tumors improves survival and
decreases tumor progression (Recinos et al.
2007). In light of this evidence, the potential
morbidity of radical resection should be weighed
against its benefits prior to surgery and in the
operating room.
Prognosis after surgery is dependent on the
tumor type and extent of resection. As we have
seen, patients with diffuse gliomas have the worst
prognosis. On the other hand those with dorsal
exophytic tumors have excellent prognoses.
Patients with focal intrinsic tumors of the medulla
frequently have a long survival following surgery
with slow progression of disease. The prognosis
after treatment of cervicomedullary tumors has
been shown to be related to preoperative function
(Recinos et al. 2007). Consequently, surgery
before the onset of severe symptoms should be
considered when possible.
Future Directions
The field of pediatric neurosurgery continues to
advance rapidly, and there are many changes on
the horizon that could alter the way pediatric
brain tumors are managed. It is clear that minimally invasive techniques, such as endoscopy
and microsurgery, will evolve with the goal of
making surgery less morbid while improving
therapeutic benefits.
Another newer area of investigation is in the
delivery of anti-neoplastic agents by the surgeon.
This field is in its infancy where children are concerned. But some possibilities have been explored,
including interstitial infusions, convection
enhanced delivery, and the use of drug-impregnated
wafers delivered to the site of the tumor (Westphal
et al. 2003; Degen et al. 2003; Souweidane
et al. 2004). Advances in this technology will

allow children to be treated with fewer systemic
side effects while simultaneously taking an aggressive approach at treating the offending tumor.
The themes of risk and treatment stratification
continue to permeate the field of pediatric neurooncology. New ways of thinking about ependymoma,
craniopharyngioma, and other tumors have led us
to conclude the there are times when gross-total
resection is not warranted, and that a tailored
approach is needed to maximize the benefits to the
patient. With improvements in imaging and operative techniques, this trend toward individualized
therapies will push treatment in the direction of
better quality of life for patients.
Neurosurgical care of children with brain
tumors is highly specialized and multi-faceted.
We have seen that children are uniquely responsive
to many therapeutic options, surgical and otherwise.
It is also clear that the expertise of a pediatric
neurosurgeon provides an added benefit to these
children. The role of the pediatric neurosurgeon
continues to grow and develop, with the hope of
producing safer and more effective care for
children with brain tumors.
References
Albright AL, Pollack IF (2004) Brainstem gliomas.
Youmans Neurol Surg 3:36633669
Albright AL, Sposto R, Holmes E, Zeltzer PM, Finlay JL,
Wisoff JH, Berger MS, Packer RJ, Pollack IF (2000)
Correlation of neurosurgical subspecialization with
outcomes in children with malignant brain tumors.
Neurosurgery 47(4):879885; discussion 8587
Aldana PR, Steinbok P (2009) Prioritizing neurosurgical
education for pediatricians: results of a survey of pediatric neurosurgeons. J Neurosurg Pediatr 4:309316
Aquilina K, Merchant TE, Rodriguez-Galindo C, Ellison
DW, Sanford RA, Boop FA (2010) Malignant transformation of irradiated craniopharyngioma in children.
J Neurosurg Pediatr 5:155161
Beutler D, Avoledo P, Reubi J-C, Macke HR, MuellerBrand J, Merlo A, Kuhne T (2005) Three-year recurrence-free survival in a patient with recurrent
medulloblastoma after resection, high-dose chemotherapy, and intrathecal Yttrium-90-labeled DOTA0D-Phe1-Tyr3-octreotode radiopeptide brachytherapy.
Cancer 103(4):869873
153

6(7):27592763
Blakeley JO, Grossman SA (2006) Management of pineal
region tumors. Curr Treat Options Oncol 7:505516
Daglioglu E, Catalepe O, Akalan N (2003) Tectal
gliomas in children: the implications for natural history and management strategy. Pediatr Neurosurg
38:223231
Degen JW, Walbridge S, Vortmeyer AO, Oldfield EH,
Lonser RR (2003) Safety and efficacy of convectionenhanced delivery of gemcitabine or carboplatin in a
malignant glioma model in rats. J Neurosurg
99(5):893898
Dhall G (2009) Medulloblastoma. J Child Neurol
24:14181430
Elliott RE, Hsieh K, Hochman T, Belitskaya-Levy I,
Wisoff J, Wisoff JH (2010) Efficacy and safety of radical
resection of primary and recurrent craniopharyngiomas
in 86 children. J Neurosurg Pediatr 5:3048
Evans AE, Jenkins RD, Sposto R, Ortega JA, Wilson CB,
Wara W, Ertel IJ, Kramer S, Chang SH, Leikin SL,
Hammond GD (1990) The treatment of medulloblastoma:
results of a prospective randomized trial of radiation
therapy with and without CCNU, vincristine, and prednisone. J Neurosurg 72(4):572582
Figarella-Branger D, Civatte M, Bouvier-Labit C,
Gouvernet J, Gambarelli D, Gentet JC, Lena G, Choux
M, Pellissier JF (2000) Prognostic factors in intracranial ependymomas in children. J Neurosurg
93(4):605613
Foreman NK, Love S, Gill SS, Coakham HB (1997)
Second-look surgery for incompletely resected fourth
ventricle ependymomas: technical case report.
Neurosurgery 40(4):856860
Heuer GG, Jackson EM, Magge SN, Storm PB (2007)
Surgical management of pediatric brain tumors. Expert
Rev Anticancer Ther 7(12 Suppl):S61S68
Kalapurakal JA (2005) Radiation therapy in the management of pediatric craniopharyngioma a review.
Childs Nerv Syst 21:808816
Karajannis M, Allen JC, Newcomb EW (2008) Treatment of
pediatric brain tumors. J Cell Physiol 217(3):584589
Leach PA, Estlin EJ, Coope DJ, Thorned JA, Kamaly-Asl
ID (2008) Diffuse brainstem gliomas in children:
should we or shouldnt we biopsy. Br J Neurosurg
22(5):619624
Luther N, Edgar MA, Dunkel IJ, Souweidane MM (2006)
Correlation of endoscopic biopsy with tumor marker
status in primary intracranial germ cell tumors.
J Neurooncol 79(1):4550
Merchant TE, Kun LE, Krasin MJ, Wallace D,
Chintagumpala MM, Woo SY, Ashley DM, Sexton M,
Kellie SJ, Ahern V, Gajjar A (2008) Multi-institution
prospective trial of reduced-dose craniospinal
irradiation (23.4 Gy) followed by conformal posterior
fossa (36 Gy) and primary site irradiation (55.8 Gy)
and dose-intensive chemotherapy for average-risk
medulloblastoma. Int J Radiat Oncol Biol Phys
70(3):782787
154
Merchant TE, Li C, Xiong X, Kun LE, Boop FA, Sanford
RA (2009) Conformal radiotherapy after surgery for
paediatric ependymoma: a prospective study. Lancet
Oncol 10(3):258266
Merchant TE, Pollack IF, Loeffler JS (2010) Brain tumors
across the age spectrum: biology, therapy and late
effects. Semin Radiat Oncol 20:5866
Mueller S, Chang S (2009) Pediatric brain tumors: current
treatment strategies and future therapeutic approaches.
Neurotherapeutics 6(3):570586
Mukherjee D, Kosztowski T, Zaidi HA, Jallo G, Carson B,
Chang DC, Quinones-Hinojosa A (2009) Disparities
in access to pediatric neurooncological surgery in the
United States. Pediatrics 124:e688e696
Nejat F, Khashab ME, Rutka JT (2008) Initial management of childhood brain tumors: neurosurgical considerations. J Child Neurol 23:11361148
Pfister S, Remke M, Benner A, Mendrzyk F, Toedt G,
Felsberg J, Wittman A, Devens F, Gerber NU,
Joos S, Kulozik A, Reifenberger G, Rutkowski S,
Wiestler OD, Radlwimmer B, Scheurlen W, Lichter
P, Korshunov A (2009) Outcome prediction in pediatric medulloblastoma based on DNA copy-number
aberrations of chromosomes 6q and 17q and
the MYC and MYCN loci. J Clin Oncol
27(10):16271636
Pollack IF, Hamilton RL, James CD, Finkelstein SD,
Burnham J, Yates AJ, Holmes EJ, Zhou T, Finlay JL
(2006) Rarity of PTEN deletions and EGFR amplification in malignant gliomas of childhood: results from
the Childrens Cancer Group 945 cohort. J Neurosurg
105(5 suppl):418424
Puget S, Garnett M, Wray A, Grill J, Habrand JL, Bodaert N,
Zerah M, Bezerra M, Renier D, Pierre-Kahn A, SainteRose C (2007) Pediatric craniopharyngiomas: classification and treatment according to the degree of hypothalamic
involvement. J Neurosurg 106(1 Suppl):312
Razzaq AA, Cohen AR (1997) Neoadjuvant chemotherapy for hypervascular malignant brain tumors of childhood. Pediatr Neurosurg 27(6):296303

Recinos PF, Sciubba DM, Jallo GI (2007) Brainstem
tumors: where are we today? Pediatr Neurosurg
43:192201
Rickert CH, Strater R, Kaatsch P, Wassmann H, Jurgens
H, Dockhorn-Dworniczak B, Paulus W (2001)
Pediatric high-grade astrocytomas show chromosomal
imbalances distinct from adult cases. Am J Pathol
158:15251532
Rutkowski S, Bode U, Deinlein F, Ottensmeier H,
Warmuth-Metz M, Soerensen N, Graf N, Emser A,
Pietsch T, Wolff JEA, Kortmann RD, Kuehl J (2005)
Treatment of early childhood medulloblastoma by
postoperative chemotherapy alone. N Engl J Med
352(10):978986
Souweidane MM (2009) The evolving role of surgery in
the management of pediatric brain tumors. J Child
Neurol 24(11):13661374
Souweidane MM, Luther N (2006) Endoscopic resection
of solid intraventricular brain tumors. J Neurosurg
105(2):271278
Souweidane MM, Ochiogrosso G, Mark EB, Edgar MA
(2004) Interstitial infusion of IL13-PE38QQR in the
rat brain stem. J Neurooncol 67(3):287293
Souweidane MM, Morgenstern PF, Christos PJ, Edgar
MA, Khakoo Y, Rutka JT, Dunkel IJ (2009)
Intraoperative arachnoid and cerebrospinal fluid sampling in children with posterior fossa brain tumors.
Neurosurgery 65(1):7278
Stummer W, Pichlmeier U, Meinel T, Wiestler OD, Zanerlla
F, Reulen HJ, ALA-Glioma Study Group (2006)
Fluorescence-guided surgery with 5-aminolevulinic
acid for resection of malignant glioma: a randomized
controlled multicentre phase III trial. Lancet Oncol
7(5):392401
Westphal M, Hilt DC, Bortey E, Delavault P, Olivares R,
Warnke PC, Whittle IR, Jskelinen J, Ram Z (2003)
A phase 3 trial of local chemotherapy with biodegradable carmustine (BCNU) wafers (Gliadel wafers) in
patients with primary malignant glioma. Neurooncology 5(2):7988
Pediatric Intraventricular Brain

Tumors: Endoscopic Neurosurgical
Techniques
15
David I. Sandberg and Faiz Ahmad
Contents
Abstract
Introduction ............................................................
155
Endoscopic Biopsy of Pediatric

Intraventricular Brain Tumors .............................
Indications ................................................................
Techniques ...............................................................
Outcomes .................................................................
156
156
157
158
Simultaneous Endoscopic Biopsy of Pediatric

Intraventricular Brain Tumors
and Cerebrospinal Fluid Diversion ......................
Indications ................................................................
Techniques ...............................................................
Outcomes .................................................................
159
159
159
160
Endoscopic Resection of Pediatric

Intraventricular Brain Tumors .............................
Indications ................................................................
Techniques ...............................................................
Outcomes .................................................................
160
160
161
161
Conclusions .............................................................
162
References ...............................................................
162
Endoscopic neurosurgical techniques are

increasingly being utilized in the management
of intraventricular brain tumors in children.
Because intraventricular tumors are deepseated within the brain, traditional surgical
approaches require a large incision and craniotomy. Modern endoscopic techniques are
employed by pediatric neurosurgeons to treat
many intraventricular brain tumors in a minimally invasive manner. When a histological
diagnosis is required in order to determine
how a particular tumor should be managed,
endoscopic biopsy is often the least invasive
means of obtaining a specimen. The majority
of children with intraventricular brain tumors
present with hydrocephalus, and endoscopic
techniques enable simultaneous treatment of
hydrocephalus along with tumor biopsy. While
the majority of intraventricular brain tumors
cannot be completely resected with current
technology, selected cases are amenable to
complete surgical resection. This chapter
reviews indications, techniques, and outcomes
associated with endoscopic management of
pediatric intraventricular brain tumors.
Introduction
D.I. Sandberg (*) F. Ahmad
Department of Neurological Surgery, University of
Miami Miller School of Medicine and Miami Childrens
Hospital, Ambulatory Care Building, Suite 3109, 3215
SW 62nd Avenue, Miami, FL 33155, USA
e-mail: dsandberg@med.miami.edu
Endoscopic techniques have greatly expanded

the ability of neurosurgeons to treat a variety of
intracranial disorders in a less invasive manner
than traditional, open surgical techniques.

155
D.I. Sandberg and F. Ahmad
156
Endoscopy was first popularized by urologists

and gynecologists. The first recorded use of
endoscopy by a neurosurgeon was in 1922, when
Walter Dandy used a cystoscope to inspect the
ventricles in two patients. Use of the endoscope
over the next four decades was rare because of
cumbersome available instruments and primitive
lighting from a lamp at the end of the endoscope.
The development of an external light source
adapted specifically for endoscopes at the Institute
of Optics of Paris increased the light intensity
immensely and revolutionized the field of endoscopy (Guiot and Comoy 1963). However, use of
the endoscope by neurosurgeons was limited
until the development of computed tomography
(CT) and magnetic resonance imaging (MRI)
greatly expanded the ability to diagnose conditions amenable to endoscopic treatment. In the
past two decades, advances in fiberoptic technology and improved available instrumentation have
greatly expanded applications of the endoscope
in brain surgery.
Currently, endoscopic techniques are frequently
utilized in the treatment of hydrocephalus, symptomatic arachnoid cysts, tumors of the ventricles
and skull base, and craniosynostosis. Advantages
of endoscopic techniques over traditional, open
craniotomies include smaller incisions, decreased
postoperative pain, less blood loss, and shorter
hospitalizations. Endoscopic techniques are typically employed in place of open approaches, but
occasionally the endoscope is used simultaneously
with conventional microsurgical methods to
improve visualization of anatomical locations that
are challenging to access.
This chapter focuses on the current role of
endoscopic neurosurgical techniques in the management of intraventricular brain tumors in children. The location of tumors within the ventricles
renders them amenable to endoscopy, as cerebrospinal fluid (CSF) is an excellent medium for
transmission of light and images. Traditional, open
neurosurgical techniques for intraventricular
tumors are often challenging because the tumors
location deep beneath the cortical surface requires
significant retraction of cerebral cortex for adequate visualization. Aggressive surgical resection
is not required for some neoplasms found within
the ventricles, which are best treated with radiation

therapy and/or chemotherapy. Endoscopic biopsy
of some intraventricular tumors can enable a pathological diagnosis to be established in a minimally
invasive manner. When intraventricular tumors
cause obstructive hydrocephalus, endoscopy
enables simultaneous tumor biopsy with cerebrospinal fluid diversion, sparing some patients from
implantation of a shunt. Selected intraventricular
tumors can be resected completely with endoscopic techniques. This chapter reviews these
applications of endoscopy in children with intraventricular brain tumors with a focus on indications, techniques, and outcomes.
Endoscopic Biopsy of Pediatric

Indications
Endoscopic biopsy is most often performed for
tumors in the lateral ventricles, at the foramen of
Monro, and in the third ventricle. Endoscopic
biopsies of paraventricular (thalamogeniculate,
mesencephalic, or basal ganglia) tumors are
generally avoided unless the ependymal surface
is clearly violated, as these tumors are more
amenable to biopsy via standard stereotactic
techniques (Yurtseven et al. 2003; Stachura et al.
2005). Tumors in the fourth ventricle are typically not approached endoscopically because
essentially all fourth ventricular tumors in children require aggressive surgical resection,
which is best achieved via open microsurgical
techniques.
If the pathology of an intraventricular brain
tumor is obvious from pre-operative imaging
studies pre-operatively, resection is necessary,
and the lesion is not amenable to complete
endoscopic resection, then open microsurgical
techniques are employed. Thus, for example, a
patient with tuberous sclerosis with an obvious
subependymal giant cell astrocytoma at the foramen of Monro causing obstructive hydrocephalus
would not be offered an endoscopic biopsy. In this
case, the patient would typically be offered an
open craniotomy, which would offer the best
15 Pediatric Intraventricular Brain Tumors: Endoscopic Neurosurgical Techniques
157
Techniques
Fig. 15.1 Coronal T1-weighted MRI demonstrating a

homogeneously enhancing lesion in the suprasellar region
and third ventricle in an 11-year-old boy
chance of complete resection with currently

available technology. Contrarily, endoscopic
biopsy is considered if the differential diagnosis
of an intraventricular lesion includes pathological entities such as germ cell tumors, central nervous system (CNS) lymphoma, or disseminated
metastatic tumors which do not require aggressive surgical resection. Endoscopic biopsy in
these cases can spare patients from an unnecessarily extensive open procedure (Souweidane
et al. 2000; Badie et al. 2004).
As an example, Fig. 15.1 illustrates a case in
which an endoscopic biopsy of an intraventricular tumor was performed to make a pathological
diagnosis and determine the appropriate treatment. In this case, an 11-year-old boy presented
with headaches and an MRI demonstrated a large,
homogeneously enhancing lesion in the suprasellar region and third ventricle. Highest on the
differential diagnosis in this case was germinoma,
a tumor which is treated by radiation therapy and/
or chemotherapy rather than surgical resection.
An endoscopic biopsy was therefore performed.
To our surprise, the pathological diagnosis was
ganglioglioma, a rare tumor in this location.
Subtotal surgical resection was subsequently
performed via an open craniotomy, but the
endoscopic biopsy enabled the appropriate
treatment plan to be formulated.
Endoscopic tumor biopsies are performed under

general anesthesia. In the majority of cases, ventricular cannulation is straightforward because
the tumor has caused obstructive hydrocephalus
and the ventricles are enlarged. Tumor biopsy in
the absence of ventricular dilation is technically
challenging and should only be performed by
an experienced endoscopic surgeon, typically
with the aid of frameless stereotactic guidance (Souweidane 2005a, b ; Cappabianca
et al. 2008 ) .
The patients head is immobilized with pin
fixation if the procedure is combined with frameless stereotaxy. Otherwise, the head is
positioned in a padded horseshoe head-holder.
If the tumor is located in the frontal horn or
body of the lateral ventricle or in the third
ventricle, surgery is performed with the
patient supine. If the tumor is located in the
occipital horn or atrium of the lateral ventricle, a posterior parietal approach is used with
the patient typically positioned in a lateral
decubitus position or prone.
An entry point and trajectory is planned such
that eloquent cortex and major blood vessels are
avoided and the least possible amount of normal
brain is traversed en route to the tumor. A small
incision (approximately 3 cm) is made and then
followed by a single burr hole. The dura mater is
opened, and the pia mater of the brain is coagulated and incised. A rigid cannula and obturator
are introduced into the lateral ventricle, and then
the obturator is removed and the endoscope is
introduced. At this point, CSF is often sampled
for cytology and/or tumor markers. The majority
of tumors are approached via a burr hole situated
several centimeters lateral to the midline at
approximately the coronal suture. Once the frontal horn of the lateral ventricle is entered, the
choroid plexus is identified at the foramen of
Monro. For tumors located in the lateral ventricle
or at the foramen of Monro, the lesion is immediately
visualized upon entry into the frontal horn.
Tumors in the third ventricle are accessed by
introducing the endoscope through the foramen
of Monro into the third ventricle. Great care is
158
taken not to injure the fornix. The majority of

intraventricular tumors can be accessed with a
straight trajectory, so a zero degree rigid endoscope is used. For tumors which cannot be
easily reached with a straight trajectory, angled
(30 or 70 degree) rigid endoscopes or flexible
endoscopes are used.
Before performing the biopsy, other accessible
areas within the ventricles are inspected for any
obvious lesions that were not apparent on preoperative imaging studies. Additional lesions
may be observed with direct inspection which are
not visible on pre-operative imaging studies.
Biopsy specimens are taken with cupped biopsy
forceps. Coagulation of the tumor prior to taking
specimens is avoided, as is may prevent an accurate pathological diagnosis from being made.
Multiple small samples are taken, preferably
from different sites within the lesion. At least one
sample is sent for frozen-section to ensure that
lesional tissue rather than normal brain has been
sampled and to obtain a preliminary diagnosis.
Bleeding from the site of tumor biopsy is controlled by irrigating with lactated ringers solution
or by tamponade using an inflated embolectomy
balloon. Persistent bleeding may be controlled
with bipolar cautery after adequate biopsy specimens have been removed.
After the biopsy has been completed and
hemostasis obtained, the endoscope is carefully
removed. A small piece of gelfoam is placed in
the cortical opening to minimize CSF egress. An
external ventricular drain may be left in place if
there has been significant intraventricular hemorrhage or if associated hydrocephalus has not
been effectively treated by endoscopic techniques. The small dural opening is either sutured
closed or left open and covered with a small
piece of gelfoam, and the skin incision is then
sutured closed.
Outcomes
In published reports, endoscopic biopsy of intraventricular brain tumors has a high rate of accurately
obtaining a pathological diagnosis with acceptably
low morbidity (Gaab and Schroeder 1998;
Yurtseven et al. 2003; Luther et al. 2005;

Depreitere et al. 2007). Reported rates of accurately making a diagnosis range widely from 57%
to 100%, but higher diagnostic yield is likely
achieved in centers where neurosurgeons frequently perform endoscopic biopsies and pathologists are experienced in making a diagnosis with
relatively small tumor specimens (Gaab and
Schroeder 1998; Yurtseven et al. 2003; Luther
et al. 2005; Depreitere et al. 2007). New neurological
deficits are extremely rare with endoscopic biopsy
of intraventricular tumors. Wound complications
and infections occur at relatively low rates, similar to those reported with other intracranial procedures. CSF is most likely when a biopsy is not
accompanied by appropriate treatment for hydrocephalus. The most feared complication from
endoscopic tumor biopsy, intraventricular hemorrhage, is relatively uncommon. Luther et al.
(2005) reported a 3.5% rate of hemorrhage in a
series of 86 patients who underwent endoscopic
brain tumor surgeries, and hemorrhage resulted
in a new neurological deficit in only one patient.
In two patients, intraventricular hemorrhage
necessitated aborting the procedure before completion of the pre-operative objective. In a series
of 46 patients with intraventricular tumors who
underwent a variety of endoscopic procedures
(biopsy, subtotal or total tumor resection, third
ventriculostomy, or stent placement), Schroeder
et al. (2004) reported one permanent complication (permanent memory loss) and seven transient
complications (infection, memory loss, mutism,
CSF leak, intraventricular hemorrhage, and trochlear nerve palsy). No surgeries in this series
were aborted due to intraventricular hemorrhage
or poor visualization. In another series of 46
patients with intraventricular and/or periventricular tumors, Tirakotai et al. (2007) reported three
transient complications (two wound infections
and one cerebrospinal fluid leak) and one new
neurological deficit related to post-operative
hemorrhage but no mortality.
In summary, in experienced centers, endoscopic biopsy can typically accomplish the objective of obtaining a pathological tumor diagnosis
with almost no risk of mortality and relatively
low morbidity rates.
159
performed to treat the hydrocephalus and the

one of the tumors was biopsied. Pathological
evaluation revealed the diagnosis of germinoma,
a disease which does not require aggressive surgical resection, and the patient was treated with
external beam radiation therapy and chemotherapy. No subsequent surgeries for CSF diversion
were required.
Techniques
Fig. 15.2 Sagittal T1-weighted MRI scan with gadolinium

demonstrating multiple enhancing intraventricular lesions
and obstructive hydrocephalus in a 17-year-old boy
Simultaneous Endoscopic Biopsy

of Pediatric Intraventricular Brain
Tumors and Cerebrospinal Fluid
Diversion
Indications
The majority of children with intraventricular
brain tumors present with symptomatic hydrocephalus because tumors obstruct CSF pathways.
Obstruction can occur at the foramen of Monro
(either unilaterally or bilaterally), at the cerebral
aqueduct, or at the fourth ventricular outlet.
Endoscopy often enables simultaneous treatment
of hydrocephalus, tumor sampling to determine a
pathologic diagnosis, and CSF sampling for tumor
markers. Achieving these objectives with a single
operation not only spares the patient from undergoing two separate surgical procedures but also
can avoid the need for ventriculoperitoneal shunting and its associated complications.
As an example, Fig. 15.2 shows the preoperative MRI scan of a 17-year-old boy who presented with headaches and vomiting. The MRI
revealed multifocal, enhancing lesions causing
obstructive hydrocephalus. In a single operation, an endoscopic third ventriculostomy was
Endoscopic third ventriculostomy (ETV) is the

most common procedure for CSF diversion performed jointly with endoscopic biopsy of an
intraventricular tumor. When these procedures
are performed simultaneously, the ETV should
be completed before the biopsy. Because patients
are typically symptomatic from hydrocephalus,
treating the hydrocephalus by ETV is the most
important priority. Performing the biopsy first
carries the risk of rendering the ETV harder to
complete if bleeding from the tumor obscures the
surgeons view.
The technical details of ETV can be reviewed
in previous publications (Sandberg 2008). After
induction of general anesthesia, a single burr hole
is made and the endoscope is inserted into the
lateral ventricle. CSF is sampled for routine analysis, cytology, and/or tumor markers. When the
ideal trajectory for the ETV differs from the ideal
trajectory for tumor biopsy, considerable thought
is given to the best entry site. For example, to
biopsy a pineal region tumor, an anterior incision
and burr hole are optimal in order to reach the
posterior third ventricle without injuring the
fornix. However, a very anterior entry site makes
the ETV technically difficult to perform, as the
straightest trajectory to the tuber cinereum is
usually via entry at around the coronal suture.
Making two burr holes is an option in these cases,
but both procedures can generally be performed
with a single burr hole chosen by splitting the difference between the ideal entry sites for the ETV
and the biopsy. If necessary, after completion of
the ETV with a rigid zero degree endoscope, an
angled endoscope (most commonly 30 degrees)
is inserted in order to better visualize the posterior third ventricle. Flexible endoscopes are also
160
available, but rigid endoscopes generally offer

superior optics.
In addition to ETV, fenestration of the septum
pellucidum (septostomy) and fenestration of
tumor cysts can be performed together with endoscopic biopsy of intraventricular tumors.
Septostomy is performed when the two lateral
ventricles do not communicate with one another,
typically due to obstruction at the foramen of Monro.
Technical details regarding septostomy and cyst
fenestration can be reviewed in previous publications (Sandberg 2008). Once the procedure for
CSF diversion has been completed, endoscopic
biopsy of the intraventricular tumor is performed
as described previously in this chapter.
Outcomes
Successful treatment of tumor-related hydrocephalus is judged by resolution of clinical symptom, decreased ventricular size and/or resolution
of transependymal CSF absorption, and avoiding
shunting procedures. By these criteria, numerous
reports have described successful treatment of
tumor-related hydrocephalus in appropriate
patients (Gaab and Schroeder 1998; Javadpour
and Mallucci 2004; Cipri et al. 2005; Luther et al.
2005; Klimo and Goumnerova 2006; OBrien
et al. 2006). In the majority of these reports, ETV
was the procedure used to treat hydrocephalus
simultaneously with endoscopic tumor management. For example, Gaab and Schroeder reported
complete resolution of hydrocephalus-related
symptoms in all 22 patients with CSF obstruction
from intraventricular lesions after endoscopic
treatment (Gaab and Schroeder 1998). In another
series, OBrien et al. (2006) managed 42 patients
with tumor-associated obstructive hydrocephalus
and noted that ETV successfully treated hydrocephalus in 68% of patients. Ray et al. (2005)
reported a similar success rate of 70% in treating
43 patients with tumor-related hydrocephalus by
ETV. Depreitere et al. (2007) reported a 64%
success rate in hydrocephalus management in
the 14 patients who underwent endoscopic third
ventriculostomy along with endoscopic biopsy.
Twenty of 46 patients with intraventricular or
periventricular tumors reported by Tirakotai et al.

(2007) had associated hydrocephalus. Nineteen
of these patients were successfully treated with
ETV, and one required endoscopic stent placement. ETV performed by Macarthur et al. (2002)
in patients with intraventricular tumors successfully treated hydrocephalus in 63 of 66 patients
(95%) in the short term and 55 of 66 patients
(83%) in the long term. Finally, in a recent series
reported by Souweidane (2005b), 22 of 26
patients with intraventricular tumors presented
with hydrocephalus. Eight of these patients were
successfully treated with ETV, four underwent
endoscopic septostomy, and four underwent cyst
decompression with successful restoration of
CSF flow. Overall, 16 of these 22 patients (73%)
underwent successful endoscopic management
and did not require shunting.
In summary, endoscopic CSF diversion is very
effective in treating obstructive hydrocephalus
caused by intraventricular tumors. Reported complication rates in the studies described above
are acceptably low. The majority of patients are
treated by ETV, but other techniques such as septostomy and cyst fenestration are often utilized
simultaneously with tumor biopsy to successfully
treat associated hydrocephalus.
Endoscopic Resection of Pediatric

Indications
In most publications to date, authors have noted
that endoscopy cannot achieve gross total removal
of most intraventricular brain tumors. According
to Cappabianca et al. (2008), the ideal tumor for
complete endoscopic resection should have the
following characteristics: soft consistency, small
size (<2 cm in greatest diameter), associated
hydrocephalus, mild to moderate vascularity, low
histological grade, singular lesion, location completely inside the ventricles, and accessibility
through a straight trajectory. Only a small minority of intraventricular tumors meet all of these
criteria, but complete resection of intraventricular tumors of various pathologies has been
reported (Abdullah and Caemaert 1995; Gaab

and Schroeder 1998; Stachura et al. 2005;
Souweidane and Luther 2006; Cappabianca et al.
2008; Idris et al. 2008).
Several factors should be considered before
attempting complete resection. These include the
tumors size, vascularity, and anatomical relationship to important intracranial structures
(Gaab and Schroeder 1998; Souweidane and
Luther 2006; Cappabianca et al. 2008). Plain CT
scans of the head may be helpful in predicting
the consistency of the tumor, as hyperdense or calcified lesions are likely to be more firm and difficult to remove than hypodense lesions (Souweidane
and Luther 2006). The experience of the surgeon
should also be taken into account, as endoscopic
tumor resection can be challenging and should not
be attempted by neurosurgeons with limited
experience performing endoscopic surgeries.
Techniques
Planning the entry point and the general approach
are performed in the same manner as described
previously for endoscopic biopsies. If frameless
stereotaxy is used to achieve the ideal entry point
and trajectory, the patients head is secured with
pin fixation. A large burr hole is made to enable
movement of the endoscope at different angles.
Such movement is important to visualize various
components of the tumor as well as surrounding
structures. Endoscopic brain tumor removal is
performed piecemeal, as en-bloc endoscopic
resection is not possible with currently available
instrumentation. Tumors are typically resected
using a combination of biopsy forceps, bipolar
coagulation, microscissors, and suction. Typically,
an initial biopsy is taken from an avascular area
without cauterization in order to ensure that a
pathological diagnosis is made. The tumor capsule may then be coagulated and incised further.
Soft intracapsular contents may be suctioned out
using a small pediatric feeding tube. Care is taken
to only use suction within the tumor capsule to
avoid suctioning CSF and thereby rapidly decompressing the ventricle. Hemostasis is achieved
using a combination of irrigation, tamponade
161
with an embolectomy balloon, and bipolar or

monopolar coagulation.
As noted previously, the majority of intraventricular tumors are not completely resectable
endoscopically due to features inherent to the
tumor as well as the limited available instrumentation and small working channels. The development of improved instruments adapted for the
endoscope, such as miniaturized ultrasonic
aspirators and non-ultrasonic aspiration devices,
will likely increase the neurosurgeons ability to
completely resect intraventricular tumors endoscopically in the future (Lekovic et al. 2006;
Oertel et al. 2008). Published experience with
these devices is limited at the present time.
Outcomes
The largest published experience with complete
resection of intraventricular brain tumors is for
colloid cysts of the anterior third ventricle
(Longatti et al. 2006). Some craniopharyngiomas
which are relatively small and largely cystic may
also be amenable to complete endoscopic resection (Cinalli et al. 2006). Resection of solid
tumors has been reported less frequently and is
limited to case reports and small series (Gaab and
Schroeder 1998; Macarthur et al. 2002; Luther
and Souweidane 2005; Souweidane and Luther
2006). Harter et al. (2006) reported complete
endoscopic resection of a dysembryoplastic neuroepithelial tumor (DNET) located in the foramen of Monro. Macarthur et al. (2002) reported
four extensive and three partial tumor resections
among 77 patients who underwent endoscopic
procedures for brain tumors. Luther and
Souweidane (2005) reported complete endoscopic removal of a posterior third ventricular
ependymoma. These authors later reported
attempting resection of solid intraventricular
brain tumors in seven patients, and gross total
resection was achieved in five patients
(Souweidane and Luther 2006). Subtotal resection was achieved in the other two patients. In 30
cases of endoscopically treated intra-ventricular
lesions, Gaab and Schroeder (1998) abandoned
two attempts at complete endoscopic resection
162
because the consistency of the tumor was too firm

for endoscopic removal. The authors were able to
completely remove all colloid cysts and epidermoid lesions and five additional tumors. They
reported no operative mortality and four patients
had complications (meningitis, memory loss or
mutism). The authors conclude that endoscopic
intervention is safe and effective and that complete
excision is possible in selected intraventricular
lesions.
Conclusions
Endoscopic neurosurgical techniques enable
minimally invasive management of selected intraventricular tumors. Endosopic tumor biopsies
and occasionally complete tumor resection can
be achieved with acceptably low morbidity via
small incisions and a single burr hole.
Simultaneous endoscopic treatment of hydrocephalus by endoscopic third ventriculostomy
and other techniques enables many patients to
avoid shunting procedures. With improvements
in technology and instrumentation, it is likely
that an increasing proportion of intraventricular
tumors will be completely resected with endoscopic techniques, thereby sparing patients more
invasive open neurosurgical procedures.
References
Abdullah J, Caemaert J (1995) Endoscopic management
of craniopharyngiomas: a review of 3 cases. Minim
Invasive Neurosurg 38(2):7984
Badie B, Brooks N, Souweidane M (2004) Endoscopic
and minimally invasive microsurgical approaches for
treating brain tumor patients. J Neurooncol
69(13):209219
Cappabianca P, Cinalli G, Gangemi M, Brunori A, Cavallo
LM, de Divitiis E, Decq P, Delitala A, Di Rocco F,
Frazee J, Godano U, Grotenhuis A, Longatti P, Mascari
C, Nishihara T, Oi S, Rekate H, Schroeder HW,
Souweidane MM, Spennato P, Tamburrini G, Teo C,
Warf B, Zymberg ST (2008) Application of neuroendoscopy to intraventricular lesions. Neurosurgery
62(Suppl 2):575597; discussion 5978
Cinalli G, Spennato P, Savarese L, Ruggiero C, Aliberti F,
Cuomo L, Cianciulli E, Maggi G (2006) The role of
transventricular neuroendoscopy in the management

of craniopharyngiomas: three patient reports and
review of the literature. J Pediatr Endocrinol Metab
19(Suppl 1):341354
Cipri S, Gangemi A, Cafarelli F, Messina G, Iacopino P,
Al Sayyad S, Capua A, Comi M, Musitano A (2005)
Neuroendoscopic management of hydrocephalus
secondary to midline and pineal lesions. J Neurosurg
Sci 49(3):97106
Dandy WE (1922) Cerebral ventriculoscopy. Bull Johns
Hopkins Hosp 33:189
Depreitere B, Dasi N, Rutka J, Dirks P, Drake J (2007)
Endoscopic biopsy for intraventricular tumors in
children. J Neurosurg Pediatr 106(5 Suppl):340346
Gaab MR, Schroeder HW (1998) Neuroendoscopic
approach to intraventricular lesions. J Neurosurg
88(3):496505
Guiot G, Comoy C (1963) Ventriculoscopy. Rev Prat
13:36553656
Harter D, Omeis I, Forman S, Braun A (2006) Endoscopic
resection of an intraventricular dysembryoplastic
neuroepithelial tumor of the septum pellucidum.
Pediatr Neurosurg 42(2):105107
Idris Z, Hallaert G, Vanhauwaert D, Kalala JR, Dewaele F,
Baert E, Roost DV, Caemaert J (2008) Frameless
neuronavigation-guided endoscopic total en-bloc
removal of a third ventricular colloid cyst: a case
report on surgical technique. Minim Invasive
Neurosurg 51(3):173177
Javadpour M, Mallucci C (2004) The role of neuroendoscopy in the management of tectal gliomas. Childs
Nerv Syst 20(1112):852857
Klimo P Jr, Goumnerova LC (2006) Endoscopic third ventriculocisternostomy for brainstem tumors. J Neurosurg
105(4 Suppl):271274
Lekovic GP, Gonzalez LF, Feiz-Erfan I, Rekate HL (2006)
Endoscopic resection of hypothalamic hamartoma
using a novel variable aspiration tissue resector.
Neurosurgery 58(1 Suppl):ONS166ONS169; discussion ONS166-9
Longatti P, Godano U, Gangemi M, Delitala A, Morace E,
Genitori L, Alafaci C, Benvenuti L, Brunori A, Cereda
C, Cipri S, Fiorindi A, Giordano F, Mascari C, Oppido
PA, Perin A, Tripodi M (2006) Cooperative study by
the Italian neuroendoscopy group on the treatment of
61 colloid cysts. Childs Nerv Syst 22(10):12631267
Luther N, Souweidane MM (2005) Neuroendoscopic
resection of posterior third ventricular ependymoma.
Case report. Neurosurg Focus 18(6A):E3
Luther N, Cohen A, Souweidane MM (2005) Hemorrhagic
sequelae from intracranial neuroendoscopic procedures for intraventricular tumors. Neurosurg Focus
19(1):E9
Macarthur DC, Buxton N, Punt J, Vloeberghs M,
Robertson IJ (2002) The role of neuroendoscopy in the
management of brain tumours. Br J Neurosurg
16(5):465470
OBrien DF, Hayhurst C, Pizer B, Mallucci CL (2006)
Outcomes in patients undergoing single-trajectory
endoscopic third ventriculostomy and endoscopic
15
Pediatric Intraventricular Brain Tumors: Endoscopic Neurosurgical Techniques
biopsy for midline tumors presenting with obstructive

hydrocephalus. J Neurosurg 105(3 Suppl):219226
Oertel J, Krauss JK, Gaab MR (2008) Ultrasonic aspiration in neuroendoscopy: first results with a new tool.
J Neurosurg 109(5):908911
Ray P, Jallo GI, Kim RY, Kim BS, Wilson S, Kothbauer K,
Abbott R (2005) Endoscopic third ventriculostomy for
tumor-related hydrocephalus in a pediatric population.
Neurosurg Focus 19(6):E8
Sandberg DI (2008) Endoscopic management of hydrocephalus in pediatric patients: a review of indications, techniques and outcomes. J Child Neurol 23(5):550560
Schroeder HW, Oertel J, Gaab M (2004) Incidence of
complications in neuroendoscopic surgery. Childs
Nerv Syst 20(1112):878883
Souweidane MM (2005a) Endoscopic management of
pediatric brain tumors. Neurosurg Focus 18(6A):E1
Souweidane MM (2005b) Endoscopic surgery for
intraventricular brain tumors in patients without
163
hydrocephalus. Neurosurgery 57(4 Suppl):312318;

discussion 3128
Souweidane MM, Luther N (2006) Endoscopic resection
of solid intraventricular brain tumors. J Neurosurg
105(2):271278
Souweidane MM, Sandberg DI, Bilsky MH, Gutin PH
(2000) Endoscopic biopsy for tumors of the third
ventricle. Pediatr Neurosurg 33(3):132137
Stachura K, Libionka W, Czepko R (2005) The use of
neuroendoscopy in the treatment of intraventricular
and paraventricular brain tumors. Neurol Neurochir
Pol 39(2):101107
Tirakotai W, Hellwig D, Bertalanffy H, Riegel T (2007)
The role of neuroendoscopy in the management of
solid or solid-cystic intra- and periventricular tumours.
Childs Nerv Syst 23(6):653658
Yurtseven T, Ersahin Y, Demirtas E, Mutluer S (2003)
Neuroendoscopic biopsy for intraventricular tumors.
Minim Invasive Neurosurg 46(5):293299
Neurosurgical Management
of Pediatric Brain Tumors
16
Mehdi Shahideh, George M. Ibrahim,

and James T. Rutka
Contents
Abstract
Introduction ............................................................
165
Low-Grade Gliomas...............................................
Management.............................................................
Prognostic Factors ....................................................
166
167
168
High-Grade Gliomas..............................................
Management.............................................................
Prognostic Factors ....................................................
168
169
169
Craniopharyngioma...............................................
Management.............................................................
Prognosis ..................................................................
169
170
171
Ependymoma ..........................................................
Management.............................................................
Prognosis ..................................................................
172
173
174
Medulloblastoma ....................................................
Management.............................................................
Prognosis ..................................................................
174
175
175
References ...............................................................
176
The management of pediatric central nervous

system neoplasms (CNS) poses a formidable
challenge to clinicians. The purpose of this
chapter is to provide a broad overview of
commonly encountered neoplasms afflicting
children: low and high grade gliomas, craniopharyngiomas, ependymomas, and medulloblastomas. Special emphasis is placed on the
role of the neurosurgeon in the treatment of
these conditions along with an overview of the
most recent diagnostic modalities. Operative
intervention in many of these neoplasms has
the potential to confer cure and/or survival
advantage in addition to providing definitive
pathological diagnosis. Evolving technologies
and medical therapies, which have had a notable impact on patient care and quality of life
are also discussed in this chapter. Finally, a
discussion of patient prognosis is provided for
each neoplasm along with factors directly
affecting outcome.
Introduction
M. Shahideh G.M. Ibrahim J.T. Rutka (*)

Division of Neurosurgery, The Hospital for Sick
Children, 555 University Ave,
Toronto, ON, Canada
e-mail: James.rutka@sickkids.ca
Primary central nervous system (CNS) neoplasms

are the most prevalent solid tumor in children
with an incidence of 3.5 cases per 100,000 per
year (Bondy et al. 2008). They are second only to
leukemia as a cause of pediatric malignancies.
These tumors demonstrate immense heterogeneity

165
M. Shahideh et al.
166
in course and aggression and can range from

curable to universally fatal. CNS neoplasms are
an important cause of nonaccidental death in
children and their incidence seems to be increasing (Bondy et al. 2008). The majority of these
tumors are infratentorial, but many have been
described throughout the neuraxis. Although the
biology of these tumors remains uncertain, ongoing research has made notable progress in the
understanding of their tumorigenesis. Five percent to 10% of childhood brain tumors are associated with a heritable syndrome. These include:
neurofibromatosis types 1 and 2, Turcot, nevoid
basal cell carcinoma syndrome, tuberous sclerosis, Li-Fraumeni, familial adenomatous polyposis and Von Hippel Lindau.
Diagnosis of pediatric brain tumors may be
difficult because of non-localizing symptoms and
the challenges of a full neurological exam in this
population. A high index of suspicion, appropriate investigations and referral are essential to
diagnosis. Several diagnostic imaging modalities
have facilitated the early diagnosis of childhood
tumors and allowed for a better appreciation of
tumor extent, behavior and recurrence. Magnetic
resonance imaging (MRI), positron emission
tomography (PET), magnetoencephalography
(MEG), diffusion tensor imaging (DTI), and
functional MRI have all been applied to the diagnosis and follow-up of CNS neoplasms. The
safety of neurosurgical intervention for CNS neoplasms has also been markedly improved by the
advent of new technologies such as neuronavigation systems, intraoperative ultrasound and MRI,
cortical mapping and multimodal intraoperative
monitoring. These adjuncts have also extended
the possibility of surgical intervention to previously non-operative tumors. In this chapter, we
focus on common pediatric CNS tumors including
gliomas, craniopharyngiomas, ependymomas and
medulloblastomas. The management of these children with a focus on neurosurgical treatment is
presented. Of notable emphasis is the need for a
multidisciplinary approach to pediatric neoplasms.
This involves the surgeon, medical and radiation
oncologists, physio and occupational therapists,
social works and of course, the patients families.
Low-Grade Gliomas
Pediatric low grade gliomas (LGG) are a heterogenous subgroup of tumors encompassing
lesions of astrocytic, oligodendroglial and
mixed glial-neuronal histology. They are classified as grade I and grade II tumors based on histological features including the extent of
necrosis, mitosis, and endothelial proliferation
(Louis 2007). Grade I corresponds to noninfiltrating pilocytic astrocytoma, while Grade II
LGG are diffuse astrocytomas. These account
for the majority of pediatric supratentorial
gliomas. The outcomes of LGG are better in
children than adults, which may be related to
the frequency with which varying tumor histologies are encountered. Pilocytic astrocytomas are the most common LGG in children
with a peak incidence at 59 years of age. These
may occur anywhere in the neuraxis, but are
mainly found in the cerebellum, optic pathway
and dorsally exophytic brain stem (Fig. 16.1).
Diffuse astrocytomas are more frequently
supratentorial, along the deep midline structures and at the cervicomedullary junction
(Louis 2007). Additionally, the molecular aberrancies in pediatric tumors may be distinct from
those in adult populations. For example, while
the majority of adult LGG harbor TP53 mutations, these are only found in 510% of childhood tumors and are associated with malignant
transformation (Broniscer et al. 2007).
The clinical presentation of children with LGG
is highly variable and based on tumor location.
Occasionally, children will manifest with generalized findings related to intracranial hypertension
(headache, nausea, vomiting, papilledema, large
head circumference, lethargy). Localizing symptoms may include seizures, endocrinopathies or
focal neurological deficit. On MRI imaging, these
lesions often appear hypointense on T1-weighted
sequences, hyperintense on T2-weighted
sequences with little or no enhancement post gadolinium injection. Pilocytic astrocytomas are usually well-circumscribed with an enhancing mural
nodule and cystic component.
16 Neurosurgical Management of Pediatric Brain Tumors
167
Fig. 16.1 Low-grade gliomas may manifest in various

locations: (a) Large optic pathway LGG extending anteriorly and laterally into the temporal lobes bilaterally.
Shown also is a photograph of intraoperative resection.
(b) Coronal contrast-enhanced T1-weighted MRI showing
a minimally enhancing heterogenous mass in the area of
the third ventricle. (c) Coronal T1-weighted MRI section

showing a right mesial temporal LGG. (d) Surgical
approach to LGG involving the splenium of the corpus
callosum. Shown is the interhemispheric approach plan
with neuronavigation guidance emphasizing cortical
venous anatomy
Management
lesions, particularly, those that do not demonstrate

radiographic progression on serial MRI scans.
Also, classic findings on MRI are usually sufficient
to diagnose optic pathway/chiasmatic gliomas.
Peri-operative adjuncts, which have improved the
safety of surgical intervention include functional
The initial goal of surgical intervention for LGG is

for pathological tissue diagnosis. Relative exceptions to this include tumors in locations not amenable to surgery, such as deep-seated brain stem
M. Shahideh et al.
168
brain mapping and intraoperative monitoring.

Often children with non-surgical LGG will undergo
chemotherapy if radiographic or clinical deterioration is documented in an effort to delay radiation
therapy and its associated neurocognitive toxicity.
As extensive a resection as possible with acceptable neurological outcome should generally be
performed (Smoots et al. 1998). In children where
95% gross total resection was achieved based on
the surgeons perceptions or MRI performed
2448 h post-operatively, 5- and 10-year survival
rates ranged from 75% to 100% (Rutka and Kuo
2004; Pollack et al. 1995). In addition to reducing
tumor burden and tissue diagnosis, resection permits management of intracranial hypertension,
prevention of neurological deficits and control of
seizures. The timing of resection is contentious as
there is no evidence for improvement in long-term
survival following early intervention (Cairncross
and Laperriere 1989). Patients who have partial
resections or residual disease may experience
recurrence or tumor progression. The management
of these patients remains controversial. These children have a 57 year progression-free survival
ranging from 45% to 65% (Shaw and Wisoff
2003). While some surgeons advocate secondlook surgery to remove residual tumor, others
favor a watch-and-wait strategy. The use of
adjuvant radiotherapy is also debatable. Conflicting
evidence exists on its impact on progression-free
survival, and it has little effect on overall survival
(Pollack et al. 1995). Often, adjuvant radiotherapy
is reserved for patients with progressive disease
and limited surgical options.
Emerging therapies, such as the use of microsurgery combined with interstitial radiosurgical
I125 seed implantation have demonstrated promising results. Long-term neurocognitive toxicity of
such therapy remains uncertain. As our understanding of the molecular underpinnings of tumor
pathogenesis progress, the potential for targeted
molecular therapy evolves.
Prognostic Factors
Prognostic factors in LGG include tumor histology
and location as well as extent of resection.
Nonpilocytic histology is more highly associated
with progression and malignant transformation.

While children have better prognoses than adults,
the prognostic role of age in children is unclear.
High-Grade Gliomas
High grade gliomas (HGG) encompass 812% of
pediatric central nervous tumors (Bondy et al.
2008). The incidence of HGG among patients
18 years and younger is 0.63 per 100,000 personyears with equal distribution among both genders. These tumors are classified primarily into
anaplastic astrocytoma (WHO grade III) and
glioblastoma multiforme (WHO grade IV). The
majority of these tumors are supratentorial, but
may arise anywhere along the neuraxis. There are
acquired and genetic factors that predispose
children to HGG. The only known acquired risk
factor is prior exposure to radiation. Various
inherited defects in cell cycle regulation have
been linked to childhood HGG including
Li-Fraumeni syndrome, neurofibromatosis type I
(NF-1) and Turcot syndrome. The majority of
childhood HGG however are of unclear etiology.
Children may present with a variety of symptoms, depending on the tumor location. These
may present with signs of intracranial hypertension, focal motor deficits, or with extrapyramidal
symptoms. Seizures are more typical of slowgrowing tumors. Intratumoral hemorrhage may
be a cause of acute deterioration in 510% of
children with HGG (Tamber and Rutka 2003).
Initial imaging with CT may reveal the underlying space occupying lesion, hydrocephalus or an
acute hemorrhage, however, the ideal diagnostic
modality is an MRI scan. High grade gliomas on
MRI may enhance heterogeneously and classically have central necrosis resulting in a ring
enhancing lesion. They are usually hypointense
on T1-weighted sequences, hyperintense on
T2-weighted sequences and are commonly associated with peri-lesional edema on fluid-attenuated
inversion recovery sequences.
Ten percent to 15% of all pediatric CNS
tumors are located in the brainstem, of which diffuse intrinsic gliomas are the most common
(Finlay and Zacharoulis 2005). These lesions are
infiltrative with a typically aggressive clinical
course. Children often present with multiple

cranial nerve palsies, ataxia, or long tract signs
classically 12 months prior to diagnosis (Finlay
and Zacharoulis 2005). These usually appear
hypointense on T1-weighted sequences and
hyperintense on T2 and fluid-attenuated inversion recovery sequences with variable gadolinum
enhancement.
Management
With the exception of diffuse pontine gliomas,
surgical intervention for pathological diagnosis
and maximal safe resection is the initial treatment
for HGG. For diffuse pontine gliomas, classic
history, examination and imaging findings are
considered by most to be sufficient for diagnosis.
Despite advances in surgical technique, radiation
and chemotherapeutic agents, outcomes for children with HGG continue to be poor with 5-year
overall survival ranging from 10% to 20% (Finlay
and Zacharoulis 2005; Broniscer et al. 2006).
Most centres advocate maximal surgical resection followed by focal radiotherapy and chemotherapy in children over the age of 3. Resection
of 90% of more of these tumors confers a survival advantage (Finlay et al. 1995). In children
under the age of 3, various chemotherapeutic
combinations are often used to delay brain irradiation (Dufour et al. 2006). This is often advantageous as younger children are more susceptible to
radiation-related complications including endocrine dysfunction, ototoxicity, neurocognitive
delay and secondary malignancy. Furthermore,
these children may have unique tumor biology
with a more indolent course.
The first study to establish the survival benefit
of chemotherapy compared to radiation alone
showed a 5-year event free survival of 46% in the
chemotherapy group versus 18% in the radiotherapy group alone using weekly vincristine
followed by maintenance cycles of prednisone,
lomustine and vincristine (Sposto et al. 1989).
A subsequent trial with eight chemotherapeutic
agents (so called eight-drugs-in-1-day-regimen)
versus prednisone, lomustine and vincristine in
addition to radiotherapy failed to show superiority
(Finlay et al. 1995). Although temozolomide
169
confers a survival advantage in adults with HGG,

its superiority over standard chemotherapeutic
combinations has not been demonstrated. For
patients with diffuse pontine gliomas, focal
radiotherapy is standard of care. This improves
symptoms in 75% of newly diagnosed children,
however prognosis remains poor with overall
survival less than 1015%. Few therapies are
available for recurrent HGG, and prognosis is
usually guarded. High-dose myeloablative chemotherapy with autologous bone marrow rescue has
been described, but is not widely in use (Finlay et al.
2008). Research at the molecular level with the aim
of identifying a therapeutic target is ongoing.
Prognostic Factors
The degree of resection is one of the most important
prognostic factors in children with HGG as stated
earlier. Also, the tumor histology is a determinant
of outcome with WHO Grade IV tumors fairing
worse than WHO Grade III. Patients under
3 years of age also have improved outcomes
compared to older children (Sanders et al. 2007;
Dufour et al. 2006). This may be because tumors
of younger children have unique molecular composition. Overexpression of p53 is also significantly associated with poor prognosis in HGG
(Pollack et al. 2002). Recognized prognostic factors for diffuse intrinsic gliomas are associated
neurological symptoms and time between onset
of symptoms and diagnosis.
Craniopharyngioma
The first description of a craniopharyngioma is
credited to Zenker for his documented observations in 1857. It was not until 1904 that Erdheim
described what he called hypophysial duct
tumors which in 1932 Cushing termed craniopharyngioma. Today, craniopharyngiomas are
the commonest nonglial tumor of childhood and
represent 613% of all childhood brain tumors
(Garre and Cama 2007). The incidence of this
neoplasm is equal in both genders with a peak
occurring at 515 years. Craniopharyngiomas are
epithelial tumors arising along the path of the
170
craniopharyngeal duct, the canal connecting the

stomodeal ectoderm with the evaginated Rathkes
pouch (Karavitaki et al. 2006). These typically
suprasellar, histologically benign WHO grade I
tumors often demonstrate locally aggressive
behaviour with a tendency to infiltrate nearby
critical parasellar structures. Craniopharyngiomas
tend be cystic in nature with 8499% being predominantly or exclusively cystic over purely solid
lesion (Zhang et al. 2002; Petito et al. 1976). The
cysts may be multiloculated and contain liquid
ranging from machinery oil to shimmering
cholesterol-laden fluid. Articles pertaining to
malignant transformation and intra-tumor bone/
teeth formation have been also been reported.
The clinical manifestations of patients with
craniopharyngiomas varies given its proximity to
visual pathways, pituitary gland, hypothalamus,
third ventricle and major vascular structures
(Muller 2008). Headaches and nausea/vomiting
tend to dominate initial presentation (Muller 2008)
however, visual disturbances (bitemporal hemianopia), growth failure, and psychiatric symptoms
are additional symptoms commonly seen in children (Muller 2008; Karavitaki and Wass 2008).
The imaging workup for craniopharyngiomas
requires utilization of both CT and MRI. CT is
used to detect calcifications, differentiate cystic
and solid tumor components and provide useful
information regarding surrounding bony anatomy.
Administration of contrast results in enhancement
of the solid tumor component as well as cyst capsule. MRI demonstrates iso- or hypointense signals for both the solid and cystic portions of the
tumor on T1 weighted images. T2 weighted images
yield mixed hypo- or hyperintense signal for the
solid component of the tumor and hyperintense
signal for the cystic segment (Fig. 16.2). Once
again, contrast enhancement is appreciated in both
solid and capsule components of the tumor.
Occasionally, angiography may be required in
tumors which encase surrounding vasculature.
Management
The management of patients with craniopharyngiomas is continuously evolving. A multimodal approach prioritizing disease control and
M. Shahideh et al.
preservation of quality of life is necessary.

There are two main pathways typically applied
when dealing with craniopharyngiomas. The first
is attempted gross total resection with or without
an adjunct while the second involves subtotal
resection plus adjunctive measures. The surgical
resection of craniopharyngiomas continues to
be challenging despite modern advancements.
Consequently, attempts at gross total resection
should be limited to carefully selected tumors,
through surgical approaches providing the optimal exposure (Karavitaki et al. 2006). Depending
on surgeon comfort, endoscopic approaches
should be considered for amendable tumors. Cure
has been reported in cases where gross total
resection was possible.
More often than not, patients presents with
purely cystic tumors or those not ideal for gross
total resection. The goal of management in these
scenarios involves reducing cyst size while preventing solid tumor progression through subtotal
resection and available adjuncts. Cystic components can be managed by means of aspiration
either by placement of an Ommaya reservoir or
stereotactic aspiration. A 10-month median cyst
reaccumulation rate of 81% has been noted.
Intracystic radiotherapy utilizing Yttrium-90 or
Phosporus-32 to control cyst progression has
yielded promising results with control rates of
96% and 88% for cystic and partially cystic
tumors respectively (Gopalan et al. 2008). Poor
secondary control of the solid tumor component
is currently a limitation of this modality.
Intracystic chemotherapy utilizing bleomycin
has shown some success; however appropriate
long term studies are still lacking (Karavitaki
et al. 2006). Recently, the use of alpha-interferon
has also been described, and holds considerable
promise.
For patients undergoing subtotal resection,
residual tumor is generally treated with external
beam radiotherapy, stereotactic radiosurgery and
more recently proton beam therapy. Gamma knife
radiotherapy although useful has limitations
related to the radiosensitivity of surrounding
structures with a therapeutic window of 912 Gy
and accompanying size restrictions. The use of
Gamma knife as a primary treatment modality has
been reported with positive results. Stereotactic
171
Fig. 16.2 Pre- and post-surgical images of a 9 year male

with 3 month history of visual decline and headache.
T1-weighted (a) sagittal and (b) coronal MRI scans
depicting a heterogenous, multicystic, supracellar lesion
which is superiorly displacing the optic chiasm and tracts.
Note the cystic component extending as far inferiorly as

the prepontine cisterns as well as the trapped left lateral
ventricle. Post-operative (c) sagittal and (d) coronal
T1-weighted contrast-enhanced MRI shows gross total
resection of the lesion and restoration of CSF flow
radiosurgery provides biological advantages by

allowing higher doses of radiation via fractionated dosing (54 Gy in 30 fractions) and thus is
associated with less morbidity. The efficacy of
proton beam therapy as an adjunct still requires
further investigation however results thus far are
encouraging. Recent literature demonstrates similar tumor control rates with decreased associated
morbidity when comparing gross total resection
to subtotal resection plus adjuncts (Yang et al.
2010). Major complications associated with surgery include endocrine dysfunction requiring prolonged endocrine hormone replacement therapy,
visual disturbances, vascular trauma, and hypothalamic disturbances resulting in eating disorders, apathy, sleep disturbances, and memory
problems. Complications of adjuvant therapies
encompass the aforementioned, albeit to a lesser
extent, with the addition of radiation-attributable

concerns such as vasculitis, radiation necrosis,
edema and the formation of de novo tumors.
Prognosis
The overall 5 and 10 year survival rates for
patients undergoing gross total resection versus
subtotal resection plus adjuvant therapy are 98%
versus 99% and 98% versus 95% respectively
(Yang et al. 2010). The 10-year recurrence-free
survival rates have been reported as 7481% for
gross total resection, 4142% for subtotal resection, and 8390% for subtotal resection plus
adjuvants (Duff 2000). When recurrences do
occur, they do so at mean or median intervals
ranging between 1 and 4.3 years.
M. Shahideh et al.
172
Table 16.1 Prognostic factors for recurrence/regrowth
and poor functional outcome in children with
craniopharyngiomas
Prognostic factors
Recurrence/re-growth
Age (<5)
Size (>4 cm)
Unfavourable site
Subtotal resection
Hypothalamic
involvement
Poor functional outcome

Age (<5)
Site and size
Hypothalamic invasion
Multiple surgical procedures
Aggressive inexperienced
surgeon
Known prognostic factors for risk of recurrence/re-growth and poor functional outcome for
craniopharyngiomas are outlined in Table 16.1.
It is important to note that the best predictor of
survival is absence of recurrence. Overall, the
management of craniopharyngiomas remains
complex involving a multidisciplinary team and
various treatment modalities. Future efforts
should not only focus on long-term tumor control and survival, but also on the reduction of
treatment-related morbidities and preserving
quality of life.
Ependymoma
Ependymomas are glial tumors first described by
Bailey in 1924. They are the third most common
central nervous system tumor in the pediatric
population accounting for 612% of brain tumors
in children and almost 2% of all childhood cancers (Kulkarni et al. 2004). Median age at diagnosis ranges from 4 to 6 years of age, (Zacharoulis
et al. 2008) with a 1.4 times increased predominance in boys. Twenty-five percent to 40% of
cases are diagnosed in children younger than
3 years of age (Zacharoulis et al. 2008). Roughly
90% of pediatric ependymomas are of intracranial
origin with about two-thirds of these occurring in
the posterior fossa. Ependymomas are thought to
arise from the neoplastic transformation of
ependymal or subependymal cells lining the ventricular system and central canal of the spinal
cord. They are generally well demarcated solid
tumors that evolve slowly and displace rather
Table 16.2 World Health Organization (WHO) classification of ependymomas

Grade
I
Classification
Subependymoma
II
III
Myxopapillary
ependymoma
Ependymoma
Anaplastic
ependymoma
Ependymoblastoma
IV
Comments
Rare and often an
incidental finding
Typically presents in the
conus or filum
Most common
Second most common
Extremely rare
than invade adjacent tissue. Ependymomas are

prone to leptomeningeal dissemination in 920%
of afflicted patients (Zacharoulis et al. 2008). The
WHO classifies ependymomas into four subtypes
described in Table 16.2.
Clinical presentation in the pediatric population can be somewhat variable depending on the
location of the tumor. Supratentorial ependymomas frequently obstruct circulation of cerebrospinal fluid thus resulting in signs and symptoms
of intracranial hypertension. With significant
cortical compression, seizures and sensorimotor
findings may occur. For infratentorial tumors,
signs and symptoms of brainstem, cranial nerve
and cerebellar compression may accompany
those of increased intracranial pressure depending on the growth characteristics of the offending lesion. Finally, the manifestation of spinal
cord pathology may range from inconspicuous
findings, such as scoliosis and back pain to
myelopathy, bowel/bladder dysfunction and
paresis.
Diagnostic imaging with CT scan often demonstrates low to isodense lesions with the possibility of cystic changes and areas of calcification.
Administration of contrast results in tumor
enhancement. MRI remains the primary modality
for assessment of ependymomas. Generally,
ependymomas demonstrate low T1 and high T2
signal intensity with an intermediate to high
FLAIR signal. Gadolinium administration yields
marked tumor enhancement. Studies utilizing
magnetic resonance spectroscopy are currently
underway to potentially differentiate various
infratentorial tumors.
173
Fig. 16.3 Pre- and post-operative images of an

ependymoma in a 12 year old female. (a) Sagittal and
(b) coronal non-enhanced T1-weighted MR images showing a large, heterogenous, cystic lesion in the fourth ventricle abutting the brainstem anteriorly and the cerebellum
posteriorly. Note the slight downward displacement of the
cerebellar tonsils. (c) Sagittal non-enhanced T2-weighted

MRI of same lesion illustrating disruption of the natural
flow of CSF with enlarged lateral and third ventricle. (d)
Sagittal non-enhanced T1 weighted post operative MRI
demonstrating gross total resection and restoration of
CSF flow
Management
when attempting gross total resection, with

supratentorial lesions being more amendable to
complete removal in comparison to other locations (Zacharoulis and Moreno 2009). Careful
judgment must be employed in order to balance
the risks of an aggressive resection with patient
outcome and quality of life. Due to the radiosensitive nature of ependymomas, radiotherapy
applied to the tumor bed post-operatively has
shown survival benefit in a number of published
studies. Current dosing ranges from 45 to 60 Gy
with craniospinal radiotherapy applied only in
The current standard of care when managing

ependymomas involves surgical resection followed by post operative radiation therapy to the
surgical bed (Fig. 16.3). The goals of surgery are
to obtain a definitive tissue diagnosis, achieve
gross total resection of the tumor, and re-establish
flow of cerebrospinal fluid. The most important
prognostic indicator of survival post operatively
is the extent of surgical resection (Zacharoulis
et al. 2008). A 4060% rate of success is achieved
M. Shahideh et al.
174
patients who demonstrate disseminated disease.

It is important to note that for children less than
3 years of age, immediate post operative radiation is not typically administered and in its
place, multi-agent chemotherapeutic agents are
utilized.
Unfortunately, ependymomas tend to be quite
chemoresistant in nature and its role in the management of afflicted patients is minimal. There
are multiple ongoing trials investigating new
agents possessing more tumor specific activity
than standard cytotoxic agents. These novel treatments are aimed at tyrosine kinase inhibition and
anti-angiogenic agents with data conveying overall results still pending.
Recurrence tends to be a reality of ependymoma
management with roughly 50% of patients having local recurrence within 1324 months
(Zacharoulis and Moreno 2009). Recurrence is
typically managed with additional surgery followed, when appropriate by further radiation
therapy. More recently, radiosurgery has been
attempted with recurrent lesions meeting specific
criteria. The result of this modality, albeit mixed,
are usually unsuccessful. Overall, survival tends
to be very poor once there is evidence of recurrence (Zacharoulis and Moreno 2009).
Prognosis
The prognostic criteria currently available for
evaluating the survival rates of children with
ependymomas takes into account multiple factors
and may differ amongst institutions. Age has
been deemed the most significant factor influencing survival. It should however be noted that
higher grade tumors are more common in younger
children and utilization of radiotherapy is limited
in those under 3 years of age which contribute
strongly to prognosis. Infratentorial ependymomas also carry a worse prognosis than supratentorial lesions once again partly due to the surgical
challenges unique to lesions. To summarize, age,
tumor location, histological grade, extent of
surgical resection, and use of radiation therapy
all directly affect overall survival. As it stands
typical 5 year overall survival ranges from 25%
to 75% depending on the prognostic criteria

(Zacharoulis and Moreno 2009).
Medulloblastoma
The term medulloblastoma was first coined by
Bailey and Cushing in 1925 after its initial
description by Dr. James Wright in 1910. Today,
medulloblastoma is the most common malignant
brain tumor afflicting children and is classified as
a WHO grade IV tumor. Overall, they account for
approximately 20% of all pediatric central nervous system tumors (1% of all adult brain tumors)
and 40% of all posterior fossa tumors (CBTRUS
20072008). The peak incidence of medulloblastoma is between 5 and 7 years of age with 10%
presenting within the first year of life and 30%
after the first decade (Chan et al. 2000).
Medulloblastomas are cerebellar tumors arising
predominantly from the cerebellar vermis and
occupying the fourth ventricle. Their histogenesis
remains a point of controversy with one hypothesis suggesting the cell of origin as being derived
from the external granular layer of the cerebellum.
Another hypothesis proposes a multi-cellular
lineage based on differential immunoreactivity
studies. Regardless, these tumors behave aggressively invading the surrounding cerebellar tissue
and occasionally the brainstem in 15% of cases.
Medulloblastomas also have a propensity for
spreading throughout the central nervous system
with the possibility of extraneural metastasis
particularly to bone and lymph nodes.
Given their anatomical location, children typically present with signs of raised intracranial
pressure with late findings encompassing those
related to cerebellar/brain stem compression/
invasion (ataxia, nystagmus, cranial nerve findings, long tract sign, etc.). In rare cases, children
may present with back pain or extremity weakness secondary to spinal cord dissemination. The
imaging workup leading to the diagnosis of
medulloblastoma typically begins with a CT scan
of the head. This lesion appears as a solid,
homogeneous, isodense to hyperdense, contrastenhancing midline cerebellar mass. Hydrocephalus
is noted in 8590% of cases. Being the study of
choice, MRI is employed to demonstrate the distinctive characteristics of medulloblastomas. T1

and T2 weighted images exhibit a heterogeneous
hypointense and hyperintense mass respectively
with post gadolinium heterogeneous enhancement.
MRI of the complete neuraxis should be performed
when possible prior to commencement of treatment for staging purposes. More recently, magnetic
resonance spectroscopy is gaining favour as a diagnostic tool in addition to its role in aiding clinical
management as new reports indicate the presence
of elevated absolute taurine concentrations on spectroscopy specific to medulloblastomas.
Management
The management of medulloblastoma includes
surgery, chemotherapy and radiation. Treatment
protocols are based on risk stratification, which
takes into account age at presentation, residual
disease, as well as evidence of disseminated disease. The mainstay therapy remains surgical
gross total resection which has clearly shown
improved survival outcomes. In addition, surgery
provides the added benefit of histological confirmation, restoring the cerebrospinal fluid flow
(whether naturally or via diversion), and assessment of the surrounding subarachnoid space for
signs of dissemination. The aim of total resection
however should not jeopardize patient safety and
neurological function.
Postoperative complications associated with
surgical resection may encompass transient or
permanent brainstem dysfunction and cerebellar
findings. One specific complication intrinsic to
posterior fossa tumor resection in the pediatric
population is cerebellar mutism syndrome. This
phenomenon typically develops within a day or
two after surgery in 824% of patients (Robertson
et al. 2006). This is characterized by mutism,
hypotonia, ataxia and mood disturbances persisting for weeks to months.
Adjuvant radiotherapy exploits the radiosensitive nature of medulloblastomas and is considered standard of care for patients greater than
3 years of age post surgical intervention. Its
use has been associated with improved survival
175
outcomes. Radiation therapy is aimed at destroying

residual cells throughout the neuraxis and can be
administered in various regimes depending on
the institution. Despite ongoing research to establish the lowest optimal regime, the sequelae of
radiotherapy still remains an active concern to
the developing nervous system. As described in
prior sections, complications include lowered
intelligence quotient, endocrine dysfunction,
diffuse or focal demyelination, white matter
necrosis, and secondary tumors. As with other
primary CNS neoplasms, children under the age
of 3 years are generally omitted from or undergo
delayed radiotherapy with the aid of new chemotherapeutic agents such as the Head Start III protocol. Other emerging radiotherapy modalities
being investigated for use with medulloblastomas
include proton beam therapy and radiosurgery.
The use of a proton beam allows for a higher proportion of tumor versus normal tissue distribution
thus limiting damage to surrounding structures
(Mueller and Chang 2009). Radiosurgery
although not a primary treatment modality, can
be successfully used for local tumor control in
patients with residual or recurrent disease.
The current treatment standard in North
America for average-risk stratified patients
(please refer to Prognostic Factors below for definition) generally includes postoperative craniospinal irradiation of 23.4 Gy, plus a boost to the
posterior fossa of 54 Gy followed by a 12 month
course of chemotherapy (Mueller and Chang
2009). Patients stratified to the poor-risk category
typically receive a similar regime with the exception of an increase dose of radiation to the craniospinal axis of 36 Gy (Mueller and Chang 2009).
There are currently no standard protocols for
those in the infant grouping however there are
ongoing trials and investigations.
Prognosis
The clinical criteria used to assign patients into
various prognostic groups as discussed above
are based on age at presentation, presence of
disseminated disease, and extent of postoperative
residual (Mueller and Chang 2009). Patients are
M. Shahideh et al.
176
Table 16.3 Prognostic stratification of children with
medulloblastoma based on age, disease burden and postoperative tumor residual
Average
risk
Poor risk
Infants
Disseminated Postoperative 5 Year

Age disease
residual
survival
85%
>3 Negative
<1.5 cm2
>3
<3
Positive
Negative/
positive
>1.5 cm2
n/a
60%
30%a
Presence of metastatic disease considerably worsens

prognosis
convection-enhanced delivery of chemotherapeutic agents will redefine the role of the surgeon in
the evolving management of CNS neoplasms.
Challenges for clinicians include management of
delayed toxicity of therapy and providing end-oflife care for children with incurable tumors.
Acknowledgement This work was supported by the
Wiley and Jack Beqaj Funds for Neurosurgical Research
at the Hospital for Sick Children.
References
typically divided into three risk categories as
summarized in Table 16.3 with their corresponding 5-year survival (Packer and Vezina 2008).
Tumor progression and recurrence occurs in up
to 30% of children within the first 2 years following treatment (Bowers et al. 2007). Following
recurrence, patients typically succumb to their
illness within a year. As the understanding and
management of medulloblastomas progress, so
does overall survival of those affected. Future
efforts should ideally exam the biological characteristics of medulloblastoma including the cell(s)
of origin and the abnormal pathways involved in
order to dramatically change disease stratification and treatment.
In conclusion, neurosurgeons will likely continue to play a significant role in the management
of pediatric CNS neoplasms. Tissue sampling for
pathological diagnosis is quintessential to the
medical management of these lesions. Furthermore,
surgical resection of tumors often confers cure or
significant survival advantage to children.
Rapid advances in imaging modalities, surgical adjuncts and adjuvant therapies have improved
the diagnosis, safety of surgical intervention and
overall survival of children with CNS neoplasms.
Future directions from a research perspective
include basic science experimentation to better
understand tumor biology, novel biologicallydirected therapies, and techniques to minimize
toxicity of available treatments. The role of evidence-based clinical trials in establishing the efficacy of novel therapies cannot be overemphasized.
Minimally-invasive neurosurgery, functional cortex mapping for maximal resection and direct
Bondy ML, Scheurer ME, Malmer B, Barnholtz-Sloan JS,

Davis FG, Ilyasova D, Kruchko C, McCarthy BJ,
Rajaraman P, Schwartzbaum JA, Sadetzki S, Schlehofer
B, Tihan T, Wiemels JL, Wrensch M, Buffler PA
(2008) Brain tumor epidemiology: consensus from the
Brain Tumor Epidemiology Consortium. Cancer
113:19531968
Bowers DC, Gargan L, Weprin BE, Mulne AF, Elterman
RD, Munoz L, Giller CA, Winick NJ (2007) Impact of
site of tumor recurrence upon survival for children
with recurrent or progressive medulloblastoma.
J Neurosurg 107:510
Broniscer A, Chintagumpala M, Fouladi M, Krasin MJ,
Kocak M, Bowers DC, Iacono LC, Merchant TE,
Stewart CF, Houghton PJ, Kun LE, Ledet D, Gajjar A
(2006) Temozolomide after radiotherapy for newly
diagnosed high-grade glioma and unfavorable lowgrade glioma in children. J Neurooncol 76:313319
Broniscer A, Baker SJ, West AN, Fraser MM, Proko E,
Kocak M, Dalton J, Zambetti GP, Ellison DW, Kun
LE, Gajjar A, Gilbertson RJ, Fuller CE (2007) Clinical
and molecular characteristics of malignant transformation of low-grade glioma in children. J Clin Oncol
25:682689
Cairncross JG, Laperriere NJ (1989) Low-grade glioma.
To treat or not to treat? Arch Neurol 46:12381239
CBTRUS (20072008) Primary brain tumors in the United
States, statistical report (20002004), Chicago
Chan AW, Tarbell NJ, Black PM, Louis DN, Frosch MP,
Ancukiewicz M, Chapman P, Loeffler JS (2000) Adult
medulloblastoma: prognostic factors and patterns of
relapse. Neurosurgery 47:623631; discussion 631622
Duff J (2000) Long-term outcomes for surgically resected
craniopharyngiomas. Neurosurgery 46:291302;
discussion 302295
Dufour C, Grill J, Lellouch-Tubiana A, Puget S,
Chastagner P, Frappaz D, Doz F, Pichon F, Plantaz D,
Gentet JC, Raquin MA, Kalifa C (2006) High-grade
glioma in children under 5 years of age: a chemotherapy
only approach with the BBSFOP protocol. Eur
J Cancer 42:29392945
Finlay JL, Zacharoulis S (2005) The treatment of high
grade gliomas and diffuse intrinsic pontine tumors of

childhood and adolescence: a historical and futuristic
perspective. J Neurooncol 75:253266
Finlay JL, Boyett JM, Yates AJ, Wisoff JH, Milstein JM,
Geyer JR, Bertolone SJ, McGuire P, Cherlow JM,
Tefft M et al (1995) Randomized phase III trial in
childhood high-grade astrocytoma comparing vincristine, lomustine, and prednisone with the eight-drugsin-1-day regimen. Childrens Cancer Group. J Clin
Oncol 13:112123
Finlay JL, Dhall G, Boyett JM, Dunkel IJ, Gardner SL,
Goldman S, Yates AJ, Rosenblum MK, Stanley P,
Zimmerman RA, Wallace D, Pollack IF, Packer RJ
(2008) Myeloablative chemotherapy with autologous
bone marrow rescue in children and adolescents with
recurrent malignant astrocytoma: outcome compared
with conventional chemotherapy: a report from the
Childrens Oncology Group. Pediatr Blood Cancer
51:806811
Garre ML, Cama A (2007) Craniopharyngioma: modern
concepts in pathogenesis and treatment. Curr Opin
Pediatr 19:471479
Gopalan R, Dassoulas K, Rainey J, Sherman JH, Sheehan
JP (2008) Evaluation of the role of Gamma Knife surgery in the treatment of craniopharyngiomas.
Neurosurg Focus 24:E5
Karavitaki N, Wass JA (2008) Craniopharyngiomas.
Endocrinol Metab Clin North Am 37:173193, ixx
Karavitaki N, Cudlip S, Adams CB, Wass JA (2006)
Craniopharyngiomas. Endocr Rev 27:371397
Kulkarni AV, Bouffet E, Drake JM (2004) Ependymal
tumors: brain and spinal tumors of childhood. Arnold,
London
Louis D (2007) WHO classification of tumors of the CNS,
4th edn. World Health Organization, Lyon
Mueller S, Chang S (2009) Pediatric brain tumors: current
treatment strategies and future therapeutic approaches.
Neurotherapeutics 6:570586
Muller HL (2008) Childhood craniopharyngioma. Recent
advances in diagnosis, treatment and follow-up. Horm
Res 69:193202
Packer RJ, Vezina G (2008) Management of and prognosis with medulloblastoma: therapy at a crossroads.
Arch Neurol 65:14191424
Petito CK, DeGirolami U, Earle KM (1976)
Craniopharyngiomas: a clinical and pathological
review. Cancer 37:19441952
Pollack IF, Claassen D, al-Shboul Q, Janosky JE,
Deutsch M (1995) Low-grade gliomas of the cerebral
177
hemispheres in children: an analysis of 71 cases.

Pollack IF, Finkelstein SD, Woods J, Burnham J, Holmes
EJ, Hamilton RL, Yates AJ, Boyett JM, Finlay JL,
Sposto R (2002) Expression of p53 and prognosis in
children with malignant gliomas. N Engl J Med
346:420427
Robertson PL, Muraszko KM, Holmes EJ, Sposto R,
Packer RJ, Gajjar A, Dias MS, Allen JC (2006)
Incidence and severity of postoperative cerebellar
mutism syndrome in children with medulloblastoma: a
prospective study by the Childrens Oncology Group.
Rutka JT, Kuo JS (2004) Pediatric surgical neuro-oncology:
current best care practices and strategies. J Neurooncol
69:139150
Sanders RP, Kocak M, Burger PC, Merchant TE, Gajjar
A, Broniscer A (2007) High-grade astrocytoma in very
young children. Pediatr Blood Cancer 49:888893
Shaw EG, Wisoff JH (2003) Prospective clinical trials of
intracranial low-grade glioma in adults and children.
Neuro-oncology 5:153160
Smoots DW, Geyer JR, Lieberman DM, Berger MS (1998)
Predicting disease progression in childhood cerebellar
astrocytoma. Childs Nerv Syst 14:636648
Sposto R, Ertel IJ, Jenkin RD, Boesel CP, Venes JL,
Ortega JA, Evans AE, Wara W, Hammond D (1989)
The effectiveness of chemotherapy for treatment of
high grade astrocytoma in children: results of a
randomized trial. A report from the Childrens Cancer
Study Group. J Neurooncol 7:165177
Tamber MS, Rutka JT (2003) Pediatric supratentorial
high-grade gliomas. Neurosurg Focus 14:E1
Yang I, Sughrue ME, Rutkowski MJ, Kaur R, Ivan ME,
Aranda D, Barani IJ, Parsa AT (2010)
Craniopharyngioma: a comparison of tumor control
with various treatment strategies. Neurosurg Focus
28:E5
Zacharoulis S, Moreno L (2009) Ependymoma: an update.
J Child Neurol 24:14311438
Zacharoulis S, Ji L, Pollack IF, Duffner P, Geyer R, Grill
J, Schild S, Jaing TH, Massimino M, Finlay J, Sposto
R (2008) Metastatic ependymoma: a multi-institutional
retrospective analysis of prognostic factors. Pediatr
Zhang YQ, Wang CC, Ma ZY (2002) Pediatric craniopharyngiomas: clinicomorphological study of 189
cases. Pediatr Neurosurg 36:8084
Pediatric Brain Tumor Biopsy

or Resection: Use of Postoperative
Nonnarcotic Analgesic Medication
17

Contents
Abstract
Introduction ............................................................
180
Postoperative Nonnarcotic Regimine

Following Brain Tumor Biopsy or Resection.......
180
References ...............................................................
181
R.S. Tubbs (*) M.M. Mortazavi A.A. Cohen-Gadol

Pediatric Neurosurgery, Children Hospital,
1600 7th S, ACC 400, Birmingham, AL 35233, USA
e-mail: Shane.tubbs@chsys.org
Introduction: Recent reports have shown the

efficacy in using scheduled non-narcotic analgesic regimens following cranial and spine
neurosurgery.
Methods: We review our experience and
the literature regarding the use of scheduled
doses of alternating acetaminophen and ibuprofen following craniotomy for tumor biopsy
or resection.
Results: From our institutional experience
with 51 children, postoperative imaging identified nine patients (17.67%) had routine, postoperative blood in the resection cavity per
both radiology and neurosurgical review. One
patient had moderate postoperative bleeding
in the tumor cavity (1.9%). No patient was
symptomatic and no patient required a return
to the operating room. Twenty-eight patients
required postoperative morphine for breakthrough pain (54.9%), 21 of which received
less than three doses (75.0%). Overall, 44 of
51 patients (86.3%) required no or minimal
narcotic medication for pain. A literature
review supports these observations.
Conclusions: A scheduled regimen of nonsteroidal anti-inflammatory drugs given in
alternating doses immediately after craniotomy for tumor biopsy or resection and
throughout hospitalization does not appear to
result in any significant post-operative hemorrhage. It appears that such a regimen may
lessen the need for postoperative narcotics, as

179
180
opioid use following surgery was minimal in

86% of our patients.
Introduction
Due to the many known drawbacks of narcotic
analgesia, our institution has previously evaluated the analgesic efficacy of scheduled non-steroidal anti-inflammatory medication after
suboccipital decompression for Chiari I malformation and lumbar laminotomy for partial dorsal
rhizotomy (Smyth et al. 2004; Tubbs et al. 2007).
Our protocol was based on a pain regimen as first
described by Hudgins and Gilreath (2001) in a
study of 17 children undergoing posterior cranial
fossa decompression for Chiari I malformation.
We have also more recently used such a treatment
in patients undergoing craniotomy for tumor
(Bauer et al. 2010). In this patient population, we
assessed an hourly neurological examination during a post-operative overnight stay in our intensive care unit. As we did not wish to over sedate
these patients with narcotics, adequate pain relief
was a concern as was the potential increased risk
of hemorrhage into the tumor bed due to the use
of a nonsteroidal anti-inflammatory drug (i.e.
ibuprofen).
Postoperative Nonnarcotic
Regimine Following Brain Tumor
Biopsy or Resection
From our previous study (Bauer et al. 2010) 51
children were studied. On postoperative imaging, nine patients (17.67%) had routine, postoperative blood in the resection cavity per both
radiology and neurosurgical review. One patient
had moderate postoperative bleeding in the
tumor cavity (1.9%). No patient was symptomatic and no patient required a return to the operating room. Twenty-eight patients required
postoperative morphine for breakthrough pain
(54.9%), 21 of which received less than three
doses (75.0%). Overall, 44 of 51 patients
(86.3%) required no or minimal narcotic medication for pain.
R.S. Tubbs et al.
Textbooks and review articles routinely state

that ibuprofen should not be used in the immediate post-operative period for patients undergoing
craniotomy for tumor, aneurysm, or hemorrhage
because of the platelet dysfunction ibuprofen has
been shown to cause in laboratory investigations
(Aitkenhead and Smith 1998; Albright et al.
2008; Berkowitz and McDonald 1987; Chiaretti
et al. 2000; Cupitt 1999; Drummond and Patel
2000; Francis 1997; Gottschalk et al. 2007;
Greenberg 2006; Ortiz-Cardona and Bendo 2007;
Quiney et al. 1996; Rahimi et al. 2006; Roberts
2005; Shirley 2000; Stoneham and Walters 1995).
One recent textbook only discusses the use of
narcotics and acetaminophen for postoperative
analgesia, without mentioning ibuprofen
(Albright et al. 2008).
There are many known drawbacks to the use
of narcotic analgesia, including potential suppression of respiration, nausea, pruritis, emesis,
urinary retention, euphoria, and constipation
(Aitkenhead and Smith 1998; Albright et al.
2008; Berkowitz and McDonald 1987; Chiaretti
et al. 2000; Cupitt 1999; Drummond and Patel
2000; Francis 1997; Gottschalk et al. 2007;
Gladding et al. 2008; Greenberg 2006; OrtizCardona and Bendo 2007; Quiney et al. 1996;
Rahimi et al. 2006; Roberts 2005; Shirley 2000;
Smyth et al. 2004; Stoneham and Walters 1995;
Verchere et al. 2002). Furthermore, the neurological examination may be clouded with opioid
use. Because of these drawbacks, it is prudent to
limit the use of narcotics in post-operative
patients. However, by limiting narcotics, under
treating pain is a significant concern.
Our previous nonopioid treatment of post neurosurgical patients as mentioned above, utilized a
non-steroidal anti-inflammatory medication, ibuprofen. Potential platelet dysfunction has been a
significant barrier in the use of non-steroidal antiinflammatory medications after a craniotomy for
tumor, vascular lesion, or hemorrhage. Multiple
textbooks and review articles have declared that
ibuprofen is contraindicated immediately after a
craniotomy (Aitkenhead and Smith 1998;
Albright et al. 2008; Berkowitz and McDonald
1987; Chiaretti et al. 2000; Cupitt 1999;
Drummond and Patel 2000; Francis 1997;
17
Pediatric Brain Tumor Biopsy or Resection: Use of Postoperative Nonnarcotic Analgesic Medication
Gottschalk et al. 2007; Greenberg 2006; OrtizCardona and Bendo 2007; Quiney et al. 1996;
Rahimi et al. 2006; Roberts 2005). These articles
reference less recent publications that use older
platelet functionality studies. Based on our review
and clinical experience, we feel that the platelet
dysfunction caused by ibuprofen is not clinically
significant. A recent publication by Gladding
et al. (2008) analyzed platelet function ex vivo
with a Platelet Function Analyzer 100 closure
time test in 24 healthy subjects. In this study, ibuprofen was not found to cause platelet dysfunction. Another study by Bozzo et al. (2001) using
a different platelet function study showed that
aspirin had a significant effect on platelet function, ketorolac had a moderate effect on platelet
function, and ibuprofen had a minimal effect on
platelet function.
Moreover, spine surgeons and general surgeons routinely give ibuprofen and ketorolac
post-operatively without an apparent increased
risk of hemorrhage (Aitkenhead and Smith 1998;
Albright et al. 2008; Drummond and Patel 2000;
Roberts 2005). There is a single report of ketorolac used after pediatric heart surgery where no
increased risk of bleeding was found (Gupta
et al. 2005).
While mounting evidence appears to exist
that ibuprofen is safe to administer to patients
immediately after a craniotomy, no prospective
trial has provided evidence of this. We do not
currently recommend this analgesic regimen as
standard of care, rather we believe we have
shown it to be reasonable safe in a retrospective
review to allow us to perform a prospective trial
at our institution (Bauer et al. 2010). We acknowledge that our previous study lacked a control
group, post operative imaging within 24 h of surgery in every patient, and standardized pain outcome tools such as the visual analog scale (Smyth
et al. 2004).
We found that a scheduled post-operative regimen of ibuprofen and acetaminophen caused no
significant post-operative hemorrhage in children
undergoing craniotomy for tumor biopsy or
resection (Bauer et al. 2010). It appears that such
a regimen lessens the need for post-operative narcotics as opioid use following surgery was
181
minimal in 86% of these patients, and no patient

was discharged on narcotic medication. We realize the lack of a control group and the lack of a
validated pain outcome tools like the Visual
Analog Scale limits the conclusions that can be
made regarding pain control. We plan a prospective trial to confirm the safety and efficacy of this
analgesic regimen.
References
Aitkenhead A, Smith G (1998) Neurosurgical anaesthesia,
3rd edn. Churchill Livingstone, New York
Albright AL, Pollack IF, Adelson PD (2008) Principles
and practice of pediatric neurosurgery, 2nd edn.
Thieme, New York
Bauer DF, Waters AM, Tubbs RS, Rozzelle CJ, Wellons
JC, Blount JP, Oakes WJ (2010) Safety and utility of
scheduled nonnarcotic analgesic medications in children undergoing craniotomy for brain tumor.
Neurosurgery 67:353355
Berkowitz RA, McDonald TB (1987) Post-operative pain
management. Indian J Pediatr 64:351367
Bozzo J, Escolar G, Hernandez MR, Galan AM, Ordinas
A (2001) Prohemorrhagic potential of dipyrone, ibuprofen, ketorolac, and aspirin: mechanisms associated
with blood flow and erythrocyte deformability. J
Cardiovasc Pharmacol 38:183190
Chiaretti A, Viola L, Piewtrini D (2000) Preemptive analgesia with tramadol and fentanyl in pediatric neurosurgery. Childs Nerv Syst 16:93100
Cupitt JM (1999) Pain and opiates following elective
craniotomy. Anaesthesia 54:1299
Drummond J, Patel P (2000) Neurosurgical anaesthesia.
In: Miller R (ed) Anaesthesia, vol 2, 5th edn. Churchill
Livingstone, New York
Francis GA (1997) Pain following craniotomy. Anaesthesia
52:604605
Gladding PA, Webster MWI, Farrell HB, Zeng ISL, Park
R, Ruijne N (2008) The antiplatelet effect of six nonsteroidal anti-inflammatory drugs and their pharmacodynamic interaction with aspirin in healthy volunteers.
Am J Cardiol 101:10601063
Gottschalk A, Berkow LC, Stevens RD (2007) Prospective
evaluation of pain and analgesic use following major
elective intracranial surgery. J Neurosurg 106:210216
Greenberg MS (2006) Handbook of neurosurgery, 6th
edn. Thieme, New York
Gupta A, Daggett C, Ludwick J, Wells W, Lewis A (2005)
Ketorolac after congenital heart surgery: does it
increase the risk of significant bleeding complications? Pediatr Anesthesia 15:139142
Hudgins RJ, Gilreath L (2001) Chiari 1 decompression as
an outpatient procedure. In: American society of pediatric neurosurgeons scientific program, 24th annual
meeting. Maui, HI
182
Ortiz-Cardona J, Bendo AA (2007) Perioperative pain
management in the neurosurgical patient. Anesthesiol
Clin 25:655674
Quiney N, Cooper R, Stoneham M, Walters F (1996) Pain
after craniotomy. A time for reappraisal? Br J
Neurosurg 10:295299
Rahimi SY, Vender JR, Macomson SD, French A, Smith
JR, Alleyne CH (2006) Postoperative pain management after craniotomy: evaluation and cost analysis.
Roberts GC (2005) Post-craniotomy analgesia: current
practices in British neurosurgical centres a survey of
post-craniotomy analgesic practices. Eur J Anaesthesiol
22:328332
Shirley P (2000) Pain relief post craniotomy: a balanced
approach? Anaesthesia 55:409410
R.S. Tubbs et al.

Smyth MD, Banks JT, Tubbs RS, Wellons JC, Oakes WJ
(2004) Efficacy of scheduled nonnarcotic analgesic
mediations in children after suboccipital craniectomy.
J Neurosurg 100(2 Suppl):183186
Stoneham M, Walters F (1995) Post operative
analgesia for craniotomy patients: current attitudes
among neuroanaesthetists. Eur J Anaesthesiol
2:571573
Tubbs RS, Law C, Davis D, Oakes WJ (2007) Scheduled
oral analgesics and the need for opiates in children
following partial dorsal rhizotomy. J Neurosurg 106
(6 suppl):439440
Verchere E, Grenier B, Mesli A, Siao D, Sesay M,
Maurette P (2002) Postoperative pain management
after supratentorial craniotomy. J Neurosurg
Anesthesiol 14:96101
Clinical Trials in Pediatric Brain

Tumors; Radiotherapy
18
Anna Skowronska-Gardas, Marzanna Chojnacka,

and Katarzyna Pedziwiatr
Contents
Abstract
Introduction ............................................................
183
Medulloblastoma ....................................................
184
Ependymoma ..........................................................
188
Low-Grade Glioma ................................................
190
High-Grade Glioma and Brain Stem Glioma......
191
Craniopharyngioma...............................................
192
Germ-Cell Tumors .................................................
192
Conclusions .............................................................
194
References ...............................................................
195
Irradiation is an essential part of the management of most pediatric brain tumors. A good
treatment strategy should consider not only
survival but also the quality of life.
The modern radiotherapy techniques as
conformal and stereotactic radiotherapy, intensity modulated radiotherapy and recently new
radiation modalities as protons, or light ions
allow obtaining a good dose distribution.
Numerous clinical trials with radiotherapy in
pediatric brain tumors as medulloblastoma,
ependymoma, astrocytoma and germ-cell
tumors were undertaken in recent years. This
chapter summarizes the results achieved in
some of them, published in the last 5 years.
Introduction
A. Skowronska-Gardas (*) M. Chojnacka

K. Pedziwiatr
Department of Radiotherapy, M. Sklodowska-Curie
e-mail: a.gardas@chello.pl
Despite the development in neurosurgery and the

introduction of new drugs in chemotherapy, irradiation is an essential part of the management in
most pediatric brain tumors. Current explorations
of sophisticated radiation technologies include
the use of 3-D neuroimaging for accurate target
localization, and improved conformality dose to
the intended target. Among radiation related toxicities, the impact of irradiation on cognitive
function and learning affects most the quality of
life. Currently radiation therapy is going through
a dynamic period. For more accurate outlining of
the target volumes and organs at risk, new

183
A. Skowronska-Gardas et al.
184
imaging tools such as spiral CT, MRI, PET, and

PET-CT have been introduced. Advanced techniques for radiation therapy treatment planning
and delivery have been implemented for clinical
use, with the aim of reducing radiation-related
side effects and make radiotherapy safer for children of all ages.
Conformal radiotherapy techniques allow the
application of a uniform high dose of irradiation
in the region of the tumor while minimizing dose
to normal tissues. In the intensity modulated radiation therapy (IMRT) with inverse dose planning,
the dose planner sets certain criteria and then
optimizes the direction of the beam and the intensity distribution. IMRT is capable of generating
complex 3-D dose distribution to conform closely
to the target volume, and produce significant dose
gradients between the target volume and adjacent
tissue structures. A good control of treatment
delivery should be applied as verification films,
or the electronic portal imaging devices (EPID)
should be used to measure the beam transmitted
through the patient, for these sophisticated methods of planning. The most advanced form of verification imaging and radiation control is obtained
by 3-D image-guided radiotherapy (IGRT). This
means that a CT image is created utilizing the
accelerator photons or on board kilo voltage
CT, and gives better quality images.
The aim of stereotactically guided conformal
radiotherapy (SCRT) is the delivery of highly
conformal dose distribution to the target volume,
using conventional fractionation schemes.
Stereotactic radiosurgery (SRS) utilizes precise
head immobilization, CT or MRI imaging and
multiple intersecting beams to deliver a large
single fraction of radiation to a small clinical
target volume (CTV). This high-precision stereotactic radiotherapy was applied for children
with localized low-grade glioma, and provides
effective local control with acceptable toxicity
for patients with recurrent primitive neuroectodermal tumors (PNETs). The value of stereotactic radiosurgery for the treatment of brain tumors
in children is well recognized; however, the wide
spread use of stereotactic radiosurgery in pediatrics has been limited due to difficulties with rigid
fixation for young children, the need for general
anesthesia and the risk of radionecrosis. Cyber

knife radiosurgery is a system of both frameless
and accurate, treatment of infants and young
children, and offers the possibility of using hypofractionated regimens and reducing the need for
general anesthesia.
Further improvements in dose distributions
require new radiation modalities: protons and/or
light ions. In some cases proton therapy could
give a further reduction in the dose to healthy tissues compared with photon or electron therapy,
especially when intensity modulated proton therapy (IMPT) is applied. The feasibility and the
potential advantage of spot-scanning proton therapy for cranio-spinal irradiation were recently
investigated. This technique appears to be feasible and safe, and difficulties of multiple-field
patching can be avoided.
Numerous clinical trials with radiotherapy in
pediatric brain tumors were undertaken in recent
years. The aim of this chapter is to review the
results of some of these trials, published in the
last 5 years, concerning the survival and/or radiotherapy technique. A comprehensive search
strategy through the indexed database Pubmed
was performed using search terms that included:
brain tumors, radiotherapy, children, and
clinical trials. Relevant articles published in
English between 2005 and May 2010 were
included.
Medulloblastoma
Medulloblastoma accounts for ~20% of brain
tumors occurring in the pediatric age group.
Current management strategies are based on
results of series of multicenter randomized clinical trials undertaken in North America by the
Childrens Oncology Group (COG) and in
Europe by the International Society of Pediatric
Oncology (SIOP). For children with an average/
standard risk medulloblastoma in order to reduce
the late effects of radiation without reducing survival, a few studies have been undertaken with a
reduced dose radiotherapy and adjuvant chemotherapy. The (COG) presented results of phase
III study of 421 children treated with 23.4 Gy
18 Clinical Trials in Pediatric Brain Tumors; Radiotherapy
craniospinal irradiation, 55.8 Gy posterior fossa

radiotherapy, plus one of two adjuvant chemotherapy regimes (lomustin, cisplatin and vincristin, or cyclophosphamid, cisplatin and vincristin).
Five-year event-free and overall survival for all
patients was 81% and 86%, respectively. Only
the neuroradiographic inaccessibility had detrimental impact on event-free survival. Patients
with frank dissemination had a 5-year event free
survival (EFS) of 36%, and 67% of failures had
some component of disseminations. This study
confirmed a possibility of reduction craniospinal
radiotherapy dose in adequately staged patients.
(Packer et al. 2006).
A similar study with primary objective to
decrease the late effects of prophylactic irradiation without reducing survival in standard-risk
childhood medulloblastoma was undertaken by
French Society for Pediatric Oncology (FSPO).
Between 1991 and 1998, 136 patients with total
or subtotal tumor resection with no metastases
and negative cranio-spinal fluid (CSF) cytology
were included in this study. Chemotherapy with
eight drugs in 1 day, followed by two courses of
etoposide plus carboplatin was administered after
surgery. Radiation therapy was delivered 90 days
after surgery, in doses of 55 Gy to the posterior
fossa and 25 Gy to the neuraxis. The overall survival and 5-year recurrence-free survival rate
were 73.8% and 64.8%, respectively. Fifty seven
percent of relapses included posterior fossa, 51%
the spine, and 53% the brain. Radiological review
showed that 4% of patients were erroneously
included due to the presence of definite residual
tumor. The correlation between the presence of a
major protocol deviation and treatment failure
was demonstrated, but these differences were not
statistically significant. However, the quality of
radiotherapy had a prognostic value in the overall
population of children with standard-risk
medulloblastoma, survival for patients with 2
major protocol deviations was significantly worse
than for those with 1 (p < 0.04). In addition, the
prognostic value of CSF cytology was a subject
of debate. Both negative MRI and CSF cytological studies are necessary to confirm standard risk
medulloblastoma. In conclusion, the reduced
dose of radiotherapy with chemotherapy can be
185
proposed for adequately included average-risk

medulloblastoma patients. (Oyharcabal-Bourden
et al. 2005).
A similar objective to limit neurocognitive
sequelae of radiotherapy was undertaken by
Merchant et al. (2008) in the multi-institutional
study. Conformal radiotherapy to less than entire
posterior fossa after craniospinal irradiation may
reduce neurocognitive sequelae for patients with
average-risk medulloblastoma. From 1996 to
2003, 86 patients in age 321 years with this
diagnosis were introduced to a prospective, multiinstitutional trial of risk-adopted radiotherapy
and dose-intensive chemotherapy. Radiotherapy
consisted of dose-reduced Craniospinal irradiation (CSI) (23.4 Gy), conformal posterior fossa
radiotherapy (36 Gy), and primary site radiotherapy (55.8 Gy). The target volume for the primary
site included the postoperative tumor bed with a
margin of 2 cm. Chemotherapy started 6 weeks
after radiotherapy and included four cycles of
high-dose cyclophosphamide, cisplatin, and vincristine, with the peripheral stem-cell support.
The estimated 5-year EFS and local control rates
were 83% and 91.7%, respectively. The cumulative incidence of posterior fossa failure was 4.9%.
Lower mean doses were given for temporal lobes,
cochleae, brain stem, and hypothalamus. The
authors concluded that irradiation of less than the
entire posterior fossa after reduced-dose CSI for
average risk medulloblastoma results in disease
control comparable to treatment of the entire posterior fossa (Merchant et al. 2008).
Hyperfractionated radiotherapy is believed to
increase biologic equivalent dose to the tumor
without increasing the biologic equivalent dose
to the normal tissue. In the French M-SFOP98
protocol, reported by Carrie et al. (2009), 48
patients with standard-risk medulloblastoma in
age 518 years were investigated. They received
hyperfractionated radiotherapy (36 Gy in 36 fractions twice a day) to the craniospinal axis followed by a boost to the tumor bed with 1.5 cm
margin, to a total dose of 68 Gy. None of patients
received any chemotherapy. The overall survival
and event-free survival after 6 years were 78% and
75% respectively. After 77 month of follow-up,
13 relapses were observed, but no relapse
186
occurred within the posterior fossa outside the

boost volume. Annual full-scale IQ decline was
two points over a 6-year period, and 50% of children had FSIQ scores within the normal range.
In conclusion, this prospective study resulted
in a very good event-free survival and in promising results on neurocognitive evaluation (Carrie
et al. 2009).
German authors presented results of the prospective randomized multicentre trial HIT 91,
with long-term outcome and clinical prognostic
factors in children with medulloblastoma. They
analyzed 280 patients of age 318 years, enrolled
from 1991 to 1997 to the randomized trial, comparing preradiation (sandwich) and postradiation (maintenance) chemotherapy, with a median
follow-up of 10 years. All patients received craniospinal irradiation of 35.2 Gy followed by a
boost to the posterior fossa of 55.2 Gy. Metastatic
deposits in the spine and supratentorial region
were boosted up to a total dose of 50 Gy.
Vincristine was given weekly concomitantly with
radiotherapy. Maintenance chemotherapy consisted of eight cycles with CCNU, vincristine,
and cisplatine. Sandwich chemotherapy consisted
of two courses, each containing four cycles of
chemotherapy (ifosfamide, etoposide, high-dose
methotrexate, cisplatin, cytarabine). The study
showed a benefit of the maintenance strategy for
all randomized children except patients with
M2/3 stages. The overall survival after maintenance compared to sandwich treatment was for
M0 patients 91% and 62%; for M1 patients 70%
and 34% and for M2/3 patients was 42% and
45%, respectively. The long-term analysis failed
to show superiority of either treatment strategy.
This large randomized study confirms that immediate radiotherapy after surgery is likely to be
crucial for optimal disease control in localized
disease (Von Hoff et al. 2009).
Outcomes for recently treated standard-risk
medulloblastoma patients are good, with eventfree survival rates of at least 7080%. However,
the outcome for patients with M2-3 disease
remains relatively poor. During the time of PNET-3
Study, it was recommended that patients with
M2-3 disease should be treated with chemotherapy
followed by full dose radiotherapy. The outcome
for these patients has been analyzed. Between

1992 and 2000, 68 patients received chemotherapy
comprising vincristine, etoposide, carboplatin, and
cyclophoshamid, followed by craniospinal radiotherapy of 35 Gy/21 fractions with a posterior
fossa boost of 20 Gy/12 fractions. Thirty-five percent irradiated patients received a metastasis boost
(mean dose 47 Gy). Overall survival rates at 3 and
5 years were 50% and 43%, respectively, eventfree survival were 39% and 34.7%, respectively.
Patients who had disease progression during chemotherapy had a significantly worse outcome than
those who had stable or responding disease (overall survival 12% vs. 51%); however, these comparisons were based on small patient numbers.
The use of preirradiation chemotherapy appears to
identify a poor prognostic group with early progression, but newer approaches such as more
intensive chemotherapy and radiotherapy need to
be explored (Taylor et al. 2005).
Hyperfractionated radiotherapy was used not
only for standard-risk patients but also in patients
with metastatic medulloblastoma. The COG
studied pre-irradiation chemotherapy followed
by hyperfractionated radiotherapy for children
with high-risk medulloblastoma/PNETs. The
feasibility and preliminary results of phase II in
this trial were presented by Allen et al. (2009).
The study included patients >3 years at diagnosis
with medulloblastoma M2-3, and >1.5 cm2 postoperative residual disease, and all patients with
noncerebellar PNET. After surgery, patients
received five alternating cycles of chemotherapy
followed by hyperfractionated craniospinal radiotherapy (40 Gy), with a boost to the primary site
(72 Gy), given in twice-daily 1 Gy fractions.
Metastases in the spinal cord were boosted up to
50 Gy and those in the cauda equine region to
56 Gy. The valid study group consisted of 124
patients, 68% of them completed the entire protocol. Three toxic deaths (2.4%) during chemotherapy occurred. The 5-year progression-free
survival and overall survival rates were 43% and
52%, respectively. These results seem to be similar to data from other contemporary, less-intensive
protocols (Allen et al. 2009).
Italian researchers tested the efficacy and toxicity of hyperfractionated accelerated radiotherapy
(HART) delivered after intensive sequential

chemotherapy. Between 1998 and 2007, 33 patients
with metastatic medulloblastoma (M > 0) received
postoperative metotrexat, etoposide, cyclophosphamid, and carboplatin in 2-month schedule, then
HART to the neuraxis of 39 Gy (1.3 Gy per fraction), and a posterior fossa boost up to 60 Gy
(1.5 Gy per fraction), given twice a day. In selected
cases, the myeloablative high-dose chemotherapy
after HART was administrated. The 5-year eventfree, progression- free and overall survival rates
were 70%, 72%, and 73%, respectively. No severe
clinical complications of radiotherapy have been
observed so far. These results compare favorably
with other trials using conventional therapies
(Gandola et al. 2009).
Some reports stress the impact of technical
quality of radiotherapy on survival in patients
with medulloblastoma. FSPO Study established
that the presence of more than one major deviation in radiotherapy protocol had a specific negative prognostic value in the overall population of
children with standard-risk medulloblastoma
(Oyharcabal-Bourden et al. 2005).
The evaluation of the potential influence of
radiotherapy quality on survival in high-risk
medulloblastoma patients was considered in Trial
9031 at the Pediatric Oncology Group. Twohundred and ten patients were treated according
to protocol guidelines and were eligible for analysis. Treatment volume (whole brain, spine, posterior fossa, and primary tumor bed) and dose
prescription deviations were assessed for each
patient. An analysis of the first site of failure was
undertaken. Fifty seven percent of patients had
one or more major deviations in their treatment
schedule. Major deviations by treatment site were:
brain (26%), spinal (7%), posterior fossa (40%),
and primary tumor bed (17%), but they did not
significantly influence overall and event-free survival. More than half of the relapses in the craniospinal axis were diffuse which suggests that even
correctly delivered radiotherapy may not be effective enough to completely treat the high metastatic
potential of high-risk medulloblastoma. Factors
other than craniospinal irradiation parameters
may have positively influenced the outcome of
these patients (Miralbell et al. 2006).
187
Similar observations were done in group of 95

children with medulloblastoma treated with threedimensional conformal radiotherapy in our department. Five-year overall and event-free survival
was 73.3% and 65.8%, respectively. In ~2/3 of
patients, recurrence was multifocal. In all patients
with isolated failures, the recurrence appeared
within the isodose level of 95100%. The portal
films did not show positional errors. The threedimensional conformal radiation therapy allows
avoiding failures related to radiotherapy uncertainties (Skowroska-Gardas et al. 2007).
Twenty five to 35% of cases of medulloblastoma, occur in children younger than 3 years of
age. The prognosis for these children treated with
surgery, radiotherapy, and chemotherapy remains
poor, and survivors are in the high risk for cognitive deficits. Trial conducted in Germany, reported
by Rutkowski et al. (2005) investigated whether
intensive postoperative chemotherapy alone
improves survival and cognitive function in young
children. After surgery children received three
cycles of intravenous chemotherapy (cyclophosphamid, vincristine, methotrexate, carboplatin,
and etoposide) and intraventricular methotrexat
was given instead of radiotherapy. Treatment was
terminated if a complete remission was achieved
after three cycles. Otherwise, children received
second surgery, radiotherapy or experimental chemotherapy as appropriate. Progression-free survival and overall survival were significantly
different among children with or without residual
tumors (82% and 93% vs. 50% and 56%) and
among children with or without macroscopic
metastases (33% and 38% vs. 68% and 77%). At
the end of the designated chemotherapy, 14
patients had residual disease and received further
treatment with chemotherapy, radiotherapy or
both, ten of them died. Tumor relapses occurred
in 9 of 29 patients without residual disease after
chemotherapy. Sixteen patients received further
treatment at progression or relapse; six of them
(38%) are alive after the inclusion of radiotherapy
as the salvage therapy. In 19 of 23 children MRI
detected asymptomatic leukoencephalopathy.
After treatment the results of all IQ tests were
significantly lower than those of healthy controls
within the same age group, but higher than those
188
of patients in a previous trial who had received

radiotherapy (Rutkowski et al. 2005).
The first UKCCSG/SIOP CNS 9204 trial was
undertaken to assess the role of a primary chemotherapy strategy in avoiding or delaying
radiotherapy in children younger than 3 years
with malignant brain tumors. Patients with
ependymoma were not included for this study.
Between 1993 and 2003, 97 children were
enrolled, one-third of them were patients with
medulloblastoma. Following maximal surgical
resection, chemotherapy was delivered for 1 year
or until disease progression. Radiotherapy was
withheld in the absence of progression.
In total, 72 (81%) of patients have progressed,
and subsequently 29 of them were irradiated.
Patients with medulloblastoma presented in
83.9% residual disease and/or metastases at diagnosis. For these patients outcome was related to
histology. The 5-year overall survival for desmoplastic/nodular medulloblastoma was 52.9% and
only 33.3% for standard medulloblastoma; eventfree survival was 35.3% and 33.3%, respectively.
All children with large cell or anaplastic medulloblastoma died within 2 years of diagnosis.
Outcomes for patients with other types of tumors
were worse, overall 5 year survival rates for highgrade glioma, choroid plexus carcinoma, atypical
rhabdoid tumor, and PNETs were 30%, 26%,
16% and 9%, respectively. Authors of this
recently published study concluded that outcome
for very young children with brain tumors
depends on the degree of surgical resection and
histological tumor type. In 45% of patients radiotherapy was avoided. Patients with desmoplastic/
nodular subtype of medulloblastoma, despite the
adverse clinical features of metastatic disease and
incomplete tumor resection, had the best prognosis (Grundy et al. 2010).
Ependymoma
Ependymoma accounts for 10% of CNS tumors
in childhood. The peak incidence is the first
3 years of life, where ependymoma accounts
for up to 30% of all childhood brain tumors.
Surgery and postoperative radiation therapy are
essential for the successful management of

ependymoma, but those who receive radiotherapy are at risk of cognitive, endocrine, and neurological side effects.
In the past, postoperative irradiation of the
craniospinal axis was recommended in patients
with anaplastic ependymoma or in those with
infratentorial tumors. In patients with low-grade
ependymomas, radiotherapy was often performed
as a local treatment. In most cases, the predominate pattern of failure was local. German authors
report results of different radiotherapeutic
approaches in a group of 57 patients with histologically confirmed ependymoma, treated in the
University Hospital in Heidelberg between 1974
and 2006. In this group, 25% of patients
were <4 years of age, and in 53% diagnosis of
anaplastic ependymoma was confirmed. In 16
patients irradiation of the craniospinal axis was
performed, with a median dose of 20 Gy. Fortyone patients were treated with local radiotherapy;
they received 45 Gy to the posterior fossa, including a boost to the tumor bed of 9 Gy. Overall survival was 71% at 5 years, 80% for low-grade and
79% for high-grade tumors. A statistically significant difference in overall survival and a failure-free survival between CSI and local
radiotherapy was not shown. Authors concluded
that local radiotherapy in patients with localized
ependymoma is equally effective for CSI. Patients
with localized ependymoma benefit from local
doses 45 Gy (Combs et al. 2008).
FSPO tested the hyperfractionated radiotherapy for children with ependymoma, with aim to
improve the therapeutic ratio. Between 1996
and 2002, 28 children >5 years of age were
included in the study. The planned dose was
60 Gy after complete resection and 66 Gy after
partial resection, delivered in fractionated doses
of 1 Gy, twice a day. In ten patients, diagnosis of
anaplastic ependymoma was established. Of 24
patients, eight died, all because of their disease.
The estimated 5-year overall and progressionfree survival rates were 62.5% and 54.2%,
respectively. Three-fourths of the patients had
normal psychomotor development. The results
of this study have shown that hyperfractionated
radiotherapy is safe, but provides no outcome
benefit with other strategies of radiotherapy

such as standard fractionated regimes (Conter
et al. 2009).
Ependymoma in childhood has a peak incidence in infants and babies. The current gold
standard for treating ependymoma in early
infancy is hard to define. In the past, surgery followed by craniospinal or focal radiotherapy has
been the standard treatment. Due to the serious
adverse effects of such treatment, alternative
treatment approaches have been introduced by
adding intensive chemotherapy to delay or omit
radiotherapy. However, treatment results have
been disappointing, with survival rates between
20% and 50%. The role of radiotherapy in anaplastic ependymoma in children under age of
3 years was evaluated in the prospective German
Brain Tumor trials HIT-SKK 87 and 92
(Timmermann et al. 2005). Fifteen children were
enrolled to the HIT-SKK 87 trial and 19 to HIT92. In the HIT-SKK 87 trial, low risk patients
(R0M0) after surgery, received maintenance chemotherapy until elective radiotherapy at age of
three was given. In high-risk patients, (R+M+)
intensive induction chemotherapy was followed
by maintenance chemotherapy and delayed radiotherapy. In the case of progression, radiotherapy
started immediately. In the HIT-SKK-92 trial,
MTX-based chemotherapy was applied. Overall
survival and progression-free survival rates after
3-years for all children were 55.9% and 27.3%,
respectively. Three-year progression- free survival for children treated with SKK 87 schedule
was 40% and with SKK-92 was 16.7%. In 13
children, no radiotherapy was administered.
Preventive radiotherapy was given to nine patients
and salvage radiotherapy to 12 children. Only
3/13 children survived without radiotherapy. At
the last follow-up, nine children were free of disease, 25 progressed, and 76% of them failed at
the tumor site only. Authors concluded that the
local control in ependymomas is the most important prognostic factor for treatment outcome.
Omission or long delay of radiotherapy should be
avoided as it jeopardizes survival, even if intensive chemotherapy has been given. In localized
disease, restriction of target volume to the primary tumor site only can achieve tumor control
189
and reduce the risk of long-term toxicity, and

radiotherapy of the neuraxis could be avoided
(Timmermann et al. 2005).
Different approach to applying radiotherapy
in children younger than 3 years was presented in
the prospective study conducted by the UKCCSG/
SIOP group. The aim of this trial was to assess
the role of a primary chemotherapy strategy in
avoiding or delaying radiotherapy in young children with ependymoma. Between 1992 and 2003,
89 children aged 3 years, with no metastatic disease were enrolled to the study. After surgical
resection, children received alternating blocks of
myelosuppressive and nonmyelosuppressive chemotherapy for an intended duration of 1 year.
Radiotherapy was withheld until progression of
tumor. For localized tumor, 50 Gy in 25 fractions
was given, for metastatic disease, whole neuraxis
radiotherapy 35 Gy in 21 fractions, followed by a
boost to the primary tumor 20 Gy in 12 fractions
(total dose 55 Gy), was recommended. For infants
the dose to neuraxis was reduced to 20 Gy, the
total boost dose was the same as for older children. Fifty of the 80 patients with nonmetastatic
disease progressed, 34 of them were irradiated.
The median age at irradiation for the whole group
was 3.6 years. The 5-year cumulative incidence
of freedom from radiotherapy was 42%. The
5-year overall survival for the whole group was
60%. The presented protocol avoided or delayed
radiotherapy in a substantial proportion of children younger than 3 years, without compromising survival. Authors suggest that primary
chemotherapy strategies have an important role
in the treatment of very young children with
intracranial ependymomas (Grundy et al. 2007).
The most promising results of treatment with
surgery and conformal radiotherapy in children
with ependymoma were presented by Merchant
et al. (2009a). The conformal methods of planning and delivering radiotherapy reduced side
effects and improved rates of local tumor control,
event-free survival, and overall survival for these
patients. This treatment approach was extended
to include children under the age of 3 years, with
the aim of improving tumor control. Between
1997 and 2007, 153 pediatric patients with localized ependymoma were treated. The tumors were
190
localized in infratentorial region in 122 cases,

85 were anaplastic, and 35 had received previous
chemotherapy. Forty nine percent of patients
were of age 3 years. These patients received
conformal radiotherapy after definitive surgery in
doses from 54 to 59.4 Gy to the residual tumor
and/or to postoperative tumor bed with a 10 mm
margin.
Seven-year local control, event-free survival
and overall survival were 87.3%, 69.1% and 81%
respectively. The incidence of secondary malignant tumor at 7 years was 2.3% and brainstem
necrosis 1.6%. This study highlights the longterm benefits of gross total resection and high
dose postoperative radiotherapy for the treatment
of children with localized ependymoma, even for
those who are younger than 3 years (Merchant
et al. 2009a).
Low-Grade Glioma
Low-grade gliomas account for 35% of all childhood CNS tumors. Resection remains the preferred treatment option for patients with cerebellar
and cerebral hemispheric tumors and frequently
is curative. However, for patients with midline
and other tumor localization, where total or nearly
total resection is not possible, the most appropriate management remains undefined. Radiotherapy
provides long lasting disease control in a significant percentage of such patients, but the radiotherapy related neurocognitive and neuroendocrine
sequelae, especially in younger patients, have led
many investigators to explore the use of chemotherapy with delayed or omitted radiotherapy.
The effectiveness of novel combination chemotherapy for children with low-grade glioma,
who relapsed after irradiation or showed visual
deterioration, was investigated. Patients were
treated for 18 months with a multi-drug regimen
of vincristine, etoposide, cyclophosphamide and
5-fluorouracil. The progression-free survival was
67.3% (Lee et al. 2006).
For children with incompletely resected symptomatic tumor, older than 58 years, high precision conformal or stereotactic radiotherapy is
recommended. Stereotactic radiotherapy uses
highly focal, precise, fractionated radiotherapy

planning. This form of treatment is made possible
with head fixation devices and modifications of
standard linear accelerator. The goal of stereotactic radiotherapy is to minimize the amount of
normal tissue irradiated without compromising
tumor control. The accuracy and precision of stereotactic radiotherapy can reduce the margin of
normal tissue receiving the prescribed dose.
Marcus et al. ( 2005) presented results of a
prospective trial for 50 children with localized
low-grade gliomas treated with stereotactic
radiotherapy. Disease progression during or after
chemotherapy or progression after surgery alone
were indications for treatment of these patients.
Stereotactic radiotherapy was delivered using a
dedicated 6 MV linear accelerator. Immobilization
was accomplished with a removable head frame.
For treatment planning, CT and MRI fusion was
used. The target volume included the preoperative tumor plus a 2-mm margin for the planning
target volume. The median total dose was 52.2 Gy
in 1.8 Gy daily fractions. The 5 and 8 years progression-free survival was 82.5% and 65%,
respectively; overall survival was 97.8% and
82%, respectively. The local progression was
observed in six patients, and in two of them with
pathologic progression to anaplastic astrocytoma.
Five patients developed CNS dissemination. Six
patients died, three due to dissemination, two of
progression and one because of secondary tumor.
Marginal failures have not been observed. This
study supports the use of stereotactic radiotherapy in the treatment of low-grade gliomas in children (Marcus et al. 2005).
The promising results of the phase II trial of
conformal radiation therapy for pediatric lowgrade glioma was presented by Merchant et al.
(2009b). Between 1997 and 2006, 78 children in
age 219 years (mediana 9 years) after biopsy or
incomplete surgery received 54 Gy of conformal
radiation therapy in 30 fractions. Gross tumor
volume involve the cystic and solid tumor
presented on MRI before radiotherapy, with
additional margin of 10 mm. Tumor locations
were diencephalon, hemisphere and cerebellum.
A median follow-up was 89 months, and 13
patients experienced disease progression. The
5- and 10-year event-free survival was 87% and

74% respectively and overall survival was 98%
and 96% respectively. The incidence of vasculopathy was about 5% at 6 years, and was higher
for patients younger than 5 years of age. Authors
concluded that this conformal irradiation would
reduce adverse effects without effecting the rate
of treatment failure in pediatric low-grade astrocytoma. However, radiotherapy should be delayed
in young patients to reduce the risk of vasculopathy (Merchant et al. 2009b).
In the study by Merchant et al. (2009c), cognitive effects of conformal radiotherapy were correlated with patients age, neurofibromatosis type
1, status, tumor location, extent of resection, and
radiation dose. Patients younger than 5 years
experienced the greatest decline in cognition.
However, adverse effects were limited and predictable for most patients (Merchant et al.
2009c).
High-Grade Glioma and Brain

Stem Glioma
High-grade malignant gliomas represent up to
10% of total pediatric brain tumors. Although
children with malignant gliomas appear to fare a
little better than adults, outcomes for these patients
remain relatively poor, with a median survival of
1542 months. The role of high-dose chemotherapy in the treatment of high-grade glioma remains
unclear. In the nationwide study, reported by
MacDonald et al. (2005), the Children Cancer
Group (USA) prospectively evaluated 102 children with malignant gliomas and postoperative
residual disease for three sets of drugs including
carboplatin/etoposide, ifosfamid/etoposide, and
cyclophosphamid/etoposid, given in four courses
before radiotherapy. Nonhematological serious
toxicities were observed in 29% of patients and
21% did not receive radiotherapy. Overall survival
rate was 24% at 5 years and event-free survival
rate for all patients was 8%. The authors conclude
that these high-dose chemotherapy regimes provide no additional clinical benefit to conventional
treatment in children with high-grade glioma
(MacDonald et al. 2005).
191
German authors tested a hypothesis, that

intensive chemotherapy in protocol HIT-GBM-C
could increase survival of pediatric patients with
high-grade glioma. Ninety-seven patients in
median age of 10 years were treated with standard fractionated irradiation and simultaneous
chemotherapy, including cisplatin, etoposid, vincristine, and ifosfamide, followed by additional
cycles of maintenance chemotherapy. Resection
was completed in 21 patients. Overall survival
rates were 56% and 19% after 12 and 60 months,
respectively. Comparing with previous, less
intensive protocols, there was no significant benefit for patients with residual tumor, but the 5-year
overall survival rate for patients with complete
resection was 63%. Authors concluded that HITGBM-C chemotherapy after complete tumor
resection was similar to previous, less intensive
protocols (Wolff et al. 2010).
The prognosis for children with newly diagnosed diffuse brain stem glioma is very poor. The
altered fractionated (hyperfractionated and accelerated) radiotherapy regimes with or without cisplatinum based chemotherapy failed to show any
clinical benefit for these patients, and conventionally fractionated local radiotherapy remains
the standard of care, leading to temporary clinical
improvement in a substantial percentage of
patients and is associated with a median survival
of 9 months. The role of chemotherapy in the
treatment of children with diffuse brain stem
glioma is not well defined. Multiple trials involving various regimes administered in conjunction
with radiotherapy have not demonstrated any
improvement in the outcome.
Temozolamide is a methylating agent that has
shown promising results in a subset of adult
patients with high-grade glioma and in children
with the same diagnosis. Jalali et al. (2010) presented results of a prospective study of radiotherapy with concurrent and adjuvant temozolamide
in children with diffuse intrinsic pontine glioma.
Between March 2005 and November 2006, 20
children (median age 8.3 years) were accrued to
the study. They were treated with focal radiotherapy to a total dose 54 Gy in 30 fractions, along
with concurrent temozolamide to a maximum of
12 cycles. Eighteen children have died from
192
progressive disease; two are living with progressive or stable disease. Median overall survival
and progression-free survival were 9 and 7 month,
respectively. Authors concluded that temozolamid with radiotherapy has not yielded any
improvement in the outcome of diffuse brain
stem tumors compared with radiotherapy alone
(Jalali et al. 2010).
Combs et al. (2009) presented surprisingly
good results of high-precision radiotherapy in
85 patients with brain-stem gliomas. Thirtyone patients were younger than 18 years.
Histopathological examination confirmed a
low-grade glioma in 57 patients. Radiation
therapy was performed as fractionated stereotactic radiotherapy to a total dose of 54 Gy in conventional fractionation of 1.8 Gy. Overall survival
was 70% at 2-years and 63% at 3-years. Median
progressionfree survival was 52 months. Younger
age had favorable impact on the outcome. The
high precision radiotherapy seems to be very
good option for patients with small, low-grade
brain stem gliomas (Combs et al. 2009).
Craniopharyngioma
Craniopharyngioma constitute up to 810% of
intracranial tumors in children. Although histologically benign, they are locally aggressive and
could invade critical structures such as optic chiasm, pituitary gland, and hypothalamus. Complete
resection is achievable in 7090% of cases, and
6080% remains relapse-free. However, incompletely resected tumors will relapse in ~70% of
cases. In children with recurrent craniopharyngioma, the use of highly conformal radiotherapy
results in very good local control with a low incidence of complications.
The prospective trial of conformal radiotherapy for pediatric patients with craniopharyngioma was conducted to determine whether the
irradiated volume could be safely reduced to
decrease effects on cognitive function. Twentyeight patients received conformal radiotherapy in
doses of 5455.8 Gy administered to the gross
tumor volume, surrounded by a 1 cm clinical target volume margin. Patients were evaluated with
neuropsychometric tests. The estimated 3-years

progression-free survival was 90.3%. The percentage of total brain, supratentorial brain or left
temporal lobe volumes receiving a dose >45 Gy
had a significant impact on longitudinal IQ. The
use of conformal radiation therapy resulted in
tumor control equivalent to that achieved using
conventionally planned radiation therapy
(Merchant et al. 2006).
The long-term results of combined proton and
photon irradiation for craniopharyngioma were
also evaluated. Fifteen patients including five
children with craniopharyngioma were treated in
part or entirely with fractionated 160 MeV proton beam therapy. Patients were treated after
documented recurrence, after initial surgery or
after subtotal resection or biopsy. Median followup was 13 years after radiotherapy. Actuarial
10-years survival rate was 72%, actuarial 5 and
10 years, local control rates were 93% and 85%,
respectively. One child shows learning difficulties and slight retardation, comparable to his preradiotherapy status. The other patients have
professional achievements within the normal
range and continued leading normal or near normal working life. Results in terms of survival and
local control are comparable with other contemporary series (Fitzek et al. 2006).
Germ-Cell Tumors
Intracranial germ-cell tumors are rare, comprise
13% of all primary CNS neoplasms. Based on
the histological components and the degree of
differentiation they are classified as germinomatous and nongerminomatous germ-cell tumors.
Germinomas comprise two-third of CNS
germ-cell tumors, and nongerminoma account
for the remaining third. The nongerminomatous
germ-cell tumors may be composed of elements
of chorioncarcinoma, endodermal sinus or yolk
sac tumor, embryonic carcinoma or teratoma.
Very few prospective studies are available, and
retrospective studies are limited, based on the
low number of patients and variability in tumors
size, location, histology, and treatment applied.
The optimal treatment for intracranial germinomas
remains controversial. Historically, the CSI has

been used to treat these radiosensitive tumors,
which led to long-term survival in >90% of
patients. However, because of the potential late
complications of CSI, there has been a trend in
recent years toward using more limited radiation
fields, either with or without chemotherapy.
The patterns of failure associated with evolving target volume in 21 patients treated with chemotherapy and focal irradiation or with
craniospinal irradiation were compared in retrospective study. Nine patients received chemotherapy prior to focal radiotherapy and 12
patients received craniospinal irradiation. The
overall survival rate at 10 years for all patients
was 86%, 89% for patients who received focal
radiotherapy and 83% for patients who received
CSI. The 10-year local control rate in the brain
for patients who received focal irradiation was
59%, compared with 100% for those who
received CSI; similarly, distant control in the
spine was 62% after focal irradiation and 100%
for CSI. Authors concluded that although focal
irradiation with chemotherapy is attractive methods, this strategy appeared to be associated with
increased rates of failures in the brain and spine
(Nguyen et al. 2006).
The basic question is whether extended field
radiotherapy could be safety replaced with the
use of upfront chemotherapy followed by
involved-field radiotherapy with a dose reduction.
Eighty-one patients with intracranial germinoma
were treated, 42 underwent chemo-radiotherapy
(CRT) and 39 underwent radiotherapy only. In
the first group, patients received one to five cycles
of platinum-based chemotherapy, followed by
involved-field radiotherapy in tumor dose of
50 Gy. In the radiotherapy group, craniospinal
radiotherapy range 2136 Gy was followed by
involved-field radiotherapy to 54 Gy. The 5- and
10-year overall survival was 100% and 92.5% for
radiotherapy alone and 92.9% and 92.9% for
chemo-radiotherapy, respectively. The 5-year
recurrence-free survival rate was 100% for radiotherapy and 88% for chemo-radiotherapy. In four
patients after up-front chemotherapy recurrences
in brain (3) and spine (1) have developed. Only
one patient achieved complete remission from
193
salvage treatment. In contrast, quality of life was

lower for patients who received radiotherapy.
These results have shown that better quality of
life provided by chemo-radiotherapy was compensated by the higher rate of relapse. The inclusion of the ventricles in involved-field radiotherapy
after upfront chemotherapy, especially for
patients with initial negative seeding, seems to be
very promising (Eom et al. 2008).
Several reports indicated that the incidence of
spinal recurrence in intracranial germinoma was
found to be too low to warrant routine spinal irradiation. Japanese authors performed the analysis
of the risk factors of spinal seeding in these
patients. Between 1980 and 2007, 165 patients
with no evidence of spinal metastases at diagnosis were treated with cranial radiotherapy. Two
third of them were children of age <18 years.
Spinal recurrences were developed in 15 patients
(9.1%). In multifocal analysis, large intracranial
disease and multifocal cranial disease were independent risk factors of spinal recurrence. These
results suggest that elective craniospinal irradiation appears to be appropriate in patients with
multifocal or large tumors only. After salvage
treatment, 3-years overall and disease-free survival was 65% and 57%, respectively; significantly better outcome have had patients treated
with combined radio- and chemotherapy (Ogawa
et al. 2008).
Chemotherapy is the only strategy for children
with germ-cell tumors, compared with regimes
with irradiation. Between 2001 and 2004, 25
patients of age 4 months 24 years were treated
with 46 cycles intensive chemotherapy only.
The 6-years event-free and overall survival was
45.6% and 75.3%, respectively. These intensive
chemotherapy regimes were less effective than
irradiation containing regimes. Presently, standard treatment for CNS germ cell tumors should
include irradiation either alone or combined with
chemotherapy for those with malignant germ cell
tumors (da Silva et al. 2010). Nongerminoma
germ-cell intracranial tumors, contrary to germinoma are relatively radio resistant and associated
with poor outcome. The combination of radiation
therapy with cisplatinum-based chemotherapy
could improve the outcome.
194
Childrens Oncology Group in phase II

prospective trial investigated preradiation chemotherapy with responsebased radiation therapy
in children with germinoma and nongerminoma
intracranial germ-cell tumors. Children with germinomas received cisplatin + etoposid, alternating with vincristine + cyclophosphamid. Children
with nongerminomatous tumors received double
doses of cisplatin and cyclophosphamid. For germinoma patients in complete response, radiotherapy was decreased to 30.6 Gy. High-risk patients
received neuraxis radiotherapy: 50.4 Gy local
and 30.6 Gy whole CSI in complete remission,
and 54 Gy local + 36 Gy whole CSI in incomplete
remission. All 12 germinoma patients and 11/14
nongerminoma patients were progression-free
after 5-years, and response (germinoma 91%,
nongerminoma 55%) and survival are encouraging after this regimen plus response-based radiotherapy (Kretschmar et al. 2007).
Conclusions
The modern radiotherapy techniques such as
conformal radiotherapy and stereotactic fractionated radiotherapy, have been shown to be an
appropriate treatment for many pediatric brain
tumors. These techniques allow applying a high
homogenous dose of irradiation in the region of
the tumor while minimizing dose to normal tissues. IMRT is capable of generating 3-D dose
distribution close to the target volume. IGRT is
the most advanced verification imaging and gives
a good control of treatment delivery. The Cyber
Knife radiosurgery system allows treating precisely infants and young children, with reduced
requirement of general anesthesia. Further
improvements in dose distribution require new
radiation modalities as protons or light ions, with
a further dose reduction to healthy tissues and
decreasing the late effects of radiation therapy.
The new approach of radiotherapy and importance of new drugs in combined treatment were
recently considered.
For patients with standard risk medulloblastoma, good results were obtained with reduced
dose radiotherapy and adjuvant chemotherapy.
Hyperfractionated radiotherapy to the craniospinal

axis, with high dose conformal radiotherapy
boost to the primary site, without chemotherapy,
resulted with high survival rates, without any
negative impact on cognitive function. Although
outcome for patients with high-risk disease, especially those with metastatic disease remains relatively poor, risk adopted radiotherapy followed
by dose intensive chemotherapy improved the
outcome of children with high-risk medulloblastoma. Some reports stress the impact of technical
quality of radiotherapy on survival in patients
with medulloblastoma, and conformal radiotherapy with improved quality control allows detection and correction of most of the major deviations.
Up to 30% of medulloblastoma cases, occur in
children younger than 3 years. The prognosis for
these children remains poor and survivors are in
the high risk for cognitive deficits. Intensive postoperative chemotherapy with intraventricular
metotrexat allows obtaining relatively good
results in children without residual tumor and/or
metastatic disease. The results of IQ in these
patients were better than in those who received
radiotherapy.
In patients with localized ependymoma, primary tumor site is the predominant site of failure,
and the craniospinal irradiation may be omitted
in these patients. Hyperfractionated radiotherapy
with or without chemotherapy, despite the high
total dose adopted, does not change the prognosis. In children under the age of 3 years, with anaplastic ependymoma, delaying radiotherapy
jeopardizes survival even after intensive chemotherapy. Limited volume conformal radiation
therapy achieves high rates of disease control and
results in stable neurocognitive outcomes, even
in very young children.
For patients with low-grade glioma, resection
remains the preferred treatment option. However,
in patients with partially resected or non-resectable
tumors, other methods of treatment such as radiotherapy or chemotherapy could be considered.
Chemotherapy seems to be effective in delaying
the need for early radiotherapy in chiasmatichypothalamic gliomas. For children with incompletely resected, symptomatic tumors, especially
older than 5 years, high precision conformal
or stereotactic radiotherapy is recommended.

Encouraging results with low risk of radiotherapy
related complications can be obtained.
High-grade glioma in children carries a somewhat better prognosis than in adults, but outcomes
for these patients remain relatively poor. The
strongest prognostic factor for these patients is
the extent of surgery, so the optimal therapy
should include the maximal possible resection.
High-dose chemotherapy regimes added to postoperative radiotherapy provide no additional
benefit.
The prognosis for children with diffuse brain
stem glioma is still very poor and conventionally
fractionated local therapy remains the standard of
care. Multiple trials involving various regimes of
chemotherapy, and new drugs as temozolamide
have not demonstrated any improvement in outcome. In small focal brain-stem low-grade
glioma, treated with high precision radiotherapy,
better results can be obtained. In patients with
incompletely resected or recurrent craniopharyngioma, the use of conformal radiotherapy or
combined photon/proton therapy results in tumor
control equivalent to that achieved using conventionally planned radiotherapy, but with lower
possibility of late side-effects.
The optimal treatment for intracranial germinoma remains controversial. The long-term multiinstitutional studies confirm the effectiveness of
radiotherapy. The incidence of spinal relapses was
low and did not justify routine spinal irradiation.
The whole ventricle is recommended as the smallest target volume for germinoma after induction
of cisplatin-based-chemotherapy. Cisplatin-basedchemotherapy and irradiation improved prognosis
for patients with non-germinomatous germ cell
tumors. Intensive chemotherapy was effective in
some of these patients, and salvage therapy,
including irradiation, was feasible in patients with
recurrent disease.
References
Allen J, Donahue B, Mehta M, Miller DC, Rorke LB,
Jakacki R, Robertson P, Sposto R, Holmes E,
Vezina G, Muraszko K, Puccetti D, Prados M, Chan
195
KW (2009) A Phase II study of preradiotherapy
chemotherapy followed by hyperfractionated radiotherapy for newly diagnosed high-risk medulloblastoma/primitive neuroectodermal tumor: a report from
the Childrens Oncology Group (CCG 9931). Int J
Carrie C, Grill J, Figarella-Branger D, Bernier V,
Padovani L, Habrand JL, Benhassel M, Mege M,
Mah M, Quetin P, Maire JP, Baron MH, Clavere P,
Chapet S, Maignon P, Alapetite C, Claude L, Laprie A,
Dussart S (2009) Online quality control, hyperfractionated radiotherapy alone and reduced boost volume
for standard risk medulloblastoma: long-term results
of MSFOP 98. J Clin Oncol 27:18791883
Combs SE, Ketler V, Welzel T, Behnisch W, Kulozik AE,
Bischof M, Hof H, Debus J, Schulz-Ertner D (2008)
Influence of radiotherapy treatment concept on the
outcome of patients with localized ependymomas. Int
Combs SE, Steck I, Schultz-Ertner D, Welzer T, Kulozik
AE, Behnisch W, Huber PE, Debus J (2009) Longterm outcome of high-precision radiotherapy in
patients with brain stem gliomas: results from a difficult-to-treat patient population using fractionated stereotactic radiotherapy. Radiother Oncol 91:6066
Conter C, Carrie C, Bernier V, Geoffray A, Pagnier A,
Gentet JC, Lellouch-Tubiana A, Chabaud S, Frappaz
D (2009) Intracranial ependymomas in children: society of pediatric oncology experience with postoperative hyperfractionated local radiotherapy. Int J Radiat
Oncol Biol Phys 74:15361542
Da Silva NS, Cappellano AM, Diez B, Cavalheiro S,
Gardner S, Wisoff J, Kellie S, Parker R, Garvin J,
Finlay J (2010) Primary chemotherapy for intracranial
germ cell tumors: results of the third international
CNS germ cell tumor study. Pediatr Blood Cancer
54:377383
Eom KY, Kim IH, Park CI, Kim HJ, Kim JH, Kim K, Kim
SK, Wang KC, Cho BG, Jung HW, Heo DS, Kang HJ,
Shin HY, Ahn HS (2008) Upfront chemotherapy ad
involved-field radiotherapy results in more relapses
than extended radiotherapy for intracranial germinomas: modification in radiotherapy volume might be
needed. Int J Radiat Oncol Biol Phys 71:667671
Fitzek MM, Linggood RM, Adams J, Munzenrider JE
(2006) Combined proton and photon irradiation for
craniopharyngioma: long term results of the early
cohort of patients treated at Harvard Cyclotron
Laboratory and Massachusetts General Hospital. Int J
Gandola L, Massimino M, Cefalo G, Solero C, Spreafico
F, Pecori E, Riva D, Collini P, Pignoli E, Giangaspero
F, Luksch R, Beretta S, Poggi G, Biassoni V, Ferrari A,
Pollo B, Favre C, Sardi I, Terenziani M, FossatiBellani F (2009) Hyperfractionated accelerated
radiotherapy in the Milan strategy for metastatic
medulloblastoma. J Clin Oncol 27:566571
Grundy RG, Wilne SA, Weston CL, Robinson K, Lashford LS,
Ironside J, Cox T, Kling Chong W, Campbell RHA,
Bailey CC, Gattamaneni R, Picton S, Thorpe N,
196
Mallucci C, English MW, Punt JAG, Walker DA,
Ellison DW, Machin D (2007) Primary postoperative
chemotherapy without radiotherapy for intracranial
ependymoma in children: the UKCCSG/SIOP prospective study. Lancet Oncol 8:696705
Grundy RG, Wilne SH, Robinson KJ, Ironside JW, Cox T,
Chong WK, Michalski A, Campbell RH, Bailey CC,
Thorp N, Pizer B, Punt J, Walker DA, Ellison DW,
Machin D (2010) Primary postoperative chemotherapy
without radiotherapy for treatment of brain tumours
other than ependymoma in children under 3 years:
results of the first UKCCSG/SIOP CNS 9204 trial. Eur
J Cancer 46:120133
Jalali R, Rau N, Arora B, Gupta T, Dutta D, Munshi A,
Sarin R, Kurkure P (2010) Prospective evaluation of
radiotherapy with concurrent and adjuvant temozolomide in children with newly diagnosed diffuse intrinsic pontine glioma. Int J Radiat Oncol Biol Phys
77:113118
Kretschmar C, Kleinberg L, Greenberg M, Burger P,
Holmes E, Wharam M (2007) Pre-radiation chemotherapy with response-based radiation therapy in children with central nervous system germ cell tumors: a
report from the Childrens Oncology Group. Pediatr
Lee MJ, Ra YS, Park JB, Goo HW, Ahn SD, Khang SK,
Song JS, Kim YJ, Ghim TT (2006) Effectiveness of
novel combination chemotherapy, consisting of 5-fluorouracil, vincristine, cyclophosphamide and etoposide, in the treatment of low-grade gliomas in children.
MacDonald TJ, Arenson EB, Alter J, Sposto R, Bevan
HE, Bruner J, Deutsch M, Kurczynski E, Luerssen T,
McGuire-Cullen P, OBrien R, Shah N, Steinbok P,
Strain J, Thomson J, Holmes E, Vezina G, Yates A,
Philips P, Packer R (2005) Phase II study of high-dose
chemotherapy before radiation in children with newly
diagnosed high-grade astrocytoma: final analysis of
Childrens Cancer Group Study 9933. Cancer
104:28622871
Marcus KJ, Goumnerova L, Billett AL, Lavally B, Scott
RM, Bishop K, Xu R, Young Poussaint T, Kieran M,
Kooy H, Pomeroy SL, Tarbell NJ (2005) Stereotactic
radiotherapy for localized low-grade gliomas in children: final results of a prospective trial. Int J Radiat
Merchant TE, Kiehna EN, Kun LE, Mulhern RK, Li C,
Xiong X, Boop FA, Sanford RA (2006) Phase II trial
of conformal radiation therapy for pediatric patients
with craniopharyngioma and correlation of surgical
factors and radiation dosimetry with change in cognitive function. J Neurosurg 104:94102
Merchant TE, Kun LE, Krasin MJ, Wallace D,
Chintagumpala MM, Woo SY, Ashley DM, Sexton
M, Kellie SJ, Ahern V, Gajjar A (2008) A multiinstitution prospective trial of reduced-dose craniospinal irradiation (23.4 Gy) followed by conformal
posterior fossa (36 Gy) and primary site irradiation
(55.8 Gy) and dose-intensive chemotherapy for
average-risk medulloblastoma. Int J Radiat Oncol
Biol Phys 70:782787
Merchant TE, Li C, Xiong X, Kun LE, Boop FA, Sanford
RA (2009a) Conformal radiotherapy after surgery for
paediatric ependymoma: a prospective study. Lancet
Oncol 10:258266
Merchant TE, Kun LE, Wu S, Xiong X, Sanford RA,
Boop FA (2009b) Phase II trial of conformal radiation
therapy for pediatric low-grade glioma. J Clin Oncol
27:35983604
Merchant TE, Conklin HM, Wu S, Lustig RH, Xiong X
(2009c) Late effects of conformal radiation therapy for
pediatric patients with low-grade glioma: prospective
evaluation of cognitive, endocrine, and hearing deficits. J Clin Oncol 27:36913697
Miralbell R, Fitzgerald TJ, Laurie F, Kessel S, Glicksman
A, Friedman HS, Urie M, Kepner JL, Zhou T, Chen Z,
Barnes P, Kun L, Tarbel NJ (2006) Radiotherapy in
pediatric medulloblastoma: quality assessment of
Pediatric Oncology Group Trial 9031. Int J Radiat
Nguyen QN, Chang EL, Allen PK, Maor MH, Alter JL,
Nahajan A, Wolff JEA, Weinberg JS, Woo SY (2006)
Focal and craniospinal irradiation for patients with
intracranial germinoma and patterns of failure. Cancer
107:22282236
Ogawa K, Yoshii Y, Shikama N, Nakamura K, Uno T,
Onishi H, Itami J, Shioyama Y, Iraha S, Ohydo A, Toita
T, Kakinohana Y, Tamaki W, Ito H, Murayama S (2008)
Spinal recurrence from intracranial germinoma: risk
factors and treatment outcome for spinal recurrence.
Int J Radiat Oncol Biol Phys 72:13471354
Oyharcabal-Bourden V, Kalifa C, Gentet JC, Frappaz D,
Edan C, Chastagner P, Sariban E, Pagnier A, Babin A,
Pichon F, Neuenschwander S, Vinchon M, Bours D,
Mosseri V, Le Gales C, Ruchoux M, Carrie C, Doz F
(2005) Standard-risk medulloblastoma treated by adjuvant chemotherapy followed by reduced-dose craniospinal radiation therapy: a French Society of Pediatric
Oncology Study. J Clin Oncol 23:47264734
Packer RJ, Gajjar A, Vezina G, Rorke-Adams L, Burger
PC, Robertson PL, Bayer L, LaFond D, Donahue BR,
Marymont MH, Muraszko K, Langston J, Sposto R
(2006) Phase III study of craniospinal radiation therapy followed by adjuvant chemotherapy for newly
diagnosed average-risk medulloblastoma. J Clin Oncol
24:42024208
Rutkowski S, Bode U, Deinlein F, Ottensmaier H,
Pietsch T, Wolff JEA, Kortmann RD, Kuehl J (2005)
352:978986
Skowroska-Gardas A, Chojnacka M, MorawskaKaczyska M, Perek D, Perek-Polnik M (2007)
Patterns of failure in children with medulloblastoma
treated with 3D conformal radiotherapy. Radiother
Oncol 84:2633
Taylor RE, Bailey CC, Robinson KJ, Weston CL, Walker
DA, Ellison D, Ironside J, Pizer BL, Lashford LS
(2005) Outcome for patients with metastatic (M2-3)
medulloblastoma treated with SIOP/UKCCSG
PNET-3 chemotherapy. Eur J Cancer 41:727734

Timmermann B, Kortmann RD, Khl J, Rutkowski S,
Dieckmann K, Meisner C, Bamberg M (2005) Role of
radiotherapy in anaplastic ependymoma in children
under age of 3 years: results of the prospective German
brain tumor trials HIT-SKK 87 and 92. Radiother
Oncol 77:278285
Von Hoff K, Hinkes B, Gerber NU, Deinlein F, Mittler U,
Urban C, Benesch M, Warmuth-Metz M, Soerensen
N, Zwiener I, Goette H, Schlegel PG, Pietsch T,
Kortmann RD, Kuehl J, Rutkowski S (2009) Long-
197
term outcome and clinical prognostic factors in children
with medulloblastoma treated in the prospective randomised multicentre trial HIT 91. Eur J Cancer
45:12091217
Wolff JE, Driever PH, Erdlenbruch B, Kortmann RD,
Rutkowski S, Pietsch T, Parker C, Metz MW, Gnekow
A, Kramm CM (2010) Intensive chemotherapy
improves survival in pediatric high-grade glioma after
gross total resection: results of the HIT- GBM-C protocol. Cancer 116:705712
Epileptic Seizures and

Supratentorial Brain Tumors
in Children
19
Roberto Gaggero, Alessandro Consales, Francesca

Fazzini, Maria Luisa Garr, and Pasquale Striano
Contents
Abstract
Introduction ............................................................
199
Epilepsy as First Sign of the Tumor .....................
200
Epilepsy After Surgery or Other

Tumor Treatments..................................................
201
Neurosurgery and Drug-Resistant Epilepsy

Secondary to Brain Tumors ..................................
202
Antiepileptic Treatment .........................................
203
Conclusion ..............................................................
204
References ...............................................................
205
R. Gaggero (*) A. Consales F. Fazzini M.L. Garr

P. Striano
Department of Neurosciences, Gaslini Childrens
Hospital, Genoa, Largo G. Gaslini 5,
16147 Genoa, Italy
e-mail: garob@iol.it
Brain tumors represent about one third of the

pediatric tumors. Epilepsy is overall associated with supratentorial brain tumors, that
prevail in children under 3 years and over
10 years. Prevalence of seizures as first tumor
symptom is 1520%. In children under the
3 years of life, epilepsy is more frequent
(until 70%), it can persists after the surgical
intervention and in some cases an epileptic
encephalopathy occurs. Epilepsy can began
after surgery in other cases (1040%), as a
consequence of brain damage. Moreover,
brain tumors are in a relevant percentage of
cases (2650%) the cause of a drug-resistant
epilepsy, that can be successfully treated
by surgery. The post surgical evolution is
more favorable for patients with some particular tumor types (DNT, gangliogliomas).
Pharmacological therapy of the epilepsy
related with tumors is difficult, for an high
frequency of side effects and of interactions
with chemotherapy. The new AEDs are
particularly promising not only for the favorable effect, but also for the lower incidence
of side effects and of interactions.
Introduction
Brain tumors represent a relevant clinical problem in children and the most frequent cause of
death for malignancy under the 16 years of life

199
200
(Wilne et al. 2007). The incidence of brain tumors

in children ranges from 2.5 to 18/100.000 cases
and makes up one third of all cases of pediatric
tumors (Porter et al. 2010). The incidence is
higher in the age from 4 to 9 years (Makino et al.
2010), with some differences related to the age.
In fact, in children between 3 and 11 years
infratentorial tumors are more frequent (57.5%),
whereas in younger infants there is a predominance of supratentorial tumors.
The most frequent supratentorial tumor is
astrocytoma (3540%), followed by medulloblastoma (1520%), anaplastic astrocytoma,
glioblastoma (15%), ependymoma (10%), and
craniopharyngioma (5%) (Rosemberg and
Fujiwara 2005). Other tumor types include germ
cell tumors (4.4%), gangliogliomas, meningiomas (2.5%), primitive neuroectodermal
tumors (PNET) (1.9%), choroid plexus tumors
(0.9%), and dysembryoplastic neuroepithelial
tumors (DNT) (0.6%). Among infratentorial
tumors, astrocytoma is the more frequent
type (38%), followed by pilocytic astrocytomas, medulloblastomas, and ependymomas
(Rosemberg and Fujiwara 2005).
As in adults, epilepsy may be the initial presenting features of a primary brain tumor in children (Liigant et al. 2001; Khan et al. 2006; Riva
et al. 2006). However, it can also emerge during
the course of therapy or as a late effect of tumor
treatment (e.g., surgery, chemotherapy, radiotherapy) (Ullrich 2009), thus adding substantial morbidity to the patients (van Breemen et al. 2007).
In addition, the medical management of epilepsy
is often difficult in this age group, particularly
due to the risk of side-effects deriving from use
of AEDs as well as to their potential interactions
with chemotherapy drugs (Khan et al. 2006; van
Breemen et al. 2007).
The mechanisms involved in the pathophysiology of tumor-related epilepsy are multiple,
complex, varied, and-at least in part-not yet well
defined. Several hypothesis have been proposed:
the direct effects of the tumor, a focal cortical
hyperexcitability, neurotransmitters abnormalities, and even the effect of metabolic products of
the tumor itself (Ullrich 2009). Indeed, at level of
R. Gaggero et al.
peri-tumoral brain tissue, morphological changes

and Ph variations, abnormalities in ion levels and
amino acids, alterations of glutaminergic receptors, enzymatic modifications, and alterations of
intercellular communication have been widely
documented (Shaller et al. 2003). Moreover, several other mechanisms can influence the expression of epilepsy in brain tumors, including genetic
changes, tumor malignancy, location, and age of
onset (Kim et al. 2001; Khan et al. 2006; Riva
et al. 2006).
Epilepsy as First Sign of the Tumor

The frequency of seizures in adults with brain
tumors is generally high, ranging from 23% to
86% (Liigant et al. 2001; Riva et al. 2006), with
highest percentages in low-grade tumors and
lower in secondary tumors (Liigant et al. 2001)
and in glioblastomas (Riva et al. 2006). Seizures
occur as the first presentation symptom in about
85% of the cases (Riva et al. 2006). Association
of epilepsy and brain tumors in children is lower:
seizures occur in 1520% of children with a cerebral tumor (Khan et al. 2006), most likely due to
the fact that infratentorial tumors are more frequent in the pediatric age. Seizures represent a
rare clinical onset of brain neoplasm, ranging
from 10% to 15% in different series (Wilne et al.
2007).
A recent review on the clinical presentation of
paediatric brain tumors (Wilne et al. 2007) reports
that most frequent early symptoms of brain
tumors are headache (33%), nausea and vomiting
(32%), abnormalities of gait (27%). Seizures are
reported only in 13% of cases and are more frequent (38%) as initial symptoms in tumors with
supratentorial location. In a group of 98 children
(Shady et al. 1994) with supratentorial brain
tumors, mainly of glial type, 18 subjects (18.3%)
were affected by seizures. In half of them seizures occurred at the onset and in 30% they were
the unique symptom. Seizures were focal in 77%
and EEG showed lateralized abnormalities in
88% of cases. In patients with seizures, tumors
were cortical in 59%, whereas only 15% of cases
19
Epileptic Seizures and Supratentorial Brain Tumors in Children
with infratentorial lesions presented seizures.

Epilepsy was associated with temporal and frontal
localization of the lesion and with some particular
types of tumors (in cases of gangliogliomas and
oligoastrocytomas the prevalence of seizures
was 88% and 86%, respectively). The cases with
seizures as first sign of the tumor showed a favorable evolution.
Another paper (Gilles et al. 1992) investigated
a large population of children and adolescents
with brain tumors and showed that among 3,291
cases only 14% presented seizures before the
diagnosis. Prevalence was higher in patients
with supratentorial lesions (22%) and in particular
in subjects aged more than 14 years. In 50% of
cases the period from the first seizure and the
surgery was >1 year. In 89% of cases with seizures
other symptoms were observed: consciousness
disorder, mental impairment, gait and motor
abnormalities. The main tumors characteristics
were supratentorial and cortical localization,
histology type (astrocitomas of different grade,
ependymoma).
A particular and different condition is represented by brain tumors in very young children
(under 3 years of life). They represent a particular
clinical problem since their histological type and
the immature stage of evolution of the nervous
system (Michasky and Garr 2004). The clinical
presentation of the tumors in this period is peculiar; the prevalence of supratentorial localization
(Rutledge et al. 1987; Chung et al. 1998; Mehrotra
et al. 2009; Gaggero et al. 2009) can explain the
high frequency of seizures as first symptom. The
prevalent tumor types in this age are astrocytoma,
primitive neuroectodermical tumor, papilloma or
carcinoma of the choroid, teratoma, medulloblastoma, dermoid tumor, embryonal rhabdomyosarcoma (Gaggero et al. 2009).
In a group (Rutledge et al. 1987) of very early
onset tumors (<1 year), 2540% of the cases
with a supratentorial location presented seizures
at the onset and continued to suffer of epilepsy
in 1015% of cases. Mainly temporal mesial
tumors were associated with seizures. In our
experience (Gaggero et al. 2009), among 28
children evaluated, 20 (71.4%) suffered from
201
epilepsy. Mean age at seizure onset was

18.7 months. In all cases, seizures recurrence was
high, generally with a daily or weekly frequency.
In 15 (75%) children, epilepsy was an early
manifestation or the presenting symptom of the
tumor. In this group, seizures were focal with or
without secondary generalization in eight cases
(53.3%) and generalized (spasms, tonic-clonic
seizures, atypical absences) in the other seven
individuals. Seizures remitted in 11 of these
cases (73%) after surgical treatment, whereas in
the other four epilepsy persisted and presented a
long lasting severe course. Five (25%) subjects
from this group suffered from occasional episodes of convulsive status epilepticus needing
benzodiazepines rescue treatment. During the
post-surgical evolution, other two (13%) children featured the clinical picture of an epileptic
encephalopathy with stagnation in development,
speech difficulties, and continuous spike-waves
during sleep (CSWSS). Post-surgical follow-up
ranged from 4 to 10 years. In this series, the outcome of epilepsy was generally good, with most
children (76.4%) being seizure free or showing
>90% improvement in seizure frequency after
surgery. However, in about 20% of the cases,
epilepsy persisted despite surgery and different
antiepileptic drugs (AEDs) regimen. Best epilepsy outcome was observed in patients with
low-grade tumors and without neurological
deficits after surgery.
Epilepsy After Surgery or Other

Tumor Treatments
In a significant number of patients (up to 45%)
with brain tumors, epilepsy stars after the diagnosis and the treatment, often as consequence of
the therapeutical procedures (Glantz et al. 2000).
Part of them is represented by post-surgical
seizures, that occur in the immediate period
post-surgery (days or weeks). These cases are
not frequent (12%) (Wang et al. 1994). In other
cases, the seizures appear before the diagnosis
and the treatment and they persist after surgery
(Gaggero et al. 2009). In our experience
202
(Gaggero et al. 2009), concerning epilepsy

associated with brain tumors in infants (under
3 years), seizures began after surgery in 5 out of
20 cases (25%) and they became severe and
drug resistant. Of these five children, four had
focal seizures, with or without secondary generalization. All these patients were affected by a
serious neurological impairment (hemiplegia,
mental disorders). Also one of these cases
presented an epileptic encephalopathy with a
continuous spike waves during slow sleep
(CSWSS) EEG. So, in these cases the epilepsy
is probably due to the brain damage.
A review of the neurological sequelae of brain
tumors in children (Ullrich 2009) considers the
effects of the different treatments: the lesions
secondary to the surgery, the side effects of the
chemotherapy, the consequences of the radiation.
All these conditions can determine an epilepsy.
Surgical resection remains the mainstay of therapy for most primary brain tumors, both for
reduction of tumor burden and to provide histological diagnosis. For many tumors, surgical
resection is the most important mode of therapy.
The deficits and long-term effects resulting from
neurosurgery are multifactorial and depend on
the tumor location, the attempted degree of resection, the age of the patient, and the presurgical
performance status. Direct neurological sequelae
from surgical removal can include localized
lesions, hydrocephalus and perioperative stroke,
with consequent clinical symptoms (ataxia, hemiparesis, and persistent neurosensorial and neurocognitive deficits).
As regards chemotherapy, it should be kept in
mind that some drug treatments may increase the
risk of seizure occurrence. Thus, cyclosporine
can be cause seizures and leucoencephalopathy
(Gaggero et al. 2006) in 10% of the cases, whereas
cisplatin, vincristine and metotrexate are associated with seizures in 1% of the patients (Ullrich
2009). Radiotherapy can determine a vasculopathy, especially in infants. In fact, in children under
3 years, these effects are more severe. The risk of
developing severe brain damage after surgery and
irradiation are peculiar in this subgroup of
patients due to incomplete brain development
(Michasky and Garr 2004).
R. Gaggero et al.
Neurosurgery and Drug-Resistant

Epilepsy Secondary to Brain Tumors
Drug-resistant epilepsy can be secondary to brain
tumors with a significant number of patients,
including children. The surgical treatment of the
drug-resistant epilepsy in pediatrics has become
frequent in the last decades. The results of this
therapy are generally good. The tumors types
vary in the different series. Prevalence of dysembryoplastic neuroepithelial tumors (DNT) ranges
from 16% to 39% (Kim et al. 2001; Cossu et al.
2008; Zentner et al. 1997). Gangliogliomas frequency was 2556% (Cossu et al. 2008; Zentner
et al. 1997). Gangliocitomas: 8% (Cossu et al.
2008); olygodendrogliomas 26% (Kim et al.
2001); low grade astrocytomas 18% (Zentner
et al. 1997; Cossu et al. 2008). The results of the
surgery on the epilepsy were reported to be positive with high frequency of seizures free patients:
from 67% to 74% (Van Oijen et al. 2006; Benifla
et al. 2006, 2008; Cossu et al. 2008).
The results are more favorable in the cases
with brain tumors in comparison with the other
etiological factors: the percentage of seizure free
patients class was 71% (Zentner et al. 1997),
90% (Kral et al. 2001), 89% (Cossu et al. 2008).
The favorable prognostic factors are an unique
lesion, a temporal location, a complete lesionectomy, an older age at onset (Benifla et al. 2006,
2008; Van Oijen et al. 2006; Cossu et al. 2008).
Some specific tumors are in particular related
with a more favorable outcome after surgery.
Several series of dysembryoplastic neuroepithelial tumors (DNT) operated for a drug resistant
epilepsy show a very high frequency of seizures
free patients in childhood and adolescence:
83100% (Nolan et al. 2004; Chang et al. 2010;
Ozien et al. 2010). In some reports, after a more
prolonged follow up, the percentage of remission trends to decrease (from 85% to 62% after
4 years) (Nolan et al. 2004). Also gangliogliomas
are related with a favorable outcome in children:
90% seizure free (Ogiwara et al. 2010), 87.5%
(Ozien et al. 2010).
In a high percentage of cases, tumors are
associated with other cerebral lesions, that can
19
represent a concomitant epileptogenic factor:

mesial temporal sclerosis 13% (Benifla et al.
2006), cortical dysplasia 7% (Benifla et al. 2006)
and 16% (Cossu et al. 2008), astrogliosis 12%
(Benifla et al. 2006).
Antiepileptic Treatment
The use of chronic antiepileptic drugs (AED) for
the treatment of the seizures is very frequent in
patients with epilepsy secondary to brain tumors.
The AEDs can be use after the first seizures
before the tumor diagnosis or in the occasion of
the surgical intervention or for the treatment of
post-surgery epilepsy. In a recent paper (Sogawa
et al. 2009), it has been reported that AEDs were
used in 10% of 334 pediatric brain tumors
patients. The most frequently used AEDs were
phenytoin and oxcarbazepine. Initial therapy was
frequently changed because lack of efficacy and
adverse effects. At last follow up the most
common antiepileptic drugs were oxcarbazepine
and levetiracetam. The patients started on
newer-generation AEDs (levetiracetam, oxcarbazepine, lamotrigine) tended to remain on the
same treatment more than did patients on older
generation antiepileptic drugs: valproate, phenytoin and phenobarbital. Recently, several other
reports confirm the prevalent efficacy and tolerability of the new AEDs, as levetiracetam, but the
studies were performed mainly in adults and in
small groups (Van Breemen et al. 2007).
A specific problem concerns the interference
of the AED with the chemotherapeutic drugs.
Several old AEDs (phenytoin, phenobarbital, carbamazepine) are metabolic inductors and they
reduced the blood levels and the efficacy of the
chemotherapy. So, enzyme inducing anticonvulsants drugs are generally not recommended
because they can lead to insufficient serum levels
of concomitant chemotherapeutic drugs (Vecht
et al. 2003). However, because many AEDs and
chemotherapeutics share common metabolic
pathways via the hepatic cytochrome P450 (CYP)
isoenzymes, there is potential for drug interactions. Likewise, chemotherapeutics can alter the
pharmacokinetics of AEDs, resulting in decreased
203
seizure control. Other agents, such as valproate,

are enzyme-inhibiting AEDs that can impede the
metabolism of other drugs, potentially increasing
the serum concentration of chemotherapeutics
(Vecht et al. 2003). A new generation of AEDs
that are not metabolized by the CYP pathway is
currently being developed. Among these, gabapentin and levetiracetam show the most promise in
treating epileptic seizures in patients with brain
tumors. Interactions between these newer AEDs
and chemotherapeutic agents have not been
reported. Thus far several drugs without enzyme
induction effect are preferred (Vecht et al. 2003).
The absence of interaction has been proved also
in children with epilepsy and brain tumors
(Ruggiero et al. 2010).
From a general point of view, the use of AEDs
in patients with epilepsy secondary to brain
tumors has not been the object of specific studies.
Their efficacy appears to be limited, their mechanisms of action are not specific for the epileptogenesis related with tumors. Some other actions
of AEDs can be useful for the tumor treatment.
For instance, it has been proposed an apoptosis
effect on the tumoral cells by valproate, with a
positive therapeutic result, but this observation
remains controversial (Blaheta et al. 2005).
Another aspect concerning the use of AEDs in
patients with epilepsy associated to brain tumors
is represented by the anticonvulsant prophylaxis
prescribed after surgery. For a long time it has
been believed that preventing seizures with antiepileptic drugs (seizure prophylaxis) was effective and necessary, but the supporting evidence
was little and mixed. The incidence of postoperative epilepsy following a subfrontal craniotomy
did not exceed 12% when examined at various
time periods during a 3-years postoperative
course (Wang et al. 1994). Antiepileptic drugs
were not warranted to reduce the incidence of
postoperative seizures after the 1-month postoperative period and should not be used for long-term
prophylactic therapy in children following a
subfrontal craniotomy.
Ten years ago the Quality Standards
Subcommittee of the American Academy of
Neurology (Glantz et al. 2000) published the
results of a meta-analysis of 12 studies, that have
R. Gaggero et al.
204
examined, either in randomized controlled trials

or cohort studies, the ability of prophylactic anticonvulsants to prevent first seizures in patients
with brain tumors. All these studies were dedicated to adult populations. None have demonstrated efficacy. Only one of the 12 studies
reported a significant difference in seizure frequency between the anticonvulsant prophylaxis
and nonprophylaxis groups, and this difference
favored the non-prophylaxis group. In contrast,
deleterious interactions with cytotoxic drugs and
corticosteroids are a major concern, and the incidence and severity of anticonvulsant side effects
appear to be appreciably higher (2040%) in
brain tumor patients than in the general population of patients receiving anticonvulsants.
Despite the lack of definitive evidence, many
physicians at that time continued to administer
anticonvulsant medication prophylactically for
preventing first seizures in patients with brain
tumors, both adults and children (Stevens 2006).
Two years ago the conclusions of a study on
antiepileptic drugs for preventing seizures in people with brain tumors (Tremont-Lukats et al.
2008), carried out according to Cochrane Criteria
talked that here was no difference between the
treatment interventions and the control groups in
preventing a first seizure in participants with
brain tumors. The evidence is neutral, neither for
nor against seizure prophylaxis, in people with
brain tumors. The decision to start an antiepileptic drug for seizure prophylaxis is ultimately
guided by assessment of individual risk factors
and careful discussion with patients. New AEDs
as levetiracetam represent a new opportunity for
the prevention of seizures.
In summary, AEDs usually show a modest
efficacy; they can be useful for preventing seizure
after surgery during the chemotherapy, whereas
the long term action is less sure. They have to be
employed in cases with partial tumor resection,
cortical location, disseminated tumors and concomitant hemorrhages (Glantz et al. 2000;
Tremont-Lukats et al. 2008; Ullrich 2009). Also
the interruption of the therapy, after surgery and a
favorable outcome, is possible (Khan et al. 2006).
Sixty-two patients discontinued AEDs at a
median time of 5.6 years from the first seizure.
Median time since AED withdrawal was

2.3 years. Seizures recurred in 17 (27%) patients.
Median seizure free period before AED withdrawal was 1.3 years. More than one tumor resection and whole-brain radiation treatment were
associated with seizure recurrence. At seizure
recurrence, control was re-established in 15
patients with AED reinstitution. Two patients
with poor drug compliance continue to have seizures. In conclusions, AED withdrawal can be
successfully achieved in majority of carefully
selected patients (Khan and Onar 2006).
Conclusion
The prevalence of brain tumors is elevated (about
one third of pediatric tumors). Supratentorial
tumors represent more than 50% of them and
they are more frequent in children under 3 years
and over 10 years. Epilepsy is overall associated
with supratentorial brain tumors. Seizures can be
the initial symptom of a brain tumor or they can
occur during the evolution. In other cases, tumors
of particular type can be the cause of a drugresistant intractable epilepsy, that can be successfully treated by surgery. Seizures occur in 1520%
of children with brain tumors. Globally, seizures
are the first sign of a brain tumor with a low prevalence (only 1015%). Seizures are focal in the
majority of the cases. A particular problem is
represented by the children under 3 years old. In
this group, epilepsy is more frequent (up to 70%);
it can persists after the surgical intervention and
in some cases it evolves into an epileptic encephalopathy. Epilepsy can began after surgery in
other cases (1040%), as a consequence of brain
damage. The prevalence is higher in children
over 3 years old.
Drug-resistant epilepsy is secondary to a
brain tumor in 2650% of cases in the pediatrics.
The results of the surgery is more favorable in
children with brain tumors and epilepsy. The
prevalence of seizure-free cases after surgery
is 8090%, compared with 6070% of the
cases with other etiology. The post surgical evolution is also more favorable for patients with
some particular tumors (DNT, gangliogliomas).
19
The pharmacological therapy of the epilepsy

related with tumors is difficult with many side
effects and interactions with chemotherapy. The
new AEDs are particularly promising for the lower
incidence of side effects and a good efficacy.
References
Benifla M, Otsubo H, Ochi A, Weiss SK, Donner EJ,
Shroff M, Hawkins C, Drake JM, Elliot I, Smith ML,
Snead OC, Rutka JT (2006) Temporal lobe surgery for
intractable epilepsy in children: an analysis of outcomes in 136 children. Neurosurgery 59:12031213
Benifla M, Rutka JT, Otsubo H, Lamberti-Pasculli M,
Elliot I, Sell E, RamachandranNair R, Ochi A, Weiss
SK, Snead OC, Donner EJ (2008) Long-term seizure
and social outcomes following temporal lobe surgery
for intractable epilepsy during childhood. Epilepsy
Res 82:133138
Blaheta RA, Michaelis M, Driever PH, Cinatl J Jr (2005)
Evolving anticancer drug valproic acid: insights into
the mechanism and clinical studies. Med Res Rev
25:383397
Chang EF, Christie C, Sullivan JE, Garci PA, Berger MS,
Barbaro NM (2010) Seizure outcomes after resection
of dysembryoplastic neuroepithelial tumor in 50
patients clinical article. J Neurosurg 5:123130
Chung SK, Wang KC, Nam DH, Cho BK (1998) Brain
tumor in the first year of life: a single institute study.
J Korean Med Sci 13:6570
Cossu M, Lo Russo G, Francione S, Mai R, Nobili L,
Sartori I, Tassi L, Citterio A, Colombo N, Bramerio
M, Galli C, Castana L, Cardinale F (2008) Epilepsy
surgery in children: results and predictors of outcome
on seizures. Epilepsia 49:6572
Gaggero R, Haupt R, Fondelli P, De Vescovi R, Marino A,
Lanino E, Dallorso S, Faraci M (2006) Intractable epilepsy secondary to cyclosporine toxicity in children
undergoing allogeneic hematopoietic bone marrow
transplantation. J Child Neurol 21:861866
Gaggero R, Consales A, Fazzini F, Mancardi MM,
Baglietto MG, Nozza P, Rossi A, Pistorio A, Tumolo
M, Cama A, Garr ML, Striano P (2009) Epilepsy
associated with supratentorial brain tumors under
3 years of life. Epilepsy Res 87:184189
Gilles FH, Sobel E, Leviton A, Hedley-Whyte ET,
Tavare CJ, Sobel RA (1992) Epidemiology of seizures
in children with brain tumors. The Childhood Brain
Tumor Consortium. J Neurooncol 12:5368
Glantz MJ, Cole BF, Forsyth PA, Recht LD, Wen PY,
Chamberlain MC, Grossman SA, Cairncross JG
(2000) Special article practice parameter: anticonvulsant prophylaxis in patients with newly diagnosed
brain tumors. Neurology 54:18861893
Khan RB, Onar A (2006) Seizure recurrence and risk
factors after antiepilepsy drug withdrawal in children
with brain tumors. Epilepsia 47:375379
205
Khan RB, Boop FA, Onar A, Sanford RA (2006) Seizures

in children with low-grade tumors: outcome after
tumor resection and risk factors for uncontrolled seizures. J Neurosurg 104(supp 6):377382
Kim SK, Wang KC, Hwang YS, Kim KJ, Cho BK (2001)
Intractable epilepsy associated with brain tumors in
children: surgical modality and outcome. Childs Nerv
Syst 17:445452
Kral T, Kuczaty S, Blmck I, Urbach H, Clusmann H,
Wiestler OD, Elger C, Schramm J (2001) Postsurgical
outcome of children and adolescents with medically
refractory frontal lobe epilepsies. Childs Nerv Syst
17:595601
Liigant A, Haldre S, Oun A, Linnamgi U, Saar A, Asserm
T, Kaasik AE (2001) Seizures disorders in patients
with brain tumors. Eur Neurol 45:4651
Makino K, Nakamura H, Yano S, Kuratsu J, Kumamoto
Brain Tumors Group (2010) Population-based epidemiological study of primary intracranial tumors in
childhood. Childs Nerv Syst 26:10291034
Mehrotra N, Shamji MF, Vassilyadi M, Ventureyra EC
(2009) Intracranial tumors in first year of life: the
CHEO experience. Childs Nerv Syst 25:15631569
Michasky A, Garr ML (2004) Infants tumors in brain and
spine. In: Walker DA, Perilongo G, Punt JAG, Taylor
R (eds) Tumors child. Arnold Publisher, London, pp
359369
Nolan MA, Sakuta R, Chuang N, Otsubo H, Rutka JT,
Snead OC, Hawkins CE, Weiss SK (2004)
Dysembryoplastic neuroepithelial tumors in childhood: long-term outcome and prognostic features.
Neurology 62:22702276
Ogiwara H, Nordii DR, Di Patri AJ, Alden TD, Bowman
C, Tomita T (2010) Pediatric epileptogenic gangliogliomas: seizure outcome and surgical results. J
Neurosurg Pediatr 5:271276
Ozien F, Gunduz A, Asan Z, Tandiveri T, Ozkara C, Yeni
N, Yalcinkaya C, Ozyurt E, Uzan M (2010)
Dysembryoplastic neuroepithelial tumors and gangliogliomas: clinical results of 52 patients. Acta
Neurochir (Wien) 152:16611671
Porter KR, Mc Carthy BJ, Freels S, Kim Y, Davis FG
(2010) Prevalence estimates for primary brain tumors
in the United States by age, gender, behavior, and histology. Neuro-oncology 2:520527
Riva M, Salmaggi A, Marchioni E, Silvani A, Tomei G,
Lorusso L, Merli R, Imbesi F, Russo A, For the
Lombardia Neurooncology Group (2006) Tumorassociated epilepsy: clinical impact and the role of
referring centres of a cohort of glioblastoma patients.
A multicentre study from the Lombardia Neurooncology Group. J Neurol Sci 27:345351
Rosemberg S, Fujiwara D (2005) Epidemiology of
pediatric tumors of the nervous system according to
the WHO 2000 classification: a report of 1195 cases
from a single institution. Childs Nerv Syst
21:940944
Ruggiero A, Rizzo D, Mastrangelo S, Battaglia D,
Attin G, Riccardi R (2010) Interactions between antiepileptic and chemotherapeutic drugs in children with
206
brain tumors: is it time to change treatment? Pediatr
Rutledge SL, Snead OC, Morawetz R, Chandra-Sekar B
(1987) Brain tumors presenting as seizure disorder in
infants. J Child Neurol 2:214219
Shady JA, Black PM, Kupsky WJ, Tarbell NJ, Scott RM,
Leong T, Holmes G (1994) Seizures in children with
supratentorial astroglial neoplasms. Pediatr Neurosurg
21:2330
Shaller B, Yaniv I, Michowitz S, Kornreich L, Schwartz
M, Goldberg-Stern H, Cohen IJ (2003) Epilepsy associated with paediatric brain tumors: the neurooncologic perspective. Pediatr Neurol 29:232235
Sogawa Y, Kan L, Levy AS, Maytal J, Shinnar S (2009)
The use of antiepileptic drugs in pediatric brain tumor
patients. Pediatr Neurol 41:192194
Stevens GH (2006) Antiepileptic therapy in patients with
central nervous system malignancies. Curr Neurol
Neurosci Rep 6:311318
Tremont-Lukats IW, Ratilal BO, Armstrong T, Gilbert
MR (2008) Antiepileptic drugs for preventing seizures
in people with brain tumors. Cochrane Database Syst
Rev 16(2):CD004424
Ullrich NJ (2009) Neurologic sequelae of brain tumors in
children. J Child Neurol 24:14461454
R. Gaggero et al.
Van Breemen MS, Wilms EB, Vecht CJ (2007) Epilepsy
in patients with brain tumours: epidemiology, mechanisms, and management. Lancet Neurol 6:421430
Van Oijen M, De Waal H, Van Rijen PC, Jennekens A,
Van Huffelen AC, Van Nieuwenhuzen O, Dutch
Collaborative Epilepsy Surgery Program (2006)
Resective epilepsy surgery in childhood: the Dutch
experience 19922002. Eur J Paediatr Neurol
10:114123
Vecht CJ, Wagner GL, Wilms EB (2003) Treating seizures
in patients with brain tumors: drug interactions
between antiepileptic and chemotherapeutic agents.
Semin Oncol 30(Suppl 19):4952
Wang EC, Geyer JR, Berger MS (1994) Incidence of postoperative epilepsy in children following subfrontal
craniotomy for tumor. Pediatr Neurosurg 21:165172
Wilne SH, Collier J, Kennedy CR, Koller K, Grundy R,
Walker D (2007) Presentation of childhood CNS
tumours: a systematic review and meta-analysis.
Lancet Oncol 8:685695
Zentner J, Hufnagel A, Wolf HK, Ostertun B, Beherens F,
Campos MG, Elger CE, Wiestler OD, Schramm J
(1997) Surgical treatment of neoplasms associated
with medically intractable epilepsy. Neurosurgery
41:378383
Postoperative Pain in Children:

Advantage of Using Nonnarcotic
Analgesic Regimen
20

Abstract
Contents
Introduction ............................................................
207
Postoperative Pain in Children: Advantage

of Using Nonnarcotic Analgesic
Regimen ..................................................................
208
References ...............................................................
209
Postoperative morphine is commonly used to

control pain in children following neurosurgical procedures. We have previously reported
our success in treating postoperative pain in
children undergoing neurosurgical procedures
including craniotomy. This cohort underwent
a scheduled regimen of acetaminophen
[10 mg/kg] and ibuprofen [10 mg/kg] alternating every 2 h. Pain scores were significantly
lower in this group compared to a retrospective review of other same and similar procedures. Additionally, the length of hospital stay
was shorter in these patients and antiemetic
requirements were lower. Based on our experience, a regimen of minor analgesic therapy,
given in alternating doses every 2 h immediately
following craniotomy and throughout hospitalization, significantly reduces postoperative
pain scores, hospitalization, and antiemetic
requirements.
Introduction
R.S. Tubbs (*) M.M. Mortazavi A.A. Cohen-Gadol

Pediatric Neurosurgery, Children Hospital,
1600 7th S, ACC 400, Birmingham, AL 35233, USA
e-mail: Shane.tubbs@chsys.org
Inadequate control of pain following surgical

procedures may result in physical and mental
stress (Warren et al. 2010). Narcotic use following surgical procedures has many drawbacks
such as potential suppression of breathing, pruritus, nausea, emesis, gastrointestinal tract stasis,
euphoria, urinary retention and constipation
(Hesselgard et al. 2006; Lawhorn et al. 1995;

207
208
Warren et al. 2010). In an attempt to maximize

pain control and minimize complications of cranial surgery, Hudgins and Gilreath (2001) found
that in 17 children undergoing a posterior cranial
fossa decompression for a Chiari I decompression that oral analgesia (acetaminophen/ibuprofen)
given every 2 h in an alternating fashion was sufficient for pain control and decreased the hospital
stay in all children (average hospital stay of
24.7 h). They also reported no complications
from this drug regimen. Similarly and more
recently, we reported a prospective study demonstrating that this same alternating use of oral
analgesia was efficacious in pain control in a
group of 50 children following posterior cranial
fossa decompression for a Chiari I malformation
(Smyth et al. 2004). These children had lower
postoperative pain scores, less incidence of
postoperative emesis and nausea and shorter
hospital stays.
Postoperative Pain in Children:

Advantage of Using Nonnarcotic
Analgesic Regimen
Postoperative pain following craniotomy has
been well studied in adults (Ortiz-Cardona and
Bendo 2007; Quiney et al. 1996; Rahimi et al.
2006; Roberts 2005; Shirley 2000; Stoneham and
Walters 1995; Verchere et al. 2002). We have
studied postoperative pain in children undergoing
a number of neurosurgical procedures including
Chiari I decompression (Smyth et al. 2004), brain
tumor biopsy and resection (Bauer et al. 2010),
and dorsal rhizotomy (Tubbs et al. 2007).
Arguably, one of the most painful procedures
to perform is the transection of dorsal rootlets. In
an earlier study (Tubbs et al. 2007), we aimed to
maximize postoperative pain relief in children
undergoing dorsal rhizotomy for spastic diplegia
where these rootlets were cut in order to decrease
patient spasticity. Postoperative pain scores and
the necessity for epidural morphine were analyzed in a group of 22 consecutive children (age
range 410 years; mean 6 years) with cerebral
palsy who underwent a partial dorsal rhizotomy
for spasticity over an 19 month period. All patients
R.S. Tubbs et al.
(10 males and 12 females) were ambulators who

underwent a two to three level lumbar laminotomy for exposure of the conus medullaris and
proximal cauda equina. The dorsal roots of L1 to
S1 were identified and stimulated. Simultaneous
electrophysiological and manual examination of
the lower extremities was performed. Dorsal
roots were transected if they were believed to be
pathologic. Approximately 7580% of all dorsal
roots at these levels was transected. Per our institution, each patient had a standard intraoperative
indwelling epidural catheter placed for the potential infusion of postoperative preservative free
morphine [Duramorph 80 mg/kg (Elkins Sinn,
Cherry Hill, NJ) and Stadol 40 mg/kg (Sandoz,
Broomfield, CO)]. Additionally, this cohort underwent a scheduled regimen of minor oral analgesic
medications (standing doses of acetaminophen
[10 mg/kg] and ibuprofen [10 mg/kg] alternating
every 2 h). Following this procedure, all patients
were maintained flat in the intensive care unit
until they were maintained 24 h without the infusion of epidural morphine. Once this criterion
was met, the epidural catheter was removed and
the patient was transferred to the ward until they
were relatively pain free and could drink and urinate. For comparison, a retrospective review of
20 consecutive patients (age range 59 years;
mean 7 years, 12 females and 8 males) undergoing this same procedure over a 15 month period
by our senior author and receiving only epidural
morphine every 34 h as needed was performed.
Pain relief was evaluated by a standard visualanalog faces pain scale (Wong and Baker 1988).
Each patient was monitored by a single nurse and
in addition to the use of a visual-analog pain
scale, vital signs were monitored to evaluate for
potential breakthrough pain. The decision to
administer epidural morphine was decided by a
pediatric intensivist.
Only one patient on the alternating oral analgesic protocol was given a single postoperative
dose of morphine epidurally. None of the remaining patients required postoperative epidural morphine. In contrast, the mean number of epidural
infusions of morphine for retrospectively
reviewed patients was 4.2 (range 26). Mean pain
scores were also significantly lower in this group
20
Postoperative Pain in Children: Advantage of Using Nonnarcotic Analgesic Regimen
compared to reviewed patients undergoing this

same procedure and receiving epidural morphine
every 34 h prn (1 versus 3.5, respectively).
Hospitalization was also shorter (3.3 compared
with 4.1 days) and antiemetic requirements lower
for these patients. No side effects were observed
in this patient group. No statistical difference was
found between age, gender or pre and postoperative vital signs in the group given alternating oral
analgesics.
Dorsal rhizotomy in children for spasticity is
associated with challenging postoperative pain
management for multiple reasons (Dews et al.
1996; Lawhorn et al. 1995). These patients are
typically young (less than 5-years-old), which
often makes the interpretation of pain difficult.
Additionally, these patients not only have postoperative pain from musculoskeletal dissection but
also from manipulation of the cauda equina,
which is more difficult to control with oral opiates (Ross 1991). Patients using an oral analgesic
protocol as described herein would have the
advantage of avoiding intravenous administration
of pain medications that are intrinsically more
dangerous than oral pain medications (de Gray
and Matta 2005). These patients also have a
decreased chance of nosocomial infections by a
potentially shorter hospital stay. Moreover, children are less able than adults to communicate
their level of discomfort, therefore, a scheduled
pain regimine may be preferred. Although epidural infusion of morphine has been shown to be
more efficacious than intermittent intravenous
doses of morphine following selective dorsal
rhizotomy in children (Malviya et al. 1999), our
results were so positive, that one could consider
not placing an epidural catheter at operation. This
would negate the possibility of complications
such as catheter migration, and medication errors
and decrease nursing responsibilities (Harris
et al. 1991).
We have found that a scheduled non-narcotic
protocol given to children following various neurosurgical procedures in children is efficacous
and in the vast majority of patients, obviates the
need for epidural or oral opiates. Nurses who are
required to administer nonnarcotic analgesics
every 2 h believed that children had less pain.
209
In our experience, administration of nonnarcotic

analgesics every 2 h results in a less lengthy
hospital stay following neurosurgery in children.
Such a regimen may decrease associated complications of postoperative pain control following
such procedures.
References
Bauer DF, Waters AM, Tubbs RS, Rozzelle CJ, Wellons
JC, Blount JP, Oakes WJ (2010) Safety and utility of
scheduled nonnarcotic analgesic medications in children undergoing craniotomy for brain tumor.
de Gray LC, Matta BF (2005) Acute and chronic pain following craniotomy: a review. Anaesthesia 60:693704
Dews TW, Schubert A, Fried A, Ebrahim Z, Oswalt K,
Paranandi L (1996) Intrathecal morphine for analgesia
in children undergoing selective dorsal rhizotomy.
J Pain Symptom Manage 11:188194
Harris MM, Kahana MD, Park TS (1991) Intrathecal morphine for postoperative analgesia in children after selective dorsal root rhizotomy. Neurosurgery 28:519522
Hesselgard K, Strmblad LG, Romner B, Reinstrup P
(2006) Postoperative continuous intrathecal pain treatment in children after selective dorsal rhizotomy with
bupivacain and two different morphine doses. Pediatr
Anesth 16:436443
Hudgins RJ, Gilreath L (2001) Chiari 1 decompression as
an outpatient procedure. In: American society of pediatric neurosurgeons scientific program, 24th annual
meeting. Maui, HI
Lawhorn CD, Boop FA, Brown RE, Andelman PD,
Schmitz ML, Kymer PJ, Shirey R (1995) Continuous
epidural morphine/butorphanol infusion following
selective dorsal rhizotomy in children. Childs Nerv
Syst 11:621624
Malviya S, Pandit UA, Merkel S, Voepel-Lewis T, Zang
L, Siewert M, Tait AR, Muraszko K (1999) A comparison of continuous epidural infusion and intermittent intravenous bolus doses of morphine in children
undergoing selective dorsal rhizotomy. Reg Anesth
Pain Med 24:438443
Ortiz-Cardona J, Bendo AA (2007) Perioperative pain
management in the neurosurgical patient. Anesthesiol
Clin 25:655674
Quiney N, Cooper R, Stoneham M, Walters F (1996)
Pain after craniotomy. A time for reappraisal? Br
Rahimi SY, Vender JR, Macomson SD, French A, Smith
JR, Alleyne CH (2006) Postoperative pain management after craniotomy: evaluation and cost analysis.
Roberts GC (2005) Post-craniotomy analgesia: current
practices in British neurosurgical centres a survey of
post-craniotomy analgesic practices. Eur J Anaesthesiol
22:328332
210
Ross D (1991) Intrathecal morphine for postoperative
analgesia in children after selective dorsal rhizotomy.
Shirley P (2000) Pain relief post craniotomy: a balanced
approach? Anaesthesia 55:409410
Smyth MD, Banks JT, Tubbs RS, Wellons JC, Oakes WJ
(2004) Efficacy of scheduled nonnarcotic analgesic
mediations in children after suboccipital craniectomy.
J Neurosurg 100(2 Suppl):183186
Stoneham M, Walters F (1995) Post operative analgesia
for craniotomy patients: current attitudes among neuroanaesthetists. Eur J Anaesthesiol 2:571573
Tubbs RS, Law C, Davis D, Oakes WJ (2007) Scheduled
oral analgesics and the need for opiates in children
R.S. Tubbs et al.

following partial dorsal rhizotomy. J Neurosurg 106
(6 suppl):439440
Verchere E, Grenier B, Mesli A, Siao D, Sesay M,
Maurette P (2002) Postoperative pain management
after supratentorial craniotomy. J Neurosurg
Anesthesiol 14:96101
Warren DT, Bowen-Roberts T, Ou C, Purdy R, Steinbok
P (2010) Safety and efficacy of continuous morphine infusions following pediatric cranial surgery
in a surgical ward setting. Childs Nerv Syst 26:
15351541
Wong DL, Baker CM (1988) Pain in children: comparison
of assessment scales. Pediatr Nurs 14:917
Pediatric Brain Tumors: Application

of Stratication Criteria to Rene
Patient Management
21
Ian F. Pollack
Contents
Abstract
Introduction ............................................................
211
Low-Grade Gliomas...............................................
212
High-Grade Gliomas..............................................
213
Brainstem Gliomas.................................................
215
Medulloblastoma/PNET ........................................
216
Brain Tumors in Infants ........................................
218
Conclusion ..............................................................
220
References ...............................................................
220
Current treatment strategies for children with

brain tumors focus on improving outcome
for tumor types that have historically been
relatively resistant to therapy, and reducing
treatment-related sequelae for children with
therapy-responsive tumors. Refinements in
clinical and molecular stratification for many
types of childhood brain tumors have facilitated these efforts to achieve risk-adapted
treatment planning. In some instances, molecular characterization approaches have also
yielded insights into new therapeutic targets.
This chapter reviews advances in stratification
approaches for several of the most common
types of childhood brain tumors, including
high- and low-grade gliomas, medulloblastomas and other primitive neuroectodermal
tumors (PNETs), and infant brain tumors, and
discusses how new information regarding the
biological features critical to tumorigenesis is
being translated into novel therapeutic
approaches for these challenging tumors.
Introduction
I.F. Pollack (*)

Department of Neurosurgery, Childrens Hospital of
Pittsburgh, 4401 Penn Avenue, Pittsburgh,
PA 15224, USA
e-mail: Ian.pllack@chp.edu
Brain tumors are the most common solid tumors

of childhood, and are currently the leading cause
of childhood cancer-related mortality (Pollack
1994). Although there have been significant
improvements during the last two decades in the
outcome for certain types of childhood brain

211
212
tumors, such as medulloblastoma, other groups,

such as malignant gliomas and diffuse intrinsic
brainstem gliomas, continue to have an extremely
poor prognosis (Jennings et al. 2002). For the
more favorable prognosis tumor subgroups, there
has been increasing recognition that cure often
comes at a cost of significant late sequelae, which
can impair long-term quality of life. Accordingly,
current management strategies focus on
attempting to improve the rate of long-term
survival in children with poor prognosis
tumors and to reduce the long-term side effects
of treatment in more favorable-risk tumors. The
current chapter outlines how these approaches
are being applied in several of the most common
groups of childhood brain tumors, specifically
low- and high-grade gliomas and primitive neuroectodermal tumors, and highlights instances
in which advances in clinical and biological
risk-based categorization and molecularly based
treatment strategies are being incorporated into
therapeutic trials.
Low-Grade Gliomas
Low-grade gliomas are a diverse group that constitutes the largest subset of childhood brain
tumors. These include pilocytic astrocytomas and
subependymal giant cell astrocytomas (SEGAs),
which generally are classified as grade I lesions,
and fibrillary and pilomyxoid astrocytomas,
which are considered grade II lesions. Until
recently, pilocytic astrocytomas were thought to
be largely devoid of consistent genetic aberrations. However, several studies during the last
few years have determined that these tumors
characteristically exhibit alterations in the BRAF
gene, most commonly involving translocations
between BRAF and KIAA, or activating mutations, such as BRAFv600E, which leads to activation of the MAP kinase signaling pathway (Jones
et al. 2008). Subependymal giant cell astrocytomas characteristically arise in the setting of
tuberous sclerosis and exhibit mutations in the
TSC1 and TSC2 genes, leading to dysregulated
activation of mTOR signaling (Lam et al. 2010).
As with the BRAF anomalies that characterize
I.F. Pollack
pilocytic tumors, these consistent genomic alterations

provide exciting new targets for molecularly
directed therapeutic strategies. In contrast to the
above two groups, the molecular basis for childhood fibrillary low-grade astrocytomas remains
less well defined. Although in adults such tumors
represent an early stage in a pathway of tumorigenesis that often ends in higher grade lesions
(Parsons et al. 2008), such a phenotype is less
commonly observed in childhood lesions.
Notwithstanding the histological and molecular diversity among the different subgroups of
low-grade gliomas, the factor that is most strongly
associated with outcome in all low-grade gliomas
is the extent of surgical tumor removal. In a large
natural history study of the Childrens Cancer
Group and Pediatric Oncology Group (CCG9891/
POG8930), 5-year progression-free survival was
more than 90% in children with low-grade
gliomas that had undergone gross total resection,
whereas approximately half of children with less
extensive tumor removal had disease progression
during that interval (Sanford et al. 2002). As a
result, a variety of surgical adjuncts, such as
image-guidance techniques, intraoperative imaging, and neurophysiologic monitoring, are sometimes employed in an effort to enhance the
likelihood of safely achieving an extensive resection. Because of the striking relationship between
resection extent and outcome, it has been difficult
to determine conclusively whether there is an
association between histology and prognosis.
Although pilocytic astrocytomas seem to have a
more favorable prognosis than fibrillary lesions,
this may reflect that pilocytic tumors, particularly
when superficially located, are often reasonably
well circumscribed and more likely to be amenable to gross total resection (Sanford et al.
2002). Likewise, superficial lesions involving the
cerebral and cerebellar cortices tend to have a
better prognosis than deep lesions involving the
thalamus, basal ganglia, optic pathways and
brainstem, probably relating to the fact that they
are more amenable to extensive removal without
excessive morbidity.
Because patients whose tumors have undergone gross total resection have a greater than
90% long-term survival rate, adjuvant therapy is
21
Pediatric Brain Tumors: Application of Stratification Criteria to Refine Patient Management
usually not required for such lesions, which

include the majority of cerebral hemisphere and
cerebellar low-grade gliomas. In contrast, the
management is much more controversial for
lesions not amenable to complete removal,
because of the high risk of recurrence. This is a
particularly common scenario for midline
lesions involving sites such as the hypothalamus
and optic pathways, which are usually not amenable to complete removal because of their
involvement in critical surrounding structures.
Management of such tumors is further complicated by the fact that they commonly arise in
young children, who are at high risk of longterm sequelae from side effects of wide-field
irradiation.
In a series of recent studies, several chemotherapy regimens have been noted to have efficacy in delaying or avoiding the need for
radiotherapy in children with progressive or highrisk incompletely resected tumors (Packer et al.
1997; Ater et al. 2007). The recently completed
A9952 study performed a phase 3 randomized
comparison of two active regimens, carboplatin/
vincristine and 6-thioguanine-procarbazinelomustine-vincristine. Although it appeared that
both regimens of this study had activity in delaying tumor progression, a substantial percentage
of patients did eventually suffer progressive disease, highlighting the need for additional treatment options (Ater et al. 2007). Accordingly, a
series of pilot studies have been initiated to
explore additional therapeutic options that might
be included in a subsequent phase 3 protocol.
ACNS0223 examined the feasibility of administering temozolomide in addition to carboplatin/
vincristine. ADVL0515 examined the use of
vinblastine as an alternative to vincristine in
the carboplatin-containing regimen, in view of
the activity of this agent when administered
alone. Finally, ACNS0221 is evaluating the efficacy of conformal radiotherapy in children older
than 10 years with progressive tumors and in
younger children with chemotherapy-refractory
tumors to determine whether this can be administered with acceptable side effects and can lead to
long-term disease control. In addition, a number
of biological agents are now being examined in
213
these tumors, including antiangiogenic agents,

such as bevacizumab and lenalidomide, and
growth signaling inhibitors, such as rapamycin
analogues and AZD6244, which are directed
against the dysregulated mTOR and MAPK pathways, respectively, and are commonly observed
in these tumors.
High-Grade Gliomas
Malignant (high-grade) gliomas are subdivided
into grade III anaplastic astrocytomas, oligodendrogliomas, and oligoastrocytomas, and grade IV
glioblastomas and gliosarcomas. Although significant research has been directed at defining the
molecular pathways of tumorigenesis in highgrade gliomas that arise in adults, comparatively
little information is available in pediatric lesions.
Adult lesions have characteristically been subdivided into so-called primary lesions that arise
de novo as grade IV tumors, which typically
exhibit amplification and often rearrangement of
the EGFR gene and deletion of PTEN, secondary
lesions that progress from low-grade fibrillary
astrocytomas to grade III and ultimately grade IV
lesions in a stepwise fashion, which typically
have mutations of TP53 and IDH1 or IDH2 as
early genetic anomalies, and oligodendroglial
tumors, which often exhibit deletions of chromosomes 1p and 19q (Parsons et al. 2008; Hartmann
et al. 2009). In this regard, our previous studies
have noted TP53 mutations in ~ half of childhood
malignant gliomas, comparable to the frequency
in adult secondary malignant astrocytomas
(Pollack et al. 2002). However, pediatric malignant gliomas rarely arise from apparent lowgrade precursors and infrequently exhibit
mutations in the IDH1 or IDH2 genes (Parsons
et al. 2008; Hartmann et al. 2009), which suggests that despite their similarities in terms of
TP53 alterations, childhood high-grade gliomas
arise by a distinct mechanism from adult secondary malignant gliomas. Childhood lesions are
also biologically distinct from adult primary
malignant gliomas because they infrequently
exhibit deletions or mutations of the PTEN gene
or amplification of EGFR (Pollack et al. 2006a).
214
In view of recent reports that highlight the

existence of multiple pathways of tumorigenesis
in adults (Verhaak et al. 2010), it is likely that pediatric lesions are not only genetically distinct from
many adult lesions, but may themselves encompass several parallel pathways of tumorigenesis.
As with adult malignant gliomas, the prognosis
for children with these tumors remains poor,
despite improved surgical techniques and application of newer approaches for delivering irradiation. Although the addition of chemotherapy with
lomustine and vincristine to postoperative irradiation was demonstrated in the Childrens Cancer
Group (CCG)-943 study to improve survival
compared to the use of irradiation alone (Sposto
et al. 1989), subsequent studies with the more
complex 8 in 1 regimen in the CCG-945 study
failed to further improve outcome (Finlay et al.
1995). In addition, use of more intensive pre- and
post-irradiation regimens did not improve
outcome and in some instances were associated
with unacceptable rates of toxicity (MacDonald
et al. 2005).
The two clinical factors that were associated
with outcome in these studies have been histology
and extent of tumor resection. Not unexpectedly,
patients with grade IV lesions (i.e., glioblastoma)
had a worse prognosis than those with anaplastic
astrocytoma or other grade III lesions, and those
with oligodendroglial tumors seemed to have a
better prognosis than other subgroups of malignant gliomas (Finlay et al. 1995). In addition,
patients with deep-seated or highly infiltrative
tumors that were not amenable to extensive resection had a worse prognosis than those with more
resectable lesions. Further analysis of the CCG945 cohort also demonstrated that a number of
molecular features correlated with a worse prognosis, including overexpression and/or mutation
of the tumor suppressor gene TP53 (Pollack et al.
2002), high expression of proliferation markers
such as MIB1, and overexpression of methylguanine DNA methyltransferase (MGMT), which
interferes with the activity of alkylating chemotherapeutic agents, such as the nitrosoureas
(Pollack et al. 2006b).
Based upon subsequent studies in adults that
noted improvements in outcome from administering
I.F. Pollack
chemotherapy with temozolomide both during

and after radiotherapy versus treatment with irradiation alone, a series of pediatric phase 2 studies
have examined this approach. The ACNS0126
study incorporated daily administration of temozolomide during irradiation followed by treatment
on a 5-day per 28-day schedule thereafter, coupled with correlative analysis of tumor MGMT
and mismatch repair status (Cohen et al. 2007).
Although outcome results from the ACNS0126
study appear to be comparable to those obtained
in adults, they were not better than those achieved
in the CCG-945 study using CCNU and vincristine.
As in the CCG-945 study, MGMT overexpression
was noted to be an adverse prognostic factor,
associated with a poor response to radiation plus
temozolomide. A follow-up study (ACNS0423)
combined both lomustine and temozolomide,
given the relatively favorable rate of 1-year survival
observed in a pilot study that used this combination (Jakacki et al. 2008). The association between
MGMT expression and outcome will also be
assessed in the ACNS0423 study, and if a strong
association is again demonstrated, this would
provide a rationale for stratifying therapy based
on these features in subsequent studies, if alkylatorbased chemotherapy is used.
A study currently under development is
proposing to examine the use of other agents in
conjunction with irradiation, followed by the
antiangiogenic agent bevacizumab plus irinotecan after irradiation. Correlative studies
incorporating microarray-based genotyping
and expression profiling are also being pursued
to parallel the recent extensive analyses that have
been completed in adult malignant gliomas
(Parsons et al. 2008; Verhaak et al. 2010) as a
way to identify genes associated with tumor
progression in pediatric malignant gliomas and
to potentially identify heretofore unrecognized
therapeutic targets. In this regard, the COG and
the Pediatric Brain Tumor Consortium (PBTC)
have initiated phase 1 and 2 studies of agents that
inhibit several of the signaling pathways implicated in glioma growth, such as platelet-derived
growth factor receptor and epidermal growth
factor receptor, signal transduction intermediates,
and inhibitors of angiogenic signaling for patients
21
with recurrent disease. Promising results would

be considered for application in newly diagnosed
patients.
Brainstem Gliomas
Brainstem gliomas are subdivided into focal and
diffuse lesions, which have dramatically different
biological behavior and therapeutic outcome.
Appropriate categorization of these tumors was
sometimes challenging in the era prior to the
availability of high-resolution imaging, but with
the advent of MRI the majority of these lesions
can now be properly separated into distinct risk
groups noninvasively. In this regard, one notable
advancement in management has been that neurosurgeons and neuro-oncologists have became
increasingly adept at identifying focal tumors,
such as dorsally exophytic brainstem gliomas and
focal lesions of the midbrain, medulla, and cervicomedullary junction, which are generally lowgrade histologically. Such tumors are typically
treated like other low-grade gliomas in that
accessible lesions, such as dorsally exophytic
brainstem gliomas, are often managed with surgical resection. If an extensive resection has been
achieved, expectant management with observation alone is often pursued. For more deep-seated
brainstem low-grade gliomas, which may not be
amenable to complete resection, the same issues
apply as noted earlier for non-brainstem low-grade
gliomas, in terms of the use of focal conformal
irradiation or chemotherapy, depending on the
age of the patient.
In contrast to the reasonably favorable prognosis of low-grade focal brainstem gliomas, the
outcome for children with diffuse intrinsic
brainstem gliomas remains exceedingly poor.
Historically, these tumors have been treated with
irradiation, which provides an interval of symptom resolution in many patients. Early therapeutic studies for these lesions examined the safety
and efficacy of escalating the dose of irradiation
using hyperfractionated delivery approaches.
Although these studies demonstrated that escalation of the radiation dose to as high as 7,800 cGy
was often tolerated, this approach had no impact
215
on progression-free or overall survival duration

(Jennings et al. 2002).
In an effort to improve outcome, a series of
studies examined the use of pre- and/or postirradiation chemotherapy for these tumors, but
unfortunately no favorable effects were observed
with a variety of agents, even when administered
at high doses (Jennings et al. 2002). Accordingly,
more recent studies have evaluated approaches
for radiosensitization or administering chemotherapy concurrently with irradiation in an
effort to potentiate the effect of radiotherapy.
Unfortunately, the ACNS0126 study of temozolomide with irradiation, which incorporated a
stratum for patients with brainstem gliomas, did
not detect a significant improvement in the rate
of long-term survival. Likewise a European study
of topotecan during radiotherapy failed to observe
a significant survival benefit. A study of the
radiosensitizer gadolinium texaphyrin during
irradiation (ACNS0222), which incorporated
the maximally tolerated dose determined by the
A09712 phase 1 study (Bradley et al. 2008), is
currently under analysis. In parallel with these
efforts, studies by the Pediatric Brain Tumor
Consortium have examined several molecularly
targeted treatment strategies in conjunction with
irradiation. Studies with the PDGFR inhibitor
imatinib (PBTC-006), the EGFR inhibitor
gefitinib (PBTC-007), and the farnesyltransferase
inhibitor zarnestra (PBTC-014) have been completed, but the results to date have been disappointing (Pollack et al. 2007). A study is currently
in progress of a conceptually different molecularly
targeted strategy, using capecitabine, a prodrug
of 5-fluoro-uracil, which may be selectively
metabolized to the active agent by the increased
thymidine phosphorylase activity seen in gliomas,
an effect further enhanced by irradiation.
One of the ongoing challenges to progress in
the management of diffuse intrinsic brainstem
gliomas has been the lack of direct biological
information from the tumor target to help identify
additional therapeutic targets. Because the vast
majority of diffuse intrinsic brainstem gliomas
can be diagnosed by imaging findings alone in
the context of appropriate clinical symptoms,
biopsies are not routinely used to establish the
216
histological diagnosis, which has limited the

availability of tissue to provide insights regarding
relevant targets for future therapeutic studies.
Most biological studies involving these tumors
have made use of archival autopsy specimens,
and examined a narrow panel of targets, such as
EGFR and p53. More recently a series of more
real-time autopsy-based analyses have been
undertaken, which have facilitated collection of
higher quality, better preserved tumor material,
allowing the application of microarray-based
DNA copy number and gene expression analyses.
In addition, several studies in Europe have incorporated stereotactically directed brainstem glioma
biopsies at diagnosis as a way to obtain biologically informative tumor material. Although this
approach remains controversial, the relatively
low rates of morbidity in these studies have
encouraged some groups in North America to
also consider the feasibility of using biopsy data
as a way of guiding the selection of molecularly
targeted therapeutic approaches in individual
patients in an effort to improve upon the dismal
rates of response and long-term survival in children
with these tumors.
Medulloblastoma/PNET
Primitive neuroectodermal tumors (PNETs),
such as medulloblastoma, pineoblastoma, and
supratentorial PNETs, are the most common
childhood malignant brain tumors. It has long
been controversial as to whether the large group
of CNS small blue cell tumors represented distinct entities based on lesion location, or were
different manifestations of a common underlying
molecular pathway of tumor development. Recent
genomic studies seem to support the former interpretation in that the pattern of genomic abnormalities and gene expression alterations of
cerebral PNETs differ from those of medulloblastomas (Pomeroy et al. 2002). However, these
studies also suggest an even higher level of complexity, in that multiple distinct molecular signatures have been noted within individual tumor
subgroups, which has called attention to the need
I.F. Pollack
for genomically based tumor classifications

(Thompson et al. 2006; Kool et al. 2008).
In this regard, Thompson et al. (2006) noted
that based on unsupervised analysis of gene
expression profiles, medulloblastomas partitioned into five subgroups, in part determined by
mutations in the Wingless (WNT) pathway and
deletion of chromosome 6 (subgroup B) and
mutations in the Sonic Hedgehog (SHH) pathway
(subgroup D), among others. Similarly, Kool
et al. (2008) identified five distinct subsets based
on analysis of expression and comparative
genomic hybridization (CGH) arrays. They
defined one subtype (A) based on alterations in
WNT signaling, which was associated with mutations in the b-catenin gene and monosomy of
chromosome 6, a second (B) based on alterations
in SHH signaling, which was associated with
mutations in PTCH1 and loss of chromosome 9p,
and three other groups (C-E) based on expression
of neuronal differentiation genes and/or photoreceptor genes. Chromosome 17 aberrations
occurred in type C or D tumors and loss of the
inactivated X-chromosome was noted in female
cases of type C, D, and E tumors. These molecular groups also differed significantly in terms of
clinicopathological features in that metastatic
disease at diagnosis was most commonly seen in
C, D, and E tumors, whereas most cases with desmoplastic histology were in molecular type B.
Patients below 3 years of age had type B, D, or E
tumors. These results were then validated in a
second independent series of medulloblastomas
(Kool et al. 2008).
An ongoing challenge relates to the need to
integrate this complexity in biological stratification with the well known clinical risk stratification that has been used for these tumors. The
latter is based on a series of studies from the
CCG, POG, and SIOP cooperative groups that
noted significant differences in outcome as a
function of metastasis status, extent of tumor
resection, tumor location, and age following
treatment with standard doses of irradiation
(~3,600 cGy to the craniospinal axis with a boost
to a dose of 5,400 cGy to the tumor bed) (Tait
et al. 1990; Zeltzer et al. 1999). In these early
21
studies, 5-year progression-free survival rates

were ~60% for children older than 3 years of age
with extensively resected, non-metastatic [M0]
posterior fossa lesions (so-called average-risk
tumors) but less than 40% in patients younger
than 3 years and those with extensive residual
disease, metastases, or non-posterior fossa tumor
location (so-called high-risk tumors). These
observations provided an impetus for studies
that stratified therapy based on these clinical
risk factors, with the goal of improving survival
in the high-risk group and reducing the longterm side effects of treatment in the average-risk
group (Taylor et al. 2005; Packer et al. 2006;
Pizer et al. 2006).
Although initial efforts to reduce sequelae in
average-risk patients by reducing the craniospinal radiation dose to 2,340 cGy were associated
with a decrease in progression-free survival
(Thomas et al. 2000), subsequent pilot studies
that combined reduced-dose radiotherapy with
adjuvant chemotherapy noted preservation of
high rates of long-term survival with potentially
fewer radiation-related cognitive and endocrine
sequelae than treatment with standard doses of
irradiation alone. To follow up on these observations, the Childrens Oncology Group (COG) initiated a randomized phase 3 study (A9961) that
was designed to compare two adjuvant chemotherapy regimens for average-risk patients. This
study confirmed that reducing the dosage of craniospinal irradiation from 3,600 to 2,340 cGy in
conjunction with chemotherapy was not associated with an unacceptable drop in survival rates
(Packer et al. 2006), and provided an impetus for
a subsequent study (ACNS0331) that is examining whether doses and volumes of irradiation can
be further reduced with intensification of adjuvant chemotherapy. This ongoing study incorporates a stratified randomization design. For
children younger than 8 years, one aspect of the
study is evaluating the feasibility of further reducing the craniospinal radiotherapy dose from 2,340
to 1,800 cGy to diminish cognitive sequelae,
which are most severe in younger children, and a
second aspect is examining the safety of decreasing the volume of posterior fossa irradiation
using conformal delivery to decrease ototoxicity.
217
In children 8 years and older, a single randomization

for the boost volume size is incorporated.
This study also includes a battery of correlative analyses to evaluate molecular features that
have been found in recent retrospective studies to
identify prognostically distinct tumor subsets
independent of clinical factors (Pomeroy et al.
2002), as well as genome-wide screening of copy
number alterations and gene expression profiles
to look for patterns of abnormalities that can
further refine prognostic classification. These
studies are designed to identify molecular markers
that can detect tumors likely to recur despite
favorable clinical features, which would constitute a basis for biologically based stratification in
future studies. Given that recent publications
from several groups have noted that molecular
subsets of medulloblastomas may in part overlap
with clinically defined subsets, but may also convey prognostic information that supplements
clinical risk stratification (Thompson et al. 2006;
Kool et al. 2008), it is likely that future studies
will incorporate critical molecular factors to further refine patient inclusion criteria in the context
of clinical risk factors.
Thus, evolution in categorization as a function
of risk remains an ongoing effort. In that regard,
a recent review of the data from A9961 demonstrated that the subset of tumors with anaplastic
histological features was associated with a significantly worse prognosis than those with classical histology (Packer et al. 2006), which has led
to the amendment of current COG PNET protocols to group anaplastic tumors with other highrisk PNETs. Further refinement in inclusion
based on analysis of molecular features that separate tumors with anaplastic features from those
with classical histology is also likely to occur in
the near future, which will help to ensure that
patients who are enrolled on protocols that entail
reduction of intensity of therapy encompass the
most favorable risk group possible.
In parallel with the above strategies for average-risk medulloblastomas, which are directed
toward reducing treatment-related sequelae while
maintaining good long-term survival rates, the
focus of study designs for high-risk PNETs has
been on increasing the percentage of children
I.F. Pollack
218
with long-term survival (Taylor et al. 2005; Pizer

et al. 2006). Although a series of studies examined
the potential benefit of administering intensive
chemotherapy prior to irradiation as a way to
enhance disease control, these studies were in
some cases associated with an unacceptable rate
of toxicity and early disease progression.
Accordingly, a series of more recent approaches
have built upon the known activity of irradiation
and chemotherapy for these tumors by administering conventional chemotherapeutic agents with
radiosensitizing properties during irradiation,
followed by administration of additional postirradiation therapy. In this regard, the CCG-99701
study which involved a dose escalation study of
carboplatin with vincristine during radiotherapy
followed by adjuvant chemotherapy after irradiation, noted long-term survival rates that appeared
substantially better than those from previous
studies (Jakacki et al. 2007).
These results provided a foundation for the
phase 3 ACNS0332 study, which includes a factorial randomized design, one aspect of which
examines whether administration of carboplatin
and vincristine with irradiation achieves superior
outcome results compared to administration of
vincristine alone, and a second aspect of which
examines whether adding isotretinoin to an adjuvant chemotherapy backbone enhances outcome
compared to administration of adjuvant chemotherapy alone. This component of the study is
based upon promising observations that isotretinoin can synergistically enhance the activity of
platinum-based chemotherapy in preclinical
models. The ACNS0332 study also includes
molecular correlative analyses as noted above for
the average-risk patients in an effort to determine
whether high-risk patients can be further stratified based on molecular risk factors.
In addition to the potential relevance of molecular features in risk stratification of medulloblastomas treated with conventional treatment
strategies, studies of the molecular factors underlying tumorigenesis have also provided insights
into novel therapeutic approaches. In this regard,
alterations of the sonic hedgehog (SHH) pathway,
which are involved in developmental regulation
during embryogenesis, have been noted in a
sizeable subset of medulloblastomas. The role of

this pathway in medulloblastoma tumorigenesis
was first noted in patients with Gorlin syndrome,
a rare autosomal disorder associated with a
number of systemic manifestations as well as a
predisposition to development of medulloblastomas. This syndrome results from mutations in the
PTCH1 gene, which encodes the receptor for
binding of the SHH protein. Subsequently, mutations of PTCH1 and several other members of the
SHH signaling pathway have been noted in
patients with sporadic medulloblastomas (Raffel
et al. 1997). The feasibility of pharmacologically
blocking this pathway was subsequently demonstrated (Romer and Curran 2005), and favorable
results of pathway blockade has been observed in
patients with metastatic medulloblastoma (Rudin
et al. 2009). A trial of the SHH inhibitor GDC0449 is currently in progress in the PBTC, and it
is conceivable that therapeutic response will be
associated with the status of SHH pathway
activation in a given tumor.
In addition, studies of other proteins involved
in neural cell developmental regulation are currently in progress, including inhibitors of histone
deacetylase function and Notch signaling. Studies
of antiangiogenic signaling inhibition, using
agents such as bevacizumab, as well as growth
signaling inhibition, using agents blocking insulinlike growth factor receptor and epidermal growth
factor receptor activation, as well as downstream
signal transduction elements, have also been
undertaken.
Brain Tumors in Infants

The management of malignant brain tumors in
children younger than 3 years has historically
incorporated somewhat different treatment
approaches than comparable tumors in older children, owing to the significant sensitivity of the
infant brain to the neurotoxic effects of irradiation (Duffner et al. 1993; Geyer et al. 2005).
Treatment protocols in the 1980s and 1990s
examined the applicability of intensive postsurgical chemotherapy as a way to delay or avoid the
use of irradiation. These studies demonstrated
21
that approximately one-third of children would

favorably respond to such treatment and not
require radiotherapy, although most would manifest disease progression within 12 years of diagnosis, which generally was fatal. In more recent
studies, a variety of strategies have been examined
in an effort to improve on these results. One
approach, which was examined in the CCG-99703
study, involved the use of a second phase of
extremely intensive myeloablative consolidation
chemotherapy following an initial course of
induction therapy. A second approach involved
the use of focal irradiation to the tumor bed in
conjunction with chemotherapy for patients with
localized disease, which was examined in children with medulloblastoma in the P9934 study.
A third approach that has been pursued has
involved the use of high-dose systemic and intrathecal methotrexate, which has been examined in the
HIT-SKK92 study of the German Pediatric
Oncology Group (Rutkowski et al. 2005). With all
three approaches, outcome results have been superior to those of prior cooperative group studies that
have employed less intensive therapy.
In addition to intensification of therapy, a second factor that has contributed to an overall
improvement in the management of infants with
brain tumors has been the refinement of tumor
classification, which in part reflects advances in
molecular stratification. In the past, malignant
infant tumors were treated on fairly homogeneous
therapeutic protocols that considered them all as
embryonal tumors. In contrast, recent studies
have advocated distinctive treatment algorithms
for ependymomas and PNET-like tumors. In
addition, it has been recognized that the PNETlike group encompasses a number of distinctive
subsets that warrant distinct management
approaches. One group previously considered to
be a variant of PNET, but now recognized to be a
separate entity includes atypical teratoid/rhabdoid
tumors (AT/RTs), which characteristically have
mutation or inactivation of the INI1 gene (as
assessed by sequencing or immunohistochemistry
analysis or by fluorescence in situ hybridization
to detect loss of this chromosome 22 locus (Biegel
et al. 1999). The immunohistochemical test for
INI1 expression has now been incorporated in the
219
screening armamentarium to facilitate rapid

identification of these tumors (Judkins et al.
2005). The importance of distinguishing these
tumors from infant medulloblastomas and PNETs
is that AT/RTs have a much lower overall survival
rate than these other groups, often exhibiting
rapid progression during and after initial chemotherapy (Geyer et al. 2005). As a result, current
treatment protocols for these tumors are examining the use of alternative chemotherapy regimens
and, as in the COG ACNS0333 study, exploring
the early implementation of irradiation in an
effort to improve on the poor prognosis of these
lesions, based on the observation that most longterm survivors in previous studies have received
early irradiation in addition to intensive multiagent chemotherapy (Tekautz et al. 2005).
Contemporary studies for infants with
medulloblastomas and other PNETs are also
increasingly stratifying therapy based on prognostic
features. The P9934 study specifically focused on
non-metastatic medulloblastomas, based on the
fact that these tumors have a significant better
prognosis than lesions with detectable metastases
at diagnosis (Geyer et al. 2005). Conversely, the
COG ACNS0334 study focuses on infants with
metastatic medulloblastoma and supratentorial
PNETs (ACNS0334), which represents a highrisk subset of tumors. This study uses chemotherapy derived from the backbone of CCG-99703 as
a component of the induction regimen and examines in a randomized fashion the efficacy and toxicity of adding methotrexate as per the HIT-SKK
regimen to this aspect of therapy followed by
consolidation therapy as per the CCG-99703
regimen. The goal of this study is to determine
whether further intensification of induction
therapy increases the likelihood of complete
tumor regression and is tolerated without unacceptable toxicity.
A somewhat different design philosophy is
being applied for infants with localized medulloblastoma, based on observations from the HITSKK92, CCG-9921, and CCG-99703 studies that
patients whose tumors undergo radiographically
complete resections and have desmoplastic features
generally had a favorable response to therapy
(Geyer et al. 2005; Rutkowski et al. 2005),
220
whereas those with residual disease whose tumors

have classical histological features have had
much lower rates of long-term disease control.
These observations have provided a rationale to
stratify therapy based on these factors, with a
goal of reducing late sequelae of therapy in the
more favorable-risk, completely resected desmoplastic group and improving survival in patients
whose tumors have more adverse prognostic
features. Efforts to design a therapeutic trial
incorporating these parameters are currently in
progress.
Conclusion
The treatment of pediatric brain tumors has come
to incorporate a host of stratification criteria,
based on histological, clinical, and molecular
factors. These refinements in risk-based treatment planning combined with advancements in
imaging technology and surgical techniques have
led to improvements in outcome for children with
several types of brain tumors, such as medulloblastoma. Unfortunately, the prognosis for children with certain types of tumors, such as diffuse
intrinsic brainstem glioma, remains suboptimal.
The increasing implementation of molecular
analysis approaches for treatment-refractory
tumor subgroups may provide new insights into
molecularly targeted treatment options that offer
the hope of improving patient outcome while
reducing the side effects of therapy.
Acknowledgment This work was supported in part by
National Institutes of Health grants P01NS40923 and
R01NS37704.
References
Ater J, Mazewski C, Roberts W, Sposto R, Zhou T, Freyer
D, Jakacki R, Kadota R, Lazarus K, Packer R, Pearce
J, Prados M, Ettinger A, Vezina G, Wisoff J, Yates A,
Pollack I (2007) Phase 3 randomized study of two chemotherapy regimens for treatment of progressive lowgrade glioma in young children: preliminary report
from the Childrens Oncology Group protocol A9952.
Neuro Oncol 9:204
I.F. Pollack
Biegel JA, Zhou J-Y, Rorke LB, Stenstrom C, Wainwright
LM, Fogelgren B (1999) Germline and acquired mutations of INI1 in atypical teratoid and rhabdoid tumors.
Cancer Res 59:7479
Bradley KA, Pollack IF, Reid JM, Adamson PC, Ames
MM, Vezina G, Blaney S, Ivy P, Zhou T, Krailo M,
Reaman G, Mehta MP (2008) Motexafin gadolinium
and involved field radiation therapy for intrinsic
pontine glioma of childhood: a Childrens Oncology
Group Phase 1 study. Neuro Oncol 10:752758
Cohen KJ, Heideman R, Zhou T, Holmes E, Burger PC,
Brat DJ, Rosenblum M, Pollack IF (2007) Should
temozolomide be the standard of care for children with
newly diagnosed high-grade gliomas? Results of the
Childrens Oncology Group ACNS0126 study. Neurooncology 9:188
Duffner PK, Horowitz ME, Krischer JP, Friedman HS,
James HE, Freeman CR, Seidel FG, Kun LE (1993)
Postoperative chemotherapy and delayed radiation in
children less than three years of age with malignant
brain tumors. N Engl J Med 328:17251731
Finlay JL, Boyett JM, Yates AJ, Milstein JM, Geyer JR,
Bertolone SJ, McGuire P, Cherlow JM, Tefft M (1995)
Randomized phase III trial in childhood high-grade
astrocytoma comparing vincristine, lomustine, and
prednisone with the eight-drugs-in-1-day regimen.
Breiger D, Broxsonm E, Donahue B, Finlay JL,
Hartmann C, Meyer J, Balss J, Capper D, Mueller W,
Christians A, Felsberg J, Wolter M, Mawrin C, Wick
W, Weller M, Herold-Mende C, Unterberg A, Jeuken
JWM, Wesseling P, Reifenberger G, von Deimling A
(2009) Type and frequency of IDH1 and IDH2 mutations are related to astrocytic and oligodendroglial
differentiation and age: a study of 1,010 diffuse
gliomas. Acta Neuropathol 118:469474
Jakacki R, Burger P, Zhou T, Holmes E, Packer R,
Goldwein J, Mehta M, Pollack I (2007) Outcome for
metastatic (M+) medulloblastoma (MB) treated with
carboplatin during craniospinal radiotherapy (CSRT)
followed by cyclophosphamide (CPM) and vincristine
(VCR). J Clin Oncol 25:75S
Jakacki R, Yates A, Blaney S, Zhou T, Timmerman RD,
Ingle AM, Flom L, Prados M, Adamson PC, Pollack I
(2008) A phase I trial of temozolomide and CCNU in
newly diagnosed high-grade gliomas of childhood: a
report from the Childrens Oncology Group. Neurooncology 10:569576
Jennings MT, Sposto R, Boyett JM, Vezina LG, Holmes
E, Berger MS, Bruggers CS, Bruner JM, Chan K-W,
Dusenbery KE, Ettinger LJ, Fitz CR, Lafond D,
21
Mandelbaum DE, Massey V, McGuire W, McNeely L,

Moulton T, Pollack IF, Shen V (2002) Preradiation
chemotherapy in primary high-risk brain stem tumors:
phase II study CCG-9941 of the Childrens Cancer
Jones DT, Kocialkowski S, Liu L, Pearson DM, Backlund
LM, Ichimura K, Collins VP (2008) Tandem duplication producing a novel oncogenic BRAF fusion gene
defines the majority of pilocytic astrocytomas. Cancer
Res 68:86738677
INI1 protein expression distinguishes atypical teratoid/
rhabdoid tumor from choroid plexus carcinoma.
Kool M, Koster J, Bunt J, Hasselt NE, Lakeman A, van
Sluis P, Troost D, Schouten-van Meeteren N, Caron
HN, Cloos J, Mrsic A, Ylstra B, Grajkowska W,
Hartmann W, Pietsch T, Ellison D, Clifford SC,
Versteeg R (2008) Integrated genomics identifies five
medulloblastoma subtypes with distinct genetic profiles, pathway signatures and clinicopathological
features. PLoS One 3(8):e3088
Lam C, Bouffet E, Tabori U, Mabbott D, Taylor M, Bartels
U (2010) Rapamycin (sirolimus) in tuberous sclerosis
associated pediatric central nervous system tumors.
MacDonald T, Arenson EB, Ater J, Sposto R, Bevan HE,
Bruner J, Deutsch M, Kurczynski E, Luerssen T,
McGuire-Cullen P, O-Brien R, Shah N (2005) Phase II
study of high-dose chemotherapy before radiation in
children with newly diagnosed high-grade astrocytoma: final analysis of Childrens Cancer Group study
9933. Cancer 104:28622871
Packer RJ, Ater J, Allen J, Phillips P, Geyer R, Nicholson HS,
Jakacki R, Kurczynski E, Needle M, Finlay J, Reaman G,
Boyett JM (1997) Carboplatin and vincristine chemotherapy for children with newly-diagnosed progressive
low-grade gliomas. J Neurosurg 86:747754
Packer R, Gajjar A, Vezina G, Rorke-Adams L, Burger P,
Robertson P, Bayer L, LaFond D, Donahue B,
Marymont M, Muraszko K, Langston JW, Sposto R
(2006) Phase III study of craniospinal radiation therapy followed by adjuvant chemotherapy regimens for
newly diagnosed average-risk medulloblastoma. J Clin
Oncol 24:42024208
Parsons DW, Jones S, Zhang X, Lin JC, Leary RJ,
Angenendt P, Mankoo P, Carter H, Siu IM, Gallia GL,
Olivi A, McLendon R, Rasheed BA, Keirm S,
Nikolskaya T, Nikolsky Y, Busam DA, Tekleab H,
Diaz LA Jr, Hartigan J, Smith DR, Strausberg RL,
Marie SK, Shinjo SM, Yan H, Riggins GJ, Bigner DD,
Karchin R, Papadopoulos N, Parmigiani G, Vogelstein
B, Velculescu VE, Kinzler KW (2008) An integrated
genomic analysis of human glioblastoma multiforme.
Science 321:18071812
221

supratentorial primitive neuro-ectodermal tumours
entered into the SIOP/UKCCSG PNET 3 study. Eur J
Cancer 42:11201128
Pollack IF (1994) Brain tumors in children. N Engl J Med
337:15001507
malignant gliomas in children. N Engl J Med
346:420427
Pollack IF, Hamilton RL, James CD, Finkelstein SD,
Burnham J, Yates AJ, Holmes EJ, Zhou T, Finlay JL
(2006a) Rarity of PTEN deletions and EGFR amplification in malignant gliomas of childhood: results from
the Childrens Cancer Group 945 cohort. J Neurosurg
Pediatr 105:34313437
Pollack IF, Hamilton RL, Sobol RW, Burnham J, Yates
AJ, Holmes EJ, Zhou T, Finlay JL (2006b) MGMT
expression strongly correlates with outcome in childhood malignant gliomas: results from the CCG-945
cohort. J Clin Oncol 24:34313437
Pollack IF, Jakacki RI, Blaney SM, Hancock ML, Kieran
MW, Phillips P, Kun LE, Friedman H, Packer R,
Banerjee A, Geyer JR, Goldman S, Young-Poussaint
T, Krasin MJ, Wang Y, Hayes M, Murgo A, Weiner S,
Boyett JM (2007) Phase I trial of imatinib in children
with newly diagnosed brainstem and recurrent malignant gliomas: a Pediatric Brain Tumor Consortium
report. Neuro Oncol 9:145160
Pomeroy S, Tamayo P, Gaasenbeek M, Sturla LM, Angelo
M, McLaughlin ME, Kim JYH, Goumnerova LC,
Black PM, Lau C, Allen JC, Zagzag D, Olson JM,
Curran T, Wetmore C, Biegel JA, Poggio T, Mukherjee
S, Rifkin A, Califano A, Stolovitzky G, Louis DN,
Mesirov JP, Lander ES, Golub TR (2002) Prediction
of central nervous system embryonal tumour outcome
based on gene expression. Nature 415:436442
Raffel C, Jenkins RB, Frederick L, Hebrink D, Alderete B,
Fults DW, James CD (1997) Sporadic medulloblastomas contain PTCH mutations. Cancer Res 57:842845
Romer J, Curran T (2005) Targeting medulloblastoma:
small-molecule inhibitors of the Sonic hedgehog
pathway as potential cancer therapeutics. Cancer Res
65:49754978
Rudin CM, Hann CL, Laterra J, Yauch RL, Callahan CA,
Fu L, Holcomb T, Stinson J, Gould SE, Coleman B,
LoRusso PM, Von Hoff DD, deSauvage FJ, Low JA
(2009) Treatment of medulloblastoma with hedgehog
pathway inhibitor GDC-0449. N Engl J Med
361:11731178
Rutkowski S, Bode U, Deinlein F, Ottensmeier H,
Pietsch T, Wolff JE, Kortmann RD, Kuehl J (2005)
352:978986
Sanford A, Kun L, Sposto R, Holmes E, Wisoff JH, Heier
L, McGuire-Cullen P (2002) Low-grade gliomas of
222
childhood: impact of surgical resection. A report from
the Childrens Oncology Group. J Neurosurg
96:427428
high grade astrocytoma in children: results of a randomized trial. A report from the Childrens Cancer
Tait DM, Thornton-Jones H, Bloom HJG, Lemerle J,
Morris-Jones P (1990) Adjuvant chemotherapy for
medulloblastoma: the first multi-centre control trial of
the International Society of Paediatric Oncology
(SIOP I). Eur J Cancer 26:464469
Taylor RE, Bailey CC, Robinson KJ, Weston CL, Walker
DA, Ellison D, Ironside J, Pizer BL, Lashford LS
(2005) Outcome for patients with metastatic (M2-3)
medulloblastoma treated with SIOP/UKCCSG
PNET-3 chemotherapy. Eur J Cancer 41:727734
Tekautz TM, Fuller CE, Blaney S, Fouladi M, Broniscer
therapy and high-dose alkylator-based chemotherapy.
Thomas RM, Deutsch M, Kepner JL, Boyett JM, Krischer
P, Aronin P, Albright AL, Allen JC, Packer RJ,
Linggood R, Mulhern R, Stehbens JA, Langston J,
Stanley P, Duffner P, Rorke L, Cherlow J, Friedman
I.F. Pollack
HS, Finlay JL, Vietti TJ, Kun LE (2000) Low-stage
medulloblastoma: final analysis of trial comparing
standard-dose with reduced dose neuraxis irradiation.
Thompson MC, Fuller C, Hogg TL, Dalton J, Finkelstein
D, Lau CC, Chintagumpala M, Adesina A, Ashley
DM, Kellie SJ, Taylor MD, Curran T, Gajjar A,
Gilbertson RJ (2006) Genomics identifies medulloblastoma subgroups that are enriched for specific
genetic alterations. J Clin Oncol 24:19241931
Verhaak RGW, Hoadley KA, Purdom E, Wang V, Qi Y,
Wilkerson MD, Miller CR, Ding L, Golub T, Mesirov
JP, Alexe G, Lawrence M, OKelly M, Tamayo P, Weir
BA, Gabriel S, Winckler W, Gupta S, Jakkula L, Feiler
HS, Hodgson JG, James CD, Sarkaria JN, Brennan C,
Kahn A, Spellman PT, Wilson RK, Speed TP, Gray
JW, Meyerson M, Getz G, Perou CM, Hayes DN
(2010) Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized
by abnormalities in PDGFR, IDH1, EGFR, and NF1.
Cancer Cell 17:98110
Zeltzer PM, Boyett JM, Finlay JL, Albright AL, Rorke
LB, Milstein JM, Allen JC, Stevens KR, Stanley P,
Li H, Wisoff JH, Geyer JR, McGuire-Cullen P,
Stehbens JA, Shurin SB, Packer RJ (1999) Metastasis
stage, adjuvant treatment, and residual tumor are
prognostic factors for medulloblastoma in children:
conclusions from the Childrens Cancer Group 921
randomized phase III study. J Clin Oncol
17:832845
Pediatric Supratentorial Primitive

Neuroectodermal Tumor: Treatment
with Chemotherapy and Radiation
22
Donna L. Johnston and Daniel L. Keene
Contents
Abstract
Introduction ............................................................
223
Role of Chemotherapy ...........................................
224
Role of Radiation Therapy ....................................
225
Conclusions .............................................................
226
References ...............................................................
226
Supratentorial primitive neuroectodermal

tumors (SPNET) are rare embryonal tumors of
the central nervous system that account for
only 2.5% of brain tumors in children. This
tumor has proven to be difficult to treat, with
published survival rates ranging from 17% to
57%. The current therapy for pediatric patients
with SPNET used by most clinicians is a combination of both chemotherapy and radiation
therapy. Studies have shown a significant
improvement in survival for patients with this
tumor type if radiation therapy is utilized. The
role of high dose chemotherapy followed by
stem cell rescue is emerging as potentially
effective therapy for these tumors. Overall,
SPNET has had a low survival rate, but more
recent studies have shown an improvement in
the survival of pediatric patients with SPNET.
Introduction
D.L. Johnston (*) D.L. Keene

Division of Neurology, Childrens Hospital of Eastern
Ontario, Ottawa, ON, Canada
e-mail: djohnston@cheo.on.ca

(SPNET) are embryonal tumors of the central
nervous system and account for only approximately 2.5% of childhood brain tumors (Gaffney
et al. 1985). The mean age at diagnosis for this
type of tumor is 24 years of age and there is no
sex predilection (Albright et al. 1995; Johnston
et al. 2008; Jakacki 1999). SPNET bear many
similarities to posterior fossa medulloblastomas
in terms of morphological features, however they

223
224
exhibit important differences with respect to

biological features, outcome and response to
therapy (Gaffney et al. 1985; Rorke et al. 1997;
Jakacki 1999; Li et al. 2005). In most cooperative
group trials, SPNETs are stratified as high-risk
medulloblastomas and treated with both chemotherapy and radiation therapy. With this approach,
the published survival rates for SPNET in children
ranges from 17% to 57% (Johnston et al. 2008;
Dirks et al. 1996; Cohen et al. 1995; Albright
et al. 1995; Yang et al. 1999; Paulino and Melian
1999; Jakacki 1999; Reddy et al. 2000; Geyer
et al. 2005; Pizer et al. 2006).
Role of Chemotherapy
No single agent has been shown to have a significant effect on the survival of SPNET, however
many studies exist which show that SPNET are at
least temporarily responsive to chemotherapy
(Cohen et al. 1995; Geyer et al. 1994, 2005;
Reddy et al. 2000). The first randomized treatment trial for pediatric SPNET used radiation
therapy and compared the use of vincristine,
CCNU, and prednisone against methylprednisolone, CCNU, procarbazine, hydroxyurea, cisplatin, cytarabine and cyclophosphamide (Cohen
et al. 1995). This study found no difference in
survival among the two treatment groups.
Subsequent trials have looked at other chemotherapy regimens for pediatric SPNET patients.
For the most part, these trials are either limited to
infants or to children over the age of 3 years
with SPNET.
One of the largest studies of infants with SPNET
showed that the eight in one chemotherapy regimen (vincristine, BCNU, procarbazine, methylprenisolone, hydroxyurea, cisplatin, cytarabine
and cyclophosphamide) in children less than
18 months of age had a 3 year progression free survival of 55% for those with SPNET not in the
pineal region, and 0% for those with SPNET in
the pineal region (Geyer et al. 1994). In this
study the majority of infants did not receive
radiation therapy.
Another study of infants less than 36 months
of age with malignant brain tumors included
D.L. Johnston and D.L. Keene
36 patients with SPNET and 10 with SPNET in

the pineal region (Geyer et al. 2005). This study
compared vincristine, cisplatin, cyclophosphamide and etoposide against vincristine, carboplatin, ifosfamide and etoposide. Radiation
therapy was only given to those patients with
residual tumor or metastatic disease at diagnosis.
No difference in survival between the two chemotherapy regimens was noted with a 5 year
event free survival in both groups of 17%.
Reddy et al. (2000) reported a study of 22
children over the age of 3 years with SPNET who
were treated with radiation therapy (3440 Gy to
craniospinal axis and 10.820 Gy boost to primary tumor site) and chemotherapy (vincristine,
CCNU and cisplatin). The 3 year progression
free survival and overall survival in this group of
patients were 47% and 59% respectively, with a
5 year progression free survival and overall survival of 37% and 53% respectively.
In our recent Canadian review of SPNET
patients, the chemotherapy regimens used were
moderate to high-intensity therapy and most
employed agents including alkaloids (vincristine
and etoposide), platinum-based compounds
(cisplatin or carboplatin), and alkylating agents
(mostly cyclophosphamide but also carmustine,
procarbazine and lomustine) (Johnston et al.
2008). In this review the mean age at diagnosis
was 4 years and the 4 year survival was 38%.
Most other studies are not specific for SPNET
patients, but include these patients with the high
risk medulloblastoma patients and treat them in a
similar fashion.
Because of the poor survival rates for pediatric
SPNET patients, recent trials have utilized high
dose chemotherapy with stem cell rescue. This
type of therapy has been shown to be effective for
pediatric SPNET patients. The largest study utilizing this mode of therapy is the Head Start study
which treated 43 children with SPNET (Fangusaro
et al. 2008). In this study patients received induction chemotherapy with vincristine, cisplatin,
cyclophosphamide and etoposide, followed by
consolidation with carboplatin, thiotepa and
etoposide with autologous hematopoietic cell
rescue. The patients only received radiation therapy if they were over the age of 6 at diagnosis or
22
Pediatric Supratentorial Primitive Neuroectodermal Tumor: Treatment with Chemotherapy
if they had disseminated disease. The reported

5 year overall survival rate was 49% with an
event free survival rate of 39%. Fifteen of the 20
surviving patients had not received radiation
therapy.
In a smaller study of 15 patients with
SPNET, Massimino et al. (2006) used high dose
methotrexate, high dose etoposide, high dose
cyclophosphamide, and high dose carboplatin
followed by craniospinal irradiation (dose 31.2
39 Gy) with hyperfractionated accelerated radiotherapy plus a focal boost (dose 59.760 Gy), and
then either maintenance therapy with vincristine/
lomustine or high dose thiotepa followed by stem
cell rescue. The reported 3 year progression free
survival, event free survival and overall survival
rates were 54%, 34% and 61% respectively. For
the patients who received the high dose thiotepa
followed by stem cell rescue, the 3 year progression free survival and overall survival rates were
70% and 87.5% respectively.
In a study of children under the age of 6 years
with malignant brain tumors who were treated
with intensive therapy followed by stem cell rescue,
14 of the 62 patients had a SPNET (Mason et al.
1998). The therapy consisted of five cycles of
chemotherapy of cisplatin, vincristine, etoposide
and cyclophosphamide followed by consolidation chemotherapy with carboplatin, thiotepa,
etoposide and stem cell rescue. The investigators
found 2 year event free survival and overall survival rates of 27% and 48% respectively. For the
patients with SPNET, these figures were 43% and
64% respectively.
Strother et al. (2001) reported a study of pediatric patients over the age of 3 years with newly
diagnosed medulloblastoma or SPNET using
topotecan followed by craniospinal radiation and
then high dose therapy with cyclophosphamide,
cisplatin and vincristine with stem cell rescue.
There were 53 patients enrolled on this study, and
the 2 year progression free survival was 93.6%
for average risk patients, and 73.7% for high
risk patients.
Finally, Chintagumpala et al. (2009), published a study of 16 pediatric patients with SPNET
utilizing high dose cyclophosphamide, cisplatin
and vincristine with stem cell support. They
225
reported a 5 year event free survival rate of 68%

and an overall survival rate of 73%.
Overall, chemotherapy has shown to provide
some response in the treatment of pediatric
patients with SPNET. The use of high dose chemotherapy with stem cell rescue warrants larger
trials to see if there is a sustained improved effectiveness for the treatment of pediatric SPNET.
Role of Radiation Therapy

The use of radiation therapy has a significant
impact on survival in most childhood brain
tumors and is usually effective at stopping tumor
growth and providing clinical and radiological
improvement in the tumor. However, the use of
radiation therapy carries with it the burden of
imposing many significant secondary effects. The
benefits, however, in most cases outweigh the
risk of these side effects. In the case of SPNET, it
has a definite role in therapy. In a recent Canadian
review of 48 pediatric patients with SPNET, only
the use of radiotherapy was found to be independently significantly associated with an improved
survival (Johnston et al. 2008). The median dose
of radiation given was 54 Gy, and 72% of patients
received radiation that included the craniospinal
axis. Other studies have shown that chemotherapy without the use of radiation therapy has a
very high relapse rate (Jakacki et al. 1995; MarecBerard et al. 2002). In another review from the
Hospital for Sick Children in Toronto, which
examined 36 children with SPNET, it was found
that only children who received radiation therapy
were long term survivors, but in comparing those
who received brain radiation only to those who
received craniospinal radiation, no significant
difference was found (Dirks et al. 1996). In one
other small study, it was found that of the 15 children
with SPNET studied, 5 of the 5 who received up
front radiation therapy survived their malignancy,
while only 5 of the 10 who did not receive up
front radiation therapy were alive at follow up
(McBride et al. 2008). Other recent studies have
also confirmed the benefit of radiation therapy in
the treatment of this tumor (Taylor et al. 2009;
Timmerman et al. 2002, 2006).
226
The SIOP/UKCCSG PNET 3 study examined

the use of radiation therapy alone, versus radiation therapy with chemotherapy for the treatment
of pediatric SPNET (Taylor et al. 2009; Pizer
et al. 2006). Patients received a mean craniospinal radiation therapy dose of 34.7 Gy and a mean
total primary dose of 53.4 Gy. The event free survival at 3 and 5 years was 50% and 47% respectively and there was no significant impact on
overall survival or event free survival for patients
treated with either radiation therapy alone or
chemotherapy prior to radiation therapy. They
also found that the duration of radiation therapy
did not impact survival.
The German brain tumor trials HIT 88/89
and 91 examined the role of radiotherapy in the
treatment of SPNET patients (Timmerman et al.
2002). Patients who received local plus craniospinal irradiation demonstrated a significantly
improved survival compared to those who
received local irradiation only (43.7% vs. 14.3%
PFS, p = 0.0012). Also, patients who received a
local irradiation dose of greater than or equal to
54 Gy had a significantly higher survival than
those who received a local irradiation dose of less
than 54 Gy (44.7% vs. 10% PFS, p = 0.0045).
Finally, patients who received a dose to the
cransiospinal axis of less than 35 Gy had a significantly lower progression free survival compared to those who received 35 Gy or more of
craniospinal irradiation (0% vs. 49.3% PFS,
p = 0.0051). Thus, their study of 63 children with
SPNET showed that there was a definite role of
radiation therapy, that there needs to be therapy
to the entire craniospinal axis to a dose of 35 Gy
or higher, and there needs to be a boost to the
primary tumor site at a dose of 54 Gy or higher.
Another small American study found that a
dose of less than 50 Gy to the primary tumor site
was associated with a significant decrease in
survival (66.7% vs. 11.1%, p = 0.002) (Paulino
and Melian 1999).
Finally, a German study examined the impact
of radiation therapy in young children with
SPNET (Timmerman et al. 2006). This study
examined children under the age of 3 years with
SPNET, and found that only one child of the 15
who did not receive radiation therapy survived,
compared to 4 of the 14 who received radiation

therapy surviving their malignancy (3 year OS
rate 28.6% vs. 6.7%, p = 0.02). They also found
that radiation therapy given as salvage therapy
for relapse was not at all effective. Overall, the
survival for patients who received chemotherapy
with preventive radiation therapy was 40%, compared to 6.7% for those who received chemotherapy without radiation therapy, and 0% for those
who received chemotherapy and salvage radiation therapy. It was concluded that there was a
definite role for radiation therapy in children
under the age of 3 years with SPNET.
Overall, studies have shown that radiation
therapy has a significant role in the therapy of
SPNET. The craniospinal radiation needed is at a
dose of 35 Gy or higher with the dose to the primary tumor site of 54 Gy or higher, in order to
have a significant effect on survival.
Conclusions
in the pediatric population are rare tumors but
can respond to therapy. Studies have shown a role
for chemotherapy in the treatment of these tumors,
and a definitive role for the use of radiation. Large
studies have shown a definite improvement in
survival when radiation therapy is used, and have
shown the need for both local and craniospinal
radiation therapy. New studies utilizing high dose
chemotherapy with stem cell transplant have
shown promise with relatively small patient
numbers, and needs further investigation. Overall,
further studies are needed to improve the survival
of these patients.
References
Albright AL, Wisoff JH, Zeltzer P, Boyett J, Rorke LB,
Stanley P, Geyer JR, Milstein JM (1995) Prognostic
factors in children with supratentorial (nonpineal)
primitive neuroectodermal tumors. Pediatr Neurosurg
22:17
Chintagumpala M, Hassall T, Palmer S, Ashley D, Wallace
D, Kasow K, Merchant TE, Krasin MJ, Dauser R,
Boop F, Krance R, Woo S, Cheuk R, Lau C, Gilbertson
22
Pediatric Supratentorial Primitive Neuroectodermal Tumor: Treatment with Chemotherapy
R, Gajjar A (2009) A pilot study of risk-adapted

radiotherapy and chemotherapy in patients with
supratentorial PNET. J Neurooncol 11:3340
Cohen BH, Zeltzer PM, Boyett JM, Geyer JR, Allen JC,
Finlay JL, McGuire-Cullen P, Milstein JM, Rorke LB,
Stanley P (1995) Prognostic factors and treatment
results for supratentorial primitive neuroectodermal
tumors in children using radiation and chemotherapy:
a Childrens Cancer Group randomized trial. J Clin
Oncol 13:16871696
Dirks PB, Harris L, Hoffman HJ, Humphreys RP, Drake
JM, Rutka JT (1996) Supratentorial primitive neuroectodermal tumors in children. J Neurooncol
29:7584
Fangusaro J, Finlay J, Sposto R, Ji L, Saly M, Zacharoulis
S, Asgharzadeh S, Abromowitch M, Olshefski R,
Halpern S, Dubowy R, Comito M, Diez B, Kellie S,
Hukin J, Rosenblum M, Dunkel I, Miller DC, Allen J,
Gardner S (2008) Intensive chemotherapy followed by
consolidative myeloablative chemotherapy with autologous hematopoietic cell rescue (AuHCR) in young
children with newly diagnosed supratentorial primitive neuroectodermal tumors (sPNETs): report of the
Head Start I and II experience. Pediatr Blood Cancer
50:312318
Gaffney CC, Sloane JP, Bradley NJ, Bloom HJG (1985)
Primitive neuroectodermal tumors of the cerebrum.
J Neurooncol 3:2333
Geyer JR, Zeltzer PM, Boyett JM, Rorke LB, Stanley P,
Albright AL, Wisoff JH, Milstein JM, Allen JC, Finlay
JL (1994) Survival of infants with primitive neuroectodermal tumors or malignant ependymomas of the
CNS treated with eight drugs in 1 day: a report from
the Childrens Cancer Group. J Clin Oncol
12:16071615
Jakacki RI (1999) Pineal and nonpineal supratentorial
primitive neuroectodermal tumors. Childs Nerv Syst
15:586591
Jakacki RI, Zeltzer PM, Boyett JM, Albright AL, Allen
JC, Geyer JR, Rorke LB, Stanley P, Stevens KR,
Wisoff J (1995) Survival and prognostic factors following radiation and/or chemotherapy for primitive
neuroectodermal tumors of the pineal region in infants
and children: a report of the Childrens Cancer Group.
Johnston DL, Keene DL, Lafay-Cousin L, Steinbok P,
Sung L, Carret AS, Crooks B, Strother D, Wilson B,
Odame I, Eisenstat DD, Mpofu C, Zelcer S, Huang A,
Bouffet E (2008) Supratentorial primitive neuroectodermal tumors: a Canadian pediatric brain tumor
consortium report. J Neurooncol 86:101108
227
Li MH, Bouffet E, Hawkins CE, Squire JA, Huang A

(2005) Molecular genetics of supratentorial primitive
neuroectodermal
tumors
and
pineoblastoma.
Neurosurg Focus 19:E3
Marec-Berard B, Jouvet A, Thiesse P, Kalifa C, Doz F,
Frappaz D (2002) Supratentorial embryonal tumors
in children under 5 years of age: a SFOP study of
treatment with postoperative chemotherapy alone.
Med Pediatr Oncol 38:8390
Mason WP, Grovas A, Halpern S, Dunkel IJ, Garvin J,
Heller G, Rosenblum M, Gardner S, Lyden D, Sands
S, Puccetti D, Lindsley K, Merchant TE, OMalley B,
Bayer L, Petriccione MM, Allen J, Findlay JL (1998)
Intensive chemotherapy and bone marrow rescue for
young children with newly diagnosed malignant brain
tumors. J Clin Oncol 16:210221
Massimino M, Gandola L, Spreafico F, Luksch R, Collini
P, Giangaspero F, Simonetti F, Casanova M, Cefalo G,
Pignoli E, Ferrari A, Terenziani M, Podda M, Meazza
C, Polastri D, Poggi G, Ravagnani F, Rossati-Bellani F
(2006) Supratentorial primitive neuroectodermal
tumors (S-PNET) in children: a prospective experience with adjuvant intensive chemotherapy and hyperfractionated radiotherapy. Int J Radiat Oncol Biol Phys
64:10311037
McBride SM, Daganzo SM, Banerjee A, Gupta N,
Lamborn KR, Prados MD, Berger MS, Wara WM,
Haas-Kogan DA (2008) Radiation is an important
component of multimodality therapy for pediatric
non-pineal supratentorial primitive neuroectodermal
tumors. Int J Radiat Oncol Biol Phys 72:13191323
Paulino AC, Melian E (1999) Medulloblastoma and
supratentorial primitive neuroectodermal tumors, an
institutional experience. Cancer 86:142148
supratentorial primitive neuron-ectodermal tumors
entered into the SIOP/UKCCSG PNET 3 study. Eur
J Cancer 42:11201128
Reddy AT, Janss AJ, Phillips PC, Weiss HL, Packer RJ
(2000) Outcome for children with supratentorial primitive neuroectodermal tumors treated with surgery,
radiation and chemotherapy. Cancer 88:21892193
Rorke LB, Trojanowski JQ, Lee VM, Zimmerman RA,
Sutton LN, Biegel JA, Goldwein JW, Packer RJ (1997)
Primitive neuroectodermal tumors of the central nervous system. Brain Pathol 2:765784
Strother D, Ashley D, Kellie SJ, Patel A, Jones-Wallace
D, Thompson S, Heideman R, Benaim E, Krance R,
Bowman L, Gajjar A (2001) Feasibility of four
consecutive high-dose chemotherapy cycles with
stem-cell rescue for patients with newly diagnosed
medulloblastoma or supratentorial primitive neuroectodermal tumor after craniospinal radiotherapy: results
of a collaborative study. J Clin Oncol 19:26962704
Taylor RE, Donachie PHJ, Weston CL, Robinson KJ,
Lucraft H, Saran F, Ellison DW, Ironside J, Walker
DA, Pizer BL (2009) Impact of radiotherapy parameters
228
on outcome for patients with supratentorial primitive
neuro-ectodermal tumours entered into the SIOP/
UKCCSG PNET 3 study. Radiother Oncol 92:8388
Timmerman B, Kortmann RD, Kuhl J, Rutkowski S,
Meisner C, Pietsch T, Deinlein F, Urban C, WarmuthMetz M, Bamberg M (2002) Role of radiotherapy in
the treatment of supratentorial primitive neuroectodermal tumors in childhood: results of the prospective
German brain tumor trials HIT 88/89 and 91. J Clin
Oncol 20:842849

Timmerman B, Kortmann RD, Kuhl J, Rutkowski S,
Meisner C, Pietsch T, Deinlein F, Urban C, WarmuthMetz M, Bamberg M (2006) Role of radiotherapy in
supratentorial primitive neuroectodermal tumor in
young children: results of the German HIT- and HITSKK92 trials. J Clin Oncol 24:15541560
Yang HJ, Nam DH, Wang KC, Kim YM, Chi JG, Cho BK
(1999) Supratentorial primitive neuroectodermal
tumor in children: clinical features, treatment outcome
and prognostic factors. Childs Nerv Syst 15:377383
Pediatric Cancer Survivors:

Neurocognitive Late Effects
23
Sarah Hile, Erica Montague, Bonnie Carlson-Green,

Paul Colte, Leanne Embry, and Robert D. Annett
Contents
Introduction .............................................................. 230
Treatment Staging ...................................................... 239

Genetic Variables ....................................................... 240
Environmental Variables ............................................ 240
Cancer Diagnoses Most Vulnerable to

Neurocognitive Late Effects .................................... 232
Leukemia.................................................................... 232
Brain Tumors ............................................................. 232
Functional Outcomes Related to Late Effects .......

Health Status ..............................................................
Underachievement/Underemployment ......................
Psychosocial Functioning ..........................................
240
240
241
241
Treatment Related Morbidity of the

Child CNS .................................................................
Neurosurgery..............................................................
Cranial Radiation Therapy (CRT) ..............................
Proton Beam Radiotherapy ........................................
Chemotherapy ............................................................
Combination CRT with Chemotherapy .....................
Interventions .............................................................
Cognitive Remediation ..............................................
Pharmacological Interventions...................................
Academic Interventions .............................................
Ecological Interventions ............................................
242
242
243
243
244
233
233
234
236
237
238
Critical Variables Affecting Susceptibility

to Late Effects........................................................... 239
Child Variables ........................................................... 239
Conclusions and Future Directions ........................ 244

References ................................................................. 245
Abstract
S. Hile, Ph.D. E. Montague, M.S.

Psychology Department, University of New Mexico,
MSC03 2220, Albuquerque, NM 87131-0001, USA
B. Carlson-Green, Ph.D.
Pediatric Neuropsychologist, Childrens Hospitals
and Clinics of Minnesota, Psychology Services,
360 Sherman Street # 200, St. Paul, MN 55102, USA
P. Colte, Psy.D.
Division of Hematology/Oncology/BMT,
Primary Childrens Medical Center,
100 North Mario Capecchi Drive, Salt Lake City,
UT 84113-1100, USA
L. Embry, Ph.D.
Pediatric Hematology/Oncology, University of Texas
Health Science Center at San Antonio,
333 N. Santa Rosa Street 8th floor, San Antonio,
TX 78207, USA
Neurocognitive deficits are a common late

effect experienced by pediatric cancer survivors and can manifest across a variety of
domains including: attention and concentration,
executive functioning, processing speed, psychomotor skills, verbal memory, visuospatial
skills, and language (Moore J, Pediatr Psychol
30:5163, 2005). Deficits have also been
found to manifest across the broader domains
R.D. Annett, Ph.D. (*)

Department of Pediatrics and Psychology, University of
New Mexico Health Sciences Center,
Albuquerque, NM 87131, USA
e-mail: rdannett@unm.edu

229
230
of global intellectual functioning and academic performance. These deficits, however,

seem to be limited to the specific diagnoses of
leukemia and brain tumors. This is largely due
to the aggressive CNS-directed treatments, as
they induce inalterable structural damage to
the brain, which has been linked back to
observable deficits in neurocognitive functioning. Preliminary research also indicates that
neurocognitive deficits can have serious implications for survivors overall functional
capabilities as well as their general quality of
life. As such, a variety of interventions
have begun to be developed in order to
address these issues. However, research surrounding these two areas is still in its infancy.
Future research needs to address how neurocognitive late effects interfere with greater
functional capabilities and general quality of
life as well as how interventions can help combat these problems.
Introduction
Childhood cancer diagnoses are still relatively
rare, with the risk of developing cancer between
birth and age 20 being only 1 in 300 (Institute of
Medicine 2003). Unfortunately, however, cancer
still represents one of the leading causes of death
in children. Prior to 1970, individuals under
20 years of age receiving a cancer diagnosis had
little hope of surviving. The past 20 years, however, have been witness to a dramatic increase of
over 20% in survival rates with overall 5-year
survival rates estimated between 70% and 80%
(Institute of Medicine 2003). Childhood cancer
survival rates are, of course, dependent upon the
specific cancer diagnosis. Some cancer diagnoses
have more promising outcomes than others. For
example, the most commonly occurring childhood cancer, leukemia, demonstrates a 5-year
survival rate over 80% (Institute of Medicine
2003). In contrast, survival rates for malignant
cancers of the brain and central nervous system
(CNS), the second most common group of childhood cancer diagnoses, only approach 71.6%
(SEER 2007).
S. Hiles et al.
These survival gains can be attributed largely

to the remarkable advances in treatment, with
some combination of surgery, chemotherapy, and/
or radiation therapy as the major treatment modalities. While childrens cancer survival rates have
increased due to new chemotherapy agents and
therapy modifications, these treatment outcomes
also have been associated with increasing concern
about the toxicities associated with their usage.
With survivors living longer, greater concern is
being paid to the quality of lives they are living.
This has resulted in a conceptual shift in the focus
of outcomes with greater attention to the high
prevalence yet low severity dysfunctions that for
many children may not manifest until after their
cancer therapy has ended. Typically referred to as
late effects (Anderson et al. 2001), cancer treatment-related toxicities can be defined as any
chronic or subsequently occurring negative physical, neurocognitive, or psychosocial outcome that
becomes apparent when a childs cancer therapy
has concluded. Health-related late effects occur in
6575% of pediatric cancer survivors and include
physical malfunctions such as coronary artery disease, congestive heart failure, reproductive difficulties, endocrine issues, renal failure, hearing or
other sensory loss (Oeffinger et al. 2006), neurocognitive deficits (Moore 2005), or problematic
psychological outcomes such as social-emotional
problems (Kazak et al. 2009). As a result,
increased attention and interest is being directed
to the overall toxicity of treatments, the accompanying morbidities, and the impact on quality of
life for children and their families.
Neurocognitive dysfunction is commonly
assessed as a late effect experienced by childhood
cancer survivors. Domains of neurocognitive
functioning are identified in Table 23.1, with
examples of the instruments used to measure
these in children. It is projected that between 50%
and 60% of survivors are at risk for developing
some form of neurocognitive dysfunction (Nathan
et al. 2007). In pediatric cancer survivors, neurocognitive dysfunction has been identified in the
domains of attention and concentration, executive functioning, processing speed, psychomotor
skills, verbal memory, visuospatial skills, and
language (Moore 2005; Robinson et al. 2010).
23
Pediatric Cancer Survivors: Neurocognitive Late Effects
231
Table 23.1 Domains of neurocognitive function and common pediatric measures for these domains
Domain
General intellectual
functioning
Definitiona
Often referred to as intelligence quotient or IQ;
typically a measure of combined domains of
cognitive skills or abilities
Executive function
Processing speed
Set of skills called upon in novel, unfamiliar

contexts; individual skills may include decision
making, planning, and acting in a purposeful,
effective manner
The selection of information for further processing;
involves filtering information and sensory stimuli,
response initiation and inhibition, and sustained
performance or vigilance
Speed of mental and graphomotor processing
Verbal comprehension
Ability to understand language meaningfully
Visual perception
Also labeled visuospatial skills; the ability to use

visual information about object location and form to
solve problems and learn
Also labeled motor coordination; the ability to make
purposeful and intentional actions; frequently
assessment focuses on fine motor coordination
The ability to encode, store, and recall/retrieve
verbal and visual information
Sustained attention and

concentration
Visuomotor function
Memory
Academic achievement
Typically involves the assessment of core academic

skills, including reading, writing, and mathematics
with the purpose of diagnosing learning disabilities
Common measures
Wechsler Intelligence Scale for
Children, Fourth Edition; StanfordBinet Intelligence Scale; Bayley
Scales of Infant Development;
Wechsler Preschool and Primary
Scale of Intelligence, Third Edition
Childrens Category Test; Wisconsin
Card Sorting Test; Delis-Kaplan
Executive Function System; Tower
Test
Continuous Performance Test; Test
of Everyday Attention for Children
WISC-IV Coding and Symbol

Search subtests; WJ-III
Achievement fluency tests
WISC-IV Comprehension;
NEPSY-II Comprehension of
Instructions; the Token Test
Beery VMI: Visual Perception
subtest
Beery VMI: Motor Coordination
subtest; Finger Tapping; Grooved
Pegboard Test
California Verbal Learning Test for
Children; Wide Range Assessment
of Learning and Memory-Second
Edition
Woodcock-Johnson Tests of
Achievement; Wide Range
Achievement Test; WIAT-III
All definitions adapted from Strauss et al. 2006
The impact of neurocognitive dysfunction on a

childs global cognitive functioning (e.g., IQ;
Montour-Proulx et al. 2005) and the ability to
learn and perform well in school can be significant (Mitby et al. 2003). For example, these
authors reported that nearly 25% of survivors of
heterogeneous childhood cancers received special
education services in school as compared with
only 8% of a sibling control group, suggesting
that neurocognitive problems adversely impact
functional outcomes for survivors. It is important
to note that the deficits listed above manifest
across a variety of cancer diagnoses. However,
different diagnoses (e.g. brain tumors vs. leukemia)
generate different patterns of dysfunction. These

patterns will be explored and described in greater
depth in the following sections. Additionally, the
domains affected and the severity of the dysfunction can be contingent upon individual characteristics (e.g., childs age at treatment, sex, race)
and treatment factors (e.g., type(s) of therapies,
diagnosis, time since diagnosis, complications),
as well as environmental contributions. As a
result, there does not appear to be a uniform profile of neurocognitive dysfunction in childhood
cancer survivors.
The objective of this chapter is to review factors associated with the type and severity of
S. Hiles et al.
232
central nervous system dysfunction as well as the

hypothesized mechanisms contributing to such
dysfunction. In addition, this chapter will review
overall functional outcomes that have been associated with neurocognitive problems and the specific rehabilitative interventions that have been
developed to target them.
Cancer Diagnoses Most Vulnerable

to Neurocognitive Late Effects
Although neurocognitive dysfunction is among
the most common late effects experienced by
pediatric cancer survivors, not all cancer diagnoses pose equal risk. Since leukemia and brain
tumors are two of the pediatric cancers most susceptible to neurocognitive late effects (Moore
2005), the following discussion will focus on
these diagnoses.
Leukemia
Leukemia is the most prevalent form of childhood
cancer, accounting for over 25% of childhood
cancer diagnoses (SEER 2007). Although the
survival rate has increased dramatically over the
past 20 years, survivors of leukemia are at a
heightened risk for experiencing a variety of neurocognitive late effects. Given that leukemic cells
have the potential to infiltrate the central nervous
system (CNS), it became common practice in the
1970s to proactively treat the CNS by administering doses of cranial radiation (CRT) and intrathecal (IT) chemotherapy agents. CNS prophylaxis
is less common in other pediatric cancers that are
less likely to have CNS involvement. Currently,
IT chemotherapy is frequently administered
without detectable CNS involvement in order to
preempt any possible CNS infiltration. Recently
in practice, the use of CRT for children with
leukemia has declined as awareness of the risk
factors (i.e., female sex, age <3 years, dose of
radiation) for long-term neurocognitive deficits
has become more apparent. In fact, the accumulating evidence of the impact of CRT to the developing brain has resulted in diminished use of
CRT in all but the most severe pediatric leukemias (e.g., high risk pre-B cell ALL, T-cell ALL),
patients with overt CNS disease or those with
high-risk cytogenetics. Treatment-related neurocognitive late effects have also been linked with
specific chemotherapy agents, such as the use of
anti-folate agents as well as the administration of
chemotherapy intrathecally (Moore 2005).
A recent meta-analysis (Campbell et al. 2007)
revealed that survivors of acute lymphoblastic
leukemia (ALL) consistently demonstrate dysfunction across the broad domains of intellectual
and academic functioning as well as a number of
specific neuropsychological domains. The most
common and severe dysfunction identified by
these authors occurred in the domains of attention, speed of information processing and executive functioning. Buizer et al. (2009) suggest a
similar pattern of late effects, also implicating
potential loss of skill in the areas of memory, verbal comprehension, visuospatial skills, and visuomotor function. These neurocognitive late effects
have been associated with pathological changes
in brain structure such as white matter changes in
the form of leukoencephalopathy (MontourProulx et al. 2005). Other studies have found
survivors of childhood leukemia to experience
difficulties in functional outcomes, including
academic achievement. Survivors of ALL have
been found to demonstrate slower progress in
academic achievement such as specific difficulties in the areas of language skills and mathematics
(Buizer et al. 2006). Additionally, survivors of
childhood ALL were also more likely to repeat a
grade in school (Buizer et al. 2006).
Brain Tumors
Brain tumors represent the next most common
pediatric cancer, second only to leukemia, with
an incidence rate of 41.4 per 1,000,000 (SEER
2007). Given the heterogeneous nature of brain
tumors, which vary by tumor location and severity, it is not surprising that survival rates also vary
significantly. For example, children diagnosed
with low grade, resectable gliomas have a 5 year
survival rate ranging from 90% to 100%. This is
23
in stark contrast to medulloblastoma (with survival rates varying from 20% to 85% depending
largely on the age of the child and risk classification) or ependymoma (with estimated survival
rates of 5075%). The survival of pediatric brain
tumors is not without significant complications.
Current research indicates that survivors of
pediatric brain tumors are at increased risk for a
variety of neurocognitive late effects. Tumor site,
age of child, amount and location of radiotherapy,
time since treatment and treatment-associated
morbidities all contribute to these effects
(Ellenberg et al. 2009). Furthermore, pediatric
brain tumors not only produce proximal changes
in a childs brain, but also alter developmental
trajectories (Ris et al. 2005).
A recent meta-analysis (Robinson et al. 2010)
revealed that children treated for brain tumors
experience clinically significant dysfunction both
in broad neurocognitive domains (e.g., general
intellectual functioning) as well as in specific
neurocognitive functions: attention/concentration,
executive functioning, information processing
speed, psychomotor skills, verbal memory, visuospatial skills, visuospatial memory, and language. Results also revealed more severe deficits
in verbal memory, language, and attention compared to the other domains assessed. Other studies
have found deficits across the areas of attention/
executive function, processing speed, visual
perceptual skills, memory, and overall intellectual functioning (Anderson et al. 2001).
Self-reported neurocognitive difficulties
among pediatric brain tumor survivors indicate
that these survivors are aware of their limitations
and struggle significantly with learning difficulties (Carlson-Green 2009). Survivors have
reported problems in the cognitive domains of
task efficiency, emotional regulation, organization and memory (Ellenberg et al. 2009).
Functional impairments have been clearly
identified in academic achievement in the
domains of reading, spelling, and math (Reddick
et al. 2003) and chronic health conditions that
impact vocational success, psychosocial adaptation, and quality of life (Ellenberg et al. 2009).
In addition to neurocognitive late effects,
survivors of pediatric brain tumors also suffer
233
from physical morbidities and medical sequelae

related to treatment or tumor location. These
physical changes are associated with psychosocial
difficulties that can adversely impact self-esteem
and quality of life. Neuroendocrine late effects,
estimated to occur in as many as 80% of all pediatric brain tumors, are the result of damage to the
hypothalamus (Anderson et al. 2001). Endocrinerelated problems may include growth-hormone
deficiency, hypothyroidism, and sex hormone
deficiency resulting in infertility. In addition,
pediatric brain tumor survivors often have
pulmonary problems, bone-loss, and are at
greater risk for developing second malignancies
(Anderson et al. 2001).
Treatment Related Morbidity

of the Child CNS
Previous research has demonstrated that the risk
for neurocognitive late effects is most common
with a specific cancer diagnosis of leukemia or
brain tumors due to the treatment strategies used
to combat these diseases. The following section
will further explicate these CNS-directed treatments and the mechanisms of action by which
they contribute to neurocognitive late effects.
Neurosurgery
While commonly implemented cancer treatments
include chemotherapy and radiation therapy, children diagnosed with brain tumors often undergo
initial treatment with neurosurgical intervention.
The objectives of this method of intervention are
twofold; first is to procure a specific diagnosis of
the tumor; second is cytoreduction, or debulking,
a technique that has shown success in improving
survival outcome (Anderson et al. 2001).
Although neurosurgical treatment is considered
a necessary strategy to maximize survival, neurosurgical resection represents a particularly
aggressive assault on the brain. As such, neurosurgical intervention has been found to yield
significant immediate and often long-term consequences to childrens CNS. Carpentieri et al.
234
(2003) found that patients treated with surgery

only for localized brain tumors demonstrated an
increased prevalence of poor performance on
neurocognitive measures of motor output, verbal
memory, and visuospatial organization. Askins
and Moore (2008) suggest that such late effects
may occur as a result of neurosurgical resection
of tumors that encroach upon critical areas of the
brain are involved with language, memory, attention, executive functioning, and global neurocognitive skills (e.g., IQ). As such, it is not just the
removal process, but also the actual location of
the tumor that plays a critical role in the degree to
which neurocognitive late effects are expressed.
For example, previous research indicates that
tumors located in the supratentorial region (i.e.,
above the cerebellum) yield significant neurocognitive morbidities such as deficits in global neurocognitive function, decreased manual dexterity,
and poor emotional regulation, as well as seizures, and diminished quality of life (Anderson
et al. 2001). This stands in contrast to the more
minimal neurocognitive dysfunction that may
result from tumors located in the hypothalamic
and infratentorial regions, which may lead instead
to physical morbidities and medical sequelae.
Left hemispheric lesions are more likely to be
associated with verbal and language-based deficits, while right hemispheric lesions are more
often associated with visual-perceptual deficits.
(Anderson et al. 2001). In contrast, Carpentieri
et al. (2003) found no differences in neurocognitive performances when infratentorial tumor
location was compared with supratentorial tumor
location. Finally, tumor location has been implicated in differences in overall quality of life
(Sands et al. 2001).
Tumor location is a risk factor that will not be
ameliorated regardless of advances in research.
However, tumor resection strategies have
improved with the advancement of newer
techniques such as stereotactic procedures. The
use of operating microscopes and stereotactic
biopsies allow for more precision, thus minimizing the amount of direct insult to the brain
(Anderson et al. 2001). The end result of such
precision is the minimization of neurocognitive
late effects. Recently, more effort is being made
S. Hiles et al.
to study specific brain tumor types, but much of

the available research on neurocognitive outcomes following surgical resection is confounded
by heterogeneous samples and the concurrent use
of other therapies, such as CRT and IT chemotherapy. More research is necessary to determine
which specific neurocognitive late effects are
likely to occur based on tumor type or location.
In summary, neurosurgical procedures for benign
or malignant brain tumors in children may be
associated with a range of neurocognitive, physical,
and psychosocial late effects, which may impact
long-term functional outcomes.
Cranial Radiation Therapy (CRT)

Despite the advances in surgical technology, neurosurgery alone does not adequately treat the vast
majority of children with brain tumors and adjunct
treatments are typically implemented. One common treatment for malignant brain tumors is high
dose radiation therapy delivered to the CNS. The
general objective of CRT is to selectively destroy
malignant cancer cells. Two types of radiation are
commonplace for the treatment of brain tumors.
Radiation is either narrowly directed at the specific tumor site or is applied to the whole brain
with an emphasis on the tumor site. In addition to
treating brain tumors, low dose CRT is also a
treatment strategy for several pediatric leukemias
(e.g., T-cell ALL; relapsed pre-B ALL) as the
CNS can be a refuge for leukemic cells. For the
treatment of leukemia, children usually receive a
lower, more diffuse dose of radiation (1,800 cGy;
Anderson et al. 2001). Historically, children had
been treated with 2,400 cGy until studies showed
that the dosage could be lowered, thereby reducing neurocognitive morbidity (Iuvone et al. 2002).
This modified approach to treatment has resulted
in increased survival statistics and has reduced the
risk of disease recurrence in the CNS (Iuvone
et al. 2002).
Unfortunately, long term survivors treated
with CRT have demonstrated neurocognitive
problems that are linked to the treatment itself.
These consequences generally occur in acute,
subacute (26 months after radiation), and late
23
stages, with the more pronounced and chronic

neurocognitive dysfunction manifesting in the
late stage, thought by many to be between 2 and
5 years post completion of therapy (Moore 2005).
Some neurocognitive problems improve when
treatment is stopped; however, many are irreversible and some are even thought to be progressive
(Moore 2005). Such irreversible dysfunction
occurs in neuropsychological processes including
attention, memory, visual construction ability, and
computation/arithmetic skills (Moore 2005).
Research also indicates that children treated with
CRT are at risk for experiencing declines in IQ
scores (Anderson et al. 2001; Reddick et al.
2003). One explanation is that the intellectual
declines observed in children with brain tumors
treated with CRT may be the result of a failure to
learn and acquire new information (presumably
secondary to working memory, attention, and
memory dysfunction), as opposed to a loss of
previously-learned information. Results of a
study of 27 children treated with CRT for acute
lymphoblastic leukemia indicated that declines
observed in IQ may be secondary to deficits in
processing speed and working memory (Schatz
et al. 2000). While it is clear that children receiving CRT treatment are at risk for experiencing
neurocognitive late effects as well as slowed
intellectual development, research regarding the
severity of such CRT-induced deficits is mixed.
Some studies have found that patients treated
with irradiation for brain tumors demonstrated
overall intellectual functioning in the average
range (Sands et al. 2001).
Despite the potentially mild presentation of
CRT-induced neurocognitive dysfunction, there
is some evidence to suggest that these late effects
can increase in severity relative to an increase in
treatment intensity and younger age at treatment.
For example, higher doses of radiation have been
associated with lower IQ scores in pediatric brain
tumor survivors, particularly in young children.
Young children treated with a lower dosage of
CRT for medulloblastoma had a 15-point (1 standard
deviation) advantage in IQ score in comparison
to children irradiated at a higher dose (Mulhern
and Butler 2004). Despite this finding, other
studies have not replicated these dose-dependent
235
effects. For example, a later study found no differences in the presentation of neurocognitive
dysfunction when mild doses of CRT (18 cGy)
were compared to moderate doses (24 cGy)
(Mulhern and Butler 2004). As such, research
remains inconclusive regarding the relationship
between treatment intensity and severity of dysfunction, though most research does indicate a
dose-dependent relationship.
Promising new research highlights the role
that 3-dimensional computer modeling may have
on improving neurocognitive outcomes through
modifications related to dose of radiation and
fractionation parameters. Ris et al. (2005) presented a promising novel technique for modeling
radiotherapy effects, Integral Biologically
Effective Dose (IBED), which may help to minimize neurocognitive late effects. IBED is a radiobiological approach which estimates cell survival
parameters in a given brain region while taking
into account factors such as radiation dose, volume, and fractionation. In the pilot study, the
IBED technique was used to model the impact of
radiation therapy on the neurocognitive domain
of attention. The authors hypothesize that IBED
may be particularly useful in the early detection
of neurocognitive dysfunction, allowing for early
intervention (Ris et al. 2005).
In addition to treatment intensity, age at treatment has also received significant attention in the
literature and has been consistently identified as a
moderator for the expression of neurocognitive
late effects (Anderson et al. 2001; Moore 2005).
Evidence suggests that the administration of CRT
during infancy can have a significantly deleterious effect on later academic and intellectual performance (Nathan et al. 2007). Research
consistently indicates that children treated with
CRT at a young age have been shown to demonstrate a significant decrease in intellectual
functioning, whereas children treated with CRT
at an older age do not show such a dramatic intellectual decline. As such, it is recommended that
CRT not be implemented as a treatment strategy
for children younger than 3 years of age (Nathan
et al. 2007).
The use of CRT has also been associated with
significant structural damage to the developing
S. Hiles et al.
236
brain. Cortical atrophy, vascular damage, and

leukoencephalopathy have all been associated
with the use of CRT (Askins and Moore 2008;
Reddick et al. 2003; Moore 2005). Approximately
50% of children treated with CRT show changes
in white matter including demyelization and
necrosis, both of which are irreversible (Askins
and Moore 2008). There is some debate as to
whether the etiology of white matter injury is
primarily caused by disease-related factors (i.e.,
glial cell death causing demyelination) or indirect effects of treatment (i.e., destruction of
precursor cells or microvascular damage).
However, the empirical evidence to date has
shown that children treated with CRT demonstrate a significant decrease in normal-appearing white matter (NAWM) volume when
compared to untreated individuals of the same
age (Reddick et al. 2003).
More recently, in an attempt to identify the etiology of neurocognitive late effects, research has
begun to explore the link between these structural
abnormalities and overall neurocognitive performance. Although children treated with CRT demonstrate a range of structural changes, research
has specifically focused on volume changes in
cortical and subcortical white matter (Reddick
et al. 2003; Moore 2005). The relationship
between white matter damage and neurocognitive deficits can be traced to the functional role of
white matter within the CNS. White matter tissue
primarily consists of myelinated neuronal axons
and acts as a conduit of signals between cortical
structures, allowing for information processing to
occur and ensuring that lower order sensory systems are connected with higher order areas, such
as those involved in executive functions. Because
white matter works to increase axonal conductivity throughout the CNS, it has been suggested
that damage can significantly influence neuronal
transmission speed (Moore 2005). Research to
date has found evidence to support this assumption, as children treated with CRT demonstrate
slower reaction times as well as a decrease in performance on neuropsychological measures
(Moore 2005). Overall, these results suggest that
some neurocognitive late effects are likely the
result of slowed cortical activity secondary to

CRT induced white-matter alterations.
Additional studies have consistently revealed a
significant relationship between declines in NAWM
and declines in IQ scores, factual knowledge, verbal and nonverbal abstract thinking, and attention
(Reddick et al. 2003; Iuvone et al. 2002). The association between white matter changes and neurocognitive dysfunction has been documented in
other populations. For example, children experiencing traumatic brain injury (TBI) also experience
white matter damage and demonstrate neurocognitive deficits similar to pediatric cancer survivors
(e.g., attention, executive functioning, processing
speed, working memory, and long-term memory)
(Askins and Moore 2008). Taken together, these
findings support a model that explains neurocognitive deficits, in part, as a consequence of CRTinduced alterations in white matter.
Proton Beam Radiotherapy

One promising advance to radiotherapy treatment
for brain tumors is proton beam radiotherapy.
Proton beam radiotherapy is a type of external
beam radiotherapy that uses ionizing radiation. It
differs from standard radiation therapy in that
nearly all of the energy is deposited more precisely into the tumor site. The relatively large
mass of the protons tends to stay focused on the
tumor shape, with few protons penetrating
beyond. Surrounding healthy tissue is spared and,
presumably, protected from the deleterious effects
resulting from standard CRT. At this time, there
are only five proton therapy centers in North
America, largely due to the size and cost of the
cyclotron or synchrotron equipment necessary to
deliver the proton therapy. In the only published
pediatric studies of proton beam radiotherapy,
positive disease outcomes have been reported
(Merchant 2009). No literature is available to
inform us of the potential neurocognitive consequences of this treatment. It is proposed that this
treatment will significantly reduce neurocognitive late effects, while also improving functional
outcomes (Merchant 2009).
23
Chemotherapy
Due to the long-term toxicity of CRT and the
problematic outcomes associated with its use,
most treatment protocols reserve CRT only for
high-risk leukemia and children with specific
types of brain tumors. With the declining use of
high dose CRT, CNS prophylaxis has begun to
rely heavily on chemotherapy. CNS chemotherapy agents are typically delivered intrathecally,
which allows the agents to surpass the bloodbrain barrier. Empirical evidence seems to indicate
that CNS-directed chemotherapy is far from
benign. Approximately two-thirds of studies on
the effects of CNS-directed chemotherapy document some type of decline in scores on measures
of global neurocognitive functioning, and 25%
report deficits in at least one area of neuropsychological functioning in treatment of pediatric
leukemia (Moleski 2000). Although the majority
of studies show some type of significant neurocognitive decline, mixed results are reported in
the literature. Some studies have observed no
neurocognitive late effects when children receiving treatment for ALL are treated with CNS prophylaxis (Kadan-Lottick et al. 2009). Overall
results seem to indicate that while children do
demonstrate slight neurocognitive dysfunction as
a result of chemotherapy, such deficits are negligible and do not reach clinical significance
(Moleski 2000).
The manifestation of neurocognitive dysfunction may depend upon the specific chemotherapy
agents used, as different agents vary in terms of
toxicity levels. Common chemotherapy agents
include: methotrexate, cytosine, arabinoside, and
systemic steroids. Of these, intrathecal methotrexate (IT-MTX) has received a significant
amount of attention and its neurotoxicity has
been well documented (Moleski 2000). IT-MTX
neurotoxicity is presumed to be a function of its
role as a folate antagonist, and it acts to impair
rapid division of cancer cells. Yet there is concern
that rapid cell division also occurs among healthy
cells, such that healthy cells may be targeted by
IT-MTX particularly in younger children. This
may produce the morbidities that have been
237
observed when high dose IT-MTX is utilized in

pediatric populations. Research exploring the
neurotoxicity of IT-MTX has found that children
commonly experience deficits in attention,
memory, fine motor/tactual perceptual skills as
well as deficits in general academic and global
neurocognitive functioning (Moleski 2000).
Additionally, research has suggested a doseresponse relationship with IT-MTX (MontourProulx et al. 2005) and an association between
IT-MTX and lower intelligence (Iuvone et al.
2002). However, this finding does not consistently manifest across all research. As such, more
research needs to be conducted in order to determine
the relationship between IT-MTX dosage, method
of delivery, severity of neurocognitive dysfunction,
and how child age may mediate the expression of
neurocognitive dysfunction.
Although IT-MTX has been studied the most
extensively in terms of neurocognitive dysfunction, other chemotherapy agents have been examined as well. Waber et al. (2000) compared the
effects of two different corticosteroids that are
commonly used in treatment protocols for ALL.
Results indicated that the administration of dexamethasone (which passes through the bloodbrain barrier) might produce more neurocognitive
sequelae than prednisone (which does not pass
the barrier). Individuals treated with dexamethasone performed poorly on measures of reading
comprehension, arithmetic calculation, and working memory. Unfortunately, however, the results
of this study have not been replicated and research
exploring the effects of corticosteroids on
neurocognitive late effects has demonstrated no
differences between children treated with dexamethasone versus prednisone (Kadan-Lottick et al.
2009). In sum, chemotherapy treatment protocols
vary in terms of the chemotherapy agents used,
the dosages, and the administration method.
Unfortunately, most research exploring the
deleterious effects of these agents has focused
almost exclusively on the administration of
IT-MTX. As such, it is necessary that
neuropsychological research procedures continue
to accompany clinical trials and assess the
outcomes that are associated with other specific
238
chemotherapy agents and the interaction with

specific features of the trial (e.g., child age) to
ensure acute and long-term safety.
Similar to CRT, research has begun to identify
specific areas of structural damage that are directly
related to the use of CNS-directed chemotherapy.
For example, intracerebral calcifications have
been found in 24% of children treated with
IT-MTX for ALL (Iuvone et al. 2002). Additionally,
intensity of treatment has been related to the manifestation of calcifications and the presence of leukoencephalopathy in children (Iuvone et al. 2002).
Damage in normal appearing white matter is not
isolated to the use of CRT, but also extends to
CNS-directed chemotherapy. White matter abnormalities have been demonstrated via neuroimaging in two thirds of children who received
CNS-directed chemotherapy and the evidence
continues to build suggesting that CNS-directed
chemotherapy induces significant structural
abnormalities in the form of diffuse white matter
damage, leukoencephalopathy, intracerebral calcifications, and cortical atrophy (Moleski 2000).
Empirical research has begun to explore the
linkage between chemotherapy-induced structural damage in the brain and neurocognitive
dysfunction. Intracerebral calcifications have
been associated with neurocognitive dysfunction
and more specifically, calcification of the basal
ganglia has been found to be predictive of poor
intellectual and memory functioning (Moleski
2000). More research has focused recently on
the relationship between white matter damage
and observed neurocognitive dysfunction.
Reddick et al. (2006) found a significant connection between white matter volume and neurocognitive dysfunction. Survivors of ALL, who
demonstrated poorer performance on neurocognitive measures of attention, intelligence, and
academic achievement, also demonstrated a
noticeable decrease in white matter. Thus, these
results support the notion that white matter damage is responsible for subtle neurocognitive dysfunction. Although research exploring the
connection between structural abnormalities and
neurocognitive late effects is still in its infancy,
preliminary results suggest that difficulties in
neurocognitive performance can be explained by
S. Hiles et al.
treatment-induced structural changes (i.e., cortical atrophy), intracerebral calcifications, and

white matter damage.
Combination CRT with Chemotherapy

The combination of CRT and CNS-directed
chemotherapy seems to produce neurocognitive
late effects across the domains of attention/
concentration, processing speed, visual perceptual skills, executive functioning, and memory
(Nathan et al. 2007). While these late effects consistently manifest across survivors of childhood
brain tumors and leukemia, they are often relatively milder for children who have been treated
for leukemia (e.g., slight declines in IQ; Anderson
et al. 2001; Sands et al. 2001). Initial research
suggests that the severity of deficit may depend
upon treatment intensity in the forms of the
number of treatment types a child undergoes
(e.g., surgery, radiation, chemotherapy), as well
as the intensity of the dosage and age at which
treatment is received. Carlson-Green et al. (1995)
observed that among pediatric brain tumor survivors, the number of different treatment modalities a child receives is a better predictor of
outcomes than merely the presence or absence of
CRT. Additionally, the influence of treatment
intensity may also manifest in terms of a combined assault on the CNS by both CRT and CNSdirected chemotherapy, as well as the synergistic
effects of multiple treatments. Some have suggested that the order of treatments may even
contribute to outcomes. A meta-analysis of
research on the neurocognitive late effects in
children diagnosed with ALL demonstrated that
those who received a combination of both CRT
and chemotherapy scored significantly lower on
measures of global neurocognitive functioning
than those who received chemotherapy alone
(Campbell et al. 2007). The increased severity of
neurocognitive deficits that result from this combination therapy has been explained in terms of
the interaction effect of both treatments (Moleski
2000). Children appear to be at greatest risk for
developing leukoencephalopathy as CRT-induced
alterations in the blood-brain barrier may enhance
23
the neurotoxic effects of IT-MTX, allowing it to

permeate throughout the CNS and affect myelination. This may be particularly true for younger
children, as the blood-brain barrier is more
permeable during infancy. Use of combination
therapies appears to maximize structural damage
to the CNS and thereby yield more significant
neurocognitive deficits (Moleski 2000).
Critical Variables Affecting

Susceptibility to Late Effects
Child Variables
In addition to cancer diagnosis and treatment protocol variables, there are a variety of individual
characteristics that influence the manifestation of
neurocognitive late effects. For example, demographic variables such as sex and the childs age at
time of diagnosis and treatment have been found to
influence the severity of neurocognitive dysfunction (Askins and Moore 2008; Moleski 2000;
Nathan et al. 2007). Research has identified girls to
be at a higher risk for developing neurocognitive
impairment (Nathan et al. 2007), even though boys
are treated longer for ALL. This is true for both
CRT and chemotherapy. The degree of neurocognitive dysfunction also appears to be negatively
correlated with age such that younger children are
at heightened risk for experiencing more severe
neurocognitive deficits. (Moore 2005).
The enhanced vulnerability of younger children
appears to be due to the biological mechanisms
by which CNS-directed therapies produce neurocognitive deficits. Both CRT and chemotherapy
induce changes in normal appearing white matter
(Reddick et al. 2006), with younger children
being particularly vulnerable to damage because
myelinated axons are not yet fully formed (Askins
and Moore 2008). As such, developing myelin is
much more vulnerable to the neurotoxic assault of
CNS-directed therapies.
A childs history of prior neurological or
developmental difficulties may alter the risk for
neurocognitive late effects. Such difficulties may
include preexisting neurological conditions
(e.g., epilepsy, Down syndrome, history of
239
Attention-Deficit/Hyperactivity Disorder, learning

disability). Perioperative events (e.g., hydrocephalus), and postoperative events (e.g., cerebellar
mutism, development of seizures) may also affect
neurocognitive outcomes. It has been shown that
individuals with a stronger history of neurological deficits demonstrate lower performances on
neurocognitive measures that tap visual-spatial
skills, memory, attention, and overall IQ.
Treatment Staging
Timing or staging of treatment may interact with
child factors to impact neurocognitive late
effects. For example, in many treatment protocols for brain tumors, CRT and IT-chemotherapy
are administered conjointly. Some research suggests that staggering therapies, changing the
order of treatments, or even delaying treatment
(such as CRT in the very young) may be beneficial in reducing neurocognitive late effects, particularly for very young, female children. An
early study comparing timing of treatment found
that females treated prior to age 5 had better outcomes if their CRT therapy was preceded by a
course of IT-MTX than if administered conjointly. Although the sample was small, mean IQ
score for females in the pre-irradiation group
was 25 points higher than the control group,
which received CRT and MTX conjointly
(Askins et al. 2009). Similarly, 17 infants diagnosed with a brain tumor prior to 36 months of
age were treated with a trial of mechlorethamine,
Vincristine, procarbazine, and prednisone
(MOPP) in place of the usual CRT (Ater et al.
1997). CRT was reserved for infants who experienced relapse. When compared to a matched
group treated with CRT, the MOPP group demonstrated IQs in the average range, while the CRT
group had lower intellectual functioning scores
that continued to decline over time. Survival
rates were similar for both groups (Ater et al.
1997). This finding has been replicated and holds
true for neurocognitive functions beyond general
intelligence, including academic achievement,
memory, and attention. Overall, these findings
suggest there may be lessening of neurocognitive
S. Hiles et al.
240
late effects for infants and toddlers from a staggered course of treatment.
Genetic Variables
In addition to demographic characteristics such
as age and sex, the observed variation in type
and severity of neurocognitive deficits may also
be due to differences in genetic composition.
Preliminary research suggests that genetic polymorphisms influence an individuals response to
therapy by affecting chemotherapy metabolism
and clearance as well as the recovery of normal
tissue and function (Nathan et al. 2007). Krajinovic
et al. (2005) identified the NOS3 894TT genotype to be significantly related to IQ declines. In
this study, children with this genotype who were
treated with CRT experienced significant declines
in intellectual functioning while children without
this genotype did not. Research regarding genetic
polymorphisms as they relate to neurocognitive
late effects is still in its infancy and neuropsychological research procedures need to continue to
be embedded within pediatric oncology trials
before any conclusive statements can be made
regarding genetic risk factors.
Environmental Variables
While it is clear that a variety of child-related
factors affect the type and severity of neurocognitive late effects, the influence of environmental
factors should not be overlooked. For example,
declines in academic functioning have been
explained by the absences from school due to
medical treatment and follow-up appointments
(Askins and Moore 2008). Additionally, parents
abilities to meet the educational needs of their child
and identify and provide appropriate resources
may modify the degree to which individuals
experience neurocognitive dysfunction (Nathan
et al. 2007). Studies of children with brain tumors
(e.g., Carlson-Green et al. 1995) and children
with traumatic brain injuries have suggested that
environmental or contextual factors such as
family stress and socioeconomic status may affect
neurocognitive outcomes as much as treatment

or other medical factors. Such environmental factors also have important implications for intervention and remediation strategies to combat
these late effects.
Functional Outcomes Related

to Late Effects
The degree to which neurocognitive late effects are
expressed depends upon a range of child-related
and environmental variables. Previous research has
targeted these topics with vigor, as studies exploring these topics are very prolific. Unfortunately,
there is a dearth of research regarding the overall functional outcomes of neurocognitive late
effects, possibly related to the relative recency of
increased survival rates. This is a topic demanding further attention. Three major domains of
functioning have emerged that are likely to be
affected by neurocognitive late effects: health
status, underachievement/underemployment and
psychosocial functioning.
Health Status
Although changes in neurocognitive dysfunction
are among the most commonly observed late
effects, cancer survivors are at risk for experiencing a wide array of cancer-related morbidities and
health complications. As identified earlier, such
morbidities can include: endocrine abnormalities, secondary cancers, physical disabilities or
deformities, coronary artery disease, congestive
heart failure, renal failure or dialysis, reproductive issues, and hearing loss (Oeffinger et al.
2006). Overall, nearly three-quarters of pediatric
cancer survivors experience some type of chronic
health condition and 42% of survivors experienced a severe, disabling, or even life-threatening
condition (Oeffinger et al. 2006). As such,
cancer survivors demonstrate significantly poorer
health in the domains of social, physical, and
emotional functioning in comparison to community-based peers. However, as more individuals
are living beyond 5 years post-treatment, significant
23
attention has been directed towards improving

the overall quality of life. One targeted area of
improvement is the health status of survivors.
Identified variables that may be related the overall health status of pediatric cancer survivors,
include: female sex, lower education, lower SES,
and ongoing fatigue.
Although studies have examined a variety of
demographic variables in relation to overall
health status and health related quality of life,
little empirical research has examined the relationship between neurocognitive dysfunction and
overall health status. Recently Krull et al. (2010)
reported that cancer survivors were prone to
experiencing psychological problems such as
attention deficits, emotional problems, externalizing behavior, and social withdrawal. These
neurocognitive and psychological variables were
observed to correlate with health status outcomes,
including adult obesity and physical inactivity.
Continued investigation is needed in order to
extend our understanding of the extent to which
neurocognitive deficits contribute to health
outcomes.
Given the significant health-related late effects,
the Childrens Oncology Group (COG) created
guidelines for long-term follow-up of pediatric
cancer survivors. These guidelines (http://www.
childrensoncologygroup.org/disc/le/)
provide
information for patients and families, as well as
healthcare providers, laying out recommendations
and timelines for physical and neuropsychological/
psychological screenings. It is becoming more
common for pediatric cancer survivors to attend
long-term follow-up clinics, where a multidisciplinary team can provide early detection of
physical- and neuro-behavioral-health related
late effecs.
Underachievement/Underemployment
Studies of pediatric cancer survivors have documented difficulties in achievement in school and
in the work force. Pediatric cancer survivors
have demonstrated greater difficulties with scholastic achievement and were significantly less
likely to complete high school when compared
241
to healthy controls (Mitby et al. 2003). Survivors

also were more likely to require special education services when compared to sibling controls.
One study from the Childhood Cancer Survivor
Study (CSSS) reported that survivors were at
higher risk for later employment difficulties and
were more likely to be unemployed (Pang et al.
2008). The highest rates of unemployment were
found in brain tumor survivors treated with more
than 3,900 cGy of cranial radiation. This suggests that for some survivors, the neurocognitive
sequelae of treatment may affect their ability to
become or remain employed. It is clear that survivors of pediatric cancers are at risk for a variety of neurocognitive late effects that are
typically accompanied by problems in academic
functioning and occupational achievement. What
remains to be seen is the degree to which neurocognitive late effects account for the observed
difficulties in academics and employment. The
link between performance on neurocognitive
measures, achievement in school and integration
within the work force may be mediated by variables previously described. Even childhood cancer patients who have minimal deleterious
treatments may fall behind in school due to frequent absences, complications from their medical care, and/or treatment-related fatigue, and
exhibit functional impairment that is subsequently observed in unemployment or underemployment.
Psychosocial Functioning
In addition to significant observed difficulties in
academic and occupational functioning, pediatric cancer survivors are at risk for experiencing
an array of both social and emotional problems
(Kazak et al. 2009). In terms of social skills/
competence, survivors have been found to demonstrate less social competence, have fewer
friends, participate in fewer peer activities, and
experience more social isolation, as compared
with healthy peers (Kazak et al. 2009). This trend
toward social isolation seems to be more evident
in survivors who demonstrate neurocognitive
deficits. (Kazak et al. 2009). Mental/emotional
S. Hiles et al.
242
health findings indicate that some survivors are

at risk for experiencing a variety of psychosocial distress. In particular, adult survivors of
childhood cancer report more negative moods,
tension, depression, and global psychological
distress (Kazak et al. 2009). Not all studies,
however, find results that are indicative of significant emotional difficulties. Zebrack et al.
(2004) found that most cancer survivors were
psychologically well, and when distress
occurred for these individuals, it seemed to be
more an artifact of diminished social functioning and social isolation. This trend toward
social isolation has been replicated across
studies.
Although little research has explored how
neurocognitive dysfunction might account for
some of the expressed difficulties in psychosocial functioning, preliminary findings are suggestive of such a relationship. Research has
repeatedly revealed that children who are
actively on treatment tend to have no more psychopathology than the general population. It
stands to reason that there may be additional
variables such as neurocognitive variables that
mediate the relationship between the cancer
experience and later psychosocial functioning.
A qualitative study examining the attitudes of
brain tumor survivors found that experiences
with neurocognitive dysfunction tended to make
survivors feel dumb. Survivors described that
they had to work twice as hard to stay on task at
school at the expense of extra-curricular activities (Carlson-Green 2009). Additionally, it has
been suggested that neurocognitive dysfunction
such as slow processing speed may affect the
ability to process social cues and nuances in
communication and thus limiting their ability to
interact effectively with peers.
Interventions
Although many of the observed neurocognitive
late effects of childhood cancer survivors may
not be severe, they are significant enough that
they disrupt school performance. Research
has focused on four interventions to address

neurocognitive late effects for pediatric cancer
survivors: cognitive remediation, pharmacological
interventions, academic interventions, and ecological interventions.
Cognitive Remediation
Cognitive remediation attempts to improve cognitive functioning following brain injury or damage. The intervention involves retraining and
overlearning tasks similar in nature to that of the
deficit. Butler et al. (2008) applied the notion of
cognitive remediation to pediatric cancer survivors and have separated the intervention into
three components: brain injury rehabilitation,
special education/educational psychology, and
clinical psychology. Pediatric cancer survivors
with documented attentional difficulties were
enrolled and took part in the intervention across a
6-month period. Children received Attention
Process Training, metacognitive strategies, and
cognitive-behavioral strategies to help withstand
distraction. Unfortunately, the study yielded
disappointingly low compliance and no significant improvements in neurocognitive functioning
(Butler et al. 2008). Based on data from the traumatic brain injury literature, cognitive remediation
could be a potentially beneficial intervention for
cancer survivors, however, additional research
needs to be done to provide further validation for
this intervention because the primary limitation
is that gains made in a laboratory setting do not
always generalize to other settings (i.e., school,
home, community, etc.).
An alternative to the focus for cognitive remediation has been to focus the intervention on early
identification and prevention. Computer-based
interventions may provide potential to deliver
cognitive remediation earlier in the childs treatment or at critical periods during cancer therapy
when the child may be most amenable to intervention. Empirical literature on pediatric traumatic brain injury suggests that there may be
optimal windows in which cognitive remediation
intervention is most successful. Several computerbased cognitive remediation interventions currently
23
exist focusing on specific neurocognitive

functions, such as attention and working
memory. These interventions are delivered in the
home, sometimes with input from a trained coach,
thereby reducing issues of children having to
travel to academic centers to participate. These
interventions have already shown promise for
helping to improve the working memory and
attention skills of children with ADHD, thus
investigators are beginning to apply the approach
within pediatric oncology. Recently, Hardy et al.
(2011) reported the results from a small feasibility
study that examined the effects of one computerbased cognitive remediation intervention for
child survivors of cancer. These preliminary findings suggest that childhood cancer survivors can
improve in working memory skills and demonstrate corresponding declines in parent-reported
attention problems.
Pharmacological Interventions
Many of the neurocognitive late effects demonstrated by cancer survivors are similar to deficits
presented by individuals diagnosed with AD/HD:
Inattentive type. In particular, both populations
experience sustained difficulties with attention,
concentration and executive functioning, although
childhood cancer survivors do not typically manifest symptoms of impulsivity or hyperactivity
related to their treatments. Due to this similarity
in symptomatology, research has begun to extend
commonly used pharmacological interventions
for AD/HD population to pediatric cancer survivors. For example, the efficacy of methylphenidate hydrochloride (MPH), a commonly used drug
treatment for children with AD/HD, is currently
being explored in pediatric cancer survivors. The
first study to explore the efficacy of MPH in cancer survivors experiencing academic achievement
deficits and problems with attention employed a
randomized, double-blind trial (Conklin et al.
2010). A significant increase in sustained attention
was witnessed in the participants who received
MPH treatment. A later study found significant
decreases in attention problems as reported by
parents and teachers (Conklin et al. 2010).
243
Unfortunately, however, the results of other

studies have not been so definitive. In sum, these
studies suggest that MPH is an effective treatment for pediatric cancer survivors. However,
we would caution that frequent monitoring of
medication effects should be undertaken when
considering MPH as a treatment for neurocognitive deficits in pediatric cancer survivors. Whether
MPH can be used effectively with children
during their cancer therapy remains open to
speculation.
Academic Interventions
In addition to the potential neurocognitive late
effects resulting directly from treatment, children
diagnosed with cancer frequently miss many days
of school instruction, resulting in delayed academic development. Particularly for children
diagnosed with ALL, treatment may last between
2 and 3 years. Families frequently resort to homebased instruction during this period. This can
take several forms, the most common of which
are home school programs, typically associated
with family-financed and private education programs, or home-bound school programs that are
associated with public education and are likely to
include a public school teacher coming to the
childs home during the week for limited periods
of time. Increasingly, families are now utilizing
online education programs run by their local
school districts.
When survivors prepare to return to school,
they may be faced with barriers, including
reduced physical abilities and fatigue, altered
cognitive skills, and delayed social skills.
Clinically, a subset of survivors may not ever
choose to reintegrate given the significant obstacles associated with their medical care. Although
few studies are available for review, the available
research suggests that school reintegration programs may help survivors by improving their
school and social adjustment (Prevatt et al. 2000).
Children and adolescents may benefit from additional informal interventions, including special
accommodations such as preferential seating at
the front of the class to limit distractibility, extra
244
time on tests, and decreased expectations for

volume of homework. As more is learned about
the functional impact of specific neurocognitive
late effect profiles, additional interventions could
be tailored to each child at different phases of
their cancer care.
Ecological Interventions
Thus far, the discussion regarding interventions
has focused on child-directed interventions.
However, it should not be forgotten that cancer
survivors function within a community that
includes caretakers, teachers, and peers. The
impact of the cancer experience on the family and
childs extended community can be significant. As
such, interventions specifically targeting parents
or caregivers, providing them with problem-solving skills training following initial diagnosis have
been developed and tested (Askins et al. 2009).
Problem-solving skills training has been operationalized with sequential individual meetings
between the parent/caregiver and interventionist
that attempt to build coping skills for the challenges that parents encounter when a child is diagnosed with cancer. Specific problem areas are
identified by the parent/caregiver and the intervention then seeks to build coping skills for that problem. Although these skills are important for coping
with the initial treatment, they may also be beneficial when the child transitions to survivorship, and
the family adjusts to long term effects of the disease and therapy. Additionally, the importance of
extending some type of intervention to the community at large should not be overlooked. Strong
communication among parents, school officials,
and health professionals is critical for improving
survivors outcomes and that more preparation of
parents to learn advocacy skills would be beneficial. As such, school reintegration programs such
as the Welcome Back educational program
offered through the Leukemia and Lymphoma
Society (www.lls.org) provide useful information
and strategies for advocacy for parents and teachers alike. Such interventions are suggested to
include an education component in which teachers and peers are informed of the neurocognitive
S. Hiles et al.
and psychosocial late effects that accompany

cancer survivorship.
Conclusions and Future Directions

In summary, children treated for leukemia or brain
tumors are at risk for developing neurocognitive
late effects in the domains of attention/concentration, processing speed, visual perceptual skills, executive functioning, and memory. These late effects
are likely the result of toxic CNS-directed treatments in the form of surgery, CRT, and chemotherapy. Additionally it has been well documented that
a variety of child-related and environmental factors
increase risk for developing these late effects.
Although the past 10 years have witnessed
significant improvement in our understanding of
treatment-related neurocognitive late effects,
many questions still remain and are insufficiently
addressed by current research. Not all pediatric
cancer survivors experience neurocognitive late
effects, and which children will experience clinically significant late effects remains somewhat
uncertain. While research has indentified some
patient-related and external factors that heighten
risk, additional research needs to be completed in
this area. A greater refinement of these factors
would help create a model for predicting neurocognitive late effects. Additionally, research
exploring the link between neurocognitive deficits and later functional outcomes is only emerging. Moving forward, future research should be
focused on delineating the relationship between
neurocognitive deficits and functional outcomes
across health, academic/employment, and psychosocial domains. The research exploring interventions for pediatric cancer survivors also
remains rather limited, though initial results are
promising. Future randomized trials examining
pediatric cancer medical therapies need to embed
measures of neurocognitive function in order to
further monitor the acute and long-term safety of
the trial. Furthermore, the addition of routine
neurocognitive measures within a trial can serve
to identify children for additional intervention to
mitigate adverse neurocognitive outcomes and
improve overall functional outcomes.
23
References
Anderson DM, Rennie KM, Ziegler RS, Neglla JP,
Robison LR, Gurney JG (2001) Medical and neurocognitive late effects among survivors of childhood
central nervous system tumors. Cancer 92:27092719
Askins M, Moore B (2008) Preventing neurocognitive
late effects in childhood cancer survivors. J Child
Neurol 23:11601171
Askins MA, Sahler OJ, Sherman SA, Fairclough DL,
Butler RW, Katz ER, Dolgin MJ, Varni JW, Noll RB,
Phipps S (2009) Report from a multi-institutional
randomized clinical trial examining computerassisted problem-solving skills training for Englishand Spanish-speaking mothers of children with newly
diagnosed cancer. J Pediatr Psychol 34:551563
Ater JL, van Eys J, Woo SY, Copeland DR, Bruner J
(1997) MOPP chemotherapy without irradiation as a
primary postsurgical therapy for brain tumors in
infants and young children. J Neurooncol 32:243252
Buizer AI, de Sonneville LM, van den Heuvel-Eibrink
MM, Veerman AJ (2006) Behavioral and educational
limitations after chemotherapy for childhood acute
lymphoblastic leukemia or Wilms tumor. Cancer
106:20672074
Buizer AI, de Sonneville LM, Veerman AJ (2009) Effects
of chemotherapy on neurocognitive function in children with acute lymphoblastic leukemia: a critical
review of the literature. Pediatr Blood Cancer
52:447454
Butler R, Copeland D, Fairclough D, Mulhern R, Katz E,
Kazak A, Noll RB, Patel SK, Sahler OJ (2008) A multicenter, randomized clinical trial of a cognitive remediation program for childhood survivors of a pediatric
malignancy. J Consult Clin Psychol 76:367378
Campbell LK, Scaduto M, Sharp W, Dufton L, Van Slyke
D, Whitlock JA, Compas B (2007) A meta-analysis of
the neurocognitive sequelae of treatment for childhood
acute lymphocytic leukemia. Pediatr Blood Cancer
49:6573
Carlson-Green B (2009) Brain tumor survivors speak out.
J Pediatr Oncol Nurs 26:266279
Carlson-Green B, Morris R, Krawiecki N (1995) Family
and illness predictors of outcome in pediatric brain
tumors. J Pediatr Psychol 20:769784
Carpentieri SC, Waber DP, Pomeroy SL, Scot RM,
Goumnerova LC, Kieran MW, Billett AL, Tarbell NJ
(2003) Neuropsychological functioning after surgery
in children treated for brain tumor. Neurosurgery
52:13481356
Conklin HM, Helton S, Ashford J, Mulhern RK, Reddick
WE, Bonner M, Japer BW, Wu S, Xiong X, Khan RB
(2010) Predicting methylphenidate response in longterm survivors of childhood cancer: a randomized,
double-blind, placebo-controlled. J Pediatr Psychol
35(2):144155
Ellenberg L, Liu Q, Gioia G, Yasui Y, Packer RJ, Mertens
A, Donaldson SS, Stovall M, Kadan-Lottick N,
245
Armstrong G, Robison LL, Zeltzer LK (2009)

Neurocognitive status in long-term survivors of childhood CNS malignancies: a report from the childhood
cancer survivor study. Neuropsychology 23:705717
Hardy KK, Willard VW, Bonner MJ (2011) Computerized
cognitive training in survivors of childhood cancer: a
pilot study. J Pediatr Oncol Nurs 28(1):2733
Institute of Medicine (2003) Childhood cancer survivorship: improving care and quality of life. The National
Academies Press, Washington, DC
Iuvone L, Mariotti P, Colosimo C, Guzzetta G, Ruggiero
A, Riccardi R (2002) Long-term cognitive outcome,
brain computed tomography scan, and magnetic resonance imaging in children cured for acute lymphoblastic leukemia. Cancer 95:25622570
Kadan-Lottick NS, Brouwers P, Breiger D, Kaleita T,
Dziura J, Liu H, Chen L, Nicoletti M, Stork L, Bostrom
B, Neglia JP (2009) A comparison of neurocognitive
functioning in children previously randomized to dexamethasone or prednisone in the treatment of childhood
acute lymphoblastic leukemia. Blood 114:17461752
Kazak AE, Alderfer MA, Rodriguez AM (2009)
Psychosocial and behavioral issues in cancer survival in
pediatric populations. In: Miller SM, Bowen DJ, Croyle
RT, Rowland JH (eds) Handbook of cancer control and
behavioral science: a resource for researchers, practitioners, and policymakers. American Psychological
Association, Washington, DC, pp 449465
Krajinovic M, Robaey P, Chiasson S, Lemieux-Blanchard
E, Rouillard M, Primeau M, Bournissen FG, Moghrabi
A (2005) Polymorphisms of genes controlling homocysteine levels and IQ score following the treatment
for childhood ALL. Pharmacogenomics 6:293302
Krull KR, Huang S, Gurney JG, Klosky JL, Leisenring W,
Termuhlen A, Ness KK, Kumar Srivastava D, Mertens
A, Stovall M, Robison LL, Hudson MM (2010)
Adolescent behavior and adult health status in childhood cancer survivors. J Cancer Surviv 4:210217
Merchant TE (2009) Proton beam therapy in pediatric
oncology. Cancer J 15:298305
Mitby PA, Robinson LL, Whitton JA, Zevon MA, Gibs
IC, Tersak JM, Meadows AT, Stovall M, Zeltzer LK,
Mertens AC, and Childhood Cancer Survivor Study
Steering Committee (2003) Utilization of special education services and educational attainment among
long-term survivors of childhood cancer. Cancer
97:1151126
Moleski M (2000) Neuropsychological, neuroanatomical,
and neurophysiologic consequences of CNS chemotherapy for acute lymphoblastic leukemia. Arch Clin
Neuropsychol 15:603630
Montour-Proulx I, Kuehn SM, Keene DL, Barrowman NJ,
Hsu E, Matzinger M, Dunlap H, Halton JM (2005)
Cognitive changes in children treated for acute
lymphoblastic leukemia with chemotherapy only
according to the Pediatric Oncology Group 9605
protocol. J Child Neurol 20:129133
Moore BD 3rd (2005) Neurocognitive outcomes in
survivors of childhood cancer. J Pediatr Psychol
30:5163
246
Mulhern RK, Butler RW (2004) Neurocognitive sequelae
of childhood cancers and their treatment. Pediatr
Rehabil 7:114
Nathan PC, Patel SK, Dilley K, Goldsby R, Harvey J,
Jacobsen C, Kadan-Lottick N, McKinley K, Millham
AK, Moore I, Okcu F, Woodman CL, Brouwers P,
Armstrong FD (2007) Guidelines for identification
of, advocacy for, and intervention in neurocognitive
problems in survivors of childhood cancer: a report
from the Childrens Oncology Group. Arch Pediatr
Adolesc Med 161:798806
Oeffinger KC, Mertens AC, Sklar CA, Kawashima T,
Hudson MM, Meadows AT, Friedman DL, Marina N,
Hobbie W, Kadan-Lottick NS, Schwartz CL,
Leisenring W, Robison LL (2006) Chronic health conditions in adult survivors of childhood cancer. N Engl
J Med 355:15721582
Pang JWY, Friedman DL, Whitton JA, Stovall M, Mertens
AC, Robison LL, Weiss NS (2008) Employment status
among adult survivors in the childhood cancer survivor study. Pediatr Blood Cancer 50:104110
Prevatt FF, Heffer RW, Lowe PA (2000) A review of
school reintegration programs for children with
cancer. J Sch Psychol 38:447467
Reddick WE, White HA, Glass JO, Wheeler GC,
Thompson SJ, Gajjar A, Leigh L, Mulhern RK (2003)
Developmental model relating white matter volume to
neurocognitive deficits in pediatric brain tumor survivors.
Cancer 97:25122519
Reddick WE, Shan ZY, Glass JO, Helton S, Xlong X, Wu
S, Bonner MJ, Howard SC, Christensen R, Khan RB,
Pui C, Mulhern RK (2006) Smaller white-matter volumes are associated with larger deficits in attention
and learning among long-term survivors of acute lymphoblastic leukemia. Cancer 106:941949
Ris MD, Ryan PM, Lamba M, Brenemen J, Cecil K,
Succop P, Ball W (2005) An improved methodology
S. Hiles et al.
for modeling neurobehavioral late-effects of radiotherapy in pediatric brain tumors. Pediatr Blood
Cancer 44:487493
Robinson KE, Kuttesch JF, Champion JE, Andreotti CF,
Hipp DW, Bettis A, Barnwell A, Compas BE (2010) A
quantitative meta-analysis of neurocognitive sequelae
in survivors of pediatric brain tumors. Pediatr Blood
Cancer 55:525531
Sands SA, Kellie SJ, Davidow AL, Diez B, Villablance J,
Weiner HL, Pietanza MC, Balmaceda C, Finlay JL
(2001) Long-term quality of life and neuropsychologic
functioning for patients with CNS germ-cell tumors:
from the First International CNS Germ-Cell Tumor
Study. Neuro Oncol 3:174183
Schatz J, Kramer JH, Ablin A, Matthay KK (2000)
Processing speed, working memory, and IQ: a developmental model of cognitive deficits following cranial
radiation therapy. Neuropsychology 14:189200
SEER (2007) Surveillance, Epidemiology, and End
Results (SEER) program public-use data (1973-2004).
National Cancer Institute, DCCPS, surveillance
research program, cancer statistics branch, released
April 2007. www.seer.cancer.gov/
Strauss E, Sherman EMS, Spreen O (2006) A compendium of neuropsychological tests, 3rd edn. Oxford
University Press, New York
Waber DP, Carpentieri SC, Klar N, Silverman LB, Schwenn
M, Hurwitz CA, Mullenix PJ, Tarbell NJ, Sallan SE
(2000) Cognitive sequelae in children treated for acute
lymphoblastic leukemia with dexamethasone or prednisone. J Pediatr Hematol Oncol 22:206213
Zebrack BJ, Zeltzer LK, Whitton J, Mertens AC, Odom L,
Berkow R, Robison LL (2004) Psychological outcomes in long-term survivors of childhood leukemia,
hodgkins disease, and non-hodgkins lymphoma: a
report from the Childhood Cancer Survivor Study.
Pediatrics 110:4252
Adult Survivors of Pediatric

Cancer: Risk of Cardiovascular
Disease
24
Eric Chow and Lillian Meacham
Contents
Abstract
Introduction ............................................................
248
Components of Cardiovascular Risk....................
248
Endocrine Factors ..................................................
251
Surveillance.............................................................
252
Future Directions ...................................................
252
Conclusion ..............................................................
254
References ...............................................................
254
E. Chow (*)
Hematology-Oncology, Seattle Childrens Hospital, Fred
Hutchinson Cancer Research Center,
Box 19024, M4-C308,
Seattle, WA 98109, USA
e-mail: ericchow@u.washington.edu
L. Meacham
Hematology/Oncology and Endocrinology,
Emory University, 2015 Uppergate Drive Northeast,
Atlanta, GA 30322, USA
With contemporary therapy, approximately

75% of children diagnosed with central nervous system (CNS) tumors are expected to be
5-year survivors. With increased survival,
there is an increased emphasis on better
understanding cancer-related late adverse
effects, including premature development of
cardiovascular disease. Due to the nature of
CNS tumors and its treatment, survivors are at
increased risk of various neuroendocrine deficiencies that increase risk of obesity, dyslipidemia, and hypertension. Radiation exposure
to the heart and some chemotherapeutic agents
also are associated with primary cardiomyopathy. Together, these factors likely contribute
to the observed increase in cardiovascular disease risk seen in both small and large cohort
studies of childhood CNS tumor survivors. In
response, cancer survivor-specific health
screening guidelines have been published in
an attempt to detect early manifestations of
cardiovascular disease and hopefully prevent
more serious morbidity. In addition to describing the magnitude and risk factors associated
with these adverse cardiovascular outcomes,
there also is ongoing research focused on
understanding genetic determinants of treatmentrelated sequelae as well as promoting healthier
lifestyles.

247
248
Introduction
Similar to many other pediatric cancers, survival
following central nervous system (CNS) tumors
has improved over recent decades. Data from the
US national Surveillance, Epidemiology, and
End Results cancer registry report an approximate 75% 5-year survival for patients diagnosed
with CNS tumors at age <20 years between 1999
and 2006 (Altekruse et al. 2010). As a result,
there has been an increased emphasis on better
understanding cancer-related late adverse effects
among long-term survivors and an effort to minimize these effects without compromising cancer
treatment efficacy. Results from the large North
American Childhood Cancer Survivor Study
(CCSS) cohort suggest that CNS tumor survivors
carry a disproportionately high burden of chronic
health conditions compared with unaffected siblings as well as other cancer diagnostic groups
(Oeffinger et al. 2006). Compared with siblings,
CNS survivors were 12-times more likely to
report both multiple chronic health conditions as
well as conditions classified as severe, life-threatening, or disabling; risk ratios that ranked either
first or second in magnitude amongst the eight
major cancer diagnostic groups included in that
cohort (Oeffinger et al. 2006). The mean age of
all survivors and siblings in this cohort was only
27 and 29 years, respectively. Data from the
CCSS and cohort studies from the United
Kingdom also show that childhood cancer survivors experience increased cardiovascular mortality and morbidity (Mertens et al. 2008; Reulen
et al. 2010). In the CCSS cohort, CNS tumor survivors specifically had a 4-fold increased risk of
cardiovascular mortality compared with age- and
gender-standardized
population
estimates
(Armstrong et al. 2009b). CNS tumor patients are
at increased risk of subsequent cardiovascular
disease due to a variety of treatment exposures
(Table 24.1). Separate from any effect of the
tumor itself, radiotherapy and surgery to the brain
can disrupt the hypothalamic-pituitary axis, leading to an increased risk of various hormone deficiencies that can contribute to obesity and
dyslipidemia. Some CNS patients also receive
E. Chow and L. Meacham
spinal radiotherapy, which can result in scatter

radiation to the heart. Various chemotherapy
exposures also can increase risk of subsequent
cardiomyopathy and hormone deficiencies.
Although the overall number of adverse cardiovascular events in childhood cancer survivors is
relatively small compared with the frequency of
tumor recurrence over time, as survivors age, the
long-term risk of death from cardiovascular disease and other chronic health conditions increases
and eventually surpasses that due to tumor recurrence (Armstrong et al. 2009a).
Components of Cardiovascular Risk

Development of obesity, dyslipidemia, hypertension, and impaired glucose metabolism
(sometimes termed insulin resistance) have
been associated with a significantly increased
risk of subsequent cardiovascular disease in the
general population (Alberti et al. 2009). The joint
development of these traits has been termed the
metabolic syndrome and is associated with an
increased risk of subsequent cardiovascular disease
greater than that associated with any of its individual components. It is estimated that individuals manifesting the metabolic syndrome are twice
as likely to develop cardiovascular disease over a
510 year period compared with individuals
without the syndrome (Alberti et al. 2009). Data
from a variety of retrospective and prospective
studies show that CNS tumor survivors are at
increased risk of obesity, dyslipidemia, and
hypertension (Heikens et al. 2000; Gurney et al.
2003b; Lustig et al. 2003; Neville et al. 2006;
Pietila et al. 2009; Meacham et al. 2009, 2010).
The data regarding development of impaired
glucose tolerance in this population are less clear.
CNS survivors also may experience an increased
risk of vasculopathy (e.g. stroke) and cardiomyopathy due to treatment-related exposures (Gurney
et al. 2003a; Bowers et al. 2006; Morris et al.
2009; Shankar et al. 2008; Mulrooney et al.
2009). Together, the development of these
conditions likely contributes to the overall
increased risk of cardiovascular mortality seen in
this population .
24 Adult Survivors of Pediatric Cancer: Risk of Cardiovascular Disease
249
Table 24.1 Summary of host and treatment factors known to be associated with cardiovascular late (CV) effects
Host factors
Tumor location
Age
Gender
Gene-treatment interactions
Treatment factors
Surgery
Radiotherapy
Chemotherapy
Anthracyclines
Alkylating agents
Glucocorticoids
Comments
Hypothalamic tumors more likely to result in disruption of hypothalamic pituitary axis
(HPA), leading to central hormone deficiencies and associated CV trait development.
Can also affect hypothalamic regulation of satiety, leading to hyperphagia and obesity
Treatment at younger age associated with greater risk
Females appear to be more susceptible
This area under active investigation
See tumor location above
Higher doses to the brain associated with increased risk of stroke (especially 18 Gy).
Doses to the HPA (especially 18 Gy) also associated with increased risk of growth
hormone deficiency and components of the metabolic syndrome (obesity, hypertension, dyslipidemia, and insulin resistance). Scatter to the neck also associated with
carotid disease. Scatter from spinal doses may affect the heart and great vessels
Associated with cardiomyopathy, in a dose-dependent fashion. Unclear if safe lower
dose threshold exists. Younger age of exposure and females appear to be at increased
risk
Higher doses associated with hypogonadism, which untreated is associated with
dyslipidemia
Associated with hypertension, insulin resistance and hyperglycemia, and obesity
Increased risk of obesity among CNS tumor

survivors has been reported across multiple
studies, although some only support an increased
risk in selected subgroups. The CCSS, which
prospectively enrolled children treated between
1970 and 1986 at 26 participating institutions in
North America, examined self-reported height
and weight from 921 adult CNS tumor survivors
(mean age 27 years; excluded nonmalignant
histologies such as craniopharyngioma) (Gurney
et al. 2003b). Investigators found that only female
survivors (n = 431) were at greater risk of obesity
as measured by body mass indices (BMI) compared with population norms, with risk greatest
among those treated at a younger age (age <9 years
versus 9 years). A retrospective longitudinal
analysis of BMI trends in 148 survivors treated at
St. Jude Childrens Research Hospital also
reported that younger age at diagnosis was associated with increased BMI over time (Lustig et al.
2003). Other smaller studies that used alternative
measures of obesity such as waist-hip ratios have
found that survivors may have increased central
adiposity (Heikens et al. 2000; Neville et al.
2006; Pietila et al. 2009). Hypothalamic tumors,
such as craniopharyngiomas, also are strongly
associated with subsequent obesity (Lustig et al.

2003). However, other histologies not always
located in the hypothalamic pituitary axis (HPA)
such as medulloblastomas and astrocytomas also
have been associated with obesity (Lustig et al.
2003; Meacham et al. 2010). In these latter
instances, these associations are more likely due
to the late adverse effects from treatment (surgery
and/or radiotherapy). Cranial radiotherapy has
been consistently associated with subsequent
obesity (Gurney et al. 2003b; Neville et al. 2006;
Pietila et al. 2009; Meacham et al. 2010). Among
female CNS patients, radiotherapy doses
4059 Gy to the HPA were associated with
increased obesity compared with unirradiated
patients (Gurney et al. 2003b), even after excluding patients with hypothalamic tumors (Lustig
et al. 2003). However, in studies of acute lymphoblastic leukemia patients, both males and
females who received 1824 Gy total cranial
radiotherapy were at increased risk of obesity
compared with those who received chemotherapy only (Garmey et al. 2008). However, similar
to the gender predisposition seen among CNS
survivors, female leukemia survivors appeared to
be at greater risk than males.
250
CNS tumor survivors also have been shown to

commonly exhibit dyslipidemia and have elevated blood pressure. In a single arm Finnish
study (n = 52; mean age 14 years), approximately
one-quarter of survivors had abnormal lipid values and elevated blood pressures (Pietila et al.
2009). In a small Dutch study (n = 26; mean age
26 years), survivors also had both higher LDL
levels and blood pressures, and lower HDL levels, compared with healthy controls (Heikens
et al. 2000). The CCSS found medulloblastoma
survivors (but not other histologies) more likely
to report being on lipid-lowering (OR 5.9; 95%
CI 3.89.1) and anti-hypertensive (OR 1.7; 95%
CI 1.02.9) medications compared with siblings
(Meacham et al. 2010). Treatment risk factors for
dyslipidemia included cranial radiotherapy and
growth hormone deficiency (Heikens et al. 2000;
Pietila et al. 2009; Meacham et al. 2010). Cranial
radiotherapy was associated with elevated blood
pressures in the Finnish study but not in the CCSS
cohort, although it is possible this discrepancy
may be due the latter studys definition of hypertension based on self-reported use of antihypertensive medications.
At present, it has not been shown that CNS
tumor survivors have a consistent risk of impaired
glucose metabolism, insulin resistance, or diabetes. An Australian study prospectively performed
glucose tolerance testing in 212 pubertal or adult
survivors of childhood cancer (<10% of this
cohort were CNS tumor survivors, however) and
found that 11% had impaired glucose metabolism
or frank diabetes mellitus (Neville et al. 2006).
Pituitary gland radiation exposure and growth
hormone deficiency were associated risk factors
in univariate analysis, although only untreated
hypogonadism was associated after multivariate
adjustment. Although exposure to cranial radiotherapy was associated with subsequent diabetes
in the CCSS (OR 1.6, 95% CI 1.02.3) compared
with siblings, CNS tumor patients (regardless of
radiotherapy exposures) were not at increased
risk in stratified analyses (Meacham et al. 2009).
In the CCSS, diabetes risk was greatest among
patients exposed to abdominal radiotherapy and
leukemia survivors treated with cranial radiotherapy and/or total body irradiation as used for
hematopoietic cell transplantation. Other studies

of leukemia and hematopoietic cell transplantation survivors also have identified an increased
prevalence of impaired glucose metabolism following these radiotherapy exposures (Oeffinger
et al. 2009; Chow et al. 2010). It is unclear
whether these discrepant results reflect the different cranial radiotherapy doses used historically
for leukemia (typically 1824 Gy) compared
with CNS tumors (typically much higher doses)
or represent other exposures unique to leukemia
survivors. However, as doses used to treat CNS
tumors become lower and more closely approximate leukemia treatment doses, it will be important to determine whether CNS tumor survivors
are at increased risk of abnormal glucose
metabolism.
Although CNS survivors are at increased risk
for some components of the metabolic syndrome,
it remains unclear whether survivors are more
likely to develop sufficient number of components to meet most definitions of this syndrome.
In the previously mentioned Finnish study, 8% of
pediatric aged survivors (n = 4) were classified as
having the metabolic syndrome (Pietila et al.
2009). All four patients had received cranial
radiotherapy and three of four were growth hormone deficient (none currently receiving supplementation). Studies of leukemia survivors have
reported an association between cranial radiotherapy and increasing number of metabolic syndrome components, correlated with degree of
growth hormone deficiency (Gurney et al. 2006).
A CCSS analysis examined the prevalence of
clustering of three or more of the following four
cardiometabolic risk factors: BMI 30 kg/m2,
and medication use for diabetes, hypertension,
and dyslipidemia (Meacham et al. 2010).
Investigators did not find CNS tumor survivors to
be at increased risk compared with siblings for
the presence of three or more of these cardiovascular risk factors, although estimates for medulloblastoma survivors were increased but imprecise
(odds ratio 3.1, 95% CI 0.86.8). One challenge to
interpreting and comparing research results has
been the variable definitions used to define the
metabolic syndrome in adults and the lack of a
formal definition among children (Alberti et al.
2009; Steinberger et al. 2009). Furthermore, investigators have raised concerns regarding the clinical utility of the metabolic syndrome concept in
pediatrics (Steinberger et al. 2009). Nevertheless,
given the relatively young age of most childhood
cancer survivors, it remains possible that as survivors age, they may begin manifesting multiple
syndrome components.
In addition to an increased risk of cardiovascular disease related to the components of the
metabolic syndrome, survivors also have specific
cancer treatment-related risks for stroke and primary heart disease (Shankar et al. 2008; Morris
et al. 2009). Data from the CCSS found that CNS
tumor patients (n = 1,871; mean age 27 years)
overall had RR 29.0, 95% CI 13.860.7 of stroke
(based on 63 events) occurring 5 years after initial cancer diagnosis compared with siblings
(Bowers et al. 2006). This corresponded to an
estimated 5.6% cumulative incidence 25 years
after diagnosis. Risk was significantly greater
among those who received cranial radiotherapy,
in a dose dependent fashion (Bowers et al. 2006),
but remained increased even among unirradiated
survivors compared with siblings (RR 12.9, 95%
CI 4.834.5). Overall, the risk of incident stroke
occurring 5 years after diagnosis was similar to
the risk within 5 years of diagnosis and as well as
risk during cancer treatment (Gurney et al.
2003a). The CCSS data also found that risk was
greatest among survivors treated with radiotherapy, surgery, and chemotherapy compared with
surgery alone or surgery plus radiotherapy, even
after adjustment for cranial radiotherapy dose.
CNS tumor survivors may also experience
increased carotid artery disease secondary to
scatter radiation to the neck from cranial or
craniospinal radiotherapy (Heikens et al. 2000).
Studies of other childhood cancer survivors and
adult cancer survivors suggest that doses 40 Gy
were most relevant (Morris et al. 2009), although
a recent small study (n = 30, mean age 28 years)
found increased carotid disease following
1829 Gy compared with controls (Meeske
et al. 2009).
Radiotherapy also has direct cardiotoxic
effects with a dose dependent effect (Shankar
et al. 2008). CNS tumor patients who receive spi-
251
nal radiotherapy may have scatter to the heart.

Anthracycline chemotherapy also has been
strongly associated with cardiomyopathy in a
dose dependent fashion (Mulrooney et al. 2009;
Tukenova et al. 2010). However, CNS tumor
patients are rarely exposed to this class of chemotherapy. Data from the CCSS found CNS tumor
survivors to have a 23 fold increased risk of self/
proxy-reported congestive heart failure, pericardial, and valvular disease, and a 6-fold increased
risk of myocardial infarction compared with siblings (Mulrooney et al. 2009). In this study, after
adjustment for anthracyclines and other covariates, heart radiation doses 15 Gy were significantly associated with these four cardiac
outcomes. In a separate French-British cohort
study (n = 4,122, all cancers; median follow-up
26 years), heart radiation doses as low as 5 Gy
was shown to be associated with increased cardiac disease compared with both population
norms as well as survivors who did not receive
any heart radiation (Tukenova et al. 2010).
Endocrine Factors
Cardiovascular disease among CNS tumor survivors
may be influenced by endocrine abnormalities as
well as direct cardiovascular injury. Survivors of
CNS tumors can experience a wide range of
endocrine abnormalities as a sequelae of the
tumor itself, as well as treatment late effects.
HPA injury has been well-documented in this
population (reviewed by Chemaitilly and Sklar
2010). Central hormone insufficiencies may lead
to an increased risk of cardiovascular disease
through a variety of mechanisms. Growth hormone deficiency is the most common deficit
observed after cranial radiotherapy, and has been
observed after doses as low as 18 Gy (Gurney
et al. 2006). Untreated, individuals with growth
hormone deficiency experience decreased linear
growth and changes in lipid and carbohydrate
metabolism. Growth hormone deficient patients
manifest dyslipidemia, decreased bone mineral
density, decreased reported vitality and altered
body composition characterized by increased
adiposity and decreased lean body mass
252
(Johannsson 2009). Although the primary indication

for GH replacement therapy in children has been
short stature and hypoglycemia, given the longterm metabolic and psychological sequelae among
deficient individuals, continued replacement therapy in adulthood may be warranted (GH Research
Society 2000; Johannsson 2009). Although GH
deficiency following cancer therapy is typically
permanent, it is still recommended that patients
be retested once they have achieved adult stature
to determine if they meet the more stringent criteria for GH deficiency for adults prior to considering life-long replacement therapy (GH Research
Society 2000). Some research suggests that adults
with partial deficiency may experience increased
cardiovascular risk as manifested by intermediate
lipid levels and thickening of carotid artery
endothelium (Murray et al. 2010). The cost and
benefits of replacement therapy in childhood cancer survivors also need to be weighed against the
possibility of an increased risk of second cancers
and other possible complications among supplemented individuals, particularly those who also
received prior radiotherapy (Ergun-Longmire
et al. 2006).
Other central hormone deficiencies that may
affect cardiovascular risk include hypothyroidism and hypogonadism. Untreated hypothyroidism is associated with dyslipidemia and weight
gain. Hypogonadism also can be associated with
dyslipidemia and altered body composition. In a
study of nearly 250 Australian childhood cancer
survivors (~10% CNS tumor survivors), untreated
hypogonadism also was reported to be a statistically significant risk factor for impaired glucose
tolerance, even after multivariate adjustment
(Neville et al. 2006). However, this finding has
not been reported by others.
Surveillance
Guided by the existing literature, the US-based
Childrens Oncology Group has developed
screening and surveillance guidelines for childhood cancer survivors based on treatment and
host factors (Morris et al. 2009; Shankar et al.
2008; Nandagopal et al. 2008). Similar efforts
also are occurring in other countries (Skinner

et al. 2006), and efforts are underway to try to
harmonize recommendations across countries
(Hudson M., personal communication). In
addition to recommendations promulgated by
oncology groups, other national groups have also
produced cardiovascular disease screening guidelines specific to children, including defining pediatric sub-populations (including cancer survivors)
that are at increased risk of future cardiovascular
disease (Kavey et al. 2006) (Table 24.2). Having
cancer survivor specific recommendations is
important given the increased risk of cardiovascular and other chronic diseases occurring in relatively young adult populations, ages at which
many general population-based screening recommendations are not yet applicable (US Preventative
Services Task Force 2010). However, in contrast
to many general population guidelines, the costeffectiveness of proposed screening studies and
screening intervals has not yet undergone more
rigorous analysis. Given the limited number of
older adult survivors of childhood cancer, the
applicability of current guidelines for older survivors also is unclear. However, most large cohort
studies suggest that the relative risk of cardiovascular and other chronic diseases does not plateau
and continues to increase compared with control
populations (Oeffinger et al. 2006; Mertens et al.
2008; Reulen et al. 2010).
Future Directions
With the ageing of existing survivors, the rates of
cardiovascular disease following cancer therapy
will likely increase and continue to be disproportionately high compared with the general population. Close follow-up of existing survivor cohorts
will help inform this risk as survivors age.
However, a limitation of survivorship research is
always the time lag between treatment and subsequent late effects and the potential challenges in
extrapolating effects of historical treatment to
contemporary protocols. With the dissemination
of multimodality therapy (surgery, radiotherapy,
chemotherapy) in pediatric oncology in the
1970s, many therapeutic agents used in the 1970s
253
Table 24.2 Published recommendations for cardiovascular disease surveillance

Screening
(condition)
US Preventive Services
Task Force (2010)
US Childrens Oncology
Group (Shankar et al.
2008; Nandagopal et al.
2008; Morris et al. 2009)
Blood pressure
(hypertension)
Annually for adults

(age 18 years)
Annually if treatment risk Regularly for all

factors present
survivors
Fasting lipids
(dyslipidemia)
Only individuals with

increased coronary heart
disease risk: males
2034 year; females
2044 year
No recommendations
2 year after completing

therapy and every 2 year
if risk factors present
EKG/echocardiogram
(cardiomyopathy)
Fasting glucose
(impaired glucose
tolerance/diabetes)
EKG 2 year after

completing therapy, and
echocardiogram every
15 years depending on
treatment exposures
No recommendation for Every 2 year in cancer
children; only asympsurvivors exposed to
tomatic adults with blood specific treatments
pressures
> 135/80 mmHg
UK Childrens
Cancer Study
Group (Skinner
et al. 2006)
Consider if
received chest
radiation or BMT
Echocardiogram
every 35 years
depending on
treatment
exposures
Survivors who
received BMT
American Heart
Association/
American
Academy of
Pediatrics (2004a;
Kavey et al. 2006)
Measured at
every healthcare
episode if >3 year
old
All cancer
survivors
No
recommendations
Baseline level for

all cancer
survivors
National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and
Adolescents
and 1980s continue to be used today. Nevertheless,

new studies will need to be developed to monitor
for any cardiovascular effects associated with
contemporary therapy that may differ in overall
intensity as well as incorporate new therapeutic
agents which may have unintended cardiovascular sequelae.
In addition, efforts are now underway to better
understand the contribution of host genetic variation in modifying risk. Polymorphisms in the leptin receptor (Ross et al. 2004) and carbonyl
reductase (Blanco et al. 2008) genes have been
associated with obesity and increased susceptibility to anthracycline-mediated cardiomyopathy
in childhood cancer survivors, respectively.
However, these findings await replication.
Selected lifestyle issues may also modify the
risk of cardiovascular disease in this population.
CCSS data found that CNS tumor survivors were
less likely to be physically active compared with

siblings and compared with other cancer survivors. Medulloblastoma survivors were the most
likely (35%) to report an inactive lifestyle (Ness
et al. 2009). Among treatment factors, cranial
radiotherapy exposure was associated with significant RRs 1.21.5 for both genders. The other
significant treatment risk factors in this analysis
were anthracycline-containing chemotherapy
regimens and lower extremity amputations. In
contrast to physical inactivity, CNS tumor survivors were less likely to smoke compared with
siblings in a separate CCSS analysis (Emmons
et al. 2002). Childhood cancer survivors also
appear to be interested in adopting healthier
lifestyles (Demark-Wahnefried et al. 2005),
and interventions designed to promote components of such lifestyles have shown some success (Emmons et al. 2009).
254
Conclusion
With the increasing proportion of childhood CNS
tumor patients being cured, survivors face a variety of late sequelae related to their cancer and
cancer therapy. These include neuroendocrine
abnormalities that can adversely affect cardiovascular health. Some tumor and therapy-related
effects also directly impact vascular integrity and
to a lesser degree, cardiac function. Awareness of
these late effects may improve adherence to survivor- and therapy-specific health surveillance
guidelines, leading to earlier detection and
possible intervention. New research is focused
on better understanding genetic susceptibility to
these cardiovascular sequelae among cancer survivors and designing effective interventions that
reduce cardiovascular morbidity associated with
cancer therapy.
References
Alberti KG, Eckel RH, Grundy SM, Zimmet PZ, Cleeman
JI, Donato KA, Fruchart JC, James WP, Loria CM,
Smith SC Jr (2009) Harmonizing the metabolic syndrome: a joint interim statement of the International
Diabetes Federation Task Force on Epidemiology and
Prevention; National Heart, Lung, and Blood Institute;
American Heart Association; World Heart Federation;
International Atherosclerosis Society; and International
Association for the Study of Obesity. Circulation
120:16401645
Altekruse SF, Kosary CL, Krapcho M, Neyman N,
Aminou R, Waldron W, Ruhl J, Howlader N, Tatalovich
Z, Cho H, Mariotto A, Eisner MP, Lewis DR, Cronin
K, Chen HS, Feuer EJ, Stinchcomb DG, Edwards BK
(eds) (2010) SEER cancer statistics review, 1975
2007. National Cancer Institute, Bethesda, MD. http://
seer.cancer.gov/csr/1975_2007/. Accessed 1 Dec 2010
Armstrong GT, Liu Q, Yasui Y, Neglia JP, Leisenring W,
Robison LL, Mertens AC (2009a) Late mortality
among 5-year survivors of childhood cancer: a summary from the Childhood Cancer Survivor Study. J
Clin Oncol 27:23282338
Armstrong GT, Liu Q, Yasui Y, Huang S, Ness KK,
Leisenring W, Hudson MM, Donaldson SS, King AA,
Stovall M, Krull KR, Robison LL, Packer RJ (2009b)
Long-term outcomes among adult survivors of childhood central nervous system malignancies in the
Childhood Cancer Survivor Study. J Natl Cancer Inst
101:946958

Blanco JG, Leisenring WM, Gonzalez-Covarrubias VM,
Kawashima TI, Davies SM, Relling MV, Robison LL,
Sklar CA, Stovall M, Bhatia S (2008) Genetic polymorphisms in the carbonyl reductase 3 gene CBR3
and the NAD(P)H:quinone oxidoreductase 1 gene
NQO1 in patients who developed anthracycline-related
congestive heart failure after childhood cancer. Cancer
112:27892795
Bowers DC, Liu Y, Leisenring W, McNeil E, Stovall M,
Gurney JG, Robison LL, Packer RJ, Oeffinger KC
(2006) Late-occurring stroke among long-term survivors of childhood leukemia and brain tumors: a report
from the Childhood Cancer Survivor Study. J Clin
Oncol 24:52775282
Chemaitilly W, Sklar CA (2010) Endocrine complications
in long-term survivors of childhood cancers. Endocr
Relat Cancer 17:R141R159
Chow EJ, Simmons JH, Roth CL, Baker KS, Hoffmeister
PA, Sanders JE, Friedman DL (2010) Increased cardiometabolic traits in pediatric survivors of acute
lymphoblastic leukemia treated with total body
irradiation. Biol Blood Marrow Transplant
16:16741681
Demark-Wahnefried W, Werner C, Clipp EC, Guill AB,
Bonner M, Jones LW, Rosoff PM (2005) Survivors of
childhood cancer and their guardians. Cancer
103:21712180
Emmons K, Li FP, Whitton J, Mertens AC, Hutchinson R,
Diller L, Robison LL (2002) Predictors of smoking
initiation and cessation among childhood cancer
survivors: a report from the childhood cancer survivor
study. J Clin Oncol 20:16081616
Emmons KM, Puleo E, Mertens A, Gritz ER, Diller L, Li
FP (2009) Long-term smoking cessation outcomes
among childhood cancer survivors in the Partnership
for Health Study. J Clin Oncol 27:5260
Ergun-Longmire B, Mertens AC, Mitby P, Qin J, Heller
G, Shi W, Yasui Y, Robison LL, Sklar CA (2006)
Growth hormone treatment and risk of second neoplasms in the childhood cancer survivor. J Clin
Endocrinol Metab 91:34943498
Garmey EG, Liu Q, Sklar CA, Meacham LR, Mertens
AC, Stovall MA, Yasui Y, Robison LL, Oeffinger
KC (2008) Longitudinal changes in obesity and
body mass index among adult survivors of childhood
acute lymphoblastic leukemia: a report from the
Childhood Cancer Survivor Study. J Clin Oncol
26:46394645
GH Research Society (2000) Consensus guidelines for the
diagnosis and treatment of growth hormone (GH) deficiency in childhood and adolescence: summary statement of the GH Research Society. J Clin Endocrinol
Metab 85:39903993
Gurney JG, Kadan-Lottick NS, Packer RJ, Neglia JP, Sklar
CA, Punyko JA, Stovall M, Yasui Y, Nicholson HS,
Wolden S, McNeil DE, Mertens AC, Robison LL
(2003a) Endocrine and cardiovascular late effects
among adult survivors of childhood brain tumors:
Childhood Cancer Survivor Study. Cancer 97:663673

Gurney JG, Ness KK, Stovall M, Wolden S, Punyko JA,
Neglia JP, Mertens AC, Packer RJ, Robison LL, Sklar
CA (2003b) Final height and body mass index among
adult survivors of childhood brain cancer: Childhood
Cancer Survivor Study. J Clin Endocrinol Metab
88:47314739
Gurney JG, Ness KK, Sibley SD, OLeary M, Dengel DR,
Lee JM, Youngren NM, Glasser SP, Baker KS (2006)
Metabolic syndrome and growth hormone deficiency
in adult survivors of childhood acute lymphoblastic
leukemia. Cancer 107:13031312
Heikens J, Ubbink MC, van der Pal HP, Bakker PJ, Fliers
E, Smilde TJ, Kastelein JJ, Trip MD (2000) Long term
survivors of childhood brain cancer have an increased
risk for cardiovascular disease. Cancer 88:21162121
Johannsson G (2009) Treatment of growth hormone deficiency in adults. Horm Res 71(Suppl 1):116122
Kavey RE, Allada V, Daniels SR, Hayman LL, McCrindle
BW, Newburger JW, Parekh RS, Steinberger J (2006)
Cardiovascular risk reduction in high-risk pediatric
patients: a scientific statement from the American
Heart Association Expert Panel on Population and
Prevention Science; the Councils on Cardiovascular
Disease in the Young, Epidemiology and Prevention,
Nutrition, Physical Activity and Metabolism, High
Blood Pressure Research, Cardiovascular Nursing,
and the Kidney in Heart Disease; and the
Interdisciplinary Working Group on Quality of Care
and Outcomes Research: endorsed by the American
Academy of Pediatrics. Circulation 114:27102738
Lustig RH, Post SR, Srivannaboon K, Rose SR, Danish
RK, Burghen GA, Xiong X, Wu S, Merchant TE
(2003) Risk factors for the development of obesity in
children surviving brain tumors. J Clin Endocrinol
Metab 88:611616
Meacham LR, Sklar CA, Li S, Liu Q, Gimpel N, Yasui Y,
Whitton JA, Stovall M, Robison LL, Oeffinger KC
(2009) Diabetes mellitus in long-term survivors of
childhood cancer. Increased risk associated with radiation therapy: a report for the childhood cancer survivor
study. Arch Intern Med 169:13811388
Meacham LR, Chow EJ, Ness KK, Kamdar KY, Chen Y,
Yasui Y, Oeffinger KC, Sklar CA, Robison LL,
Mertens AC (2010) Cardiovascular risk factors in
adult survivors of pediatric cancer a report from the
childhood cancer survivor study. Cancer Epidemiol
Biomarkers Prev 19:170181
Meeske KA, Siegel SE, Gilsanz V, Bernstein L, Nelson
MB, Sposto R, Weaver FA, Lavey RS, Mack MP,
Nelson MD Jr (2009) Premature carotid artery disease
in pediatric cancer survivors treated with neck irradiation. Pediatr Blood Cancer 53:615621
Mertens AC, Liu Q, Neglia JP, Wasilewski K, Leisenring
W, Armstrong GT, Robison LL, Yasui Y (2008) Causespecific late mortality among 5-year survivors of
childhood cancer: the Childhood Cancer Survivor
Study. J Natl Cancer Inst 100:13681379
Morris B, Partap S, Yeom K, Gibbs IC, Fisher PG, King
AA (2009) Cerebrovascular disease in childhood
255
cancer survivors: a Childrens Oncology Group

Report. Neurology 73:19061913
Mulrooney DA, Yeazel MW, Kawashima T, Mertens AC,
Mitby P, Stovall M, Donaldson SS, Green DM, Sklar
CA, Robison LL, Leisenring WM (2009) Cardiac
outcomes in a cohort of adult survivors of childhood
and adolescent cancer: retrospective analysis of the
Childhood Cancer Survivor Study cohort. BMJ
339:b4606
Murray RD, Wieringa G, Lawrance JA, Adams JE, Shalet
SM (2010) Partial growth hormone deficiency is associated with an adverse cardiovascular risk profile and
increased carotid intima-medial thickness. Clin
Endocrinol (Oxf) 73(4):508515
Nandagopal R, Laverdiere C, Mulrooney D, Hudson MM,
Meacham L (2008) Endocrine late effects of childhood cancer therapy: a report from the Childrens
Oncology Group. Horm Res 69:6574
National High Blood Pressure Education Program
Working Group on High Blood Pressure in Children
and Adolescents (2004) The fourth report on the
diagnosis, evaluation, and treatment of high blood
pressure in children and adolescents. Pediatrics
114:555576
Ness KK, Leisenring WM, Huang S, Hudson MM, Gurney
JG, Whelan K, Hobbie WL, Armstrong GT, Robison
LL, Oeffinger KC (2009) Predictors of inactive lifestyle among adult survivors of childhood cancer: a
report from the Childhood Cancer Survivor Study.
Cancer 115:19841994
Neville KA, Cohn RJ, Steinbeck KS, Johnston K, Walker
JL (2006) Hyperinsulinemia, impaired glucose tolerance, and diabetes mellitus in survivors of childhood
cancer: prevalence and risk factors. J Clin Endocrinol
Metab 91:44014407
Oeffinger KC, Mertens AC, Sklar CA, Kawashima T,
Hudson MM, Meadows AT, Friedman DL, Marina N,
Hobbie W, Kadan-Lottick NS, Schwartz CL,
Leisenring W, Robison LL (2006) Chronic health
conditions in adult survivors of childhood cancer.
N Engl J Med 355:15721582
Oeffinger KC, Adams-Huet B, Victor RG, Church TS,
Snell PG, Dunn AL, Eshelman-Kent DA, Ross R,
Janiszewski PM, Turoff AJ, Brooks S, Vega GL (2009)
Insulin resistance and risk factors for cardiovascular
disease in young adult survivors of childhood acute
lymphoblastic leukemia. J Clin Oncol 27:36983704
Pietila S, Makipernaa A, Sievanen H, Koivisto AM,
Wigren T, Lenko HL (2009) Obesity and metabolic
changes are common in young childhood brain tumor
survivors. Pediatr Blood Cancer 52:853859
US Preventative Services Task Force (2010) http://
www.uspreventiveservicestaskforce.org/. Accessed
1 Dec 2010
Reulen RC, Winter DL, Frobisher C, Lancashire ER,
Stiller CA, Jenney ME, Skinner R, Stevens MC,
Hawkins MM (2010) Long-term cause-specific mortality among survivors of childhood cancer. JAMA
304:172179
256
Ross JA, Oeffinger KC, Davies SM, Mertens AC, Langer
EK, Kiffmeyer WR, Sklar CA, Stovall M, Yasui Y,
Robison LL (2004) Genetic variation in the leptin receptor gene and obesity in survivors of childhood acute
lymphoblastic leukemia: a report from the Childhood
Cancer Survivor Study. J Clin Oncol 22:35583562
Shankar SM, Marina N, Hudson MM, Hodgson DC,
Adams MJ, Landier W, Bhatia S, Meeske K, Chen
MH, Kinahan KE, Steinberger J, Rosenthal D (2008)
Monitoring for cardiovascular disease in survivors of
childhood cancer: report from the Cardiovascular
Disease Task Force of the Childrens Oncology Group.
Pediatrics 121:e387e396
Skinner R, Wallace WH, Levitt G (eds) (2006) Therapy
based long term follow up, 2nd edn. United Kingdom
Childrens Cancer Study Group. http://www.cclg.org.
uk/treatment-research/guidelines. Accessed 5 Jan 2012

Steinberger J, Daniels SR, Eckel RH, Hayman L, Lustig
RH, McCrindle B, Mietus-Snyder ML (2009) Progress
and challenges in metabolic syndrome in children and
adolescents: a scientific statement from the American
Heart Association Atherosclerosis, Hypertension, and
Obesity in the Young Committee of the Council on
Cardiovascular Disease in the Young; Council on
Cardiovascular Nursing; and Council on Nutrition,
Physical Activity, and Metabolism. Circulation
119:628647
Tukenova M, Guibout C, Oberlin O, Doyon F, Mousannif
A, Haddy N, Guerin S, Pacquement H, Aouba A,
Hawkins M, Winter D, Bourhis J, Lefkopoulos D,
Diallo I, de Vathaire F (2010) Role of cancer treatment in long-term overall and cardiovascular
mortality after childhood cancer. J Clin Oncol
28:13081315
Part III
Gliomas
Pediatric Glioma: Role of PlateletDerived Growth Factor Receptor
25
Tobey J. MacDonald
Contents
Abstract
Introduction ............................................................
260
PDGFR Signaling ...................................................
260
PDGFR in Glioma Pathogenesis ...........................
260
PDGFR and Neural Stem Cells ............................
262
PDGFR Pathway Expression in Gliomas.............
262
PDGFR in Glioma Angiogenesis...........................
263
PDGFR in Pediatric Gliomas ................................
263
Therapeutic Targeting of PDGFR in Glioma ......
264
References ...............................................................
265
T.J. MacDonald (*)

Childrens Healthcare of Atlanta, Aflac Cancer Center
and Blood Disorders Service, Emory Childrens Center,
Emory University, 2015 Uppergate Drive NE, Suite 442,
Atlanta, GA 30322, USA
e-mail: tobey.macdonald@emory.edu
Abundant evidence exists, especially in adult

patients, implicating a critical role for PDGFR
in the pathogenesis of human gliomas, including induction of primary oncogenesis as well as
the maintenance of established glioma survival,
growth, invasive spread, angiogenesis, and progression from lower to higher grades. More
recent evidence also points towards a central
role for PDGFR in maintaining neural stem
cells in adults and in the possible recruitment of
cancer stem cells and other CNS progenitor and
facilitating cells leading to the development
and progression of gliomas. Although the data
are more limited in comparison to that described
in the adult literature, emerging evidence now
indicates that PDGFR amplification is the single most common genetic alteration in pediatric
high-grade gliomas, including diffuse intrinsic
pontine gliomas. However, the expression patterns and potential functional roles of PDGFR
in gliomas may be distinct between adults and
children. Investigations to establish the functional role and clinical relevance of PDGFR
expression and PDGFR activity in childhood
glioma are ongoing. Despite the lack of success
to date of targeted PDGFR inhibitors in clinical
trials for the treatment of high-grade gliomas,
PDGFR remains an attractive target for the
development of novel therapeutic strategies for
gliomas given the critical role that PDGFR
signaling plays in oncogenic growth- and
survival-promoting pathways.

259
260
Introduction
The platelet-derived growth factor receptor
(PDGFR) regulates several essential functions in
normal cells, including cellular proliferation, differentiation and migration, and is widely
expressed in a variety of malignancies (Andrae
et al. 2008). Importantly, PDGFR stimulation can
subsequently lead to the activation of the oncogenic Src and Ras signal transduction pathways,
which results in the downstream activation of a
wide range of Src- and Ras-dependent cellular
effectors (Heldin et al. 1998). In gliomas of adult
patients, the PDGF receptors and their cognate
ligands are often overexpressed and dysregulation of PDGFR activity has been linked to the
pathogenesis of these tumors (Smits and Funa
1998). The importance of PDGFR signaling to
glioma progression has made this receptor an
attractive candidate target for therapeutic intervention. Although there is currently only limited
information with regard to the expression and
activity of PDGFR in pediatric gliomas, emerging evidence points to a similarly critical role in
the development, growth, survival, and angiogenesis of these tumors in children.
PDGFR Signaling
Platelet-derived growth factor receptors are members of the protein-tyrosine kinase family of receptors and consist of two structurally related cell
surface receptor types, known as the alpha-receptor
(PDGFRA) and the beta-receptor (PDGFRB). The
receptors are transmembrane proteins with an
intracellular, ligand-stimulatable protein tyrosine
kinase domain (Claesson-Welsh 1996). Five
dimeric isoforms of platelet-derived growth factor
(PDGF), the ligand for PDGFR, have been
isolated: PDGF-AA, PDGF-AB, PDGF-BB,
PDGF-CC, and PDGF-DD (Lokker et al. 2002).
These disulfide-bonded chains bind with different
affinities and specificities to the receptors and, in
part, mediate different cellular functions
(Fredriksson et al. 2004). PDGF effects are mediated by both autocrine and paracrine mechanisms.
T.J. MacDonald
The receptors are activated upon ligand-induced

dimerization, whereby the receptors become
phosphorylated on tyrosine residues. Heldin et al.
(1998) first showed that these phosphorylated
residues form attachment sites for recruitment
of additional signaling molecules, which in turn
activate the Ras-Raf pathway. In addition,
PDGFR can directly activate phosphatidylinositol 3-kinase (PI3K), mitogen-activated protein
kinase (MAPK), Jak family kinase, Src family
kinase, and phospholipase C gamma signal
transduction pathways, leading to a cascade of
cellular events, which have been shown to
require intact PDGFR tyrosine kinase activity
(Vantler et al. 2006). In the developing central
nervous system, astrocyte-derived PDGF is a
potent mitogen for O-2A progenitor cells, and is
crucial for the proper timing of oligodendrocyte
differentiation and control of myelination (Hu
et al. 2008). Rubin et al. (1988) reported that in
certain cells, PDGFRB expression is inducible
such that cells in normal tissue, which do not
typically express receptors in vivo, begin to
express the receptor in inflammatory lesions or
when the cells are explanted in vitro. Thus,
PDGFR signaling cascade appears to be an
important component of both normal development and pathologic conditions, and as such, is
tightly regulated in both its expression and
activity.
PDGFR in Glioma Pathogenesis

A wide range of work from in vitro studies to
mouse models has implicated the PDGFR signal
transduction pathway in various processes of
glioma pathogenesis including proliferation, cellular migration, development, and angiogenesis.
One of the critical early findings in support of the
role for PDGFR in gliomagenesis was the demonstration that the v-sis oncogene of the simian
sarcoma virus (SSV), a retroviral homolog of the
PDGFB-chain gene, induces transformation by
an autocrine activation of PDGF receptors at the
cell surface, and that infection with SSV induces
malignant gliomas in experimental animals, suggesting a role for autocrine-activated PDGFR in
25
Pediatric Glioma: Role of Platelet-Derived Growth Factor Receptor
glioma development (Doolittle et al. 1983; Huang

et al. 1984). PDGF antagonists blocking the
PDGF autocrine activated pathway revert the
transformed phenotype of these tumor cells
(Garrett et al. 1984). Similarly, Potapova et al.
(1996) reported that T98G human glioblastoma
cells, which are nontumorigenic in athymic mice,
endogenously express predominantly PDGFRB
type receptors and continuously secrete small
amounts of PDGF-BB. Furthermore, the addition
of specific anti-PDGF-BB antibody inhibited the
proliferation of T98G cells in culture, and conversely, multiple clonal cell lines that were
constructed to stably overexpress PDGF-BB
exhibited a striking increase (200250%) in cell
proliferation rate, an enhanced colony-forming
frequency in soft agar, and reliably developed
tumors in 46 weeks.
Further support for a direct oncogenic role for
PDGF, was demonstrated by Shih et al. (2004) in
newborn mice injected intracerebrally with a
Moloney murine leukemia retrovirus carrying the
v-sis/PDGF-BB oncogene and a replication competent helper virus. In these studies, mice also
developed brain tumors with many characteristics of human gliomas, and in particular, demonstrated a dose-dependent effect of PDGF-BB on
glial tumorigenesis. Analysis of proviral integrations in these brain tumors identified almost 70
common insertion sites (CISs). These CISs
harbored known but also putative novel cancercausing genes, and microarray analysis further
identified differentially expressed genes in the
mouse brain tumors compared to normal brain; in
particular, it was observed that known tumor
genes and markers of immature cells were upregulated in the tumors. Elevation in PDGF-BB
production resulted in tumors with shortened
latency, increased cellularity, regions of necrosis,
and were considered overall to be of high-grade
glioblastoma multiforme (GBM)-like character.
Long latency tumors resembled slow-growing
oligodendrogliomas and contained significantly
less integrations in comparison to the short
latency tumors that developed 1342 weeks after
injection. Active PDGFR signaling appeared to
be required for the maintenance of the tumors
with high-grade characteristics because treatment
261
of the high-grade tumors with a small molecule

inhibitor of PDGFR activation resulted in reversion of these tumors to relatively lower grade
tumor histology. Together, these studies suggest
that PDGFR signaling quantitatively regulates
glioma tumor grade and is required to sustain
high-grade gliomagenesis.
Although both the alpha- and beta-receptors
are structurally related to the v-fms and v-kit
oncogenes, the PDGF receptors have not been
shown to have direct transforming potential. In
malignant gliomas of adult patients, one of the
most common molecular genetic alterations identified has been constitutive activation of PDGFR
autocrine signaling characterized by coexpression and colocalization of PDGFs and PDGFR
(Lokker et al. 2002). More recent evidence by
Ishii et al. (2008), and others, implicates PDGFR
in maintaining neural stem cells and recruitment
of stem cells to sites of CNS injury or tumor formation, indicating that PDGFR dysregulation
may play several critical roles in gliomagenesis.
While many researchers view brain tumors as
clonal entities derived from the cancer stem cell,
recent documentation of the importance of the
tumor microenvironment for glioma initiation
and progression revealed additional complexities
in brain tumorigenesis. The observed paracrine
effects of PDGF in animal models of gliomagenesis and the growth factor-rich environment of
brain tumors suggest that progenitor cell recruitment may play a role in gliomagenesis. In this
view, glioma formation involves recruitment of
progenitor and tumor-facilitating cells from the
adjacent brain and possibly other sites. Indeed,
Shih et al. (2004) also showed that elevated
PDGF-BB levels in experimental mouse tumors
mediated vascular smooth muscle cell recruitment that supported tumor angiogenesis.
Furthermore, Toepoel et al. (2008) identified five
distinct haplotypes for the PDGFRA promoter
region, designated H1, H2alpha, H2beta,
H2gamma, and H2delta. Of these haplotypes, H1
and H2alpha are the most common, whereby H1
drives low and H2alpha high transcriptional
activity in transient transfection assays. In a case
control study of 71 glioblastoma patients, there
was a clear underrepresentation of H1 alleles
T.J. MacDonald
262
observed, indicating that PDGFRA promoter

haplotypes may predispose to gliomas. In this
proposed model, PDGFRA is upregulated in a
haplotype-specific manner during neural stem
cell differentiation, which subsequently affects
the pool size of cells that can later undergo
gliomagenesis.
PDGFR and Neural Stem Cells

Neurons and oligodendrocytes are produced in
the adult brain subventricular zone (SVZ) from
neural stem cells (B cells), which express GFAP
and have morphological properties of astrocytes.
Jackson et al. (2006) identified B cells expressing
PDGFRA in the adult SVZ and showed that these
PDGFRA expressing B cells are functionally able
to generate neurons and oligodendrocytes in vivo.
Furthermore, conditional knock-out of PDGFRA
in a subpopulation of postnatal stem cells of the
SVZ prevented oligodendrocyte differentiation
while mature neurons were unaffected, indicating
that PDGFRA is required for oligodendrogenesis,
but not neurogenesis. Infusion of PDGF alone
was sufficient to induce SVZ B cell proliferation,
which contributed to the generation of large
hyperplasias with some features of gliomas. More
recently, Ishii et al. (2008) reported the presence
of PDGFRB expressing cells in the SVZ of newborn mice, but there were no morphological
abnormalities noted in the PDGFRB conditionally deleted mice. However, PDGFRB (/)
neurospheres demonstrated increased apoptosis,
decreased migration, and reduced neuronal
differentiation in response not only to PDGF but
also to bFGF, indicating the possibility of nonredundant signaling pathway convergence
between PDGFR and bFGFR. Together, these
findings demonstrate that activation of PDGFR
signaling occurs early in the adult stem cell lineage, and may help to regulate the balance
between oligodendrocyte and neuron differentiation, and that excessive PDGFR activation in the
SVZ in stem cells may be sufficient to induce
hallmarks associated with early stages of tumor
formation.
PDGFR Pathway Expression

in Gliomas
The various forms of PDGF, as well as the two
PDGF receptors, have been shown to be expressed
at high frequency in adult human gliomas, especially GBM. For example, in an analysis of
PDGFR expression in adult GBM by Fleming
et al. (1992), amplification and overexpression of
PDGFRA were observed in distinct subsets of
GBMs. For instance, amplification of PDGFRA
was detected in 8% GBM, whereas Western blot
analysis of the same tumors showed elevated
protein expression of PDGFRA in 24% GBM.
Increased PDGFRA protein was observed in
the presence or absence of gene amplification
and 3 out of the 4 tumors with elevated levels of
PDGFRA also overexpressed PDGFRB, which
was present as a single copy gene in all of the
tumors analyzed. Hermanson et al. (1992)
reported increased expression of PDGF-AA and
PDGF-BB chains in GBMs relative to lower
grade gliomas, while PDGFRA mRNA was found
to be heterogeneously expressed in all glioma
grades. Expression of PDGFRA in GBMs tended
to be colocalized in tumor cells with PDGF,
whereas mRNA and protein expression of
PDGFRB colocalized to PDGF-BB in GBM
endothelium. In a subsequent study of 83 adult
human gliomas by Hermanson et al. (1996), in
situ hybridization confirmed increased PDGFRA
mRNA expression in all of the tumors irrespective of stages of malignancy, although the highest
levels of PDGFRA expression were found in
GBM. Of these, 7 of 43 (16%) GBM, but none of
the other tumor grades tested, showed amplification of the PDGFRA gene, suggesting that a
mechanism other than gene amplification is
responsible for the overexpression of PDGFRA
mRNA in gliomas. Comparison of the in situ
hybridization data with genetic alterations in the
same tumors showed a significant correlation of
loss of heterozygosity on chromosome 17p with
high expression levels of PDGFRA in both highand low-grade gliomas, suggesting that PDGFRA
activity may in part be responsible for driving
25
tumor cell proliferation in early and late stages of

glioma development. The demonstrated colocalization of expression of PDGFs and PDGFRs
indicates the possibility for both autocrine and
paracrine activation of PDGFR in high-grade
gliomas.
Lokker et al. (2002) examined the expression
of all PDGF ligands and receptors in 11 glioma
cell lines and 5 primary GBM tumor tissues by
quantitative reverse transcription-PCR. The
expression of functional PDGF/PDGFR pairs
was identified in all of the samples tested.
PDGF-CC expression was ubiquitous in brain
tumor cells and tissues, but was very low or
absent in normal adult and fetal brain, while
PDGF-DD was expressed in 10 of 11 brain tumor
cell lines and 3 of 5 primary brain tumor samples.
Active PDGF autocrine signaling was evident in
all five glioblastoma cell lines tested. Together,
these data confirm that both types of PDGFRs
and all of their respective PDGF ligands are
widely expressed in adult gliomas and that the
PDGFR pathway is likely to be active in gliomas
and contributes to glioma progression.
PDGFR in Glioma Angiogenesis

Endothelial cells proliferate during brain development and are quiescent in normal adult brain
but proliferate again under pathologic conditions
such as glioma growth. The vascular phenotype
of low grade glioma is comparable to normal
brain; however, high grade gliomas are focally
highly vascularized and there is an associated
prominent endothelial cell proliferation. The
mechanisms of this change in vascular phenotype
are not fully understood; however, there is evidence that PDGFR activation and signaling play
an important role in this process as well as in
normal angiogenesis and vascular differentiation.
Specifically, increased numbers of PDGFRB
have been demonstrated on neovascular endothelial cells in several pathological conditions including wound healing, inflammation, and glioma
tumorigenesis. For example, Plate et al. (1992)
performed in situ hybridization and immunocytochemistry in normal human brain, astrocytoma
263
(WHO grade II), anaplastic oligo-astrocytoma

(WHO grade III), and GBM (WHO grade IV)
and found that PDGFRB mRNA was not detectable in the vessels of normal human brain, but
was expressed in the vasculature of low and high
grade gliomas, particularly in endothelial cell
proliferations in GBM. This study also reported
that primary cultures of endothelial cells derived
from GBM maintained PDGF receptor expression for 2 days in vitro, whereas it was not detectable in vitro in endothelial cells derived from
normal brain. In a report by Lafuente et al. (1999),
endothelium, multinucleated giant cells, perivascular tumor cells, and peritumoral microglia-like
cells of GBMs were all shown to be positive for
PDGFRB, and a correlation was found between
PDGFRB and Ki67, indicating that PDGFRB is
likely closely linked to neoangiogeneis and tumor
proliferation. Wang et al. (1999) showed expression of wild-type PDGFRB by normal aortic
endothelial cells and that PDGFRB was also able
to affect the transcriptional expression of vascular endothelial growth factor (VEGF), a critical
angiogenesis regulator and mitogen for such
cells. Together, these studies indicate that the
malignant phenotype in human gliomas is
associated with an upregulation of PDGFRB on
endothelial cells of blood vessels, which in turn,
upregulates VEGF in order to augment tumor
vascularization.
PDGFR in Pediatric Gliomas

To date, most of the evidence for the role of
PDGFR in glioma pathogenesis has been derived
from studies involving adult human gliomas.
In an attempt to confirm that similar alterations in
PDGFR expression and activity are found in
pediatric gliomas, recent studies have begun to
focus on childhood glioma tumor tissue. In a
study by Wong et al. (2006), single nucleotide
polymorphic (SNP) allele arrays were used to
analyze 28 pediatric gliomas consisting of 14
high-grade gliomas and 14 low-grade gliomas.
Amplification of PDGFRA was detected in
several cases of GBM analyzed and interestingly,
PDGFRA amplifications were located within
264
regions of LOH, indicating that in some pediatric

GBM one allele of PDGFRA was lost but the
remaining allele was amplified. Nakamura et al.
(2007) performed an extensive characterization
of 44 pediatric astrocytomas, including 16 diffuse astrocytomas (WHO grade II), 10 anaplastic
astrocytomas (WHO grade III), and 18 glioblastomas (WHO grade IV), but did not detect amplification of the PDGFRA gene in any tumor.
In a later study, Thorarinsdottir et al. (2008)
used tissue microarrays generated from untreated
tumors from 85 newly diagnosed pediatric glioma
patients (22 high-grade gliomas (HGG) and 63
low-grade gliomas (LGG)) and performed immunohistochemistry to assess the total and phosphorylated (activated) protein expression of PDGFR
and several of the major PDGFR pathway downstream effectors, PTEN, Akt, MAPK, and the
mammalian target of rapamycin (mTOR). The levels
of expression for each marker were correlated with
each other and with the clinicopathologic data,
including extent of initial tumor resection, evidence
of dissemination, tumor grade, proliferation
index, and survival, as well as with mRNA gene
expression profiles previously obtained from a
subset of these tumors. The investigators reported
that high expression of phospho-PDGFRA and
PDGFRB was observed in ~8086% of HGG
and ~4042% of LGG and that there was a significant association of activated PDGFR, and
mRNA levels of members of the PDGFR pathway with the malignant histology of childhood
gliomas. Furthermore, loss of expression of
PTEN, an inhibitor of the PDGFR signaling
pathway, was associated with worse overall
survival. In another study by Liang et al. (2008)
the protein expression profile of a series of 42
pediatric malignant gliomas was investigated.
Immunopositivity for PDGFRA was 45%,
PDGFRB 31%, and PTEN 67%. In this study,
there was no significant effect of the expression
level of any of the markers tested on the overall
survival or progression-free survival of the
patients; however, the activated forms of PDGFR
were not evaluated and they did observe that all
long-term survivors expressed PTEN.
In a more recent study of 63 pediatric highgrade gliomas, Bax et al. (2010) found PDGFRA
T.J. MacDonald
amplification to be the most frequent focal

alteration. In this series, PDGFRA amplification
was present in up to 17% of primary pediatric
GBM, 29% of diffuse intrinsic pontine glioma
(DIPG), and 50% of cases of high-grade glioma
arising as a secondary malignancy after irradiation. Notably, a number of cases without PDGFRA
amplification were still found to show overexpression of a specific PDGFRA-associated gene
signature, which was distinct from that observed
in adult cases with PDGFRA amplification.
Similarly, in a study of 11 DIPG samples,
Zarghooni et al. (2010) also found evidence of
PDGFR amplification in 4 (36%) tumor samples.
Immunohistochemistry revealed strong expression
of PDGRFA in 7 of the 11 DIPG samples and
weak expression in the remainder. The relatively
higher frequency of PDGFR amplification in
DIPG suggests that it is molecularly distinct from
other pediatric high-grade gliomas. Furthermore,
in a direct comparison of adult and pediatric
high-grade gliomas using a high resolution
genomic screening approach, Paugh et al. (2010)
confirmed that PDGFRA is the predominant target of focal amplification in pediatric high-grade
gliomas, including DIPG, whereby EGFR is the
predominant amplification in the adult cases.
Taken together, these findings suggest that
PDGFR may be a clinically relevant therapeutic
target in pediatric high-grade gliomas.
Therapeutic Targeting of PDGFR

in Glioma
Because of the wealth of evidence implicating
PDGFRs in glioma pathogenesis, including the
promotion of tumor growth by PDGF via autocrine stimulation of malignant glial cells, by
overexpression or overactivation of PDGFRs, or
by stimulation of angiogenesis within the tumor,
these mechanisms could provide possible therapeutic targets. PDGFR blockade may also lower
the interstitial fluid pressure within solid tumors
and enhance drug delivery. Indeed, extensive
experimental data highlight the potential therapeutic advantage of targeting PDGFR. However,
recent clinical data suggest that antagonism of
25
this growth factor is associated with fluid

accumulation that could obscure any clinical
benefit. The combination of PDGFR inhibitors
with chemotherapy may ultimately provide better
results and prove to be effective, but these trials
are still ongoing and valid conclusions cannot be
drawn at this time.
In experimental model systems, such as those
using murine and human glioblastoma cells,
specific PDGFR inhibitors that block PDGF
autocrine-mediated phosphorylation of PDGFR,
Akt, and MAPK signal transduction have been
shown to inhibit tumor growth and increase survival. For example, Lokker et al. (2002) demonstrated the functional significance of PDGF
autocrine signaling in C6 GBM cells by the fact
that a specific PDGFR inhibitor was able to block
soft agar colony formation of C6 cells, and when
the PDGFR inhibitor was given orally to nude
mice bearing subcutaneously injected C6 GBM
cells, it effectively reduced tumor formation by
44%. Furthermore, treatment with a specific
PDGFR antagonist inhibited survival and/or
mitogenic pathways in each of the GBM cell
lines tested. Kilic et al. (2000) was first to show
that oral administration of the PDGFR inhibitor
imatinib mesylate blocked the growth of U343
and U87 human GBM cells injected into the
brains of nude mice. Together, these findings
demonstrate the potential therapeutic utility of
PDGFR inhibitors for the treatment of GBM.
The administration of imatinib mesylate for
refractory pediatric malignant gliomas was first
reported by Pollack et al. (2007) in a phase I
cooperative group clinical trial. A phase II dose
was established by this study; however there are
no active trials investigating the efficacy of this
agent in pediatric gliomas at the present time.
Pilocytic astrocytomas, which are the most common type of childhood glioma, are typically slow
growing noninvasive tumors. Most children with
pilocytic astrocytomas survive many years with
their tumor, but some are refractory to treatment
and progress rapidly with metastasis. Because of
the reports linking signaling by the PDGFR pathways to the development of glioma progression
in adults, McLaughlin et al. (2003) reported the
treatment of a single patient with a refractory,
265
metastatic pilocytic astrocytoma with imatinib

mesylate, and interestingly observed marked,
transient regression of the tumor during treatment. Immunohistochemistry was used to assess
expression of PDGFR in this patients tumor tissue and in pilocytic astrocytomas from 19 other
patients, and it was found that PDGFR expression was detected in the tumor vasculature in
each of the panel of 20 tumors, and not in the
tumor cells, while none of the other targets of
imatinib (c-Kit and bcr-abl) were detectable. The
authors of this report concluded that the PDGFRsignaling pathway that is postulated to contribute
to the development of adult gliomas may also be
active in promoting angiogenesis in pilocytic
astrocytomas of children and that PDGFR inhibitors may be useful in treating refractory tumors.
References
Andrae J, Gallini R, Betsholtz C (2008) Role of plateletderived growth factors in physiology and medicine.
Genes Dev 22:12761312
Bax DA, Mackay A, Little SE, Carvalho D, Viana-Pereira
M, Tamber N, Grigoriadis AE, Ashworth A, Reis RM,
Ellison DW, Al-Sarraj S, Hargrave D, Jones C (2010)
A distinct spectrum of copy number aberrations in
pediatric high-grade gliomas. Clin Cancer Res
16:33683377
Claesson-Welsh L (1996) Mechanism of action of
platelet-derived growth factor. Int J Biochem Cell
Biol 28:373385
Doolittle RF, Hunkapiller MW, Hood LE, Devare SG,
Robbins KC, Aaronson SA, Antoniades HN (1983)
Simian sarcoma virus onc gene, v-sis, is derived from
the gene (or genes) encoding a platelet-derived growth
factor. Science 221:275277
Fleming TP, Saxena A, Clark WC, Robertson JT, Oldfield
EH, Aaronson SA, Ali IU (1992) Amplification and/or
overexpression of platelet-derived growth factor receptors and epidermal growth factor receptor in human
glial tumors. Cancer Res 52:45504553
Fredriksson L, Li H, Eriksson U (2004) The PDGF family: four gene products form five dimeric isoforms.
Cytokine Growth Factor Rev 15:197204
Garrett JS, Coughlin SR, Niman HL, Tremble PM, Giels
GM, Williams LT (1984) Blockade of autocrine stimulation in simian sarcoma virus-transformed cells reverses
down-regulation of platelet-derived growth factor receptors. Proc Natl Acad Sci USA 81:74667470
Heldin CH, Ostman A, Rnnstrand L (1998) Signal transduction via platelet-derived growth factor receptors.
Biochim Biophys Acta 1378:79113
266
Hermanson M, Funa K, Hartman M, Claesson-Welsh L,
Heldin CH, Westermark B, Nistr M (1992) Plateletderived growth factor and its receptors in human
glioma tissue: expression of messenger RNA and protein suggests the presence of autocrine and paracrine
loops. Cancer Res 52:32133219
Hermanson M, Funa K, Koopmann J, Maintz D, Waha A,
Westermark B, Heldin CH, Wiestler OD, Louis DN,
von Deimling A, Nistr M (1996) Association of loss
of heterozygosity on chromosome 17p with high
platelet-derived growth factor alpha receptor expression in human malignant gliomas. Cancer Res
56:164171
Hu JG, Fu SL, Wang YX, Li Y, Jiang XY, Wang XF, Qiu
MS, Lu PH, Xu XM (2008) Platelet-derived growth
factor-AA mediates oligodendrocyte lineage differentiation through activation of extracellular signal-regulated kinase signaling pathway. Neuroscience
151:138147
Huang JS, Huang SS, Deuel TF (1984) Transforming
protein of simian sarcoma virus stimulates autocrine
growth of SSV-transformed cells through PDGF cellsurface receptors. Cell 39:7987
Ishii Y, Matsumoto Y, Watanabe R, Elmi M, Fujimori T,
Nissen J, Cao Y, Nabeshima Y, Sasahara M, Funa K
(2008) Characterization of neuroprogenitor cells
expressing the PDGF beta-receptor within the subventricular zone of postnatal mice. Mol Cell Neurosci
37:507518
Jackson EL, Garcia-Verdugo JM, Gil-Perotin S, Roy M,
Quinones-Hinojosa A, VandenBerg S, Alvarez-Buylla
A (2006) PDGFR alpha-positive B cells are neural
stem cells in the adult SVZ that form glioma-like
growths in response to increased PDGF signaling.
Neuron 51:187199
Kilic T, Alberta JA, Zdunek PR, Acar M, Iannarelli P,
OReilly T, Buchdunger E, Black PM, Stiles CD
(2000) Intracranial inhibition of platelet-derived
growth factor-mediated glioblastoma cell growth by
an orally active kinase inhibitor of the 2-phenylaminopyrimidine class. Cancer Res 60:51435150
Lafuente JV, Adn B, Alkiza K, Garibi JM, Rossi M,
Cruz-Snchez FF (1999) Expression of vascular
endothelial growth factor (VEGF) and platelet derived
growth factor receptor-beta (PDGFR-beta) in human
gliomas. J Mol Neurosci 13:177185
Liang ML, Ma J, Ho M, Solomon L, Bouffet E, Rutka JT,
Hawkins C (2008) Tyrosine kinase expression in
pediatric high grade astrocytoma. J Neurooncol
87:247253
Lokker NA, Sullivan CM, Hollenbach SJ, Israel MA,
Giese NA (2002) Platelet-derived growth factor
(PDGF) autocrine signaling regulates survival and
mitogenic pathways in glioblastoma cells: evidence
that the novel PDGF-C and PDGF-D ligands may play
a role in the development of brain tumors. Cancer Res
62:37293735
McLaughlin ME, Robson CD, Kieran MW, Jacks T,
Pomeroy SL, Cameron S (2003) Marked regression of
T.J. MacDonald
metastatic pilocytic astrocytoma during treatment with
imatinib mesylate (STI-571, Gleevec): a case report
and laboratory investigation. J Pediatr Hematol Oncol
25:644648
Nakamura M, Shimada K, Ishida E, Higuchi T, Nakase H,
Sakaki T, Konishi N (2007) Molecular pathogenesis of
pediatric astrocytic tumors. Neuro-oncology 9:113123
Paugh BS, Qu C, Jones C, Liu Z, Adamowicz-Brice M,
Zhang J, Bax DA, Coyle B, Barrow J, Hargrave D,
Lowe J, Gajjar A, Zhao W, Broniscer A, Ellison DW,
Grundy RG, Baker SJ (2010) Integrated molecular
genetic profiling of pediatric high-grade gliomas
reveals key differences with the adult disease. J Clin
Oncol 28:30613068
Plate KH, Breier G, Farrell CL, Risau W (1992) Plateletderived growth factor receptor-beta is induced during
tumor development and upregulated during tumor
progression in endothelial cells in human gliomas.
Lab Invest 67:529534
Pollack IF, Jakacki RI, Blaney SM, Hancock ML, Kieran
MW, Phillips P, Kun LE, Friedman H, Packer R,
Banerjee A, Geyer JR, Goldman S, Poussaint TY,
Krasin MJ, Wang Y, Hayes M, Murgo A, Weiner S,
Boyett JM (2007) Phase I trial of imatinib in children
with newly diagnosed brainstem and recurrent malignant gliomas: a pediatric brain tumor consortium
report. Neuro-oncology 9:145160
Potapova O, Fakhrai H, Baird S, Mercola D (1996)
Platelet-derived growth factor-B/v-sis confers a tumorigenic and metastatic phenotype to human T98G
glioblastoma cells. Cancer Res 56:280286
Rubin K, Tingstrm A, Hansson GK, Larsson E,
Rnnstrand L, Klareskog L, Claesson-Welsh L, Heldin
CH, Fellstrm B, Terracio L (1988) Induction of
B-type receptors for platelet-derived growth factor in
vascular inflammation: possible implications for
development of vascular proliferative lesions. Lancet
1:13531356
Shih AH, Dai C, Hu X, Rosenblum MK, Koutcher JA,
Holland EC (2004) Dose-dependent effects of plateletderived growth factor-B on glial tumorigenesis. Cancer
Res 64:47834789
Smits A, Funa K (1998) Platelet-derived growth factor
(PDGF) in primary brain tumours of neuroglial origin.
Histol Histopathol 13:511520
Thorarinsdottir HK, Santi M, McCarter R, Rushing EJ,
Cornelison R, Jales A, MacDonald TJ (2008) Protein
expression of platelet-derived growth factor receptor
correlates with malignant histology and PTEN with
survival in childhood gliomas. Clin Cancer Res
14:33863394
Toepoel M, Joosten PH, Knobbe CB, Afink GB, Zotz RB,
Steegers-Theunissen RP, Reifenberger G, van Zoelen
EJ (2008) Haplotype-specific expression of the human
PDGFRA gene correlates with the risk of glioblastomas.
Int J Cancer 123:322329
Vantler M, Huntgeburth M, Caglayan E, Ten Freyhaus H,
Schnabel P, Rosenkranz S (2006) PI3-kinase/Aktdependent antiapoptotic signaling by the PDGF alpha
25
receptor is negatively regulated by Src family kinases.

FEBS Lett 580:67696776
Wang D, Huang HJ, Kazlauskas A, Cavenee WK (1999)
Induction of vascular endothelial growth factor expression in endothelial cells by platelet-derived growth
factor through the activation of phosphatidylinositol
3-kinase. Cancer Res 59:14641472
Wong KK, Tsang YT, Chang YM, Su J, Di Francesco
AM, Meco D, Riccardi R, Perlaky L, Dauser RC,
Adesina A, Bhattacharjee M, Chintagumpala M, Lau
267
CC (2006) Genome-wide allelic imbalance analysis of

pediatric gliomas by single nucleotide polymorphic
allele array. Cancer Res 66:1117211178
Zarghooni M, Bartels U, Lee E, Buczkowicz P, Morrison
A, Huang A, Bouffet E, Hawkins C (2010) Wholegenome 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
28:13371344
An Overview of Pediatric High-Grade

Gliomas and Diffuse Intrinsic
Pontine Gliomas
26
Rishi R. Lulla and Jason Fangusaro
Contents
Abstract
Introduction ............................................................
269
Pediatric High-Grade Gliomas .............................

Histologic and Molecular Classification
of High-Grade Gliomas..............................................
Risk Factors, Clinical Presentation
and Diagnosis ...........................................................
Treatment of High-Grade Gliomas ..........................
Prognostic Factors and Outcomes ............................
Differences Between Pediatric
and Adult High-Grade Gliomas ...............................
270
270
271
273
276
277
Diffuse Intrinsic Pontine Gliomas ........................

Clinical Presentation and Diagnosis ........................
Molecular Characterization of Diffuse
Intrinsic Pontine Gliomas ........................................
A Review of Treatment Approaches ........................
Prognostic Factors and Outcomes ............................
Future Strategies in Treatment .................................
277
277
279
279
280
280
Discussion................................................................
281
References ...............................................................
282
High-grade gliomas represent approximately

10% of all pediatric brain tumors. Diffuse
intrinsic pontine gliomas are a distinct clinical
entity in pediatrics and accounts for another
10% of central nervous system tumors in children. Pediatric patients with high-grade
gliomas or diffuse intrinsic pontine gliomas
have a very poor prognosis despite a variety of
aggressive therapies, including surgery, chemotherapy and radiation therapy. Though
these tumors behave similarly to adult highgrade gliomas, there are important differences
in the molecular biology of high-grade gliomas
between children and adults. In this chapter, we
present an overview of both pediatric high-grade
gliomas as well as diffuse intrinsic pontine
gliomas. We focus upon their epidemiology,
etiology, clinical presentation and biologic
features. In addition, we discuss treatment
approaches, prognostic factors, outcomes and
future research directions.
Introduction
R.R. Lulla (*) J. Fangusaro

Hematology/Oncology/Stem Cell Transplantation,
Childrens Memorial Hospital,
2300 Childrens Plaza, Box 30,
Chicago, IL 60614, USA
e-mail: r-lulla@northwestern.edu
Although the prognosis for children with primary

central nervous system (CNS) tumors has greatly
improved over the last few decades with 5-year
survival rates approaching 75%, primary malignant brain tumors represent the leading cause of
death from solid tumors in children (Warren
2008). High-grade gliomas represent one of the most

269
270
common CNS tumors in adults (CBTRUS 2005).

In pediatrics, however, they represent approximately 10% of all tumors, are characterized by
their aggressive behavior and account for a significant source of morbidity and mortality.
Children with high-grade gliomas present with
signs and symptoms common to other CNS
tumors including headache, nausea, vomiting,
abnormal gait or coordination and papilledema.
Treatment usually involves focal radiation therapy after a maximal tumor resection. Most
patients are also treated with adjuvant chemotherapy; however, the data on its utility remain
inconclusive. Overall survival for patients with
high-grade gliomas remains poor, even after multimodal therapy.
Diffuse intrinsic pontine glioma (DIPG) is a
distinct clinical entity in pediatrics characterized
by diffuse involvement of the pons, typically by
high-grade tumor cells. Diagnosis is usually confirmed by imaging studies alone as the tumors
have a characteristic infiltrative nature seen on
magnetic resonance imaging (MRI). Surgical
biopsies are rarely done and little is known about
the molecular characteristics of these tumors.
Despite several decades of research, the outcomes
for patients with DIPG are dismal. Attempts to
intensify radiation therapy and give adjuvant chemotherapy have not resulted in any significant
improvement in survival.
Future treatment for high-grade glioma and
DIPG is dependent on investigators obtaining a
better understanding of tumor biology and pathogenesis. Currently, several new agents with alternative mechanisms of action and novel methods
of delivering therapy are under investigation with
the goal of increasing survival and minimizing
treatment toxicity.
Pediatric High-Grade Gliomas

Histologic and Molecular Classication
of High-Grade Gliomas
The World Health Organization (WHO) classification is the most widely accepted system for
classifying and grading CNS tumors and is based
R.R. Lulla and J. Fangusaro
upon an assumption that each tumor arises from

the abnormal proliferation of a specific cell type
(Louis et al. 2007). Most pediatric high-grade
gliomas are of astrocytic origin, though gliomas
can arise from other glial cells including oligodendrocytes and ependymal cells (CBTRUS
2005). Gliomas are graded from I to IV based
upon their histologic features and malignant
potential. Grade I gliomas have a low proliferative potential and are considered benign; they are
often cured by surgery alone. Grade II gliomas
are infiltrative with cellular atypia and frequently
recur locally. In adults, grade II gliomas may
transform into higher grade tumors over time; this
phenomenon is not usually observed in children.
Grade III gliomas, such as anaplastic astrocytoma,
are characterized by hypercellularity, nuclear
atypia and brisk mitotic activity. Finally, WHO
grade IV gliomas, such as glioblastoma multiforme, demonstrate some or all of the features of
grade III tumors but also have microvascular proliferation and/or evidence of pseudopalisading
necrosis (Louis et al. 2007). In pediatric gliomas,
tumor grade is significantly correlated with outcome
(Finlay et al. 1995).
Recent data on the molecular biology and
genetics of pediatric high-grade gliomas have
contributed to a better understanding of these
tumors. The mechanisms of tumorigenesis are
different than those previously elucidated for
adult high-grade gliomas (Nakamura et al. 2007).
One hallmark of adult high-grade gliomas is
amplification and over-expression of epidermal
growth factor receptor (EGFR). Though some
pediatric high-grade gliomas do have over expression of EGFR without gene amplification (Bredel
et al. 1999), mutations in the p53 pathways are a
relatively more common molecular abnormality.
There are several unique defects that potentially
lead to alteration of this pathway including
mutation of the TP53 tumor suppressor gene or
over-expression of the p53 protein. In one study,
over-expression of the p53 protein, regardless of
the presence of a TP53 mutation, was associated
with a poorer outcome in pediatric high-grade
gliomas (Rood and MacDonald 2005).
Additionally, pediatric high-grade gliomas frequently have chromosomal abnormalities including
26
An Overview of Pediatric High-Grade Gliomas and Diffuse Intrinsic Pontine Gliomas
gains at 1p, 2q, and 21q as well as losses at 6q,

11q, and 16q (Rickert et al. 2001). The biologic
significance of these alterations and others that
have been described remains unclear as no effective targeted therapies have been developed for
pediatric high-grade gliomas.
Risk Factors, Clinical Presentation

and Diagnosis
The etiology of pediatric high-grade gliomas
remains largely unknown as most arise sporadically; however, there are a few conditions that
increase a patients risk of developing these
tumors. Most are inherited conditions associated
with defects in regulation of cell proliferation and
apoptosis caused by germline mutations (Melean
et al. 2004). Neurofibromatosis type I (NF-1) is
the most common inherited genetic disorder predisposing patients to CNS tumors. Mutations in
the NF-1 gene result in absence of neurofibromin
which normally functions as a regulator of both
the Ras proto-oncogene as well as cell growth.
Absence of the protein predisposes the cell to
uncontrolled growth and tumor formation. (Ward
and Gutmann 2005). Patients with NF-1 are at
risk for developing both benign and malignant
tumors in the nervous system. Within the CNS,
low-grade optic pathway gliomas are the most
common tumors, however, patients with NF-1 are
also at increased risk for developing high-grade
gliomas. A recent case series reported 4 cases of
high-grade glioma in a group of 145 patients with
NF-1 and CNS tumors. Some of these patients
had low grade optic pathway tumors prior to the
development of high-grade gliomas in other locations (Rosenfeld et al. 2010).
Patients affected with the Li-Fraumeni syndrome have a defect in the TP53 gene which in
turn prevents normal apoptosis and cell cycle
regulation. This mutation results in unregulated
cell proliferation and an increased risk of malignancy. The tumors most closely associated with
Li-Fraumeni syndrome are: soft tissue sarcoma,
osteosarcoma, pre-menopausal breast cancer and
brain tumors including high-grade gliomas. Other
rare conditions predisposing children to CNS
271
tumors include Turcot syndrome (mutations of

APC and the DNA mismatch repair genes
hMSH2, hMLH1 and hPMS2), tuberous sclerosis
complex (mutations of either TSC1 or TSC2) and
von Hippel-Lindau disease (mutations of VHL).
Additionally, patients with germline PTCH
mutations (basal cell nevus syndrome) are also at
increased risk of developing high-grade gliomas,
although they are most commonly affected by
medulloblastoma which is the most common
malignant brain tumor in pediatrics (Melean
et al. 2004).
Another risk factor for the development of
high-grade gliomas is exposure to previous radiation therapy. Despite a relatively low risk of secondary malignancy in all children exposed to
cranial irradiation (estimated to be between 1%
and 3% for all survivors), development of a secondary high-grade glioma has been reported in
this population. Cranial irradiation is most often
used in the treatment of childhood cancers such
as acute lymphoblastic leukemia, lymphomas
and CNS tumors (Neglia et al. 2006). A long
term-follow up report from St. Jude Childrens
Research Hospital identified a dose-dependent
tumorigenic effect from previous radiation therapy on the risk of a developing a secondary
malignancy. Further, it suggested that patients
who were treated for acute lymphoblastic leukemia at a younger age (less than 6 years old) were
at an increased risk of developing a secondary
high-grade glioma (Walter et al. 1998). A subsequent report from the Childhood Cancer Survivor
Study confirmed these findings in a larger cohort.
Additionally, when both the original cancer diagnosis and the chemotherapeutic regimen utilized
were taken into consideration, radiation exposure
was the single most important risk factor for the
development of a secondary CNS tumor. They
also reported an average latency period of 9 years
from the time of original diagnosis to development of high-grade glioma (Neglia et al. 2006).
This prolonged delay in tumor development
emphasizes the need for long term follow up and
prompt evaluation of new neurologic symptoms
in childhood cancer survivors.
Epidemiological reports have suggested that
some environmental toxins (such as aromatic
272
hydrocarbons and vinyl chloride) are associated

with an increased risk for the development of
high-grade gliomas (Warren 2008). Most recently,
similar reports have suggested an increase in the
development of gliomas associated with prolonged cellular phone usage. In addition to questionable biologic plausibility, these studies are
subject to all the pitfalls common to epidemiologic research. Neither environmental toxins nor
radiofrequency exposure from cellular phones
have been conclusively implicated in the development of high-grade gliomas in children (Bondy
et al. 2008).
Children with high-grade gliomas have clinical presentations similar to other patients with
CNS tumors. Many of these signs and symptoms
are non-specific and include persistent headache,
emesis or behavior changes that may otherwise
be attributed to common viral syndromes of
childhood. The diagnosis of a childhood brain
tumor was historically associated with long
latency periods between symptom onset and
diagnosis. However, neuroimaging techniques
have shortened this interval in recent years. In
contrast to lower grade brain tumors which often
have an average 4 month latency period between
symptoms and diagnosis, children with highgrade gliomas have a shorter duration of symptoms prior to diagnosis. This may be attributed to
the higher mitotic activity and faster growth rate
of higher grade tumors resulting in more rapid
involvement of the adjacent normal brain.
Presenting signs and symptoms vary by age,
tumor location and the presence of increased
intracranial pressure (ICP). High-grade gliomas
can occur anywhere within the CNS. In children
they are approximately equally distributed
between supratentorial and infratentorial locations. Patients with supratentorial tumors may
present with common signs of increased ICP
including headache, nausea, early morning emesis, diplopia and papilledema. More subtle findings of increased ICP include behavioral and
personality changes (Warren 2008). Supratentorial
tumors may also present with focal motor deficits, hemiplegia, dysmetria, pyramidal tract findings or chorea. Tumors located in the cerebellum
may also present with increased ICP without other
obvious focal findings, reflecting hydrocephalus

from tumor obstruction. Seizures may be the presenting symptom of a high-grade glioma depending upon the tumor location, but they are more
frequently associated with low grade gliomas,
such as gangliogliomas and dysembryoplastic
neuroepithelial tumors. Special attention should
be given to infants who present with non-specific
signs and symptoms which may include macrocephaly, lethargy, persistent vomiting, irritability
or failure to thrive (Fangusaro 2009).
Once the signs and symptoms of a brain tumor
have been identified, diagnostic imaging is usually performed. Computerized tomography (CT)
of the head is typically done first and may demonstrate a mass lesion, enlarged ventricles and/or
hemorrhage into the tumor or adjacent brain
parenychyma. As with other brain tumors, MRI
is the test of choice for high-grade gliomas. In
additional to providing well-defined tumor characteristics including location and involvement of
adjacent normal brain tissue, MRI is critical for
surgical and radiation therapy planning. There is
no specific MRI finding that can distinguish a
supratentorial high-grade glioma from other
malignant lesions; however, high-grade gliomas
often have irregular borders, nodular enhancement, mass effect and invasion into adjacent tissues. The lesions are classically hypointense on
T1- and hyperintense on T2-weighted and fluid
attenuation inversion recover (FLAIR) sequences
(Fig. 26.1). Tumor enhancement after contrast
administration is commonly seen in supratentorial high-grade gliomas but the degree of enhancement does not always correlate with tumor grade
(Warren 2008).
As neuroimaging technology continues to
improve, newer modalities may contribute to the
characterization of high-grade gliomas on MRI.
Perfusion weighted imaging provides a non-invasive
way to determine tumor angiogenesis and capillary
permeability; high-grade gliomas often demonstrate increased blood flow to lesions using this
modality. Magnetic resonance spectroscopy
(MRS) uses imaging to metabolically profile a
region of interest in the brain. In gliomas, as
tumor grade increases, there is also an increase in
the choline: N-acetyl aspartate (NAA) ratio on
26
273
Fig. 26.1 A T1-weighted MRI image (a) of a glioblastoma

multiforme demonstrates a hypointense lesion in the left
frontal and parietal lobes. Blood products are present in
the interventricular portion of the mass. On a T2/FLAIR
image from the same patient (b), there is extensive confluent

abnormal hyperintensity involving the white matter and
extending into the corpus callosum and across the
midline
MRS reflecting greater metabolic activity of

higher grade tumors. (Lemort et al. 2007). Most
recently, positive emission tomography (PET)
has been evaluated in patients with brain
tumors. High-grade gliomas have variable
uptake of the currently available tracer elements and in some cases uptake may be similar
to or slightly above the surrounding white and
grey matter, making PET images difficulty to
interpret. Newer tracers and co-registration of
PET images with MRI are under investigation
to improve the utility of this modality for both
diagnosis and response to therapy (Chen 2007).
None of the previously described imaging techniques can make a conclusive diagnosis in most
high-grade gliomas, and therefore, a surgical
resection is recommended whenever possible to
establish both a pathologic diagnosis and as an
initial therapeutic intervention.
intervention may be required immediately,

including surgery for obstructive hydrocephalus
and placement of either an external ventricular
drain or a ventriculo-peritoneal shunt. In some
cases, pressure may be relieved by a third ventirculostomy. Steroids have been a mainstay of
treatment for cerebral edema associated with
brain tumors since the early 1960s, though no
randomized controlled trials have been done to
determine their true benefit. As compared to other
corticosteroids, dexamethasone has a potent antiinflammatory effect, a relatively low mineralocorticoid effect and excellent penetration into the
CNS which makes it the preferred choice in this
setting. The recommended dosing for children
varies greatly based upon practitioner, but it typically starts between 0.5 and 1 mg per kilogram
per day divided into four doses. The dose is
subsequently weaned as the patient stabilizes.
Considering the significant side effects of steroids
(including weight gain, hypertension, elevated
blood glucose and behavior changes) and their
ability to interfere with chemotherapeutic agents
through hepatic enzyme up-regulation, every
attempt should be made to use the lowest effective
dose whenever possible (Warren 2008).
Treatment of High-Grade Gliomas

The first step in treatment of a patient with
high-grade glioma is stabilization, particularly
for patients with increased ICP. Neurosurgical
274
For the group of patients who develop seizures

at the time of their diagnosis, anticonvulsants
should be administered. Considering its favorable side effect profile and minimal interaction
with other medications, levetiracetam is often the
first choice in anticonvulsants for children with
brain tumors. However, the choice of anti-epileptic
should be made in conjunction with a pediatric
neurologist or neurosurgeon. There has been
significant controversy surrounding the use of
prophylactic anticonvulsants in patients with
newly diagnosed brain tumors. In 2000, the
American Academy of Neurology reviewed this
practice and issued a consensus statement. Their
findings suggest that there is no benefit to the
prophylactic use of anticonvulsants to prevent a
first seizure in patients with brain tumors and the
Academy does not recommend their routine use
in patients without a previous history of seizure.
Further, for patients who have been prescribed
anticonvulsants in the perioperative period, tapering
and discontinuation of the medication at the end
of the first post-operative week is recommended
provided there are no clinical indications of
seizure activity (Glantz et al. 2000).
After immediate stabilization of the patient,
surgical resection of the tumor should be performed as soon as possible. Surgery is valuable in
establishing a diagnosis, relieving elevated ICP
and decreasing the active tumor burden. Surgical
resection is often complicated given the fact that
high-grade gliomas do not have well defined
tumor borders and may diffusely infiltrate the
adjacent normal brain. For children with supratentorial high-grade gliomas, a gross total resection
should be planned if feasible. The degree of surgical resection is one of the most important prognostic factors for survival, independent of patient
age, tumor location and histology. The Childrens
Cancer Group study CCG-945 reported that children with high-grade gliomas who underwent a
tumor resection of greater than 90% had a 5-year
progression-free survival (PFS) of 35% 7% as
compared to only 17% 4% for those patients
who had a less than 90% tumor resection (Finlay
et al. 1995). Therefore, every attempt should be
made to remove as much tumor as possible without causing significant neurologic morbidity.
This can be particular challenging for lesions that

involve critical areas of the brain, deep midline
structures and many infratentorial tumors.
High-grade gliomas are infiltrative lesions
with irregular borders and microscopic tumor
invasion, and it has become clear that radiation
therapy is essential to their treatment. Even in the
setting of a complete resection, microscopic
tumor cells are present at the resection margins
and require treatment in an attempt to prevent
local recurrence. Radiation therapy is the mainstay of treatment for high-grade gliomas, particularly for children older than 3 years at the time of
their diagnosis. Those under 3 years old are more
susceptible to the deleterious late effects of radiation such as neurocognitive delay, ototoxicity,
endocrine dysfunction and secondary malignancy. Many groups elect to treat these patients
with chemotherapy up-front in an attempt to
delay or completely avoid radiation (Duffner
et al. 1993). Patients under 3 years of age have an
improved survival when compared to older
patients (Sanders et al. 2007). This suggests that
the tumors have unique molecular and/or biologic
characteristics. For older patients with high-grade
gliomas, the standard approach to external beam
radiation therapy involves a focal field including
the tumor resection cavity and a margin of surrounding normal appearing tissue. There is no
role for whole brain radiotherapy in the up-front
treatment of high-grade glioma (Buckner et al.
2007). Conventional dosing for high-grade
glioma is 5060 Gy delivered in approximately
180200 cGy fractions over a period of 6 weeks.
Previous attempts to intensify radiation therapy
with doses as high as 72 Gy have not conferred a
significant survival advantage (Warren 2008). As
would be expected, higher doses of radiation subject the child to significantly more side effects.
Chemotherapy was introduced into the treatment
for high-grade gliomas in 1976, however its role is
still disputed. There may be a subset of patients who
benefit from chemotherapy (for example, young
children in whom radiation therapy can be delayed),
but in others, many experts debate the advantage
in long term survival when compared to radiation
therapy alone. In an initial study sponsored by
the Childrens Cancer Group (CCG-943),
26
patients were randomly assigned to receive 54 Gy

of focal radiation versus the same radiotherapy
dose in combination with chemotherapy. The
chemotherapy regimen included weekly vincristine during radiation followed by maintenance
cycles of vincristine, lomustine and prednisone
(Sposto et al. 1989). The results of this study suggested a significant survival advantage for those
patients treated with both radiation and chemotherapy. The 5 year event-free survival (EFS) in
the combination arm was 46% versus 18% in the
group that received radiation therapy alone. The
data from this initial study have never been replicated. A central review of pathology done several
years after the study was completed revealed that
some of these patients actually had low-grade
gliomas. However, removal of those patients and
re-analysis of the data still suggested a statistically significant benefit from the addition of chemotherapy. As a result of this study, the currently
accepted standard-of-care for treating high-grade
gliomas has been a combination of radiation therapy and chemotherapy. Several subsequent combination studies have been done and have failed
to replicate these initial promising findings
(Fangusaro 2009).
One such study CCG-945 (the immediate successor to CCG-943), assigned all patients above
24 months to standard radiation therapy followed
by randomization to one of two chemotherapy
strategies. The conventional arm was vincristine,
prednisone and lomustine as given in the previous study and the experimental arm was 8 chemotherapeutic agents given in 1 day (the so-called
8-in-1). The 8-in-1 chemotherapy included the
three agents in the standard arm along with
hydroxyurea, cisplatin, cytarabine, dacarbazine
and procarbazine. Of note, the patients assigned
to the experimental arm received two cycles of
pre-radiation chemotherapy. Younger infants
under 24 months of age were non-randomly
assigned to 8-in-1 chemotherapy alone in an
effort to avoid or delay radiation. The results of
the study did not find any statistical difference in
the PFS for the two arms (26% 8% in the control group versus 33% 8% in the experimental
group). Similarly, there was no difference in
overall survival (OS) (Finlay et al. 1995).
275
One of the most recent cooperative group trials

(Childrens Oncology Group study ACNS0126)
replaced the conventional chemotherapy of CCG943 with temozolomide, extrapolating from
promising adult data. Temozolomide and radiation therapy is currently considered the standardof-care in adult high-grade glioma after a trial by
Stupp et al. (2005) demonstrated superiority of
this combination when compared to radiation
alone (Stupp et al. 2005). Unfortunately, most
trials done in pediatrics using temozolomide
in combination with radiation have been Phase I
and Phase II studies, and they have failed to
prove superiority of the combination versus outcomes for historical controls (Broniscer et al. 2006).
One of the barriers to chemotherapy trials in
pediatric high-grade gliomas is the fact that
researchers are reluctant to randomize patients
to receive radiation alone, especially given the
promising results first reported in CCG-943
study. At this point, it is not clear if temozolomide and radiation together are superior to radiotherapy alone. Another approach taken by some
investigators has been high-dose chemotherapy
with autologous stem cell rescue in the treatment
of high-grade gliomas. This treatment has been
studied by the Childrens Oncology Group and
remains quite controversial, but there is preliminary data that suggest that a subgroup of patients
may benefit from this approach (Finlay et al. 2008).
There are several other barriers to investigating treatment regimens in children with highgrade gliomas. There are a limited number of
agents and delivery of drugs into the CNS is often
difficult without significant toxicity. Further, as
mentioned previously, these tumors are biologically heterogeneous which limits the utility of
targeted therapies. However, there are several
novel approaches currently under investigation.
Monoclonal antibodies such as bevacizumab
(directed against vascular endothelial growth
factor) have been combined with conventional
chemotherapy agents (irinotecan) and have shown
some promising results in adults with recurrent/
progressive disease. These combinations are
currently in clinical trials for children with refractory or recurrent disease, however, response rates
have been variable (Gururangan et al. 2008).
276
Several other small molecule inhibitors are

currently under investigation and include receptor
tyrosine kinases, histone deacetylase inhibitors
and integrins. Data on the utility of these agents
are preliminary, but findings suggest there may
be a specific role in the setting of minimal residual
disease after conventional treatment (Herrington
and Kieran 2009). Another approach, convection
enhanced delivery (CED), utilizes surgical techniques to implant a catheter and deliver agents
(chemotherapy, cytotoxic proteins and others)
locally into the tumor or resection cavity. This is
a promising new technique, and is currently being
evaluated in Phase I trials. Also in early clinical
trials is the use of vaccines directed against
known glioma antigens in children with gliomas
(www.clinicaltrials.gov NCT00862199).
Prognostic Factors and Outcomes

Studies of high-grade glioma treatment in children have reported variable outcomes. Despite
aggressive multimodal therapy, the median PFS
for patients with high-grade gliomas is approximately 10 months with a median OS of 18 months.
Five-year OS ranges between 10% and 20% in
various studies (Warren 2008). There are some
important determinants of outcomes which warrant discussion. As previously mentioned, the
extent of tumor resection positively correlates
with outcomes as shown in the CCG-945 study.
Patients with greater than 90% tumor resection
have the best overall survival. Additionally, histologic grade can also predict outcome; patients
with WHO grade III tumors (anaplastic astrocytoma) have superior outcomes as compared to
those patients with WHO grade IV tumors (Finlay
et al. 1995). The CCG-945 study also investigated cytogenetic and molecular markers influencing outcomes. Patients who had p53
over-expression or mutations in the p53 gene had
a significantly lower PFS when compared to
those in whom no mutation or over-expression
was identified. Abnormal p53 was more common
in the WHO grade IV tumors, however p53 overexpression was shown to be an independent predictor of outcome for patients with any histologic
grade tumor (Pollack et al. 2002). Analysis of

methylguanine methlytransferase (MGMT) status
in this same cohort demonstrated a significantly
worse outcome for those patients with overexpression of MGMT, independent of other
prognostic factors (Pollack et al. 2006).
There is a subset of children with high-grade
glioma (those less than 3 years of age) who have
improved outcomes when compared to older
children. In a retrospective review done at St. Jude
Childrens Research Hospital, the 5-year EFS
and OS were 28.6% and 66.3%, respectively for
children under the age of three treated for highgrade gliomas (Sanders et al. 2007). These
outcomes are significantly better than those previously reported for older children. Many of the
patients included in this group experienced disease recurrences which required surgery and the
eventual use of radiation therapy. Additionally,
most of the surviving patients had severe longterm sequelae of treatment (Sanders et al. 2007).
It remains unclear, however, whether age is an
independent prognostic factor in this group.
High-grade gliomas in very young children may
have unique molecular and biologic characteristics which improve treatment success and warrant
further investigation.
Many children with high-grade gliomas will
have recurrence of their disease after initial treatment; thus far, treatment of this group has been of
limited clinical benefit. Patients usually progress
rapidly and succumb to their disease. In the CCG945 study, median survival of patients who
experienced disease recurrence after surgery,
chemotherapy and radiation was only 9 months
despite receiving aggressive relapse therapy
(Wisoff et al. 1998). High-grade gliomas tend to
recur locally, presumably as a result of microscopic invasion of tumor cells into the surrounding tissue. However, these tumors may metastasize
through the subarachnoid space. Metastases outside of the central nervous system are extremely
rare. Treatment options for recurrent disease
include surgery, additional radiation (especially
in patients who were less than 3 years of age at
the time of their diagnosis) and experimental
therapeutics as described above. Almost all of the
aforementioned and on-going studies involving
26
small molecule inhibitors in pediatrics have been

performed in the setting of recurrent disease.
Differences Between Pediatric

and Adult High-Grade Gliomas
Despite similar histological appearances, there
are several notable differences in the epidemiology, molecular features and clinical behavior of
pediatric and adult high-grade gliomas. In adults,
high-grade gliomas represent nearly 80% of all
gliomas. Most gliomas in children are low-grade,
and high-grade gliomas represent only about
1520% of CNS tumors of childhood (CBTRUS
2005). Further, most adult high-grade gliomas
are supratentorial. In contrast, approximately
50% of tumors in children will occur below the
tentorium cerebelli. Both developmental and
molecular differences between adult and pediatric gliomas are important. In adults, tumors can
arise de novo (primary glioblastoma) or as a
result of transformation of a previous low grade
tumor. This transformation phenomenon has not
been well described in pediatric high-grade glioma.
With regard to molecular classification, most adult
tumors will have EGFR amplification or PTEN
mutation. EGFR is an important drug target and its
amplification in adults may be associated with
adverse outcomes, though its role as an independent prognostic factor is still debated (Bredel et al.
1999). In contrast, EGFR amplification is variable
in pediatric tumors. Pediatric high grade gliomas
represent a much more heterogeneous group of
tumors making targeted therapy more challenging
in children (Nakamura et al. 2007).
Finally, with respect to treatment, temozolomide and radiation therapy have become the standard-of-care in adult high-grade glioma as a
result of the study by Stupp et al. (2005) which
demonstrated significantly higher OS in the combination group. This data has not been replicated
in children and may be a result of biologic differences between the age groups. The cytotoxic
effect of temozolomide is negated by the MGMT
enzyme. Many adult tumors have epigenetic
silencing of the MGMT enzyme which might
explain superior outcomes with temozolomide.
277
In contrast, some pediatric high-grade gliomas

over-express MGMT, and this may render temozolomide less effective. (Pollack et al. 2006).
Another explanation for the difference in outcomes between adults and children is that the
adult study design has not yet been replicated in
children.
Diffuse Intrinsic Pontine Gliomas

Clinical Presentation and Diagnosis
Approximately 10% of all pediatric central nervous system tumors occur in the brainstem and a
majority are histologically considered gliomas
(CBTRUS 2005). Approximately 20% of the
lesions are characterized as low-grade astrocytomas and generally behave similarly to other
low grade tumors with regard to growth rates and
clinical course. The remaining 80% of gliomas in
the brainstem occur in the pons and represent a
distinct clinical entity in pediatric neuro-oncology. These lesions diffusely infiltrate the pons
and are not amenable to surgical resection; they
are commonly referred to as diffuse intrinsic pontine gliomas (DIPG). Despite collaborative efforts
over the past few decades to identify therapeutic
options for patients with DIPG, nearly all patients
succumb to their disease (Hargrave et al. 2006).
Patients with DIPG can present at any age
from infancy through adolescence, though most
commonly the lesions occur in mid-to-late childhood at a mean age of 79 years. There is no sex
predilection and patients present with relatively
acute onset with progressive symptoms, typically
for 12 months preceding the diagnosis.
Neurological examination findings for children
with newly diagnosed DIPG can include ataxia,
cranial nerve deficits or long tract signs including
clonus and muscle spasticity (Frazier et al. 2009).
These lesions are difficult to approach surgically
and the diagnosis is usually made by imaging
studies alone. The classic DIPG appearance on
MRI is characterized by diffuse infiltration and
hypertrophy of the pons with some mass effect
on adjacent structures such as the fourth ventricle
and basilar artery (Fangusaro 2009). On T1-weighted
278
Fig. 26.2 Characteristic imaging findings of diffuse

intrinsic pontine glioma include infiltration and hypertrophy of the pons without significant enhancement on
post gadolinium T1-weighted images (a). A T2/FLAIR
sequence (b) clearly demonstrates the hyperintense, irregular mass within the pons with mass-effect on the fourth
ventricle. Evidence of hydrocephalus is seen on both
images
images, the mass usually appears hypointense

with irregular margins which is a direct reflection
of the infiltrative nature of the tumor. On
T2-weighted and FLAIR sequences, DIPGs are
hyperintense which aids in distinguishing them
from focal tumors in the brainstem (Fig. 26.2).
Another important imaging characteristic is the
absence of significant enhancement after gadolinium administration which can sometimes help
distinguish these lesions from low grade tumors
in the brainstem (Frazier et al. 2009). The role of
FDG-PET imaging in DIPG is still under investigation; however, some reports suggest that this
modality can differentiate between low-grade
and high-grade brainstem gliomas. When correlated with MRI imaging, increased tracer uptake
may be indicative of a higher-grade lesion. FDGPET has also been utilized after treatment in an
attempt to differentiate tumor recurrence from
radiation necrosis (Chen 2007).
As imaging techniques have improved, many
different classification schemes for children with
brainstem gliomas have been developed. One of
the most recent classification systems was
described by Choux et al. (2000) and described
four unique types of brainstem gliomas. Type I

lesions are diffuse throughout the brainstem,
Type II lesions are focal and intrinsic to the pons,
Type III tumors are focal but characterized by an
exophytic appearance on MRI and finally Type
IV lesions are located at the cervicomedullary
junction (Choux et al. 2000). Childhood DIPG
which are the focus of this section, are characterized Type I lesions in the classification. The other
types are characterized by more focal lesions and
may have a more favorable outcome.
The role of surgical biopsies in patients with a
newly diagnosed DIPG has been controversial.
Historically, if a patient has the classic symptoms
in addition to the characteristic imaging findings
on MRI, no surgical biopsy is required to confirm
the diagnosis. If the patient presentation or imaging findings do not fit a characteristic pattern,
most would advocate for a diagnostic biopsy.
Though most biopsied brainstem lesions are
either high or low grade gliomas, ependymomas
and primitive neuroectodermal tumors have been
previously reported (Fangusaro 2009). Since relatively few biopsies have been performed in
patients with DIPG, there is a paucity of tumor
26
tissue available to identify biological and molecular

therapeutic targets. Researchers have had to rely
on tissues obtained at autopsy which are often
previously exposed to radiation and chemotherapy which may alter tumor biology. Previous
reports have suggested that stereotactic biopsies
for patients with DIPG are safe and well-tolerated
and as such some groups are now advocating for
biopsies prior to treatment in an effort to gain
more biologic information on this rare and deadly
tumor (Cartmill and Punt 1999).
Molecular Characterization of Diffuse

Intrinsic Pontine Gliomas
Considering biopsies are rarely performed in
patients with DIPG, there is a relative paucity
of information regarding tumor biology and
molecular characterization. In a study by
Gilbertson et al. (2003) at St. Jude Childrens
Research Hospital, researchers analyzed the
molecular features of DIPG specimens obtained
by both biopsy and autopsy. This group of
tumors was heterogenous with regard to genetic
abnormalities; however, a subset of the tumors
had mutations in TP53 and others had TP53
nuclear immunoreactivity which was independent of tumor grade. Another finding was
increasing ERBB1 (epidermal growth factor
receptor) amplification and over-expression
with increasing WHO tumor grade suggesting
that this may be an important pathway in DIPG
development and an appealing therapeutic target (Gilbertson et al. 2003). A previous smaller
study evaluated only seven specimens of which
five were obtained at autopsy. Using polymerase
chain reaction techniques, the researchers identified that four of the seven samples had partial
loss of chromosome 17p which included the
TP53 gene. Another four cases had loss of the
long arm of chromosome 10. Five cases had
mutation in TP53, but none had mutations of
ERBB1 (Louis et al. 1993). Many ongoing studies are evaluating the complex biologic and
molecular composition of these tumors by evaluating both biopsy and post-mortem autopsy
specimens.
279
A Review of Treatment Approaches

Standard therapy for patients with DIPG has
historically involved focal radiation therapy
alone. Radiation often produces transient
improvement in neurologic function and longer
progression-free survival without any difference
in overall survival when compared to patients
treated with other regimens. Considering the
transient response to radiation, many groups have
previously attempted to increase the total radiation dose as high as 78 Gy by hyperfractionation.
These patients had similar outcomes compared
to those who received conventional radiation
therapy, with some increase in treatment related
toxicity such as radiation necrosis and prolonged
steroid dependency (Frazier et al. 2009). Failure
of alternative radiation dosing to improve outcomes prompted studies evaluating the efficacy
of concomitant and adjuvant chemotherapy for
children with DIPG.
A first approach evaluated pre-radiation chemotherapy for newly diagnosed patients. This
strategy has significant limitations, largely
because a majority of patients had rapid early disease progression and required cessation of chemotherapy. Additionally, many patients never
completed the prescribed chemotherapy and
radiotherapy, making the result of the studies difficult to interpret. The largest study of this type
was performed by Jennings et al. (2002) and
included over 60 patients in 2 groups from the
Childrens Cancer Group study CCG-1941. One
group received carboplatin, etoposide, and vincristine and the other group received these three
agents in addition to cyclophosphamide. All
patients then went on to receive radiation therapy.
There was no difference in response and likelihood of progressive disease between the two
groups. Additionally, no differences in PFS or
OS were reported when compared to historic
controls (Jennings et al. 2002) . Other agents that
have been utilized pre-radiation chemotherapy
include carboplatin and thiotepa (used in conjunction with etoposide and autologous stem cell
rescue). In summary, studies with up-front chemotherapy have demonstrated some initial tumor
response, but no change in patient outcomes.
280
One multi-institutional study considered the

use of chemotherapy both before and after conventional radiation therapy. Thirty-three patients
were enrolled and 16 received up front chemotherapy with irinotecan. Ten of these patients
demonstrated disease stabilization with this therapy but six had tumor progression or chemotherapy toxicity. All patients received conventional
radiation followed by monthly pulses of temozolomide. All of the patients in this study had disease progression and died, with no significant
effect on PFS or OS. (Broniscer et al. 2005).
The most extensively studied approach to chemotherapy is drug therapy in combination with
radiation therapy. A variety of agents have been
investigated in this approach, including lomustine,
vincristine, procarbazine, carboplatin, etoposide,
topotecan, thalidomide and temozolomide. The
results of all these studies have been largely
equivocal. Some patients will achieve a brief
response, some will have disease stabilization,
but almost all patients will ultimately succumb to
their disease. The concurrent administration of
chemotherapy and radiation does not significantly prolong progression-free survival or overall survival (Frazier et al. 2009). In 2006, a
German group retrospectively reviewed data
from 153 patients with pontine gliomas treated
with various regimens. Despite the fact that all
the patients in the cohort ultimately died of their
disease, there was some suggestion that both chemotherapy and radiotherapy together improved
PFS. Patients treated with both chemotherapy
and radiation had a 1-year OS of 45.8% as compared to only 34.4% for those treated with chemotherapy alone (Wagner et al. 2006). Data from
this meta-analysis continues to suggest that there
may be a small effect of chemotherapy on survival of patients with DIPG, which has fueled
ongoing research for both new agents and novel
methods to deliver chemotherapy to the tumor.
Finally, a few studies have evaluated the role
of chemotherapy after conventional radiation
therapy. In a large study by the Childrens Cancer
Group, 74 patients received radiation therapy
followed by either BCNU, vincristine and prednisone in one group or no further treatment in the
other. There was no significant difference in OS
between the two groups (Jenkin et al. 1987).

Other agents trialed in this manner have included
busulfan and thiotepa (followed by autologous
stem cell rescue), trophosphamide and etoposide.
As with the other regimens, post-radiation chemotherapy has not altered the disease course for
patients with DIPG (Frazier et al. 2009)
Prognostic Factors and Outcomes

As mentioned previously, most patients with
DIPG will ultimately succumb to their disease.
The current standard of care for patients is focal
radiation therapy with a target dose between 55
and 59 Gy. A significant percentage of patients
(nearly 75%) will experience an improvement in
their symptoms with the initiation of radiation
therapy, but most will rapidly progress. The best
1-year survival rates reported are just under 50%
and 5-year OS rates are between 0% and 15% in
various studies. There are a limited number of
prognostic factors that have been identified.
Patients who have a shorter period from symptoms onset and diagnosis of DIPG as well as
those who have significant neurologic findings
at the time of diagnosis may have poorer outcomes (Fangusaro 2009). As mentioned previously, a single meta-analysis suggests that
patients who receive both chemotherapy and
radiation therapy have an improved 1-year survival (Wagner et al. 2006). Finally, patients with
NF-1 represent a unique population and have a
more indolent course and improved outcomes
with lesions and symptoms that are consistent
with DIPG. It is unknown if these tumors may
have unique molecular and biological characteristics which explain their atypical clinical course
(Fangusaro 2009).
Future Strategies in Treatment

The continued poor prognosis for patients with
DIPG is a driving force behind the development
of new agents and methods of delivering chemotherapy in pediatric neuro-oncology. Given the
lack of success with any of the currently available
26
conventional chemotherapeutic agents, it is unlikely

that these drugs alone or in combination will
result in significant improvements in outcomes.
Several strategies to circumvent the bloodbrain
barrier and deliver chemotherapy locally are in
development and early clinical testing.
Directed delivery of drugs into the CNS is a
promising treatment for strategy for brain tumors.
Adult neuro-oncologists have experience with
biodegradable polymers loaded with chemotherapeutic agents which are directly implanted into
the tumor resection cavity. These have not been
extensively studied in children but have some
limitations for treatment of DIPG. Implantation
of the polymer into the brainstem is technically
difficult without causing significant neurologic
compromise. Further, the drugs rely on diffusion
through the tumor bed to infiltrate the surrounding tissues. In contrast, when drugs are directly
injected into the CNS, otherwise called convection enhanced delivery (CED), a positive pressure
gradient facilitates drug distribution. Thus far,
multiple animal studies have reported successful
implantation of a small catheter into the brainstem through which chemotherapeutic agents can
be delivered. Multiple agents such as BCNU, carboplatin and gemcitabine, interleukin-13 and the
modified Pseudomonas exotoxin have been safely
used in these studies to treat animals with brainstem lesions (Frazier et al. 2009). Most recently,
two cases reports from the National Institutes of
Health reported the safe use of CED in two
patients, one with DIPG. The patient was previously treated for 10 months with conventional
therapies. After CED with Pseudomonas exotoxin, some progressive neurologic signs were
noted within a few days with radiographic or
clinical tumor progression within 8 weeks (Lonser
et al. 2007). This technology still remains in its
infancy; further clinical trials are currently underway. Many factors which affect CED including
agent selection, rate and volume of delivery and
frequency of dosing need to be optimized in order
to maximize efficacy and minimize neurotoxicity. Another alternative technique for circumventing the bloodbrain barrier is the use of intranasal
chemotherapy, which has been explored in some
pre-clinical studies.
281
Most newly diagnosed patients with DIPG

are currently treated with conventional radiotherapy in addition to biologic agents usually
in the setting of a Phase I or Phase II trial.
These agents have included epigenetic drugs,
anti-angiogenics agents and monoclonal antibodies
directed against EGFR amongst others. Newer
agents that are being designed and tested
include small interfering RNAs delivered
directly to the site of the tumor. Modification
of the tumor cells by these molecules may
facilitate natural immune-mediated destruction
of the tumor cells (Frazier et al. 2009). This
technology also provides the basis for the
development of vaccines directed against the
tumor cells. The future of DIPG therapy will
likely include a combination of radiation with
targeted chemotherapy as well as agents with
complementary mechanisms of action such as
immune stimulators or tumor resistance
modifiers.
Discussion
In this chapter, we presented an overview of
both pediatric high-grade gliomas and DIPGs.
Although significant improvement in survival
for children with other brain tumors has been
made in the past few decades, outcomes for
patients with high-grade gliomas and DIPGs
remain poor. Currently, available therapies
including surgery, radiation and conventional
chemotherapy have failed to significantly alter
the OS of children with these lesions.
Researchers are attempting to further clarify
the molecular biology and mechanisms of tumorigenesis in these high-grade tumors. This has
been particularly challenging for DIPG as
untreated tumor tissues is rarely available. Both
high-grade glioma and DIPG are the focus of
ongoing research with newer agents that have
unique mechanisms of action. Coupled with
innovative methods of circumventing the
bloodbrain-barrier and delivery of agents
directly to the tumor, the future of treatment for
patients with these lesions is ongoing and
remains promising.
282
References
Bondy ML, Scheurer ME, Malmer B, Barnholtz-Sloan JS,
Davis FG, Ilyasova D, Kruchko C, McCarthy BJ,
Rajaraman P, Schwartzbaum JA, Sadetzki S, Schlehofer
B, Tihan T, Wiemels JL, Wrensch M, Buffler PA
(2008) Brain tumor epidemiology: consensus from the
brain tumor epidemiology consortium. Cancer
113:19531968
Bredel M, Pollack IF, Hamilton RL, James CD (1999)
Epidermal growth factor receptor expression and gene
amplification in high-grade non-brainstem gliomas of
childhood. Clin Cancer Res 5:17861792
Broniscer A, Iacono L, Chintagumpala M, Fouladi M,
Wallace D, Bowers DC, Stewart C, Krasin MJ, Gajjar
A (2005) Role of temozolomide after radiotherapy for
newly diagnosed diffuse brainstem glioma in children:
results of a multiinstitutional study (SJHG-98). Cancer
103:133139
Broniscer A, Chintagumpala M, Fouladi M, Krasin MJ,
Kocak M, Bowers DC, Iacono LC, Merchant TE,
Stewart CF, Houghton PJ, Kun LE, Ledet D, Gajjar A
(2006) Temozolomide after radiotherapy for newly
diagnosed high-grade glioma and unfavorable lowgrade glioma in children. J Neurooncol 76:313319
Buckner JC, Brown PD, ONeill BP, Meyer FB, Wetmore
CJ, Uhm JH (2007) Central nervous system tumors.
Mayo Clin Proc 82:12711286
Cartmill M, Punt J (1999) Diffuse brain stem glioma. A
review of stereotactic biopsies. Childs Nerv Syst
15:235237, discussion 238
CBTRUS (2005) Statistical report: primary brain and central nervous system tumors diagnosed in the United
States in 20042005. Central Brain Tumor Registry of
the United States. Hinsdale, IL, www.cbtrus.org
Chen W (2007) Clinical applications of PET in brain
tumors. J Nucl Med 48:14681481
Choux M, Lena G, Do L (2000) Brainstem tumors. In:
Choux M, DiRocco C, Hockley A (eds) Pedaitric neurosurgery. Churchill Livingstone, New York, pp
471491
Duffner PK, Horowitz ME, Krischer JP, Friedman HS,
James HE, Freeman CR (1993) Postoperative chemotherapy and delayed radiation in children less than
three years of age with malignant brain tumors. N Engl
J Med 328:17251731
Fangusaro J (2009) Pediatric high-grade gliomas and diffuse intrinsic pontine gliomas. J Child Neurol
24:14091417
Finlay JL, Boyett JM, Yates AJ, Wisoff JH, Milstein JM,
Geyer JR, Bertolone SJ, McGuire P, Cherlow JM,
Tefft M (1995) Randomized phase III trial in childhood high-grade astrocytoma comparing vincristine,
lomustine, and prednisone with the eight-drugs-in-1-day
regimen. Childrens Cancer Group. J Clin Oncol
13:112123

Finlay JL, Dhall G, Boyett JM, Dunkel IJ, Gardner SL,
Goldman S, Yates AJ, Rosenblum MK, Stanley P,
Zimmerman RA, Wallace D, Pollack IF, Packer RJ
(2008) Myeloablative chemotherapy with autologous
bone marrow rescue in children and adolescents with
recurrent malignant astrocytoma: outcome compared
with conventional chemotherapy: a report from the
Childrens Oncology Group. Pediatr Blood Cancer
51:806811
Frazier JL, Lee J, Thomale UW, Noggle JC, Cohen KJ,
Jallo GI (2009) Treatment of diffuse intrinsic brainstem gliomas: failed approaches and future strategies.
J Neurosurg Pediatr 3:259269
Gilbertson RJ, Hill DA, Hernan R, Kocak M, Geyer R,
Olson J, Gajjar A, Rush L, Hamilton RL, Finkelstein
SD, Pollack IF (2003) ERBB1 is amplified and
overexpressed in high-grade diffusely infiltrative
pediatric brain stem glioma. Clin Cancer Res
9:36203624
Glantz MJ, Cole BF, Forsyth PA, Recht LD, Wen PY,
Chamberlain MC, Grossman SA, Cairncross JG
(2000) Practice parameter: anticonvulsant prophylaxis
in patients with newly diagnosed brain tumors.
Report of the Quality Standards Subcommittee of
the American Academy of Neurology. Neurology
54:18861893
Gururangan S, Chi S, Onar A, Packer R, Poussaint T,
Gilbertson R, Friedman H, Kun L, Boyett J (2008)
Abstracts for the thirteenth annual meeting of the
Society for Neuro-Oncology: Phase II study of bevacizumab plus irinotecan in children with recurrent
malignant glioma and diffuse brainstem glioma a
Pediatric Brain Tumor Consortium Study (PBTC022). Neuro-oncology 10:759915
Hargrave D, Bartels U, Bouffet E (2006) Diffuse brainstem glioma in children: critical review of clinical
trials. Lancet Oncol 7:241248
Herrington B, Kieran MW (2009) Small molecule inhibitors in children with malignant gliomas. Pediatr Blood
Cancer 53:312317
Jenkin RD, Boesel C, Ertel I, Evans A, Hittle R, Ortega J,
Sposto R, Wara W, Wilson C, Anderson J (1987)
Brain-stem tumors in childhood: a prospective
randomized trial of irradiation with and without
adjuvant CCNU, VCR, and prednisone. A report of
the Childrens Cancer Study Group. J Neurosurg
66:227233
Jennings MT, Sposto R, Boyett JM, Vezina LG, Holmes
E, Berger MS, Bruggers CS, Bruner JM, Chan KW,
Dusenbery KE, Ettinger LJ, Fitz CR, Lafond D,
Mandelbaum DE, Massey V, McGuire W, McNeely L,
Moulton T, Pollack IF, Shen V (2002) Preradiation
chemotherapy in primary high-risk brainstem tumors:
phase II study CCG-9941 of the Childrens Cancer
Lemort M, Canizares-Perez AC, Van der Stappen A,
Kampouridis S (2007) Progress in magnetic resonance
imaging of brain tumours. Curr Opin Oncol
19:616622
26
Lonser RR, Warren KE, Butman JA, Quezado Z, Robison

RA, Walbridge S, Schiffman R, Merrill M, Walker
ML, Park DM, Croteau D, Brady RO, Oldfield EH
(2007) Real-time image-guided direct convective
perfusion of intrinsic brainstem lesions. Technical
note. J Neurosurg 107:190197
Louis DN, Rubio MP, Correa KM, Gusella JF, von
Deimling A (1993) Molecular genetics of pediatric
brain stem gliomas. Application of PCR techniques
to small and archival brain tumor specimens.
Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, Burger
PC, Jouvet A, Scheithauer BW, Kleihues P (2007) The
2007 WHO classification of tumours of the central
nervous system. Acta Neuropathol 114:97109
Melean G, Sestini R, Ammannati F, Papi L (2004) Genetic
insights into familial tumors of the nervous system.
Am J Med Genet C Semin Med Genet 129C:7484
Nakamura M, Shimada K, Ishida E, Higuchi T, Nakase H,
Sakaki T, Konishi N (2007) Molecular pathogenesis of
pediatric astrocytic tumors. Neuro-oncology 9:113123
Neglia JP, Robison LL, Stovall M, Liu Y, Packer RJ,
Hammond S, Yasui Y, Kasper CE, Mertens AC,
Donaldson SS, Meadows AT, Inskip PD (2006) New
primary neoplasms of the central nervous system in
survivors of childhood cancer: a report from the
Childhood Cancer Survivor Study. J Natl Cancer Inst
98:15281537
children with malignant gliomas. N Engl J Med
346:420427
Pollack IF, Hamilton RL, Sobol RW, Burnham J, Yates
AJ, Holmes EJ, Zhou T, Finlay JL (2006)
O6-methylguanine-DNA methyltransferase expression strongly correlates with outcome in childhood
malignant gliomas: results from the CCG-945 Cohort.
Rickert CH, Strater R, Kaatsch P, Wassmann H, Jurgens
H, Dockhorn-Dworniczak B, Paulus W (2001)
Pediatric high-grade astrocytomas show chromosomal
imbalances distinct from adult cases. Am J Pathol
158:15251532
283
Rood BR, MacDonald TJ (2005) Pediatric high-grade

glioma: molecular genetic clues for innovative therapeutic approaches. J Neurooncol 75:267272
Rosenfeld A, Listernick R, Charrow J, Goldman S (2010)
Neurofibromatosis type 1 and high-grade tumors of
the central nervous system. Childs Nerv Syst
26(5):663667
Sanders RP, Kocak M, Burger PC, Merchant TE, Gajjar
A, Broniscer A (2007) High-grade astrocytoma in very
young children. Pediatr Blood Cancer 49:888893
high grade astrocytoma in children: results of a randomized trial. A report from the Childrens Cancer
Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher
B, Taphoorn MJ, Belanger K, Brandes AA, Marosi C,
Bogdahn U, Curschmann J, Janzer RC, Ludwin SK,
Gorlia T, Allgeier A, Lacombe D, Cairncross JG,
Eisenhauer E, Mirimanoff RO (2005) Radiotherapy
plus concomitant and adjuvant temozolomide for
glioblastoma. N Engl J Med 352:987996
Wagner S, Warmuth-Metz M, Emser A, Gnekow AK,
Strater R, Rutkowski S, Jorch N, Schmid HJ, Berthold
F, Graf N, Kortmann RD, Pietsch T, Sorensen N,
Peters O, Wolff JE (2006) Treatment options in childhood pontine gliomas. J Neurooncol 79:281287
Walter AW, Hancock ML, Pui CH, Hudson MM, Ochs JS,
Rivera GK, Pratt CB, Boyett JM, Kun LE (1998)
Secondary brain tumors in children treated for acute
lymphoblastic leukemia at St Jude Childrens Research
Hospital. J Clin Oncol 16:37613767
Ward BA, Gutmann DH (2005) Neurofibromatosis 1:
from lab bench to clinic. Pediatr Neurol 32:221228
Warren KE (2008) Pediatric high-grade gliomas 2008
educational book. Am Soc Clin Oncol:472477
Wisoff JH, Boyett JM, Berger MS, Brant C, Li H, Yates
AJ, McGuire-Cullen P, Turski PA, Sutton LN, Allen
JC, Packer RJ, Finlay JL (1998) Current neurosurgical
management and the impact of the extent of resection
in the treatment of malignant gliomas of childhood: a
report of the Childrens Cancer Group trial no. CCG945. J Neurosurg 89:5259
Pediatric Low-Grade Glioma:

The Role of Neurobromatosis-1
in Guiding Therapy
27
Robert Listernick and David H. Gutmann
Contents
Abstract
Introduction ............................................................
286
Molecular Genetics
of Neurofibromatosis Type-1 .................................
286
Modeling Neurofibromatosis Type-1

Associated Brain Tumors in Mice.........................
286
Optic Pathway Gliomas .........................................

Epidemiology ...........................................................
Presenting Signs and Symptoms ..............................
Natural History.........................................................
NF1-Associated OPG Versus Sporadic OPG ...........
The Role of Screening Neuroimaging .....................
290
290
290
290
291
291
Treatment of Neurofibromatosis Type-1

Associated Optic Pathway Gliomas ......................
Watchful Waiting .....................................................
Radiation Therapy ....................................................
Surgery ...................................................................
Chemotherapy ..........................................................
291
291
292
292
292
Brainstem Gliomas.................................................
292
References ...............................................................
293
R. Listernick (*)
Division of General Academic Pediatrics,
Childrens Memorial Hospital, Box 16,
2300 Childrens Plaza, Chicago, IL 60614
e-mail: rlisternick@childrensmemorial.org
D.H. Gutmann
Department of Neurology, Washington University School
of Medicine, Campus Box 8111, 660S. Euclid Ave,
St. Louis, MO 63110, USA
e-mail: gutmannd@neuro.wustl.edu
Individuals with NF1 are prone to the

development of both benign and malignant
tumors of the central nervous system, including Schwann cell tumors (e.g. neurofibromas)
and glial cell tumors (e.g. astrocytomas). Although
the most common tumor in children with NF1
is an optic pathway glioma (OPG), astrocytomas may occur throughout the central nervous system most notably in the brainstem.
The greatest risk period for the development
of OPG is in the first 6 years of life.
Symptomatic NF1-associated OPG generally
present in one of two ways- ophthalmologic
symptoms and signs or the development of
precocious puberty. Progressive disease after
diagnosis of NF1-associated OPG requiring
treatment occurs in 2352% of cases. Although
tumors involving the post-chiasmatic optic
tracts and optic radiation have been associated
with a worse prognosis, predicting the natural
history of an individual OPG is impossible. As
low grade astrocytomas in children with NF1
may remain quiescent following diagnosis,
therapeutic decisions regarding these tumors
must incorporate the presence or absence of
NF1 into the decision-making process. Surgery
of an intraorbital OPG should be limited to
improving cosmetic appearance or reducing
the risk of corneal exposure of a proptotic eye.
Given the real risk of secondary malignancies
and cerebral vasculopathy, there is virtually no
role for radiation therapy in the treatment of

285
286
NF1-associated OPG. The combination of

carboplatin and vincristine is standard firstline therapy for symptomatic NF1-associated
OPG which require treatment. Insights derived
from studying the pathophysiology and biochemical basis for glioma formation and visual
loss have uncovered new opportunities for
therapeutic intervention including anti-neoplastic
cell therapies that target the RAS/mTOR/
STAT3 pathway, treatments that target the
tumor microenvironment, including microglia and microglia-derived glioma growth
factors, and neuro-protective strategies that
reduce the collateral damage to neurons
resulting from tumor formation or cancer
treatment.
R. Listernick and D.H. Gutmann
Neurofibromatosis type 1 (NF1) is a common

autosomal dominant disorder which affects 1 in
3,500 people worldwide. Individuals with NF1
are prone to the development of both benign and
malignant tumors of the central nervous system,
including Schwann cell tumors (e.g. neurofibromas)
and glial cell tumors (e.g. astrocytomas).
Although the most common tumor in children
with NF1 is an optic pathway glioma (OPG),
astrocytomas may occur throughout the central
nervous system, most notably in the brainstem. As
the biology of low grade astrocytomas in children
with NF1 is quite distinct from their sporadic
counterparts, therapeutic decisions regarding these
tumors must incorporate the presence or absence
of NF1 into the decision-making process. The
roles of watchful waiting, chemotherapy and
radiation therapy each require careful consideration depending upon the patients NF1 status.
neurons, astrocytes, oligodendrocytes, and

microglia in the central nervous system. Sequence
analysis of the neurofibromin predicted sequence
revealed a small 300 amino acid domain with
similarity to the catalytic region of a family of
proteins called GTPase activating proteins
(GAPs) (Xu et al. 1990). Neurofibromin, like
other GAP molecules, functions to accelerate the
conversion of active GTP-bound RAS to inactive
GDP-bound RAS. In many cell types, including
astrocytes, active GTP-bound RAS increases cell
proliferation by signaling through downstream
RAS effector molecules (Dasgupta et al. 2005a).
Consistent with the role of neurofibromin as a
GAP regulator of cell growth, tumors from
patients with NF1 exhibit high levels of RAS
activity (Basu et al. 1992; DeClue et al. 1992).
Based on these observations, initial biologicallytargeted clinical trials for NF1-associated tumors
employed inhibitors of RAS activity. While these
initial experiences did not result in tumor control,
progress in basic science laboratories revealed
that neurofibromin controls astrocyte cell growth
by negatively regulating the mammalian Target
Of Rapamycin (mTOR)/Signal Transducer
(Dasgupta et al. 2005b) and Activator of
Transcription-3 (STAT3) (Banerjee et al. 2010)
signaling pathway (Fig. 27.1). Studies using both
genetically-engineered mouse and rodent tumor
explant models demonstrated that mTOR
(Hegedus et al. 2008) and STAT3 (Banerjee et al.
2010) inhibition reduce tumor growth in vivo. In
addition to negatively controlling RAS activity,
neurofibromin has also been shown to positively
regulate intracellular cyclic AMP (cAMP) levels
(Fig. 27.1). In this regard, Nf1-deficient mouse
astrocytes exhibit lower levels of cAMP, which
serve to increase their survival in vitro (Dasgupta
et al. 2003) .
Molecular Genetics
of Neurobromatosis Type-1
Modeling Neurobromatosis Type-1

Associated Brian Tumors in Mice
The NF1 gene was identified in 1990, and found

to encode a 220250 kDa protein, termed neurofibromin (Viskochil et al. 1990). Neurofibromin
is expressed in multiple tissue types, including
Individuals with NF1 are born with one mutated

(non-functional) copy of the NF1 gene in all cells
of their body, but develop gliomas only when the
one remaining functional copy of the NF1 gene is
Introduction
27 Pediatric Low-Grade Glioma: The Role of Neurofibromatosis-1 in Guiding Therapy
287
Fig. 27.1 Neurofibromin growth control signaling

pathways. Neurofibromin functions as a negative regulator of the RAS protein by accelerating the conversion of
active GTP-bound RAS to its inactive GDP-bound form.
RAS, once activated, triggers a signaling cascade in astrocytes involving sequential Akt/mTOR/STAT3 activation.
In addition, neurofibromin positively regulates adenylyl

cyclase, which results in increased intracellular cAMP
generation. Neurofibromin loss in NF1-associated glioma
cells leads to reduced cAMP levels and increased Akt/
mTOR/Rac1/STAT3 signaling
inactivated by mutation. Since gliomas are composed

of glial-fibrillary acidic protein (GFAP)expressing cells (glia), we and others sought to
develop a mouse model of NF1-associated optic
glioma by conditionally inactivating the mouse
Nf1 gene in GFAP-immunoreactive cells.
Surprisingly, mice lacking Nf1 gene expression
in GFAP-immunoreactive cells failed to develop
brain tumors (Bajenaru et al. 2002). To more
fully recapitulate the genetics of the human condition, mice with one mutated and one functional
copy of the Nf1 gene (Nf1+/ mice) were generated
that additionally lacked neurofibromin expression

in GFAP-immunoreactive cells. Similar to children with NF1, these genetically-engineered
mice developed low-grade glial neoplasms of the
optic nerve and chiasm (Zhu et al. 2005)
(Fig. 27.2).
The fact that glioma formation required Nf1
inactivation in glial cells coupled with a brain
environment composed of Nf1+/ cells prompted
us to define the role of the tumor microenvironment on gliomagenesis. Examination of the optic
nerves in these mice prior to obvious glioma
288
Fig. 27.2 NF1 gliomagenesis in mice requires Nf1 loss

in progenitor cells coupled with a supportive Nf1+/
microenvironment. Nf1 loss in astrocytes of geneticallyengineered mice is not sufficient for glioma formation in
the brain, whereas Nf1+/ mice with glial Nf1 inactivation
develop optic glioma. These models demonstrate the

requirement for a supportive tumor microenvironment in
the formation of NF1-associated glioma. The inset shows
presence of microglia in these mouse optic gliomas
formation revealed the presence of resident brain

immune system cells, called microglia, in the
region of the developing tumor. Subsequent studies revealed that Nf1+/ microglia, unlike their
wild-type counterparts, produce increased levels
of various chemokines and growth factors
(Daginakatte and Gutmann 2007). Two of these
microglia-generated molecules have been studied
and found to increase Nf1-deficient astrocyte
growth (proliferation and survival) in vitro.
Moreover, silencing of microglia in these genetically-engineered mice reduced optic glioma
tumor growth, further supporting the critical role
of the tumor microenvironment in maintaining
glioma growth.
One of the unfortunate clinical consequences

of optic glioma formation and continued growth
is visual loss. As visual impairment is frequently
quantified in preverbal children using visual
evoked response (VEP) recording, we measured
vision in Nf1 optic glioma mice using VEP and
found reductions associated with glioma formation (Hegedus et al. 2009). In addition, this visual
reduction is accompanied by optic nerve axonal
swelling and increased death in the optic nerve
cell bodies. Subsequent studies revealed that
Nf1+/ optic nerve cells (retinal ganglion neurons) exhibit increased vulnerability to damage
compared to their normal counterparts, which
reflects reduced cAMP levels in Nf1+/ retinal
289
Fig. 27.3 Additional strategies for glioma treatment.

Treatment of NF1-associated glioma could target the neoplastic cells (RAS/mTOR/STAT3) pathway, but also the
microenvironment composed of NF1+/ cells (microglia).
Moreover, neuroprotective strategies to minimize the

collateral damage to neurons would improve outcome
(Hegedus et al. 2009)
ganglion cells (Brown et al. 2010). Following

treatment of Nf1 optic glioma mice with agents
which elevate intracellular cAMP levels, the
number of dying retinal ganglion neurons was
reduced.
Collectively, the insights derived from studying
the pathophysiology and biochemical basis for
glioma formation and visual loss have uncovered
new opportunities for therapeutic intervention
(Fig. 27.3). First, anti-neoplastic cell therapies
that target the RAS/mTOR/STAT3 pathway
should block the proliferation of the NF1deficient glioma cells. Second, treatments that
target the tumor microenvironment, including
microglia and microglia-derived glioma growth
factors, would deprive the NF1-deficient glioma
cells of critical mitogens and chemokines that
drive their continued growth. Third, neuro-protective
strategies that reduce the collateral damage to
neurons resulting from tumor formation or cancer
treatment should limit the injury to the normal
developing brain. Finally, combinations of these
290
therapies are likely to have synergistic beneficial

effects and improve the outcomes in young children
with NF1-associated optic glioma.
Optic Pathway Gliomas

Epidemiology
If all children with NF1 underwent screening
neuroimaging, OPGs would be identified in
approximately 15% of cases. However, of those,
only of these tumors would ever cause signs
or symptoms, making the true incidence of symptomatic OPG approximately 7% (Listernick et al.
1994). NF1-associated OPG are tumors of young
children with a median age at presentation of
approximately 5 years (Listernick et al. 1994;
Singhal et al. 2002). Although the greatest risk
period for the development of OPG is in the first
6 years of life, new symptomatic OPG may arise
in older children and adults. Eight NF1 patients
ranging in age from 8 to 22 years had OPGs that
appeared for the first time or progressed after
the patients seventh birthday (Listernick et al.
2004). These late-onset cases not withstanding,
it is reasonable to counsel families of children
with NF1 that the risk of developing a newly
symptomatic OPG after 7 years of age is
extremely low.
Presenting Signs and Symptoms

Symptomatic NF1-associated OPG generally
present in one of two ways- ophthalmologic
symptoms and signs or the development of precocious puberty. Young children rarely report vision
loss, necessitating the use of reliable, reproducible measures to detect visual abnormalities. The
most commonly found abnormalities on ophthalmologic examination include proptosis, decreased
visual acuity, afferent pupillary defect and optic
nerve atrophy. More detailed testing may reveal
decreased color vision and abnormal visual fields
(Listernick et al. 2007). A retrospective analysis
of NF1-associated OPG demonstrated that 59%
of the patients had visual signs at the time of
diagnosis; 72% of the symptomatic patients had

decreased visual acuity and 31% had proptosis
(Thiagalingam et al. 2004).
As many as 40% of the symptomatic children
will present with signs and symptoms of precocious puberty, predominantly accelerated linear
growth even before secondary sexual characteristics develop (Habiby et al. 1995). Therefore,
careful attention must be paid to the growth of
young children with NF1 and yearly measurements on standardized growth charts should be
performed.
Natural History
Progressive disease after diagnosis of NF1associated OPG requiring treatment occurs in
2352% of cases (Thiagalingam et al. 2004;
Nicolin et al. 2009). Although there are some
prognostic features that have been identified, predicting the natural history of an individual OPG
is impossible. Tumors involving the post-chiasmatic
optic tracts and optic radiation have been associated
with a worse prognosis (Balcer et al. 2001; Liu
et al. 2004). In addition, children with symptomatic tumors who present under the age of 1 and
older than 10 years tend to have a worse prognosis requiring treatment (Listernick et al. 2004;
Opocher et al. 2006). Recent data suggest that
diffusion-weighted and dynamic contrast imaging
may be useful markers in distinguishing aggressive from quiescent OPG, although their utility in
identifying aggressive NF1-associated OPG was
not established (Jost et al. 2008). Cases of spontaneous regression of NF-1 OPG also have been
reported (Parsa et al. 2001).
Few studies have examined the visual outcome
of children who have NF1-associated OPG. One
study demonstrated that, of 15 children with
NF1-associated OPG who received treatment
(11 chemotherapy, 4 radiotherapy), only 1 had definite and 2 had mild improvement in visual acuity
(Dalla Via et al. 2007). A large retrospective multicenter study of 110 patients with NF1-associated
OPG treated with chemotherapy documented
that while most patients had improvement or
stabilization of vision after treatment, visual acuity
worsened in one-third of the patients (personal

communication).
NF1-Associated OPG Versus

Sporadic OPG
Recognition of the inherent biologic differences
between NF1-associated and sporadic OPG is
critical for decision-making regarding treatment.
Patients with sporadic OPG are more likely to
present with signs and symptoms of increased
intracranial pressure and visual loss and to have
progressive disease (Listernick et al. 1995; Singhal
et al. 2002). Optic nerve involvement is more
common in NF1-associated OPG whereas chiasmal and post-chiasmal tumors are more frequently
seen in sporadic OPG. Bilateral intraorbital optic
nerve gliomas are pathognomonic of NF1.
The Role of Screening Neuroimaging

Early detection of OPG by screening neuroimaging would be imperative if it ultimately led to
preserved vision in children. The 1997 NF1-OPG
Task Force determined that there was no conclusive evidence that early detection of tumors
would reduce the rate of vision loss, the primary
source of morbidity (Listernick et al. 2007).
Asymptomatic OPG would be identified that
would never progress, escalating costs and parental anxiety, and exposing the child to the potential
risks of sedation. Moreover, isolated cases of the
emergence of OPG despite a systematic screening protocol have been reported. Alternative
screening tools such as visual evoked potentials
have not proven useful in young children less
than 3 years, the highest risk group. It is possible
that a new generation of technology, such as optical coherence tomography, may be of benefit
when studied. Some institutions perform screening neuroimaging in infants less than 1 year of
age diagnosed with NF1 since reliable measurements of visual acuity in this age group are difficult to obtain.
291
Treatment of Neurobromatosis
Type-1-Associated Optic Pathway
Gliomas
Although there is scant information in the literature as to what constitutes progressive disease
warranting therapy, the primary goal of treatment of OPG should be the prevention of visual
loss in young children. As such, the indications
for treatment should be clearly defined. Previous
studies have been hampered by the use of overaggressive, unnecessary treatment of tumors that
may never have progressed. Although it has not
been carefully studied as yet, radiographic features such as tumor growth or increased tumor
contrast enhancement on neuroimaging should
not be absolute indications for treatment in the
absence of visual loss. In addition, clinical progression has been variably defined as new-onset
endocrinopathy or neurologic dysfunction or a
variety of visual disturbances including deteriorating visual acuity or visual fields. Rather, all of
the above factors need to be taken into account
before entering into a course of action and, in
most cases, treatment should only be undertaken
when the patient has clear evidence of prospectively defined progressive disease.
Watchful Waiting
From the above information, it is clear that at
least 50% of the NF1-associated OPG discovered
on screening neuroimaging and many of the initially symptomatic tumors will never progress
following discovery. As such, a plan of watchful
waiting with serial ophthalmologic and MRI
examinations as previously outlined is appropriate in many circumstances (Listernick et al.
2007). Factors that may influence the decision to
treat include very young age, an inability to
obtain reliable ophthalmologic examinations or
coexisting ophthalmologic morbidities such as
the presence of a growing eyelid plexiform
neurofibroma.
292
Radiation Therapy
There is virtually no role for radiation therapy in
the treatment of NF1-associated OPG. Whereas
radiation therapy was the standard treatment of
OPG for many years, accumulated data highlight
the unacceptable endocrinologic and neurocognitive adverse sequelae. In addition, radiation
therapy poses two specific risks for NF1 patients.
In a multicenter study, 9 of 18 patients with progressive NF1-associated OPG treated with radiation
therapy were diagnosed subsequently with 12
secondary brain tumors for a relative risk of 3.04;
the greatest risk was in patients treated in childhood
(Sharif et al. 2006). Secondly, NF1 children
treated with radiation have a significant risk of
developing cerebral occlusive vasculopathy
(Grill et al. 1999).
Surgery
Surgery of an intraorbital OPG should be limited
to improving cosmetic appearance or reducing
the risk of corneal exposure of a proptotic eye.
Consequently, surgery should be reserved for
proptotic blind or near-blind eyes. There are no
data that suggest that tumors limited to the
intraorbital or intracranial optic nerve are at risk
of growing backward and infiltrating the optic
chiasm leading to bilateral visual impairment. On
occasion, hypothalamic or chiasmal tumors may
lead to hydrocephalus due to third ventricular
compression requiring surgical decompression.
Although surgical biopsy is not generally useful
for typical NF1-associated OPG, of 17 NF1 brain
tumors deemed atypical because of rapid growth,
atypical location or progressive symptoms, nine
were not classic pilocytic astrocytomas. However,
none of the high grade tumors were in the optic
pathway (Leonard et al. 2006).
Chemotherapy
The combination of carboplatin and vincristine
has been the standard first-line treatment for
pediatric low-grade gliomas for over 10 years
(Packer et al. 1997). In a retrospective study of

nine children with NF1-associated OPG treated
with carboplatin alone, improved vision was
found in four patients and significant radiologic
regression occurred in four patients (Listernick
et al. 1999). Given the significant percentage of
children who develop carboplatin allergy (as high
as 44% in one study) and the potential for an
unfavorable visual outcome as previously
described, alternate strategies have been sought.
Although the use of temozolomide alone or
the combination of procarbazine, vincristine,
6-thioguanine and chloroethylcyclohexylnitosurea (CCNU) has been shown to have activity
against low grade gliomas, the use of alkylating
agents and multidrug regimens in children with
NF1 runs the theoretical risk for the development
of secondary malignancies. A phase II study of
weekly vinblastine for treatment of recurrent or
refractory low grade gliomas demonstrated a
response rate of 42%; at a median follow-up of
31 months, 19 of 29 patients who completed
52 weeks of therapy were progression free
(Nicolin et al. 2009). Future studies of biologic
agents which target the RAS/mTOR/STAT3 pathway undoubtedly will move the treatment of
NF1-associated OPG away from traditional chemotherapeutic regimens and closer to the age of
molecular medicine.
Brainstem Gliomas
Brainstem gliomas account for approximately
20% of all pediatric central nervous system
neoplasms. Of these, the majority are diffusively
infiltrative and have an exceedingly dismal outcome with a median survival of 912 months
despite aggressive therapy. In contrast, NF1
brainstem tumors are indolent tumors rarely
requiring treatment.
NF1-brainstem gliomas must first be distinguished from the common focal areas of T2 signal hyperintensity seen on MRI in as many as
60% of children with NF1, so-called unidentified bright objects (UBOs). UBOs, which disappear with increasing age, are felt to represent
either areas of dysplasia or dysmyelination.
They display neither mass effect nor contrast

enhancement, clearly distinguishing them from
true neoplasms. In one study of 125 patients with
NF1 who underwent neuroimaging, brainstem
gliomas were identified in 23 patients, of whom
15 were asymptomatic at the time having undergone neuroimaging for either unrelated symptoms
or a remote lesion (Ullrich et al. 2007). The symptomatic tumors generally caused either cranial
nerve deficits or signs of increased intracranial
pressure due to obstructive hydrocephalus. Given
the precarious location of these tumors, only one
tumor with an exophytic component underwent
biopsy, confirming its identity as a low-grade
astrocytoma. Although 6 of the 23 patients
received some form of treatment that might have
affected the natural history of the tumor, only one
untreated patient had evidence of radiographic
and clinical progression over a median follow-up
period of 67 months. Thus, as in NF1-associated
OPG, the accumulated evidence suggests that
brainstem mass lesions in NF1 should be
approached conservatively and should be treated
only when documented clinical and radiographic
progression has occurred.
Acknowledgements We thank Ms. Samantha Higer for
generating the illustrations used in this chapter.
References
Bajenaru ML, Zhu Y, Hedrick NM, Donahoe J, Parada LF,
Gutmann DH (2002) Astrocyte-specific inactivation of
the neurofibromatosis 1 gene (NF1) is insufficient for
astrocytoma
formation.
Mol
Cell
Biol
22(14):51005113
Balcer LJ, Liu GT, Heller G, Bilaniuk L, Volpe NJ, Galetta
SL, Molloy PT, Phillips PC, Janss AJ, Vaughn S,
Maguire MG (2001) Visual loss in children with neurofibromatosis type 1 and optic pathway gliomas: relation to tumor location by magnetic resonance imaging.
Am J Ophthalmol 131(4):442445
Banerjee S, Byrd JN, Gianino SM, Harpstrite SE,
Rodriguez FJ, Tuskan RG, Reilly KM, PiwnicaWorms DR, Gutmann DH (2010) The neurofibromatosis type 1 tumor suppressor controls cell
growth by regulating signal transducer and activator
of transcription-3 activity in vitro and in vivo.
Cancer Res 70(4):13561366
Basu TN, Gutmann DH, Fletcher JA, Glover TW, Collins
FS, Downward J (1992) Aberrant regulation of ras
293
proteins in malignant tumour cells from type 1

neurofibromatosis patients. Nature 356(6371):713715
Brown JA, Gianino S, Gutmann DH (2010) Defective
cyclic AMP generation underlies the sensitivity of
central nervous system neurons to neurofibromatosis-1
heterozygosity. J Neurosci 30(16):55795589
Daginakatte GC, Gutmann DH (2007) Neurofibromatosis-1
(Nf1) heterozygous brain microglia elaborate paracrine factors that promote Nf1-deficient astrocyte and
glioma growth. Hum Mol Genet 16(9):10981112
Dalla Via P, Opocher E, Pinello ML, Calderone M,
Viscardi E, Clementi M, Battistella PA, Laverda AM,
Da Dalt L, Perilongo G (2007) Visual outcome of a
cohort of children with neurofibromatosis type 1 and
optic pathway glioma followed by a pediatric neurooncology program. Neuro-oncology 9(4):430437
Dasgupta B, Dugan LL, Gutmann DH (2003) The neurofibromatosis 1 gene product neurofibromin regulates
pituitary adenylate cyclase-activating polypeptidemediated signaling in astrocytes. J Neurosci
23(26):89498954
Dasgupta B, Li W, Perry A, Gutmann DH (2005a) Glioma
formation in neurofibromatosis 1 reflects preferential
activation of K-RAS in astrocytes. Cancer Res
65(1):236245
Dasgupta B, Yi Y, Chen DY, Weber JD, Gutmann DH
(2005b) Proteomic analysis reveals hyperactivation of
the mammalian target of rapamycin pathway in neurofibromatosis 1-associated human and mouse brain
tumors. Cancer Res 65(7):27552760
DeClue JE, Papageorge AG, Fletcher JA, Diehl SR, Ratner
N, Vass WC, Lowy DR (1992) Abnormal regulation of
mammalian p21ras contributes to malignant tumor
growth in von Recklinghausen (type 1) neurofibromatosis. Cell 69(2):265273
Grill J, Couanet D, Cappelli C, Habrand JL, Rodriguez D,
Sainte-Rose C, Kalifa C (1999) Radiation-induced cerebral vasculopathy in children with neurofibromatosis
and optic pathway glioma. Ann Neurol 45(3):393396
Habiby R, Silverman B, Listernick R, Charrow J (1995)
Precocious puberty in children with neurofibromatosis
type 1. J Pediatr 126(3):364367
Hegedus B, Banerjee D, Yeh TH, Rothermich S, Perry A,
Rubin JB, Garbow JR, Gutmann DH (2008) Preclinical
cancer therapy in a mouse model of neurofibromatosis-1 optic glioma. Cancer Res 68(5):15201528
Hegedus B, Hughes FW, Garbow JR, Gianino S, Banerjee
D, Kim K, Ellisman MH, Brantley MA Jr, Gutmann
DH (2009) Optic nerve dysfunction in a mouse model
of neurofibromatosis-1 optic glioma. J Neuropathol
Exp Neurol 68(5):542551
Jost SC, Ackerman JW, Garbow JR, Manwaring LP,
Gutmann DH, McKinstry RC (2008) Diffusionweighted and dynamic contrast-enhanced imaging as
markers of clinical behavior in children with optic
pathway glioma. Pediatr Radiol 38(12):12931299
Leonard JR, Perry A, Rubin JB, King AA, Chicoine
MR, Gutmann DH (2006) The role of surgical
biopsy in the diagnosis of glioma in individuals
with neurofibromatosis-1. Neurology 67(8):15091512
294
Listernick R, Charrow J, Greenwald M, Mets M (1994)
Natural history of optic pathway tumors in children
with neurofibromatosis type 1: a longitudinal study.
J Pediatr 125(1):6366
Listernick R, Darling C, Greenwald M, Strauss L, Charrow
J (1995) Optic pathway tumors in children: the effect
of neurofibromatosis type 1 on clinical manifestations
and natural history. J Pediatr 127(5):718722
Listernick R, Charrow J, Tomita T, Goldman S (1999)
Carboplatin therapy for optic pathway tumors in children with neurofibromatosis type-1. J Neurooncol
45(2):185190
Listernick R, Ferner RE, Piersall L, Sharif S, Gutmann
DH, Charrow J (2004) Late-onset optic pathway
tumors in children with neurofibromatosis 1.
Neurology 63(10):19441946
Listernick R, Ferner RE, Liu GT, Gutmann DH (2007) Optic
pathway gliomas in neurofibromatosis-1: controversies
and recommendations. Ann Neurol 61(3):189198
Liu GT, Brodsky MC, Phillips PC, Belasco J, Janss A,
Golden JC, Bilaniuk LL, Burson GT, Duhaime AC,
Sutton LN (2004) Optic radiation involvement in
optic pathway gliomas in neurofibromatosis. Am
J Ophthalmol 137(3):407414
Nicolin G, Parkin P, Mabbott D, Hargrave D, Bartels U,
Tabori U, Rutka J, Buncic JR, Bouffet E (2009) Natural
history and outcome of optic pathway gliomas in
children. Pediatr Blood Cancer 53(7):12311237
Opocher E, Kremer LC, Da Dalt L, van de Wetering MD,
Viscardi E, Caron HN, Perilongo G (2006)
Prognostic factors for progression of childhood optic
pathway glioma: a systematic review. Eur J Cancer
42(12):18071816
Packer RJ, Ater J, Allen J, Phillips P, Geyer R, Nicholson HS,
Jakacki R, Kurczynski E, Needle M, Finlay J, Reaman G,
Boyett JM (1997) Carboplatin and vincristine chemotherapy for children with newly diagnosed progressive
low-grade gliomas. J Neurosurg 86(5):747754

Parsa CF, Hoyt CS, Lesser RL, Weinstein JM, Strother
CM, Muci-Mendoza R, Ramella M, Manor RS,
Fletcher WA, Repka MX, Garrity JA, Ebner RN,
Monteiro ML, McFadzean RM, Rubtsova IV, Hoyt
WF (2001) Spontaneous regression of optic gliomas:
thirteen cases documented by serial neuroimaging.
Arch Ophthalmol 119(4):516529
Sharif S, Ferner R, Birch JM, Gillespie JE, Gattamaneni
HR, Baser ME, Evans DG (2006) Second primary
tumors in neurofibromatosis 1 patients treated for
optic glioma: substantial risks after radiotherapy.
J Clin Oncol 24(16):25702575
Singhal S, Birch JM, Kerr B, Lashford L, Evans DG
(2002) Neurofibromatosis type 1 and sporadic optic
gliomas. Arch Dis Child 87(1):6570
Thiagalingam S, Flaherty M, Billson F, North K (2004)
Neurofibromatosis type 1 and optic pathway
gliomas: follow-up of 54 patients. Ophthalmology
111(3):568577
Ullrich NJ, Raja AI, Irons MB, Kieran MW, Goumnerova
L (2007) Brainstem lesions in neurofibromatosis
type 1. Neurosurgery 61(4):762766, discussion
766767
Viskochil D, Buchberg AM, Xu G, Cawthon RM, Stevens
J, Wolff RK, Culver M, Carey JC, Copeland NG,
Jenkins NA et al (1990) Deletions and a translocation
interrupt a cloned gene at the neurofibromatosis type 1
locus. Cell 62(1):187192
Xu GF, Lin B, Tanaka K, Dunn D, Wood D, Gesteland R,
White R, Weiss R, Tamanoi F (1990) The catalytic
domain of the neurofibromatosis type 1 gene product
stimulates ras GTPase and complements ira mutants
of S. cerevisiae. Cell 63(4):835841
Zhu Y, Harada T, Liu L, Lush ME, Guignard F, Harada C,
Burns DK, Bajenaru ML, Gutmann DH, Parada LF
(2005) Inactivation of NF1 in CNS causes increased
glial progenitor proliferation and optic glioma formation. Development 132(24):55775588
Treatment of Pediatric OpticHypothalamic Gliomas: Prognosis
28
Luca Massimi
Contents
Abstract
Introduction ............................................................
295
Natural History ......................................................
296
Survival and Tumor Progression ..........................
297
Prognostic Factors..................................................
Age ...........................................................................
Tumor Location........................................................
NF-1 .........................................................................
Hydrocephalus .........................................................
Histological Features ...............................................
Management.............................................................
Statistical Value........................................................
300
300
300
300
301
301
301
302
Early Outcome and Complications

of Treatments ............................................................
302
Long-Term Clinical Outcome ...............................
303
References ...............................................................
305
Pediatric optic-hypothalamic gliomas (POHGs)

have a favorable prognosis with regards to the
long-term survival if compared with other
pediatric tumors of the central nervous system. Indeed, in spite of the high rate of tumor
progression, POHGs can reach a 10-year overall survival more than 80% in many series.
The reasons of such a good survival are the
indolent course and/or the favorable location
(Dodge I tumors) showed by some of them,
the lower trend of progression in case of associated neurofibromatosis type 1, and the good
response to integrated therapies.
On the other hand, the long-term clinical
outcome of affected children strongly contrasts with the survival because of the involvement of the chiasm and the hypothalamus by
the tumor and the sequelae of treatment.
Unilateral or bilateral severe loss of vision/
visual field amputation, neurological deficits,
chronic endocrine imbalance, and neuropsychological impairment actually make the
quality of life of POHGs patients poor in most
of the cases.
Introduction
L. Massimi (*)
Institute of Neurosurgery-Pediatric Neurosurgery,
A. Gemelli Hospital,
Largo A. Gemelli 8, 00168 Rome, Italy
e-mail: lmassimi@email.it

show several antithetic aspects that make their
treatment controversial and their prognosis hard
to be summarized. Actually, although the large

295
296
majority of them are biologically benign low

grade gliomas, POHGs involve highly eloquent
and/or vital neural structures so that their gross
total surgical excision can be rarely achieved and
(neo)adjuvant treatments are often required in
case of disease progression. The prognosis is
affected by the tumor location, it being excellent
in case of Dodge I tumors and significantly worse
in case of involvement of the chiasm and, in particular, the hypothalamus. Moreover, important
differences may be found when considering the
age and the presence of Neurofibromatosis 1
(NF-1), since POHGs may present an indolent
course in adolescents and/or NF-1 subjects while
they can quickly enlarge and progress in young
children (namely, infants) and/or non-NF-1
patients. Finally, the use of diverse protocols by
different Centers makes the results of management hardly comparable.
The aim of this chapter is to report on the natural history and the prognosis of POHGs, paying
attention to the main prognostic factors and to the
long-term outcome.
Natural History
There is a general agreement in the literature on
to define the behavior of POHGs as unpredictable
(Jahraus and Tarbell 2006; Massimi et al. 2007).
Indeed, about a half of them remains silent for
long periods (years or decades), or presents a
slow growth over the time. Moreover, spontaneous or surgery-induced reduction in size or even
complete disappearance of the tumor are reported
not rarely. On the other hand, 2030% of POHGs
exhibit a constant clinical and radiological progression, thus requiring a complex management. In
some instances, they can even mimic malignant
tumors presenting rapid volume enlargement,
infiltrative growth pattern, and metastatic seeding
through the CSF pathways.
The benign course of POHGs was traditionally explained with the possible hamartomatous
nature of these tumors. This hypothesis was subsequently denied based on histological findings
revealing well characterized tumors (mainly low
grade astrocytomas) provided with a proliferation
L. Massimi
activity. Surely, the absence of anaplastic features

and the low proliferative index can contribute to
their indolent course in several instances.
Moreover, the intrafascicular growth, which is
found in anterior POHGs, may represent a further
limit to the progression of this subset of POHGs,
though the role of such an anatomic location has
been questioned by some authors (Shuper et al.
1999). However, the mechanism of the delayed
growth/slow progression still remains unclear in
most cases. Some predicting factors have been
recently proposed, as increased cell apoptosis,
efficient host immune reactions, hormonal
imbalance, decelerating growth kinetics, spontaneous or surgery-related vascular occlusion
and subsequent necrosis. Ongoing studies on
these issues could provide further information
on this topic.
The unfavorable clinical course showed by
progressing POHGs is hard to be explained as
well. Actually, such an aggressive behavior can
be attributed to higher grade histotypes (pilomyxoid astrocytoma, grade III and IV astrocytomas)
only in a few cases. The large majority of POGHs
do not present any correlation between histological, clinical or radiological features, and tumor
progression. Although the mechanisms inducing
or associated to the tumor progression currently
remain obscure, some interesting hypotheses
have been formulated on the basis of recent
radiological or molecular findings. According to
clinico-radiological observations, the tumor progression could depend, at least in part, on the volume of the tumor at diagnosis since large tumors
show a slower growth compared with small ones.
Also the morphological appearance and the location of POHGs would play an important role, the
progression being more likely in globular exophytic chiasmatic/hypothalamic tumors than in
multilobular tumors extending to the optic tracts
(Shuper et al. 1999). Moreover, a quick tumor
progression would correlate with the tumor metabolic profile detected by MRI spectroscopy
(higher normalized choline concentrations) or by
positron emission tomography (increased glucose
uptake). As to the base research, some authors
proposed vascularization and angiogenesis as
possible factors favoring POHGs progression,
28
Treatment of Pediatric Optic-Hypothalamic Gliomas: Prognosis
the tumors with higher microvessels density

showing a higher rate of progression and a shorter
progression-free survival (PFS) (Bartels et al.
2006). Interesting correlations have been found
also between POHGs cells proliferation and
topoisomerase II-alpha expression (Bredel et al.
2002) or 18F-fluorothymidine uptake/breakdown
of the brain blood barrier (Muzi et al. 2006).
The natural history of POHGs in NF-1
deserves a special mention since they are reported
to show a particularly benign course. Traditionally,
NF-1 is believed to be associated to a lower risk
of POHGs progression so that affected subjects
can be often managed by simple observation.
Moreover, spontaneous tumor regression is nearly
always observed in patients with NF-1. Many
series showing lower rates and degrees of tumor
progression in NF-1 patients harboring POHGs
are reported in the literature (Dutton 1994;
Jahraus and Tarbell 2006; Steinbok et al. 2002).
Some authors, however, demonstrated rates of
tumor progression similar to those of non-NF-1
subjects. In the largest series of NF-1 children
with low grade gliomas provided so far (109
patients, 83 of whom harboring POHGs), 75% of
subjects showed tumor progression after a
5.3 years median follow-up (Driever et al. 2010).
In the same series, the 5-year event free survival
(EFS) was as low as 24%, strongly contrasting
with the 96% overall survival (OS), and POHGs
were associated with the worst EFS. Such a
behavior can produce a worsening of the visual
function in up to one fifth of NF-1 patients over
their life. As to the age, tumor progression is considered almost virtually absent in adolescents or
adults with NF-1 compared with non-NF-1
patients. Also such a low risk of negative evolution has been questioned by some authors who
detected late tumor progression in adolescents
and in up to 10% of cases more than 1 year after
the diagnosis (Segal et al. 2010). Grill et al.
(2000), however, found a statistically significant
difference regarding the patients age at tumor
progression between children with NF-1 (47%
after the 6th year of life, mean age 5.7 2.9 years)
and those without NF-1 (76% before the 6th year,
mean age 4.1 4 years).
297
Survival and Tumor Progression

In spite of their unpredictable behavior and their
unfavorable location, POHGs are generally associated to a good long-term survival. According to
the literature (Table 28.1), indeed, the 5-year OS
ranges from 70% to 100%, the 10-year OS from
50% to more than 90%, and the 15-year OS from
35% to 75%. POHGs also do better than other low
grade gliomas of the diencephalic region, as third
ventricle (10-years OS: ~65%), pineal region
(~55%), thalamus (~53%), and whole diencephalon (25%) (Yazici et al. 2011). Such excellent
rates result from both the indolent course often
showed by POGHs and by the good response to
treatments in many progressing tumors. A further
explanation is that late mortality in patients harboring POHGs is likely to be caused by factors
different from the progression of their optic pathway tumors (namely, hypothalamic dysfunction
or late sequelae of treatments). Among all these
concurrent factors, the good quality of the management and the effectiveness of treatments seem
to play the most important role. According to the
analysis performed by Oh et al. (2011) in a series
of 181 children with low grade gliomas, indeed,
POHGs showed the same OS than tumors located
elsewhere (94% after a 6.4 years median followup) in spite of a significantly worse freedom from
progression (39% versus 76%). On multivariate
analysis, the optic-hypothalamic location was the
only predictor of poor freedom from progression.
Such a pronounced tendency to progress has
been confirmed also by other authors in large
series of children with low grade astrocytomas
where POHGs represented a specific risk factors
for tumor progression (other than early age, fibrillary histotype, and extent of surgery) (Stokland
et al. 2010). In POHGs, the tumor progression
feels the effects of the different prognostic factors
and the different types of management, as showed
by the wide range of percentages associated to
the progression free survival (PFS). The 3-year
PFS actually ranges from 40% to 100% and the
5-year PFS from 10% to 85%. PFS may vary
even within the same series according to the
19
28
104
18
23
54
32
73
104
50
85
198
Fouladi et al.
(2003)
Guillamo et al.
(2003)
Khafaga et al.
(2003)
Laithier et al.
(2003)
Gnekow et al.
(2004)
Komotar et al.
(2004)
Combs et al.
(2005)
15
63 (21
NA
PMA + 42
PA)
29
Gururangan et al.
2002
Massimino et al.
(2002)
Steinbok et al.
(2002)
6.9
PMA: 18
months; PA:
58 months
3.6
1.42
5.6
4.7
<16
3.75
4.08
11.1
23
24
1.96
Age at
diagnosis
(years)
11 < 5 year
21
51
Authors
Patients
25
Grabenbauer et al.
(2000)
50
Mahoney et al.
(2000)
Gayre et al. (2001) 42
NF1
(No)
3
I: 20; IIIII: 80
III: 100
I: 2; II: 30; III:

68
II: 17; III: 83
I/II: 30, III: 70
I: 22, II/III: 88
II: 30; III: 70
II: 43 III: 57
NA
I/II: 51, III: 49
I: 67, II: 33
Dodge (%)
I: 8; II: 32; III:
60
I: 1; II/III: 49
Treatment
Biopsy in 60% and subtotal removal
40% + 100% RT (4560 Gy)
POG 8936 (carboplatin CT, biopsy only for
hypothalamic tumors)
Surveillance in 13; surgery + adjuvant CT or
RT in 26
Carboplatin in 100% (previous surgery in 16,
previous CT in 13 and RT in 10)
Biopsy in 23; CT in all the cases (cisplatinetoposide)
Observation in 13; surgery alone in 9;
surgery + RT in 5, surgery + CT in 4; CT
alone in 1
Surveillance in 37; biopsy/conservative
surgery + CT (cyclophosphamide, vincristine,
carboplatin) in 20 or + RT (52.2 Gy) in 16
Surveillance in 41; surgery in 9, RT in 28,
CT in 11
Surgery alone in 17; surgery + RT (RT 50 Gy)
in 12; RT alone in 16; observation in 5
BBSFOP in all the cases (at diagnosis in
61%, after surveillance in 39%)
Surgery in 143 (67 biopsies); CT (carboplatin-vincristine) in 123; RT in 27
PMA: surgery in 21; surgery alone in 6, CT
in 10, RT in 1, CT + RT in 4
PA: surgery in 42; surgery alone in 13; CT in
3; RT in 15; Ct + RT in 11
Surgery + RT (52.2 Gy) in all but 2 patients
(RT alone)
3.5
6.5
Time: 26 months in NA
PMA, 147 months
in PA
5 year: 61%
7.5 year: 100%; 5 3 year: 92%; 5

year: 90%
year: 72%
Time: 63 months
in PMA, 213
months in PA
5 year: 93%
3.6
5.20
Follow up
(years)
9
6 year: 36% (RT

13.6
69%; CT 12%; wait
and see 37%)
Not available
5.6
100% at follow-up
in II, 33% in III
3 year: 78%
3 year: 64%
Event free survival:

50%
10 year: 65%
Progression free
survival
10 year: 69%
5 year: 90%; 10
year: 82%
5 year: 87.5%; 10 5 year: 69%; 10
year: 75%
year: 62%
5 year: 89%
5 year: 34%
100% at
follow-up in I/II,
72% in III
6 year: 86 5%
3 year: 100%
3 year: 84%
Not available
9 year: 85%
Overall survival
10 year: 94%
Table 28.1 Synopsis of the characteristics and survival of some the main series reported in the literature in the last decade
298
L. Massimi
78
83
17
16
18
133
15
83
Surez et al.
(2006)
Hsu et al. (2008)
Jaing et al. (2008)
Nicolin et al.
(2009)
Sawamura et al.
(2009)
Driever et al.
(2010)
<17
5.89
5.5
30 months
3 year + 8
months
8.3
I/II: 38; III: 45
Observation in 40%; surgery in 25% (biopsy

in more than 50%), CT in 50%, RT in 10%
Biopsy in all the cases (except for NF-1 and

clear diagnosis); stereotactic RT in 37
(52.2 Gy)
I: 9; II: 24; III:
Gross surgical debulking in 33; adjuvant CT
67
(vincristine + carboplatin) in 10
I: 6; II/III: 94
Surgery alone in 7; surgery + RT (50 Gy) in
3; surgery + CT (actinomycin-D + Vincristine)
in 2; surgery + brachytherapy in 2; RT alone
in 2; observation in 1
II/III: 16
CT (cisplatin, etoposide and vinblastine): all
patients had progressing, unresectable
tumors (non had received RT)
I: 3; II: 1; III: 14 Surgery in all cases (12 debulking, 6 biopsies);
CT in 4 (cisplatin + etoposide), RT in 3
I: 29; II: 54; III: Observation in 64; CT (mainly carbopla50
tin + vincristine) in 32, surgery in 18,
surgery + CT in 15, surgery + RT in 2, RT in 2
II/III: 15
Biopsy in 10 (all PA); CT (cisplatin-vincristine) in 15
NA
PA pilocytic astrocytoma, PMA pilomyxoid astrocytoma
33
Ahn et al. (2006)
37
Marcus et al.
(2005)
510 year: 94.6%

(for treated
children)
100% (median
follow-up: 63
months)
12 year: 96%
5 year: 80%
81.5 months:
93.8%
7.5 year: 70%
5 year: 93%
5 year: 97.8%, 8
year: 82%
5 year: 73%
5 year: 48.3%; 10
year: 45.6% (for
treated children)
Time: 55 months
5 year: 63.3%
5 year: 55.5%
Time: 39 months
5 year: 52%
5 year: 82.5%; 8
year: 65%
5.3
6 months17 year
8.6
6.8
NA
6.05
6.9
28
299
L. Massimi
300
different therapy administered. Fouladi et al.

(2003), for example, reported a 69% 6-year PFS
in children who received RT and 12% 6-year PFS
for those who received CT (36% 6-years PFS for
the whole cohort). Unexpectedly, EFS seems to
present a more regular trend, being about 60% at
5 years and about 4060% at 10 years, without significant differences between NF-1 and non-NF-1
patients (Mahoney et al. 2000; Yazici et al. 2011).
The largest series where the tumor progression
was specifically investigated is composed by 133
children described by Nicolin et al. (2009). There,
tumor progression interested 23 out of 84 children
initially managed by clinical observation alone
(27%). After first line treatment, 34 out of 69 children underwent tumor progression (49%), the
phenomenon involving 55% of non-NF-1 patients
and 44% of NF-1 patients. Twenty children
showed one progression while the remaining 14
experienced two or more progressions (up to eight
progressions in two cases). The mean time from
treatment to first progression was 28.4 months.
Differently from the Fouladis experience, these
authors did not find significant variations of PFS
according to the different type of management
(surgery alone, chemotherapy alone, surgery plus
chemotherapy). The results in terms of survival in
the current literature are reported in Table 28.1.
Prognostic Factors
Age
The most unfavorable course in POHGs is
observed in infants or very young children. The
age less than 35 years is regarded as the most
important prognostic factor since it often results
the sole significant predictor of prognosis on statistical analyses (Grabenbauer et al. 2000; Khafaga
et al. 2003; Sawamura et al. 2009; Stokland et al.
2010). The worse prognosis in infants compared
with older children concerns both the 5-year PFS
(3545%) and the 5-year progression rate (25
40%), the OS more than 5 years (from 60% up to
20%), and the long-term outcome.
Local aggressiveness, tumor dissemination,
early hypothalamic involvement, poor clinical
conditions at diagnosis (diencephalic syndrome)

are typical of children less than 3 years where the
prognosis does not feel even the benefic effects
of the association with NF-1. Such an aggressive
behavior is thought to result from a higher rate of
malignant or pilomyxoid gliomas as well as from
an immature immune response against tumors
provided by infants and very young children. On
these grounds, infants and young children usually
undergo specific treatment protocols.
Tumor Location
Dodge I tumors are generally considered to have
the best long term survival since their favorable
location allows the surgeon to achieve a complete
tumor resection in case of disease progression.
Actually, their OS is generally close to 100%
although it does not significantly exceed that of
children with posterior-chiasmatic tumors (80
90%) in several series considering at least a
10 years follow-up (Caldarelli and Pezzotta
1999). Such an apparent discrepancy results from
the risk of tumor recurrence (530%) showed by
anteriorly-located tumors despite the surgical
resection, because of a possible chiasm invasion
undetected on neuroimaging investigations, and
on the progression of other lesions associated to
NF-1 (e. g., malignant gliomas). Of course,
Dodge III tumors are those showing the worst
long-term survival and outcome due to the infiltration of hypothalamus. Data on prognosis
according to tumor location from the extensive
review carried out by Dutton (1994) on 1,136
cases are summarized in Table 28.2.
NF-1
NF-1 is considered as a favorable prognostic factor in POHGs. As mentioned in the paragraph on
the natural history, such a belief is quite controversial since some authors observed excellent OS
and PFS compared with non-NF-1 patients
(Laithier et al. 2003; Nicolin et al. 2009) while
others did not find significant differences
(Caldarelli and Pezzotta 1999), or even noticed
28
Table 28.2 Prognosis according to tumor location (from

Dutton 1994)
Dodge I
Tumor-related mortality 5%a
Tumor progression
21%
Stabilization of vision
91%
Dodge II
29%
42%
77%
Dodge III
43%
51%
71%
Rate including non-treated patients (usually about 0% in

case of treated patients)
that NF-1 was associated to a less unfavorable

prognosis (Massimi et al. 2007). The protective
action of NF-1 on POHGs would depend on their
peculiar involvement of the single optic nerve,
the chiasmatic and the hypothalamic invasion
being less common. Moreover, when the chiasm
or the hypothalamus are involved, neuroimaging
features of NF-1-related POHGs are better than
those showed by non-NF-1 children. In these
cases, POGHs are smaller, rarely associated to
enlarging tumoral cysts, calcification, and hydrocephalus. Besides, they tend to remain stable in
size and to avoid the encasement of the Willis
circle arteries. On the other hand, the bad survival
observed in some instances could result from the
evolution of other NF-1-related lesions other than
from tumor progression.
Hydrocephalus
Although rarely found as independent prognostic
factor, hydrocephalus is a well-known disease
complicating the clinical course and contributing
to the worse prognosis in POHGs with prevalent
hypothalamic involvement (Tow et al. 2003). The
mortality and morbidity related to hydrocephalus
in shunted children actually adds up to those
associated with POHGs, thus leading to a vicious
circle. Dangerous sequelae connected to hydrocephalus result from the possible complications
of shunting devices and the relatively high number of reoperations that is required to face them.
301
represent an exception since they are almost

invariably low grade gliomas (grade I/II astrocytomas). Among them, grade I pilocytic astrocytomas are definitely more common than grade II
fibrillary or pilomyxoid astrocytomas. Therefore,
although grade II astrocytomas are associated
with a worse prognosis, the statistical relevance
of this association is usually low. Actually, only
in a few series with a relatively high number of
grade II POHGs there is a significant correlation
between histotype and OS/PFS (Komotar et al.
2004; Stokland et al. 2010). Since fibrillary astrocytoma is rare in children, pilomyxoid astrocytoma
is generally responsible of such a poor prognosis
because of its adverse characteristics, as prevalent
occurrence in infants and young children, frequent involvement of the hypothalamus, relatively high proliferation index (Ki67 > 3%), and
tendency to infiltrative growth, CSF dissemination and recurrence.
Cell proliferation is a good predictor of tumor
progression but its sensitivity in POHGs seems to
be reduced. The MIB-1 proliferation index,
indeed, does not show a strict relationship with
progression and survival of POHGs. Moreover,
topoisomerase II-a, another proliferation marker
with a good correspondence with MIB-1 index,
has been proved not to correlate with POHGs
prognosis (Bredel et al. 2002). On the other hand,
vascularity could play a role in the progression of
POHGs. Actually, a high tumor microvessel
density (MVD, > 21 vessels/1.2 mm2) was found
to be associated to a significantly higher rate of
progression than a MVD < 21 vessels/1.2 mm2
(84% versus 35%) (Bartels et al. 2006). This
parameter was also predictive of reduced PFS on
multivariate analysis stratified for extent of surgical resection. Differently, angiogenesis does not
seem to influence the tumor progression as demonstrated by the poor correlation between immunostaining for VEGF/VEGFR and PFS.
Management
Histological Features
The tumor grade usually exerts an obvious influence on the prognosis but, once again, POHGs
The kind of treatment does not make a significant

difference in the prognosis of POHGs. In the
series of Ahn et al. (2006), where the attitude was
to perform a surgical removal as much extended
302
as possible, the 5-year OS and PFS was 93.6%

and 52.4%, respectively. Similarly, the prospective study realized by the French Society of
Pediatric Oncology by using a pure chemotherapy protocol showed a 89% 5-year OS and a 34%
PFS (Laithier et al. 2003). Moreover, in the
retrospective RT-based series of Khafaga et al.
(2003), 5-year OS and PFS was 87.5% and 69%,
respectively. These examples clearly indicate that
prognosis of POHGs does not depend on a single
treatment but on the integration of the different
therapeutic options (that was implicit in each of
the previous series). Moreover, the concurrent
role played by the other prognostic factors has to
be taken into account. For example, surgicallyamenable POHGs, as Dodge I tumors, show a
very good 10-year OS (~100%) but, of course,
this is the result of a favorable tumor location
other than of the surgical excision.
The main difference about prognosis and outcome among the different treatments consists in
the burden of their correlated complications.
Consequently, surgery is currently limited to
biopsy or to partial removal of huge POHGs
needing acute decompression, while RT is preferentially used in old children with progressing
tumors or as rescue therapy. CT is considered the
best first line treatment for progressing POHGs
since it combines good survival rates and relatively low complication rates. Moreover, CT
alone has been proved to provide the same PFS as
when it is used in combination with surgery
(Nicolin et al. 2009).
Statistical Value
A systematic review of the literature to assess the
role of different prognostic factors in the progression of POHGs was carried out by Opocher et al.
(2006). Twenty-three articles were considered for
this analysis even though many of them had
important methodological limitations based on
the authors evaluation criteria. A statistical
multivariate analysis was done in 11 out of these
23 articles. According to this review, only the
age < 1 year at diagnosis was an independent
prognostic factor for tumor progression, this
L. Massimi
result being validated by 3 high-quality studies.

None of the other proposed prognostic factors,
such as sex, race, NF-1, tumor size, tumor
location, histology, presence of neurologic deficits at diagnosis, presence of hydrocephalus,
extent of surgery, RT, doses of RT, and CT
reached a solid evidence indicating a possible
relationship with the progression of POHGs.
Among them, however, NF-1 was associated with
a better PFS (compared with non-affected children) in three studies with multivariate analysis
(only one of them was a high-quality study).
Tumor location was found to affect the PFS in two
studies with multivariate analysis, though both
were burdened by methodological limitations.
Early Outcome and Complications

of Treatments
Mortality after surgery for POHGs is currently
almost nil (Ahn et al. 2006). This result is
explained by two main factors. The first one is
represented by the progresses in the surgical,
anesthesiological, and intensive care management and technologies that make neurosurgical
operations more and more safe. The second reason of such a low surgical mortality depends on
the decreased number of operations currently
performed (chemotherapy actually represents the
first line treatment in many Centers), which often
consist in relatively low-risk procedures, as
biopsy or partial tumor removal. In spite of these
figures, surgery-related morbidity may be consistent. It usually varies according to the age of the
patient, the extent of the surgical resection, and
the tumor location. Postoperative acute worsening of the visual function is the most frequent
complication, being reported in up to 3575% of
the cases, followed by endocrine dysfunction
(from 15% up to 60%), neurological deficits as
cranial nerves palsy, hemiparesis, dysphasia,
seizures (10%) and early neuro-psychological
sequelae (Massimi et al. 2007).
Hydrocephalus often complicates the clinical
picture of POHGs involving chiasm and hypothalamus due to the invasion of the third ventricle.
Therefore, ventriculo-peritoneal shunt is frequently
28
performed in this subset of patients (endoscopic

third ventriculostomy cannot be realized because
of the third ventricle involvement). The complications of CSF shunting devices in POHGs, as
obstruction, infection, or migration, do not differ
from those of children shunted for other reasons.
However, a peculiar complication, that is ascites,
can occur. Ascites actually arises from the abnormal accumulation of endoperitoneal fluid resulting
from raised protein content in the CSF due to the
secreting activity of POHGs.
Concerns about the use of RT in POHGs are
raised by the young age of the patients to deal
with and the usually large irradiation fields
required by this kind of tumors. The most feared
complication is represented by the possible
impairment of cognitive development. In their
study specifically addressing this issue, Lacaze
et al. (2003) found that cognitive outcome in children with POHGs was preserved when only CT
was administered (mean IQ: 107 17) but it was
reduced when also RT was performed (mean IQ:
88 24), although children receiving RT were
those with the worst clinic-radiological picture
since the diagnosis. A further and most severe
complication of RT is represented by radiationinduced malignancies, usually high grade
gliomas, which is found to burden patients with
POHGs in up to 10% of the cases (Marcus et al.
2005). Nevertheless, a strict correlation between
radio-induced gliomas and extension of irradiation field is not ever demonstrated (Kortmann
et al. 2003) so that the role of different etiologic
factors should be hypothesized in some instances.
Moreover, it has to be considered that also CT
may have carcinogenic effects, being able to
induce both hematologic and solid tumors
(Jahraus and Tarbell 2006).
Specific complications resulting from the irradiation of the suprasellar region are endocrine
imbalances, mainly growth hormone deficit and
hypogonadotrophic hypogonadism, followed by
impairment of visual function, Moya-Moya disease and/or other cerebral vasculopaties, and
cerebrovascular complications. Some authors
reported a higher risk of these complications in
children with NF-1 (Lacaze et al. 2003; Osztie
et al. 2001), thus reinforcing the indication to a
303
wait-and-see policy in this subgroup of patients.

It is worth noting that, currently, the role of RT in
the genesis of these complication is made smaller
by three main factors: (1) The responsibility of
surgery and CT in the development of some
RT-related complications (e. g. neuropsychological sequelae, second malignancies) as well as
the role played by NF-1 in some other of them
(e. g. development of de novo malignant gliomas
or Moya-Moya disease in non-irradiated children);
(2) The current use of RT mainly as rescue treatment in several protocols, which allows children
to receive radiations in older ages; (3) The development of modern 3D conformal techniques
(stereotactic radiosurgery, intensity-modulated
RT, proton beam RT) that makes RT much safer
than in the past.
The advantages of CT mainly result from its feasibility in the youngest children and the lower rate
of late complications compared with RT and surgery, in particular concerning the neuropsychological sequelae. However, the effects of CT on some
kind of neurological complications in POHGs, such
as cognitive disorders and seizures resulting from
leukoencephalopathy and cerebral infarctions,
cannot be excluded. Myelotoxicity (neurtropenia,
thrombocytopenia), electrolytic imbalance (mainly
hyponatremia), renal failure, cisplatin-related ototoxicity, and intollerance to some chemotherapic
drugs are the most common complications of CT.
Long-Term Clinical Outcome

Late clinical outcome in POHGs is relatively
poor if compared to the overall good survival
rates because of the negative effects of delayed
tumor progression and the sequelae of treatments.
Data on prognosis and outcome in the pediatric
series of Tow et al. (2003) are reported in
Table 28.3 as summarizing example.
The visual impairment is the most challenging
and discouraging aspect of POHGs management
other than one of the most important for the
patients quality of life. Although visual acuity can
be preserved in more than a half of the children,
being improved in 1520% and stabilized in about
4045%, a spontaneous or treatment-related
L. Massimi
304
Table 28.3 Differences in prognosis and outcome (from Tow et al. 2003)
N. of children
Mean age
NF-1 patients
Treated patients
Treatment
Mean follow-up
Survival
Visual outcome
Other outcomes
Optic nerve gliomas

(Dodge I)
16 (3 bilateral tumors)
8.7 years
50%
56% (9 cases)
Surgery, RT
15.3 years
100%
Stability or improvement, mean acuity 20/40
Seizures and moyamoya
in 1 case
Optic chiasm gliomas

(Dodge II)
16
7.1 years
63%
63% (10 cases)
RT (CT in 1 case)
15.7 years
94%
Stability in the good eye and
worsening in the bad eye, mean
acuity 20/40 in 80% of cases
New endocrine and/or
neurological deficits in 8 cases
(50%)
worsening is detected in 3035% of cases. At late

follow-up, nearly 20% of the patients are bilaterally blind or show severe bilateral visual impairment, while a severe unilateral deficit can involve
up to 30% of cases. The young age and the Dodge
III type are associated to a poor visual outcome
as well as severe optic pallor at diagnosis or during follow-up (Campagna et al. 2010). A good
clinico-radiological correlation between response
of the tumor to treatments and visual outcome is
not ever observed (Shofty et al. 2011). The visual
performance at diagnosis correlates with the final
outcome only in Dodge I tumors. The trend of the
visual function does not correlate with the survival. NF-1 seems to predict a high risk of delayed
visual loss. In the series of 20 NF-1 children
described by Dalla Via et al. (2007), 65% of cases
met the WHO criteria of hypovision and 40% had
a visual acuity less than 20% in both eyes after a
median follow-up more than 6 years.
The visual field approximately follows the
same trend than visual acuity, being stabilized in
most of cases (5560%), improved in 1520% of
cases, and worsened in about 35% of patients
(Ahn et al. 2006; Grabenbauer et al. 2000).
However, a normal, bilateral visual field is
observed in a minority of children (nearly 10%)
al late follow-up (Campagna et al. 2010). The
majority of them, indeed, presents a unilateral
Optic chiasm/hypothalamus gliomas

(Dodge III)
15
5.2 years
27%
87% (13 cases)
Surgery, RT, CT
13.5 years
73%
Stability only in non-treated subjects,
mean acuity 20/40 in < 50% of cases
High neurological morbidity
(hemiparesis, seizures, frontal lobe
syndrome), endocrine imbalance in
82% of cases
or bilateral visual field amputation (mainly

homonymous hemianopsy followed by quadrantopsia and concentric reduction).
The role of the type of treatment on the visual
outcome, on the other hand, is still quite poorly
understood. Actually, both surgery and RT are
thought to be able to stabilize or even improve
the visual function over the time (RT more than
surgery) but both of them are also reported to
cause permanent acute (surgery) or delayed visual
damages (RT) (Ahn et al. 2006; Grabenbauer
et al. 2000; Khafaga et al. 2003). Instead, CT is
generally considered the best treatment to stabilize or improve the visual outcome. However, in
the meta-analysis of the literature of the past
20 years performed by Moreno et al. (2010), only
8 among 85 articles were eligible to investigate
the effects of CT on the visual outcome and all of
them were class IV papers. Based on this analysis,
which concerns 174 children, the CT-related
improvement of the visual function was poor
(14.4% of cases, range: 045%) while the stabilization regarded about 47% of cases (range: 27100%).
However, no information on the duration of the
effects on the visual stabilization was reported so
that no conclusion on possible advantages of CT
over the other treatments could be taken. These
figures have been recently confirmed by other
authors (Shofty et al. 2011).
28
A further important aspect of late outcome

concerns the neuropsychological development.
Current management allows to preserve the
cognitive functions in about 6065% of children
that, consequently, are able to attend regular
schools. As mentioned before, the neuropsychological deficits affecting the remaining 3540%
of subjects result not only from the sole RT
administration but also from the poor clinical
conditions of some children receiving RT, from
the too young age at irradiation (<5 years) of
some of them, and from the worse clinical course
of POHGs in infants. Moreover, as demonstrated
by Carpentieri et al. (2003) in a series composed
by 106 children with brain tumor investigated
before RT administration, the alterations of the
performance IQ in these patients clearly point out
the presence of a pre-irradiation intellectual
impairment as result of the tumor growth and the
surgical operation. Other authors have afterwards
confirmed these findings, proposing the tumor as
an important risk factor even for preoperative
cognitive impairment in neuro-oncological
children (Di Rocco et al. 2011). Patients managed without RT have a better neuropsychological outcome when treated by CT alone compared
with surgery and CT, the verbal IQ and the full IQ
being more affected by surgery than the visuospatial IQ (Nicolin et al. 2009). Children requiring treatment for hydrocephalus usually show the
worst neuropsychological development while no
differences are detected between non-NF-1 and
NF-1 patients.
Similar considerations can be done for the socalled CT-related neurological complications.
Indeed, though some chemotherapeutic drugs
(namely, platinum compounds) are proved to
induce late neuro-toxicity, it could also result
from the additional effects of RT and surgery.
Surgery alone is actually responsible of about
50% of severe neurological disability as result of
acute postoperative deficits (Nicolin et al. 2009).
Therefore, the exact mechanisms for CT-induced
neuro-toxicity still remain partially unknown. An
interesting hypothesis involves both a genetic
predisposition to develop neurological disorders
(e.g., low efficiency in DNA repair, deregulated
immune response) and some deleterious effects
305
of CT in the genesis of these deficits (Ahles and

Saykin 2007).
Endocrine dysfunctions are probably the most
common reason of abnormal outcome in children
harboring POHGs. Hormone imbalances are
found in 3070% of cases, involving up to
90100% of the patients in case of hypothalamic
tumor invasion (Grabenbauer et al. 2000). As for
the neurological deficits, endocrine function
strictly correlates with the tumor location and the
type of treatment. Moreover, endocrine impairment is demonstrated in up to 35% of patients
before that any treatment is started (Khafaga et al.
2003) so that its trend may be affected simply by
the tumor growth/progression. Growth hormone
deficiency, precocious or delayed puberty, hypothyroidism, pan-hypopituitarism, insipid diabetes, diencephalic syndrome with failure to thrive
are the most common sequelae of such a progression. Apart from the diencephalic syndrome,
which usually shows a significant improvement
after therapy, all the other endocrine problems
tend to worsen after treatment. For this reason,
many patients need chronic hormone replacement in their life, and fail to reach normal weight
and height values.
References
Ahles TA, Saykin AJ (2007) Candidate mechanisms for
chemotherapy-induced cognitive changes. Nat Rev
Cancer 7:192201
Ahn Y, Cho BK, Kim SK, Chung YN, Lee CS, Kim IH,
Yang SW, Kim HS, Kim HJ, Jung HW, Wang KC
(2006) Optic pathway glioma: outcome and prognostic factors in a surgical series. Childs Nerv Syst
22:11361142
Bartels U, Hawkins C, Jing M, Ho M, Dirks P, Rutka J,
Stephens D, Bouffet E (2006) Vascularity and angiogenesis as predictors of growth in optic pathway/
hypothalamic gliomas. J Neurosurg 104:314320
Bredel M, Slavc I, Birner P, Haberler C, Strobel T, Budka
H, Rossler K, Hainfellner JA (2002) DNA topoisomerase IIalpha expression in optic pathway gliomas
of childhood. Eur J Cancer 38:393400
Caldarelli M, Pezzotta S (1999) Optic pathway and hypothalamic tumors. In: Choux M, Di Rocco C, Hockley
AD, Walker ML (eds) Pediatric neurosurgery.
Churchill Livingstone, London, pp 509529
Campagna M, Opocher E, Viscardi E, Calderone M,
Severino M, Cermakova I, Perilongo G (2010) Optic
306
pathway glioma: long-term visual outcome in children
without neurofibromatosis type 1. Pediatr Blood
Cancer 55:10831088
Carpentieri SC, Weber DP, Pomeroy SL, Scott RM,
Goumnerova LC, Kieran MW, Billet AL, Tarbell NJ
(2003) Neuropsychological functioning after surgery
in children treated for brain tumor. Neurosurgery
52:13481357
Combs SE, Shulz-Ertner D, Moschos D, Thilmann C,
Huber PE, Debus J (2005) Fractionated sterotactic
radiotherapy of optic pathway gliomas: tolerance and
long term outcome. Int J Radiat Oncol Biol Phys
62:814819
Dalla Via P, Opocher E, Pinello ML, Calderone M,
Viscardi E, Clementi M, Battistella PA, Laverda AM,
Da Dalt L, Perilongo G (2007) Visual outcome of a
cohort of children with neurofibromatosis type 1 and
optic pathway glioma followed by a pediatric neurooncology program. Neuro-oncology 9:430437
Di Rocco C, Chieffo D, Frassanito P, Caldarelli M,
Massimi L, Tamburrini G (2011) Heralding cerebellar
mutism: evidence for presurgical language impairment
as primary risk factor in posterior fossa tumors.
Cerebellum. doi:10.1007/s12311-011-0273-2
Driever PH, von Hornstein S, Pietsch T, Kortmann R,
Warmuth-Metz M, Emser A, Gnekow AK (2010)
Natural history and management of low-grade glioma
in NF-1 children. J Neurooncol 100:199207
Dutton JJ (1994) Gliomas of the anterior visual pathway.
Surviv Ophtalmol 38:427452
Fouladi M, Wallace D, Langston JW, Mulhern R, Rose
SR, Gajjar A, Sanford RA, Merchant TE, Jenkins JJ,
Kun LE, Heideman RL (2003) Survival and functional
outcome in children with hypothalamic/chiasmatic
tumors. Cancer 97:10841092
Gayre GS, Scott IU, Feuer W, Saunders TG, Siatowski
RM (2001) Long-term outcome in patients with anterior visual pathway gliomas. J Neuroophtalmol
21:17
Gnekow AK, Kortmann RD, Pietsch T, Emser A
(2004) Low grade chiasmatic- hypothalamic glioma
carboplatin and vincristine chemotherapy effectively
defers radiotherapy within a comprehensive treatment strategy report from the multicenter treatment
study for children and adolescents with a low grade
glioma HIT-LGG 1996 of the Society of Pediatric
Oncology and Hematology (GPOH). Klin Padiatr
216:331342
Grabenbauer GG, Schuchardt U, Buchfelder M, Rodel
CM, Gusek G, Marx M, Doerr HG, Fahlbusch R, Huk
WJ, Wenzel D, Sauer R (2000) Radiation therapy of
optico-hypothalamic gliomas (OHG) radiographic
response, vision and late toxicity. Radiother Oncol
54:239245
Grill J, Laithier V, Rodriguez D, Raquin MA, Pierre-Kahn
A, Kalifa C (2000) When do children with optic pathway tumours need treatment? An oncological perspective in 106 patients treated in a single centre. Eur
J Paediatr 159:692696
L. Massimi
Guillamo SJ, Crange A, Kalifa C, Grill J, Rodriguez D,
Doz F, Barbarot S, Zerah M, Sanson M, Bastuji-Garin
S, Wolkenstein P, Rseau NF (2003) Prognostic
factors of CNS tumors in neurofribomatosis 1 (NF-1).
A retrospective study of 104 patients. Brain
126:152160
Gururangan S, Cavazos CM, Ashley D, Herndon JE,
Bruggers CS, Moghrabi A, Scarcella DL, Watral M,
Tourt-Uhlig S, Reardon D, Friedman HS (2002) Phase
II study of carboplatin in children with progressive
low-grade gliomas. J Clin Oncol 20:29512958
Hsu TR, Wong TT, Chang FC, Ho DM, Tang RB, Thien
PF, Chang KP (2008) Responsiveness of progressive
optic pathway tumors to cisplatin-based chemotherapy
in children. Childs Nerv Syst 24:14571461
Jahraus CD, Tarbell NJ (2006) Optic pathways gliomas.
Jaing TH, Lin KL, Tsay PK, Hsueh C, Hung PC, Wu CT,
Tsen CK (2008) Treatment of optic pathway hypothalamic gliomas in childhood: experience with 18 consecutive cases. J Pediatr Hematol Oncol 30:222224
Khafaga Y, Hassounah M, Kandil A, Kanaan I, Allam A,
El Husseiny G, Kofide A, Belal A, Al Shabanah M,
Schultz H, Jenkin D (2003) Optic gliomas: a retrospective analysis of 50 cases. Int J Radiat Oncol Biol
Phys 56:807812
Komotar RJ, Burger PC, Carson BS, Brem H, Olivi A,
Goldthwaite PT, Tihan T (2004) Pilocytic and pilomyxoid
hypothalamic/chiasmatic
astrocytomas.
Kortmann RD, Timmermann B, Taylor RE, Scarzello G,
Plasswilm L, Paulsen F, Jeremic B, Gnekow AK,
Dieckmann K, Kay S, Bamberg M (2003) Current and
future strategies in radiotherapy of childhood lowgrade glioma of the brain. Part I: treatment modalities
of radiation therapy. Strahlenther Onkol 179:509520
Lacaze E, Kieffer V, Streri A, Lorenzi C, Gentaz E,
Habrand JL, Dellatolas G, Kalifa C, Grill J (2003)
Neuropsychological outcome in children with optic
pathway tumours when first-line treatment is chemotherapy. Br J Cancer 89:20382044
Laithier V, Grill J, Le Deley MC, Ruchoux MM, Couanet
D, Doz F, Pichon F, Rubie H, Frappaz D, Vannier JP,
Babin-Boilletot A, Sariban E, Chastagner P, Zerah M,
Raquin MA, Hartmann O, Kalifa C (2003) Progressionfree survival in children with optic pathway tumors:
dependence on age and the quality of the response to
chemotherapy results of the first French prospective
study for the French Society of Pediatric Oncology.
Mahoney DH, Cohen ME, Friedman HS, Kepner JL,
Gemer L, Langston JW, James HE, Duffner PK, Kun
LE (2000) Carboplatin is effective therapy for young
children with progressive optic pathway tumors: a
Pediatric Oncology Group phase II study. Neurooncology 2:213220
Marcus KJ, Goumnerova L, Billett AL, Lavally B, Scott
RM, Bishop K, Xu R, Young Poussaint T, Kieran M,
Kooy H, Pomeroy SL, Tarbell NJ (2005) Stereotactic
28
radiotherapy for localized low-grade gliomas in children:

final results of a prospective trial. Int J Radiat Oncol
Biol Phys 61:374379
Massimi L, Tufo T, Di Rocco C (2007) Management of
optic-hypothalamic gliomas in children: still a challenging problem. Expert Rev Anticancer Ther
7:15911610
Massimino M, Spreafico F, Cefalo G, Riccardi R, TesoroTess JD, Gandola L, Riva D, Ruggiero A, Valentini L,
Mazza E, Genitori L, Di Rocco C, Navarria P, Casanova
M, Ferrari A, Luksch R, Terenziani M, Balestrini MR,
Colosimo C, Fossati-Bellani F (2002) High response
rate to cisplatin/etoposide regimen in childhood
low-grade glioma. J Clin Oncol 20:42094216
Moreno L, Bautista F, Ashley S, Duncan C, Zacharoulis S
(2010) Does chemotherapy affect the visual outcome
in children with optic pathway glioma? A systematic
review of the evidence. Eur J Cancer 46:22532259
Muzi M, Spence AM, OSullivan F, Mankoff DA, Wells
JM, Grierson JR, Link JM, Krohn KA (2006) Kinetic
analysis of 3-deoxy-3-18F-fluorothymidine in
patients with gliomas. J Nucl Med 47:16121621
Nicolin G, Parkin P, Mabbott D, Hargrave D, Bartels U,
Tabori U, Rutka J, Buncic JR, Bouffet E (2009) Natural
history and outcome of optic pathway gliomas in
children. Pediatr Blood Cancer 53:12311237
Oh KS, Hung J, Robertson PL, Garton HJ, Muraszko KM,
Sandler HM, Hamstra DA (2011) Outcomes of multidisciplinary management in pediatric low-grade
gliomas. Int J Radiat Oncol Biol Phys. doi:10.1016/j.
ijrobp. 2011.01.019
Opocher E, Kremer LCM, Da Dalt L, van de Wetering
MD, Viscardi E, Caron HN, Perilongo G (2006)
Prognostic factors for progression of childhood opticpathway glioma: a systematic review. Eur J Cancer
42:18071816
Osztie E, Vrallyay P, Doolittle ND, Lacy C, Jones G,
Nickolson HS, Neuwelt EA (2001) Combined intraarterial carboplatin, intraarterial etoposide phosphate,
and IV cytoxan chemotherapy for progressive
optic-hypothalamic gliomas in young chidlren. AJNR
307
Sawamura Y, Kamoshima Y, Kato T, Tajima T, Tsubaki J

(2009) Chemotherapy with cisplatin and vincristine
for optic pathway/hypothalamic astrocytoma in young
children. Jpn J Clin Oncol 39:277283
Segal L, Darvish-Zargar M, Dilenge ME, Ortenberg J,
Polomeno R (2010) Optic pathway gliomas in patients
with neurofibromatosis type 1: follow-up of 44
patients. J AAPOS 14:155158
Shofty B, Ben-Sira L, Freedman S, Yalon M, Dvir R,
Weintraub M, Toledano H, Constantini S, Kesler A
(2011) Visual outcome following chemotherapy for
progressive optic pathway gliomas. Pediatr Blood
Cancer 57:481485
Shuper A, Kornreich L, Michowitz S, Hoerv G,
Schwarz M, Weitz R, Zaizov R, Cohen IJ (1999)
Visual pathway tumor: a heterogeneous tumor with
a variable clinical course. Pediatr Hematol Oncol
16:407414
Steinbok P, Hentschel S, Almqvist P, Cochrane DD,
Poskitt K (2002) Management of optic chiasmatic/
hypothalamic astrocytomas in children. Can J Neurol
Sci 29:132138
Stokland T, Liu JF, Ironside JW, Ellison DW, Taylor R,
Robinson KJ, Picton SV, Walker DA (2010) A multivariate analysis of factors determining tumor progression in childhood low-grade glioma: a population-based
cohort study (CCLG CNS9702). Neuro-oncology
12:12571268
Surez JC, Viano JC, Zunino S, Herrera EJ, Gomez J,
Tramunt B, Marengo I, Hiramatzu E, Miras M, Pena
M, Sonzini Astudillo B (2006) Management of optic
pathway gliomas: new therapeutical option. Childs
Nerv Syst 22:679684
Tow SL, Chandela S, Miller NR, Avellino AM (2003)
Long-term outcome in children with gliomas of the
anterior
visual
pathway.
Pediatr
Neurol
28:262270
Yazici N, Varan A, Akalan N, Soylemezoglu F, Zorlu F,
Kutluk T, Akyuz C, Buyukpamuku M (2011)
Diencephalic tumors in children: a 30-year experience
of a single institution. Childs Nerv Syst
27:12511256
Pediatric Low-Grade Gliomas:

Advantage of Using Lower Doses
of Cisplatin/Etoposide
29
Maura Massimino, Veronica Biassoni,

and Elisabetta Schiavello
Contents
Abstract
Introduction ............................................................
310
Patients Eligibility
and Chemotherapy Regimen ................................
310
Evaluation of Response..........................................
310
Evaluation of Toxicity ............................................
311
Statistical Analysis .................................................
311
Results .....................................................................
311
Toxicity ....................................................................
315
Discussion................................................................
315
References ...............................................................
318
M. Massimino (*) V. Biassoni E. Schiavello

Pediatric Unit, Fondazione IRCCS Istituto Nazionale
Tumori, Via Venezian 1, 20133 Milan, Italy
e-mail: Maura.massimino@istitutotumori.mi.it
In this chapter we describe the experience

gained using cisplatin and etoposide in childhood progressive low-grade glioma. After
successfully adopting this combination in ten
3-day courses for progressive low-grade
gliomas, the subsequent protocol, reducing
the daily doses of cisplatin and etoposide,
aimed at achieving the same response and
3-year PFS rates with lower neurotoxicity and
myelotoxicity. In this second series we have
treated 37 pts; 9 were metastatic cases.
Treatment was prompted by radiological
evidence of progression and/or clinical deterioration. A MRI volume reduction was appreciable in 65%; response was maximal
12 months after starting treatment. The 3-year
EFS and OS rates were 65% and 97%,
respectively for the whole series. Clinical,
neurological or functional improvements
were seen in 26/37 cases. Audiological toxicity caused alterations in 4/34 cases. The
previous protocol had achieved volume reductions in 70% of cases, causing audiological
damage (data updated) in 11/31 (p = 0.023),
with 3-year PFS and OS rates of 70% and
100%, respectively. The updated 3-year PFS
and OS rates were 85% and 96%, respectively,
for those children treated in Milan. Lower
doses of cisplatin/etoposide are still effective
in progressive low-grade glioma, with less
acute and persistent morbidity.

309
M. Massimino et al.
310
Introduction
For 20 years now, chemotherapy has been used in
children with progressive low-grade glioma
(LGG) not amenable to further surgery, or with
large symptomatic tumors, with a view to delaying radiotherapy, especially in younger patients
with large tumors or diencephalic syndrome
(Packer et al. 1993; Gajjar et al. 1997; Petronio
et al. 1991). Many schedules have been described
so far, mainly including carboplatin, vincristine,
etoposide, nitrosoureas, cisplatin, actinomycin-D,
cyclophosphamide, procarbazine and 6-thioguanine
(Packer et al. 1993; Gajjar et al. 1997; Petronio
et al. 1991; Laithier et al. 2003), and, more
recently, also vinblastine (Lafay-Cousin et al.
2005), and imatinib (McLaughlin et al. 2003).
Carboplatin alone and carboplatin containing
regimens have been the most widely used due to
the response obtained and the low neurotoxicity
even if cisplatin based schedules have induced
higher number of responses (Gajjar et al. 1993;
Hsu et al. 2008) with the prize to be paid to potential
higher nephro- and oto-toxicities.
Having obtained satisfactory results with a
regimen of cisplatin (30 mg/m/d) and etoposide
(150 mg/m/d) in ten 3-day courses (Massimino
et al. 2002), we launched a new institutional protocol with lower daily doses of both drugs. The
design of the protocol was a non-inferiority study
whose primary endpoint was to maintain the
same rates of response and 3-year progressionfree survival with a lower neurotoxicity (based on
audiograms or other age-tailored acoustic tests)
and myelotoxicity. We report the results obtained
in this new series of patients.
Patients Eligibility
and Chemotherapy Regimen
All children under 18 years of age with radiological
or symptomatic evidence of progressive lowgrade glioma not amenable to surgery were
eligible for treatment irrespective of the patients
age or the tumors origin and dissemination. A
histological assessment was required except in
cases of optochiasmatic tumors in NF1, or if

surgery was considered dangerous and the clinical presentation and radiological images were
clearly indicative of a low-grade glioma. Patients
already treated with other chemotherapy regimens or radiotherapy were not included for the
purposes of this analysis. Cisplatin was given as
a 2-h infusion at 25 mg/m2/d on days 13, then
etoposide was infused in 30 min at a dose of
100 mg/m2/d, again on days 13, after the cisplatin infusion; the drugs were preceded and followed by a 2-h hydration. In children under 1 year
old or weighing less than 10 kg, the doses were
calculated according to their weight. The intervals between the first four cycles were 4 weeks
long, then the next three cycles were at 5-week
intervals, and the remaining three cycles were at
6-week intervals, the aim being to cover a period
of approximately 1011 months, thus delivering
ten cycles in all with a total dose for cisplatinum
of 750 mg/m2 and for etoposide of 3,000 mg/m2.
The protocol was approved by our institutions
ethical and scientific committee. Formal consent
was obtained from the patients parents. The
therapy was administered in all children on an
outpatient basis, except for the first course, which
was given during a hospital stay if the patient was
fitted with a central venous catheter.
Evaluation of Response
Initial staging included magnetic resonance
imaging (MRI) of the brain and spine in all
patients. When multiple sites of disease emerged
at MRI, a cytological examination of the spinal
fluid was performed. The follow-up during the
treatment was based on neurological and clinical
assessments before each chemotherapy cycle,
with MRI of the tumor site after the first three
cycles and then every 3 months thereafter for the
first year during treatment. Tumor volume was
calculated from both T2-weighted and enhanced
T1-weighted MRI scans; thorough neurological
and clinical assessments were repeated every
4 months for the first 3 years after completing the
treatment, then every 6 months up to the fifth
year, then yearly. In the event of neurofunctional
29
Pediatric Low-Grade Gliomas: Advantage of Using Lower Doses of Cisplatin/Etoposide
impairment, specific tests (visual field and acuity,

visual-evoked potentials [VEP]) were prescribed
at the same time as the radiological evaluation.
Visual field assessments and visual-evoked potentials were performed at 6-month intervals,
wherever feasible, together with an audiometric
evaluation (vocal, behavioral audiometry, auditory
brainstem responses or acoustic oto-emission
testing, depending on the patients age) during
the first 5 years after the glioma was diagnosed.
Endocrinological tests were also performed routinely and repeated if a patient showed clinical
signs of any endocrinological alterations.
Radiological response was assessed according to
International Society of Paediatric Oncology criteria and protocols, i.e. complete response was
defined as no evidence of disease, partial response
meant a radiographically evident reduction in
size of more than 50%, and stable disease meant
no signs of tumor progression. We also considered objective response: a reduction in size of
unequivocal residual tumor manifestation
between 50% and 25% (product of the two largest perpendicular diameters) (Gnekow 1995).
Stable disease was considered a positive response
and chemotherapy was consequently continued.
All patients were still being actively followed up
at the time of this report.
Evaluation of Toxicity
Audiotoxicity was monitored in all patients during
treatment with full tone audiometry, acoustic
potentials or acoustic oto-emission testing
according to age, and thereafter, after treatment
end, as already described in the previous paragraph. When full tone audiometry was feasible,
stopping rules for toxicity during treatment
followed those commonly adopted for medulloblastoma when applying a platinum-containing
regimen, with an evaluation before every course
in children with NF1 and after two to three
chemotherapy courses in children without NF1.
In case of a loss of perception over 40 dB at 4,000
8,000 Hz, the cisplatinum was to be replaced by
carboplatinum given at a dose of 400 mg/m2 day 1
of the subsequent course. In absence of standard
311
criteria for stopping chemotherapy after other

ototoxicity evaluation exams, all other patients
were evaluated on a case-by-case base for treatment continuation.
Life tables were obtained for all patients. Overall
survival and event-free survival were evaluated
according to the Kaplan-Meier method (Kaplan
and Meyer 1958). Class comparisons were
obtained with the log-rank test, considering the
time of response from the date of starting chemotherapy, and the time of relapse as when MRI
revealed tumor growth, or a clinical deterioration
became apparent, whichever came first.
Results
Patient accrual lasted from November 2001 to
December 2007, enrolling 22 males and 15
females with a median age at the time of treatment of 72 months (range 6198 months). The
median follow-up as at the time of writing this
report was 48 months. Main demographic features are shown in Table 29.1. When chemotherapy was begun, all children suffered from
radiologically-evident progressive disease and/or
worsening symptoms; the median time elapsing
from the tumors diagnosis to starting treatment
was 18 months (range 1 month 10 years). Seven
children had neurofibromatosis type 1 (NF1) and
one other child had an unknown neurocutaneous
syndrome with cervico-medullary glioma,
spinal and subcutaneous lipomas and mental
retardation; all the other patients had sporadic
tumors. The tumor originated in the optochiasmaticdiencephalic region in 23 children: 22 had a
Dodge 3 extension and one had a Dodge 1 tumor.
Five children had tumors in the cervico-medullary
region, while 5 gliomas were mesencephalothalamic, and 2 were multinodular with no clear
primary site, 1 glioma occupied the whole
spine and 1 was temporal. In all, nine children
presented with nodular metastases and one of them
had a diagnosis of NF1. All patients with metastases
M. Massimino et al.
312
Table 29.1 Demographic features of the series
Patient
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Sex
m
m
m
f
f
f
m
m
m
f
m
m
f
f
f
m
m
m
f
m
m
m
m
f
m
m
f
m
m
m
m
f
m
f
f
f
f
Age
7.06
7.00
10.08
3.09
1.06
9.03
6.07
0.07
7.00
13.00
6.00
0.06
14.00
0.05
12.00
2.00
0.06
6.00
11.00
2.00
6.00
7.00
1.10
6.00
12.06
7.10
6.05
6.00
13.02
2.06
7.00
6.00
1.04
7.10
4.02
1.00
4.01
Histology
Fibrillary
Pilocytic
Fibrillary
Pilocytic
Pilomixioid
Ganglioglioma
Pilocytic
Ganglioglioma
Pilocytic
Pilocytic
Pilocytic
Pilocytic
Pilocytic
Ganglioglioma
Pilocytic
Pilocytic
Pilocytic
Pilocytic
Pilocytic
Pilocytic
Pilocytic
Pilocytic
Pilomixoid
Pilocytic
NF1
y
y
y
n
y
n
n
n
n
n
n
n
n
n
n
n
n
y
n
n
y
n
n
n
n
n
n
n
n
n
n
n
n
n
n
n
y
Site
Opto-chiasm atic
Multi-nodular
Opto-chiasm atic
Mesencephalo-thalamic
Opto-chiasm atic
Cervico-medullary
Opto-chiasm atic
Opto-chiasm atic
Opto-chiasm atic
Opto-chiasm atic
Opto-chiasm atic
Opto-chiasm atic
Temporal
Opto-chiasm atic
Cervico-medullary
Opto-chiasm atic
Opto-chiasm atic
Multi-nodular
Opto-chiasm atic
Opto-chiasm atic
Opto-chiasm atic
Opto-chiasm atic
Opto-chiasm atic
Cervico-medullary
Opto-chiasm atic
Opto-chiasm atic
Cervico-medullary
Opto-chiasm atic
Pan-medullary
Opto-chiasm atic
Cervico-medullary
Opto-chiasm atic
Opto-chiasm atic
Meta
n
y
n
n
n
n
n
n
n
n
n
y
n
y
n
n
y
n
n
y
n
n
n
y
n
n
n
y
n
y
n
y
n
n
n
n
n
Legenda: y=yes; n=no
presented secondary nodules firstly evident in

brain and then confirmed also in spine MRI study,
but for one girl who had sciatalgic symptoms
prompting to spine MRI evaluation. Liquoral
seeding was never documented by cephalorachidian cytology. Twenty-four children underwent surgery, enabling a histological diagnosis in
all cases (partial tumor resections in 18 cases and
biopsies in 6). The centrally-reviewed histological diagnoses were: pilocytic astrocytoma in 17

cases, pilomyxoid astrocytoma in 2, ganglioglioma in 3, and fibrillary astrocytoma in 2.
Concordance of diagnoses was appreciated in
22/24 cases: two tumors were undergraded to
pilocytic astrocytoma after a previous diagnosis
of grade 3 astrocytoma. The other 13 children
29
were diagnosed on clinical grounds alone. Twelve

of 13 were affected by optochiasmatic tumors
while the last one was a 23 month-child affected
by a cervico-medullary tumor. He had a short history of rapid loss of limb strength and gait competence, that was sent urgently to primary
adjuvant treatment after multidisciplinary discussion concluding that the diagnosis of a low-grade
glioma was highly probable and that the clinical
improvement due to chemotherapy could overcome that of surgery that was quite risky. The
signs and symptoms prompting the diagnoses
were eyesight deficiencies in 23, pyramidal deficiencies in 15, diencephalic syndrome in 4, cranial nerve palsy in 2, monolateral proptosis in 1
and untreatable seizures in one. Other functional
alterations were precocious puberty in three
cases and panhypopituitarism in three, obesity
in one, and isolated growth hormone deficiency
in one; one child developed diabetes insipidus
after surgery.
After chemotherapy, radiology documented
tumor reduction in 24/37 cases, fulfilling the criteria for partial remission in 17 of them. Maximal
response so far was recorded a median 12 months
(range 231 months) after starting the treatment.
Table 29.2 shows how the main signs and symptoms evolved during and after treatment. As for
sight deficits we detail here in brief. Of the 23
313
children with optochiasmatic tumors, 17 could be

adequately evaluated with consistent examinations at the beginning and at the end of the chemotherapy schedule as above described: 7 were
improved, while the other 10 were stable. Of the
six remaining patients: 3 could only receive a
behavioral sight evaluation at the beginning of
their disease and were deemed ameliorated during and after treatment; 2 children were blind
after surgery and the last one received instrumental evaluation only at the end of treatment for lack
of compliance.
Three-year progression-free (PFS) and overall
survival (OS) were 66% 8.5% and 97% 3%,
respectively, for all patients (Fig. 29.1). For the
purpose of this report, we updated the follow-up
Table 29.2 Signs and symptoms outcome
Sight deficits
23
Dyencephalyc sy.
Proptosis
Pyramidal deficits
4
1
15
Cran nerves palsy

Seizures
As a total,
2
1
of 37
Improvement 7,
stable others
Improvement 4
Improvement 1
Improvement 12
Worse 2
Improvement 2
Improvement 1
Improvement 26
Stable 9
Worse 2
973%
Cum. Survival
.8
OS
65.58.5%
.6
PFS
.4
.2
0
Fig. 29.1 PFS and OS of all

series
12
24
36
48
Time
60
72
84
M. Massimino et al.
314
Fig. 29.2 PFS according to
metastases
Cum. Survival
.8
75
9%
39
17%
.6
.4
P = 0.014
.2
PFS of pts without metastases
PFS of pts with metastases
0
0
12
24
36
48
60
72
84
Time
on all 28 non pre-treated patients in our first

reported series, whose median follow-up was
thus 113 months: their 3-year PFS was 86% 7%
and their OS 96% 3.5% (P ns). If we compare
only the patients with no metastatic deposits, the
3-year PFS rates are 85% 7% and 75% 9% for the
first and the second series, respectively (P ns), and
the 3-year OS are 96.3% 4% and 96.3% 3%.
So far, 13 children have had adverse events,
occurring a median 26 months after starting the
treatment. Tumors progressed in 12 children and
one (a child with mental retardation and an
unidentified neurocutaneous syndrome) died of
an ab ingestis pneumonia while on treatment.
Another 4 children have died at 39, 49, 61 and
62 months after treatment beginning: two due to
tumor bleeding after re-surgery, one because of a
glioblastoma developed 44 months after irradiation for fibrillary astrocytoma progression, and
one during radiotherapy for rapid tumor progression. One other child developed a glioblastoma
14 months after irradiation for progressive fibrillary astrocytoma and is receiving second-line
chemotherapy. Tumor progression was local in
nine patients and in metastatic nodules in three,
all presenting with metastatic disease already at
diagnosis.
Children with metastatic disease had a significantly lower 3-year PFS than those without
metastases, i.e. 39% 17% in the former and
75% 9% in the latter (P = 0.0149) (Fig. 29.2).
Children with a clinical diagnosis alone had a
significantly higher 3-year PFS (92% 7%) than
those with a histological diagnosis (48% 12%)
(P = 0.0036). When patients were grouped according to their histological diagnoses, the 3-year PFS
was 52% 14% for patients with pilocytic astrocytoma, and 36% 20% for those with ganglioglioma, pilomixoid and fibrillary astrocytoma
(P = 0.004) (Fig. 29.3).
Tumor reduction also correlated with OS at
3 years, which was 100% for children achieving a
tumor reduction and 92% 8% for the others
(P = 0.066). No other factors influenced OS. Age
under 1 year or over 5 years, neurofibromatosis type
1, male or female gender, tumor origin in the optochiasmatic region, and debulking surgery as opposed
to biopsy had no statistical influence on PFS or OS.
It is however worth to underline that only one of the
seven children with NF1 relapsed so far and this
child had at diagnosis ubiquitarious tumor deposits.
Putting together the first and the second series, the
PFS at 3 years for 15 children with NF1 was
93% 6% and their OS at 3 years was 100%.
29
Fig. 29.3 PFS according to

histologies
92 7%
52 14%
.8
Cum. Survival
315
.6
.4
36 20%
.2
PFS of patients with clinical diagnoses
PFS of patients with pilocytic astro.
PFS ofpatientswithother histologies
0
0
12
Toxicity
The leukocyte nadir was always over 2.0 109/L
after the first course and was consequently not
checked again between courses. No infections
were documented. No child had to stop treatment
due to any toxicity. Audiological toxicity was
evaluated according to the Brock grading system
(Brock et al. 1991) after completing all planned
treatment in 34 children: grade 1 audiological
alterations were documented in one child, grade
0 in two, and acoustic potential deficiencies in
one; toxicity for the two patients graded with 0
meant a 20 dB loss at 4,000 Hz for both children.
In the first series, alterations had been recorded in
8/25 cases, classified as grade 1 in 5, grade 2 in 1
and grade 0 in 2 (p < 0.05); toxicity for the two
patients graded with 0 meant a 20 dB loss at
8,000 Hz for both children. All patients were
assessed by a physiatrist at diagnosis and tailored
rehabilitation was provided, as appropriate.
Endocrine tests were included in the diagnostic
work-up and in the subsequent follow-up. Growth
hormone replacement was prescribed according
to need, even in the case of residual disease if the
treatment was completed and the tumor remained
24
36
48
Time
60
72
84
stable for at least 912 months, or on the strength

of a case-by-case multidisciplinary discussion.
Discussion
In previous reports on the efficacy of chemotherapy in preventing the progression or inducing the
regression of pediatric low-grade gliomas, many
authors describe institutional, national or international trials using different drug combinations in
cases in which the site, diffusion or clinical
aggressiveness of the disease (i.e. rapid recurrence after surgery) were considered unamenable
to surgery alone, or when radiotherapy was considered too harmful in view of the tumors extent
and the patients young age (Packer et al. 1993;
Gajjar et al. 1997; Petronio et al. 1991; Laithier
et al. 2003). Numerous studies have been published in recent years addressing the chemotherapeutic management of low-grade gliomas as a
whole, a category into which most optic gliomas
would fit. Perhaps the foremost of these was published in two reports by Packer and colleagues
(1993, 1997) which included patients with newly
diagnosed progressive disease and patients with
316
recurrent disease. In them, the authors describe

the use of concurrent carboplatin and vincristine
in a 10-week induction phase, followed by
48 weeks of maintenance carboplatin/vincristine
on a slightly modified schedule for patients who
exhibited an objective radiographic response to
treatment and at least clinical stability in patients
who had stable clinical or neuroradiographic disease. This regimen resulted in a progression-free
survival of 75% at 2 years and 68% at 3 years.
Children aged 5 years or younger had a notably
more favorable overall rate of response. This is
presently the most commonly used regimen for
low-grade gliomas.
Given that most children presenting with evolutive features have tumors arising in visual pathways, causing more or less severe eyesight
deficiencies, we were concerned to find that,
although better results than classical regimen
one were achieved in terms of tumor response
and PFS in our first reported series treated with
cisplatin and etoposide, 28% of the children
treated developed hearing impairments, albeit at
high frequencies with no social impact
(Massimino et al. 2002). Another issue was the
potentially cumulative toxicity of etoposide,
especially in cases with neurofibromatosis. Hence
our second pauci-institutional trial forming the
object of this report, using considerably reduced
doses of cisplatin (17% less) and etoposide (34%
less) for each course.
We recorded much the same results, in terms
of 3-year PFS and OS, when we compared the
two series of patients. A tumor volume reduction
(evaluated using the same criteria in the two
series) was seen in 20/28 patients in the first
series treated at our institute, and in 24/37 treated
with the second protocol, confirming a very similar response rate. The time to maximal response
differed considerably between the two series,
however, being a median 8 months for the first
and 12 months for the second. This finding
prompts us to suggest that crucial cumulative
quantities of cisplatin (around 700750 mg/sqm)
and etoposide (around 3,0003,600 mg/sqm) are
needed to obtain the maximal response with this
chemotherapy regimen. We can also assume that
higher cumulative doses are responsible for a
M. Massimino et al.
greater toxicity, but fail to improve or extend

tumor response.
Stewart et al. have studied several chemotherapy
drugs with respect to the concentrations reached
in human intracranial tumours (Stewart et al.
1994a, b). Concentrations in intracranial tumours
appeared to be similar to those in extracranial
tumours for cisplatin, phosphonacetyl-L-aspartate,
4-(9-acridinylamino)-methanesulphon-m-aniside
(AMSA), pentamethylmeamine, doxorubicin and
vinblastine. Low drug concentrations in normal
brain and cephalospinal fluid did not preclude
high concentrations in intracranial tumours.
Analysis of variance, however, revealed that
platinum concentrations varied significantly
(P < 0.05) across the entire category as a function
of tumour type (ranking order: medulloblastomalymphoma-meningioma group > brain metastases >
extracranial tumours > gliomas), tumour histopathology (others > adenocarcinomas > squamous
carcinomas > gliomas), and tumour source
(non-small-cell lung cancer (NSCLC), tumours
from other extracranial sources and the
medulloblastoma-lymphoma-meningioma
group > head and neck cancers > gliomas). Of
note, tumour location (intracranial vs. extracranial) appeared to have far less of an impact on
tumour platinum concentration than did tumour
histopathology and source. Tumour grade may
have had an effect on tumour platinum accumulation for gliomas, but it did not have a significant
impact for other tumour types. In the case of lowgrade glioma of childhood we have clinical evidence of an efficacious penetration of the global
dose of cisplatinum that we have adopted. The
tumor reduction after chemotherapy meant a
trend to better PFS and OS as compare to lack of
response with volume stability as already showed
by other series (Lafay-Cousin et al. 2005).
Differently from the previous series (Massimino
et al. 2002) we could not observe any prognostic
significance for patient age neither under 1 nor
over 5 years. A possible explanation can be found
in the different numbers in subgroup stratification i.e. a higher patient number over 5 years and
a lower patient number under 1 year in the second
series here reported. It is worth mentioning,
moreover, that the series discussed in the present
29
paper contained an unusually large proportion of

patients with metastases, representing over 24%
of all the consecutively-treated children, whereas
the rates reported in the literature are under 10%
even in these modern times of MRI (Gajjar et al.
1995; Fernandez et al. 2003). This rate is significantly higher than that in the previous series
reported (Massimino et al. 2002). PFS was significantly lower for the subgroup of metastatic
patients, despite 7/9 of them responding to the
treatment, which emphasizes the need to find a
different treatment, possibly including other
drugs or different approaches, if genetics reveal a
particular profile for tumors with this pattern of
presentation, as recently demonstrated for pilocytic astrocytoma originating at different sites
(Sharma et al. 2007). Disease control also differed significantly for clinically versus histologically diagnosed cases: patients with clinical/
radiological diagnoses and pilocytic astrocytoma
having a more favorable outcome. A more favorable prognosis after a diagnosis on clinical and
radiological grounds can only be explained by
the typical signs that lead to the diagnosis of a
pilocytic astrocytoma, that can be identified by
most experienced neuroradiologists (Kornreich
et al. 2001). On the other hand, it is fairly obvious
that the group of patients with grade 2 tumors,
such as pilomyxoid astrocytoma and fibrillary
astrocytoma, have a worse prognosis (Fisher
et al. 2008). In our experience, the two children
with fibrillary astrocytoma fared very badly,
their tumor evolving into glioblastoma. Recent
literature has shown that grade 1 and grade 2
astrocytoma are molecularly clearly distinguished also in children (Korshunov et al. 2009)
and therefore the evolution from fibrillary
astrocytoma to glioblastoma could be related
either to radiation treatment or to the tumor
spontaneous evolution as it is more common in
adults than in children (De Carli et al. 2009; Yan
et al. 2009). The rapid progression and death of
a third patient with a pilocytic astrocytoma of
visual pathway, already presenting at diagnosis
with huge metastases, was quite unexpected, but
similar evolutions are not absent in literature
reports and recent histopathological stratification of pilocytic astrocytoma can give reason
317
of cases with particular ominous outcome

(Tibbetts et al. 2009).
Besides the radiological response, we evaluated also the symptoms and signs that had brought
our patients to diagnosis and that improved in
70% of them. As Gutmann also pointed out, given
the slow growth rate and infiltrative nature of
low-grade glioma in children, a reduction in overall tumor volume cannot be the only endpoint in
clinical trials for these diseases (Gutmann 2008).
Changes in clinical variables may be even more
difficult to evaluate (Laithier et al. 2003) due, for
instance, to the lack of standardized methods of
ophthalmological evaluation, to the young age of
the patients involved, to the difficulties of conducting complex endocrinological and/or neurofunctional tests at peripheral treatment centers. As
pediatric oncologists, we are obliged to face a
number of issues relating to the final outcome of
our patients with low-grade glioma, who may be
referred for diagnosis in emergency conditions,
with life-threatening disease in cases of diencephalic syndrome, or with a slow-growing disease as
is more often the case of patients with NF1
(Walker 2003). This particular subset of patients
are per se at risk of developing subsequent tumors
among which juvenile chronic myelomonocytic
leukemia (Emanuel 2008). We are therefore, on
one hand very cautious in deciding when and
how treating these children, being aware of the
long life they have in front and the possibility to
develop and suffer from side-effects but, on the
other hand, when treatment beginning is decided,
it is desirable to offer an efficacious and possible
definitive strategy as suggested by a control of
disease documented at 3 years in over 90% of
our patients. Children with NF1 deserve a keen
hematologic follow-up even if secondary leukemias after epipodophyllotoxins have not been
associated with any genetic predisposition also
including germline NF1 mutations; we have
moreover consistently lowered etoposide cumulative dose thus reaching a moderate dose
with low correlated risk of secondary leukemias
as defined by the NCI in their analysis (Smith
et al. 1999).
In the majority of patients whose disease is
diagnosed at over 1 year of age, a low-grade
M. Massimino et al.
318
glioma is more like a chronic condition than a

true tumor with a definite beginning and end to
the time when it is a threat to life, so we are also
obliged to prevent our treatments from harming
our patients. We have demonstrated that a good
response rate can be maintained with a cisplatin
and etoposide regimen that correlates with a satisfying PFS, but also that we can safely reduce
the dosage of these drugs and thereby reduce the
risk of ototoxicity in a significant number of
patients by comparison with the original cisplatin/etoposide schedule. Good responses to a cisplatin containing regimen have been also
demonstrated by other authors like Gajjar et al.
referring 1 complete and 6 partial remission on
13 patients affected by low-grade progressing
glioma (Gajjar et al. 1993). They also documented sensorineural hearing loss for median
cumulative dose of 444 mg/m2 that is a quite
lower dose than the one we have adopted, but we
do not have the information of the schedule
adopted that is determinant in ototoxicity generation (Rybak 2007). Recent publications have
moreover highlighted the possibility to preventing hearing loss by cisplatin with the use exogenous agents. Numerous attempts have been
made at upstream protection of the cochlea with
a variety of antioxidant compounds before
death pathways. Among them, several antioxidants containing thiol groups attenuate cisplatin ototoxicity including sodium thiosulfate,
diethyldithiocarbamate, D or L-methionine,
methylthiobenzoic
acid,
lipoic
acid,
N-acetylcysteine, tiopronin, glutathione ester and
amifostine (Fouladi et al. 2008). Among this,
sodium thiosulfate seems the easiest to be administered and it is suggested that it can permit
maintaining dose-intensity of cisplatin by preventing myelosuppression and renal toxicity
(Robbins et al. 1996). We are also confident that
our treatment has not added any significant cognitive impairment to our patients, as already
demonstrated in 18 patients from the first series
who had chiasmatic-hypothalamic tumors (Riva
et al. 2009).
In conclusion, we can confirm that the cisplatin/
etoposide regimen is still effective at lower cumulative doses, which carry a lower hematological
and audiological toxicity. This regimen should be

intensified for metastatic tumors, however, and it
is not indicated for the fibrillary histotype. Apart
from this and other efficacious chemotherapy
regimens, new radiotherapeutic and neurosurgical techniques have now become a reality to treat
these tumors, though only at referral centers for
the time being, and pediatric neuro-oncologists
need to be aware of these new possibilities and
always discuss their cases in a multidisciplinary
setting in order to give patients every chance of
suitable treatment (Puget et al. 2007; Jahraus and
Tarbell 2006; Hug et al. 2002).
References
Brock PR, Bellman SC, Yeomans EC, Pinkerton CR,
Pritchard J (1991) Cisplatin ototoxicity in children: a
practical grading system. Med Pediatr Oncol
19:295300
De Carli E, Wang X, Puget S (2009) IDH1 and IDH2
mutations in gliomas. N Engl J Med 360(8):765773
Emanuel PD (2008) Juvenile myelomonocytic leukemia
and chronic myelomonocytic leukemia. Leukemia
22:13351342
Fernandez C, Figarella-Branger D, Girard N, Bouvier-Labit
C, Gouvernet J, Paz Paredes A, Lena G (2003) Pilocytic
astrocytomas in children: prognostic factors a retrospective study of 80 cases. Neurosurgery 53:544555
Fisher PG, Tihan T, Goldthwaite PT, Wharam MD, Carson
BS, Weingart JD, Repka MX, Cohen KJ, Burger PC
(2008) Outcome analysis of childhood low-grade
astrocytomas. Pediatr Blood Cancer 51:245250
Fouladi M, Chintagumpala M, Ashley D, Kellie S,
Gururangan S, Hassall T, Gronewold L, Stewart CF,
Wallace D, Broniscer A, Hale GA, Kasow KA,
Merchant TE, Morris B, Krasin M, Kun LE, Boyett
JM, Gajjar A (2008) Amifostine protects against
cisplatin-induced ototoxicity in children with averagerisk medulloblastoma. J Clin Oncol 26:37493755
Gajjar A, Heideman RL, Kovnar EH, Langston JA,
Sanford RA, Douglass EC, Jenkins JJ, Horowitz ME,
Kun LE (1993) Response of pediatric low grade
gliomas to chemotherapy. Pediatr Neurosurg
19:113118
Gajjar A, Bhargava R, Jenkins JJ, Heideman R, Sanford
RA, Langston JW, Walter AW, Kuttesch JF, Muhlbauer
M, Kun LE (1995) Low-grade astrocytoma with
neuraxis dissemination at diagnosis. J Neurosurg
83:6771
Gajjar A, Sanford RA, Heideman R, Jenkins JJ, Walter A,
Li Y, Langston JW, Muhlbauer M, Boyett JM, Kun LE
(1997) Low-grade astrocytoma: a decade of experience at St. Jude Childrens Research Hospital. J Clin
Oncol 15:27922799
29
Gnekow AK (1995) Recommendations of the brain tumor

subcommittee for the reporting of trials: SIOP brain
tumor subcommittee International Society of
Pediatric Oncology. Med Pediatr Oncol 24:104108
Gutmann DH (2008) Using neurofibromatosis-1 to better
understand and treat pediatric low-grade glioma.
J Child Neurol 23:11861194
Hsu TR, Wong TT, Chang FC, Ho DM, Tang RB, Thien
PF, Chang KP (2008) Responsiveness of progressive
optic pathway tumors to cisplatin-based chemotherapy
in children. Childs Nerv Syst 24:14571461
Hug EB, Muenter MW, Archambeau JO, DeVries A,
Liwnicz B, Loredo LN, Grove RI, Slater JD (2002)
Conformal proton radiation therapy for pediatric lowgrade astrocytomas. Strahlenther Onkol 178:1017
Jahraus CD, Tarbell NJ (2006) Optic pathway gliomas.
Kaplan EL, Meyer P (1958) Non-parametric estimation
from incomplete observations. J Am Stat Assoc
53:457481
Kornreich L, Blaser S, Schwarz M, Shuper A, Vishne TH,
Cohen IJ, Faingold R, Michovitz S, Koplewitz B,
Horev G (2001) Optic pathway glioma: correlation of
imaging findings with the presence of neurofibromatosis. AJNR Am J Neuroradiol 22:19631969
Korshunov A, Meyer J, Capper D, Christians A, Remke
M, Witt H, Pfister S, von Deimling A, Hartmann C
(2009) Combined molecular analysis of BRAF and
IDH1 distinguishes pilocytic astrocytoma from diffuse
astrocytoma. Acta Neuropathol 118:401405
Lafay-Cousin L, Holm S, Qaddoumi I, Nicolin G, Bartels
U, Tabori U, Huang A, Bouffet E (2005) Weekly vinblastine in pediatric low-grade glioma patients with
carboplatin allergic reaction. Cancer 103:26362642
Laithier V, Grill J, Le Deley MC, Ruchoux MM, Couanet
D, Doz F, Pichon F, Rubie H, Frappaz D, Vannier JP,
Babin-Boilletot A, Sariban E, Chastagner P, Zerah M,
Raguin MA, Hartmann O, Kalifa C, French Society of
Pediatric (2003) Progression-free survival in children
with optic pathway tumors: dependence on age and the
quality of the response to chemotherapy results of
the first French prospective study for the French
Society of Pediatric Oncology. J Clin Oncol
21:45724578
Massimino M, Spreafico F, Cefalo G, Riccardi R, TesoroTess JD, Gandola L, Riva D, Ruggiero A, Valentini L,
Mazza E, Genitori L, Di Rocco C, Navarria P, Casanova
M, Ferrari A, Luksch R, Terenziani M, Balestrini MR,
Colosimo C, Fossati-Bellani F (2002) High response
rate to cisplatin/etoposide regimen in childhood
low-grade glioma. J Clin Oncol 20:42094216
McLaughlin ME, Robson CD, Kieran MW, Jacks T,
Pomeroy SL, Cameron S (2003) Marked regression of
metastatic pilocytic astrocytoma during treatment with
imatinib mesylate (STI-571, Gleevec): a case report
and laboratory investigation. J Pediatr Hematol Oncol
25:644648
Packer RJ, Lange B, Ater J, Nicholson HS, Allen J, Walker
R, Prados M, Jakacki R, Reaman G, Needles MN,
319
Phillips P, Ryan J, Boyett JM, Geyer R, Finlay J (1993)

Carboplatin and vincristine for recurrent and newly
diagnosed low-grade gliomas of childhood. J Clin
Oncol 11:850856
Packer RJ, Ater J, Allen J, Phillips P, Geyer R, Nicholson
HS, Jakacki R, Kurczynski E, Needle M, Finlay J,
Reaman G, Boyett JM (1997) Carboplatin and vincristine for children with newly diagnosed progressive
low-grade gliomas. J Neurosurg 86:747754
Petronio J, Edwards SB, Prados M, Freyberger S, Rabbitt
J, Silver P, Levin VA (1991) Management of chiasmal
and hypothalamic gliomas of infancy and childhood
with chemotherapy. J Neurosurg 74:701708
Puget S, Crimmins DW, Garnett MR, Grill J, Oliveira R,
Boddaert N, Wray A, Lelouch-Tubiana A, Roujeau T,
Di Rocco F, Zerah M, Sainte-Rose C (2007) Thalamic
tumors in children: a reappraisal. J Neurosurg
106:354362
Riva D, Massimino M, Giorgi C, Nichelli F, Erbetta A,
Usilla A, Vago C, Bulgheroni S (2009) Cognition
before and after chemotherapy alone in children with
chiasmatic-hypothalamic tumors. J Neurooncol
92:4956
Robbins KT, Fontanesi J, Wong FS, Vicario D, Seagren S,
Kumar P, Weisman R, Pellitteri P, Thomas JR, Flick P,
Palmer R, Weir A 3rd, Kerber C, Murry T, Ferguson R,
Los G, Orloff L, Howell SB (1996) A novel organ
preservation protocol for advanced carcinoma of the
larynx and pharynx. Arch Otolaryngol Head Neck
Surg 122:853857
Rybak LP (2007) Mechanisms of cisplatin ototoxicity and
progress in otoprotection. Curr Opin Otolaryngol
Head Neck Surg 15:364369
Sharma MK, Mansur DB, Reifenberger G, Perry A,
Leonard JR, Aldape KD, Albin MG, Emnett RJ,
Loeser S, Watson MA, Nagarajan R, Gutmann DH
(2007) Distinct genetic signatures among pilocytic
astrocytomas relate to their brain region origin. Cancer
Res 67:890900
Smith MA, Rubinstein L, Anderson JR, Arthur D,
Catalano PJ, Freidlin B, Heyn R, Khayat A, Krailo M,
Land VJ, Miser J, Shuster J, Vena D (1999) Secondary
leukemia or myelodisplastic syndrome after treatment
with epipodophyllotoxins. J Clin Oncol 17:275281
Stewart DJ, Molepo JM, Eapen L, Montpetit VA, Goel R,
Wong PT, Popovic P, Taylor KD, Raaphorst GP
(1994a) Cisplatin and radiation in the treatment of
tumors of the central nervous system: pharmacological considerations and results of early studies. Int
Stewart DJ, Molepo JM, Green RM, Montpetit VA,
Hugenholtz H, Lamothe A, Mikhael NZ, Redmond
MD, Gadia M, Goel R (1994b) Factors affecting
platinum concentrations in human surgical tumour
specimens after cisplatin. Br J Cancer 71:598604
Tibbetts KM, Emnett RJ, Gao F, Perry A, Gutmann DH,
Leonard JR (2009) Histopathologic predictors of
pilocytic astrocytoma event-free survival. Acta
Neuropathol 117:657665
320
Walker D (2003) Recent advances in optic nerve glioma
with a focus on the young patient. Curr Opin Neurol
16:657664
Yan H, Parsons DW, Jin G, McLendon R, Rasheed BA,
Yuan W, Kos I, Batinic-Haberle I, Jones S, Riggins GJ,
M. Massimino et al.
Friedman H, Friedman A, Reardon D, Herndon J,
Kinzler KW, Velculescu VE, Vogelstein B, Bigner DD
(2009) IDH1 and IDH2 mutations in gliomas. N Engl
J Med 360:765773
Pediatric Paragangliomas:
Role of Germline Mutation
in Succinate Dehydrogenase
30
Pinki K. Prasad and Elizabeth Yang
Contents
Abstract
Introduction ............................................................
322
Genetics ...................................................................
Succinate Dehydrogenase ........................................
Syndromes Associated with PGL ............................
322
322
323
SDH Tumor Suppressor Mechanism....................
324
Clinical Presentation ..............................................
325
Diagnosis .................................................................
325
Staging and Prognosis............................................
327
Treatment ................................................................
327
Follow-Up ................................................................
329
References ...............................................................
330
P.K. Prasad
Department of Pediatrics, Division of HematologyOncology, Vanderbilt University School of Medicine
397 PRB, Nashville, TN 37232, USA
e-mail: pprasa@Isuhsc.edu
E. Yang (*)
Childrens National Medical Center, Center for Cancer
and Blood Disorders of Northern Virginia, 6565
Arlington Blvd, Falls Church, VA 22042, USA
e-mail: eyang@childrensnational.org
Paragangliomas are parasympathetic or

sympathetic extra adrenal tumors of neural
crest origin and constitute a fascinating group
of genetically driven tumors. Sporadic paragangliomas occur, but paragangliomas often
arise as part of paraganglioma syndromes
(PGL), with linkage to mutation in subunits of
the succinate dehydrogenase complex (SDH).
The SDH enzyme complex is part of the Kreb
cycle and is Complex II of the mitochondria
electron transport chain. Subunits A and B are
catalytic, while C and D subunits anchor the
complex to the inner mitochondrial membrane. Mutations in the iron sulfur protein
subunit SDHB are associated with PGL4,
which consists of sympathetic tumors with
variable malignant potential. SDHC mutations
are responsible for the head and neck tumors
in PGL3. SDHD mutations cause parasympathetic head and neck tumors in PGL1 and
demonstrate genomic imprinting. Tumors
bearing SDH gene mutations tend to be more
aggressive and are associated with decreased
survival compared to tumors without SDH
mutations. SDH gene mutations highlight how
a mitochondrial enzyme functions as a tumor
suppressor. The mechanism of tumorigenesis
is stabilization of the master regulator of
hypoxia HIF1a and activation of the hypoxic
response. The clinical presentations usually
are manifestations of catechol hypersecretion.
Urine and/or plasma fractionated catecholamines

321
P.K. Prasad and E. Yang
322
constitute biochemical diagnosis. Standard

radiographic workup includes CT, MRI,
and radioactive meta-iodobenzylguanidine
(MIBG) scintography if available, as well as
emerging techniques capitalizing on the catechol precursor uptake property of these tumors.
Surgical removal is the mainstay of therapy
for PGLs and requires pre-op adrenergic
blockade together with experienced surgical
and anesthesia teams. Early detection can lead
to timely tumor removal and prevention of
metastasis, underscoring the importance of
genetic testing in patients presenting with
paragangliomas and their family members.
Introduction
Paragangliomas (PGL) are tumors that arise from
neural crest cells, comprising a group of dominantly inherited disorders. These disorders are
characterized by the development of highly vascularized extra-adrenal tumors arising in the
sympathetic or parasympathetic ganglia.
Paragangliomas can be benign or malignant, and
functional or non-functional; malignancy is
defined by the presence of metastasis and functional refers to whether the tumor secretes catechols. Functional PGL are usually sympathetic
in origin, and are distinguished from pheochromocytomas by their extra adrenal site of origin.
Paragangliomas are located anywhere from the
base of the skull to the pelvis, but most commonly
arise in the head and neck or in the abdomen near
the renal system. A classic site for paragangliomas is the carotid body at the bifurcation of
the carotid artery. The vast majority of PGL arising in the head and neck are parasympathetic and
nonfunctional, whereas most intraabdominal
PGL are secretory chromaffin tumors. Upwards
of 50% of PGL are familial and are associated
with a paraganglioma syndrome (Neumann
et al. 2004).
Approximately 1020% of PGL cases are
diagnosed during childhood. The average age of
diagnosis is 11 years with a slight predominance
in boys, particularly in boys under the age of 10
(Beltsevich et al. 2004; Neumann et al. 2002).
In this chapter, we examine pediatric

paragangliomas, the genetic predispositions to
PGL syndromes, and illustrate the clinical
approach to pediatric paraganglioma with a particular patient case.
Genetics
Succinate Dehydrogenase
In the mitochondria, adenosine triphosphate
(ATP) is generated as a result of the interaction of
the Krebs cycle and oxidative phosphorylation.
Succinate dehydrogenase (SDH) is a nuclearly
encoded dual function mitochondrial enzyme
complex that is part of the Krebs cycle and is also
Complex II of the mitochondrial electron transport chain. The SDH complex consists of four
protein subunits: SDHA and SDHB form the catalytic domain, while SDHC and SDHD anchor the
complex to the inner mitochondrial membrane.
Mutations of the B, C, and D subunits were discovered in 2000 and 2001 in association with either
familial PGL syndromes or sporadic paraganglioma
and pheochromocytoma (Baysal et al. 2001).
Evidence suggests that tumorigenesis in PGL
syndromes is linked to activation of hypoxia
related pathways, as mitochondrial Complex II
plays a role in oxygen sensing and signaling
(Gimenez-Roqueplo et al. 2001, 2002).
SDHB gene is located on chromosome 1p35
26 and consists of 8 exons. The SDHB product is
the iron sulfur protein subunit in Complex II. A
number of germline mutations have been reported
in the SDHB gene, including missense and
nonsense mutations in exons 2, 3, 4, 6, and 7,
frameshift mutations, as well as mutations in two
introns (Astuti et al. 2001; Bayley et al. 2009;
Pawlu et al. 2005). The SDHB mutation carriers
have almost twice the risk of developing intraabdominal PGLs compared to SDHD mutation
carriers, consistent with higher prevalence of
malignant disease in SDHB mutation-positive
patients (Benn et al. 2006; Gimenez-Roqueplo
et al. 2003; Neumann et al. 2004) . A study
of 84 patients revealed a strong association of
SDHB mutations with ectopic site and recurrent
30
Pediatric Paragangliomas: Role of Germline Mutation in Succinate Dehydrogenase
malignancies (Gimenez-Roqueplo et al. 2003).

This indicates that SDHB mutations generate
more aggressive types of PGL disease and may
require more rigorous follow-up. We have
reported a pediatric patient with a SDHB mutation
in exon 2, who presented with three simultaneous
tumors in the renal bed (Prasad et al. 2009).
Recognition of SDHB related disease in individual patients could be delayed due to several
factors. Despite an autosomal dominant pattern
of inheritance, penetrance of the disease is incomplete and age dependent (Benn et al. 2006). Many
index patients with SDHB related tumors do not
have a family history of PGL even though SDHB
mutations may exist in family members. Younger
patients may be diagnosed before their SDHB
mutation-positive parents come to medical attention. Such an example is our previously reported
13 year-old PGL4 patient whose father was
asymptomatic at age 58, but was found to carry
the SDHB mutation (Prasad et al. 2009).
Mutations in SDHC and SDHD are also
associated with paragangliomas. SDHC gene is
located on chromosome 1q21 and mutations in
this gene are associated with solitary tumors in
the head and neck area (Schiavi et al. 2005). The
SDHD gene, located on chromosome 11q23,
consists of 4 exons encoding the small subunit of
cytochrome B. Missense and nonsense mutations
are known in all 4 exons and a mutation in one
intron leads to a splicing defect (Pawlu et al.
2005). Familial PGL caused by SDHD mutations
demonstrate genomic imprinting; consequently, a
disease phenotype is manifested only after paternal transmission. Maternal alleles are also transferred, but the first generation progeny are
asymptomatic carriers, consistent with autosomal
dominant transmission with maternal genomic
imprinting (Pigny et al. 2008).
The SDH genes can be characterized as tumor
suppressor genes. The relevant germline mutations in SDHB and SDHD result in loss of function of one allele and tumors exhibit the somatic
loss of the wild type allele (Baysal et al. 2002;
Benn et al. 2003; Neumann et al. 2002; Prasad
et al. 2009). Loss of heterozygosity (LOH) in
SDHB tumors has been demonstrated by linkage probes, by fluorescence in situ hybridization
323
using probes for chromosome 1p, and by direct

sequencing (Gimenez-Roqueplo et al. 2002;
Prasad et al. 2009). Tumors without LOH have
also been reported.
Although sporadic non-malignant paragangliomas are curable, metastasis and the presence
of genetic mutations are associated with decreased
survival. In a retrospective study of 54 patients
with malignant paragangliomas or pheochromocytomas, 23 of whom had SDHB mutations and
31 of whom were SDH mutation-negative, the
presence of SDHB mutation significantly predicted decreased survival (Amar et al. 2007). The
5-year survival after the diagnosis of first metastasis was 0.67 for patients without SDHB mutations, but was 0.36 for SDHB mutation-positive
patients, with a relative mortality risk of 2.7 independently associated with SDHB mutations.
Median survival in this study was 42 months for
SDHB mutation carriers and 244 months for
patients without SDHB mutations. Patients who
had germline mutations were younger and more
frequently had bilateral or extra-adrenal tumors.
The median time from the diagnosis of primary
tumor to metastasis was significantly shorter as
well, with 4 months in patients with SDHB mutations and 20 months in patients without mutations
(Amar et al. 2007).
Syndromes Associated with PGL

Paragangliomas often occur as sporadic tumors
but they may also develop as part of a hereditary tumor syndrome. Hereditary PGLs are
related to familial PGL syndrome, von Hippel
Landau syndrome, and less frequently
Neurofibromatosis I. The familial PGL syndromes are numbered 1 through 4. PGL1, PGL3
and PGL4 result from germline mutations in
genes that encode subunits of the mitochondrial
succinate dehydrogenase enzyme (SDH),
while mutation in a cofactor for SDH has been
identified in a PGL2 kindred (Burnichon et al.
2009; Favier et al. 2005; Pasini and Stratakis
2009; Timmers et al. 2009).
PGL1 syndrome primarily consists of
parasympathetic head and neck tumors and is
324
associated with SDHD gene mutations. SDHD

mutation has a 68% penetrance of phenotype by
age 40 (Benn et al. 2006; Burnichon et al. 2009).
PGL2 has not been well characterized, though
recently the mutation for this syndrome has been
mapped to a gene at 11q13.1, named SDH5 (or
SDHAF2), which is involved in the flavination of
SDHA and is required for the full function of the
SDH complex (Hao et al. 2009). PGL3 is associated with the SDHC gene mutation and is almost
exclusively associated with parasympathetic
head and neck tumors (Niemann and Muller
2000). PGL4 is associated with SDHB mutations. PGL4 encompasses patients with sympathetic PGL of varying malignant potential
arising within the abdomen, pelvis, or thorax.
Over 60% of PGL4 patients will be affected by
an abdominal or thoracic tumor by age 60 years
(Benn et al. 2006; Burnichon et al. 2009).
SDH Tumor Suppressor Mechanism

Linkage of SDH gene mutations to PGL syndromes
showed for the first time that mitochondrial
enzymes can be tumor suppressors (Kirches
2009). SDH mutations affect the same pathway
as the tumor suppressor VHL in von-Hippel-Lindau
disease, which highlights the selective pressure
for tumor growth in chronically hypoxic conditions. The mechanism of tumorigenesis is the
stabilization of the oxygen-dependent hypoxiainducible factor (HIF-1a), which is the master
regulator of the hypoxic response. In normoxia,
HIF-1 is a short-lived protein rapidly cleared by
the ubiquitin system. HIF-1a binds to ubiquitin
only when HIF-1a is hydroxylated at two prolines by a prolyl hydroxylase (PHD) whose
activity depends on oxygen and the relative concentrations of a-ketoglutarate and succinate. In
the absence of oxygen, PHD is inactive; HIF-1a
is not hydroxylated, does not bind ubiquitin, and
is stabilized (Lee et al. 2005). The abundance of
HIF-1a results in the induction of the hypoxic
response, which includes neoangiogenesis, as
well as adaptation to survival and proliferation in
hypoxia by shifting to more aerobic glycolytic
SDH loss of function
PHD inactivation
VHL loss of function
HIF-1 stabilization
Hypoxic response
Growth advantage in hypoxic environment
tumorigenesis
Fig. 30.1 SDH loss of function: mechanism of tumorigenesis. Both SDH and VHL loss-of-function mutations
cause stabilization of the master regulator HIF-1a, leading to activation of the hypoxic response and
tumorigenesis
metabolism. While the VHL gene product is a

subunit of the E3-ubiquitin ligase that degrades
HIF-1a and loss of VHL leads to inability of
ubiquitin to bind HIF-1, loss of SDH results in
inactivation of the prolyl hydroxylase whose
activity is necessary for HIF-1 degradation. Both
result in constitutively elevated HIF-1 levels,
conferring growth advantage in the hypoxic condition commonly found in solid tumors
(Fig. 30.1).
As Complex II of the mitochondria electron
transport chain, the SDH enzyme complex converts succinate to fumarate, and releases an electron to ubiquinone. Since succinate is the substrate
for SDH, loss of function mutations in SDH result
in the accumulation of succinate. The direct competitive feedback inhibition of PHD by succinate,
and resultant increased HIF-1, has been demonstrated in cancer cell lines and pheochromocytoma models (Lee et al. 2005; Selak et al. 2005).
Cell-permeable esterified a-ketoglutarate has
been used in cell line and xenograft models to
reactivate PHD in hypoxia, which destabilized
30
HIF-1 and ultimately resulted in cell death

(Tennant et al. 2009). These studies support the
hypothesis that the hypoxic response is necessary
for tumor survival and raise the interesting possibility of using derivatized a-ketoglutarate as a
cancer therapeutic.
Clinical Presentation
SDHB associated PGL is characterized by a high
malignant potency warranting aggressive therapy,
strict follow-up, and family screening. The diagnosis may be delayed by a negative family history
or an atypical clinical presentation with signs and
symptoms that are predominantly related to tumor
growth rather than catecholamine excess. The biochemical phenotype usually consists of hypersecretion of norephinpehrine and/or dopamine, but
10% of tumors are biochemically silent. The clinical expression of these tumors in individual
patients cannot be predicted by the type and location of the SDHB mutation (Timmers et al. 2007).
The clinical presentation of a sympathetic or
a functional PGL in childhood depends on catecholamine secretion and release. Children
usually present with sustained hypertension or
paroxysmal episodes of headaches, palpitations
and diaphoresis (Pham et al. 2006). Other
symptoms include pallor, orthostatic hypotension, syncope, tremor, and anxiety. Symptoms
can be nonspecific and include blurred vision,
abdominal pain, diarrhea, weight loss, hyperglycemia, polyuria and polydipsia, low-grade
fever, behavioral problems, and decline in
school performance (Prasad et al. 2009; Sullivan
et al. 2005). Children can also present with
symptoms due to tumor burden or incidental
radiographic findings.
Clinical Vignette
We present a 13 year-old boy who sought medical attention because of an insignificant bump
on his chest, but was noted to be hypertensive
with blood pressures consistently over 160/100.
325
He reported a history of episodic severe headaches,

vomiting, facial flushing, profuse sweatiness, and
inability to tolerate any exercise, including going
upstairs. His episodic symptoms were misconstrued as delinquent behavior and he was sent to
an alternative school. Two years prior, he had
been worked up extensively for similar symptoms
at a different hospital, but had normal radiologic
scans and serum studies. He had been treated
with a beta-adrenergic blocker for his high blood
pressure.
Diagnosis
Once there is a clinical suspicion of PGL, a biochemical diagnosis should be sought. The diagnosis of PGL has been simplified by assays that can
quantify levels of catecholamines and their metabolites in blood and urine. Currently, the diagnostic
test of choice is fractionated plasma and/or 24 h
urine catecholamines (dopamine, norephinephrine and epinephrine), metanephrine, norephinephrine, vanillymandelic acid (VMA) and
homovanillic acid (HVA) (Waguespack et al.
2010). Fractionated metanephrines are highly
sensitive tests, approaching 100% sensitivity for
the diagnosis of sympathetic chromaffin tumors
(Lenders et al. 1995; Weise et al. 2002).
Paragangliomas classically secrete nor-epinephrine, while pheochromocytomas secrete epinephrine. An elevation of these analytes greater than
4-fold above the reference range is associated with
an almost 100% probability of the presence of a
catecholamine-secreting tumor (Eisenhofer et al.
2003). Drugs known to interfere with these assays
include acetaminophen, tricyclic antidepressants,
phenoxybenzamines and decongestants; and
should be discontinued prior to testing (Lenders
et al. 2005). A major secretory protein present in
chromaffin granules called Chromogranin A is a
useful marker in the rare SDHB related PGL that is
biochemically silent (Timmers et al. 2007, 2008).
If biochemical diagnosis of catecholamine
excess is established, then radiographic studies
should be undertaken. Often, radiographic studies are done prior to a biochemical diagnosis due
326
Fig. 30.2 Three simultaneous intraabdominal paragangliomas. Reconstruction of patients abdominal CT

scan, revealing three intraabdominal tumors, one of which
displaces the renal vein cephalad (T tumor, Ao aorta, RV

renal vein, *renal artery). (Reprinted from Prasad et al.
(2009), with permission from Elsevier)
to concern about abdominal pain or a tumor is

found incidentally on a radiographic study. The
initial test of choice is either computed tomography (CT) or magnetic resonance imaging (MRI)
of the abdomen and pelvis followed by neck and
chest scans (Pacak et al. 2007). Functional imaging with radioactive meta-iodobenzylguanidine
(MIBG) scintography has also been used to
detect metastatic disease and to assess tumor
avidity. MIBG is a highly specific test that can
confirm the catecholamine-secreting nature of a
tumor. Other radionuclide imaging techniques
include Indium-111 labeled octreotide somatostatin receptor scans (SRS), 6-fluorodopamine
(18F-DA), and fluoro-dihydroxyphenylalanine
(18F-DOPA) scans, but experience and availability
are limited with these modalities (Reynolds and
Lewington 2008). More recently, fluorodeoxyglucose positron emission tomography (FDG PET)
has been used to detect metastatic disease, though
there is not sufficient literature to determine if this
imaging modality is useful in PGL.
Clinical Vignette
Our 13-year-old patient underwent CT scanning
for evaluation of hypertension. Three concurrent
inatraabdominal tumors were found at the level
of the renal hilum, two on the right, with one
above the renal artery and one below, and one
tumor was on the left of the aorta (Fig. 30.2).
Biochemical diagnosis was immediately sought.
Both urine and plasma contained 34 times the
normal level of nor-epinephrine but epinephrine
level was normal, suggesting PGL rather than
pheochromocytoma. The patients peripheral
blood was tested for mutations in VHL, the Ret
oncogene (MEN2), SDHD, and SDHB. A deletion of the cytosine nucleotide 88 in exon 2 codon
30 of SDHB causing a frameshift was identified
in one allele. Thus, this patient carries a heterozygous germline mutation in SDHB, has PGL4
syndrome, and presented with three PGLs.
Presumably tumor growth started more than
2 years prior when the patient initially exhibited
30
symptoms, but the tumor mass too small to be

detected. Testing of first-degree relatives revealed
that the father, then 58 years old, carried the
same SDHB mutation, but he had no elevation of
catecholamines and was completely asymptomatic. The patients brother had normal SDHB
genes (Fig. 30.3).
Staging and Prognosis

Currently there is not a clinical staging system
for malignant PGL. The prognosis for a completely resected tumor is excellent, though life
expectancy for malignant disease is generally
determined by the location of metastatic disease.
Patients with liver or lung metastases have an
overall 5-year survival between 34% and 60%
(Pacak et al. 2007). Children with malignant
tumors have a 5-year disease specific survival
rate closer to 80% and a 10-year survival around
31%, Survival is more variable in children who
have malignant disease, with a mean survival of
157 32 months (Pham et al. 2006).
Clinical Vignette
In our patient diagnosed with three tumors at the
same time, we could not determine for certain
whether the three tumors represented separate
genetic events, or whether there was one primary
tumor and the other two tumors were metastatic
events. Given that the three tumors were in close
vicinity in the intraabdominal region where PGLs
are known to arise, and that no disease at distant
sites were detected, we favored the hypothesis that
the three tumors were independent primary events
rather than malignant metastatic disease.
Nevertheless, developing three tumors in a presumed span of 2 years suggests aggressive disease.
Treatment
Once the diagnosis of PGL is confirmed, medical
therapy should be initiated. Currently the mainstay of treatment is surgical resection of all
327
tumors. Pre-operative biopsy is not indicated and

is potentially dangerous. Weeks prior to resection,
it is imperative that medical therapy be initiated,
in order to minimize complications that may arise
from acute catecholamine surges during induction of anesthesia and surgical manipulation of
tumor. Universal guidelines for medical management of a catecholamine-secreting tumor do not
exist. Review of medical and surgical management revealed that 76% of children who underwent surgery for a pheochromocytoma or a PGL
underwent preoperative alpha adrenergic blockade followed by beta adrenergic receptor blockade (Pham et al. 2006). The primary agent used
in children is the noncompetitive alpha blocker
phenoxybenzamine (Hack 2000). The starting
dose of phenoxybenzamine in children is 0.25
1.0 mg/kg per day or 10 mg once daily; the dose
is increased every few days until the patients
symptoms and blood pressure are controlled and
mild orthostasis has been induced (Kaufman
et al. 1983). On the second or third day of alphaadrenergic blockade, patients are encouraged to
start a diet high in sodium content because of the
catecholamine-induced volume contraction and
the orthostasis associated with alpha-adrenergic
blockade. High sodium intake causes intravascular volume expansion, which may be contraindicated in patients with congestive heart failure or
renal insufficiency.
After adequate alpha-adrenergic blockade has
been achieved, beta-adrenergic blockade is initiated, which typically occurs 23 days preoperatively. The beta-adrenergic blocker should not be
started first, because blockade of vasodilatory
peripheral beta-adrenergic receptors with unopposed alpha-adrenergic receptor stimulation can
lead to a further elevation in blood pressure. The
clinician should exercise caution if the patient is
asthmatic or has congestive heart failure. Chronic
catecholamine excess can produce a cardiomyopathy that may become evident with the initiation
of beta-adrenergic blockade, resulting in acute
pulmonary edema. Therefore, beta-adrenergic
blocker should be administered cautiously at a
low dose, and increased as necessary to control
tachycardia. In most cases, the patient is ready
for surgery in 1014 days after starting the alpha-
328
Fig. 30.3 Sequence data of SDHB LOH. Dideoxy sequence

tracings of SDHB exon 2 in the sense (forward) direction
from DNA isolated from the probands peripheral blood
(upper panel), abdominal tumor (middle panel), and
peripheral blood of a first degree relative (lower panel).
The vertical bar on the base call line of each sequence tracing is the intron 1/exon 2 junction, and the arrows indicate
the start of the c.88delC frameshift variant. Note that the
two peripheral blood tracings are comparable while the

tumor shows relative loss of signal compared to peripheral
blood for one nucleotide at each position where the frameshift results in heterozygosity. A blind linear compilation
of nucleotides with diminished signal gave the sequence of
the normal SDHB exon 2 allele, consistent with significant
loss of this allele in tumor tissue. Reprinted from Prasad
et al. (2009), with permission from Elsevier)
30
adrenergic blockade. Other approaches to alpha

blockade that have been tried include prazosin,
doxazosin, and calcium channel blockers such as
nifedipine and nicardipine (Kocak et al. 2002;
Lebuffe et al. 2005).
Patients with unresectable malignant disease

or distant metastatic disease are more difficult to
manage. Radiation therapy may be preferable to
surgery for large tumors that are associated with
extensive bone destruction or intracranial involvement, for which removal would be associated
with considerable operative morbidity. The
importance of radiation dose was shown in one
series, in which local failure rates were much
higher with doses less than 40 Gy (25% for
<40 Gy versus 12% 40 Gy) (Kim et al. 1980).
Other systemic treatment modalities are
only palliative in nature and include MIBG,
somatostatin analogs, and chemotherapy.
Cyclophosphamide, vincristine and dacarbazine
(CVD) with and without doxorubicin have been
used in PGL and have shown to be associated
with some tumor regression. In a long-term follow-up analysis, combination chemotherapy with
CVD produced a complete response rate of 11%
and a partial response rate of 44%. Median survival was 3.8 years for patients whose tumors
responded to therapy and 1.8 years for patients
whose tumors did not respond (Huang et al.
2008).
Clinical Vignette
When our patient presented, he was already on
the beta blocker atenolol, at 50 mg twice a day. In
preparation for surgical resection, phenoxybenzamine was started at 5 mg twice a day and
atenolol was decreased to 25 mg twice a day concurrently. The dose of phenoxybenzamine was
gradually increased to 50 mg twice a day over a
3-week period before the blood pressure reached
acceptable range and the patient was deemed
safe for surgery (see table below).
In this patient, all three tumors were successfully resected in one surgical operation that
lasted 16 h. Any intraoperative handling of the
tumors caused significant fluctuations in blood
pressure, requiring immediate anti-hypertensives
and pressor drips. Thus, surgery required delicate manipulations and proceeded cautiously
and slowly, illustrating the importance of experienced surgical and anesthestic teams in the resection of catchol-secreting PGLs. After surgery, our
patients blood pressure promptly normalized to
79/44, and he was rapidly weaned off all adrenergic blockade. 1 week after surgery, urine
catecholamines were within normal limits.
10
Follow-Up
What are the clinical implications of findings
associated with SDHB mutations? Amar et al.
(2005) and Brouwers et al. (2006) recommend
that all patients with metastatic paragangliomas
Day
BP
systolic
160 175 153 144 156 164 152 148 168
199 171
BP
diastolic
102 121
86
83
79
94
88
82
92
111
329
11
12
13
14
15
16
165 135 167 156 152
17
18
19
152
150
144 141 141 144
20
21
22
99
98
70
82
90
90
81
82
72
75
75
80
Phenoxy
benzamine
(mg twice
daily)
10
10
10
10
10
10
10
10
12.5
20
20
25
30
30
30
30
40
40
50
50
Atenolol
(mg twice
a day)
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
330
be offered testing for SDHB mutations and that

such testing should receive priority over testing
for other disease-associated genes (Amar et al.
2005; Brouwers et al. 2006). Amar et al. (2007)
discussed that a sporadic presentation does not
exclude the presence of SDHB mutations, because
SDHB-related PGL may have low penetrance;
thus it is important to offer all patients with PGL
genetic testing for the SDHB mutation (Amar
et al. 2007). Brouwers et al. (2006) examined a
number of patients with PGL and concluded that
mutations of the SDHB gene are responsible for
about half of all malignancies originating from
extra-adrenal PGL. Discovery of the SDHB
mutation may have important implications for
triage, screening and medical management of
other asymptomatic mutation-positive family
members. (Benn et al. 2006). Routine screening
for catecholamine-secreting tumors for at-risk
patients can lead to early detection and tumor
removal, preventing subsequent development of
a fatal malignancy.
Carriers of SDH mutations may need close
follow-up, but a consensus protocol has not yet
been established. Little is known about the relationship between particular mutations and clinical
phenotypes; thus further studies correlating genotype to phenotype are required. A study reported
that SDHB mutations have 50% penetrance by
35 years of age and 77% by 50 years of age
(Neumann et al. 2004). For surveillance, regular
blood pressure measurements, physical examination specially dedicated to ruling out abdominal
and head and neck masses, combined with regular
urinary catecholamine measurements seem to be
reasonable for monitoring patients and carriers.
Clinical Vignette
For our patient, MRI of neck, chest, abdomen,
and pelvis obtained at 3 months post-op was
negative for disease. We planned to follow urine
catecholamines every 3 months and obtain MRI if
any elevation were detected. However, the patient
was lost to follow up after 6 months. When he resurfaced 2 years later for an unrelated complaint,
he was still asymptomatic.
Two of the three resected tumors were

sequenced for the SDHB gene. In both, only the
mutant SDHB allele was detected, consistent with
LOH as the mechanism of tumorigenesis. We
were unable to determine definitively whether the
tumors represented the same or separate secondary genetic events by examining the sequences
surrounding the mutation or other polymorphisms. This patients PGL4 was clearly aggressive, but not likely to be malignant, thus, it still
portends a good outcome with aggressive therapy. This young mans case poignantly illustrated
many of the salient genetic and clinical aspects of
paraganglioma syndromes.
References
Amar L, Bertherat J, Baudin E, Ajzenberg C, Bressac-de
Paillerets B, Chabre O, Chamontin B, Delemer B,
Giraud S, Murat A, Niccoli-Sire P, Richard S, Rohmer
V, Sadoul JL, Strompf L, Schlumberger M, Bertagna
X, Plouin PF, Jeunemaitre X, Gimenez-Roqueplo AP
(2005) Genetic testing in pheochromocytoma or functional paraganglioma. J Clin Oncol 23:88128818
Amar L, Baudin E, Burnichon N, Peyrard S, Silvera S,
Bertherat J, Bertagna X, Schlumberger M, Jeunemaitre
X, Gimenez-Roqueplo AP, Plouin PF (2007) Succinate
dehydrogenase B gene mutations predict survival in
patients with malignant pheochromocytomas or paragangliomas. J Clin Endocrinol Metab 92:38223828
Astuti D, Latif F, Dallol A, Dahia PL, Douglas F, George
E, Skoldberg F, Husebye ES, Eng C, Maher ER (2001)
Gene mutations in the succinate dehydrogenase subunit SDHB cause susceptibility to familial pheochromocytoma and to familial paraganglioma. Am J Hum
Genet 69:4954
Bayley JP, Grimbergen AE, van Bunderen PA, van der
Wielen M, Kunst HP, Lenders JW, Jansen JC, Dullaart
RP, Devilee P, Corssmit EP, Vriends AH, Losekoot M,
Weiss MM (2009) The first Dutch SDHB founder
deletion
in
paraganglioma-pheochromocytoma
patients. BMC Med Genet 10:34
Baysal BE, Rubinstein WS, Taschner PE (2001)
Phenotypic dichotomy in mitochondrial complex II
genetic disorders. J Mol Med 79:495503
Baysal BE, Willett-Brozick JE, Lawrence EC, Drovdlic
CM, Savul SA, McLeod DR, Yee HA, Brackmann
DE, Slattery WH 3rd, Myers EN, Ferrell RE,
Rubinstein WS (2002) Prevalence of SDHB, SDHC,
and SDHD germline mutations in clinic patients with
head and neck paragangliomas. J Med Genet
39:178183
Beltsevich DG, Kuznetsov NS, Kazaryan AM, Lysenko
MA (2004) Pheochromocytoma surgery: epidemio-
30
logic peculiarities in children. World J Surg

28:592596
Benn DE, Croxson MS, Tucker K, Bambach CP,
Richardson AL, Delbridge L, Pullan PT, Hammond J,
Marsh DJ, Robinson BG (2003) Novel succinate
dehydrogenase subunit B (SDHB) mutations in familial phaeochromocytomas and paragangliomas, but an
absence of somatic SDHB mutations in sporadic phaeochromocytomas. Oncogene 22:13581364
Benn DE, Gimenez-Roqueplo AP, Reilly JR, Bertherat J,
Burgess J, Byth K, Croxson M, Dahia PL, Elston M,
Gimm O, Henley D, Herman P, Murday V, NiccoliSire P, Pasieka JL, Rohmer V, Tucker K, Jeunemaitre
X, Marsh DJ, Plouin PF, Robinson BG (2006) Clinical
presentation and penetrance of pheochromocytoma/
paraganglioma syndromes. J Clin Endocrinol Metab
91:827836
Brouwers FM, Eisenhofer G, Tao JJ, Kant JA, Adams KT,
Linehan WM, Pacak K (2006) High frequency of SDHB
germline mutations in patients with malignant catecholamine-producing paragangliomas: implications for
genetic testing. J Clin Endocrinol Metab 91:45054509
Burnichon N, Rohmer V, Amar L, Herman P, Leboulleux
S, Darrouzet V, Niccoli P, Gaillard D, Chabrier G,
Chabolle F, Coupier I, Thieblot P, Lecomte P, Bertherat
J, Wion-Barbot N, Murat A, Venisse A, Plouin PF,
Jeunemaitre X, Gimenez-Roqueplo AP (2009) The
succinate dehydrogenase genetic testing in a large prospective series of patients with paragangliomas. J Clin
Eisenhofer G, Goldstein DS, Walther MM, Friberg P,
Lenders JW, Keiser HR, Pacak K (2003) Biochemical
diagnosis of pheochromocytoma: how to distinguish
true- from false-positive test results. J Clin Endocrinol
Metab 88:26562666
Favier J, Briere JJ, Strompf L, Amar L, Filali M,
Jeunemaitre X, Rustin P, Gimenez-Roqueplo AP
(2005) Hereditary paraganglioma/pheochromocytoma
and inherited succinate dehydrogenase deficiency.
Horm Res 63:171179
Gimenez-Roqueplo AP, Favier J, Rustin P, Mourad JJ,
Plouin PF, Corvol P, Rotig A, Jeunemaitre X (2001)
The R22X mutation of the SDHD gene in hereditary
paraganglioma abolishes the enzymatic activity of
complex II in the mitochondrial respiratory chain and
activates the hypoxia pathway. Am J Hum Genet
69:11861197
Gimenez-Roqueplo AP, Favier J, Rustin P, Rieubland C,
Kerlan V, Plouin PF, Rotig A, Jeunemaitre X (2002)
Functional consequences of a SDHB gene mutation in
an apparently sporadic pheochromocytoma. J Clin
Gimenez-Roqueplo AP, Favier J, Rustin P, Rieubland C,
Crespin M, Nau V, Khau Van Kien P, Corvol P, Plouin
PF, Jeunemaitre X (2003) Mutations in the SDHB gene
are associated with extra-adrenal and/or malignant
phaeochromocytomas. Cancer Res 63:56155621
Hack HA (2000) The perioperative management of children with phaeochromocytoma. Paediatr Anaesth
10:463476
331
Hao HX, Khalimonchuk O, Schraders M, Dephoure N,

Bayley JP, Kunst H, Devilee P, Cremers CW, Schiffman
JD, Bentz BG, Gygi SP, Winge DR, Kremer H, Rutter
J (2009) SDH5, a gene required for flavination of succinate dehydrogenase, is mutated in paraganglioma.
Science 325:11391142
Huang H, Abraham J, Hung E, Averbuch S, Merino M,
Steinberg SM, Pacak K, Fojo T (2008) Treatment of
malignant pheochromocytoma/paraganglioma with
cyclophosphamide, vincristine, and dacarbazine: recommendation from a 22-year follow-up of 18 patients.
Cancer 113:20202028
Kaufman BH, Telander RL, van Heerden JA, Zimmerman
D, Sheps SG, Dawson B (1983) Pheochromocytoma
in the pediatric age group: current status. J Pediatr
Surg 18:879884
Kim JA, Elkon D, Lim ML, Constable WC (1980) Optimum
dose of radiotherapy for chemodectomas of the middle
ear. Int J Radiat Oncol Biol Phys 6:815819
Kirches E (2009) Mitochondrial and nuclear genes of
mitochondrial components in cancer. Curr Genomics
10:281293
Kocak S, Aydintug S, Canakci N (2002) Alpha blockade
in preoperative preparation of patients with pheochromocytomas. Int Surg 87:191194
Lebuffe G, Dosseh ED, Tek G, Tytgat H, Moreno S,
Tavernier B, Vallet B, Proye CA (2005) The effect of
calcium channel blockers on outcome following the
surgical treatment of phaeochromocytomas and paragangliomas. Anaesthesia 60:439444
Lee S, Nakamura E, Yang H, Wei W, Linggi MS, Sajan
MP, Farese RV, Freeman RS, Carter BD, Kaelin WG
Jr, Schlisio S (2005) Neuronal apoptosis linked to
EglN3 prolyl hydroxylase and familial pheochromocytoma genes: developmental culling and cancer.
Cancer Cell 8:155167
Lenders JW, Keiser HR, Goldstein DS, Willemsen JJ, Friberg
P, Jacobs MC, Kloppenborg PW, Thien T, Eisenhofer G
(1995) Plasma metanephrines in the diagnosis of pheochromocytoma. Ann Intern Med 123:101109
Lenders JW, Eisenhofer G, Mannelli M, Pacak K (2005)
Phaeochromocytoma. Lancet 366:665675
Neumann HP, Bausch B, McWhinney SR, Bender BU,
Gimm O, Franke G, Schipper J, Klisch J, Altehoefer C,
Zerres K, Januszewicz A, Eng C, Smith WM, Munk R,
Manz T, Glaesker S, Apel TW, Treier M, Reineke M,
Walz MK, Hoang-Vu C, Brauckhoff M, Klein-Franke A,
Klose P, Schmidt H, Maier-Woelfle M, Peczkowska M,
Szmigielski C (2002) Germ-line mutations in nonsyndromic pheochromocytoma. N Engl J Med 346:14591466
Neumann HP, Pawlu C, Peczkowska M, Bausch B,
McWhinney SR, Muresan M, Buchta M, Franke G,
Klisch J, Bley TA, Hoegerle S, Boedeker CC, Opocher
G, Schipper J, Januszewicz A, Eng C (2004) Distinct
clinical features of paraganglioma syndromes associated with SDHB and SDHD gene mutations. JAMA
292:943951
Niemann S, Muller U (2000) Mutations in SDHC cause
autosomal dominant paraganglioma, type 3. Nat Genet
26:268270
332
Pacak K, Eisenhofer G, Ahlman H, Bornstein SR,
Gimenez-Roqueplo AP, Grossman AB, Kimura N,
Mannelli M, McNicol AM, Tischler AS (2007)
Pheochromocytoma: recommendations for clinical
practice from the First International Symposium.
October 2005. Nat Clin Pract Endocrinol Metab
3:92102
Pasini B, Stratakis CA (2009) SDH mutations in tumorigenesis and inherited endocrine tumours: lesson from
the phaeochromocytoma-paraganglioma syndromes. J
Intern Med 266:1942
Pawlu C, Bausch B, Neumann HP (2005) Mutations of the
SDHB and SDHD genes. Fam Cancer 4:4954
Pham TH, Moir C, Thompson GB, Zarroug AE, Hamner
CE, Farley D, van Heerden J, Lteif AN, Young WF Jr
(2006) Pheochromocytoma and paraganglioma in
children: a review of medical and surgical management at a tertiary care center. Pediatrics
118:11091117
Pigny P, Vincent A, Cardot Bauters C, Bertrand M, de
Montpreville VT, Crepin M, Porchet N, Caron P
(2008) Paraganglioma after maternal transmission of a
succinate dehydrogenase gene mutation. J Clin
Prasad P, Kant JA, Wills M, OLeary M, Lovvorn H 3rd,
Yang E (2009) Loss of heterozygosity of succinate
dehydrogenase B mutation by direct sequencing in
synchronous paragangliomas. Cancer Genet Cytogenet
192:8285
Reynolds S, Lewington V (2008) Radionuclide imaging
of phaeochromocytoma and paraganglioma. Oncol
News 3:2124
Schiavi F, Boedeker CC, Bausch B, Peczkowska M,
Gomez CF, Strassburg T, Pawlu C, Buchta M,
Salzmann M, Hoffmann MM, Berlis A, Brink I,
Cybulla M, Muresan M, Walter MA, Forrer F, Valimaki
M, Kawecki A, Szutkowski Z, Schipper J, Walz MK,
Pigny P, Bauters C, Willet-Brozick JE, Baysal BE,
Januszewicz A, Eng C, Opocher G, Neumann HP
(2005) Predictors and prevalence of paraganglioma

syndrome associated with mutations of the SDHC
gene. JAMA 294:20572063
Selak MA, Armour SM, MacKenzie ED, Boulahbel H,
Watson DG, Mansfield KD, Pan Y, Simon MC,
Thompson CB, Gottlieb E (2005) Succinate links TCA
cycle dysfunction to oncogenesis by inhibiting HIFalpha prolyl hydroxylase. Cancer Cell 7:7785
Sullivan J, Groshong T, Tobias JD (2005) Presenting signs
and symptoms of pheochromocytoma in pediatricaged patients. Clin Pediatr (Phila) 44:715719
Tennant DA, Frezza C, MacKenzie ED, Nguyen QD,
Zheng L, Selak MA, Roberts DL, Dive C, Watson DG,
Aboagye EO, Gottlieb E (2009) Reactivating HIF prolyl hydroxylases under hypoxia results in metabolic
catastrophe and cell death. Oncogene 28:40094021
Timmers HJ, Kozupa A, Eisenhofer G, Raygada M,
Adams KT, Solis D, Lenders JW, Pacak K (2007)
Clinical presentations, biochemical phenotypes, and
genotype-phenotype correlations in patients with succinate dehydrogenase subunit B-associated pheochromocytomas and paragangliomas. J Clin Endocrinol
Metab 92:779786
Timmers HJ, Pacak K, Huynh TT, Abu-Asab M, Tsokos M,
Merino MJ, Baysal BE, Adams KT, Eisenhofer G (2008)
Biochemically silent abdominal paragangliomas in
patients with mutations in the succinate dehydrogenase
subunit B gene. J Clin Endocrinol Metab 93:48264832
Timmers HJ, Gimenez-Roqueplo AP, Mannelli M, Pacak
K (2009) Clinical aspects of SDHx-related pheochromocytoma and paraganglioma. Endocr Relat Cancer
16:391400
Waguespack SG, Rich T, Grubbs E, Ying AK, Perrier ND,
Ayala-Ramirez M, Jimenez C (2010) A current review
of the etiology, diagnosis, and treatment of pediatric
pheochromocytoma and paraganglioma. J Clin
Endocrinol Metab 95(5):20232037
Weise M, Merke DP, Pacak K, Walther MM, Eisenhofer G
(2002) Utility of plasma free metanephrines for detecting childhood pheochromocytoma. J Clin Endocrinol
Metab 87:19551960
Index
A
Aaronson, S.A., 262
Abbott, R., 160
Acar, M., 265
Acute lymphoblastic leukemia (ALL)
CNS prophylaxis, 237
survivors, 232, 238
Adamowicz-Brice, M., 264
Adams, K.T., 329, 330
Adamson, P., 94
Adn, B., 263
Adesina, A.M., 132, 216, 263
Adult survivors, pediatric cancer
ageing, existing survivors, 253
CCSS, 248
components
CCSS, 250
CNS tumor survivors, 250
hypothalamic tumors, 249
insulin resistance, 248
obesity, CNS tumor survivors, 249
radiotherapy, 251
endocrine factors, 251252
host and treatment factors, 248, 249
neuroendocrine abnormalities, 254
surveillance, 252, 253
treatment risk factors, 253
AEDs See Antiepileptic drugs (AEDs)
Afink, G.B., 261
Agarwal, B., 19
Ahern, V., 185
Ahmad, F., 155
Ahn, Y., 299, 301
Ajzenberg, C., 329
Alapetite, C., 185
Alarcon-Vargas, D., 19
Alberta, J.A., 265
Ali, I.U., 262
Ali, S.Z., 53
Alkiza, K., 263
ALL. See Acute lymphoblastic leukemia (ALL)
Allam, A., 298, 302
Allen, J., 186, 311, 315

Allgeier, A., 275, 277
Almqvist, P., 298
Al-Sarraj, S., 264
Al Shabanah, M., 298, 302
Alter, J., 191
Alvarez-Buylla, A., 262
Amar, L., 329, 330
Ambrosini-Spaltro, A., 137
Ambros, P., 7, 15
Amirkhan, R.H., 138
Annett, R.D., 229
Antiepileptic drugs (AEDs)
chronic, 203
epilepsy, 203
pharmacokinetics, 203
reinstitution, 204
Antonelli, M., 133
Apparent diffusion coefficient (ADC) maps
DWI, 45
hypointense signal, 48
Arcella, A., 133
Ardon, H., 19
Arenson, E.B., 191
Aronson, L., 16, 35
Arora, B., 191
Arslanoglu, A., 16, 35
Asgharzadeh, S., 18
Ashdown, B.C., 124
Ashley, D.M., 185, 216, 225, 298
Ashley, S., 304
Ashworth, A., 264
Askins, M., 234
Ater, J., 311, 315
AT/RTs. See Atypical teratoid/rhabdoid tumors (AT/RTs)
Atypical teratoid/rhabdoid tumors (AT/RTs)
clinical presentation, CNS, 45
combination, trastuzumab, 19
divergent tissue types, 4
EMA and RTPS, 32
epidemiology, 4
identification, 1314

DOI 10.1007/978-94-007-2957-5, Springer Science+Business Media Dordrecht 2012
333
Index
334
Atypical teratoid/rhabdoid tumors (AT/RTs) (cont.)
lymphoid nuclei, 135
malignant rhabdoid tumor, 4
meningeal dissemination, 3236
microscopic pathology, 67
molecular biology, 15
molecular pathology, 78
PNET, 134
PNET/MB, 3132
postoperative imaging, 17
preoperative imaging
CT, 1516
MRI, 1617
proliferation rates, 32
sequential courses, 1819
SMARCB1, 19
treatment, 1718
Aurias, A., 7, 15
Avellino, A.M., 303
Aygun, N., 16, 35
B
Babin-Boilletot, A., 298
Baird, S., 261
Baker, S.J., 264
Balestrini, M.R., 298
Ball, W., 235
Balss, J., 131
Banerjee, A., 265
Bannwart, F., 4
Bansal, K.K., 3
Barbarot, S., 298
Baron, M.H., 185
Barrett, J.W., 19
Barrow, J., 264
Bartels, M., 122
Bartels, U., 264, 299, 300
Bastuji-Garin, S., 298
Battistella, P.A., 304
Baudin, E., 329, 330
Baumann, C., 138
Bautista, F., 304
Bax, D.A., 264
Bayar, E., 1718
Beckwith, J.B., 4, 14
Behnisch, W., 192
Belal, A., 298, 302
Belanger, K., 275, 277
Belitskaya-Levy, I., 151
Bell, J.C., 19
Benaim, E., 225
Bendel, A., 18
Benhassel, M., 185
Berger, M.S., 279
Berkow, R., 242
Bernier, V., 185
Bertagna, X., 329, 330
Bertalanffy, H., 158, 160
Bertherat, J., 329, 330
Betensky, R.A., 135

Bevan, H.E., 191
Bhattacharjee, M., 263
Biassoni, V., 309
Biegel, J.A., 7, 9, 14, 15, 18, 19, 31, 33, 134, 135
Biggs, P.J., 14
Billett, A.L., 190, 233234, 299, 304, 305
Bing, F., 39
Biopsy/resection, pediatric brain tumor
narcotic analgesia, 180
non-steroidal anti-inflammatory
medication, 180
postoperative nonnarcotic regimine
nonopioid treatment, 180
platelet dysfunction ibuprofen, 180
platelet function, 181
visual analog scale, 181
Bishop, J.A., 53
Bishop, K., 190, 299
Black, P.M., 265
Blaney, S.M., 18, 94, 265
BMI. See Body mass indices (BMI)
Bode, U., 187
Body mass indices (BMI)
CCSS analysis, 250
obesity risk measurement, 249
Bogdahn, U., 275, 277
Bolwell, B., 125
Bonner, M.J., 238, 243
Bonnin, J.M., 14
Boop, F.A., 189, 190, 225
Boshoff, C., 139
Bouffet, E., 1517, 35, 135, 264, 299, 300
Bournissen, F.G., 240
Bowers, D.C., 13, 1618, 35, 121
Bowman, L., 225
Boyett, J.M., 265, 279, 311, 315
Bozzo, J., 181
Brain stem gliomas, 191192
Brain tumors
diagnostic principles
brainstem, 63
of childhood, 64
complexity, 66
dissemination, CT and MRI, 6263
ependymomas, 6566
grade I and II gliomas, 66
H&E and proliferation markers, 65
JPA, 64
medulloblastomas, 6667
microscopic examination, 65
modern scanners, 65
MRS, 6465
neuro-imaging, 64
optic pathway/suprasellar, 64
pathways and proteins, 66
pediatric and adult neurooncology, 63
posterior fossa, 63
sPNETs and ATRTs, 66
variants, medulloblastoma, 66
Index
volumetric measurements, 65
epidemiology, 6162
gliomas
brainstem, 215216
high-grade, 231214
low-grade, 212213
infants
AT/RTs, 219
embryonal tumors, 219
medulloblastoma, 219220
treatment approaches, 218219
medulloblastoma/PNET, 216218
mice, NF1
deficient astrocyte growth, 288
GEAP, 287
microglia, 288
neruofibromin expression, 287
Nf1 loss, 288
optic glioma formation and VEP, 288
therapies, 290
treatment, 289
prognosis, 212
therapeutic principles
analysis, 71
chemotherapy, 6971
multidisciplinary teams, 67
neurosurgery, 6768
pre-clinical data, 71
radiotherapy, 6869
vaccine trials and clinical investigations, 71
treatment, 220
Brandes, A.A., 275, 277
Braun, A., 161
Brazauskas, R., 125
Breier, G., 263
Bremer, J., 132
Brem, H., 298
Brenemen, J., 235
Bressac-de Paillerets, B., 329
Bressenot, A ., 138
Briner, J., 4
Brodeur, G., 95
Brokinkel, B., 133
Broniscer, A., 18, 264
Brouwers, F.M., 329, 330
Bruggers, C.S., 279, 298
Bruner, J.M., 191, 279
Brunori, A., 160
Buchdunger, E., 265
Buchfelder, M., 298
Buczkowicz, P., 264
Budka, H., 131
Buizer, A.I., 232
Buncic, J.R., 299, 300
Bunt, J., 216
Burger, P.C., 14, 16, 18, 35, 135, 298
Burnichon, N., 330
Butler, R., 242
Buttarelli, F.R., 133
Buxton, N., 160, 161
335
C
Cairncross, J.G., 275, 277
Calderone, M., 304
Camacho, D.L., 17
Cameron, S., 265
Cao, Y., 261, 262
Cappabianca, P., 160
Capper, D., 131133
Carlson-Green, B., 229, 238
Caron, H.N., 216, 302
Carpentieri, S.C., 233234, 237, 304, 305
Carrie, C., 185
Carson, B.S., 298
Carvalho, D., 264
Casanova, M., 225, 298
Castellote, A., 120
Castillo, M., 17
Cavallo, L.M., 160
Cavazos, C.M., 298
Cavenee, W.K., 263
Cecil, K., 235
Cefalo, G., 225, 298
Cell-cycle regulators, rhabdomyosarcoma
frequency, cell-cycle regulators,
27, 28
fusion genes, 29
immunohistochemical staining,
RB expression, 28
Ki-67 protein and PI3K-Akt pathway, 27
MYCN, 27
PAX7-FKHR-positive, 29
p53 pathway
MDM2, 2627
nuclear accumulation, 26
upregulation, p21CIP1 and 14-3-3s, 26
RB pathway
allelic imbalance, 13q12-14, 25
CDKs and cyclin, 25
CIP/KIP family, 2526
cylin D-CDK4-CDK6 complexes, 25
INK4a/ARF family, 26
long arm, chromosome 13, 24
protein expression, 25
target genes, E2F, 24
Central nervous system (CNS)
AT/RT, 3
child
chemotherapy, 237238
CRT, 234236, 238239
neurosurgery, 233234
proton beam radiotherapy, 236
embryonal neoplasms, 48
hSNF5 series, 7
leukemic cells, 232
rhabdoid tumors, 5
tumor, 40
Cerebellar pontine angle (CPA)
ependymoma, 40
internal auditory canal, 4344
Cerebrospinal fluid (CSF), 5, 9, 156
336
Cerebrovascular disorders
cerebral hematoma, 120
cranial radiation, 121
sinus thrombosis, 120, 121
stroke-like migraine, 122
telangiectasias, 121
thrombocytopenia, 120
venous infarcts, 121
Chabre, O., 329
Chako, A.G., 33
Chako, G., 33
Challagulla, K., 19
Chamontin, B., 329
Chandela, S., 303
Chandrasoma, P.T., 130, 139
Chang, C.Y., 1517
Chang, F.C., 1517, 299
Chang, K.P., 299
Chang, Y.M., 263
Chan, K.W., 186, 279
Chan, Y., 122
Chapet, S., 185
Chastagner, P., 298
Chemotherapy
carboplatin and vincristine, 70
defined, 69
HGG, 70
intrinsic and extrinsic tumor resistance, 69
low-grade gliomas (LGG), 310
metronomic dosing schedule, 70
NF1, 292
prognosis, infants, 71
treatment, brain tumors, 69
uses and diagnosis, 70
Cheng, L., 136
Cheng, Y.C., 1517
Cheuk, R., 225
Chiasson, S., 240
Chik, K., 122
Childrens oncology group (COG) protocol, 6768, 217
Chintagumpala, M.M., 132, 185, 216, 225, 263
Chi, S.N., 18
Cho, B.K., 299, 301
Chojnacka, M., 183
Choroid plexus carcinoma, 56, 135
Choux, M., 278
Chow, E.J., 247
Christensen, R., 238
Christians, A., 131
Christophe, C., 119
Chu, A.Y., 140
Chung, Y.N., 299, 301
Chu, W.C.W., 122
Cinalli, G., 160
Claes, L., 19
Claesson-Welsh, L., 260, 262
Clarkson, K.S., 139
Clark, W.C., 262
Claude, L., 185
Clavere, P., 185
Index
Clementi, M., 304
Clifford, S.C., 135, 216
Clinical vignette, PGL
clinical presentation, 325
diagnosis, 326327
follow-up, 330
staging and prognosis, 327
treatment, 329
Cloos, J., 216
Cnaan, A., 95
CNS. See Central nervous system (CNS)
CNS imaging, childhood leukemia
anterior lumbosacral radiculopathy, 120
arachnoid trabeculae, 119
cerebrovascular disorders, 118
chemical meningitis, 119
chloroma, 119
cranial radiation, 125
description, 117118
fat suppression techniques, 119
granulocytic sarcoma, 119
magnetic resonance imaging (MRI), 118
neurotoxic effects, 126
occipital meningeal leukemia, 118
side effects
cerebrovascular disorders, 120122
GHVD, 124
hematopoietic cell transplant (HCT), 125
infections, 123124
malignant neoplasm, 124125
PRES, 123
white matter changes, 122
CNS PNET. See CNS primitive neuroectodermal tumor
(CNS PNET)
CNS primitive neuroectodermal tumor (CNS PNET)
APC gene, 76
B-catenin, 76
breast carcinoma, 84
colon cancer, 84
desmoplastic tumors, 85
developmental signaling pathways, 76
medulloblastomas, 76, 77
method, activation, 85
methodology, 7778
neuroepithelial cells, 76
nuclear localization, 83
pathogenesis, 85
pathway activation, 83
single base substitutions, 76
WNT/B-catenin pathway status in CNS PNET
immunohistochemistry, 7880
mutational analysis, 81
survival analysis, 8182
Cochrane, D.D., 298
Cohen, A., 158
Cohen-Gadol, A.A., 179, 207
Cohen, K.J., 16, 18, 35
Cohen, M.E., 298
Collini, P., 225
Colosimo, C., 298
Index
Colte, P., 229
Combs, S.E., 192, 238
Conklin, H.M., 191
Consales, A., 199
Copeland, D., 242
Cornelison, R., 264
Cosgrove, M., 130, 139
Couanet, D., 298
Coyle, B., 264
CPA. See Cerebellar pontine angle (CPA)
Cranial radiation therapy (CRT)
chemotherapy, 238239
IBED, 235
IQ scores, 235
neurocognitive dysfunction, 235
objective, 234
structural damage, brain development, 235236
Craniopharyngioma
adamantinomatous type, 150
de novo tumors, 171
disease progression, 150
malignant transformation, 170
prognostic factors, recurrence/regrowth, 172
radiotherapy, 192
transphenoidal resection, 151
visual decline and headache., 170, 171
Crange, A., 298
CRT. See Cranial radiation therapy (CRT)
Cruz-Snchez, F.F., 263
CSF. See Cerebrospinal fluid (CSF)
Curran, T., 216
Curschmann, J., 275, 277
Cyber Knife radiosurgery system, 194
D
Da Dalt, L., 302, 304
Dai, C., 261
Dalla Via, P., 304
Dalton, J., 18, 216
Dan, B., 119
Dancey, J., 94
Dasi, N., 160
Dauser, R.C., 132, 225, 263
Deb, P., 134
Debus, J., 192, 238
Decq, P., 160
de Divitiis, E., 160
De Girolami, U., 133
Deinlein, F., 187
Delattre, O., 7, 15
Delemer, B., 329
Delitala, A., 160
Dellatolas, G., 303
Depreitere, B., 160
de Sonneville, L.M., 232
Deutsch, M., 191
De Vleeschouwer, S., 19
Dewitz, R., 122
Diagnosis, AT/RTs
337
clinical features, 40
CPA, 40
differential diagnosis
CNS, 48
CT scanner and MRI findings, 48
ependymoma, 49
meningioma, 49
pilocytic astrocytoma, 49
PNET/MBs, 4849
histopathology
mesenchymal and epithelial cells, 47
rhabdoid cells, 47
teratoid components, 47
immunohistochemistry and cytogenetic study
EMA and VMA, 47
nuclear expression, INI1, 42, 47
PGL, 325327
postoperative imaging, 46
preoperative imaging techniques
CT scanner, 4244
MRI diffusion, 45
MRI perfusion, 45
MRI sequences, 4345
MRI spectroscopy, 4546
presentation, cases, 40, 41
primitive neuroectodermal cells, 40
radiologist, 40
Diencephalic syndrome, 310
Differential diagnosis, AT/RTs
anaplastic/rhabdoid forms, 57
choroid plexus carcinoma, 56
description, 53
germ cells, 57
glioblastoma, 5657
medulloblastoma, 56
microscopic findings
cytologic smear, 54, 55
EMA and GFAP, 55
imunohistochemistry, 55
large malignant cells, 54, 55
medulloblastoma, 54, 55
neuronal-appearing cells, 55
rhabdoid cells, 54
sagittal view MRI scan, 54
teratoid, 54
morphologic appearances and liberal use, INI1, 57
mutation, hSNF5/INI1/SMARCB1, 54
Diffuse intrinsic pontine gliomas
clinical presentation and diagnosis
ataxia and cranial nerve deficits, 277
biopsies, 278279
brainstem, 277
characteristic imaging, 278
lesions, 277
low-grade astrocytomas, 277
MRI, 277278
radiation and chemotherapy, 279
role, FDG-PET, 278
molecular characterization, 279
prognosis and outcomes, 280
Index
338
Diffuse intrinsic pontine gliomas (cont.)
treatment
blood-brain barrier, 281
chemotherapy, carboplatin and thiotepa, 279
clinical trials, 281
drugs, 280, 281
effects, PES and OS, 280
implementation, polymer, 281
meta-anlysis, 280
natural immune-mediated destruction, 281
pediatric and adult neuro-oncology, 280, 281
pre-and post-radiation, 279, 280
Pseudomonas exotoxin, 281
therapy, 279
Diffusion tensor imaging (DTI), 166
Diffusion weighted-image (DWI)
ADC maps, 45
restriction, diffusion, 42, 45
Di Francesco, A.M., 263
Dirks, P., 160
Di Rocco, C., 298
Di Rocco, F., 160
DNET. See Dysembryoplastic neuroepithelial
tumors (DNET)
Doerr, H.G., 298
Do, L., 278
Donahue, B., 186
Dong, S., 135
Donofrio, V., 133
Down syndrome, 239
Doz, F., 298
Drake, J., 160
Driever, P.H., 299
DTI. See Diffusion tensor imaging (DTI)
Duffner, P.K., 14, 298
Dumesnil, R., 122
Duncan, C., 304
Dunham, C.P., 33
Dunn, M.E., 9
Dusenbery, K.E., 279
Dussart, S., 185
Dutta, D., 191
DWI. See Diffusion weighted-image (DWI)
Dysembryoplastic neuroepithelial tumors (DNET)
drug resistant epilepsy, 202
prevalence, 202
E
Eapen, L., 316
Eberhart, C.G., 135
Edgar, M.A., 136
EFS. See Event free survival (EFS)
Ehrenforth, S., 120
Eisenhauer, E., 275, 277
Eisenhofer, G., 329, 330
Elder, D.E., 140
Electronic portal imaging devices (EPID), 184
El Husseiny, G., 298, 302
Elliott, R.E., 151
Ellison, D.W., 216, 264

Elmi, M., 261, 262
EMA. See Epithelial membrane antigen (EMA)
Embry, L., 229
Emser, A., 187, 298, 299
Endoscopic third ventriculostomy (ETV)
noncommunicating hydrocephalus, 145
palliative measure, 145
Ependymoma
diagnostic imaging, 172
management, 173174
prognosis, 174
radiotherapy, 188190
supratentorial, 172
EPID. See electronic portal imaging devices (EPID)
Epileptic seizures and supratentorial brain tumors
and adolescents, 201
antiepileptic treatment
AED, 203
anticonvulsant prophylaxis
and nonprophylaxis groups, 204
chemotherapeutic drugs, 203
seizure prophylaxis, 204
frequency, 200
mechanisms, pathophysiology, 200
neurosurgery and drug-resistant epilepsy, 202203
paediatric, 200
pharmacological therapy, 205
post-surgical evolution, 201
surgery/tumor treatments, 201202
tumor types, 200, 201
Epithelial membrane antigen (EMA)
microcystic/angiomatous meningiomas, 138
perivascular cytoplasmic processes, 132
Escolar, G., 181
Ettinger, L.J., 279
ETV. See Endoscopic third ventriculostomy (ETV)
Eusebi, V., 137
Evans, A., 95
EVD. See External ventricular drain (EVD)
Event free survival (EFS), 14
External ventricular drain (EVD)
acute hydrocephalus, 145
upward herniation syndromes, 145
Eyesight deficiencies, children, 316
F
Fahlbusch, R., 298
Fairclough, D., 242
Fakhrai, H., 261
Farrell, H.B., 181
Fazzini, F, 199
Fellstrm, B., 260
Felsberg, J., 133
Ferrari, A., 225, 298
Ferster, A., 119
Feuer, W., 298
Figarella-Branger, D., 185
Finkelstein, D., 216
Index
Finkelstein, S.D., 279
Finlay, J., 18, 311, 315
Fisher, B., 275, 277
Fisher, M.J., 18
Fitz, C.R., 279
Fitzgibbons, P.L., 130, 139
Flanagan, A.M., 139
Fleischhack, G., 120
Fleming, T.P., 262
Fluid attenuated inversion recovery (FLAIR)
sequences, 44
T2 echo-planar (EP), 43
Fluorodeoxyglucose positron emission tomography
(FDG PET), 326
Fogelgren, B., 7, 15, 134
Forman, S., 161
Forsyth, P.A., 19
Fossati-Bellani, F., 298
Fouladi, M., 18, 298, 300
Frappaz, D., 298
Frazee, J., 160
French, P., 135
Friedman, H.S., 14, 265, 298
Frhwald, M., 31
Fuhrer, K., 139
Fujimori, T., 261, 262
Fuller, C.E., 18, 216
Funa, K., 261, 262
G
Gaab, M.R., 158, 160, 161
Gadia, M., 316
Gaggero, R., 199
Gajjar, A., 18, 185, 216, 225, 264, 279, 298, 300
Galan, A.M., 181
Gandola, L., 225, 298
Gangemi, M., 160
Garcia-Verdugo, J.M., 262
Gardner, S.L., 18
Garen, P.D., 14
Gargan, L., 16, 17, 35
Garibi, J.M., 263
Garr, M.L., 199
Garton, H.J., 297
Garvin, A.J., 14
Gayre, G.S., 298
GCTs. See Germ-cell tumors (GCTs)
Gemer, L., 298
Genitori, L., 298
Gentaz, E., 303
Germ-cell tumors (GCTs), 57, 64, 192194
Geyer, J.R., 265
Geyer, R., 279, 311, 315
GFAP. See Glial fibrillary acidic protein (GFAP)
GHVD. See Graft-Versus-Host Disease (GHVD)
Giangaspero, F., 133, 225
Giese, N.A., 263, 265
Gilbertson, R.J., 18, 94, 216, 225, 279
Gil-Perotin, S., 262
339
Gilreath, L., 180, 208
Gimenez-Roqueplo, A.P., 329, 330
Gimi, B., 16, 17, 35
Giraud, S., 329
Gladding, P.A., 181
Glass, J.O., 122, 238
Glial fibrillary acidic protein (GFAP)
diagnosis, glioma, 131
fetal ependymal cells, 132
neoplastic cells, 130
Glioblastoma, 5657
Gliomas
brainstem, 215216
high-grade
ACNS0126, CCG-943 and CCG-945 study, 214
adult lesions, 213
irradiation, 214
low-grade
chemotherapy regimens, 213
gross total resection, 212213
pilocytic astrocytomas and SEGAs, 212
protocol, 213
Gnekow, A.K., 298, 299
Godano, U., 160
Goel, D., 3
Goel, R., 316
Goldman, S., 18, 265
Goldthwaite, P.T., 298
Gomez, J., 299
Good, C., 122
Gorlia, T., 275, 277
Gorlin syndrome, 218
Goumnerova, L.C., 18, 190, 233234, 299, 304, 305
Grabenbauer, G.G., 298
Graf, N., 187
Graft-Versus-Host Disease (GHVD), 124
Grajkowska, W., 216
Green, A., 18
Green, R.M., 316
Grigoriadis, A.E., 264
Grill, J., 185, 297, 298, 303
Grotenhuis, A., 160
Grundy, R.G., 75, 264
Guillamo, S.J., 298
Guo, W.Y., 1517
Gupta, A., 134
Gupta, T., 191
Gurney, J.G., 241
Gururangan, S., 298
Gusek, G., 298
Gutmann, D.H., 135, 285, 317
H
Habrand, J.L., 185, 303
Hale, G., 18
Hamilton, M., 94
Hamilton, R.L., 135, 279
Hamoudi, A.B., 1718
Hamstra, D.A., 297
Index
340
Hancock, M.L., 265
Handgretinger, R., 7, 15
Hansson, G.K., 260
Hardy, K.K., 243
Hargrave, D., 264, 299, 300
Harter, D., 161
Harter, P.N., 132
Hartman, M., 262
Hartmann, C., 131133
Hartmann, O., 298
Hartmann, W., 216
Hassall, T., 225
Hasselt, N.E., 216
Hassounah, M., 298, 302
Hattab, E.M., 136
Hattingen, E., 122
Hawkins, C.E., 1517, 35, 135, 264
Hayes, M., 265
Hayhurst, C., 160
HCG See Human chorionic gonadotropin (HCG)
Hebrink, D., 130
Heideman, R., 225
Heideman, R.L., 298, 300
Heldin, C.H., 260, 262
Hellwig, D., 158, 160
Helton, S., 238
Henderson, S., 139
Hentschel, B., 133
Hentschel, S., 298
HER family receptors
cancer
cell cycle, 91
protein overexpression, 92
tumorigenesis, 91
characteristics
cell surface, 91
ligand binding, 91
transmembrane growth factor receptors, 90
coexpression profiling, 97
development, peripheral nervous system, 91
EGFR
cytometric analysis, 93
in vitro and in vivo studies, 94
Membranous EGFR expression, 93
neural cell cultures, 92
HER2
ethylnitrosurea-induced rat neuroblastoma, 94
labeling in scattered neuroblastic cells, 94
prognostic factor, 95
HER3, 95
HER4
adult and fetal tissues, 95
adverse prognostic factor, 9697
cytoplasmic staining, 96
Western immunoblotting, 95
Hermanson, M., 262
Hernandez, M.R., 181
Hernan, R., 279
Herndon, J.E., 298
Herpers, M.J., 131
Herrera, E.J., 299

HGGs See High-grade gliomas (HGGs)
1H High resolution magic angle spinning
NMR spectroscopy (HR-MAS)
biochemical changes, 113
chemometric methods, 114
choline metabolism, 113
frozen section analysis, 109
glial and PNET brain tumour, 111
metabolic response, transfection, 109
methods
cell line samples, 110
fitting and multivariate analysis, 110
tissue samples, 109110
vertical bore spectrometer, 110
multivariate techniques, 109
MYCN oncogene, 113
neuroblastoma and medulloblastoma cell lines,
112, 113
NMR technique, 109
phosphorylethanolamine, 111
principal component scores, 111, 112
retinoic acid treatment, 113
semi-automated methods, 109
taurine, 113
Hielscher, T., 135
High-grade gliomas (HGGs)
chemotherapy, 68
clinical presentations, 272
cortical pediatric, 67
CT and MRI, 272
diagnosis, 271, 272
differences, pediatric and adult, 277
EGFR pathway, 67
epidemiology, 271272
genetic factors, 168
germline mutations, 271
lesions classification, 272
Li-Fraumeni syndrome, 271
management, 169
molecular classification
astrocytic origin, 270
features, grade I-IV, 270
mechanisms and EGFR, 270
mutation, TP53, 270271
WHO, 270
MRS, 272273
neurofibromin function, 271
PET, 273
prognosis, 169
prognostic factors and outcomes, 276277
PTCH mutations, 271
risk factors, 271
signs and symptoms, 272
temozolomide administration, adults, 70
treatment, 274276
Turcot syndrome, 168
Higuchi, T., 264
Hilden, J.M., 9, 18
Index
Hile, S., 229
Hill, D.A., 279
Hiramatzu, E., 299
Hishiki, T., 95
Hoang, M.P., 138
Hochman, T., 151
Ho, D.M., 1517, 299
Hogg, T.L., 216
Holland, E.C., 261
Hollenbach, S.J., 263, 265
Holmes, E., 186, 191, 279
Ho, M., 264
Ho, R., 95
Horn, B., 18
Horowitz, M.M., 125
Hosoi, H., 19
Howard, R.G., 121
Howard, S.C., 238
Howman-Giles, R., 99
HPA. See Hypothalamic pituitary axis (HPA)
HR-MAS. See 1H High resolution magic angle spinning
NMR spectroscopy (HR-MAS)
Hsieh, K., 151
Hsueh, C., 299
Hsu, TR., 299
Huang, A., 264
Huang, H.J., 263
Huang, S., 241
Huber, P.E., 192, 238
Hudgins, R.J., 180, 208
Hudson, M.M., 241
Hugenholtz, H., 316
Hughes, D., 93, 95
Huk, W.J., 298
Human chorionic gonadotropin (HCG)
cortical dysplasia, 134
oligodendroglioma, 136
Hung, J., 297
Hung, P.C., 299
Hurwitz, C.A., 237
Hu, X., 261
Hwang, E.I., 61
Hyperfractionated accelerated radiotherapy (HART),
186187
Hypothalamic pituitary axis (HPA), 249, 251
I
Iannarelli, P., 265
IBED. See Integral biologically effective dose (IBED)
Ibrahim, G.M., 165
ICP. See Intracranial pressure (ICP)
Iehara, T., 19
Immunohistochemistry, adult and pediatric brain tumors
AT/RT, 134135
brachyury and diagnosis, chordoma
immunoreactivity, 139
vestigial notochordal remnants, 138
carcinoma, melanoma and sarcoma
cytokeratin (CK), 139
341
MITF, 140
S-100, 139
choroid plexus carcinoma, 135
CNS, 129
gliomas
DNT, 132
EMA, 132
GFAP, 131
gliosarcoma, 132
pleomorphic xanthoastrocytoma (PXA), 133
glioneuronal neoplasms
CD34, 134
pilocytic astrocytoma, 133
hemangioblastoma, 138
hemangiopericytomas (HPC) and SFT, 137
intracranial germ cell tumors
HCG, 136
PLAP, 136
medulloblastoma subtypes
sonic hedgehog (SHH) pathway, 135
tissue microarrays, 136
multiple schwannoma syndromes, 135
non-neoplastic tissue
inflammatory cells, 130
isocitrate dehydrogenase 1 (IDH1), 131
TP53 gene, 130
primary CNS lymphomas, 137
IMRT. See Intensity modulated radiation therapy (IMRT)
Ingle, A., 94
Integral biologically effective dose (IBED), 235
Intensity modulated radiation therapy (IMRT), 184
Intracranial pressure (ICP), 145
Intrathecal methotrexate (IT-MTX)
administration, 237
ALL, 238
neurotoxicity, 237
Intraventricular brain tumors
cerebrospinal fluid diversion
ETV, 159
hydrocephalus, 160
indications, 159
septostomy, 160
CSF, 156
endoscopic biopsy
CNS lymphoma, 157
coagulation, 158
coronal T1-weighted MRI, 157
obstructive hydrocephalus, 157
pathological diagnosis, 158
standard stereotactic techniques, 156
endoscopic resection
colloid cysts, 162
DNET, 161
histological grade, 160
ultrasonic aspirators, 161
vascularity, 161
intracranial disorders, 155
technology and instrumentation, 162
Ishida, E., 264
Ishii, Y., 261, 262
Index
342
Israel, M.A., 263, 265
IT-MTX. See Intrathecal methotrexate (IT-MTX)
Itoh, T., 132
Iwasaki, Y., 132
Izycka-Swieszewska, E., 89
J
Jackson, E.L., 262
Jacks, T., 265
Jacques, T.S., 139
Jaing, T.H., 299
Jakacki, R., 94, 186, 311, 315
Jakacki, R.I., 265
Jalali, R., 191
Jales, A., 264
Jallo, G.I., 160
James, C.D., 130
James, H.E., 298
Janss, A.J., 18, 224
Janzer, R.C., 275, 277
Jeibmann, A., 132
Jenkin, D., 298, 302
Jenkins, J.J., 298, 300
Jennings, M.T., 279
Jeunemaitre, X., 329, 330
Johnson, D.P., 122
Johnston, D.L., 223
Jones, C., 264
Jones-Wallace, D., 225
Joud, A., 138
Judkins, A.R., 33, 135
Jung, H.W., 299, 301
Junker, R., 120
Juvenile pilocytic astrocytomas (JPA), 67
K
Kalifa, C., 297, 298, 303
Kalpana, G.V., 19
Kamoshima, Y., 299
Kanaan, I., 298, 302
Kandil, A., 298, 302
Kant, J.A., 321, 329, 330
Kaplan, E.L., 311
Kasow, K., 225
Katnick, R., 14
Kato, T., 299
Katsumi, Y., 19
Katz, E., 242
Kazak, A., 242
Kazlauskas, A., 263
Keene, D.L., 223
Kellie, S.J., 185, 216, 225
Kepner, J.L., 14, 298
Kesler, S.R., 125
Keyvani, K., 133
Khafaga, Y., 298, 302
Khan, R.B., 238
Kieffer, V., 303

Kieran, M.W., 18, 190, 233234, 265, 299,
304, 305
Kieslich, M., 122
Kikuchi, K., 19
Kilic, T., 265
Kim, B.S., 160
Kim, H.J., 299, 301
Kim, H.S., 299, 301
Kim, I.H., 299, 301
Kim, R.Y., 160
Kim, S.K., 299, 301
Kim, Y.H., 133
Klareskog, L., 260
Klar, N., 237
Kleihues, P., 4, 133
Klein, O., 138
Kleinschmidt-DeMasters, B., 135
Klopfenstein, K.J., 1718
Klosky, J.L., 241
Knobbe, C.B., 261
Kocak, M., 279
Koch, H.G., 120
Kofide, A., 298, 302
Kohashi, K., 23
Komotar, R.J., 298
Konishi, N., 264
Kool, M., 216
Koopmann, J., 262
Koovakkattu, D., 125
Kooy, H., 190, 299
Koral, K., 13, 16, 17, 35
Korshunov, A., 135
Kortmann, R.D., 187, 298, 299
Kosnik, E., 1718
Koster, J., 216
Kothbauer, K., 160
Koutcher, J.A., 261
Krailo, M., 94
Krajinovic, M., 240
Kramm, C.M., 19
Krance, R., 225
Krasin, M.J., 18, 185, 225, 265
Krawiecki, N., 238
Kremer, L.C.M., 302
Krishna, M., 139
Kros, J.M., 131
Krull, K.R., 241
Kuehl, J., 187
Kulozik, A.E., 192
Kumar Srivastava, D., 241
Kun, L.E., 14, 18, 185, 189, 190, 265,
298, 300
Kurczynski, E., 191, 315
Kurkure, P., 191
Kuroda, H., 19
Kurtkaya-Yapicier, O., 130
Kurtzberg, J., 9
Kuwahara, Y., 19
Index
L
Lacaze, E., 303
Lacombe, D., 275, 277
Lafond, D., 279
Lafuente, J.V., 263
Laithier, V., 297, 298
Lake, D., 17
Lakeman, A., 216
Lamba, M., 235
Lamothe, A., 316
Lanfermann, H., 117, 122
Lange, B., 311
Lange, J., 7, 15
Langston, J.W., 298, 300
Laningham, F.H., 122
Laprie, A., 185
Larsson, E., 260
Lau, C.C., 132, 216, 225, 263
Lavally, B., 190, 299
Laverda, A.M., 304
Le Deley, M.C., 298
Lee, C.S., 299, 301
Lee, E., 264
Lefkowitz, I.B., 14, 53
Lehrnbecher, T., 122
Leisenring, W., 121, 241
Lemieux-Blanchard, E., 240
Lena, G., 278
Leukoencephalopathy vs. transient abnormalities
methotrexate (MTX), 122
radiation therapy, 122
white matter changes, 122
LGGs. See Low-grade gliomas (LGGs)
Liang, M.L., 264
Li, C., 121, 189
Li-Fraumeni syndrome, 271
Linehan, W,M., 329, 330
Lin, K.L., 299
Lirng, J.F., 1517
Listernick, R., 285
Little, S.E., 264
Liu, Y., 121
Liu, Z., 264
Loeffler, M., 133
Lokker, N.A., 263, 265
London, K., 99
Longatti, P., 160
Longee, D.C., 9
Lorenzi, C., 303
Louis, D.N., 262
Lovvorn, H. 3rd., 321
Lowe, J., 264
Low-grade gliomas (LGGs)
carboplatin and vincristine, 70
cisplatin/etoposide use
carboplatin, vinblastine and imatinib, 310
chemotherapeutic management, low-grade
gliomas, 315
diencephalic syndrome, 310
343
disease control, 317
evaluation, toxicity, 311
eyesight deficiencies, children, 316
intracranial tumours, 316
low drug concentrations, 316
outcomes, 311314
PFS, 317
prevention, hearing loss, 318
radiological responses, 317
risk, ototoxicity, 318
statistical analysis, 311
toxicity, 315
tumour location, 316
volume reduction, tumor, 316
complete resection, 67
focal radiotherapy, 68
grade I and grade II tumors, 166
mutation, BRAF, 67
neurocognitive toxicity, 168
NF1
autosomal dominant disorder, 286
brainstem, 292293
growth control signaling pathways, 287
molecular genetics, 286
optic pathway gliomas, 290291
treatment, 291292
optic pathway/chiasmatic gliomas, 167
prognostic factors, 168
Lucaya, J., 120
Ludwin, S.K., 275, 277
Luerssen, T., 191
Luksch, R., 225, 298
Lun, X., 19
Lustig, R.H., 191
Luther, N., 158, 161
M
Mabbott, D., 299, 300
Macarthur, D.C., 160, 161
Maccollin, M., 135
MacDonald, T.J., 191, 259, 264
Mackay. A., 264
Mack, S., 135
Madani, A., 119
Magnetic resonance imaging (MRI), AT/RTs
axial T2 weighted image, 1617
characteristic, 36
contrast enhancement, 3536
diffusion, 45
diffusion weight, 36
edema and tumor borders, 35
heterogeneous signal, T1 SE and T2 SE, 4245
in vivo proton, 45
intravenous gadolinium, 16
mean ADC value, AT/RTs, 17
medulloblastoma, 16
MRS, 36
344
Magnetic resonance imaging (MRI), AT/RTs (cont.)
NAA, 46
perfusion, 45
signal intensities, 3435
spinal drop metastases, 17
tumor size, 35
Magnetic resonance spectroscopy (MRS), 272273
Mah, M., 185
Mahoney, D.H., 298
Maignon, P., 185
Maintz, D., 262
Maire, J.P., 185
Ma, J., 264
Majhail, N.S., 125
Malignant peripheral nerve sheath tumors (MPNST), 100
Mallucci, C.L., 160
Mandelbaum, D.E., 279
Mani, S., 19
Manley, P.E., 18
Marcus, K.J., 18, 190, 299
Marengo, I., 299
Mariani, L., 133
Marie, B., 138
Maris, J., 95
Marosi, C., 275, 277
Martin, S.E., 130, 139
Marx, M., 298
Mascari, C., 160
Mason, W.P., 275, 277
Massey, V., 279
Massimi, L., 295
Massimino, M., 225, 298, 309
Matsumoto, Y., 261, 262
Mazewski, C., 18
Mazza, E., 298
MB/PNETs. See Medulloblastoma/primitive
neuroectodermal tumors (MB/PNETs)
McCarter, R., 264
McCowage, G., 18
McFadden, G., 19
McGuire-Cullen, P., 191
McGuire, W., 279
McLaughlin, M.E., 265
McNeely, L., 279
McNeil, E., 121
Meacham, L., 247
Meazza, C., 225
Meco, D., 263
Medulloblastoma/primitive neuroectodermal tumors
(MB/PNETs), 14
Medulloblastomas
adjuvant radiation, 149150
chemotherapy, 150
primitive neuroectodermal tumors, 148
Meerbaum, S., 18
Mege, M., 185
Mehta, M., 186
Mehta, V., 132
Meningeal dissemination, AT/RTs
Index
ADC-image, 33, 36
bifocals and spinal cord, 33
bony destruction, coronal CT, 33, 35
elevated intracranial pressure, signs, 32
frequency, meningeal dissemination, 33
infra/supratentorial position, 32
ratio, 32
Merchant, T.E., 18, 185, 189191, 225, 298, 300
Mercola, D., 261
Mertens, A.C., 241, 242
Metabolic syndrome, 248, 250251
Metabolite profile differences, childhood brain tumors
description, 108
gene expression, 108
HR-MAS (see 1H High resolution magic angle
Spinning NMR spectroscopy (HR-MAS))
medulloblastomas, 108
tumour genetics, 108
Meta-iodobenzylguanidine (MIBG) scintography, 326
Methylphenidate hydrochloride (MPH)
efficacy, 243
pediatric cancer survivors, 243
Metreweli, C., 121
Meye, A., 27
Meyer, J., 131
Meyermann, R., 131, 133
Meyer, P., 311
Microphthalmia transcription factor (MITF), 140
Mikhael, N.Z., 316
Miller, D.C., 186
Miller, N.R., 303
Minturn, J., 95
Miras, M., 299
Mirimanoff, R.O., 275, 277
Mitosis/karryorhexis index (MKI)
EGFR expression, 93
Shimada classification system, 90
Mittelbronn, M., 133
Moertel, C.L., 9
Moghrabi, A., 240, 298
Moharir, M., 99
Molepo, J.M., 316
Molyneux, A.J., 139
Montagne, K., 138
Montague, E., 229
Montpetit, V.A., 316
Moore, B., 234
Moreno, L., 304
Morgenstern, P.F., 143
Morrison, A., 264
Morris, R., 238
Mortazavi, M.M., 179, 207
Moschos, D., 238
Moulton, T., 279
Moya-Moya disease, 303
MPH. See Methylphenidate hydrochloride (MPH)
MPNST. See Malignant peripheral nerve
sheath tumors (MPNST)
MRS. See MR-spectroscopy (MRS)
Mrsic, A., 216
Index
MR-spectroscopy (MRS), 36
Mueller, W., 131
Mulhern, R.K., 238, 242, 298, 300
Mullenix, P.J., 237
Multiple schwannoma syndromes, 135
Munshi, A., 191
Muraszko, K.M., 186, 297
Murat, A., 329
Murgo, A., 265
N
Nabeshima, Y., 261, 262
Nagashima, K., 132
Nakamura, M., 264
Nakase, H., 264
Nakazato, Y., 133
Navarria, P., 298
NBs. See Neuroblastic tumor (NBs)
Needles, M.N., 311, 315
Ness, K.K., 241
Neuroblastic tumor (NBs)
genesis and progression, cancers, 90
HER, cancer, 9192
HER family, neuroblastic tumors
coexpression profiling, 97
EGFR, 9294
HER2, 9495
HER3, 95
HER4, 9597
HER, peripheral nervous system
aberrant structure, 91
heart, 91
HER receptor characteristics, 9091
mitosis/ karryorhexis index (MKI), 90
NMYC amplification, 90
pediatric embryonal neoplasms, 89
spontaneous regression, 90
treatment modalities, 90
Neurocognitive deficits, cancer survivors
child CNS, 233239
diagnoses
brain tumors, 232233
leukemia, 232
domains, neurocognitive function and measures,
230, 231
functional outcomes, late effects
health status, 240241
neurocognitive, 240
psychosocial functioning, 241242
underachievement/underemployment, 241
health-related late effects, 230
interventions
academic, 243244
cognitive remediation, 242243
ecological, 244
pharmacological, 243
objectives, 231232
randomized trials, 244
variables affecting, late effects
345
child, 239
environmental, 240
genetic, 240
treatment staging, 239240
Neurofibromatosis type 1
autosomal dominant neurocutaneous syndrome, 99
cellular proliferation, 100
cerebral glucose utilisation
characterisation, CNS lesions, 100, 101
FDG PET scans, 100, 101
gliomas, locations, 9
imaging modalities, 100
MPNST, 100
OPG, 101102
T2 hyperintense lesions (T2H), MRI, 103
Neurosurgeon role, children with brain tumors
age, patient, 146
biopsy, 146147
brain stem tumors, 152
craniopharyngioma, 150151
cytoreductive surgery, 146
ependymoma, 150
gliomas, 148
HGG, 148
hydrocephalus
CSF, 145
ETV, 145
EVD, 145
ICP, 145
symptoms, infancy, 144
ventriculoperitoneal (VP) shunt, 145
LGG, 148
medulloblastoma, 148150
neurooncological care, 144
neurosurgical care, 153
operative techniques, 146
pediatric neurooncology, 143
pineal region tumors, 151152
second look surgery, 147
staging, 147
sub-specialization, 144
tissue procurement, 147148
Neurosurgery
brain mapping, 68
COG protocol, 6768
description, 67
LGGs, 67
Neurosurgical management, pediatric brain tumors
CNS neoplasms, 165
DTI, 166
ependymoma, 172174
HGG, 168169
LGG, 166168
medulloblastoma
extraneural metastasis, 174
MRI, 175
prognosis, 175176
radiation therapy, 175
MRI, 166
346
Niccoli-Sire, P., 329
Nicholson, H.S., 311, 315
Nicolin, G., 299, 300
Nishihara, T., 160
Nissen, J., 261, 262
Nistr, M., 262
Nobusawa, S., 133
Noll, R.B., 242
Non-small-cell lung cancer (NSCLC), 316
Northcott, P.A., 135
North, K., 99
Nowak-Gottl, U., 120
O
Oakley, G.J., 139
OBrien, D.F., 160
OBrien, R., 191
Oda, Y., 23
Odell, E., 139
Odom, L., 242
Oeffinger, K.C., 121
Oertel, J., 158
Oghaki, H., 133
Oh, D., 130, 139
Ohgaki, H., 133
Oh, K.S., 297
Ohnishi, A., 132
Oi, S., 160
Oldfield, E.H., 262
OLeary, M., 321
Oliv, T., 120
Olivi, A., 298
Olson, J., 279
Olson, T.A., 1718
Omeis, I., 161
OPGs. See Optic pathway gliomas (OPGs)
Opocher, E., 302, 304
Optic pathway gliomas (OPGs)
FDG accumulation, 101102
FDG PET appearance, symptomatic OPG, 102
radiotherapy, 103
Optic pathway/suprasellar tumors
craniopharyngiomas, 64
GCTs, 64
Ordinas, A., 181
OReilly, T., 265
Ortega, J.J., 120
Ostman, A., 260
Otabe, O., 19
Ottensmaier, H., 187
Overall survival (OS), 14
P
Pacak, K., 329, 330
Packer, R.J., 9, 14, 31, 53, 61, 121, 191, 224, 265,
311, 315
Padovani, L., 185
Paediatric oncology group (POG) protocols, 8
Index
Palmer, N.F., 4, 14
Palmer, S., 225
Paragangliomas (PGL)
clinical presentation, 325
definition, malignancy, 322
diagnosis
biochemical, 325326
clinical vignette, 326327
CT/MRI, 326
drugs, 325
FDG PET, 326
intrabominal parangaliomas, 326
MIBG scintography, 326
sequence data, SDBG LOH, 326, 328
testing, 325
disorders, 322
fullow-up, 329330
function, 322
genetics
SDH, 322323
syndromes, 323324
staging and prognosis, 327
treatment
alpha-adrenergicblockade, 327
anethesis and surgical manipulation, 327
beta-adrenergicblockade, 327, 329
biopsy, 327
clinical vignette, 327, 329
surgery, 327
tumor suppressors mechanism, SDH, 324325
Parkin, P., 299, 300
Park, R., 181
Parmar, H., 1517, 35
Pasha, T.L., 140
Patel, A., 225
Patel, S.K., 242
Patil, S., 135
Paugh, B.S., 264
Paulus, W., 133
PBTC. See Pediatric Brain Tumor Consortium (PBTC)
PDGFR. See Platelet-derived growth factor receptor
(PDGFR)
Pediatric Brain Tumor Consortium (PBTC)
and COG, 214
SHH inhibitor GDC-0449, 218
Pediatric oncology group (POG), 14
description, 295296
Dodge I, 296
low grade, 296
natural history
defnition, behavior, 296
hamartomatous, 296
mechanisms, 296
MRI, 296
NF-1 and PFS, 297
prediction, 296
progression, 296297
radiological features, 296
survival, 297
Index
NF-1, 296
outcome and complications
advantages, CT, 303
biopsy, 302
carcinogenic effects, 303
CSF shunting devices, 303
differences, prognosis and outcome, 304
hydrocephalus, 302303
impairment and irradiation, 303
long-term outcomes, 303304
mortality, 302
postoperative acute worsening, 302
role, RT, 303
prognostic factors
age, 300
histological features, 301
hydrocephalus, 301
management, 301302
NF-1, 300301
statistical value, 302
tumor location, 300, 301
survival and tumor progression
investigation, 300
mortality, 297
PFS, 297298
risk factors, 297
synopsis, 298299
therapy administration, 300
Pedziwiatr, K., 183
Peet, A., 107
Pena, M., 299
Perilongo, G., 302, 304
Perlaky, L., 263
Perlman, E.J., 14
Perry, A., 33, 135
Peyrard, S., 330
Pfister, S., 135
PGL. See Paragangliomas (PGL)
Phillips, P.C., 191, 224, 265, 311, 315
Pichon, F., 298
Pierre-Kahn, A., 297
Pieters, R.S., 1718
Pietsch, T., 133, 187, 216, 298, 299
Pignoli, E., 225
Pinello, M,L., 304
Piqueras, J., 120
Pizer, B., 160
Placental alkaline phosphatase (PLAP)
embryonal carcinomas, 136
placental syncytiotrophoblasts, 136
PLAP. See Placental alkaline phosphatase (PLAP)
Plate, K.H., 263
Platelet-derived growth factor receptor (PDGFR)
angiogenesis
endothelial cells, 263
in situ hybridization, 263
VEGF, 263
glioma pathogenesis
CISs, 261
haplotypes, promoter region, 261262
347
in vitro studies, 260
molecular genetic alterations, 261
T98G human glioblastoma cells, 261
and neural stem cells, 262
pathway expression
amplification and overexpression, PDGFRA, 262
functional PDGF/PDGFR pairs, 263
pediatric gliomas
immunohistochemistry, 264
PDGFRA amplification, 264
SNP allele arrays, 263
signaling
alpha and beta-receptor, 260
dimeric isoforms, 260
ligand-induced dimerization, 260
therapeutic targeting
autocrine signaling, C6 GBM cells, 265
experimental model systems, 265
metastatic pilocytic astrocytoma, 265
PDGFR blockade, 264
Plouin, P.F., 329, 330
PNET/MBs. See Primitive neuroectodermal
tumors/medulloblastomas (PNET/MBs)
PNETs. See Primitive neuroectodermal tumors (PNETs)
Podda, M., 225
Podo, F., 113
POG. See Pediatric oncology group (POG)
Poggi, G., 225
POHGs. See Pediatric optic-hypothalamic gliomas
(POHGs)
Polastri, D., 225
Pollack, I.F., 211, 265, 279
Pomeroy, S.L., 135, 190, 233234, 265, 299, 304, 305
Popovic, P., 316
Porto, L., 117, 122
Positive emission tomography (PET), 273
Poskitt, K., 298
Posterior fossa, 63
Posterior reversible encephalopathy syndrome (PRES)
cyclosporine A (CsA), 123
MR imaging, 123
neuroimaging abnormalities, 123
pediatric patients, 123
Postoperative pain
craniotomy, 208
description, 207208
dorsal rhizotomy, 209
dorsal roots, 208
epidural morphine, 208209
neurosurgical procedures, 209
Potapova, O., 261
Poussaint, T.Y., 265
Powers, J.M., 14
Prados, M., 186, 311
Prasad, P.K., 321
Prayson, R.A., 130, 139
Preibisch, C., 122
PRES. See Posterior reversible
encephalopathy syndrome (PRES)
Presneau, N., 139
348
Primeau, M., 240
Primitive neuroectodermal tumor/medulloblastomas
(PNET/MBs)
AT/RT, 40
CT and MRI, 4849
hSNF5 gene, 8
leptomeningeal spread and portions, 6
multiform glioblastoma, 42
progression-free survival, 8
rCBV values, 45
Primitive neuroectodermal tumors (PNETs)
anaplastic histological features, 217
average-risk and high-risk tumors, 216217
disease control, 217
genomic studies, 216
isotretinoin, 218
sequelae, 217
SHH pathway, 218
WNT signaling, 216
Primitive neuroectodermal tumors/medulloblastomas
(PNET/MBs), 3132
Puccetti, D., 186
Pui, C.H., 122, 238
Punt, J., 160, 161
Pusch, S., 131
Q
Qu, C., 264
Quetin, P., 185
Quinones-Hinojosa, A., 262
R
Raaphorst, G.P., 316
Radiotherapy, pediatric brain tumors
ependymoma
HIT-SKK 87 and 92 trial, 189
hyperfractionated radiotherapy, 188
postoperative irradiation, 188
survival rate, 190
germ-cell tumors
chemotherapy, 193
nongerminomatous, 192, 194
spinal recurrence, intracranial germinoma, 193
high-grade and brain stem glioma
HIT-GBM-C, 191
nonhematological serious toxicity, 191
survival, 1952
temozolamide, 191192
low-grade gliomas
CT and MRI fusion, 190
resection, 190
stereotactic radiotherapy, 190
vasculopathy, 191
medulloblastoma
craniospinal irradiation (CSI), 185, 186
French Society for Pediatric Oncology (FSPO), 185
HART, 186187
Index
hyperfractionated radiotherapy, 185186
intravenous chemotherapy, 187
multicenter randomized clinical trials, COG, 184
preirradiation chemotherapy, 186
reduction craniospinal radiotherapy dose, 185
survival rate, 188
UKCCSG/SIOP CNS 9204 trial, 188
Raghunathan, A., 129
Raquin, M.A., 297, 298
Rau, N., 191
Ravagnani, F., 225
Ray, P., 160
rCBV. See Relative cerebral blood volume (rCBV)
Reaman, G., 311, 315
Reardon, D., 298
Reddick, W.E., 122, 238
Reddy, A.T., 224
Redmond, M.D., 316
Reifenberger, G., 133, 261
Reis, R.M., 264
Rekate, H., 160
Relative cerebral blood volume (rCBV), 45
Reuss, D., 132
Rhabdoid tumor predisposition syndrome (RTPS), 32
Rhabdomyosarcoma
cell-cycle regulators alteration (see Cell-cycle
regulators, rhabdomyosarcoma)
description, 24
RB and p53 pathways, 24
regulate signal transduction pathways, 24
WHO classification, 24
Riccardi, R., 263, 298
Richard, S., 329
Richards, K.N., 93, 95
Riegel, T., 158, 160
Risau, W., 263
Ris, M.D., 235
Riva, D., 298
Rizzo, J.D., 125
Robaey, P., 240
Robertson, I.J., 160, 161
Robertson, J.T., 262
Robertson, P.L., 186, 297
Robison, L.L., 121, 241, 242
Robson, C.D., 265
Rodel, C.M., 298
Rodriguez, D., 297, 298
Roebuck, D.J., 124
Rogers, H.A., 75
Rohmer, V., 329
Rollins, N.K., 16, 17, 35
Rnnstrand, L., 260
Rorke-Adams, L.B., 18
Rorke, L.B., 7, 9, 14, 15, 18, 31, 53, 134, 135, 186
Roseau, N.F., 298
Rosenblum, M.K., 135, 136, 261
Rose, S.R., 298, 300
Rossati-Bellani, F., 225
Rossi, M., 263
Rouillard, M., 240
Index
Rousseau-Merck, M.F., 7, 15
Roy, M., 262
RTPS. See Rhabdoid tumor predisposition
syndrome (RTPS)
Rubie, H., 298
Rubin, J., 18
Rubin, K., 260
Rubinstein, L.J., 14
Ruchoux, M.M., 298
Ruggiero, A., 298
Ruijne, N., 181
Rumboldt, Z., 17
Rushing, E.J., 264
Rush, L., 279
Russo, P., 15
Rutka, J.T., 1517, 35, 135, 160, 165, 264, 299, 300
Rutkowski, S., 19, 187
Ruymann, F.B., 1718
Ryan, J., 311
Ryan, P.M., 235
S
Sadoul, J.L., 329
Sahler, O.J., 242
Sahm, F., 131, 132
Sainz, P., 120
Sakaki, T., 264
Sallan, S.E., 237
Snchez-Toledo, J., 120
Sandberg, D.I., 155
Sanford, R.A., 189, 190, 298, 300
Sanson, M., 298
Santi, M., 264
Sarat, C.P., 134
Sariban, E., 298
Sarin, R., 191
Sarkar, C., 134
Sasahara, M., 261, 262
Sauer, R., 298
Saunders, T.G., 298
Sawa, H., 132
Sawamura, Y., 132, 299
Saxena, A., 262
Scarcella, D.L., 298
Schackert, G., 133
Scheithauer, B.W., 9, 18, 130
Schiavello, E., 309
Schittenhelm, J., 131, 132
Schlumberger, M., 329, 330
Schobess, R., 120
Schouten-van Meeteren, N., 216
Schroeder, H.W., 158, 160, 161
Schuchardt, U., 298
Schultz-Ertner, D., 192
Schultz, H., 298, 302
Schwabe, D., 120
Schwenn, M., 237
Scot, R.M., 233234
Scott, I.U., 298
349
Scott, R.M., 190, 299, 304, 305
Seethala, R.R., 139
SEGAs. See Subependymal giant cell astrocytomas
(SEGAs)
Senger, D.L., 19
Sevenet, N., 7, 15
Sexton, M., 185
SFT. See Solitary fibrous tumor (SFT)
Shahideh, M., 165
Shah, N., 191
Shan, Z.Y., 238
Sharma, M.C., 134
Shen, V., 279
Sherman, E.M.S., 231
Sherrod, A ., 130, 139
SHH. See Sonic hedgehog (SHH)
Shih, A.H., 261
Shimada, K., 264
Shroff, M., 1517, 35
Shulz-Ertner, D., 238
Siatowski, R.M., 298
Siegel, K., 14
Silvera, S., 330
Silverman, L.B., 237
Simonetti, F., 225
Simon, M., 133
Skowronska-Gardas, A., 183
Sobecks, R.M., 125
Socie, G., 125
Soerensen, N., 187
Solitary fibrous tumor (SFT), 137
Solomon, L., 264
Sonic hedgehog (SHH), 218
Sonzini Astudillo, B., 299
Souweidane, M.M., 143, 158, 160, 161
Speleman, F., 93, 95
Spennato, P., 160
SPNETs. See Supratentorial primitive neuroectodermal
tumors (sPNETs)
Sposto, R., 186, 191, 279
Spreafico, F., 225, 298
Spreen, O., 231
Stawski, R., 133
Steck, I., 192
Steegers-Theunissen, R.P., 261
Stefanko, S.Z., 131
Steinbok, P., 191, 298
Stemmer-Rachamimov, A.O., 135
Stenstrom, C., 7, 15, 134
Stewart, D.J., 316
Stiles, C.D., 265
Stovall, M., 121, 241
Strain, J., 191
Strauss, E., 231
Streri, A., 303
Striano, P., 199
Strompf, L., 329
Strother, D.R., 14, 225
Stupp, R., 275, 277
Sturdgess, I.C., 139
Index
350
Surez, J.C., 299
Subependymal giant cell astrocytomas (SEGAs), 212
Succinate dehydrogenase (SDH), PGL
Krebs cycle, 322
non-maligant paragangliomas, 323
recognition, SDBH diseases, 323
SDHC and SDHD, 323
subunits and mutation, 322323
tumor suppressors mechanism
description, 324
inhibition, PHD, 324325
loss of function, VHL, 324
stabilization, HIF-1a, 324
Succop, P., 235
Sugimoto, T., 19
Su, J., 263
Sullivan, C.M., 263, 265
Sun, B., 19
(SPNETs)
chemotherapy
cycles, five, 225
infants, 224
survival rates, 224
defined, 223224
radiation therapy
craniospinal, 225
German brain tumor trials, 226
uses, 225, 226
Sure, U., 133
Sutton, L., 14
SWI/SNF-related, Matrix-associated, Actin-dependent
Regulator of Chromatin, subfamily B,
member 1 (SMARCB1), 19
T
Tabori, U., 299, 300
Taguchi, T., 23
Tajima, T., 299
Takahashi, Y., 23
Takei, H., 132
Tamber, N., 264
Tamburrini, G., 160
Tamura, S., 19
Tanaka, H., 125
Tanaka, Y., 133
Tang, R.B., 299
Tan, L., 15
Tao, J.J., 329, 330
Taphoorn, M.J., 275, 277
Tarbell, N.J., 190, 233234, 237, 299,
304, 305
Taubert, H., 27
Taylor, K.D., 316
Taylor, M.D., 135, 216
Tekautz, T.M., 18
Tekhtani, D., 16, 35
Teng, M.M., 1517
Teo, C., 160

Terask, J., 94
Terenziani, M., 225, 298
Termuhlen, A., 241
Terracio, L., 260
Tesoro-Tess, J.D., 298
Thien, P.F., 299
Thilmann, C., 238
Thompson, M.C., 216
Thompson, S., 225
Thomson, J., 191
Thorarinsdottir, H.K., 264
Tien, R.D., 124
Tihan, T., 14, 298
Timmins, C.F., 35
Timmons, C.F., 16, 17
Tingstrm, A., 260
Tirabosco, R., 139
Tirakotai, W., 158, 160
Toepoel, M., 261
Tourt-Uhlig, S., 298
Tow, S.L., 303
Toxicity
LGG, 315
low-grade gliomas (LGG), 315
Tramunt, B., 299
Trevino, J., 93, 95
Tripathi, M., 134
Troost, D., 216
Tsang, Y.T., 263
Tsay, P.K., 299
Tsen, C.K., 299
Tsubaki, J., 299
Tsuchiya, K., 19
Tsuda, M., 132
Tubbs, R.S., 179, 207
Tu, P.H., 136
Turner, C.D., 18
U
Ullrich, N.J., 18
V
Valentini, L., 298
Van Calenbergh, F., 19
van, den, Bent, M.J., 275, 277
VandenBerg, S., 262
van der Kwast, T.H., 131
van de Wetering, M.D., 302
Van Eden, C.G., 131
Van Gool, S.W., 19
Vannier, J.P., 298
Van Roy, N., 93, 95
van Sluis, P., 216
van Zoelen, E.J., 261
Vsquez, E., 120
Veerman, A.J., 232
Index
Versteege, I., 7, 15
Versteeg, R., 216
Vezina, G., 186, 191
Vezina, L.G., 279
Viana-Pereira, M., 264
Viano, J.C., 299
Vignaud, J.M., 138
Vimentin membrane antigen (VMA), 47
Viscardi, E., 302, 304
Vital, A., 133
Vloeberghs, M., 160, 161
von Deimling, A., 131133, 262
von Hippel Landau syndrome, 323
von, Hornstein, S.
Voss, S., 94
Vujovic, S., 139
W
Waayer-Van, B.M., 131
Waber, D.P., 233234, 237
Waha, A., 262
Wainwright, L.M., 7, 15, 134
Walker, R., 311
Wallace, D., 18, 185, 225, 298, 300
Walter, A.W., 18, 125
Wang, D., 263
Wang, K.C., 299, 301
Wang, L., 19
Wang, Q., 95
Wang, Y., 265
Wang, Z., 125
Warf, B., 160
Warmuth-Metz, M., 31, 187, 299
Watanabe, R., 261, 262
Watanabe, T., 133
Watral, M., 298
Watterson, J., 9
Weber, D.P., 304, 305
Webster, M.W.I., 181
Weinbreck, N., 138
Weiner, S., 265
Weiss, H.L., 224
Weller, M., 133, 275, 277
Welsh, C.T., 17
Welzer, T., 192
Wenzel, D., 298
Weprin, B., 16, 17, 35
Wermes, C., 120
Westermark, B., 262
Westphal, M., 133
Whitton, J., 242
Wick, W., 133
Wiestler, O.D., 262
Willard, V.W., 243
Wills, M., 321
Wilson, J.D., 136
Wilson, M., 107
Wilson, S., 160
351
Wingard, J.R., 125
Wisoff, J., 151
Wisoff, J.H., 151
Witt, H., 135
WNT/B-catenin pathway, CNS PNET
immunohistochemistry
MKI67 levels, 79, 80
nuclear staining, 79
primary tumors, 78
mutational analysis
amino acid substitutions, 81
tumor types, 81
survival analysis
Fishers exact test, 81
Kaplan Meier curves, 81, 82
Wolff, J.E.A., 187
Wolkenstein, P., 298
Wollf, J.E., 19
Wong, K.K., 132, 263
Wong, P.T., 316
Wong, T.T., 1517, 299
Woo, S.Y., 185, 225
Wozniak, A., 89
Wrede, K., 133
Wu, C.T., 299
Wrl, P., 27
Wu, S., 190, 191, 238
Wu, Y., 19
X
Xiong, X., 189191, 238
Xu, R., 190, 299
Xu, X., 140
Y
Yachnis, A.T., 135
Yang, E., 321
Yang, S.W., 299, 301
Yao, X., 18
Yates, A., 191
Yeh, T.H., 135
Yeung, D.K.W., 122
Ylstra, B., 216
Yogeswaren, S.T., 132
Young Poussaint, T., 190, 299
Yousem, D.M., 16, 35
Yu, I.T., 14
Z
Zacharoulis, S., 304
Zage, P., 93, 95
Zanella, F., 122
Zarghooni, M., 264
Zdunek, P.R., 265
Zebrack, B.J., 242
Zeltzer, L.K., 242
Index
352
Zeng, I.S.L., 181
Zentgraf, H., 131
Zerah, M., 298
Zhang, F., 15
Zhang, J., 264
Zhang, P.J., 140
Zhang, Z., 19
Zhao, H., 95
Zhao, W., 264

Zhou, H., 19, 135
Zhou, J.Y., 7, 15, 134
Zimmerman, M.A., 18
Zotz, R.B., 261
Zunino, S., 299
Zweidler-McKay, P.A., 93, 95
Zymberg, S.T., 160

Pediatric Cancer v2 Teratoid Rhabdoid (UnitedVRG)

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

Pediatric Cancer v2 Teratoid Rhabdoid (UnitedVRG)

Uploaded by

Copyright:

Available Formats

Pediatric Cancer

Pediatric Cancer, Volume 2

For further volumes:

Diagnosis, Therapy, and Prognosis

Although touched by technology, surgical pathology always

One Point of View

This is the second volume in the series, Pediatric Cancer. Specifically, it

Endoscopic neurosurgical techniques for treating pediatric intraventricular

relevant brain malignancy. There exists a tremendous, urgent demand by the

Pediatric Atypical Teratoid/Rhabdoid Tumors

Pediatric Atypical Teratoid/Rhabdoid Tumor.............................

Pediatric Rhabdomyosarcoma: Role of Cell Cycle

Pediatric Atypical Teratoid/Rhabdoid Tumors:

Part II Brain Tumors (General)

Pediatric Brain Tumors (An Overview) .......................................

Pediatric CNS Primitive Neuroectodermal Tumor:

Neuroblastic Tumors Status and Role of HER

Children with Neurofibromatosis Type 1:

Metabolite Profile Differences in Childhood Brain

Central Nervous System Imaging in Childhood Leukemia ....... 117

Immunohistochemistry in the Differential Diagnosis

Children with Brain Tumors: Role of the Neurosurgeon ........... 143

Pediatric Intraventricular Brain Tumors: Endoscopic

Neurosurgical Management of Pediatric Brain Tumors ............ 165

Pediatric Brain Tumor Biopsy or Resection:

Clinical Trials in Pediatric Brain Tumors; Radiotherapy.......... 183

Epileptic Seizures and Supratentorial Brain

Postoperative Pain in Children: Advantage of Using

Pediatric Brain Tumors: Application of Stratification

Pediatric Supratentorial Primitive Neuroectodermal Tumor:

Pediatric Cancer Survivors: Neurocognitive Late Effects ......... 229

Adult Survivors of Pediatric Cancer:

Part III Gliomas

Pediatric Glioma: Role of Platelet-Derived Growth

An Overview of Pediatric High-Grade Gliomas

Pediatric Low-Grade Glioma: The Role

Treatment of Pediatric Optic-Hypothalamic Gliomas:

Pediatric Low-Grade Gliomas: Advantage of Using

Pediatric Paragangliomas: Role of Germline

Pediatric Neuroblastoma: Treatment With Oral Irinotecan

Faiz Ahmad Department of Neurological Surgery, University of Miami

Leanne Embry Pediatric Hematology/Oncology, University of Texas Health

Heinrich Lanfermann Institut fr Neuroradiologie, Klinikum der Johann

Kathryn North Department of Nuclear Medicine, Childrens Hospital at

Luciana Porto Institut fu r Neuroradiologie, Klinikum der Johann Wolfgang

Pediatric Atypical Teratoid/

Clinical Presentation of CNS AT/RT ....................

Imaging of AT/RT ..................................................

Gross and Microscopic Pathology

Molecular Pathology of AT/RT .............................

Treatment of AT/RT: Surgery ...............................

Treatment of AT/RT: Chemotherapy ...................

Treatment of AT/RT: Radiotherapy .....................

Pediatric Atypical teratoid/rhabdoid tumors

K.K. Bansal (*) D. Goel

Pediatric Atypical teratoid/rhabdoid tumors

M.A. Hayat (ed.), Pediatric Cancer, Volume 2,

Beckwith and Palmer In 1978 first described a

K.K. Bansal and D. Goel

(Rorke et al. 1996; Gyure 2005) although it is not

Clinical Presentation of CNS AT/RT

Pediatric Atypical Teratoid/Rhabdoid Tumors (An Overview)

multifocal (Rorke et al. 1996; Bhattacharjee et al.