CLINICAL PRACTICE

COMPANION
An NEJM Special Collection
 Limited Time Free Clinical Practice Companion

We are pleased to present you with a collection of recent articles relevant to todays practicing
physician. NEJM is committed to helping physicians improve patient care by providing the best
current information in an understandable and clinically useful format . These regular features, Clinical
Practice, Clinical Therapeutics, and Current Concepts, demonstrate how to diagnose and manage
diseases, keep you up to date with the latest in clinical care, and highlight the best treatments and
management options. This collection will only be available for a limited time, with free access available
by clicking through from the Special Collections tab in My NEJM.
 Table of Contents

Table of Contents

CLINICAL CLINICAL CURRENT OTHER CLINICAL

4 PRACTICE 51 THERAPEUTICS 100 CONCEPTS 131 RESOURCES
VIDEOS IN CLINICAL MEDICINE
CLINICAL PRACTICE CLINICAL THERAPEUTICS REVIEW ARTICLE

5 Gout 52 Iron-Chelating
101 Point-of-Care 132 Clinical Evaluation
of the Knee
T. Neogi Therapy for Transfusional Ultrasonography
February 3, 2011 Iron Overload C.L. Moore and J.A. Copel T.L. Schraeder and others
July 22, 2010
G.M. Brittenham February 24, 2011
CLINICAL PRACTICE January 13, 2011
15 Calcium
INTERACTIVE MEDICAL CASE
Kidney Stones REVIEW ARTICLE
E.M. Worcester CLINICAL THERAPEUTICS
110 Myocardial Infarction 137 Lying Low
J.J. Ross and others
and F.L. Coe
September 2, 2010
63 Bisphosphonates
for Osteoporosis
Due to Percutaneous
Coronary Intervention February 10, 2011
M.J. Favus A. Prasad and J. Herrmann
CLINICAL PRACTICE November 18, 2010 February 3, 2011
25 Emergency
of Asthma
Treatment
CLINICAL THERAPEUTICS REVIEW ARTICLE
S.C. Lazarus
August 19, 2010
72 Ranibizumab Therapy
for Neovascular
122 MDR Tuberculosis
Critical Steps for
Age-Related Macular Prevention and Control
CLINICAL PRACTICE Degeneration E. Nathanson and others
35 Early Alzheimers Disease
R. Mayeux
J.C. Folk and E.M. Stone
October 21, 2010
September 9, 2010

June 10, 2010

CLINICAL THERAPEUTICS
CLINICAL PRACTICE
Helicobacter pylori
80 Dietary Therapy
in Hypertension
43 Infection F.M. Sacks and
K.E.L. McCol H. Campos
April 29, 2010 June 3, 2010

CLINICAL THERAPEUTICS

91 Mitral-Valve Repair
for Mitral-Valve Prolapse
S. Verma and T.G. Mesana
December 3, 2009

BACK to TOC
 Clinical Practice

Clinical Practice articles begin with the presentation of a single case and continue to provide a
complete description of diagnostic and treatment strategies, therapeutic options, areas of uncertainty,
treatment guidelines everything you need to know about the current state of knowledge about
a common condition. These articles are also available in audio format, so you can listen at your
computer or download articles for transfer to any iPod or MP3 player.

BACK to TOC
4
 The n e w e ng l a n d j o u r na l of m e dic i n e

clinical practice

Gout
Tuhina Neogi, M.D., Ph.D.

This Journal feature begins with a case vignette highlighting a common clinical problem.
Evidence supporting various strategies is then presented, followed by a review of formal guidelines,
when they exist. The article ends with the authors clinical recommendations.

A 54-year-old man with crystal-proven gout has a history of four attacks during the
previous year. Despite receiving 300 mg of allopurinol daily, his serum urate level is
7.2 mg per deciliter (428 mol per liter). He is moderately obese and has hypertension,
for which he receives hydrochlorothiazide, and his serum creatinine level is 1.0 mg
per deciliter (88 mol per liter). How should his case be managed?

The Cl inic a l Probl em

Symptoms and Prevalence

Gout is a type of inflammatory arthritis induced by the deposition of monosodium From the Section of Clinical Epidemiolo-
urate crystals in synovial fluid and other tissues. It is associated with hyperurice- gy Research and Training Unit, Boston
University School of Medicine; and the
mia, which is defined as a serum urate level of 6.8 mg per deciliter (404 mol per Department of Epidemiology, Boston
liter) or more, the limit of urate solubility at physiologic temperature and pH.1 Hu- University School of Public Health
mans lack uricase and thus cannot convert urate to soluble allantoin as the end prod- both in Boston. Address reprint requests
to Dr. Neogi at the Clinical Epidemiology
uct of purine metabolism. Hyperuricemia that is caused by the overproduction of Unit, Boston University School of Medi-
urate or, more commonly, by renal urate underexcretion is necessary but not suffi- cine, 650 Albany St., Suite X-200, Boston,
cient to cause gout. In one cohort study, gout developed in only 22% of subjects with MA 02118, or at tneogi@bu.edu.

urate levels of more than 9.0 mg per deciliter (535 mol per liter) during a 5-year N Engl J Med 2011;364:443-52.
period.2 Copyright 2011 Massachusetts Medical Society.
Gout has two clinical phases. The first phase is characterized by intermittent
acute attacks that spontaneously resolve, typically over a period of 7 to 10 days, with
asymptomatic periods between attacks. With inadequately treated hyperuricemia,
transition to the second phase can occur, manifested as chronic tophaceous gout,
which often involves polyarticular attacks, symptoms between attacks, and crystal
deposition (tophi) in soft tissues or joints. Although the prevalence of tophaceous An audio version Click here to
gout varies among populations, in one study, tophi were detected in three quarters of this article access audio
of patients who had had untreated gout for 20 years or more.3 Recurrent attacks are is available at version.
NEJM.org
common. In one study, approximately two thirds of patients with at least one gout
attack in the previous year had recurrent attacks.4
An estimated 6.1 million adults in the United States have had gout.5 The preva-
lence increases with age and is higher among men than among women, with a
ratio of 3 or 4 to 1 overall.5-7 However, this sex disparity decreases at older ages, at
least in part because of declining levels of estrogen, which has uricosuric effects in
women. The rising incidence and prevalence of gout are probably related to the aging
of the population, increasing levels of obesity, and dietary changes.6,7

Risk Factors
The use of thiazide diuretics, cyclosporine, and low-dose aspirin (<1 g per day) can
cause hyperuricemia, whereas high-dose aspirin (3 g per day) is uricosuric. Factors

n engl j med 364;5 nejm.org february 3, 2011 443

BACK to TOC
5
 The n e w e ng l a n d j o u r na l
that are associated with hyperuricemia and gout
include insulin resistance, the metabolic syndrome,
obesity, renal insufficiency, hypertension, conges-
tive heart failure, and organ transplantation.8,9 The
uricosuric effects of glycosuria in diabetes may re-
duce the risk of gout.10 Rare X-linked inborn errors
of metabolism can cause gout.8 Genomewide as-
sociation studies have identified common poly-
morphisms in several genes involved in renal urate
transport that are associated with gout, includ-
ing SLC2A9, ABCG2, SLC17A3, and SLC22A12.11,12
The risk of incident gout is increased in persons
with an increased intake of dietary purines (par-
ticularly meat and seafood), ethanol (particularly
beer and spirits), soft drinks, and fructose13-16
and is decreased in those with an increased intake
of coffee, dairy products, and vitamin C (which
lower urate levels).15,17,18
Triggers for recurrent flares include recent di-
uretic use, alcohol intake, hospitalization, and
surgery.19,20 Urate-lowering therapy, which reduces
the risk of gout attacks in the long term, can
trigger attacks in the early period after its initia-
tion, presumably as a result of mobilization of
bodily urate stores.21,22
S t r ategie s a nd E v idence
The diagnostic standard remains synovial fluid
or tophus aspiration with identification of nega-
tively birefringent monosodium urate crystals
under polarizing microscopy. Crystals are detect-
able during attacks and also potentially between
attacks, primarily in previously inflamed joints
in patients with hyperuricemia.23 However, crys-
tal evaluation is not performed routinely in clini-
cal practice.15 Hyperuricemia may not be present
during acute gout attacks and therefore may not
be a helpful criterion for diagnosis. A typical pre-
sentation that is strongly suggestive of the diag-
nosis includes rapid development of severe pain
(i.e., within 24 hours), erythema, and swelling in
a characteristic joint distribution for example,
in the first metatarsophalangeal joint (podagra).
In a population with a 0.5% prevalence of gout
overall, a patient with hyperuricemia and this pre-
sentation has an 82% chance of having gout.23
The differential diagnosis of acute gout in-
cludes other crystal-induced arthritides (e.g., cal-
cium pyrophosphate dihydrate) and a septic joint.
Joint aspiration with Grams staining and culture
must be performed if a septic joint is suspected,
444 n engl j med 364;5
BACK to TOC
of m e dic i n e
even if monosodium urate crystals are identified.
Older adults, particularly women, may present
with polyarticular involvement, which may be mis-
taken for rheumatoid arthritis; a tophus may be
mistaken for a rheumatoid nodule. Tophaceous
deposits that are not clinically apparent may be
visualized by plain radiography or another imag-
ing method. A diagnosis of gout should prompt
evaluation for potentially modifiable risk factors
(e.g., dietary habits) and associated coexisting ill-
nesses (e.g., hypertension and hyperlipidemia) that
may require intervention.
T r e atmen t Op t ions
Acute Gout
The main aim of therapy for acute gout is rapid
relief of pain and disability caused by intense in-
flammation. Options for managing acute attacks
include the use of nonsteroidal antiinflammatory
drugs (NSAIDs), colchicine, glucocorticoids, and
possibly corticotropin.24 The choice of agent, dose,
and duration of therapy is guided by consideration
of coexisting illnesses that preclude the safe use
of a particular regimen, as well as the severity of
the gout. Adjunctive measures include applying
ice to and resting the affected joint.25
NSAIDs and colchicine are first-line agents
for acute attacks (Table 1).24 Oral colchicine has
long been used, although it has only recently (in
2009) been approved by the Food and Drug Ad-
ministration (FDA) for use in patients with acute
gout. In a randomized trial, colchicine (at a dose
of 1.2 mg at the onset of a flare, followed by
0.6 mg 1 hour later) was significantly more likely
than placebo to result in a reduction in pain of
50% or more 24 hours later (rates, 37.8% and
15.5%, respectively).26 This regimen had efficacy
similar to that of a high-dose regimen (1.2 mg,
then 0.6 mg per hour for 6 hours), with fewer
gastrointestinal side effects. This study did not
address treatment after the first 24 hours.
The relative efficacy of colchicine as compared
with NSAIDs is unknown. In head-to-head studies,
various NSAIDs have had similar benefits for acute
gout, and a controlled trial showed the efficacy of
tenoxicam over placebo.24,27
When the use of NSAIDs or colchicine is poor-
ly tolerated or contraindicated, glucocorticoids or
corticotropin may be used, although evidence for
the use of intraarticular and intramuscular glu-
cocorticoids and corticotropin is limited by a lack
nejm.org february 3, 2011
6
BACK to TOC
Table 1. Pharmacologic Management Options for Acute Gout Attacks.

Examples of Regimens from Alternative Regimens for

Drug Randomized Clinical Trials Complete Attack Resolution* Precautions
Nonsteroidal antiinflammatory Avoid in patients with renal or hepatic insufficiency, bleeding dis-
drug order, congestive heart failure, or allergy; associated with an
increased risk of adverse thrombotic and gastrointestinal
events; may be administered with a proton-pump inhibitor
in patients at risk for gastrointestinal events.
Naproxen 500 mg orally twice daily for 5 days 375500 mg orally twice daily for 3 days,
then 250375 mg orally twice daily
for 47 days or until attack resolves
Indomethacin 50 mg orally three times daily for 2 days, 50 mg orally three times daily for 3 days,
then 25 mg orally three times daily then 25 mg orally three times daily
for 3 days for 47 days or until attack resolves
Colchicine 1.2 mg orally at first sign of gout flare, Consider additional acute gout regimen Avoid (or use lower dose) in older adults and those with renal
followed by 0.6 mg orally 1 hr later to continue managing attack insufficiency, hepatic dysfunction, or known gastrointestinal

n engl j med 364;5

1224 hr after colchicine regimen symptoms; adjust dose (and avoid in patients with renal
(e.g., 0.6 mg of colchicine twice daily, or hepatic impairment) if used in conjunction with
a nonsteroidal antiinflammatory P-glycoprotein or CYP3A4 inhibitors (e.g., cyclosporine, clar-
drug regimen, or an oral glucocorti- ithromycin, certain antiretroviral agents, certain antifungal

nejm.org
coid regimen until attack resolves) agents, certain calcium-channel blockers, and grapefruit
juice); avoid for gout-flare therapy in patients with renal or he-
patic impairment who are already receiving colchicine pro-
clinical pr actice

phylaxis; monitor for gastrointestinal symptoms, myotoxicity,

and blood dyscrasias (details are available at www.fda.gov).
Oral glucocorticoids Prednisolone, 3035 mg daily for 5 days Prednisone, 3060 mg daily for 2 days Use caution in patients with hyperglycemia or congestive heart
(prednisone or (depending on severity of attack), failure; may be used in patients with moderate-to-severe renal

february 3, 2011
prednisolone) then reduce by 510 mg every 2 days impairment.
(depending on starting dose) in
10-day taper

* Longer durations of therapy may be necessary for patients with long-standing disease and severe flares.
There are no published trials establishing the efficacy of celecoxib, the only selective cyclooxygenase-2 inhibitor available in the United States, for use in acute gout.
Although there are insufficient data to recommend the use of intraarticular glucocorticoid injection, it may be a useful alternative for attacks that are limited to one or two joints and
amenable to aspiration and in the absence of joint sepsis.

445

7
 The n e w e ng l a n d j o u r na l of m e dic i n e

of data from blinded, randomized, placebo- hyperuricemia, and antihyperuricemic therapies

controlled trials24,27-29 (Table 1). Monoarticular are not without risk. Recommendations that are
attacks are often managed with the use of intra- based on both consensus and evidence support
articular glucocorticoids. In two randomized, the consideration of urate-lowering therapy in pa-
placebo-controlled trials of a 5-day course of oral tients with hyperuricemia who have at least two
prednisolone (one evaluating a dose of 30 mg gout attacks per year or tophi (as determined by
daily and the other a dose of 35 mg daily), the either clinical or radiographic methods).38 How-
efficacy of prednisolone was equivalent to that ever, the severity and frequency of flares, the pres-
of standard regimens of indomethacin (vs. the ence of coexisting illnesses (including nephroli-
30-mg dose of prednisolone) and naproxen (vs. thiasis), and patient preference are additional
the 35-mg dose).30,31 considerations.24 Urate-lowering therapy should
The dose and duration of therapy for acute not be initiated during acute attacks but rather
gout should be sufficient to eradicate the profound started 2 to 4 weeks after flare resolution, with a
inflammatory response. Although randomized tri- low initial dose that is increased as needed over a
als have generally studied the effects of short period of weeks to months, and with close mon-
courses of treatment on pain reduction, clinical itoring of urate levels, renal function, and adverse
experience suggests that 7 to 10 days of treatment effects. The dose should be adjusted as necessary
may be necessary to ensure the resolution of to maintain a serum urate level below 6 mg per
symptoms. Increased doses of antiinflammatory deciliter (357 mol per liter), which is associated
drugs are typically prescribed for the first few with a reduced risk of recurrent attacks and to-
days, with a reduction in the dose once symptoms phi.22,39,40 It is uncertain whether a more stringent
begin to improve.32 Flares should be treated target of less than 5 mg per deciliter (297 mol per
without interruption of urate-lowering therapy. liter) results in greater disease control.41,42 Ther-
A medications in the pocket strategy should be apy is generally continued indefinitely.
considered for patients with established gout so Three classes of drugs are approved for low-
that therapy can be started promptly at the onset ering urate levels: xanthine oxidase inhibitors,
of symptoms that are consistent with typical at- uricosuric agents, and uricase agents (Table 2 and
tacks. Fig. 2). Xanthine oxidase inhibitors block the
There is evidence that attacks of gout are synthesis of uric acid and can be used regardless
caused by the activation of the NLRP3 inflamma- of whether there is overproduction of urate. In
some by urate crystals, leading to the release of this class of drugs, the one most commonly pre-
interleukin-133 (Fig. 1). For this reason, inter- scribed to lower urate levels is allopurinol, which
leukin-1 antagonists are being studied as poten- is effective in decreasing flares and tophi, particu-
tial options for patients in whom other treatments larly among patients in whom target urate levels
are not feasible.34 In a randomized trial, the fullyare achieved.22,39 Although allopurinol has an ac-
human monoclonal antibody canakinumab sig- ceptable side-effect profile in most patients, a mild
nificantly reduced pain from acute gout, as com- rash develops in approximately 2%.22,39,43 Severe
pared with 40 mg of intramuscular triamcino- allopurinol hypersensitivity is much less common
lone acetonide, 72 hours after administration of but can be life-threatening. Allopurinol desensi-
the study drug.35 Anakinra and rilonacept im- tization can be attempted in patients with mild
proved acute and chronic gout symptoms, respec- cutaneous reactions, but its safety in those with
tively, in two small, uncontrolled pilot studies; more serious reactions is unknown.44 The major-
however, rilonacept did not significantly reduce ity of patients receive 300 mg of allopurinol daily,
pain, as compared with indomethacin, in a ran- but this dose is often inadequate to achieve target
domized trial.34,36,37 More data are needed to as- urate levels. Daily doses up to 800 mg may be used
sess the potential role of these agents. in patients with normal renal function. The dose
is typically reduced in patients with renal impair-
Hyperuricemia ment, owing to concerns about an increased risk
Pharmacologic Approaches of hypersensitivity in such patients. However, stud-
The purpose of lowering serum urate levels is to ies have not shown an association between dose
prevent acute flares and development of tophi. and risk of hypersensitivity, and a reduced dose
However, gout does not develop in all patients with may contribute to suboptimal gout control.43

446 n engl j med 364;5 nejm.org february 3, 2011

BACK to TOC
8
 clinical pr actice
In 2009, another xanthine oxidase inhibitor,
febuxostat, was approved by the FDA for the
treatment of hyperuricemia in patients with
gout. As compared with a daily dose of 300 mg
of allopurinol, febuxostat at daily doses of 80
mg and 120 mg was 2.5 and 3 times as likely, re-
spectively, to achieve serum urate levels of less
than 6 mg per deciliter in a 52-week trial.22 Dur-
ing the initial 8 weeks of the study, the fre-
quency of gout attacks was higher among pa-
tients receiving 120 mg of febuxostat than
among those receiving either 80 mg of febuxo-
stat or 300 mg of allopurinol, but there was no
significant difference among the three groups
for the remainder of the trial. In another study
involving patients with renal impairment (de-
fined as a creatinine clearance of 30 to 89 ml
per minute), daily doses of 80 mg and 40 mg of
febuxostat were superior to 300 mg of allopuri-
nol (or 200 mg in patients with moderate renal
impairment) for lowering serum urate to a level
below 6 mg per deciliter.39 There was no in-
crease in cardiovascular risk or hypersensitivity
associated with the use of either dose of febuxo-
stat, as compared with allopurinol, although the
trial was not powered for such comparisons.
Postmarketing surveillance is needed to better
understand the risks and benefits of febuxostat. Its
efficacy as compared with increased doses of al-
lopurinol is not known, nor is its safety in per-
sons with allopurinol hypersensitivity.
Uricosuric drugs (including probenecid, sulfin-
pyrazone, and benzbromarone) block renal tubu-
lar urate reabsorption. Although these drugs can
be used in patients with underexcretion of urate
(accounting for up to 90% of patients with gout),
they are used less frequently than xanthine oxi-
dase inhibitors and are contraindicated in pa-
tients with a history of nephrolithiasis. Benzbro-
marone (not available in the United States) may
be used in patients with mild-to-moderate renal
insufficiency but is potentially hepatotoxic, where-
as the other two drugs are generally ineffective
in patients with renal impairment. In two open-
label, randomized trials, benzbromarone was
equivalent to allopurinol (the latter at a daily dose
of as much as 600 mg) and superior to probenecid
(among patients in whom target urate levels were
not achieved with 300 mg of allopurinol) in lower-
ing serum urate to 5 mg per deciliter or less.41,45
Uricase converts uric acid into soluble allan-
toin. Pegloticase, a polyethylene glycolated (peg-
n engl j med 364;5
BACK to TOC
Monosodium
urate crystals
Resident
macrophage
lineage
cells
+ +
NLRP3 C5b-9
inflammasome
Prointerleukin-1 Other mediators (e.g., TNF-,
interleukin-6 and 8,
leukotrienes, alarmins)
Interleukin-1
Interleukin-1
Endothelial receptor
or synovial cell Neutrophil recruitment,
activation, and release
of additional
Adhesion molecules, inflammatory
chemokines mediators
Figure 1. Mechanisms of Inflammation in Gout.
In acute gout, monosodium urate crystals that have undergone phagocyto-
sis activate the NLRP3 inflammasome, leading to secretion of interleukin-
1. In turn, this secretion can induce further production of interleukin-1
and other inflammatory mediators and further the activation of synovial lin-
ing cells and phagocytes. Monosodium urate crystals also induce many
other inflammatory cytokines (e.g., tumor necrosis factor [TNF-], inter-
leukin-6 and 8, leukotrienes, and alarmins) by mechanisms that are both
dependent on and independent of interleukin-1. Experimental models of
gout have demonstrated a role for the activation of the terminal comple-
ment pathway (C5b-9 membrane attack complex) induced by monosodium
urate crystals. Binding of interleukin-1 to the interleukin-1 receptor results
in signal transduction, leading to altered expression of adhesion molecules
and chemokines, which together with the other inflammatory events re-
sults in the neutrophil recruitment that is a major driver of the intense in-
flammation in gout. In chronic gout, with low-grade synovitis and frequent-
ly recurring or nonresolving flares, these inflammatory processes are
probably ongoing with potentially continued release of inflammatory medi-
ators, including interleukin-1, in the presence of persistent monosodium
urate crystals.
ylated) modified porcine recombinant uricase,
was approved by the FDA in 2010 for chronic gout
that is refractory to conventional treatments. The
approval was based on data from two double-
blind, randomized, placebo-controlled, 6-month
trials showing the drugs urate-lowering and to-
phus-reducing effects. However, pegloticase must
be administered intravenously, and infusion reac-
tions were common.46 Rasburicase, which is ap-
proved for use in preventing the tumor lysis syn-
nejm.org february 3, 2011 447
9
 448

BACK to TOC
Table 2. Pharmacologic Options for Hyperuricemia Therapy in Gout.*

Drug Example of Regimen Considerations or Precautions

Urate-lowering therapy Aim to maintain serum urate levels below 6 mg per deciliter, which requires regu-
lar monitoring and may require dose adjustments. Accompany the initiation
of therapy with flare prophylaxis.
Xanthine oxidase inhibitor Use in patients with urate overproduction or underexcretion. Avoid use (or moni-
tor closely) in patients receiving azathioprine or 6-mercaptopurine because
The

these drugs are metabolized by xanthine oxidase.

Allopurinol Starting dose: 50100 mg orally daily; increase dose every 24 wk
to achieve serum urate target, with dose based on creatinine
clearance; average daily dose, 300 mg, although many patients
require higher doses
Use with caution in patients with renal insufficiency (based on creatinine clear-
ance). The maximal dose may be as high as 800 mg daily, but there are limited
data for doses above 300 mg daily. A mild rash occurs in approximately 2% of
patients, and the risk is potentially increased by coadministration of ampicillin,
amoxicillin, thiazide diuretics, or ACE inhibitors. Allopurinol hypersensitivity is
rare, occurring in approximately 0.1% of patients, but can be fatal (rate of

n engl j med 364;5

death, 20%). If the target serum urate level is not achieved, consider dose es-
calation beyond the level suggested by guidelines in patients with renal impair-
ment (with close monitoring) or consider the use of an alternative therapy
(e.g., febuxostat). Allopurinol can increase the anticoagulant effect of warfarin.

nejm.org
Febuxostat Starting dose: 40 mg orally daily; increase to 80 mg orally daily Use as a second-line agent for patients who have contraindications or an inade-
after 24 wk to achieve serum urate target, if necessary quate response to allopurinol or uricosuric therapy. Although no dose adjust-
n e w e ng l a n d j o u r na l

ment is required for patients with mild-to-moderate renal or hepatic insuffi-

of

ciency, there are insufficient data for use in patients with a creatinine clearance
of <30 ml per minute or severe hepatic impairment. Currently contraindicated
for use with theophylline. Febuxostat has a higher cost than allopurinol.

february 3, 2011
Uricosuric agent (probenecid) Starting dose: 250 mg orally daily; increase by 500 mg per mo to a Avoid in patients with a history of nephrolithiasis and a creatinine clearance of
maximal dose of 23 g per day (2 divided doses) in patients <30 ml per minute. Adequate hydration is required to reduce risk of nephroli-
m e dic i n e

with normal renal function to achieve serum urate target
thiasis. The use of this drug can increase serum penicillin levels. Evaluate for
renal uric acid excretion in patients with a family history of early onset of gout,
onset of gout at <25 yr, or a history of nephrolithiasis, since this may identify
patients with an overproduction of urate in whom uricosuric therapy should be
avoided because of the risk of nephrolithiasis.

10
 clinical pr actice

drome, is not appropriate for use in patients with

antibodies against pegloticase. Anaphylaxis occurs in 5% of patients (vs. 0% in

See Table 1 for precautions, particularly taking into account potential for increased

See Table 1 for precautions, particularly taking into account potential for increased
and use caution in patients with congestive heart failure (insufficient safety data;
gout because of its immunogenicity and short

treatment for longer than 4 weeks. Do not use in patients with G6PD deficiency,
to above 6 mg per deciliter, particularly on two consecutive occasions) or with
patients without a therapeutic response (in whom serum urate levels increase
Use for chronic gout in adults whose disease is refractory to conventional therapy

symptoms with the use of a xanthine oxidase inhibitor at maximum medically

tions (26%, vs. 5% in placebo group) even with premedication, particularly in

some exacerbations in clinical trials). Cost is higher than for other therapies.
appropriate dose, or other contraindication). There is a risk of infusion reac-

Aim to reduce the risk of flare during initial decrease in urate levels, presumably

toxic effects with prolonged therapy. This drug has not been formally tested
(e.g., lack of normalization of serum urate, inadequate control of signs and

related to rapid mobilization of bodily urate stores. The duration of therapy

placebo group). No data are available regarding retreatment after stopping
half-life.

is not well defined but treatment for at least 6 mo or until tophi resolve is

but has been used for prophylaxis in trials of urate-lowering therapies.

Lifestyle, Nutrition, and Adjunctive Therapies
Observational data indicate that nonpharmaco-
logic approaches, such as avoiding alcohol or mod-
ifying ones diet, can reduce serum urate levels
but may not be sufficient to control established
gout.24 In one randomized trial involving persons
without gout, 500 mg of vitamin C per day for
2 months resulted in serum urate levels that were
0.5 mg per deciliter (30 mol per liter) lower than

toxic effects with prolonged therapy.

in those receiving placebo.47 The intake of dairy
milk reduced serum urate levels by approximately
10% during a 3-hour period in a small, random-
ized, crossover trial involving healthy volunteers.48
Whether these approaches would have similar ef-
fects in persons with gout, or with a longer dura-
recommended.

tion of therapy, is not known. Losartan and feno-

fibrate, which have uricosuric effects, may be
Benzbromarone and sulfinpyrazone are available in a limited number of countries but not in the United States.
considered in patients with gout who have hyper-
tension or hypertriglyceridemia, respectively,49 al-
though it is not known whether their use reduces
the frequency of gout attacks.
with antihistamines and glucocorticoids; start gout-flare pro-
Intravenous infusion of 8 mg every 2 wk; requires premedication

* ACE denotes angiotensin-converting enzyme, and NSAID nonsteroidal antiinflammatory drug.

Flare Prophylaxis during Initiation of Urate-Lowering

Therapy
Because rapid lowering of urate levels is associat-
ed with gout flares, with an increased risk asso-
phylaxis 7 days before initiating treatment

ciated with therapies that more effectively lower

0.6 mg orally once or twice daily as tolerated

urate levels,22,46 prophylaxis against acute flares

is advised during the initiation of urate-lowering
therapy (Table 2).24 In a study of patients with nor-
mal renal function who were starting allopurinol
Naproxen, 250 mg twice daily

therapy, oral colchicine (at a dose of 0.6 mg twice

daily for an average of 5.2 months) significantly
Febuxostat at a dose of 120 mg is available in Europe.

reduced the likelihood of gout attacks and less-

ened the severity of flares that did occur, as com-
pared with placebo.21 Diarrhea was common,
resulting in a once-daily regimen of colchicine for
many patients. Thus, the general recommenda-
tion for flare prophylaxis is to use colchicine at a
Flare prophylaxis during initiation
of urate-lowering therapy

dose of 0.6 mg once or twice daily, with dose ad-

justments as needed for renal impairment, poten-
tial drug interactions, or intolerance. Although
Uricase (pegloticase)

NSAIDs are also used for prophylaxis, there are

few studies that support their use.24 For patients
without tophi, prophylaxis should be continued
Colchicine

for 6 months. The optimal duration for those with

NSAID

tophi is uncertain; ongoing prophylaxis until to-

phus resolution may be necessary.

n engl j med 364;5 nejm.org february 3, 2011 449

BACK to TOC
11
 The n e w e ng l a n d j o u r na l of m e dic i n e

Risk factors for recurrent gout flares may differ

Exogenous
purines
Dietary
restriction
Endogenous
purines
Xanthine
oxidase
inhibitors
Body Recombinant
urate pool uricase
Renal
insufficiency
or failure
Renal Uricosuric
tubules agents
Figure 2. Management Strategies in Patients
with Hyperuricemia.
Hyperuricemia can be targeted at many levels. Restric-
tion of exogenous purine intake through dietary modi-
fications or the use of xanthine oxidase inhibitors to
block uric acid synthesis from endogenous purine me-
tabolism can reduce the amount of urate that contrib-
utes to the total-body urate pool. Modified uricase
agents reduce the total-body urate pool by converting
uric acid into soluble allantoin. In patients with normal
renal function, uricosuric agents can promote renal
elimination of urate, thereby reducing total-body urate
pools. However, decreased renal urate excretion in pa-
tients with renal impairment leads to increased total-
body urate stores.
A r e a s of Uncer ta in t y
Data are limited regarding the safety and effica-
cy of combination therapies for the treatment of
gout (e.g., the use of a xanthine oxidase inhibitor
and a uricosuric agent for hyperuricemia or the
use of multiple drugs for acute gout attacks). The
safety and cost-effectiveness of new agents for
gout, including inhibitors of urate transporter
1 and purine nucleoside phosphorylase, which are
under development, and interleukin-1 antagonists,
require further study. Preliminary data have sug-
gested the potential efficacy of the interleukin-1
antagonists canakinumab and rilonacept for flare
prophylaxis.34
from those that predispose patients to the initial
attack. Whether factors that lower serum urate
levels over the long term in persons without gout
would have similar effects with short-term or
episodic exposure in persons with gout requires
clarification.
It is not known to what level urate can be
safely lowered. Observational data have suggested
associations between low urate levels and an in-
creased risk of Parkinsons disease,50 but it is
unclear whether the low levels are a cause or
consequence of disease. The optimal duration of
urate-lowering therapy is also uncertain, and such
therapy is recommended indefinitely at this time.
In one study, the withdrawal of urate-lowering
therapy was associated with prolonged symptom-
free intervals (3 to 4 years) in a cohort of 89 pa-
tients after long-term control of urate levels (<7 mg
per deciliter), flares, and tophi resolution,51 but
further study is needed.
Finally, the concept of asymptomatic hyperuri-
cemia as a benign condition is being challenged.
Experimental data suggest that urate may con-
tribute to vascular remodeling and hypertension,
although it remains uncertain whether urate plays
a causal role in cardiovascular disease.9
Guidel ine s
The American College of Rheumatology is cur-
rently developing guidelines for the management
of gout. The European League against Rheumatism
and the British Society for Rheumatology have
published guidelines for the evaluation and man-
agement of gout on the basis of trial data (when
available) and expert consensus.23,24,42 The pres-
ent recommendations are largely consistent with
these guidelines.
C onclusions a nd
R ec om mendat ions
In patients presenting with suspected gout, the
diagnosis should be confirmed by examination of
synovial fluid or tophus aspirate for monosodium
urate crystals. Management should be tailored to
the stage of disease and coexisting illnesses. The
patient who is described in the vignette has crystal-
proven gout, with multiple attacks and a serum
urate level of more than 6 mg per deciliter despite
receipt of allopurinol at a dose of 300 mg per day.

450 n engl j med 364;5 nejm.org february 3, 2011

BACK to TOC
12
 clinical pr actice

Since his renal function is normal, the allopuri-
nol dose should be increased (e.g., 100-mg incre-
ments every 2 to 4 weeks until the target urate level
is reached), with monitoring of renal function and
serum urate levels and assessment for potential
adverse reactions. Colchicine prophylaxis (0.6 mg
once or twice daily) is reasonable while the dose
of allopurinol is escalated. If target serum urate
levels cannot be achieved or if the patient has seri-
ous side effects at higher allopurinol doses, the
use of either febuxostat or a uricosuric agent is
another option, given his normal renal function.
The patient should understand that the intake
of alcohol and an excessive amount of meat or
seafood and sugar-sweetened drinks may contrib-
ute to elevated urate levels and should be mini-
mized. He should be advised to keep well hydrated
and to lose weight. Associated cardiovascular risk
factors should be identified and treated. Although
the use of hydrochlorothiazide may contribute to
the increased urate level, I would not necessarily
change that medication if it is effectively con-
trolling his blood pressure, and I would advise
him to take the diuretic consistently, since inter-
mittent use may precipitate flares. The addition of
losartan for the hypertension might be considered.
He should be advised to maintain his urate-low-
ering regimen during flares, which can be man-
aged with colchicine. Follow-up is necessary to
ensure that appropriate serum urate levels are
achieved and maintained and to monitor the pa-
tient for adverse effects.
Dr. Neogi reports serving as a core expert panel leader for the
American College of Rheumatology Gout Treatment Guidelines.
No other potential conflict of interest relevant to this article was
reported.
Disclosure forms provided by the author are available with the
full text of this article at NEJM.org.
I thank Drs. Saralynn Allaire, Hyon Choi, and Yuqing Zhang
for their review of the first draft of the manuscript and Dr. Rob-
ert Terkeltaub for his review of an earlier version of Figure 1.

References

1. Loeb JN. The influence of temperature
on the solubility of monosodium urate.
Arthritis Rheum 1972;15:189-92.
2. Campion EW, Glynn RJ, DeLabry LO.
Asymptomatic hyperuricemia: risks and
consequences in the Normative Aging
Study. Am J Med 1987;82:421-6.
3. Gutman AB. The past four decades of
progress in the knowledge of gout, with
an assessment of the present status. Ar-
thritis Rheum 1973;16:431-45.
4. Neogi T, Hunter DJ, Chaisson CE, Al-
lensworth-Davies D, Zhang YQ. Frequency
and predictors of inappropriate manage-
ment of recurrent gout attacks in a longi-
tudinal study. J Rheumatol 2006;33:104-9.
5. Lawrence RC, Felson DT, Helmick
CG, et al. Estimates of the prevalence of
arthritis and other rheumatic conditions
in the United States: part II. Arthritis
Rheum 2008;58:26-35.
6. Arromdee E, Michet CJ, Crowson CS,
OFallon WM, Gabriel SE. Epidemiology
of gout: is the incidence rising? J Rheuma-
tol 2002;29:2403-6.
7. Wallace KL, Riedel AA, Joseph-Ridge
N, Wortmann R. Increasing prevalence of
gout and hyperuricemia over 10 years
among older adults in a managed care
population. J Rheumatol 2004;31:1582-7.
8. Choi HK, Mount DB, Reginato AM.
Pathogenesis of gout. Ann Intern Med
2005;143:499-516.
9. Feig DI, Kang DH, Johnson RJ. Uric
acid and cardiovascular risk. N Engl J Med
2008;359:1811-21. [Erratum, N Engl J Med
2010;362:2235.]
10. Rodrguez G, Soriano LC, Choi HK.
Impact of diabetes against the future risk
of developing gout. Ann Rheum Dis
2010;69:2090-4.
11. Dehghan A, Kottgen A, Yang Q, et al.
Association of three genetic loci with uric
acid concentration and risk of gout: a ge-
nome-wide association study. Lancet 2008;
372:1953-61.
12. Taniguchi A, Urano W, Yamanaka M,
et al. A common mutation in an organic
anion transporter gene, SLC22A12, is a
suppressing factor for the development of
gout. Arthritis Rheum 2005;52:2576-7.
13. Choi HK, Atkinson K, Karlson EW,
Willett W, Curhan G. Alcohol intake and
risk of incident gout in men: a prospective
study. Lancet 2004;363:1277-81.
14. Choi HK, Willett W, Curhan G. Fruc-
tose-rich beverages and risk of gout in
women. JAMA 2010;304:2270-8.
15. Choi HK, Atkinson K, Karlson EW,
Willett W, Curhan G. Purine-rich foods,
dairy and protein intake, and the risk of
gout in men. N Engl J Med 2004;350:1093-
103.
16. Choi HK, Curhan G. Soft drinks, fruc-
tose consumption, and the risk of gout in
men: prospective cohort study. BMJ 2008;
336:309-12.
17. Choi HK, Gao X, Curhan G. Vitamin C
intake and the risk of gout in men: a pro-
spective study. Arch Intern Med 2009;
169:502-7.
18. Choi HK, Willett W, Curhan G. Coffee
consumption and risk of incident gout in
men: a prospective study. Arthritis Rheum
2007;56:2049-55.
19. Hunter DJ, York M, Chaisson CE,
Woods R, Niu J, Zhang Y. Recent diuretic
use and the risk of recurrent gout attacks:
the online case-crossover gout study.
J Rheumatol 2006;33:1341-5. [Erratum,
J Rheumatol 2006;33:1714.]
20. Zhang Y, Woods R, Chaisson CE, et al.
Alcohol consumption as a trigger of re-
current gout attacks. Am J Med 2006;
119(9):800.e13-800.e18.
21. Borstad GC, Bryant LR, Abel MP,
Scroggie DA, Harris MD, Alloway JA. Col-
chicine for prophylaxis of acute flares
when initiating allopurinol for chronic
gouty arthritis. J Rheumatol 2004;31:
2429-32.
22. Becker MA, Schumacher HR Jr, Wort-
mann RL, et al. Febuxostat compared with
allopurinol in patients with hyperurice-
mia and gout. N Engl J Med 2005;353:
2450-61.
23. Zhang W, Doherty M, Pascual E, et al.
EULAR evidence based recommendations
for gout. Part I: Diagnosis. Report of a
task force of the EULAR Standing Com-
mittee for International Clinical Studies
Including Therapeutics (ESCISIT). Ann
Rheum Dis 2006;65:1301-11.
24. Zhang W, Doherty M, Bardin T, et al.
EULAR evidence based recommendations
for gout. Part II: Management. Report of a
task force of the EULAR Standing Com-
mittee for International Clinical Studies
Including Therapeutics (ESCISIT). Ann
Rheum Dis 2006;65:1312-24.
25. Schlesinger N, Detry MA, Holland
BK, et al. Local ice therapy during bouts
of acute gouty arthritis. J Rheumatol 2002;
29:331-4.
26. Terkeltaub RA, Furst DE, Bennett K,
Kook KA, Crockett RS, Davis MW. High
versus low dosing of oral colchicine for

n engl j med 364;5 nejm.org february 3, 2011 451

BACK to TOC
13
 clinical pr actice

early acute gout flare: twenty-four-hour out-
come of the first multicenter, randomized,
double-blind, placebo-controlled, parallel-
group, dose-comparison colchicine study.
Arthritis Rheum 2010;62:1060-8.
27. Sutaria S, Katbamna R, Underwood
M. Effectiveness of interventions for the
treatment of acute and prevention of re-
current gout a systematic review. Rheu-
matology (Oxford) 2006;45:1422-31.
28. Axelrod D, Preston S. Comparison of
parenteral adrenocorticotropic hormone
with oral indomethacin in the treatment
of acute gout. Arthritis Rheum 1988;31:
803-5.
29. Janssens HJ, Lucassen PL, Van de Laar
FA, Janssen M, Van de Lisdonk EH. System-
ic corticosteroids for acute gout. Cochrane
Database Syst Rev 2008;2:CD005521.
30. Janssens HJ, Janssen M, van de Lis-
donk EH, van Riel PL, van Weel C. Use of
oral prednisolone or naproxen for the
treatment of gout arthritis: a double-
blind, randomised equivalence trial. Lan-
cet 2008;371:1854-60.
31. Man CY, Cheung IT, Cameron PA,
Rainer TH. Comparison of oral predniso-
lone/paracetamol and oral indomethacin/
paracetamol combination therapy in the
treatment of acute goutlike arthritis: a
double-blind, randomized, controlled tri-
al. Ann Emerg Med 2007;49:670-7.
32. Mandell BF, Edwards NL, Sundy JS,
Simkin PA, Pile JC. Preventing and treat-
ing acute gout attacks across the clinical
spectrum: a roundtable discussion. Cleve
Clin J Med 2010;77:Suppl 2:S2-S25.
33. Martinon F, Ptrilli V, Mayor A, Tardi-
vel A, Tschopp J. Gout-associated uric acid
crystals activate the NALP3 inflamma-
some. Nature 2006;440:237-41.
34. Neogi T. Interleukin-1 antagonism in
acute gout: is targeting a single cytokine
the answer? Arthritis Rheum 2010;62:
2845-9.
35. So A, De Meulemeester M, Pikhlak A,
et al. Canakinumab for the treatment of
acute flares in difficult-to-treat gouty ar-
thritis: results of a multicenter, phase II,
dose-ranging study. Arthritis Rheum 2010;
62:3064-76.
36. So A, De Smedt T, Revaz S, Tschopp J.
A pilot study of IL-1 inhibition by anakin-
ra in acute gout. Arthritis Res Ther
2007;9:R28.
37. Terkeltaub R, Sundy JS, Schumacher
HR, et al. The interleukin 1 inhibitor
rilonacept in treatment of chronic gouty
arthritis: results of a placebo-controlled,
monosequence crossover, non-randomised,
single-blind pilot study. Ann Rheum Dis
2009;68:1613-7.
38. Mikuls TR, MacLean CH, Olivieri J, et
al. Quality of care indicators for gout man-
agement. Arthritis Rheum 2004;50:937-43.
39. Becker MA, Schumacher HR, Espinoza
LR, et al. The urate-lowering efficacy and
safety of febuxostat in the treatment of the
hyperuricemia of gout: the CONFIRMS
trial. Arthritis Res Ther 2010;12:R63.
40. Perez-Ruiz F, Liot F. Lowering serum
uric acid levels: what is the optimal target
for improving clinical outcomes in gout?
Arthritis Rheum 2007;57:1324-8.
41. Reinders MK, Haagsma C, Jansen TL,
et al. A randomised controlled trial on the
efficacy and tolerability with dose escala-
tion of allopurinol 300-600 mg/day versus
benzbromarone 100-200 mg/day in patients
with gout. Ann Rheum Dis 2009;68:892-7.
42. Jordan KM, Cameron JS, Snaith M, et
al. British Society for Rheumatology and
British Health Professionals in Rheuma-
tology guideline for the management of
gout. Rheumatology (Oxford) 2007;46:
1372-4.
43. Dalbeth N, Kumar S, Stamp L, Gow P.
Dose adjustment of allopurinol according
to creatinine clearance does not provide
adequate control of hyperuricemia in pa-
tients with gout. J Rheumatol 2006;33:
1646-50.
44. Fam AG, Dunne SM, Iazzetta J, Paton
TW. Efficacy and safety of desensitization
to allopurinol following cutaneous reac-
tions. Arthritis Rheum 2001;44:231-8.
45. Reinders MK, van Roon EN, Jansen
TL, et al. Efficacy and tolerability of urate-
lowering drugs in gout: a randomised
controlled trial of benzbromarone versus
probenecid after failure of allopurinol.
Ann Rheum Dis 2009;68:51-6.
46. Sundy JS, Becker MA, Baraf HS, et al.
Reduction of plasma urate levels following
treatment with multiple doses of pegloti-
case (polyethylene glycol-conjugated uri-
case) in patients with treatment-failure
gout: results of a phase II randomized
study. Arthritis Rheum 2008;58:2882-91.
47. Huang HY, Appel LJ, Choi MJ, et al.
The effects of vitamin C supplementation
on serum concentrations of uric acid: re-
sults of a randomized controlled trial.
Arthritis Rheum 2005;52:1843-7.
48. Dalbeth N, Wong S, Gamble GD, et al.
Acute effect of milk on serum urate con-
centrations: a randomised controlled cross-
over trial. Ann Rheum Dis 2010;69:1677-
82.
49. Takahashi S, Moriwaki Y, Yamamoto
T, Tsutsumi Z, Ka T, Fukuchi M. Effects of
combination treatment using anti-hyper-
uricaemic agents with fenofibrate and/or
losartan on uric acid metabolism. Ann
Rheum Dis 2003;62:572-5.
50. Kutzing MK, Firestein BL. Altered
uric acid levels and disease states. J Phar-
macol Exp Ther 2008;324:1-7.
51. Perez-Ruiz F, Atxotegi J, Hernando I,
Calabozo M, Nolla JM. Using serum urate
levels to determine the period free of
gouty symptoms after withdrawal of long-
term urate-lowering therapy: a prospective
study. Arthritis Rheum 2006;55:786-90.
Copyright 2011 Massachusetts Medical Society.

CLINICAL TRIAL REGISTRATION

The Journal requires investigators to register their clinical trials
in a public trials registry. The members of the International Committee
of Medical Journal Editors (ICMJE) will consider most reports of clinical
trials for publication only if the trials have been registered.
Current information on requirements and appropriate registries
is available at www.icmje.org/faq_clinical.html.

452 n engl j med 364;5 nejm.org february 3, 2011

BACK to TOC
14
 The n e w e ng l a n d j o u r na l of m e dic i n e

clinical practice

Calcium Kidney Stones

Elaine M. Worcester, M.D., and Fredric L. Coe, M.D.

This Journal feature begins with a case vignette highlighting a common clinical problem.
Evidence supporting various strategies is then presented, followed by a review of formal guidelines,
when they exist. The article ends with the authors clinical recommendations.

A 43-year-old man presents for evaluation of recurrent kidney stones. He passed his
first stone 9 years earlier and has had two additional symptomatic stones. Analysis of
the first and the last stones showed that they contained 80% calcium oxalate and 20%
calcium phosphate. Analysis of a 24-hour urine collection while the patient was not
receiving medications revealed a calcium level of 408 mg (10.2 mmol), an oxalate level
of 33 mg (367 mol), and a volume of 1.54 liters; the urine pH was 5.6. The patient had
been treated with 20 to 40 mmol of potassium citrate daily since he passed his first
stone. How should he be further evaluated and treated?

The Cl inic a l Probl em

From the Nephrology Section, Depart-
ment of Medicine, University of Chicago,
Chicago. Address reprint requests to Dr.
Worcester at the Nephrology Section,
MC 5100, University of Chicago, 5841 S.
Maryland Ave., Chicago, IL 60637, or at
eworcest@bsd.uchicago.edu.
N Engl J Med 2010;363:954-63.
Copyright 2010 Massachusetts Medical Society.
Click here to An audio version
access audio of this article
version. is available at
NEJM.org
In the United States, the prevalence of kidney stones has risen over the past 30
years.1 By 70 years of age, 11.0% of men and 5.6% of women will have a symptom-
atic kidney stone. The risk among white persons is approximately three times that
among black persons. About 80% of stones are composed of calcium oxalate with
variable amounts of calcium phosphate. Diagnosis of a calcium stone requires
analysis after passage or removal of the stone. After passage of a first stone, the risk
of recurrence is 40% at 5 years and 75% at 20 years. Among patients with recurrent
calcium stones who have served as control subjects in randomized, controlled trials
of interventions, new stones formed in 43 to 80% of subjects within 3 years.2-9
Hospitalizations, surgery, and lost work time that are associated with kidney stones
cost more than $5 billion annually in the United States.10 Stone formation is associ-
ated with increased rates of chronic kidney disease and hypertension,11,12 increases
that are not completely explained by obesity, which is a risk factor for each of these
conditions.13
Although many inherited and systemic diseases are associated with calcium
kidney stones,14 most such stones are idiopathic. The majority of patients with
idiopathic stones have at least one metabolic abnormality, as identified by 24-hour
urine testing. Prevention requires evaluation to identify systemic disease and modi-
fiable factors.

Patho gene sis

Physicochemical Factors
Supersaturation, often expressed as the ratio of urinary calcium oxalate or calcium
phosphate concentration to its solubility, is the driving force in stone formation. At
levels of supersaturation below 1, crystals dissolve, whereas at supersaturation levels
above 1, crystals can nucleate and grow, promoting stone formation. Supersatura-
tion is generally higher in patients with recurrent kidney stones than in those
without the condition, and the type of stone that is formed correlates with urinary

954 n engl j med 363;10 nejm.org september 2, 2010

BACK to TOC
15
 clinical pr actice
supersaturation. Calcium oxalate supersaturation
is independent of urine pH, but calcium phosphate
supersaturation increases rapidly as urine pH rises
from 6 to 7. Since calcium oxalate stones form
over an initial calcium phosphate layer,15 treat-
ment optimally should lower the supersaturation
of both types. Most 24-hour analyses of kidney-
stone risk that are performed at specialized labo-
ratories include calculated supersaturation values.
Urine also contains substances that can ac-
celerate or retard urinary crystallization.16 The
only such substance that can be modified in
practice at this time is citrate, which can slow
the growth of calcium crystals.17
Anatomic abnormalities, in particular those
that result in urinary stasis (such as ureteropelvic
junction obstruction, horseshoe kidney, or poly-
cystic kidney), may precipitate or worsen stone
formation.18 Patients with a single functioning
kidney are at particular risk, since stone passage
with ureteral obstruction can result in acute kid-
ney failure.
Metabolic Factors
Imbalances between excretions of calcium, oxa-
late, and water create supersaturation. Hypercal-
ciuria, the most common metabolic abnormality
found in patients with recurrent calcium stones,
is most often familial and idiopathic19 and is
strongly influenced by diet. Gut calcium absorp-
tion is increased in persons with idiopathic hyper-
calciuria, but serum calcium values remain un-
changed, since absorbed calcium is promptly
excreted.20 On a low-calcium diet, such persons
often excrete more calcium than they eat,21 and
urinary calcium excretion also rises markedly af-
ter the intake of calcium-free nutrients such as
simple oral glucose; in such cases, the only source
possible is bone. Although hypercalciuria is some-
times divided into subtypes (absorptive, resorptive,
and renal leak), this classification is not helpful
in guiding treatment. However, measurement of
serum calcium is indicated to identify patients
with primary hyperparathyroidism.
The level of oxalate excretion is modestly
higher among patients with recurrent calcium
stones than among those without the condition,
possibly because of increased oxalate absorption
in the gut.22 The intake of ascorbic acid and a
high level of protein may increase oxalate pro-
duction.23 Because calcium binds with oxalate in
the gut and hinders its absorption, oxalate is
n engl j med 363;10
BACK to TOC
more readily absorbed when dietary calcium is
low.23 This may be why a low-calcium diet does
not successfully prevent stone recurrence.24
Citrate chelates calcium in the urine, decreas-
ing supersaturation and reducing the growth of
crystals17; hypocitraturia is a risk factor for stone
formation. Distal renal tubular acidosis, hypo-
kalemia, and the use of carbonic anhydrase in-
hibitors (e.g., topiramate) lead to hypocitraturia,
but the cause of this condition in most patients
with recurrent kidney stones is unknown.25
Hyperuricosuria, often from high dietary intake
of purines, is thought to promote the formation
of calcium stones by reducing the solubility of
calcium oxalate.26
Histopathology
Intraoperative papillary biopsy specimens ob-
tained from patients with recurrent kidney stones
show that the pattern of crystal deposition differs
according to the type of stone. Idiopathic calci-
um oxalate stones form over regions of intersti-
tial calcium phosphate deposits (Randalls plaque)
on the papillary surface,27 whereas idiopathic
calcium phosphate stones are associated with
crystal deposits in inner medullary collecting
ducts that contain mainly apatite,28,29 sometimes
mixed with other crystals. (For additional details,
see the Supplementary Appendix, available with
the full text of this article at NEJM.org.)
S t r ategie s a nd E v idence
Patients with recurrent calcium stones should be
evaluated to rule out systemic disease and guide
preventive therapy. Evaluation includes history
taking directed at detecting potential causes of
stones (Table 1). All stones should be analyzed to
classify the type and to detect conversion from
one stone type to another for example, from
calcium oxalate to struvite in the presence of in-
fection or to calcium phosphate if the urinary pH
rises in response to treatment.30
Computed tomography (CT) without the use
of contrast material provides information regard-
ing the presence, size, and location of stones, as
well as ruling out anatomic abnormalities and
providing a baseline for assessing whether sub-
sequent stones that are passed are old or new
(with the latter indicating a need for improved
preventive treatment). Given the expense and
radiation exposure of CT, renal ultrasonography
nejm.org september 2, 2010 955
16
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 1. Key Coexisting Medical Conditions, Medication Use, Diet, and Other Factors Associated with Calcium Kidney Stones.

Variable Features Type of Kidney Stone

Calcium Calcium
Oxalate Phosphate
Medical or surgical history
Bowel disease Chronic diarrhea, malabsorption Yes
Intestinal surgery Small-bowel resection, ileostomy Yes
Bariatric surgery Duodenal switch, Roux-en-Y gastric bypass Yes
Sarcoidosis Yes Yes
Gout Yes
Renal tubular acidosis Yes
Bone disease or fracture Primary hyperparathyroidism, idiopathic Yes Yes
hypercalciuria, myeloma
Immobilization Trauma, prolonged illness Yes Yes
Hyperthyroidism Untreated, iatrogenic Yes Yes
Renal anomaly Urinary stasis Yes Yes
Medications
Topiramate Seizures, migraine Yes
Calcium supplements Antacids, dietary supplement Yes Yes
Carbonic anhydrase inhibitor Glaucoma Yes
Alkali Bicarbonate, citrate Yes
Vitamin D Yes Yes
Occupational or recreational factor
Dehydration Hot environment, inability to drink Yes Yes
Dietary factor
Oxalate loads Nuts, spinach, ascorbic acid Yes
Excess salt Prepared foods, snack foods Yes Yes
Eating disorders Vomiting, use of laxatives Yes Yes
Strange diets* Protein powder, sugar loads Yes Yes
Family history
History of kidney stones in a first- Idiopathic hypercalciuria, primary hyperoxaluria Yes Yes
degree relative

* Strange diets include very restrictive choices of food or the use of a large number or amount of supplements.

or abdominal plain radiography may be used in
follow-up imaging of known stones, although
these methods are less sensitive than CT.
Metabolic testing should be done after the
resolution of the acute episode of stone passage,
when patients have resumed their usual diet and
activity. Evaluation includes a blood test to screen
for hypercalcemia, chronic kidney disease, and
renal tubular acidosis. Analysis of a 24-hour
urine collection to detect metabolic abnormali-
ties should preferably be performed twice, since
mineral excretions may vary from day to day.31
Tables 2 and 3 provide a suggested framework
for testing and interpretation. Whether to evalu-
ate patients after a single kidney-stone episode
is controversial, although it seems prudent to rule
out systemic disease, especially in patients with
a first stone before adulthood.
T r e atmen t
Management of Symptomatic Stones
Stones that have formed in kidneys do not re-
quire removal or fragmentation unless they cause
obstruction, infection, serious bleeding, or persis-
tent pain. Ureteral stones of less than 10 mm in

956 n engl j med 363;10 nejm.org september 2, 2010

BACK to TOC
17
 clinical pr actice

Table 2. Diagnostic Testing for Patients with Recurrent Kidney Stones.*

Measurement Normal Value or Range for Adults Purpose

Blood testing
Calcium 8.810.3 mg/dl Detection of primary hyperparathyroidism, excessive vitamin D
intake, sarcoidosis
Phosphate 2.55.0 mg/dl Detection of primary hyperparathyroidism
Creatinine 0.61.2 mg/dl Detection of chronic kidney disease
Bicarbonate 2028 mmol/liter Detection of renal tubular acidosis
Chloride 95105 mmol/liter Detection of renal tubular acidosis
Potassium 3.54.8 mmol/liter Detection of renal tubular acidosis, eating disorders, gastro-
intestinal disease
Urine collection over 24-hour period
Volume >1.5 liter/day Detection of low volume as cause of stones
Calcium <300 mg/day for men, <250 mg/day for women; Detection of hypercalciuria
<140 mg/g creatinine/day
Oxalate <40 mg/day Detection of hyperoxaluria
pH 5.86.2 Calculation of calcium phosphate and uric acid supersatura-
tion, diagnosis of renal tubular acidosis
Phosphate 5001500 mg/day Calculation of calcium phosphate supersaturation
Citrate >450 mg/day for men, >550 mg/day for women Detection of low citrate level and diagnosis of renal tubular aci-
dosis; calculation of calcium phosphate supersaturation
Uric acid <800 mg/day for men, <750 mg/day for women Detection of hyperuricosuria as cause of stones; calculation of
uric acid supersaturation
Sodium 50150 mmol/day Diet counseling; calculation of supersaturation
Potassium 20100 mmol/day Use of potassium salts; calculation of supersaturation
Magnesium 50150 mg/day Detection of malabsorption; calculation of
supersaturation
Sulfate 2080 mmol/day Calculation of supersaturation; measure of net acid production
Ammonium 1560 mmol/day Calculation of supersaturation
Creatinine 2024 mg/kg/day for men, 1519 mg/kg/day Comparison of actual with predicted creatinine to assess the
for women completeness of the urine collection
Protein catabolic rate 0.81.0 g/kg/day Estimation of protein intake
Calculated supersaturation
Calcium oxalate 610 Guidance of treatment
Calcium phosphate 0.52 Guidance of treatment
Other screening tests
Urinary cystine screening Negative Detection of cystinuria
Stone analysis Basic classification of condition

* Blood testing for renal tubular acidosis, chronic kidney disease, and hypercalcemia, along with urinary cystine screening and kidney-stone
analysis, are appropriate for all patients with recurrent kidney stones. Collection of urine over a 24-hour period is appropriate if medical
prevention of kidney-stone formation is planned. To convert the values for calcium to millimoles per day, multiply by 0.025. To convert
the values for phosphate to millimoles per day, multiply by 0.0323. To convert the values for creatinine to micromoles per day multiply by
0.00884. To convert the values for urinary oxalate to micromoles per day, multiply by 11.11. To convert the values for urinary citrate to mmol
per day, multiply by 0.0052. To convert the values for urinary uric acid to millimoles per day, multiply by 0.00595. To convert the values for
urinary magnesium to mmol per day, multiply by 0.0411. To convert the values for urinary urea nitrogen to moles per day, multiply by 0.0357.
The protein catabolic rate is calculated by multiplying the urea nitrogen excretion in grams per day by 6.25 and dividing by body weight in
kilograms.
Supersaturation is expressed as the ratio of urinary calcium oxalate or calcium phosphate concentration to its solubility.
Urinary cystine was tested with the use of the cyanide nitroprusside test. A negative test means that the cystine concentration is less than
75 mg per liter.

n engl j med 363;10 nejm.org september 2, 2010 957

BACK to TOC
18
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 3. Primary Causes of Calcium Kidney Stones and Their Treatments.*

Cause Key Abnormality

Serum Serum Parathyroid Urine Urine Urine Urine
Calcium Hormone Calcium pH Citrate Oxalate
Idiopathic calcium oxalate Normal Normal Normal or Normal Normal or Normal or
increased decreased increased
Idiopathic calcium phosphate Normal Normal Normal or Increased Normal or Normal
increased decreased

Primary hyperparathyroidism Increased Increased Increased Increased Normal Normal

Sarcoidosis Increased Decreased Increased Normal Normal Normal
Lithium use Increased Increased Increased Increased Normal Normal
Oral supplementation with Normal or Normal Increased Normal Normal Normal
calcium or vitamin D increased
Ileostomy Normal or Normal Decreased Decreased Decreased Normal
decreased
Short-bowel syndrome Normal or Normal Decreased Decreased Decreased Increased
decreased

Bariatric surgery Normal or Normal Decreased Normal Decreased Increased

decreased
Renal tubular acidosis Normal Normal Normal or Increased Decreased Normal
increased

* NA denotes not available because histologic analyses have not been reported for patients with the listed conditions.

diameter may be followed with conservative treat-
ment in the absence of fever, infection, or renal
failure, if pain is controlled. Opioid analgesics
and nonsteroidal antiinflammatory agents are
both effective for pain control in acute colic.
Therapy with drugs that block 1-adrenergic re-
ceptors or calcium-channel blockers may facili-
tate passage of ureteral stones.32 In general, stones
larger than 10 mm in diameter will not pass, and
those smaller than 5 mm will; stones from 5 mm
to 10 mm have variable outcomes. Stones in the
distal ureter are more likely to pass than those
located more proximally.
If stones do not pass, there are several surgi-
cal options for removal.33 Data to guide surgical
recommendations are derived largely from meta-
analyses of small trials. For ureteral stones, the
treatment of choice is either shock-wave litho-
tripsy or ureteroscopy with laser lithotripsy. Stone-
free rates are better with ureteroscopy, but com-
plication rates are higher, including sepsis and
ureteral injury. For stones lodged in the kidney,
the size, location, and presumed composition
play a role in determining treatment. Not all
stone types fragment equally well; for example,
calcium oxalate monohydrate and brushite stones
are more resistant to fragmentation than calci-
um oxalate dihydrate or apatite stones. Shock-
wave lithotripsy and ureteroscopy are frequently
used for smaller stones. Percutaneous nephro-
lithotomy may be used for single large stones
(above 2 cm) or a large or obstructing stone bur-
den. This procedure requires general anesthesia
and hospitalization and carries more risk of com-
plications, including bleeding and infection, than
other techniques but can result in a stone-free
kidney.34 Open or laparoscopic procedures are
occasionally used for stone removal in challeng-
ing cases.
Prevention of Idiopathic Calcium Oxalate
Stones
Prevention of recurrent stones requires decreas-
ing urinary supersaturation, which is generally
achieved by raising urine volume and lowering
calcium and oxalate excretion. It should be rec-
ognized that urinary abnormalities are graded
risk factors, and thresholds for the definition of
normal urinary function are not absolute cut-
offs.35 Table 4 summarizes treatment strategies.

958 n engl j med 363;10 nejm.org september 2, 2010

BACK to TOC
19
 clinical pr actice

Supersaturation Tissue Changes Treatment

Urine Urine Calcium Calcium Uric Interstitial Collecting-Duct
Uric Acid Volume Oxalate Phosphate Acid Plaque Plugging
Normal or Normal or High Normal Normal Increased Not present Thiazide for idiopathic hypercalciuria;
increased decreased or high or high potassium citrate for calcium oxalate
Normal Normal High High Normal Normal Increased (and perhaps calcium phosphate)
stones; allopurinol for hyperuricosu-
ria; sodium restriction and possible
protein or oxalate restriction; in-
creased fluid intake

Normal Normal High High Low Increased Increased Parathyroid surgery

Normal Normal High High Normal NA NA Glucocorticoids, possible ketoconazole
Normal Normal High High Normal NA NA Discontinuation of lithium
Normal Normal High High Low NA NA Discontinuation of supplements

Normal Decreased High Low High Increased Increased

Normal Decreased High Low High Normal Increased Fluids, alkali; supplements to reduce
urine oxalate excretion for the short-
bowel syndrome and bariatric surgery

Normal Normal High Normal Normal Normal Increased

Normal Increased Normal High Low Normal Increased Alkali, possible thiazides

(For additional details regarding treatment trials,
see Table 1 in the Supplementary Appendix.)
A randomized trial of increased fluid intake
that was targeted to maintain a daily urine vol-
ume of more than 2 liters showed a significant
reduction in recurrent stone passage among pa-
tients with first-time kidney stones.36 A target
urine volume of 2 to 2.5 liters is reasonable and
can be achieved by an increased intake of fluids,
especially water, although most low-sodium, low-
carbohydrate fluids are acceptable in moderation
for this purpose.
In a randomized, controlled trial involving
Italian men with hypercalciuria,24 a diet that was
low in animal protein (52 g per day), sodium
(50 mmol per day), and oxalate (200 mg per day)
with normal calcium intake (1200 mg per day)
was associated with a reduction in stone forma-
tion of almost 50% over a period of 5 years, as
compared with a diet that was low in calcium
(400 mg per day) and oxalate. In contrast, in a
U.S. trial, a low-protein diet did not reduce stone
recurrence during a 4.5-year period, but compli-
ance with the diet was poor and dietary sodium
was not restricted.37 A low-sodium diet can sig-
nificantly decrease excretion of both calcium and
oxalate,38 but data on the effect of a sodium-
restricted diet alone on stone recurrence are
lacking. Calcium restriction should be avoided in
patients with hypercalciuria, since it may result
in a reduction in bone mineral density39 and an
increased rate of fracture.40
Thiazide-type diuretics decrease urine calcium
excretion, and in randomized, controlled trials,
these medications significantly reduced recur-
rence rates of calcium stones by more than 50%
during a 3-year period, as compared with place-
bo.2,4,6,9 Long-acting agents like chlorthalidone
and indapamide are effective with once-daily
doses, whereas twice-daily doses are recom-
mended for hydrochlorothiazide.
Hyperoxaluria may occur when dietary calci-
um is low or oxalate intake is unusually high or
(less commonly) when oxalate is overproduced.
Dietary oxalate restriction to less than 100 mg
per day and the avoidance of an intake of ascor-
bic acid above 100 mg per day are prudent if hy-
peroxaluria is present. Foods that are very high
in oxalate include spinach, rhubarb, wheat bran,
chocolate, beets, miso, tahini, and most nuts.
(A list of the oxalate content of various foods is
available at www.ohf.org under Resources.) Marked

n engl j med 363;10 nejm.org september 2, 2010 959

BACK to TOC
20
 The n e w e ng l a n d j o u r na l of m e dic i n e

hyperoxaluria should prompt consideration of

obtain calcium from dietary sources

Need to monitor urine pH and calcium

malabsorption or one of the primary hyperox-

(may be desirable); allergy and sun

Need to avoid fluids containing excess

Hypokalemia, reduced blood pressure

phosphate supersaturation; avoid

Difficulty in maintaining diet; should
aluria syndromes.41
Potential Complications

Two randomized trials have shown substan-

tial reductions in stone recurrence among patients

and avoid supplements

supersaturation of >1
salt or carbohydrates

with hypocitraturia who were treated with po-

Allergy (may be severe)

tassium alkali three times daily.3,7 One trial of
sodiumpotassium citrate had negative results.8

sensitivity
Potassium alkali may be safely combined with
thiazide9,42 when indicated, but no trials have com-
pared the combination against either agent alone
for the prevention of stone recurrence.
Hyperuricosuria can decrease the solubility of
treatment for patients with a single

calcium oxalate and increase the incidence of

Useful for all patients; possible sole

patients with hyperoxaluria; and

be useful for some with normo-

Recommendations for sodium and

patients with hypercalciuria or

for calcium in all patients with

calcium oxalate stones. Allopurinol (at a dose of

Patients with hyperuricosuria and

hyperuricosuria; for oxalate in

Patients with hypercalciuria; may

protein especially useful in

300 mg daily) decreased stone recurrence in a

Selection Criteria

Patients with hypocitraturia

randomized trial involving patients with idio-

pathic calcium oxalate stones who had hyperuri-
cosuria.5 A reduction in the intake of protein
calcium stones
calcium stones
stone episode

(and therefore purine) is also prudent but has

calciuria

not been explicitly tested among patients with

hyperuricosuria and recurrent kidney stones.
In long-term clinical follow-up, preventive
treatment resulted in persistent reductions in
Table 4. Treatment Recommendations for the Prevention of Idiopathic Calcium Kidney Stones in Adults.

stone recurrence during a period of 20 years or

<0.81 g of animal protein/kg/day;

100300 mg/day (may be taken once

hydrochlorothiazide, 12.525 mg

Potassium citrate, 1020 mmol two

Adequate to maintain urine volume

more.43 However, compliance tended to wane

oxalate, <100 mg/day; calcium,

indapamide, 1.252.5 mg/day;

Sodium, <100 mmol/day; protein,

Chlorthalidone, 12.550 mg/day;

over time, with rates of nonadherence approach-

ing 20% per year.44
or three times daily
Dose

8001000 mg/day

Prevention of Calcium Phosphate Stones

Most calcium stones consist of more than 90%
>2 liters daily

twice daily

calcium oxalate with trace amounts of calcium

phosphate, but the proportion of calcium phos-
daily)

phate in stones has increased over time.45,46 Idio-

pathic calcium phosphate stones (more than 50%
calcium phosphate) are more common among
tion, which may improve solubility
calcium; inhibits growth of calcium
maintains bone mineral, prevents
Lowers supersaturation by decreasing

Lowers supersaturation by decreasing

Lowers supersaturation by chelating

women and are associated with alkaline urine

Lowers urinary uric acid concentra-
Lowers supersaturation by dilution

calcium and oxalate excretion;

pH, a condition whose cause is not well under-

Mechanism of Action

stood. Mild abnormalities in urine acidification

may be present, although metabolic acidosis is
uncommon.46,47 Some patients convert from the
calcium excretion

of calcium salts

formation of calcium oxalate stones to the forma-

hyperoxaluria

tion of calcium phosphate stones. In one study,

of solutes

crystals

such patients had a urine pH that was more alka-

line (>6.2) at baseline than those who continued
to produce calcium oxalate stones.30 Calcium
phosphate stones are associated with poorer
stone-free rates after percutaneous nephrolitho-
Potassium alkali

tomy and with more shock-wave lithotripsy treat-

Thiazide-type

ments than are calcium oxalate stones.46,48

diuretic

Allopurinol
Treatment

Among patients with calcium phosphate stones,

Fluids

treatment is similar to that of patients with cal-

Diet

cium oxalate stones except that potassium al-

960 n engl j med 363;10 nejm.org september 2, 2010

BACK to TOC
21
 clinical pr actice
kali should be used cautiously because it raises
urine pH, potentially worsening calcium phos-
phate supersaturation. Levels of urine pH and
citrate and the degree of supersaturation should
be assessed after starting therapy. If the citrate
level does not rise and the degree of supersatura-
tion worsens, the medication is unlikely to be of
benefit.
A r e a s of Uncer ta in t y
Treatment trials for calcium stones have not
looked specifically at outcomes in patients with
calcium phosphate stones. Dietary recommenda-
tions to increase fluids, lower salt and protein
intake, and maintain a normal intake of calcium
are supported by an Italian randomized trial,24
but no women were included in this study, and it
is unclear whether many Americans can comply
with the necessary dietary pattern sufficiently
to successfully prevent stones. The Dietary Ap-
proaches to Stop Hypertension (DASH)Sodium
diet, when modified to remove high-oxalate foods,
replicates many of the features of the study diet
and may provide a model to follow, but its effects
on stone recurrence have not been explicitly stud-
ied.49 Stone formation is associated with an in-
creased risk of bone disease, chronic kidney
disease, and hypertension, but it is not known
whether effective stone prevention decreases these
risks.
Guidel ine s
Guidelines of the American Urological Associa-
tion (www.auanet.org) recommend that patients
who require surgery for ureteral stones should be
informed about benefits and risks of all current
treatment approaches. Shock-wave lithotripsy and
ureteroscopy with laser lithotripsy are both con-
sidered acceptable first-line treatments, although
ureteroscopy achieves greater stone-free rates.
Percutaneous access and open or laparoscopic
surgery are used as needed for selected cases.
The guidelines do not address evaluation or treat-
ment to prevent recurrent stones.
C onclusions a nd
R ec om mendat ions
Preventive treatment to decrease stone recur-
rence is indicated for patients with recurrent cal-
cium stones, such as the patient in the vignette.
n engl j med 363;10
BACK to TOC
If systemic disease is not present, treatment
should focus on metabolic abnormalities uncov-
ered during the workup, such as hypercalciuria,
hypocitraturia, hyperuricosuria, or hyperoxalu-
ria. Although data comparing specific supersatu-
ration targets are lacking, a logical strategy is to
lower calcium oxalate and calcium phosphate
supersaturation to the low end of the normal
range.
Patients should be advised to increase fluid
intake to at least 2 liters daily and reduce sodi-
um intake to 2300 mg and protein intake to 0.8
to 1 g per kilogram of body weight per day, since
these dietary interventions have reduced stone
recurrence in randomized trials. Calcium intake
should not be reduced below the recommended
intake for sex and age and should be supplied by
food rather than by supplements, which may in-
crease the risk of stone formation. In many pa-
tients, medication is also needed; the choice of
medication is influenced by the metabolic ab-
normalities identified, the type of stone, and the
preference of patients.
The stones of the patient in the vignette con-
tain 20% phosphate, despite a low urine pH
while the patient was not receiving medications;
the increased phosphate level may reflect his
previous treatment with citrate. Both hypocitra-
turia and hypercalciuria may contribute to his
stone formation. In addition to the recommen-
dations above, we would initiate therapy with a
thiazide-type diuretic (e.g., 25 mg of chlorthali-
done daily) to lower the urinary calcium level.
A reduction in sodium intake will also reduce a
thiazide-induced loss of potassium.
A follow-up 24-hour urine collection and se-
rum chemical analysis should be performed in
4 to 6 weeks to assess the efficacy of treatment
and possible side effects, particularly hypokale-
mia, which can worsen hypocitraturia. If potas-
sium supplementation is needed, it may be added
as potassium alkali, but the urine pH level and
the level of calcium phosphate supersaturation
should be monitored. If the level of calcium
phosphate supersaturation rises and is consistent-
ly above 1, potassium chloride should be substi-
tuted. Primary treatment with potassium alkali
would be an alternative to a thiazide but may not
lower the level of urinary calcium phosphate
supersaturation as effectively. Ongoing attention
is warranted at follow-up visits to monitor wheth-
er the patient is adhering to preventive recom-
mendations.
nejm.org september 2, 2010 961
22
 The n e w e ng l a n d j o u r na l
Drs. Worcester and Coe report receiving consulting fees from
LabCorp. No other potential conflict of interest relevant to this
article was reported.
Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.
References
1. Stamatelou KK, Francis ME, Jones
CA, Nyberg LM, Curhan GC. Time trends
in reported prevalence of kidney stones in
the United States: 1976-1994. Kidney Int
2003;63:1817-23.
2. Borghi L, Meschi T, Guerra A, No-
varini A. Randomized prospective study
of a nonthiazide diuretic, indapamide, in
preventing calcium stone recurrences.
J Cardiovasc Pharmacol 1993;22:Suppl 6:
S78-S86.
3. Ettinger B, Pak CY, Citron JT, Thomas
C, Adams-Huet B, Vangessel A. Potassium-
magnesium citrate is an effective prophy-
laxis against recurrent calcium oxalate
nephrolithiasis. J Urol 1997;158:2069-73.
4. Ettinger B, Citron JT, Livermore B,
Dolman LI. Chlorthalidone reduces calci-
um oxalate calculous recurrence but mag-
nesium hydroxide does not. J Urol 1988;
139:679-84.
5. Ettinger B, Tang A, Citron JT, Liver-
more B, Williams T. Randomized trial of
allopurinol in the prevention of calcium
oxalate calculi. N Engl J Med 1986;315:
1386-9.
6. Laerum E, Larsen S. Thiazide prophy-
laxis of urolithiasis: a double-blind study
in general practice. Acta Med Scand 1984;
215:383-9.
7. Barcelo P, Wuhl O, Servitge E, Rousaud
A, Pak CYC. Randomized double-blind
study of potassium citrate in idiopathic
hypocitraturic calcium nephrolithiasis.
J Urol 1993;150:1761-4.
8. Hofbauer J, Hbarth K, Szabo N, Mar-
berger M. Alkali citrate prophylaxis in
idiopathic recurrent calcium oxalate uro-
lithiasis a prospective randomized
study. Br J Urol 1994;73:362-5.
9. Fernndez-Rodrguez A, Arrabal-Mar-
tn M, Garca-Ruiz MJ, Arrabal-Polo MA,
Pichardo-Pichardo S, Zuluaga-Gmez A.
Papel de las tiazidas en la profilaxis de la
litiasis calcica recidivante. Actas Urol Esp
2006;30:305-9.
10. Saigal CS, Joyce G, Timilsina AR. Di-
rect and indirect costs of nephrolithiasis
in an employed population: opportunity
for disease management? Kidney Int 2005;
68:1808-14.
11. Rule AD, Bergstralh EJ, Melton LJ III,
Li X, Weaver AL, Lieske JC. Kidney stones
and the risk for chronic kidney disease.
Clin J Am Soc Nephrol 2009;4:804-11.
12. Madore F, Stampfer MJ, Rimm EB,
Curhan GC. Nephrolithiasis and risk of
hypertension. Am J Hypertens 1998;11:
46-53.
13. Taylor EN, Stampfer MJ, Curhan GC.
962 n engl j med 363;10
BACK to TOC
of m e dic i n e
We thank Andrew Evan, Ph.D., for providing original versions
of the figures in the Supplementary Appendix and guidance in
the preparation of the manuscript.
Obesity, weight gain, and the risk of kid-
ney stones. JAMA 2005;293:455-62.
14. Coe FL, Evan A, Worcester E. Kidney
stone disease. J Clin Invest 2005;115:2598-
608.
15. Evan AP, Coe FL, Lingeman JE, et al.
Mechanism of formation of human cal-
cium oxalate renal stones on Randalls
plaque. Anat Rec (Hoboken) 2007;290:
1315-23.
16. De Yoreo JJ, Qiu SR, Hoyer JR. Molec-
ular modulation of calcium oxalate crys-
tallization. Am J Physiol Renal Physiol
2006;291:F1123-F1131.
17. Tiselius HG, Berg C, Fornander AM,
Nilsson MA. Effects of citrate on the dif-
ferent phases of calcium oxalate crystal-
lization. Scanning Microsc 1993;7:381-9.
18. Gambaro G, Fabris A, Puliatta D,
Lupo A. Lithiasis in cystic kidney disease
and malformations of the urinary tract.
Urol Res 2006;34:102-7.
19. Worcester EM, Coe FL. New insights
into the pathogenesis of idiopathic hyper-
calciuria. Semin Nephrol 2008;28:120-32.
20. Worcester EM, Gillen DL, Evan AP, et
al. Evidence that postprandial reduction
of renal calcium reabsorption mediates
hypercalciuria of patients with calcium
nephrolithiasis. Am J Physiol Renal Phys-
iol 2007;292:F66-F75.
21. Coe FL, Favus MJ, Crockett T, et al.
Effects of low-calcium diet on urine cal-
cium excretion, parathyroid function and
serum 1,25(OH)2D3 levels in patients
with idiopathic hypercalciuria and in nor-
mal subjects. Am J Med 1982;72:25-32.
22. Voss S, Hesse A, Zimmermann DJ,
Sauerbruch T, von Unruh GE. Intestinal
oxalate absorption is higher in idiopathic
calcium oxalate stone formers than in
healthy controls: measurements with the
[(13)C2]oxalate absorption test. J Urol
2006;175:1711-5.
23. Taylor EN, Curhan GC. Determinants
of 24-hour urinary oxalate excretion. Clin
J Am Soc Nephrol 2008;3:1453-60.
24. Borghi L, Schianchi T, Meschi T, et al.
Comparison of two diets for the preven-
tion of recurrent stones in idiopathic hy-
percalciuria. N Engl J Med 2002;346:77-
84.
25. Hamm LL, Hering-Smith KS. Patho-
physiology of hypocitraturic nephrolithia-
sis. Endocrinol Metab Clin North Am 2002;
31:885-93.
26. Coe FL, Kavalach AG. Hypercalciuria
and hyperuricosuria in patients with cal-
cium nephrolithiasis. N Engl J Med 1974;
291:1344-50.
nejm.org september 2, 2010
27. Miller NL, Gillen DL, Williams JC Jr,
et al. A formal test of the hypothesis that
idiopathic calcium oxalate stones grow
on Randalls plaque. BJU Int 2009;103:966-
71.
28. Evan AP, Lingeman JE, Coe FL, et al.
Crystal-associated nephropathy in patients
with brushite nephrolithiasis. Kidney Int
2005;67:576-91.
29. Evan AE, Lingeman JE, Coe FL, et al.
Histopathology and surgical anatomy of
patients with primary hyperparathyroid-
ism and calcium phosphate stones. Kid-
ney Int 2008;74:223-9.
30. Parks JH, Coe FL, Evan AP, Worcester
EM. Urine pH in renal calcium stone
formers who do and do not increase stone
phosphate content with time. Nephrol Dial
Transplant 2009;24:130-6.
31. Parks JH, Goldfisher E, Asplin JR, Coe
FL. A single 24-hour urine collection is
inadequate for the medical evaluation of
nephrolithiasis. J Urol 2002;167:1607-12.
32. Hollingsworth JM, Rogers MA, Kauf-
man SR, et al. Medical therapy to facili-
tate urinary stone passage: a meta-analy-
sis. Lancet 2006;368:1171-9.
33. Wignall GR, Canales BK, Denstedt JD,
Monga M. Minimally invasive approaches
to upper urinary tract urolithiasis. Urol
Clin North Am 2008;35:441-54.
34. Keeley FX Jr, Assimos DG. Clinical tri-
als of the surgical management of uro-
lithiasis: current status and future needs.
Adv Chronic Kidney Dis 2009;16:65-9.
35. Curhan GC, Willett WC, Speizer FE,
Stampfer MJ. Twenty-four-hour urine
chemistries and the risk of kidney stones
among women and men. Kidney Int 2001;
59:2290-8.
36. Borghi L, Meschi T, Amato F, Briganti
A, Novarini A, Giannini A. Urinary vol-
ume, water and recurrences of idiopathic
calcium nephrolithiasis: a 5-year random-
ized prospective study. J Urol 1996;155:
839-43.
37. Hiatt RA, Ettinger B, Caan B, Quesen-
berry CP Jr, Duncan D, Citron JT. Ran-
domized controlled trial of low animal
protein, high fiber diet in the prevention
of recurrent calcium oxalate kidney stones.
Am J Epidemiol 1996;144:25-33.
38. Nouvenne A, Meschi T, Prati B, et al.
Effects of a low-salt diet on idiopathic
hypercalciuria in calcium-oxalate stone
formers: a 3-mo randomized controlled
trial. Am J Clin Nutr 2010;91:565-70.
39. Heilberg IP, Weisinger JR. Bone dis-
ease in idiopathic hypercalciuria. Curr
Opin Nephrol Hypertens 2006;15:394-402.
23

40. Lauderdale DS, Thisted RA, Wen M,
Favus MJ. Bone mineral density and frac-
ture among prevalent kidney stone cases
in the Third National Health and Nutri-
tion Examination Survey. J Bone Miner
Res 2001;16:1893-8.
41. Hoppe B, Beck BB, Milliner DS. The
primary hyperoxalurias. Kidney Int 2009;
75:1264-71.
42. Odvina CV, Preminger GM, Lindberg
JS, Moe OW, Pak CY. Long-term combined
treatment with thiazide and potassium
citrate in nephrolithiasis does not lead to
hypokalemia or hypochloremic metabolic
alkalosis. Kidney Int 2003;63:240-7.
Fox
n engl j med 363;10
BACK to TOC
clinical pr actice
43. Parks JH, Coe FL. Evidence for durable
kidney stone prevention over several de-
cades. BJU Int 2009;103:1238-46.
44. Parks JH, Asplin JR, Coe FL. Patient
adherence to long-term medical treatment
of kidney stones. J Urol 2001;166:2057-
60.
45. Mandel N, Mandel I, Fryjoff K, Rej-
niak T, Mandel G. Conversion of calcium
oxalate to calcium phosphate with recur-
rent stone episodes. J Urol 2003;169:2026-
9.
46. Parks JH, Worcester EM, Coe FL, Evan
AP, Lingeman JE. Clinical implications of
abundant calcium phosphate in routinely
nejm.org september 2, 2010
analyzed kidney stones. Kidney Int 2004;
66:777-85.
47. Gault MH, Parfrey PS, Robertson WG.
Idiopathic calcium phosphate nephro-
lithiasis. Nephron 1988;48:265-73.
48. Kacker R, Meeks JJ, Zhao L, Nadler
RB. Decreased stone-free rates after per-
cutaneous nephrolithotomy for high cal-
cium phosphate composition kidney stones.
J Urol 2008;180:958-60.
49. Taylor EN, Fung TT, Curhan GC.
DASH-style diet associates with reduced
risk for kidney stones. J Am Soc Nephrol
2009;20:2253-9.
Copyright 2010 Massachusetts Medical Society.
Robert Raichelson, M.D.
963
24
 The n e w e ng l a n d j o u r na l of m e dic i n e

clinical practice

Emergency Treatment of Asthma

Stephen C. Lazarus, M.D.

This Journal feature begins with a case vignette highlighting a common clinical problem.
Evidence supporting various strategies is then presented, followed by a review of formal guidelines,
when they exist. The article ends with the authors clinical recommendations.

A 46-year-old woman who has had two admissions to the intensive care unit (ICU) for
asthma during the past year presents with a 4-day history of upper respiratory illness
and a 6-hour history of shortness of breath and wheezing. An inhaled corticosteroid
has been prescribed, but she takes it only when she has symptoms, which is rarely.
She generally uses albuterol twice per day but has increased its use to six to eight times
per day for the past 3 days. How should this case be managed in the emergency de-
partment?

The Cl inic a l Probl em

Asthma is one of the most common diseases in developed countries and has a From the Division of Pulmonary and Crit-
worldwide prevalence of 7 to 10%.1 It is also a common reason for urgent care and ical Care Medicine and the Cardiovascu-
lar Research Institute, University of Cali-
emergency department visits. From 2001 through 2003 in the United States, asthma fornia, San Francisco, San Francisco.
accounted for an average 4210 deaths annually and an average annual total of ap- Address reprint requests to Dr. Lazarus
proximately 504,000 hospitalizations and 1.8 million emergency department visits.2 at the University of California, San Fran-
cisco, 505 Parnassus Ave., San Francisco,
The average annual rate of emergency department visits for asthma was 8.8 per 100 CA 94143-0111, or at lazma@ucsf.edu.
persons with current asthma. Rates were higher among children than among
adults (11.2 vs. 7.8 visits per 100 persons), among blacks than among whites (21 vs. N Engl J Med 2010;363:755-64.
Copyright 2010 Massachusetts Medical Society.
7 visits per 100 persons), and among Hispanics than among non-Hispanics (12.4
vs. 8.4 visits per 100 persons). Women made twice the number of emergency depart-
ment visits as men.2 Approximately 10% of visits result in hospitalization.1
Asthma is a heterogeneous disease, with varied triggers, manifestations, and
An audio version Click here to
responsiveness to treatment. Some patients with acute severe asthma presenting to of this article access audio
the emergency department have asthma that responds rapidly to aggressive therapy, is available at version.
and they can be discharged quickly; others require admission to the hospital for NEJM.org
more prolonged treatment. The reasons for this difference in responsiveness to
treatment include the degree of airway inflammation, presence or absence of mucus
plugging, and individual responsiveness to 2-adrenergic and corticosteroid medi-
cations. The major challenge in the emergency department is determining which
patients can be discharged quickly and which need to be hospitalized.

S t r ategie s a nd E v idence

Initial Assessment in the Emergency Department

Patients presenting to the emergency department with asthma should be evaluated
and triaged quickly to assess the severity of the exacerbation and the need for urgent
intervention (Fig. 1). A brief history should be obtained, and a limited physical ex-
amination performed. This assessment should not delay treatment; it can be per-
formed while patients receive initial treatment. Clinicians should search for signs

n engl j med 363;8 nejm.org august 19, 2010 755

BACK to TOC
25
 The n e w e ng l a n d j o u r na l of m e dic i n e

Triage patient immediately

Take brief history to ascertain risk factors
Previous intubation or ICU admission
2 Hospitalizations or 3 emergency department visits in past yr
>2 Canisters of short-acting 2-adrenergic agonist per mo
Coexisting conditions
Assess vital signs and perform brief physical examination
Observe for breathlessness
Measure respiratory rate and heart rate, check for pulsus paradoxus
Note whether accessory muscles of respiration used
Perform chest examination

Initiate treatment with oxygen to

achieve SaO2 90%

Assess for mild-to-moderate exacerbation Assess for severe exacerbation

FEV1 or PEF 40% FEV1 or PEF <40%
Patient talks in sentences Patient talks in words or phrases but
Pulse 120 beats/min not sentences
Minimal or no pulsus paradoxus Pulse >120 beats/min, respiratory
SaO2 90% rate >30 breaths per min
Pulsus paradoxus (decrease in systolic
arterial pressure by >25 mm Hg on
inspiration)
Loud wheezes or silent chest
SaO2 <90% or PaO2 <60 mm Hg

Initiate treatment Initiate treatment

Short-acting 2-adrenergic agonist
administered by means of a metered-
dose inhaler with valved holding
chamber or a nebulizer, up to 3 doses
in first hr
Oral corticosteroids if no immediate
response or if patient recently received
systemic corticosteroids
High-dose short-acting 2-adrenergic
agonist plus ipratropium bromide
administered by means of a metered-
dose inhaler with valved holding
chamber or a nebulizer every 20 min
or continuously for 1 hr
Oral corticosteroids

Reassess history, symptoms, vital signs, results

of physical examination, PEF, and SaO2
after 6090 min of treatment

Figure 1. Initial Assessment of a Patient Presenting to the Emergency Department with Asthma.
Adapted from the National Asthma Education and Prevention Program Expert Panel Report 3.3 FEV1 denotes forced
expiratory volume in 1 second, ICU intensive care unit, PaCO2 partial pressure of arterial carbon dioxide, PaO2 partial
pressure of arterial oxygen, PEF peak expiratory flow, and SaO2 arterial oxygen saturation.

of life-threatening asthma (e.g., altered mental
status, paradoxical chest or abdominal movement,
or absence of wheezing), which necessitate ad-
mission. Attention should be paid to factors that
are associated with an increased risk of death
from asthma, such as previous intubation or ad-
mission to an ICU, two or more hospitalizations
for asthma during the past year, low socioeco-
nomic status, and various coexisting illnesses.3
The measurement of lung function (e.g., forced
expiratory volume in 1 second [FEV1] or peak ex-
piratory flow [PEF]) can be helpful for assessing

756 n engl j med 363;8 nejm.org august 19, 2010

BACK to TOC
26
 clinical pr actice
the severity of an exacerbation and the response
to treatment but should not delay the initiation of
treatment. Laboratory and imaging studies should
be performed selectively, to assess patients for
impending respiratory failure (e.g., by measuring
the partial pressure of arterial carbon dioxide
[PaCO2]), suspected pneumonia (e.g., by obtaining
a complete blood count or a chest radiograph), or
certain coexisting conditions such as heart dis-
ease (e.g., by obtaining an electrocardiogram).
Treatment in the Emergency Department
All patients should be treated initially with sup-
plementary oxygen to achieve an arterial oxygen
saturation of 90% or greater, inhaled short-acting
2-adrenergic agonists, and systemic corticoste-
roids (Fig. 1). The dose and timing of these agents
and the use of additional pharmacologic therapy
depend on the severity of the exacerbation.
2-Adrenergic Agonists
Inhaled short-acting 2-adrenergic agonists should
be administered immediately on presentation,
and administration can be repeated up to three
times within the first hour after presentation.
The use of a metered-dose inhaler with a valved
holding chamber is as effective as the use of a
pressurized nebulizer in randomized trials,4,5
but proper technique is often difficult to ensure
in ill patients. Most guidelines recommend the
use of nebulizers for patients with severe exacerba-
tions; metered-dose inhalers with holding cham-
bers can be used for patients with mild-to-mod-
erate exacerbations, ideally with supervision from
trained respiratory therapists or nursing person-
nel (see the Supplementary Appendix and a
Video, both available at NEJM.org, for descrip-
tions of how to use inhalers with and inhalers
without a holding chamber, respectively). The
dose administered by means of metered-dose in-
halers for exacerbations is substantially greater
than that used for routine relief: four to eight
puffs of albuterol can be administered every 20
minutes for up to 4 hours and then every 1 to
4 hours as needed (Table 1). Albuterol can be
delivered by means of a nebulizer either inter-
mittently or continuously. A meta-analysis of re-
sults from six randomized trials indicated that
intermittent administration and continuous ad-
ministration have similar effects on both lung
function and the overall rate of hospitalization,6
n engl j med 363;8
BACK to TOC
whereas a Cochrane review of findings from
eight trials suggested that continuous adminis-
tration resulted in greater improvement in PEF
and FEV1 and a greater reduction in hospital ad-
missions, particularly among patients with severe
asthma.7
Albuterol is the inhaled 2-adrenergic agonist
most widely used for emergency management.
Levalbuterol, the R-enantiomer of albuterol, has
been shown to be effective at half the dose of
albuterol, but randomized trials conducted in the
emergency department have not consistently
shown a clinical advantage of levalbuterol over
racemic albuterol.8,9 Pirbuterol and bitolterol are
effective for mild or moderate exacerbations, but
a higher dose is required than with albuterol or
levalbuterol, and their use for severe exacerba-
tions has not been studied.
Oral or parenteral administration of 2-
adrenergic agonists is not recommended, since
neither has been shown to be more effective than
inhaled 2-adrenergic agonists, and both are
associated with an increased frequency of side
effects. The long-acting inhaled 2-adrenergic
salmeterol has not been studied for the treat-
ment of exacerbations, though trials with formot-
erol (ClinicalTrials.gov numbers, NCT00819637
and NCT00900874) are under way.
Anticholinergic Agents
Because of its relatively slow onset of action, in-
haled ipratropium is not recommended as mono-
therapy in the emergency department but can be
added to a short-acting 2-adrenergic agonist for
a greater and longer-lasting bronchodilator ef-
fect.10,11 In patients with severe airflow obstruc-
tion, the use of ipratropium together with a 2-
adrenergic agonist in the emergency department,
as compared with a 2-adrenergic agonist alone,
has been shown to reduce rates of hospitaliza-
tion by approximately 25%,12,13 although there is
no apparent benefit of continuing ipratropium
after hospitalization.
Systemic Corticosteroids
In most patients with exacerbations that neces-
sitate treatment in the emergency department,
systemic corticosteroids are warranted. The ex-
ception is the patient who has a rapid response
to initial therapy with an inhaled 2-adrenergic
agonist. Although most randomized, controlled
nejm.org august 19, 2010 757
27
 758
BACK to TOC
Table 1. Medications for Treatment of Asthma Exacerbation in the Emergency Department.*
Drug and Available Formulation
Short-acting 2-adrenergic agonists
Albuterol
Metered-dose inhaler (90 g/puff)
Nebulizer solution (0.63 mg/3 ml, 1.25
mg/3 ml, 2.5 mg/3 ml, or 5.0 mg/ml)
Levalbuterol
Metered-dose inhaler (45 g/puff)
Nebulizer solution (0.63 mg/3 ml, 1.25
n engl j med 363;8
mg/0.5 ml, or 1.25 mg/3 ml)
Bitolterol
nejm.org
Metered-dose inhaler (370 g/puff)
Nebulizer solution (2 mg/ml)
august 19, 2010
Pirbuterol, metered-dose inhaler (200 g/puff) Same as for albuterol, metered-dose inhaler; pirbuterol
Anticholinergic agents
Ipratropium bromide
Metered-dose inhaler (18 g/puff)
Nebulizer solution (0.25 mg/ml)
28
Dose Comments
Adverse effects include tachycardia, palpitations, tremor, and hypo-
kalemia.
48 puffs every 20 min up to 4 hr, then every 14 hr as needed
2.55 mg every 20 min over the first hr, then 2.510 mg every For optimal delivery, dilute solution to a minimum of 3 ml at a gas flow
14 hr as needed or 1015 mg/hr continuously of 68 liters/min. Use large-volume nebulizers for continuous admin-
The
istration.
Same as for albuterol, metered-dose inhaler; levalbuterol
administered in half the milligram dose of albuterol has
similar efficacy and safety
1.252.5 mg every 20 min over the first hr, then 1.255 mg
every 14 hr as needed; levalbuterol administered at half
the milligram dose of albuterol has similar efficacy and
safety; continuous nebulization has not been evaluated
Has not been studied in patients with severe asthma exacerbations. Not
available in the United States.
n e w e ng l a n d j o u r na l
Same as for albuterol, metered-dose inhaler; bitolterol thought
of
to be half as potent as albuterol on a milligram basis
Same as for albuterol, nebulizer solution; bitolterol thought
to be half as potent as albuterol on a milligram basis
Has not been studied in patients with severe asthma exacerbations.
thought to be half as potent as albuterol on a milligram
m e dic i n e
basis
Adverse effects include dry mouth, cough, and blurred vision.
Should not be used as first-line therapy; should be added to short-acting
2-adrenergic agonist therapy for severe exacerbations. The addition
of ipratropium to a short-acting 2-adrenergic agonist has not been
shown to provide further benefit once the patient is hospitalized.
8 puffs every 20 min as needed, for up to 3 hr
0.5 mg every 20 min for 1 hr (three doses), then as needed;
can be used with albuterol in one nebulizer
 clinical pr actice

trials of corticosteroids in patients seen in the

Adverse effects include adrenal suppression, growth suppression, osteo-

paired. The total course of systemic corticosteroids for an asthma ex-

porosis, muscle weakness, hypertension, weight gain, diabetes, cata-

emergency department and those admitted to the

acerbation necessitating an emergency department visit or hospital-

May be used for up to 3 hr during initial management of severe exacerba-
tions. The addition of ipratropium to albuterol has not been shown

probably is no need to taper, especially if patients are concurrently

There is no known advantage of higher doses of corticosteroids to treat
severe asthma exacerbations or of intravenous administration over

there is no need to taper the dose; for courses of 710 days, there
hospital have been small, these studies individu-

oral therapy, provided that gastrointestinal absorption is not im-

ization may be 3 to 10 days. For corticosteroid courses of <1 wk,

to provide further benefit once the patient is hospitalized. ally14,15 and collectively16-18 show that the use, as
compared with nonuse, of systemic corticoste-
roids is associated with a more rapid improvement
in lung function, fewer hospitalizations, and a

racts, Cushings syndrome, and dermal thinning.

lower rate of relapse after discharge from the
emergency department. Because comparisons of
oral prednisone and intravenous corticosteroids
have not shown differences in the rate of im-

receiving inhaled corticosteroids.

provement of lung function or in the length of
the hospital stay,19-21 the oral route is preferred for
patients with normal mental status and without
conditions expected to interfere with gastrointes-
tinal absorption. Although the optimal dose of
corticosteroid is not known, pooled data from
controlled trials involving patients seen in the
emergency department or admitted to the hospi-
tal have shown no significant advantage of doses
greater than 100 mg per day of prednisone equiv-
alent.19,20,22-25 The most recent guidelines from
peak expiratory flow reaches 70% of predicted value or a
4080 mg/day in one dose or two divided doses, given until

the National Asthma Education and Prevention

Program (NAEPP) (Expert Panel Report 3) recom-
mend the use of 40 to 80 mg per day in one dose
or two divided doses.3
* Adapted from the National Asthma Education and Prevention Program Expert Panel Report 3.3
Nebulizer solution (each 3-ml vial contains 3 ml every 20 min for 3 doses, then as needed
8 puffs every 20 min as needed, up to 3 hr

Inhaled Corticosteroids
Although high-dose inhaled corticosteroids are
often used to treat worsening of asthma control
and to try to prevent exacerbations, the evidence
does not support the use of inhaled corticoster-
personal best value

oids as a substitute for systemic corticosteroids

in the emergency department.26 Inhaled cortico-
steroids are, however, preferred for long-term
asthma control. At the time of discharge from
the emergency department, these agents should
be continued in patients who have been taking
them for long-term control and should be pre-
Levalbuterol is the R-enantiomer of albuterol.

scribed for patients who have not previously taken

Metered-dose inhaler (each puff contain-

Systemic corticosteroids: prednisone, pred-

0.5 mg of ipratropium bromide and
ing 18 g of ipratropium and 90 g

nisolone, and methylprednisolone

them. In a randomized, controlled trial of 1006

consecutively enrolled patients with acute asthma
Ipratropium bromide and albuterol

treated in a Canadian emergency department,

the addition at discharge of inhaled budesonide
2.5 mg of albuterol)

(for 21 days) to treatment with oral corticoste-

roids (for 5 to 10 days) was associated with a 48%
of albuterol)

reduction in the rate of relapse at 21 days and

with improvement in the quality of life with re-
spect to asthma (as measured by the Asthma
Quality of Life Questionnaire) and symptoms, as
compared with treatment with oral corticoste-
roids alone.27

n engl j med 363;8 nejm.org august 19, 2010 759

BACK to TOC
29
 The n e w e ng l a n d j o u r na l of m e dic i n e

Reassess history, symptoms, vital signs, results

of physical examination, PEF, and SaO2 after
6090 min of treatment

Patient has continued mild-to-moderate Patient has continued severe

exacerbation exacerbation

Continue treatment Continue treatment

Oxygen to achieve SaO2 90% Oxygen to achieve SaO2 90%
Short-acting 2-adrenergic agonist Short-acting 2-adrenergic agonist plus
administered by means of a ipratropium bromide administered
metered-dose inhaler with valved by means of a metered-dose inhaler
holding chamber or a nebulizer, with valved holding chamber or a
every 60 min nebulizer, every hr or continuously
Oral corticosteroids Oral corticosteroids
Continue treatment 13 hr, provided Consider magnesium sulfate or heliox
there is improvement

Within <4 hr, make decision to admit or discharge

Patient has good response Patient has incomplete response Patient has poor response
FEV1 or PEF 70% sus- FEV1 or PEF 4069% FEV1 or PEF <40%
tained for 60 min Mild-to-moderate symptoms PaCO2 42 mm Hg
No distress Severe symptoms
Normal examination Drowsiness, confusion

Discharge Discharge or admit, on the basis Admit

of risk factors, likelihood of ad-
herence, and home environment

Figure 2. Continued Management of Asthma in the Emergency Department.

Adapted from the National Asthma Education and Prevention Program Expert Panel Report 3.3 Heliox is a mixture
of helium and oxygen, usually 79% and 21%, respectively, whose density is about one third that of air. FEV1 denotes
forced expiratory volume in 1 second, PaCO2 partial pressure of arterial carbon dioxide, PEF peak expiratory flow,
and SaO2 arterial oxygen saturation.

Treatments That Are Not Recommended Assessment of Response to Treatment

Although methylxanthines were once a standard
treatment for asthma in the emergency depart-
ment, it is now clear that their use increases the
risk of adverse events without improving out-
comes.28 Antibiotics should not be used routinely
but rather should be reserved for patients in whom
bacterial infection (e.g., pneumonia or sinusitis)
seems likely. Similarly, neither aggressive hydra-
tion nor administration of mucolytic agents is
recommended for acute exacerbations.3
Patients should be reassessed after the first treat-
ment with an inhaled bronchodilator and again
at 60 to 90 minutes (i.e., after three treatments).3
This assessment should include a survey of symp-
toms, a physical examination, and measurement
of FEV1 or PEF (Fig. 2). For the most severe exac-
erbations, this repeat assessment should prob-
ably include the measurement of arterial blood
gases. Most patients will have clinically signifi-
cant improvement after one dose of an inhaled

760 n engl j med 363;8 nejm.org august 19, 2010

BACK to TOC
30
 clinical pr actice
bronchodilator, and 60 to 70% will meet the cri-
teria for discharge from the emergency depart-
ment (see below) after three doses.29-31 The de-
gree of subjective and objective improvement that
occurs in response to treatment predicts the need
for hospitalization.32-38 In a study of 720 patients
treated in 36 Australian emergency departments,
the need for hospital admission among patients
assessed as having moderate asthma, as well as
the need for ICU care of patients assessed as hav-
ing severe asthma, was better predicted by the
assessment of asthma severity after 1 hour of
treatment than by the initial assessment in the
emergency department.38
Indications for Admission
After treatment in the emergency department for
1 to 3 hours, patients who have an incomplete or
poor response, defined as an FEV1 or PEF of less
than 70% of the personal best or predicted value,
should be evaluated for admission to the hospi-
tal. Patients who have an FEV1 of less than 40%,
persistent moderate-to-severe symptoms, drows-
iness, confusion, or a PaCO2 of 42 mm Hg or
greater should be admitted. Patients who have an
FEV1 of 40 to 69% and mild symptoms should be
assessed individually for risk factors for death,
ability to adhere to a prescribed regimen, and
the presence of asthma triggers in the home. The
NAEPP Expert Panel Report 3 suggests that the
decision to admit or discharge a patient should
be made within 4 hours after presentation to the
emergency department.3
Management of Respiratory Insufficiency
Patients with altered mental status, exhaustion,
or hypercapnia should be considered for immedi-
ate intubation and ventilatory support. Because of
high positive intrathoracic pressures, intubation
and ventilation may lead to hypotension and
barotrauma. Care should be taken to ensure ade-
quate intravascular volume, and to avoid high
airway pressures. A strategy of permissive hyper-
capnia, achieved by adjusting the ventilator to
correct hypoxemia while avoiding high airway
pressures, was associated in an observational
study with decreased mortality among patients
with status asthmaticus,39 and this approach has
become standard.
Guidelines suggest that once a decision has
been made in the emergency department to in-
tubate a patient, the procedure should be semi-
elective and performed under controlled condi-
n engl j med 363;8
BACK to TOC
tions (vs. performed as an emergency procedure
by the first available staff). Randomized trials
have shown a benefit from noninvasive positive-
pressure ventilation for acute exacerbations of
chronic obstructive pulmonary disease, but most
information used to guide the ventilation strategy
for treating acute asthma comes from case re-
ports or noncontrolled studies. A randomized
crossover study that compared the use of bilevel
positive airway pressure for 2 hours with stan-
dard care in children with acute asthma showed a
significantly lower respiratory rate and improved
scores on a questionnaire regarding asthma
symptoms with bilevel positive airway pressure
but no significant difference in arterial oxygen
saturation, transcutaneous carbon dioxide levels,
or other outcomes.40 In a randomized, sham-
controlled trial of the use of bilevel positive air-
way pressure in 30 adults with acute asthma,
bilevel positive airway pressure was associated
with a higher FEV1 value at 4 hours and a lower
rate of hospitalization (17.6%, vs. 62.5% with
sham treatment).41 These data suggest that non-
invasive positive-pressure ventilation could be
considered for patients who decline intubation
and for selected patients who are likely to co-
operate with mask therapy, but more data are
needed to recommend this approach.
Discharge from the Emergency Department
Patients may be discharged if the FEV1 or PEF
after treatment is 70% or more of the personal
best or predicted value and if the improvements
in lung function and symptoms are sustained for
at least 60 minutes.3 After discharge, patients
should continue to use inhaled short-acting 2-
adrenergic agonists as needed and should be given
oral corticosteroids for 3 to 10 days3 (Table 2).
Inhaled corticosteroids can be started at any time
during treatment of the exacerbation, but initia-
tion at the time of discharge, if not before, is
prudent to reduce the risk of relapse.27,42,43
Education of Patients
The need for treatment in the emergency depart-
ment often reflects inadequate maintenance ther-
apy and insufficient knowledge of how to deal
with a worsening of asthma control. Presenta-
tion to the emergency department provides a
unique opportunity to educate patients about
medications, inhaler technique, and steps that
can reduce exposure to household triggers of
allergic reaction and to ensure that discharged
nejm.org august 19, 2010 761
31
 The n e w e ng l a n d j o u r na l
Table 2. Recommendations for Discharge from the Emergency Department.*
Medications
Continue inhaled short-acting 2-adrenergic agonists every 12 hr, as needed
Continue oral corticosteroids at a dose of 4080 mg/day for 310 days
If course is <1 wk, no need to taper the dose
If course is 710 days, probably no need to taper, especially if patients are
concurrently receiving inhaled corticosteroids
Continue or start an inhaled corticosteroid at a medium dose (e.g., beclo-
methasone [HFA], 240480 g/day; budesonide [DPI], 6001200 g/day;
or fluticasone [DPI], 300500 g/day)
Education
Review purposes and doses of asthma medications with patient
Review inhaler technique with patient
Teach patient to monitor for signs and symptoms of poor asthma control
Provide patient with an asthma action plan
Follow-up
Advise patient to call primary care provider within 35 days after discharge
Schedule a follow-up appointment with provider to occur within 14 wk
* DPI denotes dry-powder inhaler, and HFA hydrofluoroalkane formulation.
patients have an asthma action plan and instruc-
tions for monitoring their symptoms and imple-
menting their plan. A follow-up appointment
should be scheduled with the patients primary
care provider or with an asthma specialist to oc-
cur 1 to 4 weeks after discharge. Guidelines also
recommend that patients be encouraged to con-
tact their asthma care provider within 3 to 5 days
after discharge, when the risk of relapse is great-
est,3 although data are lacking to show that this
action improves outcomes.
A r e a s of Uncer ta in t y
In patients with severe asthma that is refractory
to standard treatment, intravenous magnesium
sulfate is widely used,44 but there is controversy
regarding its efficacy. A meta-analysis of 1669
patients in 24 studies who received either intra-
venous magnesium sulfate (used in 15 studies) or
nebulized magnesium sulfate (used in 9 studies)
showed that intravenous treatment was weakly
associated with improved lung function in adults
but had no significant effect on hospital admis-
sions; in children, the use of intravenous magne-
sium sulfate significantly improved lung func-
tion and reduced rates of hospital admission. The
effect of nebulized magnesium sulfate is less
762 n engl j med 363;8
BACK to TOC
of m e dic i n e
substantiated.45 Expert opinion46 and guidelines3
suggest that clinicians consider the use of intra-
venous magnesium sulfate in patients who have
severe exacerbations and whose FEV1 or PEF re-
mains less than 40% of the personal best or
predicted value after initial treatments. The re-
sults of a large multicenter trial in the United
Kingdom47 (Current Controlled Trials number,
ISRCTN04417063) comparing treatment with in-
travenous or nebulized magnesium sulfate and
standard treatment in patients with severe asthma
are expected in 2011.
Heliox is a mixture of helium and oxygen,
usually 79% and 21%, respectively, with a den-
sity about one third that of air, that reduces
airflow resistance within regions of the bron-
chial tree where turbulent flow predominates. It
is thought to reduce the work of breathing and
to improve delivery of aerosolized medications.
However, its role in the management of acute
severe asthma is unclear. A Cochrane analysis of
544 patients in 10 trials led to the conclusion
that heliox might be beneficial in patients with
severe airflow obstruction who have not had a
response to initial treatment,48 and current
guidelines reflect this conclusion.3
Since the administration of oral leukotriene
inhibitors results in increases in the FEV1 within
1 to 2 hours,49,50 there has been interest in using
these agents in the emergency department, but
their usefulness in that setting is unclear. In a
randomized, placebo-controlled trial of intrave-
nous montelukast in 583 adults whose FEV1 re-
mained at 50% or less of the predicted value after
60 minutes of standard care, the use of montelu-
kast significantly improved the FEV1 at 60 minutes
but did not reduce the rate of hospitalization.51
Guidel ine s
The NAEPP and the Global Initiative for Asthma
have developed and updated evidence-based guide-
lines for the diagnosis and management of asth-
ma.3,52 The recommendations in this article are
consistent with these guidelines.
C onclusions a nd
R ec om mendat ions
The patient described in the vignette has chronic
uncontrolled asthma necessitating daily rescue
use of albuterol, but she has not been receiving
nejm.org august 19, 2010
32
 clinical pr actice

daily controller therapy. Her history of ICU admis-
sions and excessive albuterol use indicate that she
is at increased risk for death related to asthma.
Treatment with oxygen, aerosolized albuterol
and ipratropium, and systemic corticosteroids
should be initiated. The patient should be moni-
tored closely and her signs and symptoms re-
assessed frequently, and a decision to admit or
discharge her should be made within 4 hours
after presentation. If she is discharged from the
emergency department, she should be educated
about medications, inhaler technique, and steps
for monitoring symptoms and for managing ex-
acerbations. Emergency department staff should
provide her with a discharge plan, schedule a
follow-up appointment, and ensure that she has
adequate medications or prescriptions to last un-
til that appointment. Because of her previous ad-
missions to the ICU and her history of consis-
tently poor asthma control, referral to an asthma
specialist would be prudent.
No potential conflict of interest relevant to this article was
reported.
Disclosure forms provided by the author are available with the
full text of this article at NEJM.org.

References
1. Rowe BH, Voaklander DC, Wang D,
et al. Asthma presentations by adults to
emergency departments in Alberta, Can-
ada: a large population-based study. Chest
2009;135:57-65.
2. Moorman JE, Rudd RA, Johnson CA,
et al. National surveillance for asthma
United States, 19802004. MMWR Sur-
veill Summ 2007;56:1-54.
3. National Heart, Lung, and Blood In-
stitute, National Asthma Education and
Prevention Program. Expert Panel Report
3: guidelines for the diagnosis and man-
agement of asthma: full report 2007. (Ac-
cessed July 23, 2010, at http://www.nhlbi
.nih.gov/guidelines/asthma/asthgdln.pdf.)
4. Cates CJ, Crilly JA, Rowe BH. Holding
chambers (spacers) versus nebulisers for
beta-agonist treatment of acute asthma.
Cochrane Database Syst Rev 2006;2:
CD000052.
5. Dhuper S, Chandra A, Ahmed A, et al.
Efficacy and cost comparisons of bron-
chodilatator administration between me-
tered dose inhalers with disposable spac-
ers and nebulizers for acute asthma in an
inner-city adult population. J Emerg Med
2008 December 10 (Epub ahead of print).
6. Rodrigo GJ, Rodrigo C. Continuous vs
intermittent beta-agonists in the treatment
of acute adult asthma: a systematic review
with meta-analysis. Chest 2002;122:160-5.
7. Camargo CA Jr, Spooner CH, Rowe
BH. Continuous versus intermittent beta-
agonists in the treatment of acute asth-
ma. Cochrane Database Syst Rev 2003;4:
CD001115.
8. Carl JC, Myers TR, Kirchner HL,
Kercsmar CM. Comparison of racemic
albuterol and levalbuterol for treatment of
acute asthma. J Pediatr 2003;143:731-6.
9. Qureshi F, Zaritsky A, Welch C, Mead-
ows T, Burke BL. Clinical efficacy of race-
mic albuterol versus levalbuterol for the
treatment of acute pediatric asthma. Ann
Emerg Med 2005;46:29-36.
10. Rodrigo GJ, Rodrigo C. First-line ther-
apy for adult patients with acute asthma
receiving a multiple-dose protocol of ipra-
tropium bromide plus albuterol in the
emergency department. Am J Respir Crit
Care Med 2000;161:1862-8.
11. Gelb AF, Karpel J, Wise RA, Cassino C,
Johnson P, Conoscenti CS. Bronchodilator
efficacy of the fixed combination of ipra-
tropium and albuterol compared to albu-
terol alone in moderate-to-severe persis-
tent asthma. Pulm Pharmacol Ther 2008;
21:630-6.
12. Plotnick LH, Ducharme FM. Com-
bined inhaled anticholinergics and beta2-
agonists for initial treatment of acute
asthma in children. Cochrane Database
Syst Rev 2000;4:CD000060.
13. Rodrigo GJ, Castro-Rodriguez JA.
Anticholinergics in the treatment of chil-
dren and adults with acute asthma: a sys-
tematic review with meta-analysis. Thorax
2005;60:740-6. [Errata, Thorax 2008;63:
1029, 2006;61:274, 458.]
14. Fanta CH, Rossing TH, McFadden ER
Jr. Glucocorticoids in acute asthma: a criti-
cal controlled trial. Am J Med 1983;74:
845-51.
15. Littenberg B, Gluck EH. A controlled
trial of methylprednisolone in the emer-
gency treatment of acute asthma. N Engl
J Med 1986;314:150-2.
16. Rodrigo G, Rodrigo C. Corticosteroids
in the emergency department therapy of
acute adult asthma: an evidence-based
evaluation. Chest 1999;116:285-95.
17. Rowe BH, Edmonds ML, Spooner CH,
Diner B, Camargo CA Jr. Corticosteroid
therapy for acute asthma. Respir Med
2004;98:275-84.
18. Krishnan JA, Davis SQ, Naureckas ET,
Gibson P, Rowe BH. An umbrella review:
corticosteroid therapy for adults with acute
asthma. Am J Med 2009;122:977-91.
19. Harrison BD, Stokes TC, Hart GJ,
Vaughan DA, Ali NJ, Robinson AA. Need
for intravenous hydrocortisone in addition
to oral prednisolone in patients admitted
to hospital with severe asthma without
ventilatory failure. Lancet 1986;1:181-4.
20. Ratto D, Alfaro C, Sipsey J, Glovsky
MM, Sharma OP. Are intravenous corti-
costeroids required in status asthmaticus?
JAMA 1988;260:527-9.
21. Jnsson S, Kjartansson G, Gslason D,
Helgason H. Comparison of the oral and
intravenous routes for treating asthma
with methylprednisolone and theophyl-
line. Chest 1988;94:723-6.
22. Engel T, Dirksen A, Frlund L, et al.
Methylprednisolone pulse therapy in acute
severe asthma: a randomized, double-blind
study. Allergy 1990;45:224-30.
23. Bowler SD, Mitchell CA, Armstrong
JG. Corticosteroids in acute severe asthma:
effectiveness of low doses. Thorax 1992;
47:584-7.
24. Emerman CL, Cydulka RK. A random-
ized comparison of 100-mg vs 500-mg
dose of methylprednisolone in the treat-
ment of acute asthma. Chest 1995;107:
1559-63.
25. Marquette CH, Stach B, Cardot E, et al.
High-dose and low-dose systemic cortico-
steroids are equally efficient in acute
severe asthma. Eur Respir J 1995;8:22-7.
[Erratum, Eur Respir J 1995;8:1435.]
26. Edmonds ML, Camargo CA Jr, Pollack
CV Jr, Rowe BH. Early use of inhaled
corticosteroids in the emergency depart-
ment treatment of acute asthma. Cochrane
Database Syst Rev 2003;3:CD002308.
27. Rowe BH, Bota GW, Fabris L, Ther-
rien SA, Milner RA, Jacono J. Inhaled
budesonide in addition to oral corticoste-
roids to prevent asthma relapse following
discharge from the emergency depart-
ment: a randomized controlled trial. JAMA
1999;281:2119-26.
28. Parameswaran K, Belda J, Rowe BH.
Addition of intravenous aminophylline to
beta2-agonists in adults with acute asth-
ma. Cochrane Database Syst Rev 2000;4:
CD002742.
29. Karpel JP, Aldrich TK, Prezant DJ,
Guguchev K, Gaitan-Salas A, Pathiparti R.
Emergency treatment of acute asthma
with albuterol metered-dose inhaler plus
holding chamber: how often should treat-
ments be administered? Chest 1997;112:
348-56.
30. Strauss L, Hejal R, Galan G, Dixon L,
McFadden ER Jr. Observations on the ef-
fects of aerosolized albuterol in acute

n engl j med 363;8 nejm.org august 19, 2010 763

BACK to TOC
33
 clinical pr actice

asthma. Am J Respir Crit Care Med 1997;
155:454-8.
31. Rodrigo C, Rodrigo G. Therapeutic
response patterns to high and cumulative
doses of salbutamol in acute severe asth-
ma. Chest 1998;113:593-8.
32. Rodrigo G, Rodrigo C. Early predic-
tion of poor response in acute asthma
patients in the emergency department.
Chest 1998;114:1016-21.
33. Chey T, Jalaludin B, Hanson R, Leeder
S. Validation of a predictive model for
asthma admission in children: how accu-
rate is it for predicting admissions? J Clin
Epidemiol 1999;52:1157-63.
34. McCarren M, Zalenski RJ, McDermott
M, Kaur K. Predicting recovery from acute
asthma in an emergency diagnostic and
treatment unit. Acad Emerg Med 2000;7:
28-35.
35. Karras DJ, Sammon ME, Terregino CA,
Lopez BL, Griswold SK, Arnold GK. Clini-
cally meaningful changes in quantitative
measures of asthma severity. Acad Emerg
Med 2000;7:327-34.
36. Smith SR, Baty JD, Hodge D III. Vali-
dation of the pulmonary score: an asthma
severity score for children. Acad Emerg
Med 2002;9:99-104.
37. Gorelick MH, Stevens MW, Schultz
TR, Scribano PV. Performance of a novel
clinical score, the Pediatric Asthma Se-
verity Score (PASS), in the evaluation of
acute asthma. Acad Emerg Med 2004;11:
10-8.
38. Kelly AM, Kerr D, Powell C. Is severity
assessment after one hour of treatment
better for predicting the need for admis-
sion in acute asthma? Respir Med 2004;98:
777-81.
39. Darioli R, Perret C. Mechanical con-
trolled hypoventilation in status asthmati-
cus. Am Rev Respir Dis 1984;129:385-7.
40. Thill PJ, McGuire JK, Baden HP, Green
TP, Checchia PA. Noninvasive positive-
pressure ventilation in children with lower
airway obstruction. Pediatr Crit Care Med
2004;5:337-42. [Erratum, Pediatr Crit Care
Med 2004;5:590.]
41. Soroksky A, Stav D, Shpirer I. A pilot
prospective, randomized, placebo-con-
trolled trial of bilevel positive airway pres-
sure in acute asthmatic attack. Chest 2003;
123:1018-25.
42. Blais L, Ernst P, Boivin JF, Suissa S.
Inhaled corticosteroids and the prevention
of readmission to hospital for asthma. Am
J Respir Crit Care Med 1998;158:126-32.
43. Sin DD, Man SF. Low-dose inhaled
corticosteroid therapy and risk of emer-
gency department visits for asthma. Arch
Intern Med 2002;162:1591-5.
44. Jones LA, Goodacre S. Magnesium sul-
phate in the treatment of acute asthma:
evaluation of current practice in adult
emergency departments. Emerg Med J
2009;26:783-5.
45. Mohammed S, Goodacre S. Intrave-
nous and nebulised magnesium sulphate
for acute asthma: systematic review and
meta-analysis. Emerg Med J 2007;24:823-
30.
46. Rowe BH, Camargo CA Jr. The role of
magnesium sulfate in the acute and chron-
ic management of asthma. Curr Opin
Pulm Med 2008;14:70-6.
47. NIHR Health Technology Assessment
Program. The 3Mg Trial: randomised con-
trolled trial of intravenous or nebulised
magnesium sulphate or standard therapy
for acute severe asthma. 2010. (Accessed
July 23, 2010, at http://www.hta.ac.uk/
project/1619.asp.)
48. Rodrigo G, Pollack C, Rodrigo C,
Rowe BH. Heliox for nonintubated acute
asthma patients. Cochrane Database Syst
Rev 2006;4:CD002884.
49. Liu MC, Dub LM, Lancaster J. Acute
and chronic effects of a 5-lipoxygenase
inhibitor in asthma: a 6-month random-
ized multicenter trial. J Allergy Clin Im-
munol 1996;98:859-71.
50. Dockhorn RJ, Baumgartner RA, Leff
JA, et al. Comparison of the effects of in-
travenous and oral montelukast on airway
function: a double blind, placebo con-
trolled, three period, crossover study in
asthmatic patients. Thorax 2000;55:260-5.
51. Camargo CA Jr, Gurner DM, Smith-
line HA, et al. A randomized placebo-
controlled study of intravenous montelu-
kast for the treatment of acute asthma.
J Allergy Clin Immunol 2010;125:374-80.
52. Bateman ED, Hurd SS, Barnes PJ, et al.
Global strategy for asthma management
and prevention: GINA executive summary.
Eur Respir J 2008;31:143-78.
Copyright 2010 Massachusetts Medical Society.

COLLECTIONS OF ARTICLES ON THE JOURNALS WEB SITE

The Journals Web site (NEJM.org) sorts published articles into
more than 50 distinct clinical collections, which can be used as convenient
entry points to clinical content. In each collection, articles are cited in reverse
chronologic order, with the most recent first.

764 n engl j med 363;8 nejm.org august 19, 2010

BACK to TOC
34

From the Taub Institute for Research on
Alzheimers Disease and the Aging Brain
and the Gertrude H. Sergievsky Center,
Columbia University, New York. Address
reprint requests to Dr. Mayeux at the
Taub Institute for Research on Alzheimers
Disease and the Aging Brain and the Ger-
trude H. Sergievsky Center, Columbia
University, 630 W. 168th St., New York,
NY 10032, or at rpm2@columbia.edu.
This article (10.1056/NEJMcp0910236) was
updated on September 15, 2010, at NEJM
.org.
N Engl J Med 2010;362:2194-201.
Copyright 2010 Massachusetts Medical Society.
Click here to An audio version
access audio of this article
version. is available at
NEJM.org
2194
BACK to TOC
The n e w e ng l a n d j o u r na l of m e dic i n e
clinical practice
Early Alzheimers Disease
Richard Mayeux, M.D.
This Journal feature begins with a case vignette highlighting a common clinical problem.
Evidence supporting various strategies is then presented, followed by a review of formal guidelines,
when they exist. The article ends with the authors clinical recommendations.
A 72-year-old man who is still managing investments at a brokerage firm seeks consul-
tation at the urging of his wife for increasing difficulty with memory over the past
2 years. Clients have expressed concern about his occasional lapses in memory. His
wife reports that he frequently repeats questions about social appointments and be-
comes angry when she points this out. The physical examination is normal, but the
patient has difficulty remembering elements of a brief story and adding a small
amount of change. He has a score of 28 out of 30 on the MiniMental State Examina-
tion, indicating slightly impaired cognitive function.1 Early Alzheimers disease is
suspected. How should the patient be further evaluated and treated?
The Cl inic a l Probl em
Alzheimers disease is the most frequent cause of dementia in Western societies,
affecting an estimated 5 million people in the United States and 17 million world-
wide.2 The annual incidence worldwide increases from 1% between the ages of 60
and 70 years to 6 to 8% at the age of 85 years or older.3 In countries in which sur-
vival to the age of 80 years or older is not uncommon, the proportion of persons in
this age group with Alzheimers disease now approaches 30% and is expected to
continue to increase substantially.4 The disease onset is insidious, and manifesta-
tions evolve over a period of years from mildly impaired memory to severe cognitive
loss. A transitional state, referred to as mild cognitive impairment, often precedes
the earliest manifestations of Alzheimers disease.5 The course of Alzheimers dis-
ease is inevitably progressive and terminates in mental and functional incapacity
and death. Plateaus sometimes occur in which the degree of cognitive impairment
is stable for 1 or 2 years, but progression usually resumes thereafter.
An inability to retain recently acquired information is typically the initial symp-
tom, whereas memory for remote events is relatively spared until later. With disease
progression, impairment in other areas of cognition (e.g., language, abstract rea-
soning, and executive function or decision making) occurs to varying degrees and
typically coincides with difficulty at work or in social situations or household ac-
tivities. Changes in mood and affect often accompany the decline in memory.6 Delu-
sions and psychotic behavior are not typically presenting signs but can occur at any
time during the disease course.7 The occurrence of psychosis during the initial
stages of dementia suggests other diagnoses, such as dementia with Lewy bodies.
At autopsy, the most frequent pathological features in the brains of patients with
Alzheimers disease include extracellular beta-amyloid protein in diffuse plaques
and in plaques containing elements of degenerating neurons, termed neuritic
plaques.8 Intracellular changes include deposits of hyperphosphorylated tau pro-
tein, a microtubule assembly protein, in the form of neurofibrillary tangles. These
pathological lesions first appear in the entorhinal regions of the hippocampus and
n engl j med 362;23 nejm.org june 10, 2010
35
 clinical pr actice
then become widespread. Over time, there is wide-
spread loss of neurons and synapses. The patho-
genic mechanisms that are responsible for the
development of these changes are unknown.
A family history of dementia is one of the most
consistently reported risk factors for Alzheimers
disease.3 There are rare cases of families with
autosomal dominant inheritance of Alzheimers
disease that develops between the ages of 30 and
50 years; about half these cases result from mu-
tations in genes encoding amyloid precursor pro-
tein, presenilin 1, or presenilin 2.9 Studies of these
mutated genes have led to the assertion that Alz-
heimers disease is caused by the generation and
aggregation of beta-amyloid peptide, which then
forms neuritic plaques. Although several hundred
families carry these mutations, they account for
less than 1% of cases.
First-degree relatives of patients with late-onset
disease have approximately twice the expected
lifetime risk of the disease. The disease is also
more often concordant among monozygotic twins
than among dizygotic twins.10 Individuals from
families that have many members with late-onset
Alzheimers disease are at increased risk for de-
mentia, but the distribution of cases is rarely con-
sistent with mendelian inheritance.
The genetic variant encoding apolipoprotein
(APOE) 4 is the only well-established mutation
associated with the late-onset form of Alzhei-
mers disease.11 Risks that are associated with
the APOE 4 allele peak between the ages of 60
and 80 years. As compared with the absence of
the APOE 4 allele, the presence of one such al-
lele is associated with a doubling or tripling of the
lifetime risk of disease, and the presence of two
copies is associated with an increase in risk by a
factor of five or more. Associations between Alz-
heimers disease and variants in sortilin-related
receptor 1 (SORL1),12 clusterin, phosphatidylinos-
itol-binding clathrin assembly protein, and a com-
plement component (3b/4b) receptor have been
reported,13,14 but mechanisms underlying these
associations remain uncertain.
S t r ategie s a nd E v idence
Impaired memory is typically one of the first signs
of Alzheimers disease, but difficulty recalling the
names of friends or recent events is also common
among normal elderly persons. The clinician is
thus faced with the difficulty of distinguishing
between normal aging and the early stages of
n engl j med 362;23
BACK to TOC
Alzheimers disease. Mild cognitive impairment is
an intermediate state in which persons have more
memory problems than would be considered nor-
mal for their age, but their symptoms are not as
severe as the symptoms of Alzheimer disease and
they do not have functional impairment.5 Alz-
heimers disease develops at a much higher fre-
quency among persons with mild cognitive im-
pairment than among those with normal aging.
Determining when patients have reached the very
early stage of Alzheimers disease is not easy, par-
ticularly because it is likely that a preclinical stage
of Alzheimers disease exists in which senile
plaques, neuritic plaques, and neurofibrillary tan-
gles occur in sufficient numbers to meet standard
neuropathological criteria for Alzheimers disease
in the absence of overt symptoms or signs of de-
mentia.15 Other causes of memory impairment
must also be considered, such as cerebrovascular
disease, hydrocephalus, hypothyroidism, vitamin
B12 deficiency, central nervous system infection,
a cognitive disorder related to human immuno-
deficiency virus infection, adverse effects of pre-
scribed medications, substance abuse, and cancer.
A substantial decline in verbal memory and
executive function (e.g., the ability to perform se-
quential tasks) typically occurs at the onset of Alz-
heimers disease but may be difficult to document
without formal neuropsychological testing (Fig. 1).
Reduced independence in daily activities (often
recognized by the patients family) is one of the
strongest predictors of disease.16 Functional sta-
tus can be measured by the Clinical Dementia
Rating (CDR) scale, which evaluates cognitive and
functional performance on a scale ranging from
0 to 3, with higher scores indicating a greater se-
verity of impairment.17 This assessment requires
a collateral source of information gathering con-
cerning the patients ability to function indepen-
dently but can be performed in the primary care
setting and is particularly useful for clinicians who
do not have ready access to formal neuropsycho-
logical testing. The assessment requires 30 to 45
minutes to administer, and training is provided
online. (Additional details are available in the
Supplementary Appendix, available with the full
text of this article at NEJM.org.) The CDR score
was the strongest predictor of Alzheimers disease
in a study involving community volunteers without
dementia, and scores on a functional rating scale
that is based on the CDR effectively identified
patients in the early stages of Alzheimers disease
in a clinical setting.18 Formal neuropsychological
nejm.org june 10, 2010 2195
36
 The n e w e ng l a n d j o u r na l of m e dic i n e

CDR = 0 CDR = 0.5 CDR = 1.0

Normal Aging Preclinical Alzheimers Disease Early Alzheimers Disease

MIld Cognitive Impairment

CSF tau and

phosphorylated tau
Increases with time

Dependence on
assistance in
daily activities

FDG-PET parietal
metabolism
CSF beta-amyloid
Decreases with time peptide
Increasing
amyloid
l id plaques
l andd Hippocampal size
on MRI
neurofibrillary tangles
Neuropsychological
test performance

Time

Figure 1. Sequence of Pathological, Clinical, Physiological, and Radiologic Changes from Normal Aging to Early
Alzheimers Disease.
Changes from normal aging to preclinical Alzheimers disease to early Alzheimers disease (yellow to green) are
shown. The most frequent pathological feature of Alzheimers disease is the presence of extracellular beta-amyloid
protein in diffuse plaques, along with intracellular changes that include deposits of hyperphosphorylated tau protein
in the form of neurofibrillary tangles. These changes correspond to scores on the Clinical Dementia Rating (CDR)
scale, which ranges from 0 to 3, with 0 indicating no impairment, 0.5 very mild impairment, 1.0 mild impairment,
2.0 moderate impairment, and 3.0 severe impairment. CSF denotes cerebrospinal fluid, FDG-PET 18F-fluorodeoxy-
glucosepositron-emission tomography, and MRI magnetic resonance imaging.

testing that shows a substantial decline in verbal
memory and executive function supports the di-
agnosis of Alzheimers disease18,19 but requires a
trained professional for administration and inter-
pretation.
Occasionally, patients with early Alzheimers
disease present with impaired language or per-
ceptual dysfunction rather than memory loss.20
Over time, both memory impairment and func-
tional decline become apparent in such patients.
Patients with early disease are at increased risk
for motor vehicle accidents. The American Acad-
emy of Neurology21 recommends that clinicians
perform a careful assessment of driving ability,
including asking the caregiver to rate the patients
driving ability and reviewing any traffic citations
and accidents. Cognitive assessments that include
visual perception and sequential-task performance
may also be helpful in assessing the capacity to
drive.22 Many state motor vehicle agencies have
simulated driving laboratories or are willing to
assess driving ability for a nominal fee. Informa-
tion regarding resources for evaluating poten-
tially impaired drivers is available through the
National Highway Traffic Safety Administration
(www.nhtsa.dot.gov).
T r e atmen t Op t ions
Drug Therapies
Cholinesterase inhibitors (donepezil, rivastigmine,
and galantamine) and the N-methyl-d-aspartate
receptor antagonist memantine are the only treat-
ments for Alzheimers disease that have been ap-
proved by the Food and Drug Administration23
(Table 1). Randomized, placebo-controlled clini-
cal trials of cholinesterase inhibitors have includ-
ed patients with mainly mild-to-moderate Alz-

2196 n engl j med 362;23 nejm.org june 10, 2010

BACK to TOC
37
 clinical pr actice

Table 1. Drug Therapy for Alzheimers Disease.

Common Adverse
Medication Dose Side Effects Comments
Donepezil (Aricept) 5 mg/day at bedtime with or without Nausea, vomiting, loss of appetite, Available in a single daily dose
food for 4 to 6 weeks; 10 mg/day weight loss, diarrhea, dizziness,
there-after, if tolerated muscle cramps, insomnia and
vivid dreams

Rivastigmine (Exelon) 3 mg daily, split into morning and
evening doses with meals; dose
increased by 3 mg/day every
4 weeks as tolerated, with a max-
imum daily dose of 12 mg
Nausea, vomiting, loss of appetite, Available as a patch
weight loss, diarrhea, indiges-
tion, dizziness, drowsiness,
headache, diaphoresis,
weakness

Galantamine (Razadyne) 8 mg daily, split into morning and
evening doses with meals; dose
increased by 4 mg every 4 weeks,
as tolerated, with a maximum
daily dose of 16 to 24 mg
Nausea, vomiting, loss of appetite, Available as an extended-
weight loss, diarrhea, dizziness, release capsule
headache, fatigue

Memantine (Namenda) 5 mg/day with or without food; dose
increased by 5 mg every week,
with a maximum daily dose of
20 mg
Constipation, dizziness, headache,
pain (nonspecific)
Often used as an adjunct to
cholinesterase inhibitors;
not recommended alone
for treatment of early
disease

heimers disease and have shown significant but
clinically marginal benefits with respect to cog-
nition, daily function, and behavior.24-26 The con-
dition of patients who are taking these drugs re-
mains stable for a year or more and then may
decline, though at a rate that is slower than that
among untreated patients.
Although there are few studies directly compar-
ing the three cholinesterase inhibitors, a system-
atic review and meta-analysis of data from 27
randomized trials concluded that there were no
significant differences in effects on cognitive per-
formance among these medications.27 During the
study period (usually, 3 to 6 months), the use of
each of these drugs as prescribed at a standard
dose resulted in a mean improvement of 2 to
3 points on the Alzheimers Disease Assessment
Scale for cognition (a scale ranging from 0 to 70,
with higher scores indicating worse cognition)
or a decreased rate of decline, as compared with
the placebo group (approximately a 3-point differ-
ence, with a minimal clinically important differ-
ence of 4 points).
On the basis of 14 studies that measured daily
function, donepezil was modestly but significantly
more effective than rivastigmine. Donepezil was
likewise modestly but significantly better than
rivastigmine and galantamine with regard to be-
havior, as measured by the Neuropsychiatric In-
ventory (on a scale ranging from 1 to 144, with
higher scores indicating a greater severity of dis-
ease). Patients receiving donepezil had a mean
reduction of 4.3 points in the baseline score, as
compared with a reduction of 1.4 for those re-
ceiving the other agents. The likelihood of an
overall improvement in score was 1.9 times as
great with donepezil as with placebo, 1.2 times
as great with rivastigmine as with placebo, and
1.6 times as great with galantamine as with pla-
cebo. Adverse effects (including nausea, vomiting,
diarrhea, dizziness, and weight loss) were frequent
with all three medications, although slightly less
frequent with donepezil than with the other medi-
cations.
Initial randomized trials of memantine in-
volving patients with moderate-to-severe disease
showed a small but significant reduction in cog-
nitive deterioration.28 Subsequent randomized tri-
als involving patients with mild-to-moderate dis-
ease showed that memantine resulted in marginal
benefits over a period of 6 months, with absolute
changes in cognitive and functional measures of
1 percentage point.29 However, studies that were
limited to patients with mild or early-stage dis-
ease have shown no significant benefit of meman-
tine therapy.30 Memantine has also been used in
patients with late-stage disease in combination
with cholinesterase inhibitors, such as donepezil,

n engl j med 362;23 nejm.org june 10, 2010 2197

BACK to TOC
38
 The n e w e ng l a n d j o u r na l of m e dic i n e

with modest improvements (a relative change in stage of Alzheimers disease. In one study, 25% of
score of 2 to 5%) on the Severe Impairment Bat- patients with Alzheimers disease were reported
tery and the activities of daily living inventory of to have received the diagnosis of depression at
the Alzheimers Disease Cooperative Study.31 the time of or just before the onset of symptoms
More data are needed to guide the optimal of the disease.40 In patients in whom pharmaco-
timing of treatment of early Alzheimers disease. therapy is considered appropriate, selective sero-
In a small, randomized, placebo-controlled trial tonin-reuptake inhibitors are commonly used; tri-
of donepezil, patients in whom Alzheimers dis- cyclic antidepressants are generally avoided, since
ease had been diagnosed within the preceding their anticholinergic effects can cause or exacer-
year showed improvement in cognitive perfor- bate confusion.41
mance over a period of 24 weeks.32 In an open- Psychosis that is characterized by hallucina-
label study, patients who were treated early in the tions and delusions may occur infrequently in pa-
disease course had improvement that was only tients with early Alzheimers disease. The occur-
slightly greater than that of patients who began rence of agitation, delusions, hallucinations, and
treatment later.26 In another observational study, irritability early in the disease course also raises
a duration of treatment with cholinesterase inhibi- the possibility of an alternative diagnosis, such as
tors or memantine of at least 3 years was associ- dementia with Lewy bodies. Treatment with con-
ated with a significantly slower rate of decline in ventional or atypical antipsychotic agents may be
cognitive ability and daily function.33 helpful, but such drugs should be used with cau-
In practice, subjective reports of improvement tion because of the potential adverse effects (e.g.,
in patients receiving cholinesterase inhibitors or parkinsonism, extrapyramidal signs, sedation, and
memantine are common, but objective improve- confusion).42
ments are modest, if detectable at all. A rational
approach is to try a cholinesterase inhibitor first, Caregiver Support
switching to another agent in the same class if Persons who live with and provide care for pa-
the initial agent is ineffective or if intolerable side
tients with Alzheimers disease, even in the early
effects emerge.23 Memantine may be added to any phases of the disease, often report emotional
of the cholinesterase inhibitors in patients who stress, in part related to the need to give up vaca-
have little or no improvement with cholinesterase tions, hobbies, or even work to care for the pa-
inhibitor monotherapy. tient. Caregivers should routinely be offered coun-
seling and support. Resources for caregivers and
Other Strategies patients are available through the Alzheimers As-
The use of nonsteroidal antiinflammatory drugs, sociation (www.alz.org).
estrogen therapy, antioxidant vitamins, or statins
has been proposed for the prevention of Alzhei- A r e a s of Uncer ta in t y
mers disease, but the results of randomized trials
have been inconsistent or negative.34-37 Similarly, Further study of brain-imaging methods and bio-
the efficacy of commonly used complementary markers that may facilitate the identification of
therapies (e.g., ginkgo biloba, acetyl-L-carnitine, patients with early Alzheimers disease is needed.
lecithin, huperzine A, piracetam, curcumin, peri- Focal atrophy on magnetic resonance imaging
winkle, and phosphatidylserine) has not been (MRI) of the inferior temporal region, particularly
shown in randomized trials.38 A review of nine the hippocampus, has been shown to predict the
randomized clinical trials of cognitive training conversion from mild cognitive impairment to
and rehabilitation therapies that were used to ad- Alzheimers disease.43 However, there is no stan-
dress loss of memory and other intellectual func- dard technique to quantify atrophy in the clinical
tions showed no significant effects.39 setting, and the diagnostic sensitivity and speci-
ficity of MRI are unclear.
Management of Psychiatric Symptoms Studies have shown that evidence of decreased
Behavioral and psychiatric symptoms typically in- metabolism and perfusion in the parietal lobes
crease with disease progression. However, depres- on 18F-fluorodeoxyglucosepositron-emission to-
sion and anxiety are frequent even in the early mography (FDGPET) is as accurate as evidence

2198 n engl j med 362;23 nejm.org june 10, 2010

BACK to TOC
39
 clinical pr actice

of focal atrophy on MRI in predicting progression
from mild cognitive impairment to Alzheimers
disease.43,44 However, PET scanning is costly and
not widely available at present, and its role in di-
agnosis remains uncertain. PET imaging with the
use of amyloid-binding compounds, such as car-
bon 11labeled Pittsburgh compound B (PIB),45
has been reported to identify patients with early
Alzheimers disease.46 Some normal elderly per-
sons without dementia have PIB retention simi-
lar to that observed in patients with Alzheimers
disease, but progression to Alzheimers disease
occurs more rapidly in persons with mild cogni-
tive impairment who have PIB retention than in
those without retention, indicating that amyloid
deposition may be an early biomarker of incipi-
ent disease.47,48
Measurement of markers in cerebrospinal fluid
has also been proposed to identify early Alzheim-
ers disease. Among persons with mild cognitive
impairment, reduced levels of beta-amyloid pep-
tide and increased levels of total tau and tau
phosphorylated at threonine 181 have predicted
the diagnosis of Alzheimers disease.49,50 Assess-
ment requires lumbar puncture, and the thresh-
old diagnostic levels of these markers have var-
ied across studies.51,52 These measures are now
commercially available with clinical interpreta-
tion, but their role in practice remains unclear.
Guidel ine s
The European Federation of Neurological Socie-
ties has published recommendations for the di-
agnosis and management of Alzheimers disease.53
On the basis of available randomized trials, treat-
ment with cholinesterase inhibitors is recom-
mended even for mild or early disease; no spe-
cific cholinesterase inhibitor is recommended over
another. The American Academy of Neurology pub-
lished practice recommendations in 200154 that
have not yet been updated. In 2006, the American
Association for Geriatric Psychiatry published
practice recommendations that also emphasize
treatment with approved medications for cogni-
tive symptoms, as well as symptomatic treatment
for neuropsychiatric manifestations, such as de-
pression and psychosis, and attention to issues
related to safety, such as driving, living alone, and
medication administration.55
C onclusions a nd
R ec om mendat ions
The 72-year-old patient who is described in the
vignette has a history of memory and functional
impairment, with a relatively high MiniMental
State Examination1 score and a normal neurologic
examination. Basic blood chemical analysis and
measures of thyrotropin should be performed,
along with additional laboratory studies as deemed
clinically relevant. Brain MRI to rule out other
brain diseases and assess atrophy and a detailed
neuropsychological assessment are warranted to
make a preliminary diagnosis. If the diagnosis of
Alzheimers disease is established, I would discuss
with the patient and caregiver potential safety is-
sues, including the current living situation and
driving, and I would initiate treatment with one
of the cholinesterase inhibitors, probably donepe-
zil (starting at 5 mg each night at bedtime). I would
plan a follow-up visit in 4 to 6 weeks to assess the
side effects and efficacy of the medication (both
subjective and objective) by repeating the Mini
Mental State Examination. At that time, the dose
of the cholinesterase inhibitor could be increased
to 10 mg daily if the drug has been well toler-
ated. The patient should be closely followed clin-
ically, with repeated neuropsychological assess-
ment within 2 years.
Dr. Mayeux reports receiving an honorarium from Quintiles
for serving on a data and safety monitoring board for a trial of a
product manufactured by Eli Lilly. No other potential conflict of
interest relevant to this article was reported.
Disclosure forms provided by the author are available with the
full text of this article at NEJM.org.

References
1. Folstein MF, Folstein SE, McHugh PR.
Mini-Mental State: a practical method
for grading the cognitive state of patients
for the clinician. J Psychiatr Res 1975;
12:189-98.
2. Ferri CP, Prince M, Brayne C, et al.
Global prevalence of dementia: a Delphi
consensus study. Lancet 2005;366:2112-7.
3. Mayeux R. Epidemiology of neurode-
generation. Annu Rev Neurosci 2003;26:
81-104.
4. Ziegler-Graham K, Brookmeyer R,
Johnson E, Arrighi HM. Worldwide varia-
tion in the doubling time of Alzheimers
disease incidence rates. Alzheimers De-
ment 2008;4:316-23.
5. Petersen RC, Smith GE, Waring SC,
Ivnik RJ, Tangalos EG, Kokmen E. Mild
cognitive impairment: clinical character-
ization and outcome. Arch Neurol 1999;
56:303-8. [Erratum, Arch Neurol 1999;56:
760.]
6. Ownby RL, Crocco E, Acevedo A, John
V, Loewenstein D. Depression and risk for

n engl j med 362;23 nejm.org june 10, 2010 2199

BACK to TOC
40
 The n e w e ng l a n d j o u r na l
Alzheimer disease: systematic review, meta-
analysis, and metaregression analysis. Arch
Gen Psychiatry 2006;63:530-8.
7. Lopez OL, Becker JT, Sweet RA, et al.
Psychiatric symptoms vary with the sever-
ity of dementia in probable Alzheimers
disease. J Neuropsychiatry Clin Neurosci
2003;15:346-53.
8. Duyckaerts C, Delatour B, Potier MC.
Classification and basic pathology of Alz-
heimer disease. Acta Neuropathol 2009;
118:5-36.
9. Rogaeva E, Kawarai T, George-Hyslop
PS. Genetic complexity of Alzheimers
disease: successes and challenges. J Alz-
heimers Dis 2006;9:Suppl:381-7.
10. Gatz M, Reynolds CA, Fratiglioni L, et
al. Role of genes and environments for
explaining Alzheimer disease. Arch Gen
Psychiatry 2006;63:168-74.
11. Mucke L. Neuroscience: Alzheimers
disease. Nature 2009;461:895-7.
12. Rogaeva E, Meng Y, Lee JH, et al. The
neuronal sortilin-related receptor SORL1
is genetically associated with Alzheimer
disease. Nat Genet 2007;39:168-77.
13. Harold D, Abraham R, Hollingworth
P, et al. Genome-wide association study
identifies variants at CLU and PICALM as-
sociated with Alzheimers disease. Nat
Genet 2009;41:1088-93.
14. Lambert JC, Heath S, Even G, et al.
Genome-wide association study identifies
variants at CLU and CR1 associated with
Alzheimers disease. Nat Genet 2009;41:
1094-9.
15. Price JL, McKeel DW Jr, Buckles VD, et
al. Neuropathology of nondemented ag-
ing: presumptive evidence for preclinical
Alzheimer disease. Neurobiol Aging 2009;
30:1026-36.
16. Hsiung GY, Alipour S, Jacova C, et al.
Transition from cognitively impaired not
demented to Alzheimers disease: an analy-
sis of changes in functional abilities in a
dementia clinic cohort. Dement Geriatr
Cogn Disord 2008;25:483-90.
17. Morris JC. Clinical dementia rating:
a reliable and valid diagnostic and staging
measure for dementia of the Alzheimer
type. Int Psychogeriatr 1997;9:Suppl 1:
173-8.
18. Dickerson BC, Sperling RA, Hyman
BT, Albert MS, Blacker D. Clinical predic-
tion of Alzheimer disease dementia across
the spectrum of mild cognitive impairment.
Arch Gen Psychiatry 2007;64:1443-50.
19. Aggarwal NT, Wilson RS, Beck TL,
Bienias JL, Bennett DA. Mild cognitive im-
pairment in different functional domains
and incident Alzheimers disease. J Neu-
rol Neurosurg Psychiatry 2005;76:1479-84.
20. Alladi S, Xuereb J, Bak T, et al. Focal
cortical presentations of Alzheimers dis-
ease. Brain 2007;130:2636-45.
21. Iverson DJ, Gronseth GS, Reger MA,
Classen S, Dubinsky RM, Rizzo M. Prac-
2200 n engl j med 362;23
BACK to TOC
of m e dic i n e
tice parameter update: evaluation and
management of driving risk in dementia:
report of the Quality Standards Subcom-
mittee of the American Academy of Neu-
rology. Neurology 2010;74:1316-24.
22. Dawson JD, Anderson SW, Uc EY,
Dastrup E, Rizzo M. Predictors of driving
safety in early Alzheimer disease. Neurol-
ogy 2009;72:521-7.
23. Farlow MR, Cummings JL. Effective
pharmacologic management of Alzhei-
mers disease. Am J Med 2007;120:388-97.
24. Rsler M, Anand R, Cicin-Sain A, et
al. Efficacy and safety of rivastigmine in
patients with Alzheimers disease: inter-
national randomised controlled trial. BMJ
1999;318:633-8.
25. Wilcock GK, Lilienfeld S, Gaens E. Ef-
ficacy and safety of galantamine in pa-
tients with mild to moderate Alzheimers
disease: multicentre randomised controlled
trial. BMJ 2000;321:1445-9.
26. Winblad B, Wimo A, Engedal K, et al.
3-Year study of donepezil therapy in Alz-
heimers disease: effects of early and con-
tinuous therapy. Dement Geriatr Cogn
Disord 2006;21:353-63.
27. Hansen RA, Gartlehner G, Webb AP,
Morgan LC, Moore CG, Jonas DE. Efficacy
and safety of donepezil, galantamine, and
rivastigmine for the treatment of Alz-
heimers disease: a systematic review and
meta-analysis. Clin Interv Aging 2008;3:
211-25.
28. Reisberg B, Doody R, Stffler A,
Schmitt F, Ferris S, Mbius HJ. Meman-
tine in moderate-to-severe Alzheimers
disease. N Engl J Med 2003;348:1333-41.
29. McShane R, Areosa Sastre A, Mina-
karan N. Memantine for dementia. Co-
chrane Database Syst Rev 2006;2:
CD003154.
30. Bakchine S, Loft H. Memantine treat-
ment in patients with mild to moderate
Alzheimers disease: results of a ran-
domised, double-blind, placebo-controlled
6-month study. J Alzheimers Dis 2008;13:
97-107.
31. Tariot PN, Farlow MR, Grossberg GT,
Graham SM, McDonald S, Gergel I. Me-
mantine treatment in patients with mod-
erate to severe Alzheimer disease already
receiving donepezil: a randomized con-
trolled trial. JAMA 2004;291:317-24.
32. Seltzer B, Zolnouni P, Nunez M, et al.
Efficacy of donepezil in early-stage Alz-
heimer disease: a randomized placebo-con-
trolled trial. Arch Neurol 2004;61:1852-6.
33. Rountree SD, Chan W, Pavlik VN,
Darby EJ, Siddiqui S, Doody RS. Persistent
treatment with cholinesterase inhibitors
and/or memantine slows clinical progres-
sion of Alzheimer disease. Alzheimers Res
Ther 2009;1:7.
34. Aisen PS, Thal LJ, Ferris SH, et al. Ro-
fecoxib in patients with mild cognitive
impairment: further analyses of data
nejm.org june 10, 2010
from a randomized, double-blind, trial.
Curr Alzheimer Res 2008;5:73-82.
35. Feldman HH, Doody RS, Kivipelto M,
et al. Randomized controlled trial of ator-
vastatin in mild to moderate Alzheimer
disease: LEADe. Neurology 2010;74:956-64.
36. Hogervorst E, Yaffe K, Richards M,
Huppert FA. Hormone replacement therapy
to maintain cognitive function in women
with dementia. Cochrane Database Syst
Rev 2009;1:CD003799.
37. Isaac MG, Quinn R, Tabet N. Vitamin
E for Alzheimers disease and mild cogni-
tive impairment. Cochrane Database Syst
Rev 2008;3:CD002854.
38. Kelley BJ, Knopman DS. Alternative
medicine and Alzheimer disease. Neurol-
ogist 2008;14:299-306.
39. Clare L, Woods RT, Moniz Cook ED,
Orrell M, Spector A. Cognitive rehabilita-
tion and cognitive training for early-stage
Alzheimers disease and vascular demen-
tia. Cochrane Database Syst Rev 2003;4:
CD003260.
40. Panza F, Frisardi V, Capurso C, et al.
Late-life depression, mild cognitive im-
pairment, and dementia: possible contin-
uum? Am J Geriatr Psychiatry 2010;18:98-
116.
41. Starkstein SE, Mizrahi R, Power BD.
Depression in Alzheimers disease: phe-
nomenology, clinical correlates and treat-
ment. Int Rev Psychiatry 2008;20:382-8.
42. Schneider LS, Tariot PN, Dagerman
KS, et al. Effectiveness of atypical antipsy-
chotic drugs in patients with Alzheimers
disease. N Engl J Med 2006;355:1525-38.
43. Schroeter ML, Stein T, Maslowski N,
Neumann J. Neural correlates of Alz-
heimers disease and mild cognitive im-
pairment: a systematic and quantitative
meta-analysis involving 1351 patients.
Neuroimage 2009;47:1196-206.
44. Yuan Y, Gu ZX, Wei WS. Fluorodeoxy-
glucose-positron-emission tomography,
single-photon emission tomography, and
structural MR imaging for prediction of
rapid conversion to Alzheimer disease in
patients with mild cognitive impairment:
a meta-analysis. AJNR Am J Neuroradiol
2009;30:404-10.
45. Klunk WE, Engler H, Nordberg A, et
al. Imaging brain amyloid in Alzheimers
disease with Pittsburgh Compound-B. Ann
Neurol 2004;55:306-19.
46. Perrin RJ, Fagan AM, Holtzman DM.
Multimodal techniques for diagnosis and
prognosis of Alzheimers disease. Nature
2009;461:916-22.
47. Fripp J, Bourgeat P, Acosta O, et al.
Appearance modeling of 11C PiB PET im-
ages: characterizing amyloid deposition in
Alzheimers disease, mild cognitive im-
pairment and healthy aging. Neuroimage
2008;43:430-9.
48. Okello A, Koivunen J, Edison P, et al.
Conversion of amyloid positive and nega-
41
 clinical pr actice

tive MCI to AD over 3 years: an 11C-PIB
PET study. Neurology 2009;73:754-60.
49. Hansson O, Zetterberg H, Buchhave P,
Londos E, Blennow K, Minthon L. Asso-
ciation between CSF biomarkers and in-
cipient Alzheimers disease in patients with
mild cognitive impairment: a follow-up
study. Lancet Neurol 2006;5:228-34.
50. Mattsson N, Zetterberg H, Hansson O,
et al. CSF biomarkers and incipient Alz-
heimer disease in patients with mild cog-
nitive impairment. JAMA 2009;302:385-93.
51. Smach MA, Charfeddine B, Ben Oth-
man L, et al. Evaluation of cerebrospinal
fluid tau/beta-amyloid(42) ratio as diag-
nostic markers for Alzheimer disease. Eur
Neurol 2009;62:349-55.
52. Vemuri P, Wiste HJ, Weigand SD, et al.
MRI and CSF biomarkers in normal, MCI,
and AD subjects: diagnostic discrimina-
tion and cognitive correlations. Neurolo-
gy 2009;73:287-93.
53. Waldemar G, Dubois B, Emre M, et al.
Recommendations for the diagnosis and
management of Alzheimers disease and
other disorders associated with dementia:
EFNS guideline. Eur J Neurol 2007;14(1):
e1-e26.
54. Knopman DS, DeKosky ST, Cummings
JL, et al. Practice parameter: diagnosis of
dementia (an evidence-based review): re-
port of the Quality Standards Subcommit-
tee of the American Academy of Neurology.
Neurology 2001;56:1143-53.
55. Lyketsos CG, Colenda CC, Beck C, et
al. Position statement of the American As-
sociation for Geriatric Psychiatry regard-
ing principles of care for patients with de-
mentia resulting from Alzheimer disease.
Am J Geriatr Psychiatry 2006;14:561-72.
Copyright 2010 Massachusetts Medical Society.

POSTING PRESENTATIONS AT MEDICAL MEETINGS ON THE INTERNET

Posting an audio recording of an oral presentation at a medical meeting on the
Internet, with selected slides from the presentation, will not be considered prior
publication. This will allow students and physicians who are unable to attend the
meeting to hear the presentation and view the slides. If there are any questions
about this policy, authors should feel free to call the Journals Editorial Offices.

n engl j med 362;23 nejm.org june 10, 2010 2201

BACK to TOC
42
 The n e w e ng l a n d j o u r na l of m e dic i n e

clinical practice

Helicobacter pylori Infection

Kenneth E.L. McColl, M.D.

This Journal feature begins with a case vignette highlighting a common clinical problem.
Evidence supporting various strategies is then presented, followed by a review of formal guidelines,
when they exist. The article ends with the authors clinical recommendations.

A 29-year-old man presents with intermittent epigastric discomfort, without weight

loss or evidence of gastrointestinal bleeding. He reports no use of aspirin or non-
steroidal antiinflammatory drugs (NSAIDs). Abdominal examination reveals epi-
gastric tenderness. A serologic test for Helicobacter pylori is positive, and he receives a
10-day course of triple therapy (omeprazole, amoxicillin, and clarithromycin). Six
weeks later, he returns with the same symptoms. How should his case be further
evaluated and managed?

The Cl inic a l Probl em

Helicobacter pylori, a gram-negative bacterium found on the luminal surface of the From the Division of Cardiovascular and
gastric epithelium, was first isolated by Warren and Marshall in 19831 (Fig. 1). It Medical Sciences, University of Glasgow,
Gardiner Institute, Glasgow, United King-
induces chronic inflammation of the underlying mucosa (Fig. 2). The infection is dom. Address reprint requests to Dr. Mc-
usually contracted in the first few years of life and tends to persist indefinitely un- Coll at the Division of Cardiovascular and
less treated.2 Its prevalence increases with older age and with lower socioeconomic Medical Sciences, University of Glasgow,
Gardiner Institute, 44 Church St., Glas-
status during childhood and thus varies markedly around the world.3 The higher gow G11 6NT, United Kingdom, or at
prevalence in older age groups is thought to reflect a cohort effect related to poorer k.e.l.mccoll@clinmed.gla.ac.uk.
living conditions of children in previous decades. At least 50% of the worlds human
This article (10.1056/NEJMcp1001110) was
population has H. pylori infection.2 The organism can survive in the acidic environ- updated on August 4, 2010, at NEJM.org.
ment of the stomach partly owing to its remarkably high urease activity; urease
converts the urea present in gastric juice to alkaline ammonia and carbon dioxide.4 N Engl J Med 2010;362:1597-604.
Copyright 2010 Massachusetts Medical Society.
Infection with H. pylori is a cofactor in the development of three important upper
gastrointestinal diseases: duodenal or gastric ulcers (reported to develop in 1 to 10%
of infected patients), gastric cancer (in 0.1 to 3%), and gastric mucosa-associated
lymphoid-tissue (MALT) lymphoma (in <0.01%). The risk of these disease outcomes
in infected patients varies widely among populations. The great majority of patients
An audio version Click here to
with H. pylori infection will not have any clinically significant complications. of this article access audio
is available at version.
Gastric and Duodenal Ulcers NEJM.org
In patients with duodenal ulcers, the inflammation of the gastric mucosa induced by
the infection is most pronounced in the nonacid-secreting antral region of the
stomach and stimulates the increased release of gastrin.5 The increased gastrin
levels in turn stimulate excess acid secretion from the more proximal acid-secreting
fundic mucosa, which is relatively free of inflammation.5,6 The increased duodenal
acid load damages the duodenal mucosa, causing ulceration and gastric metaplasia.
The metaplastic mucosa can then become colonized by H. pylori, which may contrib-
ute to the ulcerative process. Eradication of the infection provides a long-term cure
of duodenal ulcers in more than 80% of patients whose ulcers are not associated
with the use of NSAIDs.7 NSAIDs are the main cause of H. pylorinegative ulcers.

n engl j med 362;17 nejm.org april 29, 2010 1597

BACK to TOC
43
 The n e w e ng l a n d j o u r na l of m e dic i n e

Figure 1. Helicobacter pylori.

H. pylori is a gram-negative bacterium with a helical Figure 2. Gastric-Biopsy Specimen Showing Helico-
rod shape. It has prominent flagellae, facilitating its bacter pylori Adhering to Gastric Epithelium and
penetration of the thick mucous layer in the stomach. Underlying Inflammation.
H. pylori is visible as small black rods (arrows) on the
epithelial surface and within the glands. The underly-
Ulceration of the gastric mucosa is believed to ing mucosa shows inflammatory-cell infiltrates.
be due to the damage to the mucosa caused by
H. pylori. As with duodenal ulcers, eradicating the
infection usually cures the disease, provided that specimens of the gastric mucosa often reveal the
the gastric ulcer is not due to NSAIDs.8 presence of H. pylori and associated inflammation,
although this finding is also common in persons
Gastric Cancer without upper gastrointestinal symptoms. Most
Extensive epidemiologic data suggest strong asso- randomized trials of therapy for H. pylori eradica-
ciations between H. pylori infection and noncar- tion in patients with nonulcer dyspepsia have
dia gastric cancers (i.e., those distal to the gastro- shown no significant benefit regarding symp-
esophageal junction).9 The infection is classified toms; a few have shown a marginal benefit,16,17
as a human carcinogen by the World Health Or- but this can be explained by the presence of un-
ganization.10 The risk of cancer is highest among recognized ulceration.18 There is thus little evi-
patients in whom the infection induces inflam- dence that chronic H. pylori infection in the ab-
mation of both the antral and fundic mucosa and sence of gastric or duodenal ulceration causes
causes mucosal atrophy and intestinal metapla- upper gastrointestinal symptoms.
sia.11 Eradication of H. pylori infection reduces the The prevalence of H. pylori infection is lower
progression of atrophic gastritis, but there is little among patients with gastroesophageal reflux dis-
evidence of reversal of atrophy or intestinal meta- ease (GERD)19 and those with esophageal adeno-
plasia,12 and it remains unclear whether eradica- carcinoma (which may arise as a complication of
tion reduces the risk of gastric cancer.13 GERD) than among healthy controls.20 H. pylori
associated atrophic gastritis, which reduces acid
Gastric MALT Lymphoma secretion, may provide protection against these
Epidemiologic studies have also shown strong as- diseases. A recent meta-analysis showed no sig-
sociations between H. pylori infection and the pres- nificant association between H. pylori eradication
ence of gastric MALT lymphomas.14 Furthermore, and an increased risk of GERD.21
eradication of the infection causes regression of
most localized gastric MALT lymphomas.15 S t r ategie s a nd E v idence
Other Gastrointestinal Conditions Candidates for Testing for H. pylori
At least 50% of persons who undergo endoscopy Infection
for upper gastrointestinal symptoms have no evi- Since the vast majority of patients with H. pylori
dence of esophagitis or gastric or duodenal ulcer- infection do not have any related clinical disease,
ation and are considered to have nonulcer or routine testing is not considered appropriate.22,23
functional dyspepsia. In such patients, biopsy Definite indications for identifying and treating

1598 n engl j med 362;17 nejm.org april 29, 2010

BACK to TOC
44
 clinical pr actice
the infection are confirmed gastric or duodenal
ulcers and gastric MALT lymphoma.22,23 Testing
for infection, and subsequent eradication, also
seems prudent after resection of early gastric can-
cers.24 In addition, European guidelines recom-
mend eradicating H. pylori infection in first-degree
relatives of patients with gastric cancer and in
patients with atrophic gastritis, unexplained iron-
deficiency anemia, or chronic idiopathic thrombo-
cytopenic purpura, although the data in support
of these recommendations are scant.23
Patients with uninvestigated, uncomplicated
dyspepsia may also undergo testing for H. pylori
infection by means of a nonendoscopic (noninva-
sive) method22,23,25; eradication therapy is pre-
scribed for patients with positive test results.
The rationale for this strategy is that in some pa-
tients with dyspepsia, underlying H. pyloriinduced
ulcer disease is causing their symptoms. This
nonendoscopic strategy is not appropriate for
patients with accompanying alarm symptoms
(e.g., weight loss, persistent vomiting, or gastro-
intestinal bleeding) or for older patients (45 or
55 years of age, depending on the specific set
of guidelines) with new-onset dyspepsia, in whom
endoscopy is warranted.22,23,25 The nonendoscopic
strategy is also not generally recommended for
patients with NSAID-associated dyspepsia, since
NSAIDs can cause ulcers in the absence of H. py-
lori infection.
An attraction of the test-and-treat strategy is
that it avoids the discomfort and costs of endos-
copy. However, because only a minority of pa-
tients with dyspepsia who have a positive H. pylori
test have underlying ulcer disease,26,27 most pa-
tients treated by means of the test-and-treat strat-
egy incur the inconvenience, costs, and potential
side effects of therapy without a benefit. In a
placebo-controlled trial of empirical treatment
involving 294 patients with uninvestigated dys-
pepsia and a positive H. pylori breath test, the
1-year rate of symptom resolution was 50% in
those receiving H. pylorieradication therapy, as
compared with 36% of those receiving placebo
(P = 0.02)28; 7 patients would need to receive
eradication therapy for 1 patient to have a benefit.
A greater benefit would be expected if treatment
were limited to patients with an increased prob-
ability of having an ulcer. However, neither the
characteristics of the symptoms nor the presence
of other risk factors for ulcer (e.g., male sex,
smoking, and family history of ulcer disease) are
n engl j med 362;17
BACK to TOC
particularly useful in clinical practice for identi-
fying patients with ulcer dyspepsia and those
with nonulcer dyspepsia.29
In randomized trials comparing a noninvasive
test-and-treat strategy with early endoscopy26,27
or with proton-pumpinhibitor therapy,30,31 the
three strategies resulted in a similar degree of
symptom improvement, but early endoscopy was
more expensive than the other two strategies.32
However, the test-and-treat strategy is unlikely
to be cost-effective in populations with a preva-
lence of H. pylori infection below 20%.33 Infor-
mation is lacking on the longer-term outcomes
of these strategies.
Tests for H. pylori Infection
Table 1 summarizes the various tests for H. pylori
infection.
Nonendoscopic Tests
Serologic testing for IgG antibodies to H. pylori is
often used to detect infection. However, a meta-
analysis of studies of several commercially avail-
able quantitative serologic assays showed an
overall sensitivity and specificity of only 85% and
79%, respectively.34 The appropriate cutoff values
vary among populations, and the test results are
often reported as positive, negative, or equivocal.
Also, this test has little value in confirming erad-
ication of the infection, because the antibodies
persist for many months, if not longer, after
eradication.
The urea breath test involves drinking 13C-
labeled or 14C-labeled urea, which is converted
to labeled carbon dioxide by the urease in H. py-
lori. The labeled gas is measured in a breath
sample. The test has a sensitivity and a specific-
ity of 95%.35 The infection can also be detected
by identifying H. pylorispecific antigens in a
stool sample with the use of polyclonal or mono-
clonal antibodies (the fecal antigen test).36 The
monoclonal-antibody test (which also has a
specificity and a sensitivity of 95%36) is more
accurate than the polyclonal-antibody test. For
both the breath test and the fecal antigen test,
the patient should stop taking proton-pump in-
hibitors 2 weeks before testing, should stop tak-
ing H2 receptor antagonists for 24 hours before
testing, and should avoid taking antimicrobial
agents for 4 weeks before testing, since these
medications may suppress the infection and re-
duce the sensitivity of testing.
nejm.org april 29, 2010 1599
45
 The n e w e ng l a n d j o u r na l
Table 1. Tests for Helicobacter pylori Infection.*
Test Advantages
Nonendoscopic
Serologic test Widely available; the least expensive of available
tests
Urea breath test High negative and positive predictive values;
useful before and after treatment
Fecal antigen test High negative and positive predictive values with
monoclonal-antibody test; useful before and
after treatment
Endoscopic
Urease-based tests Rapid, inexpensive, and accurate in selected
patients
Histologic assessment Good sensitivity and specificity
Culture Excellent specificity; provides opportunity to test
for antibiotic sensitivity
* PPI denotes proton-pump inhibitor.
Endoscopic Tests
H. pylori infection can be detected on endoscopic
biopsy of the gastric mucosa, by means of several
techniques. The biopsy specimens are usually
taken from the prepyloric region, but an addi-
tional biopsy specimen obtained from the fundic
mucosa may increase the tests sensitivity, espe-
cially if the patient has recently been treated with
a proton-pump inhibitor.
The urease-based method involves placement
of the endoscopic biopsy specimen in a solution
of urea and pH-sensitive dye. If H. pylori is pres-
ent, its urease converts the urea to ammonia, in-
creasing the pH and changing the color of the
dye. Recommendations for avoiding proton-pump
inhibitors, H2 receptor antagonists, and anti-
microbial therapy before testing apply to this test
as well, to minimize the chance of false negative
results.37 The test has a sensitivity of more than
90% and a specificity of more than 95%.35
Another means of diagnosis involves routine
histologic testing of a biopsy specimen; if there
is H. pylori infection, the organism and associ-
ated gastritis are apparent on sections stained
with hematoxylin and eosin or Giemsa. Although
culturing of the organism is also possible and
permits testing for sensitivity to antimicrobial
agents, facilities for the culture of H. pylori are
not widely available and the method is relatively
insensitive.
1600 n engl j med 362;17
BACK to TOC
of m e dic i n e
Disadvantages
Positive result may reflect previous rather than current in-
fection; not recommended for confirming eradication
False negative results possible in the presence of PPIs or with
recent use of antibiotics or bismuth preparations; consid-
erable resources and personnel required to perform test
Process of stool collection may be distasteful to patient;
false negative results possible in the presence of PPIs or
with recent use of antibiotics or bismuth preparations
False negative results possible in the presence of PPIs or
with recent use of antibiotics or bismuth preparations
Requires trained personnel
Variable sensitivity; requires trained staff and properly
equipped facilities
Treatment of H. pylori Infection
Various drug regimens are used to treat H. pylori
infection (Table 2). Most include two antibiotics
plus a proton-pump inhibitor or a bismuth prep-
aration (or both). The most commonly used ini-
tial treatment is triple therapy consisting of a
proton-pump inhibitor plus clarithromycin and
amoxicillin, each given twice per day for 7 to 14
days. Metronidazole is used in place of amoxicil-
lin in patients with a penicillin allergy.
The recommended duration of triple therapy
is typically 10 to 14 days in the United States and
7 days in Europe.22,23 A recent meta-analysis of
21 randomized trials showed that the rate of
eradication was increased by 4 percentage points
with the use of triple therapy for 10 days as com-
pared with 7 days and by 5 percentage points
with the use of triple therapy for 14 days as com-
pared with 7 days38 absolute differences that
are statistically significant but of marginal clin-
ical significance.
Another possible initial therapy in areas with
a high prevalence of clarithromycin-resistant
H. pylori infection (i.e., >20%) is quadruple ther-
apy comprising the use of a proton-pump inhibi-
tor, tetracycline, metronidazole, and a bismuth salt
for 10 to 14 days23; however, bismuth salts are
not available in some countries. A recent meta-
analysis of 93 studies showed a higher rate of
eradication with quadruple therapy that included
nejm.org april 29, 2010
46
 clinical pr actice
both clarithromycin and metronidazole than
with triple therapy that included both these
agents in populations with either clarithromy-
cin or metronidazole resistance.39
An alternative initial regimen is 10-day se-
quential therapy, involving a proton-pump inhibi-
tor plus amoxicillin for 5 days followed by a
proton-pump inhibitor plus clarithromycin and
tinidazole for 5 more days. This regimen was
reported to achieve an eradication rate of 93%,
as compared with a rate of 77% with standard
triple therapy, in a meta-analysis of 10 random-
ized trials in Italy.40 However, in a trial in Spain,
the eradication rate among patients randomly
assigned to receive sequential therapy was only
84%, indicating a need to confirm its efficacy
before it is used widely.41
Confirmation of Eradication
It is important to confirm the eradication of
H. pylori infection in patients who have had an
H. pyloriassociated ulcer or gastric MALT lym-
phoma or who have undergone resection for early
gastric cancer.22,23 In addition, to avoid repeated
treatment of patients whose symptoms are not
attributable to H. pylori, follow-up testing is indi-
cated in patients whose symptoms persist after
H. pylori eradication treatment for dyspepsia. Eradi-
cation may be confirmed by means of a urea
breath test or fecal antigen test; these are per-
formed 4 weeks or longer after completion of ther-
apy, to avoid false negative results due to suppres-
sion of H. pylori.22 Eradication can also be confirmed
by testing during repeat endoscopy (Table 1) for
patients in whom endoscopy is required.
Management of Persistent Infection
after Treatment
Before prescribing a second course of therapy, it is
important to confirm that the infection is still pres-
ent and consider whether additional antimicrobial
treatment is appropriate. Further attempts at erad-
ication are indicated in patients with confirmed
ulcer or gastric MALT lymphoma or after resec-
tion for early gastric cancer. However, if the ini-
tial therapy was for uninvestigated dyspepsia,
which is associated with a low likelihood of un-
derlying ulcer and symptomatic benefit from erad-
ication, the appropriateness of further eradication
therapy is unclear; data from studies designed to
determine the optimal management of such cases
are lacking. Options for treatment include em-
n engl j med 362;17
BACK to TOC
Table 2. Regimens Used to Treat Helicobacter pylori Infection.
Standard initial treatment (use one of the following three options)
Triple therapy for 714 days
PPI, healing dose twice/day*
Amoxicillin, 1 g twice/day
Clarithromycin, 500 mg twice/day
Quadruple therapy for 1014 days
PPI, healing dose twice/day*
Tripotassium dicitratobismuthate, 120 mg four times/day
Tetracycline, 500 mg four times/day
Metronidazole, 250 mg four times/day
Sequential therapy
Days 15
PPI, healing dose twice/day*
Amoxicillin, 1 g twice/day
Days 610
PPI, healing dose twice/day*
Clarithromycin, 500 mg twice/day
Tinidazole, 500 mg twice/day
Second-line therapy, if triple therapy involving clarithromycin was used initially
(use one or the other)
Triple therapy for 714 days
PPI, healing dose once/day*
Amoxicillin, 1 g twice/day
Metronidazole, 500 mg (or 400 mg) twice/day
Quadruple therapy, as recommended for initial therapy
* Examples of healing doses of proton-pump inhibitors (PPIs) include the follow-
ing regimens, all twice per day: omeprazole at a dose of 20 mg, esomeprazole
at a dose of 20 mg, rabeprazole at a dose of 20 mg, pantoprazole at a dose of
40 mg, and lansoprazole at a dose of 30 mg. In some studies, esomeprazole
has been given at a dose of 40 mg once per day.
If the patient has an allergy to amoxicillin, substitute metronidazole (at a dose
of 500 mg or 400 mg) twice per day and (in initial triple therapy only) use
clarithromycin at reduced dose of 250 mg twice per day.
Quadruple therapy is appropriate as first-line treatment in areas in which the
prevalence of resistance to clarithromycin or metronidazole is high (>20%) or
in patients with recent or repeated exposure to clarithromycin or metronidazole.
Alcohol should be avoided during treatment with metronidazole or tinidazole,
owing to the potential for a reaction resembling the reaction to disulfiram with
alcohol use.
pirical acid-inhibitory therapy, endoscopy to check
for underlying ulcer or another cause of symp-
toms, and repeat use of the noninvasive test-and-
treat strategy. The possibility that symptoms may
be due to a different cause (e.g., biliary tract,
pancreatic, musculoskeletal, or cardiac disease or
psychosocial stress) should routinely be consid-
ered. If another course of therapy is administered
to eradicate H. pylori infection, the importance of
adherence to the treatment regimen should be
nejm.org april 29, 2010 1601
47
 The n e w e ng l a n d j o u r na l of m e dic i n e

for culturing H. pylori and performing sensitivity

Table 3. Guidelines for Evaluation and Management of Helicobacter pylori
Infection.* testing and experience with alternative treatments
for the infection. Several regimens have been re-
American College Maastricht III ported to be effective as salvage therapy in case
of Gastroenterology Consensus Report
series. For example, retreatment after treatment
Criteria for testing
failure with a triple regimen consisting of levo-
Active gastric or duodenal ulcer, his- Same as American College of Gastro- floxacin or rifabutin, along with a proton-pump
tory of active gastric or duodenal enterology criteria, with the follow-
ulcer not previously treated for ing additional criteria: gastric can- inhibitor and amoxicillin, has been associated
H. pylori infection, gastric MALT cer in first-degree relative, atrophic with high rates of eradication.45-47 However, cau-
lymphoma, history of endoscop- gastritis, unexplained iron-deficien- tion is warranted in the use of rifabutin, which
ic resection of early gastric can- cy anemia, or chronic idiopathic
cer, or uninvestigated dyspepsia thrombocytopenic purpura may lead to resistance of mycobacteria in pa-
Criteria for test-and-treat strategy tients with preexisting mycobacterial infection.
Age <55 yr and no alarm symptoms Age <45 yr and no alarm symptoms
Duration of therapy A r e a s of Uncer ta in t y
1014 Days 7 Days
Data from randomized trials are lacking to guide
* The American College of Gastroenterology guidelines are reported by Chey, the care of patients whose symptoms persist after
Wong, and the Practice Parameters Committee of the American College of completion of H. pylori eradication therapy for
Gastroenterology22; the Maastricht III consensus report guidelines are reported
by Malfertheiner and colleagues.23 MALT denotes mucosa-associated lymphoid
uninvestigated dyspepsia. The effect of eradica-
tissue. tion of H. pylori infection on the risk of gastric
Eradication of H. pylori in patients with chronic idiopathic thrombocytopenic cancer is unclear but is currently under study.
purpura has been reported to increase the platelet count, although the data
are limited.
The age cutoff varies among countries, depending on the prevalence of upper Guidel ine s
gastrointestinal cancer.
Alarm symptoms include dysphagia, weight loss, evidence of gastrointestinal
bleeding, and persistent vomiting.
The American College of Gastroenterology guide-
lines22 and the Maastricht guidelines23 differ
slightly in their recommendations for testing and
emphasized, since poor adherence may underlie treatment of H. pylori infection (Table 3).
the failure of initial therapy.
The choice of second-line treatment is influ-
C onclusions a nd
enced by the initial treatment (Table 2). Treatment R ec om mendat ions
failure is often related to H. pylori resistance to
clarithromycin or metronidazole (or both agents). The noninvasive test-and-treat strategy for H. py-
If initial therapy did not include a bismuth salt, lori infection is reasonable for younger patients
bismuth-based quadruple therapy is commonly who have upper gastrointestinal symptoms but
used as second-line therapy, with eradication not alarm symptoms, like the patient in the vi-
rates in case series ranging from 57 to 95%.42 gnette. Noninvasive testing can be performed
Triple therapies have also been tested as second- with the use of the urea breath test, fecal antigen
line therapies in patients in whom initial therapy test, or serologic test; the serologic test is the
failed. A proton-pump inhibitor used in combina- least accurate. Triple therapy with a proton-pump
tion with metronidazole and either amoxicillin inhibitor, clarithromycin, and amoxicillin or
or tetracycline is recommended in patients previ- metronidazole remains an appropriate first-line
ously treated with a proton-pump inhibitor, therapy, provided that there is not a high local
amoxicillin, and clarithromycin.23,43 Clarithro- rate of clarithromycin resistance. Recurrence or
mycin should be avoided as part of second-line persistence of symptoms after eradication therapy
therapy unless resistance testing confirms that for uninvestigated dyspepsia is much less likely
the H. pylori strain is susceptible to the drug.44 to indicate that treatment has failed than to indi-
Patients in whom H. pylori infection persists cate that the symptoms are unrelated to H. pylori
after a second course of treatment and for whom infection. Further eradication therapy should not
eradication is considered appropriate should be be considered unless persistent H. pylori infection
referred to a specialist with access to facilities is confirmed. Data are lacking to inform the op-

1602 n engl j med 362;17 nejm.org april 29, 2010

BACK to TOC
48

timal management of recurrent or persistent dys- other potential reasons for the symptoms should
pepsia after noninvasive testing and treatment of also be reconsidered.
H. pylori infection. Options include symptomatic
acid-inhibitory therapy, endoscopy to check for and Nycomed and consulting fees from Sacoor. No other poten-
underlying ulcer or another cause of symptoms,
and repeat of the H. pylori test-and-treat strategy; full text of this article at NEJM.org.
References
1. Warren JR, Marshall BJ. Unidentified
curved bacilli on gastric epithelium in ac-
tive chronic gastritis. Lancet 1983;1:1273-5.
2. Everhart JE. Recent developments in
the epidemiology of Helicobacter pylori.
Gastroenterol Clin North Am 2000;29:
559-79.
3. Woodward M, Morrison C, McColl K.
An investigation into factors associated
with Helicobacter pylori infection. J Clin
Epidemiol 2000;53:175-81.
4. Marshall BJ, Barrett LJ, Prakash C,
McCallum RW, Guerrant RL. Urea pro-
tects Helicobacter (Campylobacter) pylori
from the bactericidal effect of acid. Gas-
troenterology 1990;99:697-702.
5. el-Omar EM, Penman ID, Ardill JES,
Chittajallu RS, Howie C, McColl KEL.
Helicobacter pylori infection and abnor-
malities of acid secretion in patients with
duodenal ulcer disease. Gastroenterology
1995;109:681-91.
6. Gillen D, el-Omar EM, Wirz AA, Ar-
dill JES, McColl KEL. The acid response to
gastrin distinguishes duodenal ulcer pa-
tients from Helicobacter pylori-infected
healthy subjects. Gastroenterology 1998;
114:50-7.
7. Hentschel E, Brandsttter G, Drag-
osics B, et al. Effect of ranitidine and
amoxicillin plus metronidazole on the
eradication of Helicobacter pylori and the re-
currence of duodenal ulcer. N Engl J Med
1993;328:308-12.
8. Axon ATR, OMorin CA, Bardhan
KD, et al. Randomised double blind con-
trolled study of recurrence of gastric ulcer
after treatment for eradication of Helico-
bacter pylori infection. BMJ 1997;314:
565-8.
9. Hansen S, Melby KK, Aase S, Jellum E,
Vollset SE. Helicobacter pylori infection
and risk of cardia cancer and non-cardia
gastric cancer: a nested case-control study.
Scand J Gastroenterol 1999;34:353-60.
10. Infection with Helicobacter pylori. In:
IARC Working Group on the Evaluation
of Carcinogenic Risks to Humans. IARC
monographs on the evaluation of carcino-
genic risks to humans. Vol. 61. Schisto-
somes, liver flukes and Helicobacter pylori.
Lyon, France: International Agency for
Research on Cancer, 1994:177-240.
11. Uemura N, Okamoto S, Yamamoto S,
et al. Helicobacter pylori infection and the
development of gastric cancer. N Engl J
Med 2001;345:784-9.
n engl j med 362;17
BACK to TOC
clinical pr actice
Dr. McColl reports receiving lecture fees from AstraZeneca
tial conflict of interest relevant to this article was reported.
Disclosure forms provided by the author are available with the
12. Leung WK, Lin SR, Ching JY, et al.
Factors predicting progression of gastric
intestinal metaplasia: results of a ran-
domised trial on Helicobacter pylori eradi-
cation. Gut 2004;53:1244-9.
13. Malfertheiner P, Sipponen P, Naumann
M, et al. Helicobacter pylori eradication
has the potential to prevent gastric can-
cer: a state-of-the-art critique. Am J Gas-
troenterol 2005;100:2100-15.
14. Parsonnet J, Hansen S, Rodriguez L,
et al. Helicobacter pylori and gastric lym-
phoma. N Engl J Med 1994;330:1267-71.
15. Fischbach W, Goebeler-Kolve ME,
Dragosics B, Greiner A, Stolte M. Long
term outcome of patients with gastric
marginal zone B cell lymphoma of mu-
cosa associated lymphoid tissue (MALT)
following exclusive Helicobacter pylori
eradication therapy: experience from a
large prospective series. Gut 2004;53:
34-7.
16. Moayyedi P, Soo S, Deeks J, et al. Sys-
tematic review and economic evaluation
of Helicobacter pylori eradication treat-
ment for non-ulcer dyspepsia. BMJ 2000;
321:659-64.
17. Laine L, Schoenfeld P, Fennerty MB.
Therapy for Helicobacter pylori in patients
with nonulcer dyspepsia: a meta-analysis
of randomized, controlled trials. Ann In-
tern Med 2001;134:361-9.
18. McColl KEL. Absence of benefit of
eradicating Helicobacter pylori in patients
with nonulcer dyspepsia. N Engl J Med
2000;342:589.
19. Raghunath A, Hungin AP, Wooff D,
Childs S. Prevalence of Helicobacter py-
lori in patients with gastro-oesophageal
reflux disease: systematic review. BMJ
2003;326:737-9.
20. de Martel C, Llosa AE, Farr SM, et al.
Helicobacter pylori infection and the risk
of development of esophageal adenocarci-
noma. J Infect Dis 2005;191:761-7.
21. Yaghoobi M, Farrokhyar F, Yuan Y,
Hunt RH. Is there an increased risk of
GERD after Helicobacter pylori eradica-
tion? A meta-analysis. Am J Gastroenterol
2010 January 19 (Epub ahead of print).
22. Chey WD, Wong BCY, Practice Param-
eters Committee of the American College
of Gastroenterology. American College of
Gastroenterology guideline on the man-
agement of Helicobacter pylori infection.
Am J Gastroenterol 2007;102:1808-25.
23. Malfertheiner P, Megraud F, OMorain
nejm.org april 29, 2010
C, et al. Current concepts in the manage-
ment of Helicobacter pylori infection: the
Maastricht III consensus report. Gut 2007;
56:772-81.
24. Fukase K, Kato M, Kikuchi S, et al.
Effect of eradication of Helicobacter py-
lori on incidence of metachronous gastric
carcinoma after endoscopic resection of
early gastric cancer: an open-label, ran-
domised controlled trial. Lancet 2008;372:
392-7.
25. Talley NJ, Vakil N, Practice Parame-
ters Committee of the American College
of Gastroenterology. Guidelines for the
management of dyspepsia. Am J Gastro-
enterol 2005;100:2324-37.
26. McColl KEL, Murray LS, Gillen D, et
al. Randomised trial of endoscopy with
testing for Helicobacter pylori compared
with non-invasive H pylori testing alone
in the management of dyspepsia. BMJ
2002;324:999-1002. [Errata, BMJ 2002;325:
479, 580.]
27. Lassen AT, Pedersen FM, Bytzer P,
Schaffalitzky de Muckadell OB. Helico-
bacter pylori test-and-eradicate versus
prompt endoscopy for management of
dyspeptic patients: a randomised trial.
Lancet 2000;356:455-60.
28. Chiba N, Van Zanten SJO, Sinclair P,
Ferguson RA, Escobedo S, Grace E. Treat-
ing Helicobacter pylori infection in pri-
mary care patients with uninvestigated
dyspepsia: the Canadian adult dyspepsia
empiric treatmentHelicobacter pylori pos-
itive (CADET-Hp) randomised controlled
trial. BMJ 2002;324:1012-6.
29. The Danish Dyspepsia Study Group.
Value of the unaided clinical diagnosis in
dyspepsia patients in primary care. Am J
Gastroenterol 2001;96:1417-21.
30. Jarbol DE, Kragstrup J, Stovring H,
Havelund T, Schaffalitzky de Muckadell
OB. Proton pump inhibitor or testing for
Helicobacter pylori as the first step for pa-
tients presenting with dyspepsia? A cluster-
randomized trial. Am J Gastroenterol
2006;101:1200-8.
31. Manes G, Menchise A, de Nucci C,
Balzano A. Empirical prescribing for dys-
pepsia: randomised controlled trial of test
and treat versus omeprazole treatment.
BMJ 2003;326:1118.
32. Delaney BC, Innes MA, Deeks J, et al.
Initial management strategies for dyspep-
sia. Cochrane Database Syst Rev 2001;3:
CD001961.
1603
49
 clinical pr actice

33. Delaney BC, Moayyedi P, Forman D.
Initial management strategies for dyspep-
sia. Cochrane Database Syst Rev 2003;2:
CD001961.
34. Loy CT, Irwig LM, Katelaris PH, Talley
NJ. Do commercial serological kits for
Helicobacter pylori infection differ in ac-
curacy? A meta-analysis. Am J Gastroen-
terol 1996;91:1138-44.
35. Vaira D, Vakil N. Blood, urine, stool,
breath, money, and Helicobacter pylori.
Gut 2001;48:287-9.
36. Gisbert JP, Pajares JM. Stool antigen
test for the diagnosis of Helicobacter py-
lori infection: a systematic review. Helico-
bacter 2004;9:347-68.
37. Midolo P, Marshall BJ. Accurate diag-
nosis of Helicobacter pylori: urease tests.
Gastroenterol Clin North Am 2000;29:
871-8.
38. Fuccio L, Minardi ME, Zagari RM,
Grilli D, Magrini N, Bazzoli F. Meta-analy-
sis: duration of first-line proton-pump
inhibitor based triple therapy for Helico-
bacter pylori eradication. Ann Intern Med
2007;147:553-62.
39. Fischbach L, Evans EL. Meta-analysis:
the effect of antibiotic resistance status
on the efficacy of triple and quadruple
first-line therapies for Helicobacter pylori.
Aliment Pharmacol Ther 2007;26:343-57.
40. Jafri NS, Hornung CA, Howden CW.
Meta-analysis: sequential therapy appears
superior to standard therapy for Helico-
bacter pylori infection in patients naive to
treatment. Ann Intern Med 2008;148:923-
31. [Erratum, Ann Intern Med 2008;149:
439.]
41. Snchez-Delgade J, Calvet X, Bujanda L,
Gisbert JP, Tit L, Castro M. Ten-day se-
quential treatment for Helicobacter pylori
eradication in clinical practice. Am J Gas-
troenterol 2008;103:2220-3.
42. Gisbert JP, Pajares JM. Helicobacter
pylori rescue regimen when proton pump
inhibitor-based triple therapies fail. Ali-
ment Pharmacol Ther 2002;16:1047-57.
43. Realdi G, Dore MP, Piana A, et al. Pre-
treatment antibiotic resistance in Helico-
bacter pylori infection: results of three
randomized controlled studies. Helico-
bacter 1999;4:106-12.
44. Lamouliatte H, Mgraud F, Delchier
JC, et al. Second-line treatment for failure
to eradicate Helicobacter pylori: a random-
ized trial comparing four treatment strat-
egies. Aliment Pharmacol Ther 2003;18:
791-7.
45. Gisbert JP, Morena F. Systematic review
and meta-analysis: levofloxacin-based res-
cue regimens after Helicobacter pylori
treatment failure. Aliment Pharmacol Ther
2006;23:35-44.
46. Saad RJ, Schoenfeld P, Kim HM, Chey
WD. Levofloxacin-based triple therapy
versus bismuth-based quadruple therapy
for persistent Helicobacter pylori infec-
tion: a meta-analysis. Am J Gastroenterol
2006;101:488-96.
47. Qasim A, Sebastian S, Thornton O,
et al. Rifabutin- and furazolidone-based
Helicobacter pylori eradication therapies
after failure on standard first- and second-
line eradication attempts in dyspepsia pa-
tients. Aliment Pharmacol Ther 2005;21:
91-6.
Copyright 2010 Massachusetts Medical Society.

FULL TEXT OF ALL JOURNAL ARTICLES ON THE WORLD WIDE WEB

Access to the complete contents of the Journal on the Internet is free to all subscribers. To use this Web site, subscribers should
go to the Journals home page (NEJM.org) and register by entering their names and subscriber numbers as they appear on their
mailing labels. After this one-time registration, subscribers can use their passwords to log on for electronic access to the entire
Journal from any computer that is connected to the Internet. Features include a library of all issues since January 1993 and
abstracts since January 1975, a full-text search capacity, and a personal archive for saving articles and search results of interest.
All articles can be printed in a format that is virtually identical to that of the typeset pages. Beginning 6 months after
publication, the full text of all Original Articles and Special Articles is available free to nonsubscribers.

1604 n engl j med 362;17 nejm.org april 29, 2010

BACK to TOC
50
 Clinical Therapeutics

Clinical Therapeutics articles provide practical guidance for the use of specific medications, devices,
and procedures in patient care. Topics include: how the therapy is used, evidence that supports (or
fails to support) its use, adverse effects and areas of uncertainty, guidelines from major professional
societies, and author recommendations.

BACK to TOC
51

From the Division of Pediatric Hematol-
ogy, Department of Pediatrics, Columbia
University College of Physicians and Sur-
geons, New York. Address reprint re-
quests to Dr. Brittenham at the Division
of Pediatric Hematology, Department of
Pediatrics, Columbia University Medical
Center, Rm. CHN 10-08, 3959 Broadway,
New York, NY 10032, or at gmb31@
columbia.edu.
N Engl J Med 2011;364:146-56.
Copyright 2011 Massachusetts Medical Society.
146
BACK to TOC
The n e w e ng l a n d j o u r na l of m e dic i n e
clinical therapeutics
Iron-Chelating Therapy
for Transfusional Iron Overload
Gary M. Brittenham, M.D.
This Journal feature begins with a case vignette that includes a therapeutic recommendation. A discussion
of the clinical problem and the mechanism of benefit of this form of therapy follows. Major clinical studies,
the clinical use of this therapy, and potential adverse effects are reviewed. Relevant formal guidelines,
if they exist, are presented. The article ends with the authors clinical recommendations.
A 16-year-old boy with sickle cell anemia undergoes routine screening with transcra-
nial Doppler ultrasonography to assess the risk of stroke. This examination shows an
abnormally elevated blood-flow velocity in the middle cerebral artery. The hemoglo-
bin level is 7.2 g per deciliter, the reticulocyte count is 12.5%, and the fetal hemoglo-
bin level is 8.0%. Long-term treatment with red-cell transfusion is initiated to prevent
stroke. A hematologist recommends prophylactic iron-chelating therapy.
The Cl inic a l Probl em
Long-term treatment with red-cell transfusion effectively prevents stroke and other
complications of sickle cell anemia1 and can sustain patients with chronic con-
genital and acquired refractory anemia, including thalassemia major, Diamond
Blackfan anemia, myelodysplastic syndromes, myelofibrosis, aplastic anemia, and
other disorders. In the United States, 10,000 to 20,000 patients with sickling disor-
ders receive repeated transfusions. An estimated 4000 to 5000 patients with myelo-
dysplastic syndromes and other forms of acquired refractory anemia require red-
cell transfusions. The number of patients with transfusion-dependent thalassemia
in the United States is smaller probably less than 1000.2 However, globally, al-
most 100,000 patients with thalassemia syndromes undergo transfusions. The ma-
jority of these patients are in low- and middle-income countries.3
Because humans lack any effective means to excrete excess iron, long-term trans-
fusion alone inexorably produces the clinical problem of iron overload. In patients
with thalassemia who undergo transfusion from infancy, iron-induced liver dis-
ease and endocrine disorders develop during childhood and are almost inevitably
followed in adolescence by death from iron-induced cardiomyopathy.4 In patients
with sickle cell anemia, although iron-induced complications appear to develop later,
eventually, liver disease with cirrhosis as well as cardiac and pancreatic iron deposi-
tion can develop.5,6 The annual per-patient costs of care for complications of iron
overload are estimated at $15,000 to $20,000.7,8
PATHOPH YSIOL O GY A ND EFFEC T OF THER A PY
At the end of their life span, transfused red cells are phagocytosed by reticuloendo-
thelial macrophages in the liver, bone marrow, and spleen (Fig. 1). Their hemoglobin
is digested, and the iron is freed from heme and released into the cytosol. Early in
the course of long-term transfusion, most of this additional iron can be stored
within reticuloendothelial macrophages. Gradually, limits on the capacity of macro-
phages to retain iron result in the release of excess iron into plasma.9 Transferrin
n engl j med 364;2 nejm.org january 13, 2011
52
 clinical ther apeutics
binds the released iron, with an increase in the
plasma iron concentration and transferrin satu-
ration. As the transferrin saturation increases,
hepatocytes are recruited to serve as storage sites
for the excess iron.
With continued transfusion, macrophages and
hepatocytes can no longer retain all the surplus
iron. Iron then enters plasma in amounts that
exceed the transport capacity of circulating trans-
ferrin. As a consequence, nontransferrin-bound
iron appears in the plasma (Fig. 1) as a hetero-
geneous assortment of iron complexes that ap-
pear to be the major mediators of extrahepatic
tissue damage in transfusional iron overload.10
Nontransferrin-bound plasma iron enters spe-
cific cells, particularly hepatocytes, cardiomyo-
cytes, anterior pituitary cells, and pancreatic beta-
cells. In these cells, iron accumulation leads to
the generation of reactive oxygen species, result-
ing in damage to lipids, proteins, DNA, and
subcellular organelles, including lysosomes and
mitochondria. This injury may result in cellular
dysfunction, apoptosis, and necrosis.
Therapy with chelating agents that form a
complex with iron and promote its excretion can
clear plasma nontransferrin-bound iron, remove
excess iron from cells, and maintain or return
body iron to safe levels (Fig. 1). (An interactive
graphic depicting iron supply and storage in
sickle cell anemia with long-term red-cell trans-
fusion and iron-chelating therapy is available
with the full text of this article at NEJM.org.)
Two iron-chelating agents are approved for use
in North America (Table 1): parenteral deferox-
amine mesylate (Desferal, Novartis) and oral
deferasirox (Exjade, Novartis).
Deferoxamine is a siderophore (an iron-bind-
ing compound) produced by the bacterium Strep-
tomyces pilosus. It is poorly absorbed after oral ad-
ministration and is rapidly cleared; consequently,
subcutaneous or intravenous administration is
necessary. One molecule of deferoxamine binds a
single atom of iron, forming a feroxamine com-
plex that is virtually inert metabolically. Plasma
iron chelated with deferoxamine is eliminated pre-
dominantly by the kidneys. Hepatocytes efficiently
take up deferoxamine, which then chelates hepato-
cellular iron, with the feroxamine excreted in the
bile. Within cells, deferoxamine is localized to ly-
sosomes, where it induces autophagy of cytosolic
ferritin. Lysosomal degradation of cytosolic ferritin
releases iron that is bound by deferoxamine, and
the chelated iron is then cleared from the cell.14
n engl j med 364;2 nejm.org
BACK to TOC
In contrast to deferoxamine, the synthetic che-
lator deferasirox is well absorbed from the gastro-
intestinal tract and is cleared from the circulation
slowly.13,15 Two molecules of deferasirox are need-
ed to bind a single atom of iron. Like deferox-
amine, deferasirox forms complexes with plasma
iron, but deferasiroxiron complexes are elimi-
nated predominantly through a hepatobiliary
route. Hepatocytes readily take up deferasirox,
which chelates hepatocellular iron. The defera-
siroxiron complexes are then excreted in the
bile.15 Within cells, deferasirox chelates cyto-
solic iron, leading to ferritin degradation by the
proteasome.14
A third iron chelator, the synthetic oral agent
deferiprone (Ferriprox, Apotex; Kelfer, Cipla), is
not approved for use in the United States or
Canada. In the European Union and some other
countries, it is approved specifically for patients
with thalassemia major when deferoxamine is
contraindicated or inadequate (Table 1).
CL INIC A L E V IDENCE
The use of deferoxamine therapy antedates the
common use of randomized, controlled trials to
establish the efficacy of medical treatments. Only
one small, randomized trial has compared chela-
tion plus deferoxamine with no therapy; this trial An interactive Click here
graphic is to access
enrolled 20 children with -thalassemia. After a available at interactive
mean of 5.8 years of treatment with intramuscu- NEJM.org graphic.
Figure 1 (see next two pages). Iron Supply and Storage in Sickle Cell
Anemia with Long-Term Red-Cell Transfusion and Iron-Chelating Therapy.
In all four panels, the area of the red and blue circles is roughly proportional
to the amount of iron in each pool, and the width of the arrows is roughly
proportional to the daily magnitude of the iron flux. Panel A shows normal
iron supply and storage in a healthy person without sickle cell disease (i.e.,
with hemoglobin A). The major pathway of internal iron exchange is a unidi-
rectional flow from plasma transferrin to the erythroid marrow to circulating
red cells to reticuloendothelial macrophages and back to plasma transferrin
(orange arrows). In the circulating plasma, virtually all iron is bound to trans-
ferrin. Panel B shows that in sickle cell anemia, hemolysis shortens the aver-
age life span of the red cell from about 4 months to 5 or 6 weeks, increasing
red-cell catabolism by reticuloendothelial macrophages and increasing iron
delivery to the erythroid marrow by 6 to 8 times the normal rate. There is lit-
tle ineffective erythropoiesis, and iron absorption from the gastrointestinal
tract is not increased. Panel C shows that long-term red-cell transfusion de-
creases erythroid marrow activity to 2 to 3 times the normal rate but results
in accumulation of iron in reticuloendothelial macrophages and hepatocytes.
Eventually, the capacity for safe storage is exceeded, with the appearance of
plasma nontransferrin-bound iron (dashed arrow) and its progressive depo-
sition in the heart and endocrine organs. In Panel D, the green arrows show
that iron-chelating therapy with deferasirox can clear plasma nontransferrin-
bound iron and remove excess iron from the liver, heart, and other organs,
with subsequent excretion through the bile into the stool.
january 13, 2011 147
53
 The n e w e ng l a n d j o u r na l of m e dic i n e

A Normal Iron Supply and Storage

Reticuloendothelial
macrophages
Functional iron
Storage iron Heart and Muscle and other
endocrine organs parenchymal cells

Transferrin-
bound iron
Ci l tii
Circulating Gastrointestinal
red cells tract

Hepatocytes
Erythroid
marrow

B Sickle Cell Anemia

Reticuloendothelial
macrophages
Heart and Muscle and other
endocrine organs parenchymal cells

Transferrin-
bound iron
Circulating
Gastrointestinal
red cells
tract

Hepatocytes
Erythroid
marrow

148 n engl j med 364;2 nejm.org january 13, 2011

BACK to TOC
54
 clinical ther apeutics

C Sickle Cell Anemia with Long-Term Nontransferrin-bound iron

Red-Cell Transfusion
Reticuloendothelial
macrophages
Heart and Muscle and other
Transfusion parenchymal cells
endocrine organs

Transferrin-
bound iron
Circullating
Circulating
ti Gastrointestinal
red cells tract

Hepatocytes
Erythroid
marrow

D Sickle Cell Anemia with Transfusion Chelator-bound iron

and Iron-Chelating Therapy
Reticuloendothelial
macrophages Muscle and other
Heart and parenchymal cells
Transfusion
endocrine organs

Transferrin-
bound iron
Circulating
ti Gastrointestinal
red cells tract

Hepatocytes
Erythroid
marrow

n engl j med 364;2 nejm.org january 13, 2011 149

BACK to TOC
55
 The n e w e ng l a n d j o u r na l of m e dic i n e

lar deferoxamine, the mean hepatic iron concen-

plasma zinc level, progression of hepatic fibrosis asso-

tration was 25.9 mg per gram of liver tissue (dry

Agranulocytosis and neutropenia; gastrointestinal distur-

major, when deferoxamine therapy is contraindicated
Transfusional iron overload in patients with thalassemia

bances, arthropathy, increased liver-enzyme levels, low

ciated with increase in iron overload or hepatitis C11

Not approved in United States or Canada; approved in
weight) in the deferoxamine group and 42.2 mg
per gram in the control group.16 At 14 years, one
death had occurred in the deferoxamine group and
six deaths had occurred in the control group.17
In lieu of randomized trials, observational
studies have investigated the effects of deferox-
Deferiprone

Europe11 and other countries

amine in the management of transfusion-related
iron overload. One study involved 977 children
with transfusion-dependent thalassemia major
Bidentate, 3:1 complex

Predominantly urinary
Oral, three times daily

who survived beyond the first decade of life.18

75100 mg/kg/day

or inadequate
Subsequent survival was examined according to
5-year birth cohorts beginning in 1960; deferox-
amine was introduced in 1975. The survival rate
23 hr

increased progressively in each 5-year cohort

(see Fig. 1 in the Supplementary Appendix, avail-
serum creatinine level; potentially fatal renal able at NEJM.org). The survival rate was signifi-
Gastrointestinal disturbances, rash, increase in

and hepatic impairment or failure, gastro- cantly higher among children born after 1975
Approved in United States, Canada, Europe,

than among those in previous cohorts.

Deferasirox has been compared with deferox-
amine in a few short-term trials sponsored by
Novartis.19-23 In the largest of these trials, 586
Deferasirox

children with -thalassemia were randomly as-

intestinal hemorrhage13
Transfusional iron overload

signed to either agent, with dosing according to

Tridentate, 2:1 complex

and other countries

Predominantly biliary

the baseline hepatic iron concentration.19 The pri-

2040 mg/kg/day

mary end point was the percentage of subjects

Oral, once daily

with either a maintained or reduced hepatic iron

concentration at 1 year; this end point was
816 hr

reached in 52.9% of patients assigned to de-

ferasirox and in 66.4% of patients assigned to
deferoxamine. This result, which did not meet a
changes, allergy, respiratory distress syndrome
disturbances, growth retardation and skeletal
Irritation at the infusion site, ocular and auditory

prespecified noninferiority target, was attributed

to the relative underdosing of deferasirox. The
Approved in United States, Canada, Europe,

with higher-than-recommended doses12

Subcutaneous or intravenous, 810 hr/day,

total hepatic iron concentration decreased by a

mean of 2.4 mg per gram (dry weight) in the
deferasirox group and by 2.9 mg per gram in the
Deferoxamine

deferoxamine group. No trial has established the

Transfusional iron overload

long-term effectiveness of deferasirox in prevent-

Hexadentate, 1:1 complex

and other countries

ing organ toxicity or improving survival.

Table 1. Iron-Chelating Agents in Clinical Use.

Deferiprone has also been compared with de-

Biliary and urinary
2550 mg/kg/day

57 days/wk

feroxamine in several small, randomized trials.24

As is the case with deferasirox, no long-term
2030 min

trials have been performed to evaluate the effect

of deferiprone on organ function or survival.
Chelator-iron complex

CL INIC A L USE
Route of elimination
Regulatory approval

Iron-chelating therapy should be considered in

Plasma half-life

Adverse effects
Administration

all patients who require long-term red-cell trans-

Usual dose

Indication
Variable

fusion. Such patients include those with sickle

cell disease, myelodysplastic syndromes, thalas-
semia major, DiamondBlackfan anemia, aplastic

150 n engl j med 364;2 nejm.org january 13, 2011

BACK to TOC
56
 clinical ther apeutics
anemia, and other congenital and acquired forms
of refractory anemia.
There are alternatives to chelation in some
patients. Some of the underlying disorders re-
quiring transfusion may be cured by hematopoi-
etic stem-cell transplantation. In some patients
with sickle cell disease, exchange transfusion may
reduce or obviate the need for iron chelation. In-
frequently, phlebotomy may be an option for the
removal of excess iron in the event of cure or re-
mission of a refractory anemia. Iron-chelating
therapy itself may sometimes decrease or elimi-
nate the need for transfusion in patients with my-
elodysplasia25 or myelofibrosis. Chelation therapy
may not be needed in patients with myelodyspla-
sia or other acquired refractory anemias who have
an estimated survival of less than 1 year.26
With the exception of these groups, iron-
chelating therapy is indicated in almost all pa-
tients requiring long-term red-cell transfusion.
Iron chelation is contraindicated in patients who
are hypersensitive to the chelating agent or ex-
cipients in the chelator formulation, and it re-
quires specialized management in patients with
several renal disease or anuria. Chelators should
be avoided or used with great caution in patients
who are pregnant or breast-feeding.
Ideally, iron-chelating therapy should be initi-
ated prophylactically, before clinically significant
iron accumulation has occurred. Treatment should
begin when patients have received between 10 and
20 red-cell transfusions. Patients who have already
undergone repeated transfusion without sufficient
chelation can also be successfully treated, but they
may require more intensive regimens (see below).
Evaluation of the patient before the initiation
or adjustment of iron-chelating therapy includes a
detailed characterization of the underlying dis-
order, with thorough documentation of the his-
tory of transfusion and chelation; determination
of the body iron load by measurement of the
hepatic iron and serum ferritin concentrations;
estimation of the rate of transfusional iron load-
ing; and assessment of cardiac iron deposition.
The techniques for assessing cardiac iron over-
load, transfusional iron loading, and body iron
burden are described in the Supplementary Ap-
pendix. The extent of any existing iron-induced
hepatic, cardiac, or endocrine dysfunction should
be established, and in children and adolescents,
growth and maturation should be assessed. Nu-
tritional evaluation with correction of deficien-
cies is recommended.27
n engl j med 364;2
BACK to TOC
In the United States and Canada, the choice
of an iron chelator for transfusional iron over-
load is either parenteral deferoxamine or oral
deferasirox. The decision is best made with the
patient and, if the patient is a child, with his or
her parents. Despite the lack of data on long-
term effectiveness, most patients now opt for
deferasirox because of the ease of oral adminis-
tration. Deferasirox is preferred for prophylactic
or maintenance therapy. Deferoxamine, which
has been proved to reverse iron-induced heart
disease and increase long-term survival,28 may
be indicated if deferasirox is ineffective in a par-
ticular patient, and it may be favored for severe
iron overload, especially with cardiac involvement.
Conversely, deferasirox may be the better choice
in patients who are unable to tolerate subcutane-
ous infusions of deferoxamine. Deferasirox also
may be substituted for deferoxamine after suc-
cessful clearance of cardiac iron. Deferiprone is
available in the United States on a compassion-
ate-use basis, usually in combination with defer-
oxamine, in patients in whom iron-induced
heart failure has developed or who are at high
risk for the development of heart failure.29,30
Deferoxamine is administered subcutaneously
or intravenously, usually with a portable pump, for
8 to 10 hours each day, 5 to 7 days per week.
Subcutaneous administration is preferred except
in patients with severe cardiac iron deposition, for
whom continuous intravenous deferoxamine ther-
apy is recommended.28,31 Deferasirox is adminis-
tered orally once daily, and deferiprone is admin-
istered orally three times daily.
The dose of an iron-chelating agent is deter-
mined by three principal factors: the presence or
absence of cardiac iron overload, the rate of
transfusional iron loading, and the body iron
burden (see the Supplementary Appendix and
Table 2). In brief, if cardiac iron overload is pres-
ent, ridding the heart of the excess iron becomes
the critical therapeutic goal. In the absence of
cardiac iron overload, the long-term objective is
to maintain the body iron at a level that permits
safe storage while avoiding chelator toxicity. The
greater the rate of transfusional iron loading,
the greater the dose of an iron chelator that will
be needed to control the accumulation of iron.
During treatment, tests to monitor chelator-
associated toxicity should be performed, depend-
ing on the potential adverse effects of the specific
agent to be used (see Table 1 and the Adverse
Effects section below). In patients who receive
nejm.org january 13, 2011 151
57
 152

BACK to TOC
Table 2. Usual Doses of Deferoxamine or Deferasirox for Transfusional Iron Overload.*

Hepatic Iron
Concentration No Cardiac Iron Overload (T2*, 20 msec) Cardiac Iron Overload (T2*, <20 msec)

Daily Transfusional Iron Intake Mild to Moderate Severe

<0.3 mg/kg of body weight 0.3 to 0.5 mg/kg of body weight >0.5 mg/kg of body weight T2*, 10 to <20 msec T2*, <10 msec
The

15 mg/g, dry weight Deferoxamine: 4050 mg/kg/ Deferoxamine: 4050 mg/kg/ Deferoxamine: 4050 mg/kg/ Deferoxamine: 50 mg/kg/day Deferoxamine: 50 mg/kg/day
day, 8 to 10 hr/day, 6 or 7 day, 8 to 10 hr/day, 6 or 7 day, 8 to 10 hr/day, 6 or 7 by continuous intravenous by continuous intravenous
days/wk, by subcutaneous days/wk, by subcutaneous days/wk, by subcutaneous infusion28; deferasirox: oral infusion28; deferasirox: oral
infusion; deferasirox: oral infusion; deferasirox: oral infusion; deferasirox: oral dose of 40 mg/kg daily, but dose of 40 mg/kg daily, but
dose of 3040 mg/kg daily dose of 3040 mg/kg daily dose of 3040 mg/kg daily uncertain efficacy in reduc- uncertain efficacy in reduc-
ing cardiac iron ing cardiac iron
7 to <15 mg/g, dry Deferoxamine: 3040 mg/kg/ Deferoxamine: 4050 mg/kg/ Deferoxamine: 4050 mg/kg/ Deferoxamine: 4050 mg/kg/ Deferoxamine: 4050 mg/kg/

n engl j med 364;2

weight day, 8 to 10 hr/day, 5 days/ day, 8 to 10 hr/day, 6 or 7 day, 8 to 10 hr/day, 6 or 7 day, 8 to 10 hr/day, 6 or 7 day by continuous infu-
wk, by subcutaneous infu- days/wk, by subcutaneous days/wk, by subcutaneous days/wk, by subcutaneous sion28; deferasirox: oral
sion; deferasirox: oral dose infusion; deferasirox: oral infusion; deferasirox: oral infusion; deferasirox: oral dose of 3040 mg/kg daily
of 2030 mg/kg daily dose of 3040 mg/kg daily dose of 3040 mg/kg daily dose of 3040 mg/kg daily

nejm.org
3 to <7 mg/g, dry Deferoxamine: 3040 mg/kg/ Deferoxamine: 3040 mg/kg/ Deferoxamine: 3040 mg/kg/ Deferoxamine: 4050 mg/kg/ Deferoxamine: 4050 mg/kg/
weight day, 8 to 10 hr/day, 5 days/ day, 8 to 10 hr/day, 5 days/ day, 8 to 10 hr/day, 5 days/ day, 8 to 10 hr/day, 6 or 7 day by continuous infu-
n e w e ng l a n d j o u r na l

wk, by subcutaneous infu- wk, by subcutaneous infu- wk, by subcutaneous infu- days/wk, by subcutaneous sion28; deferasirox: oral
of

sion; deferasirox: oral dose sion; deferasirox: oral dose sion; deferasirox: oral dose infusion; deferasirox: oral dose of 3040 mg/kg daily
of 2030 mg/kg daily of 2030 mg/kg daily of 2030 mg/kg daily dose of 3040 mg/kg daily
<3 mg/g, dry weight Deferoxamine: suspend thera- Deferoxamine: suspend thera- Deferoxamine: suspend thera- Deferoxamine: adjust intrave- Deferoxamine: adjust intrave-

january 13, 2011

py; deferasirox: suspend py; deferasirox: suspend py; deferasirox: suspend nous or subcutaneous dose nous or subcutaneous dose
therapy therapy therapy according to therapeutic according to therapeutic
m e dic i n e

index32; deferasirox: adjust index32; deferasirox: adjust

oral dose, monitoring renal oral dose, monitoring renal
and hepatic function and and hepatic function and
blood count blood count

* To minimize interference with growth and skeletal development, the dose of deferoxamine in young children should not exceed 25 to 30 mg per kilogram of body weight. The dose
should be adjusted according to the therapeutic index.32 The bioavailability of deferasirox may affect the response. T2* denotes the cardiac effective transverse relaxation time on mag-
netic resonance imaging.

58
 clinical ther apeutics

deferoxamine, these tests include annual assess- the dose as the hepatic iron concentration ap-
ments of auditory function and vision. In patients proaches normal levels. Although treatment with
who receive deferasirox, serum creatinine, serum deferoxamine may reduce endocrine complica-
aminotransferases, and bilirubin levels and com- tions of iron overload, such as a delay of puberty,
plete blood counts should be assessed monthly. the chelator itself can interfere with growth,35
In patients who receive deferiprone, weekly as- apparently as a result of skeletal dysplasia.36 To
sessment of complete blood counts and monthly minimize this effect, the dose of deferoxamine in
assessments of serum aminotransferases should young children should not exceed 25 to 30 mg per
be performed. kilogram.37
The effectiveness of iron-chelating therapy is In the registration trial of deferasirox de-
best monitored by periodic measurements of car- scribed above,19 gastrointestinal disturbances
diac iron concentrations by magnetic resonance occurred in approximately 15% of patients, rash
imaging (MRI), of cardiac function, and of he- in 11%, and increases in serum creatinine levels
patic iron concentrations and by review of the in 38%. Similar rates have been observed in
actual rate of transfusional iron loading once subsequent trials.20-22 In January 2010, on the
or twice a year, depending on the severity of the basis of postmarketing studies, the Food and
iron overload (see the Supplementary Appendix). Drug Administration required a change in the
Dose adjustments can be made according to the prescribing information for deferasirox. The new
guidelines in Table 2. Serum ferritin concentra- information states that the drug could cause
tions are usually measured at least quarterly. potentially fatal renal and hepatic impairment or
As at baseline, hepatic, cardiac, and endocrine failure as well as gastrointestinal hemorrhage.13
function should be assessed periodically along These adverse effects were reported to occur
with nutritional status, and, in children and more frequently in older patients and in patients
adolescents, growth and maturation should be with high-risk myelodysplastic syndromes, throm-
monitored. bocytopenia, or underlying renal or hepatic im-
In the United States, on the basis of 2006 pairment.
wholesale acquisition prices, the annual costs of The most common adverse effects of deferi-
deferoxamine for iron-chelating therapy have been prone are diarrhea and gastrointestinal effects,
estimated to range from $6,824 to $29,209, plus arthropathy (including severe arthritis with clin-
$9,286 for infusion, and the estimated annual ically significant disability), increased levels of
costs of deferasirox range from $24,404 to $53,095, serum liver enzymes, and progression of hepatic
with the actual cost depending on dose and body fibrosis associated with an increase in iron over-
weight.7 Administration of a chelator will be load or hepatitis C.11 The most serious adverse
needed as long as transfusion is continued and effects are agranulocytosis (incidence, 1.1%) and
will be lifelong in most patients. neutropenia (incidence, 4.9%); weekly monitor-
ing of the neutrophil count is recommended.
A DV ER SE EFFEC T S Neurologic abnormalities have been reported
with higher-than-recommended doses of deferi-
Discomfort or pain at the site of subcutaneous prone.38
infusion develops in almost all patients treated
with subcutaneous deferoxamine, and induration A r e a s of Uncer ta in t y
or erythema develops in some patients. These
symptoms can often be mitigated with topical Several areas of uncertainty exist with regard to
anesthetic or glucocorticoid creams. Visual and the optimal approach to iron-chelating therapy.
auditory toxicity associated with deferoxamine First, a variety of binary combinations of chelat-
has been reported, and in one series involving 89 ing agents are being examined in off-label uses,
patients treated with this agent, 13 presented with with either synchronous or sequential adminis-
vision loss, deafness, or both.33 Subsequent stud- tration. The clinical usefulness of such combina-
ies suggest a much lower incidence of toxicity,34 tion therapies is unclear at present, in the ab-
and these risks can be minimized by not exceed- sence of unequivocal evidence of the superiority
ing doses of 50 mg per kilogram of body weight of any specific combination over treatment with
in patients with iron overload and by decreasing a single agent.24

n engl j med 364;2 nejm.org january 13, 2011 153

BACK to TOC
59
 The n e w e ng l a n d j o u r na l
Second, MRI performed to evaluate the iron
content of the liver, heart, and other organs has
become the method of choice for guiding iron-
chelating therapy, but calibrated methods are not
available for all patients. A 1-year study of de-
ferasirox involving patients with various types of
anemias used an alternative approach; the inves-
tigators based the initial dose on the rate of
transfusional iron loading and subsequently ad-
justed the dose according to measurements of
serum ferritin levels and safety markers.23 The
long-term efficacy and safety of this strategy are
uncertain.
Third, in patients with myelodysplastic syn-
dromes who have undergone long-term transfu-
sion and for whom prolonged survival is antici-
pated, iron-chelating therapy may be appropriate.
In individual patients, the benefit of such treat-
ment may vary, given the morbidity associated
with chelation, the variable prognosis for the
underlying disorder, and the latency period be-
tween the onset of the transfusion and the de-
velopment of clinical manifestations of iron
overload. Data are lacking from prospective,
randomized trials examining the clinical cir-
cumstances in which morbidity and mortality
improve with iron-chelating therapy in these
patients39; one such trial is currently recruiting
patients.40
Fourth, there is evidence to suggest that iron
overload is associated with lower rates of cardio-
myopathy, endocrinopathy, and other conditions
among patients with sickle cell disease than
among patients with thalassemia.41 The effects
of the systemic inflammatory state on iron han-
dling in patients with sickle cell disease,42 as
well as differences in the rate and duration of
transfusion and in the age of the patient at the
initiation of long-term transfusion, the extent of
ineffective erythropoiesis, and gastrointestinal
iron absorption, may be involved. It is unclear
whether these data provide sufficient grounds to
recommend a higher body iron threshold for
chelation therapy in patients with sickle cell
disease, especially given the potential risks of
iron-induced liver disease.
Finally, there is a lack of certainty with respect
to the optimal hepatic iron concentrations (Table
2) for minimizing the risk that hepatic fibrosis
will progress to cirrhosis and its ultimate compli-
cation, hepatocellular carcinoma.43,44
154 n engl j med 364;2
BACK to TOC
of m e dic i n e
Guidel ine s
Guidelines and consensus statements on the man-
agement of sickle cell disease,45,46 thalasse-
mia,32,47 DiamondBlackfan anemia,48 aplastic
anemia,49 and myelodysplastic syndromes26,50 all
include recommendations for iron-chelating
therapy for transfusional iron overload. All these
guidelines are generally consistent with the ap-
proach outlined in this review, although there
are variations in the individual recommenda-
tions, depending in part on the year of publica-
tion and the specific underlying disorder. For
example, the guidelines for iron chelation in
thalassemia32,47 generally endorse measurement
of serum ferritin levels as a useful way to moni-
tor iron overload, whereas the guidelines for the
management of sickle cell disease45,46 emphasize
that ferritin levels can be altered by liver disease
and inflammation. The guidelines on the man-
agement of thalassemia by the Italian Society of
Hematology47 endorse deferoxamine as first-line
therapy over the oral chelators, whereas most of
the other guidelines do not state an explicit pref-
erence in this regard.
R ec om mendat ions
After discussing the need for iron-chelating ther-
apy with the patient and his family, I would de-
scribe the advantages and disadvantages of sub-
cutaneous deferoxamine and oral deferasirox so
that a genuinely informed decision can be made.
At present, the most frequent choice is oral de-
ferasirox. Before the initiation of treatment, I
would obtain information about the number of
previous transfusions and the rate of ongoing
transfusion and would arrange for cardiac T2*
and hepatic transverse relaxation rate (R2) mea-
surements, an MRI evaluation of cardiac function,
and echocardiographic and electrocardiographic
studies. Auditory and ophthalmic testing, includ-
ing slit-lamp examination and dilated-fundus ex-
amination, should be performed. Laboratory tests
should include a complete blood count with a
differential count and measurement of serum
creatinine, serum aminotransferases, bilirubin
levels, and iron indexes. After transfusion of a
total of 10 to 20 units of blood, or with the he-
patic iron concentration between 3 and 7 mg per
gram, I would administer once-daily oral therapy
nejm.org january 13, 2011
60
 clinical ther apeutics

with deferasirox at a dose of 20 mg per kilogram.
With good support and careful monitoring, this
regimen, adjusted as needed, should provide
long-term protection against the complications
of transfusional iron overload in this patient.
Supported by grants from the St. Giles Foundation, the Food
and Drug Administration (R01 FD003702), and the National In-
stitutes of Health (R37 DK049108; R01 DK066251).
Dr. Brittenham reports that his institution filed a patent
application in April 2010 for a rapid MRI method. No other
potential conflict of interest relevant to this article was re-
ported.
Disclosure forms provided by the author are available with the
full text of this article at NEJM.org.

References
1. Adams RJ, McKie VC, Hsu L, et al.
Prevention of a first stroke by transfu-
sions in children with sickle cell anemia
and abnormal results on transcranial
Doppler ultrasonography. N Engl J Med
1998;339:5-11.
2. Vichinsky EP, MacKlin EA, Waye JS,
Lorey F, Olivieri NF. Changes in the epide-
miology of thalassemia in North Ameri-
ca: a new minority disease. Pediatrics
2005;116(6):e818-e825.
3. Weatherall DJ. The inherited diseases
of hemoglobin are an emerging global
health burden. Blood 2010;115:4331-6.
4. Weatherall DJ, Clegg JB. The thalas-
semia syndromes. 4th ed. Oxford, Eng-
land: Blackwell Science, 2001.
5. Darbari DS, Kple-Faget P, Kwagyan J,
Rana S, Gordeuk VR, Castro O. Circum-
stances of death in adult sickle cell disease
patients. Am J Hematol 2006;81:858-63.
6. Finch CA, Lee MY, Leonard JM. Con-
tinuous RBC transfusions in a patient
with sickle cell disease. Arch Intern Med
1982;142:279-82.
7. Delea TE, Sofrygin O, Thomas SK,
Baladi JF, Phatak PD, Coates TD. Cost ef-
fectiveness of once-daily oral chelation
therapy with deferasirox versus infusional
deferoxamine in transfusion-dependent
thalassaemia patients: US healthcare sys-
tem perspective. Pharmacoeconomics
2007;25:329-42.
8. Delea TE, Edelsberg J, Sofrygin O, et
al. Consequences and costs of noncompli-
ance with iron chelation therapy in pa-
tients with transfusion-dependent thalas-
semia: a literature review. Transfusion
2007;47:1919-29.
9. Fillet G, Beguin Y, Baldelli L. Model of
reticuloendothelial iron metabolism in
humans: abnormal behavior in idiopathic
hemochromatosis and in inflammation.
Blood 1989;74:844-51.
10. Breuer W, Hershko C, Cabantchik ZI.
The importance of non-transferrin bound
iron in disorders of iron metabolism.
Transfus Sci 2000;23:185-92.
11. European Medicines Agency. Ferriprox
(deferiprone): summary of product charac-
teristics. European public assessment
report 2009, rev. ed. 14. (http://www.ema
.europa.eu/humandocs/Humans/EPAR/
ferriprox/ferriprox.htm.)
12. Novartis Pharmaceuticals. Desferal
(deferoxamine mesylate) prescribing infor-
mation. 2010. (http://www.desferal.com.)
13. Novartis Pharmaceuticals. Exjade (de-
ferasirox) prescribing information. 2010.
(http://www.exjade.com.)
14. De Domenico I, Ward DM, Kaplan J.
Specific iron chelators determine the
route of ferritin degradation. Blood 2009;
114:4546-51.
15. Waldmeier F, Bruin GJ, Glaenzel U, et
al. Pharmacokinetics, metabolism, and
disposition of deferasirox in beta-thalas-
semic patients with transfusion-depen-
dent iron overload who are at pharmaco-
kinetic steady state. Drug Metab Dispos
2010;38:808-16.
16. Barry M, Flynn DM, Letsky EA, Ris-
don RA. Long-term chelation therapy in
thalassaemia major: effect on liver iron
concentration, liver histology, and clini-
cal progress. BMJ 1974;2:16-20.
17. Modell B, Letsky EA, Flynn DM, Peto
R, Weatherall DJ. Survival and desferriox-
amine in thalassaemia major. Br Med J
(Clin Res Ed) 1982;284:1081-4.
18. Borgna-Pignatti C, Rugolotto S, De
Stefano P, et al. Survival and complica-
tions in patients with thalassemia major
treated with transfusion and deferox-
amine. Haematologica 2004;89:1187-93.
19. Cappellini MD, Cohen A, Piga A, et al.
A phase 3 study of deferasirox (ICL670), a
once-daily oral iron chelator, in patients
with beta-thalassemia. Blood 2006;107:
3455-62.
20. Piga A, Galanello R, Forni GL, et al.
Randomized phase II trial of deferasirox
(Exjade, ICL670), a once-daily, orally-
administered iron chelator, in comparison
to deferoxamine in thalassemia patients
with transfusional iron overload. Haema-
tologica 2006;91:873-80.
21. Porter J, Galanello R, Saglio G, et al.
Relative response of patients with myelo-
dysplastic syndromes and other transfu-
sion-dependent anaemias to deferasirox
(ICL670): a 1-yr prospective study. Eur J
Haematol 2008;80:168-76.
22. Vichinsky E, Onyekwere O, Porter J, et
al. A randomised comparison of defera-
sirox versus deferoxamine for the treat-
ment of transfusional iron overload in
sickle cell disease. Br J Haematol 2007;
136:501-8.
23. Cappellini MD, Porter J, El-Beshlawy
A, et al. Tailoring iron chelation by iron
intake and serum ferritin: the prospective
EPIC study of deferasirox in 1744 patients
with transfusion-dependent anemias. Hae-
matologica 2010;95:557-66.
24. Roberts DJ, Brunskill SJ, Doree C,
Williams S, Howard J, Hyde CJ. Oral de-
feriprone for iron chelation in people with
thalassaemia. Cochrane Database Syst Rev
2007;3:CD004839.
25. Jensen PD, Heickendorff L, Pedersen
B, et al. The effect of iron chelation on
haemopoiesis in MDS patients with trans-
fusional iron overload. Br J Haematol
1996;94:288-99.
26. Gattermann N. Overview of guide-
lines on iron chelation therapy in patients
with myelodysplastic syndromes and
transfusional iron overload. Int J Hema-
tol 2008;88:24-9.
27. Claster S, Wood JC, Noetzli L, et al.
Nutritional deficiencies in iron overload-
ed patients with hemoglobinopathies. Am
J Hematol 2009;84:344-8.
28. Davis BA, Porter JB. Long-term out-
come of continuous 24-hour deferoxamine
infusion via indwelling intravenous cathe-
ters in high-risk beta-thalassemia. Blood
2000;95:1229-36.
29. Cohen AR. Compassionate use of de-
feriprone for patients with thalassemia and
iron-induced heart disease. (ClinicalTrials
.gov identifier no. NCT00293098.) (http://
www.clinicaltrials.gov/ct2/show/
NCT00293098?term=deferiprone+
thalassemia&rank=2.)
30. Borgna-Pignatti C, Cappellini MD, De
Stefano P, et al. Cardiac morbidity and
mortality in deferoxamine- or deferiprone-
treated patients with thalassemia major.
Blood 2006;107:3733-7.
31. Anderson LJ, Westwood MA, Holden
S, et al. Myocardial iron clearance during
reversal of siderotic cardiomyopathy with
intravenous desferrioxamine: a prospec-
tive study using T2* cardiovascular mag-
netic resonance. Br J Haematol 2004;127:
348-55.
32. Iron overload. In: Cappellini MD, Co-
hen A, Eleftheriou A, Piga A, Porter J,
Taher A. Guidelines for the clinical man-
agement of thalassemia. 2nd ed. rev.
Nicosia, Cyprus: Thalassaemia Interna-
tional Federation, November 2008:33-63.
33. Olivieri NF, Buncic JR, Chew E, et al.
Visual and auditory neurotoxicity in pa-
tients receiving subcutaneous deferox-
amine infusions. N Engl J Med 1986;314:
869-73.
34. Cohen A, Martin M, Mizanin J, Kon-
kle DF, Schwartz E. Vision and hearing
during deferoxamine therapy. J Pediatr
1990;117:326-30.
35. De Sanctis V, Roos M, Gasser T, et al.

n engl j med 364;2 nejm.org january 13, 2011 155

BACK to TOC
61
 clinical ther apeutics

Impact of long-term iron chelation thera-
py on growth and endocrine functions in
thalassemia. J Pediatr Endocrinol Metab
2006;19:471-80.
36. Chan YL, Pang LM, Chik KW, et al.
Patterns of bone diseases in transfusion-
dependent homozygous thalassemia ma-
jor: predominance of osteoporosis and
desferrioxamine-induced bone dysplasia.
Pediatr Radiol 2002;32:492-7.
37. Olivieri NF, Brittenham GM. Iron-
chelating therapy and the treatment of
thalassemia. Blood 1997;89:739-61. [Er-
ratum, Blood 1997;89:2621.]
38. Beau-Salinas F, Guitteny MA, Dona-
dieu J, Jonville-Bera AP, Autret-Leca E.
High doses of deferiprone may be associ-
ated with cerebellar syndrome. BMJ 2009;
338:a2319.
39. Leitch HA. Controversies surround-
ing iron chelation therapy for MDS. Blood
Rev 2010 October 26 (Epub ahead of
print).
40. Novartis Pharmaceuticals. Myelodys-
plastic syndromes (MDS) event free sur-
vival with iron chelation therapy study
(TELESTO). (ClinicalTrials.gov identifier no.
NCT00940602.) (http://www.clinicaltrials
.gov/ct2/show/NCT00940602?term=telesto
&rank=1.)
41. Vichinsky E, Butensky E, Fung E, et al.
Comparison of organ dysfunction in
transfused patients with SCD or beta
thalassemia. Am J Hematol 2005;80:70-4.
42. Walter PB, Fung EB, Killilea DW, et al.
Oxidative stress and inflammation in
iron-overloaded patients with beta-thal-
assaemia or sickle cell disease. Br J Hae-
matol 2006;135:254-63.
43. Borgna-Pignatti C, Vergine G, Lom-
bardo T, et al. Hepatocellular carcinoma
in the thalassaemia syndromes. Br J Hae-
matol 2004;124:114-7.
44. Ko C, Siddaiah N, Berger J, et al. Prev-
alence of hepatic iron overload and asso-
ciation with hepatocellular cancer in end-
stage liver disease: results from the
National Hemochromatosis Transplant
Registry. Liver Int 2007;27:1394-401.
45. Division of Blood Diseases and Re-
sources NHLBI. The management of sick-
le cell disease. Bethesda, MD: National
Heart, Lung, and Blood Institute, 2002.
(http://www.nhlbi.nih.gov/health/prof/
blood/sickle/sc_mngt.pdf.)
46. Standards for the clinical care of adults
with sickle cell disease in the UK. London:
Sickle Cell Society, 2008. (http://www
.sicklecellsociety.org/pdf/CareBook.pdf.)
47. Angelucci E, Barosi G, Camaschella
C, et al. Italian Society of Hematology
practice guidelines for the management
of iron overload in thalassemia major and
related disorders. Haematologica 2008;
93:741-52.
48. Vlachos A, Ball S, Dahl N, et al. Diag-
nosing and treating Diamond Blackfan
anaemia: results of an international clini-
cal consensus conference. Br J Haematol
2008;142:859-76.
49. Marsh JC, Ball SE, Cavenagh J, et al.
Guidelines for the diagnosis and manage-
ment of aplastic anaemia. Br J Haematol
2009;147:43-70.
50. Bennett JM. Consensus statement on
iron overload in myelodysplastic syn-
dromes. Am J Hematol 2008;83:858-61.
Copyright 2011 Massachusetts Medical Society.

SPECIALTIES AND TOPICS AT NEJM.ORG

Specialty pages at the Journals Web site (NEJM.org) feature articles in cardiology,
endocrinology, genetics, infectious disease, nephrology, pediatrics, and many other
medical specialties. These pages, along with collections of articles on clinical and
nonclinical topics, offer links to interactive and multimedia content and feature
recently published articles as well as material from the NEJM archive (18121989).

156 n engl j med 364;2 nejm.org january 13, 2011

BACK to TOC
62
 The n e w e ng l a n d j o u r na l of m e dic i n e

clinical therapeutics

Bisphosphonates for Osteoporosis

Murray J. Favus, M.D.
This Journal feature begins with a case vignette that includes a therapeutic recommendation. A discussion
of the clinical problem and the mechanism of benefit of this form of therapy follows. Major clinical studies,
the clinical use of this therapy, and potential adverse effects are reviewed. Relevant formal guidelines,
if they exist, are presented. The article ends with the authors clinical recommendations.

A 67-year-old woman was referred by her primary care physician for treatment of os-
teoporosis and progressive bone loss. One year before the visit, the patient had dis-
continued hormone-replacement therapy. She had subsequently begun to experience
midback pain and lost 3.8 cm (1.5 in.) in height. A dual-energy x-ray absorptiometry
(DXA) scan showed bone mineral density T scores of 3.1 at the lumbar spine and 2.8
at the femoral neck, which are consistent with a diagnosis of osteoporosis. One year
later, a second scan showed a further decrease of 5.4% in bone mineral density at the
lumbar spine (Fig. 1), as well as a compression fracture of the 11th thoracic vertebra
(Fig. 2). Results of blood and urine tests ruled out the common secondary causes of
osteoporosis. To prevent additional vertebral fractures, oral bisphosphonate therapy
was recommended.

The Cl inic a l Probl em

Osteoporosis is a systemic skeletal disorder that is characterized by the loss of bone From the Department of Medicine, Uni-
tissue, disruption of bone architecture, and bone fragility, leading to an increased versity of Chicago, Chicago. Address re-
print requests to Dr. Favus at the Depart-
risk of fractures.1 Bone loss and low bone mass are asymptomatic until fractures ment of Medicine, University of Chicago,
occur. Estrogen deficiency after menopause is the most common cause of osteopo- 5841 S. Maryland Ave., Chicago, IL 60637, or
rosis, but secondary causes2 must be ruled out before treatment is undertaken at mfavus@medicine.bsd.uchicago.edu.
(Table 1). N Engl J Med 2010;363:2027-35.
Osteoporosis is the most common metabolic bone disease and the most common Copyright 2010 Massachusetts Medical Society.
cause of fractures in older adults in the United States. Ten million people in the
United States have osteoporosis, and an additional 33 million people have low bone
mass (osteopenia) and are at increased risk for fractures.4,5 More than 2 million
fractures occur each year as a result of osteoporosis or osteopenia, including
300,000 hip fractures, 547,000 vertebral fractures, and 135,000 pelvic fractures. Post-
menopausal white women have a 40% lifetime risk of at least one osteoporotic
fracture.4
Osteoporotic hip fractures are associated with the highest morbidity and mor-
tality. Up to 50% of patients with such fractures have permanently impaired mobil-
ity, and 25% lose the skills necessary to live independently.6,7 A recent meta-analysis
showed that among older men and women, the rate of death from any cause is
increased by a factor of 5 to 8 during the first 3 months after a hip fracture.8

Pathoph ysiol o gy a nd Effec t of Ther a py

Estrogen deficiency due to either spontaneous or surgical menopause9 increases the

production by bone marrow stromal cells and osteoblasts of the receptor activator
of nuclear factor B ligand (RANKL), which, in turn, increases the binding of
RANKL10 to the osteoclast cell-surface receptor nuclear factor B (RANK). Increased

n engl j med 363;21 nejm.org november 18, 2010 2027

BACK to TOC
63
 The n e w e ng l a n d j o u r na l
binding of RANKL to RANK initiates the prolif-
eration of osteoclast precursors and their differ-
entiation into mature osteoclasts.10-12 The expand-
ed osteoclast population increases bone turnover
and the depth and number of resorption pits
(Fig. 3). Later in the course of menopause, age-
related bone loss and accompanying changes in
the properties of bone material exacerbate the
bone loss and fragility associated with estrogen
deficiency.10 At the microscopical level, the in-
creased number and activity of osteoclasts disrupt
trabecular connectivity and increase cortical po-
rosity.9,11 Resorption pits are incompletely filled,
since osteoblastic new bone formation does not
keep pace with rates of bone resorption.10 Re-
duced bone density and bone quality compromise
the mechanical weight-bearing properties of the
skeleton and confer a predisposition to fractures
occurring either spontaneously or when falls cause
mechanical overload.11
Bisphosphonates reduce fractures by suppress-
ing bone resorption.12,13 The molecular structure
of the bisphosphonates (P-C-P) is analogous to
that of the naturally occurring pyrophosphates
(P-O-P), with two short side chains (R1 and R2)
attached to the C core.
Bisphosphonates Pyrophosphates
R1
O O O O
OH P C P OH OH P O P OH
OH OH OH OH
R2
The R1 side chain determines bone-binding af-
finity, and the R2 side chain determines antire-
sorption potency. Bisphosphonates that are ap-
proved for use in the United States (alendronate,
ibandronate, risedronate, and zoledronate) have
nitrogen-containing R2 side chains14 that en-
hance antiresorptive and antifracture potency.
Variations in the structure of the side chains deter-
mine the strength with which the biphospho-
nate binds to bone, the distribution through
bone, and the amount of time it remains in the
bone after treatment is discontinued.15
In bone, bisphosphonates accumulate in the
hydroxyapatite mineral phase, and the concentra-
tion of the bisphosphonates is increased by a factor
of 8 at sites of active bone resorption.14,16,17 The
bound nitrogen-containing bisphosphonates en-
ter osteoclasts and reduce resorption through inhi-
2028 n engl j med 363;21
BACK to TOC
of m e dic i n e
bition of farnesyl pyrophosphate synthase (FPPS),
an enzyme in the mevalonate-to-cholesterol path-
way.18,19 Inhibition of FPPS interferes with iso-
prenylation of small guanosine triphosphatases
(GTPases) at the ruffled border of the osteoclasts
and disrupts the attachment of osteoclasts to the
bone surface, which stops resorption and pro-
motes early cell death.16,20
Cl inic a l E v idence
Three of the most important phase 3 trials of the
use of bisphosphonates for the treatment of os-
teoporosis are described below. In these trials, a
reduction in the rate of fractures was the primary
end point, and increases in bone mineral density
at the lumbar spine and a reduction in markers of
bone turnover were secondary end points.
In the Fracture Intervention Trial (FIT),21 2027
postmenopausal women at high risk for fracture,
with low bone density at the femoral neck and at
least one vertebral fracture, were randomly as-
signed to either placebo or alendronate, at a dose
of 5 mg daily for 24 months, followed by 10 mg
daily for the final 12 months of the trial. At 36
months, 15.0% of the women who received the
placebo and 8.0% of the women who were treated
with alendronate had sustained one or more new
vertebral fractures, as assessed by radiography
(P = 0.001). New hip fractures occurred in 2.1%
of the women in the placebo group and 1.1% of
the women in the alendronate group (P = 0.05).
In the Vertebral Efficacy with Risedronate
Therapy (VERT) trial,22 2458 postmenopausal
women with at least one vertebral fracture and a
T score at the lumbar spine of 2.0 or less were
randomly assigned to either placebo or risedro-
nate at a dose of 2.5 mg or 5 mg daily. During the
course of the trial, data from other studies sug-
gested that a dose of 2.5 mg was less effective
than a dose of 5 mg; therefore the 2.5-mg group
was discontinued. In the two remaining groups,
the rate of new vertebral fractures after 3 years
was 11.3% among subjects treated with 5 mg of
risedronate daily, as compared with 16.3% in the
placebo group (P = 0.003). In a subsequent trial,
risedronate was shown to be effective in reducing
the rate of hip fractures as well.23
The efficacy of zoledronic acid in the treatment
of osteoporosis was evaluated in the Health Out-
comes and Reduced Incidence with Zoledronic
Acid Once Yearly trial (HORIZON; ClinicalTrials
.gov number, NCT00049829).24 In this trial,
nejm.org november 18, 2010
64
 clinical ther apeutics

A Anteroposterior Spine Bone Density B Reference: L1L4

1.44 2
Normal
1.32 1

Young-Adult T Score
1.19 0

BMD (g/cm2)
1.07 1
0.95 Osteopenia 2 Region BMD Young-Adult Age-Matched
g/cm2 % T score % z score
0.82 3 L1 0.805 71 2.7 81 1.5
0.70 4 L2 0.743 61 3.9 70 2.7
L3 0.706 59 4.1 67 2.9
Osteoporosis
0.58 5 L4 0.786 66 3.4 75 2.2
L1L4 0.758 64 3.5 73 2.3
0.00
0 20 30 40 50 60 70 80 90 100
Age (yr)

C Trend: L1L4
2
Change from Baseline (%)

1
0
1
2
3 Change from
Date Measured Age BMD Baseline
4 yr g/cm2 % %/yr
5 04/13/2005 66.7 0.758 4.9 5.4
05/14/2004 65.8 0.798 Baseline Baseline
6
0 65 66 67
Age (yr)

Figure 1. Dual-Energy X-Ray Absorptiometry (DXA) Scan of the Lumbar Spine.

Bone mineral density (BMD) was measured with the use of DXA in the 67-year-old woman described in the vignette (Panel A). The wom-
ans BMD is analyzed from lumbar spine L1 through L4 (Panel B). The BMD (rectangle) meets World Health Organization criteria for os-
teoporosis (T score of less than 2.5), as shown in the reference graph. The T score is the standard-deviation change in BMD from the
theoretical peak bone mass this woman had in her mid-20s to the current BMD. The z score is the standard-deviation difference be-
tween the mean BMD of a population matched for age, race, and sex and the patients current BMD. The current study (Panel C) shows
that her BMD (solid rectangle) is 5.4% lower than that indicated by the previous scan (open rectangle). A clinically significant change is
a change of more than 2.8%.

7765 postmenopausal women with osteoporosis
(T score of 2.5 or less or 1.5 or less with evi-
dence of vertebral fracture) were randomly as-
signed to either zoledronic acid, at a dose of 5 mg
administered at baseline, 12 months, and 24
months, or placebo. At 36 months, the absolute
rate of new vertebral fractures as assessed by stan-
dard radiography was 3.3% in the zoledronic acid
group, as compared with 10.9% in the placebo
group (P<0.001). There were 52 new hip fractures
(1.4%) in the zoledronic acid group, as compared
with 88 (2.5%) in the placebo group (P<0.001).
Randomized, placebo-controlled trials of other
oral bisphosphonates, including ibandronate,25
clodronate,26 and etidronate,27 have shown that
these drugs also have efficacy in reducing the risk
of new vertebral fractures. However, because these
trials were not powered to show efficacy for the
treatment of hip fractures, the clinical usefulness
of these agents for preventing hip fractures is
currently unknown. Pamidronate has been used
to treat a variety of bone diseases in children and
adults. However, no randomized, placebo-con-
trolled trial has been performed with sufficient
power to assess the efficacy of the drug for the
treatment of hip fracture in women with post-
menopausal osteoporosis.
Cl inic a l Use
All postmenopausal women with measurements
of bone mineral density at either the spine or the
hip that meet World Health Organization (WHO)
criteria for osteoporosis (T score of less than 2.5)

n engl j med 363;21 nejm.org november 18, 2010 2029

BACK to TOC
65
 The n e w e ng l a n d j o u r na l
A B
T4
T6
T11
T12
L4
Figure 2. Dual-Energy X-Ray Absorptiometry (DXA) Scan of the Thoracic
and Lumbar Spine.
Vertebral deformation at the thoracic and lumbar spine was assessed in the
patient described in the vignette, with the use of the vertebral-fracture as-
sessment on the same DXA scanner as that used for the measurement of
bone mineral density in Figure 1. Panel A shows thoracic kyphosis due to a
75% loss of height of T11. For comparison, Panel B shows a normal verte-
bral-fracture assessment in a 58-year-old woman without osteoporosis.
should receive long-term therapy with an agent
that has been proven to prevent fractures. In con-
trast, it may be difficult to decide whom to treat
among the large number of patients who have os-
teopenia (T score of 1.0 to 2.5). Many postmeno-
pausal women in whom fractures develop have
osteopenia rather than osteoporosis; in these wom-
en, the fractures may occur because of the contri-
butions of risk factors that are independent of bone
mineral density.28 I often use the WHO Fracture
Risk Assessment Tool (FRAX; www.sheffield.ac
.uk/FRAX/) to assist in making treatment deci-
sions. FRAX is a calculator algorithm that incor-
porates risk factors with measurements of bone
mineral density, generating a quantitative esti-
mate of the 10-year probability of a major osteo-
porotic fracture (hip, vertebral, humerus, or fore-
2030 n engl j med 363;21
BACK to TOC
of m e dic i n e
arm) or of a hip fracture alone in patients who
have not yet begun therapy. In general, I initiate
pharmacologic treatment in patients who have a
10-year probability of a hip fracture that exceeds
3% or a 10-year probability of a major osteopo-
rotic fracture that exceeds 20%.29
In addition to weighing the objective evidence,
I consider the patients lifestyle. I am more likely
to initiate treatment for low bone mass in a pa-
tient who wishes to continue participating in
sports or recreational activities such as cycling,
tennis, skiing, and running. Such patients are
likely to have a greater risk of falls and fractures
than are sedentary patients.
A major consideration in selecting therapy is
the risk of hip fracture. All treatments that have
been approved by the Food and Drug Administra-
tion (FDA) have shown efficacy in reducing the
rates of vertebral fracture, but not all have been
clearly shown to reduce the rate of hip fractures.
If bone mineral density at the hip is low, I usu-
ally select an agent for which there are trials
showing efficacy in preventing hip fractures. I
recommend either alendronate or risedronate if
the patient is capable of taking an oral agent. If
the patient cannot tolerate oral bisphosphonates,
then I may select intravenous zoledronic acid. If
bone density at the hip is normal or only mildly
reduced, I may select oral or intravenous ibandro-
nate, which has not been shown to be effective
in reducing the risk of hip fracture.
Alternatives to bisphosphonates include the
anabolic agent teriparatide (parathyroid hormone
1-34), which reduces the risk of vertebral and non-
vertebral fractures but, among subjects in a large,
pivotal trial, did not reduce the risk of hip frac-
ture alone.30 Teriparatide is also more expensive
than the bisphosphonates and requires daily sub-
cutaneous injection. Estrogen is effective in de-
creasing the risk of vertebral and hip fractures in
postmenopausal women31 but may confer in-
creased risks of breast cancer and cardiovascular
disease. Raloxifene is an oral selective estrogen-
receptor modulator (SERM) that decreases the risk
of vertebral fractures by 40 to 49%, but it may not
reduce the risk of nonvertebral fractures.32 Cal-
citonin administered by means of a nasal spray
is an antiresorptive agent that has limited effi-
cacy in reducing the risk of vertebral fractures
and lacks efficacy in preventing hip fracture.33
Oral bisphosphonates must be taken after an
overnight fast either once weekly (alendronate at
a dose of 70 mg or risedronate at a dose of 35 mg),
nejm.org november 18, 2010
66
 clinical ther apeutics
once monthly (ibandronate at a dose of 150 mg
or risedronate at a dose of 150 mg), or on 2 con-
secutive days once monthly (risedronate at a dose
of 75 mg). The tablets are taken with 6 to 8 oz of
tap water. The patient should remain upright for
at least 30 minutes after taking the drug to mini-
mize gastroesophageal reflux. To optimize ab-
sorption, food, medications, and liquids other
than tap or filtered water should be avoided for
at least 30 to 45 minutes to allow for dissolution
of the tablet and gastric emptying.
Intravenous bisphosphonates include ibandro-
nate (at a dose of 3 mg every 3 months) and zole-
dronic acid (at a dose of 5 mg every 12 months).
They are usually administered in an outpatient
facility that has the resources for administering
and monitoring intravenous infusions.
Oral and intravenous bisphosphonates are con-
traindicated in patients who have had a prior al-
lergic reaction to a bisphosphonate or who have
an estimated creatinine clearance of 35 ml per
minute or less, vitamin D depletion (serum 25-
hydroxyvitamin D levels should be more than 30 ng
per milliliter before initiating bisphosphonates),
osteomalacia (vitamin D depletion or deficiency
causing defective mineralization), or hypocalce-
mia. Oral bisphosphonates are contraindicated in
patients who have impaired swallowing or esoph-
ageal disorders such as achalasia, esophageal vari-
ces, or severe gastroesophageal reflux or who are
unable to sit up for at least 30 minutes after tak-
ing the medication. There are no known interac-
tions between bisphosphonates and other medi-
cations.
After initiating bisphosphonate therapy, I typi-
cally reevaluate the patient in 1 month to assess
tolerance and thereafter at 3 months, 6 months,
and 1 year. At 3 months and 6 months, I obtain
measurements of bone-turnover markers, such as
osteocalcin or serum C-terminal telopeptide of
type 1 collagen (CTX). At 1 year, and every 2 years
thereafter, I repeat the assessment of bone min-
eral density with the use of DXA. An increase in
bone mineral density is not required for a therapy
to be considered effective, but a substantial de-
cline in bone mineral density requires further
evaluation.
Poor adherence to therapy should be suspected
if the patient has an otherwise unexplained de-
cline in bone mineral density, a new fracture,
continued bone loss, or high rates of bone turn-
over that persist after 12 months of therapy. When
I suspect poor adherence, I ask the patient wheth-
n engl j med 363;21
BACK to TOC
Table 1. Common Secondary Causes of Osteoporosis and Laboratory Evalua-
tions.*
Possible Cause of Osteoporosis Laboratory Test
Vitamin D deficiency Measurement of serum 25-hydroxyvitamin
D level
Primary hyperparathyroidism Measurement of fasting serum calcium
and parathyroid hormone levels
Celiac disease Measurement of serum tissue transgluta-
minase, total IgA, and gliadin levels
Idiopathic hypercalciuria Measurement of 24-hour urine calcium
excretion after discontinuation of
calcium supplements
Hyperthyroidism Measurement of serum thyrotropin and
total thyroxine levels
Myeloma Serum and urine immunoelectrophoresis
* Additional information regarding secondary causes of osteoporosis can be
found in Tannenbaum et al.2 and Jamal et al.3
er he or she has had any side effects and attempt
to document the patients use of the drug by
measuring markers of bone turnover. Evidence of
treatment failure in a patient with good adher-
ence to an oral bisphosphonate regimen requires
a change to either intravenous zoledronic acid or
another class of medications such as anabolic
agents (e.g., teriparatide).
The optimal duration of bisphosphonate ther-
apy remains unresolved. However, on the basis
of available data, it seems likely that discontinu-
ing therapy after 5 years, at least for a temporary
drug holiday, is not harmful and may be advan-
tageous.34 Patients with mildly reduced bone min-
eral density may be the most suitable candidates
for a 1-year to 2-year drug holiday, because the
risk of fracture will be low if bone loss occurs
while the person is not receiving therapy.
Generic alendronate was introduced in 2008
and is less expensive than other agents, with cost
ranging from $4 to $40 per month. The cost of
risedronate ranges from $60 to $120 per month;
generic risedronate will become available in the
near future. The cost of oral ibandronate ranges
from $90 to $130 per month. One infusion of zole-
dronic acid is estimated to cost $1,300; intrave-
nous ibandronate costs about $1,300 per year.
A dv er se Effec t s
An acute-phase reaction characterized by fever,
myalgia, bone pain, and weakness35 occurs in 20%
of patients after an initial intravenous infusion of
nejm.org november 18, 2010 2031
67
 The n e w e ng l a n d j o u r na l of m e dic i n e

Figure 3. Cellular Elements Involved in Postmenopausal Trabecular Bone Turnover before and during Bisphosphonate Therapy.
Panel A shows the untreated postmenopausal state, in which osteoclast-mediated bone resorption occurs at a high rate, exceeding os-
teoblast-driven bone formation and leading to net bone loss. Panel B shows the initial events associated with bisphosphonate therapy,
including the localization and concentration of bisphosphonates in bone through binding to sites of active bone resorption. Panel C
shows the effects of bisphosphonates after 6 months of therapy. The number of osteoclasts has decreased owing to early apoptosis. As
a result, bone resorption is decreased, and osteoblasts and bone formation are also decreased. The bisphosphonate concentration is re-
duced around previous resorption pits. A lower steady-state rate of bone turnover, similar to premenopausal rates, is established.

bisphosphonates and in a very small number of
patients during oral therapy. Erosive esophagitis,
ulceration, and bleeding have been associated with
daily oral alendronate or risedronate therapy but
occur rarely with current (nondaily) regimens.
Heartburn, chest pain, hoarseness, and vocal-cord
irritation36 may occur with weekly (alendronate
or risedronate) or monthly (ibandronate or rised-
ronate) therapy. A relationship between esopha-
geal cancer and oral bisphosphonates, suggested
on the basis of a small number of case reports,
has not been substantiated.37
Transient renal toxic effects can occur after
rapid intravenous administration.38,39 Slow infu-
sion rates (no less than 15 minutes) and lower
doses minimize peak drug serum levels and the
risk of renal damage. Bisphosphonates are not
recommended when creatinine clearance is less
than 35 ml per minute.38 Dose reductions may be
required for patients with stage III chronic kid-
ney disease (as defined by an estimated glomeru-
lar filtration rate between 59 and 30 ml per minute
per 1.73 m2 of body-surface area). Mild transient
hypocalcemia is a rare complication of intrave-
nous bisphosphonate therapy that may require an
interruption in treatment,40 but once the serum
calcium level has returned to the normal range,
therapy can be resumed. Severe hypocalcemia is
a contraindication for continued administration.
Osteonecrosis of the jaw is a rare but serious
complication of long-term bisphosphonate ther-
apy that may appear either spontaneously or after
an oral surgical procedure. Exposed mandibu-
lar or maxillary dead bone, nonhealing mucosa,
and chronic infection may persist for weeks to
years.41-43 More than 95% of cases of osteone-
crosis of the jaw occur in patients who are re-
ceiving zoledronic acid or pamidronate for the
treatment of myeloma, breast cancer, or other
bone cancers at doses 10 to 12 times as high as
those used for the treatment of osteoporosis.42,44
Case reports suggest that atypical femoral frac-
tures (in the subtrochanteric and mid-diaphyseal
portions of the femur) may be more common
during bisphosphonate therapy.37,45 Recent data
from a cross-sectional study of femur fractures
recorded in the Danish national health registry46
and a pooled post hoc analysis of the trials that

2032 n engl j med 363;21 nejm.org november 18, 2010

BACK to TOC
68
 clinical ther apeutics

studied the effects of alendronate and zoledron- Gynecologists,54 and the North American Meno-
ic acid on the incidence of fractures47 showed no pause Society6 agree that persons with osteopo-
relationship between the use of bisphosphonates rosis (bone mineral density T score of less than
and atypical femur fractures. However, these re- 2.5) or low bone mass and hip or vertebral frac-
ports are not definitive, and the possibility of a tures should receive treatment. These guidelines
relationship continues to be investigated. also suggest that persons with T scores higher
than 1.5 should not receive therapy unless there
A r e a s of Uncer ta in t y is clinical evidence of osteoporosis. Thus, contro-
versy remains regarding the indications for treat-
The optimal duration of bisphosphonate therapy ment among people with mild reductions in bone
remains uncertain. Recent retrospective studies density. The guidelines include oral bisphospho-
and case reports suggest that long-term bisphos- nates among the first-line therapies for osteopo-
phonate therapy may result in the suppression of rosis but do not name specific FDA-approved drugs.
bone turnover and confer a predisposition to in-
creased bone fragility, with an increased risk for R ec om mendat ions
atypical femur fractures.37 Markers of bone turn-
over underestimate the extent of suppressed bone The patient described in the vignette is at high
formation,12,48 and their usefulness in monitoring risk for additional fractures on the basis of her
long-term safety may therefore be limited. An ac- history of vertebral compression fracture and a
cumulation of microcracks in bone-biopsy speci- bone mineral density T score in the osteoporosis
mens was found in one study of patients receiv- range. A drug with efficacy in preventing hip and
ing alendronate therapy when the analysis was spinal fractures is required, and I would treat the
adjusted for potential confounders such as age patient with either alendronate or risedronate for
and bone mineral density at the femoral neck49 5 years. After 5 years of treatment, I would decide
but not in another study of long-term alendro- whether a drug holiday might be appropriate for
nate therapy (mean, 6.5 years).50 Prospective stud-this patient, taking into consideration the fact
ies are needed to estimate the long-term risk of that she is at high risk for recurrent fracture. I
side effects associated with bisphosphonate ther- would suggest a calcium intake of 1200 mg per
apy, including osteonecrosis of the jaw and atyp- day from dietary sources, with calcium supple-
ical femur fractures. Until a better estimate of the
ments as a second choice. I would also measure
risk of these complications emerges, one must the serum 25-hydroxyvitamin D level and select
balance the long-term risk of these uncommon an appropriate level of vitamin D intake, encour-
complications against the known efficacy of the age regular weight-bearing exercise, and empha-
agents in reducing rates of common osteoporotic size the importance of adherence to procedures
fractures. It is also not known whether these com- for taking the medication. I would use measure-
plications can be minimized by periodic rotation ments of bone mineral density to monitor her re-
of treatment from one class of agents to another. sponse to therapy 12 months after treatment is
initiated and then at 24-month intervals as needed.
Guidel ine s A decline in bone mass or another low-trauma
fracture would require careful review of the treat-
Guidelines for the management of osteoporosis ment plan and possible selection of another agent.
published by the National Osteoporosis Founda- Dr. Favus reports receiving honoraria and consulting fees
tion,51 the American Association of Clinical En- from CVS Caremark and Amgen. No other potential conflict of
interest relevant to this article was reported.
docrinologists,52 the American College of Physi- Disclosure forms provided by the author are available with the
cians,53 the American College of Obstetricians and full text of this article at NEJM.org.

References
1. NIH Consensus Development Panel
on Osteoporosis Prevention, Diagnosis, and
Therapy. Osteoporosis prevention, diag-
nosis, and therapy. JAMA 2001;285:785-95.
2. Tannenbaum C, Clark J, Schwartzman
K, et al. Yield of laboratory testing to iden-
tify secondary contributors to osteoporo-
sis in otherwise healthy women. J Clin
Endocrinol Metab 2002;87:4431-7.
3. Jamal SA, Leiter RE, Bayoumi AM,
Bauer DC, Cummings SR. Clinical utility
of laboratory testing in women with os-
teoporosis. Osteoporos Int 2005;16:534-40.
4. Bone health and osteoporosis: a re-
port of the Surgeon General. Rockville,

n engl j med 363;21 nejm.org november 18, 2010 2033

BACK to TOC
69
 The n e w e ng l a n d j o u r na l
MD: Department of Health and Human
Services, 2004.
5. Advisory Group on Osteoporosis. Re-
port. London: Department of Health, 1994.
6. Management of osteoporosis in post-
menopausal women: 2010 position state-
ment of the North American Menopause
Society. Menopause 2010;17:25-54.
7. Cooper C, Atkinson EJ, Jacobsen SJ,
OFallon WM, Melton LJ III. Population-
based study of survival after osteoporotic
fractures. Am J Epidemiol 1993;137:1001-5.
8. Haentjens P, Magaziner JS, Coln-
Emeric CS, et al. Meta-analysis: excess
mortality after hip fracture among older
women and men. Ann Intern Med 2010;
152:380-90.
9. Raisz LG. Pathogenesis of osteoporo-
sis: concepts, conflicts, and prospects.
J Clin Invest 2005;115:3318-25.
10. Khosla S. Pathogenesis of osteoporo-
sis. In: Robertson RP, ed. Translational
endocrinology & metabolism: osteoporo-
sis update. Chevy Chase, MD: The Endo-
crine Society, 2010:55-86. (http://www
.endojournals.org/translational/issues/
osteoporosis.pdf.)
11. Seeman E, Delmas PD. Bone quality
the material and structural basis of
bone strength and fragility. N Engl J Med
2006;354:2250-61.
12. Hofbauer LC, Schoppet M. Bone biol-
ogy underlying therapeutic approaches. In:
Robertson RP, ed. Translational endocri-
nology & metabolism: osteoporosis update.
Chevy Chase, MD: The Endocrine Society,
2010:117-48. (http://www.endojournals.org/
translational/issues/osteoporosis.pdf.)
13. Rodan G, Reszka A, Golub E, Rizzoli
R. Bone safety of long-term bisphospho-
nate treatment. Curr Med Res Opin 2004;
20:1291-300.
14. McClung MR. Bisphosphonates. En-
docrinol Metab Clin North Am 2003;32:
253-71.
15. Russell RG, Watts NB, Ebetino FH,
Rogers MJ. Mechanisms of action of
bisphosphonates: similarities and differ-
ences and their potential influence on clini-
cal efficacy. Osteoporos Int 2008;19:733-
59.
16. Sato M, Grasser W, Endo N, et al.
Bisphosphonate action: alendronate local-
ization in rat bone and effects on osteo-
clast ultrastructure. J Clin Invest 1991;
88:2095-105.
17. Masarachia P, Weinreb M, Balena R,
Rodan GA. Comparison of the distribu-
tion of 3H-alendronate and 3H-etidronate
in rat and mouse bones. Bone 1996;19:
281-90.
18. Bergstrom JD, Bostedor RG, Masara-
chia PJ, Reszka AA, Rodan GA. Alendro-
nate is a specific, nanomolar inhibitor of
farnesyl diphosphate synthase. Arch Bio-
chem Biophys 2000;373:231-41.
19. Fisher JE, Rogers MJ, Halasy JM, et al.
Alendronate mechanism of action: gera-
nylgeraniol, an intermediate in the meval-
2034 n engl j med 363;21
BACK to TOC
of m e dic i n e
onate pathway, prevents inhibition of os-
teoclast formation, bone resorption, and
kinase activation in vitro. Proc Natl Acad
Sci U S A 1999;96:133-8.
20. Luckman SP, Hughes DE, Coxon FP,
Graham R, Russell RG, Rogers MJ. Nitro-
gen-containing bisphosphonates inhibit
mevalonate pathway and prevent post-
translational prenylation of GTP-binding
proteins, including Ras. J Bone Miner Res
1998;13:581-9.
21. Black DM, Cummings SR, Karpf DB,
et al. Randomised trial of effect of alen-
dronate on risk of fracture in women with
existing vertebral fractures. Lancet 1996;
348:1535-41.
22. Harris ST, Watts NB, Genant HK, et
al. Effects of risedronate treatment on
vertebral and nonvertebral fractures in
women with postmenopausal osteoporo-
sis: a randomized controlled trial. JAMA
1999;282:1344-52.
23. McClung MR, Geusens P, Miller PD, et
al. Effect of risedronate on the risk of hip
fracture in elderly women. N Engl J Med
2001;344:333-40.
24. Black DM, Delmas PD, Eastell R, et al.
Once-yearly zoledronic acid for treatment
of postmenopausal osteoporosis. N Engl J
Med 2007;356:1809-22.
25. Chestnut CH III, Skag A, Christiansen
C, et al. Effects of oral ibandronate ad-
ministered daily or intermittently on frac-
ture risk in postmenopausal osteoporo-
sis. J Bone Miner Res 2004;19:1241-9.
26. McCloskey E, Selby P, Davies M, et al.
Clodronate reduces vertebral fracture risk
in women with postmenopausal or sec-
ondary osteoporosis: results of a double-
blind, placebo-controlled 3-year study.
J Bone Miner Res 2004;19:728-36.
27. Wells GA, Cranney A, Peterson J, et al.
Etidronate for the primary and secondary
prevention of osteoporotic fractures in
postmenopausal women. Cochrane Data-
base Syst Rev 2008;1:CD003376.
28. Cranney A, Tugwell P, Adachi J, et al.
Meta-analyses of therapies for postmeno-
pausal osteoporosis. III. Meta-analysis of
risedronate for the treatment of post-
menopausal osteoporosis. Endocr Rev
2002;23:517-23.
29. Dawson-Hughes B, Looker AC, Toste-
son AN, Johansson H, Kanis JA, Melton LJ
III. The potential impact of new National
Osteoporosis Foundation guidance on
treatment patterns. Osteoporos Int 2010;
21:41-52.
30. Neer RM, Arnaud CD, Zanchetta JR,
et al. Effect of parathyroid hormone (1-34)
on fractures and bone mineral density in
postmenopausal women with osteoporo-
sis. N Engl J Med 2001;344:1434-41.
31. Cauley JA, Robbins J, Chen Z, et al.
Effects of estrogen plus progestin on risk
of fracture and bone mineral density: the
Womens Health Initiative randomized
trial. JAMA 2003;290:1729-38.
32. Cranney A, Tugwell P, Zytaruk N, et al.
nejm.org november 18, 2010
Meta-analyses of therapies for postmeno-
pausal osteoporosis. IV. Meta-analysis of
raloxifene for the prevention and treat-
ment of postmenopausal osteoporosis.
Endocr Rev 2002;23:524-8.
33. Cranney A, Tugwell P, Zytaruk N, et
al. Meta-analyses of therapies for post-
menopausal osteoporosis. VI. Meta-analy-
sis of calcitonin for the treatment of post-
menopausal osteoporosis. Endocr Rev
2002;23:540-51.
34. Black DM, Schwartz AV, Ensrud KE, et
al. Effects of continuing or stopping alen-
dronate after 5 years of treatment: the
Fracture Intervention Trial Long-term Ex-
tension (FLEX): a randomized trial. JAMA
2006;296:2927-38.
35. Adami S, Bhalla AK, Dorizzi R, et al.
The acute-phase response after bisphos-
phonate administration. Calcif Tissue Int
1987;41:326-31.
36. Ribeiro A, DeVault KR, Wolfe JT III,
Stark ME. Alendronate-associated esoph-
agitis: endoscopic and pathologic fea-
tures. Gastrointest Endosc 1998;47:525-8.
37. Watts NB, Diab DL. Long-term use of
bisphosphonates in osteoporosis. J Clin
Endocrinol Metab 2010;95:1555-65.
38. Lewiecki EM, Miller PD. Renal safety
of intravenous bisphosphonates in the
treatment of osteoporosis. Expert Opin
Drug Saf 2007;6:663-72.
39. Boonen S, Sellmeyer DE, Lippuner K,
et al. Renal safety of annual zoledronic
acid infusions in osteoporotic postmeno-
pausal women. Kidney Int 2008;74:641-8.
40. Maalouf NM, Heller HJ, Odvina CV,
Kim PJ, Sakhaee K. Bisphosphonate-
induced hypocalcemia: report of 3 cases
and review of the literature. Endocr Pract
2006;12:48-53.
41. Ruggiero SL, Mehrotra B, Rosenberg
TJ, Engroff SL. Osteonecrosis of the jaws
associated with the use of bisphospho-
nates: a review of 63 cases. J Oral Maxil-
lofac Surg 2004;62:527-34.
42. Khosla S, Burr D, Cauley J, et al.
Bisphosphonate-associated osteonecrosis
of the jaw: report of a task force of the
American Society for Bone and Mineral
Research. J Bone Miner Res 2007;22:1479-
91.
43. Bilezikian JP. osteonecrosis of the jaw
do bisphosphonates pose a risk? N Engl
J Med 2006;355:2278-81.
44. Woo SB, Hellstein JW, Kalmar JR. Sys-
temic review: bisphosphonates and osteo-
necrosis of the jaws. Ann Intern Med 2006;
144:753-61. [Erratum, Ann Intern Med
2006;145:235.]
45. Shane E. Evolving data about subtro-
chanteric fractures and bisphosphonates.
N Engl J Med 2010;362:1825-7.
46. Abrahamsen B, Eiken P, Eastell R.
Subtrochanteric and diaphyseal femur
fractures in patients treated with alendro-
nate: a register-based national cohort
study. J Bone Miner Res 2009;24:1095-
102.
70
 clinical ther apeutics

47. Black DM, Kelly MP, Genant HK, et al.
Bisphosphonates and fractures of the sub-
trochanteric or diaphyseal femur. N Engl
J Med 2010;362:1761-71.
48. Odvina CV, Zerwekh JE, Rao DS,
Maalouf N, Gottschalk FA, Pak CYC. Se-
verely suppressed bone turnover: a poten-
tial complication of alendronate therapy.
J Clin Endocrinol Metab 2005;90:1294-301.
49. Stepan JJ, Burr DB, Pavo I, et al. Low
bone mineral density is associated with
bone microdamage accumulation in post-
menopausal women with osteoporosis.
Bone 2007;41:378-85.
50. Chapurlat RD, Arlot M, Burt-Pichat B,
et al. Microcrack frequency and bone re-
modeling in postmenopausal osteoporotic
women on long-term bisphosphonates: a
bone biopsy study. J Bone Miner Res 2007;
22:1502-9. [Erratum, J Bone Miner Res
2008;23:1153.]
51. Clinicians guide to prevention and
treatment of osteoporosis. Washington,
DC: National Osteoporosis Foundation,
2008.
52. Hodgson SF, Watts NB, Bilezikian JP,
et al. American Association of Clinical
Endocrinologists medical guidelines for
clinical practice for the prevention and
treatment of postmenopausal osteoporo-
sis: 2001 edition, with selected updates for
2003. Endocr Pract 2003;9:544-64. [Erra-
tum, Endocr Pract 2004;10:90.]
53. Qaseem A, Snow V, Shekelle P, Hop-
kins R Jr, Forciea MA, Owens DK. Phar-
macologic treatment of low bone density
or osteoporosis to prevent fractures: a clini-
cal practice guideline from the American
College of Physicians. Ann Intern Med
2008;149:404-15.
54. American College of Obstetricians
and Gynecologists Womens Health Care
Physicians. Osteoporosis. Obstet Gynecol
2004;104:Suppl:66S-76S.
Copyright 2010 Massachusetts Medical Society.

JOURNAL ARCHIVE AT NEJM.ORG

Every issue of the Journal is now available at NEJM.org, beginning with the first
article published in January 1812. The entire archive is fully searchable, and browsing
of titles and tables of contents is easy and available to all. Individual subscribers are
entitled to free 24-hour access to 50 archive articles per year. Access to content in
the archive is available on a per-article basis and is also being provided through
many institutional subscriptions.

n engl j med 363;21 nejm.org november 18, 2010 2035

BACK to TOC
71

From the Department of Ophthalmology
and Visual Sciences (J.C.F., E.M.S.) and
the Howard Hughes Medical Institute
(E.M.S.), University of Iowa Carver Col-
lege of Medicine, Iowa City. Address re-
print requests to Dr. Stone at the Depart-
ment of Ophthalmology and Visual
Sciences, University of Iowa, 375 Newton
Rd., 4111 MERF, Iowa City, IA 52242, or at
edwin-stone@uiowa.edu.
N Engl J Med 2010;363:1648-55.
Copyright 2010 Massachusetts Medical Society.
1648
BACK to TOC
The n e w e ng l a n d j o u r na l of m e dic i n e
clinical therapeutics
Ranibizumab Therapy for Neovascular
Age-Related Macular Degeneration
James C. Folk, M.D., and Edwin M. Stone, M.D., Ph.D.
This Journal feature begins with a case vignette that includes a therapeutic recommendation. A discussion
of the clinical problem and the mechanism of benefit of this form of therapy follows. Major clinical studies,
the clinical use of this therapy, and potential adverse effects are reviewed. Relevant formal guidelines,
if they exist, are presented. The article ends with the authors clinical recommendations.
A
n ophthalmologist refers a 66-year-old man for consultation
with a retinal specialist. The patient has had blurred and distorted vision in
his right eye for the past 2 weeks. Examination reveals drusen beneath the
retina in both eyes. His right eye also has a subretinal hemorrhage just temporal to
the center of the macula, as well as subretinal fluid (Fig. 1A). A fluorescein angio-
gram shows subfoveal neovascularization (Fig. 1B). Optical coherence tomography
reveals fluid beneath and within the layers of the retina (Fig. 1C). The retinal spe-
cialist diagnoses neovascular age-related macular degeneration (AMD) and recom-
mends intraocular injections of ranibizumab.
The Cl inic a l Probl em
AMD is the leading cause of blindness in the United States. All patients initially
have a form of the disorder called dry AMD, characterized by the development and
accumulation of drusen, which are localized deposits of extracellular material that
appear as yellow spots in the retina on ophthalmoscopy. As dry AMD progresses,
focal areas of atrophy of the retinal pigment epithelium appear. Wet, or neovascu-
lar, AMD then develops in some patients with established dry AMD. Wet AMD is
characterized by the growth of abnormal vessels beneath the retinal pigment epi-
thelium and between the retinal pigment epithelium and the overlying retina.
The Eye Diseases Prevalence Research Group estimates that in the year 2000,
a total of 1.2 million residents of the United States had neovascular AMD, 973,000
had dry AMD with atrophy of the retinal pigment epithelium, and 7.3 million had
large drusen (125 m in diameter) and were therefore at increased risk for atrophy
or neovascularization. The same group estimates that these numbers will increase
by more than 50% by the year 2020 as a result of aging of the population.1
Many patients with AMD have moderate vision loss, between 20/50 and 20/100
in the better eye. These patients have quality-of-life measurements that are 32%
below normal, similar to those among patients with severe angina or hip fractures.2
A person with very severe neovascular disease may have 20/800 vision in the better
eye and will have a reduction in quality of life of 60%, similar to that of a patient
who is bedridden with a catastrophic stroke.2 AMD is estimated to cost the United
States $30 billion a year.2
Pathoph ysiol o gy a nd Effec t of Ther a py
Despite decades of intensive investigation, the molecular mechanisms underlying
the pathogenesis of AMD are still obscure. It is likely that in most patients, a num-
n engl j med 363;17 nejm.org october 21, 2010
72
 clinical ther apeutics

ber of genetic factors3-12 and environmental fac-

tors (e.g., smoking)13 contribute incrementally to A
the development of the disease. Genomewide as-
sociation studies of patients with AMD and con-
trol subjects have shown that the largest single
genetic factor contributing to AMD is a variant of
codon 402 in the gene encoding complement fac-
tor H. This observation strongly supports the long-
held belief that the immune system is an impor-
tant contributor to AMD.3-5,14,15
The photoreceptor cells of the retina depend
on the underlying retinal pigment epithelium for
phagocytosis of their continuously renewed out-
er segments, as well as for reisomerization of their
light-sensitive, vitamin-Abased chromophore16
B
(Fig. 2A). Both the photoreceptors and the reti-
nal pigment epithelium depend on the chorio-
capillaris for oxygen, nutrients, and removal of
metabolic waste products.17 In patients with AMD,
drusen form between the retinal pigment epithe-
lium and Bruchs membrane. In some patients
with AMD, there is apoptotic atrophy of the reti-
nal pigment epithelium and choriocapillaris in the
macula (the central region of the fundus).14,18
Neovascularization occurs in about 10% of
patients with AMD for reasons that are still un-
known but that may be due in part to injury or
degeneration of Bruchs membrane. This compli-
cation is responsible for most (perhaps as much C
as 90%) of the severe vision loss caused by this
disease.19 In neovascular AMD, abnormal capil-
laries grow from the choroid through Bruchs
membrane and into spaces beneath the retinal
pigment epithelium and retina (Fig. 2B). As these
Figure 1. Clinical Characteristics of Neovascular
vessels proliferate, they leak serum or blood, Age-Related Macular Degeneration.
which causes swelling beneath and within the A color photograph of the right eye at presentation
retina and loss of visual acuity. If left unchecked, shows scattered yellow drusen, multilayered hemor-
the vessels eventually cause a subretinal fibrotic rhage, and serous fluid in the macula (Panel A). A late-
scar and permanent loss of vision.20,21 phase fluorescein angiogram of the right eye shows hy-
Vascular endothelial growth factor A (VEGF-A) perfluorescence of occult choroidal neovascularization
involving the center of the macula (Panel B). The dark
is closely associated with the growth and perme- area to the left is caused by hemorrhage, which blocks
ability of neovascular vessels.22-26 Ranibizumab the underlying fluorescence. A time-domain optical co-
is the Fab fragment of a recombinant, human- herence tomogram taken through the center of the
ized, monoclonal antibody that binds to all right macula reveals fluid (arrow) within and beneath
forms of VEGF-A. The inhibition of VEGF-A re- the retina (Panel C).
duces the permeability of the neovascular ves-
sels, as well as their further growth.27 In tu- and leakage are stopped, the retinal swelling
mors, long-term VEGF blockade results in vessel usually diminishes and vision can stabilize or
maturation and remodeling, decreased numbers improve. Studies in primate models have shown
of endothelial cells, increased coverage of the that when ranibizumab is injected into the eye,
vessel wall by pericytes, and a more stable, non- effective retinal concentrations are maintained
leaking vessel28 (Fig. 2C). When vessel growth for about 1 month.27

n engl j med 363;17 nejm.org october 21, 2010 1649

BACK to TOC
73

A Normal Retina
Photoreceptors
Retina
Retinal pigment epithelium
Choroid Bruchs membrane
B Neovascular AMD
Fluid accumulation
within retinal layers
Fluid accumulation
beneath the retina
Fluid leakage from
immature vessels
C Ranibizumab Therapy
Reattachment of photoreceptors
to retinal pigment epithelium
Shrinkage and
maturation of vessels
1650
BACK to TOC
The n e w e ng l a n d j o u r na l of m e dic i n e
Figure 2. Therapeutic Action of Ranibizumab.
In the normal retina, the retinal pigment epithelium
is attached to Bruchs membrane and the choroid sup-
plies oxygen and nutrients to the outer layers of the
retina (Panel A). In patients with neovascular age-related
macular degeneration (AMD), new blood vessels grow
from choroidal capillaries through Bruchs membrane
and proliferate beneath the retinal pigment epithelium
(Panel B). Vascular endothelial growth factor A (VEGF-A)
binds with its receptor and promotes vascular growth
and leakage. Fluid or blood accumulates beneath the
retina, detaching it from the retinal pigment epithelium.
Fluid can also accumulate within the retinal layers. Rani-
bizumab binds to VEGF-A and prevents it from binding
to its receptor (Panel C). Long-term blockade of VEGF-A
causes shrinkage and maturing of the vessels so that
Choriocapillaris
they no longer leak. The accumulation of fluid within
and beneath the retina resolves, and the photoreceptors
reattach to the underlying retinal pigment epithelium.
Cl inic a l E v idence
Two main patterns of choroidal neovasculari-
zation are seen on fluorescein angiography in
patients with AMD. Classic neovascularization
fluoresces brightly in the early phases of the
angiogram and leaks profusely in the late phas-
VEGF-A es, whereas occult neovascularization fills more
VEGF-A slowly and leaks much less.29 Large, random-
receptor ized, controlled trials have evaluated the benefit
of ranibizumab for the treatment of both forms
of neovascularization.
In the Minimally Classic/Occult Trial of the
Anti-VEGF Antibody Ranibizumab in the Treat-
New blood- ment of Neovascular Age-Related Macular De-
vessel growth
generation (MARINA) trial (ClinicalTrials.gov
number, NCT00056836), investigators compared
monthly intraocular injections of ranibizumab
with placebo in patients who had AMD with the
predominantly occult type of choroidal neovas-
cularization.30 At 1 year, only 5% of patients
who were treated with either 0.3 mg or 0.5 mg
of ranibizumab had lost 15 letters of vision
Ranibizumab (about three lines on a standard eye chart), as
compared with 38% of control subjects. Of the
patients who were treated with ranibizumab,
34% of those receiving the 0.5-mg dose and 25%
of those receiving the 0.3-mg dose gained 15
letters of vision, as compared with only 5% of
control subjects.30 Ranibizumab had a positive
treatment effect in all subgroups of patients
with neovascular AMD.31
n engl j med 363;17 nejm.org october 21, 2010
74
 clinical ther apeutics
In the Anti-VEGF Antibody for the Treatment
of Predominantly Classic Choroidal Neovascu-
larization in Age-Related Macular Degeneration
(ANCHOR) study (NCT00061594), investigators
compared monthly intraocular injections of ra-
nibizumab with photodynamic therapy in patients
who had AMD with the predominantly classic type
of neovascularization.32 Photodynamic therapy
involves the intravenous injection of verteporfin,
a photosensitizing dye, followed by application of
a nonthermal red laser light (at a wavelength of
689 nm) to the area of neovascularization.33 The
red laser light stimulates the dye, which causes the
formation of singlet oxygen and secondary dam-
age to the vessels. Only 4% of patients who were
treated with 0.5 mg of ranibizumab and 5% of
those treated with 0.3 mg lost 15 letters of vision
at 1 year, as compared with 36% of patients who
were treated with laser-activated verteporfin.
Among patients who were treated with ranibi-
zumab, 40% of those receiving the 0.5-mg dose
and 36% of those receiving the 0.3-mg dose
gained 15 letters of vision, as compared with
only 6% in the verteporfin group.
Cl inic a l Use
In the absence of anti-VEGF therapy, neovascular
AMD usually results in a substantial loss of vision.
An alternative treatment is thermal laser photo-
coagulation, but this therapy damages the over-
lying and surrounding retina and is associated
with a high rate of recurrent neovascularization.
At 24 months after laser treatment of neovascu-
larization that extended beneath the center of the
macula, only 1% of treated eyes had better than
20/100 vision.34 Another alternative is photody-
namic therapy, but the ANCHOR study showed
that treatment with ranibizumab is superior.32
Bevacizumab is an agent that is commonly
used as an alternative to ranibizumab. Bevaci-
zumab is also a monoclonal antibody that binds
to VEGF-A, but the Food and Drug Administra-
tion (FDA) has approved it only for the treatment
of colon cancer. The similarity between the two
molecules has led many clinicians to hypothesize
that the two drugs might be equally effective for
the treatment of neovascular AMD, although this
hypothesis has not been formally confirmed.
When bevacizumab is used for the treatment of
AMD, a pharmacy can split a vial into small
n engl j med 363;17
BACK to TOC
doses that can be administered intraocularly.
Generally, the dose that is used is 1.25 mg, which
costs about $75, as compared with approximately
$2,000 for a dose of ranibizumab. It should be
emphasized that this use of bevacizumab for neo-
vascular AMD is considered off-label therapy.
Ranibizumab is injected into the eye in an out-
patient clinic (see video, available with the full text A video showing Click here
of this article at NEJM.org). Hospitalization is not injection of to access
ranibizumab video.
needed. We perform injections in a room specially is available at
designed for minor surgery. Informed consent NEJM.org
should be obtained before the procedure. We be-
gin the injection procedure by giving anesthetic
drops of proparacaine hydrochloride 0.5% and
tetracaine hydrochloride 0.5%, followed by an an-
tibiotic drop. Many retinal specialists do not use
subconjunctival anesthesia, but for most of our
patients, we inject about 0.2 ml of 1% lidocaine
without epinephrine beneath the conjunctiva in
the superotemporal quadrant, 3 mm posterior to
the corneal limbus. After waiting about 4 min-
utes, we place a lid speculum into the eye to hold
the lids open and drop 5% povidoneiodine onto
the conjunctiva.
We use calipers to mark the injection site
3.5 mm posterior to the limbus (the outer border
of the cornea). We use a 30-gauge needle on a
1-mm tuberculin syringe to inject 0.05 ml (0.5 mg
of ranibizumab) into the middle of the vitreous
cavity, then place a sterile cotton swab firmly
over the injection site as the needle is withdrawn
to prevent backflush of fluid through the injec-
tion site. We then remove the lid speculum and
gently irrigate the eye to remove any residual
povidoneiodine.
We give an additional drop of antibiotic and
instruct the patient to use the drops twice more
that day and then four times per day for the next
3 days. Some retinal specialists no longer use pro-
phylactic antibiotic drops because they believe that
the drops do not reduce the risk of infectious
endophthalmitis. We tell the patient to call if vi-
sion decreases or the eye becomes painful. We
give the patient a sheet that contains the instruc-
tions for using the antibiotic drops, warnings, and
numbers to call if there are problems. The patient
makes an appointment to return in 1 month.
Intraocular hemorrhage is rare after injections
of ranibizumab. As noted, the needle is small (30
gauge), and the injection site is chosen to avoid
major ocular blood vessels. The injection of the
nejm.org october 21, 2010 1651
75
 The n e w e ng l a n d j o u r na l
drug temporarily raises the intraocular pressure
to or even above the vascular perfusion pressure
of the eye, which effectively tamponades any
bleeding until clotting occurs. Therefore, there is
no need for a patient to stop taking warfarin or
antiplatelet drugs. There are no known interac-
tions with other drugs, and no adjustments need
to be made for other medical conditions.
At the 1-month follow-up appointment, the
patients vision is checked, and the eye is exam-
ined for signs of inflammation, new hemor-
rhage, or fluid within or under the retina. We
then give a second injection and repeat the pro-
cess again 1 month later, for a total of three
monthly treatments. At subsequent visits, we
give additional ranibizumab injections, but we
carefully extend the follow-up period between
the visits by 1 week if the eye remains stable,
with no fluid or hemorrhage.35 Once the follow-
up interval reaches 3 months and the eye is
stable, consideration can be given to stopping
the injections, but the patient should still be
seen every 3 to 4 months to detect any late recur-
rences. Patients should be reminded frequently
to call if vision loss or distortion occurs.
If a patient has only a partial response or no
response to serial injections, it may be appropri-
ate to stop the treatment. This decision should
be made jointly with the patient. It often de-
pends somewhat on whether the better eye is
being treated. Patients often wish to forgo injec-
tions of modest benefit if the poorer eye is being
treated, whereas they may be reluctant to stop if
the injections are being given in the better eye.
Brown and colleagues36 calculated that the
cost of treating a patient with monthly injections
of ranibizumab over a 2-year period, as described
in the MARINA trial, would be $52,652. Of this
total (which includes incidental procedural costs),
$44,812 would be the cost of the drug, and $5,981
the physicians fees.36
A dv er se Effec t s
The injection of either anesthetic or ranibizumab
can cause mild subconjunctival hemorrhage. The
blood is cosmetically unappealing but resolves
without sequelae. Many clinicians use topical an-
esthetics only, but in our experience, most pa-
tients who are treated in this manner feel some
pain when the needle pierces the wall of the eye.
1652 n engl j med 363;17
BACK to TOC
of m e dic i n e
Any intraocular injection can cause bacterial
endophthalmitis, a serious complication that
threatens vision. In the MARINA study, the pre-
sumed rate of endophthalmitis (presumed be-
cause not all eyes showed positive cultures) was
1.0% (infection in 5 of 477 patients); the rate per
injection was 0.05% (infection associated with
5 of 10,443 injections).30 Patients in whom endo-
phthalmitis develops have vision loss or an in-
crease in floaters, almost always within the first
week after the injection. The eye is red and may
be painful. Patients with such symptoms after
ranibizumab injection should be seen promptly
by their retinal specialist.
In the MARINA and ANCHOR trials, the re-
spective risks of nonocular hemorrhage were 9%
and 6% in the treated groups versus 6% and 2%
in the comparator groups, which may indicate a
systemic anti-VEGF effect.30,32 In the MARINA
trial, rates of stroke were 2.5% in the 0.5-mg
group and 1.3% in the 0.3-mg group versus 0.8%
in the sham group. Rates of myocardial infarc-
tion were 1.3% in the 0.5-mg group and 2.5%
in the 0.3-mg group versus 1.7% in the sham
group.30 Similar effects were seen in the 2-year
results from the ANCHOR trial.37 None of these
differences were statistically significant.
As noted above, bevacizumab is also an anti
VEGF-A monoclonal antibody. In a meta-analysis
of trials of bevacizumab in patients with meta-
static cancer, arterial thromboembolic events,
including cardiac ischemia and stroke, occurred in
3.3% of patients who had received the drug, as
compared with 2.0% who had not.38 The total-
body dose of bevacizumab that was used in these
trials was more than 1000 times the dose used
in the treatment of AMD.
A r e a s of Uncer ta in t y
A major area of uncertainty is whether bevaci-
zumab is as effective as ranibizumab in the treat-
ment of neovascular AMD. This issue has been
evaluated in several small studies, with encour-
aging results.39-42 However, these findings are
provisional. To answer the question definitively,
a large, randomized clinical trial, called the
Comparison of AMD Treatments Trial (CATT;
NCT00593450), is currently being conducted. Re-
sults from this trial are expected in the spring of
2011.43 Since the cost of the dose of bevacizumab
nejm.org october 21, 2010
76
 clinical ther apeutics

that is commonly used in AMD is so much less R ec om mendat ions

than the cost of the standard dose of ranibizum-
ab ($75 vs. $2,000), the majority of our patients
choose to be treated with bevacizumab on the
basis of currently available data.
Another area of uncertainty is the optimal
treatment schedule. Some specialists administer
ranibizumab on a fixed monthly schedule,
whereas others administer the agent only when
fluid is present on optical coherence tomogra-
phy or when leakage is seen on fluorescein an-
giography.44 In CATT, investigators are compar-
ing these two approaches to treatment, as well
as comparing ranibizumab with bevacizumab. If
the study shows that the difference in outcome
between the fixed and as-needed approaches is
minimal, patients and physicians may opt for
less-frequent injections. All the groups in CATT
are being followed monthly, but most patients
dislike having to return for a visit every month.
An additional trial may be warranted in the fu-
ture, comparing monthly visits with gradual ex-
tension of the follow-up interval for patients
whose condition is stable. Some pharmacoki-
netic data, though not conclusive, support the
hypothesis that bevacizumab may have a longer
intraocular half-life than ranibizumab.45
Guidel ine s
The American Academy of Ophthalmology includ-
ed ranibizumab in its September 2008 Preferred
Practice Pattern Guidelines for Age-Related Macu-
lar Degeneration, grading such therapy as level A
(most important to the care process) and level I
(greatest strength of evidence).46 The same guide-
lines also listed bevacizumab as a recommended
treatment, grading it level A for importance but
only level III (the lowest level) with respect to
strength of evidence. The guidelines state that
when using bevacizumab, the ophthalmologist
should provide appropriate informed consent with
respect to the off-label status. The guidelines
committee of the European Society of Retina Spe-
cialists make similar recommendations in a state-
ment published in August 2007.47
The patient described in the vignette is an appro-
priate candidate for either ranibizumab or beva-
cizumab therapy. We would explain to the pa-
tient that only ranibizumab has been approved by
the FDA for use in AMD. We would also explain
that the efficacy of bevacizumab is thought to be
similar to that of ranibizumab, although such
efficacy has not been evaluated in definitive tri-
als, and that bevacizumab is substantially less
expensive.
Regardless of the patients choice of agent,
we would carry out the initial injection in our
outpatient minor surgery unit, as described in
the Clinical Use section. After the first injec-
tion of either bevacizumab or ranibizumab, we
would ask the patient to return for another in-
jection in 4 weeks. After the second injection,
we would ask him to return in 4 weeks if he
received ranibizumab and in 6 weeks if he re-
ceived bevacizumab. We would continue injec-
tions at these intervals until the retina was dry.
We would then gradually extend the interval
between follow-up visits, depending on the ex-
amination findings.
In our experience, most treatment failures are
due to missed follow-up visits. Patients should be
reminded repeatedly to call if their vision wors-
ens. Physicians should have a system in place for
calling patients who have been lost to follow-up
because such patients have a high risk of recur-
rent neovascularization with permanent scarring
and vision loss. Finally, ophthalmologists should
be aware of the patients overall medical condi-
tion and should communicate clearly with the
patients other physicians when warranted.
Supported by the Robert C. Watzke Research Fund, Research
to Prevent Blindness, the Foundation Fighting Blindness, a grant
(EY016822) from the National Eye Institute, and the Howard
Hughes Medical Institute.
No potential conflict of interest relevant to this article was
reported. Disclosure forms provided by the authors are avail-
able with the full text of this article at NEJM.org.
We thank Robert F. Mullins, Ph.D., for his critical review
and advice; Alton Szeto Illustration for the drawings that
served as the design for Figure 2; and Patricia Duffel for her
reference help.

References
1. Friedman DS, OColmain BJ, Muoz B,
et al. Prevalence of age-related macular
degeneration in the United States. Arch
Ophthalmol 2004;122:564-72.
2. Brown GC, Brown MM, Sharma S, et al.
The burden of age-related macular degen-
eration: a value-based medicine analysis.
Trans Am Ophthalmol Soc 2005;103:173-
86.
3. Edwards AO, Ritter R III, Abel KJ,
Manning A, Panhuysen C, Farrer LA.
Complement factor H polymorphism and
age-related macular degeneration. Science
2005;308:421-4.
4. Hageman GS, Anderson DH, Johnson

n engl j med 363;17 nejm.org october 21, 2010 1653

BACK to TOC
77
 The n e w e ng l a n d j o u r na l
LV, et al. A common haplotype in the com-
plement regulatory gene factor H (HF1/
CFH) predisposes individuals to age-
related macular degeneration. Proc Natl
Acad Sci U S A 2005;102:7227-32.
5. Haines JL, Hauser MA, Schmidt S, et al.
Complement factor H variant increases
the risk of age-related macular degenera-
tion. Science 2005;308:419-21.
6. Hammond CJ, Webster AR, Snieder H,
Bird AC, Gilbert CE, Spector TD. Genetic
influence on early age-related maculopa-
thy: a twin study. Ophthalmology 2002;
109:730-6.
7. Heiba IM, Elston RC, Klein BE, Klein R.
Sibling correlations and segregation analy-
sis of age-related maculopathy: the Beaver
Dam Eye Study. Genet Epidemiol 1994;11:
51-67. [Erratum, Genet Epidemiol 1994;
11:571.]
8. Klaver CC, Wolfs RC, Assink JJ, van
Duijn CM, Hofman A, de Jong PT. Genetic
risk of age-related maculopathy: popula-
tion-based familial aggregation study.
Arch Ophthalmol 1998;116:1646-51.
9. Klein RJ, Zeiss C, Chew EY, et al.
Complement factor H polymorphism in
age-related macular degeneration. Science
2005;308:385-9.
10. Rivera A, Fisher SA, Fritsche LG, et al.
Hypothetical LOC387715 is a second major
susceptibility gene for age-related macular
degeneration, contributing independently
of complement factor H to disease risk.
Hum Mol Genet 2005;14:3227-36.
11. Seddon JM, Cote J, Page WF, Aggen
SH, Neale MC. The US twin study of age-
related macular degeneration: relative
roles of genetic and environmental influ-
ences. Arch Ophthalmol 2005;123:321-7.
12. Zareparsi S, Buraczynska M, Bran-
ham KE, et al. Toll-like receptor 4 variant
D299G is associated with susceptibility to
age-related macular degeneration. Hum
Mol Genet 2005;14:1449-55.
13. Thornton J, Edwards R, Mitchell P, Har-
rison RA, Buchan I, Kelly SP. Smoking and
age-related macular degeneration: a review
of association. Eye (Lond) 2005;19:935-44.
14. Hageman GS, Luthert PJ, Victor
Chong NH, Johnson LV, Anderson DH,
Mullins RF. An integrated hypothesis that
considers drusen as biomarkers of im-
mune-mediated processes at the RPE-
Bruchs membrane interface in aging and
age-related macular degeneration. Prog
Retin Eye Res 2001;20:705-32.
15. Penfold PL, Madigan MC, Gillies MC,
Provis JM. Immunological and aetiologi-
cal aspects of macular degeneration. Prog
Retin Eye Res 2001;20:385-414.
16. Bok D. The retinal pigment epitheli-
um: a versatile partner in vision. J Cell Sci
Suppl 1993;17:189-95.
17. Hayreh SS, Baines JA. Occlusion of
the posterior ciliary artery. II. Chorio-ret-
inal lesions. Br J Ophthalmol 1972;56:736-
53.
1654 n engl j med 363;17
BACK to TOC
of m e dic i n e
18. Zarbin MA. Current concepts in the
pathogenesis of age-related macular de-
generation. Arch Ophthalmol 2004;122:598-
614.
19. Klein R, Wang Q, Klein BE, Moss SE,
Meuer SM. The relationship of age-related
maculopathy, cataract, and glaucoma to
visual acuity. Invest Ophthalmol Vis Sci
1995;36:182-91.
20. Green WR. Histopathology of age-relat-
ed macular degeneration. Mol Vis 1999;5:
27.
21. Green WR, Enger C. Age-related mac-
ular degeneration histopathologic studies:
the 1992 Lorenz E. Zimmerman Lecture.
Ophthalmology 1993;100:1519-35.
22. Frank RN, Amin RH, Eliott D, Puklin
JE, Abrams GW. Basic fibroblast growth
factor and vascular endothelial growth
factor are present in epiretinal and cho-
roidal neovascular membranes. Am J
Ophthalmol 1996;122:393-403.
23. Kvanta A, Algvere PV, Berglin L, Sere-
gard S. Subfoveal fibrovascular mem-
branes in age-related macular degenera-
tion express vascular endothelial growth
factor. Invest Ophthalmol Vis Sci 1996;
37:1929-34.
24. Lopez PF, Sippy BD, Lambert HM,
Thach AB, Hinton DR. Transdifferentiat-
ed retinal pigment epithelial cells are im-
munoreactive for vascular endothelial
growth factor in surgically excised age-
related macular degeneration-related cho-
roidal neovascular membranes. Invest
Ophthalmol Vis Sci 1996;37:855-68.
25. Miller JW, Adamis AP, Shima DT, et
al. Vascular endothelial growth factor/
vascular permeability factor is temporally
and spatially correlated with ocular an-
giogenesis in a primate model. Am J
Pathol 1994;145:574-84.
26. Otani A, Takagi H, Oh H, et al. Vascu-
lar endothelial growth factor family and
receptor expression in human choroidal
neovascular membranes. Microvasc Res
2002;64:162-9.
27. Ferrara N, Damico L, Shams N, Low-
man H, Kim R. Development of ranibizum-
ab, an anti-vascular endothelial growth fac-
tor antigen binding fragment, as therapy for
neovascular age-related macular degenera-
tion. Retina 2006;26:859-70.
28. Jain RK. Normalization of tumor vas-
culature: an emerging concept in anti-
angiogenic therapy. Science 2005;307:58-
62.
29. Subfoveal neovascular lesions in age-
related macular degeneration: guidelines
for evaluation and treatment in the mac-
ular photocoagulation study. Arch Oph-
thalmol 1991;109:1242-57.
30. Rosenfeld PJ, Brown DM, Heier JS, et al.
Ranibizumab for neovascular age-related
macular degeneration. N Engl J Med
2006;355:1419-31.
31. Boyer DS, Antoszyk AN, Awh CC,
Bhisitkul RB, Shapiro H, Acharya NR.
nejm.org october 21, 2010
Subgroup analysis of the MARINA study
of ranibizumab in neovascular age-relat-
ed macular degeneration. Ophthalmology
2007;114:246-52.
32. Brown DM, Kaiser PK, Michels M, et
al. Ranibizumab versus verteporfin for
neovascular age-related macular degen-
eration. N Engl J Med 2006;355:1432-44.
33. Photodynamic therapy of subfoveal
choroidal neovascularization in age-related
macular degeneration with verteporfin:
one-year results of 2 randomized clinical
trials TAP report. Arch Ophthalmol
1999;117:1329-45. [Erratum, Arch Oph-
thalmol 2000;118:488.]
34. Laser photocoagulation of subfoveal
neovascular lesions in age-related macu-
lar degeneration: results of a randomized
clinical trial. Arch Ophthalmol 1991;109:
1220-31.
35. Spaide R. Ranibizumab according to
need: a treatment for age-related macular
degeneration. Am J Ophthalmol 2007;143:
679-80.
36. Brown MM, Brown GC, Brown HC,
Peet J. A value-based medicine analysis
of ranibizumab for the treatment of sub-
foveal neovascular macular degeneration.
Ophthalmology 2008;115:1039-45.
37. Brown DM, Michels M, Kaiser PK, Heier
JS, Sy JP, Ianchulev T. Ranibizumab versus
verteporfin photodynamic therapy for neo-
vascular age-related macular degeneration:
two-year results of the ANCHOR study.
Ophthalmology 2009;116:57-65.
38. Ranpura V, Hapani S, Chuang J, Wu S.
Risk of cardiac ischemia and arterial
thromboembolic events with the angiogen-
esis inhibitor bevacizumab in cancer pa-
tients: a meta-analysis of randomized con-
trolled trials. Acta Oncol 2010;49:287-97.
39. Bashshur ZF, Haddad ZA, Schakal AR,
Jaafar RF, Saad A, Noureddin BN. Intravit-
real bevacizumab for treatment of neovas-
cular age-related macular degeneration:
the second year of a prospective study. Am
J Ophthalmol 2009;148(1):59-65.
40. Fong DS, Custis P, Howes J, Hsu JW.
Intravitreal bevacizumab and ranibizum-
ab for age-related macular degeneration:
a multicenter, retrospective study. Oph-
thalmology 2010;117:298-302.
41. Ehlers JP, Spirn MJ, Shah CP, et al. Ra-
nibizumab for exudative age-related mac-
ular degeneration in eyes previously treat-
ed with alternative vascular endothelial
growth factor inhibitors. Ophthalmic Surg
Lasers Imaging 2010;41:182-9.
42. Tufail A, Patel PJ, Egan C, et al. Beva-
cizumab for neovascular age related mac-
ular degeneration (ABC Trial): multicen-
tre randomised double masked study. BMJ
2010;340:c2459.
43. Martin DF, Maguire MG, Fine SL.
Identifying and eliminating the road-
blocks to comparative-effectiveness re-
search. N Engl J Med 2010;363:105-7.
44. Lalwani GA, Rosenfeld PJ, Fung AE, et
78

al. A variable-dosing regimen with intra-
vitreal ranibizumab for neovascular age-
related macular degeneration: year 2 of
the PrONTO Study. Am J Ophthalmol
2009;148:43-58.
45. Krohne TU, Eter N, Holz FG, Meyer
CH. Intraocular pharmacokinetics of be-
vacizumab after a single intravitreal injec-
Palmyra, Syria
n engl j med 363;17
BACK to TOC
clinical ther apeutics
tion in humans. Am J Ophthalmol 2008;
146:508-12.
46. American Academy of Ophthalmology
Retina Panel. Preferred practice pattern
guidelines. age-related macular degenera-
tion. San Francisco: American Academy
of Ophthalmology, 2008. (http://one.aao
.org/CE/PracticeGuidelines/PPP_Content
nejm.org october 21, 2010
.aspx?cid=f413917a-8623-4746-b441-
f817265eaf b4.)
47. Schmidt-Erfurth UM, Richard G, Au-
gustin A, et al. Guidance for the treat-
ment of neovascular age-related macular
degeneration. Acta Ophthalmol Scand
2007;85:486-94.
Copyright 2010 Massachusetts Medical Society.
Christian Mller, M.D.
1655
79

From the Department of Nutrition, Har-
vard School of Public Health (F.M.S., H.C.);
and Channing Laboratory and Cardiology
Division, Department of Medicine, Brigham
and Womens Hospital and Harvard
Medical School (F.M.S.) all in Boston.
Address reprint requests to Dr. Sacks at
the Department of Nutrition, Harvard
School of Public Health, Bldg. 1, 2nd Fl.,
Boston, MA 02115, or at fsacks@hsph
.harvard.edu.
N Engl J Med 2010:362:2102-12.
Copyright 2010 Massachusetts Medical Society.
2102
BACK to TOC
The n e w e ng l a n d j o u r na l of m e dic i n e
clinical therapeutics
Dietary Therapy in Hypertension
Frank M. Sacks, M.D., and Hannia Campos, Ph.D.
This Journal feature begins with a case vignette that includes a therapeutic recommendation. A discussion
of the clinical problem and the mechanism of benefit of this form of therapy follows. Major clinical studies,
the clinical use of this therapy, and potential adverse effects are reviewed. Relevant formal guidelines,
if they exist, are presented. The article ends with the authors clinical recommendations.
A 57-year-old woman presents to an outpatient clinic for evaluation of hypertension.
She has no history or symptoms of cardiovascular disease and reports having gained
15 kg over the past 30 years. Her blood pressure is 155/95 mm Hg, her weight 86 kg,
her height 165 cm, her body-mass index (BMI, the weight in kilograms divided by the
square of the height in meters) 31, and her waist circumference 98 cm. Her serum
triglyceride level is 175 mg per deciliter (2.0 mmol per liter), high-density lipoprotein
cholesterol 42 mg per deciliter (1.1 mmol per liter), low-density lipoprotein cholesterol
110 mg per deciliter (2.8 mmol per liter), and glucose 85 mg per deciliter (4.7 mmol per
liter). Her clinical profile is thus consistent with the metabolic syndrome.1 She is a
nonsmoker, is sedentary, and eats a diet that is high in white bread, processed meats,
and snacks and drinks containing sugars and sodium and is low in fruits and vege-
tables. She is interested in adopting a healthier lifestyle.
The Cl inic a l Probl em
Hypertension is defined as a systolic blood pressure of 140 mm Hg or higher or a
diastolic blood pressure of 90 mm Hg or higher.2 However, morbidity increases
among persons whose blood pressure is above 115/75 mm Hg. High blood pressure
is associated with an increased risk of stroke, myocardial infarction, heart failure,
renal failure, and cognitive impairment.2-4 Systolic blood pressure above 115 mm Hg
is the most important determinant of the risk of death worldwide,2 being respon-
sible for 7.6 million cardiovascular deaths annually.3
From 1960 through 1991, blood pressure decreased in the United States, and
after the first 10 years of this interval, the rate of cardiovascular deaths decreased.2
Effective hypertension screening and treatment were probably the reason for these
beneficial trends. However, from 1990 through 2002, blood pressure increased.5,6
Intake of fruits and vegetables and adherence to healthful dietary patterns de-
clined during this period7,8 and the prevalence of abdominal obesity increased9;
both trends have contributed to hypertension.
Among most populations in industrialized countries, the prevalence of hyper-
tension increases dramatically with age; in the United States it rises from about
10% in persons 30 years of age to 50% in those 60 years of age.6 However, some
persons, including strict vegetarians,10-12 populations whose diet consists mostly
of vegetable products,11,13 and those whose sodium intake is low,13-15 have virtu-
ally no increase in hypertension with age.
S t r ategie s a nd E v idence
Pathophysiology and Effect of Therapy
Essential hypertension is the name for hypertension that cannot be attributed to a
specific renal or adrenal disease, such as chronic renal failure or an adrenal tumor;
n engl j med 362;22 nejm.org june 3, 2010
80
 clinical ther apeutics
the vast majority of patients with hypertension
have essential hypertension. The pathophysiology
of essential hypertension is complex, with much
remaining to be discovered (Fig. 1, and Section 1
in the Supplementary Appendix, available with
the full text of this article at NEJM.org). The
three cornerstones of dietary treatment of hyper-
tension a healthful dietary pattern, reduced
sodium intake, and reduced body fat influence
the pathophysiology of hypertension at many of
its points of control.
High sodium intake is strongly correlated with
the development of hypertension.16-18 Sodium
intake initiates an autoregulatory sequence that
leads to increased intravascular fluid volume and
cardiac output, peripheral resistance, and blood
pressure. The elevation in blood pressure results
in a phenomenon called pressure natriuresis, in
which increased renal perfusion pressure leads
to increased excretion of fluid and sodium. In
essential hypertension, however, sodium excre-
tion is impaired. It is hypothesized that in most
cases essential hypertension is a genetic disorder
involving many individual genes, each of which
influences the bodys handling of sodium to
varying degrees18 and becomes expressed in the
context of an unhealthful dietary environment,
particularly one characterized by excessive intake
of salt.
Numerous other factors contribute to the
pathophysiology of hypertension. Especially in
the elderly, large conduit arteries such as the
aorta and carotid arteries become stiff and less
compliant, increasing systolic blood pressure.19
Proliferation of smooth-muscle cells and endo-
thelial dysfunction occur in resistance vessels,
including small arteries and arterioles, causing
vasoconstriction and increasing peripheral vascu-
lar resistance.20-22 Although the systemic renin
angiotensinaldosterone axis is often suppressed
in the presence of elevated blood pressure, angio-
tensin II activity is increased locally in various
tissues, including the kidneys, vascular endothe-
lium, and adrenal glands.23,24 Increased activity
in the sympathetic nervous system may also be
a factor.25-30 Both aging19,31-33 and obesity25-30
contribute to the pathogenesis of hypertension
through several mechanisms (Fig. 1, and Section
1 in the Supplementary Appendix).
Two effective interventions for lowering blood
pressure in patients with hypertension are reduc-
ing sodium intake and reducing weight. Reduc-
tions in dietary salt lessen the amount of sodium
n engl j med 362;22
BACK to TOC
the kidney has to excrete to restore normal blood
volume. Compliance in the aorta and carotid
artery in older patients with hypertension is im-
proved when sodium intake is reduced.34 Reduc-
tion in sodium intake also improves arterial vaso-
dilatation.21,22 Weight loss moderates activation
of the reninangiotensinaldosterone axis35,36
and the sympathetic nervous system37,38 and di-
minishes sodium retention.39 Decreases in ab-
dominal visceral fat also improve the function-
ing of both conduit and resistance vessels.40
In addition to sodium restriction and weight
loss, several other dietary modifications that are
collectively termed a healthful dietary pattern
have been shown to reduce blood pressure. Al-
though the mechanisms of these diets have not
been fully clarified, adherence to these diets has
been found to reset the pressurenatriuresis curve
so that a lower pressure suffices to excrete so-
dium and reduce blood volume,41 reduce aortic
stiffness,42 and improve vasodilatation in small
resistance vessels.43,44 As compared with the
typical U.S. diet, the kinds of dietary patterns
that have been proved to lower blood pressure
emphasize fruits, vegetables, and low-fat dairy
products; include whole grains, poultry, fish, and
nuts; make use of unsaturated vegetable oils;
and contain smaller amounts of red meat, sweets,
and sugar-containing beverages.45,46 Clinical trials
of such diets have not usually emphasized the
identification of specific nutrients or single foods
that lower blood pressure but rather have used
epidemiologic data to define dietary patterns,
such as Mediterranean-style diets47,48 and vegetar-
ian diets.11,12 (see Section 2 in the Supplemen-
tary Appendix for a discussion of the effects of
specific foods and nutrients on blood pressure).
Clinical Evidence
The most carefully studied and established health-
ful dietary patterns are the Dietary Approaches
to Stop Hypertension (DASH) diet,45,49 variants
of that diet,46,50 and variations of the Mediterra-
nean diet.51,52 In the original DASH trial,49 459
adults whose systolic blood pressure was less than
160 mm Hg and whose diastolic blood pressure
was 80 to 95 mm Hg, 133 of whom had hyperten-
sion, were randomly assigned to a control diet
typical of the average U.S. diet, a diet rich in
fruits and vegetables, or a combination diet rich
in fruits, vegetables, and low-fat dairy products
and relatively low in saturated and total fat. So-
dium intake and body weight were maintained at
nejm.org june 3, 2010 2103
81
 The n e w e ng l a n d j o u r na l of m e dic i n e

A High-sodium, high-calorie diet Increased

sympathetic
nervous system
activity
Intrinsic renal factors
(genetic and prenatal)
regulate sodium excretion
Large conduit arteries
become less compliant

High sodium level activates local

angiotensin II in heart and arteries

Increased blood pressure Increased cardiac output

Smooth-muscle cell
Abnormal pressure natriuresis
proliferation and
and sodium retention
rearrangement
Endothelial-cell
Increased tissue angiotensin II dysfunction in small
in kidneys and adrenal glands resistance vessels
Increased peripheral
resistance

Abdominal fat further increases

conduit artery stiffness, sympathetic
nervous system activity, and
angiotensin II levels

B
Low-sodium, low-calorie diet
Weight loss reduces
sympathetic nervous
system activity

Weight loss, low sodium intake,

and healthy diet reduce stiffness
of large conduit arteries
Decreased blood pressure

Healthy diet improves renal

sodium excretion

Weight loss, low sodium intake, and

healthy diet improve function of small
resistance vessels and decrease peripheral resistance

Decreased abdominal fat

Figure 1. Mechanisms Linked to Increases in Blood Pressure and the Therapeutic Effects of Healthful Dietary Patterns, Sodium Reduction,
and Weight Loss.

2104 n engl j med 362;22 nejm.org june 3, 2010

BACK to TOC
82
 clinical ther apeutics
constant levels. After 8 weeks, among the par-
ticipants with hypertension, the diet rich in fruits
and vegetables reduced systolic and diastolic
blood pressure by 7.2 and 2.8 mm Hg more, re-
spectively, than the control diet (P<0.001 and
P = 0.01, respectively). The combination diet re-
sulted in greater reductions (11.4 and 5.5 mm Hg,
respectively, as compared with the control diet;
P<0.001 for each). The effects were less pro-
nounced among participants who did not have
hypertension at baseline.
In a subsequent trial, the effect of various
levels of sodium intake was studied in the con-
text of the DASH diet in 412 participants with
blood pressure levels at enrollment similar to
those of participants in the original DASH trial.53
Patients were randomly assigned to either the
DASH combination diet (now commonly termed
the DASH diet) or a control diet. Participants in
each group were then given a diet with high,
intermediate, and low levels of sodium (3.5, 2.3,
and 1.2 g per day, respectively) for 30 days each
in random order. Body weight was held constant
by adjusting total caloric intake. Reducing sodi-
um intake resulted in a significant incremental
reduction in both systolic and diastolic blood
pressure in both groups (Fig. 2).
In a secondary analysis from the sodium trial,
the blood-pressurelowering effects of the DASH
diet and low sodium were each accentuated as
age increased54 (Fig. 3). Systolic blood pressure
was 12 mm Hg higher among participants be-
tween 55 and 76 years of age than among those
between 21 and 41 years of age when they were
given a typical U.S. diet that was high in sodium.
This difference in systolic blood pressure is simi-
lar to that in the U.S. population when the same
age groups are compared.55 In marked contrast,
systolic blood pressure was the same among
older and younger participants when they were
given the DASH diet with low sodium content.
This finding suggests that the typical rise in
blood pressure that occurs with age during adult
life may be prevented or reversed if the low-
sodium DASH diet is followed.
Women, blacks, and those with the metabolic
syndrome have a mildly enhanced reduction in
blood pressure in response to a low-sodium
diet.53,54,56,57 It is not possible to identify indi-
vidual patients for whom sodium reduction is
especially effective58 (see Section 3 in the Sup-
plementary Appendix).
Two reduced-carbohydrate versions of the
n engl j med 362;22
BACK to TOC
145
Higher to lower sodium
Control diet 2.1 (0.1 to 4.0) Control: 8
Mean Systolic Blood Pressure (mm Hg)
DASH: 7
140
8.0 (4.9 to 6.0 (4.0 to 7.9)
11.1) 7.5 (4.2 to
10.8)
135 DASH diet
1.6 (0.6 to 3.8) 6.7 (3.5 to
9.8)
130
5.1 (3.0 to 7.3)
125 Lower-sodium DASH vs. higher-sodium control: 15
0
High (3.5 g) Intermediate (2.3 g) Low (1.2 g)
Dietary Sodium
Figure 2. Sodium Reduction, the DASH Diet, and Changes in Systolic Blood
Pressure.
The figure shows the additive beneficial effects of the DASH diet and reduced
intake of sodium on systolic blood pressure in patients with mild hyperten-
sion who were older than 45 years of age. The participants were a subgroup
of those in the study of the effects of the DASH diet and reductions in dietary
sodium,53 who were randomly assigned to follow a DASH diet (33 partici-
pants) or a typical U.S. diet (37 participants) for 90 days. During that period,
each group consumed three versions of the diet adjusted for daily sodium
content. The participants in each group consumed each of the sodium-
adjusted diets for 30 days in a crossover design; body weight was held con-
stant. The two downward-sloping arrows on the left depict the effect of in-
termediate sodium intake as compared with higher sodium intake, and the
two downward-sloping arrows on the right depict the effect of lower sodium
intake as compared with intermediate sodium intake. The dotted lines show
the effect of the DASH diet as compared with the typical U.S. diet at each
level of dietary sodium. Numbers shown represent the mean changes with
95% confidence intervals. Adapted from Bray et al.54
DASH diet were studied in 164 adults enrolled in
the Optimal Macronutrient Intake Trial to Pre-
vent Heart Disease (OmniHeart).46,50 One diet
higher in unsaturated fat and another higher in
protein were compared with a diet similar to the
standard DASH diet but slightly higher in carbo-
hydrates. As compared with the high-carbohy-
drate diet, the high-protein diet reduced mean
systolic blood pressure in participants with hy-
pertension by 3.5 mm Hg and mean diastolic
blood pressure by 2.4 mm Hg (P = 0.006 and
P = 0.008, respectively).50 The comparable effects
of the diet high in unsaturated fat were 2.9 and
1.9 mm Hg, respectively (P = 0.02 for both). As
with the DASH diet itself, these effects were less
pronounced in participants who did not have
hypertension at baseline.
The traditional Mediterranean diet47,48 has
many similarities to DASH-type diets, especially
nejm.org june 3, 2010 2105
83
 The
145
Mean Systolic Blood Pressure (mm Hg)
140
135
130
125
120
115
0 blood pressure and P = 0.05 for diastolic blood
2341 4247
Age (yr)
Figure 3. Effects of a Low-Sodium DASH Diet
on Systolic Blood Pressure with Increasing Age.
A total of 412 participants were randomly assigned to
follow a DASH diet (208 participants) or a typical U.S.
diet (control group, 204 participants) for 90 days. Dur-
ing that period, each group consumed three versions
of the diet adjusted for daily sodium content: high (3.5 g),
intermediate (2.3 g), and low (1.2 g). The participants
in each group consumed each of the sodium-adjusted
diets for 30 days in a crossover design; body weight
was held constant. Mean (SD) systolic blood pressure
is depicted for the DASH group during the period of
low sodium intake and for the control group during the
period of high sodium intake, according to age, at the
end of the 30-day period; there were 45 to 58 partici-
pants per group in each of the four age ranges shown.
The slope for the control group during the period of
high sodium intake was 0.3 mm Hg per year, spanning
30 years. The slope for the DASH-diet group during
the period of low sodium intake was 0 mm Hg per year.
I bars denote 95% confidence intervals. Data are from
Sacks et al.53
to the diet from the OmniHeart study that was
higher in unsaturated fat. In controlled trials
involving patients with the metabolic syndrome51
or type 2 diabetes,52 a reduced-carbohydrate Med-
iterranean diet lowered blood pressure and im-
proved serum lipid levels more than a low-fat
diet. In these trials, unlike the DASH trials,
weight was not held constant through caloric
adjustment; in both cases, patients assigned to
the Mediterranean diet lost more weight than
those assigned to the low-fat diet.
Epidemiologic studies generally support evi-
dence from clinical trials on the effects of di-
etary management, as do community-based and
clinic-based intervention programs (see Sections
4 and 5 in the Supplementary Appendix).
The effect of adding weight loss to the DASH
diet was evaluated in 144 adults in the Exercise
2106
BACK to TOC
n e w e ng l a n d j o u r na l of m e dic i n e
and Nutrition Interventions for Cardiovascular
Health (ENCORE) study.59 Participants were ran-
Typical diet,
high sodium domly assigned to a control diet, to the DASH
diet alone, or to a reduced-calorie modification
of the DASH diet. At 4 months, blood pressure
was reduced by 3.4/3.8 mm Hg in the control
group, by 11.2/7.5 mm Hg in the group given the
DASH diet alone (P<0.001 for both systolic and
diastolic pressures as compared with the control
DASH diet, diet), and 16.1/9.9 mm Hg with the DASH diet
low sodium
plus weight management (P = 0.02 for systolic
4854 5576
pressure as compared with the DASH diet alone).
Clinical Use
Dietary management is appropriate for all pa-
tients with hypertension. In addition, patients with
prehypertension (systolic blood pressure between
120 and 139 mm Hg or diastolic blood pressure
between 80 and 89 mm Hg) should adopt the
same dietary changes, given the benefit of di-
etary therapy at these blood-pressure levels.
Drug therapy plays an essential role in treat-
ing hypertension. The Seventh Report of the Joint
National Committee on Prevention, Detection,
Evaluation, and Treatment of High Blood Pres-
sure emphasizes that in patients for whom life-
style modification (including dietary therapy,
physical activity, and moderation of alcohol con-
sumption) does not reduce blood pressure below
140/90 mm Hg (or 130/80 mm Hg for patients
with diabetes or chronic renal disease), drug
therapy should be implemented and modified
over time given a patients response.2 However,
medication should not supplant dietary manage-
ment; rather, the two forms of treatment should
be considered complementary. The DASH diet is
effective in combination with angiotensin-recep-
tor blockers.60 Sodium reduction is highly effec-
tive in older patients with hypertension who are
taking antihypertensive medicines61 and in those
with resistant hypertension taking several anti-
hypertensive agents.62
We guide patients in adopting a healthful diet
with the use of a chart or table such as that
shown in Table 1. In simple terms, we encourage
patients to eat poultry, fish, nuts, and legumes
instead of red meat; low-fat and nonfat dairy
products instead of full-fat dairy products; veg-
etables and fruit instead of snacks and desserts
high in sugars; breads and pastas made from
whole grain instead of white flour; fruit itself
n engl j med 362;22 nejm.org june 3, 2010
84
 clinical ther apeutics
rather than fruit juice; and polyunsaturated and
monounsaturated cooking oils such as olive, cano-
la, soybean, peanut, corn, sunflower, or safflower
rather than butter, coconut oil, or palm-kernel
oil. Table 1 provides information about the num-
ber of servings and portion sizes for each type
of food that should be consumed in 1 week.
Adopting a healthful dietary approach means
making the correct choices at the market so that
the most healthful foods will be available at
home. The recommendations in Table 1 include
a food-shopping guide. In the United States, it is
common to place healthful foods at the periphery
of the market; most weekly shopping should be
concentrated there. Use of canned and processed
foods should be limited, unless their salt content
has been reduced or virtually eliminated. For con-
venience, low-sodium, frozen, or canned vegeta-
bles can be substituted for fresh ones. Sections
of the market that contain sweetened beverages,
candies, and cookies should be avoided entirely.
Sodium restriction is central to the dietary
management of hypertension. Patients should be-
come familiar with reading the food labels that
specify the sodium content of packaged and
processed foods.63 Processed foods are often high
in sodium. A low sodium diet is sometimes less
palatable for patients who are accustomed to a
high-sodium diet; however, tastes adapt quickly,
and studies have shown that low-sodium diets
can be as acceptable to patients as higher-sodium
diets.64 Herbs, spices, and citrus fruit (juice or
peel) and other acidic ingredients such as vinegar
can be added to dishes to compensate for low
sodium content and may even be preferred over
foods with higher amounts of sodium.
Patients should not skip meals, should con-
sume one third of their daily food intake at break-
fast, and should limit eating in restaurants to no
more than once weekly. Eating in many restau-
rants subverts the goal of a low-sodium diet,
since one serving of some soups, sandwiches,
fried chicken, or pizza can far exceed the total
recommended daily amount of sodium.65 The
health care reform law includes a requirement
that all chain restaurants with more than 20
locations provide information for consumers
regarding the amount of sodium and other di-
etary components in menu items.66
Compliance with dietary therapy is better, and
success rates in achieving blood-pressure control
are higher, when accompanied by active guidance
n engl j med 362;22
BACK to TOC
or counseling of the patient by clinicians or an-
cillary medical personnel with expertise in dietary
management.67-72 We always recommend that pa-
tients record their dietary intake for 1 or 2 weeks
and discuss this record with a dietitian, who will
provide specific meal plans. This is especially
important when weight loss is needed. Follow-
up with a dietitian is essential, whether arranged
in individual or group appointments. In addition,
numerous Web sites73-76 and books77-80 can pro-
vide patients with further information and guid-
ance on healthful diets.
The costs associated with dietary treatment
of hypertension are relatively modest. In one
study in the Boston area conducted in 2006, the
cost of the DASH meal plan was $31 per week in
areas with low socioeconomic status and $40 per
week in areas with high socioeconomic status;
perceived affordability was similar for patients
interviewed in clinics in both areas.81 An initial
consultation with a dietitian costs approximately
$150, and follow-up consultations about $100.
Coverage of this service by health insurance or
employer programs varies.
Adverse Effects
Adverse events generally occurred less frequently
in persons following the DASH diet and its vari-
ants or Mediterranean diets49,52,53 (see Section 6
in the Supplementary Appendix).
A r e a s of Uncer ta in t y
One crucial frontier of dietary research is that of
devising and evaluating effective behavioral and
community-based interventions. In the DASH
trials, dietary modifications were studied over a
short time span, and participants were carefully
monitored for compliance. Compliance is an es-
sential element in the long-term dietary treatment
therapy of hypertension, and we need to learn
what components of behavioral interventions
lead to adherence.82 In addition, no large, long-
term, clinical-outcomes trial of these diets has
been performed, although one long-term obser-
vational study of an earlier randomized trial and
one relatively short-term randomized trial report-
ed a decrease in the incidence of cardiovascular
events with sodium reduction (see Section 7 in the
Supplementary Appendix).83,84 However, we be-
lieve that it is not necessary to conduct a large-
scale, randomized trial to address this question in
nejm.org june 3, 2010 2107
85
 2108

BACK to TOC
Table 1. Recommended Weekly and Occasional Food Purchases for One Person Following a Healthful Diet Containing 2100 kcal and 1500 mg of Sodium per Day.*

Servings Serving Total Amount

Type of Food per Wk Size Purchased per Wk Recommendations
Weekly purchases
Market periphery Do most weekly shopping in this section
The

Vegetables
Leafy greens
Salad greens 4 1 cup 12 bags or heads Lettuce, mixed spring greens, spinach bunch (about 1 lb)
Other greens 4 1/2 cup 12 bunches Kale, collard greens, mustard greens (about 1 lb)
Cruciferous 3 1/2 cup 12 heads Broccoli, cabbage, cauliflower (about 1 lb)
Colorful 15 1/2 cup 812 individual items Tomatoes, carrots, squash, peppers, sweet potatoes, corn, egg-

n engl j med 362;22

plant, avocados (about 3 lb)
Other 3 1/2 cup 1/2 lb Celery, green beans, peas, lima beans, sprouts
Fruits

nejm.org
Fresh 20 1 medium or 1520 individual items Apples, pears, grapes, bananas, peaches, plums, oranges, tan-
n e w e ng l a n d j o u r na l

1/2 cup gerines, berries, cantaloupe, pineapple

of

chopped
Dried 8 1/4 cup 1 bag Raisins, apricots, prunes, cherries (about 1/2 lb)

june 3, 2010
Juice 4 1 glass (8 oz) 1 qt Orange, grapefruit, unsweetened carrot
Herbs, alliums, and other seasonings Use freely Thyme, ginger, garlic, onion, bay leaf, lemon juice
m e dic i n e

Meat, poultry, and fish

Fish and shellfish 2 68 oz 1 lb Cod, sea bass, halibut; fresh or canned salmon, tuna, or sar-
dines; mollusks, shrimp, crabmeat
Poultry 2 68 oz 1 lb Turkey, chicken, low-sodium cold cuts
Red meats 1 24 oz 1/4 lb Beef, pork, lamb, low-sodium cold cuts
Dairy products
Milk 10 1 glass (8 oz) 1/2 gallon Choose low-fat or nonfat products
Yogurt 3 1 cup 1 container Choose low-fat or nonfat products (about 32 oz)
Cheese 4 1 slice 1/4 lb Soft or hard

86
BACK to TOC
Processed-food aisles Choose only low-sodium products
Nuts (whole or butter) 10 1 oz 1 bag or jar Walnuts, almonds, peanuts (about 1/2 lb)
Legumes 3 1 cup 1 can or bag Chickpeas, lentils, black beans (about 1 lb)
Olives 2 1/2 cup 1 jar Black, green, stuffed (about 1/4 lb)
Spices Use freely Black pepper, cayenne, cinnamon, paprika
Baked goods 20 1 slice 1 bag Bread, rolls, pancakes, waffles (about 1 1/2 lb); choose whole-
grain products
Tomato products 4 2/3 cup 2 jars or cans Sauce, juice, whole or diced (about 12 oz per jar or can)
Chips and other snacks 3 1/2 cup 3 bags Tortilla chips, popcorn, pretzels (about 1 1/2 oz per bag)
Chocolate or sweets 1 1 oz 1 bar or similar amount Granola bars, chocolate bars (about 1 oz)
Other food aisles (sweetened beverages, candy, cookies) Skip these aisles
Less frequent purchases
Breakfast cereals 2 1/2 cup 1 1/2 cups Oats, bran, whole wheat flakes, other whole grains
Pasta, rice, and grains 3 1 cup 1/2 cup Pasta, brown rice, bulgur, quinoa, wheat berries
(cooked)

n engl j med 362;22

Cooking oils 12 1 tbs 3/4 cup Canola, corn, sunflower, olive, soybean
Table fats 16 1 tsp 1/3 cup Soft, oil-based spreads free of trans fat
Salad dressings and mayonnaise 21 1 tsp 1/2 cup Choose low-sodium items

nejm.org
Sugars 24 1 tsp 1/2 cup Table sugar, jelly, honey, maple syrup
Desserts 1 1/2 cup 1/2 cup Ice cream, sorbet, frozen yogurt, other (4 oz)
clinical ther apeutics

Eggs 3 1 3 Large eggs

Salt 7 1/3 tsp 2 1/3 tsp Salt for cooking or added at the table

june 3, 2010
* Patients should observe the following general recommendations: dont skip meals, and consume one third of daily calorie intake at breakfast; limit eating out to once weekly and
choose meals with a low salt content just one slice of pizza, a turkey sandwich, or a pasta dish can easily contain 2000 mg of sodium. Examples of conversion from standard to met-
ric measures: 1 oz equals 28 g; 1 teaspoon, 5 g; 1 cup leafy greens, about 75 g.
Unsalted frozen or canned vegetables can be substituted for fresh vegetables.
Choose at least four different types of vegetables from this category.
Also visit the processed-food aisle as needed for other food items in the less frequent purchases category.
Look for lower-sodium, unsalted, or reduced-salt items. Compare brands and choose those with lower sodium content. The total amount of sodium consumed in a week from pro-
cessed foods or eating out should not exceed 2000 mg.
Weekly allowances are provided for items that are generally purchased less than once a week. The amounts for weekly intake should be set aside in individual containers to make it
easier to keep track of how much is consumed.

2109

87
 The n e w e ng l a n d j o u r na l of m e dic i n e

view of the known benefits of healthful diets with mended for the management of hypertension,
regard to blood pressure and other risk factors. and it is therefore reasonable to assume that di-
etary change could normalize her blood pres-
Guidel ine s sure. The patient should be given written instruc-
tions on how to adopt a healthful diet such as the
We recommend the American Heart Association DASH diet, a reduced-carbohydrate version of the
guidelines for cardiovascular health and the di- DASH diet, or a Mediterranean-style diet. The in-
etary management of hypertension.85,86 These structions should include ways to substantially
guidelines endorse foods and approaches to diet reduce sodium intake. We also recommend a
similar to those included in the DASH diet and small consistent daily reduction in caloric intake
cite intake of 65 mmol, or 1.5 g, of sodium per of 200 to 300 kcal per day, coupled with an in-
day as optimal. In addition, a target BMI of less crease in physical activity. Her physician should
than 25 is recommended. Finally, the guidelines schedule a consultation with a dietitian, including
recommend no more than two alcoholic drinks a regular schedule of follow-up visits. The pa-
per day for men and one for women and people tient should monitor her blood pressure at home,
of lighter weight. (One drink is equivalent to 12 oz with an automated machine, at least once a
of beer, 5 oz of wine, or 1.5 oz of 80-proof liquor, month, preferably more frequently. A trial of inten-
each of which represents approximately 14 g of sive dietary treatment is warranted for 6 months
ethyl alcohol.) to try to achieve the targeted goal for blood pres-
sure (systolic blood pressure <140 mm Hg, dia-
stolic blood pressure <90 mm Hg) before medi-
C onclusions a nd
R ec om mendat ions cation is introduced.
No potential conflict of interest relevant to this article was
The diet of the patient described in our vignette reported. Disclosure forms provided by the authors are available
is very different from the healthful diets recom- with the full text of this article at NEJM.org.

References
1. Alberti KG, Eckel RH, Grundy SM, et al.
Harmonizing the metabolic syndrome:
a joint interim statement of the Interna-
tional Diabetes Federation Task Force on
Epidemiology and Prevention; National
Heart, Lung, and Blood Institute; Ameri-
can Heart Association; World Heart Feder-
ation; International Atherosclerosis Soci-
ety; and International Association for the
Study of Obesity. Circulation 2009;120:
1640-5.
2. Chobanian AV, Bakris GL, Black HR,
et al. Seventh report of the Joint National
Committee on Prevention, Detection, Eval-
uation, and Treatment of High Blood
Pressure. Hypertension 2003;42:1206-52.
3. Lawes CM, Vander Hoorn S, Rodgers A.
Global burden of blood-pressure-related
disease, 2001. Lancet 2008;371:1513-8.
4. Rao MV, Qiu Y, Wang C, Bakris G.
Hypertension and CKD: Kidney Early
Evaluation Program (KEEP) and National
Health and Nutrition Examination Survey
(NHANES), 1999-2004. Am J Kidney Dis
2008;51:Suppl 2:S30-S37.
5. Hajjar I, Kotchen JM, Kotchen TA. Hy-
pertension: trends in prevalence, incidence
and control. Annu Rev Public Health
2006;27:465-90.
6. Ezzati M, Oza S, Danaei G, Murray CJ.
Trends and cardiovascular mortality ef-
fects of state-level blood pressure and
uncontrolled hypertension in the United
States. Circulation 2008;117:905-14.
7. Mellen PB, Gao SK, Vitolins MZ, Goff
DC Jr. Deteriorating dietary habits among
adults with hypertension: DASH dietary
accordance, NHANES 1988-1994 and 1999-
2004. Arch Intern Med 2008;168:308-14.
8. Blanck HM, Gillespie C, Kimmons JE,
Seymour JD, Serdula MK. Trends in fruit
and vegetable consumption among U.S.
men and women, 1994-2005. Prev Chron-
ic Dis 2008;5:A35.
9. Okosun IS, Prewitt TE, Cooper RS.
Abdominal obesity in the United States:
prevalence and attributable risk of hyper-
tension. J Hum Hypertens 1999;13:425-30.
10. Saile F. Uber den einflus der vege-
tarischen ernahrung auf den blutdruck.
Med Klin 1930;25:929-31.
11. Sacks FM, Rosner B, Kass EH. Blood
pressure in vegetarians. Am J Epidemiol
1974;100:390-8.
12. Sacks FM, Kass EH. Low blood pres-
sure in vegetarians: the effects of specific
foods and nutrients. Am J Clin Nutr 1988;
48:Suppl:795-800.
13. Epstein FH, Eckoff RD. The epidemi-
ology of high blood pressure geo-
graphic distributions and etiological fac-
tors. In: Stamler J, Stamler R, Pullman TN,
eds. The epidemiology of hypertension.
New York: Grune & Stratton, 1967:155-66.
14. Rodriguez BL, Labarthe DR, Huang B,
Lopez-Gomez J. Rise in blood pressure
with age: new evidence of population dif-
ferences. Hypertension 1994;24:779-85.
15. Elliott P, Stamler J, Nichols R, et al.
Intersalt revisited: further analyses of 24
hour sodium excretion and blood pressure
within and across populations. BMJ 1996;
312:1249-53. [Erratum, BMJ 1997;315:458.]
16. Adrogu HJ, Madias NE. Sodium and
potassium in the pathogenesis of hyper-
tension. N Engl J Med 2007;356:1966-78.
17. Primary hypertension: pathogenesis.
In: Kaplan NM. Kaplans clinical hyper-
tension. 9th ed. Philadelphia: Lippincott
Williams & Wilkins, 2006:50-121.
18. OShaughnessy KM, Karet FE. Salt
handling and hypertension. Annu Rev Nutr
2006;26:343-65.
19. Blacher J, Safar ME. Large-artery stiff-
ness, hypertension and cardiovascular risk
in older patients. Nat Clin Pract Cardio-
vasc Med 2005;2:450-5.
20. Berk BC. Biology of the vascular wall in
hypertension. In: Brenner BM, ed. Brenner
& Rectors the kidney. 7th ed. Vol. 2. Phil-
adelphia: Saunders Elsevier, 2004:1999-
2022.
21. Gates PE, Strain WD, Shore AC. Hu-
man endothelial function and microvas-
cular ageing. Exp Physiol 2009;94:311-6.
22. de Jongh RT, Sern EH, Ijzerman RG,
Stehouwer CD. Microvascular function:
a potential link between salt sensitivity,
insulin resistance and hypertension. J Hy-
pertens 2007;25:1887-93.
23. Lee MA, Bhm M, Paul M, Ganten D.
Tissue renin-angiotensin systems: their

2110 n engl j med 362;22 nejm.org june 3, 2010

BACK to TOC
88

role in cardiovascular disease. Circulation
1993;87:Suppl:IV-7IV-13.
24. Re RN. Mechanisms of disease: local
renin-angiotensin-aldosterone systems and
the pathogenesis and treatment of cardio-
vascular disease. Nat Clin Pract Cardio-
vasc Med 2004;1:42-7.
25. Landsberg L, Young JB. Diet and the
sympathetic nervous system: relationship
to hypertension. Int J Obes 1981;5:Suppl
1:79-91.
26. Esler M, Straznicky N, Eikelis N,
Masuo K, Lambert G, Lambert E. Mecha-
nisms of sympathetic activation in obesity-
related hypertension. Hypertension 2006;
48:787-96.
27. Victor RG, Shafiq MM. Sympathetic
neural mechanisms in human hyperten-
sion. Curr Hypertens Rep 2008;10:241-7.
28. Feldstein C, Julius S. The complex in-
teraction between overweight, hyperten-
sion, and sympathetic overactivity. J Am
Soc Hypertens 2009;3:353-65.
29. Hall JE, Hildebrandt DA, Kuo J. Obe-
sity hypertension: role of leptin and sym-
pathetic nervous system. Am J Hypertens
2001;14:103S-115S.
30. Hall JE. The kidney, hypertension, and
obesity. Hypertension 2003;41:625-33.
31. Seals DR, Moreau KL, Gates PE, Es-
kurza I. Modulatory influences on ageing
of the vasculature of healthy humans. Exp
Gerontol 2006;41:501-7.
32. Zandi-Nejad K, Luyckx VA, Brenner
BM. Adult hypertension and kidney dis-
ease: the role of fetal programming. Hyper-
tension 2006;47:502-8.
33. Sealey JE, Blumenfeld JD, Bell GM,
Pecker MS, Sommers SC, Laragh JH. On
the renal basis for essential hypertension:
nephron heterogeneity with discordant
renin secretion and sodium excretion
causing a hypertensive vasocontriction-
volume relationship. J Hypertens 1988;6:
763-77.
34. Safar ME, Temmar M, Kakou A, Lacol-
ley P, Thornton SN. Sodium intake and
vascular stiffness in hypertension. Hyper-
tension 2009;54:203-9.
35. Engeli S, Bhnke J, Gorzelniak K, et al.
Weight loss and the renin-angiotensin-
aldosterone system. Hypertension 2005;45:
356-62.
36. Ho JT, Keogh JB, Bornstein SR, et al.
Moderate weight loss reduces renin and
aldosterone but does not influence basal
or stimulated pituitary-adrenal axis func-
tion. Horm Metab Res 2007;39:694-9.
37. Straznicky NE, Lambert EA, Lambert
GW, Masuo K, Esler MD, Nestel PJ. Effects
of dietary weight loss on sympathetic ac-
tivity and cardiac risk factors associated
with the metabolic syndrome. J Clin Endo-
crinol Metab 2005;90:5998-6005.
38. Straznicky NE, Lambert EA, Nestel PJ,
et al. Sympathetic neural adaptation to
hypocaloric diet with or without exercise
training in obese metabolic syndrome sub-
jects. Diabetes 2010;59:71-9.
n engl j med 362;22
BACK to TOC
clinical ther apeutics
39. Rocchini AP, Key J, Bondie D, et al.
The effect of weight loss on the sensitivity
of blood pressure to sodium in obese ado-
lescents. N Engl J Med 1989;321:580-5.
40. Pierce GL, Beske SD, Lawson BR, et al.
Weight loss alone improves conduit and
resistance artery endothelial function in
young and older overweight/obese adults.
Hypertension 2008;52:72-9.
41. Akita S, Sacks FM, Svetkey LP, Conlin
PR, Kimura G. Effects of the Dietary Ap-
proaches to Stop Hypertension (DASH)
diet on pressure-natriuresis relationship.
Hypertension 2003;42:8-13.
42. Al-Solaiman Y, Jesri A, Zhao Y, Morrow
JD, Egan BM. Low-sodium DASH reduces
oxidative stress and improves vascular
function in salt-sensitive humans. J Hum
Hypertens 2009;23:826-35.
43. Rallidis LS, Lekakis J, Kolomvotsou A,
et al. Close adherence to a Mediterranean
diet improves endothelial function in sub-
jects with abdominal obesity. Am J Clin
Nutr 2009;90:263-8.
44. McCall DO, McGartland CP, McKinley
MC, et al. Dietary intake of fruits and veg-
etables improves microvascular function in
hypertensive subjects in a dose-dependent
manner. Circulation 2009;119:2153-60.
45. Sacks FM, Obarzanek E, Windhauser
MM, et al. Rationale and design of the
Dietary Approaches to Stop Hypertension
trial (DASH): a multicenter controlled-
feeding study of dietary patterns to lower
blood pressure. Ann Epidemiol 1995;5:
108-18.
46. Swain JF, McCarron PB, Hamilton EF,
Sacks FM, Appel LJ. Characteristics of the
dietary patterns tested in the Optimal Mac-
ronutrient Intake Trial to Prevent Heart
Disease (OmniHeart): options for a heart-
healthy diet. J Am Diet Assoc 2008;108:
257-65.
47. Trichopoulou A. Mediterranean diet:
the past and the present. Nutr Metab Car-
diovasc Dis 2001;11:Suppl:1-4.
48. Kokkinos P, Panagiotakos DB, Poly-
chronopoulos E. Dietary influences on
blood pressure: the effect of the Mediter-
ranean diet on the prevalence of hyperten-
sion. J Clin Hypertens (Greenwich) 2005;
7:165-70.
49. Appel LJ, Moore TJ, Obarzanek E, et al.
The effect of dietary patterns on blood
pressure: results from the Dietary Ap-
proaches to Stop Hypertension trial.
N Engl J Med 1997;336:1117-24.
50. Appel LJ, Sacks FM, Carey VJ, et al.
Effects of protein, monounsaturated fat,
and carbohydrate intake on blood pressure
and serum lipids: results of the Omni-
Heart randomized trial. JAMA 2005;294:
2455-64.
51. Esposito K, Marfella R, Ciotola M, et al.
Effect of a Mediterranean-style diet on
endothelial dysfunction and markers of
vascular inflammation in the metabolic
syndrome: a randomized trial. JAMA
2004;292:1440-6.
nejm.org june 3, 2010
52. Esposito K, Maiorino MI, Ciotola M,
et al. Effects of a Mediterranean-style diet
on the need for antihyperglycemic drug
therapy in patients with newly diagnosed
type 2 diabetes: a randomized trial. Ann
Intern Med 2009;151:306-14. [Erratum,
Ann Intern Med 2009;151:591.]
53. Sacks FM, Svetkey LP, Vollmer WM,
et al. Effects on blood pressure of reduced
dietary sodium and the Dietary Approach-
es to Stop Hypertension (DASH) diet.
N Engl J Med 2001;344:3-10.
54. Bray GA, Vollmer WM, Sacks FM,
Obarzanek E, Svetkey LP, Appel LJ. A fur-
ther subgroup analysis of the effects of
the DASH diet and three dietary sodium
levels on blood pressure: results of the
DASH-Sodium Trial. Am J Cardiol 2004;
94:222-7. [Erratum, Am J Cardiol 2010;105:
579.]
55. Goff DC, Howard G, Russell GB, La-
barthe DR. Birth cohort evidence of popu-
lation influences on blood pressure in the
United States, 1887-1994. Ann Epidemiol
2001;11:271-9.
56. Uzu T, Kimura G, Yamauchi A, et al.
Enhanced sodium sensitivity and disturbed
circadian rhythm of blood pressure in es-
sential hypertension. J Hypertens 2006;24:
1627-32.
57. Chen J, Gu D, Huang J, et al. Meta-
bolic syndrome and salt sensitivity of
blood pressure in non-diabetic people in
China: a dietary intervention study. Lan-
cet 2009;373:829-35.
58. Obarzanek E, Proschan MA, Vollmer
WM, et al. Individual blood pressure re-
sponses to changes in salt intake: results
from the DASH-Sodium trial. Hyperten-
sion 2003;42:459-67.
59. Blumenthal JA, Babyak MA, Hinder-
liter A, et al. Effects of the DASH diet
alone and in combination with exercise
and weight loss on blood pressure and car-
diovascular biomarkers in men and women
with high blood pressure: the ENCORE
study. Arch Intern Med 2010;170:126-35.
60. Conlin PR, Erlinger TP, Bohannon A,
et al. The DASH diet enhances the blood
pressure response to losartan in hyper-
tensive patients. Am J Hypertens 2003;16:
337-42.
61. Whelton PK, Appel LJ, Espeland MA,
et al. Sodium reduction and weight loss in
the treatment of hypertension in older
persons: a randomized controlled trial of
nonpharmacologic interventions in the
elderly (TONE). JAMA 1998;279:839-46.
62. Pimenta E, Gaddam KK, Oparil S, et al.
Effects of dietary sodium reduction on
blood pressure in subjects with resistant
hypertension. Hypertension 2009;54:475-
81.
63. Grimes CA, Riddell LJ, Nowson CA.
Consumer knowledge and attitudes to
salt intake and labelled salt information.
Appetite 2009;53:189-94.
64. Karanja N, Lancaster KJ, Vollmer WM,
et al. Acceptability of sodium-reduced re-
2111
89
 clinical ther apeutics

search diets, including the Dietary Ap-
proaches to Stop Hypertension diet, among
adults with prehypertension and stage 1
hypertension. J Am Diet Assoc 2007;107:
1530-8.
65. Salt-assault: brand-name comparisons
of processed foods. 2nd ed. Washington,
DC: Center for Science and the Public
Interest, 2008:1-31. (Accessed May 7,
2010, at http://www.cspinet.org/new/pdf/
saltupdatedec08.pdf.)
66. Pear R. New health initiatives put spot-
light on prevention. New York Times. April
5, 2010:A10. (Accessed May 7, 2010, at
http://www.nytimes.com/2010/04/05/
health/policy/05health.html.)
67. Rankins J, Sampson W, Brown B,
Jenkins-Salley T. Dietary Approaches to
Stop Hypertension (DASH) intervention
reduces blood pressure among hyperten-
sive African American patients in a neigh-
borhood health care center. J Nutr Educ
Behav 2005;37:259-64.
68. Svetkey LP, Stevens VJ, Brantley PJ, et al.
Comparison of strategies for sustaining
weight loss: the weight loss maintenance
randomized controlled trial. JAMA 2008;
299:1139-48.
69. Hsieh YC, Hung CT, Lien LM, et al.
A significant decrease in blood pressure
through a family-based nutrition health
education programme among community
residents in Taiwan. Public Health Nutr
2008;12:570-7.
70. Wood DA, Kotseva K, Connolly S, et
al. Nurse-coordinated multidisciplinary,
family-based cardiovascular disease pre-
vention programme (EUROACTION) for
patients with coronary heart disease and
asymptomatic individuals at high risk of
cardiovascular disease: a paired, cluster-
randomised controlled trial. Lancet 2008;
371:1999-2012.
71. Sacks FM, Bray GA, Carey VJ, et al.
Comparison of weight-loss diets with dif-
ferent compositions of fat, protein, and
carbohydrates. N Engl J Med 2009;360:
859-73.
72. Eriksson MK, Franks PW, Eliasson M.
A 3-year randomized trial of lifestyle inter-
vention for cardiovascular risk reduction
in the primary care setting: the Swedish
Bjrkns study. PLoS One 2009;4(4):e5195.
73. Oldways Web site. (Accessed May 7,
2010, at http://www.oldwayspt.org.)
74. Mediterranean Foods Alliance Web
site. (Accessed May 7, 2010, at http://
mediterraneanmark.org/index.html.)
75. MayoClinic.com. DASH diet recipes.
(Accessed May 7, 2010, at http://www
.mayoclinic.com/health/dash-diet-recipes/
RE00089.)
76. DASH for Health Web site. (Accessed
May 7, 2010, at http://www.dashforhealth
.com.)
77. Katzen M, Willett W. Eat, drink, and
weigh less: a flexible and delicious way to
shrink your waist without going hungry.
New York: Hyperion, 2006.
78. Your guide to lowering your blood
pressure with DASH. Bethesda, MD: Na-
tional Heart, Lung, and Blood Institute,
2006.
79. Moore T, Svetkey L, Lin PH, Karanja
N. The DASH diet for hypertension. New
York: Free Press, 2001:1-264.
80. American Heart Association. No-fad
diet: a personal plan for healthy weight
loss. New York: Clarkson Potter, 2005.
81. Young CM, Batch BC, Svetkey LP. Ef-
fect of socioeconomic status on food
availability and cost of the Dietary Ap-
proaches to Stop Hypertension (DASH)
dietary pattern. J Clin Hypertens (Green-
wich) 2008;10:603-11.
82. Williamson DA, Anton SD, Han H, et
al. Adherence is a multi-dimensional con-
struct in the POUNDS LOST trial. J Behav
Med 2010;33:35-46.
83. Cook NR, Cutler JA, Obarzanek E, et al.
Long term effects of dietary sodium reduc-
tion on cardiovascular disease outcomes:
observational follow-up of the Trials of
Hypertension Prevention (TOHP). BMJ
2007;334:885-8.
84. Chang HY, Hu YW, Yue CS, et al. Ef-
fect of potassium-enriched salt on cardio-
vascular mortality and medical expenses
of elderly men. Am J Clin Nutr 2006;83:
1289-96.
85. Appel LJ, Brands DW, Daniels SR,
Karanja N, Elmer PJ, Sacks FM. Dietary
approaches to prevent and treat hyperten-
sion: a scientific statement from the Amer-
ican Heart Association. Hypertension
2006;47:296-308.
86. Lichtenstein AH, Appel LJ, Brands M,
et al. Diet and lifestyle recommendations
revision 2006: a scientific statement from
the American Heart Association Nutrition
Committee. Circulation 2006;114:82-96.
[Errata, Circulation 2006;114(1):e27, 114(23):
e629.]
Copyright 2010 Massachusetts Medical Society.

COLLECTIONS OF ARTICLES ON THE JOURNALS WEB SITE

The Journals Web site (NEJM.org) sorts published articles into
more than 50 distinct clinical collections, which can be used as convenient
entry points to clinical content. In each collection, articles are cited in reverse
chronologic order, with the most recent first.

2112 n engl j med 362;22 nejm.org june 3, 2010

BACK to TOC
90
 The n e w e ng l a n d j o u r na l of m e dic i n e

clinical therapeutics

Mitral-Valve Repair for Mitral-Valve Prolapse

Subodh Verma, M.D., Ph.D., and Thierry G. Mesana, M.D., Ph.D.
This Journal feature begins with a case vignette that includes a therapeutic recommendation. A discussion
of the clinical problem and the mechanism of benefit of this form of therapy follows. Major clinical studies,
the clinical use of this therapy, and potential adverse effects are reviewed. Relevant formal guidelines,
if they exist, are presented. The article ends with the authors clinical recommendations.

A 55-year-old man with a holosystolic murmur of increasing intensity has been seen
regularly by his family physician for the past 3 years. He is referred to a cardiologist.
The patient reports no shortness of breath, chest pain, or palpitations. An electrocar-
diogram shows normal sinus rhythm. A transthoracic echocardiogram reveals se-
vere, anteriorly directed mitral regurgitation with isolated prolapse of the middle
scallop of the posterior leaflet. Flow reversal is detected in the pulmonary veins. The
calculated regurgitant volume is 75 ml, the regurgitant fraction 63%, and the effec-
tive regurgitant orifice 53 mm2, features consistent with severe mitral regurgitation.
The transthoracic echocardiogram also shows mildly depressed left ventricular func-
tion (ejection fraction, 58%), slightly elevated left ventricular dimensions (end-systolic
dimension, 42 mm), and normal right ventricular systolic pressure. The patient is
referred to a cardiac surgeon for consideration of mitral-valve repair.

The Cl inic a l Probl em

Mitral-valve prolapse is defined as the displacement of some portion of one or both From the Division of Cardiac Surgery, St.
leaflets of the mitral valve into the left atrium during systole. In developed coun- Michaels Hospital, University of Toronto,
Toronto (S.V.); and the Division of Cardiac
tries, it is the most common cause of chronic mitral regurgitation; in a study of the Surgery, Ottawa Heart Institute, Univer-
Framingham Offspring Study cohort, the prevalence of mitral-valve prolapse was sity of Ottawa, Ottawa (T.G.M.). Address
2.5%.1 More than 150 million people worldwide may be affected.2-4 The disorder reprint requests to Dr. Verma at St. Mi-
chaels Hospital, University of Toronto,
has both genetic and acquired forms, and several chromosomal loci for autosomal 30 Bond St., Toronto, ON M5B 1W8, Can-
dominant mitral-valve prolapse have been identified.5-9 Although mitral-valve pro- ada, or at subodh.verma@sympatico.ca.
lapse is more common in women, more men are referred for surgery4; whether this
N Engl J Med 2009;361:2261-9.
reflects a difference between the sexes in the morphologic features or natural his- Copyright 2009 Massachusetts Medical Society.
tory of the disorder or referral bias is unclear.
The natural history of mitral-valve prolapse is heterogeneous and is largely
determined by the severity of mitral regurgitation. Although a majority of patients
remain asymptomatic and may have a near-normal life expectancy, approximately
5 to 10% have progression to severe mitral regurgitation.10,11 Left untreated, mitral-
valve prolapse with severe mitral regurgitation results in limiting symptoms, left
ventricular dysfunction, heart failure, pulmonary hypertension, and atrial fibrilla-
tion. Spontaneous rupture of mitral chordae may occur, and endocarditis and stroke
are serious complications. The mortality rate of persons who have mitral-valve
prolapse with severe mitral regurgitation is approximately 6 to 7% per year.12,13

Pathoph ysiol o gy a nd the Effec t of Ther a py

The mitral valve and subvalvular apparatus include the annulus, valve leaflets, chor-
dae tendineae, papillary muscles, and left ventricular wall. The valve has anterior
and posterior leaflets, and each leaflet typically consists of three discrete segments
or scallops. These are designated P1, P2, and P3 in the posterior mitral-valve leaflet,

n engl j med 361;23 nejm.org december 3, 2009 2261

BACK to TOC
91
 The n e w e ng l a n d j o u r na l of m e dic i n e

A
Aortic valve

Left coronary
artery Right fibrous
trigone

Bundle of
His
Left fibrous
trigone

AC Anterior leaflet PC

Circumflex
artery
Posterior leaflet
Coronary
Annulus sinus

B C Annulus

Leaflet

Secondary
cord
AC A3 PC
A1 Primary
A2 P3 cord
P1

P2
Papillary
muscle

Figure 1. The Mitral Valve.

The mitral valve has anterior and posterior leaflets, which are separated by the anterior commissure (AC) and the
posterior commissure (PC) (Panel A). The leaflets are inserted on the circumference of the mitral annulus, which is
in continuity with the aortic annulus and the left and right fibrous trigones. The circumflex coronary artery, coronary
sinus, aortic valve, and bundle of His are all close to the mitral valve. Panel B shows the mitral-valve leaflets, each of
which usually consists of three discrete segments or scallops. These are designated A1, A2, and A3 for the anterior
leaflet and P1, P2, and P3 for the posterior leaflet. The valve leaflets each receive chordae tendineae from the anter-
olateral and posteromedial papillary muscles (Panel C). Primary chordae are attached to the free edge of the valve
leaflet, and secondary chordae are attached to the ventricular surface of the leaflet.

and A1, A2, and A3 in the anterior leaflet (Fig. 1). muscles. Competence of the mitral valve relies on
The valve leaflets receive chordae tendineae from coordinated interaction of the valve and subvalvu-
the anterolateral and posteromedial papillary lar apparatus. During systole, the papillary mus-

2262 n engl j med 361;23 nejm.org december 3, 2009

BACK to TOC
92
 clinical ther apeutics
cles contract, increasing tension on the chordae
tendineae and preventing the valve leaflets from
everting into the left atrium.
Mitral-valve prolapse is characterized predom-
inantly by myxomatous degeneration. In younger
patients, the disease is often manifested by excess
leaflet tissue and is known as Barlows syndrome,
the most extreme form of myxomatous degen-
eration. On the other hand, in older patients, the
prolapsing mitral valve tends not to have excess
leaflet tissue, an entity known as fibroelastic
deficiency. Both conditions can lead to leaflet
prolapse and chordal elongation or rupture, repre-
senting the spectrum of degenerative mitral-valve
disease.14 These anatomic abnormalities result
in the mitral orifice not closing completely dur-
ing systole, causing regurgitation. Annular dila-
tation may also develop over time, leading to
further progression of mitral regurgitation.
Patients with mild-to-moderate mitral regur-
gitation from mitral-valve prolapse may remain
asymptomatic and without clinical deterioration
for many years. However, increasing severity of
mitral regurgitation, even among asymptomatic
patients, imposes a volume load on the left ven-
tricle, which, if sustained over time, results in
ventricular dilatation, hypertrophy, neurohumoral
activation, and heart failure. In addition, elevation
in the mean left atrial pressure leads to left
atrial enlargement, atrial fibrillation, pulmonary
congestion, and pulmonary hypertension.
The goal of surgical correction for mitral-valve
prolapse is to restore a competent mitral valve.
There are two options for surgical correction of
severe mitral regurgitation due to mitral-valve
prolapse: valve replacement or valve repair.
Mitral-valve replacement can be performed
with the use of either a mechanical or a biologic
prosthesis. However, there are several drawbacks
to mitral-valve replacement. These include the
need for lifelong anticoagulation therapy and the
risk of thromboembolism with the use of me-
chanical valves; the risk of prosthetic-valve dete-
rioration and failure with the use of bioprosthetic
valves; and the risk of prosthetic-valve endo-
carditis. In addition, if the chordae tendineae are
severed during surgery, the ventricular wall is no
longer anchored to the valve apparatus, and the
tethering effect of the chordae is lost. As a re-
sult, left ventricular wall stress increases and left
ventricular function deteriorates.15-19 The goals
of mitral-valve repair are to obtain a proper line
of coaptation on both leaflets, to correct annular
n engl j med 361;23 nejm.org
BACK to TOC
dilatation, and to preserve (or repair, if neces-
sary) the subvalvular apparatus.
Cl inic a l E v idence
We are unaware of any randomized trials that
have compared medical management to surgery
for severe mitral regurgitation due to mitral-valve
prolapse. However, evidence from observational
series strongly suggests that surgical interven-
tion is beneficial.12,20-22 One study evaluated the
effect of early surgery on long-term outcomes in
221 patients who had mitral regurgitation with
flail leaflets.20 The 63 patients undergoing sur-
gery within 1 month after diagnosis had a sig-
nificantly better 10-year survival rate than those
whose mitral regurgitation was managed con-
servatively (79% vs. 65%; adjusted risk ratio, 0.30;
95% confidence interval [CI], 0.12 to 0.71;
P = 0.008). In another report, 394 patients with
mitral regurgitation and flail leaflets were stud-
ied.21 During a median follow-up period of 3.9
years, the linearized mortality rate associated with
nonsurgical management was 2.6% per year.
Mitral-valve surgery was performed in 315 pa-
tients (repair in 250, replacement in 65). Surgical
intervention was independently associated with a
reduced risk of death (adjusted hazard ratio for
death, 0.42; 95% CI, 0.21 to 0.84; P = 0.01).
To our knowledge, there are also no random-
ized trials comparing mitral-valve repair with
replacement. Again, however, data from obser-
vational studies suggest a benefit of mitral re-
pair.23-26 A meta-analysis of 29 studies compared
mitral-valve repair with replacement for various
conditions, including myxomatous degenera-
tion.23 Mitral-valve replacement was associated
with lower survival than was repair (hazard ratio
for death, 1.58; 95% CI, 1.41 to 1.78).
In a study from Finland, mitral-valve repair
was compared with replacement in 184 consecu-
tive patients who were followed for a mean of 7.3
years.24 There was a significant survival benefit
for the patients who underwent mitral-valve
repair as compared with those who underwent
replacement (5-year survival, 81.2% vs. 73.5%),
which persisted after adjustment for baseline
propensity score (P = 0.02). In contrast, in a report
from the Cleveland Clinic, 3286 patients who
underwent an isolated primary operation for
degenerative mitral-valve disease (mitral repair,
93%; mitral replacement, 7%) between 1985 and
2005 were studied.25 Propensity scoring was
december 3, 2009 2263
93
 The n e w e ng l a n d j o u r na l
used to select 195 matched pairs for analysis.
Among the propensity-matched patients, there
was no significant difference in survival at 5, 10,
or 15 years.
Cl inic a l Use
Patients with mitral-valve prolapse should have a
careful assessment of symptoms and should un-
dergo electrocardiography (primarily to evaluate
cardiac rhythm) and transthoracic echocardiog-
raphy to assess the mechanism and severity of
mitral regurgitation, as well as left ventricular
size and function. A semiquantitative scale is
often used to grade mitral regurgitation: 1+ (trace),
2+ (mild), 3+ (moderate), and 4+ (severe). How-
ever, quantitative Doppler assessments are recom-
mended to define severe mitral regurgitation
more precisely; these variables include a regurgi-
tant volume of at least 60 ml, a regurgitant frac-
tion of at least 50%, and an effective regurgitant
orifice of at least 40 mm2.27
Patients who have severe mitral regurgitation
with symptoms or with left ventricular dysfunc-
tion (ejection fraction, <60%), dilatation (left
ventricular end-systolic dimension, >40 mm), or
both should be offered surgery.28,29 Likewise,
asymptomatic patients without left ventricular
dysfunction or dilatation but with atrial fibrilla-
tion or pulmonary hypertension should be con-
sidered for surgery. Asymptomatic persons with
mild-to-moderate mitral regurgitation and no
evidence of left ventricular dysfunction or dilata-
tion should be observed until the development of
either symptoms or severe mitral regurgitation.
Before the advent of mitral-valve repair, valve
replacement was the preferred procedure for se-
vere mitral regurgitation. Valve replacement may
still be preferred in certain situations, such as in
patients with advanced age, infective endocardi-
tis, a requirement for a combined or complex
surgical procedure, or extensive calcifications of
the leaflets or annulus. In such cases, chordal-
sparing valve replacement for mitral regurgitation
may be a suitable alternative to repair.
Individual and institutional experience is cru-
cial in determining the likelihood of success of
a repair procedure. High-volume centers have the
lowest mortality rates and the highest propor-
tion of patients undergoing mitral-valve repair
rather than replacement.30 In counseling the pa-
tient, the surgeon should precisely evaluate the
likelihood of successful repair in light of his or
2264 n engl j med 361;23
BACK to TOC
of m e dic i n e
her own experience and may recommend a sec-
ond opinion. If there is a possibility that intra-
operative conversion to mitral replacement may
be necessary, the decision between a mechanical
valve and a bioprosthesis should be discussed
with the patient before the operation.
Mitral-valve surgery is not recommended in
patients with clinically significant coexisting
conditions, such as advanced respiratory, hepatic,
or renal dysfunction, or those with marked extra-
cardiac arteriopathy or recent cerebrovascular
events. Depressed left ventricular function is an
independent predictor of poor outcomes but is
not a contraindication to mitral-valve repair.31 In
patients with coexisting coronary artery disease,
mitral-valve repair combined with coronary-
artery bypass surgery should be the procedure of
choice.32 Two validated scoring systems for de-
termining risk during cardiac surgery are com-
monly used to determine perioperative risk.33,34
We routinely perform intraoperative trans-
esophageal echocardiography during all mitral-
valve repair procedures.28,29 Transesophageal
echocardiography provides precise anatomic and
functional information that is helpful in plan-
ning the operation, including the extent of leaf-
let deformity, the mechanism and severity of
mitral regurgitation, the condition of the subval-
vular apparatus, the diameter of the mitral an-
nulus, left atrial dimensions, and ventricular
function.35
Successful mitral-valve repair encompasses
four general principles.36 First, repair must re-
store an adequate surface of coaptation of both
leaflets in systole.14,37 Second, full leaflet motion
should be restored or preserved. Third, to pre-
vent progressive dilatation, an annuloplasty ring
or band should be used to reinforce the repair by
stabilizing the annulus. Mitral-valve repair with-
out annuloplasty reinforcement is not recom-
mended. Last, the surgeon should ensure that no
more than trace-to-mild mitral regurgitation is
present at the completion of the repair.
In patients with isolated prolapse of the pos-
terior middle scallop (P2), which is encountered
in the majority of patients with degenerative mi-
tral regurgitation, repair usually involves limited
resection of this scallop, including the removal
of the minimum number possible of adjacent
chordae and supporting apparatus. The remain-
ing segments of the posterior leaflet, namely P1
and P3, are then brought together (Fig. 2). If
excessive posterior-leaflet tissue is present, the
nejm.org december 3, 2009
94
 clinical ther apeutics

A Isolated prolapse of B
the posterior middle scallop
Excessive
tissue
A1 B1

P1 P3
P2

Incision line
Incision lines

A2 B2

Resected Plication Resected

area sutures area

A3 B3

Reapproximation Completed
of remaining tissue sliding plasty

A4 B4

Annuloplasty Annuloplasty
ring ring

Figure 2. Mitral-Valve Prolapse.

The most common leaflet abnormality seen in mitral-valve prolapse is isolated prolapse of the posterior middle
scallop (P2) (Panel A1). In patients with isolated prolapse of P2, repair usually involves limited resection of this
scallop by means of a quadrangular or triangular incision (Panel A2). The remaining parts of the posterior leaflet,
namely P1 and P3, are then brought together (Panel A3). After the leaflet repair is complete, an annuloplasty ring
or band is used to reinforce and stabilize the annulus, thus preventing progressive dilatation (Panel A4). If excessive
posterior leaflet tissue is present (Panel B1), the height of the posterior leaflet is reduced by incising P1 and P3 from
the annulus (Panel B2), followed by reapproximation of the free edges (sliding plasty) (Panel B3). After the leaflet
repair is complete, an annuloplasty ring or band is inserted (Panel B4).

n engl j med 361;23 nejm.org december 3, 2009 2265

BACK to TOC
95
 The n e w e ng l a n d j o u r na l
height of the posterior leaflet is reduced by inci-
sions in P1 and P3, followed by reapproximation
of the free edges (sliding plasty) (Fig. 2). Finally,
the annulus, which is distorted or dilated or
both, is stabilized with an annuloplasty ring or
band (Fig. 2). Limited resection, artificial chordal
replacement (with Gore-Tex expanded polytetra-
fluoroethylene sutures), or both may be appro-
priate, followed by annuloplasty reinforcement,
in cases of mitral-valve prolapse without redun-
dant leaflet tissue.
Repairs of the anterior leaflet, either in isola-
tion or with concomitant posterior leaflet repair,
are more complex procedures that are best han-
dled by surgeons who are experienced in mitral
repair. Various techniques may be used, including
limited triangular resection of the anterior leaf-
let, chordal transposition, chordal shortening,
artificial (Gore-Tex) chordal replacement, and
edge-to-edge repair10,21,28,38-41 (Fig. 3).
The repair is assessed initially by visual inspec-
tion and by injecting saline through the mitral
valve to look for regurgitation (the saline test),
and then by intraoperative transesophageal echo-
cardiography after the patient is weaned from
cardiopulmonary bypass. Patients should not
leave the operating theater with more than 1+
mitral regurgitation on transesophageal echo-
cardiography.36,42 Since anesthesia may result in
substantial changes in preload and afterload, it
is important to perform the intraoperative trans-
esophageal echocardiography under conditions
that approximate postoperative conditions in a
patient who is awake. This can be achieved by
adjusting inotropes and vasopressors to raise the
afterload and blood pressure.
After mitral-valve repair, the left ventricle
must be able to eject the entire stroke volume
into the aorta. This constitutes a substantial in-
crease in afterload as compared with ejection
into the left atrium. Therefore, afterload reduc-
tion is important to maintain optimal cardiac
output. In addition, because myocardial dysfunc-
tion may be present (even in patients with an
apparently normal preoperative ejection frac-
tion),43 inotropic support may be necessary to
improve contractility. Patients with a low preop-
erative ejection fraction and heart failure may
require more intensive treatment to allow the left
ventricle to recover, including temporary pacing,
intraaortic balloon counterpulsation, or in rare
cases, support with a ventricular assist device.
2266 n engl j med 361;23
BACK to TOC
of m e dic i n e
In the absence of preoperative atrial fibrilla-
tion, and if normal sinus rhythm is maintained
throughout hospital admission, aspirin alone may
be sufficient for patients who had mitral-valve
repair with ring annuloplasty. Otherwise, patients
typically undergo anticoagulation with warfarin
for 3 months, with a target international normal-
ized ratio of 2.0 to 2.5. Antibiotic prophylaxis for
dental procedures is recommended in all patients
receiving an annuloplasty ring or other prosthetic
material.44
There are currently no standard recommenda-
tions regarding postoperative echocardiographic
follow-up after mitral-valve repair. It is customary
at our center to perform transthoracic echocar-
diography once before discharge and again at 6 to
8 weeks after discharge. Usually patients are then
transferred to the care of their cardiologist and
family physician, and we recommend that echo-
cardiography be performed annually thereafter.
We estimate that the overall costs for mitral-
valve repair, including hospital admission, profes-
sional fees, operating time, and prosthetic mate-
rial (annuloplasty ring or band), are currently
approximately $40,000 at our institution. Data
from the Nationwide Inpatient Sample indicate
that the mean estimated institutional cost for
mitral repair in the United States increased from
$28,405 in 2001 to $38,642 in 2005.45
A dv er se Effec t s
Mitral-valve repair is associated with an operative
mortality of 3% or less.22,38,46-50 This figure is
nearer 1% in high-volume centers.30 The most
common cause of death is heart failure. Predic-
tors of death include advanced age, poorer New
York Heart Association class, atrial fibrillation,
lower preoperative ejection fraction, greater pre-
operative left ventricular end-systolic dimension,
and coexisting conditions including diabetes, renal
disease, chronic lung disease, and obesity.22,38,51,52
In an analysis from the Society of Thoracic
Surgeons National Adult Cardiac Surgery Data-
base,52 major postoperative complications before
discharge included prolonged (>24 hours) venti-
latory support (7.3% of patients), renal failure
(2.6%), and stroke (1.4%). Reoperation during
initial hospitalization was required in 6.3% of
patients. Thromboembolism after mitral-valve
repair occurs in approximately 5% of patients
within the first 5 years after surgery.38,39
nejm.org december 3, 2009
96
 clinical ther apeutics

Figure 3. Special Repair Techniques.

A Chordal replacement
Ruptured chordae tendineae can be replaced with an
artificial substitute (Gore-Tex expanded polytetrafluoro-
ethylene sutures) (Panel A). Ruptured or surgically severed
primary chordae can be replaced with secondary chordae,
a process called chordal transfer (Panel B). Edge-to-edge
repair (Panel C) is performed by sewing the anterior
and posterior leaflets together at the central points of
their middle scallops, which corrects the prolapse while
leaving two functional valve orifices on each side. Ruptured
chorda tendinea

Intraoperative conversion to mitral-valve re-

placement occurs in 2 to 10% of cases. Systolic Gore-Tex
anterior motion of the mitral valve may occur sutures
postoperatively if leaflet coaptation is not opti-
mal, and mitral stenosis can occur if the annu-
loplasty ring is too small. Other rare adverse
effects of mitral-valve repair include damage to
important structures around the mitral appara-
tus, such as the circumflex coronary artery, the B Chordal transfer
aortic valve, and the bundle of His.
The most important late complication of mi-
tral-valve repair is recurrent mitral regurgitation,
which may occur in as many as 30% of pa-
tients.36 Reoperation to treat recurrent mitral re- Transfer
Missing primary
gurgitation after primary repair is required in ap-
chorda tendinea
proximately 0.5 to 1.5% of patients per year.49,51

A r e a s of Uncer ta in t y Secondary
chorda tendinea

We are unaware of any randomized trials that

have compared mitral-valve repair with mitral-
valve replacement for mitral-valve prolapse, and it
is unlikely that such a trial will be conducted.
Therefore, the current recommendation for mitral-
valve repair in the treatment of severe degenera-
tive mitral regurgitation is based on observation-
al data. C Edge-to-edge repair
It is unclear whether asymptomatic patients Annuloplasty
who have severe mitral regurgitation without left ring
ventricular dysfunction or dilatation, atrial fibril- Anterior leaflet
lation, or pulmonary hypertension should under-
go early surgery. Some investigators have found
evidence of reduced morbidity and mortality with
surgery and recommend early intervention,22,50
whereas others have found that watchful waiting
does not seem to result in worse outcomes.46
Posterior leaflet
The guidelines of the American Heart Associa-
tion (AHA) and the American College of Cardiol-
ogy (ACC) recommend mitral-valve repair for such
patients if the operative success rate is expected
to exceed 90%.28,29 Conversely, the European So-

n engl j med 361;23 nejm.org december 3, 2009 2267

BACK to TOC
97
 The n e w e ng l a n d j o u r na l of m e dic i n e

ciety of Cardiology (ESC) recommends watchful
waiting.53
There is growing experience with minimally
invasive mitral-valve repair performed through a
right minithoracotomy. In a single-center series
involving 1339 patients, the 30-day mortality rate
was 2.4%, the 5-year survival rate was estimated
to be 82.6%, and the reoperation rate was 3.7%.54
These results are similar to those obtained with
traditional mitral-valve repair. This approach re-
quires further evaluation with respect to wide-
spread generalizability and cost-effectiveness; it
is currently performed at only a few specialized
centers.
Guidel ine s
The ACC and the AHA established guidelines for
the management of valvular disease in 2006, with
an update in 2008.28,29 These guidelines gave a
class I recommendation to mitral-valve surgery
for chronic severe mitral regurgitation in the
presence of symptoms, a left ventricular ejection
fraction of less than 60%, or an end-systolic di-
mension of more than 40 mm. Mitral-valve repair
was recommended over replacement for most pa-
tients (class I recommendation). The guidelines
advise that such persons be referred to surgical
centers at which the surgeons are experienced in
mitral-valve repair. The ESC guidelines of 2007
made similar recommendations.53 As noted above,
the societies differ somewhat in terms of their
recommendations for patients who have asymp-
tomatic mitral-valve prolapse with severe mitral
regurgitation but normal left ventricular volumes
and function; the ACCAHA guidelines give a
class IIA recommendation in this regard.
R ec om mendat ions
The patient in the vignette is asymptomatic but
has signs of ventricular dysfunction and elevated
left ventricular dimensions. He should therefore
be offered mitral-valve surgery and should be re-
ferred to a center with demonstrated expertise in
mitral-valve repair. His operative risk should be
formally assessed with the use of one of the vali-
dated risk-scoring algorithms. Intraoperative trans-
esophageal echocardiography should be per-
formed to provide a detailed anatomical and
functional assessment at the time of surgery that
would permit a final decision to be made about
the specifics of the operative procedure. Unless
severe deformity of the valve leaflets or subvalvu-
lar apparatus is present, we would recommend
mitral-valve repair rather than replacement. Since
mitral-valve prolapse is often genetically trans-
mitted, it may be worth considering echocardio-
graphic screening of first-degree relatives.
Dr. Mesana reports receiving honoraria from Medtronic. No oth-
er potential conflict of interest relevant to this article was reported.
We thank Dr. Gilbert Tang for assistance in preparation of the
manuscript.

References
1. Freed LA, Levy D, Levine RA, et al.
Prevalence and clinical outcome of mitral-
valve prolapse. N Engl J Med 1999;341:1-7.
2. The changing spectrum of valvular
heart disease pathology. In: Braunwald E,
ed. Harrisons advances in cardiology.
New York: McGraw-Hill, 2002:317-23.
3. Hayek E, Gring CN, Griffin BP. Mitral
valve prolapse. Lancet 2005;365:507-18.
4. Avierinos JF, Inamo J, Grigioni F,
Gersh B, Shub C, Enriquez-Sarano M. Sex
differences in morphology and outcomes
of mitral valve prolapse. Ann Intern Med
2008;149:787-95.
5. Shell WE, Walton JA, Clifford ME,
Willis PW III. The familial occurrence of
the syndrome of mid-late systolic click
and late systolic murmur. Circulation
1969;39:327-37.
6. Devereux RB, Brown WT, Kramer-Fox
R, Sachs I. Inheritance of mitral valve pro-
lapse: effect of age and sex on gene ex-
pression. Ann Intern Med 1982;97:826-32.
7. Disse S, Abergel E, Berrebi A, et al.
Mapping of a first locus for autosomal
dominant myxomatous mitral-valve pro-
lapse to chromosome 16p11.2-p12.1. Am J
Hum Genet 1999;65:1242-51.
8. Freed LA, Acierno JS Jr, Dai D, et al.
A locus for autosomal dominant mitral
valve prolapse on chromosome 11p15.4.
Am J Hum Genet 2003;72:1551-9.
9. Nesta F, Leyne M, Yosefy C, et al. New
locus for autosomal dominant mitral
valve prolapse on chromosome 13: clini-
cal insights from genetic studies. Circula-
tion 2005;112:2022-30.
10. Barlow JB, Pocock WA. Mitral valve
prolapse, the specific billowing mitral
leaflet syndrome, or an insignificant non-
ejection systolic click. Am Heart J 1979;
97:277-85.
11. Abrams J. Mitral valve prolapse: a plea
for unanimity. Am Heart J 1976;92:413-5.
12. Ling LH, Enriquez-Sarano M, Seward
JB, et al. Clinical outcome of mitral regur-
gitation due to flail leaflet. N Engl J Med
1996;335:1417-23.
13. Rosen SE, Borer JS, Hochreiter C, et al.
Natural history of the asymptomatic/
minimally symptomatic patient with se-
vere mitral regurgitation secondary to
mitral valve prolapse and normal right
and left ventricular performance. Am J
Cardiol 1994;74:374-80.
14. Filsoufi F, Salzberg SP, Aklog L, Ad-
ams DH. Acquired disease of the mitral
valve. In: Selke F, Swanson S, del Nido P,
eds. Sabiston and Spencer surgery of the
chest. 7th ed. Philadelphia: Elsevier Saun-
ders, 2005:1299-333.
15. Pitarys CJ II, Forman MB, Panayiotou
H, Hansen DE. Long-term effects of exci-
sion of the mitral apparatus on global and
regional ventricular function in humans.
J Am Coll Cardiol 1990;15:557-63.
16. Goldman ME, Mora F, Guarino T,
Fuster V, Mindich BP. Mitral valvuloplasty
is superior to valve replacement for pres-
ervation of left ventricular function: an
intraoperative two-dimensional echocar-
diographic study. J Am Coll Cardiol 1987;
10:568-75.
17. Yacoub M, Halim M, Radley-Smith R,
McKay R, Nijveld A, Towers M. Surgical
treatment of mitral regurgitation caused
by floppy valves: repair versus replacement.
Circulation 1981;64:II-210II-216.

2268 n engl j med 361;23 nejm.org december 3, 2009

BACK to TOC
98

18. David TE, Uden DE, Strauss HD. The
importance of the mitral apparatus in left
ventricular function after correction of
mitral regurgitation. Circulation 1983;68:
II-76II-82.
19. Okita Y, Miki S, Ueda Y, Tahata T,
Sakai T, Matsuyama K. Comparative evalu-
ation of left ventricular performance after
mitral valve repair or valve replacement
with or without chordal preservation.
J Heart Valve Dis 1993;2:159-66.
20. Ling LH, Enriquez-Sarano M, Seward
JB, et al. Early surgery in patients with
mitral regurgitation due to flail leaflets:
a long-term outcome study. Circulation
1997;96:1819-25.
21. Grigioni F, Tribouilloy C, Avierinos JF,
et al. Outcomes in mitral regurgitation
due to flail leaflets: a multicenter Euro-
pean study. JACC Cardiovasc Imaging
2008;1:133-41.
22. Enriquez-Sarano M, Avierinos JF,
Messika-Zeitoun D, et al. Quantitative de-
terminants of the outcome of asymptom-
atic mitral regurgitation. N Engl J Med
2005;352:875-83.
23. Shuhaiber J, Anderson RJ. Meta-analy-
sis of clinical outcomes following surgi-
cal mitral valve repair or replacement. Eur
J Cardiothorac Surg 2007;31:267-75.
24. Jokinen JJ, Hippelinen MJ, Pitknen
OA, Hartikainen JE. Mitral valve replace-
ment versus repair: propensity-adjusted
survival and quality-of-life analysis. Ann
Thorac Surg 2007;84:451-8.
25. Gillinov AM, Blackstone EH, Nowicki
ER, et al. Valve repair versus valve replace-
ment for degenerative mitral valve dis-
ease. J Thorac Cardiovasc Surg 2008;135:
885-93.
26. Zhao L, Kolm P, Borger MA, et al.
Comparison of recovery after mitral valve
repair and replacement. J Thorac Cardio-
vasc Surg 2007;133:1257-63.
27. Zoghbi WA, Enriquez-Sarano M, Foster
E, et al. Recommendations for evaluation
of the severity of native valvular regurgi-
tation with two-dimensional and Doppler
echocardiography. J Am Soc Echocardiogr
2003;16:777-802.
28. American College of Cardiology, Amer-
ican Heart Association Task Force on Prac-
tice Guidelines. ACC/AHA 2006 guide-
lines for the management of patients with
valvular heart disease: a report of the
American College of Cardiology/Ameri-
can Heart Association Task Force on Prac-
tice Guidelines (Writing Committee to
revise the 1998 guidelines for the man-
agement of patients with valvular heart
disease) developed in collaboration with
the Society of Cardiovascular Anesthesiol-
ogists endorsed by the Society for Cardio-
vascular Angiography and Interventions
and the Society of Thoracic Surgeons.
J Am Coll Cardiol 2006;48(3):e1-e148.
29. Bonow RO, Carabello BA, Chatterjee
K, et al. 2008 Focused update incorporat-
ed into the ACC/AHA 2006 guidelines for
the management of patients with valvular
n engl j med 361;23
BACK to TOC
clinical ther apeutics
heart disease: a report of the American
College of Cardiology/American Heart
Association Task Force on Practice Guide-
lines (Writing Committee to revise the
1998 guidelines for the management of
patients with valvular heart disease): en-
dorsed by the Society of Cardiovascular
Anesthesiologists, Soceity for Cardiovas-
cular Angiography and Interventions, and
Society of Thoracic Surgeons. J Am Coll
Cardiol 2008;52(13):e1-e142.
30. Gammie JS, OBrien SM, Griffith BP,
Ferguson TB, Peterson ED. Influence of
hospital procedural volume on care pro-
cess and mortality for patients undergo-
ing elective surgery for mitral regurgita-
tion. Circulation 2007;115:881-7.
31. Talwalkar NG, Earle NR, Earle EA,
Lawrie GM. Mitral valve repair in patients
with low left ventricular ejection fractions:
early and late results. Chest 2004;126:709-
15.
32. Gillinov AM, Faber C, Houghtaling
PL, et al. Repair versus replacement for
degenerative mitral valve disease with co-
existing ischemic heart disease. J Thorac
Cardiovasc Surg 2003;125:1350-62.
33. Society of Thoracic Surgeons (STS) on-
line risk calculator. (Accessed November
6, 2009, at http://209.220.160.181/
STSWebRiskCalc261/.)
34. European System for Cardiac Opera-
tive Risk Evaluation (EUROSCORE). (Ac-
cessed November 6, 2009, at http://www.
euroscore.org/calc.html.)
35. Shernan SK. Perioperative transesoph-
ageal echocardiographic evaluation of the
native mitral valve. Crit Care Med 2007;
35:Suppl:S372-S383.
36. Filsoufi F, Carpentier A. Principles of
reconstructive surgery in degenerative mi-
tral valve disease. Semin Thorac Cardio-
vasc Surg 2007;19:103-10.
37. Adams DH, Filsoufi F. Another chap-
ter in an enlarging book: repair degen-
erative mitral valves. J Thorac Cardiovasc
Surg 2003;125:1197-9.
38. Gillinov AM, Cosgrove DM III, Black-
stone EH, et al. Durability of mitral valve
repair for degenerative disease. J Thorac
Cardiovasc Surg 1998;116:734-43.
39. Duran CG, Pomar JL, Revuelta JM, et
al. Conservative operation for mitral in-
sufficiency: critical analysis supported by
postoperative hemodynamic studies of 72
patients. J Thorac Cardiovasc Surg 1980;
79:326-37.
40. Mesana TG, Ibrahim M, Kulik A, et al.
The hybrid flip-over technique for ante-
rior leaflet prolapse repair. Ann Thorac
Surg 2007;83:322-3.
41. Mesana T, Ibrahim M, Hynes M.
A technique for annular placation to facili-
tate sliding plasty after extensive mitral
valve posterior leaflet resection. Ann Tho-
rac Surg 2005;79:720-2.
42. David TE. Outcomes of mitral valve
repair for mitral regurgitation due to de-
generative disease. Semin Thorac Cardio-
vasc Surg 2007;19:116-20.
nejm.org december 3, 2009
43. Starling MR, Kirsh MM, Montgomery
DG, Gross MD. Impaired left ventricular
contractile function in patients with long-
term mitral regurgitation and normal ejec-
tion fraction. J Am Coll Cardiol 1993;22:
239-50.
44. Wilson W, Taubert KA, Gewitz M, et
al. Prevention of infective endocarditis:
guidelines from the American Heart As-
sociation: a guideline from the American
Heart Association Rheumatic Fever, Endo-
carditis, and Kawasaki Disease Commit-
tee, Council on Cardiovascular Disease in
the Young, and the Council on Clinical
Cardiology; Council on Cardiovascular
Surgery and Anesthesia, and the Quality
of Care and Outcomes Research Interdis-
ciplinary Working Group. Circulation
2007;116:1736-54.
45. Barnett SD, Ad N. Surgery for aortic
and mitral valve disease in the United
States: a trend of change in surgical prac-
tice between 1998 and 2005. J Thorac Car-
diovasc Surg 2009;137:1422-9.
46. Rosenhek R, Rader F, Klaar U, et al.
Outcome of watchful waiting in asymp-
tomatic severe mitral regurgitation. Cir-
culation 2006;113:2238-44.
47. Mohty D, Orszulak TA, Schaff HV,
Avierinos JF, Tajik JA, Enriquez-Sarano M.
Very long-term survival and durability of
mitral valve repair for mitral valve pro-
lapse. Circulation 2001;104:Suppl 1:I-1I-7.
48. Akins CW, Hilgenberg AD, Buckley
MJ, et al. Mitral valve reconstruction ver-
sus replacement for degenerative or is-
chemic mitral regurgitation. Ann Thorac
Surg 1994;58:668-75.
49. Lee EM, Shapiro LM, Wells FC. Supe-
riority of mitral valve repair in surgery
for degenerative mitral regurgitation. Eur
Heart J 1997;18:655-63.
50. Kang D-H, Kim JH, Rim JH, et al.
Comparison of early surgery versus con-
ventional treatment in asymptomatic se-
vere mitral regurgitation. Circulation
2009;119:797-804.
51. Suri RM, Schaff HV, Dearani JA, et al.
Survival advantage and improved durabil-
ity of mitral repair for leaflet prolapse
subsets in the current era. Ann Thorac
Surg 2006;82:819-26.
52. OBrien SM, Shahian DM, Filardo G,
et al. The Society of Thoracic Surgeons
2008 cardiac surgery risk models: part 2
isolated valve surgery. Ann Thorac
Surg 2009;88:Suppl:S23-S42.
53. Vahanian A, Baumgartner H, Bax J, et
al. Guidelines on the management of val-
vular heart disease: The Task Force on the
Management of Valvular Heart Disease of
the European Society of Cardiology. Eur
Heart J 2007;28:230-68.
54. Seeburger J, Borger MA, Falk V, et al.
Minimal invasive mitral valve repair for
mitral regurgitation: results of 1339 con-
secutive patients. Eur J Cardiothorac Surg
2008;34:760-5.
Copyright 2009 Massachusetts Medical Society.
2269
99
 Current Concepts

Current Concepts articles present current approaches to a wide variety of clinical problems. These
concise reviews are intended for the practicing physician.

BACK to TOC
100
 The n e w e ng l a n d j o u r na l of m e dic i n e

review article

Current Concepts

Point-of-Care Ultrasonography
Christopher L. Moore, M.D., and Joshua A. Copel, M.D.

U
ltrasonography is a safe and effective form of imaging that From the Departments of Emergency
has been used by physicians for more than half a century to aid in diagnosis Medicine (C.L.M.) and Obstetrics, Gynecol-
ogy, and Reproductive Sciences (J.A.C.),
and guide procedures. Over the past two decades, ultrasound equipment Yale University School of Medicine, New
has become more compact, higher quality, and less expensive, which has facilitated Haven, CT. Address reprint requests to
the growth of point-of-care ultrasonography that is, ultrasonography performed Dr. Moore at the Department of Emer-
gency Medicine, Yale University School of
and interpreted by the clinician at the bedside. In 2004, a conference on compact Medicine, 464 Congress Ave., Suite 260,
ultrasonography hosted by the American Institute of Ultrasound in Medicine (AIUM) New Haven, CT 06519, or at chris.moore@
concluded that the concept of an ultrasound stethoscope is rapidly moving from yale.edu.
the theoretical to reality. This conference included representatives from 19 medical N Engl J Med 2011;364:749-57.
organizations; in November 2010, the AIUM hosted a similar forum attended by 45 Copyright 2011 Massachusetts Medical Society.
organizations.1-3 Some medical schools are now beginning to provide their students
with hand-carried ultrasound equipment for use during clinical rotations.4
Although ionizing radiation from computed tomographic (CT) scanning is in-
creasingly recognized as a potentially major cause of cancer, ultrasonography has
been used in obstetrics for decades, with no epidemiologic evidence of harmful ef-
fects at normal diagnostic levels.5,6 However, ultrasonography is a user-dependent
technology, and as usage spreads, there is a need to ensure competence, define the
benefits of appropriate use, and limit unnecessary imaging and its consequenc-
es.7-10 This article provides an overview of the history and current status of compact,
point-of-care ultrasonography, with examples and discussion of its use.

His t or y of Ultr a sonogr a ph y a nd the Ba sic Technol ogy

Medical ultrasonography was developed from principles of sonar pioneered in

World War I,11 and the first sonographic images of a human skull were published
in 1947.12 The first ultrasound images of abdominal disease were published in
1958,13 and ultrasonography was widely adopted in radiology, cardiology, and ob-
stetrics over the next several decades. Although clinicians from other specialties
occasionally reported using ultrasonography, point-of-care ultrasonography did not
really begin to progress until the 1990s, when more compact and affordable ma-
chines were developed. The early portable machines were hampered by poor image
quality, but in 2010, many point-of-care units can nearly match the imaging quality
of the larger machines.
Ultrasound is defined as a frequency above that which humans can hear, or
more than 20,000 Hz (20 kHz). Therapeutic ultrasound, designed to create heat
using mechanical sound waves, is typically lower in frequency than diagnostic
ultrasound and is not discussed in this article. The frequency of diagnostic ultra-
sound is in the millions of Hertz (MHz). Lower-frequency ultrasound has better
penetration, but at lower resolution. Higher-frequency ultrasound provides better
images, but it does not visualize deep structures well. A typical transabdominal or

n engl j med 364;8 nejm.org february 24, 2011 749

BACK to TOC
101
 The n e w e ng l a n d j o u r na l
cardiac probe has a frequency in the range of 2 to
5 MHz, whereas some dermatologic ultrasound
probes have frequencies as high as 100 MHz.
Ultrasonography uses a crystal a quartz
or composite piezoelectric material that gen-
erates a sound wave when an electric current is
applied. When the sound wave returns, the ma-
terial in turn generates a current. The crystal thus
both transmits and receives the sound. Early ultra-
sonography used a single crystal to create a one-
dimensional image known as A-mode. The stan-
dard screen image that machines now generate is
known as B-mode (also called two-dimensional
or gray-scale ultrasonography), and is created by
an array of crystals (often 128 or more) across
the face of the transducer. Each crystal produces
a scan line that is used to create an image or
frame, which is refreshed many times per second
to produce a moving image on the screen (Fig. 1).
Additional modes, including three-dimensional,
four-dimensional, Doppler, and tissue Doppler
modes, are now commonly available but are not
addressed in this article.
Ultrasound penetrates well through fluid and
solid organs (e.g., liver, spleen, and uterus); it
does not penetrate well through bone or air,
limiting its usefulness in the skull, chest, and
areas of the abdomen where bowel gas obscures
the image. Fluid (e.g, blood, urine, bile, and as-
cites), which is completely anechoic, appears
black on ultrasound images, making ultrasonog-
raphy particularly useful for detecting fluid and
differentiating cystic or vascular areas from solid
structures.
Two-dimensional ultrasound is used to visual-
ize a plane that is then shown on the screen.
This plane may be directed by the user in any
anatomical plane on the patient: sagittal (or longi-
tudinal), transverse (or axial), coronal (or frontal),
or some combination (oblique). An indicator on
the probe is used to orient the user to the orien-
tation of the plane on the screen. By convention,
in general and obstetrical imaging, the indicator
corresponds to the left side of the screen as it is
viewed. Cardiology uses the opposite convention
for echocardiography, with the indicator corre-
sponding to the right of the screen. Users should
be aware of these conventions when conducting
integrated examinations that include both gen-
eral and cardiac imaging.14
750 n engl j med 364;8
BACK to TOC
of m e dic i n e
P oin t- of- C a r e A ppl ic at ions
Point-of-care ultrasonography is defined as ultra-
sonography brought to the patient and performed
by the provider in real time. Point-of-care ultra-
sound images can be obtained nearly immedi-
ately, and the clinician can use real-time dynam-
ic images (rather than images recorded by a
sonographer and interpreted later), allowing find-
ings to be directly correlated with the patients
presenting signs and symptoms.15 Point-of-care
ultrasonography is easily repeatable if the patients
condition changes. It is used by various special-
ties in diverse situations (Table 1) and may be
broadly divided into procedural, diagnostic, and
screening applications.
Procedural Guidance
Ultrasound guidance may improve success and
decrease complications in procedures performed
by multiple specialties, including central and
peripheral vascular access, thoracentesis, paracen-
tesis, arthrocentesis, regional anesthesia, incision
and drainage of abscesses, localization and re-
moval of foreign bodies, lumbar puncture, biop-
sies, and other procedures.16
Procedural guidance may be static or dynamic.
With static guidance, the structure of interest is
identified, and the angle required by the needle
is noted, with the point of entry marked on the
skin. In dynamic procedures, ultrasonography
visualizes the needle in real time. Static guid-
ance may initially be easier to perform, but prop-
erly performed dynamic guidance provides more
accurate guidance and is generally preferred by
experienced users.
In response to the 1999 Institute of Medicine
report To Err Is Human, the Agency for Healthcare
Research and Quality listed use of real-time
ultrasound guidance during central line insertion
to prevent complications as 1 of the 12 most
highly rated patient safety practices designed to
decrease medical errors.17 The use of ultrasound
to guide central venous access has been shown
to reduce the failure rate, the risk of complica-
tions, and the number of attempts, as compared
with the landmark technique, particularly in the
case of less experienced users or patients with
more complex conditions.18,19 The evidence for
these benefits of ultrasound guidance is greatest
nejm.org february 24, 2011
102
 Current Concepts

A B

Indicator to
patients right

Inferior Splenic
Indicator vena cava vein

Spine Aorta

Sagittal plane

Indicator

Coronal
plane

Transverse plane

Figure 1. Basic (B-Mode) Two-Dimensional Ultrasound Image.

A typical ultrasound transducer, shown in Panel A, has 128 or more crystals arranged across the face of the probe. Each crystal trans-
mits and receives bursts of sound (typically in the megahertz range), creating a scan line. The scan lines together make up a frame,
which is refreshed many times per second and displayed on a two-dimensional screen to create a moving image. As shown in Panel B,
the plane of the ultrasound can be directed in any anatomical plane or between planes. By convention, in abdominal imaging, the probe
indicator (a bump or groove on the probe) is to the left of the screen and is generally directed toward the patients right side in a trans-
verse plane. The ultrasound image shown is a transverse image of the abdominal aorta. The indicator is directed to the patients right
side, corresponding to the left side of the screen. The aorta is black (fluid-filled) and located just anterior to the vertebral bodies. (See
also Video 4, available with the full text of this article at NEJM.org.)

n engl j med 364;8 nejm.org february 24, 2011 751

BACK to TOC
103
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 1. Selected Applications of Point-of-Care Ultrasonography, According to Medical Specialty.*

Specialty Ultrasound Applications

Anesthesia Guidance for vascular access, regional anesthesia, intraoperative monitoring
of fluid status and cardiac function
Cardiology Echocardiography, intracardiac assessment
Critical care medicine Procedural guidance, pulmonary assessment, focused echocardiography
Dermatology Assessment of skin lesions and tumors
Emergency medicine FAST, focused emergency assessment, procedural guidance
Endocrinology and endocrine surgery Assessment of thyroid and parathyroid, procedural guidance
General surgery Ultrasonography of the breast, procedural guidance, intraoperative assessment
Gynecology Assessment of cervix, uterus, and adnexa; procedural guidance
Obstetrics and maternalfetal medicine Assessment of pregnancy, detection of fetal abnormalities, procedural guidance
Neonatology Cranial and pulmonary assessments
Nephrology Vascular access for dialysis
Neurology Transcranial Doppler, peripheral-nerve evaluation
Ophthalmology Corneal and retinal assessment
Orthopedic surgery Musculoskeletal applications
Otolaryngology Assessment of thyroid, parathyroid, and neck masses; procedural guidance
Pediatrics Assessment of bladder, procedural guidance
Pulmonary medicine Transthoracic pulmonary assessment, endobronchial assessment, proce-
dural guidance
Radiology and interventional radiology Ultrasonography taken to the patient with interpretation at the bedside,
procedural guidance
Rheumatology Monitoring of synovitis, procedural guidance
Trauma surgery FAST, procedural guidance
Urology Renal, bladder, and prostate assessment; procedural guidance
Vascular surgery Carotid, arterial, and venous assessment; procedural assessment

* FAST denotes focused assessment with sonography for trauma.

for the internal jugular site, with less evidence for
the femoral and subclavian sites and in pediatric
patients.20
A needle may be imaged dynamically with the
use of either an in-plane or out-of-plane ultra-
Videos showing
point-of-care sound approach (Fig. 2, and Video 1, available at
ultrasonography NEJM.org). For vascular access, an in-plane ap-
are available at proach corresponds to the long axis of the vessel.
NEJM.org
An in-plane, or long-axis, approach is generally
Click here preferred for dynamic vascular access, particu-
to access larly for central venous access, because the en-
videos. tire length of the needle, including the tip, can
be visualized throughout the procedure. How-
ever, it may be more difficult to keep the needle
in view with the use of an in-plane approach,
and for smaller vessels, it may be challenging to
image the entire vessel in the long axis.
An out-of-plane approach is perpendicular to
the needle and corresponds to the short axis of
the vessel. The advantage of this approach is
that the needle can be centered over the middle
of the vessel. It is also easier to keep the vessel
and the needle in view in the short axis. However,
an out-of-plane approach may underestimate the
depth of the needle tip if the ultrasound plane
cuts across the shaft of the needle, proximal to
the tip. A detailed description of ultrasound-
guided central venous access of the internal
jugular vein is provided by Ortega et al. as part of
the Journals Videos in Clinical Medicine series.21
Diagnostic Assessment
The concept of a focused (limited, or goal-
directed) examination is important in point-of-
care ultrasonography. Clinicians from diverse
specialties can become very adept at using ultra-

752 n engl j med 364;8 nejm.org february 24, 2011

BACK to TOC
104
 Current Concepts

A In-plane view of the needle

(long axis of the vessel)

Needle
shaft Needle tip

B Out-of-plane view of the needle

(short axis of the vessel)

Needle in
cross-section

Reverberation
artifact

Figure 2. Ultrasound Guidance for Vascular Access and Other Procedures Involving Needles.
Panel A shows a long-axis, in-plane view of the needle. Although it may be more difficult to keep the needle and structure of interest
in view, the long-axis view is advantageous because it shows the entire needle, including the tip (ultrasound image at right). Panel B shows
a short-axis approach, with the characteristic target sign of the needle in the vessel lumen. The ultrasound image also shows a rever-
beration artifact, which occurred in this case when the ultrasound beam struck a metallic object. The artifact appears as closely spaced,
tapering lines below the needle. Although the visualized portion of the needle is centered in the lumen, the disadvantage of the short
axis is that the plane of the ultrasound may cut through the needle shaft proximally, underestimating the depth of the tip. (See also
Video 1.)

sonography to examine a particular organ, dis-
ease, or procedure that is directly relevant to their
area of expertise, whereas imaging specialists
typically perform more comprehensive examina-
tions (Table 1).
Point-of-care ultrasonography may involve the
use of a series of focused ultrasonographic ex-
aminations to efficiently diagnose or rule out
certain conditions in patients presenting with
particular symptoms or signs, such as hypoten-

n engl j med 364;8 nejm.org february 24, 2011 753

BACK to TOC
105
 The n e w e ng l a n d j o u r na l
A Skin and
Rib subcutaneous
tissue
Pleural
Rib line
shadow
A line
(normal
reverberation
artifact)
B Rib Rib
Pleural line
Rib
Rib
shadow
shadow
B lines
(lung rockets)
Figure 3. Ultrasound Images of the Pleural Line
in a Healthy Patient and in a Patient with Alveolar
Interstitial Syndrome.
In Panel A, a high-frequency linear probe is placed
with the indicator toward the patients head (screen
left), in the midclavicular line at approximately the third
intercostal space. At the posterior edge of the rib, a hy-
perechoic (bright) pleural line is seen, which is the inter-
face between the visceral and parietal pleura. In a mov-
ing image of a normal lung, shimmering or sliding
would be seen at the pleural line, indicating that the
visceral pleura is closely associated with the parietal
pleura. An A line (a normal reverberation artifact) is
also seen. In Panel B, a phased-array sector probe is
placed at the same anatomical location on a different
patient. This sector image is much deeper, but it shows
the same structures, as well as pathological B lines,
artifacts that extend to the bottom of the screen (lung
rockets). This patient had alveolar interstitial syndrome
from congestive heart failure. (See also Video 3.)
sion, chest pain, or dyspnea. In patients with
trauma, this approach is known as FAST (focused
assessment with sonography for trauma). Point-
of-care ultrasonography allows immediate, dy-
namic, and repeated assessments in these situ-
ations and has the potential for detecting
conditions such as pneumothorax in which ultra-
sonography was traditionally thought to be un-
helpful. Here we focus on an integrated point-of-
care examination for trauma (FAST), as well as
754 n engl j med 364;8
BACK to TOC
of m e dic i n e
the use of point-of-care ultrasonography for pul-
monary assessment.
FAST Examination
FAST was a term coined at an international con-
sensus conference in 1996 to describe an inte-
grated, goal-directed, bedside examination to
detect fluid, which is likely to be hemorrhage in
cases of trauma.22 The extended FAST (e-FAST)
also includes examination of the chest for pneu-
mothorax.23
The e-FAST examination combines five fo-
cused examinations for the detection of: free
intraperitoneal fluid, free fluid in the pelvis,
pericardial fluid, pleural effusion, and pneumo-
thorax. Peritoneal fluid is detected using views
of the hepatorenal space (Morisons pouch),
splenorenal space, and retrovesicular spaces. The
thorax is evaluated for fluid at the flanks and for
pneumothorax anteriorly. The pericardium may
be evaluated for effusion, particularly in cases of
penetrating trauma (see Video 2).
A FAST examination may be completed in
less than 5 minutes and has been shown to have
a sensitivity of 73 to 99%, a specificity of 94 to
98%, and an overall accuracy of 90 to 98% for
clinically significant intraabdominal injury in
trauma.24 The use of the FAST examination has
been shown to reduce the need for CT or diag-
nostic peritoneal lavage and to reduce the time
to appropriate intervention, resulting in a shorter
hospital stay, lower costs, and lower overall mor-
tality, although more rigorous study of patient-
centered outcomes is recommended.25,26 A com-
plete or partial FAST examination may also be
helpful in evaluating patients who do not have
trauma for ascites, intraperitoneal hemorrhage,
pleural effusion, pneumothorax, or pericardial
effusion.
Pulmonary Ultrasonography
The use of ultrasound to detect pneumothorax
was first described in a horse in 1986, and then
in humans shortly afterward.27 In a normal lung,
the visceral and parietal pleura are closely associ-
ated, and ultrasound shows shimmering or sliding
at the pleural interface during respiration (Fig. 3,
and Video 3). The absence of sliding indicates a
pneumothorax. A small pneumothorax may be
missed with the use of ultrasonography, and pa-
tients with blebs or scarring may have false
positive findings.28 However, for assessing pa-
nejm.org february 24, 2011
106
 Current Concepts
tients with trauma for pneumothorax, ultrasonog-
raphy has been shown to be more than twice as
sensitive as conventional supine chest radiogra-
phy for detecting occult pneumothorax (pneumo-
thorax seen only on CT), with similarly high spec-
ificity (>98%).23 The presence of a lung point
sign, where the visceral pleura intermittently
comes in contact with the parietal pleura, is nearly
100% specific for the detection of pneumothorax.
Comet tails are an ultrasound artifact that
arises when ultrasound encounters a small air
fluid interface. In 1997, Lichtenstein et al. de-
scribed the sonographic identification of alveo-
lar interstitial syndrome, diagnosed on the basis
of comet tails that extend from the pleural line to
the bottom of the screen, also known as B lines
(Fig. 3B). Alveolar interstitial syndrome is an
ultrasonographic finding in several different con-
ditions.29 In an acute condition, alveolar inter-
stitial syndrome usually represents pulmonary
edema, but it may also be seen in the acute re-
spiratory distress syndrome and more chronic
interstitial diseases and may be a focal finding
in infectious or ischemic processes. Characteris-
tics of the artifacts may be helpful in distin-
guishing these conditions.
Ultrasonography has been shown to be more
accurate than auscultation or chest radiography
for the detection of pleural effusion, consolida-
tion, and alveolar interstitial syndrome in the
critical care setting.30 In the emergency care set-
ting, the presence of B lines on pleural ultraso-
nography predicts fluid overload, adding diag-
nostic accuracy to the physical examination and
measurement of brain natriuretic peptide.31 The
presence of B lines has been shown to be dy-
namic, disappearing in patients undergoing he-
modialysis.31,32
Screening
Screening with ultrasonography is attractive be-
cause it is noninvasive and lacks ionizing radia-
tion. Ultrasonography has been described as a
screening test for cardiovascular and gynecolog-
ic disease, and compact ultrasonography has
been incorporated into mobile screening labs.33
However, the benefits of screening must be
weighed against the harms, particularly false pos-
itive findings that lead to unnecessary testing,
intervention, or both. The U.S. Preventive Ser-
vices Task Force (USPSTF) has specifically rec-
ommended that ultrasonography not be used for
n engl j med 364;8
BACK to TOC
routine screening for carotid stenosis, peripheral
vascular disease, or ovarian cancer in the general
population (class D recommendation inef-
fective or harms outweigh benefits), although
research is ongoing to determine whether more
narrowly defined populations may benefit from
such screening.34
In 2005, the USPSTF gave a class B recom-
mendation for one-time ultrasound screening
for abdominal aortic aneurysm in men between
the ages of 65 and 75 years who had ever smoked,
leading to the incorporation of screening for
abdominal aortic aneurysm into Medicare reim-
bursement.35,36 The USPSTF reports that ultraso-
nography has a sensitivity of 95% and a specific-
ity of nearly 100% when performed in a setting
with adequate quality assurance.
Imaging of the abdominal aorta is performed
with a curvilinear probe of 2 to 5 MHz. With the
patient in a supine position, gentle pressure is
applied to move bowel gas out of the way. The
aorta should be imaged as completely as possi-
ble from the proximal (celiac trunk) to the distal
bifurcation and should include assessment of
the iliac arteries when possible. It should be
measured at its maximum diameter from out-
side wall to outside wall in two planes, trans-
verse and longitudinal. Challenges include en-
suring that the aorta is imaged, not the inferior
vena cava or another fluid-filled structure, and
ensuring that the entire diameter is measured
(Fig. 1, and Video 4).
Ultrasonography of the abdominal aorta has
been shown to be fairly straightforward to learn
as a focused examination, and screening by pri-
mary care providers using point-of-care ultraso-
nography may provide an economical method
for wider screening, although more study is
needed in this area.
Point- of- C a r e Ultr a sonogr a ph y
in O ther Se t t ings
Point-of-care ultrasonography is increasingly be-
ing used in resource-limited settings. The World
Health Organization states that plain radiogra-
phy and ultrasonography, singly or in combina-
tion, will meet two thirds of all imaging needs in
developing countries.37 Ultrasonography has been
used at the Mount Everest base camp to diagnose
high-altitude pulmonary edema, and ultrasonog-
raphy is the only diagnostic imaging technique
nejm.org february 24, 2011 755
107
 The n e w e ng l a n d j o u r na l of m e dic i n e

used on the International Space Station, where in some cases obviating the need for more re-
astronauts obtain images that are interpreted on source-intensive imaging performed by a consult-
earth.38,39 The use of hand-carried ultrasono- ing radiologist.47 However, indiscriminate use of
graphic devices has been described in prehospi- ultrasonography could lead to further unneces-
tal settings, including ambulance and disaster sary testing, unnecessary interventions in the
settings, as well as in battlefield medicine (the case of false positive findings, or inadequate
scenario for which hand-carried ultrasonogra- investigation of false negative findings. More
phy was initially developed).40-42 The e-FAST ex- imaging could simply lead to increased expense
amination for internal bleeding and pneumotho- without added benefit, or might even be harmful.
rax has been the most extensively described As a user-dependent technology, point-of-care
application in the prehospital setting (Video 2). ultrasonography requires consideration of appro-
priate training and quality assurance. In addition,
P ol ic y C onsider at ions methodologically rigorous studies are needed to
assess patient-centered outcomes for point-of-
From 2000 to 2006, physician fees billed for care ultrasonography.25,48-50
medical imaging in the United States more than
doubled, with the proportion of billing for in- C onclusions
office imaging rising from 58 to 64%.43 Al-
though the rate of imaging increased among both The use of point-of-care ultrasonography will
radiologists and nonradiologists, the rate of in- continue to diffuse across medical specialties
crease was faster among nonradiologists.44,45 and care situations. Future challenges include
Most of this increase was related to advanced gaining a better understanding of when and how
imaging (CT, magnetic resonance imaging, and point-of-care ultrasonography can be used effec-
nuclear medicine), but certain applications of tively, determining the training and assessment
ultrasonography by nonradiologists (particularly that will be required to ensure competent use of
breast and cardiac applications) increased at a the technology, and structuring policy and re-
very rapid rate.46 imbursement to encourage appropriate and effec-
With appropriate use, point-of-care ultrasonog- tive use.
raphy can decrease medical errors, provide more Dr. Moore reports receiving consulting fees from SonoSite
efficient real-time diagnosis, and supplement or and Philips; and Dr. Copel, speaking fees from Siemens, World
replace more advanced imaging in appropriate Class CME, the Institute for Advanced Medical Education, and
Educational Symposia, grant support from Philips, and reim-
situations. In addition, point-of-care ultrasonog- bursement for travel expenses from Philips and Esaote and serving
raphy may allow more widespread, less-expen- as a paid member of the editorial board of Contemporary OB/GYN at
sive screening for defined indications. It may be modernmedicine.com. No other potential conflict of interest
relevant to this article was reported.
particularly cost-effective in a reimbursement Disclosure forms provided by the authors are available with
scheme based on episodes of care (bundling), the full text of this article at NEJM.org.

References
1. Greenbaum LD, Benson CB, Nelson
LH III, Bahner DP, Spitz JL, Platt LD. Pro-
ceedings of the Compact Ultrasound
Conference sponsored by the American
Institute of Ultrasound in Medicine. J Ultra-
sound Med 2004;23:1249-54.
2. Alpert JS, Mladenovic J, Hellmann DB.
Should a hand-carried ultrasound ma-
chine become standard equipment for
every internist? Am J Med 2009;122:1-3.
3. AIUM Ultrasound Practice Forum,
2010: point-of-care use of ultrasound.
(http://www.aium.org/advertising/
2010Forum.pdf.)
4. Rao S, van Holsbeeck L, Musial JL, et al.
A pilot study of comprehensive ultra-
sound education at the Wayne State Uni-
versity School of Medicine: a pioneer year
review. J Ultrasound Med 2008;27:745-9.
5. Brenner DJ, Hall EJ. Computed tomog-
raphy an increasing source of radiation
exposure. N Engl J Med 2007;357:2277-
84.
6. Barnett SB. Routine ultrasound scan-
ning in first trimester: what are the risks?
Semin Ultrasound CT MR 2002;23:387-
91.
7. Filly RA. Is it time for the sonoscope?
If so, then lets do it right! J Ultrasound
Med 2003;22:323-5.
8. Adler RS. The use of compact ultra-
sound in anesthesia: friend or foe. Anesth
Analg 2007;105:1530-2.
9. Bree RL, Benson CB, Bowie JD, et al.
The role of radiology in the era of com-
pact ultrasound systems: SRU Conference,
October 14 and 15, 2003. Ultrasound Q
2004;20:19-21.
10. Greenbaum LD. It is time for the so-
noscope. J Ultrasound Med 2003;22:321-2.
11. Dussik KT. On the possibility of us-
ing ultrasound waves as a diagnostic aid.
Z Neurol Psychiat 1942;174:153-68. (In
German.)
12. Edler I, Lindstrm K. The history of
echocardiography. Ultrasound Med Biol
2004;30:1565-644.
13. Donald I, Macvicar J, Brown TG. Inves-
tigation of abdominal masses by pulsed
ultrasound. Lancet 1958;1:1188-95.
14. Moore C. Current issues with emer-

756 n engl j med 364;8 nejm.org february 24, 2011

BACK to TOC
108

gency cardiac ultrasound probe and im-
age conventions. Acad Emerg Med 2008;
15:278-84.
15. Gluckman JL, Mann W, Portugal LG,
Welkoborsky H-J. Real-time ultrasonog-
raphy in the otolaryngology office set-
ting. Am J Otolaryngol 1993;14:307-13.
16. Nicolaou S, Talsky A, Khashoggi K,
Venu V. Ultrasound-guided interventional
radiology in critical care. Crit Care Med
2007;35:Suppl:S186-S197.
17. Making health care safer: a critical
analysis of patient safety practices. Rock-
ville, MD: Agency for Healthcare Re-
search and Quality. (AHRQ publication
no. 01-E058.)
18. Randolph AG, Cook DJ, Gonzales CA,
Pribble CG. Ultrasound guidance for
placement of central venous catheters:
a meta-analysis of the literature. Crit Care
Med 1996;24:2053-8.
19. Hind D, Calvert N, McWilliams R, et al.
Ultrasonic locating devices for central ve-
nous cannulation: meta-analysis. BMJ
2003;327:361-8.
20. Kumar A, Chuan A. Ultrasound guid-
ed vascular access: efficacy and safety. Best
Pract Res Clin Anaesthesiol 2009;23:299-
311.
21. Ortega R, Song M, Hansen CJ, Barash
P. Ultrasound-guided internal jugular vein
cannulation. N Engl J Med 2010;362(16):
e57. (Video available at NEJM.org.)
22. Scalea TM, Rodriguez A, Chiu WC, et
al. Focused Assessment with Sonography
for Trauma (FAST): results from an inter-
national consensus conference. J Trauma
1999;46:466-72.
23. Kirkpatrick AW, Sirois M, Laupland
KB, et al. Hand-held thoracic sonography
for detecting post-traumatic pneumotho-
races: the Extended Focused Assessment
with Sonography for Trauma (EFAST).
J Trauma 2004;57:288-95.
24. Melanson SW. The FAST Exam: a re-
view of the literature. In: Jehle D, Heller
MB, eds. Ultrasonography in trauma: the
FAST Exam. Dallas: American College of
Emergency Physicians, 2003:127-45.
25. Melniker LA, Leibner E, McKenney
MG, Lopez P, Briggs WM, Mancuso CA.
Randomized controlled clinical trial of
point-of-care, limited ultrasonography for
trauma in the emergency department: the
first Sonography Outcomes Assessment
Program trial. Ann Emerg Med 2006;48:
227-35.
RECEIVE IMMEDIATE NOTIFICATION WHEN AN ARTICLE
are published Online First, sign up at NEJM.org.
n engl j med 364;8
BACK to TOC
Current Concepts
26. Hosek WT, McCarthy ML. Trauma ul-
trasound and the 2005 Cochrane Review.
Ann Emerg Med 2007;50:619-20.
27. Rantanen NW. Diseases of the thorax.
Vet Clin North Am Equine Pract 1986;2:
49-66.
28. Dulchavsky SA, Schwarz KL, Kirkpat-
rick AW, et al. Prospective evaluation of
thoracic ultrasound in the detection of
pneumothorax. J Trauma 2001;50:201-5.
29. Lichtenstein D, Mzire G, Biderman
P, Gepner A, Barr O. The comet-tail arti-
fact: an ultrasound sign of alveolar-inter-
stitial syndrome. Am J Respir Crit Care
Med 1997;156:1640-6.
30. Lichtenstein D, Goldstein I, Mour-
geon E, Cluzel P, Grenier P, Rouby JJ.
Comparative diagnostic performances of
auscultation, chest radiography, and lung
ultrasonography in acute respiratory dis-
tress syndrome. Anesthesiology 2004;100:
9-15.
31. Liteplo AS, Marill KA, Villen T, et al.
Emergency Thoracic Ultrasound in the
Differentiation of the Etiology of Short-
ness of Breath (ETUDES): sonographic
B-lines and N-terminal pro-brain-type na-
triuretic peptide in diagnosing congestive
heart failure. Acad Emerg Med 2009;16:
201-10.
32. Noble VE, Murray AF, Capp R, Sylvia-
Reardon MH, Steele DJR, Liteplo A. Ultra-
sound assessment for extravascular lung
water in patients undergoing hemodialy-
sis. Chest 2009;135:1433-9.
33. Payne JW. Screening with holes in it?
Washington Post. July 19, 2005. (http://
www.washingtonpost.com/wp-dyn/
content/article/2005/07/18/
AR2005071801175.html.)
34. Screening for carotid artery stenosis.
Rockville, MD: U.S. Preventive Services
Task Force, December 2007. (http://www
.uspreventiveservicestaskforce.org/uspstf/
uspsacas.htm.)
35. U.S. Preventive Services Task Force.
Screening for abdominal aortic aneurysm:
recommendation statement. Ann Intern
Med 2005;142:198-202.
36. Thompson SG, Ashton HA, Gao L,
Scott RA. Screening men for abdominal
aortic aneurysm: 10 year mortality and
cost effectiveness results from the ran-
domised Multicentre Aneurysm Screening
Study. BMJ 2009;338:2307-18.
37. Aide-memoire for diagnostic imaging
services. Geneva: World Health Organi-
IS PUBLISHED ONLINE FIRST
To be notified by e-mail when Journal articles
nejm.org february 24, 2011
zation, 2009. (http://whqlibdoc.who.int/
aide-memoire/a71903.pdf.)
38. Otto C, Hamilton DR, Levine BD, et al.
Into thin air: extreme ultrasound on Mt
Everest. Wilderness Environ Med 2009;
20(3):283-9.
39. Sargsyan AE, Hamilton DR, Jones JA,
et al. FAST at MACH 20: clinical ultra-
sound aboard the International Space Sta-
tion. J Trauma 2005;58:35-9.
40. Nelson BP, Chason K. Use of ultra-
sound by emergency medical services: a
review. Int J Emerg Med 2008;1:253-9.
41. Brooks AJ, Price V, Simms M. FAST on
operational military deployment. Emerg
Med J 2005;22:263-5.
42. Beck-Razi N, Fischer D, Michaelson
M, Engel A, Gaitini D. The utility of fo-
cused assessment with sonography for
trauma as a triage tool in multiple-casual-
ty incidents during the second Lebanon
war. J Ultrasound Med 2007;26:1149-56.
43. Medicare imaging payments. Wash-
ington, DC: Government Accountability
Office, 2008. (GAO-08-452.)
44. Hillman BJ, Olson GT, Griffith PE, et
al. Physicians utilization and charges for
outpatient diagnostic imaging in a Medi-
care population. JAMA 1992;268:2050-4.
45. Maitino AJ, Levin DC, Parker L, Rao
VM, Sunshine JH. Practice patterns of ra-
diologists and nonradiologists in utiliza-
tion of noninvasive diagnostic imaging
among the Medicare population. Radiol-
ogy 2003;228:795-801.
46. Miller ME. MedPAC recommenda-
tions on imaging services. Washington,
DC: Subcommittee on Health Committee
on Ways and Means, 2005.
47. Schoen C, Guterman S, Shih A, et al.
Bending the curve: options for achieving
savings and improving value in U.S.
health spending. Washington, DC: The
Commonwealth Fund, 2007.
48. Rose JS. Ultrasonography and out-
comes research: one small step for man-
kind or another drop in the bucket? Ann
Emerg Med 2006;48:237-9.
49. Liu SS, Ngeow JE, Yadeau JT. Ultra-
sound-guided regional anesthesia and
analgesia: a qualitative systematic review.
Reg Anesth Pain Med 2009;34:47-59.
50. Vance S. The FAST scan: are we im-
proving care of the trauma patient? Ann
Emerg Med 2007;49:364-6.
Copyright 2011 Massachusetts Medical Society.
757
109
 The n e w e ng l a n d j o u r na l of m e dic i n e

review article

Current Concepts

Myocardial Infarction Due to Percutaneous

Coronary Intervention
Abhiram Prasad, M.D., and Joerg Herrmann, M.D.

A
pproximately 1.5 million patients undergo percutaneous coro- From the Department of Internal Medi-
nary intervention (PCI) in the United States every year.1 Depending on local cine and the Division of Cardiovascular
Diseases, Mayo Clinic, Rochester, MN.
practices and the diagnostic criteria used, 5 to 30% of these patients (75,000 Address reprint requests to Dr. Prasad at
to 450,000) have evidence of a periprocedural myocardial infarction.2,3 At the higher the Mayo Clinic, 200 First St. SW, Roch-
estimate, the incidence of these events is similar to the annual rate of major spon- ester, MN 55905, or at prasad.abhiram@
mayo.edu.
taneous myocardial infarction.1 Thus, many cardiologists and internists are likely
to encounter patients with coronary artery disease who have sustained a periproce- N Engl J Med 2011;364:453-64.
dural myocardial infarction. However, the clinical significance of these events and Copyright 2011 Massachusetts Medical Society.

their management remain a matter of considerable controversy and uncertainty

(Table 1).4-6 Questions that often arise include the following: Do we need to routinely
screen patients for periprocedural myocardial infarction? Which patients should be
observed in the hospital for a prolonged period after periprocedural myocardial
infarction? What are the therapeutic implications, and what should we tell patients
who sustained a periprocedural myocardial infarction despite an otherwise suc-
cessful procedure? Is a periprocedural myocardial infarction prognostically equiva-
lent to a spontaneous myocardial infarction? Is periprocedural myocardial infarc-
tion a valid end point in clinical trials? The aim of this review is to address these
questions and to provide a current perspective on this issue.

Defini t ions a nd Pr edic t or s of P CI-R el ated

M yonecrosis

Current PCI guidelines give a class I recommendation for the measurement of car-
diac biomarkers (the MB fraction of creatine kinase [CK-MB], cardiac troponin, or
both) in patients who have signs or symptoms suggestive of myocardial infarction
during or after PCI and for those who have undergone complicated procedures.7 In
addition, a class IIa recommendation is given for routine measurements of cardiac
biomarkers 8 to 12 hours after the procedure. In either case, a new CK-MB or tro-
ponin I or T rise greater than 5 times the upper limit of normal would constitute a
clinically significant periprocedural MI [myocardial infarction].7 The more recent
consensus document on the universal definition of myocardial infarction specifi-
cally classifies cardiac-biomarker levels that are more than 3 times the upper refer-
ence limit as indicative of a periprocedural myocardial infarction and recommends
measurement of cardiac troponin as the preferred biomarker.8 Given the availabil-
ity of high-sensitivity cardiac troponin assays, this guideline establishes the thresh-
old for a diagnosis of periprocedural myocardial infarction at very low levels of
myonecrosis.
The predictors of periprocedural myocardial infarction can be broadly catego-
rized as patient-, lesion-, and procedure-related risk factors.2 The major risk factors,
in terms of both frequency and potency, are complex lesions (e.g., the presence of
thrombus, stenosis of a saphenous-vein graft, or a type C lesion), complex procedures

n engl j med 364;5 nejm.org february 3, 2011 453

BACK to TOC
110
 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 1. Evidence for and against the Clinical Significance of Periprocedural Myocardial Infarction.*

Evidence for Clinical Significance Evidence against Clinical Significance

Patients with elevated cardiac biomarkers after PCI have evidence of
focal infarction on cardiac imaging
A large number of studies have shown a correlation between PMI
and adverse clinical outcomes (see the Supplementary Appendix,
available with the full text of this article at NEJM.org), and these
studies greatly outnumber those that do not
There is a positive correlation between the magnitude of postproce-
dural biomarker elevation and the likelihood of adverse out-
comes
Studies have shown that pre-PCI interventions such as statin ther-
apy reduce the frequency of PMI and improve long-term out-
comes
Virtually all data correlating PMI to adverse clinical outcomes are
derived from retrospective studies that have shown associations
but not causal relationships
Retrospective studies are generally limited because they cannot
adequately adjust for all possible confounding variables with
respect to baseline clinical, angiographic, and procedural char-
acteristics that may determine the likelihood of both PMI and
adverse outcomes
Most studies did not use high-sensitivity cardiac troponin assays;
when these assays were used, the studies did not apply the cur-
rently recommended 99th percentile cutoff value for the upper
limit of the normal range
In most cases, PMI results in minimal myonecrosis and therefore
does not substantially impair cardiac function one of the
most important determinants of outcome in coronary artery
disease

* PCI denotes percutaneous coronary intervention, and PMI periprocedural myocardial infarction.

(e.g., treatment of multiple lesions or use of rota-
tional atherectomy), and associated complica-
tions (e.g., abrupt vessel closure, side-branch
occlusion, distal embolization, or no reflow).2,9-12
In contrast, patient-related factors, such as ad-
vanced age, diabetes mellitus, renal failure,
multivessel disease, and left ventricular dysfunc-
tion, are the important determinants of clinical
outcomes after PCI.2,9-11 The occurrence of peri-
procedural ischemic symptoms, particularly chest
pain at the end of the procedure, or electrocardio-
graphic evidence of ischemia defines the sub-
group of patients most likely to have periproce-
dural myocardial infarction.11,13
Mech a nisms of P CI-R el ated
M yonecrosis
Large periprocedural myocardial infarcts are usu-
ally due to angiographically visible complications;
however, this is generally not the case in the vast
majority of patients with elevated biomarker lev-
els after PCI.6,14,15 Cardiac magnetic resonance
imaging (MRI) has confirmed two distinct loca-
tions for procedural myonecrosis: adjacent to the
site of the intervention, where the injury is most
likely due to epicardial side-branch occlusion, and
downstream from the intervention site, where it
is most likely due to compromise of the micro-
vascular circulation (Fig. 1).2,16 Acute myocardial P os tpro cedur a l M yonecrosis
injury occurs with equal frequency at the two lo-
cations and is detected on MRI in 25% of pa- In the CK-MB and early cardiac troponin era, nu-
tients after PCI, with a mean infarct size of ap- merous studies evaluated the clinical significance
proximately 5% of the left ventricular mass.3 The
size of distal infarcts correlates directly with the
extent to which the plaque volume is reduced (em-
bolized) by PCI, since more debris is sent down-
stream, but this is not so for the proximal type of
injury. Moreover, the composition of the plaque
influences the extent of periprocedural myonecro-
sis. PCI for plaques with large necrotic cores leads
to greater degrees of myonecrosis, whereas fibrous
plaques are relatively inert in this regard.17,18
Embolization of plaque material has been de-
tected on intracoronary Doppler ultrasonography
during PCI. Although it occurs at each phase of
the intervention, embolization is most pronounced
during stent implantation.19 Even though the num-
ber of microemboli correlates positively with the
severity of myocardial microvascular dysfunction
and myonecrosis, there is considerable overlap
with regard to the magnitude of plaque micro-
embolization between patients with and those
without periprocedural myocardial infarction.19,20
This finding suggests that factors other than the
burden of plaque microembolization influence the
likelihood of periprocedural myocardial infarc-
tion, such as the release of vasoactive factors from
the atherosclerotic plaque, platelet activation, and
preexisting vulnerability of the myocardium.2
T r a di t iona l F o cus on

454 n engl j med 364;5 nejm.org february 3, 2011

BACK to TOC
111
 current concepts

Figure 1. Mechanisms Underlying Periprocedural Myocardial Infarction.

Cardiac-biomarker elevation before percutaneous coronary intervention (PCI) is primarily due to spontaneous rupture of vulnerable
plaques, epicardial thrombosis, and subsequent myocardial injury. In the absence of abrupt, PCI-related epicardial-artery closure, peri-
procedural myocardial infarction is related to either side-branch occlusion or iatrogenic plaque rupture by balloons and stents, which
promotes microvascular injury owing to distal embolization, the release of vasoactive peptides, or both.

of cardiac-biomarker elevations after PCI, and
these studies have been systematically reviewed
in a previous publication.2 The general conclu-
sion from the retrospective analyses was that a
CK-MB elevation higher than 5 times the upper
limit of normal was independently associated with
an increased risk of in-hospital adverse cardiac
events, whereas lower levels did not appear to
influence in-hospital outcomes significantly (Ta-
ble 2).21,26,40-43 Data indicating a relationship be-
tween the CK-MB level and long-term survival were
less consistent. The results of several studies sug-

n engl j med 364;5 nejm.org february 3, 2011 455

BACK to TOC
112
 Table 2. Large Cohort Studies of Cardiac-Biomarker Elevation after Percutaneous Coronary Intervention (PCI), and Outcomes.*

456

BACK to TOC
No. of Incidence Type of Length of Multivariate Adjusted Long-Term
Study Patients of ACS Intervention Incidence In-Hospital Outcomes Follow-up Outcomes
% % mo
Stone et al.21 7147 68.9 PCI Increased risk of death (odds ratio, 8 for 24 Increased risk of death (hazard ratio, 2.2 for
CK-MB >8 ULN; 67 for Q-wave MI) CK-MB >8 ULN; 9.9 for Q-wave MI)
CK-MB, >432 ng/ml 29.6
CK-MB, >32 ng/ml 7.7
Q-wave MI 0.6
Dangas et al.22 4085 67.3 PCI NA 12 Increased risk of death or MI (odds ratio,
1.5 for CK-MB >5 ULN)
CK-MB, 420 ng/ml 24.1
CK-MB, >20 ng/ml 12.8
Ajani et al.23 1326 85.4 PCI NA 12 Increased risk of death or MI (odds ratio,
The

1.57 for CK-MB >3 ULN)

CK-MB, 412 ng/ml 25
CK-MB, >12 ng/ml 20
Kini et al.24 1675 NA PCI Increased risks of chest pain (58% 133 No association
vs. 912%), heart failure (35%
vs. 67%), and increased length

n engl j med 364;5

of stay for CK-MB >5 ULN vs.
other elevations and normal levels
CK-MB, 13 ULN 12.6

nejm.org
CK-MB, >35 ULN 3.7
CK-MB, >5 ULN 2.4
n e w e ng l a n d j o u r na l
of

Brener et al.25 3478 64.0 Stent placement NA 1515 Increased risk of death (odds ratio, 1.89
for CK-MB >3 ULN; 6.36 for CK-MB
>10 ULN)

february 3, 2011
CK-MB, >8.826.4 ng/ml 14.5
(CK-MB, >13 ULN)
m e dic i n e

CK-MB, >26.444 ng/ml 4.5

(CK-MB, >35 ULN)
CK-MB, >4488 ng/ml 2.9
(CK-MB, >510 ULN)
CK-MB, >88 ng/ml 2.4
(CK-MB, >10 ULN)
Ellis et al.26 8409 63.0 PCI Increased length of stay for CK-MB 4 Increased risk of death (1.2, 1.9, and 8.9%
>1 ULN for CK-MB <1, 15, and >5 ULN, re-
spectively)
CK-MB, 8.840 ng/ml 13.6
CK-MB, >40 ng/ml 3.6 48 Increased risk of death (15.1, 16.9, and 19.4%
for CK-MB <1, 15, and >5 ULN, re-
spectively)

113
 BACK to TOC
Brener et al.27 3573 NA PCI NA 36 Increased risk of death (hazard ratio, 1.1 for
CK-MB >10 ULN)
CK-MB, >8.826.4 ng/ml 21
(CK-MB, >13 ULN)
CK-MB, >26.444 ng/ml 6
(CK-MB, >35 ULN)
CK-MB, >4488 ng/ml 6
(CK-MB, >510 ULN)
CK-MB, >88 ng/ml 5
(CK-MB, >10 ULN)
Hong et al.28 1693 79.0 Saphenous-vein Increased need for balloon pump (7.8% 12 Increased risk of death (hazard ratio, 3.3 for
graft PCI vs. 1.1%) and repeat PCI (4.2% vs. CK-MB >5 ULN)
1.2%) for CK-MB elevation vs. no ele-
vation
CK-MB, 420 ng/ml 32.1
CK-MB, >20 ng/ml 15.2

n engl j med 364;5

Andron et al.29 3864 30.4 PCI NA 642 Increased risk of death (hazard ratio, 1.3,
1.76, and 2.26 for CK-MB 13, >35 and
>5 ULN, respectively)

nejm.org
CK-MB, 412 ng/ml 19.9
CK-MB, >1220 ng/ml 4.4
current concepts

CK-MB, >20 ng/ml 5.1

Jang et al.30 1807 40.9 Drug-eluting No association 137 Increased risk of death (0.5, 1.1, and 2.6%
stenting PCI for CK-MB <1, 15, and >5 ULN, re-

february 3, 2011
spectively)
CK-MB, 525 ng/ml 14.6
CK-MB, >25 ng/ml 6.4
Natarajan et al.31 1128 61.0 PCI Increased risk of major cardiac events 12 No association
(3.8 for cTnI 5 ULN)
cTnI, 14 ULN 7.6
cTnI, 5 ULN 9.1
Nallamothu et al.32 1157 36.5 PCI NA 117 Increased risk of death (hazard ratio, 2.4 for
cTnI 8 ULN, 8.9 for Q-wave MI)
cTnI, 25.9 ng/ml 16.0
cTnI, 69.9 ng/m 4.6
cTnI, 1015.9 ng/ml 2.0
cTnI, 16.0 ng/ml 6.5
Q-wave MI 0.3

457

114
 458

BACK to TOC
Table 2. (Continued.)

No. of Incidence Type of Length of Multivariate Adjusted Long-Term

Study Patients of ACS Intervention Incidence In-Hospital Outcomes Follow-up Outcomes
% % mo
33
Prasad et al. 1949 47.9 PCI Increased length of stay 26 Increased risk of death (hazard ratio, 1.2 per
log2 increase in cTnT)
cTnT, 0.03 ng/ml 19.6
Hubacek et al.34 1208 31.0 PCI NA 24 No association
Increase in cTnT >0.1 ng/ml 20
The

Feldman et al.35 1601 43.3 PCI No association 258 Increased risk of death (hazard ratio, 1.6)
cTnI, 0.15 ng/ml 51.9
De Labriolle et al.36 3200 0.0 PCI NA 12 No association
cTnI, >0.30 ng/ml 23.4
Cavallini et al.37 2362 45.1 PCI NA 24 No association

n engl j med 364;5

cTnI, 0.150.45 ng/ml 19.7
cTnI, >0.45 ng/ml 19.8
Fuchs et al.38 1129 70.9 PCI Increased risk of major adverse cardio- 8 No association

nejm.org
vascular events (odds ratio, 2.1 for
cTnI >3 ULN)
n e w e ng l a n d j o u r na l
of

cTnI, 0.150.45 ng/ml 15.2

cTnI, >0.45 ng/ml 15.4
CK-MB, >4 ng/ml 40.8

february 3, 2011
Cavallini et al.39 3494 50.8 PCI NA 24 Increased risk of death (odds ratio, 1.04 per
peak CK-MB ratio unit)
m e dic i n e

cTnI, >0.15 ng/ml 44.2

CK-MB, >5 ng/ml 16.0

* Plusminus values are means SD. Only data from studies that included at least 1000 patients, long-term outcome data, and concentration-based biomarker analysis are shown.
Hazard ratios were determined by means of a multivariate Cox proportional-hazards regression model, if available; otherwise, odds ratios were determined by multivariate analysis.
Acute coronary syndromes (ACS) included angina at rest and urgent priority interventions. CK-MB denotes the MB fraction of creatine kinase, cTnI cardiac troponin I, cTnT cardiac tro-
ponin T, MI myocardial infarction, NA not available, and ULN upper limit of the normal range.
Outcomes other than evolving myocardial infarction are shown. P<0.05 for all comparisons.
In this study, the final analysis was based on mass immunoassay.
This ratio was calculated by dividing the maximum post-PCI level by the ULN or baseline level of CK-MB.

115
 current concepts
gested that any elevation in CK-MB was associ-
ated with reduced long-term survival and that
there was a direct correlation between the mag-
nitude of myonecrosis and mortality.26,39,41,42 In
contrast, other studies have shown that only large
myocardial infarctions, variably defined as a
CK-MB level exceeding 5 or 8 times the upper
limit of normal or the presence of new Q waves,
were predictive of a poor long-term outcome, es-
pecially if they were related to an unsuccessful
revascularization procedure (Table 2).21,40,43,44
Studies evaluating the relationship between the
postprocedural cardiac troponin level and long-
term mortality, in general, have not excluded pa-
tients with acute coronary syndromes, many of
whom would have had abnormal cardiac-biomark-
er levels at baseline.31,32,35,39,45-47 Thus, the report-
ed frequency of postprocedural elevations in car-
diac troponin has been highly variable, and
although some studies showed that the serum
concentration of cardiac troponin was an inde-
pendent predictor of survival, others did not (Ta-
ble 2). The inconsistent findings were most likely
due to heterogeneity of the inclusion criteria, varia-
tions in the sensitivity and specificity of the bio-
marker assays, different sample sizes, and dif-
ferences in the duration of follow-up. Two recent
meta-analyses concluded that an elevated cardiac
troponin level after PCI does provide prognostic
information.48,49 Both analyses were influenced by
studies from our catheterization laboratories on
postprocedural cardiac troponin T elevations in
which we had reached a similar conclusion.33,50
However, the studies included in the meta-analyses
(including our own) had used cardiac troponin
cutoff values for normal that were higher than the
currently recommended 99th percentile, thereby
limiting the accuracy of their conclusions.8
Focus on Pr epro cedur a l R isk
To date, virtually all studies of periprocedural
myocardial infarction have been limited by the
lack of precision with which they determined
preprocedural risk. Contemporary cardiac tropo-
nin assays have greatly enhanced our ability to
detect myonecrosis before and after PCI.46,51 In a
recent analysis, using the currently recommend-
ed 99th percentile value as the cutoff for a nor-
mal cardiac troponin T level, we found that ap-
proximately one third of patients who underwent
nonemergency PCI had evidence of preprocedural
n engl j med 364;5
BACK to TOC
myonecrosis.6 These patients had a greater ath-
erosclerotic burden and more unstable disease
than patients without evidence of preprocedural
myonecrosis, a finding that is consistent with
previous reports.52 Applying the universal defini-
tion of myocardial infarction to patients with
normal preprocedural cardiac troponin T levels,
another one third of patients sustained a peripro-
cedural myocardial infarction after the proce-
dure when cardiac troponin T was used to detect
myonecrosis, as compared with only 1 in 15 pa-
tients when CK-MB was used.6 The preprocedural
rather than postprocedural cardiac-biomarker level
was a powerful independent predictor of short-
term and long-term mortality.6 Similar findings
have been reported in two additional recent stud-
ies that used cardiac troponin I within the frame-
work of the universal definition of myocardial
infarction36,37 and in an analysis from the Evalu-
ation of Drug Eluting Stents and Ischemic Events
(EVENT) registry.53
These observations may seem surprising,
since one might argue that the clinical effect of
myocardial infarction should be the same re-
gardless of its cause. However, most periproce-
dural myocardial infarcts are very small in rela-
tion to the magnitude of myonecrosis, especially
in patients with stable coronary artery disease.
Among patients with normal preprocedural car-
diac troponin values, less than 5% have CK-MB
values that are higher than 5 times the upper
reference limit after PCI, and Q-wave infarctions
are rare (<0.1%). Instead, CK-MB levels that are
higher than 5 times the upper reference limit are
generally observed in patients with elevated pre-
procedural cardiac troponin T.6 Thus, it is likely
that in the older studies that explored the effect
of periprocedural myocardial infarction on out-
comes, a large proportion of the patients who
had been classified as biomarker-negative on the
basis of CK or CK-MB levels at the time of PCI
actually had nonST-segment elevation myocar-
dial infarction according to contemporary defi-
nitions. This conclusion is supported by the high
proportion of patients (about 50% on average) who
had acute coronary syndromes in the previous
studies (Table 2, and the Supplementary Appen-
dix, available with the full text of this article at
NEJM.org).
In summary, recent studies reveal that the pre-
procedural cardiac troponin level is a powerful
independent predictor of prognosis after PCI.
nejm.org february 3, 2011 459
116
 The n e w e ng l a n d j o u r na l
Moreover, these studies suggest that the associa-
tion between postprocedural myonecrosis and
outcomes after an otherwise successful PCI is,
in general, a reflection of the preprocedural risk,
which can be estimated by measuring baseline
cardiac troponin levels with the use of contem-
porary high-sensitivity assays in conjunction with
the clinical and angiographic characteristics of the
patient.
Pro gnos t ic Signific a nce
of Per ipro cedur a l v er sus
Sp on ta neous E v en t s
On the basis of the traditional concept of peri-
procedural myocardial infarction described above,
this complication has often been equated with
spontaneous myocardial infarction in clinical tri-
als.54 The validity of this assumption has not been
examined in detail, and it has been confounded by
the variable definitions of periprocedural myocar-
dial infarction used in the past. The current uni-
versal definition of myocardial infarction attempts
to address this issue by introducing a specific
category (type 4a) for periprocedural myocardial
infarction to distinguish it from spontaneous
myocardial infarction (types 1 and 2).8
Akkerhuis and colleagues compared the effect
of periprocedural myocardial infarction as detect-
ed by CK-MB elevation with that of spontaneous
myocardial infarction on 6-month mortality in a
heterogeneous group of patients who had acute
coronary syndromes without ST-segment eleva-
tion; the data were derived from five different
clinical-trial databases.55 The authors reported a
positive correlation between CK-MB levels and
mortality in both groups, although the absolute
mortality was significantly higher among patients
who had spontaneous myocardial infarction than
among those who had periprocedural myocardial
infarction. The authors concluded that the clini-
cal significance of periprocedural myocardial in-
farction should be considered similar to the ad-
verse consequences of spontaneous myocardial
infarction. However, the study was conducted in
the era of balloon angioplasty, before the wide-
spread use of stents, and the analysis was not
adjusted for confounding clinical variables.
To address these limitations, an analysis was
conducted of data from the Acute Catheterization
and Urgent Intervention Triage Strategy (ACUITY)
trial (Clinical.Trials.gov number, NCT00093158)
460 n engl j med 364;5
BACK to TOC
of m e dic i n e
involving 7773 patients with moderate-to-high-
risk, nonST-segment elevation acute coronary
syndromes who underwent PCI.15 Periprocedural
myocardial infarction and spontaneous myocar-
dial infarction during follow-up developed in 6.0%
and 2.6% of the cohort, respectively. Among pa-
tients with either type of myocardial infarction,
as compared with those without myocardial in-
farction, unadjusted mortality at 1 year was sig-
nificantly higher. After adjustment for differences
in baseline and procedural characteristics between
the two groups, spontaneous myocardial infarc-
tion was a powerful independent predictor of an
increased risk of death, whereas periprocedural
myocardial infarction was not significantly as-
sociated with an increased risk of death. Similar
observations have been made among patients with
diabetes and stable coronary artery disease in the
Bypass Angioplasty Revascularization Investiga-
tion 2 Diabetes (BARI 2D) trial (NCT00006305),
in which a first spontaneous, symptomatic myo-
cardial infarction was associated with higher mor-
tality, as compared with myocardial infarction
induced by percutaneous or surgical revascular-
ization.56
Taken together, contemporary studies indicate
that spontaneous myocardial infarction is a pow-
erful predictor of mortality. Periprocedural myo-
cardial infarction, although frequent, is a marker
of atherosclerotic burden and procedural complex-
ity, but in most cases, it does not have important
independent prognostic significance in stable cor-
onary artery disease or in nonST-elevation acute
coronary syndromes. Although large periproce-
dural myocardial infarcts may affect prognosis,
they rarely occur in the absence of procedural com-
plications or in patients with normal baseline
cardiac troponin levels.
A r e a s of Uncer ta in t y
There is a pressing need for the interventional
community and the associated professional or-
ganizations to examine the new data and provide
more practical guidelines for defining periproce-
dural myocardial infarction. This process should
include an assessment of the appropriateness of
relying on biomarkers alone and of the low thresh-
old used for the universal definition, as compared
with a definition that includes clinical criteria such
as symptoms or evidence of ischemia or infarc-
tion on electrocardiography or cardiac imaging.
nejm.org february 3, 2011
117
 current concepts

Figure 2. Recommendations for the Prevention

and Management of Periprocedural Myocardial
Pre-PCI Troponin Level
Infarction.
ACS denotes acute coronary syndromes, CK-MB MB
fraction of creatine kinase, and GP glycoprotein. Normal Elevated

Since most of the data on periprocedural myo- Proceed with PCI

Recognize increased risk and
inform patient
cardial infarction are derived from patients with Consider initiation of antiplatelet
normal levels of cardiac biomarkers before the therapy
(e.g., GP IIb/IIIa inhibitors, clo-
procedure (predominantly those with stable or pidogrel) and high-dose statin
unstable angina), clearer guidelines are needed therapy
Manage ACS according to standard
with regard to whether periprocedural myocardial guidelines
infarction can be diagnosed in patients with non Treat decompensated heart failure,
if present
ST-elevation myocardial infarction in whom bio- Provide optimal care for coexisting
markers are rising before PCI and, if so, what di- conditions such as renal dysfunc-
tion, anemia, and diabetes
agnostic criteria should be used. This is probably
not feasible in contemporary practice, since PCI
is often performed within 24 hours after hospital PCI PCI
admission. Another practical issue that needs to If appropriate, use distal protection If appropriate, use distal protection
for vein-graft interventions for vein-graft interventions
be addressed is whether the class IIa recommen-
dation to routinely measure biomarkers after PCI is No procedural Procedural
complications complications
still appropriate and, if so, what the therapeutic or symptoms
implications of an elevated post-PCI level would of ischemia,
or both
be. A recent report from the National Cardiovas-
cular Data Registry indicates that the majority of
hospitals in the United States do not routinely
Observe in hospital according to stan- Allow longer in-hospital observation
measure cardiac biomarkers at the time of PCI.14 dard local practice Measure post-PCI CK-MB or troponin
The improved understanding of the clinical Consider measuring post-PCI troponin at 816 hr or as clinically indicated
or CK-MB at 816 hr (e.g., for those
significance of periprocedural myocardial infarc- with complex PCI)
tion has important implications for the design
of future randomized trials (i.e., periprocedural
myocardial infarction and spontaneous myocar-
dial infarction should not be considered equiva- Post-PCI Troponin Level
If troponin elevated, consider measuring CK-MB
lent clinical end points). This issue has most
recently been relevant with respect to the inter-
pretation of data from the Cangrelor versus Stan- Normal Elevated
dard Therapy to Achieve Optimal Management
of Platelet Inhibition (CHAMPION) PLATFORM Provide standard post-PCI care CK-MB >5 upper reference limit or
and secondary prevention for equivalent magnitude of tropo-
trial (NCT00385138).54 In that study, the majority coronary artery disease nin elevation or new Q waves
of patients had acute coronary syndromes with- Prolong in-hospital observation
(by at least 1 day)
out ST-segment elevation and underwent PCI with- Assess for left ventricular dys-
in 24 hours after presentation. This did not allow function
If indicated, repeat angiography
a reliable distinction between spontaneous myo- to identify procedural compli-
cardial infarction and periprocedural myocardi- cations and need for inter-
vention
al infarction, and led the investigators to con- Troponin greater than upper reference
clude that the result of the trial calls into limit and CK-MB <5 upper refer-
ence limit
question the definition of periprocedural MI used. Prolong duration of observation only
Differentiating spontaneous myocardial infarc- if clinically indicated (e.g., proce-
dural complication)
tion from periprocedural myocardial infarction Intensify secondary prevention for
will be increasingly difficult in clinical practice, coronary artery disease to ensure
optimal management
since most invasively managed cases involve car-
diac catheterization during a period when pre-

n engl j med 364;5 nejm.org february 3, 2011 461

BACK to TOC
118
 The n e w e ng l a n d j o u r na l of m e dic i n e

procedural biomarker levels would generally be
rising. Thus, we would caution against including
myocardial infarction as a component of the pri-
mary composite end point in future clinical tri-
als of PCI in acute coronary syndromes or using
it as a surrogate for long-term outcomes, although
one might reasonably consider it as a secondary
efficacy end point or a safety end point.
Impl ic at ions for Pr ac t ice
Our recommendation is that cardiac troponin lev-
els be routinely measured before PCI is performed
(Fig. 2). A normal preprocedural level of cardiac
troponin will assist in risk stratification by iden-
tifying patients in whom PCI can be performed
with very low risk and who may be considered for
early discharge from the hospital. In addition, a
pre-PCI elevation in cardiac troponin identifies
high-risk patients with complex or thrombotic le-
sions who may benefit from the preprocedural
initiation of potent antiplatelet therapies and
statins to improve outcomes.2,57,58 Post-PCI levels
should be routinely measured in patients who
have undergone complex procedures, who have
suboptimal angiographic results, or who have pro-
cedural complications, as well as in those who
have signs or symptoms of myocardial ischemia,
in order to quantify the extent of myocardial in-
jury. However, a reasonable case can be made for
not routinely measuring postprocedural cardiac
troponin levels in uncomplicated, successful PCI,
since it is not likely that in such cases relevant
additional information can be gained that will be
independent of the preprocedural risk and proce-
dural outcomes. The role of postprocedural mon-
itoring of biomarkers for risk stratification in the
secondary prevention of coronary artery disease or
as a metric of quality remains to be established.
There are no established cutoff values for car-
diac troponin that define a large periprocedural
myocardial infarct, and until such values can be
clearly identified, a CK-MB level that is more
than 5 times the upper reference limit, the pres-
ence of new Q waves, or both would appear to
be reasonable criteria for defining a periproce-
dural myocardial infarction as extensive. We be-
lieve that, in general, this definition can reliably
be applied only to patients with normal cardiac
troponin levels before PCI. In the absence of data
that can be used to help direct practice, we rec-
ommend that patients with large periprocedural
myocardial infarction be monitored in the hospi-
tal for an additional day because of the reported
risks of arrhythmias, hemodynamic instability,
heart failure, and death (Table 2, and the Supple-
mentary Appendix). For the purpose of preproce-
dural consent, one should discuss the frequency
of a large periprocedural myocardial infarction
(<5%) with the patient and inform the patient if it
occurs after the intervention.
The care of patients with acutely elevated pre-
procedural cardiac troponin who sustain major
periprocedural myonecrosis should, in general, be
based on the guidelines for managing acute coro-
nary syndromes. Patients whose condition unex-
pectedly deteriorates soon after PCI (e.g., those
with recurrent and unrelenting chest pain, par-
ticularly in combination with ST-segment shifts
or echocardiographic evidence of ischemia or peri-
cardial effusion) should undergo repeat coronary
angiography. The goal is to identify procedural
complications that are amenable to further in-
tervention, such as acute stent thrombosis, coro-
nary dissection, or perforation, to limit myone-
crosis and relieve symptoms. In most cases, this
involves repeat PCI; it is rare in current practice
for patients to require cardiac surgery.
Perhaps the most important implication for the
long-term care of the vast majority of patients with
periprocedural myocardial infarction is the realiza-
tion that they represent a higher-risk cohort owing
to a greater disease burden and more unstable
disease. These patients should therefore be target-
ed for optimal secondary prevention based on the
current guidelines. Occasionally, patients with
stable coronary artery disease have extensive peri-
procedural myocardial infarction. The long-term
care of such patients should be similar to that for
patients with spontaneous myocardial infarction.
Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.

References
1. Lloyd-Jones D, Adams RJ, Brown TM,
et al. Heart disease and stroke statistics
2010 update: a report from the Ameri-
can Heart Association. Circulation 2010;
121(7):e46-e215.
2. Herrmann J. Peri-procedural myocar-
dial injury: 2005 update. Eur Heart J
2005;26:2493-519.
3. Porto I, Selvanayagam JB, Van Gaal
WJ, et al. Plaque volume and occurrence
and location of periprocedural myocardi-
al necrosis after percutaneous coronary
intervention: insights from delayed-
enhancement magnetic resonance imag-
ing, Thrombolysis in Myocardial Infarc-
tion myocardial perfusion grade analysis,
and intravascular ultrasound. Circulation
2006;114:662-9.

462 n engl j med 364;5 nejm.org february 3, 2011

BACK to TOC
119

4. Bhatt DL, Topol EJ. Does creatinine
kinase-MB elevation after percutaneous
coronary intervention predict outcomes
in 2005? Periprocedural cardiac enzyme
elevation predicts adverse outcomes. Cir-
culation 2005;112:906-15.
5. Cutlip DE, Kuntz RE. Does creatinine
kinase-MB elevation after percutaneous
coronary intervention predict outcomes
in 2005? Cardiac enzyme elevation after
successful percutaneous coronary inter-
vention is not an independent predictor of
adverse outcomes. Circulation 2005;112:
916-22.
6. Prasad A, Rihal CS, Lennon RJ, Singh
M, Jaffe AS, Holmes DR Jr. Significance of
periprocedural myonecrosis on outcomes
after percutaneous coronary intervention:
an analysis of preintervention and post-
intervention troponin T levels in 5487 pa-
tients. Circ Cardiovasc Interv 2008;1:10-9.
7. Smith SC Jr, Feldman TE, Hirshfeld
JW Jr, et al. ACC/AHA/SCAI 2005 guide-
line update for percutaneous coronary
intervention: a report of the American
College of Cardiology/American Heart
Association Task Force on Practice Guide-
lines (ACC/AHA/SCAI Writing Committee
to Update 2001 Guidelines for Percutane-
ous Coronary Intervention). Circulation
2006;113(7):e166-e286.
8. Thygesen K, Alpert JS, White HD.
Universal definition of myocardial infarc-
tion. Eur Heart J 2007;28:2525-38.
9. Mandadi VR, DeVoe MC, Ambrose JA,
et al. Predictors of troponin elevation af-
ter percutaneous coronary intervention.
Am J Cardiol 2004;93:747-50.
10. Blankenship JC, Haldis T, Feit F, et al.
Angiographic adverse events, creatine
kinase-MB elevation, and ischemic end
points complicating percutaneous coro-
nary intervention (a REPLACE-2 substudy).
Am J Cardiol 2006;97:1591-6.
11. Cai Q, Skelding KA, Armstrong AT Jr,
Desai D, Wood GC, Blankenship JC. Pre-
dictors of periprocedural creatine kinase-
myocardial band elevation complicating
elective percutaneous coronary interven-
tion. Am J Cardiol 2007;99:616-20.
12. van Gaal WJ, Ponnuthurai FA, Selva-
nayagam J, et al. The Syntax score pre-
dicts peri-procedural myocardial necrosis
during percutaneous coronary interven-
tion. Int J Cardiol 2009;135:60-5.
13. Alcock RF, Roy P, Adorini K, et al. In-
cidence and determinants of myocardial
infarction following percutaneous coro-
nary interventions according to the re-
vised Joint Task Force definition of tropo-
nin T elevation. Int J Cardiol 2010;140:
66-72.
14. Wang TY, Peterson ED, Dai D, et al.
Patterns of cardiac marker surveillance
after elective percutaneous coronary in-
tervention and implications for the use of
periprocedural myocardial infarction as a
quality metric: a report from the National
Cardiovascular Data Registry (NCDR).
J Am Coll Cardiol 2008;51:2068-74.
n engl j med 364;5
BACK to TOC
current concepts
15. Prasad A, Gersh BJ, Bertrand ME, et
al. Prognostic significance of periproce-
dural versus spontaneously occurring
myocardial infarction after percutaneous
coronary intervention in patients with
acute coronary syndromes: an analysis
from the ACUITY (Acute Catheterization
and Urgent Intervention Triage Strategy)
trial. J Am Coll Cardiol 2009;54:477-86.
16. Srinivasan M, Rihal C, Holmes DR,
Prasad A. Adjunctive thrombectomy and
distal protection in primary percutaneous
coronary intervention: impact on micro-
vascular perfusion and outcomes. Circu-
lation 2009;119:1311-9.
17. Uetani T, Amano T, Ando H, et al. The
correlation between lipid volume in the
target lesion, measured by integrated
backscatter intravascular ultrasound, and
post-procedural myocardial infarction in
patients with elective stent implantation.
Eur Heart J 2008;29:1714-20.
18. Hong YJ, Mintz GS, Kim SW, et al. Im-
pact of plaque composition on cardiac
troponin elevation after percutaneous cor-
onary intervention: an ultrasound analy-
sis. JACC Cardiovasc Imaging 2009;2:458-
68.
19. Bahrmann P, Werner GS, Heusch G,
et al. Detection of coronary microemboli-
zation by Doppler ultrasound in patients
with stable angina pectoris undergoing
elective percutaneous coronary interven-
tions. Circulation 2007;115:600-8.
20. Herrmann J, Haude M, Lerman A, et
al. Abnormal coronary flow velocity re-
serve after coronary intervention is asso-
ciated with cardiac marker elevation. Cir-
culation 2001;103:2339-45.
21. Stone GW, Mehran R, Dangas G,
Lansky AJ, Kornowski R, Leon MB. Dif-
ferential impact on survival of electrocar-
diographic Q-wave versus enzymatic
myocardial infarction after percutaneous
intervention: a device-specific analysis of
7147 patients. Circulation 2001;104:642-7.
22. Dangas G, Mehran R, Feldman D, et
al. Postprocedure creatine kinase-MB el-
evation and baseline left ventricular dys-
function predict one-year mortality after
percutaneous coronary intervention. Am J
Cardiol 2002;89:586-9.
23. Ajani AE, Waksman R, Sharma AK, et
al. Usefulness of periprocedural creati-
nine phosphokinase-MB release to pre-
dict adverse outcomes after intracoronary
radiation therapy for in-stent restenosis.
Am J Cardiol 2004;93:313-7.
24. Kini A, Marmur JD, Kini S, et al. Cre-
atine kinase-MB elevation after coronary
intervention correlates with diffuse ath-
erosclerosis, and low-to-medium level el-
evation has a benign clinical course: im-
plications for early discharge after coronary
intervention. J Am Coll Cardiol 1999;
34:663-71.
25. Brener SJ, Ellis SG, Schneider J, Topol
EJ. Frequency and long-term impact of
myonecrosis after coronary stenting. Eur
Heart J 2002;23:869-76.
nejm.org february 3, 2011
26. Ellis SG, Chew D, Chan A, Whitlow
PL, Schneider JP, Topol EJ. Death follow-
ing creatine kinase-MB elevation after
coronary intervention: identification of
an early risk period: importance of cre-
atine kinase-MB level, completeness of re-
vascularization, ventricular function, and
probable benefit of statin therapy. Circu-
lation 2002;106:1205-10.
27. Brener SJ, Lytle BW, Schneider JP, Ellis
SG, Topol EJ. Association between CK-MB
elevation after percutaneous or surgical
revascularization and three-year mortali-
ty. J Am Coll Cardiol 2002;40:1961-7.
28. Hong MK, Mehran R, Dangas G, et al.
Creatine kinase-MB enzyme elevation fol-
lowing successful saphenous vein graft
intervention is associated with late mor-
tality. Circulation 1999;100:2400-5.
29. Andron M, Stables RH, Egred M, et al.
Impact of periprocedural creatine kinase-
MB isoenzyme release on long-term mor-
tality in contemporary percutaneous cor-
onary intervention. J Invasive Cardiol 2008;
20:108-12.
30. Jang JS, Hong MK, Park DW, et al. Im-
pact of periprocedural myonecrosis on
clinical events after implantation of drug-
eluting stents. Int J Cardiol 2008;129:368-
72.
31. Natarajan MK, Kreatsoulas C, Ve-
lianou JL, Mehta SR, Pericak D, Goodhart
DM. Incidence, predictors, and clinical
significance of troponin-I elevation with-
out creatine kinase elevation following
percutaneous coronary interventions. Am
J Cardiol 2004;93:750-3.
32. Nallamothu BK, Chetcuti S, Mukher-
jee D, et al. Prognostic implication of tro-
ponin I elevation after percutaneous coro-
nary intervention. Am J Cardiol 2003;91:
1272-4.
33. Prasad A, Singh M, Lerman A, Lennon
RJ, Holmes DR Jr, Rihal CS. Isolated ele-
vation in troponin T after percutaneous
coronary intervention is associated with
higher long-term mortality. J Am Coll
Cardiol 2006;48:1765-70.
34. Hubacek J, Basran RS, Shrive FM,
Shewchuk L, Goodhart DM, Anderson TJ.
Prognostic implications of C-reactive pro-
tein and troponin following percutane-
ous coronary intervention. Can J Cardiol
2009;25(2):e42-e47.
35. Feldman DN, Minutello RM, Berg-
man G, Moussa I, Wong SC. Relation of
troponin I levels following nonemergent
percutaneous coronary intervention to
short- and long-term outcomes. Am J Car-
diol 2009;104:1210-5.
36. De Labriolle A, Lemesle G, Bonello L,
et al. Prognostic significance of small tro-
ponin I rise after a successful elective per-
cutaneous coronary intervention of a na-
tive artery. Am J Cardiol 2009;103:639-45.
37. Cavallini C, Verdecchia P, Savonitto S,
et al. Prognostic value of isolated tropo-
nin I elevation after percutaneous coro-
nary intervention. Circ Cardiovasc Interv
2010;3:431-5.
463
120
 current concepts

38. Fuchs S, Kornowski R, Mehran R, et
al. Prognostic value of cardiac troponin-I
levels following catheter-based coronary
interventions. Am J Cardiol 2000;85:1077-
82.
39. Cavallini C, Savonitto S, Violini R, et
al. Impact of the elevation of biochemical
markers of myocardial damage on long-
term mortality after percutaneous coro-
nary intervention: results of the CK-MB and
PCI study. Eur Heart J 2005;26:1494-8.
40. Kugelmass AD, Cohen DJ, Moscucci
M, et al. Elevation of the creatine kinase
myocardial isoform following otherwise
successful directional coronary atherec-
tomy and stenting. Am J Cardiol 1994;74:
748-54.
41. Abdelmeguid AE, Topol EJ, Whitlow
PL, Sapp SK, Ellis SG. Significance of
mild transient release of creatine kinase-
MB fraction after percutaneous coronary
interventions. Circulation 1996;94:1528-36.
42. Kong TQ, Davidson CJ, Meyers SN,
Tauke JT, Parker MA, Bonow RO. Prog-
nostic implication of creatine kinase ele-
vation following elective coronary artery
interventions. JAMA 1997;277:461-6.
43. Saucedo JF, Mehran R, Dangas G, et
al. Long-term clinical events following
creatine kinasemyocardial band isoen-
zyme elevation after successful coronary
stenting. J Am Coll Cardiol 2000;35:1134-
41.
44. Jeremias A, Baim DS, Ho KK, et al.
Differential mortality risk of postproce-
dural creatine kinase-MB elevation fol-
lowing successful versus unsuccessful stent
procedures. J Am Coll Cardiol 2004;44:
1210-4.
45. Cantor WJ, Newby LK, Christenson
RH, et al. Prognostic significance of ele-
vated troponin I after percutaneous coro-
nary intervention. J Am Coll Cardiol
2002;39:1738-44.
46. Kizer JR, Muttrej MR, Matthai WH, et
al. Role of cardiac troponin T in the long-
term risk stratification of patients under-
going percutaneous coronary interven-
tion. Eur Heart J 2003;24:1314-22.
47. Kini AS, Lee P, Marmur JD, et al. Cor-
relation of postpercutaneous coronary
intervention creatine kinase-MB and tro-
ponin I elevation in predicting mid-term
mortality. Am J Cardiol 2004;93:18-23.
48. Nienhuis MB, Ottervanger JP, Bilo HJ,
Dikkeschei BD, Zijlstra F. Prognostic val-
ue of troponin after elective percutaneous
coronary intervention: a meta-analysis.
Catheter Cardiovasc Interv 2008;71:318-
24.
49. Testa L, Van Gaal WJ, Biondi Zoccai
GG, et al. Myocardial infarction after per-
cutaneous coronary intervention: a meta-
analysis of troponin elevation applying
the new universal definition. QJM 2009;
102:369-78.
50. Herrmann J, Von Birgelen C, Haude
M, et al. Prognostic implication of cardiac
troponin T increase following stent im-
plantation. Heart 2002;87:549-53.
51. Miller WL, Garratt KN, Burritt MF,
Lennon RJ, Reeder GS, Jaffe AS. Baseline
troponin level: key to understanding the
importance of post-PCI troponin eleva-
tions. Eur Heart J 2006;27:1061-9.
52. Heeschen C, van Den Brand MJ,
Hamm CW, Simoons ML. Angiographic
findings in patients with refractory unsta-
ble angina according to troponin T status.
Circulation 1999;100:1509-14.
53. Jeremias A, Kleiman NS, Nassif D, et
al. Prevalence and prognostic signifi-
cance of preprocedural cardiac troponin
elevation among patients with stable cor-
onary artery disease undergoing percuta-
neous coronary intervention: results from
the Evaluation of Drug Eluting Stents
and Ischemic Events registry. Circulation
2008;118:632-8.
54. Bhatt DL, Lincoff AM, Gibson CM, et
al. Intravenous platelet blockade with
cangrelor during PCI. N Engl J Med
2009;361:2330-41.
55. Akkerhuis KM, Alexander JH, Tardiff
BE, et al. Minor myocardial damage and
prognosis: are spontaneous and percuta-
neous coronary intervention-related events
different? Circulation 2002;105:554-6.
56. Chaitman BR, Hardison RM, Adler D,
et al. The Bypass Angioplasty Revascular-
ization Investigation 2 Diabetes random-
ized trial of different treatment strategies
in type 2 diabetes mellitus with stable
ischemic heart disease: impact of treat-
ment strategy on cardiac mortality and
myocardial infarction. Circulation 2009;
120:2529-40.
57. Herrmann J, Lerman A, Baumgart D,
et al. Preprocedural statin medication re-
duces the extent of periprocedural non-
Q-wave myocardial infarction. Circula-
tion 2002;106:2180-3.
58. Di Sciascio G, Patti G, Pasceri V, Gas-
pardone A, Colonna G, Montinaro A. Ef-
ficacy of atorvastatin reload in patients on
chronic statin therapy undergoing percu-
taneous coronary intervention: results of
the ARMYDA-RECAPTURE (Atorvastatin
for Reduction of Myocardial Damage
During Angioplasty) randomized trial.
J Am Coll Cardiol 2009;54:558-65.
Copyright 2011 Massachusetts Medical Society.

NEW NEJM APPLICATION FOR iPHONE

The NEJM Image Challenge app brings a popular online feature to the smartphone.
Optimized for viewing on the iPhone and iPod Touch, the Image Challenge app lets
you test your diagnostic skills anytime, anywhere. The Image Challenge app
randomly selects from 300 challenging clinical photos published in NEJM,
with a new image added each week. View an image, choose your answer,
get immediate feedback, and see how others answered.
The Image Challenge app is available at the iTunes App Store.

464 n engl j med 364;5 nejm.org february 3, 2011

BACK to TOC
121

From the Stop TB Department, World
Health Organization, Geneva. Address
reprint requests to Dr. Jaramillo at the
Stop TB Dept., World Health Organiza-
tion, CH-1211 Geneva, Switzerland, or at
jaramilloe@who.int.
N Engl J Med 2010;363:1050-8.
Copyright 2010 Massachusetts Medical Society.
1050
BACK to TOC
The n e w e ng l a n d j o u r na l of m e dic i n e
review article
current concepts
MDR Tuberculosis Critical Steps
for Prevention and Control
Eva Nathanson, M.Sc., Paul Nunn, F.R.C.P., Mukund Uplekar, M.D.,
Katherine Floyd, Ph.D., Ernesto Jaramillo, M.D., Ph.D., Knut Lnnroth, M.D., Ph.D.,
Diana Weil, M.Sc., and Mario Raviglione, M.D.
M
ultidrug-resistant (MDR) tuberculosis is defined as disease
caused by strains of Mycobacterium tuberculosis that are at least resistant to
treatment with isoniazid and rifampicin; extensively drug-resistant (XDR)
tuberculosis refers to disease caused by multidrug-resistant strains that are also
resistant to treatment with any fluoroquinolone and any of the injectable drugs
used in treatment with second-line anti-tuberculosis drugs (amikacin, capreomycin,
and kanamycin). MDR tuberculosis and XDR tuberculosis are serious threats to the
progress that has been made in the control of tuberculosis worldwide over the past
decade.1,2
In 2008, an estimated 440,000 cases of MDR tuberculosis emerged globally.1
India and China carry the greatest estimated burden of MDR tuberculosis, together
accounting for almost 50% of the worlds total cases. More than three quarters of
the estimated cases of MDR tuberculosis occur in previously untreated patients.
The proportion of MDR cases among new cases and previously treated cases of
tuberculosis reported globally from 1994 through 2009 ranged from 0 to 28.3% and
from 0 to 61.6%, respectively (Fig. 1). The highest proportions of MDR cases,
and the most severe drug-resistance patterns, appear in the countries of the former
Soviet Union. By 2009, a total of 58 countries had reported at least one case of
XDR tuberculosis. In eight countries, reported cases of XDR tuberculosis account
for more than 10% of all cases of MDR tuberculosis, and six of these countries
were part of the former Soviet Union. By far the largest number of cases of XDR
tuberculosis has been reported from South Africa (10.5% of all cases of MDR tuber-
culosis in that country), owing to rapid spread among people infected with the
human immunodeficiency virus (HIV).
National programs are failing to diagnose and treat MDR tuberculosis. Globally,
just under 30,000 cases of MDR tuberculosis were reported to the World Health
Organization (WHO) in 2008 (7% of the estimated total), of which less than one
fifth were managed according to international guidelines. The vast majority of the
remaining cases probably are not diagnosed or, if diagnosed, are mismanaged.
This problem remains despite the evidence that management of MDR tuberculosis
is cost-effective3 and that treatment of MDR tuberculosis, and even treatment of
XDR tuberculosis, is feasible in persons who are not infected with HIV.4,5
In some countries, the incidence of tuberculosis is rising, and the incidence of
MDR tuberculosis appears to be rising even faster (e.g., in Botswana and South
Korea).6 However, in Estonia, Hong Kong, the United States, and Orel and Tomsk
Oblasts (in the Russian Federation), the incidence of tuberculosis is falling, and
the incidence of MDR tuberculosis appears to be falling even faster.1,6 This trend is
n engl j med 363;11 nejm.org september 9, 2010
122
 current concepts

0 to <3%
3 to <6%
6 to <12%
12 to <18%
18%
No data available

Figure 1. Distribution of the Proportion of Cases of MDR Tuberculosis among New Cases of Tuberculosis, 19942009.
The following 27 countries are responsible for 85% of the worlds estimated cases of MDR tuberculosis and are classified as countries
with a high burden of MDR tuberculosis: China, India, Russia, Pakistan, Bangladesh, South Africa, Ukraine, Indonesia, Philippines, Nige-
ria, Uzbekistan, Democratic Republic of Congo, Kazakhstan, Vietnam, Ethiopia, Myanmar, Tajikistan, Azerbaijan, Moldova, Kyrgyzstan,
Belarus, Georgia, Bulgaria, Lithuania, Armenia, Latvia, and Estonia. Adapted from the 2010 report on MDR and XDR tuberculosis from
the WHO.1

the result of high-quality care and control prac-
tices that result in high rates of case detection
and cure, drug-susceptibility testing for all pa-
tients, and the provision of appropriate treat-
ment for all patients carrying drug-resistant
strains. In short, preventing initial infection
with MDR tuberculosis and managing the treat-
ment of existing cases appropriately are the keys
to containing the spread of this disease.
The WHO-recommended Stop TB Strategy7
provides the framework for treating and caring
for those who are sick and controlling the epi-
demic of drug-susceptible and drug-resistant dis-
ease. The DOTS approach, which underpins the
Stop TB Strategy, calls for political commitment
to national programs designed to control disease
by means of early diagnosis with the use of
bacteriologic testing, standardized treatment with
supervision and patient support, and provision
and management of the drugs used in treatment;
the approach also includes the monitoring of
treatment and evaluation of its effectiveness. Be-
tween 1995 and 2008, a total of 36 million people
were treated successfully with the use of the
DOTS approach, and 6 million lives were saved.8
Specific guidelines for controlling drug-suscep-
tible and drug-resistant disease already exist,9,10
and the Global Plan to Stop TB, 2006 through
2015, developed by the Stop TB Partnership,
specifies the scale at which these interventions
need to be funded and implemented to achieve
global targets.11 However, to date, planning,
funding, and implementation are falling far be-
hind the milestones that have been set.
Prompted by concern that political support for
the management of MDR tuberculosis is insuf-
ficient, WHO, the Bill and Melinda Gates Foun-
dation, and the Chinese Ministry of Health orga-
nized a ministerial conference in Beijing in April
2009.12 The report from the conference in Beijing
and the subsequent resolution (number 62.15)
approved by the World Health Assembly in May
2009 state that significant changes in several
components of the health care system must be
made if MDR tuberculosis is to be eliminated.13,14
This review assesses the critical factors imped-
ing control and discusses the solutions required
to address them.

n engl j med 363;11 nejm.org september 9, 2010 1051

BACK to TOC
123
 The n e w e ng l a n d j o u r na l
Cr i t ic a l W e a k ne sse s a nd How
t o A ddr ess Them
Prevention is better than cure. Thus, the top pri-
ority for the control and, ultimately, elimination
of MDR tuberculosis is prevention of its emer-
gence.15 Once MDR tuberculosis has emerged,
however, urgent measures are required to curb
its effects on efforts to control the disease. The
major obstacles and approaches to controlling
MDR tuberculosis are described below and sum-
marized in Table 1. Three topics of great impor-
tance the global shortage of health care work-
ers,16 the need for improvements in surveillance
systems,1 and the urgent need for intensified re-
search on new diagnostic tests, drugs, and vac-
cines17 have been well described elsewhere
and are beyond the scope of this article.
Financing Control and Care
To achieve the goal of universal access to diagno-
sis and treatment described in the Global Plan to
Stop TB, 1.3 million cases of MDR tuberculosis
in the 27 countries with the highest burden of
MDR disease will need to be treated between
2010 and 2015.1 The total estimated cost of such
treatment is several billion U.S. dollars, an amount
far in excess of the existing level of funding. The
national strategic plans in these countries must
incorporate the preparation of ambitious budgets
for the prevention and control of MDR tuberculo-
sis. These plans must be consistent with poli-
cies on health care financing, including social-
protection schemes (the delivery of commodities
to reduce the social vulnerability of poor popu-
lations), and with broader planning and financ-
ing frameworks. These countries especially
the middle-income countries among them
must mobilize their domestic resources. In 2001,
the WHO Commission on Macroeconomics and
Health indicated that these middle-income coun-
tries could finance all, or almost all, of their
health care needs.18 While maximizing the use
of domestic resources, they should also target re-
sources available from international financing
organizations, such as the Global Fund to Fight
AIDS (Acquired Immunodeficiency Syndrome),
Tuberculosis, and Malaria and UNITAID, an or-
ganization that provides grants allowing coun-
tries to purchase diagnostic tests and drugs used
in the treatment of HIVAIDS, malaria, and tu-
berculosis. The failure to adequately fund a re-
1052 n engl j med 363;11
BACK to TOC
of m e dic i n e
sponse to MDR tuberculosis would have cata-
strophic consequences in terms of both human
lives and tuberculosis control in general.
Abolishing Financial Barriers
Health expenditures that account for more than
40% of household income (after deducting the
cost of basic subsistence) have been defined as
catastrophic.19 In virtually all countries with a
high burden of MDR tuberculosis, treatment costs
(per course of treatment) for one person are more
than 100% of the gross national income per cap-
ita (the cost of second-line anti-tuberculosis drugs
alone is typically $2,000 to $4,000 per patient).1
Collective financing mechanisms are therefore
required to guarantee universal access to health
care. The main source of funding should be do-
mestic resources, such as contributions from tax-
es, payroll deductions, or mandatory insurance
premiums.20,21 Most countries in Africa, Asia,
and the Middle East have not attained universal
health coverage,22 although there are exceptions.
Lessons need to be drawn from universal health-
financing schemes applied in such diverse set-
tings as Mexico, Rwanda, and Thailand, where
access to care may facilitate early detection and
treatment of all tuberculosis cases.
Even before universal health coverage is
achieved, immediate steps can be taken to reduce
catastrophic health expenditures for patients with
tuberculosis and their households.23 These steps
include decentralization of services to reduce the
indirect costs that patients seeking care incur,
provision of patient incentives and social support
to promote adherence to treatment, and subsidi-
zation of care provided in the private sector that
is in line with guidelines from national tubercu-
losis programs.
Engaging All Care Providers
A substantial proportion of patients with tuber-
culosis or MDR tuberculosis seek care with pro-
viders who are not linked to national tuberculo-
sis programs.24,25 In five countries with a high
burden of MDR tuberculosis, more than half of
all sales of first-line anti-tuberculosis drugs oc-
cur in the private sector, and the proportion is
even higher for sales of second-line drugs.26
Many physicians in the private sector and some
in the public sector do not follow internationally
recommended treatment regimens for tuberculo-
sis, use medicines of questionable quality, and ne-
nejm.org september 9, 2010
124
 BACK to TOC
Table 1. Critical Challenges in the Control of MDR Tuberculosis and XDR Tuberculosis and Potential Solutions Supported by the WHO.*
Goal
Finance control and treatment
for MDR-TB and XDR-TB
Abolish financial barriers
Engage all care providers in appro-
priate MDR-TB prevention and
control
Optimize MDR-TB and XDR-TB
management and care
n engl j med 363;11
Address laboratory crisis
nejm.org
Ensure access to quality-assured
anti-TB drugs
september 9, 2010
Restrict availability of anti-TB
drugs
Prioritize TB infection control
Address global health workforce
crisis
Improve surveillance systems
Invest in research and develop-
ment of new diagnostic tests,
drugs, and vaccines
* The international group known as UNITAID purchases and distributes diagnostic tests and drugs used in the treatment of HIVAIDS, malaria, and tuberculosis. AIDS denotes acquired
immunodeficiency syndrome, MDR-TB multidrug-resistant tuberculosis, TB tuberculosis, WHO World Health Organization, and XDR-TB extensively drug-resistant tuberculosis.
1053
125
Problem
Estimated cost for 20102015 is $16.2 billion (in U.S $), increasing
annually from <$1.3 billion in 2010 to $4.4 billion in 2015; fund-
ing needed is already in excess of the planned national MDR-TB
budgets for 2010
In countries with a high burden of MDR-TB, treatment costs for a
single patient constitute more than 100% of the gross national
income per capita
A substantial proportion of patients seek care from providers who do
not follow internationally recommended standards of treatment
Persons with infectious MDR-TB and XDR-TB remain in the com-
munity for long periods of time because of delayed diagnosis and
initiation of treatment with second-line anti-TB drugs; hospitaliza-
tion of patients with MDR-TB or XDR-TB poses problem of noso-
comial transmission and is costly and inconvenient for patients
In 2008, in 27 countries with the highest burden of MDR-TB, only
1% of patients with newly diagnosed TB and 3% of patients with
previously treated TB underwent drug-susceptibility testing
Use of counterfeit and poor-quality anti-TB drugs, which can lead to
development and amplification of drug resistance, is well docu-
mented, but there is no accurate estimate of the scale of the
problem
Wide availability of anti-TB drugs over the counter in retail pharma-
cies encourages self-treatment and the purchase of inadequate
quantities and combinations of medicines
Ongoing transmission of MDR-TB and XDR-TB occurs in health
care facilities and congregate settings because of inadequate in-
fection control
Shortage of trained staff to effectively manage the 1.6 million MDR-
TB cases expected by 2015 is exacerbated in many low-income
countries by active recruitment of staff by industrialized countries
Estimates of the burden of drug-resistant TB globally and by country
remain incomplete and less than accurate
Tools for prevention, diagnosis, and treatment of TB and drug-resis-
tant TB are obsolete
Proposed Solution
Maximize use of domestic resources while targeting resources from the
Global Fund to Fight AIDS, Tuberculosis, and Malaria, UNITAID, and
other external funding mechanisms
Improve health care financing schemes to strive for universal access to
prevention and control; decentralize services to reduce indirect costs;
promote patient adherence to treatment; and subsidize care provided
in the private sector
Engage diverse providers (public, voluntary, private, and corporate) to align
TB-management practices with WHO International Standards for TB Care
Ensure timely diagnosis and treatment initiation for patients with MDR-TB
or XDR-TB; implement appropriate models of care, preferably outpa-
tient, to ensure patient-centered care, avoid disease transmission in
health care facilities, and make rational use of financial resources
Strengthen laboratory services by using new molecular technologies
current concepts
Secure affordable, quality-assured anti-TB drugs using national procure-
ment mechanisms while building up a reliable second-line anti-TB
drug market, with manufacturers investing in increased volumes and
improved quality
Restrict drug availability to accredited providers by combining government
policy, agreement with providers and industry on improved marketing
practices, and optimization of the protocol for drug management and
supply specified by the national TB program
Engage wide range of stakeholders across the health system (e.g., hospital
administrators, architects, engineers, and health care workers), includ-
ing those concerned with control of other infections with airborne po-
tential, such as influenza, to implement infection-control policies
Revise or update strategic plans for increasing the TB health care work-
force (including private health care providers) to improve basic TB
control and to scale up MDR-TB control
Establish or strengthen continuous surveillance systems for drug-resistant TB
Ensure collaboration between development and technical agencies to fa-
cilitate development and field testing of new tools for prevention, diag-
nosis, and treatment of TB
 The n e w e ng l a n d j o u r na l
glect essential principles of case management.27,28
Such practices lead to the development, amplifi-
cation, and spread of drug resistance. In addi-
tion, collaboration with public and private hospi-
tals warrants special attention.29
Guidance on implementing a mix of public
and private approaches to tuberculosis care is
available,30 and many national tuberculosis pro-
grams have begun to incorporate diverse sources
of care, including public, voluntary, private, and
corporate providers. Nonetheless, only a fraction of
the tuberculosis cases diagnosed by practitioners
outside the public sector are registered with or
referred to national tuberculosis programs.31,32
These approaches should therefore be scaled up
and applied to the prevention and management
of MDR tuberculosis as well. National tubercu-
losis programs need to play a stewardship role and
provide guidelines, training, technical and finan-
cial support, and the supervision needed to align
the practices of private providers with the Inter-
national Standards for TB Care.33 Effective en-
gagement of diverse care providers will require
national tuberculosis programs to both augment
their own capacities and strengthen private pro-
vider networks to enable them to shoulder their
responsibility for managing tuberculosis and MDR
tuberculosis. Professional associations need to act
as intermediaries between national tuberculosis
programs and private providers. Nongovernmen-
tal organizations have introduced successful pro-
grams for the management of MDR tuberculosis
in a number of countries and are key players in
scaling up diagnosis and treatment.34,35
But collaborative approaches and appropriate
incentives alone may not enlist the support of all
relevant care providers some regulation may
be necessary. In some countries with a high bur-
den of tuberculosis, providers are not required to
notify the government when a new case of tuber-
culosis has been diagnosed. And even in coun-
tries where notification is required, systems have
not been established to ensure that the require-
ment is met. Case notification for both tubercu-
losis and MDR tuberculosis must be made man-
datory; providers who follow best practices should
be certified and accredited and should be offered
access to free supplies of quality-assured anti-
tuberculosis drugs for their patients.30 Sustain-
able engagement of all care providers will require
national tuberculosis programs to work in close
1054 n engl j med 363;11
BACK to TOC
of m e dic i n e
partnership with health professionals, represen-
tatives of the pharmaceutical industry, pharma-
cists, and drug regulatory authorities, in addition
to consumer and patient associations.
Optimizing Disease Management and Care
Transmission of drug-resistant tuberculosis oc-
curs in the community,36 as indicated by the high
frequencies of MDR tuberculosis among previ-
ously untreated patients in some countries. In
most countries with limited resources, patients
with MDR or XDR tuberculosis must complete
two unsuccessful courses of treatment with first-
line anti-tuberculosis drugs before being eligible
for treatment with second-line drugs.37 Moreover,
in many countries, treatment of MDR tuberculo-
sis is started only after the diagnosis has been
confirmed, a process that takes months when
conventional methods are used. As a result, per-
sons with infectious MDR or XDR tuberculosis
remain in the community for long periods of
time. Prompt diagnosis and treatment of tuber-
culosis and MDR tuberculosis can keep the case
reproduction number of MDR strains below their
replacement rate and perhaps even below that
of non-MDR strains.6
Outbreaks of MDR tuberculosis have occurred
in hospitals, and patients with tuberculosis who
are hospitalized have a higher risk of acquiring
MDR tuberculosis than do those who are treated
as outpatients.38,39 Treating MDR tuberculosis in
a hospital is more expensive than doing so on an
ambulatory basis. Hospital treatment is also more
socially and economically disruptive for most
patients.40 In addition, the number of hospital
beds may become insufficient as countries ex-
pand treatment for MDR tuberculosis. Despite
the complexities involved in caring for patients
with MDR tuberculosis, including lengthy ther-
apy with poorly tolerated drugs, clinic-based or
community-based care has proved to be feasible
and effective in several countries, including Ne-
pal41 and Peru.42 However, the effectiveness of
outpatient care depends on the availability of pri-
mary care facilities, qualified health care work-
ers, and social support networks to promote ad-
herence to treatment. Countries need to select
the model of care that is right for them, taking
into account the personal rights and needs of
patients and communities,43 the numbers of pa-
tients who have both MDR tuberculosis and
nejm.org september 9, 2010
126
 current concepts

HIVAIDS, the social circumstances of patients,44 risk of the development of drug resistance.51 The
the health care infrastructure, and the ability of use of counterfeit and poor-quality anti-tubercu-
the country to mobilize resources. losis drugs, which can lead to the development
and amplification of drug resistance, is well doc-
Responding to the Laboratory Crisis umented, but there is no accurate estimate of the
Weak laboratory capacity remains a serious im- scale of the problem.52,53 International quality
pediment to prompt diagnosis and better control standards have been developed but are often ig-
of MDR tuberculosis.1 The goal of universal ac- nored, and an insufficient number of manufac-
cess to drug-susceptibility testing has not yet been turers have been approved under the WHO Pre-
achieved. In 2008, drug-susceptibility testing was qualification Programme.54
performed in only 1% of new tuberculosis cases To effectively prevent and manage MDR tu-
and 3% of previously treated cases in the 27 berculosis, countries need to secure affordable,
countries with the highest burden of MDR tuber- quality-assured, anti-tuberculosis drugs through
culosis. national procurement mechanisms. Affordable
Today, rapid molecular tests for MDR tubercu- and quality-assured, second-line anti-tuberculo-
losis are available.45 For instance, one new auto- sis drugs can also be accessed through the WHO
mated rapid test for rifampicin resistance holds Green Light Committee, which ensures manage-
promise for easier detection of MDR tuberculosis ment of MDR tuberculosis that is in line with
even in community settings.46 The implementa- international quality standards in 70 countries.1
tion of this and other rapid tests, especially in However, of particular concern for efforts to in-
countries with a high prevalence of concurrent crease the scale of MDR tuberculosis manage-
HIV infection and MDR tuberculosis, can prevent ment is the insufficient supply of quality-assured,
fatal delays in detection.47 The establishment of second-line anti-tuberculosis drugs.13 As of April
quality-assured diagnostic capacity, including rapid 2010, only two manufacturers that produce three
diagnostic technologies to identify MDR tubercu- of the seven second-line anti-tuberculosis drugs
losis, is feasible in resource-limited settings.48 Use on the WHO Model List of Essential Medicines
of the new molecular technologies offers one of had been approved by the WHO Prequalification
the best avenues for improving overall diagnostic Programme.54 Building up a reliable market of
capacity in the laboratory.49 At present, however, second-line anti-tuberculosis drugs, with manu-
the adoption of the new rapid tests will not elimi- facturers investing in increased volumes and im-
nate the need for conventional drug-susceptibility proved quality, requires more accurate forecast-
testing with the use of solid or liquid culture. Con- ing of demand. In addition, national authorities
ventional susceptibility testing is required to de- need to expedite the enrollment of many more
termine susceptibility to drugs other than rifam- patients under proper management conditions.
picin and isoniazid.9 While countries expand
laboratory capacity and introduce the new rapid Restricting Drug Availability
tests, targeted drug-susceptibility testing should be Anti-tuberculosis drugs are widely available over
performed in specific groups of patients at risk for the counter in retail pharmacies in many coun-
drug resistance. Expansion of diagnostic capacity tries.55 This encourages self-treatment and the
for MDR tuberculosis must be coupled with ac- purchase of inadequate quantities and combina-
cess to second-line anti-tuberculosis drugs. Efforts tions of medicines. Even when the drugs are pre-
to shorten the time required for diagnosis must scribed, those prescribing the drugs outside na-
occur in tandem with measures that minimize or- tional tuberculosis programs may not abide by
ganizational delay to ensure prompt initiation of recommended regimens, and some patients may
treatment. purchase only part of the prescription because of
financial constraints.56 Prescription and dispens-
Ensuring Access to Quality-Assured Drugs ing of medicines in general, and of antibiotics in
In 2007, only 15% of reported new cases of tuber- particular, are poorly monitored and regulated
culosis were treated with fixed-dose combina- in most countries.57 Even when regulations exist,
tions of anti-tuberculosis drugs,50 despite their their enforcement is often insufficient.
logistic advantages and potential to reduce the An essential step toward improved prevention

n engl j med 363;11 nejm.org september 9, 2010 1055

BACK to TOC
127
 The n e w e ng l a n d j o u r na l
of MDR tuberculosis is to encourage the engage-
ment of private and public providers with nation-
al tuberculosis programs on a voluntary basis.30
A more forceful approach would be to restrict the
right to prescribe and dispense the drugs to the
national tuberculosis program itself or to pro-
viders that have been accredited by the program.
Either approach would require a combination of
new government policy and dialogue with care
providers, including pharmacists, and the phar-
maceutical industry. Such measures undertaken
by national tuberculosis programs to optimize
drug management and supply have been success-
ful in some countries, including Brazil, Ghana,
Syria, and Tanzania. Consumers also need to
be aware of the risks of poor prescribing prac-
tices and, as discussed above, the clinical and
public health threats posed by substandard medi-
cines.52,57 Demand-driven efforts to push for more
accountability and enforcement of regulations
by national authorities may be highly effective.
Further advances in social responsibility and im-
proved marketing practices on the part of drug
manufacturers are also essential, along with
supportive government measures.
Prioritizing Control of Tuberculosis
Infection
As a result of inadequate measures of infection
control, there is ongoing transmission of MDR
tuberculosis and XDR tuberculosis in health care
facilities and congregate settings (e.g., prisons).38
To date, virtually no country with a high burden
of tuberculosis has implemented systematic mea-
sures to reduce the risk of tuberculosis transmis-
sion in health facilities.1 Health care workers,
especially those working in tuberculosis hospi-
tals and in resource-limited settings, are at sub-
stantially higher risk of contracting tuberculosis
and MDR tuberculosis than the general popula-
tion.58,59
All health care facilities that admit patients
with tuberculosis or patients suspected of having
tuberculosis should implement tuberculosis-con-
trol measures that complement general measures
of infection control, especially those which tar-
get other airborne infections.60 Home-based and
community treatment of MDR tuberculosis should
be promoted. To curb the increased risk of noso-
comial tuberculosis and MDR tuberculosis among
health care workers, some countries have added
1056 n engl j med 363;11
BACK to TOC
of m e dic i n e
tuberculosis to the list of recognized occupa-
tional hazards.59 Infection control requires en-
gagement with a wide range of stakeholders
across the health care system, including hospital
administrators, architects, engineers, doctors,
nurses, and laboratory staff. On the policy level,
infection control requires collaborative action
among those concerned with infections with
airborne potential, such as influenza.
The Urgen t Need for Ac tion
Critical weaknesses in current approaches to the
treatment and control of MDR tuberculosis and
XDR tuberculosis have been identified and are
being addressed at the global level. In 2009, the
Beijing Call for Action13 and the passage of World
Health Assembly Resolution 62.1514 signaled a
major step forward in coordinated planning for
the treatment and control of MDR tuberculosis
and in the commitment to achieve universal ac-
cess to diagnosis and treatment by 2015 for pa-
tients who have the disease. Resolutions, however,
are useful only insofar as they stimulate the ap-
propriate policymakers in governments to act on
them. By October 2009, 20 of the 27 countries
with the highest burden of MDR tuberculosis were
updating their national tuberculosis-control plans
to include a component addressing MDR tuber-
culosis, in compliance with the World Health As-
sembly resolution. Furthermore, for the countries
that have received grants from the Global Fund
in its ninth round of grants, funding requested
for the management of MDR tuberculosis was by
far the largest requested for all aspects of tuber-
culosis control: more than $500 million (in U.S.
dollars) was requested for the management of
MDR tuberculosis in 28 countries over a period
of 5 years.
Every one of the recommendations in this ar-
ticle for improving the treatment and control of
MDR tuberculosis requires action beyond national
tuberculosis control programs, sometimes in the
political environment outside the health care sys-
tem. This is a highly ambitious but necessary
agenda for health authorities in the affected
countries and for the global health community.
The steps involved in controlling MDR tubercu-
losis are also important steps toward strengthen-
ing health care systems, including progress in
achieving universal health care coverage. If this
nejm.org september 9, 2010
128
 current concepts

policy agenda is not pursued with urgency, the
human and financial costs to societies will be
profound.
Supported by a grant from the Bill and Melinda Gates Foun-
dation.
No potential conflict of interest relevant to this article was
reported.
Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.
We thank Karin Bergstrm, Lopold Blanc, Robert Matiru,
Andrea Pantoja, Fabio Scano, Karin Weyer, and Matteo Zignol
for their support in developing the background documents for
the ministerial meeting in Beijing in April 2009 and for review-
ing earlier versions of the manuscript.

References
1. World Health Organization. Multi-
drug and extensively drug-resistant TB
(M/XDR-TB): 2010 global report on sur-
veillance and response. 2010. (Accessed
August 16, 2010, at http://whqlibdoc.who
.int/publications/2010/9789241599191_eng
.pdf.)
2. Raviglione MC, Smith IM. XDR tuber-
culosis implications for global public
health. N Engl J Med 2007;356:656-9.
3. Tupasi TE, Gupta R, Quelapio MI, et al.
Feasibility and cost-effectiveness of treat-
ing multidrug-resistant tuberculosis: a co-
hort study in the Philippines. PLoS Med
2006;3(9):e352.
4. Orenstein EW, Basu S, Shah NS, et al.
Treatment outcomes among patients with
multidrug-resistant tuberculosis: system-
atic review and meta-analysis. Lancet In-
fect Dis 2009;9:153-61.
5. Sotgiu G, Ferrara G, Matteelli A, et al.
Epidemiology and clinical management
of XDR-TB: a systematic review by TBNET.
Eur Respir J 2009;33:871-81.
6. Dye C. Doomsday postponed? Prevent-
ing and reversing epidemics of drug resis-
tant tuberculosis. Nat Rev Microbiol 2009;
7:81-7.
7. Raviglione MC, Uplekar MW. WHOs
new Stop TB Strategy. Lancet 2006;367:
952-5.
8. Lnnroth K, Castro KG, Chakaya JM,
et al. Tuberculosis control and elimina-
tion 2010-50: cure, care, and social devel-
opment. Lancet 2010;375:1814-29.
9. World Health Organization. Treat-
ment of tuberculosis: guidelines. 4th ed.
2009 (WHO/HTM/TB/2009.420). Geneva,
2010.
10. World Health Organization. Guide-
lines for the programmatic management
of drug-resistant tuberculosis: emergency
update 2008. (Accessed August 16, 2010,
at http://whqlibdoc.who.int/publications/
2008/9789241547581_eng.pdf.)
11. Stop TB. Partnership: global plan to
stop TB, 2006-2015. (Accessed August 16,
2010, at http://www.stoptb.org/global/
plan/default.asp.)
12. Cheng MH. Ministerial meeting agrees
plan for tuberculosis control. Lancet 2009;
373:1328.
13. World Health Organization. Global tu-
berculosis control and patient care: a min-
isterial meeting of high M/XDR-TB burden
countries. (Accessed August 16, 2010, at
http://www.who.int/tb_beijingmeeting/en/
index.html.)
14. Idem. Prevention and control of multi-
drug-resistant tuberculosis and extensive-
ly drug-resistant tuberculosis. World
Health Assembly resolution 62.15. May
2009. (Accessed August 16, 2010, at
http://apps.who.int/gb/ebwha/pdf_files/
A62/A62_R15-en.pdf.)
15. Reichman LB. Unsexy tuberculosis.
Lancet 2009;373:28.
16. The world health report 2006: work-
ing together for health. Geneva: World
Health Organization, 2006. (Accessed Au-
gust 16, 2010, at http://www.who.int/whr/
2006/whr06_en.pdf.)
17. Fauci AS. Opinion: fighting TB should
be priority. New York: Msnbc.com, 2009.
(Accessed August 16, 2010, at http://www
.msnbc.msn.com/id/33890464/ns/
health-infectious_diseases/.)
18. Macroeconomics and health: investing
in health for economic development. Re-
port of the Commission on Macroeco-
nomics and Health. Geneva: World Health
Organization, 2001. (Accessed August
16, 2010, at http://whqlibdoc.who.int/
publications/2001/924154550X.pdf.)
19. Xu K, Evans DB, Kawabata K, Zeram-
dini R, Klavus J, Murray CJL. Household
catastrophic health expenditure: a multi-
country analysis. Lancet 2003;362:111-7.
20. Social health protection: an ILO strat-
egy towards universal access to health care.
Social security policy briefings. Paper 1.
Geneva: International Labour Organiza-
tion, 2008. (Accessed August 16, 2010, at
http://www.ilo.org/public/english/
protection/secsoc/downloads/policy/
policy1e.pdf.)
21. World Health Organization. Social
health insurance: sustainable health fi-
nancing, universal coverage and social
health insurance. April 2005. (Accessed
August 16, 2010, at http://apps.who.int/gb/
ebwha/pdf_files/WHA58/A58_20-en.pdf.)
22. Garrett L, Chowdhury AM, Pablos-
Mndez A. All for universal health cover-
age. Lancet 2009;374:1294-9.
23. Hanson C, Floyd K, Weil D. Tubercu-
losis in the poverty alleviation agenda. In:
Raviglione M, ed. TB: a comprehensive
international approach. New York: Infor-
ma Healthcare, 2006:1097-114.
24. Pantoja A, Floyd K, Unnikrishnan KP,
et al. Economic evaluation of public-pri-
vate mix for tuberculosis care and con-
trol, India. I. Socio-economic profile and
costs among tuberculosis patients. Int J
Tuberc Lung Dis 2009;13:698-704.
25. Lnnroth K, Aung T, Maung W, Kluge
H, Uplekar M. Social franchising of TB
care through private GPs in Myanmar: an
assessment of treatment results, access,
equity and financial protection. Health
Policy Plan 2007;22:156-66.
26. Pathway to patients: charting the dy-
namics of the global TB drug market.
New York: TB Alliance, 2007. (Accessed
August 16, 2010, at http://www.tballiance
.org/downloads/publications/Pathway_to_
Patients_Compendium_FINAL.pdf.)
27. Loveday M, Thomson L, Chopra M,
Ndlela Z. A health systems assessment of
the KwaZulu-Natal tuberculosis pro-
gramme in the context of increasing drug
resistance. Int J Tuberc Lung Dis 2008;
12:1042-7.
28. Uplekar M, Pathania V, Raviglione M.
Private practitioners and public health:
weak links in tuberculosis control. Lancet
2001;358:912-6.
29. Uplekar M. Stopping tuberculosis:
time to turn urgent attention to hospitals.
Int J Tuberc Lung Dis 2008;12:986.
30. World Health Organization. Engaging
all health care providers in TB control:
guidance on implementing public-private
mix approaches. 2006. (Accessed August
16, 2010, at http://whqlibdoc.who.int/hq/
2006/WHO_HTM_TB_2006.360_eng.pdf.)
31. Lnnroth K, Uplekar M, Blanc L. Hard
gains through soft contracts: productive
engagement of private providers in tuber-
culosis control. Bull World Health Organ
2006;84:876-83.
32. Irawati SR, Basri C, Arias MS, et al.
Hospital DOTS linkage in Indonesia:
a model for DOTS expansion into govern-
ment and private hospitals. Int J Tuberc
Lung Dis 2007;11:33-9.
33. Hopewell PC, Pai M, Maher D, Uplekar
M, Raviglione MC. International stan-
dards for tuberculosis care. Lancet Infect
Dis 2006;6:710-25.
34. Van Deun A, Maug AK, Salim MA, et
al. Short, highly effective and inexpensive
standardized treatment of multidrug-resis-
tant tuberculosis. Am J Respir Crit Care
Med 2010 May 4 (Epub ahead of print).
35. Cox HS, Kalon S, Allamuratova S, et al.
Multidrug-resistant tuberculosis treatment

n engl j med 363;11 nejm.org september 9, 2010 1057

BACK to TOC
129

outcomes in Karakalpakstan, Uzbekistan:
treatment complexity and XDR-TB among
treatment failures. PLoS ONE 2007;2(11):
e1126.
36. Marais BJ, Victor TC, Hesseling AC,
et al. Beijing and Haarlem genotypes are
overrepresented among children with drug-
resistant tuberculosis in the Western Cape
Province of South Africa. J Clin Microbiol
2006;44:3539-43.
37. Basu S, Friedland GH, Medlock J, et al.
Averting epidemics of extensively drug-
resistant tuberculosis. Proc Natl Acad Sci
U S A 2009;106:7672-7.
38. Nodieva A, Jansone I, Broka L, Pole I,
Skenders G, Baumanis V. Recent nosoco-
mial transmission and genotypes of multi-
drug-resistant Mycobacterium tuberculo-
sis. Int J Tuberc Lung Dis 2010;14:427-33.
39. Gelmanova IY, Keshavjee S, Golub-
chikova VT, et al. Barriers to successful
tuberculosis treatment in Tomsk, Russian
Federation: non-adherence, default and
the acquisition of multidrug resistance.
Bull World Health Organ 2007;85:703-
11.
40. Heller T, Lessells RJ, Wallrauch CG,
et al. Community-based treatment for
multidrug-resistant tuberculosis in rural
KwaZulu-Natal, South Africa. Int J Tuberc
Lung Dis 2010;14:420-6.
41. Malla P, Kanitz EE, Akhtar M, et al.
Ambulatory-based standardized therapy
for multi-drug resistant tuberculosis: ex-
perience from Nepal, 2005-2006. PLoS
ONE 2009;4:e8313.
42. Shin S, Furin J, Bayona J, Mate K, Kim
JY, Farmer P. Community-based treatment
of multidrug-resistant tuberculosis in
Lima, Peru: 7 years of experience. Soc Sci
Med 2004;59:1529-39.
43. Singh JA, Upshur R, Padayatchi N.
XDR-TB in South Africa: no time for de-
The Journal welcomes consideration of new submissions for Images in Clinical
Medicine. Instructions for authors and procedures for submissions can be found
on the Journals Web site at NEJM.org. At the discretion of the editor, images that
are accepted for publication may appear in the print version of the Journal,
1058 n engl j med 363;11
BACK to TOC
current concepts
nial or complacency. PLoS Med 2007;4(1):
e50.
44. Floyd K, Hutubessy R, Samyshkin Y,
et al. Health-systems efficiency in the
Russian Federation: tuberculosis control.
Bull World Health Organ 2006;84:43-51.
45. World Health Organization. Molecu-
lar line probe assays for rapid screening
of patients at risk of multidrug-resistant
tuberculosis (MDR-TB). Policy statement.
June 27, 2008. (Accessed August 16, 2010,
at http://www.who.int/tb/dots/laboratory/
lpa_policy.pdf.)
46. Boehme CC, Nabeta P, Hillemann D,
et al. Rapid molecular detection of tuber-
culosis and rifampin resistance. N Engl J
Med 2010;363:1005-15.
47. Gandhi NR, Shah NS, Andrews JR,
et al. HIV coinfection in multidrug- and
extensively drug-resistant tuberculosis re-
sults in high early mortality. Am J Respir
Crit Care Med 2009;181:80-6.
48. Paramasivan CN, Lee E, Kao K, et al.
Experience establishing tuberculosis lab-
oratory capacity in a developing country
setting. Int J Tuberc Lung Dis 2010;14:59-
64.
49. Birx D, de Souza M, Nkengasong JN.
Laboratory challenges in the scaling up of
HIV, TB, and malaria programs: the inter-
action of health and laboratory systems,
clinical research, and service delivery. Am
J Clin Pathol 2009;131:849-51.
50. Global tuberculosis control: epidemi-
ology, strategy, financing. WHO report
2009. Geneva: World Health Organization,
2009. (WHO/HTM/TB/2009.411.) (Accessed
August 16, 2010, at http://www.who.int/tb/
publications/global_report/en/index.html.)
51. Blomberg B, Fourie B. Fixed-dose com-
bination drugs for tuberculosis: applica-
tion in standardised treatment regimens.
Drugs 2003;63:535-53.
IMAGES IN CLINICAL MEDICINE
the electronic version, or both.
nejm.org september 9, 2010
52. Caudron JM, Ford N, Henkens M,
Mac C, Kiddle-Monroe R, Pinel J. Sub-
standard medicines in resource-poor set-
tings: a problem that can no longer be
ignored. Trop Med Int Health 2008;13:
1062-72.
53. Newton PN, Green MD, Fernndez
FM, Day NP, White NJ. Counterfeit anti-
infective drugs. Lancet Infect Dis 2006;6:
602-13.
54. World Health Organization. Prequali-
fication programme. (Accessed August
16, 2010, at http://apps.who.int/prequal/.)
55. Kobaidze K, Salakaia A, Blumberg
HM. Over the counter availability of anti-
tuberculosis drugs in Tbilisi, Georgia, in
the setting of a high prevalence of MDR-
TB. Interdiscip Perspect Infect Dis 2009;
2009:513-609.
56. Uplekar M, Juvekar S, Morankar S,
Rangan S, Nunn P. Tuberculosis patients
and practitioners in private clinics in In-
dia. Int J Tuberc Lung Dis 1998;2:324-9.
57. Cars O, Hgberg LD, Murray M, et al.
Meeting the challenge of antibiotic resis-
tance. BMJ 2008;337:a1438.
58. Skodric-Trifunovic V, Markovic-Denic
L, Nagorni-Obradovic L, Vlajinac H, Woeltje
KF. The risk of occupational tuberculosis
in Serbian health care workers. Int J Tu-
berc Lung Dis 2009;13:640-4.
59. Naidoo S, Jinabhai CC. TB in health
care workers in KwaZulu-Natal, South Af-
rica. Int J Tuberc Lung Dis 2006;10:676-
82.
60. WHO policy on TB infection control
in health-care facilities, congregate set-
tings and households. Geneva: World
Health Organization, 2009. (Accessed
August 16, 2010, at http://whqlibdoc.who
.int/publications/2009/9789241598323_eng
.pdf.)
Copyright 2010 Massachusetts Medical Society.
130
 Other Clinical Resources

Videos in Clinical Medicine allow you to watch common clinical procedures on your computer or
handheld device. Peer-reviewed for accuracy and chaptered for easy reference, these brief videos
provide a concise review of a procedure, including preparation, equipment, and more.
Interactive Medical Cases allow readers to follow along with experts managing an actual case, from
presentation through outcome. You are presented with patient information, then asked multiple-
choice questions. Videos, lab results and brief commentary explain important concepts.

BACK to TOC
131
 The n e w e ng l a n d j o u r na l of m e dic i n e

videos in clinical medicine

Clinical Evaluation of the Knee

Teresa L. Schraeder, M.D., Richard M. Terek, M.D., and C. Christopher Smith, M.D.

The knee is one of the most complex joints in the body. Not surprisingly, knee From Mt. Auburn Hospital, Cambridge,
problems are one of the most common musculoskeletal symptoms evaluated by the MA (T.L.S.); Warren Alpert Medical School
(T.L.S.) and the Department of Ortho-
primary care physician.1 paedics (R.M.T.), Brown University, Prov-
idence, RI; and the Division of General
Anatomy Medicine and Primary Care, Beth Israel
Deaconess Medical Center, and Harvard
The knee joint consists of three bones the femur, tibia, and patella with three Medical School (C.C.S.) both in Bos-
areas of articulation. The important soft-tissue structures of the knee include the ton. Address reprint requests to Dr. Terek
four major ligaments the anterior cruciate, posterior cruciate, and medial and at the Department of Orthopaedic Sur-
gery, Brown University, 2 Dudley St.,
lateral collateral ligaments the joint capsule, the medial and lateral menisci, the Suite 200, Providence, RI 02905, or at
quadriceps tendon, and the patellar tendon. The menisci are fibrocartilaginous richard_terek@brown.edu.
structures located between the tibiofemoral articulations; they increase stability
N Engl J Med 2010;363(4):e5.
and distribute contact forces on the articular cartilage (Fig. 1 and 2). Copyright 2010 Massachusetts Medical Society.

Common Knee Injuries

All the bones and soft tissues of the knee are subject to injury. The most common
problems that are reported in a physicians office are related to soft-tissue inflam-
mation, injury, or arthritis. Injuries to soft tissue and injuries resulting from over-
use are often caused by reproducible mechanisms of physical trauma or forces.
Fractures are less common and include fractures of the tibial plateau, the femoral
condyles, or the patella.
Knowledge of the mechanisms of knee injury can be essential to making an
accurate diagnosis. An injury caused by valgus stress to the knee can result in
medial collateral ligament strain or rupture, and an injury caused by varus stress
Click here to access the video.
can result in lateral collateral ligament strain or rupture. Abrupt noncontact decel-
eration or twisting and pivoting with simultaneous valgus stress to the knee can
cause rupture of the anterior cruciate ligament, whereas abrupt posterior translation
of the tibia can result in rupture of the posterior cruciate ligament. Twisting and
pivoting of the knee while it is bearing weight can cause a meniscal tear.
Overuse injuries are often manifested as pain in the structure subject to repetitive
stress. For example, repetitive jumping can lead to patellar tendonitis, also called
jumpers knee, whereas repeated application of direct pressure to the patella through
kneeling can cause prepatellar bursitis, sometimes referred to as housemaids knee.

History
Obtaining a complete history is the first step in determining the cause of knee pain.
The key elements include location and characterization of pain, mechanism of in-
jury, sound of a pop at the time of injury (which can indicate a ligamentous tear
or fracture), immediate or delayed swelling, recent infections, ability to bear weight,
locking sensation or instability (or incidents of subluxation), and prior injuries to
the joint.

n engl j med 363;4 nejm.org july 22, 2010 e5

BACK to TOC
132
 The new england journal of medicine

To assess nontraumatic causes of subacute or chronic knee pain (in addition to

questions about location and characterization of pain), the clinician should inquire
about fever, morning stiffness, pain after exercise, tick bites (to assess risk of Lyme
disease), infections (including sexually transmitted diseases, such as gonorrhea), his-
tory of gout or psoriasis, and activities contributing to long-term overuse.

General Examination of the Knee

Gait is an important element of the physical examination of the knee. The clinician
should always evaluate the patients gait and weight-bearing abilities, since the
findings can help distinguish knee pathology from pain referred from the hip,
lower back, or foot.
When clinically appropriate, a useful part of the clinical evaluation of knee pain
is observing the patient execute a duckwalk, which requires the patient to squat
and attempt to walk in that position. The duckwalk requires an intact ligamentous
system and a knee free of significant meniscal pathology, effusions, and bony
abnormalities, such as arthritis. Thus, if a patient can perform a duckwalk, he or
she probably has no ligamentous instability, large effusions, or meniscal tears.
Figure 1. Anatomy of the Right Knee,
The alignment of the knees should also be evaluated, with the patient standing
Anterior View.
with feet together. Look for a varus (bow-legged) or valgus (knock-kneed) defor-
mity. Such deformities can predispose the patient to osteoarthritis or indicate the
presence of significant osteoarthritis.2
Quadriceps atrophy can indicate disuse after an injury. If there is any suspicion
of quadriceps atrophy the circumference of each thigh should be measured to
confirm or rule out any changes to the underlying musculature.
It is also important to carefully assess the skin around the knee. Abnormalities
such as hematoma, rash (e.g., psoriasis), abrasions or lacerations, and cellulitis
provide important diagnostic clues.

Palpation of the Knee

Figure 2. Anatomy of the Right Knee,
Posterior View.
Always examine both knees, beginning with the unaffected knee. This approach
will reassure the patient and limit any apprehension about the examination, and it
will also establish a baseline against which the affected knee can be compared.
Using the back of the hand, assess for warmth as an indicator of inflammation.
Next, with the knee at a 90-degree angle, palpate the knee, beginning with the
anterior structures. Start by placing your thumbs on the tibial tuberosity and move
superiorly. As you move superiorly, palpate the patellar tendon and its insertion at
the inferior pole of the patella; pain in this area, especially in an athlete, might
indicate patellar tendonitis.
In a patient with a direct trauma to the knee, carefully palpate for areas of
tenderness that might indicate a fracture. In such patients, five specific findings
or factors should prompt consideration of radiographic imaging to rule out a trau-
matic fracture: an age of 55 years or older, tenderness at the head of the fibula,
isolated patellar tenderness, inability to flex the knee to 90 degrees, or inability to
bear weight and complete at least four steps. These five factors constitute the
Ottawa knee rules, a validated decision-making tool with a sensitivity of 100%.3,4
Radiography should also be considered if an injury to the anterior cruciate liga-
ment is suspected; such an injury can be associated with avulsion fractures of the
lateral tibial plateau.
In a patient with focal tenderness, erythema, and warmth and swelling anterior
to the patella, acute prepatellar bursitis should be considered. Patients with this
condition, which can be septic and may require aspiration or drainage, typically
have a history of recurrent kneeling or of direct trauma.

n engl j med 363;4 nejm.org july 22, 2010

BACK to TOC
133
 Clinical Evaluation of the Knee

Pain, swelling, and a palpable defect at the insertion of the quadriceps tendon
into the superior aspect of the patella suggests rupture of a quadriceps tendon.
This injury may be accompanied by a pop when it occurs, followed by dimin-
ished or complete absence of extensor strength.
The clinician should identify the inferior pole of the patella and move medially
to examine the medial joint line. Pain along the medial joint line might represent
medial compartment osteoarthritis, injury to the medial collateral ligament, or a
medial meniscal tear.
Tenderness at the midpoint between the anterior aspect of the medial joint line
and the tibial tuberosity may indicate pes anserine bursitis (located at the insertion
of the hamstring tendons into the tibia). Pes anserine bursitis is often found in
runners with tight hamstrings and in patients with valgus deformity; in the latter
case, it is often associated with osteoarthritis of the knee.
The clinician should also examine the lateral joint line for tenderness, which
can be caused by lateral compartment osteoarthritis, injury to the lateral collateral
ligament, or a lateral meniscal tear. Focal pain at the lateral femoral condyle is
suggestive of iliotibial band syndrome.
Palpation of the popliteal fossa can reveal a tender, fluid-filled mass called a
Bakers cyst. This results from a posterior extension of knee-joint effusions and
often accompanies osteoarthritis.

Assessment of Effusion
The absence of the normal indentations on the peripatellar grooves on either side
of the patella may indicate the presence of a large intraarticular effusion. Two ma-
neuvers can help confirm the presence of an intraarticular effusion.
In the first maneuver, with the knee extended, use the nondominant hand to
squeeze the intraarticular fluid from the suprapatellar region into the space be-
tween the patella and femur. With the dominant hand, exert pressure superiorly
from the tibia while using your index finger to push the patella against the patel-
lofemoral groove. When an effusion is present, you can easily ballot the patella.
The second maneuver used to assess for an effusion should also be performed
with the knee in extension. Gently milk the fluid into the suprapatellar pouch by
moving your hand proximally along the medial aspect of the patella. Next milk or
compress the fluid from the suprapatellar pouch to the medial knee by moving
your hand from the superior lateral region to the inferior lateral region; if there is
an effusion, compressing the lateral regions will cause a bulge to appear medial
to the patella in the areas that are naturally concave.
Comparison with the unaffected knee is essential. If an effusion is present,
arthrocentesis may be necessary for diagnostic or therapeutic reasons.5

Range of Motion
Both active and passive range of motion of the knee should be tested. The normal
range of extension is 0 to 10 degrees, and the normal range of flexion is 130 to
150 degrees. The location and movement of the patella should be noted: watch for
any signs of abnormal tracking, crepitus, or pain. If retropatellar pain and crepitus
occur while the patella is being compressed against the trochlea during active ex-
tension, patellofemoral syndrome or patellofemoral arthritis (chondromalacia)
should be considered. Pain with active range of motion but painfree passive range
of motion suggest a soft-tissue disorder such as tendonitis. Pain that is equal on
both passive and active range of motion is more likely to suggest an intraarticular
process.

n engl j med 363;4 nejm.org july 22, 2010

BACK to TOC
134
 The new england journal of medicine

Assessment of the Medial Collateral and Lateral Collateral Ligaments

Injury to the medial or lateral collateral ligaments typically involves direct trauma
to the contralateral side of the knee for example, a direct blow to the lateral side
results in valgus stress and injury to the medial collateral ligament. To assess the
medial collateral ligament, apply valgus stress to the knee. With the knee flexed to
25 degrees, place one hand on the outer aspect of the knee to apply medial pressure,
and the other hand on the inner aspect of the distal tibia to apply lateral pressure;
you are testing for tenderness or laxity along the medial collateral ligament.6,7 Simi-
larly, the lateral collateral ligament can be tested by applying lateral pressure to the
inner knee and medial pressure to the outer ankle or lower leg, which causes varus
stress to the knee.
When assessing the medial or lateral collateral ligaments, tenderness along the
ligament, but less than 5 mm of laxity and a solid end point, indicates a first-
degree sprain. In a second-degree sprain, a solid end point is maintained but there
is increased laxity when the knee is tested at 25 degrees of flexion and no laxity in
full extension. In a third-degree sprain or a complete tear of the ligament, there will
be a soft end point and more than 10 mm of laxity when the knee is at 25 degrees
of flexion; if there is also laxity with full extension, there may be additional dam-
age to the cruciate ligament.6

Assessment of the Anterial Cruciate Ligament

The anterior drawer test and the Lachman test provide information about the integ-
rity of the anterior cruciate ligament.
In the anterior drawer test, the patient should be supine, with the knee flexed
to 90 degrees and the foot placed flat on the table. Stabilize the foot (you can sit on
the end of the patients foot) and place your thumbs on the tibial plateau and your
fingers around the calf, relaxing the hamstrings; pull forward to test the anterior
cruciate ligament. If the ligament is intact, it should abruptly stop the anterior mo-
tion of the tibia with a solid end point. The affected and unaffected legs should
have similar degrees of anterior translation.
The Lachman test is a more sensitive and specific test for assessment of the
anterior cruciate ligament.7,8 In this maneuver, the patient is supine and asked to
relax the hamstrings. Use one hand to stabilize the femur while placing the knee
at 20 degrees of flexion. With the other hand, grasp the proximal tibia and briskly
pull the tibia forward. If there is more than 6 to 8 mm of laxity, more laxity than
in the unaffected knee, or a soft end point, the ligament may be torn.9 If you are
unable to firmly grasp and stabilize the femur, you can modify the Lachman maneu-
ver by placing your knee under the patients knee, firmly pressing down on the distal
femur with one hand, and pulling the tibia anteriorly with your other hand.
A large hemorrhagic effusion of rapid onset frequently accompanies anterior
cruciate ligament tears and bony fractures and contributes to the patients discom-
fort. Arthrocentesis can be of both diagnostic and therapeutic benefit and can also
facilitate a more accurate examination.

Assessment of the Posterior Cruciate Ligament

To test the integrity of the posterior cruciate ligament, perform the posterior drawer
test and assess for evidence of a tibial sag.
As in the anterior drawer test, the patient should be supine and the knee flexed
to 90 degrees. The foot should be flat on the table. Stabilize the foot (you can sit
on the end of the patients foot) and place your thumbs on the tibial tubercle and
your fingers around the calf, then briskly push the tibia posteriorly to test the
posterior cruciate ligament. If the ligament is intact, there should be a solid end
point and little posterior translation of the tibia.

n engl j med 363;4 nejm.org july 22, 2010

BACK to TOC
135
 Clinical Evaluation of the Knee

The tibial sag test is another test of the integrity of this ligament.6 Have the References
1. Cherry DK, Woodwell DA, Rechstein-
patient flex both knees at 90 degrees and place both feet flat on the table; then er EA. National Ambulatory Medical Care
observe the alignment of the tibial plateau. Normally, the tibial plateau extends 1 Survey: 2005 summary. Advance data
cm beyond the femoral condyle. If the affected tibia is displaced posteriorly on the from Vital and Health Statistics. No. 387.
(Accessed June 28, 2010, at http://www
femur, or sags, as compared with the unaffected tibial plateau, the posterior cruci- .cdc.gov/nchs/data/ad/ad387.pdf.)
ate ligament may be ruptured. 2. Sharma L, Song J, Felson DT, Cahue S,
Shamiyeh E, Dunlop DD. The role of knee
alignment in disease progression and
Assessment of the Meniscus
functional decline in knee osteoarthritis.
Patients with meniscal injuries may report clicking, catching, or locking of the JAMA 2001;286:188-95. [Erratum, JAMA
knee. In addition, they frequently have an effusion of delayed onset, appearing 2001;286:792.]
3. Stiell IG, Greenberg GH, Wells GA, et
hours or even days after the injury. al. Derivation of a decision rule for the
In addition to assessment of the knee for joint-line tenderness, there are two use of radiography in acute knee injuries.
common maneuvers that can be used to assess for possible meniscal tears. In the Ann Emerg Med 1995;26:405-13.
4. Stiell IG, Greenberg GH, Wells GA,
McMurray test, the patient is supine. To test the medial meniscus, place one hand et al. Prospective validation of a decision
over the anterior aspect of the knee, with fingers and thumb on the medial and rule for the use of radiography in acute
lateral joint lines. Grasp the patients heel with the other hand and externally ro- knee injuries. JAMA 1996;275:611-5.
5. Thomsen TW, Shen S, Shaffer RW,
tate the tibia, using the first hand to apply valgus force at the knee during passive Setnik GS. Arthrocentesis of the knee.
flexion and extension. The maneuver is repeated when applying internal rotation N Engl J Med 2006;354(19):e19. (Available
and varus stress to test the lateral meniscus. Clicking, catching, or popping at the at NEJM.org.)
6. Malanga GA, Andrus S, Nadler SF,
joint line during early extension or midextension may indicate a meniscal tear.10 McLean J. Physical examination of the
In the Apley compression test, or grind test, the patient lies prone and the knee knee: a review of the original test descrip-
is flexed to 90 degrees. Stabilize the thigh by placing your knee or hand firmly on tion and scientific validity of common
orthopedic tests. Arch Phys Med Rehabil
top of the patients posterior thigh. Grasp the foot and apply a downward com- 2003;84:592-603.
pressive force while rotating the tibia internally and externally. Pain on compres- 7. Solomon DH, Simel DL, Bates DW,
sion is considered positive for a meniscal tear.7,11 Katz JN, Schaffer JL. Does this patient
have a torn meniscus or ligament of the
Suspicion of a meniscal tear should prompt a careful assessment of the ante- knee? JAMA 2001;286:1610-20.
rior cruciate ligament; likewise, suspicion of a torn anterior cruciate ligament 8. Jackson JL, OMalley PG, Kroenke K.
should prompt a careful assessment of the meniscus. Injuries to these two struc- Evaluation of acute knee pain in primary
care. Ann Intern Med 2003;139:575-88.
tures often occur together. 9. Torg JS, Conrad W, Kalen V. Clinical
diagnosis of anterior cruciate ligament
Summary instability in the athlete. Am J Sports Med
1976;4:84-93.
In summary, a thorough history and physical examination are the first steps in 10. McMurray TP. The semilunar carti-
evaluating knee pain and making an accurate diagnosis, after which decisions lages. Br J Surg 1942;29:407-14.
about imaging studies (radiography, magnetic resonance imaging, and ultrasonog- 11. Apley AG. The diagnosis of meniscus
injuries: some new clinical methods.
raphy), treatment, and referral to specialists can be made. J Bone Joint Surg Am 1947;29:78-84.
Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. Copyright 2010 Massachusetts Medical Society.

n engl j med 363;4 nejm.org july 22, 2010

BACK to TOC
136
 HOME ARTICLES ISSUES SPECIALTIES & TOPICS FOR AUTHORS Keyword, Title, Author, or Citation

INTERACTIVE MEDICAL CASE The n e w e ng l a n d j o u r na l of m e dic i n e

Graham T. McMahon, M.D., M.M.Sc., Editor, Joel T. Katz, M.D., Associate Editor, Bruce D. Levy, M.D., Associate Editor, TOOLS
Joseph Loscalzo, M.D., Ph.D., Associate Editor
interactive medical case Downlo
Lying Low CME

John J. Ross, M.D., Anand Vaidya, M.D., and Ursula B. Kaiser, M.D.
N Engl J Med 2011; 364:e10 February 10, 2011
Lying Low
Case John J. Ross, M.D., Anand Vaidya, M.D., and Ursula B. Kaiser, M.D.

TOPICS

Hematolo
An 88-year-old woman presented to the
Oncology
emergency room with confusion. She
Medical E
began having transient episodes of Training, a
confusion, dizziness, tremors, and anxiety Primary C
a year earlier. These episodes were Gastroent
unpredictable, lasting for minutes and then General
abating spontaneously, and had been Endocrino
increasing in frequency since they began. General
Neurology
The patient felt well between episodes and
Neurosurg
reported no abnormal sensation . . .
General
Geriatrics
Learn more about Interactive Medical Cases
General
Surgery G
Psychiatr
Emergenc
Click here to access the Interactive Medical Case.
General

To save your progress and resume later, Sign-in or Play as a guest. Please note that your work will not
Create a FREE account. Your work is saved be saved. Play now.
automatically upon sign-in. TRENDS:

Hand H
About Interactive Medical Cases Feedback | Take Our Survey March 31,

Each interactive case presents an evolving patient history and a series of questions and exercises designed to test Giant S
your diagnostic and therapeutic skills. You will receive immediate feedback on your answers and treatment choices, March 31,
along with the opportunity to compare your final score with those of your peers. Video, animation, and interactive
content allow you to learn more about mechanisms, diagnostic tests, and treatments. Bocepre
Genotyp
March 31,
Signing in at the beginning of the case allows you to save your progress and resume later. Work by users
who are not signed in will not be saved.
A link at the end of the case leads to a CME examination; you can claim up to 2 AMA PRA Category 1
Credits for participating.
Please share your opinions about the case by taking our survey.

Pilot Phase: For a limited time, the Interactive Medical Cases will be in a pilot phase, during which this content and
service will be available free. Your feedback is welcome and will be used to improve the interface and functionality of
future Interactive Medical Cases.

See All Interactive Medical Cases

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

BACK to TOC
137
CONTENT: Home Current Issue Articles Issue Index Specialties & Topics Multimedia & Images Archive 1812-1989
INFORMATION FOR: Authors Reviewers Subscribers Institutions Media Advertisers