No title

The

NEW ENGLA ND JOURNAL

of

MEDICINE

Perspective march 24, 2011

Under Siege — The Individual Mandate for Health Insurance and Its Alternatives Jonathan Oberlander, Ph.D.

T

he battle over the Patient Protection and Affordable Care Act (ACA) rages on: in January, House Republicans passed legislation repealing the ACA, but the measure failed to clear the Democratic-

controlled Senate. Although Republicans’ best chance to overturn the law won’t come until 2013 — and then only if they win a majority in both Congressional houses and the presidency in 2012 — they may meanwhile pursue targeted repeal of controversial provisions. No provision is currently more beleaguered than the individual mandate to obtain health insurance or pay a penalty. Many analysts view this mandate as crucial to ensuring that healthier people join state-based insurance exchanges: since the law prohibits insurers from charging higher premiums to or turning away people with preexisting conditions, exchanges would other-

wise attract disproportionately sicker, costlier enrollees. That adverse selection would drive up premium costs and threaten exchanges’ stability. Reformers had hoped the mandate would also confer political advantages. Some Republicans, including former Massachusetts governor Mitt Romney, previously supported the policy. In a 2006 opinion piece, Romney defended Massachusetts’ decision to impose penalties on people who didn’t purchase insurance as a “personal responsibility principle. Some of my libertarian friends balk at what looks like an individual mandate,” he wrote. “But remember, someone has to pay for the health care that must, by

n engl j med 364;12

nejm.org

march 24, 2011

law, be provided: either the individual pays or the taxpayers pay. A free ride on government is not libertarian.”1 The Massachusetts mandate has in fact encouraged healthy people to obtain coverage.2 But Massachusetts proved to be only an oasis of bipartisanship. Among Republicans in Washington, pro-mandate arguments about personal responsibility gave way to concerns over individual liberty and the political priority of handing the Obama administration a defeat. The mandate now confronts a legal and political backlash. Florida’s Roger Vinson recently became the second federal judge to deem it unconstitutional, and the issue appears headed for the Supreme Court. Some state legislatures are seeking to block the mandate’s implementation. A few Democratic senators, including Claire McCaskill (MO) and Ben Nelson (NE), 1085

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

PERS PE C T IV E

who are up for reelection in 2012, say they’d like to find alternatives to the mandate. And while conservatives rail against the mandate and politically vulnerable moderates run away from it, many liberals support it reluctantly because of concerns that insurance remains unaffordable and that the requirement is a gift to the insurance industry. Increasingly, Democrats may wonder whether the provision is an albatross that should be jettisoned to save reform. Inasmuch as the health care reform debate is defined by the mandate, Democrats have a problem. The policy is highly unpopular — 76% of Americans view it unfavorably3 — and makes reform seem punitive. Although Democrats can highlight other consumer-friendly provisions that target health insurers, in the mandate fight they are allied with the insurance industry. Furthermore, President Barack Obama could in 2012 be in the uncomfortable position of defending a mandate that he argued against in the 2008 Democratic primary. (The mandate’s prominence also complicates Romney’s potential bid for the Republican presidential nomination.) Still, it’s not clear how much Democrats would gain politically by dumping the mandate. Many Republicans also oppose the ACA’s requirements that employers offer coverage or pay a penalty, the Medicaid expansion, increased government regulatory authority over the insurance industry, reductions in projected Medicare savings, and the cost of the subsidies for the uninsured to purchase insurance. Opponents of health care reform are attacking the mandate because it’s the most politically vulnerable part of the 1086

The Individual Mandate and Its Alternatives

law. Were it to fall, they would not endorse the ACA, but move on to attack other controversial provisions. If the mandate cannot be sustained, alternatives exist. The country could adopt single-payer, tax-financed national health insurance as a universal entitlement. There would be no need for penalties and no worries about adverse selection, since the government would operate one risk pool. Enrollment would be much easier to administer and universal coverage would be ensured. Yet single payer would require compulsory participation, new taxes, and the political transformation necessary to displace private insurance. It remains infeasible. Staying within the reform’s existing boundaries, one alternative is to limit enrollment in the exchanges to a fixed period each year and impose premium penalties for eligible people who choose to wait and buy coverage later — and to make the penalty apply not just the first time they purchase insurance, but across their lifetimes.4 This model is employed by the Medicare Part D program for prescription-drug coverage, which the Bush administration and many Congressional Republicans supported. However, to induce healthy uninsured people to sign up, the late penalty might have to be substantial, in which case this arrangement would be operating similarly to the mandate. As the health care economist Len Nichols points out, if we don’t have the political will to impose a strong penalty in conjunction with an individual mandate, we probably wouldn’t have the will to impose one as part of a fixed enrollment system. Furthermore, it’s not clear how a lifetime late penalty would work

n engl j med 364;12

nejm.org

when people’s insurance coverage could shift over time among private plans in the exchange, private plans outside the exchange, Medicaid, employer-based coverage, and eventually, Medicare. Another alternative would be to automatically enroll people in health insurance plans, permitting them to opt out. Auto-enrollment would occur primarily at the workplace but could also happen at state offices such as the Division of Motor Vehicles. This model could have bipartisan appeal: a 2009 health care reform bill cosponsored by Republican Senators Richard Burr (NC) and Tom Coburn (OK) and Republican Congressmen Paul Ryan (WI) and Devin Nunes (CA) relied on autoenrollment.5 Auto-enrollment can be combined with a premium penalty for people who opt out but later decide to purchase coverage. Medicare Part B, which pays for physicians’ services, works this way. Once again, though, the lateenrollment penalty might have to be substantial for auto-enrollment to effectively induce healthier people to pay for insurance coverage. Inertia alone may not be a sufficiently strong force to get younger, healthier workers to stay insured, given high and rising insurance premiums. Auto-enrollment may also not work as well outside the workplace or for workers’ dependents.4 Substituting either of these alternatives for the mandate would, as economist Jonathan Gruber argues, attenuate the ACA’s benefits, resulting in an increased number of uninsured Americans and higher premiums in the nongroup insurance market (as healthier people decline coverage).4 Implementing the rest of the ACA without any substitute

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

PE R S PE C T IV E

policy would similarly reduce coverage gains and destabilize insurance pools. The substantive case for the mandate is still strong, even if its political and legal foundations are shaken. It’s also unclear whether past Republican support for alternative policies would be sustained or, as with the individual mandate, evaporate in the heat of the political spotlight. Insofar as Republicans continue to support alternatives, they will be interested in them as part of a broader conservative health care reform package, not a means of bolstering the ACA. Republicans’ good chances of winning control of the Senate in 2012 further reduce their incentives to cooperate now on an alternative plan. What reformers may need even more than a policy alternative to the mandate is an alter-

The Individual Mandate and Its Alternatives

native rationale, especially since any provision meant to ensure broad participation in insurance pools must include financial penalties. The mandate’s defenders could again invoke the rhetoric of personal responsibility. Or they could emphasize that the mandate makes possible the insurance reforms that guarantee the availability of coverage to sick people.3 Alternatively, reformers could appeal to the mandate’s communitarian foundations, arguing that there are some public programs — such as Social Security and Medicare — that produce invaluable social benefits and that succeed because everyone participates in them. Ultimately, the furor over the mandate underscores the reality that solidarity remains elusive in U.S. health policy.

Disclosure forms provided by the author are available with the full text of this article at NEJM.org. From the University of North Carolina, Chapel Hill. This article (10.1056/NEJMp1101240) was published on February 16, 2011, at NEJM.org. 1. Romney M. Health care for everyone? We’ve found a way. Wall Street Journal. April 11, 2006. 2. Chandra A, Gruber J, McKnight R. The importance of the individual mandate — evidence from Massachusetts. N Engl J Med 2011;364:293-5. 3. Kaiser Family Foundation/Harvard School of Public Health. The public’s health care agenda for the 112th Congress. January 2011. (http://www.kff.org/kaiserpolls/upload/ 8134-F.pdf.) 4. Gruber J. Health care reform without the individual mandate. Washington, DC: Center for American Progress, February 9, 2011. (http://www.americanprogress.org/issues/ 2011/02/gruber_mandate.html.) 5. Coburn T. Individual auto-enrollment: an alternative to an individual mandate. (http:// coburn.senate.gov/public/index.cfm?a=Files .Serve&File_id=e87f06bf-d429-4eac-8e7eade046b8b882.) Copyright © 2011 Massachusetts Medical Society.

Early Accelerated Approval for Highly Targeted Cancer Drugs Bruce A. Chabner, M.D.

T

he striking results of recent phase 1 trials of targeted cancer drugs have provoked serious discussion about shortening the road to drug approval. A typical cancer drug takes 7 years from entry into human trials to approval by the Food and Drug Administration (FDA), which requires proof of efficacy in “well-controlled clinical trials.” In these randomized phase 3 trials, a new treatment alone or added to a drug combination is compared with a “standard” drug or combination. The most convincing end point for such trials is improved survival, although the FDA has accepted surrogates such as tumor progression, response rate,

or rarely, symptomatic relief. These trials are invaluable for establishing the benefits of new drugs in instances where reasonable alternative therapies exist. But phase 3 trials are expensive and time-consuming, usually taking at least 2 to 3 years to reach survival end points. The news of a highly successful new compound in phase 1 or 2 rapidly reaches physicians and patients, creating demand for early access. For drugs aimed at diseases with limited effective treatments, delaying access during phase 2 and 3 trials creates difficult ethical issues for regulatory agencies and pharmaceutical sponsors and agonizing decisions

n engl j med 364;12

nejm.org

march 24, 2011

for physicians and patients. As Miller and Joffe recently emphasized, the concept of equipoise (uncertainty as to which treatment carries greater benefit), underlies the ethical basis for randomization.1 Strong early results in phase 1 or 2 may tip the balance and argue against delaying access by performing phase 3 trials, but Miller and Joffe contend that such trials may nonetheless be justified by the need to establish new drugs’ long-term efficacy. New understanding of the molecular and genetic lesions that cause cancer has sharpened the discussion, however, by enabling rapid development of drugs that 1087

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

PERS PE C T IV E

Accelerated Approval for Targeted Cancer Drugs

Mutations in NRAS

Activation of PDGFR-β and IGF-1 receptor

Dimerization of BRAF and CRAF PLX4032 inhibits BRAF

Increased expression of the P13K–AKT signaling pathway

BRAF V600E — protein from mutated BRAF

Increase in CRAF

MEK Increased expression of COT ERK Mechanism of Action of PLX4032 in Melanoma. BRAF dimerizes with its partner CRAF and acts as a relay point for signals for normal cell growth and survival. The BRAF V600E mutation, found in 60% of melanomas, signals constitutively and drives the malignant behavior of the tumor. PLX4032, an inhibitor of BRAF, kills BRAF-dependent tumors and has proved to be highly effective in phase 1 trials.1 Further mutations in BRAF or activating mutations in NRAS, changes in expression of mitogen-activated protein kinase kinase kinase 8 (MAP3K8, or COT) or elements of the PI3 kinase pathway, or activation of alternative signaling pathways such as platelet-derived growth factor receptor β (PDGFR-β) have led to resistance to BRAF inhibition in experimental settings (labels in red) and in clinical settings (labels in green) and form the basis for new therapeutic strategies. IGF-1 denotes insulin-like growth factor 1. The image of CRAF was produced with the use of the Chimera software and is based on the 3omv crystal in the Protein Data Bank (www.wwpdb.org).

specifically target tumors bearing aberrant signaling pathways.2 Inhibitors of the BCR-ABL kinase in chronic myelogenous leukemia (CML) and inhibitors of the epidermal growth factor receptor (EGFR) in EGFR-mutated non– small-cell lung cancer (NSCLC)3 have replaced cytotoxic chemotherapy for these indications. Two new drugs have produced excellent results in phase 1 trials against cancers that responded poorly to standard treatments: PLX4032 yielded an 81% response rate in 38 patients with BRAF mutated melanoma (see diagram),4 and crizotinib had a 57% response rate in 82 patients with the EML4-ALK fusion in NSCLC.5 Rates of disease control (response 1088

or stable disease for at least 8 weeks) exceeded 90% in both trials, with minimal toxicity. Because only a fraction of patients with NSCLC and melanoma have tumors harboring the aberrant pathways, a biomarker test for the mutations was required to select appropriate trial subjects. Drugs such as PLX4032 and crizotinib normally undergo a second phase 2 trial (involving 30 to 100 patients) to confirm their activity as second- or third-line therapy. Because of their early promise, both of these drugs entered phase 3 trials directly after phase 1. Patients with newly diagnosed BRAF-mutation–positive melanoma were randomly assigned to PLX4032 or dacarba-

n engl j med 364;12

nejm.org

zine, a standard agent with a response rate of 15%; crossover from the control group to the experimental group at the time of disease progression was not allowed, to preserve the integrity of the survival end point. (The trial has now completed enrollment and reached its survival end point and is being amended to allow crossover.) In its phase 3 trial for newly diagnosed NSCLC, crizotinib is being compared with standard chemotherapy, which has a 30 to 40% response rate, a limited survival benefit, and substantial toxicity. Patients in the control group who have disease progression may receive crizotinib by entering a phase 2 trial. Because we can now define

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

PE R S PE C T IV E

patient subgroups with high response rates in phase 1 trials, performance of phase 3 trials for these drugs raises important issues. If patients with incurable disease who have the right biomarker for response are informed of these impressive early results, they will want and perhaps deserve access to the new drug and may not accept random assignment to a modestly effective and toxic standard agent. The phase 3 trial may lack equipoise in the eyes of both physicians and patients. There are alternatives for expediting access to new drugs for patients who are not eligible for or accepted into phase 3 trials. A phase 2 trial may be opened simultaneously for patients with progressive disease after initial therapy. Such trials are open for both PLX4032 and crizotinib, but they accrue limited numbers of patients and are conducted in only a few centers. In addition, compassionate-use protocols allow sponsors to distribute experimental drugs on a case-by-case basis. This mechanism was widely used to provide early access to experimental AIDS drugs, with few reports of unexpected adverse drug reactions. But sponsors often view compassionate use as competing with trial accrual and relieving pressure on the FDA for marketing approval. Neither crizotinib nor PLX4032 was available through an expanded-access protocol when its phase 3 trial opened, though such protocols are now available in a limited number of centers. A mechanism for early FDA approval, “accelerated approval,” was introduced in 1992. Reserved for serious or life-threatening illness not effectively treated by approved medications, this mechanism allows drugs to be registered

Accelerated Approval for Targeted Cancer Drugs

on the basis of surrogate end points in phase 2 — most commonly, a clearly demonstrable tumor response rate (often 20 to 30%). Accelerated approval is awarded with the stipulation that definitive trials, with a survival end point, must be conducted after approval. Of the 23 oncologic drugs given accelerated approval between 1993 and 2008, two were ultimately withdrawn from the U.S. market — gemtuzumab because of toxicity and gefitinib because of lack of ef ficacy. A 2008 study of accelerated approvals, from the Government Accountability Office, concluded that the FDA hasn’t effectively enforced requirements for postapproval trials. However, my review of accelerated approvals reveals that most such agents are integral to standard cancer treatment. Indeed, the more cogent question is whether accelerated approval can take place after phase 1. Given trialists’ ability to define patient subgroups with responsive tumors in phase 1 trials, I propose that for diseases lacking therapies that meaningfully extend survival, the FDA should set flexible standards permitting accelerated approval of new drugs after phase 1. These standards could be satisfied by the results of expanded and targeted phase 1 drug testing, as in the case of PLX4032. This strategy requires the early evaluation and validation of a companion biomarker for patient selection. Conceivably, non-targeted agents could also satisfy criteria for approval after phase 1. High response rates (>50%), high disease-control rates (>75%), and an acceptable toxicity profile in a biomarker-defined population of 75 to 100 subjects should be sufficient for acceler-

n engl j med 364;12

nejm.org

march 24, 2011

ated approval if there’s a clear unmet need. Randomized comparisons with minimally effective treatments or placebo should not be required. Specific end points for early approval should be maximally flexible and adjusted according to the targeted disease and the effectiveness and toxicity of alternative therapies. More extensive study would be required for less effective or more toxic experimental agents. The post–phase-1 approval of new agents meeting these goals is highly unlikely to have significant negative consequences such as ineffective treatment or unforeseen, overwhelming toxic effects. Early approval would allow rapid general access to treatment, while further evaluation focused on defining optimal doses, schedules, and drug combinations; long-term benefits; toxic effects; and resistance mechanisms. When striking clinical results have been demonstrated in a sizable, readily identifiable patient population in phase 1, the journey to drug approval should not be prolonged. Disclosure forms provided by the author are available with the full text of this article at NEJM.org. From the Massachusetts General Hospital Cancer Center, Boston. 1. Miller FG, Joffe S. Equipose and the dilemma of randomized clinical trials. N Engl J Med 2011;364:476-80. 2. McClellan M, Benner J, Schilsky R, et al. An accelerated pathway for targeted cancer therapies. Nat Rev Drug Discov 2011;10:7980. 3. Sequist LV, Martins RG, Spigel D, et al. First-line gefitinib in patients with advanced non-small-cell lung cancer harboring somatic EGFR mutations. J Clin Oncol 2008;26: 2442-9. [Erratum, J Clin Oncol 2008;26:3472.] 4. Flaherty KT, Puzanov I, Kim KB, et al. Inhibition of mutated, activated BRAF in metastatic melanoma. N Engl J Med 2010;363:809-19. 5. Kwak EL, Bang Y-J, Camidge DR, et al. Anaplastic lymphoma kinase inhibition in non–small-cell lung cancer. N Engl J Med 2010;363:1693-703. [Erratum, N Engl J Med 2011;364:588.] Copyright © 2011 Massachusetts Medical Society.

1089

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

PERS PE C T IV E

Tapping the Unmet Potential of Health IT

Tapping the Unmet Potential of Health Information Technology Ann S. O’Malley, M.D., M.P.H.

H

ealth information technology (HIT) holds promise for facilitating vast improvements in care and, ultimately, in the health of Americans,1,2 but achieving that potential remains a daunting task. A recent article in the Los Angeles Times described the new phenomenon of hiring computer-savvy undergraduate “scribes” to take notes for physicians during patient encounters and enter the information into electronic health records (EHRs) — a practice that suggests how far we must go to develop EHRs that clinicians will embrace. Of course, the most highly trained professional in the room need not be the one to enter data into the computer, especially during an emergency, but the perceived need for scribes and providers’ experiences using EHRs3 raise important questions about both the efficiency of care processes and the usability of current EHRs. Although EHRs laudably provide immediate access to patient data and electronic messaging functions, clinicians have been frustrated by the difficulty of using them to support care delivery and coordination.3 Transforming EHRs into effective clinical tools rather than a means of capturing information primarily for documentation and billing purposes will require progress on multiple fronts. Clinical processes must evolve so as to improve care and be more responsive to patients’ needs, and HIT’s capabilities must evolve along with them. HIT has particular potential in such areas as coordination of care, workflow efficiency and use of teams, clinical decision support, and population health management — all 1090

areas offering glimpses of both the potential and the challenges associated with improved HIT use. Few providers today, for example, can truly coordinate care — integrating care, in consultation with patients and their relatives and caregivers, across all of a patient’s conditions, needs, clinicians, and health care settings.2-4 Outpatient practices and inpatient facilities lack well-developed processes for exchanging information, both within their own walls and during care transitions. Poor care coordination negatively affects patients — particularly those with multiple chronic conditions who account for an overwhelming proportion of U.S. health care expenditures. HIT, especially if widely implemented, can facilitate coordination by making information electronically available at the point of care. As clinical care processes become more effective and efficient, they can inform new HIT capabilities that will better support coordination. For example, providers need to develop consistent notification processes to ensure timely communication about care transitions. Medical and nursing professional societies could work with HIT vendors to develop standardized notification procedures, which could be implemented through refined criteria for the “meaningful use” of HIT.5 HIT can also better support care coordination through the development of referral-tracking systems, improved approaches to reconciling patients’ medications, and expansion of “problem list” capabilities — to avoid cluttering of lists with redundant information, for example, and permit n engl j med 364;12

nejm.org

sorting and searching of lists and linking of listed problems to relevant portions of progress notes, assessments, and treatment plans.3 In addition, improving care coordination and health outcomes requires teamwork. Taking into account each team member’s training, skill set, and expertise when delegating tasks and defining roles is critical to improving efficiency, for both primary care teams within a practice and interspecialty teams sharing patients’ care. HIT can support team-based care with tools enabling team members to identify patient care goals and document and monitor progress using a shared care plan. In outpatient practices, providers report that electronic messaging and notification of staff about patient care tasks facilitate communication about delegation and task completion. Electronic messaging can also enable real-time communication with specialists — if they’re on the same system — to determine whether a consultation is necessary.2,3 Coordinating care for patients with complex health conditions who see multiple physicians can also be supported by better HIT interoperability.3 The primary care team may be in the best position to coordinate a patient’s care, but it will often need information from other providers.4 Most current EHRs don’t adequately support data exchange across providers and settings, so practices communicate with outsiders primarily on paper.3 To support information exchange, EHRs must present data in standard ways, and separate organizations providing services for the same patient need to share information securely.

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

PE R S PE C T IV E

The Health Information Technology for Economic and Clinical Health (HITECH) Act envisions the secure exchange of data across providers and settings occurring through the Nationwide Health Information Network, which will provide a common platform and protocols.5 Some states and communities are developing regional health information exchanges, but most are still in their infancy. It is hoped that these local and national efforts might eventually be linked. Some have raised concerns, however, about the sustainability of these exchanges absent a stronger business case to support them. Clearly, HIT alone cannot transform our health care system: financial incentives must be realigned to reward patient-centered care. Current fee-for-service reimbursement encourages EHR use for documentation of billable events rather than for tasks important to the quality of care, such as coordination. Payment innovations such as bundled payments and accountable care organizations aim to encourage providers to share accountability for outcomes. Such payment reform would offer clinicians incentives to demand HIT capabilities that better support the clinical tasks required to improve patients’ health and would make these activities important to the success of health care organizations. Fee-for-service reimbursement also makes it difficult for clinicians to take the time to listen to patients’ concerns. Unfortunately, adding an EHR to the clinical encounter can further distract clinicians from patients (which is one of the reasons that some emergency departments use scribes). To counter this tendency,

Tapping the Unmet Potential of Health IT

payment reform could be accompanied by training for clinicians, residents, and medical students in effective communication with patients in the presence of an EHR. HIT can also provide tools to help inform decision making with regard to diagnosis (with clinical prediction rules), prevention (reminders), disease management (registries), and treatment (electronic prescribing tools).2 The use of computerized medication orders, generated with the help of decision-support tools, is associated with reductions in adverse drug events.2 But most current commercial EHRs don’t provide or link to decision-support systems, particularly for managing chronic care or selecting preference-sensitive treatments. To permit the development of adequate decision-support tools, the evidence base must be expanded and actively maintained. Finally, primary care practices increasingly must focus not just on individual patients but on whole populations, as they strive to function as medical homes.2 A new orientation and effective methods for shifting practices from reactive and acute-symptom care to approaches including proactive, planned care for both healthy and chronically ill populations will be needed. Without HIT, it’s difficult to provide effective population-based care and report quality metrics, but most commercial EHRs currently cannot help identify which patients in a population may need particular services. Registries are another critical tool for population health management and an area where HIT applications could be better developed and integrated with EHRs. Regardless of HIT’s potential advantages, clinicians in the coun-

n engl j med 364;12

nejm.org

march 24, 2011

try’s many small primary care practices can be overwhelmed by it and will need to be convinced that EHRs are affordable, enhance efficiency, and improve care. Then, they will need extensive, ongoing support. Under HITECH, HIT Regional Extension Centers will provide technical assistance, guidance, and information on EHR adoption and meaningful use, particularly to such practices.5 These centers could also compile clinicians’ feedback for policymakers and vendors, providing an infrastructure for shared learning. Unfortunately, the United States currently lacks an adequately trained workforce to support practices in these areas. Continued research on clinical care processes, the design and use of HIT, and payment reform, as well as ongoing support for clinicians, will be key to the effective and meaningful use of HIT. Today’s EHRs do not sufficiently support aspects of care delivery that are vital to improving care and controlling costs. Disclosure forms provided by the author are available with the full text of this article at NEJM.org. From the Center for Studying Health System Change, Washington, DC. 1. Goldzweig CL, Towfigh A, Maglione M, Shekelle PG. Costs and benefits of health information technology: new trends from the literature. Health Aff (Millwood) 2009;28: w282-w293. 2. Bates DW, Bitton A. The future of health information technology in the patient-centered medical home. Health Aff (Millwood) 2010;29:614-21. 3. O’Malley AS, Grossman JM, Cohen GR, Kemper NM, Pham HH. Are electronic medical records helpful for care coordination? Experiences of physician practices. J Gen Intern Med 2010;25:177-85. 4. Stille CJ, Jerant A, Bell D, Meltzer D, Elmore JG. Coordinating care across diseases, settings, and clinicians: a key role for the generalist in practice. Ann Intern Med 2005;142: 700-8. 5. Blumenthal D, Tavenner M. The “meaningful use” regulation for electronic health records. N Engl J Med 2010;363:501-4. Copyright © 2011 Massachusetts Medical Society.

1091

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

new england journal of medicine The

established in 1812

march 24, 2011

vol. 364

no. 12

Tiotropium versus Salmeterol for the Prevention of Exacerbations of COPD Claus Vogelmeier, M.D., Bettina Hederer, M.D., Thomas Glaab, M.D., Hendrik Schmidt, Ph.D., Maureen P.M.H. Rutten-van Mölken, Ph.D., Kai M. Beeh, M.D., Klaus F. Rabe, M.D., and Leonardo M. Fabbri, M.D., for the POET-COPD Investigators*

A BS T R AC T BACKGROUND

Treatment guidelines recommend the use of inhaled long-acting bronchodilators to alleviate symptoms and reduce the risk of exacerbations in patients with moderate-tovery-severe chronic obstructive pulmonary disease (COPD) but do not specify whether a long-acting anticholinergic drug or a β2-agonist is the preferred agent. We investigated whether the anticholinergic drug tiotropium is superior to the β2-agonist salmeterol in preventing exacerbations of COPD. METHODS

In a 1-year, randomized, double-blind, double-dummy, parallel-group trial, we compared the effect of treatment with 18 µg of tiotropium once daily with that of 50 µg of salmeterol twice daily on the incidence of moderate or severe exacerbations in patients with moderate-to-very-severe COPD and a history of exacerbations in the preceding year. RESULTS

A total of 7376 patients were randomly assigned to and treated with tiotropium (3707 patients) or salmeterol (3669 patients). Tiotropium, as compared with salmeterol, increased the time to the first exacerbation (187 days vs. 145 days), with a 17% reduction in risk (hazard ratio, 0.83; 95% confidence interval [CI], 0.77 to 0.90; P<0.001). Tiotropium also increased the time to the first severe exacerbation (hazard ratio, 0.72; 95% CI, 0.61 to 0.85; P<0.001), reduced the annual number of moderate or severe exacerbations (0.64 vs. 0.72; rate ratio, 0.89; 95% CI, 0.83 to 0.96; P = 0.002), and reduced the annual number of severe exacerbations (0.09 vs. 0.13; rate ratio, 0.73; 95% CI, 0.66 to 0.82; P<0.001). Overall, the incidence of serious adverse events and of adverse events leading to the discontinuation of treatment was similar in the two study groups. There were 64 deaths (1.7%) in the tiotropium group and 78 (2.1%) in the salmeterol group.

From the Hospital of the Universities of Giessen and Marburg, Marburg (C.V.); Boehringer Ingelheim, Ingelheim (B.H., T.G., H.S.); and insaf Respiratory Research Institute, Wiesbaden (K.M.B.) — all in Germany; the Institute for Medical Technology Assessment (IMTA), Erasmus University, Rotterdam (M.P.M.H.R.-M.); and Leiden University Medical Center, Leiden (K.F.R.) — both in the Netherlands; and the University of Modena and Reggio Emilia, Modena, Italy (L.M.F.). Address reprint requests to Dr. Fabbri at the Section of Respiratory Diseases, Department of Oncology, Hematology, and Pulmonary Diseases, University of Modena and Reggio Emilia, Policlinico di Modena, Largo del Pozzo 71, I-41124 Modena, Italy, or at [email protected]. *The investigators in the Prevention of Exacerbations with Tiotropium in COPD (POET-COPD) trial are listed in the Supplementary Appendix, available at NEJM.org. N Engl J Med 2011;364:1093-103. Copyright © 2011 Massachusetts Medical Society.

CONCLUSIONS

These results show that, in patients with moderate-to-very-severe COPD, tiotropium is more effective than salmeterol in preventing exacerbations. (Funded by Boehringer Ingelheim and Pfizer; ClinicalTrials.gov number, NCT00563381.)

n engl j med 364;12

nejm.org

march 24, 2011

1093

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

C

hronic obstructive pulmonary disease (COPD) is a leading cause of disability and death worldwide.1-3 Exacerbations of COPD indicate instability or worsening of the patient’s clinical status and progression of the disease and have been associated with the development of complications, an increased risk of subsequent exacerbations, a worsening of coexisting conditions, reduced health status and physical activity, deterioration of lung function, and an increased risk of death.4-7 The prevention of exacerbations therefore constitutes a major goal of treatment.1,2 Therapy with a long-acting anticholinergic drug or a long-acting β2-agonist is recommended as first-line maintenance therapy in patients with moderate-to-very-severe COPD,1,2 since both of these drugs reduce symptoms, improve quality of life and lung function, and reduce the risk of exacerbations and hospitalizations.8-12 However, treatment guidelines do not specify whether a long-acting anticholinergic drug or a β2-agonist is the preferred agent.1,2 Comparative studies have indicated that tiotropium is associated with a greater reduction in the risk of exacerbations and exacerbation-related hospitalizations than is salmeterol, although the differences were not significant.13,14 These were short-term studies (3 to 6 months in duration) and were not designed and powered to detect a difference in the risk of exacerbations. The Prevention of Exacerbations with Tiotropium in COPD (POET-COPD) trial was specifically designed to directly compare the effects of tiotropium with those of salmeterol on the risk of moderate and severe exacerbations. A placebo group was not included in the study, since there is substantial evidence of the superiority of both tiotropium and salmeterol over placebo.8,12 Furthermore, a comparison of two active-treatment groups is in line with the recently growing relevance of comparative-effectiveness research to guidance regarding treatment decisions.15,16

Me thods Study Design and Oversight

We conducted a 1-year, randomized, double-blind, double-dummy, parallel-group trial at 725 centers in 25 countries to compare the effect of tiotropium (Spiriva, Boehringer Ingelheim) with that of salmeterol (Serevent, GlaxoSmithKline) on moder1094

n engl j med 364;12

of

m e dic i n e

ate and severe exacerbations of COPD (hereinafter called exacerbations) in patients with moderateto-very-severe COPD.17 The study was conducted in accordance with the provisions of the Declaration of Helsinki (1996) and Good Clinical Practice guidelines. All patients provided written informed consent before any study procedure was performed. The scientific steering committee (which was made up of two of the academic investigators and an external clinical researcher) and three employees of Boehringer Ingelheim developed the design and concept of the study, approved the statistical plan, had full access to the data, and interpreted the data. Onsite monitoring and site management were supported by a contract research organization (PAREXEL). The first draft of the manuscript and subsequent revisions were written by all the authors, and all the authors made the decision to submit the manuscript for publication. The statistical analysis was performed by an employee of the sponsor. All the authors had full access to the data and vouch for the accuracy and completeness of the data and the analyses, as well as the fidelity of the study to the protocol. (The protocol, including the statistical analysis plan, is available with the full text of this article at NEJM.org.) An independent ethics committee or institutional review board at each participating center reviewed and approved the protocol before commencement of the study. In addition, an independent data and safety monitoring board and a mortality adjudication committee were established (Section 10 in the Supplementary Appendix, available at NEJM.org). End Points

The primary end point was the time to the first exacerbation of COPD. The time to the first exacerbation was selected as the primary end point because it is less likely to be affected by the introduction of additional therapies or by the occurrence of multiple exacerbations in some patients.17 Secondary and safety end points included time-to-event end points, number-of-event end points, serious adverse events, and death (Section 2 in the Supplementary Appendix). An exacerbation was defined as an increase in or new onset of more than one symptom of COPD (cough, sputum, wheezing, dyspnea, or chest tightness), with at least one symptom lasting 3 days or more and leading the patient’s attending physician to initiate treatment with

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Tiotropium vs. Salmeterol for COPD Exacerbations

systemic glucocorticoids, antibiotics, or both (criterion for moderate exacerbation) or to hospitalize the patient (criterion for severe exacerbation). The determination of the end of the exacerbation was made on the basis of the clinical assessment of the investigator. Data on exacerbations (according to the trial definition), as well as health care resources used to treat these exacerbations, were collected by means of a questionnaire that was administered during regular clinic visits and telephone contacts. When an investigator reported a case of pneumonia, he or she was questioned as to whether the event had been confirmed by imaging. Patients

Patients were eligible for inclusion in the study if they were at least 40 years of age and had a smoking history of 10 pack-years or more, a diagnosis of COPD, a forced expiratory volume in 1 second (FEV1) after bronchodilation of ≤70% of the predicted value,18 a ratio of FEV1 to forced vital capacity (FVC) of ≤70%, and a documented history of at least one exacerbation leading to treatment with systemic glucocorticoids or antibiotics or hospitalization within the previous year. Spirometry (FEV1 and FVC) was performed at the screening visit according to the guidelines of the American Thoracic Society19 and was used only for the assessment of the severity of COPD. Postbronchodilator measurements were performed 30 minutes after the patient inhaled 400 µg of albuterol. Daily peak flow was recorded over the course of 4 months in a subgroup of patients, in conjunction with a genotyping analysis (for details, see Section 5 in the Supplementary Appendix); those data are not reported here. Full details regarding the exclusion criteria are provided in Section 6 in the Supplementary Appendix. Procedures

given instruction in the use of the HandiHaler and pressurized, metered-dose inhaler devices at visits 1 (screening) and 2 (randomization). Concomitant medication at baseline was defined as the therapy the patients were receiving at the time of the screening visit (visit 1). During the run-in period, patients who were receiving tiotropium were required to switch to 40 µg of ipratropium four times a day, and this therapy was discontinued at the time of randomization. Patients who were receiving a long-acting β2-agonist were permitted to continue the use of that medication during the run-in period. Patients receiving fixeddose combinations of long-acting β2-agonists and inhaled glucocorticoids were instructed to switch to inhaled glucocorticoid monotherapy at the start of the treatment phase of the study. Patients were allowed to continue their usual medications for COPD, except for anticholinergic drugs and longacting β2-agonists, during the double-blind treatment phase. After randomization, clinic visits were scheduled at months 2, 4, 8, and 12, and monthly telephone calls were scheduled between visits. Patients completed a daily diary, and records were reviewed at each study visit to assess adherence to treatment and to determine whether respiratory symptoms met the criteria for exacerbation. Adherence was not systematically assessed during the trial. During clinic visits and monthly telephone calls, a questionnaire was administered to collect details regarding exacerbations of COPD. Adverse events leading to the discontinuation of treatment and serious adverse events including fatal events were recorded at the time of each clinic visit. Patients who prematurely discontinued treatment were followed for vital status (i.e., whether they were alive and, if they had died, the primary cause of death) until the end of the planned treatment period of 360 days. Information on vital status was considered to be complete for patients who attended all trial visits through day 360 and for those who prematurely discontinued study medication but whose vital status was confirmed at day 360. Details of the randomization procedures and of the procedures for concealing the treatment assignments are provided in Section 8 in the Supplementary Appendix.

After a 2-week run-in period, eligible patients were randomly assigned to receive, for 1 year, either 18 µg of tiotropium once daily, delivered through the HandiHaler inhalation device (Boehringer Ingelheim), plus placebo twice daily, delivered through a pressurized, metered-dose inhaler, or 50 µg of salmeterol twice daily, delivered through a pressurized, metered-dose inhaler, plus placebo once daily, delivered through the Handi- Statistical Analysis Haler device (for details, see Section 7 in the We estimated that with a sample size of approxiSupplementary Appendix). All the patients were mately 6800 patients (3400 in each treatment n engl j med 364;12

nejm.org

march 24, 2011

1095

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

group), the study would have 80% power to detect a 10% reduction with tiotropium as compared with salmeterol in the risk of a first exacerbation, with a two-sided test for the null hypothesis of a hazard ratio of 1 at a significance level of 0.05. A prespecified reestimation of the sample size (with the treatment assignments concealed) on the basis of the predicted event rate was performed toward the end of the originally planned recruitment phase and resulted in an increase of the sample size to a total of 7350 patients (Section 9 in the Supplementary Appendix). The efficacy and safety analyses included all the patients who underwent randomization and who received at least one dose of the study medication. Primary and secondary time-to-event end points were analyzed with the use of a Cox proportional-hazards regression model including terms for (pooled) center and treatment; pooling was performed to account for study centers that recruited fewer than four patients. P values were calculated with the use of the Wald chi-square statistic. Kaplan–Meier plots were constructed, and log-rank tests were also performed. Number-of-event end points were compared between study groups with the use of Poisson regression with correction for overdispersion and adjustment for treatment exposure. To allow for a clear distinction between events, individual episodes of exacerbations had to be separated by a gap of at least 7 days. In keeping with the design of the study, exacerbations were not systematically followed up after a patient’s premature discontinuation of the trial medication.17 Hence, in the efficacy analysis, only exacerbations with onset during the time a patient was receiving treatment were included.7,20 Patients who withdrew from the trial prematurely without having had an exacerbation were considered as having had no exacerbation, and in the time-to-event analysis, their data were censored at the time of withdrawal. In the analyses of secondary end points, no corrections for multiple testing have been made. Subgroup analyses were performed for timeto-event end points and for number-of-event end points with the use of the models described above, with additional terms for subgroups and for interactions of subgroups with the study treatment. A post hoc subgroup analysis was performed according to patients who received inhaled glucocorticoids on a consistent basis during 1096

n engl j med 364;12

of

m e dic i n e

the study treatment period versus patients who received no inhaled glucocorticoids during the treatment period. Incidence rates of serious adverse events were calculated as the number of patients with events divided by the time at risk. The rate of death from any cause was analyzed with the use of Cox regression, with treatment as a covariate. A Kaplan–Meier analysis was also performed.

R e sult s Patients

Patients were enrolled between January 2008 and April 2009. A total of 7384 patients underwent randomization, and 7376 patients (3707 in the tiotropium group and 3669 in the salmeterol group) received at least one dose of the study medication (Fig. 1). The baseline characteristics of the patients, including coexisting conditions, were balanced between the treatment groups (Table 1, and Section 11 in the Supplementary Appendix). Fewer patients in the tiotropium group than in the salmeterol group withdrew from the study prematurely: 585 patients (15.8%) vs. 648 patients (17.7%) (hazard ratio with tiotropium, 0.88; 95% confidence interval [CI], 0.78 to 0.98; P = 0.02). The Kaplan–Meier plot for the time to the discontinuation of treatment is shown in Figure 2A. The collection of vital status up to day 360 was complete for 99.1% of the patients. Exacerbations

There were 4411 individual episodes of exacerbation among 2691 patients; 44% of the patients with an exacerbation had moderate COPD at the trial onset (stage II COPD, according to the classification of the Global Initiative for Chronic Obstructive Lung Disease [GOLD],1 which specifies four stages of COPD ranging from stage I, indicating mild disease, to stage IV, indicating very severe disease). The time to the first exacerbation (the primary end point) was increased by 42 days with tiotropium as compared with salmeterol (187 days vs. 145 days, representing the time until at least 25% of the patients [first quartile] had a first exacerbation), corresponding to a 17% reduction in risk with tiotropium (hazard ratio, 0.83; 95% CI, 0.77 to 0.90; P<0.001). Figure 2B shows the Kaplan–Meier plot for the time to the first exacerbation. Given the fact that less than 50% of the patients had an exacerbation (2691 of

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Tiotropium vs. Salmeterol for COPD Exacerbations

8293 Patients were assessed for eligibility

909 Withdrew during screening or did not meet entry criteria

7384 Underwent randomization

3711 Were assigned to receive tiotropium

3673 Were assigned to receive salmeterol

4 Did not receive tiotropium

4 Did not receive salmeterol

3707 Were included in efficacy and safety analyses

3669 Were included in efficacy and safety analyses

648 Discontinued salmeterol 292 Had adverse event 24 Had lack of efficacy 209 Withdrew consent 74 Were nonadherent to protocol 15 Were lost to follow-up 34 Had other reason

585 Discontinued tiotropium 264 Had adverse event 32 Had lack of efficacy 192 Withdrew consent 66 Were nonadherent to protocol 7 Were lost to follow-up 24 Had other reason

3122 Completed study

3021 Completed study

Figure 1. Screening, Randomization, and Follow-up.

7376 patients [36.5%]), it was not possible to calculate the median time to the first exacerbation; therefore, the time to the first exacerbation in the first quartile of patients was calculated instead. Tiotropium as compared with salmeterol significantly reduced the risk of moderate exacerbations by 14% (hazard ratio, 0.86; 95% CI, 0.79 to 0.93; P<0.001) and of severe exacerbations by 28% (hazard ratio, 0.72; 95% CI, 0.61 to 0.85; P<0.001). The Kaplan–Meier plot for the time to a first severe exacerbation is shown in Figure 2C. In addition, tiotropium reduced the risk of exacerbations leading to treatment with systemic glucocorticoids by 23% (hazard ratio, 0.77; 95% CI, 0.69 to 0.85; P<0.001), exacerbations leading to treatment with antibiotics by 15% (hazard ratio, 0.85; 95% CI, 0.78 to 0.92; P<0.001), and exacerbations leading to treatment with both systemic

n engl j med 364;12

glucocorticoids and antibiotics by 24% (hazard ratio, 0.76; 95% CI, 0.68 to 0.86; P<0.001) (Section 3 in the Supplementary Appendix). The annual rate of exacerbations was 0.64 in the tiotropium group and 0.72 in the salmeterol group, corresponding to an 11% reduction in the rate of exacerbations with tiotropium (rate ratio, 0.89; 95% CI, 0.83 to 0.96; P = 0.002). Treatment with tiotropium significantly reduced the annual rate of moderate exacerbations by 7% (0.54 vs. 0.59; rate ratio, 0.93; 95% CI, 0.86 to 1.00; P = 0.048) and the annual rate of severe exacerbations by 27% (0.09 vs. 0.13; rate ratio, 0.73; 95% CI, 0.66 to 0.82; P<0.001) (Section 3 in the Supplementary Appendix). In addition, tiotropium reduced the rate of exacerbations leading to treatment with systemic glucocorticoids by 18% (0.33 vs. 0.41; rate ratio, 0.82; 95% CI, 0.76 to 0.90;

nejm.org

march 24, 2011

1097

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

Table 1. Baseline Characteristics of the Patients.* Characteristic Male sex (%) Age (yr)

Tiotropium (N = 3707)

Salmeterol (N = 3669)

74.4

74.9

62.9±9.0

62.8±9.0

Smoking status Current smoker (%) Smoking history (pack-yr) Duration of COPD (yr)†

48.0

48.3

38.8±20.0

37.8±19.2

8.0±6.7

7.9±6.5

47.8

49.6

of

m e dic i n e

Figure 2 (facing page). Kaplan–Meier Curves for the Primary and Selected Secondary Outcomes. Kaplan–Meier curves are shown for the probability of premature discontinuation of the study medication (Panel A), the probability of a first exacerbation of chronic obstructive pulmonary disease (COPD) (Panel B), and the probability of a first severe exacerbation of COPD leading to hospitalization (Panel C) in the tiotropium and salmeterol groups. The hazard ratios are based on a Cox proportional-hazards regression model including terms for (pooled) center and treatment. CI denotes confidence interval.

GOLD stage (%)‡ II

leading to treatment with both systemic glucocorticoids and antibiotics by 20% (0.23 vs. 0.28; rate ratio, 0.80; 95% CI, 0.73 to 0.88; P<0.001) Spirometry after bronchodilation§ (Section 3 in the Supplementary Appendix). FEV1 (liters) 1.41±0.47 1.41±0.45 The effects of tiotropium as compared with FEV1 (% of predicted value) 49.2±13.3 49.4±13.1 salmeterol on the time to a first exacerbation and FVC (liters) 2.71±0.81 2.75±0.82 the annual rate of exacerbations per patient were Ratio of FEV1 to FVC (%) 52.5±10.8 52.4±11.2 consistent across prespecified subgroups accordPulmonary medications (%) ing to age, sex, smoking status (current vs. noncurrent smoker), severity of COPD (GOLD stage), Any 90.0 89.9 body-mass index, and use or no use of inhaled Anticholinergic drug glucocorticoids at baseline (Fig. 3, and Section 4 Tiotropium 30.5 30.3 in the Supplementary Appendix). Patients with a Short-acting 29.3 29.6 low body-mass index or very severe COPD seemed β2-Agonists to benefit most from tiotropium therapy (Fig. 3). Long-acting¶ 51.5 51.5 However, the P values for the tests of an interacShort-acting 52.5 53.4 tion between treatment effect and subgroup were Glucocorticoids 0.17 for the subgroup according to body-mass index and 0.05 for the subgroup according to Inhaled¶ 53.6 53.3 GOLD stage. In a post hoc analysis, a similar reWith tiotropium 18.7 18.2 duction in the risk of an exacerbation with tiotroWith long-acting β2-agonists 43.3 43.5 pium as compared with salmeterol was observed Oral 2.4 2.3 among the 2932 patients who used concomitant Methylxanthines 23.0 21.2 inhaled glucocorticoids during the study-treatment period (hazard ratio, 0.91; 95% CI, 0.82 to * Plus–minus values are means ±SD. COPD denotes chronic obstructive pul1.02), as well as among the 4046 patients who did monary disease, FEV1 forced expiratory volume in 1 second, and FVC forced vital capacity. not use inhaled glucocorticoids at any time dur† Data on duration of COPD were missing for 15 patients in the tiotropium ing the study-treatment period (hazard ratio, group and 5 in the salmeterol group. ‡ The severity of COPD was defined according to the classification of the Global 0.81; 95% CI, 0.72 to 0.91). In a subgroup analysis Initiative for Chronic Obstructive Lung Disease (GOLD), which specifies four of patients who were receiving inhaled glucocorstages of COPD ranging from stage I, indicating mild disease, to stage IV, inditicoids at baseline but did not receive them during cating very severe disease. There were 23 patients with GOLD stage I COPD — the study-treatment period versus patients who 0.2% of the patients in the tiotropium group and 0.4% in the salmeterol group. § Pulmonary function testing was performed at the screening visit (visit 1). Data were receiving inhaled glucocorticoids at baseline on FVC were missing for 1 patient in the tiotropium group. and continued to receive them during the study¶ This medication was used either alone or in a fixed combination. treatment period, the annual exacerbation rate in the tiotropium group was 0.67 (95% CI, 0.57 to P<0.001), exacerbations leading to treatment with 0.79) among the 395 patients who discontinued antibiotics by 10% (0.53 vs. 0.59; rate ratio, 0.90; the use of inhaled glucocorticoids, as compared 95% CI, 0.84 to 0.97; P = 0.004), and exacerbations with 0.78 (95% CI, 0.73 to 0.85) among the 1452 III

43.1

42.1

IV

8.9

7.9

1098

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Tiotropium vs. Salmeterol for COPD Exacerbations

Probability of Premature Discontinuation of Trial Medication

A 1.0

0.20

0.9

0.15

0.8

Hazard ratio, 0.88 (95% CI, 0.78–0.98) P=0.02 by log-rank test

Salmeterol Tiotropium

0.10

0.7 0.6

0.05

0.5

0.00

0.4 0.3

0 30 60 90 120 150 180 210 240 270 300 330 360

0.2 0.1 0.0

0

30

60

90

120

150

180

210

240

270

300

330

360

Days No. at Risk Tiotropium Salmeterol

3707 3592 3501 3429 3382 3330 3299 3268 3225 3186 3158 3138 2841 3669 3541 3436 3337 3291 3209 3181 3151 3111 3074 3054 3037 2703

B Probability of COPD Exacerbation

1.0

0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00

0.9 0.8 0.7 0.6 0.5 0.4 0.3

Hazard ratio, 0.83 (95% CI, 0.77–0.90) P<0.001 by log-rank test Salmeterol Tiotropium

0 30 60 90 120 150 180 210 240 270 300 330 360

0.2 0.1 0.0

0

30

60

90

120

150

180

210

240

270

300

330

360

Days No. at Risk Tiotropium Salmeterol

3707 3369 3136 2955 2787 2647 2561 2455 2343 2242 2169 2107 1869 3669 3328 3028 2802 2605 2457 2351 2251 2137 2050 1982 1915 1657

C 1.0

0.15

Hazard ratio, 0.72 (95% CI, 0.61–0.85) P<0.001 by log-rank test

Probability of Severe COPD Exacerbation

0.9 0.10

0.8

Salmeterol

0.7

Tiotropium

0.05

0.6 0.5

0.00

0.4 0.3

0 30 60 90 120 150 180 210 240 270 300 330 360

0.2 0.1 0.0

0

30

60

90

120

150

180

210

240

270

300

330

360

Days No. at Risk Tiotropium Salmeterol

3707 3564 3453 3359 3285 3217 3177 3125 3066 3017 2977 2948 2663 3669 3502 3362 3244 3172 3080 3032 2982 2921 2870 2834 2806 2489

n engl j med 364;12

nejm.org

march 24, 2011

1099

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

Subgroup

Tiotropium

n e w e ng l a n d j o u r na l

Salmeterol

of

m e dic i n e

P Value for Interaction

Hazard Ratio (95% CI)

no. of patients/total no. Age (yr) <55 ≥55 to <65 ≥65 Sex Male Female COPD severity stage (GOLD) Stage II Stage III Stage IV Smoking status Noncurrent smoker Current smoker BMI <20 ≥20 to <25 ≥25 to <30 ≥30 Use of inhaled glucocorticoids at baseline Yes No

0.76 237/655 484/1462 556/1590

258/665 522/1426 634/1578

0.88 (0.74–1.05) 0.87 (0.77–0.98) 0.83 (0.74–0.93)

913/2759 364/948

1016/2747 398/922

0.86 (0.78–0.94) 0.84 (0.73–0.97)

561/1781 589/1597 127/329

635/1833 627/1545 152/291

0.88 (0.79–0.99) 0.86 (0.77–0.97) 0.64 (0.50–0.81)

678/1929 599/1778

746/1896 668/1773

0.84 (0.75–0.93) 0.87 (0.78–0.97)

105/286 455/1230 424/1276 293/915

134/271 501/1254 468/1284 311/860

0.66 (0.51–0.85) 0.89 (0.79–1.02) 0.87 (0.76–0.99) 0.85 (0.72–1.00)

0.83

0.05

0.64

0.17

0.41 785/1986 492/1721

0.87 (0.79–0.96) 0.82 (0.73–0.92)

839/1955 575/1714 0.4

0.6

0.8

Tiotropium Better

1.0

1.2

1.4

Salmeterol Better

Figure 3. Subgroup Analysis of the Primary End Point. The number of patients who had at least one exacerbation of chronic obstructive pulmonary disease (COPD) with onset during the study-treatment period is shown according to subgroup. Hazard ratios were calculated with the use of Cox regression with terms for treatment. Horizontal lines represent 95% confidence intervals. The size of the squares is proportional to the size of the subgroup. The severity of COPD was defined according to the classification of the Global Initiative for Chronic Obstructive Lung Disease (GOLD), which specifies four stages of COPD ranging from stage I, indicating mild disease, to stage IV, indicating very severe disease. Noncurrent smokers included former smokers and one person who had never smoked. The body-mass index (BMI) is the weight in kilograms divided by the square of the height in meters.

patients who continued to receive them; the annual exacerbation rate in the salmeterol group was 0.86 (95% CI, 0.74 to 0.99) among the 416 patients who discontinued the use of inhaled glucocorticoids, as compared with 0.81 (95% CI, 0.75 to 0.88) among the 1401 patients who continued to receive them. SAFETY

A total of 545 patients (14.7%) in the tiotropium group and 606 (16.5%) in the salmeterol group reported a serious adverse event during the studytreatment period (Table 2). The most common serious adverse event with a frequency of 0.5% or greater was an exacerbation of COPD, which occurred in 270 patients (7.3%) in the tiotropium 1100

n engl j med 364;12

group and in 335 (9.1%) in the salmeterol group (Section 12 in the Supplementary Appendix). A total of 180 cases of pneumonia were reported, of which 158 (87.8%) were radiologically confirmed (70 in the tiotropium group and 88 in the salmeterol group). There were more patients with at least one radiologically confirmed episode of pneumonia among those who received concomitant medication with inhaled glucocorticoids for at least 1 day during the study-treatment period than among those who received no inhaled glucocorticoid during the study-treatment period — 89 of 3330 patients (2.7%), of whom 72 required hospitalization, as compared with 59 of 4046 patients (1.5%), of whom 46 required hospitalization.

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Tiotropium vs. Salmeterol for COPD Exacerbations

Table 2. Incidence Rates of Serious Adverse Events, According to System Organ Class.*

Serious Adverse Events

Tiotropium (N = 3707)

Respiratory, thoracic, and mediastinal events

Salmeterol (N = 3669)

no. (%)

rate/100 patient-yr

no. (%)

rate/100 patient-yr

Rate Ratio for Tiotropium vs. Salmeterol (95% CI)

300 (8.1)

8.66

366 (10.0)

10.99

0.79 (0.68–0.92)

Infections

96 (2.6)

2.69

109 (3.0)

3.15

0.85 (0.65–1.12)

Cardiac events

98 (2.6)

2.73

85 (2.3)

2.44

1.12 (0.84–1.50)

Neoplasms

51 (1.4)

1.42

43 (1.2)

1.23

1.15 (0.77–1.73)

Vascular events

37 (1.0)

1.03

25 (0.7)

0.71

1.44 (0.87–2.39)

Gastrointestinal events

32 (0.9)

0.89

32 (0.9)

0.92

0.97 (0.59–1.58)

Nervous system events

28 (0.8)

0.78

29 (0.8)

0.83

0.94 (0.56–1.58)

General events†

16 (0.4)

0.44

27 (0.7)

0.77

0.57 (0.31–1.07)

Injury, poisoning, and procedural complications

22 (0.6)

0.61

19 (0.5)

0.54

1.13 (0.61–2.08)

Musculoskeletal and connective-tissue events

10 (0.3)

0.28

22 (0.6)

0.63

0.44 (0.21–0.93)

* Listed are incidence rates per 100 patient-years and incidence rate ratios of serious adverse events that occurred from the beginning of the study-treatment period until 30 days after the last dose of study drug was received and that were reported by at least 0.5% of the patients in either study group. The adverse events are categorized according to the system organ classes in the Medical Dictionary for Regulatory Activities. † This category includes the diagnostic terms “death” and “sudden death.”

There were 142 deaths during the planned treatment period of 360 days (including deaths among patients who had withdrawn from the study prematurely and whose vital status was recorded at 360 days): 64 in the tiotropium group and 78 in the salmeterol group (hazard ratio with tiotropium, 0.81; 95% CI, 0.58 to 1.13). Additional information is provided in Section 13 in the Supplementary Appendix.

Discussion Tiotropium, as compared with salmeterol, significantly increased the time to the first moderate or severe exacerbation of COPD and significantly decreased the annual rate of exacerbations among patients with moderate-to-very-severe COPD. The benefit with tiotropium was seen consistently in all the major subgroups that were considered in this trial and was independent of the concomitant use of inhaled glucocorticoids. This 1-year study was designed and powered for the end point of moderate and severe exacerbations, one of the most relevant patient-related outcomes, with important effects on patients’ families, caregivers, health care providers, and payers.4-6 Any exacerbation that can be avoided

n engl j med 364;12

would be beneficial from the perspective of both the patient and the health care system and constitutes a major treatment goal in COPD.1,2 Previous large, long-term trials have shown that both salmeterol and tiotropium reduce the rate of exacerbations.8,12 However, to date, there has been insufficient evidence from direct comparisons of the two drugs; therefore, current guidelines do not favor one long-acting agent over the other for patients with COPD.1,2 The Kaplan–Meier analyses of the time to the first exacerbation show that the benefit with tiotropium as compared with salmeterol became evident as early as approximately 1 month after the initiation of treatment and was maintained over the entire 1-year study period. Thus, it appears to be unlikely that the difference in favor of tiotropium was due to early discontinuation of treatment among patients in the salmeterol group who did not have a response to that drug. Tiotropium and salmeterol have been shown to reduce airflow limitation and hyperinflation but may also directly or indirectly have an effect on various aspects of lung inflammation.21,22 However, the relevance of these mechanisms to the observed differences in the end points related to exacerbations remains to be determined. Whether the

nejm.org

march 24, 2011

1101

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

observed differences might be due to differences in the aerosolizing systems, the particle size of the aerosols, or the distribution of the drug in the lung is also unknown. The annual exacerbation rates in this study were lower than those in large trials involving patients with COPD, such as the Trial of Inhaled Steroids and Long-acting β2 Agonists (TRISTAN)23 and the Towards a Revolution in COPD Health trial (TORCH; ClinicalTrials.gov number, NCT00268216),8 were similar to those in the Understanding Potential Long-Term Impacts on Function with Tiotropium trial (UPLIFT, NCT00144339),12 and were higher than those in a recent 1-year study comparing the efficacy of two long-acting β2-agonists.24 This variability may reflect differences in inclusion criteria and in the concomitant medications, such as inhaled glucocorticoids, that patients were allowed to receive. In our trial, consistent with current guideline recommendations, concomitant therapy with inhaled glucocorticoids was allowed but was not mandatory, because the patient population included a substantial proportion of patients with moderate COPD (GOLD stage II). Approximately 40% of the patients received concomitant therapy with inhaled glucocorticoids on a consistent basis during the study-treatment period. In a post hoc analysis, treatment with tiotropium decreased the risk of exacerbations more than did treatment with salmeterol both in patients who were receiving inhaled glucocorticoids and in those who were not receiving them, suggesting that the benefit of tiotropium was independent of the use of inhaled glucocorticoids. In addition, the rate of exacerbations among patients in the tiotropium group who were receiving inhaled glucocorticoids at baseline but did not continue receiving them during the studytreatment period was not higher than the rate among those who were receiving inhaled glucocorticoids at baseline and continued to receive them during the study-treatment period. This finding is consistent with the results of the COPD and Seretide: a Multi-Center Intervention and Characterization (COSMIC) study, which showed that withdrawal of fluticasone for 1 year after a 3-month run-in period with a fixed combination of fluticasone and salmeterol was not associated with an increase in moderate or severe exacerbations.25 Differences between study groups in the proportion of patients discontinuing the study treat1102

n engl j med 364;12

of

m e dic i n e

ment have been seen in other studies involving patients with COPD and are most often attributed to relative differences in the efficacy, safety, or both of the agents used in the study.7,12,26,27 Similarly, we observed a significantly higher rate of premature discontinuation of treatment in the salmeterol group than in the tiotropium group. However, as compared with the between-group differences that have been seen in placebo-controlled studies, the absolute difference was quite small (1.9 percentage points). Both tiotropium and salmeterol have safety profiles that have been well described in the literature.28-31 Overall, the incidence of serious adverse events, adverse events leading to treatment discontinuation, and fatal events were similar across treatments. In summary, among patients with moderateto-very-severe COPD and a history of exacerbation, tiotropium was more effective than salmeterol in all the exacerbation end points that were assessed and across all major subgroups. The results of this large trial provide data on which to base the choice of long-acting bronchodilator therapy for maintenance treatment of COPD. Dr. Vogelmeier reports receiving consulting fees and travel support from Boehringer Ingelheim, payment for board membership from AstraZeneca, Boehringer Ingelheim, GlaxoSmithKline, Mundipharma, Novartis, and Nycomed, fees for expert testimony and grant support from Talecris Biotherapeutics, and lecture fees from AstraZeneca, Boehringer Ingelheim, GlaxoSmithKline, Janssen, Merck, Novartis, Nycomed, and Talecris Biotherapeutics; Dr. Hederer, being formerly employed by Boehringer Ingelheim; Drs. Glaab and Schmidt, being currently employed by Boehringer Ingelheim; Dr. Rutten-van Mölken, receiving grant support from Boehringer Ingelheim and having consulting fees, travel support, and fees for participation in review activities from Boehringer Ingelheim paid to her institution on her behalf; Dr. Beeh, having consulting fees or honoraria, grant support, and payments for development of educational materials from Boehringer Ingelheim and grant support from Almirall, Novartis, Mundipharma, Cytos Biotechnology, GlaxoSmithKline, Pfizer, and Revotar Biopharmaceuticals paid to his institution on his behalf; Dr. Rabe, receiving consulting fees and honoraria and payment for board membership from AstraZeneca, Boehringer Ingelheim, Chiesi Farmaceutici, GlaxoSmithKline, Merck, Novartis, Nycomed, and Pfizer and grant support from Altana, AstraZeneca, Boehringer Ingelheim, GlaxoSmithKline, Novartis, and Roche; and Dr. Fabbri, receiving travel support from Boehringer Ingelheim, consulting fees from AstraZeneca, Boehringer Ingelheim, Chiesi Farmaceutici, GlaxoSmithKline, Merck, Novartis, Nycomed, Pfizer, and Sigma-Tau Pharmaceuticals, and lecture fees from AstraZeneca, Boehringer Ingelheim, Chiesi Farmaceutici, GlaxoSmithKline, Merck, Novartis, Nycomed, and Pfizer and having grant support from AstraZeneca, Boehringer Ingelheim, Chiesi Farmaceutici, GlaxoSmithKline, Menarini, Merck, Nycomed, Pfizer, Schering-Plough (now Merck), SigmaTau Pharmaceuticals, and Union Chimique Belge paid to his institution on his behalf. 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.

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Tiotropium vs. Salmeterol for COPD Exacerbations We thank the patients who participated in the trial; Dr. Klaus Fichtner, Vera Drews, and Nicole Bader for administrative trial support; Dr. Steven Kesten, Dr. Fee Rühmkorf, Dr. Harald Kögler, Dr. Susanne Stowasser, Dr. Brigitta Monz, and Dagmar Selim for scientific advice; Dr. Inge Leimer and Achim Müller for statistical advice; Michael Betke-Hornfeck for technical trial support; Christine Meissner and Christina Raabe for data management References 1. Global Initiative for Chronic Obstructive Lung Disease. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: 2009 update. (http://www.goldcopd.com.) 2. National Institute for Health and Clinical Excellence. Chronic obstructive pulmonary disease: management of chronic obstructive pulmonary disease in adults in primary and secondary care. London: National Clinical Guideline Centre, 2010. (http://guidance.nice.org.uk/ CG101/Guidance/pdf/English.) 3. Niewoehner DE. Outpatient management of severe COPD. N Engl J Med 2010; 362:1407-16. 4. Donaldson GC, Seemungal TAR, Bhowmik A, Wedzicha JA. Relationship between exacerbation frequency and lung function decline in chronic obstructive pulmonary disease. Thorax 2002;57:84752. [Erratum, Thorax 2008;63:753.] 5. Soler-Cataluña JJ, Martínez-García MÁ, Román Sánchez P, Salcedo E, Navarro M, Ochando R. Severe acute exacerbations and mortality in patients with chronic obstructive pulmonary disease. Thorax 2005;60:925-31. 6. Wedzicha JA, Seemungal TA. COPD exacerbations: defining their cause and prevention. Lancet 2007;370:786-96. 7. Hurst JR, Vestbo J, Anzueto A, et al. Susceptibility to exacerbation in chronic obstructive pulmonary disease. N Engl J Med 2010;363:1128-38. 8. Calverley PMA, Anderson JA, Celli B, et al. Salmeterol and fluticasone propionate and survival in chronic obstructive pulmonary disease. N Engl J Med 2007; 356:775-89. 9. Casaburi R, Mahler DA, Jones PW, et al. A long-term evaluation of once-daily inhaled tiotropium in chronic obstructive pulmonary disease. Eur Respir J 2002;19: 217-24. 10. Mahler DA, Donohue JF, Barbee RA, et al. Efficacy of salmeterol xinafoate in the treatment of COPD. Chest 1999;115: 957-65. 11. Rennard SI, Anderson W, ZuWallack R, et al. Use of a long-acting inhaled β2adrenergic agonist, salmeterol xinafoate,

support; Bernd Damian and Rafal Falkowski for programming support; Dr. Silke Wienecke, Declan Tobin, and Karen Ryan from Parexel for onsite monitoring and site management support; Natalie Barker and Claire Scarborough, from Parexel MedCom, for editorial and technical support in the preparation of the manuscript; and Dr. Franklin Cerasoli, Dr. Idelle Weisman, and Dr. Lalitha Aiyer for review of the manuscript.

in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2001;163:1087-92. 12. Tashkin DP, Celli B, Senn S, et al. A 4-year trial of tiotropium in chronic obstructive pulmonary disease. N Engl J Med 2008;359:1543-54. 13. Briggs DD Jr, Covelli H, Lapidus R, Bhattycharya S, Kesten S, Cassino C. Improved daytime spirometric efficacy of tiotropium compared with salmeterol in patients with COPD. Pulm Pharmacol Ther 2005;18:397-404. 14. Brusasco V, Hodder R, Miravitlles M, Korducki L, Towse L, Kesten S. Health outcomes following treatment for six months with once daily tiotropium compared with twice daily salmeterol in patients with COPD. Thorax 2003;58:399404. [Erratum, Thorax 2005;60:105.] 15. Mushlin AI, Ghomrawi H. Health care reform and the need for comparative-effectiveness research. N Engl J Med 2010;362(3):e6. (Available at NEJM.org.) 16. Sawicki PT. Communal responsibility for health care — the example of benefit assessment in Germany. N Engl J Med 2009;361(20):e42. (Available at NEJM.org.) 17. Beeh KM, Hederer B, Glaab T, et al. Study design considerations in a large COPD trial comparing effects of tiotropium with salmeterol on exacerbations. Int J Chron Obstruct Pulmon Dis 2009;4:11925. 18. Quanjer PH, Tammeling GJ, Cotes JE, Pedersen OF, Peslin R, Yernault JC. Lung volumes and forced ventilatory flows: Report Working Party Standardization of Lung Function Tests, European Community for Steel and Coal: official statement of the European Respiratory Society. Eur Respir J Suppl 1993;16:5-40. 19. American Thoracic Society. Standardization of spirometry, 1994 update. Am J Respir Crit Care Med 1995;152:1107-36. 20. Keene ON, Vestbo J, Anderson JA, et al. Methods for therapeutic trials in COPD: lessons from the TORCH trial. Eur Respir J 2009;34:1018-23. 21. Bateman ED, Rennard S, Barnes PJ, et al. Alternative mechanisms for tiotropium. Pulm Pharmacol Ther 2009;22:533-42.

n engl j med 364;12

nejm.org

22. Johnson M, Rennard S. Alternative

mechanisms for long-acting beta(2)adrenergic agonists in COPD. Chest 2001; 120:258-70. 23. Calverley P, Pauwels R, Vestbo J, et al. Combined salmeterol and fluticasone in the treatment of chronic obstructive pulmonary disease: a randomised controlled trial. Lancet 2003;361:449-56. [Erratum, Lancet 2003;361:1660.] 24. Dahl R, Chung KF, Buhl R, et al. Efficacy of a new once-daily long-acting inhaled beta2-agonist indacaterol versus twice-daily formoterol in COPD. Thorax 2010;65:473-9. 25. Wouters EF, Postma DS, Fokkens B, et al. Withdrawal of fluticasone propionate from combined salmeterol/fluticasone treatment in patients with COPD causes immediate and sustained disease deterioration: a randomised controlled trial. Thorax 2005;60:480-7. 26. Calverley PM, Spencer S, Willits L, Burge PS, Jones PW. Withdrawal from treatment as an outcome in the ISOLDE study of COPD. Chest 2003;124:1350-6. 27. Niewoehner DE, Rice K, Cote C, et al. Prevention of exacerbations of chronic obstructive pulmonary disease with tiotropium, a once-daily inhaled anticholinergic bronchodilator: a randomized trial. Ann Intern Med 2005;143:317-26. 28. Celli B, Decramer M, Leimer I, Vogel U, Kesten S, Tashkin DP. Cardiovascular safety of tiotropium in patients with COPD. Chest 2010;137:20-30. 29. Kesten S, Celli B, Decramer M, Leimer I, Tashkin D. Tiotropium HandiHaler in the treatment of COPD: a safety review. Int J Chron Obstruct Pulmon Dis 2009; 4:397-409. 30. Ferguson GT, Funck-Brentano C, Fischer T, Darken P, Reisner C. Cardiovascular safety of salmeterol in COPD. Chest 2003;123:1817-24. 31. Calverley PMA, Anderson JA, Celli B, et al. Cardiovascular events in patients with COPD: TORCH study results. Thorax 2010;65:719-25. Copyright © 2011 Massachusetts Medical Society.

march 24, 2011

1103

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

original article

Pioglitazone for Diabetes Prevention in Impaired Glucose Tolerance Ralph A. DeFronzo, M.D., Devjit Tripathy, M.D., Ph.D., Dawn C. Schwenke, Ph.D., MaryAnn Banerji, M.D., George A. Bray, M.D., Thomas A. Buchanan, M.D., Stephen C. Clement, M.D., Robert R. Henry, M.D., Howard N. Hodis, M.D., Abbas E. Kitabchi, M.D., Ph.D., Wendy J. Mack, Ph.D., Sunder Mudaliar, M.D., Robert E. Ratner, M.D., Ken Williams, M.Sc., Frankie B. Stentz, Ph.D., Nicolas Musi, M.D., and Peter D. Reaven, M.D., for the ACT NOW Study

A BS T R AC T BACKGROUND From the Texas Diabetes Institute and University of Texas Health Science Center (R.A.D., D.T., N.M.) and KenAnCo Biostatistics (K.W.) — both in San Antonio; Phoenix Veterans Affairs (VA) Health Care System, Phoenix, AZ (D.C.S., P.D.R.); College of Nursing and Health Innovation, Arizona State University, Phoenix (D.C.S.); SUNY Health Science Center at Brooklyn, Brooklyn, NY (M.B.); Pennington Biomedical Research Center–Louisiana State University, Baton Rouge (G.A.B.); University of Southern California Keck School of Medicine, Los Angeles (T.A.B., H.N.H., W.J.M.); Division of Endocrinology and Metabolism, Georgetown University, Washington, DC (S.C.C.); VA San Diego Healthcare System and University of California at San Diego, San Diego (R.R.H., S.M.); University of Tennessee, Division of Endocrinology, Diabetes, and Metabolism, Memphis (A.E.K., F.B.S.); and Medstar Research Institute, Hyattsville, MD (R.E.R.). Address reprint requests to Dr. DeFronzo at the Diabetes Division, University of Texas Health Science Center, 7703 Floyd Curl Dr., San Antonio, TX 78229, or at albarado@ uthscsa.edu. Drs. DeFronzo and Tripathy contributed equally to this article. N Engl J Med 2011;364:1104-15. Copyright © 2011 Massachusetts Medical Society.

Impaired glucose tolerance is associated with increased rates of cardiovascular disease and conversion to type 2 diabetes mellitus. Interventions that may prevent or delay such occurrences are of great clinical importance. METHODS

We conducted a randomized, double-blind, placebo-controlled study to examine whether pioglitazone can reduce the risk of type 2 diabetes mellitus in adults with impaired glucose tolerance. A total of 602 patients were randomly assigned to receive pioglitazone or placebo. The median follow-up period was 2.4 years. Fasting glucose was measured quarterly, and oral glucose tolerance tests were performed annually. Conversion to diabetes was confirmed on the basis of the results of repeat testing. RESULTS

Annual incidence rates for type 2 diabetes mellitus were 2.1% in the pioglitazone group and 7.6% in the placebo group, and the hazard ratio for conversion to diabetes in the pioglitazone group was 0.28 (95% confidence interval, 0.16 to 0.49; P<0.001). Conversion to normal glucose tolerance occurred in 48% of the patients in the pioglitazone group and 28% of those in the placebo group (P<0.001). Treatment with pioglitazone as compared with placebo was associated with significantly reduced levels of fasting glucose (a decrease of 11.7 mg per deciliter vs. 8.1 mg per deciliter [0.7 mmol per liter vs. 0.5 mmol per liter], P<0.001), 2-hour glucose (a decrease of 30.5 mg per deciliter vs. 15.6 mg per deciliter [1.6 mmol per liter vs. 0.9 mmol per liter], P<0.001), and HbA1c (a decrease of 0.04 percentage points vs. an increase of 0.20 percentage points, P<0.001). Pioglitazone therapy was also associated with a decrease in diastolic blood pressure (by 2.0 mm Hg vs. 0.0 mm Hg, P = 0.03), a reduced rate of carotid intima–media thickening (31.5%, P = 0.047), and a greater increase in the level of high-density lipoprotein cholesterol (by 7.35 mg per deciliter vs. 4.5 mg per deciliter [0.4 mmol per liter vs. 0.3 mmol per liter], P = 0.008). Weight gain was greater with pioglitazone than with placebo (3.9 kg vs. 0.77 kg, P<0.001), and edema was more frequent (12.9% vs. 6.4%, P = 0.007). CONCLUSIONS

As compared with placebo, pioglitazone reduced the risk of conversion of impaired glucose tolerance to type 2 diabetes mellitus by 72% but was associated with significant weight gain and edema. (Funded by Takeda Pharmaceuticals and others; ClinicalTrials.gov number, NCT00220961.) 1104

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Pioglitazone for Diabetes Prevention in IGT

T

ype 2 diabetes mellitus affects 21 million Americans,1 and its prevalence is increasing.2 Microvascular and macrovascular complications are common in type 2 diabetes mellitus and are related to both the severity and the duration of hyperglycemia.3 The natural history of type 2 diabetes mellitus has been well defined,4 starting with a genetic predisposition and progression from normal glucose tolerance with insulin resistance to impaired glucose tolerance and eventually type 2 diabetes mellitus with the superimposition of beta-cell failure. Because hyperglycemia plays a central role in the microvascular and macrovascular complications of diabetes,3,5,6 it is possible that interventions that prevent or delay hyperglycemia may effectively prevent or delay these long-term complications. Studies have shown that the rate of conversion of impaired glucose tolerance to type 2 diabetes mellitus is reduced with lifestyle modification7; the use of metformin,7 thiazolidinediones,8-11 or acarbose12; and bariatric surgery.13 The greatest reductions in conversion rates have been observed with weight loss, the use of thiazolidinediones, and surgery. Troglitazone was reported to be associated with a 55% decrease in the rate of conversion to diabetes among women with prior gestational diabetes,11 but this agent is no longer available. In one study,9 the use of rosiglitazone decreased the risk of diabetes in adults with impaired glucose tolerance by 62%; given concerns about cardiovascular safety,14 however, the Food and Drug Administration has restricted the use of rosiglitazone therapy to patients in whom glycemic control cannot be achieved with other medications and who cannot take pioglitazone. We undertook the present study to examine the effect of pioglitazone on diabetes risk and cardiovascular risk factors in adults with impaired glucose tolerance.

Me thods Patients

We recruited male and female patients who were 18 years of age or older and had impaired glucose tolerance (defined as a 2-hour glucose level of 140 to 199 mg per deciliter [7.8 to 11.0 mmol per liter] during a single oral glucose-tolerance test)15 and a body-mass index (BMI, the weight in kilograms divided by the square of the height in meters) of 25 or more. Patients were eligible for

n engl j med 364;12

enrollment if they had a fasting plasma glucose level between 95 and 125 mg per deciliter (5.3 and 6.9 mmol per liter) and at least one other risk factor for diabetes.16 The complete list of inclusion and exclusion criteria is provided in Table 1 in the Supplementary Appendix, available with the full text of this article at NEJM.org. The criteria have also been published previously.16 During the course of recruitment, the glycemic inclusion criteria were modified to include patients with a fasting plasma glucose level between 90 and 125 mg per deciliter (5.0 and 6.9 mmol per liter) if their 2-hour plasma glucose level during the oral glucose-tolerance test was between 170 and 199 mg per deciliter (9.4 and 11.1 mmol per liter)17; the change was made in recognition of the high risk of diabetes in such persons. The first participant was recruited in January 2004, with the screening ultimately including 1827 potentially eligible patients with impaired glucose tolerance (Fig. 1). The enrollment of 602 participants was completed in March 2006. Participants were followed until they reached the primary end point of diabetes, withdrew from the study, were lost to follow-up, or completed the end of the study. Study Design

The study design and protocol, which have been described previously,16 are available at NEJM.org. Eight centers participated in the study, which was approved by the institutional review board at each site. Written informed consent was obtained from all participants. The first author designed the study and, along with the coauthors, wrote the first draft and revisions and approved the final version; he also holds the data at the University of Texas Health Science Center in San Antonio. The study was conducted in accordance with the protocol. All authors made the decision to submit the manuscript for publication. All results were transmitted to the Data Coordinating Center in Phoenix, Arizona, where they were recorded and audited and then sent to the Data Analysis Center in San Antonio. Takeda Pharmaceuticals provided financial support for the study but had no access to the data. After eligibility for the study was ascertained, participants underwent randomization according to center and sex and received 30 minutes of dietary instruction consistent with the goals of the Diabetes Prevention Program,7 which was re-

nejm.org

march 24, 2011

1105

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

1827 Patients were assessed for eligibility

1225 Were excluded 277 Had FPG out of range 187 Declined to participate 118 Had medical exclusions 12 Had too few risk factors 4 Had behavioral or logistic problems 627 Had OGTT–FPG or 2-hr glucose out of range

602 Underwent randomization

303 Were assigned to receive pioglitazone

299 Were assigned to receive placebo

90 Did not have a final visit for the following reasons: 9 Had weight gain 6 Relocated 12 Had work or schedule conflict 18 Had lack of interest 7 Did not give reason 28 Were lost to follow-up 5 Had unrelated illness or adverse events 5 Had other reason

71 Did not have a final visit for the following reasons: 3 Had weight gain 17 Relocated 1 Had work or schedule conflict 9 Had lack of interest 10 Did not give reason 22 Were lost to follow-up 3 Had unrelated illness or adverse events 6 Had other reason

213 Completed study

228 Completed study

Figure 1. Enrollment, Randomization, and Follow-up of Study Patients. FPG denotes fasting plasma glucose, and OGTT oral glucose-tolerance test.

inforced on follow-up visits. Once enrolled, participants were asked to fast overnight before undergoing an oral glucose-tolerance test at 8 a.m. the next day. Samples were collected every 15 minutes for 2 hours for measurements of glucose, insulin, and C-peptide.16 Additional baseline assessments included measurements of blood pressure, height, weight, waist circumference, and level of HbA1c; a lipid profile; screening blood tests; urinalysis, with calculation of the ratio of microalbumin to creatinine; and electrocardiography. At seven centers, high-resolution B-mode ultrasonography was performed at baseline, 15 to 18 months after baseline, and at the end of the study, to assess the far wall of the right distal 1106

n engl j med 364;12

common carotid artery (for details, see the Methods section in the Supplementary Appendix).18,19 Body-fat content was measured with the use of a dual-energy x-ray absorptiometer (DXA) (Hologic). Participants initially received 30 mg of pioglitazone per day or placebo. One month after randomization, the dose of pioglitazone was increased to 45 mg per day. Participants returned at 2, 4, 6, 8, 10, and 12 months during the first year of the study and once every 3 months thereafter. At each visit, weight, blood pressure, and pulse were measured and the extent of edema was graded (with an increase in edema defined as an increase of two grades or more from baseline) (for details, see Table 2 in the Supplementary

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Pioglitazone for Diabetes Prevention in IGT

Appendix). Fasting plasma glucose was also measured at each follow-up visit. The levels of HbA1c, alanine aminotranferase, and aspartate aminotransferase were measured every 6 months, and the oral glucose-tolerance test was repeated annually. All measurements obtained at baseline were repeated at the end of the study. Conversion of Impaired Glucose Tolerance to Diabetes

The primary outcome was the development of diabetes15 (defined as a fasting plasma glucose level ≥126 mg per deciliter [≥7.9 mmol per liter] or a 2-hour glucose level ≥200 mg per deciliter [11.1 mmol per liter]); an oral glucose-tolerance test was performed to confirm the diagnosis. If the diagnosis was not confirmed, patients continued with their assigned therapy. Diabetes was not confirmed but was considered to have developed in five patients receiving pioglitazone and five patients receiving placebo. At the final visit, six patients had a single oral glucose-tolerance test with results that met the diagnostic criteria for diabetes; four of the six were started on antidiabetic medication by their physician. Measurements and Calculations

Insulin sensitivity was derived from plasma glucose and insulin measurements obtained during the oral glucose-tolerance test (Matsuda index)20 and from the results of an intravenous glucosetolerance test with frequent sampling.21 Beta-cell function was calculated as the index of insulin secretion factored by insulin resistance (∆I0–120/ ∆G0–120 × Matsuda index) during the oral glucosetolerance test, where ∆I0–120/∆G0–120 represents the mean incremental concentrations of plasma insulin and glucose during the 120-minute oral glucose-tolerance test.22 Beta-cell function was also calculated as the product of insulin secretion and insulin sensitivity (∆I0–10 × SI) during the intravenous glucose-tolerance test with frequent sampling. Laboratory methods are described in detail in the Methods section in the Supplementary Appendix.

glitazone and placebo groups and compared with the use of the log-rank test.23 The Cox proportional-hazards model was used to estimate the effect of pioglitazone on the primary outcome.24 Data for patients who were lost to follow-up or who withdrew were censored at the time of the last visit. Statistical tests were two-sided, with an alpha level of 0.05. Data are reported as means ±SE. We calculated that enrollment of approximately 600 patients was required to achieve 90% power, if treatment with pioglitazone decreased the rate of conversion from impaired glucose tolerance to type 2 diabetes mellitus by 50%.16 For analyses of secondary outcomes, which included changes in levels of fasting plasma glucose, 2-hour glucose, and HbA1c, between-group comparisons of changes in repeated or continuous measures were performed with the use of general linear mixed models, with data transformed to logarithms when appropriate. The statistical heterogeneity of treatment effects within subgroups was assessed. No adjustment was made for multiple comparisons, and subgroup analyses were not prespecified. Two approaches were used to assess whether patients who completed the study differed from those who withdrew. The first approach involved a withdrawal-free survival analysis of time to withdrawal, with the final study visit used as the censoring variable. Data for patients who underwent an oral glucose-tolerance test at the end-ofstudy visit were censored at 3 years. All other patients were counted as having withdrawn as of the last study visit. On the basis of this analysis, the hazard ratio for withdrawal in the pioglitazone group as compared with the placebo group was 1.125 (P = 0.42). In the second approach, missing data (for the two study groups combined) were assessed for each continuous measure with analysis of variance, stratified according to whether the measure was missing at each subsequent visit. Since neither approach produced statistically significant evidence of bias due to missing data, the primary and secondary analyses were performed without data imputation.

Statistical Analysis

The development of diabetes,15 the primary outcome, was assessed by means of life-table analysis of the time from randomization to incident diabetes. Separate life-table estimated cumulative incidence curves were calculated for the pion engl j med 364;12

R e sult s Participants

The mean age of the 602 study participants was 52.3±0.5 years, and 58% were women (Table 1). The mean BMI was 34.5±0.4. A total of 407 pa-

nejm.org

march 24, 2011

1107

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

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. Baseline Characteristics of the Patients.* Pioglitazone (N = 303)

Characteristic Isolated impaired glucose tolerance — no.

Placebo (N = 299)

P Value

98

97

0.98

205

202

0.98

58/42

58/42

0.96

79

75

White

156

171

Black

57

44

Other

11

9

Both impaired glucose tolerance and impaired fasting glucose tolerance — no. Ratio of women to men — % Race or ethnic group — no.†

0.45

Hispanic

Family history of diabetes — no. (%)

177 (58.4)

186 (62.2)

0.34

History of gestational diabetes mellitus — no. (% of women)

24 (14)

37 (21)

0.07

Mean age — yr

53.0±0.4

51.5±0.7

0.12

18–39 yr

36

42

0.155

40–59 yr

32

28

0.33

Age group — %

≥60 yr

32

30

0.66

Height — cm

166±0.5

167±0.6

0.47

Mean BMI

33.0±0.4

34.5±0.4

0.44

BMI group — no. (%) <30

79 (26.0)

76 (25.4)

0.95

30–35

109 (36.0)

100 (33.4)

0.71

>35

114 (37.6)

122 (40.8)

0.65

Waist circumference — cm Men

110.5±1.1

112.2±1.3

0.70

Women

103.1±0.9

103.7±1.0

0.31

HbA1c — %

5.5±0.4

5.5±0.4

0.23

Fasting plasma glucose — mg/dl

105±0.4

105±0.4

0.72

2-Hr plasma glucose — mg/dl

168±1

168±1

0.80

Fasting plasma insulin — mU/liter

10.5±0.5

10.7±0.6

0.84

Total cholesterol

169±2

173±2

0.22

LDL cholesterol

104±2

108±2

0.20

HDL cholesterol

40±1

41±1

0.57

122±3

121±3

0.84

551±13

528±13

0.17

Systolic

127±0.9

128±0.9

0.57

Diastolic

74±0.6

74±0.6

0.99

Lipid levels — mg/dl

Triglycerides Fasting free fatty acids (µmol/liter) Blood pressure — mm Hg

* Plus–minus values are means ±SE. No intergroup differences were significant. To convert the values for glucose to millimoles per liter, multiply by 0.05551. To convert the values for insulin to picomoles per liter, multiply by 6.945. To convert the values for cholesterol to millimoles per liter, multiply by 0.02586. To convert the values for triglycerides to millimoles per liter, multiply by 0.01129. BMI denotes body-mass index (the weight in kilograms divided by the square of the height in meters), HDL high-density lipoprotein, and LDL low-density lipoprotein. † Race or ethnic group was self-reported.

1108

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Pioglitazone for Diabetes Prevention in IGT

tients had both impaired fasting glucose and impaired glucose tolerance, and 195 had isolated impaired glucose tolerance. Baseline levels of HbA1c, fasting plasma glucose, and 2-hour glucose were 5.50±0.04%, 105±0.3 mg per deciliter (5.8±0.02 mmol per liter), and 168±1 mg per deciliter (9.32±0.06 mmol per liter), respectively. None of the baseline clinical, anthropometric, or laboratory variables differed significantly between the placebo group and the pioglitazone group (Table 1).

Cumulative Hazard

0.4

n engl j med 364;12

Placebo

0.3

0.2

0.1

Follow-up

During a median follow-up period of 2.4 years (mean, 2.2), diabetes developed in 50 of the 299 patients in the placebo group (16.7%) and in 15 of the 303 patients in the pioglitazone group (5.0%). The annual average incidence of diabetes, calculated on the basis of person-years, was 7.6% in the placebo group and 2.1% in the pioglitazone group (P<0.001) (Fig. 2). The hazard ratio for development of diabetes in the pioglitazone group was 0.28 (95% confidence interval, 0.16 to 0.49; P<0.001). Adjustment for baseline characteristics did not alter the hazard ratio. The number of people who would need to be treated to prevent one case of diabetes was 8 for 2.2 years of the trial and 18 for 1 year. Among the patients who completed the study, 103 of those in the pioglitazone group (48%) and 65 of those in the placebo group (28%) had normal glucose tolerance (P<0.001). A total of 161 patients did not complete the study (71 in the placebo group and 90 in the pioglitazone group). The median follow-up time for these patients was 7.6 months. Baseline characteristics of the patients who did not complete the study were similar to those of the 441 patients who completed the study (i.e., those who had conversion to type 2 diabetes mellitus during the study or who completed the oral glucosetolerance test at the end of the trial). Reasons for not completing the study included weight gain (in 9 patients in the pioglitazone group and 3 in the placebo group); patients also left for reasons unrelated to the study medication (Fig. 1). The rate of adherence to the study regimen, assessed by means of pill counts, was greater than 80% in both groups. At the end of the study, 64% of the patients in the treatment group were taking pioglitazone at a daily dose of 45 mg and 81% of those in the placebo group were taking the corresponding placebo dose. The major reasons for

Hazard ratio, 0.28 (95% CI, 0.16–0.49) P<0.001

0.0

Pioglitazone

0

6

12

18

24

30

36

42

48

Months since Randomization No. at Risk Placebo Pioglitazone

299 303

259 262

228 244

204 228

191 218

134 140

83 87

17 24

Figure 2. Kaplan–Meier Plot of Hazard Ratios for Time to Development of Diabetes.

not increasing the dose of pioglitazone from 30 to 45 mg per day or for not maintaining the 45-mg dose were weight gain and edema. Effects of Pioglitazone

Protection from diabetes with pioglitazone was of similar magnitude (with no significant heterogeneity) in subgroups defined by sex, age, weight, race or ethnic group, and fasting glucose level, as well as in patients with both impaired glucose tolerance and impaired fasting glucose and those with isolated impaired glucose tolerance (Fig. 3). There was no evidence of heterogeneity of the response according to the baseline level of HbA1c. Greater reductions in fasting and 2-hour glucose levels were achieved in the pioglitazone group than in the placebo group (P<0.001 for both comparisons), with a between-group difference of 3.5±1.1 mg per deciliter (0.2±0.06 mmol per liter) and 14±3 mg per deciliter (0.8±0.17 mmol per liter), respectively, at the end of the study (Fig. 4A and 4B). Levels of HbA1c differed between the groups throughout the study (P<0.001), increasing by 0.20±0.02% in the placebo group, with no change in the pioglitazone group. Body weight, BMI, and body fat increased in the placebo group (96.7±1.2 to 97.3±1.3 kg, 34.5±0.4 to 34.7±0.4, and 39.0±0.7 to 39.3±0.7%, respectively) and in the pioglitazone group (94.9±1.2 to 98.7±1.3 kg,

nejm.org

march 24, 2011

1109

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

Subgroup

n e w e ng l a n d j o u r na l

Pioglitazone

Placebo

of

m e dic i n e

P Value for Heterogeneity

Hazard Ratio (95% CI)

incidence rate (per 100 person-yr) All Sex Male Female Age <50 yr 50–59 yr ≥60 yr Race or ethnic group White Hispanic Black BMI <30 30–35 >35 Waist circumference Men, <100 cm Men, ≥100 cm Women, <90 cm Women, ≥90 cm Fasting plasma glucose 95–105 mg/dl 106–115 mg/dl 116–125 mg/dl Glucose test Isolated IGT IFG and IGT

2.1

7.6

1.3 2.8

6.4 8.4

3.4 2.1 0.8

8.5 9.4 4.5

2.1 2.2 2.8

7.0 7.0 11.2

2.6 1.1 2.8

5.1 7.5 8.4

1.3 1.1 3.4 2.6

3.6 7.5 11.2 8.3

1.5 2.6 6.3

5.1 9.6 18.5

1.8 2.3

3.7 9.3

0.66

0.75

0.98

0.75

0.48

0.84

0.88

0.0

0.5

1.0

Pioglitazone Better

1.5

3.5

4.0

Placebo Better

Figure 3. Incidence Rates, Hazard Ratios, and Tests for Heterogeneity for All Patients and Selected Subgroups. The figure shows incidence rates per 100 person-years and corresponding hazard ratios and confidence intervals for the effects of pioglitazone as compared with placebo on the conversion of impaired glucose tolerance to diabetes. The x axis is interrupted to allow for better visual presentation. BMI denotes body-mass index (the weight in kilograms divided by the square of the height in meters), IFG impaired fasting glucose, and IGT impaired glucose tolerance. To convert the values for glucose to millimoles per liter, multiply by 0.05551.

34.1±0.4 to 35.5±0.4, and 40.0±0.8 to 41.9±0.7%, respectively), but the increments were greater with pioglitazone (P<0.001 for all comparisons). Systolic blood pressure declined slightly in both groups, but the difference in decline between the groups was not significant. Diastolic blood pressure was consistently lower in the pioglitazone group (P = 0.01). As compared with placebo, pioglitazone reduced levels of both alanine aminotranferase and aspartate aminotransferase (P<0.001). The change in high-density lipoprotein (HDL) cholesterol was greater with pioglitazone (40±1 to 48±1 mg per deciliter [2.2±0.06 to 2.7±0.06 mmol per liter]) than with placebo 1110

n engl j med 364;12

(41±1 to 45.1±0.7 mg per deciliter [2.3±0.06 to 2.5±0.04 mmol per liter]) (P = 0.008 for the difference between groups). Triglyceride levels declined significantly with pioglitazone (129±7.5 to 110±4.0 mg per deciliter [7.2±0.42 to 6.1±0.22 mmol per liter], P = 0.007) but not with placebo (124±4.6 to 113±4.0 mg per deciliter [6.9±0.25 to 6.30±0.22 mmol per liter], P = 0.90); the difference between groups was not significant. Neither pioglitazone nor placebo altered levels of low-density lipoprotein cholesterol. Insulin sensitivity as measured with the Matsuda index increased more with pioglitazone than with placebo (4.31±0.24 to 7.65±0.34 vs.

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Pioglitazone for Diabetes Prevention in IGT

Placebo

−5

−15

Pioglitazone

P<0.001 0

12

24

36

−10 −20

P<0.001 0

12

Month

0.0 P<0.001 12

24

36

Mean Percent Change

Mean Percent Change

2.5

0

1.0

−2.0 −3.0 −4.0

P=0.16 0

12

P<0.001

−10 24

36

Mean Percent Change

Mean Percent Change

24

36

P<0.001 0

12

36

4.0 2.0 0.0 −2.0 −4.0 −6.0

P=0.01 0

12

24

36

I Carotid Intima–Media Thickness

30 20 10 0

P<0.001

−10 −20

24

Month

H AST

12

−7.5

Month

G ALT

0

−5.0

F Diastolic Blood Pressure

0.0 −1.0

Month

30 20 10 0

0.0 −2.5

Month

E Systolic Blood Pressure

5.0

−20

36

2.5

Month

D Weight

−2.5

24

Mean Percent Change

−10

0

0

12

Month

24

Month

36

Mean Percent Change

0

C HbA1c Mean Percent Change

B 2-Hr Plasma Glucose Mean Percent Change

Mean Percent Change

A Fasting Plasma Glucose

P=0.047 4.0 2.0 0.0 0

12

24

36

Month

Figure 4. Effects of Pioglitazone as Compared with Placebo. Over the course of the study, mean percentage changes and standard errors in continuous measures were calculated with the use of a linear, mixed-repeated-measures model fit to all available data for each measure. As compared with placebo, treatment with pioglitazone (dashed lines) had beneficial effects on fasting plasma glucose levels (Panel A), 2-hour plasma glucose levels (Panel B), and HbA1c levels (Panel C) and on systolic and diastolic blood pressure (Panels E and F, respectively), alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels (Panels G and H, respectively), and carotid intima–media thickness (Panel I). Weight gain was greater with pioglitazone than with placebo (Panel D). (Body-mass index was calculated at each examination with the use of height measured at baseline; as a result, the percentage change in BMI is identical to the percentage change in weight.) A total of 365 patients (placebo group, 186; pioglitazone group, 179) completed the follow-up examination at 15 to 18 months for measurement of carotid intima–media thickness, and 336 patients (placebo group, 173; pioglitazone group, 163) completed the final examination for measurement of carotid intima–media thickness. P values are shown for the interaction between time and study group, indicating whether the slopes differ significantly over time.

4.31±0.30 to 5.23±0.31, P<0.001). Insulin sensitivity as determined with the use of an intravenous glucose-tolerance test with frequent sampling in a subgroup of 191 patients was not altered in either group. The index of insulin secretion factored by insulin resistance, calculated on the basis of the oral glucose-tolerance test (I0–120/∆G0–120 × Matsuda index), increased more with pioglitazone than with placebo (3.43±0.12 to 5.44±0.31 vs. 3.81±0.30 to 4.20±0.20, P<0.005). Similarly, the insulin secretion–insulin resistance

n engl j med 364;12

index, calculated with data from the frequentsampling intravenous glucose-tolerance test, increased more with pioglitazone than with placebo (848±65 to 1186±113 vs. 824±47 to 832±57, P<0.01). Carotid intima–media thickening increased more slowly in the pioglitazone group than in the placebo group throughout the study (Fig. 4I). The differences between groups were 16.4% at the study midpoint and 31.5% at the end of the study (P = 0.047 for the overall difference between

nejm.org

march 24, 2011

1111

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

groups). The ratio of urinary microalbumin to creatinine was low at baseline; during the study it fell slightly and similarly in the two groups (1.85±0.2 to 1.53±0.2 mg per gram in the pioglitazone group vs. 1.47±0.3 to 1.25±0.3 mg per gram in the placebo group, P = 0.20).

curred in the placebo group, and three deaths occurred in the pioglitazone group (one unexplained sudden death, one death from biliary carcinoma, and one death from a carcinoid tumor). Nine fractures occurred in 8 of the patients receiving pioglitazone (3%) and eight fractures occurred in 7 of the patients receiving placebo Adverse Events (2.6%) (Table 4 in the Supplementary Appendix). Adverse events occurred in 121 patients in the pla- All fractures were associated with trauma. cebo group and 151 patients in the pioglitazone group (P = 0.03) (Table 2). Edema increased at Discussion some point during the trial in 19 patients receiving placebo (6.4%) and 39 patients receiving pio- Although they are considered to have prediabetes, glitazone (12.9%) (P = 0.007). Events related to the patients in the upper third of the range for imcardiovascular system numbered 23 in the placebo paired glucose tolerance are at or close to the group (7.7%) and 26 in the pioglitazone group maximum level of insulin resistance and have (8.6%) (Table 3 in the Supplementary Appendix), lost approximately 80% of beta-cell function.25,26 with 1 case of congestive heart failure in each Histologic studies suggest that the beta-cell mass group (0.3%). One unexplained sudden death oc- in patients with impaired fasting glucose is significantly reduced as compared with persons who have normal fasting glucose,27 and two thirds of Table 2. Number and Type of Adverse Events.* the patients in our study had impaired fasting Pioglitazone Placebo glucose. Moreover, at least 10% of patients with Adverse Event (N = 303) (N = 299) impaired glucose tolerance have background diano. of events betic retinopathy,28,29 and peripheral neuropathy is Cancer 3 8 common in these patients.30 Because the characteristic pathophysiological defects of type 2 diaCardiovascular system 26 23 betes mellitus and the microvascular complicaCentral nervous system 6 5 tions of diabetes are already evident in patients Death 3 1 with impaired glucose tolerance, it is reasonable Digestive system 13 12 to consider interventions at this stage to prevent Edema† 39 19 the development of overt diabetes. Elective surgery 22 16 Lifestyle interventions effectively reduce the Endocrine system 1 3 conversion of impaired glucose tolerance to diabetes31-34 and remain the primary approach to Immune system 2 4 prevention of type 2 diabetes mellitus. However, Musculoskeletal system 12 13 many people remain at risk for type 2 diabetes Ophthalmologic system 0 1 mellitus despite attempts at lifestyle changRespiratory system 9 6 es.7,31,32,34 Metformin reduces the risk of converReproductive system 4 4 sion to type 2 diabetes mellitus by 31%, without Skin 6 3 weight gain.7 Thiazolidinediones also effectively reduce the risk of development of type 2 diabetes Urogenital system 5 3 mellitus in patients with impaired glucose tolerWeight gain‡ 205 128 ance.8-11 In our study, pioglitazone decreased Total 356 249 the rate of conversion to diabetes by 72%, a change that was slightly larger than that ob* For the comparison of placebo and pioglitazone regarding frequency of edema, cardiovascular events, and total served with other thiazolidinediones (52 to 62%) events, P = 0.007, P = 0.80, and P = 0.03, respectively. The and lifestyle modification (58%). Although subtotal number of adverse events — excluding edema — group analyses were not prespecified in our did not differ significantly between groups (P = 0.52). † Edema was defined as an increase above baseline by two study design, pioglitazone reduced the risk of or more grades on one or more distinct study visits. conversion to diabetes in patients with isolated ‡ Weight gain was defined as a gain of more than 1 kg. impaired glucose tolerance, in those with both 1112

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Pioglitazone for Diabetes Prevention in IGT

impaired fasting glucose and impaired glucose tolerance, in both men and women, and in all age and weight groups. The proportion of patients who had a return to normal glucose tolerance was greater with pioglitazone than with placebo. The mean weight gain in patients treated with pioglitazone was 3.6 kg. However, the greater the weight gain, the greater the improvements in betacell function and insulin sensitivity, and thus the greater the reduction in HbA1c.26,35,36 The effect of weight gain on cardiovascular risk cannot be ascertained, but in the Prospective Pioglitazone Clinical Trial in Macrovascular Events (PROactive) (ClinicalTrials.gov number, NCT00174993),37 a weight gain of 3.6 kg was not associated with an increase in the composite cardiovascular end point. It is not feasible to conduct studies of microvascular outcomes in patients with impaired glucose tolerance because of the large sample and long study duration required. However, if development of diabetic hyperglycemia can be delayed or prevented, it is plausible that the onset of microvascular complications might be slowed. During the course of this study, fasting and 2-hour glucose levels and HbA1c levels were significantly lower in the pioglitazone group than in the placebo group. The small difference in levels of HbA1c between the two groups is not surprising, given the low baseline level of HbA1c in patients with impaired glucose tolerance. However, further glycemic separation between groups over time would presumably have a beneficial effect on microvascular disease. Pioglitazone was associated with lower diastolic blood pressure, higher levels of HDL cholesterol, and reduced rates of carotid intima–media thickening, as compared with placebo. Carotid intima–media thickening is highly correlated with cardiovascular events, and changes in this measure over time have predictive value beyond that of standard markers of risk.18,38 Such results suggest that pioglitazone may provide some protection against the development of atherosclerotic cardiovascular disease, which is consistent with reports of reductions in the volume of coronary plaque39 and in mortality, nonfatal myocardial infarction, and stroke, the secondary end points in the PROactive.37 Loss to follow-up was relatively high in both study groups (24% in the placebo group and 30% in the pioglitazone group, not a significant difn engl j med 364;12

ference). Since withdrawal rates and baseline characteristics were similar between groups, biased results seem unlikely. The modest difference in levels of HbA1c between groups suggests that the reduced progression of carotid intima–media thickening with pioglitazone may reflect improvements in other metabolic variables (Fig. 4). Although pioglitazone is a well-documented insulin sensitizer, insulin resistance was reduced according to the Matsuda index of insulin sensitivity measured during the oral glucose-tolerance test but not according to measurement of insulin sensitivity during the frequent-sampling intravenous glucose-tolerance test. These disparate results may partly reflect the greater variation among centers in the results of the intravenous glucose-tolerance test with frequent sampling, which is more difficult to perform. Edema and weight gain were greater with pioglitazone than with placebo, as reported previously,8,40 and edema largely accounted for the increase in adverse events associated with pioglitazone. Physician-reported congestive heart failure developed in only one patient in each group. Although an increased incidence of fractures has been reported with the use of thiazolidinediones, in this study, the incidence was similar in both groups, and all fractures were related to trauma. In summary, treatment with pioglitazone in patients with impaired glucose tolerance reduced the risk of diabetes, although pioglitazone was associated with significant weight gain and edema. Treatment of 18 participants for 1 year prevented one case of diabetes. Use of pioglitazone improved diastolic blood pressure, HDL cholesterol levels, and serum levels of alanine aminotranferase and aspartate aminotransferase, and it slowed progression of carotid intima–media thickening. The influence of these effects on long-term diabetic complications remains to be determined. Supported by Takeda Pharmaceuticals, grants from the General Clinical Research Center (GCRC) at the University of Tennessee Health Science Center (MO1-RR-00221) and the GCRC at the University of Southern California Keck School of Medicine (MO1-RR-00043), and by the Veterans Affairs institutions in Phoenix and San Diego, which contributed their resources and the use of their facilities. Dr. Banerji reports receiving consulting fees from BMS, Novartis, Boehringer Ingelheim, Sanofi-Aventis, Merck, and Roche, and lecture fees from Merck and Sanofi-Aventis; Dr. Buchanan reports receiving consulting fees and lecture fees from Takeda and reports that the University of Southern California Keck

nejm.org

march 24, 2011

1113

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

School of Medicine has received grant support from Takeda; Dr. DeFronzo reports receiving payments for board membership from Amylin, Takeda, ISIS, and Boehringer Ingelheim and reports that the University of Texas Health Science Center at San Antonio has received grant support from Takeda, Amylin, and Eli Lilly; Dr. Henry reports receiving consulting fees, lecture fees, and payment for expert testimony from Takeda; Dr. Musi reports receiving consulting fees from Merck, Daiichi-Sankyo, Takeda, and Novartis; Dr. Ratner reports that the Medstar Research Institute has received consulting fees from Amylin, NovoNordisk, Sanofi-Aventis, and Genentech–Roche and grant support from Amylin, NovoNordisk, GlaxoSmithKline, Bayhill, Halozyme, and Integrium; Dr. Reaven reports receiving consulting fees from BMS, lecture fees from Merck, and payment for the devel-

of

m e dic i n e

opment of educational presentations from Amylin; Dr. Reaven reports that the Carl T. Hayden Veterans Affairs Medical Center has received grant support from Amylin; and Dr. Tripathy reports receiving grant support from Takeda Pharmaceuticals. 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. We thank our nurses and other technical staff for their expert help, without whom this study would not have been possible; the 602 patients who participated in this study; and Joel Michalek, Ph.D., and Lee Ann Zarzabal, Ph.D., of the Department of Epidemiology and Biostatistics at the University of Texas Health Science Center at San Antonio for help in the initial statistical analyses.

References 1. Cowie CC, Rust KF, Byrd-Holt DD, et

al. Prevalence of diabetes and impaired fasting glucose in adults in the U.S. population: National Health And Nutrition Examination Survey 1999-2002. Diabetes Care 2006;29:1263-8. 2. Diabetes programme: facts and figures. Geneva: World Health Organization, 2007. (http://www.who.int/diabetes/facts/ en.) 3. Stratton IM, Adler AI, Neil HA, et al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. BMJ 2000;321:405-12. 4. DeFronzo RA. Banting Lecture: from the triumvirate to the ominous octet: a new paradigm for the treatment of type 2 diabetes mellitus. Diabetes 2009;58:773-95. 5. Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-Year follow-up of intensive glucose control in type 2 diabetes. N Engl J Med 2008;359:1577-89. 6. Nathan DM, Cleary PA, Backlund JY, et al. Intensive diabetes treatment and cardiovascular disease in patients with type 2 diabetes. N Engl J Med 2005;353: 2643-53. 7. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002;346:393403. 8. Xiang AH, Peters RK, Kjos SL, et al. Effect of pioglitazone on pancreatic betacell function and diabetes risk in Hispanic women with prior gestational diabetes. Diabetes 2006;55:517-22. 9. Gerstein HC, Yusuf S, Bosch J, et al. Effect of rosiglitazone on the frequency of diabetes in patients with impaired glucose tolerance or impaired fasting glucose: a randomised controlled trial. Lancet 2006;368:1096-105. [Erratum, Lancet 2006;368:1770.] 10. Knowler WC, Hamman RF, Edelstein SL, et al. Prevention of type 2 diabetes with troglitazone in the Diabetes Prevention Program. Diabetes 2005;54:1150-6. 11. Buchanan TA, Xiang AH, Peters RK, et al. Preservation of pancreatic beta-cell function and prevention of type 2 diabetes by pharmacological treatment of insu-

1114

lin resistance in high-risk Hispanic women. Diabetes 2002;51:2796-803. 12. Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M. Acarbose for prevention of type 2 diabetes mellitus: the STOP-NIDDM randomised trial. Lancet 2002;359:2072-7. 13. Buchwald H, Estok R, Fahrbach K, et al. Weight and type 2 diabetes after bariatric surgery: systematic review and meta-analysis. Am J Med 2009;122(3):248. e5-256.e5. 14. Nissen SE, Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med 2007;356:2457-7. [Erratum, N Engl J Med 2007;357:100.] 15. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2008;31:Suppl:S55-S60. 16. DeFronzo RA, Banerji M, Bray GA, et al. Actos Now for the prevention of diabetes (ACT NOW) study. BMC Endocr Disord 2009;9:17. 17. Edelstein SL, Knowler WC, Bain RP, et al. Predictors of progression from impaired glucose tolerance to NIDDM: an analysis of six prospective studies. Diabetes 1997;46:701-10. 18. Hodis HN, Mack WJ, LaBree L, et al. The role of carotid arterial intima-media thickness in predicting clinical coronary events. Ann Intern Med 1998;128:262-9. 19. Hodis HN, Mack WJ, Lobo RA, et al. Estrogen in the prevention of atherosclerosis: a randomized, double-blind, placebo-controlled trial. Ann Intern Med 2001; 135:939-53. 20. Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 1999;22:1462-70. 21. Bergman RN, Phillips LS, Cobelli C. Physiologic evaluation of factors controlling glucose tolerance in man: measurement of insulin sensitivity and beta-cell glucose sensitivity from the response to intravenous glucose. J Clin Invest 1981; 68:1456-67. 22. Abdul-Ghani MA, Jenkinson CP, Richardson DK, Tripathy D, DeFronzo RA. Insulin secretion and action in subjects with

n engl j med 364;12

nejm.org

impaired fasting glucose and impaired glucose tolerance: results from the Veterans Administration Genetic Epidemiology Study. Diabetes 2006;55:1430-5. 23. Kalbfleisch J, Prentice RL. The statistical analysis of failure time data. New York: John Wiley, 1980. 24. Lachin JM, Wei LJ. Estimators and tests in the analysis of multiple nonindependent 2 x 2 tables with partially missing observations. Biometrics 1988;44:51328. [Erratum, Biometrics 1988;44:923.] 25. Ferrannini E, Gastaldelli A, Miyazaki Y, Matsuda M, Mari A, DeFronzo RA. Beta-cell function in subjects spanning the range from normal glucose tolerance to overt diabetes: a new analysis. J Clin Endocrinol Metab 2005;90:493-500. 26. Gastaldelli A, Ferrannini E, Miyazaki Y, Matsuda M, Mari A, DeFronzo RA. Thiazolidinediones improve beta-cell function in type 2 diabetic patients. Am J Physiol Endocrinol Metab 2007;292:E871-E883. 27. Butler AE, Janson J, Bonner-Weir S, Ritzel R, Rizza RA, Butler PC. Beta-cell deficit and increased beta-cell apoptosis in humans with type 2 diabetes. Diabetes 2003;52:102-10. 28. Diabetes Prevention Program Research Group. The prevalence of retinopathy in impaired glucose tolerance and recent-onset diabetes in the Diabetes Prevention Program. Diabet Med 2007; 24:137-44. 29. Cheng YJ, Gregg EW, Geiss LS, et al. Association of A1C and fasting plasma glucose levels with diabetic retinopathy prevalence in the U.S. population: implications for diabetes diagnostic thresholds. Diabetes Care 2009;32:2027-32. 30. Ziegler D, Rathmann W, Dickhaus T, Meisinger C, Mielck A. Prevalence of polyneuropathy in pre-diabetes and diabetes is associated with abdominal obesity and macroangiopathy: the MONICA/KORA Augsburg Surveys S2 and S3. Diabetes Care 2008;31:464-9. 31. Pan XR, Li GW, Hu YH, et al. Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance: the Da Qing IGT and Diabetes Study. Diabetes Care 1997;20:537-44. 32. Li G, Zhang P, Wang J, et al. The long-

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Pioglitazone for Diabetes Prevention in IGT term effect of lifestyle interventions to prevent diabetes in the China Da Qing Diabetes Prevention Study: a 20-year follow-up study. Lancet 2008;371:1783-9. 33. Knowler WC, Fowler SE, Hamman RF, et al. 10-Year follow-up of diabetes incidence and weight loss in the Diabetes Prevention Program Outcomes Study. Lancet 2009;374:1677-86. [Erratum, Lancet 2009;374:2054.] 34. Lindström J, Ilanne-Parikka P, Peltonen M, et al. Sustained reduction in the incidence of type 2 diabetes by lifestyle intervention: follow-up of the Finnish Diabetes Prevention Study. Lancet 2006;368: 1673-9. 35. Miyazaki Y, Mahankali A, Matsuda M,

et al. Improved glycemic control and enhanced insulin sensitivity in type 2 diabetic subjects treated with pioglitazone. Diabetes Care 2001;24:710-9. 36. Miyazaki Y, De Filippis E, Bajaj M, et al. Predictors of improved glycemic control with rosiglitazone therapy in type 2 diabetic patients: a practical approach for the primary care physician. Br J Diabetes Vasc Dis 2005;5:28-35. 37. Dormandy JA, Charbonnel B, Eckland DJ, et al. Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive Study (PROspective pioglitAzone Clinical Trial In macroVascular Events): a randomised controlled trial. Lancet 2005;366:1279-89.

n engl j med 364;12

nejm.org

38. Crouse JR III. Thematic review series:

patient-oriented research — imaging atherosclerosis: state of the art. J Lipid Res 2006;47:1677-99. 39. Nissen SE, Nicholls SJ, Wolski K, et al. Comparison of pioglitazone vs glimepiride on progression of coronary atherosclerosis in patients with type 2 diabetes: the PERISCOPE randomized controlled trial. JAMA 2008;299:1561-73. 40. Mazzone T, Meyer PM, Feinstein SB, et al. Effect of pioglitazone compared with glimepiride on carotid intima-media thickness in type 2 diabetes: a randomized trial. JAMA 2006;296:2572-81. Copyright © 2011 Massachusetts Medical Society.

march 24, 2011

1115

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

original article

Mercury Exposure and Risk of Cardiovascular Disease in Two U.S. Cohorts Dariush Mozaffarian, M.D., Dr.P.H., Peilin Shi, Ph.D., J. Steven Morris, Ph.D., Donna Spiegelman, Sc.D., Philippe Grandjean, M.D., David S. Siscovick, M.D., M.P.H., Walter C. Willett, M.D., Dr.P.H., and Eric B. Rimm, Sc.D.

A BS T R AC T Background From the Division of Cardiovascular Medicine (D.M.) and Channing Laboratory (D.M., W.C.W., E.B.R.), Brigham and Women’s Hospital and Harvard Medical School; and the Departments of Epidemiology (D.M, P.S., D.S., W.C.W., E.B.R.), Nutrition (D.M., W.C.W., E.B.R.), Biostatistics (D.S.), and Environmental Health (P.G.), Harvard School of Public Health — all in Boston; the University of Missouri Research Reactor, Columbia (J.S.M.); and the Cardiovascular Health Research Unit, Departments of Medicine and Epidemiology, University of Washington, Seattle (D.S.S.). Address reprint requests to Dr. Mozaffarian, 665 Huntington Ave., Bldg. 2-319, Boston, MA 02115, or at [email protected]. N Engl J Med 2011;364:1116-25.

Exposure to methylmercury from fish consumption has been linked to a potentially increased risk of cardiovascular disease, but evidence from prior studies is equivocal. Beneficial effects of the ingestion of fish and selenium may also modify such effects. Methods

Among subjects from two U.S. cohorts (a total of 51,529 men and 121,700 women) whose toenail clippings had been stored, we prospectively identified incident cases of cardiovascular disease (coronary heart disease and stroke) in 3427 participants and matched them to risk-set–sampled controls according to age, sex, race, and smoking status. Toenail mercury and selenium concentrations were assessed with the use of neutron-activation analysis. Other demographic characteristics, cardiovascular risk factors, fish consumption, and lifestyle habits were assessed by means of validated questionnaires. Associations between mercury exposure and incident cardiovascular disease were evaluated with the use of conditional logistic regression.

Copyright © 2011 Massachusetts Medical Society.

Results

Median toenail mercury concentrations were 0.23 µg per gram (interdecile range, 0.06 to 0.94) in the case participants and 0.25 µg per gram (interdecile range, 0.07 to 0.97) in the controls. In multivariate analyses, participants with higher mercury exposures did not have a higher risk of cardiovascular disease. For comparisons of the fifth quintile of mercury exposure with the first quintile, the relative risks were as follows: coronary heart disease, 0.85 (95% confidence interval [CI], 0.69 to 1.04; P = 0.10 for trend); stroke, 0.84 (95% CI, 0.62 to 1.14; P = 0.27 for trend); and total cardiovascular disease, 0.85 (95% CI, 0.72 to 1.01; P = 0.06 for trend). Findings were similar in analyses of participants with low selenium concentrations or low overall fish consumption and in several additional sensitivity analyses. Conclusions

We found no evidence of any clinically relevant adverse effects of mercury exposure on coronary heart disease, stroke, or total cardiovascular disease in U.S. adults at the exposure levels seen in this study. (Funded by the National Institutes of Health.)

1116

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Mercury Exposure and Risk of Cardiovascular Disease

C

ontroversy has arisen over the risks and benefits of fish consumption in adults. Fish intake is inversely associated with the risk of coronary heart disease, especially fatal coronary heart disease, and ischemic stroke.1 Fish are also the major source of exposure to methylmercury.2,3 Chronic, low-level methylmercury exposure appears to cause subtle but measurable neurodevelopmental delay in infants, and it is recommended that women of childbearing age, pregnant or nursing mothers, and infants and young children eat no more than two servings of fish per week and also limit their intake of selected species of fish that are especially high in methylmercury content.4 In adults, however, the main health concern is potential cardiovascular toxicity, as suggested by results of experiments in animals and limited studies in humans.2,5 Prior clinical studies of mercury exposure and cardiovascular diseases have been relatively small, and the results have been inconsistent.6-11 Thus, government agencies, the Institute of Medicine, and risk–benefit analyses have identified the effect of methylmercury exposure on cardiovascular disease as an important area of uncertainty that warrants further investigation, since current data are not sufficient to quantitatively or qualitatively determine the potential effects.1,12-15 We prospectively investigated the relationships between mercury exposure and incident cardiovascular disease in two large U.S. cohorts. Because the trace element selenium provides protection against mercury toxicity in some experimental models,1,2 we also evaluated selenium exposure as a potential effect modifier.

Me thods Population and Study Design

The designs of the Health Professionals Followup Study (HPFS) and Nurses’ Health Study (NHS) have been described previously.16,17 The HPFS is a prospective cohort study that enrolled 51,529 male U.S. health professionals 40 to 75 years of age in 1986. The NHS is a prospective cohort study that enrolled 121,700 female U.S. registered nurses 30 to 55 years of age in 1976. Participants in both cohorts are followed by means of biennial questionnaires on medical history, risk factors, lifestyle, and disease incidence. We performed a nested case–control study involving participants from both cohorts. The study

n engl j med 364;12

was designed by the authors and approved by the human subjects committees of all participating institutions. In prior analyses,18-22 we found that concentrations of mercury and selenium in toenails are excellent biomarkers of usual methylmercury and selenium exposure. Toenail clippings were provided by 68% of HPFS participants in 1987 and by 52% of NHS participants during the period from 1982 through 1984. Demographic, risk-factor, and lifestyle characteristics of these participants were similar to those of participants who did not provide clippings (data not shown). About two thirds of the HPFS participants were dentists, and they were excluded from this analysis owing to occupational exposure to inorganic mercury during dental-amalgam procedures.18 All participants provided implied consent by returning completed questionnaires and toenail samples. Cases and Controls

Participants with incident cardiovascular disease (defined as nonfatal myocardial infarction, fatal coronary heart disease, or stroke) were identified from among HPFS and NHS participants who had provided toenail samples. Methods for ascertainment of cardiovascular events in the two cohorts have been described previously.16,17,23 When cardiovascular disease outcomes were reported, we obtained permission from participants (or from relatives in cases of fatal events) to review their medical records. Physicians who were unaware of other questionnaire information used standardized criteria to confirm and classify the events. Deaths were ascertained from relatives, postal authorities, and the National Death Index, and the cause of death was classified on the basis of medical records, death certificates, and autopsy findings. Permission to review medical records was granted for 95% of the requests. A diagnosis of myocardial infarction was confirmed on the basis of standardized criteria, which included typical symptoms plus either diagnostic electrocardiographic changes or elevated cardiac enzyme levels.24,25 Deaths were ascertained by contact with family members or through the National Death Index. Fatal heart disease was confirmed on the basis of medical records or autopsy reports or, if heart disease was listed as the cause of death, on the basis of the death certificates and evidence of previous heart disease in the records. Stroke was diagnosed according to

nejm.org

march 24, 2011

1117

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

standard criteria, consisting of a constellation of neurologic deficits of sudden or rapid onset that lasted at least 24 hours or until death.23,26 Stroke subtypes were also classified as previously described23,26 (see the Supplementary Appendix, available with the full text of this article at NEJM.org). For each case participant, a control participant was selected randomly from those with stored toenail samples who were free of cardiovascular disease at the time of the case event (risk-set sampling). Controls were matched one to one with case subjects according to age (within 1 year), sex (cohort-specific), race, smoking status (current smoker, former smoker [matched on number of years since stopping], or never smoked), and month when toenail sample was returned to us. Mercury and Selenium Exposures

Total mercury and selenium concentrations were assessed in the stored toenails by means of neutron-activation analysis (University of Missouri Research Reactor). Details of the analytic methods used and information regarding validation of these measures are provided in the Supplementary Appendix. Covariate Data Collection

Data on demographic characteristics, risk factors, and lifestyle habits were collected by means of validated, self-administered questionnaires, with the use of the closest preceding questionnaire administered before the collection of toenail samples from each participant. Smoking status was assessed, including the number of years since quitting in the case of former smokers. Hypertension and hypercholesterolemia were self-reported, with the validity of these reports confirmed on random sampling of medical records. A supplementary questionnaire was used to confirm self-reported cases of diabetes according to established criteria,27 and 98% of these cases were validated on comparison with medical records. Information on weight and height was obtained; self-reported weight was validated against technician-measured weight (r = 0.96). Physical activity was assessed in terms of metabolic equivalents (METs) with the use of validated questionaires.28 Usual dietary habits were assessed by means of validated semiquantitative food-frequency questionnaires that inquired about usual consumption of foods, beverages, and supplements during the previous year.29,30 1118

n engl j med 364;12

of

m e dic i n e

Statistical Analysis

Associations of mercury concentrations with incident cardiovascular disease were evaluated with the use of multivariate-adjusted conditional logistic regression. Given risk-set sampling, this model provides a direct estimation of the hazard ratio (hereafter referred to as relative risk). Mercury concentrations were evaluated in quintiles as indicator variables, with the use of sex-specific cutoff points among controls. Tests for trend involved assigning participants the median value in their quintile of exposure and evaluating this as a continuous variable. Tests for interaction involved multiplying this variable by the effect modifier of interest and using the Wald test to calculate the P value associated with the multiplicative interaction term. A potential nonlinear dose–response relationship was evaluated by visual inspection of relative risks across deciles of exposure. Analyses were performed separately in each cohort and then combined on the basis of the absence of significant effect modification (multiplicative interaction) by sex (P≥0.05). Power calculations are provided in the Supplementary Appendix. Potential confounding was assessed with the use of multivariate models adjusted for matching characteristics, other major risk factors for cardiovascular disease, fish or n−3 fatty acid consumption, and additional dietary factors associated with mercury concentrations. Multivariate modeling was guided by the principle of parsimony and by the clinical relevance of covariates, the observed strength of association between covariates and exposure or outcome, and the percent change in the risk estimate when covariates were included. Missing data for covariates (which accounted for less than 1% of all data) were imputed by means of multiple imputation.31 We performed prespecified sensitivity analyses to minimize potential misclassification due to exposure changes over time, restricting analyses to events within 10 years of toenail sampling and to participants with no substantial change in their fish consumption (i.e., a change of no more than two quintiles in either direction) from baseline to the end of follow-up. Stratified subgroup analyses were performed with the use of unconditional logistic regression adjusted for matching factors and other covariates. All reported P values are two-tailed, with values less than 0.05 indicating statistical significance. All analyses were performed with the use of SAS software, version 9.1 (SAS Institute).

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Mercury Exposure and Risk of Cardiovascular Disease

R e sult s Study Population

We identified 3427 participants with incident cases of cardiovascular disease: 1532 nonfatal myocardial infarctions, 831 fatal cases of coronary heart disease, and 1064 strokes. These case participants were matched with 3427 controls who had not had cardiovascular disease events during the same period of follow-up. The median followup interval from the time of toenail sampling to the time of a cardiovascular disease event was 11.3 years (interquartile range, 6.4 to 15.3); followup time was identical for controls, based on the risk-set sampling method. Baseline characteristics are shown in Table 1. As expected, cardiovascular risk factors were more prevalent among case participants than among controls at baseline. Approximately two thirds of the study participants were women, reflecting the larger size of the NHS cohort as compared with the HPFS cohort and the exclusion of dentists in the HPFS cohort from the analysis. Mean (±SD) ages were 61.1±9.0 years for men and 53.8±6.1 years for women. Median toenail mercury concentrations were 0.30 µg per gram (interdecile range, 0.07 to 1.26) in case participants and 0.31 µg per gram (interdecile range, 0.07 to 1.31) in controls among men and 0.21 µg per gram (interdecile range, 0.06 to 0.77) in case participants and 0.23 µg per gram (interdecile range, 0.07 to 0.76) in controls among women. Mercury Exposure and Cardiovascular Risk Factors

Mercury concentrations correlated modestly with fish consumption (r = 0.39, P<0.001) and with estimated dietary intake of eicosapentaenoic acid and docosahexaenoic acid (EPA–DHA) (r = 0.39, P<0.001), as expected, given the predominance of seafood as a source of methylmercury exposure but also given the considerable variation in methylmercury and n−3 fatty acid content among fish species.1,3 Concentrations of mercury did not correlate with those of selenium (r = 0.03), a finding that is consistent with the multiple, varied dietary sources of selenium. In bivariate (unadjusted) analyses at baseline among the controls, higher mercury concentrations were associated with a more frequent prevalence of hypercholesterolemia, slightly lower bodymass index, modestly higher levels of physical activity, greater alcohol use, and lower total enn engl j med 364;12

ergy intake (Table 1 in the Supplementary Appendix). Mercury concentrations were also positively associated with dietary factors related to fish consumption and higher dietary intake of EPA– DHA, including slightly lower intakes of saturated fat, monounsaturated fat, trans fat, and dietary cholesterol and slightly higher intakes of protein and polyunsaturated fat. Mercury concentrations were not significantly associated with age, smoking status, family history, or presence or absence of hypertension or diabetes. Mercury Exposure and Cardiovascular Events

After adjustment for matching factors, participants with higher mercury exposure did not have a higher risk of cardiovascular events (Table 2). In fact, those with higher mercury concentrations had a lower incidence of coronary heart disease (P = 0.006 for trend), stroke (P = 0.09 for trend), and total cardiovascular disease (P = 0.002 for trend). These inverse associations were not significant after further adjustment for other cardiovascular disease risk factors plus estimated dietary EPA–DHA (Table 2). Further adjustment for consumption of saturated fat, monounsaturated fat, polyunsaturated fat, trans fat, dietary cholesterol, and total energy had little effect on the results: the adjusted relative risks for comparison of the fifth quintile of mercury exposure with the first quintile (“extreme-quintile relative risks”) were 0.85 (95% confidence interval [CI], 0.69 to 1.06) for coronary heart disease, 0.83 (95% CI, 0.60 to 1.15) for stroke, and 0.87 (95% CI, 0.73 to 1.03) for total cardiovascular disease. Adjustment for fish consumption instead of dietary EPA–DHA also did not alter the findings (data not shown). The results were also similar for mercury concentrations evaluated in deciles (Table 2 in the Supplementary Appendix). In separate analyses according to sex, the trend toward a lower incidence of cardiovascular disease with higher mercury concentrations was seen for women but not for men (Table 3 in the Supplementary Appendix). Interaction tests for sex, however, were not significant (P = 0.12, P = 0.14, and P = 0.05 for tests of interaction with coronary heart disease, stroke, and total cardiovascular disease, respectively). When coronary heart disease subtypes were evaluated, mercury exposure was not associated with the risk of nonfatal myocardial infarction (extreme-quintile relative risk, 0.84 [95% CI, 0.65 to 1.08]; P = 0.10 for trend) or fatal coronary heart disease (extreme-quintile relative risk, 0.85

nejm.org

march 24, 2011

1119

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

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. Baseline Characteristics of Case Participants with Incident Cardiovascular Disease and of Controls.* Characteristic

Men

Women

Case Participants (N = 1211)

Controls (N = 1211)

P Value

Case Participants (N = 2216)

Controls (N = 2216)

P Value

61.1±9.0

61.1±9.0

0.96

53.8±6.1

53.8±6.1

0.86

Never smoked

40.3

42.4

0.30

35.5

35.5

1.00

Age (yr)† Smoking status (%)† Former smoker

44.7

45.9

0.54

25.2

25.7

0.70

Current smoker

11.6

10.5

0.36

39.3

38.8

0.74

Family history of MI (%)

39.0

34.1

0.01

27.4

20.6

<0.001

Hypertension (%)

36.9

21.5

<0.001

13.5

8.1

<0.001

Hypercholesterolemia (%)

13.4

12.1

0.33

6.6

4.2

<0.001

7.0

3.4

<0.001

3.0

0.5

<0.001

Diabetes mellitus (%) Body-mass index‡

26.3±3.3

25.5±3.0

0.89

25.9±5.7

24.6±4.7

Physical activity (METS/wk)

15.8±21.3

19.4±26.4

<0.001

11.7±16.2

13.5±18.6

0.001

0.8±1.2

0.9±1.2

0.08

0.5±0.9

0.6±0.9

0.03

Alcohol (drinks/wk)

<0.001

Toenail selenium (µg/g)

0.92±0.61

0.92±0.6

0.99

0.78±0.22

0.78±0.25

0.34

Toenail mercury (µg/g)

0.51±2.13

0.44±0.47

0.24

0.29±0.49

0.33±0.63

0.04

2.1±1.9

2.1±1.8

0.89

1.8±1.6

1.8±1.6

0.65

Dietary intake Fish (servings/wk) EPA and DHA (mg/wk) Total energy intake (kcal/day)

270±239

264±220

0.49

184±162

184±151

0.89

2024±623

2063±640

0.13

1742±536

1727±530

0.38

Fat (% energy) Total

32.5±6.4

32.6±6.3

0.72

34.8±6.4

34.6±6.4

0.22

Saturated

11.3±2.9

11.3±2.8

0.85

12.7±3.1

12.6±3.0

0.05

Monounsaturated

12.5±2.8

12.5±2.7

0.69

12.9±2.9

12.8±2.9

0.16

5.8±1.6

5.8±1.5

0.42

6.3±1.8

6.4±1.8

0.14

Polyunsaturated Trans Protein (% energy)

1.3±0.5

1.3±0.5

0.78

1.9±0.7

1.9±0.6

0.12

18.3±3.4

18.3±3.3

0.97

18.0±3.6

17.9±3.4

0.48

Cholesterol (mg/day)

314±153

320±159

0.32

312±138

308±141

0.40

Whole grains (g/day)

20.5±19.2

20.8±18.0

0.74

15.3±15.9

15.8±13.7

0.28

* Plus–minus values are means ±SD. DHA denotes docosahexaenoic acid, EPA eicosapentaenoic acid, METS metabolic equivalents, and MI myocardial infarction. † Age and smoking status were matching factors. ‡ The body-mass index is the weight in kilograms divided by the square of the height in meters.

[95% CI, 0.59 to 1.24]; P = 0.41 for trend). Mercury exposure was also not associated with the risk of any stroke subtype (see the Supplementary Appendix). Sensitivity Analyses

Because selenium above a threshold of risk may provide protection against some forms of mercury toxicity, we restricted analyses to participants with lower selenium concentrations. Mercury ex1120

n engl j med 364;12

posure was not associated with a higher risk of total cardiovascular disease, coronary heart disease, or stroke among participants with selenium levels in the lowest quartile (<0.70 µg per gram) or the lowest decile (<0.64 µg per gram) (Table 3). Mercury exposure was also not associated with a higher risk in analyses stratified according to fish consumption (Table 4). Results were also similar in analyses stratified according to the presence or absence of hypertension, high

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Mercury Exposure and Risk of Cardiovascular Disease

Table 2. Relative Risk of Cardiovascular Disease, According to Quintile of Toenail Mercury, Among Case Participants and Matched Controls in Two Prospective Cohorts of Men and Women.* No. of Case Participants

Variable

P Value for Trend

Sex-Specific Quintile of Toenail Mercury 1

2

3

4

5

0.09

0.17

0.25

0.38

0.95

542

506

446

450

419

Model 1†

1.00 (reference)

0.97 (0.81–1.15)

0.82 (0.69–1.00)

0.81 (0.68 – 0.97)

0.78 (0.65–0.94)

0.006

Model 2‡

1.00 (reference)

1.00 (0.83–1.20)

0.89 (0.73–1.08)

0.87 (0.72–1.06)

0.85 (0.69–1.04)

0.10

233

226

209

209

187

Model 1†

1.00 (reference)

0.91 (0.70–1.19)

0.89 (0.68–1.17)

0.94 (0.72–1.23)

0.77 (0.59–1.02)

0.09

Model 2‡

1.00 (reference)

0.95 (0.72–1.26)

0.95 (0.71–1.28)

0.98 (0.73–1.31)

0.84 (0.62–1.14)

0.27

775

732

655

659

606

Model 1†

1.00 (reference)

0.95 (0.82–1.10)

0.84 (0.73–0.98)

0.85 (0.74–0.99)

0.78 (0.67–0.91)

0.002

Model 2‡

1.00 (reference)

0.98 (0.84–1.15)

0.91 (0.77–1.07)

0.91 (0.77–1.07)

0.85 (0.72–1.01)

0.06

Mean mercury (µg/g) Coronary heart disease

2363

No. of cases Multivariate RR (95% CI)

Stroke

1064

No. of cases Multivariate RR (95% CI)

Total cardiovascular disease

3427

No. of cases Multivariate RR (95% CI)

* Values for quintiles represent mean mercury levels. Quintiles were not constructed with the data from men and women combined but were sex-specific, and the relative risks (RR) for each were then combined. CI denotes confidence interval. † In Model 1, the RR is based on conditional logistic regression with risk-set sampling, in which the odds ratio directly estimates the hazard ratio or RR, with matching factors of age, sex, race, month of toenail receipt, and smoking status (never smoked, former smoker, or current smoker). ‡ In Model 2, the RR was further adjusted for body-mass index (quintiles), physical activity (metabolic equivalents per week, quintiles), alcohol intake (drinks per week, quintiles), diabetes (yes or no), hypertension (yes or no), elevated cholesterol level (yes or no), and estimated dietary intake of eicosapentaenoic acid and docosahexaenoic acid (mg per week, quintiles).

cholesterol, or diabetes or, among women, use or nonuse of hormone-replacement therapy (data not shown). The results of additional sensitivity analyses are provided in the Supplementary Appendix.

Discussion In our study, mercury exposure as assessed by an objective biomarker measurement was not associated with an increased risk of cardiovascular disease among men or women in two separate U.S. cohorts. An increased risk with greater mercury exposure was also not evident among participants with lower selenium concentrations, in n engl j med 364;12

analyses restricted to the first 10 years of followup and analyses stratified according to the duration of follow-up, or in analyses restricted to those participants without substantial changes in fish consumption over time and analyses stratified according to the level of fish consumption. These findings provide no support for clinically relevant adverse effects of typical levels of dietary methylmercury exposure on cardiovascular disease in U.S. adults. Higher mercury exposures were actually associated with trends toward lower cardiovascular disease risk, although these trends were not significant in the fully adjusted models. To our

nejm.org

march 24, 2011

1121

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

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. Odds Ratios for Cardiovascular Disease (CVD) According to Quintile of Toenail Mercury in Case Participants with Lower Selenium Levels, for Men and Women Combined from Two Prospective Cohorts. No. of Case Participants

Variable

P Value for Trend

Sex-Specific Quintile of Toenail Mercury* 1

2

3

4

5

odds ratio (95% confidence interval) Subjects in lowest quartile of selenium levels† Coronary heart disease

631

1.00 (reference)

0.94 (0.65–1.37)

0.72 (0.50–1.05)

0.71 (0.48–1.05)

0.84 (0.55–1.27)

0.46

Stroke

254

1.00 (reference)

0.70 (0.39–1.27)

0.88 (0.49–1.57)

0.59 (0.31–1.12)

0.40 (0.20–0.79)

0.006

Total CVD

885

1.00 (reference)

0.87 (0.64–1.18)

0.78 (0.58–1.07)

0.70 (0.50–0.96)

0.68 (0.48–0.96)

0.03

Coronary heart disease

242

1.00 (reference)

0.99 (0.54–1.81)

0.74 (0.40–1.36)

0.77 (0.40–1.48)

0.79 (0.40–1.57)

0.49

Stroke

111

1.00 (reference)

1.02 (0.39–2.69)

1.02 (0.40–2.54)

0.81 (0.28–2.32)

0.62 (0.38–1.17)

0.30

Total CVD

353

1.00 (reference)

0.94 (0.57–1.55)

0.80 (0.49–1.30)

0.78 (0.46–1.34)

0.67 (0.38–1.17)

0.14

Subjects in lowest decile of selenium levels‡

* Quintile cutoff points are based on the overall control population (see Table 3 in the Supplementary Appendix). An unconditional logisticregression model was used, as appropriate, for stratified subgroup analyses. Values were adjusted for age, sex, race, month of toenail receipt, smoking status (never smoked, former smoker, or current smoker), body-mass index (quintiles), physical activity (metabolic equivalents per week, quintiles), alcohol use (drinks per week, quintiles), diabetes (yes or no), hypertension (yes or no), elevated cholesterol level (yes or no), and estimated dietary intake of eicosapentaenoic acid and docosahexaenoic acid (mg per week, quintiles). † These subjects had selenium values below 0.70 µg per gram. ‡ These subjects had selenium values below 0.64 µg per gram.

knowledge, there is no biologic explanation for why mercury would induce cardiovascular benefits. These results plausibly reflect the extent to which mercury levels are an indirect, but nonetheless objective, biomarker of fish consumption and its correlates and thus probably provide independent information on how much fish a person consumes, even after adjustment for estimated consumption. Trends toward lower risk with higher mercury exposure appeared to be confined to women, but this sex difference was not significant and is probably due to chance. Trends toward lower cardiovascular disease risk with higher mercury levels have also been seen in some prior studies.7,11 Of six prior studies of the relationship between mercury exposure and cardiovascular disease,6-11 only two showed positive 1122

n engl j med 364;12

associations.6,7 The largest study (684 cases) included only nonfatal myocardial infarction and was retrospective,6 raising concern about possible selection bias. A smaller, prospective study (282 cases) showed a positive association with total coronary events but without a clear dose– response relationship or significant associations with coronary or cardiovascular mortality.7 The remaining four studies were prospective and did not show significant associations; however, they included participants with occupational exposure to mercury vapor,8 the health effects of which differ from those of methylmercury12; they assessed erythrocyte mercury levels, which reflect a more recent exposure than do toenail or hair concentrations9; or they had small numbers of cases (<100).10,11 Several of the prior studies also

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Mercury Exposure and Risk of Cardiovascular Disease

Table 4. Odds Ratios for Total Cardiovascular Disease, According to Quintile of Toenail Mercury and Stratum of Fish Consumption, for Men and Women Combined from Two Prospective Cohorts. Fish Consumption*

No. of Case Participants

P Value for Trend

Quintile of Toenail Mercury† 1

2

3

4

5

odds ratio (95% confidence interval) Total <1 serving/wk

1500

1.00 (reference)

0.99 (0.81–1.21)

0.91 (0.73–1.13)

0.80 (0.63–1.01)

0.90 (0.69–1.18)

0.20

1 to <2 servings/wk

992

1.00 (reference)

0.85 (0.63–1.14)

0.98 (0.73–1.32)

0.84 (0.62–1.13)

0.74 (0.54–1.02)

0.07

≥2 servings/wk

935

1.00 (reference)

1.11 (0.74–1.66)

0.79 (0.54–1.17)

1.17 (0.80–1.70)

0.96 (0.66–1.39)

0.86

2475

1.00 (reference)

1.00 (0.85–1.18)

0.93 (0.78–1.10)

0.88 (0.73–1.05)

0.93 (0.76–1.13)

0.32

1 to <2 servings/wk

483

1.00 (reference)

0.72 (0.42–1.22)

0.69 (0.42–1.15)

0.89 (0.54–1.47)

0.58 (0.35–0.95)

0.08

≥2 servings/wk

469

1.00 (reference)

0.98 (0.54–1.79)

0.81 (0.46–1.44)

0.97 (0.56–1.69)

0.81 (0.47–1.39)

0.38

2121

1.00 (reference)

0.99 (0.83–1.19)

0.92 (0.76–1.11)

0.87 (0.71–1.06)

0.86 (0.69–1.06)

0.10

0.5 to <1 servings/wk

932

1.00 (reference)

0.91 (0.66–1.25)

0.90 (0.65–1.24)

0.88 (0.64–1.22)

0.74 (0.54–1.03)

0.06

≥1 servings/wk

374

1.00 (reference)

1.05 (0.54–2.64)

0.88 (0.47–1.63)

1.29 (0.71–2.35)

1.38 (0.76–2.48)

0.08

Tuna or other dark-meat fish <1 serving/wk

Other fish <0.5 serving/wk

* Total fish consumption was the sum of the consumption of tuna or other dark-meat fish and the consumption of other fish. Strata were set at logical cutoff points that provided reasonable numbers of cases per stratum. † Quintile cutoff points are based on the overall control population (see Table 3 in the Supplementary Appendix). Thus, in every stratum of fish consumption, higher quintiles reflect subjects with similarly high mercury exposure. In the case of low fish consumption (e.g., <1 serving per week), higher quintiles would be consistent with more exclusive consumption of relatively contaminated fish (i.e., similar methylmercury exposure from fewer fish meals, indicating a greater proportion of more contaminated fish in the diet). Values are based on unconditional logistic regression, as appropriate, for stratified subgroup analyses and have been adjusted for age, sex, race, month of toenail receipt, smoking status (never, former, or current), body-mass index (quintiles), physical activity (metabolic equivalents per week, quintiles), alcohol (drinks per week, quintiles), diabetes (yes or no), hypertension (yes or no), elevated cholesterol (yes or no), and estimated dietary intake of eicosapentaenoic acid and docosahexaenoic acid (mg per week, quintiles). See Tables 5 and 6 in the Supplementary Appendix for stratified results for coronary heart disease and stroke, which were evaluated separately.

did not evaluate stroke6-8,11 or include women.6-8 The investigation we describe here was designed to overcome these limitations. With respect to generalizability, it is important to consider how mercury exposures in the present study compare with those in prior studies and with average population exposures. In our highest exposure quintile, the median toenail mercury concentration was 0.68 µg per gram, and in our highest decile, 1.00 µg per gram, corresponding to hair concentrations of about 1.84 and 2.70 µg per gram, respectively, calculated from a reported toenail-to-hair ratio of n engl j med 364;12

mercury of about 0.37.32-35 These exposure levels are similar to those seen in two smaller studies, in which mercury levels were positively associated with coronary heart disease risk,6,7 and are also similar to higher U.S. exposures (in the 95th percentile).36 Differences in population selenium levels have been hypothesized to explain discrepant findings of prior studies with respect to mercury and cardiovascular risk — in particular, a study from Finland.7 Before soil supplementation was begun in the 1980s, selenium levels in Finland were among the lowest in Europe (mean serum level,

nejm.org

march 24, 2011

1123

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

<70 µg per liter).37 In the Finnish mercury study, average serum selenium levels at baseline (from 1984 through 1989, after soil supplementation began) were higher (117 µg per liter)7 but still below average U.S. levels (138 µg per liter).38 In our study, we found no evidence of an increased risk with higher mercury levels, even among participants with selenium levels in the lowest decile (<0.64 µg per gram in toenails, approximately equivalent to <91 µg per liter in serum39). We also found no evidence that mercury was harmful among participants in different strata of fish consumption, including those with low fish consumption, in whom higher mercury levels would suggest more exclusive consumption of mercury-contaminated fish. Our analysis cannot exclude the possibility of mercury-related cardiovascular toxicity at higher exposures than those observed in our cohorts or in the setting of frank selenium deficiency, which would be rare in U.S. cohorts. Ecologic or small cross-sectional studies in more highly exposed populations in the Amazon,40 the Faroe Islands,32 and Asia41,42 suggest that methylmercury exposure may be associated with higher blood pressure or lower parasympathetic activity; ecologic evidence of an increased risk of clinical cardiovascular events is lacking.43 Our analysis has potential limitations. Although toenail concentrations of mercury provide an excellent biomarker of integrated, usual methylmercury exposure during the previous year, changes in dietary exposure over time could attenuate true relationships toward null. Toenail mercury concentration serves as a marker of fish consumption, and our findings may be partly confounded by the beneficial effects of fish intake, despite adjustment for responses to the dietary questionnaire; this might account for

of

m e dic i n e

trends toward lower risk. Although the findings were similar in the two independent cohorts and there is little reason to believe that biologic effects of methylmercury in these populations would be different from those in the general population of women and men, these cohorts comprised largely white, educated U.S. adults, potentially limiting generalizability. The absence of any association between mercury exposure and increased cardiovascular disease risk in adults should not alter ongoing public health and policy efforts to reduce mercury contamination in fish and the environment, which could still have the potential to offset, at least in part, the net cardiovascular benefits of fish consumption. Our findings should also not alter advisories directed toward women who are or may become pregnant or who are nursing, since methylmercury exposure from consumption of specific fish species could cause neurodevelopmental harm, or at least partly offset the neurodevelopmental benefits of fish consumption, in their children. In summary, this prospective study of two large cohorts of men and women in the United States showed no evidence of a relationship between mercury exposure and increased cardiovascular disease risk. Supported by grants from the National Institute of Environmental Health Sciences (R01-ES014433 and ES013692), the National Heart, Lung, and Blood Institute (HL34594, HL088521, and HL35464), and the National Cancer Institute (CA87969 and CA55075). Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. We thank the NHS and HPFS participants and our expert advisory panel for helpful comments on this research, including David W.K. Acheson, Managing Director, Food and Import Safety, Leavitt Partners (formerly associate commissioner for foods, Food and Drug Administration), and Rita Schoeny, senior science advisor, Office of Water, Environmental Protection Agency.

References 1. Mozaffarian D, Rimm EB. Fish in-

take, contaminants, and human health: evaluating the risks and the benefits. JAMA 2006;296:1885-99. 2. Mozaffarian D. Fish, mercury, selenium and cardiovascular risk: current evidence and unanswered questions. Int J Environ Res Public Health 2009;6:1894916. 3. Groth E III. Ranking the contributions of commercial fish and shellfish varieties to mercury exposure in the United States: implications for risk communication. Environ Res 2010;110:226-36.

1124

4. What you need to know about mercury

in fish and shellfish: 2004 EPA and FDA advice for women who might become pregnant, women who are pregnant, nursing mothers, young children. Washington, DC: Food and Drug Administration, 2004. (http://www.cfsan.fda.gov/~dms/ admehg3.html.) 5. Yaginuma-Sakurai K, Murata K, Shimada M, et al. Intervention study on cardiac autonomic nervous effects of methylmercury from seafood. Neurotoxicol Teratol 2010;32:240-5. 6. Guallar E, Sanz-Gallardo MI, van’t

n engl j med 364;12

nejm.org

Veer P, et al. Mercury, fish oils, and the risk of myocardial infarction. N Engl J Med 2002;347:1747-54. 7. Virtanen JK, Voutilainen S, Rissanen TH, et al. Mercury, fish oils, and risk of acute coronary events and cardiovascular disease, coronary heart disease, and allcause mortality in men in eastern Finland. Arterioscler Thromb Vasc Biol 2005; 25:228-33. 8. Yoshizawa K, Rimm EB, Morris JS, et al. Mercury and the risk of coronary heart disease in men. N Engl J Med 2002; 347:1755-60.

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Mercury Exposure and Risk of Cardiovascular Disease 9. Wennberg M, Bergdahl IA, Stegmayr

B, et al. Fish intake, mercury, long-chain n-3 polyunsaturated fatty acids and risk of stroke in northern Sweden. Br J Nutr 2007;98:1038-45. 10. Ahlqwist M, Bengtsson C, Lapidus L, Gergdahl IA, Schütz A. Serum mercury concentration in relation to survival, symptoms, and diseases: results from the prospective population study of women in Gothenburg, Sweden. Acta Odontol Scand 1999;57:168-74. 11. Hallgren CG, Hallmans G, Jansson JH, et al. Markers of high fish intake are associated with decreased risk of a first myocardial infarction. Br J Nutr 2001;86: 397-404. 12. Mercury Study Report to Congress. Washington, DC: Environmental Protection Agency, 1997. (http://www.epa.gov/ mercury/report.htm.) 13. Rice DC. The US EPA reference dose for methylmercury: sources of uncertainty. Environ Res 2004;95:406-13. 14. König A, Bouzan C, Cohen JT, et al. A quantitative analysis of fish consumption and coronary heart disease mortality. Am J Prev Med 2005;29:335-46. 15. Joint FAO/WHO expert consultation on the risks and benefits of fish consumption — executive summary. Geneva: Food and Agriculture Organization of the United Nations, World Health Organization, 2010. (http://www.who.int/foodsafety/chem/ meetings/jan2010/en/index.html.) 16. Hu FB, Bronner L, Willett WC, et al. Fish and omega-3 fatty acid intake and risk of coronary heart disease in women. JAMA 2002;287:1815-21. 17. Mozaffarian D, Ascherio A, Hu FB, et al. Interplay between different polyunsaturated fatty acids and risk of coronary heart disease in men. Circulation 2005; 111:157-64. 18. Joshi A, Douglass CW, Kim HD, et al. The relationship between amalgam restorations and mercury levels in male dentists and nondental health professionals. J Public Health Dent 2003;63:52-60. 19. Garland M, Morris JS, Rosner BA, et al. Toenail trace element levels as biomarkers: reproducibility over a 6-year period. Cancer Epidemiol Biomarkers Prev 1993; 2:493-7. 20. MacIntosh DL, Williams PL, Hunter DJ, et al. Evaluation of a food frequency questionnaire–food composition approach

for estimating dietary intake of inorganic arsenic and methylmercury. Cancer Epidemiol Biomarkers Prev 1997;6:1043-50. 21. Longnecker MP, Stampfer MJ, Morris JS, et al. A 1-y trial of the effect of highselenium bread on selenium concentrations in blood and toenails. Am J Clin Nutr 1993;57:408-13. 22. Longnecker MP, Stram DO, Taylor PR, et al. Use of selenium concentration in whole blood, serum, toenails, or urine as a surrogate measure of selenium intake. Epidemiology 1996;7:384-90. 23. Iso H, Rexrode KM, Stampfer MJ, et al. Intake of fish and omega-3 fatty acids and risk of stroke in women. JAMA 2001; 285:304-12. 24. Rose GA, Blackburn H. Cardiovascular survey methods. 2nd ed. World Health Organization monograph series no. 56. Geneva: World Health Organization, 1982. 25. Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarction redefined — a consensus document of The Joint European Society of Cardiology/ American College of Cardiology Committee for the redefinition of myocardial infarction. J Am Coll Cardiol 2000;36: 959-69. 26. Walker AE, Robins M, Weinfeld FD. The National Survey of Stroke: clinical findings. Stroke 1981;12:I13-I44. 27. National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979;28:1039-57. 28. Hu FB, Stampfer MJ, Colditz GA, et al. Physical activity and risk of stroke in women. JAMA 2000;283:2961-7. 29. Feskanich D, Rimm EB, Giovannucci EL, et al. Reproducibility and validity of food intake measurements from a semiquantitative food frequency questionnaire. J Am Diet Assoc 1993;93:790-6. 30. Salvini S, Hunter DJ, Sampson L, et al. Food-based validation of a dietary questionnaire: the effects of week-to-week variation in food consumption. Int J Epidemiol 1989;18:858-67. 31. Schafer JL. Analysis of incomplete multivariate data. New York: Chapman and Hall, 1997. 32. Choi AL, Weihe P, Budtz-Jørgensen E, et al. Methylmercury exposure and adverse cardiovascular effects in Faroese whaling men. Environ Health Perspect 2009;117:367-72.

n engl j med 364;12

nejm.org

33. Suzuki T, Watanabe S, Matsuo N.

Comparison of hair with nail as index media for biological monitoring of mercury. Sangyo Igaku 1989;31:235-8. 34. Morton J, Mason HJ, Ritchie KA, White M. Comparison of hair, nails and urine for biological monitoring of low level inorganic mercury exposure in dental workers. Biomarkers 2004;9:47-55. 35. Ohno T, Sakamoto M, Kurosawa T, Dakeishi M, Iwata T, Murata K. Total mercury levels in hair, toenail, and urine among women free from occupational exposure and their relations to renal tubular function. Environ Res 2007;103:191-7. 36. McDowell MA, Dillon CF, Osterloh J, et al. Hair mercury levels in U.S. children and women of childbearing age: reference range data from NHANES 1999-2000. Environ Health Perspect 2004;112:1165-71. 37. Varo P, Alfthan G, Ekholm P, Aro A, Koivistoinen P. Selenium intake and serum selenium in Finland: effects of soil fertilization with selenium. Am J Clin Nutr 1988;48:324-9. 38. Bleys J, Navas-Acien A, Laclaustra M, et al. Serum selenium and peripheral arterial disease: results from the National Health and Nutrition Examination Survey, 2003-2004. Am J Epidemiol 2009;169:9961003. 39. Mason MM, Morris JS, Spate VL, et al. Comparison of whole blood, plasma and nails as monitors for the dietary intake of selenium. J Radioanal Nucl Chem 1998; 236:29-34. 40. Fillion M, Mergler D, Sousa Passos CJ, Larribe F, Lemire M, Guimarães JR. A preliminary study of mercury exposure and blood pressure in the Brazilian Amazon. Environ Health 2006;5:29. 41. Yorifuji T, Tsuda T, Kashima S, Takao S, Harada M. Long-term exposure to methylmercury and its effects on hypertension in Minamata. Environ Res 2010; 110:40-6. 42. Lim S, Chung HU, Paek D. Low dose mercury and heart rate variability among community residents nearby to an industrial complex in Korea. Neurotoxicology 2010;31:10-6. 43. Chan HM, Egeland GM. Fish consumption, mercury exposure, and heart diseases. Nutr Rev 2004;62:68-72. Copyright © 2011 Massachusetts Medical Society.

march 24, 2011

1125

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

original article

Carbamazepine-Induced Toxic Effects and HLA-B*1502 Screening in Taiwan Pei Chen, Ph.D., Juei-Jueng Lin, M.D., Chin-Song Lu, M.D., Cheung-Ter Ong, M.D., Peiyuan F. Hsieh, M.D., Chih-Chao Yang, M.D., Chih-Ta Tai, M.D., Shey-Lin Wu, M.D., Cheng-Hsien Lu, M.D., Yung-Chu Hsu, M.D., Hsiang-Yu Yu, M.D., Long-Sun Ro, M.D., Chung-Ta Lu, M.D., Chun-Che Chu, M.D., Jing-Jane Tsai, M.D., Yu-Hsiang Su, M.D., Sheng-Hsing Lan, M.D., Sheng-Feng Sung, M.D., Shu-Yi Lin, M.S., Hui-Ping Chuang, B.S., Li-Chen Huang, B.S., Ying-Ju Chen, M.S., Pei-Joung Tsai, M.S., Hung-Ting Liao, M.S., Yu-Hsuan Lin, M.S., Chien-Hsiun Chen, Ph.D., Wen-Hung Chung, M.D., Ph.D., Shuen-Iu Hung, Ph.D., Jer-Yuarn Wu, Ph.D., Chi-Feng Chang, Ph.D., Luke Chen, Ph.D., Yuan-Tsong Chen, M.D., Ph.D., and Chen-Yang Shen, Ph.D., for the Taiwan SJS Consortium*

A bs t r ac t Background The authors’ affiliations are listed in the Appendix. Address reprint requests to Dr. Shen or Dr. Y.-T. Chen at the Institute of Biomedical Sciences, Academia Sinica, 128 Academia Rd., Section 2, Nankang, Taipei 115, Taiwan, or at bmcys@ibms .sinica.edu.tw or [email protected] .edu.tw. *Other members of the Taiwan Stevens– Johnson Syndrome (SJS) Consortium are listed in the Supplementary Appendix, available at NEJM.org. N Engl J Med 2011;364:1126-33. Copyright © 2011 Massachusetts Medical Society.

Carbamazepine, an anticonvulsant and a mood-stabilizing drug, is the main cause of the Stevens–Johnson syndrome (SJS) and its related disease, toxic epidermal necrolysis (TEN), in Southeast Asian countries. Carbamazepine-induced SJS–TEN is strongly associated with the HLA B*1502 allele. We sought to prevent carbamazepine-induced SJS–TEN by using HLA-B*1502 screening to prospectively identify subjects at genetic risk for the condition. Methods

From 23 hospitals in Taiwan, we recruited 4877 candidate subjects who had not taken carbamazepine. We genotyped DNA purified from the subjects’ peripheral blood to determine whether they carried the HLA-B*1502 allele. Those testing positive for HLA-B*1502 (7.7% of the total) were advised not to take carbamazepine and were given an alternative medication or advised to continue taking their prestudy medication; those testing negative (92.3%) were advised to take carbamazepine. We interviewed the subjects by telephone once a week for 2 months to monitor them for symptoms. We used the estimated historical incidence of SJS–TEN as a control. Results

Mild, transient rash developed in 4.3% of subjects; more widespread rash developed in 0.1% of subjects, who were hospitalized. SJS–TEN did not develop in any of the HLA-B*1502–negative subjects receiving carbamazepine. In contrast, the estimated historical incidence of carbamazepine-induced SJS–TEN (0.23%) would translate into approximately 10 cases among study subjects (P<0.001). Conclusions

The identification of subjects carrying the HLA-B*1502 allele and the avoidance of carbamazepine therapy in these subjects was strongly associated with a decrease in the incidence of carbamazepine-induced SJS–TEN. (Funded by the National Science Council of Taiwan and the Taiwan Drug Relief Foundation.) 1126

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

HLA-B*1502 Screening and Carbamazepine ther apy

T

he Stevens–Johnson syndrome (SJS) and its related disease, toxic epidermal necrolysis (TEN), are two of the most serious adverse reactions caused by drugs. SJS is characterized by high fever, malaise, and a rapidly developing, blistering exanthema of macular papules and target-like lesions, accompanied by mucosal involvement. This condition is associated with a rate of death of approximately 5%. TEN has a similar presentation, with even more extensive skin detachment and a death rate of 25 to 35%.1 Carbamazepine, an anticonvulsant and specific analgesic agent for trigeminal neuralgia, is the most common cause of SJS–TEN in Southeast Asian countries.2 We previously reported that carbamazepine-induced SJS–TEN is strongly associated with the HLA-B*1502 allele in Han Chinese populations.3 This association was subsequently confirmed in persons from Hong Kong, Malaysia, Thailand, and India and in descendants of immigrants from Southeast Asia, regions in which the HLA-B*1502 allele is prevalent.4-9 Another HLA allele, HLA-A*3101, has been associated with carbamazepine-induced hypersensitivity reactions. We have observed its association with the relatively mild maculopapular exanthema.10 More recently, a genomewide association study has shown an association between the HLAA*3101 allele and SJS–TEN in Japanese persons,11 and a report in this issue of the Journal shows the association between this allele and a range of hypersensitivity reactions, including SJS–TEN, in persons of European descent.12 Among persons of Han Chinese descent, carbamazepine-induced SJS–TEN almost never occurs in noncarriers of the HLA-B*1502 allele, evidence that this allele is directly involved in the pathogenesis of the condition. Carbamazepine directly binds to HLA-B molecules on antigenpresenting T cells and contributes to cell death mediated by cytotoxic T cells in persons with SJS–TEN.13 HLA-B*1502 can directly present carbamazepine to cytotoxic T cells without antigen processing. More important, carbamazepinespecific T-cell–mediated cytotoxicity is restricted to HLA-B*1502.14 The risk of carbamazepine-induced SJS–TEN is significantly higher among persons of Chinese origin who carry the HLA-B*1502 allele than among those who do not carry the allele (odds ratio, 1357; 95% confidence interval, 193 to

n engl j med 364;12

8838; P = 1.6×10−41).10 If HLA-B*1502 were used as a marker to predict carbamazepine-induced SJS–TEN, the test would have a high sensitivity (98.3%) and specificity (95.8%). On the basis of an incidence of carbamazepine-induced SJS–TEN of 0.25%, this allele would have a negative predictive value of 99.9% and a positive predictive value of 5.6%. The use of HLA-B*1502 genotyping to prevent carbamazepine-induced SJS–TEN in routine clinical practice thus seems warranted. We conducted a study to determine whether prospective screening by means of HLA-B*1502 genotyping before deciding on carbamazepine treatment reduces the incidence of carbamazepine-induced SJS–TEN.

Me thods Study Design

We recruited subjects from 23 participating hospitals throughout Taiwan (see the Supplementary Appendix, available with the full text of this article at NEJM.org). To ensure that all investigational sites complied with all applicable regulations and protocol requirements, an independent contract research organization (CRO Service Division, Formosa Biomedical Technology) monitored the study. We had 9 points of interaction with HLAB*1502–negative subjects and 10 points of interaction with HLA-B*1502 carriers: the first screening visit, a second clinic visit for HLA-B*1502 carriers only, and weekly telephone interviews during the 2-month follow-up period. Subjects between the ages of 6 months and 99 years who had not previously received carbamazepine and who would normally have received it at the time of screening were invited to participate in the study. We excluded subjects who had a history of carbamazepine allergy, those who had undergone bone marrow transplantation, and those who were not of Han Chinese descent. Status with respect to Han Chinese descent was determined with the use of a multiple-choice questionnaire that asked the subjects to identify the ethnic group of their parents and grandparents. We prescribed and provided carbamazepine to all subjects at the time of the screening visit, but we requested that they defer taking the drug until we had obtained and relayed to them their

nejm.org

march 24, 2011

1127

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

genetic test results. Blood samples were collected and transferred to a central laboratory (Institute of Biomedical Sciences, Academia Sinica) for HLA-B*1502 genotyping, and the results were reported to the participating physicians within 2 to 3 days. We asked HLA-B*1502–positive subjects to return to the clinic within 3 days, at which point we explained to them the risk of carbamazepineinduced SJS–TEN and recommended alternative drugs or therapy. HLA-B*1502–negative subjects (who also were counseled about the risk of SJS– TEN) were started on carbamazepine. Because the onset of SJS–TEN occurs within 2 months after the initiation of carbamazepine therapy,10 we interviewed all subjects by telephone during the 2-month period after the screening visit (for HLA-B*1502–negative subjects) or after the second clinic visit (for HLA-B*1502 carriers) to monitor them for symptoms of adverse drug reactions, including SJS–TEN. We asked subjects to return to the hospital immediately to be evaluated by a dermatologist in the event that early symptoms of SJS–TEN developed. We monitored all subjects (aside from those who were lost to follow-up) throughout the study. We performed the study in accordance with Good Clinical Practice Standards and the provisions of the Declaration of Helsinki. The study protocol was approved by the research ethics committee at Academia Sinica and by the institutional review board at each participating hospital. Written informed consent was obtained from all subjects or from parents or guardians for subjects who were under 21 years of age. The study was conducted in accordance with the protocol. Genotyping of HLA-B *1502

of

m e dic i n e

samples were also tested in parallel by HLA sequence-specific oligonucleotide reverse line blot (Dynal Biotech). In each of these 2000 samples, the results that were obtained with the use of the PG1502 DNA kit were consistent with those obtained with the use of the line-blot assay. Annual Incidence

The estimated number of SJS–TEN cases was based on diagnostic code 695.1 for erythema multiforme in the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM). We obtained the number of subjects with this code from the National Health Insurance Research Database (NHIRD), provided by the Bureau of National Health Insurance of the Department of Health. We calculated the annual incidence of carbamazepine-induced SJS–TEN in Taiwan as the annual number of cases of SJS–TEN caused by carbamazepine divided by the annual number of new carbamazepine recipients. A previous 5-year retrospective study (1997–2002), in which a dermatologist reviewed the medical records of 700 subjects with ICD-9-CM code 695.1 in the ChangGung Memorial Hospital system, suggested that among 230 subjects with SJS–TEN, 60 had taken carbamazepine; SJS–TEN had developed in the remaining 170 subjects after they had received other drugs.2 We therefore assumed that 32.9% of subjects with ICD-9-CM diagnostic code 695.1 had SJS–TEN and that the disease would be caused by carbamazepine in 26.1% of these subjects. Thus, we calculated the annual number of cases of carbamazepine-induced SJS–TEN as the number of subjects with ICD-9-CM diagnostic code 695.1 multiplied by 32.9% (the estimated proportion of subjects in Taiwan with ICD-9-CM diagnostic code 695.1 who would have SJS or TEN) multiplied by 26.1% (the estimated proportion of cases of carbamazepine-induced SJS–TEN among all those with SJS–TEN in Taiwan). Since it usually takes at least 14 days for SJS– TEN to develop after the initiation of carbamazepine treatment, we obtained data for years 2001, 2002, 2003, and 2004 from the NHIRD regarding the number of persons who received new prescriptions for at least 14 days of carbamazepine each year.

We obtained 2 ml of whole blood from each subject in a Monovettes tube, stored the sample at 4 to 12°C, and sent it to Academia Sinica on the same day that we obtained it. Genomic DNA was isolated with the use of the QIAamp DNA purification system (Qiagen). The presence or absence of the HLA-B*1502 allele was determined with the use of the PG1502 DNA detection kit (Pharmigene), according to the manufacturer’s instructions. The kits are based on real-time polymerasechain-reaction assay, with sequence-specific Statistical Analysis primers for HLAB*1502. To validate the results On the basis of the prevalence of the HLA-B*1502 of the PG1502 DNA detection kit, the first 2000 allele (8%) in the Taiwanese population,3 we de1128

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

HLA-B*1502 Screening and Carbamazepine ther apy

4877 Patients were enrolled

22 Had a protocol violation

4855 Were included in the study

4483 (92.3%) Were HLA-B*1502–negative

372 (7.7%) Were HLA-B*1502–positive

283 Did not take carbamazepine 5 Were lost to follow-up 125 Were lost to follow-up

367 Did not take carbamazepine (215 were given alternative medicines)

4120 Took carbamazepine

Figure 1. Enrollment and Outcomes. Carbamazepine was prescribed and provided for all subjects at the time of the screening visit, but they were asked 1st AUTHOR: Shen to defer taking the drug until the results of genetic testing were available toRETAKE: them. All subjects, regardless of HLA-B 2nd status, were followed for 2 months,FIGURE: with weekly telephone interviews. 1 of 1 3rd

Revised

ARTIST: ts TYPE:

Line

Combo

4-C

H/T

SIZE 6 col 33p9

termined that 4419 subjects would provide a diabetes-related neuropathic pain (11.7%), tinniAUTHOR, PLEASE NOTE: power of 99% to detect a reduction in has thebeen incitusand(3.6%), bipolar disorder or other psychiFigure redrawn type has and been reset. Please check carefully. dence of carbamazepine-induced SJS–TEN from atric disorders (2.8%) (Table 1). 2 0.25% (i.e., 25 cases per 10,000 JOB:new 36412recipients ) ISSUE: 03-24-11 to 0.03%. We used Fisher’s exact to compare the Screening for HLA-B *1502 rate of carbamazepine-induced SJS–TEN in the Of the 4855 subjects who were included in the prospective screening population with the his- study, 372 (7.7%) were found to carry the HLAtorical incidence. All reported P values are two- B*1502 allele and were advised not to take carbatailed, and a P value of less than 0.05 was consid- mazepine. These subjects were prescribed alternaered to indicate statistical significance. tive medications or advised to continue taking their prestudy medication; all were monitored for adverse events. Of these subjects, 5 were lost to R e sult s follow-up, 215 took an alternative medication, Subjects and 152 continued taking their prestudy mediFrom July 2007 through April 2010, we enrolled cation (Fig. 1). Alternative medications included 4877 subjects, of whom 4855 underwent geno- gabapentin, valproic acid, oxcarbazepine, imipratyping and were included in the 2-month follow- mine, clonazepam, and lamotrigine. The remaining up (Fig. 1). There were about equal numbers of 4483 subjects (92.3%) were HLA-B*1502–negamen and women, with a mean age of 56.5 years tive. Of these subjects, 238 did not take carbam(range, 0.6 to 98.2) for all subjects (Table 1). In- azepine and 125 were lost to follow-up, leaving dications for carbamazepine treatment included 4120 HLA-B*1502–negative subjects who took epilepsy (14.2% of subjects), neuralgia (54.1%), the drug and were monitored (Fig. 1). n engl j med 364;12

nejm.org

march 24, 2011

1129

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

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. Characteristics of the Subjects. HLA-B*1502–Positive (N = 372)

HLA-B*1502–Negative (N = 4483)

Total (N = 4855)

Male

193 (51.9)

2132 (47.6)

2325 (47.9)

Female

179 (48.1)

2351 (52.4)

2530 (52.1)

Mean

55.7

56.5

56.5

Range

4.3–91.4

0.6–98.2

0.6–98.2

Characteristic Sex — no. (%)

Age — yr

Indication for carbamazepine — no. (%) Epilepsy Neuralgia Diabetes-related neuropathic pain

57 (15.3)

632 (14.1)

689 (14.2)

195 (52.4)

2430 (54.2)

2625 (54.1)

53 (14.2)

515 (11.5)

568 (11.7)

8 (2.2)

168 (3.7)

176 (3.6)

Tinnitus Bipolar or other psychiatric disorder

12 (3.2)

122 (2.7)

134 (2.8)

Other conditions*

47 (12.6)

647 (14.4)

694 (14.3)

* These conditions include herpes zoster, unspecified cerebral-artery occlusion, meralgia paresthetica, and multiple sclerosis.

Adverse Events

Among all 4855 subjects, mild and transient rash and itching developed in 211 (4.3%), including 21 subjects who had a combination of rash, itching, and localized blisters and 26 subjects who had limited oral ulcers (Table 2). Of the 211 subjects with rash or itching, 5 were HLA-B*1502–positive and presented with symptoms after taking alternative drugs (gabapentin, lamotrigine, naproxen, imipramine, or prednisolone) (Table 2). In addition, 7 subjects had more severe cutaneous symptoms — maculopapular eruption (in 3 subjects), hypersensitivity syndrome (in 2 subjects), and urticaria (in 2 subjects) — for which they were hospitalized. One of the two subjects with urticaria was HLA-B*1502–positive and had taken oxcarbazepine before study enrollment. None of the subjects received a diagnosis of SJS–TEN. Other adverse events included fever, sore throat, fatigue, dizziness, insomnia, and gastrointestinal symptoms. These events occurred in both HLA-B*1502–positive and HLA-B*1502– negative subjects. Historical Incidence

The NHIRD data showed that carbamazepine was prescribed for at least 14 days for 50,917 persons in 2002, 48,522 in 2003, and 49,670 in 2004 who had not previously received carbamazepine, at least dating back to the beginning of the previous calendar year (Table 3). We therefore calcu1130

n engl j med 364;12

lated that the incidence of SJS–TEN in 2002, 2003, and 2004 in Taiwan was 0.24%, 0.22%, and 0.23%, respectively. We used the mean (0.23%) as the historical incidence for further analysis. Incidence of SJS–TEN after Genetic Screening

On the basis of the estimated historical incidence of 0.23%, 10 cases of SJS or TEN would be expected among the 4120 subjects who took carbamazepine in our study. However, there were no cases of either SJS or TEN in any of the subjects, a result that differed significantly from the historical incidence (P<0.001 by Fisher’s exact test) (Table 3).

Discussion Our findings suggest that screening patients for the HLA-B*1502 allele before the initiation of carbamazepine treatment and withholding carbamazepine from HLA-B*1502–positive patients can reduce the incidence of carbamazepineinduced SJS–TEN among Han Chinese. In estimating the historical incidence of this condition, we defined new carbamazepine recipients as those who had not received carbamazepine during the previous year and who were prescribed carbamazepine for at least 14 days in the year of interest, because carbamazepine-induced SJS–TEN is a delayed hypersensitivity reaction that usually takes at least 14 days to develop. However, even if

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

HLA-B*1502 Screening and Carbamazepine ther apy

Table 2. Adverse Events during the 2-Month Follow-up. HLA-B*1502–Positive HLA-B*1502–Negative with Alternative Medication with Carbamazepine (N = 215) (N = 4120)

Adverse Event

Total

number of events Mild cutaneous events Rash and itching

5*

206

Rash, itching, and blisters

1†

20

211 21

Rash, itching, and oral ulcers

0

14

14

Rash, itching, blisters, and oral ulcers

0

7

7

Itching, blisters, and oral ulcers

0

2

2

Blisters and oral ulcers

0

3

3

Maculopapular eruption

0

3

3

Severe cutaneous events Hypersensitivity syndrome

0

2

2

Urticaria

1‡

1

2

Stevens–Johnson syndrome or toxic epidermal necrolysis

0

0

0

1

92

93

Other adverse events§ Fever

4

126

130

Fatigue

Sore throat

16

818

834

Dizziness

10

497

507

Insomnia

5

197

202

Gastrointestinal symptoms

4

185

189

* Among these 5 subjects, the alternative drugs were gabapentin, lamotrigine, naproxen, imipramine, and prednisolone. † This subject had rash, itching, and blisters after taking gabapentin as an alternative treatment. These symptoms were mild and disappeared in 7 days. ‡ This subject had taken oxcarbazepine before study enrollment. § Subjects may have had more than one adverse event. Adverse events with a low frequency are not listed.

we had included all new carbamazepine recipients, regardless of the duration of treatment, as historical controls, the difference in the incidence of SJS–TEN would still be significant (P = 0.01, P = 0.02, and P = 0.02 for 2002, 2003, and 2004, respectively). Because we estimated the historical incidence of carbamazepine-induced SJS–TEN on the basis of data obtained from the NHIRD, the reliability of these data is critical for the validity of our estimation. The NHIRD was established in Taiwan when the government launched the National Health Insurance system in 1995. This mandatory single-payer health insurance system, which is administered by the Taiwanese government, provides health care for almost all people in Taiwan, with enrollment of 99.5% of the population in 2008. Of the health care facilities in Taiwan, 92.5% have been contracted by the Nan engl j med 364;12

tional Health Insurance system. The NHIRD data were therefore likely to be comprehensive. To estimate both the percentage of subjects with SJS–TEN among those with ICD-9-CM diagnostic code 695.1 (indicating erythema multiforme) and the percentage of cases of carbamazepineinduced SJS–TEN in Taiwan, we based our review on the medical records of 700 cases with diagnostic code 695.1 during a 5-year period at the Chang-Gung Memorial Hospital, the largest hospital system in Taiwan, with several regional centers. This hospital provides health care for about 12% of the Taiwanese population, and its patients are thought to be representative of the general population. It is possible that some of the drug-related adverse reactions we observed were early SJS lesions or that early withdrawal of carbamazepine may have prevented a more severe SJS–TEN or

nejm.org

march 24, 2011

1131

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

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. Historical Incidence of Carbamazepine-Induced SJS–TEN in 2002, 2003, and 2004, as Compared with the Incidence among Study Subjects.* Variable New recipients of carbamazepine (no.) Subjects with ICD-9-CM diagnostic code 695.1 (no.)

2002

2003

2004

50,917

48,522

49,670

1441

1261

1354

Carbamazepine-induced SJS–TEN (no.)

123

108

116

Incidence of carbamazepine-induced SJS–TEN (%)

0.24

0.22

0.23

<0.001

<0.001

<0.001

P value for comparison between historical incidence and incidence among study subjects†

* ICD-9-CM denotes International Classification of Diseases, 9th Revision, Clinical Modification, and SJS–TEN Stevens– Johnson syndrome and toxic epidermal necrolysis. † All P values were calculated with the use of Fisher’s exact test.

TEN-like reaction. However, we think that this is unlikely, since once patients are sensitized by carbamazepine and have early blisters or ulcers, SJS– TEN progresses, even after the withdrawal of the drug.15-17 SJS–TEN did not develop in any of the subjects who completed the 2-month follow-up. Adverse cutaneous reactions, including blisters, oral lesions, and rash, that occurred in the subjects were mild, localized, and transient. In some subjects, blisters developed only after the rash subsided (i.e, blisters were sporadic and tended not to occur at the same time as rash). Furthermore, many HLA-B*1502–negative subjects resumed taking carbamazepine without a recurrence of skin lesions. More important, SJS– TEN did not develop in these subjects, a finding that is consistent with the concept that the incidence of carbamazepine-induced SJS–TEN in HLA-B*1502–negative persons is very low. Our results suggest the value of HLA-B*1502 screening to prevent carbamazepine-induced SJS–TEN. However, as for any new pharmacogenomic test, it is important to document the use and safety of the alternative medications.18 Of the 367 HLA-B*1502 carriers, 215 (58.6%) were given an alternative medication, such as gabapentin, valproic acid, oxcarbazepine, imipramine,

clonazepam, or lamotrigine; the remainder continued to take their prestudy medication. Among the 215 HLA-B*1502 carriers who took alternative drugs, the only symptom seen during the 2-month follow-up was mild, transient rash in 5 subjects (2.3%). A strong association between HLA-B*1502 and carbamazepine-induced SJS–TEN has been found in Asian populations other than the Han Chinese,4-9 including Malay, Thai, and South Asian Indians. In Malaysia, Thailand, and India, studies have shown that carbamazepine was the major cause of drug-induced SJS–TEN. Since the contribution of HLA-B*1502 to carbamazepine-induced SJS–TEN has been proved to be causal,10,13,14,17 we speculate that in these countries, in which HLA-B*1502 is relatively prevalent, HLA-B*1502 screening could provide a benefit. Supported by grants from the National Research Program for Genomic Medicine, National Science Council, Taiwan (for the National Clinical Core and the National Genotyping Core) and from the Academy–Industry Collaboration Program (NSC953112-B-001-023 and NSC97-3112-B-001-007); and the Taiwan Drug Relief Foundation (TDRF9707 and TDRF9809). Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. We thank the subjects who participated in this study and the research and nursing staff for their meticulous data collection.

Appendix The authors’ affiliations are as follows: the Institute of Biomedical Sciences, Academia Sinica, Taipei (P.C., S.-Y.L., H.-P.C., L-.C.H., Y.-J.C., P-.J.T., H.-T.L., Y.-H.L., C.-H.C., S.-I.H., J.-Y.W., Y.-T.C., C.-Y.S.); Chu Shang Show Chwan Hospital, Nantou (J.-J.L.); the Departments of Neurology (C.-S.L, L.-S.R., C.-C.C.) and Dermatology (W.-H.C.), Chang-Gung Memorial Hospital at Linkou and College of Medicine, Chang-Gung University, Kweishan; Chia-Yi Christian Hospital, Chia-Yi (C.-T.O., Y.-C.H., Y.-H.S., S.-F.S.); Taichung Veterans General Hospital, Taichung, and National Chi Nan University, Nantou (P.-F.H.); National Taiwan University Hospital, Taipei (C.-C.Y.); Kaohsiung Medical University Chung-Ho Memorial Hospital and Kaohsiung Medical University, Kaohsiung (C.-T.T., S.-H.L.); Changhua Christian Hospital, Changhua (S.-L.W.); Chang-Gung Memorial Hospital at Kaohsiung, Kaohsiung (C-.H.L.); Neurology Department, Taipei Veterans General Hospital and National Yang-Ming University, Taipei (H.-Y.Y.); College of Public Health (C.-Y.S.) and China Medical University Hospital (C.-T.L.), China Medical University, Taichung; National Cheng Kung University Hospital, Tainan (J.-J.T.); Pharmigene (C.-F.C., L.C.); and the Department of Pediatrics, Duke University Medical Center, Durham, NC (Y.-T.C.).

1132

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

HLA-B*1502 Screening and Carbamazepine ther apy References 1. Roujeau JC, Stern RS. Severe adverse

cutaneous reactions to drugs. N Engl J Med 1994;331:1272-85. 2. Hung SI, Chung WH, Chen YT. HLA-B genotyping to detect carbamazepineinduced Stevens-Johnson syndrome: implications for personalized medicine. Personalized Medicine 2005;2:225-37. 3. Chung WH, Hung SI, Hong HS, et al. Medical genetics: a marker for StevensJohnson syndrome. Nature 2004;428:486. 4. Man CB, Kwan P, Baum L, et al. Association between HLA-B*1502 allele and antiepileptic drug-induced cutaneous reactions in Han Chinese. Epilepsia 2007;48: 1015-8. [Erratum, Epilepsia 2008;49:941.] 5. Lim KS, Kwan P, Tan CT. Association of HLA-B*1502 allele and carbamazepineinduced severe adverse cutaneous drug reaction among Asians, a review. Neurol Asia 2008;13:15-21. 6. Locharernkul C, Loplumlert J, Limotai C, et al. Carbamazepine and phenytoin induced Stevens-Johnson syndrome is associated with HLA-B*1502 allele in Thai population. Epilepsia 2008;49:2087-91. [Erratum, Epilepsia 2009;50:971.] 7. Tassaneeyakul W, Tiamkao S, Jantararoungtong T, et al. Association between HLA-B*1502 and carbamazepine-induced

severe cutaneous adverse drug reactions in a Thai population. Epilepsia 2010;51: 926-30. 8. Mehta TY, Prajapati LM, Mittal B, et al. Association of HLA-B*1502 allele and carbamazepine-induced Stevens-Johnson syndrome among Indians. Indian J Dermatol Venereol Leprol 2009;75:579-82. 9. Lonjou C, Thomas L, Borot N, et al. A marker for Stevens-Johnson syndrome: ethnicity matters. Pharmacogenomics J 2006;6:265-8. 10. Hung SI, Chung WH, Jee SH, et al. Genetic susceptibility to carbamazepineinduced cutaneous adverse drug reactions. Pharmacogenet Genomics 2006;16:297306. 11. Ozeki T, Mushiroda T, Yowang A, et al. Genome-wide association study identifies HLA-A*3101 allele as a genetic risk factor for carbamazepine-induced cutaneous adverse drug reactions in Japanese population. Hum Mol Genet 2011;20:1034-41. 12. McCormack M, Alfirevic A, Bourgeois S, et al. HLA-A*3101 and carbamazepineinduced hypersensitivity reactions in Europeans. N Engl J Med 2011;364:1134-43. 13. Yang CW, Hung SI, Juo CG, et al. Human leukocyte antigen-B*1502-bound peptides: implications for the pathogenesis

of carbamazepine-induced Stevens-Johnson syndrome. J Allergy Clin Immunol 2007;120:870-7. 14. Wei CY, Huang HW, Chen YT, Hung SI. The role of HLA-B*1502 in carbamazepine-induced Stevens-Johnson syndrome. In: Programs and abstracts of the TaiwanACGA 2010 International Conference on Genetic and Genomic Medicine, Taipei, May 2–5, 2010. 15. Garcia-Doval I, LeCleach L, Bocquet H, Otero XL, Roujeau JC. Toxic epidermal necrolysis and Stevens-Johnson syndrome: does early withdrawal of causative drugs decrease the risk of death? Arch Dermatol 2000;136:323-7. 16. Fritsch PO, Sidoroff A. Drug-induced Stevens-Johnson syndrome/toxic epidermal necrolysis. Am J Clin Dermatol 2000; 1:349-60. 17. Chung WH, Hung SI, Yang JY, et al. Granulysin is a key mediator for disseminated keratinocyte death in Stevens-Johnson syndrome and toxic epidermal necrolysis. Nat Med 2008;14:1343-50. 18. Flockhart DA, Skaar T, Berlin DS, Klein TE, Nguyen AT. Clinically available pharmacogenomics tests. Clin Pharmacol Ther 2009;86:109-13. Copyright © 2011 Massachusetts Medical Society.

MY NEJM IN THE JOURNAL ONLINE

Individual subscribers can store articles and searches using a feature on the Journal’s Web site (NEJM.org) called “My NEJM.” Each article and search result links to this feature. Users can create personal folders and move articles into them for convenient retrieval later.

n engl j med 364;12

nejm.org

march 24, 2011

1133

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

original article

HLA-A*3101 and Carbamazepine-Induced Hypersensitivity Reactions in Europeans Mark McCormack, B.A., Ana Alfirevic, M.D., Ph.D., Stephane Bourgeois, Ph.D., John J. Farrell, M.S., Dalia Kasperaviþinjtơ, Ph.D., Mary Carrington, Ph.D., Graeme J. Sills, Ph.D., Tony Marson, M.B., Ch.B,, M.D., Xiaoming Jia, M.Eng., Paul I.W. de Bakker, Ph.D., Krishna Chinthapalli M.B., B.S., Mariam Molokhia, M.B., Ch.B., Ph.D., Michael R. Johnson, D.Phil., Gerard D. O’Connor, M.R.C.P.I., Elijah Chaila, M.R.C.P.I., Saud Alhusaini, M.B., Kevin V. Shianna, Ph.D., Rodney A. Radtke, M.D., Erin L. Heinzen, Ph.D., Nicole Walley, B.S., Massimo Pandolfo, M.D., Ph.D., Werner Pichler, M.D., B. Kevin Park, Ph.D., Chantal Depondt, M.D., Ph.D., Sanjay M. Sisodiya, M.D., Ph.D., David B. Goldstein, Ph.D., Panos Deloukas, Ph.D., Norman Delanty, B.M., Gianpiero L. Cavalleri, Ph.D., and Munir Pirmohamed, Ph.D., F.R.C.P.

A bs t r ac t Background The authors’ affiliations are listed in the Appendix. Address reprint requests to Dr. Alfirevic at the Wolfson Centre for Personalised Medicine, Department of Molecular and Clinical Pharmacology, Institute of Translational Medicine, University of Liverpool, Block A: Waterhouse Bldgs., 1-5 Brownlow St., Liverpool L69 3GL, United Kingdom, or at ana.alfirevic@ liv.ac.uk. Mr. McCormack, Dr. Alfirevic, Dr. Cavalleri, and Dr. Pirmohamed contributed equally to this article. N Engl J Med 2011;364:1134-43. Copyright © 2011 Massachusetts Medical Society.

Carbamazepine causes various forms of hypersensitivity reactions, ranging from maculopapular exanthema to severe blistering reactions. The HLA-B*1502 allele has been shown to be strongly correlated with carbamazepine-induced Stevens–Johnson syndrome and toxic epidermal necrolysis (SJS–TEN) in the Han Chinese and other Asian populations but not in European populations. Methods

We performed a genomewide association study of samples obtained from 22 subjects with carbamazepine-induced hypersensitivity syndrome, 43 subjects with carbamazepine-induced maculopapular exanthema, and 3987 control subjects, all of European descent. We tested for an association between disease and HLA alleles through proxy single-nucleotide polymorphisms and imputation, confirming associations by high-resolution sequence-based HLA typing. We replicated the associations in samples from 145 subjects with carbamazepine-induced hypersensitivity reactions. Results

The HLA-A*3101 allele, which has a prevalence of 2 to 5% in Northern European populations, was significantly associated with the hypersensitivity syndrome (P = 3.5×10−8). An independent genomewide association study of samples from subjects with maculopapular exanthema also showed an association with the HLAA*3101 allele (P = 1.1×10−6). Follow-up genotyping confirmed the variant as a risk factor for the hypersensitivity syndrome (odds ratio, 12.41; 95% confidence interval [CI], 1.27 to 121.03), maculopapular exanthema (odds ratio, 8.33; 95% CI, 3.59 to 19.36), and SJS–TEN (odds ratio, 25.93; 95% CI, 4.93 to 116.18). Conclusions

The presence of the HLA-A*3101 allele was associated with carbamazepine-induced hypersensitivity reactions among subjects of Northern European ancestry. The presence of the allele increased the risk from 5.0% to 26.0%, whereas its absence reduced the risk from 5.0% to 3.8%. (Funded by the U.K. Department of Health and others.) 1134

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

HLA-A*3101 and Carbamazepine-Induced Hypersensitivity

C

arbamazepine is one of the most commonly prescribed drugs for the treatment of epilepsy, as well as trigeminal neuralgia and bipolar disorder. A minority of treated persons have hypersensitivity reactions that vary in prevalence and severity,1 with some forms associated with substantial morbidity and mortality. The mildest form, maculopapular exanthema, occurs in 5 to 10% of treated persons of European ancestry and resolves spontaneously after drug discontinuation. More severe reactions, such as the hypersensitivity syndrome, are associated with mortality of up to 10%2 and include symptoms such as rash, fever, eosinophilia, hepatitis, and nephritis. The most severe reactions, such as the Stevens–Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN), are characterized by a blistering rash affecting a variable percentage of the body-surface area. The rate of death increases with the degree of epidermal detachment. TEN is the rarest of these phenotypes and is associated with mortality of up to 30%. According to the labeling of carbamazepine, as mandated by the Food and Drug Administration (FDA), the estimated incidence of SJS–TEN is 1 to 6 cases in 10,000 persons of European ancestry who are exposed to the drug. Genomewide approaches are increasingly used to identify genetic predisposing factors for druginduced hypersensitivity reactions and druginduced liver injury.3 For example, the HLAB*1502 allele has been identified as a clinically important predictor of SJS–TEN in Asians of Han Chinese descent who are candidates for treatment with carbamazepine.4 This finding led the FDA to require a warning label for carbamazepine to indicate the need for genotyping for HLAB*1502 before the drug is prescribed. This change seems warranted, given the prospective study described in this issue of the Journal,5 which shows that withholding carbamazepine from subjects of Han Chinese ancestry who carry the HLA-B*1502 allele substantially decreased the rate of the development of SJS–TEN. The value of HLA-B*1502 in predicting the risk of SJS–TEN has been shown in other Asian populations, including those of Thailand, Malaysia, and India.4,6-8 This effect is specific for SJS–TEN, as compared with the full spectrum of carbamazepine-related hypersensitivity reactions, which led to the suggestion that different phenotypes of carbamaze-

n engl j med 364;12

pine-induced hypersensitivity may have distinct genetic predictors. Furthermore, HLA-B*1502 is rare (with a prevalence of less than 2%) in populations of European descent, in which carbamazepine-induced SJS–TEN occurs at a lower incidence than in Asian populations.9-12 Using a candidate-gene approach, previous studies have shown weak or moderate signals of association between maculopapular exanthema or the DRESS (drug reaction with eosinophilia and systemic symptoms) syndrome and single-nucleotide polymorphisms (SNPs) in the major histocompatibility complex (MHC) region,13-15 and these associations have not been consistently replicated.13,14 We present data from two independent studies that suggest the relevance of an HLA variant to the clinical spectrum of carbamazepine-related hypersensitivity reactions, including maculopapular exanthema, the hypersensitivity syndrome, and SJS–TEN.

Me thods Case Subjects

Recruitment

We recruited case subjects at centers collaborating with the University of Liverpool and Walton Centre for Neurology (both hereinafter referred to as “the Liverpool collaborators”) or at centers affiliated with the EPIGEN consortium (for details, see the Supplementary Appendix, available with the full text of this article at NEJM.org). All case subjects and control subjects were of European ancestry as determined by either self-report or genetic-marker analysis.13 Many of these subjects have been described previously in the context of hypersensitivity reactions.9,13,16 Hypersensitivity Syndrome

The group with the hypersensitivity syndrome consisted of 26 subjects: 23 who were recruited through the Liverpool collaborators and 3 who were recruited through the EPIGEN consortium. The hypersensitivity syndrome was defined as the presence of rash or liver involvement within 3 months after the initiation of carbamazepine treatment, accompanied by a minimum of two of the following manifestations: a prolonged recovery phase, despite withdrawal of the drug; fever; involvement of other internal organs (liver, kidney, lung, central nervous system, heart, muscle, thy-

nejm.org

march 24, 2011

1135

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

roid, or lymphoid system); or the presence of hematologic abnormalities, such as eosinophilia and atypical lymphocytosis.17 We considered the DRESS syndrome to be the same as the hypersensitivity syndrome.

of

m e dic i n e

Clinical Controls

The clinical control group consisted of 257 subjects with epilepsy who had been taking carbamazepine for at least 3 months with no clinical or biochemical signs of hypersensitivity. Of these subjects, 44 were recruited by the Liverpool collaborators and 213 were recruited from clinics Acute Generalized Exanthematous Pustulosis We recruited one subject with acute generalized affiliated with EPIGEN. exanthematous pustulosis (AGEP), who had a typical pustular reaction with high fever,18 from Genomewide Genotyping a collaborating clinic in Switzerland. We then Genomewide analysis of samples that were inianalyzed the data from this subject together with tially available from the 22 Liverpool subjects data from the subjects with the hypersensitivity with carbamazepine-induced hypersensitivity syndrome was performed with the use of the Illusyndrome. mina Infinium 1.2M chip at the Wellcome Trust Sanger Institute. Samples from 43 case subjects Maculopapular Exanthema The group with maculopapular exanthema con- and 200 control subjects from the EPIGEN consisted of 106 subjects: 57 who were recruited sortium were genotyped with the use of the Illuthrough the Liverpool collaborators and 49 who mina 610K Quad platform at the Duke University were recruited through the EPIGEN consortium. Center for Human Genome Variation. (Details Maculopapular exanthema was defined as rash regarding quality controls and principal-compowithout systemic symptoms that required the dis- nents analysis are provided in the Supplementary continuation of carbamazepine within 3 months Appendix.) after the initiation of drug therapy. Imputation of HLA Alleles

SJS–TEN

The SJS–TEN group consisted of 12 subjects: 8 who were recruited through the Liverpool collaborators and 4 who were recruited through the EPIGEN consortium. The diagnostic criteria suggested by the RegiSCAR Project (a research network for severe cutaneous adverse drug reactions)11,19 were used for this group (see the Supplementary Appendix). Control Subjects

Population Controls

Genotype data from the Wellcome Trust Case Control Consortium (WTCCC) (www.wtccc.org .uk) were used as a population control group. The Liverpool collaborators used data from a genomewide association study of samples from 2691 subjects in the U.K. National Blood Services Collection. EPIGEN used samples from a homogeneous subgroup of 1296 subjects in the 1958 British Birth Cohort that were selected by principal component analysis. We did not screen either set of population controls for carbamazepine-related adverse drug reactions. (For details about our access to these data, see the Supplementary Appendix.)

1136

n engl j med 364;12

We based the imputation of classic HLA alleles in samples from the 22 Liverpool subjects with the hypersensitivity syndrome and control subjects from the U.K. National Blood Service on a dense reference panel of SNP data and four-digit HLA types in 2767 unrelated subjects of European descent, according to a method described recently.20 We imputed classic HLA alleles in the samples from the EPIGEN consortium, using MACH 1.0 on the Web site of the Center for Statistical Genetics (www.sph.umich.edu/csg/abecasis/MACH).21 The 1458 subjects in the HLA-typed 1958 British Birth Cohort constituted the reference population. (See the Supplementary Appendix for further details.) We used different approaches to impute the HLA alleles in the Liverpool and EPIGEN subjects, because the two studies were originally independent. HLA-A High-Resolution Genotyping

To confirm the accuracy of imputation, highresolution, sequence-based HLA-A typing of samples from 22 subjects with the hypersensitivity syndrome and from a subgroup of 44 clinical control subjects (i.e., subjects who did not have adverse reactions to carbamazepine) was per-

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

HLA-A*3101 and Carbamazepine-Induced Hypersensitivity

formed by Histogenetics. We determined the HLA-A alleles in samples from 49 subjects with maculopapular exanthema and a subgroup of 213 clinical control subjects from the EPIGEN consortium, using sequence-specific polymerase-chainreaction assay primers (details available on request). Direct genotyping showed 100% concordance with the imputed alleles. HLA-A* 3101 Proxy SNP Genotyping

Genotyping of the rs1061235 SNP in samples from all Liverpool subjects (including case subjects with maculopapular exanthema, the hypersensitivity syndrome, or SJS–TEN and clinical control subjects) was performed by KBioscience. This allele has previously been shown to be in complete linkage disequilibrium (r2 = 1) with the HLA-A*3101 allele in populations of European descent.22 Statistical Analysis

Statistical analyses were performed with the use of PLINK (version 1.05),23 RevMan (version 5) (http://ims.cochrane.org/revman), and Haploview24 software packages. For the genomewide association studies, both the Liverpool collaborators and the EPIGEN consortium used logistic regression with an additive model of inheritance. They adjusted for population stratification by including significant principal components as covariates in the logistic-regression model and used a P value of less than 5.0×10−8 to indicate genomewide significance. Pooled analysis of the different groups of subjects was performed to estimate odds ratios and confidence intervals from summary data with the use of a random-effects model. Between-study heterogeneity was tested by calculating I2 and Tau.2 The z-score directionality or estimated risk was checked for consistency with observed data.

R e sult s Association with the Hypersensitivity Syndrome

We first conducted an association analysis involving 1.2M SNPs in samples from the 22 Liverpool subjects with carbamazepine-induced hypersensitivity syndrome and from 2691 healthy controls from the U.K. National Blood Services Collection. (The full set of results can be obtained from the European Genotype Archive, access num-

n engl j med 364;12

ber EGAS00000000037.) We identified a strong signal in the MHC region on chromosome 6, with several SNPs around HLA-A reaching genomewide significance (P = 3.5×10−8) (Fig. 1A and 1B). This variant was seen in 40.0% of case subjects but in only 4.9% of control subjects. The top hit (rs1061235) had previously been shown to be a proxy for the HLA-A*3101 allele in persons of European descent.22 Therefore, we focused on this locus as a predictor of response through high-resolution, sequence-based HLA-A typing, which confirmed the absolute correlation (r2 = 1) between rs1061235 and HLA-A*3101 that had been observed previously.22 No other classic HLA-A alleles reached nominal significance (Table 1 in the Supplementary Appendix). During the course of this study, samples from 4 additional subjects with the hypersensitivity syndrome and from 1 additional subject with AGEP became available from the Liverpool and EPIGEN centers. Of these 5 subjects, 2 (including the single subject with AGEP) tested positive for HLA-A*3101. A comparison of all 27 subjects with the 257 control subjects without adverse drug reactions produced an odds ratio of 12.41 (95% confidence interval [CI], 1.27 to 121.03; P = 0.03) (Fig. 2). Association with Maculopapular Exanthema

EPIGEN investigators performed a genomewide association study of samples from 43 subjects with maculopapular exanthema and 1296 healthy control subjects from the 1958 British Birth Cohort. (The full set of results can be obtained from dbGAP, access number EGAS00000000036.) HLAA*3101 was the most strongly associated allele (P = 1.11×10−6) and was observed in 27% of case subjects and 4% of control subjects, although the effect was not significant genomewide (Fig. 3 in the Supplementary Appendix). However, we had access to another 63 case subjects with maculopapular exanthema and 257 control subjects without adverse drug reactions. Our combined analysis of all 106 subjects with maculopapular exanthema and the 257 clinical control subjects showed a significant allelic association with HLA-A*3101 (P = 8.0×10−7), generating an odds ratio of 8.33 (95% CI, 3.59 to 19.36) (Fig. 2). Although the analyses of samples from subjects with maculopapular exanthema and those with the hypersensitivity syndrome were carried out in the Nether-

nejm.org

march 24, 2011

1137

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

A Carbamazepine-Induced Hypersensitivity

P Value (−log10)

7

6

0

1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 181920 22 21

Chromosome

B SNPs in the HLA-A Region 10

80

Typed

9

0.80 to 1.00 0.50 to <0.80 O O

P Value (−log10)

7 O

6

HLA-A*3101

60

O

5 4 3

1 0

O

O O

O O

O O OO

O OO O OO OOOO O O

O

O

O O O

OO O O OOOO

O

O O

O

O O

40

OOO

O O

OO

O OO OO O O OOOO O O O O O O O O O O O OO O OOO OOOO O O O OO O O O O O O O O O O OOO O O O OO O O O O O OO OO OOO OOOO OOO OOOOO O OO O O O O O O O O OO O O OOO O O O O OO OO O O O OOO O OOOOO OOO O O OOOO O O O O O

O O

O

20

Recombination Rate (cM/Mb)

8

2

LD

0.20 to <0.50 <0.20

Imputed O 0.80 to 1.00 O 0.50 to <0.80 O

0.20 to <0.50

O

<0.20

O

OO OO O OO OO O OO OO O OO O OO OO OO O O OO O OO O O O O O O OO O OOOOO OO OOOO OOOO O OO OOOOOOOO OO OOO OOOO O OO OO O OO O O O OO O O OO O O O

0 HLA-A 29.7

30.0

30.3

Chromosome 6 (positions in Mb)

Figure 1. Results of a Genomewide Association Study of Samples from Case Subjects with Carbamazepine-Induced Hypersensitivity Syndrome and Control Subjects. Panel A shows a Manhattan plot for logistic regression with a strong signal in the HLA-A region on chromosome 6. Each dot represents a P value for the comparison of 22 case subjects with the hypersensitivity syndrome and 2691 healthy control subjects from the Wellcome Trust Case Control Consortium. Panel B shows P values for single-nucleotide polymorphisms (SNPs) in the HLA-A region. There is a strong association between carbamazepine-induced hypersensitivity syndrome and the HLA-A*3101 allele, with identification of the allele in 40.0% of case subjects and only 4.9% of control subjects. The x axis shows the SNP position on chromosome 6 (National Center for Biotechnology Information build 36). The left y axis shows the negative log10 of P values for the comparison between case subjects and control subjects, as calculated with logistic regression. The right y axis shows the recombination rate on chromosome 6 between 29 and 31 Mb. The diamonds show the degree of linkage disequilibrium (LD) in the samples.

1138

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

HLA-A*3101 and Carbamazepine-Induced Hypersensitivity

lands and Liverpool, respectively, the effect size to immune-mediated adverse reactions, such as of HLA-A*3101 on risk was consistent across the drug-induced hypersensitivity4,25-27 and liver inEPIGEN and U.K. groups. jury.3,28 The most prominent example thus far is the strong predictive value of the HLA-B*5701 Association with SJS–TEN allele for hypersensitivity to the drug abacavir, On the basis of our hypothesis that the HLA- used for the treatment of human immunodefiA*3101 allele is associated with carbamazepine- ciency virus infection.25,26,29 FDA guidelines now induced SJS–TEN, we genotyped HLA-A*3101 in recommend HLA-B*5701 testing in advance of a group of 12 subjects with SJS–TEN who were abacavir prescription. Testing results in a reducrecruited from both the Liverpool and EPIGEN- tion in the incidence of abacavir hypersensitiviaffiliated centers. Of these 12 subjects, 5 (42%) ty27,30 and is cost-effective.29,31 carried the allele, as compared with 10 (4%) of the Our study now provides strong evidence sug257 clinical control subjects (odds ratio, 25.93; gesting that HLA-A*3101 is a predictor of carba95% CI, 4.93 to 116.18; P = 8.0×10−5) (Fig. 2). mazepine hypersensitivity in Europeans. First, A pooled analysis of all 145 subjects with despite the small number of subjects who were carbamazepine-induced hypersensitivity and 257 included in the analysis, this signal had genomecontrol subjects without adverse drug reactions wide significance. Second, the signal was obshowed a strong association between hypersen- served across independent groups of case subjects sitivity and the HLA-A*3101 allele (odds ratio, with different phenotypes of carbamazepine hy9.12; 95% CI, 4.03 to 20.65; P = 1.0×10−7). On the persensitivity, with the use of independent sets basis of this pooled collection, we calculated of control subjects. Third, HLA-A*3101 has prethat the presence of HLA-A*3101 had a sensitiv- viously been shown to be associated with multiple ity of 26% and a specificity of 96% as a predic- carbamazepine-hypersensitivity phenotypes, intor of carbamazepine-associated hypersensitivity. cluding SJS–TEN in Japanese subjects15,32 and Given that carbamazepine hypersensitivity has a with carbamazepine-induced maculopapular exprevalence of 5% (1:20), application of the test anthema in subjects of Han Chinese ancestry.14 criteria from our pooled analysis would increase The one subject in our study who had AGEP, a the post-test probability to 26% (approximately reaction characterized by sterile pustules, neutro1:4) on the basis of the positive likelihood ratio philic inflammation, and high levels of interleuof 6.74 (95% CI, 3.51 to 13.00). Since the nega- kin-8 in skin,33 also was found to carry the tive likelihood ratio was 0.77 (95% CI, 0.67 to HLA-A*3101 allele. The prevalence of this allele is 0.84), a negative HLA-A*3101 test would reduce 2 to 5% in Northern European populations, 2% the probability of hypersensitivity from 5.0% to in Han Chinese populations, and 9% in Japanese 3.8% (1:26 ratio). populations.32,34,35 The finding that HLA-A*3101 is associated with carbamazepine-induced hypersensitivity in populations of different ancestries Discussion parallels the finding of HLA-B*5701-associated We found that a variation in the presence of the abacavir hypersensitivity.30,36 HLA-A*3101 allele is an important predictor of HLA-B*1502, a predictor of carbamazepine the full spectrum of carbamazepine-induced hy- hypersensitivity reactions in Asian populations, persensitivity reactions in persons of European seems to be phenotype-specific in that it predescent. Such reactions range from the relatively dicts the development of SJS–TEN but not the mild but nevertheless troublesome maculopapu- development of the hypersensitivity syndrome or lar exanthema to more severe reactions, such as maculopapular exanthema.14 This finding has led the hypersensitivity syndrome and SJS–TEN. Al- to the suggestion that the pathogenesis of carbathough the presence of the HLA-A*3101 allele is mazepine-induced SJS–TEN is distinct from that neither necessary nor sufficient for the develop- of the hypersensitivity syndrome. This hypothement of hypersensitivity to carbamazepine, it is sis is based on the fact that class I alleles are associated with a significantly increased risk. usually associated with specific CD8+ cytotoxic Our data add to growing evidence of the role T-cell reactions, which is consistent with the of different HLA alleles in predisposing patients known pathologic features of SJS–TEN.37 How-

n engl j med 364;12

nejm.org

march 24, 2011

1139

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

Subgroup

Carbamazepine Reaction No. Positive for the HLA-A*3101 Allele

Total No.

n e w e ng l a n d j o u r na l

Controls No. Positive for the HLA-A*3101 Allele

Hypersensitivity syndrome 0 EPIGEN 3 9 10 UK 24 1 10 Subtotal 27 10 Heterogeneity: tau2=1.03, chi2=1.58; 1 df, P=0.21; I2=37%. Test for overall effect: z=2.17; P=0.03 Maculopapular exanthema EPIGEN 13 49 9 UK 10 57 1 Subtotal 23 106 10 Heterogeneity: tau2=0.00, chi2=0.01; 1 df, P=0.92; I2=0%. Test for overall effect: z=4.93; P=8.0×10−7 Stevens–Johnson syndrome EPIGEN 2 4 9 UK 3 8 1 Subtotal 5 12 10 Heterogeneity: tau2=0.00, chi2=0.01; 1 df, P=0.93; I2=0%. Test for overall effect: z=3.94; P=8.0×10−5 All phenotypes 15 9 EPIGEN 56 23 1 UK 89 38 10 Subtotal 145 Heterogeneity: tau2=0.00, chi2=0.35; 1 df, P=0.55; I2=0%. Test for overall effect: z=5.30; P=1.0×10−7

of

m e dic i n e

Study Weighting (%)

Odds Ratio (95% CI)

Total No. 213 44 257

39.4 60.6 100

3.08 (0.15–63.89) 30.71 (3.61–261.65) 12.41 (1.27–121.03)

213 44 257

83.8 16.2 100

8.19 (3.26–20.56) 9.15 (1.12–74.47) 8.33 (3.59–19.36)

213 44 257

58.2 41.8 100

22.67 (2.86–179.73) 25.80 (2.24–297.58) 25.93 (4.93–116.18)

213 44 257

83.9 16.1 100

8.29 (3.40–20.23) 14.98 (1.95–115.09) 9.12 (4.03–20.65)

0.002

0.1

1.0

HLA-A*3101 Less Likely

10.0

500.0

HLA-A*3101 More Likely

Figure 2. Distribution of the HLA-A*3101 Allele across the Spectrum of Clinical Phenotypes of Case Subjects with CarbamazepineInduced Hypersensitivity and Control Subjects without Adverse Reactions to Carbamazepine. Subjects were recruited at centers collaborating with the University of Liverpool and Walton Centre for Neurology (UK) or centers affiliated with the EPIGEN consortium. The sample obtained from one patient with acute generalized exanthematous pustulosis (AGEP) was analyzed with the samples for the group of case subjects with the hypersensitivity syndrome. Since there were no observations of the HLA-A*3101 allele in the three case subjects with the hypersensitivity syndrome in the EPIGEN cohort, 0.5 was added to each value in a two-by-two contingency table to estimate an odds ratio. One patient with the Stevens–Johnson syndrome and toxic epidermal necrolysis (SJS–TEN) was of mixed European and Thai ancestry. Study weighting (indicated by different sizes of squares) refers to the proportion of subjects who were recruited from each study cohort. Diamonds indicate pooled odds ratios. The horizontal lines indicate 95% confidence intervals. The abbreviation df denotes degrees of freedom, and I2 is the percentage of total variation that is due to heterogeneity rather than chance.

ever, our data suggest that the class I allele HLAA*3101 may be important for an increased range of phenotypes, including maculopapular exanthema and the hypersensitivity syndrome, at least in European populations. Persons with carbamazepine-induced hypersensitivity syndrome have been shown to have several types of drug-specific T cells in peripheral blood, including CD4+, CD8+, and CD4−CD8+ cells.38 Although there is 1140

n engl j med 364;12

no obvious relationship between HLA-A*3101 and HLA-B*1502, reports have indicated a possible overlap in peptide-binding specificity.39 Given the degree of linkage disequilibrium across the extended MHC region, we cannot be certain that either HLA-B*1502 or HLA-A*3101 is causal. We used samples from healthy volunteers in the WTCCC as control subjects for both genomewide association studies. Given that the

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

HLA-A*3101 and Carbamazepine-Induced Hypersensitivity

hypersensitivity syndrome occurs in 1 in 5000 to 1 in 10,000 persons, it is unlikely that more than one or two population control subjects would be at risk for hypersensitivity with exposure to carbamazepine. The larger WTCCC collections therefore afforded increased power over that provided by comparison of case subjects with the clinical control subjects. Nevertheless, in analyses involving the 257 clinical control subjects, the association with HLA-A*3101 remained strong for both the hypersensitivity syndrome (odds ratio, 12.41; 95% CI, 1.27 to 121.03) and SJS (odds ratio, 25.93; 95% CI, 4.93 to 116.18), although the effect size was smaller with maculopapular exanthema (odds ratio, 8.33; 95% CI, 3.59 to 19.36). Additional studies in populations of European descent and other ancestry are required to further characterize this association. The International Serious Adverse Event Consortium is coordinating a global effort to recruit subjects with drug-induced severe adverse events (www .saeconsortium.org). Efforts such as this will facilitate the analysis of larger groups of subjects, which may well uncover other genetic variants associated with a smaller effect size or population-specific variants. Additional work is also required to determine whether the effect of the HLA-A*3101 allele is specific to carbamazepine or whether it also applies to other drugs. A prospective study that is based on the results we present here would facilitate the routine clinical use of HLA-A*3101 testing in carbamazepine treatment. In this issue of the Journal, a report describes how prospective HLA-B*1502 typing can prevent carbamazepine-induced SJS–TEN in Han Chinese populations.5 We propose that HLA-A*3101 is clinically relevant as a marker to predict hypersensitivity reactions. On the assumption that the prevalence of carbamazepine-induced hypersensitivity is 5.0%, the presence of the HLA-A*3101 allele increases the risk of hypersensitivity to 26.0%, whereas its absence reduces the risk to 3.8%. Among Japanese patients, the estimated reduction is from 2.9% to 1.1%. The sensitivity and specificity values estimated here for Europeans (26% and 96%) and by others for Japanese populations (61% and 88%)32 vary. Among a number of possible reasons for these differences are differences in study size, in the frequency of HLA-A*3101 (and potential interacting variation), and in the

n engl j med 364;12

effect size of the allele across the two populations. Nonetheless, these numbers are similar to those underlying the test criteria calculated for the presence of the HLA-B*5701 allele and abacavir hypersensitivity (sensitivity, 45.5%; specificity, 97.6%; and a risk reduction from 7.8% to 2.7%). On the basis of our data, we have calculated that 83 patients would need to be screened to prevent one case of carbamazepine hypersensitivity; the number in the Japanese population is 56. However, it is important to note that this calculation is based on a conservative estimate of the prevalence of carbamazepine hypersensitivity (5%). The prevalence of carbamazepine-induced hypersensitivity that was determined in the context of the Standard and New Antiepileptic Drugs (SANAD) trial (Current Controlled Trials number, ISRCTN38354748) (also involving subjects of European descent) was 10%.40 On the basis of this prevalence, the number of persons who would need to be screened to prevent one instance of carbamazepine-induced hypersensitivity is 39. We therefore suggest that consideration be given to adding the association with HLAA*3101 to the drug label for carbamazepine. Supported by grants to the Liverpool collaborators from the Department of Health, the National Health Service Chair of Pharmacogenetics, the Medical Research Council Centre for Drug Safety Science, the Wolfson Foundation, the Wellcome Trust Sanger Institute, and the National Institute for Health Research (to Dr. Pirmohamed); by grants to the EPIGEN consortium from the Medical Research Council (G0400126), the Wellcome Trust (084730), University College London Hospitals Charity, Clinical Research and Development Committee (F136), and the National Institute for Health Research (08-08-SCC); by an award (2009/001) from Brainwave–the Irish Epilepsy Association; by the Medical Research Charities Group of Ireland, the Health Research Board, the National Society for Epilepsy, Fonds National de la Recherche Scientifique, and Fonds Erasme pour la Recherche Médicale, Université Libre de Bruxelles (Belgium); by a Translational Research Scholars award from the Health Research Board of Ireland (to Mr. McCormack); by the Department of Health National Institute for Health Research Biomedical Research Centres; and by a contract from the National Cancer Institute (HHS-N261200800001E) and the Intramural Research Program at the National Cancer Institute, National Institutes of Health. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. We thank the subjects and the physicians who recruited them; U.K. Eudragene co-coordinator Paul McKeigue; Drs. Ines Salado, Alfonso Carvajal, Luisa Ibanez, Jean-Louis Montastruc, Maryse Lapeyre-Mestre, Emmanuelle Bondon-Guitton, Aatif M. Husain, William B. Gallantine, and Mohamad Mikati for their help in recruiting subjects; the pharmacovigilance centers involved in case ascertainment in France and Spain, the Association Française des Centres Régionaux de Pharmacovigilance and Sistema Español de Farmacovigilancia; and Agencia Española de Medicamentos y Productos Sanitarios.

nejm.org

march 24, 2011

1141

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

Appendix The authors’ affiliations are as follows: Molecular and Cellular Therapeutics, the Royal College of Surgeons in Ireland (M. McCormack, S.A., N.D., G.L.C.), and the Division of Neurology, Beaumont Hospital (G.D.O., E.C., N.D.) — both in Dublin; the Department of Molecular and Clinical Pharmacology, University of Liverpool (A.A., G.J.S., T.M., B.K.P., M. Pirmohamed), and Walton Centre for Neurology (T.M.) — both in Liverpool; Wellcome Trust Sanger Institute, Hinxton (S.B., P.D.); and the National Society for Epilepsy, Chalfont-St-Peter, Buckinghamshire (K.C., S.M.S.) — all in the United Kingdom; the Department of Medicine, Boston University (J.F.F.), Ragon Institute of Massachusetts General Hospital, Massachusetts Institute of Technology, and Harvard University (M.C.), Harvard–MIT Division of Health Sciences and Technology (X.J.), and the Division of Genetics, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School (P.I.W.B.) — all in Boston; the Department of Clinical and Experimental Epilepsy, University College London, Institute of Neurology, Queen Square (D.K., K.C., S.M.S.), the Department of Primary Care and Public Health Sciences, Division of Health and Social Care Research, King’s College (M. Molokhia), and the Centre for Neuroscience, Department of Medicine, Imperial College London (M.R.J.) — all in London; the Cancer and Inflammation Program, Laboratory of Experimental Immunology, SAIC–Frederick, National Cancer Institute at Frederick, Frederick, MD (M.C.); the Program in Medical and Population Genetics, Broad Institute of Harvard and MIT, Cambridge, MA (P.I.W.B.); Julius Center for Health Sciences and Primary Care and the Department of Medical Genetics, Division of Biomedical Genetics, University Medical Center, Utrecht, the Netherlands (P.I.W.B.); the Center for Human Genome Variation (K.V.S., E.L.H., N.W., D.B.G.) and the Department of Medicine (Neurology) (R.A.R.), Duke University Medical School, Durham, NC; the Department of Neurology, Hôpital Erasme, Université Libre de Bruxelles, Brussels (M. Pandolfo, C.D.); and the Department of Rheumatology, Clinical Immunology, and Allergology, University of Bern, Bern, Switzerland (W.P.). References 1. Roujeau JC. Clinical heterogeneity of

drug hypersensitivity. Toxicology 2005; 209:123-9. 2. Syn WK, Naisbitt DJ, Holt AP, Pirmohamed M, Mutimer DJ. Carbamazepineinduced acute liver failure as part of the DRESS syndrome. Int J Clin Pract 2005;59: 988-91. [Erratum, Int J Clin Pract 2005;59: 1371.] 3. Daly AK, Donaldson PT, Bhatnagar P, et al. HLA-B*5701 genotype is a major determinant of drug-induced liver injury due to flucloxacillin. Nat Genet 2009;41: 816-9. 4. Chung WH, Hung SI, Hong HS, et al. Medical genetics: a marker for StevensJohnson syndrome. Nature 2004;428:486. 5. Chen P, Lin J-J, Lu C-S, et al. Carbamazepine-induced toxic effects and HLAB*1502 screening in Taiwan. N Engl J Med 2011;364:1126-33. 6. Ding WY, Lee CK, Choon SE. Cutaneous adverse drug reactions seen in a tertiary hospital in Johor, Malaysia. Int J Dermatol 2010;49:834-41. 7. Locharernkul C, Loplumlert J, Limotai C, et al. Carbamazepine and phenytoin induced Stevens-Johnson syndrome is associated with HLA-B*1502 allele in Thai population. Epilepsia 2008;49:2087-91. [Erratum, Epilepsia 2009;50:971.] 8. Mehta TY, Prajapati LM, Mittal B, et al. Association of HLA-B*1502 allele and carbamazepine-induced Stevens-Johnson syndrome among Indians. Indian J Dermatol Venereol Leprol 2009;75:579-82. 9. Alfirevic A, Jorgensen AL, Williamson PR, Chadwick DW, Park BK, Pirmohamed M. HLA-B locus in Caucasian patients with carbamazepine hypersensitivity. Pharmacogenomics 2006;7:813-8. 10. Lonjou C, Borot N, Sekula P, et al. A European study of HLA-B in StevensJohnson syndrome and toxic epidermal necrolysis related to five high-risk drugs. Pharmacogenet Genomics 2008;18:99-107. 11. Lonjou C, Thomas L, Borot N, et al. A marker for Stevens-Johnson syndrome:

1142

ethnicity matters. Pharmacogenomics J 2006;6:265-8. 12. Ferrell PB Jr, McLeod HL. Carbamazepine, HLA-B*1502 and risk of StevensJohnson syndrome and toxic epidermal necrolysis: US FDA recommendations. Pharmacogenomics 2008;9:1543-6. 13. Alfirevic A, Mills T, Harrington P, et al. Serious carbamazepine-induced hypersensitivity reactions associated with the HSP70 gene cluster. Pharmacogenet Genomics 2006;16:287-96. 14. Hung SI, Chung WH, Jee SH, et al. Genetic susceptibility to carbamazepineinduced cutaneous adverse drug reactions. Pharmacogenet Genomics 2006;16:297306. 15. Kashiwagi M, Aihara M, Takahashi Y, et al. Human leukocyte antigen genotypes in carbamazepine-induced severe cutaneous adverse drug response in Japanese patients. J Dermatol 2008;35:683-5. 16. Pirmohamed M, Lin K, Chadwick D, Park BK. TNFalpha promoter region gene polymorphisms in carbamazepine-hypersensitive patients. Neurology 2001;56: 890-6. 17. Kardaun SH, Sidoroff A, Valeyrie-Allanore L, et al. Variability in the clinical pattern of cutaneous side-effects of drugs with systemic symptoms: does a DRESS syndrome really exist? Br J Dermatol 2007;156:609-11. 18. Speeckaert MM, Speeckaert R, Lambert J, Brochez L. Acute generalized exanthematous pustulosis: an overview of the clinical, immunological and diagnostic concepts. Eur J Dermatol 2010;20:425-33. 19. Auquier-Dunant A, Mockenhaupt M, Naldi L, Correia O, Schröder W, Roujeau JC. Correlations between clinical patterns and causes of erythema multiforme majus, Stevens-Johnson syndrome, and toxic epidermal necrolysis: results of an international prospective study. Arch Dermatol 2002;138:1019-24. 20. The International HIV Controllers Study (2010). The major genetic deter-

n engl j med 364;12

nejm.org

minants of HIV-1 control affect HLA class I peptide presentation. Science 2010; 330(6010):1551-7. 21. Li Y, Willer CJ, Ding J, Scheet P, Abecasis GR. MaCH: using sequence and genotype data to estimate haplotypes and unobserved genotypes. Genet Epidemiol 2010:34:816-34. 22. de Bakker PI, McVean G, Sabeti PC, et al. A high-resolution HLA and SNP haplotype map for disease association studies in the extended human MHC. Nat Genet 2006;38:1166-72. 23. Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007;81:559-75. 24. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005;21:263-5. 25. Hetherington S, Hughes AR, Mosteller M, et al. Genetic variations in HLA-B region and hypersensitivity reactions to abacavir. Lancet 2002;359:1121-2. 26. Mallal S, Nolan D, Witt C, et al. Association between presence of HLA-B*5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet 2002;359:727-32. 27. Mallal S, Phillips E, Carosi G, et al. HLA-B*5701 screening for hypersensitivity to abacavir. N Engl J Med 2008;358:56879. 28. Donaldson PT, Daly AK, Henderson J, et al. Human leucocyte antigen class II genotype in susceptibility and resistance to co-amoxiclav-induced liver injury. J Hepatol 2010;53:1049-53. 29. Hughes DA, Vilar FJ, Ward CC, Alfirevic A, Park BK, Pirmohamed M. Costeffectiveness analysis of HLA B*5701 genotyping in preventing abacavir hypersensitivity. Pharmacogenetics 2004;14: 335-42. 30. Saag M, Balu R, Phillips E, et al. High sensitivity of human leukocyte antigenB*5701 as a marker for immunologically

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

HLA-A*3101 and Carbamazepine-Induced Hypersensitivity confirmed abacavir hypersensitivity in white and black patients. Clin Infect Dis 2008;46:1111-8. 31. Schackman BR, Scott CA, Walensky RP, Losina E, Freedberg KA, Sax PE. The cost-effectiveness of HLA-B*5701 genetic screening to guide initial antiretroviral therapy for HIV. AIDS 2008;22:2025-33. 32. Ozeki T, Mushiroda T, Yowang A, et al. Genome-wide association study identifies HLA-A*3101 allele as a genetic risk factor for carbamazepine-induced cutaneous adverse drug reactions in Japanese population. Hum Mol Genet 2011;20: 1034-41. 33. Britschgi M, Steiner UC, Schmid S, et al. T-cell involvement in drug-induced acute generalized exanthematous pustulosis. J Clin Invest 2001;107:1433-41. 34. Schmidt AH, Baier D, Solloch UV, et al. Estimation of high-resolution HLA-A,

-B, -C, -DRB1 allele and haplotype frequencies based on 8862 German stem cell donors and implications for strategic donor registry planning. Hum Immunol 2009;70:895-902. 35. Wen SH, Lai MJ, Yang KL. Human leukocyte antigen-A, -B, and -DRB1 haplotypes of cord blood units in the Tzu Chi Taiwan Cord Blood Bank. Hum Immunol 2008;69:430-6. 36. Zucman D, Truchis P, Majerholc C, Stegman S, Caillat-Zucman S. Prospective screening for human leukocyte antigenB*5701 avoids abacavir hypersensitivity reaction in the ethnically mixed French HIV population. J Acquir Immune Defic Syndr 2007;45:1-3. 37. Lerch M, Pichler WJ. The immunological and clinical spectrum of delayed drug-induced exanthems. Curr Opin Allergy Clin Immunol 2004;4:411-9.

38. Wu Y, Farrell J, Pirmohamed M, Park

BK, Naisbitt DJ. Generation and characterization of antigen-specific CD4+, CD8+, and CD4+CD8+ T-cell clones from patients with carbamazepine hypersensitivity. J Allergy Clin Immunol 2007;119: 973-81. 39. Chelvanayagam G. A roadmap for HLA-A, HLA-B, and HLA-C peptide binding specificities. Immunogenetics 1996; 45:15-26. 40. Marson AG, Al-Kharusi AM, Alwaidh M, et al. The SANAD study of effectiveness of carbamazepine, gabapentin, lamotrigine, oxcarbazepine, or topiramate for treatment of partial epilepsy: an unblinded randomised controlled trial. Lancet 2007;369:1000-15. Copyright © 2011 Massachusetts Medical Society.

SPECIALTIES AND TOPICS AT NEJM.ORG

Specialty pages at the Journal’s 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 (1812–1989).

n engl j med 364;12

nejm.org

march 24, 2011

1143

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

Review article Genomic Medicine W. Gregory Feero, M.D., Ph.D., and Alan E. Guttmacher, M.D., Editors

Genomics and Drug Response Liewei Wang, M.D., Ph.D., Howard L. McLeod, Pharm.D., and Richard M. Weinshilboum, M.D. From the Division of Clinical Pharmacology, Department of Molecular Pharmacology and Experimental Therapeutics, Mayo Medical School, Mayo Clinic, Rochester, MN (L.W., R.M.W.); and the UNC Institute for Pharmacogenomics and Individualized Therapy, University of North Carolina, Chapel Hill (H.L.M). Address reprint requests to Dr. Weinshilboum at the Division of Clinical Pharmacology, Department of Molecular Pharmacology and Experimental Therapeutics, Mayo Clinic, 200 First St. SW, Rochester, MN 55905, or at [email protected]. N Engl J Med 2011;364:1144-53. Copyright © 2011 Massachusetts Medical Society.

P

harmacogenomics is the study of the role of inherited and acquired genetic variation in drug response.1 Clinically relevant pharmacogenetic examples, mainly involving drug metabolism, have been known for decades, but recently, the field of pharmacogenetics has evolved into “pharmacogenomics,” involving a shift from a focus on individual candidate genes to genomewide association studies. Such studies are based on a rapid scan of markers across the genome of persons affected by a particular disorder or drug-response phenotype and persons who are not affected, with tests for association that compare genetic variation in a case–control setting.2 An example is provided in this issue of the Journal: McCormack and colleagues, testing for genomewide association, identified an HLA allele that is associated with hypersensitivity reactions to the anticonvulsant and mood-stabilizing drug carbamazepine in persons of European descent.3 Pharmacogenomics facilitates the identification of biomarkers that can help physicians optimize drug selection, dose, and treatment duration and avert adverse drug reactions. In addition, pharmacogenomics can provide new insights into mechanisms of drug action and as a result can contribute to the development of new therapeutic agents. In 2003, two reviews of pharmacogenetics were published the Journal.4,5 Since then, both genomic science and its application to drug response have undergone major advances.6 Here we review some of those advances, with an emphasis on discovery through genomewide association studies. We describe examples that highlight principles of pharmacogenomics that are relevant to a wide variety of drugs. The Food and Drug Administration (FDA) has altered drug labels and issued warnings about pharmacogenomic variation affecting drug response, raising the issue of the level of evidence required to show clinical utility7 and the respective roles of regulatory agencies such as the FDA and of academic and professional societies in the evaluation of pharmacogenetic analyses for the clinic.

C a r diova scul a r Drugs Many drugs have proven efficacy in the treatment and prevention of cardiovascular disease. Not uncommonly, these drugs have narrow therapeutic indexes that are influenced by genetic variation — a hallmark of drugs for which pharmacogenomic approaches are likely to provide substantial clinical benefit. The anticoagulant agents warfarin and clopidogrel are high on the list of widely prescribed cardiovascular drugs with narrow therapeutic indexes. The pharmacogenomic features of these drugs illustrate the rapid evolution of our understanding of the role of inheritance in the variation in drug efficacy and the risk of adverse drug reactions. In the case of both agents, the application of classic candidate-gene pharmacogenetics 1144

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

genomic medicine

has identified important genomic markers of variation in efficacy and adverse reactions, observations that were subsequently confirmed in genomewide association studies. The FDA acted quickly on these data by relabeling warfarin and adding a warning box on the labeling for clopidogrel. Data supporting the clinical utility of routine use of pharmacogenetic testing for both these drugs are evolving.8-10 Warfarin is the most widely prescribed oral anticoagulant in North America and much of Europe.11 Despite the availability of the international normalized ratio (INR), a laboratory test that is universally used to measure the anticoagulant effect of warfarin, serious adverse responses, including hemorrhage and undesired coagulation, continue to complicate therapy, making warfarin one of the drugs most often responsible for emergency room visits.12,13 Chemically, warfarin is a racemic mixture (i.e., one that is composed of two enantiomorphic isomers). S-warfarin is three to five times as potent as R-warfarin as an anticoagulant, has a shorter half-life, and is metabolized predominantly by a cytochrome P-450 enzyme, CYP2C9.11 Two common CYP2C9 allozymes (see Glossary) have only a fraction of the level of enzyme activity of the wild-type allozyme CYP2C9*1: 12% for CYP2C9*2 and 5% for CYP2C9*3.11,14 More than a decade ago, it was reported that patients who required a low final dose of warfarin on the basis of INR values often carried one or two of these two common CYP2C9 variant alleles and were at increased risk for hemorrhage during warfarin therapy, presumably because they metabolized the drug more slowly.14 Those observations were confirmed, but it quickly became clear that the presence of CYP2C9 polymorphisms did not explain most of the variation in the final warfarin dose. Pharmacogenetic studies of warfarin changed dramatically in 2004 when the target for warfarinbased anticoagulants, vitamin K epoxide reductase complex subunit 1 (VKORC1), was identified,15,16 and single-nucleotide polymorphisms (SNPs) in VKORC1 were shown to be associated with the dose of warfarin required to achieve a target INR value.17 In 2009, a genomewide association study looked for associations between several hundred thousand SNPs and warfarin dose in about 1000 Swedish patients who were taking warfarin. The results showed two major signals in and around CYP2C9 and VKORC1 (Fig. 1A).18 When the authors removed the effects of those signals through

n engl j med 364;12

Glossary Allozyme: Alternate versions of an enzyme determined by genetic variants (alleles) present at a genetic locus. Gene cluster: Two or more genes in close physical proximity in the genome that encode similar gene products. Genomewide association study: An approach used in genetics research to look for associations between large numbers (typically hundreds of thousands) of specific genetic variations (most commonly single-nucleotide polymorphisms) and particular diseases. Single-nucleotide polymorphism: A single-nucleotide variation in a genetic sequence; a common form of variation in the human genome.

multiple regression adjustment, they observed an additional signal, implicating another cytochrome P450 gene (CYP4F2) (Fig. 1B). CYP4F2 was subsequently shown to catalyze vitamin K oxidation.19 The variant CYP4F2 allozyme shows decreased ability to catalyze the reaction, and as a result persons who carry the relevant genetic variant in CYP4F2 might require an increase in the warfarin dose (Fig. 1C). CYP2C9, VKORC1, and CYP4F2 have also been implicated in a genomewide association study of the administration of acenocoumarol, an anticoagulant related to warfarin.20 Taken together, CYP2C9 and VKORC1 genotypes explain about 30 to 40% of the total variation in the final warfarin dose.21 These observations raise the possibility that testing patients for variations in CYP2C9 and VKORC1 might provide information that could enhance clinical algorithms currently used to guide the administration of warfarin. To examine the potential clinical utility of testing for CYP2C9 and VKORC1 genotypes, in addition to INR monitoring and routine use of clinical algorithms, the International Warfarin Pharmacogenetics Consortium recently investigated the anticoagulant response to warfarin, as well as CYP2C9 and VKORC1 genotype data, for about 4000 persons of various ancestral origins. The investigators compared therapeutic outcomes with the application of standard clinical algorithms that included age, sex, and INR values and outcomes with the use of an algorithm that included CYP2C9 and VKORC1 genotype information and concluded that the addition of genotype information enhanced outcomes, especially for patients who required unusually high or low warfarin doses.22 CYP4F2 was not included in this algorithm but has been included in several algorithms developed more recently.23,24 Consistent with this conclusion are the results of a study comparing nearly 900 pa-

nejm.org

march 24, 2011

1145

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

n e w e ng l a n d j o u r na l

The

of

m e dic i n e

A Association with CYP2C9 and VKORC1 CYP2C9

−Log10 P Value

78 31

VKORC1

17

*2*3 composite *3

14

*2

7 0

1

2

3

4

5

6

7

8

9

10

11

12

Chromosome

13 14 15 16 17 18 19 21 X 20 22

Y

M

−Log10 P Value

B Association with CYP4F2 CYP4F2

14 7 0

1

2

3

4

5

6

7

8

9

10

11

12

Chromosome

13 14 15 16 17 18 19 21 X 20 22

Y

M

C Warfarin and the Vitamin K Cycle

7-OH-warfarin

CYP2C9

Warfarin

Vitamin K dihydroquinone

Carbon dioxide

and quinone reductase

Hydroxyvitamin K

7-OH-warfarin

CYP4F2

CYP2C9

Clotting protein

VKORC1

Vitamin K– dependent carboxylation

Vitamin K quinone VKORC1

Vitamin K epoxide

Warfarin

γ-Carboxyglutamyl clotting factors

Figure 1. Warfarin Pharmacogenomics. Panels A and B show Manhattan plots of P values (negative log10) for the association between single-nucleotidepolymorphisms (SNPs) across the genome and the final warfarin dose. The horizontal line indicates a P value of 1.5×10 −7, which is the level of genomewide statistical significance. In Panel A, the results of univariate regression analysis highlight SNP signals in or near CYP2C9 and VKORC1. In Panel B, the results of multivariate regression analysis with adjustment for the contributions of CYP2C9 and VKORC1 show the CYP4F2 signal on chromosome 19. (Data are from Takeuchi et al.18) The label *2 indicates the nonsynonymous SNP rs1799853, *3 indicates the nonsynonymous SNP rs1057910, and the *2*3 composite indicates the SNP rs4917639. M denotes mitochondrial SNPs. Panel C shows the sites of action of warfarin in the vitamin K cycle, as well as the roles of CYP2C9, CYP4F2, and VKORC1 in this process.

tients for whom genetic information on CYP2C9 and VKORC1 was made available to prescribing physicians with a matched historical control group of patients who were started on warfarin therapy without genetic information.25 Six months after the initiation of warfarin therapy, hospitalizations for hemorrhage were 28% less common in the group of patients for whom genetic information on CYP2C9 and VKORC1 had been supplied to prescribing physicians than in the control group (Fig. 2). 1146

n engl j med 364;12

The FDA revised the label on warfarin in February 2010, providing genotype-specific ranges of doses and suggesting that genotypes be taken into consideration when the drug is prescribed. The wide availability of CYP2C9 and VKORC1 genotyping and the release of both Web-based and personal decision-support tools have facilitated the clinical use of this information. Nevertheless, the clinical adoption of genotype-guided administration of warfarin has been slow, even though the evidence supporting such adoption is similar to

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

genomic medicine

n engl j med 364;12

A Hospitalization for Any Cause Risk of Hospitalization (%)

30 25

Hazard ratio, 0.69 (95% CI, 0.58–0.82) P<0.001 Historical control

20 15

Intervention group

10 5 0

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180

Days after Onset of Treatment

B Hospitalization for Bleeding or Thromboembolism 10

Risk of Hospitalization (%)

the evidence supporting currently used clinical variables, such as age, drug interactions, and ancestral origin. Some observers have expressed a need for prospective assessment of the value of this genetic information in warfarin therapy, and several prospective clinical trials are ongoing.26 Alternative anticoagulant therapies are also being developed that might replace warfarin, perhaps in patients with genotypes associated with extreme variation in warfarin response.27 Clopidogrel inhibits adenosine diphosphate (ADP)–stimulated platelet activation by binding irreversibly to a specific platelet receptor of ADP, P2Y12, thus inhibiting platelet aggregation.28,29 Dual antiplatelet therapy — clopidogrel and aspirin — has been shown to decrease the risk of subsequent ischemic vascular events.30-32 However, clopidogrel is a prodrug that requires metabolic activation in a reaction catalyzed by another cytochrome P-450 enzyme, CYP2C19. Like CYP2C9, CYP2C19 is genetically polymorphic with a common SNP that results in a truncated protein product with little enzymatic activity.33 Several studies have shown that genetic variation in CYP2C19 resulting in a paucity of activity is associated with decreased clopidogrel metabolic activation, a decreased antiplatelet effect, and an increased likelihood of a cardiovascular event.34,35 These observations have been confirmed in a genomewide association study.34 Early in 2010, the FDA added a boxed warning to prescribing information for clopidogrel, stating that persons with a CYP2C19 variant encoding a form of the enzyme associated with a low rate of metabolism might require dose adjustment or the use of a different drug.36 After this FDA action, the American Heart Association and the American College of Cardiology issued a joint endorsement of CYP2C19 genotyping for patients at moderate or high risk for cardiovascular events who are treated with clopidogrel.37 This genetic test is widely available in the United States. However, enthusiasm for its use has been muted, owing to a lack of clarity with regard to the optimal treatment of patients who carry a CYP2C19 variant, as shown by data from two large, randomized trials in which CYP2C19 genotyping did not have a significant effect on the incidence of cardiovascular events among patients with acute coronary syndromes or atrial fibrillation.38,39 On the other hand, in a recent meta-analysis of data from nine pharmacogenetic studies of clopido-

8

Hazard ratio, 0.72 (95% CI, 0.53–0.97) P=0.03 Historical control

6 4

Intervention group

2 0

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180

Days after Onset of Treatment

Figure 2. Risk of Hospitalization among Patients Who Underwent VKORC1 and CYP2C9 Genotyping, as Compared with a Historical Control Group, 6 Months after the Initiation of Warfarin Therapy. RETAKE: 1st AUTHOR: Weinshilbuom (Wang) Shown are the rates of hospitalization for any cause (Panel A) and 2ndfor bleedFIGURE: 2 of 3 (Panel B). There was a significant benefit 3rd ing or thromboembolism for paRevised tients who had undergone genotyping for the presence of VKORC1 and ARTIST: ts CYP2C9 variants that have been significantly associated SIZE with the risk of 4 col Line areCombo 4-C etH/T TYPE: Data over-anticoagulation. from Epstein al.25 22p3 AUTHOR, PLEASE NOTE: Figure has been redrawn and type has been reset. Please check carefully.

grel involving 9685 patients who had an acute JOB: 36412 or were undergoing percutaISSUE: coronary syndrome neous coronary intervention, there was a significant association between homozygosity or heterozygosity for CYP2C19 reduced-function alleles and an increased risk of death from cardiovascular causes, myocardial infarction, or stroke.9 At present, it is unclear whether genotyping to predict the response to clopidogrel is clinically useful. Several studies are under way to assess the effect of dose adjustment for clopidogrel in patients who carry CYP2C19 variant alleles.8

03-24-11

Agen t s Used for Infec t ious Dise a se s Genomewide association studies have confirmed the identity of genetic variants in previously im-

nejm.org

march 24, 2011

1147

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

plicated candidate genes that contribute to clinically important outcomes, including severe idiosyncratic adverse reactions and variation in drug efficacy. In the next set of examples, the results of pharmacogenomic studies were unanticipated. Hepatotoxicity is the most common reason for the termination of clinical trials investigating the efficacy of new drugs, accounting for approximately 33% of such terminations, and is a major reason for postmarketing drug withdrawal.40 Floxacillin, an antibiotic used in Europe and Australia to treat staphylococcal infections, has been associated with an unusual form of cholestatic hepatitis, with an estimated incidence of approximately 8.5 cases per 100,000 patients.41-44 A multicenter genomewide association study, reported in 2009, analyzed the genotypes of 51 persons with floxacillin-induced hepatic injury and 282 matched controls.45 A SNP in the major histocompatibility complex and closely linked with HLA-B*5701 showed very strong association with hepatic injury. The association between the presence of HLA-B*5701 and hypersensitivity reactions to abacavir, a nucleoside analogue used to treat human immunodeficiency virus type 1 infection, had already been reported,46-48 which resulted in the FDA modification of the abacavir label to include a recommendation that patients undergo genotyping for HLA-B*5701 before the initiation of therapy.49 Rare but severe adverse events represent a major reason why drugs are withdrawn after FDA approval. Although it was possible to attempt a replication of the association between the variant in HLA-B*5701 and floxacillin-induced hepatitis,45 it is often difficult to gather enough cases of rare adverse drug reactions to apply genomewide techniques. This situation presents a challenge for regulators. To date, the FDA has generally chosen to include pharmacogenetic information relevant to rare severe adverse events on drug labels — even when the association between the variant and drug response has not been replicated — so as to warn prescribers of potential risk.50 This approach places a burden on clinicians to use their own judgment regarding the need for pharmacogenetic testing before prescribing a drug. In contrast with unreplicated tests for association are prospective trials of genotyping to avoid adverse pharmacogenetic effects. One such study is reported in this issue of the Journal,51 in which investigators observed no instances of the Stevens– Johnson syndrome or toxic epidermal necrolysis in 1148

n engl j med 364;12

of

m e dic i n e

a sample of nearly 5000 Taiwanese candidates for carbamazepine therapy, among whom carbamazepine had been withheld from carriers of the HLA-B*1502 allele, which has been reported to be associated with the Stevens–Johnson syndrome in Han Chinese.52 Another pharmacogenomic example involving agents used to treat infectious diseases concerns the treatment of chronic infection with hepatitis C virus (HCV), which develops in approximately 80% of patients who are infected with the virus and is a major cause of liver failure.53,54 Successful treatment of chronic HCV infection involves a sustained virologic response, which is defined by an undetectable level of HCV RNA in plasma. Unfortunately, only 40 to 50% of patients who are infected with HCV genotype 1 have a sustained virologic response when receiving the current standard of care for the treatment of chronic HCV infection — injections of pegylated interferon alfa together with oral ribavirin for 48 weeks.53,54 The ability to identify patients with a differential response to pegylated interferon alfa is important in the current era of new anti-HCV drugs because pegylated interferon alfa remains the backbone of therapy, to which many of these new agents are added. Recently, in three independent genomewide association studies55-57 involving patients with chronic HCV infection who were treated with pegylated interferon alfa and ribavirin, there was an association between a variant in IL28B, the gene encoding interleukin28B, and the drug response. In one of these studies, peripheral-blood mononuclear cells from patients carrying the variant allele that was associated with a poor response had comparatively low levels of IL28B expression.56 IL28B encodes a protein that is thought to be involved in suppressing the replication of a number of viruses, including HCV.55-58 This example shows how pharmacogenomic genomewide association studies not only have identified biomarkers of response to pegylated interferon alfa but also have provided insights that might be used to determine therapeutic approaches to this chronic infection and to select a drug target for therapeutic development.

A n t ineopl a s t ic Drugs The field of cancer pharmacogenomics is complicated by the fact that two genomes are involved: the germline genome of the patient and the somatic genome of the tumor. Obviously, the tumor

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

genomic medicine

genome plays a critical role in the variation in response to antineoplastic therapy. Prominent examples include HER2 overexpression or amplification in patients with breast cancer and the response of these tumors to trastuzumab59,60 and increased sensitivity to the epidermal growth factor receptor (EGFR) antagonist gefitinib among patients with non–small-cell lung cancer who have activating mutations in the gene encoding EGFR.61,62 A recent example involves melanoma and a mutation in BRAF encoding a serine–threonine protein kinase. Since a specific inhibitor, PLX4032, targets the mutant activated kinase, there is a pharmacogenetic effect in that PLX4032 prolongs survival in patients carrying the mutation. This clinical finding was based on the discovery of a BRAF mutation through the sequencing of a large number of kinase genes in tumors.63,64 On the other hand, germline SNPs in the gene encoding the enzyme thiopurine S-methyltransferase (TPMT) can result in increased sensitivity to mercaptopurine as a result of decreased metabolism,1,4,5 whereas the number of TA dinucleotide repeats in the promoter of UGT1A1 in germline DNA can increase the toxic effects of irinotecan, also as a result of decreased metabolism.1,65 There are now many examples of pharmacogenetic tests paired with anticancer drugs that are considered part of routine oncologic care (Table 1). The fact that clinically relevant pharmacogenomic variation in both the tumor genome and the patient’s germline genome can influence the response to antineoplastic therapy is illustrated in Figure 3, with gefitinib and irinotecan as examples.

A rom ata se Inhibi t or s Genetic polymorphisms in a patient’s germline genome can also play an important role in variation in the response to cancer therapy. Endocrine therapy of breast cancer66,67 offers a striking example of how a genomewide association study has lead to the identification of a mechanism that would seem to be responsible for a serious drug-induced adverse reaction that limits therapeutic options for some patients. The tumors of approximately 70% of postmenopausal women with breast cancer express the estrogen receptor. The blockade of this receptor with tamoxifen or the blockade of estrogen synthesis through the inhibition of aromatase (which catalyzes estrogen synthesis) halves the recurn engl j med 364;12

Table 1. Anticancer Drugs Approved by the Food and Drug Administration (FDA) with Labeling Regarding Pharmacogenomic Biomarkers.* Type of Biomarker and Associated Drug Biomarker with pharmacokinetic effect TPMT Mercaptopurine Thioguanine UGT1A1 Irinotecan Nilotinib Biomarker with pharmacodynamic effect EGFR Cetuximab Erlotinib Gefitinib Panitumumab KRAS Cetuximab Panitumumab ABL Imatinib Dasatinib Nilotinib C-Kit (KIT) Imatinib HER2/neu (ERBB2) Lapatinib Trastuzumab Estrogen receptor Tamoxifen * Data are from the FDA’s pharmacogenetics Web site (www.fda.gov/Drugs/ScienceResearch/ResearchAreas/ Pharmacogenetics/ucm083378.htm). The biomarkers have been separated into pharmacokinetic effect (drug metabolism) and pharmacodynamic effect (drug target). Biomarkers for cytogenetic alterations have been excluded.

rence rate.66-68 However, the administration of an aromatase inhibitor can also result in severe musculoskeletal pain that leads women (10 to 20% in some studies) to terminate therapy.67 In a genomewide association study that used DNA samples from a large clinical trial of aromatase inhibitors to treat women with breast cancer (called MA.27) (ClinicalTrials.gov number, NCT00968214), there was an association between musculoskeletal pain and variants in the gene cluster encoding T-cell leukemia–lymphoma (TCL) proteins. The marker showing the strongest (although not significant) association created a new estrogen-response element close to TCL1A.69 Functional studies showed that the markers that were associated with susceptibility to musculoskeletal pain were also as-

nejm.org

march 24, 2011

1149

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

A

n e w e ng l a n d j o u r na l

of

m e dic i n e

Tumor genome EGF

EGF receptor Gefitinib

Tumor genome Lung cancer

Gefitinib (kinase inhibitor)

EGFR activating mutations

Increased tumor sensitivity to gefitinib

B

Germline genome

Low expression of UGT1A1 and low level of glucuronidation Germline genome SN-38 (active drug)

Irinotecan UGT1A1*28 (TA)7TAA Healthy liver Irinotecan (prodrug)

SN-38 (active drug)

UGT1A1 (TA)6TAA

SN-38 glucuronide (inactive metabolite)

Colon cancer Germline genome High expression of UGT1A1 and high level of glucuronidation

Figure 3. Cancer Pharmacogenomics and Tumor and Germline Genomes. Both the tumor genome (e.g., in the case of gefitinib therapy) and the patient’s germline genome (e.g., in the case of irinotecan therapy) can contribute to pharmacogenomic variation in response to antineoplastic drugs. The tumor COLOR FIGURE genome plays a critical role in the response to gefitinib (Panel A), since the sensitivity of non–small-cell lung cancer Rev5 epidermal growth 03/04/11 to this drug is enhanced by activating mutations in the kinase domain of the gene encoding factor receptor (EGFR).58,59 Tumor EGFR encoding activating mutations within the kinase domain results in enhanced tuAuthor Dr. Weinshilboum mor sensitivity to gefitinib. The rate of toxic effects associated with irinotecan (diarrhea is Fig # and myelosuppression) 3 increased in patients with seven TA dinucleotide repeats rather than the more common Title six repeats in the promoter region of UGT1A1 encoding a UDP-glucuronosyltransferase in germline DNA, resulting in lower enzyme activity and ME a decreased rate of drug metabolism (Panel B).1,62 DE Phimister Daniel Muller Artist

sociated with increased TCL1A expression after estrogen exposure. TCL1A regulates the expression of interleukin-17 receptor A, an experimental target for the treatment of patients with rheumatoid arthritis.70 These observations, if confirmed, may provide new insight into the relationship between estrogens and joint pain. 1150

n engl j med 364;12

AUTHOR PLEASE NOTE:

Figure has been redrawn and type has been reset This example illustrates several challenges and Please check carefully opportunities associated with pharmacogenomic Issue date 03/24/2011 studies and their application to clinical practice. First, associations that are uncovered by genomewide association studies require replication if there are appropriate sample sets. However, MA.27 is a large clinical trial of aromatase inhibitors, span-

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

genomic medicine

ning 8 years at a cost of more than $35 million. Therefore, identifying a large and appropriate sample to test for replication will be difficult. In cases in which replication samples are not available or are difficult to obtain, pharmacogenomic studies may benefit from the use of functional validation to help verify the results of genomewide studies. For example, the biologic plausibility that is provided by the functional data (i.e., the association between phenotype-associated markers and TCL1A expression) increases confidence that the genetic association is driven by biology rather than chance. A final consideration is the clinical context. Because aromatase inhibitors have only a slight benefit over tamoxifen in the treatment of breast cancer, and tamoxifen is much less expensive than aromatase inhibitors, a clear therapeutic alternative is available for patients at increased risk for musculoskeletal pain. Therefore, a genetic test with sufficient predictive power to identify such patients might be clinically useful.

Cl inic a l T r a nsl at ion The use of genotyping to inform clinical decisions about drug use is not widely practiced. The slow pace of the clinical application of pharmacogenomics has many causes. Obviously, the most important issue is the need to establish clinical utility in order to support the value of genotyping. In the absence of such evidence, payers will be unlikely to provide reimbursement for routine use of pharmacogenetic testing, and tests will remain inaccessible to the majority of patients. There seems to be little consensus on the level or nature of data required to establish clinical utility.7 No matter what level of evidence is required for each situation, it will be necessary to develop simple clinical algorithms to aid physicians in their interpretation and use of genetic data. This goal may be best achieved through the development of point-of-care tools embedded in electronic medical record systems. Even with such tools, physicians and other health care providers need to be aware of this area of biomedical science in order to apply the information clinically. A major effort will be required to educate all members of the health care team about clinical genomics.

n engl j med 364;12

In recent years, the FDA has aggressively pursued drug-label modification when excess risk can be convincingly linked to a genetic marker. Several of the examples have been described here; many more are listed in the FDA’s Table of Pharmacogenomic Biomarkers in Drug Labels (www .fda.gov/Drugs/ScienceResearch/ResearchAreas/ Pharmacogenetics/ucm083378.htm). Warnings that the FDA has issued about the prescription of clopidogrel and abacavir without testing of the relevant genotype are examples of the agency’s increasingly activist stance.

C onclusions There has been a good deal of comment in the scientific literature71-74 and the popular press75 about the slow pace of the application of genomics to clinical medicine. We hope that we have provided some reassurance that advances resulting from the application of genomic science to drug therapy may be helpful in drug selection and administration and reduce the odds of adverse drug reactions. Challenges that are associated with the replication of study findings and the development of proof of the clinical significance of implicated variants underscore the importance of functional experiments to test for biologic plausibility and to extend our understanding of drug mechanisms. Finally, a blend of scientific, regulatory, and psychological factors must be addressed if pharmacogenomic tests are to become a routine part of clinical practice. The FDA-mandated incorporation of pharmacogenomic information in drug labeling will remain an important step in the acceptance of pharmacogenomics in clinical practice. Perhaps equally important will be the willingness of physicians to reexamine suboptimal pharmacologic management programs. Supported in part by grants (K22-CA130828, R01-CA138461, R01-GM28157, R01-CA132780, U01-HG005137, U19-GM61388, U01-GM63340, and UL1RR025747) from the National Institutes of Health, a PhRMA Foundation Center of Excellence in Clinical Pharmacology Award, and an American Society for Pharmacology and Experimental Therapeutics (ASPET)–Astellas Award. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. We thank Luanne Wussow, Linda Pelleymounter, and Katarzyna Ellsworth for their assistance in the preparation of the manuscript.

nejm.org

march 24, 2011

1151

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

References 1. Weinshilboum RM, Wang L. Pharma-

cogenetics and pharmacogenomics: development, science, and translation. Annu Rev Genomics Hum Genet 2006;7:223-45. 2. Manolio TA. Genomewide association studies and assessment of the risk of disease. N Engl J Med 2010;363:166-76. 3. McCormack M, Alfirevic A, Bourgeois S, et al. HLA-A*3101 and carbamazepineinduced hypersensitivity reactions in Europeans. N Engl J Med 2011;364:1134-43. 4. Weinshilboum R. Inheritance and drug response. N Engl J Med 2003;348: 529-37. 5. Evans WE, McLeod HL. Pharmacogenomics — drug disposition, drug targets, and side effects. N Engl J Med 2003;348: 538-49. 6. Feero WG, Guttmacher AE, Collins FS. Genomic medicine — an updated primer. N Engl J Med 2010;362:2001-11. 7. Woodcock J. Assessing the clinical utility of diagnostics used in drug therapy. Clin Pharmacol Ther 2010;88:765-73. 8. Fuster V, Sweeny JM. Clopidogrel and the reduced-function CYP2C19 genetic variant: a limited piece of the overall therapeutic puzzle. JAMA 2010;304:1839-40. 9. Mega JL, Simon T, Collet J-P, et al. Reduced-function CYP2C19 genotype and risk of adverse clinical outcomes among patients treated with clopidogrel predominantly for PCI: a meta-analysis. JAMA 2010;304:1821-30. 10. Paré G, Mehta SR, Yusuf S, et al. Effects of CYP2C19 genotype on outcomes of clopidogrel treatment. N Engl J Med 2010;363:1704-14. 11. Daly AK, King BP. Pharmacogenetics of oral anticoagulants. Pharmacogenetics 2003;13:247-52. 12. Budnitz DS, Shehab N, Kegler SR, Richards CL. Medication use leading to emergency department visits for adverse drug events in older adults. Ann Intern Med 2007;147:755-65. 13. Wysowski DK, Nourjah P, Swartz L. Bleeding complications with warfarin use: a prevalent adverse effect resulting in regulatory action. Arch Intern Med 2007; 167:1414-9. 14. Aithal GP, Day CP, Kesteven PJ, Daly AK. Association of polymorphisms in the cytochrome P450 CYP2C9 with warfarin dose requirement and risk of bleeding complications. Lancet 1999;353:717-9. 15. Li T, Chang CY, Jin DY, Lin PJ, Khvorova A, Stafford DW. Identification of the gene for vitamin K epoxide reductase. Nature 2004;427:541-4. 16. Rost S, Fregin A, Ivaskevicius V, et al. Mutations in VKORC1 cause warfarin resistance and multiple coagulation factor deficiency type 2. Nature 2004;427:53741. 17. Rieder MJ, Reiner AP, Gage BF, et al. Effect of VKORC1 haplotypes on tran-

1152

scriptional regulation and warfarin dose. N Engl J Med 2005;352:2285-93. 18. Takeuchi F, McGinnis R, Bourgeois S, et al. A genome-wide association study confirms VKORC1, CYP2C9, and CYP4F2 as principal genetic determinants of warfarin dose. PLoS Genet 2009;5(3):e1000433. 19. McDonald MG, Rieder MJ, Nakano M, Hsia CK, Rettie AE. CYP4F2 is a vitamin K1 oxidase: an explanation for altered warfarin dose in carriers of the V433M variant. Mol Pharmacol 2009;75:1337-46. 20. Teichert M, Eijgelsheim M, Rivadeneira F, et al. A genome-wide association study of acenocoumarol maintenance dosage. Hum Mol Genet 2009;18:3758-68. 21. Manolopoulos VG, Ragia G, Tavridou A. Pharmacogenetics of coumarinic oral anticoagulants. Pharmacogenomics 2010; 11:493-6. 22. The International Warfarin Pharmacogenetics Consortium. Estimation of the warfarin dose with clinical and pharmacogenetic data. N Engl J Med 2009;360: 753-64. [Erratum, N Engl J Med 2009;361: 1613.] 23. Sagreiya H, Berube C, Wen A, et al. Extending and evaluating a warfarin dosing algorithm that includes CYP4F2 and pooled rare variants of CYP2C9. Pharmacogenet Genomics 2010;20:407-13. 24. Zambon C-F, Pengo V, Padrini R, et al. VKORC1, CYP2C9 and CYP4F2 geneticbased algorithm for warfarin dosing: an Italian retrospective study. Pharmacogenomics 2010;12:15-25. 25. Epstein RS, Moyer TP, Aubert RE, et al. Warfarin genotyping reduces hospitalization rates: results from the MM-WES (Medco-Mayo Warfarin Effectiveness Study). J Am Coll Cardiol 2010;55:280412. 26. Ginsburg GS, Voora D. The long and winding road to warfarin pharmacogenetic testing. J Am Coll Cardiol 2010;55: 2813-5. 27. Kanagasabapathy P, Chowdary P, Gatt A. Alternatives to warfarin — the next generation of anticoagulants. Cardiovasc Ther 2010 July 8 (Epub ahead of print). 28. Hollopeter G, Jantzen HM, Vincent D, et al. Identification of the platelet ADP receptor targeted by antithrombotic drugs. Nature 2001;409:202-7. 29. Savi P, Pereillo JM, Uzabiaga MF, et al. Identification and biological activity of the active metabolite of clopidogrel. Thromb Haemost 2000;84:891-6. 30. Yusuf S, Zhao F, Mehta SR, et al. Effects of clopidogrel in addition to aspirin in patients with acute coronary syndromes without ST-segment elevation. N Engl J Med 2001;345:494-502. [Errata, N Engl J Med 2001;345:1506, 1716.] 31. Sabatine MS, Cannon CP, Gibson CM, et al. Addition of clopidogrel to aspirin and fibrinolytic therapy for myocardial in-

n engl j med 364;12

nejm.org

farction with ST-segment elevation. N Engl J Med 2005;352:1179-89. 32. King SB III, Smith SC Jr, Hirshfeld JW Jr, et al. 2007 Focused Update of the ACC/ AHA/SCAI 2005 Guideline Update for Percutaneous Coronary Intervention: a report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines: 2007 Writing Group to Review New Evidence and Update the ACC/AHA/SCAI 2005 Guideline Update for Percutaneous Coronary Intervention, writing on behalf of the 2005 Writing Committee. Circulation 2008;117:261-95. [Erratum, Circulation 2008;117(6):e161.] 33. de Morais SM, Wilkinson GR, Blaisdell J, Nakamura K, Meyer UA, Goldstein JA. The major genetic defect responsible for the polymorphism of S-mephenytoin metabolism in humans. J Biol Chem 1994; 269:15419-22. 34. Shuldiner AR, O’Connell JR, Bliden KP, et al. Association of cytochrome P450 2C19 genotype with the antiplatelet effect and clinical efficacy of clopidogrel therapy. JAMA 2009;302:849-57. 35. Mega JL, Close SL, Wiviott SD, et al. Cytochrome P-450 polymorphisms and response to clopidogrel. N Engl J Med 2009; 360:354-62. 36. FDA drug safety communication: reduced effectiveness of Plavix (clopidogrel) in patients who are poor metabolizers of the drug. (http://www.fda.gov/drugs/ drugsafety/PostmarketDrugSafety InformationforPatientsandProviders/ ucm203888.htm.) 37. Holmes DR Jr, Dehmer GJ, Kaul S, Leifer D, O’Gara PT, Stein CM. ACCF/AHA clopidogrel clinical alert: approaches to the FDA “boxed warning”: a report of the American College of Cardiology Foundation Task Force on clinical expert consensus documents and the American Heart Association endorsed by the Society for Cardiovascular Angiography and Interventions and the Society of Thoracic Surgeons. J Am Coll Cardiol 2010;56:321-41. 38. Wallentin L, James S, Storey RF, et al. Effect of CYP2C19 and ABCB1 single nucleotide polymorphisms on outcomes of treatment with ticagrelor versus clopidogrel for acute coronary syndromes: a genetic substudy of the PLATO trial. Lancet 2010;376:1320-8. 39. Paré G, Mehta SR, Yusuf S, et al. Effects of CYP2C19 genotype on outcomes of clopidogrel treatment. N Engl J Med 2010;363:1704-14. 40. Mann R, Andrews E, eds. Pharmacovigilance. 2nd ed. Chichester, England: John Wiley and Sons, 2006. 41. Olsson R, Wiholm BE, Sand C, Zettergren L, Hultcrantz R, Myrhed M. Liver damage from flucloxacillin, cloxacillin and dicloxacillin. J Hepatol 1992;15:154-61. 42. Fairley CK, McNeil JJ, Desmond P, et

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

genomic medicine al. Risk factors for development of flucloxacillin associated jaundice. BMJ 1993; 306:233-5. [Erratum, BMJ 1993;307:1179.] 43. Russmann S, Kaye JA, Jick SS, Jick H. Risk of cholestatic liver disease associated with flucloxacillin and flucloxacillin prescribing habits in the UK: cohort study using data from the UK General Practice Research Database. Br J Clin Pharmacol 2005;60:76-82. 44. Pachkoria K, Lucena MI, Crespo E, et al. Analysis of IL-10, IL-4 and TNF-alpha polymorphisms in drug-induced liver injury (DILI) and its outcome. J Hepatol 2008;49:107-14. [Erratum, J Hepatol 2009; 50:636.] 45. Daly AK, Donaldson PT, Bhatnagar P, et al. HLA-B*5701 genotype is a major determinant of drug-induced liver injury due to flucloxacillin. Nat Genet 2009;41: 816-9. 46. Hetherington S, McGuirk S, Powell G, et al. Hypersensitivity reactions during therapy with the nucleoside reverse transcriptase inhibitor abacavir. Clin Ther 2001;23:1603-14. 47. Mallal S, Nolan D, Witt C, et al. Association between presence of HLAB*5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet 2002;359: 727-32. 48. Mallal S, Phillips E, Carosi G, et al. HLA-B*5701 screening for hypersensitivity to abacavir. N Engl J Med 2008;358: 568-79. 49. Information for healthcare professionals: abacavir (marketed as Ziagen) and abacavir-containing medications. (http://www.fda.gov/Drugs/DrugSafety/ PostmarketDrugSafetyInformationfor PatientsandProviders/ucm123927.htm.) 50. Ferrell PB Jr, McLeod HL. Carbamazepine, HLA-B*1502 and risk of StevensJohnson syndrome and toxic epidermal necrolysis: US FDA recommendations. Pharmacogenomics 2008;9:1543-6. 51. Chen P, Lin J-J, Lu C-S, et al. HLA-

B*1502 screening to prevent carbamazepine-induced Stevens–Johnson syndrome. N Engl J Med 2011;364:1126-33. 52. Chung WH, Hung SI, Hong HS, et al. Medical genetics: a marker for StevensJohnson syndrome. Nature 2004;428:486. 53. NIH consensus statement on management of hepatitis C. NIH Consens State Sci Statements 2002;19(3):1-46. 54. Shepard CW, Finelli L, Alter MJ. Global epidemiology of hepatitis C virus infection. Lancet Infect Dis 2005;5:558-67. 55. Suppiah V, Moldovan M, Ahlenstiel G, et al. IL28B is associated with response to chronic hepatitis C interferon-alpha and ribavirin therapy. Nat Genet 2009;41: 1100-4. 56. Tanaka Y, Nishida N, Sugiyama M, et al. Genome-wide association of IL28B with response to pegylated interferon-alpha and ribavirin therapy for chronic hepatitis C. Nat Genet 2009;41:1105-9. 57. Ge D, Fellay J, Thompson AJ, et al. Genetic variation in IL28B predicts hepatitis C treatment-induced viral clearance. Nature 2009;461:399-401. 58. O’Brien TR. Interferon-alfa, interferonlambda and hepatitis C. Nat Genet 2009; 41:1048-50. 59. Eisenhauer EA. From the molecule to the clinic — inhibiting HER2 to treat breast cancer. N Engl J Med 2001;344: 841-2. 60. Slamon DJ, Leyland-Jones B, Shak S, et al. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med 2001;344:783-92. 61. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non–small-cell lung cancer to gefitinib. N Engl J Med 2004;350:2129-39. 62. Paez JG, Jänne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004;304:1497-500. 63. Davies H, Bignell GR, Cox C, et al.

Mutations of the BRAF gene in human cancer. Nature 2002;417:949-54. 64. Flaherty KT, Puzanov I, Kim KB, et al. Inhibition of mutated, activated BRAF in metastatic melanoma. N Engl J Med 2010; 363:809-19. 65. Iyer L, King CD, Whitington PF, et al. Genetic predisposition to the metabolism of irinotecan (CPT-11): role of uridine diphosphate glucuronosyltransferase isoform 1A1 in the glucuronidation of its active metabolite (SN-38) in human liver microsomes. J Clin Invest 1998;101:847-54. 66. Freedman RA, Winer EP. Adjuvant therapy for postmenopausal women with endocrine-sensitive breast cancer. Breast 2010;19:69-75. 67. Janni W, Hepp P. Adjuvant aromatase inhibitor therapy: outcomes and safety. Cancer Treat Rev 2010;36:249-61. 68. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15year survival: an overview of the randomised trials. Lancet 2005;365:1687-717. 69. Ingle JN, Schaid DJ, Goss PE, et al. Genome-wide associations and functional genomic studies of musculoskeletal adverse events in women receiving aromatase inhibitors. J Clin Oncol 2010;28: 4674-82. 70. Miossec P, Korn T, Kuchroo VK. Interleukin-17 and type 17 helper T cells. N Engl J Med 2009;361:888-98. 71. The human genome at ten. Nature 2010;464:649-50. 72. Collins F. Has the revolution arrived? Nature 2010;464:674-5. 73. Venter JC. Multiple personal genomes await. Nature 2010;464:676-7. 74. Varmus H. Ten years on — the human genome and medicine. N Engl J Med 2010;362:2028-9. 75. Wade N. A decade later, genetic map yields few new cures. New York Times. June 12, 2010. Copyright © 2011 Massachusetts Medical Society.

SPECIALTIES AND TOPICS AT NEJM.ORG

Specialty pages at the Journal’s 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 (1812–1989).

n engl j med 364;12

nejm.org

march 24, 2011

1153

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

images in clinical medicine

Pyogenic Liver Abscess B

A

Chin-Wei Yu, M.D. Ching-Hsing Lee, M.D. Chang Gung Memorial Hospital Taoyuan, Taiwan

A

44-year-old man with diabetes mellitus presented to our hospital after 4 days of fever and abdominal pain. The initial evaluation revealed tachycardia (heart rate, 137 beats per minute), hypotension (blood pressure, 81/44 mm Hg), and abdominal discomfort in the right upper quadrant. There was no rebound tenderness. A lesion with heterogeneous radiodensity was noted in the right upper abdomen on chest radiography (Panel A, arrowheads). Computed tomographic imaging revealed an intrahepatic lesion containing gas and fluid (Panel B, arrowheads). A pyogenic liver abscess was suspected. Blood cultures ultimately grew Klebsiella pneumoniae, which can be gas-producing. Diabetes is an important risk factor for this condition. Despite fluid resuscitation and treatment with inotropic agents and antibiotics, the patient’s clinical condition deteriorated, and he died within 48 hours after admission.

Copyright © 2011 Massachusetts Medical Society.

1154

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

case records of the massachusetts general hospital Founded by Richard C. Cabot Nancy Lee Harris, m.d., Editor Eric S. Rosenberg, m.d., Associate Editor Jo-Anne O. Shepard, m.d., Associate Editor Alice M. Cort, m.d., Associate Editor Sally H. Ebeling, Assistant Editor Christine C. Peters, Assistant Editor

Case 9-2011: A 37-Year-Old Man with Flushing and Hypotension Mandakolathur R. Murali, M.D., Mariana C. Castells, M.D., Ph.D., James Y. Song, M.D., David M. Dudzinski, M.D., J.D., and Robert P. Hasserjian, M.D.

Pr e sen tat ion of C a se Dr. Leana S. Wen (Emergency Medicine): A 37-year-old man was admitted to the hospital because of flushing and hypotension. That morning, sneezing, rhinorrhea, scratchy throat, and subjective fever had developed. After lunch, he took an over-thecounter cold preparation that included acetylsalicylic acid, dextromethorphan hydrobromide, and phenylephrine bitartrate. Thirty minutes later, fatigue, weakness, nausea, epigastric pain, facial flushing, and “beefy red” ears developed, accompanied by two episodes of nonbloody emesis. Light-headedness, diaphoresis, and nearsyncope developed. He was caught by a coworker as he fell to the ground, without head trauma or loss of consciousness. Emergency medical services were called. On evaluation, his skin was flushed; the systolic blood pressure was 50 mm Hg, the pulse 56 beats per minute and regular, the respiratory rate 16 breaths per minute, and the oxygen saturation 100% while he was breathing supplemental oxygen by means of a nonrebreather face mask. An electrocardiogram (ECG) showed sinus bradycardia with T-wave inversions. Ondansetron and normal saline were administered intravenously. He was transported to the emergency department at this hospital. On arrival, the patient reported pleuritic chest pain and worsening diffuse abdominal discomfort. He had had multiple similar episodes during the previous 12 years, with flushing, conjunctival injection, vomiting, and diarrhea. These episodes had increased in frequency in the past year, from approximately twice a year to once every 2 months. The symptoms, which were usually provoked by physical exertion, mental stress, or intense emotion, lasted up to 12 hours and were followed by weakness of 3 to 4 days’ duration. Evaluations at other hospitals had shown systolic pressures as low as 60 mm Hg. The episodes were attributed to dehydration and stress, and they were treated with intravenous normal saline (≤5 liters). The patient had no other illnesses and no known allergies to medications and took no other medications. Kiwifruit had caused throat constriction and vomiting. He worked in a health-related field and participated in triathlons. He drank alcohol occasionally, had smoked in the past, and did not use illicit drugs or over-thecounter herbal preparations. There were no recent exposures to ill persons, seafood,

n engl j med 364;12

nejm.org

march 24, 2011

From the Divisions of Rheumatology, Allergy, and Immunology (M.R.M.) and Cardiology (D.M.D.); and the Departments of Radiology (J.Y.S.) and Pathology (R.P.H.), Massachusetts General Hospital; the Division of Rheumatology, Allergy, and Immunology, Brigham and Women’s Hospital (M.C.C.); and the Departments of Medicine (M.R.M., M.C.C., D.M.D.), Radiology (J.Y.S.), and Pathology (R.P.H.), Harvard Medical School — all in Boston. N Engl J Med 2011;364:1155-65. Copyright © 2011 Massachusetts Medical Society.

1155

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

or unusual foods. He was of mixed European and Caribbean ancestry. There was no family history of renal disease, anaphylaxis, or rheumatologic diseases. On examination, the patient was alert and oriented. The temperature was 36.2°C (over the temporal artery), the blood pressure 57/33 mm Hg, the pulse up to 116 beats per minute and irregular, the respiratory rate 26 breaths per minute, and the oxygen saturation 93% while he was breathing 4 liters of oxygen by nasal cannula. There was scleral injection and diffuse blanching erythema over the upper body; the extremities were warm and well perfused, with normal capillary refill. The examination was otherwise normal. Levels of sodium, chloride, carbon dioxide, D -dimer, magnesium, amylase, lipase, creatine kinase MB isoenzymes, and troponin T; tests of liver function; the activated partial-thromboplastin time; and the red-cell indexes were normal. Screening for troponin I and serum toxins was negative. The results of other laboratory tests are shown in Table 1. An ECG showed atrial fibrillation with a ventricular rate of 102 beats per minute, with STsegment elevations (2 mm, convex) in leads aVR and V1 and downsloping ST-segment depressions (3 to 4 mm) in leads I, II, III, aVF, and V3 through V6 (Fig. 1 in the Supplementary Appendix, available with the full text of this article at NEJM.org). Normal saline was transfused rapidly through two peripheral intravenous catheters, and 325 mg of aspirin was administered. The patient then had increased flushing. Fifty minutes after his arrival, the blood pressure was 140/81 mm Hg. A chest radiograph showed low lung volumes with interstitial prominence and was otherwise normal. During the next 7 hours, the temperature rose to 38.1°C and abdominal pain increased, with tenderness in the upper abdomen. Normal saline, potassium chloride, magnesium sulfate, ondansetron, and metoclopramide were administered intravenously, and his nausea transiently improved. Urinalysis showed yellow clear urine, with trace ketones, trace urobilinogen, 1+ bilirubin, and 2+ albumin, with 20 to 50 white cells (reference range, 0 to 2) and a few squamous cells (reference range, no cells) per high-power field. There were more than 100 hyaline casts (reference range, 0 to 5) and 10 to 20 granular casts (reference range, none) and mucin per lowpower field. The urinalysis was otherwise nor1156

n engl j med 364;12

of

m e dic i n e

mal. The oxygen saturation increased to 100% while the patient was breathing ambient air. Eight hours after the patient’s arrival, an ECG showed sinus rhythm, at 100 beats per minute, and no ST-segment depressions (Fig. 2 in the Supplementary Appendix. Computed tomography (CT) of the abdomen and pelvis with intravenous and oral contrast material showed hypodense lesions in the liver and spleen that were consistent with hemangiomas; the study was otherwise normal. Ciprofloxacin was begun. He was admitted to the cardiac telemetry unit. The temperature rose to 39.4°C, systolic blood pressure decreased to 80 to 90 mm Hg, oxygen saturation decreased to 89% while he was breathing ambient air, and vomiting occurred. Oxygen supplementation (4 liters) was restarted, and he was transferred to the medical intensive care unit. Two hours later (18 hours after the patient’s arrival at the hospital), the systolic blood pressure decreased to 75 mm Hg. Diphenhydramine, ranitidine, phenylephrine, hydrocortisone, epinephrine, acetaminophen, vancomycin, ceftriaxone, and oseltamivir were administered. Transthoracic echocardiography showed an estimated ejection fraction of 76%, a mildly dilated right ventricle, right ventricular hypertrophy with normal right ventricular function, right atrial dilatation, trace mitral regurgitation, trace pulmonary insufficiency, and findings consistent with mild pulmonary-valve stenosis. Examination of a blood smear showed no malarial forms, and testing for antibodies to the human immunodeficiency virus and Borrelia burgdorferi, a nasal swab for influenza virus types A and B, and stool specimens for white cells, ova and parasites, rotavirus, and Clostridium difficile toxin were negative. The erythrocyte sedimentation rate and tests of thyroid function were normal; other results are shown in Table 1. Hypotension resolved within 2 hours, and phenylephrine was discontinued. The patient was transferred to the medical floor on the third day. During the next 3 days, tests for antinuclear antibodies, hepatitis viruses (A, B, and C), and Helicobacter pylori were negative, cultures of specimens of blood and urine remained sterile, and a stool culture grew normal enteric flora. Antibiotics were discontinued. Abdominal ultrasonography showed two lesions consistent with hemangiomas in the liver and nonspecific thickening of the gallbladder wall. On the sixth day, a test result was received.

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

case records of the massachusetts gener al hospital

Table 1. Laboratory Data.* Reference Range, Adults†

On Admission

12 Hr after Admission

18 Hr after Admission

38 Hr after Admission

Hematocrit (%)

41.0–53.0 (men)

40.8

34.4

32.6

32.1

Hemoglobin (g/dl)

13.5–17.5 (men)

13.3

11.4

10.8

10.6

4500–11,000

13,800

18,200

22,300

11,200

Variable

White-cell count (per mm3) Differential count (%) Neutrophils

40–70

59

87

92

85

Lymphocytes

22–44

30

1

4

10

Monocytes

4–11

2

0

4

3

Eosinophils

0–8

1

0

0

2

Band forms

0–10

7

12

0

0

0

1

0

0

0

327,000, large forms

225,000

294,000

270,000

Metamyelocytes Platelet count (per

mm3)

150,000–400,000

Prothrombin time Seconds

10.8–13.4

16.7

International normalized ratio

17.8

1.5

Potassium (mmol/liter) Urea nitrogen (mg/dl) Creatinine (mg/dl)

2.6

4.2

3.9

4.0

8–25

11

22

26

11

0.60–1.50

Estimated glomerular filtration rate (ml/min/1.73 m2) Glucose (mg/dl)

1.6

3.4–4.8

1.35

2.30

2.20

1.14

>60

>60

34

36

>60

70–110

122

170

168

143

Protein (g/dl) Total

6.0–8.3

5.8

5.7

5.7

5.9

Albumin

3.3–5.0

3.4

3.5

3.4

3.2

Globulin

2.6–4.1

2.4

2.2

2.3

2.7

Phosphorus (mg/dl)

2.6–4.5

2.8

1.7

2.9

2.0

Calcium (mg/dl)

8.5–10.5

9.2

7.8

7.7

7.9

Lactic acid (mmol/liter)

0.5–2.2

1.8

3.6

2.7

1.4

NT-pro-BNP (pg/ml)

0–450 (<50 yr)

C-reactive protein (mg/liter)

679

<8.0

57.3

<10.0 (between 8 p.m. and 8 a.m.)

29.6

C3

86–184

47

C4

20–58

13

Cortisol (µg/dl) Complement (mg/dl)

* To convert the values for urea nitrogen to millimoles per liter, multiply by 0.357. To convert the values for creatinine to micromoles per liter, multiply by 88.4. To convert the values for glucose to millimoles per liter, multiply by 0.05551. To convert the values for calcium to millimoles per liter, multiply by 0.250. To convert the values for phosphorus to millimoles per liter, multiply by 0.3229. To convert the values for cortisol to nanomoles per liter, multiply by 27.59. NT-pro-BNP denotes N-terminal fragment of pro−brain (B-type) natriuretic peptide. † Reference values are affected by many variables, including the patient population and the laboratory methods used. The ranges used at Massachusetts General Hospital are for adults who are not pregnant and do not have medical conditions that could affect the results. They may therefore not be appropriate for all patients.

Dr. David M. Dudzinski: The initial ECG (Fig. 1 in the Supplementary Appendix) shows atrial fibrilDr. Mandakolathur R. Murali: Dr. Dudzinski, would lation with a rapid ventricular response and an you describe the cardiac studies? Ashman beat (a wide QRS complex after a short

Differ en t i a l Di agnosis

n engl j med 364;12

nejm.org

march 24, 2011

1157

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

Videos of echocardiographic studies are available at NEJM.org

1158

n e w e ng l a n d j o u r na l

RR interval following a long RR interval) and diffuse ST-segment depression with ST-segment elevation in leads V1 and aVR. Explanations for profound ST-segment changes include epicardial coronary disease, vasospasm, metabolic disturbances, cardiomyopathy, and subendocardial ischemia — especially in a patient with a critically low blood pressure.1 Echocardiographic studies were obtained for evaluation of persistent hypotension (Videos 1 and 2, available at NEJM.org, and Fig. 3 in the Supplementary Appendix). There was mild right ventricular hypertrophy and dilatation with diastolic septal flattening, right atrial dilatation, moderate tricuspid regurgitation, mild pulmonic stenosis with doming of the valve,2 and an estimated right ventricular systolic pressure of 43 mm Hg. These findings led to the consideration of carcinoid heart disease, but in carcinoid heart disease, pulmonic regurgitation is common (80%), and almost all patients have thickened, shortened tricuspid leaflets; neither of these findings was present in this case.3 Echocardiographic examination repeated after the patient was euvolemic was unchanged except for a reduction in tricuspid regurgitation to trace, a decrease in right ventricular systolic pressure to 29 mm Hg, and resolution of right ventricular dilatation. Although a pulmonary embolus was not formally ruled out, there was no echocardiographic explanation for this patient’s persistent hypotension. Dr. Murali: I am aware of the diagnosis in this case. In arriving at a diagnosis, a systematic analysis of the events in a timeline (both historical and clinical evolution) could form a useful matrix. Episodes of flushing, light-headedness, vomiting, diarrhea, and hypotension began 12 years before admission, precipitated by physical and mental stress. Hypotension lasted about 12 hours, but generalized weakness lasted 3 to 4 days. From a twice-yearly occurrence, these episodes escalated to once every 2 months in the year preceding this admission. This hospitalization was preceded by an upper respiratory infection, and the episode occurred 30 minutes after ingestion of a pill containing acetylsalicylic acid. The prominent features were facial flushing, “beefy red” ears, vomiting, and chest tightness, with ECG changes, diaphoresis, and hypotension leading to nearsyncope. The theme that emerges from this time-

n engl j med 364;12

of

m e dic i n e

line is a recurrent and escalating systemic disease with flushing and hypotension as the salient clinical features. Although the historical timeline provides the scaffold on which to build the differential diagnosis, the evolution of the clinical features in the emergency department and telemetry unit leads us to a clinical diagnosis. The cardinal features on examination were scleral injection, diffuse blanching erythema over the upper body, warm and well-perfused skin, hypotension, atrial fibrillation and rapid ventricular response, and STsegment changes suggestive of ischemia — in the absence of hives, angioedema, and wheezing. Administration of acetylsalicylic acid during resuscitation exacerbated the flushing, without wheezing or urticaria. Therapy with intravenous fluids, diphenhydramine, ranitidine, phenylephrine, hydrocortisone, acetaminophen, and antimicrobial agents resulted in resolution of the hypotension. Thus, disorders characterized by episodic flushing, emesis, hypotension, constitutional features of weakness, and depression, with escalation in severity and precipitation by physical and mental stress and acetylsalicylic acid, need to be considered in the differential diagnosis. Flushing

Flushing and hypotension are clues to the diagnosis. Flushing is a sensation of warmth accompanied by transient erythema, usually over the face, neck, ears, chest, and limbs and is due to vasodilatation with increased cutaneous blood flow. Vasodilatation is mediated either by neurogenic (autonomic) regulation of cutaneous vascular smooth muscle or by direct action of vasodilator stimuli such as histamine, substance P, and prostaglandins (e.g., PGD2) on endothelial cells.4,5 The resultant vasodilatation and increased vascular permeability contribute to distributive shock, which occurred in this patient.6 The effects of histamine acting through H1 and H2 receptors include vasodilatation and vascular permeability, thus explaining the patient’s flushing and hypotension. Histamine acting through the H3 receptor affects neurogenic vasodilatation through local neuron–mast-cell feedback loops, and its effect on the central nervous system includes alterations in emotions and memory as well as cognitive abnormalities, possibly explaining the depression and weakness that followed the acute episodes.7

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

case records of the massachusetts gener al hospital

A further clue to the diagnosis comes from the characteristics of the flushing — that is, whether it is associated with sweating (wet flushing) or not (dry flushing). Flushing due to neurogenic stimuli is accompanied by sweating, because of autonomic innervation of the eccrine sweat glands. This process occurs with fever, exercise, heat exposure, and menopause, as well as with neurologic disorders with autonomic dysfunction such as post-encephalitic syndromes and Parkinson’s disease. These entities are easily excluded in this case. Vasodilator substances, in contrast, cause dry flushing, which is what this patient had.4,5 Vasodilator stimuli can be exogenous or endogenous, and we need to decide which applies in this case (Table 2). Anaphylaxis

Could this patient’s illness be a manifestation of recurrent anaphylaxis due to a food allergy, a medication, or even an idiopathic cause? Anaphylaxis is a systemic mast-cell–activation or basophilactivation syndrome in which mediators are released from normal mast cells, usually because of exposure to an exogenous substance such as food or medication. In 2003, the World Allergy Organization proposed that anaphylaxis be defined as a “severe, life-threatening generalized or systemic hypersensitivity reaction” that may be immunologic (whether it is IgE-mediated or not) or nonimmunologic (Table 2). Kiwifruit shares antigens with latex, and the patient’s reaction to kiwifruit is suggestive of an oral allergy syndrome; however, his flushing and hypotensive episodes were not precipitated by the ingestion of kiwifruit. Cutaneous hives, angioedema, or both are present in more than 90% of patients with anaphylaxis, and respiratory tract involvement (hoarseness, stridor, rhinitis, or wheezing) is present in 40 to 60%.7 In the absence of these manifestations and with no documented association of previous episodes with specific foods, medications, or latex, anaphylaxis due to any of these causes is unlikely in this case. Acetylsalicylic acid can potentiate mast-cell activation resulting from other causes, exacerbating flushing in this patient. This leaves us with idiopathic anaphylaxis — a syndrome of recurrent anaphylaxis that is not associated with a known trigger.8 Before we accept this as a cause, we need to examine possible endogenous causes of this patient’s symptoms.

n engl j med 364;12

Sepsis

The rapid clinical deterioration (with fever, flushing, tachycardia, and hypotension with a poor response to fluid therapy) and the neutrophilic leukocytosis with “bandemia” (an elevated level of band forms of white cells) and an increased prothrombin time suggest a systemic response to inflammatory cytokines, as is seen in sepsis.9 Rightupper-quadrant tenderness, oliguria, lesions in the liver, and the thickened gallbladder wall suggested an infectious process. This suspicion led to immediate therapy with broad-spectrum antibiotics. The patient was admitted during the H1N1 influenza pandemic, and his associated upper Table 2. Causes of Vasodilator-Mediated Flushing Accompanied by Hypotension.* Causes

Types and Associated Features

Exogenous Medications

Nicotinic acid, calcium-channel blockers, phosphodiesterase-5 inhibitors, vancomycin, angiotensin-converting−enzyme inhibitors

Foods Pharmacologic

Capsaicin, ethanol, sulfites, and monosodium glutamate

Toxic

Scombroidosis from histamine and cis-urocanic acid formed in bacteriacontaminated spoiled fish such as tuna, mackerel, and mahimahi

Anaphylaxis Immunologic

IgE- and FcεR1-mediated stimuli such as allergy to penicillin, insect venom, latex, heterologous serum, and chimeric monoclonal antibodies; non−IgEmediated reactions to stimuli such as blood products, acetylsalicylic acid, radiocontrast mediums, and some drugs

Nonimmunologic

Physical exercise, cold stimuli, opiates, curare

Primary or idiopathic

Undetected cause

Endogenous Sepsis Carcinoid syndrome Mastocytosis Medullary carcinoma of thyroid VIP-secreting tumors Pheochromocytoma Idiopathic systemic capillary leak syndrome * VIP denotes vasoactive intestinal peptide.

nejm.org

march 24, 2011

1159

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

respiratory infection led his care team to administer oseltamivir. Inflammatory cytokines lead to hepatic production of acute-phase proteins such as C-reactive protein, the level of which was elevated in this patient.9 However, the recurrent nature and temporal profile of these episodes implicate a noninfectious systemic disease. Negative serologic tests and cultures over the next 3 days ruled out infectious causes of this illness. Flushing Syndromes Produced by Endocrine Tumors

Neuroendocrine tumors, including pheochromocytoma, vasoactive intestinal peptide–producing tumors, medullary thyroid carcinoma, and carcinoid tumors, may all secrete substances that cause flushing, hypotension or hypertension, diarrhea, and respiratory symptoms in various combinations. The characteristics of the flushing episodes are atypical for pheochromocytoma, the absence of a thyroid nodule makes medullary thyroid carcinoma unlikely, and the remitting symptoms and protracted course make VIPoma unlikely. Analysis to detect mediators of these syndromes was nonetheless warranted and was negative. carcinoid syndrome

Could this patient have the carcinoid syndrome, which is characterized by cutaneous flushing, diarrhea, wheezing, and cardiac valvular lesions? Episodes are often precipitated by the ingestion of alcohol and chocolate and do not occur after exercise or the use of acetylsalicylic acid, as in this patient. Facial telangiectasia and cyanosis and pellagra-like skin changes may be seen in chronic cases.10 None of these features were noted in this patient. Echocardiographic findings raised the question of carcinoid heart disease, but the findings were also consistent with cardiac changes in a triathlete or volume overload during resuscitation. The abdominal CT was negative for ileal or appendicular masses, and the liver lesions were consistent with hemangiomas, not metastases. These features make the carcinoid syndrome unlikely, but it should be ruled out with a 24-hour urine test for 5-hydroxyindoleacetic acid. Idiopathic Systemic Capillary Leak Syndrome

Idiopathic systemic capillary leak syndrome is characterized by hypotension, hypoalbuminemia, and hemoconcentration. This condition is often preceded by an upper respiratory infection and 1160

n engl j med 364;12

of

m e dic i n e

features of distributive shock initially with warm, flushed skin. It is a diagnosis of exclusion, and the biochemical profile is absent in this patient.11 Mastocytosis

This patient’s presentation is consistent with a mast-cell activation syndrome, but it is not typical of anaphylaxis. The clinical features of flushing and hypotension with involvement of the cardiovascular, gastrointestinal, and nervous systems in the absence of urticaria, angioedema, and upperairway involvement suggest systemic mastocytosis, a mast-cell neoplasm (Table 3). In the absence of hematologic abnormalities, hepatosplenomegaly, and tissue dysfunction, the clinical diagnosis is indolent systemic mastocytosis. Determination of the serum tryptase level is essential to establish the diagnosis of systemic mastocytosis and differentiate it from anaphylaxis. Some other laboratory features could be explained by the diagnosis of mastocytosis. Active monomers of β-tryptase can generate anaphylatoxins C3a, C4a, and C5a in vitro.13,14 This patient had decreased levels of C3 and C4, which could be due to an elevated tryptase level. The elevated C-reactive protein level might be a reflection of the systemic effect of interleukin-1 secreted from the activated mast cells. Bone marrow examination will be important in confirming the diagnosis. Dr. Eric S. Rosenberg (Pathology): May we have the medical students’ diagnosis? A Harvard Medical Student: We considered allergic and immunologic, hematologic, vascular, neurogenic, and endocrine disorders. Our differential diagnosis included a carcinoid syndrome, pheochromocytoma, the hypereosinophilic syndrome, allergic reaction, thyrotoxicosis, and mastocytosis. We thought mastocytosis the most likely diagnosis. Dr. Rosenberg: Dr. Vassallo, would you tell us what your clinical impression was and what diagnostic tests were performed? Dr. Milo Vassallo (Allergy and Immunology): I interpreted the clinical history to be most consistent with recurrent episodes of anaphylaxis. In the differential diagnosis, I considered idiopathic anaphylaxis, but urticaria and respiratory symptoms were never prominent in this patient. The flushing improved rapidly in the medical intensive care unit with treatment with antihistamines, implicating histamine and mast cells rather than

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

case records of the massachusetts gener al hospital

Table 3. World Health Organization Categories of Mastocytosis.* Category

Definition

Cutaneous mastocytosis

Mast-cell infiltrates with four clinical variants; criteria for systemic mastocytosis not met (see below)

Urticaria pigmentosa Diffuse cutaneous mastocytosis Cutaneous mastocytoma Telangiectasia macularis eruptiva perstans Systemic mastocytosis

Major criterion: multifocal, dense aggregates of mast cells (≥15) in sections of bone marrow, another extracutaneous organ or organs, or both; minor criteria (one required; three required if major criterion absent): >25% of mast cells morphologically atypical, expression of CD2 or CD25 by mast cells, presence of KIT codon 816 mutation, serum total tryptase >20 mg/ml

Indolent systemic mastocytosis

No mast-cell−related organ dysfunction; no associated hematologic non−mast-cell lineage disorder (see below); skin lesions usually present

Systemic mastocytosis with an associated hematologic non−mast-cell lineage disorder

Evidence of one of the following: myelodysplastic myeloproliferative neoplasm, myelodysplastic syndrome, acute myeloid leukemia, or lymphoid neoplasm (lymphoma or plasma-cell myeloma)

Aggressive systemic mastocytosis

Organ dysfunction due to mast-cell infiltration (in bone marrow, liver, spleen, gastrointestinal tract, or bones); skin lesions usually absent

Mast-cell leukemia

>10% immature mast cells in blood or >20% in bone marrow

Extracutaneous mastocytoma

Solitary mast-cell tumor without cytologic atypia

Mast-cell sarcoma

Solitary mast-cell tumor with high-grade cytologic atypia

* Data are from Horny et al.12

other causes of flushing such as carcinoid and gastrointestinal tract tumors. The diagnostic test was analysis of the serum tryptase level. This result reflects both the total mast-cell burden and mast-cell activation. There are two assayable forms of tryptase: total tryptase, which is preformed in the mast-cell granules, and β-tryptase, which is formed by proteolysis on activation and degranulation of mast cells.15 The total tryptase level is useful in the diagnosis of anaphylaxis only if measured within hours after the onset of symptoms. The patient’s total tryptase level in a specimen of blood drawn approximately 3 hours after the onset of symptoms was 2983 ng per milliliter (reference range, <11.5); the β-tryptase level was 1350 ng per milliliter (reference range, <1). A total serum tryptase level measured 6 days later, when the patient was asymptomatic, was 54.8 ng per milliliter, and the β-tryptase level was less than 1 ng per milliliter; this result indicates an increased mastcell burden.

n engl j med 364;12

These findings provided support for a diagnosis of indolent systemic mastocytosis. The test to confirm the diagnosis was bone marrow biopsy and aspiration.

Cl inic a l Di agnosis Indolent systemic mastocytosis complicated by anaphylaxis due to aspirin.

Dr . M a nda kol athur R . Mur a l i’s Di agnosis Acute manifestation of indolent systemic mastocytosis triggered by upper respiratory infection and aspirin.

Pathol o gic a l Discussion Dr. Robert P. Hasserjian: The bone marrow–biopsy specimen contained multiple large aggregates of pale cells with oval nuclei, accounting for about

nejm.org

march 24, 2011

1161

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

A

B

C

D

of

m e dic i n e

Figure 1. Findings on Examination of Bone Marrow Aspirate. The bone marrow–biopsy specimen contains multiple aggregates of pale cells that are often located adjacent to bone trabeculae surrounding a central core of mature lymphocytes (Panel A, hematoxylin and eosin). The pale cells have oval nuclei and abundant cytoplasm and are admixed with numerous eosinophils (Panel B, hematoxylin and eosin); on the aspirate smear, the rare mast cells are abnormally spindled with uneven cytoplasmic granulation (inset, Wright–Giemsa stain). Immunohistochemical analysis shows that the mast cells express the lineage-specific marker tryptase (Panel C) and aberrantly express CD25 (Panel D).

20% of the overall marrow cellularity (Fig. 1A and 1B). The remaining marrow was markedly hypercellular and showed trilineage maturing hematopoiesis with increased eosinophils. Immunohistochemical studies revealed that the abnormal cells were mast cells expressing mast-cell tryptase and CD117 (the product of the KIT gene) (Fig. 1C). Unlike normal mast cells, these cells expressed CD25 (Fig. 1D) and CD2. The bone marrow aspirate showed normal hematopoiesis with no dysplasia or increased blasts. Rare abnormal spindleshaped mast cells were present (Fig. 1B, inset). Flow cytometry showed no abnormal cells. Cytogenetic analysis revealed a normal karyotype (46,XY). Fluorescence in situ hybridization (FISH) showed no rearrangement of the Fip1-like 1 gene (FIP1L1) or the gene that encodes platelet-derived growth 1162

n engl j med 364;12

factor receptor α (PDGFRA). Subsequently, polymerase-chain-reaction assay of a peripheral-blood sample revealed a point mutation at codon 816 (Asp-816→Val) of the KIT gene. The diagnosis of systemic mastocytosis is based on both clinical and pathological features (Table 3).15 The presence of multiple large mastcell aggregates in the bone marrow of this patient constitutes a major criterion for the diagnosis of systemic mastocytosis. In addition, there was prominent mast-cell atypia (spindle-shaped cells with hypogranulation), aberrant expression of CD2 and CD25, a KIT codon 816 mutation, and elevated serum tryptase levels, fulfilling all diagnostic criteria for systemic mastocytosis. Bone marrow biopsy is an important diagnostic test for systemic mastocytosis, since the large

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

case records of the massachusetts gener al hospital

mast-cell aggregates are highly characteristic of the disease. As in this case, mast cells may be rare in the aspirate smears because of fibrosis associated with the mast-cell aggregates.15 When the diagnosis of systemic mastocytosis was suggested on the basis of the bone marrow examination in this patient, additional considerations arose in the differential diagnosis. Systemic mastocytosis is associated with a clonal non–mast-cell hematologic neoplasm (e.g., a myelodysplastic syndrome or chronic myelomonocytic leukemia) or acute myeloid leukemia in 30 to 40% of cases (Table 3).16,17 In these cases, the prognosis is typically determined by that of the associated non–mast-cell neoplasm. Many of these myeloid neoplasms have been shown to share KIT mutation and cytogenetic abnormalities with the mast-cell proliferation, indicating an origin from a common precursor cell.18,19 In this patient, there was no morphologic evidence of an associated myelodysplastic syndrome, myeloproliferative neoplasm, or acute leukemia, and the karyotype was normal. Another important consideration in the diagnosis of systemic mastocytosis is the exclusion of a myeloid neoplasm with FIP1L1–PDGFRA rearrangement. This neoplasm is characterized by diffusely increased bone marrow mast cells, increased serum tryptase levels, and eosinophilia, and it may mimic systemic mastocytosis. Many patients with this neoplasm have a response to specific targeted therapies, such as imatinib mesylate, that are ineffective in systemic mastocytosis.16 The presence of aggregated (rather than diffusely scattered) mast cells, a KIT mutation, and a negative FISH study ruled out a myeloid neoplasm with FIP1L1–PDGFRA rearrangement. Dr. Dudzinski: The prominent yet transient STsegment changes observed on the admission ECG may be a manifestation of epicardial coronary vasospasm due to supraphysiologic concentrations of histamine from mast-cell degranulation. This mechanism, variously termed hypersensitivity coronary syndrome, allergic angina, or the Kounis syndrome, has been described in response to a host of allergic stimuli and can be manifested as ECG changes and chest pain.20-22 Dr. Vassallo: Upper gastrointestinal endoscopy on the sixth day showed duodenal ulcers. The patient remained in stable condition and was discharged on the eighth day. He was prescribed ranitidine and cetirizine for prophylactic mann engl j med 364;12

agement of flushing and pruritus and epinephrine autoinjectors for use during acute reactions. Two days later, CT of the chest was performed. Dr. James Y. Song: CT of the chest performed after the administration of intravenous contrast material for evaluation of pulmonary embolism (Fig. 2A) shows loss of corticomedullary differentiation and coarse trabeculae in the thoracic

A

B

C

Figure 2. Imaging Studies of the Chest. An axial CT scan of the chest (Panel A) obtained after the administration of contrast material shows loss of corticomedullary differentiation and heterogeneous osteosclerosis within the 10th thoracic vertebral body (arrow). The bilateral fifth ribs show patchy bone sclerosis (Panel B, arrows) and there is a focal lytic lesion (≤7 mm in diameter) within the left posterior eighth rib (Panel C, arrow). An additional lytic lesion was described in the right third rib (not shown). The diffuse sclerotic and lytic appearance is characteristic of mastocytosis.

nejm.org

march 24, 2011

1163

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

spine. The ribs show scattered regions of patchy sclerosis (Fig. 2B) alternating with normal marrow attenuation, and there is a focal lytic lesion within the left eighth rib (Fig. 2C). This mixed lytic and sclerotic infiltrative appearance in the axial skeleton is consistent with mastocytosis. In retrospect, similar subtle changes were probably present in the spine on the abdominal and pelvic CT scan obtained on admission. Dr. Rosenberg: Dr. Castells, can you discuss the care of this patient with systemic mastocytosis?

Discussion of M a nagemen t

of

m e dic i n e

The management of mastocytosis depends on the clinicopathological subtype. Reduction of the mast-cell burden is indicated in cases of aggressive mastocytosis and mast-cell leukemia, and treatment of the non–mast-cell hematologic disorder is indicated in cases of mastocytosis with such a disorder. This patient has indolent systemic mastocytosis, and treatment in these patients is aimed at avoiding triggers, limiting release of mediators, and blocking their actions. For this patient, I recommended H1- and H2histamine–receptor blockade (cetirizine and ranitidine); oral disodium cromoglycate, which blocks the release of mediators from mast cells25; leukotriene-receptor blockade (montelukast); and a proton-pump inhibitor (omeprazole).26 He carries two epinephrine self-injectable devices to treat hypotensive events. Imatinib would not be useful, since its site of action is abrogated by the D816V mutation in this patient. Unfortunately, the patient has been noncompliant with his medications, despite extensive counseling, and he has had two further episodes of flushing and hypotension during the 15 months since the diagnosis was established. Both episodes were treated in the emergency department with intravenous fluids and antihistamines. Dr. Steven E. Goldfinger (Gastroenterology): I note that the patient was once given epinephrine for his flushing. The fact that profound hypotension did not develop tends to rule out carcinoid tumors, because epinephrine can precipitate a carcinoid crisis, which could result in death.

Dr. Mariana C. Castells: With its variable clinical phenotypes,23 systemic mastocytosis is a great masquerader, and as in this patient, the onset of symptoms can precede the diagnosis by many years — a median of 9.5 years in one study.24 Exercise, alcohol, trauma, infections, contrast dyes, medications including nonsteroidal antiinflammatory drugs (NSAIDs) and antibiotics, anesthesia, and surgery can induce release of mastcell mediators, triggering symptomatic episodes. When I questioned this patient, he said that his triggers were alcohol, stress, emotions, infections, foods with high fat content, and NSAIDs. He also had chronic symptoms, including daily flushing, chronic fatigue, depression, anxiety, bone pain, fractures, and chest pain that had led to multiple cardiac evaluations. Urticaria pigmentosa is associated with indolent systemic mastocytosis in more than 80% of all cases; scratching of these reddish brown macules, which are scattered over the body except A nat omic a l Di agnosis for the palms, soles, and scalp, triggers urtication and erythema, known as Darier’s sign. If Indolent systemic mastocytosis. skin findings are overlooked or subtle, the otherThis case was presented at the Medicine Grand Rounds. wise nonspecific symptoms may lead to a fruitDr. Castells reports receiving grant support from Ovations for less search for other causes. When I examined the Cure and consulting and other fees from Schering-Plough; this patient after discharge, I noted several brown Dr. Hasserjian, consulting fees from Genzyme; and Dr. Dudzinlesions on the anterior chest, with a positive ski, honoraria from Lippincott Williams & Wilkins. No other potential conflict of interest relevant to this article was reported. Darier’s sign, findings compatible with urticaria Disclosure forms provided by the authors are available with pigmentosa. the full text of this article at NEJM.org. References 1. Habibzadeh MR, Ewy GA. ST-seg-

ment elevation in lead aVR greater than that in lead V(1) secondary to diffuse myocardial ischemia from prolonged hypotension. Clin Cardiol 2007;30:251-4. 2. Weyman AE, Hurwitz RA, Girod DA, Dillon JC, Feigenbaum H, Green D. Cross-

1164

sectional echocardiographic visualization of the stenotic pulmonary valve. Circulation 1977;56:769-74. 3. Pellikka PA, Tajik AJ, Khandheria BK, et al. Carcinoid heart disease: clinical and echocardiographic spectrum in 74 patients. Circulation 1993;87:1188-96.

n engl j med 364;12

nejm.org

4. Izikson L, English JC III, Zirwas MJ.

The flushing patient: differential diagnosis, workup, and treatment. J Am Acad Dermatol 2006;55:193-208. 5. Wilkin JK. Flushing reactions: consequences and mechanisms. Ann Intern Med 1981;95:468-76.

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

case records of the massachusetts gener al hospital 6. Kaliner M, Sigler R, Summers R, Shel-

hamer JH. Effects of infused histamine: analysis of the effects of H-1 and H-2 histamine receptor antagonists on cardiovascular and pulmonary responses. J Allergy Clin Immunol 1981;68:365-71. 7. Lieberman PL. Anaphylaxis. In: Adkinson NF Jr, Bochner BS, Busse WW, Holgate ST, Lemanske RF Jr, Simons FER, eds. Middleton’s allergy: principles & practice. 7th ed. Vol. 2. Philadelphia: Mosby, 2009:1027-49. 8. Ditto AM, Harris KE, Krasnick J, Miller MA, Patterson R. Idiopathic anaphylaxis: a series of 335 cases. Ann Allergy Asthma Immunol 1996;77:285-91. 9. Bone RC. Immunologic dissonance: a continuing evolution in our understanding of the systemic inflammatory response syndrome (SIRS) and the multiple organ dysfunction syndrome (MODS). Ann Intern Med 1996;125:680-7. 10. Modlin IM, Kidd M, Latich I, Zikusoka MN, Shapiro MD. Current status of gastrointestinal carcinoids. Gastroenterology 2005;128:1717-51. 11. Dhir V, Arya V, Malav IC, Suryanarayanan BS, Gupta R, Dey AB. Idiopathic systemic capillary leak syndrome (SCLS): case report and systematic review of cases reported in the last 16 years. Intern Med 2007;46:899-904. 12. Horny H-P, Metcalfe DD, Bennett JM, et al. Mastocytosis. In: Swerdlow S, Campo E, Harris NL, et al., eds. WHO classifi-

cation of tumours of haematopoietic and lymphoid tissues. Lyon, France: IARC Press, 2008:53-63. 13. Fukuoka Y, Schwartz LB. Active monomers of human beta-tryptase have expanded substrate specificities. Int Immunopharmacol 2007;7:1900-8. 14. Schwartz LB, Kawahara MS, Hugli TE, Vik D, Fearon DT, Austen KF. Generation of C3a anaphylatoxin from human C3 by human mast cell tryptase. J Immunol 1983;130:1891-5. 15. Schwartz LB. Diagnostic value of tryptase in anaphylaxis and mastocytosis. Immunol Allergy Clin North Am 2006;26: 451-63. 16. Pardanani A, Lim KH, Lasho TL, et al. Prognostically relevant breakdown of 123 patients with systemic mastocytosis associated with other myeloid malignancies. Blood 2009;114:3769-72. 17. Travis WD, Li CY, Yam LT, Bergstralh EJ, Swee RG. Significance of systemic mast cell disease with associated hematologic disorders. Cancer 1988;62:965-72. 18. Miranda RM, Tang G, Jones DM, Huh YO, Medeiros LJ, Wang SA. Systemic mastocytosis with associated clonal hematological non-mast cell lineage disorder (SM-AHNMD) shows distinct hematologic features and reveals a common clonal origin of myeloid and mast cells. Mod Pathol 2010;23:Suppl:311A. abstract. 19. Sotlar K, Colak S, Bache A, et al. Variable presence of KITD816V in clonal hae-

matological non-mast cell lineage diseases associated with systemic mastocytosis (SM-AHNMD). J Pathol 2010;220:586-95. 20. Kounis NG. Kounis syndrome (allergic angina and allergic myocardial infarction): a natural paradigm? Int J Cardiol 2006;110:7-14. 21. Kounis NG, Kounis GN, Kouni SN, Soufras GD, Niarchos C, Mazarakis A. Allergic reactions following implantation of drug-eluting stents: a manifestation of Kounis syndrome? J Am Coll Cardiol 2006;48:592-3. 22. Rohr SM, Rich MW, Silver KH. Shortness of breath, syncope, and cardiac arrest caused by systemic mastocytosis. Ann Emerg Med 2005;45:592-4. 23. Castells M, Austen KF. Mastocytosis: mediator-related signs and symptoms. Int Arch Allergy Immunol 2002;127:147-52. 24. Horan RF, Austen KF. Systemic mastocytosis: retrospective review of a decade’s clinical experience at the Brigham and Women’s Hospital. J Invest Dermatol 1991;96:Suppl:5S-14S, 60S-65S. 25. Soter NA, Austen KF, Wasserman SI. Oral disodium cromoglycate in the treatment of systemic mastocytosis. N Engl J Med 1979;301:465-9. 26. Escribano L, Akin C, Castells M, Schwartz LB. Current options in the treatment of mast cell mediator-related symptoms in mastocytosis. Inflamm Allergy Drug Targets 2006;5:61-77. Copyright © 2011 Massachusetts Medical Society.

Lantern Slides Updated: Complete PowerPoint Slide Sets from the Clinicopathological Conferences Any reader of the Journal who uses the Case Records of the Massachusetts General Hospital as a teaching exercise or reference material is now eligible to receive a complete set of PowerPoint slides, including digital images, with identifying legends, shown at the live Clinicopathological Conference (CPC) that is the basis of the Case Record. This slide set contains all of the images from the CPC, not only those published in the Journal. Radiographic, neurologic, and cardiac studies, gross specimens, and photomicrographs, as well as unpublished text slides, tables, and diagrams, are included. Every year 40 sets are produced, averaging 50-60 slides per set. Each set is supplied on a compact disc and is mailed to coincide with the publication of the Case Record. The cost of an annual subscription is $600, or individual sets may be purchased for $50 each. Application forms for the current subscription year, which began in January, may be obtained from the Lantern Slides Service, Department of Pathology, Massachusetts General Hospital, Boston, MA 02114 (telephone 617-726-2974) or e-mail [email protected].

n engl j med 364;12

nejm.org

march 24, 2011

1165

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

edi t or i a l

Choice of Bronchodilator Therapy for Patients with COPD Jadwiga A. Wedzicha, M.D. Chronic obstructive pulmonary disease (COPD) is a progressive airway inflammatory condition that is associated with accelerated decline of lung function and is characterized by worsening dyspnea with episodes of increased number and severity of symptoms, termed exacerbations.1 The main objectives of managing COPD are a reduction in the severity of symptoms and the prevention of exacerbations.2 Bronchodilator therapy is central to the management of COPD; this treatment produces modest increases in lung function as measured by spirometry and a reduction in the dynamic hyperinflation that leads to the dyspnea of COPD.3 Patients with COPD often present to their physician with dyspnea, and bronchodilators are often the first therapy prescribed. Initially, short-acting bronchodilators, such as albuterol, may be used, but in patients with persistent symptoms, long-acting bronchodilators provide more uniform relief. Currently there are two classes of long-acting bronchodilators available — long-acting β2-agonists and long-acting anticholinergic agents, and both classes of agents, as compared with placebo, have been shown to provide relief from symptoms in patients with COPD. Although exacerbations are more common with increasing severity of the disease, a substantial number of patients, even those with moderate COPD (defined as COPD in which the patient’s forced expiratory volume in 1 second [FEV1] is 50% or more of the predicted value) are susceptible to frequent exacerbations.4 Patients who have frequent exacerbations are also likely to have more symptoms, worse health status,5 faster disease progression,6 and an increased risk of death7; therefore, it is important to make the diagnosis of COPD in these patients accurately and to treat them effectively. Subgroup analyses of recent large COPD trials have shown that treatment with longn engl j med 364;12

acting β2-agonists or long-acting anticholinergic agents, in addition to decreasing the frequency of exacerbations in patients with severe disease, also reduce exacerbations in patients with moderate COPD.8,9 Since long-acting bronchodilators benefit these patients, it is important to ask which long-acting bronchodilator should be the initial choice for patients with COPD, especially those with moderate disease. The National Institute for Health and Clinical Excellence of England and Wales, in its 2010 update of COPD treatment guidelines, reviewed all studies that compared long-acting β2-agonists and long-acting anticholinergic agents and came to the conclusion that there was no evidence to favor one treatment over another.10 In this issue of the Journal, Vogelmeier and colleagues report the results of a study that makes some progress in addressing this important question.11 The Prevention of Exacerbations with Tiotropium in COPD trial (POET-COPD; ClinicalTrials.gov number, NCT00563381), a large international study, compared the effect of a longacting anticholinergic agent (tiotropium, at a dose of 18 µg once daily) with a long-acting β2-agonist (salmeterol, at a dose of 50 µg twice daily) over the course of 1 year in 7376 patients with COPD who had an FEV1 of 70% of the predicted value or less. The focus of the study was the reduction of exacerbations, and the enrollment criteria included a documented history of at least one exacerbation in the previous year requiring therapy or hospitalization. The primary end point was the time to the first exacerbation, and secondary and safety end points included other exacerbation outcomes and serious adverse events, including death. The results showed that, as compared with salmeterol, tiotropium prolonged the time to the first exacerbation, with a 17% reduction in the

nejm.org

march 24, 2011

1167

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

editorial

risk of an exacerbation. There was also a greater reduction with tiotropium than with salmeterol in the other exacerbation outcomes, such as the annual number of moderate exacerbations (requiring treatment with systemic glucocorticoids, antibiotics, or both) and the annual number of severe exacerbations (requiring hospitalization). The incidence of serious adverse effects, including cardiovascular complications and deaths, was similar between the two groups. An important point to note about this trial was that it was not a direct comparison of a longacting β2-agonist with a long-acting anticholinergic agent, since concomitant medications were allowed; more than 50% of the patients were receiving, on a regular basis, inhaled glucocorticoids that also reduce exacerbations.8 The fact that the patients had had at least one treated exacerbation in the previous year may explain the high use of concomitant medications among the patients in this trial. However, post hoc analyses showed that the effect of tiotropium with respect to the reduction in exacerbations was similar regardless of the severity of the COPD and regardless of whether the patients were receiving inhaled glucocorticoid therapy on a regular basis. Large multicenter studies involving patients with COPD are often difficult to perform, and previous large trials have attempted to study a wide range of outcomes. In contrast, the focus on COPD exacerbations in the study by Vogelmeier et al. enabled the data on exacerbations to be carefully collected and validated. This trial thus provides a good model for future COPD trials, which should be focused on a specific and relevant disease outcome. A strength of the study was that the reporting of exacerbation outcomes was detailed and was supplemented with daily diary cards to confirm exacerbations and the intervals between exacerbations, thus providing more accurate event rates. The main implications of this trial are for the initial care of symptomatic patients with moderate disease and a history of recent exacerbations. The trial evidence suggests that with respect to exacerbation outcomes, tiotropium, administered once daily, is superior to salmeterol, administered twice daily.11 However, novel, once-daily, long-acting β2-agonists such as indacaterol are now becoming available in Europe, and there is some evidence that outcomes with indacaterol are similar to those with tiotropium.12 There is 1168

n engl j med 364;12

no evidence for the superiority of tiotropium in patients with mild COPD (those in whom the FEV1 is >70% of the predicted value) or symptomatic patients with moderate COPD but without a history of exacerbations. However, in patients with progressive COPD, combinations of inhaled long-acting β2-agonists, long-acting anticholinergic agents, glucocorticoids, and new antiinflammatory agents such as oral phosphodiesterase-4 inhibitors may be indicated. Future trials involving patients with COPD will need to study which therapies and which specific combinations are optimal for which COPD phenotypes and disease severities, so that we can reduce the adverse effects of this disabling disease. Disclosure forms provided by the author are available with the full text of this article at NEJM.org. From the Academic Unit of Respiratory Medicine, University College London Medical School, University College London, London. 1. Wedzicha JA, Seemungal TAR. COPD exacerbations: defin-

ing their cause and prevention. Lancet 2007;370:786-96.

2. Global Initiative for Chronic Obstructive Lung Disease.

Global strategy for the diagnosis, management and prevention of chronic obstructive pulmonary disease: 2009 update. (http:// www.goldcopd.com.) 3. O’Donnell DE, Revill SM, Webb KA. Dynamic hyperinflation and exercise intolerance in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2001;164:770-7. 4. Hurst JR, Vestbo J, Anzueto A, et al. Susceptibility to exacerbation in chronic obstructive pulmonary disease. N Engl J Med 2010;363:1128-38. 5. Seemungal TAR, Donaldson GC, Paul EA, Bestall JC, Jeffries DJ, Wedzicha JA. Effect of exacerbation on quality of life in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998;157:1418-22. 6. Donaldson GC, Seemungal TAR, Bhowmik A, Wedzicha JA. The relationship between exacerbation frequency and lung function decline in chronic obstructive pulmonary disease. Thorax 2002;57:847-52. [Erratum, Thorax 2008;63:753.] 7. Soler-Cataluña JJ, Martínez-García MA, Román Sánchez P, Salcedo E, Navarro M, Ochando R. Severe acute exacerbations and mortality in patients with chronic obstructive pulmonary disease. Thorax 2005;60:925-31. 8. Jenkins CR, Jones PW, Calverley PM, et al. Efficacy of salmeterol/fluticasone propionate by GOLD stage of chronic obstructive pulmonary disease: analysis from the randomised, placebocontrolled TORCH study. Respir Res 2009;10:59. 9. Decramer M, Celli B, Kesten S, Lystig T, Mehra S, Tashkin DP. Effect of tiotropium on outcomes in patients with moderate chronic obstructive pulmonary disease (UPLIFT): a prespecified subgroup analysis of a randomised controlled trial. Lancet 2009;374:1171-8. 10. CG101 chronic obstructive pulmonary disease (update): full guideline. London: National Institute for Health and Clinical Excellence, 2010. (http://guidance.nice.org.uk/CG101/Guidance.) 11. Vogelmeier C, Hederer B, Glaab T, et al. Tiotropium versus salmeterol for the prevention of exacerbations of COPD. N Engl J Med 2011;364:1093-103. 12. Donohue JF, Fogarty C, Lötvall J, et al. Once-daily bronchodilators for chronic obstructive pulmonary disease: indacaterol versus tiotropium. Am J Respir Crit Care Med 2010;182:155-62. Copyright © 2011 Massachusetts Medical Society.

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

s ounding b oa r d

Specialization, Subspecialization, and Subsubspecialization in Internal Medicine Christine K. Cassel, M.D., and David B. Reuben, M.D. At a time when most authorities believe that the country desperately needs more generalists, the American Board of Internal Medicine (ABIM) is adding new subspecialties. Specifically, in the past 2 years the ABIM has launched certification in the fields of hospice and palliative care and advanced heart failure and has begun a process for internal-medicine certification with a focused practice in hospital medicine. The ABIM has also approved the subspecialty of adult congenital heart disease to move forward to the American Board of Medical Specialties (ABMS) for final approval. In addition, the ABIM has received requests from specialty societies to approve several new subspecialties, including medical informatics, clinical pharmacology, vascular medicine, addiction medicine, and obesity medicine. Each of these applications raises issues of a societal nature (i.e., the benefits to the public of having clear standards for emerging areas of medical specialization) versus issues of practicality (e.g., the cost of creating and maintaining certification examinations and the ongoing worry about fragmentation of care). These issues have been coupled with the concerns of different specialties that favored or opposed particular subspecialty designations. Most important, we receive clear but contradictory messages from physicians: on the one hand, “recognize what I do” (i.e., create a subspecialty for my niche practice); on the other hand, “stop fragmenting an already overfragmented system.” Meanwhile, accountability standards for physicians are proliferating, and maintaining board certification is requiring more of physicians.1 Other certifying boards, such as those of surgery and pediatrics, face similar pressures. The perspective of the ABIM may be generalized to other disciplines. To understand the ABIM’s decisionmaking process for adding new specialties, it is

n engl j med 364;12

important to examine the historical and current forces behind the drive for additional examinations and performance assessments.

His tory of Specializ ation Specialization has characterized scientific progress in medicine for nearly two centuries. With the centennial celebration of the hugely influential Flexner Report,2 great attention has been paid to Flexner’s call for educational reform. Underlying his call for more research-based education was the growth of a scientific consensus of allopathic medicine and the physiology-based and biochemistry-based understanding of human illness leading to ever more specialization in the pursuit of greater knowledge and expertise. Articles that appeared in the Journal 3,4 in 1936 and 1950 identified the creation of board certification of specialty status as an “ingenious” way for the profession, independent of government, to control the “dangers of specialism.” The authors were most concerned about unqualified practitioners claiming to be specialists and were particularly alarmed by the practice of specialist advertising, considered to be unethical at the time. As medical scientists specialized and devoted their intellectual energies to understanding more and more about narrower topic areas, general practitioners differentiated into physicians with specific areas of expertise, devoting some or all of their work to that specific area. The first medical specialty to create its own assessment board was ophthalmology in 1917. Prompted by the growth of optometry as a separate discipline, the American Medical Association and the American Ophthalmological Society created an independent board of specialists to create standards that would recognize physicians whose knowledge and skills

nejm.org

march 24, 2011

1169

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

1972 Endocrinology, diabetes, and metabolism

1941 Allergy (moved to ABAI in 1971) Cardiovascular disease

n e w e ng l a n d j o u r na l

m e dic i n e

of

1987 1988 Critical care Geriatric medicine medicine 1992 Clinical cardiac electrophysiology

Hematology Infectious disease

1993 Sports medicine

Nephrology Gastroenterology

1930

1940

1950

1994 Adolescent medicine

Rheumatology 1973 Medical oncology

Tuberculosis (pulmonary disease after 1946–1948)

1960

2010 Advanced heart failure and transplant cardiology

1970

1980

2006 Transplant hepatology 2007 Sleep medicine

1999 Interventional cardiology

1990

2000

2008 Hospice and palliative medicine

2010

2020

Decade Figure 1. Timeline of Subspecialties Approved by the American Board of Internal Medicine. The American Board of Allergy and Immunology (ABAI) was founded in 1971. COLOR FIGURE

demonstrated expertise in identifying and treating disorders of the eye. Four specialties created a federation called the ABMS in 1933, which encompassed 10 specialties by 1935; the ABIM was added in 1936. By the 1970s, there were 20 specialties, including primary surgical boards in orthopedics, urology, neurosurgery, plastic surgery, and colorectal surgery. Internal medicine kept the unifying requirement of training in general internal medicine but underwent just as much subspecialization during that time. The ABIM approved requests for 4 subspecialties in the 1940s, another 6 in the 1970s, and 10 more since then (Fig. 1). Some of these subsubspecialties are built on further specialization of large subspecialties such as cardiology and pulmonary medicine, whereas others, including sleep medicine, sports medicine, and geriatric medicine, span multiple disciplines and allow different pathways to certification from other boards. Throughout these decades, some leaders voiced concern that the growing fragmentation of medical care5 would result in the loss or undervaluing of the personal or generalist physician, who was perceived as being essential to good patient care. With more and more specialization, they worried that the generalists’ practice would become too limited in scope and an unattractive choice for residents. This concern spawned the creation of a new kind of specialty in 1969 — a generalist discipline in family med-

1170

n engl j med 364;12

2/24/11 icine (called family practice at the time). It also Author Cassel led to calls for strengthening 1 and repopulating Fig # Title5 At the same time, general internal medicine. other leaders in medicineME sawForti growing specialJMD ization as strengthening DE internal medicine.6 Artist Knoper Draft 3

AUTHOR PLEASE NOTE: Figure has been redrawn and type has been reset Please check carefully

Current Driving For ce s Issue date

3/24/11

Now we are witnessing a resurgence of interest in new specialty designations and a simultaneous eruption of concern about diminishing strength and numbers in primary care specialties. What are the driving forces in the current environment? Do these forces make a stronger case for the value of new specialties? Do they outweigh the concerns about cost, fragmentation, and physician burden? What factors are determinative in these decisions? As in earlier times, the progress of biomedical science continues to be a major factor in the emergence of new subspecialties. Advanced heart failure and transplant hepatology are two examples growing from the need for extended medical care before and after transplantation, including device management in the case of cardiology. Creation of these subspecialties reflects the recognition that there are some specific populations of patients who would benefit from highly focused knowledge and skills obtained by additional training and certification beyond that of a general cardiologist or gastroenterologist. These tertiary subspecialists also require high

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

sounding board

patient volume to maintain proficiency in their skills and are therefore usually based at referral centers. Certification for these subspecialties is limited to specialists with training in the underlying organ-system specialty, which must be maintained. Conversely, disciplines such as geriatric medicine, palliative medicine, and hospital medicine are based on clinical needs and the organization and delivery of care rather than on scientific and technical expertise in a specific organ system. Many new and emerging subspecialties are cross-disciplinary; sleep medicine crosses six different specialties, and palliative medicine is a subspecialty option for 10 different primary specialties. We anticipate that medical informatics will probably extend just as broadly across specialties. The demand for board certification comes in part from the profession itself, with the majority of survey respondents reporting that “professional image” was the primary reason they sought or renewed certification.7 All the proposals for new subspecialties have come to the board initially from specialty societies, often with support from patient groups. In the current environment of growing demands for higher standards and greater transparency, health plans and hospitals are also seeking ways to ensure that physicians have the knowledge and skills required for patients to have reasonable confidence in their capabilities, and board certification is one of the criteria frequently used for this purpose. Some health care organizations and medical groups that employ physicians require that they be certified to join the medical staff. Although the public values board certification,8 most people do not understand what criteria it represents or fully appreciate that different kinds of organizations offer certificates that represent varying degrees of rigor and clinical relevance. As evidence of interest in this type of information, most Internet-based “report cards” include statements about a physician’s board certification. It is not surprising that many new organizations have emerged that offer their own versions of a certificate. We believe that the criteria and requirements for certification in a medical specialty should be public and transparent, and both the ABMS and the ABIM post such information on their respective Web sites.

n engl j med 364;12

Criteria Used to E s ta blish Ne w Specialtie s of Internal Medicine The policies of the ABIM for establishing new areas of specialization have been “repeatedly and exhaustively re-examined”6 over the years, resulting in two successive documents (in 1993 and 2006) to guide the board in deliberations about new subspecialties. The criteria currently used in considering a request for new subspecialty status in internal medicine are articulated in the 2006 report entitled New and Emerging Disciplines in Internal Medicine — 2 (NEDIM–2).9 These criteria focus on evidence that the new discipline has a definable body of knowledge and a substantial number of clinical training programs, with the reasonable expectation that clinical services in the subspecialty will play a beneficial role in patient care (Table 1). Such designations have usually required at least 1 year of accredited training. Subspecialty applications from clinical pharmacology, vascular medicine, addiction medicine, and obesity medicine have not been approved to date because they failed to meet one or more of these criteria or they were deemed insufficiently mature, as reflected by the number of training programs or practitioners in the field.

Change s in Pr ac tice over the Cour se of a C areer In the introduction to the 1998 updated version of her classic book American Medicine and the Public Interest: A History of Specialization,10 Rosemary Stevens opined, “Arguably, the structure of the medical profession is moving toward a system of specialties defined by the job market rather than by the professional system of specialist qualifications.” If you add “defined by the practice area in which the physician focuses and in which the patient expects expertise,” Stevens’s prediction of a specialtymedicine structure may be correct. Clear evidence of specific competencies is a core component of the consumer demand for transparency, and physician specialists themselves are asking to be certified in narrower areas of expertise than they were originally trained in. In response to these developments, maintenance of certification (MOC) — the process through which time-limited certificates are renewed — might be becoming more fluid, reflecting what NEDIM–2 calls “rec-

nejm.org

march 24, 2011

1171

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

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. Criteria for Subspecialty Certification and for Recognition of Focused Practice.* Criteria for Subspecialty Certification Represents a unique body of knowledge that cannot be fully incorporated into the parent discipline Has clinical applicability (i.e., the clinical practice is distinctly different from the parent discipline) Is based on and contributes to the research base of medicine Offers evidence that the discipline improves patient care Requires supervision and direct observation in formal training settings, generally lasting at least 12 months Involves complex technology or specific site-of-care skills Has positive value for certification in the new discipline that outweighs any negative impact on the practice of general internal medicine or an existing subspecialty Criteria for Recognition of Focused Practice through Maintenance of Certification (MOC) Includes large numbers of internists who focus their practice in the discipline, while others may not practice in the focused area at all Meets an important social need for the discipline and offers evidence that focusing practice in the discipline improves patient care * Criteria are adapted from New and Emerging Disciplines in Internal Medicine — 2 (NEDIM–2).9

ognition of focused practice.” The first example of this is the ABIM’s offering of an identified hospitalist pathway for maintaining certification in internal medicine with a focused practice in hospital medicine. It is a rigorous pathway, with clear requirements for a large enough number of patients, quality and safety metrics, and an examination on hospital-based medicine. Although this pathway is approved by ABMS as a pilot offering, the ABIM and other boards, such as those in ophthalmology and radiology, have received additional requests for this kind of focused MOC option. Greater specificity in certification and recognition of focused practice should be appealing to consumers who want to know the areas in which their physicians are skilled and experienced; taken to its logical conclusion, however, it could unravel the traditional specialty-medicine structure that Stevens described. For example, should an endocrinologist who has focused only on diabetes and, to maintain certification, has focused only on demonstrating current knowledge and performance in management of diabetes still be considered an endocrinologist? The ABMS’s role has been to define these specialty areas and, in the process, to respect the breadth of competence the public could expect from that specialist. Taking the focused-practice trend too far would mean additional, smaller areas of competence, although such a focus may

1172

n engl j med 364;12

be of greater interest to the patient seeking care for a specific condition. If this approach were carried out thoughtfully, the original certificate would identify the areas of formal training with the understanding that the focus of practice could change over the course of a career. Both the original certification and the focused-practice areas would be available to the public on the Web site. In the current consumer-centered environment, it is difficult to argue against making more specific and more evidence-based information available to the public about the specialists who provide their care. Indeed, the Center for Medicare and Medicaid Services Physician Compare database, and perhaps other publicly available databases, argues for the most meaningful information (i.e., beyond administrative and claims data) to be part of the information included.11 Our challenge is to weigh the public interest in light of the traditional meaning of board certification and the benefits of making more specific physician information available against the costs of developing accurate and rigorous assessment tools. With physician practice evolving over the course of a career in medicine, the focused-practice approach may be better for consumers and may also be a way to respond to physicians’ requests to be evaluated on their actual clinical practice in addition to their designated area of training.12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

sounding board

Effec t s on Gener alis t Pr ac tice Despite concerns expressed by generalists in response to new subspecialty designations, it seems unlikely that primary care or generalist disciplines would be strengthened by a moratorium on creating new subspecialties. Rather, the payment system and organization of medical practice are much more important contextual drivers for advancing primary care.13,14 Generalist disciplines themselves are developing new knowledge standards as they evolve, including the new hospitalist designation geriatric medicine and the probable need for stronger emphasis on systems and information science for generalist physicians in a medical home. Medical students get the mistaken message that generalist disciplines are less intellectually exciting, when in fact, it is increasingly difficult to keep up with the breadth of knowledge needed in these fields, including clinical, technical, and managerial skills. Medical group leaders, payers, and hospitals need to be able to identify physicians with these skills as new practice models are formed. Some have called for a specific recognition, perhaps through focused practice in MOC, to recognize the specific systems, health information technology, and team-based competencies required for effective practice in a medical home or accountable care organization.15 Ultimately, primary care will be fostered through payment reform, systems reform, and stronger standards, not through diminished competition from the subspecialties.

Conclusions New specialties can benefit both patients and physicians. However, a proliferation of specialties without adequate justification may simply confuse the public without creating a social good. Use of specified criteria, such as those articulated in the ABIM’s NEDIM reports, can lead to rational decision-making that balances the potential benefit of recognizing more specific expertise with the detriment of fragmentation of the profession. This approach extends beyond traditional specialization, which requires formal training, to the recognition of new areas of expertise that physicians gain while in practice — that is, focused practice. Although the demands for new subspecialties come from physician

n engl j med 364;12

groups themselves, some of their members also complain about the burden of more requirements. To be meaningful, the criteria for establishing new specialties must be rigorous, but to be workable, they need to be aligned with other measurement and reporting requirements, such as those in pay-for-performance programs, state licensing processes, and hospital privilege issuing and credentialing. Certifying boards should continue their work with other accrediting and standard-setting organizations to make these designations as meaningful as possible to patients and physicians alike. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. From the American Board of Internal Medicine, Philadelphia (C.K.C., D.B.R.); and the Division of Geriatrics, David Geffen School of Medicine, University of California, Los Angeles (D.B.R.). 1. Levinson K, King TE Jr, Goldman L, Goroll AH, Kessler B.

American Board of Internal Medicine maintenance of certification program. N Engl J Med 2010;362:948-52. 2. Flexner A. Medical education in the United States and Canada: a report to the Carnegie Foundation for the Advancement of Teaching. Bulletin no. 4. New York: Carnegie Foundation for the Advancement of Teaching, 1910. 3. The certification of specialists. N Engl J Med 1936;215:468-9. 4. Fitz R. The rise of the practice of internal medicine as a specialty. N Engl J Med 1950;242:569-74. 5. Petersdorf RG. The doctors’ dilemma. N Engl J Med 1978; 299:628-34. 6. Reitemeier RJ, Benson JA Jr. Two views of “the doctors’ dilemma”: as the Board sees it. N Engl J Med 1978;299:1308-9. 7. Lipner RS, Bylsma WH, Arnold GK, Fortna GS, Tooker J, Cassel CK. Who is maintaining certification in internal medicine — and why? A national survey 10 years after initial certification. Ann Intern Med 2006;144:29-36. 8. Brennan TA, Horwitz RI, Duffy FD, Cassel CK, Goode LD, Lipner RS. The role of physician specialty board certification status in the quality movement. JAMA 2004;292:1038-43. 9. Final report of the Committee on Recognizing New and Emerging Disciplines in Internal Medicine (NEDIM) — 2. Philadelphia: American Board of Internal Medicine, 2006. (http:// www.abim.org/pdf/nedim-2-report.pdf.) 10. Stevens R. American medicine and the public interest: a history of specialization. Berkeley: University of California Press, 1998:xxvi. 11. Obama B. Memorandum for the heads of executive departments and agencies: transparency and open government. Washington, DC: The White House, 2010. (http://www.whitehouse. gov/the_press_office/Transparency_and_Open_Government.) 12. Drazen JM, Weinstein DF. Considering recertification. N Engl J Med 2010;362:946-7. 13. Bodenheimer T, Lo B, Casalino L. Primary care physicians should be coordinators, not gatekeepers. JAMA 1999;281: 2045-9. 14. Baron RJ, Cassel CK. 21st-century primary care: new physician roles need new payment models. JAMA 2008;299:1595-7. 15. What should outpatient internists do? Weblog: db’s medical rants. May 1, 2010. (http://www.medrants.com/archives/5472.) Copyright © 2011 Massachusetts Medical Society.

nejm.org

march 24, 2011

1173

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

clinical implications of basic research

In Utero Hematopoietic Stem-Cell Transplantation — A Match for Mom Ornella Parolini, Ph.D. Once researchers recognized that adult stem cells can generate multiple cell types and contribute to tissue homeostasis, it became conceivable to exploit this potential to treat genetic or acquired disorders characterized by tissue degeneration or organ dysfunction. The concept of regenerative medicine was thus born, with the general aim of transplanting donor stem cells to replace or repair defective cells of the host. Unfortunately, in the case of an HLA mismatch between donor and recipient, transplantation is hampered by the risks of immunologic recognition and rejection of the graft. However, a recent article by Nijagal and colleagues1 revives the discussion of the potential advantages of transplanting stem cells into the fetus early in gestation. Because in utero stem-cell transplantation can be carried out when the immune system is immature, it provides the theoretical opportunity to induce fetal tolerance of the foreign cells and thereby avoid rejection and the need for immunosuppressive therapy. For these reasons, this potential clinical approach is attractive for any disorder that is amenable to stemcell transplantation and that can be prenatally diagnosed. However, despite successful results of in utero transplantation in animal models, achieved for the most part with hematopoietic stem cells (HSCs), positive outcomes of this procedure in humans have been limited to cases of inherited immunodeficiency diseases.2-4 One of the main hurdles to widespread application and success of in utero stem-cell transplantation is the difficulty in achieving adequate levels of engraftment. Nijagal and colleagues tested their hypothesis that maternal cells trafficking into the fetus impede effective in utero stem-cell transplantation by mounting a sort of “immune protection” of the fetus, in which maternal cells effect the

1174

n engl j med 364;12

rejection of cells allogeneic to both mother and fetus. Using mouse models, these researchers tested their hypothesis in a series of experiments to evaluate the role of the maternal immune response in limiting engraftment. First they found that in utero transplantation of fetal HSCs elicited an increase in trafficking of maternal T cells to the fetal blood. To test the hypothesis that maternal cells play a pivotal role in the fetal engraftment of allogeneic cells, they transferred allogeneic fetal HSCs into fetuses of mothers with experimentally induced B-cell or T-cell deficiency (Fig. 1). Levels of engraftment were significantly higher in fetuses of mothers with T-cell deficiency than in fetuses of wildtype mothers or of mothers with B-cell deficiency. Finally, when in utero stem-cell transplantation was performed with HSCs matched to the mother, similar levels of engraftment were observed in fetal recipients of syngeneic and allogeneic fetal grafts. The evidence put forward is clear and striking, although further research is warranted to confirm these findings, as well as to determine whether maternal T cells are critical for the success of in utero stem-cell transplantation in other animal models. Differences in placentation between animal species might predict differences in maternal-cell trafficking and thus the extent to which the observed phenomenon in mice is relevant to other animals. However, considering that the same types of hemochorial placentation and fetal–maternal chimerism have been described in humans, the conclusion of Nijagal et al. may be relevant to humans. The clinical relevance of this study lies in the potential for improving engraftment with the use of cells that are either harvested from or matched to the mother’s cells (Fig. 1). Although collecting stem cells from pregnant women presents

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

clinical implications of basic research

A

Figure 1. Intrauterine Transplantation and Maternal T Cells. A recent study by Nijagal and colleagues1 highlights a key role of maternal T cells in limiting engraftment after in utero transplantation, in which allogeneic fetal hematopoietic stem cells (HSCs) were transplanted into the fetuses of wild-type mice (Panel A) and of mice lacking B cells (Panel B) or T cells (Panel C). Only T-cell deficiency resulted in significant improvement in engraftment levels, as compared with the levels in wildtype mothers. A notable level of engraftment was also observed after transplantation of fetal HSCs that were genetically matched to the mother (Panel D). Taken together, these results suggest that maternal T cells are responsible for failure of engraftment. It is therefore possible that the clinical success of in utero stem-cell transplantation could be improved by transplanting cells harvested from or matched to the mother’s cells (Panel E).

Allogeneic fetal hematopoietic stem cells Wild-type mother

B

Allogeneic fetal hematopoietic stem cells Mother lacking B cells

C

Allogeneic fetal hematopoietic stem cells Mother lacking T cells

D

Significant improvement in engraftment levels

Matched maternal cells Wild-type mother

Significant improvement in engraftment levels

some obstacles, they are not insurmountable, and such an approach may confer advantages that would extend the clinical applicability of in utero stem-cell transplantation beyond transplantation of HSCs to include other cell types — such as mesenchymal stromal cells — for the treatment of congenital diseases outside the hematopoietic system. Finally, the notion that maternal T cells are responsible for the rejection of allogeneic transplanted HSCs and are therefore critical to the success of engraftment underscores the complex and paradoxical relationship between mother and fetus during pregnancy. Disclosure forms provided by the author are available with the full text of this article at NEJM.org From Centro di Ricerca E. Menni, Fondazione Poliambulanza– Istituto Ospedaliero, Brescia, Italy.

E

1. Nijagal A, Wegorzewska M, Jarvis E, Le T, Tang Q, Macken-

Maternal cells

zie TC. Maternal T COLOR cells limit FIGURE engraftment after in utero hematopoietic cell transplantation in mice. J Clin Invest 2011;121:582Rev4 02/24/11 92. 2. Wengler A, Frusca T, et al. In-utero transAuthor GS, Lanfranchi Dr. Parolini plantation Fig # of parental 1 CD34 haematopoietic progenitor cells in a patient with X-linked severe combined immunodeficiency Title (SCIDXI). Lancet 1996;348:1484-7. ME AW, Roncarolo MG, Puck JM, et al. Treatment of 3. Flake X-linked immunodeficiency by in utero transDE severe combined Phimister plantation bone marrow. N Engl J Med 1996;335: Daniel Muller Artist of paternal 1806-10. AUTHOR PLEASE NOTE: 4. Peranteau Endo M, Flake AW. Evidence for Figure hasWH, been redrawn and Adibe type hasOO, been reset Please check an immune barrier after incarefully utero hematopoietic-cell transplantation. Blood Issue date2007;109:1331-3.

Matched maternal cells

03/24/2011

Copyright © 2011 Massachusetts Medical Society.

n engl j med 364;12

nejm.org

march 24, 2011

1175

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

c or r e sp ondence

Retraction: A Genomic Strategy to Refine Prognosis in Early-Stage Non–Small-Cell Lung Cancer. N Engl J Med 2006;355:570-80. to the editor: We would like to retract our article, “A Genomic Strategy to Refine Prognosis in Early-Stage Non–Small-Cell Lung Cancer,”1 which was published in the Journal on August 10, 2006. Using a sample set from a study by the American College of Surgeons Oncology Group (ACOSOG) and a collection of samples from a study by the Cancer and Leukemia Group B (CALGB), we have tried and failed to reproduce results supporting the validation of the lung metagene model described in the article. We deeply regret the effect of this action on the work of other investigators. Anil Potti, M.D. Chapel Hill, NC

University of Washington Seattle, WA

Holly K. Dressman, Ph.D. Duke University Medical Center Durham, NC

Andrea Bild, Ph.D. University of Utah Salt Lake City, UT

Jason Koontz, M.D. Duke University Medical Center Durham, NC

Robert Kratzke, M.D.

Sayan Mukherjee, Ph.D.

University of Minnesota Minneapolis, MN

Duke University Medical Center Durham, NC

Mark A. Watson, M.D., Ph.D. Washington University School of Medicine St. Louis, MO

this week’s letters 1176 Retraction: A Genomic Strategy to Refine Prognosis in Early-Stage Non–Small-Cell Lung Cancer 1177 Apixaban vs. Enoxaparin after Hip Replacement 1178 Oral Rivaroxaban for Symptomatic Venous Thromboembolism 1178 Drug-Eluting or Bare-Metal Stents in Large Coronary Arteries 1181 Safety and Effectiveness of a 2009 H1N1 Vaccine in Beijing

1176

Rebecca Petersen, M.D.

n engl j med 364;12

Michael Kelley, M.D. Geoffrey S. Ginsburg, M.D., Ph.D. Mike West, Ph.D. David H. Harpole, Jr., M.D. Joseph R. Nevins, Ph.D. Duke University Medical Center Durham, NC [email protected] 1. Potti A, Mukherjee S, Petersen R, et al. A genomic strategy

to refine prognosis in early-stage non–small-cell lung cancer. N Engl J Med 2006;355:570-80. This letter (10.1056/NEJMc1101915) was published on March 2, 2011, at NEJM.org.

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

correspondence

Apixaban vs. Enoxaparin after Hip Replacement To the Editor: Lassen and colleagues (Dec. 23 issue)1 should be congratulated for striving to reduce thromboembolic events after hip replacement, but their choice of primary efficacy outcome, which combines clinically important variables (death and symptomatic venous thromboembolism) with venographically proven deep-vein thrombosis, provokes concern. Neither mortality nor the rate of symptomatic thromboembolism differed significantly between the apixaban and enoxaparin groups; only subclinical deep-vein thrombosis was significantly more common after treatment with enoxaparin as compared with apixaban. This composite outcome measure is therefore misleading. Furthermore, nearly 30% of the participants could not be evaluated for the primary end point. Although this factor was prespecified, it highlights the limitations of including subclinical deep-vein thrombosis in the primary outcome. In addition, since the drugs were not started simultaneously, assessment of their relative safety profile is compromised; commencing thromboprophylaxis before making the surgical incision will induce more bleeding. Finally, data on surgical-site infections and wound hematoma should be reported. Complications as a result of thromboprophylaxis occur more frequently than major thromboembolic events. Clinically relevant outcome measures and provision of comprehensive safety data are required if perioperative thromboprophylaxis is to change. Simon S. Jameson, M.R.C.S. Northumbria Healthcare NHS Trust Ashington, United Kingdom [email protected]

thrombosis, especially proximal deep-vein thrombosis, is a clinically relevant outcome because it is the source of fatal pulmonary embolism. Most deaths from pulmonary embolism occur as sudden death and are not preceded by symptoms in the leg or cardiorespiratory system. Reducing the source of these events is clinically relevant. The apixaban regimen reduced major venous thromboembolism (proximal deep-vein thrombosis or pulmonary embolism) as compared with a standard-care, regulatory-approved regimen of enoxaparin. Our comparison of these regimens is valid. Systematic literature reviews indicate that this preoperative dose of enoxaparin, administered 12 hours before surgery, does not add to bleeding.1,2 The apixaban regimen did not increase bleeding as compared with enoxaparin. This conclusion is further strengthened when only those bleeding events that occur after the first dose of apixaban — which may be attributable to apixaban — are considered. Both surgical-site infections and wound hematomas were uncommon (<2% overall), and occurred with similar frequency in both treatment groups (see the Supplementary Appendix, available with the full text of this letter at NEJM.org). Thus, our study provides clinically relevant data about the relative risk– benefit profile of the apixaban and enoxaparin regimens. Michael Rud Lassen, M.D. Copenhagen University Hospital Glostrup, Denmark [email protected]

Gary E. Raskob, Ph.D. University of Oklahoma Health Sciences Center Oklahoma City, OK

Robert D. Sanders, B.Sc., F.R.C.A.

Alexander Gallus, M.D.

Imperial College London London, United Kingdom

Flinders University Adelaide, SA, Australia

Mike R. Reed, M.D. Northumbria Healthcare NHS Trust Ashington, United Kingdom No potential conflict of interest relevant to this letter was reported. 1. Lassen MR, Gallus A, Raskob GE, Pineo G, Chen D, Ramirez

LM. Apixaban versus enoxaparin for thromboprophylaxis after hip replacement. N Engl J Med 2010;363:2487-98.

The Authors Reply: We respectfully disagree with Jameson and colleagues. Subclinical deep-vein

n engl j med 364;12

Since publication of their article, the authors report no further potential conflict of interest. 1. Hull RD, Brant RF, Pineo GF, Stein PD, Raskob GE, Valentine

KA. Preoperative versus postoperative initiation of low-molecular-weight heparin prophylaxis against venous thromboembolism in patients undergoing elective hip replacement. Arch Intern Med 1999;159:137-41. 2. Strebel N, Prins M, Agnelli G, Büller HR. Preoperative or postoperative start of prophylaxis for venous thromboembolism with low-molecular-weight heparin in elective surgery? Arch Intern Med 2002;162:1451-6.

nejm.org

march 24, 2011

1177

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

Oral Rivaroxaban for Symptomatic Venous Thromboembolism To the Editor: In their article on the use of oral rivaroxaban for the treatment of acute deep-vein thrombosis (DVT), the EINSTEIN investigators (Dec. 23 issue)1 report that rivaroxaban had noninferior efficacy, as compared with subcutaneous enoxaparin followed by a vitamin K antagonist. In a separate double-blind, randomized superiority study comparing extended rivaroxaban therapy with placebo, rivaroxaban was claimed to be superior. However, the net clinical benefit was not predefined as a separate secondary end point in the protocol and at ClinicalTrials.gov. The combination of major and clinically relevant nonmajor bleeding was defined as the principal safety outcome. Although the post hoc definition of net clinical benefit is questionable, it is nevertheless clinically relevant. The greatest threats of DVT and its treatment are pulmonary embolism and major bleeding, whereas recurrent DVT and clinically relevant bleeding are primarily important from the perspective of quality of life and cost.2,3 Therefore, we would like to know the clinical benefit of rivaroxaban when it is defined as the rate of recurrent venous thromboembolism combined with the principal safety outcome, as well as when it is defined as the rate of recurrent pulmonary embolism and major bleeding. Gijs W. Landman, M.D. Isala Clinics Zwolle, the Netherlands [email protected]

2. Lancaster TR, Singer DE, Sheehan MA, et al. The impact of

long-term warfarin therapy on quality of life: evidence from a randomized trial. Arch Intern Med 1991;151:1944-9. [Erratum, Arch Intern Med 1992;152:825.] 3. Gould MK, Dembitzer AD, Sanders GD, Garber AM. Lowmolecular-weight heparins compared with unfractionated heparin for treatment of acute deep venous thrombosis: a cost-effectiveness analysis. Ann Intern Med 1999;130:789-99.

The authors reply: Although the net clinical benefit was not reported as an a priori outcome in the study summary at ClinicalTrials.gov, it was included in the study’s statistical analysis plan, which was approved before enrollment of the first patient and was made available to health authorities. This composite outcome gives equal weight to the primary efficacy outcomes and major bleeding, because of their similar overall clinical importance in terms of need for rehospitalization, long-term consequences (including the postthrombotic syndrome, thromboembolic pulmonary hypertension, and damage to various organ systems), and case fatality rates. This overall risk– benefit outcome was in favor of rivaroxaban, with a hazard ratio of 0.67 (95% confidence interval, 0.47 to 0.95), and was consistent for the various subgroups. Harry R. Buller, M.D. Academic Medical Center Amsterdam, the Netherlands [email protected]

Anthonie W.A. Lensing, M.D.

Reinold O.B. Gans, M.D., Ph.D. University Medical Center Groningen, the Netherlands No potential conflict of interest relevant to this letter was reported. 1. The EINSTEIN Investigators. Oral rivaroxaban for sympto-

matic venous thromboembolism. N Engl J Med 2010;363:2499-510.

Bayer HealthCare Pharmaceuticals Wuppertal, Germany

Martin H. Prins, M.D. Maastricht University Maastricht, the Netherlands Since publication of their article, the authors report no further potential conflict of interest.

Drug-Eluting or Bare-Metal Stents in Large Coronary Arteries To the Editor: Kaiser et al. (Dec. 9 issue)1 shed important light on the issue of the safety and efficacy of drug-eluting stents in large coronary arteries. The characteristics of the patients and their lesions were very well matched between the groups, as disclosed in detail in the article. How-

1178

n engl j med 364;12

ever, it is not clear whether there was any significant between-group difference in the mean reference-vessel diameter or post-procedural luminal diameter, since intravascular ultrasonography (IVUS) and quantitative angiography were not used in the study. These measurements are

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

correspondence

important predictors of restenosis, both with the use of bare-metal stents and drug-eluting stents.2 On the assumption that the stents were expanded as much as possible, as described in the study protocol,3 the diameters of the stents could give some idea about these measurements. It would be interesting to know whether there was any significant difference in the diameters of stents between the groups. Cemil Izgi, M.D. Gaziosmanpasa Hospital Istanbul, Turkey [email protected] No potential conflict of interest relevant to this letter was reported. 1. Kaiser C, Galatius S, Erne P, et al. Drug-eluting versus bare-

metal stents in large coronary arteries. N Engl J Med 2010;363: 2310-9. 2. Dangas GD, Claessen BE, Caixeta A, Sanidas EA, Mintz GS, Mehran R. In-stent restenosis in the drug-eluting stent era. J Am Coll Cardiol 2010;56:1897-907. 3. Pfisterer M, Bertel O, Bonetti PO, et al. Drug-eluting or baremetal stents for large coronary vessel stenting? The BASKET-PROVE (PROspective Validation Examination) trial: study protocol and design. Am Heart J 2008;155:609-14.

To the Editor: The main conclusion of Kaiser et al. was that there was no significant difference in the occurrence of the primary end point (death from cardiac causes or nonfatal myocardial infarction) with drug-eluting stents, as compared with bare-metal stents. However, the comparison of sirolimus-eluting stents with bare-metal stents showed a hazard ratio of 0.54 for the primary end point in favor of the sirolimus-eluting stent. The 95% confidence interval (0.31 to 0.93) suggests that the result is significantly different, but the P value is 0.13. I presume that this was the result of adjustment of data for multiple comparisons. I wonder whether this statistical adjustment should have been performed, since there was only one primary end point and the predefined sample size was calculated for two separate comparisons.1 Could the difference be statistically significant in favor of the sirolimus stent? Also, given the seemingly beneficial effect of sirolimus, I would also be interested in a post hoc analysis combining both drug-eluting stents and bare-metal stents. Marnix W. van Bemmel, M.D. Deventerziekenhuis Zwolle, the Netherlands No potential conflict of interest relevant to this letter was reported.

n engl j med 364;12

1. Pfisterer M, Brunner-La Rocca H-P, Rickenbacher P, et al.

Long-term benefit-risk balance of drug-eluting stents in daily practice: does stent diameter matter? Three-year follow-up of BASKET. Eur Heart J 2009;30:16-24.

To the Editor: The early studies of drug-eluting stents versus bare-metal stents were conducted in a blinded fashion.1 In these studies, no differences in rates of subacute stent thrombosis, myocardial infarction, or death were shown, findings that were confirmed in a meta-analysis2 that was performed to address the safety concerns raised by the Basel Stent Cost-Effectiveness Trial (BASKET).3 The prospective validation study by Kaiser et al. (BASKET–PROVE) further evaluated the risk. Unexpectedly, the early risk was higher with the use of bare-metal stents than with drug-eluting stents. In this study, the operators were not unaware of treatment assignment. It was found that the rate of early (<14 days) death and myocardial infarction was increased with the use of bare-metal stents, suggesting that factors associated with implantation could be involved. It is possible that operators during the trial, concerned about safety with the use of drug-eluting stents, were performing better post-dilatation procedures, used higherpressure dilatations, or used higher final balloon diameters in the groups receiving drug-eluting stents. Can the authors give figures for the proportion of patients having high-pressure (>20 atm) post-dilatation procedures, the proportion of patients who underwent IVUS, the highest implantation pressure used, and the nominal diameter of the biggest balloon used at implantation? Evald H. Christiansen, M.D., Ph.D. Aarhus University Hospital, Skejby Aarhus, Denmark [email protected] No potential conflict of interest relevant to this letter was reported. 1. Moses JW, Leon MB, Popma JJ, et al. Sirolimus-eluting stents

versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2003;349:1315-23. 2. Stone GW, Moses JW, Ellis SG, et al. Safety and efficacy of sirolimus- and paclitaxel-eluting coronary stents. N Engl J Med 2007;356:998-1008. 3. Brunner-La Rocca HP, Kaiser C, Pfisterer M. Targeted stent use in clinical practice based on evidence from the Basel Stent Cost Effectiveness Trial (BASKET). Eur Heart J 2007;28:719-25.

To the Editor: Kaiser et al. define a large coronary artery as one requiring a stent of 3.0 mm or more in diameter. However, patients requiring

nejm.org

march 24, 2011

1179

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

stents larger than 4.0 mm were excluded because the Cypher stent was not available in that size. Thus, their results represent mainly the outcome of a population with “non-small” coronary arteries who were treated with either drug-eluting or baremetal stents. Moreover, the mean stent diameter in each group is not reported. It would be interesting to know the outcome of patients treated with a bare-metal stent of 3.5 mm diameter. The results of this study should not be extrapolated to patients with larger coronary arteries (i.e., with a diameter of >3.5 mm) in which a bare-metal stent might have equivalent efficacy to a drug-eluting stent.1,2 Adel Aminian, M.D. Jacques Lalmand, M.D. Centre Hospitalier Universitaire de Charleroi Charleroi, Belgium [email protected] No potential conflict of interest relevant to this letter was reported. 1. Steinberg DH, Mishra S, Javaid A, et al. Comparison of effec-

tiveness of bare metal stents versus drug-eluting stents in large (> or = 3.5 mm) coronary arteries. Am J Cardiol 2007;99:599-602. 2. Quizhpe AR, Feres F, de Ribamar Costa J Jr, et al. Drugeluting stents vs bare metal stents for the treatment of large coronary vessels. Am Heart J 2007;154:373-8.

of

m e dic i n e

tween everolimus-eluting and bare-metal stents, but sirolimus-eluting stents were somewhat smaller than everolimus-eluting and bare-metal stents (−0.05 mm and −0.06 mm, respectively; P<0.05) despite having higher maximal dilatation pressures (P<0.001). This difference was most likely due to the stainless-steel platform of this stent, as compared with a cobalt–chromium platform in the other two stents. These small differences were clinically negligible, as shown by the virtually identical 2-year outcomes of the two drugeluting stents. Thus, the early risk of death or myocardial infarction in patients receiving baremetal stents that was observed in our study and in several previous studies1-3 cannot be attributed to differences in stent size or maximal pressures applied. We have not yet reported an analysis stratified by final stent size. Although this analysis was planned a priori, it was a secondary aim only and could not be included in the report of the primary end-point results. For the primary analyses, adjustments for multiple comparisons were necessary because the control group (i.e., patients treated with bare-metal stents) was used as comparator for both drug-eluting stents. Finally, it is possible that the newest generations of bare-metal stents may perform even better than the standard stents used, particularly in patients with largevessel stenting, but the same is true for newer generations of drug-eluting stents. However, not all newer generations of stents may show improved outcomes, as we found in our study.

The authors reply: The correspondents’ main questions relate to final stent diameters achieved in the different groups and the maximal dilatation pressures applied. Stents with a nominal diameter of 3.0 mm were used in 72% of patients, with stents of 3.5 mm used in 23% of patients (2% had smaller stents, and 3% had larger stents for Christoph Kaiser, M.D. bailout indications). Stents with a diameter of Matthias Pfisterer, M.D. 4.0 mm or more were not used because of the Hospital Basel unavailability of such sirolimus-eluting stents. University Basel, Switzerland 1 However, in the initial BASKET population, only [email protected] 23 of 988 consecutive patients (2.3%) needed Since publication of their article, the authors report no furstents with a diameter of 4.0 mm or more for ther potential conflict of interest. native-vessel stenting, indicating that this patient group is clinically of little relevance. The mean 1. Kaiser C, Brunner-La Rocca HP, Buser PT, et al. Incremental (±SD) maximal stent diameter per patient was cost-effectiveness of drug-eluting stents compared with a thirdgeneration bare-metal stent in a real-world setting: randomised 3.5±0.3 mm in patients receiving sirolimus-elut- Basel Stent Kosten Effektivitäts Trial (BASKET). Lancet 2005; ing stents, 3.5±0.4 mm in those receiving evero- 366:921-9. limus-eluting stents, and 3.5±0.4 mm in those 2. Park DW, Yun SC, Lee SW, et al. Stent thrombosis, clinical events, and influence of prolonged clopidogrel use after placereceiving bare-metal stents; the mean maximal ment of drug-eluting stent data from an observational cohort balloon pressures were 15.5±3.5 bar, 14.8±3.4 bar, study of drug-eluting versus bare-metal stents. JACC Cardiovasc Interv 2008;1:494-503. and 14.5±3.2 bar, respectively. 3. Lagerqvist B, James SK, Stenestrand U, Lindbäck Nilsson T, There were no significant differences in maxi- Wallentin L. Long-term outcomes with drug-eluting stents versus mal stent sizes or maximal pressures applied be- bare-metal stents in Sweden. N Engl J Med 2007;356:1009-19.

1180

n engl j med 364;12

nejm.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

correspondence

Safety and Effectiveness of a 2009 H1N1 Vaccine in Beijing To the Editor: The study by Wu et al. (Dec. 16 issue)1 has serious methodologic issues that limit the interpretation of their estimate of the effectiveness of a 2009 pandemic influenza A (H1N1) vaccine. First, the cumulative incidence of influenza was 2.5 times as high in the unvaccinated cohort as in the vaccinated cohort even before a protective effect from the vaccine could have been achieved (Fig. 2 of the article). Because of bias or confounding, these two cohorts did not have the same risk of influenza. In addition, many students may have been infected with the pandemic virus before the study began. Vaccine trials in China that preceded this study reported that 18 to 25% of children who were 12 to 18 years of age were positive for the pandemic H1N1 antibody.2 At one Chinese school, 32% of the students were infected.3 A subsequent serologic survey conducted in Beijing showed that 19.4% of children and young people between 6 and 17 years of age were seropositive for the pandemic virus.4 Yet in the study by Wu et al., less than 0.30% of the unvaccinated cohort and 0.01% of the vaccinated cohort were infected. Numerous cases must have been missed in both cohorts before the study began and while it was being conducted. Nicholas S. Kelley, M.S.P.H. Michael T. Osterholm, Ph.D., M.P.H.

vaccinated students had all been participants in China’s 60th National Day Celebration, whereas most of the unvaccinated students had not been participants. Generally, the students in the celebration were healthier and better educated about the prevention of infection, and most were living in urban districts — all of which might lead to an overestimation of the effectiveness of the vaccine. Although aware of these differences between the two groups, the authors took no measures to eliminate them. In fact, a total of about 80,000 primaryschool or middle-school students participated in the celebration, including 25,037 vaccinated students and 55,000 unvaccinated students. I wonder why these 55,000 unvaccinated students were not selected as the control group, which might have eliminated the interference of the above-mentioned factors. Xiang-Dong Mu, M.D. Peking University First Hospital Beijing, China [email protected] No potential conflict of interest relevant to this letter was reported.

The Authors Reply: In our study, both the unvaccinated cohort and the vaccinated cohort came from 245 schools. The vaccinated cohort consisted of students who participated in the 60th NaUniversity of Minnesota Minneapolis, MN tional Day Celebration and were vaccinated during [email protected] a 5-day period in September 2009. The unvaccinatEdward A. Belongia, M.D. ed cohort included students who did and those Marshfield Clinic Research Foundation who did not participate in the celebration. The stuMarshfield, WI dents in the vaccinated cohort attended the same No potential conflict of interest relevant to this letter was reschools at the same times and had backgrounds ported. similar to those in the unvaccinated cohort. Hence, 1. Wu J, Xu F, Lu L, et al. Safety and effectiveness of a 2009 both cohorts should have had similar exposure H1N1 vaccine in Beijing. N Engl J Med 2010;363:2416-23. 2. Liang XF, Wang HQ, Wang JZ, et al. Safety and immunoge- to influenza viruses. nicity of 2009 pandemic influenza A H1N1 vaccines in China: The incidence data for our study were collecta multicentre, double-blind, randomised, placebo-controlled ed from October 9 through November 15, 2009. trial. Lancet 2010;375:56-66. [Erratum, Lancet 2010;375:1694.] 3. Li T, Liu Y, Di B, et al. Epidemiological investigation of an Yang et al. performed their study from Novemoutbreak of pandemic influenza A (H1N1) 2009 in a boarding ber 27 through December 23, 2009, and their school: serological analysis of 1570 cases. J Clin Virol 2011;50: subjects were patients from outpatient sections 235-9. 4. Yang P, Huang F, Shi W, et al. A survey on serological epide- of six hospitals (two children’s hospitals and four miology of influenza A (H1N1) 3009 in Beijing. Zhonghua Liu general hospitals) in Beijing. Since Yang et al. Xing Bing Xue Za Zhi 2010;31:485-8. (In Chinese.) used convenience sampling and the study subjects did not represent the entire student populaTo the Editor: Wu et al. evaluate the effective- tion, there is no conflict in this regard between ness of a 2009 H1N1 vaccine. In this study, the our report and theirs. In addition, there is some

n engl j med 364;12

nejm.org

march 24, 2011

1181

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

corrections

debate about the cross-reactive antibody response between H1N1 and pandemic H1N1. The data from Chi et al. indicate that natural infection with the 2006 and 2008 seasonal H1N1 viruses in Taiwan may have induced a cross-reactive antibody response to the 2009 pandemic H1N1 virus.1 We detected a similar phenomenon. We tested 301 serum samples from volunteers after H5N1 vaccination in Beijing in 2007 and observed a cross-reaction in 2.9% of the group between 3 and 11 years of age and in 11.3% of the group between 12 and 17 years of age.2 There was also a report from the United States of a possible crossreactive antibody response to the H1N1 and pandemic H1N1 viruses.3 Our experience with seasonal flu vaccine and pandemic H1N1 flu vaccine also indicated the existence of this phenomenon. Thus, measurement of serum antibodies probably did not fully reflect the actual infection status of the population.4 We agree that some cases may have been missed in our study, such as silent infection and mild cases. However, this phenomenon should have affected both the vaccinated and unvaccinated subjects. Jiang Wu, M.D. Min Lu, Ph.D. Ying Deng, M.D. Beijing Centers for Disease Control and Prevention Beijing, China [email protected] Since publication of their article, the authors report no further potential conflict of interest. 1. Chi CY, Liu CC, Lin CC, et al. Preexisting antibody response

against 2009 pandemic influenza H1N1 viruses in the Taiwanese population. Clin Vaccine Immunol 2010;17:1958-62. 2. Clinical trial interim report of pandemic influenza A/H1N1 vaccine: serial number CR-PanFlu-4001, version 2009;0:32. 3. Hancock K, Veguilla V, Lu X, et al. Cross-reactive antibody responses to the 2009 pandemic H1N1 influenza virus. N Engl J Med 2009;361:1945-52. 4. Wu J, Zhong X, Li CK, et al. Optimal vaccination strategies for 2009 pandemic H1N1 and seasonal influenza vaccines in humans. Vaccine 2011;29:1009-16.

• Letters not related to a Journal article must not exceed 400 words. • A letter can have no more than five references and one figure or table. • A letter can be signed by no more than three authors. • Financial associations or other possible conflicts of interest must be disclosed. Disclosures will be published with the letters. (For authors of Journal articles who are responding to letters, we will only publish new relevant relationships that have developed since publication of the article.) • Include your full mailing address, telephone number, fax number, and e-mail address with your letter. • All letters must be submitted at authors.NEJM.org. Letters that do not adhere to these instructions will not be considered. We will notify you when we have made a decision about possible publication. Letters regarding a recent Journal article may be shared with the authors of that article. We are unable to provide prepublication proofs. Submission of a letter constitutes permission for the Massachusetts Medical Society, its licensees, and its assignees to use it in the Journal’s various print and electronic publications and in collections, revisions, and any other form or medium.

corrections Glycemic Control in the ICU (December 23, 2010;363:2540-6). In the Studies of Intensive Insulin Therapy subsection of Management (page 2541), the first sentence in the third paragraph should have cited reference 8, rather than 9, and the second sentence in the final paragraph (page 2543) should have cited references 6 and 8, rather than 7 and 9. In the Glucose Variability subsection of Areas of Uncertainty (page 2543), the second sentence in the first paragraph should have ended, “. . . as is hypoglycemia that develops spontaneously,” rather than “. . . regardless of the glucose level.” Finally, in the Hypoglycemia subsection of Areas of Uncertainty (page 2544), the second sentence of the first paragraph should have cited reference 27, rather than 30. The article is correct at NEJM.org. Emergence of a Multidrug-Resistant Pandemic Influenza A (H1N1) Virus (September 30, 2010;363:1381-2). In the third paragraph (page 1382), the date mentioned should have been January 4 rather than February 9, and in Figure 1 (page 1381), the dates on the x axis should have included Jan. 4, 2010, rather than Feb. 9, 2010. The article is correct at NEJM.org. Case 4-2009: A 39-Year-Old Pregnant Woman with Fever after a Trip to Africa (January 29, 2009;360:508-16). In the third paragraph (page 508), the fifth sentence should have begun, “She took mefloquine weekly,” rather than “She took mefloquine daily.” The article is correct at NEJM.org.

Correspondence Copyright © 2011 Massachusetts Medical Society.

INSTRUCTIONS FOR LETTERS TO THE EDITOR

THE JOURNAL’S WEB AND E-MAIL ADDRESSES:

Letters to the Editor are considered for publication, subject to editing and abridgment, provided they do not contain material that has been submitted or published elsewhere. Please note the following: • Letters in reference to a Journal article must not exceed 175 words (excluding references) and must be received within 3 weeks after publication of the article.

1182

n engl j med 364;12

For letters to the Editor: authors.NEJM.org For information about the status of a submitted manuscript: authors.NEJM.org To submit a meeting notice: [email protected]

nejm.org

The Journal’s Web pages: NEJM.org

march 24, 2011

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

images in clinical medicine

Podoconiosis

A

36-year-old man presented to an Ethiopian clinic with a 20-year Whitney Lapolla, M.D. history of skin nodules, pain, and edema involving his legs and feet. He was Stephen K. Tyring, M.D., Ph.D. otherwise healthy and worked as a farmer. Circulating filarial antigen tests Center for Clinical Studies for the presence of Wuchereria bancrofti were negative; he was not tested for other Webster, TX types of filaria. This clinical presentation prompted a diagnosis of probable podo- [email protected] coniosis (also known as nonfilarial elephantiasis or mossy foot). This locally endemic, noninfectious condition is caused by the long-term exposure of susceptible persons to irritant volcanic soil. Colloid particles are thought to be absorbed through the skin and taken up by macrophages, leading to lymphatic fibrosis and elephantiasis. Affected persons are typically barefoot agricultural workers in the highland tropics. Social stigma associated with this condition is widespread; patients are banned from schools, churches, and marriage. Economic productivity is often impaired. Podoconiosis is preventable with fastidious shoe wearing and foot hygiene. Treatment is limited to compression bandaging and elevation. The patient was instructed to wear shoes, but additional nodules continued to develop on uncovered areas of his sandaled feet.

Copyright © 2011 Massachusetts Medical Society.

n engl j med 364;12

nejm.org

march 24, 2011

e23

The New England Journal of Medicine Downloaded from nejm.org at MEDIZINISCHE UNIVERSITAT WIEN on March 26, 2011. For personal use only. No other uses without permission. Copyright © 2011 Massachusetts Medical Society. All rights reserved.