|
|
| |||||||||||||||||||||||||||||||||||
Of course, the system doesn't always perform as well as it should. In recent years, the FDA has been less stringent about allowing the participation of committee members who have commercial conflicts of interest, despite evidence that such ties help shape opinions.1 Science may also be defeated at later stages of the policymaking process, as when the FDA commissioner overruled the nearly unanimous recommendations of both internal and external scientists about the safety of emergency contraception (the morning-after pill). Unfortunately, the upcoming reauthorization of user-fee funding at the agency will lock in 5 more years of dependence on support from pharmaceutical manufacturers to pay the salaries of the FDA's internal drug-review staff, renewing concern about the relationships between science, commerce, and federal policy and making the independence of external advisers even more critical.2
But though the quiet voice of science may often be no match for powerful vested interests or ideology, some encouraging signs may be in the air. The same reauthorization bill, disappointing in so many respects, may tighten somewhat the conflict-of-interest rules for outside advisers. In addition, some recent committee decisions provide interesting contrasts with recommendations made several years ago about similar drugs.
During the 1999 approval process for rofecoxib (Vioxx, Merck), FDA internal reviewers noted early signals of a possible increase in cardiovascular risk. But the advisory committee focused instead on the hope that the drug would have less gastrointestinal toxicity than other nonsteroidal antiinflammatory drugs (NSAIDs), though such an advantage had not yet been convincingly demonstrated. When a quadrupling of the rate of myocardial infarction was documented a year later in a clinical trial comparing rofecoxib with naproxen, the agency allowed the company to imply that this was because of the cardioprotective effect of naproxen.
Fast-forward to 2007, when the same company sought FDA approval for etoricoxib (Arcoxia), a new drug in the same class. Merck had initially proposed — and the agency had approved — a study comparing etoricoxib with diclofenac, an NSAID that many worried carried its own cardiac risk. Several years and millions of dollars later, Merck presented the FDA with trial evidence that etoricoxib caused roughly the same number of cardiac events as diclofenac. But this time, the FDA allowed its sharpest internal critic, David Graham, to present data to the advisory committee on the implications of approving another cyclooxygenase-2 inhibitor with potentially dangerous cardiac side effects. Had the company used a more appropriate comparator, naproxen, Graham and the others argued, the increased cardiovascular risk would have been clear (see table). The committee voted overwhelmingly against approval. If diclofenac also presented a cardiac risk, the committee agreed, then the FDA should not approve a new product with no proven advantage that might confer the same hazard.3 Not only did this decision fly in the face of the original study design, it was also a sharp departure from the conventional FDA view that a new drug in an established class need not be any safer or more effective than its predecessors. The committee's vote was so lopsided that the agency, still embarrassed at having missed the risk of myocardial infarction associated with rofecoxib during 5 years of widespread use, could not but agree with its recommendation.
|
A different fate has met the most recent entry into the lucrative diet-drug market, rimonabant. Advocates said it could reduce appetite, improve lipid levels, and help people quit smoking. Faced with such a potential embarrassment of riches, its manufacturer, Sanofi-Aventis, entered the blockbuster lottery by trying for an appetite suppressant indication. Rimonabant worked slightly better than dexfenfluramine — clinical trials demonstrated a weight loss of 13 lb more than that achieved with placebo — but the drug also seemed to raise the risk of depression (sometimes severe), suicidality, anxiety, and insomnia. Despite these adverse effects, FDA staff issued an encouraging "approvable" letter in February 2006. But when an advisory committee reviewed the evidence in June 2007, it voted unanimously against the drug.5 Soon thereafter, the manufacturer withdrew its application.
The agency is now addressing yet another follow-on drug as the glitazone story evolves. In 1997, the first entry in this class of antidiabetic agents, troglitazone (Rezulin, Parke-Davis), was observed to cause fulminant hepatic necrosis, sometimes fatal. Regulatory authorities throughout the world quickly concluded that the product had an indefensible risk–benefit ratio, and it was withdrawn from the market, often within just a few months of approval. Yet the FDA and its advisory committee were swayed by the arguments of the manufacturer and kept it in use in the United States for 2 years after it had been made unavailable in nearly every other country.4 Now, a decade later, troglitazone's younger sibling, rosiglitazone (Avandia, GlaxoSmithKline), has been implicated in raising the risks of congestive heart failure and myocardial infarction, without impressive evidence of a countervailing advantage in clinical outcomes. An advisory committee meeting on July 30, 2007, did not fuel hopes for a new era of data-driven reform. The committee voted, 20 to 3, that rosiglitazone increases cardiac ischemic risk in type 2 diabetes but then recommended, by a 22-to-1 vote, that the drug remain in use. The decision was more suggestive of Rezulin redux (and of Redux) than it was of resolve. Although Avandia has been prescribed widely since 1999, several participants noted that neither the manufacturer nor the FDA had carried out enough safety studies to permit a clear conclusion.
The approval, prescribing, and safety surveillance of prescription drugs involve a complicated mix of science, regulatory law, clinical judgment, business, and politics. It is not easy to ensure that science dominates in such a heady brew, but despite missteps such as the latest move with respect to rosiglitazone, an open model holds more promise for data-driven public decision making than those followed in the energy, finance, and defense sectors, among others. As Congress persists in allowing industry funding to dominate the FDA budget (and, many fear, its perspective as well), it will be especially important for the scientific community to remain independent, conduct rigorous analyses, and make its voice heard clearly to ensure that drug-review decisions are driven solely by the data.
Dr. Avorn reports having served, pro bono, as an expert witness for plaintiffs in Vioxx-related lawsuits.
Source Information
Dr. Avorn is a professor of medicine at Harvard Medical School and chief of the Division of Pharmacoepidemiology and Pharmacoeconomics at Brigham and Women's Hospital — both in Boston.
References
| |||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
