GAITHERSBURG, MD–An advisory panel to the U.S. Food and Drug Administration (FDA) today overwhelmingly recommended that a multibillion dollar diabetes drug should stay on the market, despite the chance that it may cause serious heart problems. Millions of people have taken the drug, Avandia, to help control blood sugar levels in the 8 years that it has been publicly available.
Controversy has swirled around Avandia since May. It was then that Steven Nissen, chair of cardiovascular medicine at the Cleveland Clinic in Ohio, published in the New England Journal of Medicine an analysis of 42 clinical trials and concluded the drug boosted heart disease risk by 43% (Science 15 June, p. 1550). GlaxoSmithKline subsequently revealed that it had conducted a review itself and found an increased risk of 31%, submitting that information to FDA last year. The agency, which also conducted its own review and arrived at a comparable increased risk, came under fire for not acting earlier.
At today's meeting here, more than a dozen physicians, patients, and health advocates urged the advisory committees to act in one direction or the other. Two of them, Sidney Wolfe of Public Citizen, a public health watchdog group in Washington, D.C., and David Egilman of Brown University, argued that if Avandia were up for approval today, it would be rejected. Some endocrinologists and patients implored FDA and its advisers to keep the drug available and endorsed its efficacy.
A split within FDA was apparent. Speaking at the hearing, David Graham of FDA's Office of Surveillance and Epidemiology, which tracks the safety of drugs, estimated that Avandia has so far caused serious cardiac problems, including deaths, in 66,000 to 200,000 people. Graham called for the drug to be pulled; his boss, Gerald Del Pan, who directs that office, shares that point of view. Graham believes that Avandia offers no unique benefits compared to other drugs in battling diabetes, but all indications point to it notably increasing risks of heart attack and sudden death.
But members of FDA's Office of New Drugs, which approved Avandia in the first place, were more circumspect. Like Glaxo executives and some other researchers unaffiliated with the company, they worry that the meta-analyses may be misleading. Just a few heart attacks, for example, can have a substantial effect on the risk increase observed, as can the methodology used to analyze the data, some argued. Sanjay Kaul, a cardiologist at Cedars-Sinai Medical Center in Los Angeles, noted that he and his colleague George Diamond's re-analysis of the trials in Nissen's meta-analysis revealed that the data are "very fragile … and subject to interpretation."
There was much head-scratching among the 23 advisers, who help guide FDA about endocrinology drugs and drug safety, about whether the demonstrated benefits of Avandia outweigh its still-uncertain risks. Many also expressed frustration with data they deemed inadequate, such as short-term trials in relatively healthy people. Eric Holmboe, director of clinical performance services of the American Board of Internal Medicine in Philadelphia, estimated that anywhere from seven to 3000 people might need to be treated with Avandia for one to have a serious cardiac problem linked to the drug. "It depends on how you crunch this data," he said. Still, the data were solid enough that the advisory committee voted 20 to 3 that Avandia boosts the risk of heart disease in type 2 diabetes. However, the committee nevertheless sided against FDA's drug safety office, voting 22-1 in recommending that Avandia stay on the market. Now, FDA will have to decide whether to follow its advisers' recommendations-- normally, it does--and whether to add additional warnings to Avandia or demand that Glaxo alter ongoing studies, or launch new ones.
The picture is also murky for the only other drug in the same class as Avandia, Actos, made by Japan's Takeda Pharmaceuticals. In some studies, Actos does not appear to boost cardiac risks and may even be beneficial. However, an observational study presented today by the Department of Defense, in which it culled its database for patients on Avandia or Actos, found no difference in the heart effects of the drugs. Takeda is currently conducting a meta-analysis of its Actos trials similar to what has been done for Avandia.