Next month, the Obama-era US Food and Drug Administration (FDA) will confront what is likely to be the most defining moment in its short history. In a public meeting scheduled for mid-July, a committee of external experts will advise agency leaders—who may or may not accept their advice—on whether to ban a controversial diabetes drug implicated in causing heart attacks, when an equally effective competitor is also on the market, minus the heart attack risks.

The extensive media coverage and drawn-out agency decision making over the blockbuster drug, Avandia (rosiglitazone), made by UK-based GlaxoSmithKline, point to an increasingly noticeable power struggle within the agency. That struggle pits government epidemiologists who track the safety of drugs once they are on the market against their colleagues in an FDA division five times as large where regulators approve new drugs. Critics argue that these regulators are prone to inaction when drugs they have approved turn out to be unsafe. Those on the other side of the argument retort that every drug presents a balance of risks and benefits, and that the agency should not hastily restrict marketed drugs without compelling new evidence of harm to patients.

Two FDA epidemiologists who surveyed the literature reported in 2008 that roughly 500 excess heart attacks would be prevented each month if every diabetic taking Avandia were switched to its only direct competitor, Actos (pioglitazone), made by Takeda Pharmaceutical in Osaka, Japan. They recommended Avandia's immediate removal from the market.

The epidemiologists, David Graham and Kate Gelperin, also called for a halt in a large ongoing trial comparing Avandia head to head with Actos, which shows no evidence of boosting heart attack risk. (Both drugs do cause some patients to develop congestive heart failure, a distinct condition from heart attack.) They said that the trial, called TIDE, is “unethical and exploitative” because it puts patients in the Avandia arm at unacceptable risk, given the availability of a safer, equally effective, alternative. The agency has not acted on either recommendation.

After the epidemiologists' conclusions were publicized in a February report compiled by staff of the Senate Finance Committee—whose senior Republican, Senator Charles Grassley, has taken a keen interest in the policing of drugs and drug companies—FDA Commissioner Margaret Hamburg told senators that the agency will exhaustively consider all of the available evidence on the cardiovascular safety of Avandia. She added that, on the basis of this review and input from a double-barreled advisory committee of diabetes and drug safety experts, the FDA would make a decision about the drug's future.

For the new FDA, “Avandia is certainly a great test case,” says Steven Nissen, a cardiologist at the Cleveland Clinic in Ohio, whose 2007 meta-analysis in The New England Journal of Medicine (NEJM) first publicly reported an increased heart attack risk with Avandia (N. Engl. J. Med. 356, 2457–2471, 2007). “Will the FDA reconfigure itself as a public health agency, or will it continue to primarily be in the business of promoting pharmaceuticals?”

Other experts say that the black-and-white choice presented by Nissen is oversimplified and that data on Avandia's heart attack risks are inconclusive, making it important for the TIDE trial to proceed.

“While the original [NEJM] article still raises concerns, subsequent data have not further elevated that concern,” says David Kendall, the chief scientific and medical officer of the American Diabetes Association (ADA). The ADA has not changed the position on Avandia it adopted after the publication of the 2007 NEJM article. It advised patients on the drug to continue taking it or to consult with their physicians if the lack of clarity on Avandia's heart risks concerns them.