The meeting's first day brought an exhaustive and often testy dissection of research on Avandia, whose generic name is rosiglitazone. Drawing on an analysis prepared by a team at Duke University Medical School, FDA staff experts said that an early clinical trial turned up no convincing evidence that Avandia raises a patient's risk of heart attack or death.
FDA officials called for the "readjudication" of research on Avandia in 2010 after a team led by Cleveland Clinic cardiologist Dr. Steven Nissen analyzed the collective results of 42 smaller clinical trials. Nissen's team concluded that diabetics taking the medication were at 43% greater risk of suffering heart attack or death than were patients who controlled their diabetes with the older drug metformin, sometimes in combination with the diabetes drug sulfonylurea.
That damning analysis led the FDA in 2010 to issue stiff warnings about the use of rosiglitazone, which ended the drug's spectacular record of commercial success. Once the world's top-selling medication for Type 2 diabetes, Avandia had its marketing license revoked in Europe in 2010. In the United States, only 3,000 people take the medication today, under close monitoring by the FDA and the drug's manufacturer, the British pharmaceutical firm GlaxoSmithKline.
On Thursday, members of the FDA advisory panel reviewing Avandia's safety profile are to vote on whether to recommend that the FDA relax those restrictions, maintain or extend them, or pull the drug from the U.S. market altogether.
On Wednesday, FDA staff experts offered a spirited defense of the agency's reliance on a clinical trial conducted by GlaxoSmithKline. The trial at the heart of the controversy was called RECORD, short for Rosiglitazone Evaluated for Cardiovascular Outcomes and Regulation of Glycemia in Diabetes.
Duke's reevaluation of the RECORD trial not only upheld RECORD's finding that patients taking rosiglitazone had fewer heart attacks and deaths than did those on metformin and sulfonylurea, FDA staffers testifed, it also appeared to clear GlaxoSmithKline of widespread charges of impropriety in the conduct of that research.
"After readjudication, we do not find a cardiovascular 'signal' of concern in RECORD," wrote the Duke team. "The mortality data are, in fact, reassuring."
FDA scientist Dr. Eugenio Andraca-Carrera said the Duke review turned up "no evidence of systematic bias" in GlaxoSmithKline's collection of data comparing heart attacks and deaths among subjects taking Avandia and those taking a combination of metformin and sulfonylurea to control their diabetes. The pattern of judgments made by the company's researchers and reviewed by the Duke clinical trial specialists "does not suggest a fraudulent intent" to favor Avandia, Dr. Karen Mahoney, diabetes team leader for the FDA review, said Wednesday.
But the Duke readjudication came under harsh criticism from Dr. Thomas A. Marciniak, medical team leader for the FDA's division of cardiovascular and renal products. In testimony before the advisory committee Wednesday, Marciniak charged that the readjudication conducted by the Duke School of Medicine team was hampered by its reliance on outcomes data supplied by the researchers who conducted the RECORD trial for GlaxoSmithKline.
In a review of the original trial, Marciniak suggested that those researchers had deleted, reinterpreted or withheld patient outcomes that would reflect poorly on Avandia. That bias was made possible by the trial's design, in which researchers and patients alike knew whether they were taking Avandia (the trial drug) or metformin plus sulfonylurea (the comparison therapy). Trials meeting the gold standard of biomedical research would normally prevent subjects and researchers from knowing whether they were on the trial drug or the agent against which it is being compared.
In July 2012, as part of a $3-billion settlement with the Justice Department, GlaxoSmithKline pleaded guilty to failing to report certain safety data about Avandia to the FDA, including data from the RECORD study. The firm paid a criminal fine of $242,612,800 in connection with that plea. In a related civil settlement, GlaxoSmithKline agreed to $657 million for supplying physicians and other healthcare providers with "false and misleading representations about Avandia's safety profile."
In an interview Wednesday, Nissen of the Cleveland Clinic called the FDA proceedings a "bizarre" effort to salvage its reputation in the wake of an oversight gaffe that resulted in an estimated 50,000 unnecessary deaths or heart attacks among patients taking Avandia.
"The FDA was intensely embarrassed" and is intent on conducting a "bureaucratic whitewash" that is a waste of taxpayer funds, Nissen said.