GAITHERSBURG, Md -- Internal divides between Food and Drug Administration officials over the safety of the diabetes drug Avandia were on public display today, as an FDA advisory panel began wrestling with warnings that the drug increases the risks of heart attacks and other cardiovascular problems.
The drug, which won approval in 1999 to treat type-2 diabetes, has been under assault since 2007, when a broad study assessing results of numerous clinical trials showed elevated dangers. Since then, those red flags have been reproduced by at least two more "meta-analyses’’ studying data from a variety of clinical trials, as well as an FDA-sponsored review of large volumes of patient outcome information collected by Medicare.
FDA, which is warring with itself over how to respond, has asked its advisory panel to make a recommendation tomorrow on whether the drug should be pulled from the market.
The outcome will signal FDA’s latest approach to drug safety, and what it believes constitutes sufficient evidence of risk to withdraw a drug after it has been approved.
The Institute of Medicine last week produced a report saying "observational studies’’ of large volumes of patient outcome data may in some instances be even better than clinical trials at assessing safety of a marketed drug. But for many specialists, clinical trials in which large populations of patients are tested and monitored remain the ``gold standard’’ for learning a drug’s effects.
In defending Avandia, GlaxoSmithKline has continued to rely on several such clinical trials comparing Avandia with other treatments. Problems with the integrity of one of those trials, a key study called RECORD, which GlaxoSmithKline sponsored, were debated heavily today before the advisory panel.
An official of the FDA branch that approves new drugs recognized serious shortcomings with RECORD, but concluded that the study should not be discounted completely. He said the trial did not confirm the warnings that have been building up since 2007.
"The findings on (heart attack) seem inconclusive,’’ said the official, Dr. Ellis Unger, a deputy director in the office of new drugs.
But the RECORD trial was strongly criticized today by another FDA official, who wrote in an exhaustive report earlier this month that some individual incidents of patient deaths and apparent cardiovascular events had been improperly left out of data.
"This type of handling of events is completely unacceptable,’’ Marciniak said of RECORD trial investigators’ decisions to leave out certain adverse events. When those adverse events were folded back into the results, he said, "RECORD suggests … (Avandia) increases the risk’’ for heart attack.
Unger alluded to the disagreements between FDA officials in his remarks before the panel. "There’s no clear picture here. Otherwise, you guys would be back at home,’’ he told the panel.
A third FDA official, Dr. David Graham, associate director of science and medicine, was scheduled to testify this afternoon before the panel. Graham has repeatedly called for Avandia’s immediate removal from the market, based on his assessment that the drug does pose clear and serious risks. Also testifying this morning was Dr. Steven Nissen, a Cleveland Clinic cardiologist who published the 2007 analysis that got the ball rolling against Avandia. He said Avandia never should have been approved by the FDA in the first place.
In its haste to replace another diabetes drug that had been withdrawn from the market for safety reasons, "the mistake got made,’’ Nissen testified. "The genie got out of the bottle, and we’ve been trying to put the genie back in ever since.’’
Meanwhile today, the leading members of the Senate Finance Committee released a batch of documents gathered during litigation over Avandia’s safety that detail evidence of the company’s previous knowledge of safety issues.
The documents include internal GlaxoSmithKline emails that Republican Chuck Grassley and Democratic Chairman Max Baucus said indicated the company sought to hide signs of Avandia’s health risks.
One email was from a GlaxoSmithKline official who alluded to the company’s desire to avoid a head-to-head test between Avandia and Actos, its primary competitor, to measure low-density lipoprotein, known as "bad’’ cholesterol. Instead, GlaxoSmithKline tested Actos only, and compared it to historical data from previous Avandia tests.
"This was done for the US business, way under the radar, and we lost both,’’ the GlaxoSmithKline official wrote. "Per Sr. Mgmt. request, these data should not see the light of day to anyone outside of GSK.’’
GlaxoSmithKline said in response today that the email and other documents released by the committee had been taken out of context.
About Political Intelligence
Glen Johnson is Politics Editor at boston.com and lead blogger for "Political Intelligence." He moved to Massachusetts in the fourth grade, and has covered local, state, and national politics for over 25 years. E-mail him at email@example.com. Follow him on Twitter @globeglen.