FDA Panel Splits Over Avandia Ban

20-12 Vote to Strengthen Avandia Warning, but Many Panelists Want Ban

From the WebMD Archives

July 14, 2010 -- Avandia should carry more severe label warnings and perhaps require patient education, most members of an FDA advisory panel today voted.

However, 12 of the panel's 33 members advised the FDA to take the drug off the market. Several other panel members said they nearly voted the same way.

Only three panelists voted to leave Avandia on the market with its current label. Many panelists who voted not to ban Avandia said they did so because of worries that a small number of patients may need the drug if other treatments fail.

Ten of the panelists want Avandia maker GlaxoSmithKline to develop patient and doctor education programs to raise awareness of the drug’s risks. Avandia and Actos, the other approved drug in the same class, both increase risk of heart failure.

But there is growing but inconclusive evidence that Avandia also increases the risk of heart attack, fracture, infection, and high cholesterol.

"The evidence is much less convincing than we would like it to be … but there is meaningful evidence of Avandia harm, and I have not heard meaningful evidence of benefit that cannot be obtained by another drug. The public would be best served by not having the drug available," said panelist Richard Platt, MD, chair of population medicine at Harvard University.

"I absolutely could not vote for withdrawal," said panelist Arthur J. Moss, MD, professor of cardiology at the University of Rochester, N.Y. "I wanted to send a message to the FDA because a proper trial should have been done in 1993 when they first had an investigational new drug application [for Avandia], and that trial still needs to be done.

Critics: Avandia Clinical Trial Unethical

The split vote puts the decision on whether to ban Avandia back in the hands of the FDA. Although the FDA is likely to go with the panel vote, it may decide to ban the drug based on the reservations expressed by the majority of panelists.

The dramatic panel vote followed a discussion in which the 33 panel members agonized over their decision. They are fully aware that their deliberations will be a major factor in the way the FDA approves all new drugs in the future.


That's because the FDA itself is split over the Avandia question. The split is over a philosophical question: Should safety concerns take priority, or should the FDA focus on the need of patients (and drugmakers) for new treatments?

It's not an easy question, especially as the Avandia controversy has called into question the quality of data submitted to the FDA. Avandia maker GlaxoSmithKline has come under fire for hiding unfavorable data on Avandia and for irregularities in handling and interpreting clinical trial data.

The panel also voted:

  • 18-6 that available data suggest Avandia increases heart attack risk more than do other classes of diabetes drugs. Nine panelists said they did not have enough data to decide.
  • 21-3 that available data suggest Avandia increases heart attack risk more than does Actos. Nine panelists said they did not have enough data to decide.
  • 21-1 that there is not enough data to show that Avandia increases risk of death more than other classes of diabetes drugs. Eleven panelists said they did not have enough data to decide.

The panel also voted 19-11 to continue with a clinical trial, nicknamed TIDE, comparing Avandia to Actos. Critics of the study say the trial is unethical, as it is basically asking patients to take Avandia to see if it harms them. But most panelists, frustrated by two days of inconclusive study presentations, want to see some answers -- even if it takes years.

Oddly, panel members who voted to ban Avandia voted to continue the TIDE trial. One such panelist was Marvin Konstam, MD, professor of cardiology at Tufts University, Boston.

"The panel vote was to leave the drug on the market, and so we will need this study," Konstam said. "You need to inform the patient enrolled in the trial that the only reason Avandia is in the study is to see if it harms you. That is a problem."

WebMD Health News Reviewed by Laura J. Martin, MD on July 14, 2010



FDA advisory panel meeting, July 13-14, 2010.

Richard Platt, MD, chair of population medicine, Harvard University; member, FDA advisory panel.

Arthur J. Moss, MD, professor of cardiology, University of Rochester, N.Y.; member, FDA advisory panel.

Marvin Konstam, MD, professor of cardiology, Tufts University, Boston; member, FDA advisory panel.

© 2010 WebMD, LLC. All rights reserved.