FDA Advisers Urge Risk Assessment When Prescribing Rosiglitazone (Avandia)

GAITHERSBURG, Md., July 31 -- Patients with severe type 2 diabetes who use insulin or those with established heart disease are probably not good candidates for rosiglitazone (Avandia).

This appeared to be the key take-home message of a special FDA advisory panel meeting here on the cardiovascular safety of the drug.

Twenty of the 23 FDA advisers who spent all day Monday deliberating rosiglitazone's pluses and minuses agreed that available data supported a conclusion that it increases the cardiac ischemic risk in type 2 diabetes.

At the same time, the members of the FDA's endocrinologic and metabolic drugs advisory committee and the drug safety and risk management advisory committee, meeting jointly, also agreed that the drug should stay on the market.

GlaxoSmithKline, the drug-maker, immediately issued a press release that called the advisers' action a positive result for patients. GSK also interpreted the vote as an agreement that there was a "suggestion" of increased risk.

Interpreting the risk in individual patients is likely to be challenging, said Clifford Rosen, M.D., the Bangor, Me., endocrinologist who chaired the meeting.

A series of subset analyses of data from clinical trials presented to the advisers found a small but statistically significant increased risk of myocardial infarction among diabetes patients taking insulin and those using nitrates, said FDA statistician Joy Mele.

There was also a suggestion of increased risk among patients using an ACE-inhibitor and patients using rosiglitazone in combination with metformin.

Dr. Rosen said the committee was "extremely concerned" about the use of rosiglitazone in patients who "use lots of nitrates, patients using insulin, patients with ischemic heart disease."

He said that patients who fall into one or more of those categories should not be using rosiglitazone. "If patients are on the drug, they probably should be taken off and put on another drug. There are other drugs."

One alternative would be pioglitazone (Actos), a drug that David Graham, M.D., M.P.H., the associate director of FDA's Office of Surveillance and Epidemiology, said may not have the same risk as rosiglitazone.

To support that contention Dr. Graham and Kate Gelperin, M.D., M.P.H., a medical officer in Dr. Graham's division, presented unpublished data from an observational study that compared myocardial infarction rates for patients using rosiglitazone and pioglitazone. The data indicated that pioglitazone reduced the MI risk by 22% (adjusted HR 0.78 (95% CI 0.63-0.96).

The pioglitazone data were, however, not fully reviewed by the FDA and were a last minute addition to the FDA presentations, which upset many of the committee members. Most agreed with committee member Judith M. Kramer, M.D., M.S., of Duke University who said she was not comfortable making a decision about rosiglitazone compared to pioglitazone because the findings were presented as "just a couple of slides."

Those data are scheduled for publication later this week in the journal Pharmacoepidemiology and Drug Safety.

David Schade, M.D., of the University of New Mexico, was one of the advisers who said the evidence didn't prove an increased risk of ischemic heart disease with rosiglitazone.

He said, however, that the data suggested that the drug appeared to be safest among patients with new-onset diabetes or those who have insulin resistance or elevated fasting glucose levels but have not yet progressed to frank diabetes.

"Over time, however, as diabetes progresses the disease changes," Dr. Schade said. Patients whose diabetes has progressed "so that they have coronary heart disease or heart failure may have an increased risk with this drug."

Although sparing rosiglitazone by recommending that it be kept on the market, the advisers also confirmed the validity of a much-debated meta-analysis by Steven Nissen, M.D., of the Cleveland Clinic.

For more than three months Dr. Nissen's meta-analysis has been the subject of both praise and criticism, with the critics usually downplaying the value of a meta-analysis.

The panel, by contrast, agreed that the strongest evidence of ischemic heart disease risk with rosiglitazone came from three meta-analyses of randomized clinical trials-one by GSK, another by FDA staffers, and the one by Dr. Nissen.

Dr. Nissen's analysis, which was published online May 21 by the New England Journal of Medicine, was widely regarded as the spark that turned mild concern about the cardiovascular safety of the drug into a cause clbre. The analysis, which included data from 42 trials, found a 43% increase in risk of myocardial infarction with rosiglitazone.

During the hearing Dr. Nissen played the role of the 800-pound gorilla. He was a designated consultant, but he was relegated to a seat in the audience, while the advisers, representatives of the GlaxoSmithKline, and FDA staffers all had seats inside a chained off enclave.

And during the meeting Dr. Nissen was under an official gag order that permitted him to speak only when spoken to and then only to answer questions about his meta-analysis. He was specifically forbidden from commenting on any data presented during the meeting.

Before the meeting Dr. Nissen hinted broadly that he thought the drug should be pulled from the market, but in an interview immediately after the daylong session he said he was pleased with the advisers' actions. Noting that the advisers agreed with his finding of excess risk with rosiglitazone, he said the recommended warnings will be helpful "physicians follow those warnings."

The exact nature of the new warnings-if in fact the FDA decides to follow the advisers' recommendations-is, however, not at all clear.

One problem was that the lack of specificity on the part of the committee members. Almost all used the terms black box and boxed warning and some said there should be contraindications on the label.

However, Robert Meyer, M.D., director of new drug evaluation at the FDA's Center for Drug Evaluation and Research, said at a press conference that it wasn't clear to him that the advisors were recommending black box warnings.

In any case, Dr. Meyer pointed out that "while FDA might follow the advice given by advisory committees" the exact actions-and the timing of those actions-are up to the FDA.