Journal Advises Calm in Rosiglitazone (Avandia) Uproar

LONDON, May 24 -- The editors of The Lancet have cautioned against a rush to judgment in the face of a meta-analysis released Monday that linked the diabetes drug rosiglitazone (Avandia) to a 43% increase in relative risk of myocardial infarction.

In an unsigned editorial posted online Wednesday, The Lancet agreed that results of a meta-analysis published by the New England Journal of Medicine "certainly raise a signal of concern" that should be pursued, but it urged a wait-and-see approach.

Charactering the tone of the NEJM paper as "one of urgency," the British journal asserted, "To avoid unnecessary panic among patients, a calmer and more considered approach to the safety of rosiglitazone is needed. Alarmist headlines and confident declarations help nobody."

The FDA, physicians, and patients "can reasonably await the results of RECORD, a phase III trial designed specifically to study cardiovascular outcomes," said the editorial.

Until those data are available, "it would be premature to over-interpret a meta-analysis that the authors and NEJM editorialists all acknowledge contains important weaknesses," said The Lancet.

The odds ratio for MI was 1.43 (95% confidence interval 1.03-1.98, P=0.03), said Steven E. Nissen, M.D., of the Cleveland Clinic, lead author of the meta-analysis.

But the absolute number of rosiglitazone-associated events was small -- 86 MIs in the rosiglitazone patients versus 72 in the controls and 39 cardiovascular deaths versus 22 cardiovascular deaths in the control arm -- so even a slight change could tip the balance toward a null finding.

Nonetheless, Dr. Nissen said in an interview that he was confident that the cardiovascular risk was real and that the signal was lurking from the very beginning. There were, he said, an excess number of cardiovascular events in the registry data given to the FDA during the initial approval process. "Although those events did not reach statistical significance, the data were all going the wrong way."

Dr. Nissen concluded, "If I had been a member of the FDA advisory panel that reviewed this drug in 1999, I would have voted against approval and would have asked for more studies to assess cardiovascular risk."

According to a report in today's New York Times at least one other physician identified a risk early on -- John R. Buse, M.D., of the University of North Carolina. He wrote to the FDA in 2000 to inform the agency of "a worrisome trend in cardiovascular deaths and adverse events."

Dr. Buse, who is president-elect of the American Diabetes Association, told the newspaper that his opinion of rosiglitazone has not changed since 2000.

Rep. Henry Waxman (D-Calif.) will conduct a hearing into the FDA's oversight of the safety of rosiglitazone. Dr. Buse, along with Dr. Nissen, are expected to be star witnesses at that hearing, which is scheduled for June 6.

The weaknesses in the meta-analysis involve the data, which were obtained from the FDA, a clinical trials registry website maintained by GlaxoSmithKline, which markets rosiglitazone, and published trials. There were no patient-level data and no time-to-event studies.

At a press conference on Monday, Robert J. Meyer, M.D., of the FDA's Center for Drug Evaluation and Research, said the agency had reviewed data that contradicted the findings of the Dr. Nissen's analysis. Yet he conceded that early findings from the FDA's own meta-analysis also found an increased cardiovascular risk.

GlaxoSmithKline was also critical of the Nissen findings. Nonetheless, it acknowledged that a meta-analysis conducted by its own researchers had identified an increased cardiovascular risk, although not as great as the risk reported in the NEJM.