Drug-Maker Defends Rosiglitazone in Letter to The Lancet

KING OF PRUSSIA, Pa., May 30 -- Despite reports linking rosiglitazone (Avandia) to increased risk of cardiac events, the data safety monitoring boards assessing that risk in three studies of the drug agreed that the trials should continue.

Ronald L. Krall, M.D., chief medical officer of GlaxoSmithKline, maker of rosiglitazone, disclosed the decisions of the safety panels in a letter released online today by The Lancet.

The disclosure that all three trials -- RECORD, BARI 2D, and ACCORD -- will continue as planned could be construed as the most reassuring news about rosiglitazone since May 21 when the New England Journal of Medicine published online a meta-analysis that found a 43% increase in relative risk of myocardial infarction for patients using rosiglitazone.

Or, it may simply indicate the studies have reached their pre-specified safety stopping points.

Dr. Krall said the relative safety of rosiglitazone could also be gleaned from a balanced cohort observational study of 33,363 patients in a managed care database who began treatment with oral antidiabetic drugs from 2000 to 2004. That study, commissioned by GSK, assessed the risk of MI, coronary revascularization or both among patients taking rosiglitazone, metformin, or sulfonylurea monotherapy, or combinations of those drugs.

"The incidence of composite cardiovascular endpoint was 1.75 events per 100 patient-years for rosiglitazone-containing regimen and 1.76 events per 100 patient-years for the non-rosiglitazone-containing regiment (hazard ratio 0.93, 95% CI 0.80-1.10), he wrote.

Dr. Krall concluded that the company believed "these studies provide clear evidence of the cardiovascular safety of rosiglitazone and that the estimates of cardiovascular morbidity from the meta-analyses are not robust."

The system for drug approval and safety monitoring of drugs, he said, "is working," said Dr. Krall.

The meta-analysis in the NEJM was conducted by Steven Nissen, M.D., of the Cleveland Clinic, and colleagues.

That GSK statement will be closely scrutinized on June 6 when Rep. Henry Waxman (D-Calif.) opens a congressional hearing into FDA oversight of rosiglitazone from its initial approval in 1999 through post-marketing surveillance. Dr. Nissen will testify.

The FDA has already confirmed that its own ongoing meta-analysis of rosiglitazone studies will also show an increased cardiac event rate for patients using rosiglitazone. But the agency has also maintained that its review of preliminary data from RECORD (Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of glycaemia in Diabetes), provided evidence that contradicted the meta-analyses.

A week ago, The Lancet published an unsigned editorial-also released online-urging calm until results of prospective trials, such as RECORD are available.

Dr. Krall, who said he was responding to The Lancet editorial, conceded that the company's own meta-analyses conducted in 2005 and 2006 found hazard ratios "in the same direction" as Dr. Nissen's meta-analysis.

But he pointed out that there was "a very low frequency of events (0.6%), and the absolute difference in rates of myocardial infarctions between rosiglitazone and controls is less than 0.1%."

Dr. Krall said data from ADOPT (A Diabetes Outcomes Progression Trial) and DREAM (Diabetes Reduction Assessment with ramipril and rosiglitazone Medication) offer more reliable evidence of rosiglitazone safety than findings from meta-analyses.

In support of that assertion, he presented data from a GSK post hoc analysis of the ADOPT database. "Our analysis, which adjusted for medication exposure, found that such events were rare in this population and that all treatments were comparable," he wrote.

ADOPT compared rosiglitazone to metformin or glyburide. The hazard ratios "varied from 0.58 to 1.52 and 95% CIs for all comparisons included unity," Dr. Krall wrote.

Initial published findings from the DREAM trial, which used a 2 x2 design that randomized patients to rosiglitazone plus placebo, rosiglitazone plus ramipril, ramipril plus placebo or placebo-placebo found no significant difference in cardiovascular endpoints. But a post hoc cell-level intention-to-treat analysis found an increase in cardiovascular events among patients in the rosiglitazone-ramipril arm, a finding that Dr. Krall said "is currently unexplained."