The US Food and Drug Administration has asked the makers of two antidiabetes drugs—rosiglitazone (marketed as Avandia), made by GlaxoSmithKline, and pioglitazone (Actos), made by Takeda—to place “black box” warnings, the most serious kind, on their labels.
The new labels warn of an increased risk of congestive heart failure, because rosiglitazone and related drugs can cause fluid retention. Andrew von Eschenbach, the FDA's commissioner, announced the warning at a hearing of the US House of Representatives' Committee on Oversight and Government Reform last week to examine the FDA's role in evaluating the safety of rosiglitazone.
The new labels do not address the question of whether these drugs pose an increased risk of heart attacks and strokes.
The cardiovascular risk was raised last month by an article and accompanying editorial in the New England Journal of Medicine (doi: 10.1056/NEJMoa072761). Steven Nissen and Kathy Wolski of the Cleveland Clinic did a meta-analysis of publicly available data and concluded that rosiglitazone was associated with a significant 43% increase in the risk of heart attack (BMJ 2007;334:1073, 26 May, doi: 10.1136/bmj.39224.364630.DB).
On 5 June the New England Journal of Medicine published three online editorials related to “continued uncertainty” about rosiglitazone, together with an interim analysis of the cardiovascular outcomes associated with rosiglitazone (doi: 10.1056/NEJMoa073394). These will appear in print in the 5 July issue.
The interim analysis, by Philip Home of the Newcastle Diabetes Centre and Newcastle University, Newcastle upon Tyne, and colleagues, covers just under four years of the planned six year study. They wrote that the results were “inconclusive regarding . . . risk of hospitalization or death from cardiovascular causes.” They found a small but not significant increase in the risk of hospitalisation and death from cardiovascular causes with use of the drug (hazard ratio 1.1 (95% confidence interval 0.9 to 1.3)). The hazard ratio for fatal or non-fatal myocardial infarctions was 1.2 (0.8 to 1.8)).
In one of the editorials Bruce Psaty of the University of Washington in Seattle and Curt Furberg of Wake Forest University in North Carolina said the trial had weaknesses: doctors and patients knew which drugs they were getting, the rate of events in a high risk population was exceptionally low (perhaps because of poor reporting), and a high number of patients were lost to follow-up.
At the House of Representatives committee hearing, the chairman, Henry Waxman (Democrat, California), asked why the FDA had not done post-marketing studies concerning the possible cardiovascular risks of rosiglitazone. He noted that concerns about the cardiovascular risks of the drug had been expressed since its approval in 1999 and that the FDA's medical reviewer had called for a post-marketing study to examine possible “deleterious effects on the heart.”
“Unfortunately, at that point, FDA dropped the ball,” Mr Waxman said. He said a post-marketing study was designed to show whether the drug provided long term control of blood sugar concentrations—which it did—but this study did not assess whether the drug increased the risk of heart attacks.
He said there had not been a good post-marketing study of the drug's safety, despite many calls for studies of rosiglitazone's possible cardiovascular risks. In addition, GlaxoSmithKline had sent the FDA a memo saying that an internal company analysis showed that the drug might be associated with an increased risk of “myocardial ischemia,” although that was accompanied by a second study that did not show a raised risk.
The FDA “never required the manufacturer to conduct a thorough post-market study,” Mr Waxman said. “Avandia is a case study of the need for reform of the nation's drug safety laws.”
John Buse, of the University of North Carolina, and the incoming president of the American Diabetes Association, said that SmithKlineBeecham (now part of GlaxoSmithKline) had tried to intimidate him when he spoke out with his concerns about rosiglitazone's cardiovascular safety.
Dr Buse said in his prepared testimony that he had spoken at least twice in June 1999 about “a trend toward increases in serious cardiovascular events and cardiovascular deaths with Avandia as compared to active comparators. Neither was statistically significant. I could not find such evidence with Rezulin [Parke-Davis's troglitazone] and Actos.”
He said that employees of SmithKlineBeecham had told him in telephone calls that “there were some in the company who felt that my actions were scurrilous enough to attempt to hold me liable for a loss in market capitalisation [share value] . . . In the end I offered to help the company with further studies and signed a clarifying statement drafted by SKB which was to be used with the investment community.”
Dr Buse said, “My concern [in March 2000] was that the entire glitazone class was in danger if Rezulin was withdrawn from the market without robustly understanding the safety of the newer agents. At that time, about half the patients with diabetes in my practice were still inadequately controlled. What I needed was more ways to treat diabetes, not fewer.”
Dr von Eschenbach said the manufacturer had submitted a pooled analysis showing that “the overall incidence of myocardial ischaemia in rosiglitazone treated subjects relative to the comparators was 1.99% versus 1.51% with a hazard ratio of 1.31% (95% confidence interval 1.01 to 1.70).
“This risk equates to a more than 30% excess risk of myocardial ischaemic events in rosiglitazone treated patients . . . It does not mean that diabetics taking the drug have a 30% risk of having a heart attack.”
GlaxoSmithKline, the manufacturer, strongly defended the safety of the drug at the hearing, in press releases, and on its website (www.gsk.com). It said that the editorials in the New England Journal of Medicine “are selective in their use of data to support a biased view and therefore do a disservice to patients, physicians, science, and public health.”
Moncef Slaoui, chairman of research and development at GlaxoSmithKline, said that the safety of rosiglitazone was similar to that of other oral antidiabetes drugs.
Dr von Eschenbach said, “The data are inconsistent, and conclusions are not clear.” He said the FDA would convene an advisory committee to review all the data it has received from manufacturers and make recommendations about whether further regulatory action is needed.
See Editorial doi: 10.1136/bmj.39245.502546.BE.