Intended for healthcare professionals


FDA panel advises easing restrictions on rosiglitazone

BMJ 2013; 346 doi: (Published 10 June 2013) Cite this as: BMJ 2013;346:f3769
  1. Miriam E Tucker
  1. 1Bethesda, Maryland

A joint Food and Drug Administration advisory panel has voted to ease the current stringent restriction on prescription of the antidiabetes drug rosiglitazone (marketed by GlaxoSmithKline as Avandia), despite lingering concerns about its cardiovascular safety.

The three year old restrictions, formally known as a “risk evaluation and mitigation strategy (REMS) with elements to assure safe use (ETASU) program,” called for certification of the physician and the dispensing pharmacy and for medication guides for patients and physicians.1 They also limited the drug’s use to patients who were already taking it before the restrictions were placed and to patients for whom no other glucose lowering agents are acceptable.

The vote took place on the second day of a two day joint meeting of the FDA’s endocrinologic and metabolic drugs advisory committee and its drug safety and risk management advisory committee. The vote was split, with just half of the total of 26 members opting to keep the drug on the market with a modification of the restrictions. Other panelists preferred either that the restrictions be lifted entirely (seven members), that they be kept in place without change (five), or that the drug be taken off the US market (one).

Concern had been raised about a signal for increased risk of myocardial infarction with rosiglitazone after the publication of a meta-analysis of 42 randomized trials by Steven E Nissen and Kathy Wolski in 2007 and their 2010 update involving 56 trials.2 3

None of the trials included in those meta-analyses was specifically designed to evaluate cardiovascular outcomes. The only one to do so was GlaxoSmithKline’s 4447 patient rosiglitazone evaluated for cardiovascular outcomes and regulation of glycemia in diabetes (RECORD) trial, which found no overall increased risk of cardiovascular events with rosiglitazone in comparison with a combination of sulfonylurea and metformin.4

However, concerns were raised at a 2010 FDA advisory committee hearing regarding the reliability and interpretability of the RECORD trial. At the time that the FDA imposed its restrictions on rosiglitazone it also called for an independent readjudication of RECORD.

That readjudication, conducted by the Duke Clinical Research Institute, supported the conclusion of the original RECORD trial. The institute’s associate director, Kenneth Mahaffey, reported at the current hearing that none of the hazard ratios in the readjudication—for cardiovascular death, myocardial infarction, or stroke; cardiovascular death; myocardial infarction; stroke; and all causes of death—were significant.

Panel members said that they were somewhat reassured by the readjudication, but they also expressed concern that the trial’s open label design, its use of active comparators, and its many missing data points could be masking risks of the drug. Several panelists called for a new trial that would ideally be placebo controlled, larger, and better monitored for outcomes, but others doubted that such a trial would be feasible or ethical.

One panel member, Marvin A Konstam, chief physician executive at the cardiovascular center at Tufts Medical Center, Boston, and professor of medicine at Tufts University, Boston, suggested that the FDA consider eliminating or softening the restrictions but keeping the medication guides.

He said, “The difference between now and 2010 is the readjudication of RECORD. Those results do not remove my concern about the cardiovascular safety of rosiglitazone, but in my mind they move the needle . . . to shift the burden of decision making for prescribing this drug to the physician.”

But panel member Maria Suarez-Almazor, a professor at the University of Texas MD Anderson Cancer Center, voted to continue the current restrictions. She said, “I felt there were consistent signals . . . although they didn’t achieve statistical significance. I don’t see clear benefit of this drug over the other in the class [pioglitazone].”

She added, “There might be a small proportion of patients for whom this drug may be necessary . . . I think the REMS has been successful in restricting its use.”

William R Hiatt, professor of medicine in the division of cardiology at the University of Colorado, Aurora, voted to remove the restrictions entirely. He said, “In general this drug doesn’t look any different than any other diabetes drug in terms of its cardiac risk. That said, it seems to me that if we want to move forward and learn something, the majority voting to retain a REMS program essentially negates the opportunity for a randomized controlled clinical trial, because I think it affects the perception of the ethics of doing such a trial.”

No timetable has been set for the FDA’s decision.


Cite this as: BMJ 2013;346:f3769


  • Feature: Rosiglitazone: what went wrong? (BMJ 2010;341:c4848, doi:10.1136/bmj.f4848); News: GSK fights compensation claims from UK patients who took rosiglitazone (BMJ 2013;346:f695, doi:10.1136/bmj.f695)