RECORD Interim Analysis of Rosiglitazone Safety: No Clear-Cut Answers

Shelley Wood

June 06, 2007

June 6, 2007 (Boston, MA) - One of the authors of an interim analysis from the Rosiglitazone Evaluated for Cardiac Outcomes and Regulation of Glycemia in Diabetes (RECORD) study [ 1], showing no statistically significant differences in the overall risk of hospitalization or death from cardiovascular causes, says that the findings should be "moderately reassuring" for physicians prescribing rosiglitazone (Avandia, GlaxoSmithKline). Other experts who wrote editorials accompanying the publication of the study take a much dimmer view.

"The jury may still be out with regard to the cardiotoxicity of rosiglitazone, but when it comes to patient safety, 'first do no harm' should outweigh any presumption of innocence," Dr David M Nathan (Massachusetts General Hospital, Boston) writes in one of three accompanying editorials [ 2].

Likewise, Drs Bruce M Psaty (University of Washington, Seattle) and Curt D Furberg (Wake Forest University, Winston-Salem, NC) observe [ 3]: "Glycemic control and durability appear to be the major benefits. Rosiglitazone is also associated with significant weight gain, an adverse effect on LDL cholesterol, an increased risk of heart failure, an increased risk of fractures in women, and an apparent increase in the risk of MI. Patients should not stop treatment on their own, but if they have concerns they should consult their physicians. Together, patients and physicians can decide whether they wish to suspend the use of rosiglitazone."

But RECORD investigator Dr Stuart Pocock (London School of Hygiene and Tropical Medicine, UK), who is also an author on the interim RECORD analysis, points to the lack of difference in overall cardiovascular risk and no increased deaths.

"The overall risk in the rosiglitazone group is close to what you get in the control group, and therefore it's moderately reassuring--but not conclusive--on the subject of overall cardiovascular risk," he told heart wire . "On death and CV death, there are actually fewer patients dying in the rosiglitazone group."

The mortality numbers are not statistically significant, but according to Pocock, "they're on the right side of the fence and therefore reassuring on the subject of whether there is an excess mortality with rosiglitazone, and that's ultimately the bottom line."

RECORD numbers

The RECORD analysis and accompanying editorials are published online June 5, 2007 in the New England Journal of Medicine. The study, led by Dr Philip D Home (Newcastle University, Newcastle upon Tyne, UK), is an "unplanned interim analysis" prompted by the meta-analysis by Dr Steven Nissen and Kathy Wolski (Cleveland Clinic, OH) [ 4]--also accompanied by a Psaty/Furberg editorial [ 5] and published in the Journal last month. That meta-analysis pointed to an increase in MI and cardiovascular death with rosiglitazone and, as reported by heart wire , produced a media tsunami: GlaxoSmithKline shares plummeted, as did prescriptions for the drug. Subsequently, the chair of the US Congress Committee on Oversight and Government Reform, Rep Henry A Waxman (D-CA), has announced that the committee would be holding a hearing to review the FDA's role in evaluating the safety of rosiglitazone, to take place Wednesday morning, while the FDA, meanwhile, now plans to review the safety of the drug at a meeting of advisory panel members from its committees on endocrine and metabolic drugs and on drug safety and risk management, scheduled for July 30.

Full results from RECORD trial, launched in 2000 and designed to specifically address the cardiovascular safety of rosiglitazone, willnot be available until 2009. As such, the interim analysis published today only includes 3.75 years of follow-up and, in the words of the authors, has "limited statistical power to detect treatment differences."

In all, 217 diabetic patients in the rosiglitazone arm of the study and 202 patients in the control group--taking metformin or sulfonylurea--experienced the adjudicated primary end point of hospitalization or death from cardiovascular causes: a statistically nonsignificant difference. In an analysis including end points that have not yet been adjudicated, the hazard ratio for hospitalizations or death from CV causes was 1.11 (p=0.26). In an analysis of adjudicated events only and in an analysis that combined adjudicated and pending-adjudication events, there were no statistically significant differences in rates of death (cardiovascular or all-cause), acute MI, or a composite of death from cardiovascular causes, MI, and stroke. In both analyses, however, patients taking rosiglitazone were more than twice as likely to develop congestive heart failure, a statistically significant finding.

The HF findings are "consistent" with previous studies of thiazolidinediones, the authors note, but "of concern," warranting warnings to patients with heart failure. By contrast, the MI findings pointing to an increased risk with rosiglitazone ranging from 16% to 23% "are somewhat lower than those reported in the meta-analysis by Nissen and Wolski," Home et al write. They acknowledge that the study lost statistical power due in part to withdrawal of patients from the trial, higher-than-expected numbers of patients who were lost to follow-up, and unexpectedly low rates of the primary end point in both groups.

To heart wire , Pocock commented: "We felt we had to release this interim information to add quality and quantity" to the ongoing rosiglitazone debate. For MI, he added, "the data are consistent both with no effect and with the Nissen findings--ie, the observed hazard ratio is somewhere between the two."

Editorialists unconvinced

The editorials all take a decidedly less sanguine view.

Psaty and Furberg, as well as Nathan, point to a number of the RECORD trial's failings, among them its unblinded randomization, the very low rate of events among such a high-risk diabetic population, and the choice of a "weak" composite end point.

"This interim analysis fails to provide exculpatory evidence," Nathan concludes.

Yet a third editorial [ 6], this one from Journal editors Drs Jeffrey M Drazen, Stephen Morrissey, and Gregory D Curfman, underscores the difficulty of determining drug safety and makes a plug for "the need for open and honest disclosure."

Their tone may in part reflect a response to criticisms of the Journal, including those of former FDA deputy commissioner Dr Scott Gottlieb, that have appeared in some of the media coverage of the rosiglitazone controversy.

The inconclusive interim results and gloomy editorials will no doubt add fat to the fire in the heated media coverage of rosiglitazone's fall from grace. Scrutiny now turns to the congressional committee hearing tomorrow. Contacted by heart wire , Nissen, who has been invited to participate in the proceedings, said only that he thought the three editorials were "excellent."

According to disclosures cited in the Journal , Home donates all drug company consulting and lecture fees to institutions he works with; Pocock receives consulting fees and grant support from GlaxoSmithKline; and Nathan receives lecture fees from Pfizer, GlaxoSmithKline, and Novartis.

  1. Home PD, Pocock SJ, Beck-Nielsen H, et al. Rosiglitazone evaluated for cardiovascular outcomes--an interim analysis. N Engl J Med 2007; DOI: 0.1056/NEJMoa073394. Available at:

  2. Nathan DM. Rosiglitazone and cardiotoxicity: weighing the evidence. N Engl J Med 2007; DOI: 10.1056/NEJMe078117. Available at:

  3. Psaty BM, Furberg CD. The record on rosiglitazone and the risk of myocardial infarction. N Engl J Med 2007; DOI: 10.1056/ NEJMe078116. Available at:

  4. Nissen SE and Wolski K. Effect of rosiglitazone on the risk of myocardial infarction and death from cardiovascular causes. N Engl J Med 2007; DOI:10.1056/NEJMoa072761. Available at:

  5. Psaty B and Furberg C. Rosiglitazone and cardiovascular risk. N Engl J Med 2007;DOI 10.1056/NEJMe078099. Available at:

  6. Drazen JM, Morrissey S, Curfman GD. Rosiglitazone--continued uncertainty about safety. N Engl J Med 2007; DOI: 10.1056/NEJMe078118. Available at:

The complete contents of Heart wire , a professional news service of WebMD, can be found at, a Web site for cardiovascular healthcare professionals.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: