New Avandia Study Doesn't Prove Safety

Editorial Says Avandia ‘Potentially Dangerous’; GSK Says Study ‘Reassuring’

Medically Reviewed by Louise Chang, MD on June 05, 2007
From the WebMD Archives

June 5, 2007 - Preliminary data from a heart-safety study of Avandia -- released on the eve of a Congressional hearing into the matter -- do not settle questions raised about the diabetes drug's possible heart attack risks.

Avandia maker GlaxoSmithKline says the study findings are "reassuring." But three new editorials from TheNew England Journal of Medicine suggest that doctors and patients will not be reassured.

The GSK-funded RECORD study, managed by an independent steering committee, enrolled some 4,500 type 2 diabetes patients from Europe, Asia, and Australia. All the patients take the standard diabetes drugs metformin or sulfonylurea. Half added Avandia to their treatment, and half received a combination of metformin and sulfonylurea.

The study is supposed to continue until most patients have been treated for six years. However, a recent NEJM study that combined all available data suggested that Avandia might raise a person's risk of heart attack by 42%. In response to the resulting furor, the study authors -- and the NEJM'seditors -- decided to release data collected after patients had completed only 3.75 years of the study.

The study's findings, as reported by Newcastle University researcher Philip D. Home, DPhil, and colleagues:

  • There is no evidence that Avandia increases death from all causes of heart disease or death from all causes. Avandia patients did slightly worse than other patients, but the difference is not statistically significant, meaning it could be due to chance.
  • Avandia more than doubles a person's risk of heart failure, although that risk is still low. This result was expected, as the class of drugs to which Avandia belongs has this effect.
  • The study data are "insufficient" to determine whether the drug increases a person's risk of heart attack. The data "are consistent with as much as a 19% improvement, and as much as an 86% worsening, in risk," Home and colleagues report.

In a telephone news conference held to discuss the findings, GSK chief medical officer Ronald Krall, MD, said the study findings exonerate Avandia.

"These results are reassuring to doctors and to patients, both those with diabetes and those taking Avandia," Krall said. "These results add to the weight of evidence ... showing that Avandia is comparable in safety to the standard diabetes drugs."

But editorials published alongside the study paint a far different picture.

The study has several flaws, notes David M. Nathan, MD, director of the diabetes center at Massachusetts General Hospital and professor of medicine at Harvard Medical School in Boston.

"The interim results of the RECORD trial do not provide any assurance of the safety of treatment with [Avandia]," Nathan writes. "Physicians may find it difficult to explain to patients why they are starting treatment with a potentially dangerous drug when other choices with longer and better safety records are available."

The other editorial is by Bruce M. Psaty, MD, PhD, of the University of Washington in Seattle and Curt D. Furberg, MD, PhD, of Wake Forest University in Winston-Salem, N.C. Psaty and Furberg's previous editorial on Avandia questioned the FDA's wisdom in approving the drug.

Now Psaty and Furberg say the very low number of heart disease events in the RECORD trial is suspicious. Coupled with the study's high dropout rate, they suggest this leaves the study without enough statistical power to achieve its goal of proving that Avandia is not inferior in heart safety to standard diabetes drugs.

Moreover, they suggest that when the study findings are combined with data from previous studies, Avandia is still associated with significant risk of heart attack.

"The level of risk, a hazard ratio of 1.33 (a 33% increase in heart attack risk), is substantial and approximately equivalent in magnitude -- but in the opposite direction - to the health benefits of lipid-lowering statin drugs," Psaty and Furberg write.

And an editorial by NEJM Editor Jeffrey M. Drazen, MD, and colleagues, states "even a small increase in cardiovascular risk in a fragile population of patients with type 2 diabetes is of considerable concern.

"There is continued uncertainty about the cardiovascular safety of [Avandia]," Drazen and colleagues conclude.

Krall and Glaxo Vice President of Clinical Development Murray Stewart, FRCP, say that these conclusions distort the RECORD findings.

"It would be very unfortunate if doctors tried to make decisions based on these very small differences between Avandia and standard diabetes drugs," Krall said. "The best interpretation of these findings would be that there are no differences between Avandia and sulfonylurea with metformin."

"I am actually pleased the numbers are small," Stewart said. "This shows that good diabetes care is making a difference. People in this study kept their glucose down. This is very reassuring data."

  • Does Avandia increase heart attack risk? Our expert is not yet convinced. Read nurse Laurie Anderson’s blog to find out why.

Show Sources

SOURCES: Home, P.D. The New England Journal of Medicine, published online, June 5, 2007. Psaty, B.M. and Furberg, C.D. The New England Journal of Medicine, published online, June 5, 2007. Nathan, D.M. The New England Journal of Medicine, published online, June 5, 2007. Drazen, J.M. The New England Journal of Medicine, published online, June 5, 2007. News conference with Ronald Krall, MD, chief medical officer; and Murray Stewart, FRCP, vice president of clinical development, GlaxoSmithKline.

© 2007 WebMD, Inc. All rights reserved. View privacy policy and trust info