By DrRich
In the May 21 issue of the New England Journal of Medicine, Cleveland Clinic cardiologist Steven Nissen has published a meta-analysis that combined the results from 42 clinical trials using Avandia (rosiglitazone, GlaxoSmithKline) in the treatment of diabetes. His meta-analysis suggests that the use of Avandia is associated with a 43 percent higher risk of heart attack (myocardial infarction), and a 64 percent higher risk of cardiovascular death, than patients treated with other methods. This is a striking report, especially because it is thought that hundreds of thousands of diabetics have received prescriptions for Avandia.
According to the May 21 Wall Street Journal, the Food and Drug Administration has not yet formulated a response to this new data. Spokesmen for GlaxoSmithKline stress that ongoing studies may show Dr.Nissen's analysis to be incorrect, and that the risk profile of Avandia has not yet been fully resolved.
Avandia is one of several drugs that have been developed from the thiazolidinedione group. These drugs work by stimulating a number of genes that, among other things, modify glucose and lipid metabolism. Avandia and Actos (pioglitazone, Takeda) are the only two thiazolidinediones currently marketed for diabetes. They reduce blood glucose levels by increasing sensitivity to insulin. A third drug from this class, Rezulin, was removed from the market because of liver toxicity. Avandia is known to cause edema, weight gain, and an increase in LDL cholesterol, and doctors have been warned not to use it in patients with heart failure. But until now there have been few allegations that the drug could increase the risk of heart attacks or cardiovascular death. Indeed, some studies suggested that Avandia could reduce the rate of atherosclerosis (hardening of the arteries).
DrRich Comments:
There are inherent problems with conducting any meta-analyses, and these problems can lead to erroneous results. So Dr. Nissen's study cannot be considered definitive, and it will be difficult for anyone, including the FDA, to give clear-cut guidance based on this study. Strong opinions will doubtless be expressed on both sides of the issue, and the scientifically level-headed will doubtless point out that the question of Avandia's safety will finally be resolved only with new analyses of existing data, and with the genesis of new data (i.e., with more studies). And this will just take time.
Unfortunately, many doctors and patients need to decide what to do right now. There are 4 reasons this new report should not be ignored.
1) Thiazolidinediones are drugs that have a host of possible effects, and not all of them are good. At least one prior drug from this class (Pargluva) was shown to increase heart attacks, and the drug was never approved. Further, of more than 50 drugs from this class that have been developed, the vast majority never reached market due to various toxicities discovered in initial testing; information as to the types of toxicities seen with these drugs has not been revealed. It remains possible, therefore, that increased heart attacks may be a "class effect" with thiazolidinediones.
2) Dr. Nissen, while surely not a popular man today with segments of the drug industry, has a track record of being right when he calls attention to previously unnoticed drug toxicities. He was one of the doctors who called early attention to the cardiovascular toxicities associated with Vioxx and Pargluva, for instance. To the extent that conducting meta-analyses is an art, he has shown in the past that he is a pretty good artist.
3) Avandia, like most diabetic drugs, was approved based on short-term studies demonstrating its effectiveness in controlling blood glucose levels. It remains highly effective at doing that. But none of the studies done with this drug were aimed specifically at looking for cardiovascular side effects (or benefits, for that matter). If Avandia does cause cardiac problems, none of the studies used for its approval would be expected to have discovered these problems. (In the meta-analysis, data from all these studies and several others were pooled.)
4) Many other options are available for patients with diabetes. Few diabetics are truly dependent on Avandia for good diabetic control.
Taking all of these reasons into account, it would seem advisable for patients taking Avandia to talk to their doctors about whether they really ought to continue the drug while waiting for the FDA to make a formal recommendation, or for the scientifically adept to go through the tedious process of figuring out whether it really is safe.
Sources:
Nissen, SE. Wolski K. Effect of Rosiglitazone on the Risk of Myocardial Infarction and Death from Cardiovascular Causes. New England Journal of Medicine, on-line. May 21, 2007. http://content.nejm.org/cgi/content/full/NEJMoa072761#T5
Matthews, AW. Analysis of Avandia Studies Finds Heart-Attack Risks. Wall Street Journal, May 21, 2007. http://online.wsj.com/article/SB117976261302809593.html?mod=home_whats_news_us
