Ups and Downs of Thiazolinediones for Diabetes Assessed by Dueling Meta-Analyses

CLEVELAND, Sept. 11 -- Meta-analyses of the two FDA-approved thiazolinediones for diabetes -- rosiglitazone (Avandia) and pioglitazone (Actos) -- left neither drug unscathed.

One-analysis revealed a 42% increased risk of non-fatal myocardial infarction (P=0.02) among patients with type 2 diabetes who used rosiglitazone, The other found that pioglitazone reduced the risk of MI, stroke, or death by 18% (P=0.02).

But both drugs were associated with significant increases in risk of heart failure. For rosiglitazone, the increased risk was more than 200% (P

The pioglitazone study, led by A. Michael Lincoff, M.D. of the Cleveland Clinic, and colleagues, looked at ischemic events and death in 19 randomized trials of pioglitazone compared with standard therapy for glucose control in patients with type 2 diabetes. The rosiglitazone analysis by Sonal Singh, M.D., of Wake Forest in Winston-Salem, N.C., and colleagues, examined data from four long-term studies of rosiglitazone or standard therapy in patients with type 2 diabetes.

The fact that the rosiglitazone analysis included only long-term studies that lasted at least 12 month, gave this paper an edge on the meta-analysis of 42 trials published last May by Steven Nissen, M.D., and Kathy Wolski, M.P.H. of the Cleveland Clinic.

The earlier Nissen-Wolski meta-analysis found an almost identical increased relative risk of non-fatal MI (43% (P=0.03), but the paper was criticized because it was based on short-term trials.

The pioglitazone analysis, meanwhile, was based on individual patient-level data from 19 trials with treatment duration ranging from four months to three and a half years.

GlaxoSmithKline, which makes rosiglitazone, said in a statement that the conclusions drawn from these meta-anlayses do not confirm a difference in the safety profile of rosiglitazone and pioglitazone.

GSK said the analyses, "reflect limitations that are common to all meta-analyses, by the authors' own admission. These analyses do not yield data robust enough to guide doctors in selecting appropriate diabetes treatments for their patients. Comparisons between different meta-analyses with different endpoints and patient populations are even more unreliable."

On the flip side, Robert Spanheimer, M.D., senior director of diabetes and metabolism at Takeda in Deerfield, Ill., which makes pioglitazone, said the meta-analysis "confirms the results of the prospective PROactive study and expands the patient population."

But Dr. Spanheimer cautioned that because 10% reduction in MI and stroke in the PROactive was not statistically significant, the trial did not meet its primary endpoint. "Therefore, we cannot talk about efficacy in reducing macrovascular events."

Nonetheless, Dr. Spanheimer said the results of this meta-anlaysis, along with observational data, and the PROactive results allowed the company to be "confident that [pioglitazone] does not increase the risk of ischemic events."

Rosiglitazone, GSK said,"is the most widely studied oral medication for type 2 diabetes, and is an important option for physicians who often need to prescribe several different diabetes medicines in combination to help their patients maintain blood sugar control."

GSK concluded by noting that the FDA "is engaged in a full, objective analysis of the science, and will make its independent recommendations on the appropriate use of oral anti-diabetic medicines. GSK continues to support Avandia as safe and effective when used appropriately."

Although the FDA is reviewing safety data on pioglitazone, Dr. Spanheimer said he was confident that it would not have additional cardiovascular warnings added to its label.

On the basis of their analysis of the long-term rosiglitazone data, Dr. Singh and colleagues concluded that the potential health benefit of rosiglitazone were related to "a modest mean 1% reduction of hemoglobin A_1c " that should be weighed against "an approximate doubling in risk of heart failure and a 42% increase in MI without any effect on cardiovascular mortality."

The four trials included in the Wake Forest analysis included 6,421 patients taking rosiglitazone and 7,870 controls. The duration of follow-up was one to four years. There were 94 MIs in the rosiglitazone group versus 83 in the control group (RR 1.42, 95% CI 1.06-1.91, P=0.02). Additionally, 102 patients in the rosiglitazone arm developed heart failure versus 62 patients in the control arm (RR, 2.09, 95% CI 1.52-2.88, P