Actos, Avandia and the power of “routine medical practice”

Executive Summary

Elderly diabetics treated with GlaxoSmithKline's Avandia (rosiglitazone) have a 15 percent greater mortality risk than those using Takeda's Actos (pioglitazone), according to an observational study in the Nov. 24 Archives of Internal Medicine. Brigham and Women's Hospital's Wolfgang Winkelmayer, et al., reviewed Medicare Part A and B files and claims data on 28,361 New Jersey and Pennsylvania residents over age 65 prescribed Avandia or Actos. No differences in the risk of myocardial infarction and stroke were detected, however, and GSK noted in a statement that the study "may be biased due to imbalances in comorbid conditions ... between the two treatment groups." The firm said it stands by its data from RECORD, an ongoing long-term randomized trial that shows no statistically significant differences in cardiovascular deaths between those taking Avandia and those on placebo. Physicians and patients should wait for final RECORD data, available within a year, before making final conclusions about CV risks of the drug, GSK said. Avandia scripts have plummeted since a meta-analysis showed increased risks, and GSK is hoping that a gold-standard trial will restore some glow to the product. But Jan Vandenbroucke, Leiden University Medical Center, and Bruce Psaty, University of Washington, suggest there is no need to delay passing judgment. In a commentary on the Avandia study published Nov. 26 in the Journal of the American Medical Association, Vandenbroucke and Psaty argue that while randomized controlled trials are best for elucidating a drug's efficacy, "in contrast, data from routine medical practice may very well be used to investigate adverse effects of drugs.