Winners and losers at the rosiglitazone gamble: A meta-analytical approach at the definition of the cardiovascular risk profile of rosiglitazone

The treatment with rosiglitazone could be associated with increased risk for myocardial infarction (MI). This meta-analysis is aimed at identifying moderators of the effect of rosiglitazone on the risk of MI and chronic heart failure (CHF) in type 2 diabetic patients. The risk ratio (RR) of MI and CHF was calculated for each trial as the ratio of incidence density in rosiglitazone and comparator groups. A total of 86 trials were included. After adjusting for trial duration, RR for MI showed a significant inverse correlation with mean baseline HbA1c, triglycerides, and LDL-cholesterol (r = −0.24, −0.45, and −0.33, respectively; all p < 0.05). Conversely, rosiglitazone-associated risk of MI was increased in trials with higher mean BMI or greater proportion of insulin-treated patients (r = 0.26 and 0.42, respectively; p < 0.05). Lower triglyceride levels were also associated with a higher rosiglitazone-induced risk of CHF (r = −0.23, p < 0.05). Treatment with rosiglitazone could have divergent effects on cardiovascular risk, depending on the characteristics of the patients. Benefits could outweigh harms in patients with poor glycemic control and worse lipid profile; conversely, the drug could increase the risk of MI in obese or insulin-treated patients.