LAUSR

AIM We performed a meta-analysis of epidemiological studies evaluating exposure to pioglitazone and the risk for bladder cancer and compared these results to the drug's effects on cardiovascular disease (CVD) and non-alcoholic steatohepatitis (NASH). METHODS Pubmed and Embase were searched for cohort and case control studies for all years through 2016. RESULTS Data from 17 papers were analyzed. In cohort studies, 822 of 357,888 pioglitazone-exposed patients (0.23%) developed bladder cancer while 7691 of 2,898,682 unexposed (0.26%) did. In case control studies, 3219 of 1,146,916 patients (0.28%) developed bladder cancer. A random effects model showed no significant association between ever vs never use or with cumulative doses of pioglitazone. However, there was a significant association with 1-2 years (HR = 1.28 [1.08-1.55]) and >2 years (HR = 1.42 [1.14-1.77]) of exposure. The numbers needed to treat for one additional case of bladder cancer ranged from 899 to 6380 while to benefit CVD and NASH, 4-256 and 2-12, respectively. CONCLUSIONS Given the very small prevalence of bladder cancer in diabetic patients exposed (or not) to pioglitazone (<0.3%) and the much greater beneficial effects of the drug on CVD and NASH, the use of pioglitazone should be resurrected.