Rising experimental evidence suggests that thiazolidinediones (TZDs) exert a protective effect on breast cancer. However, studies concerning this issue were inconsistent and limited. Hence, we performed a meta-analysis with data from currently available studies to evaluate the effect of TZDs on breast cancer risk among diabetic women. We comprehensively searched for all pertinent studies addressing TZDs use and breast cancer risk published before January 1, 2016, in PubMed, Clinical Trials, and Cochrane Library. Data synthesis was performed in a random-effects model using Stata version 12.0 (Stata Corp, College Station, Texas). Fourteen independent studies were eventually selected in this meta-analysis, including 5 randomized controlled clinical trials (RCTs), 7 cohort studies, and 2 case-control studies. No significant associations of TZD use and risk of breast cancer were observed in the RCTs (pooled risk ratio [RR]: 0.77, 95% confidence interval (CI), 0.39-1.53, I2 = 26%) or case-control studies (pooled odds ratio, 0.99, 95% CI, 0.76-1.28, I2 = 31%). A 19% reduction in breast cancer risk (pooled RR: 0.81, 95% CI, 0.66-0.99, I2 = 72%) was found in the cohort studies. However, after removing the study with the smallest event number and the greatest effect size, the association became nonsignificant with greatly decreased heterogeneity (pooled RR: 0.94, 95% CI, 0.86-1.03, I2 = 16%). This meta-analysis did not find any significant association between TZDs use and risk of breast cancer among diabetic women.
Keywords: TZDs; antidiabetic drugs; breast cancer; diabetes; meta-analysis; systematic review.
Copyright © 2017 John Wiley & Sons, Ltd.