Abstract
PURPOSE: This study investigated prostate cancer risk associated with pioglitazone use. MATERIALS AND METHODS: The Taiwan's National Health Insurance database was used to create a propensity score-matched cohort of male patients with type 2 diabetes mellitus newly diagnosed in 1999-2005 and aged ≥25 years at baseline. The matched cohort included 20437 ever users and 20437 never users of pioglitazone. The patients were followed up for the incidence of prostate cancer until December 31, 2011. Hazard ratios (HRs) were created from Cox regression weighted on propensity score. RESULTS: Prostate cancer was diagnosed in 121 ever users of pioglitazone (incidence: 175.84 per 100,000 person-years) and 143 never users of pioglitazone (incidence: 216.66 per 100,000 person-years). When ever users were compared to never users of pioglitazone, the HR was 0.815 (95% confidence interval [CI], 0.639-1.039; p=0.0987). When ever users were categorized into tertiles of cumulative duration of pioglitazone therapy (<6.83, 6.83-20.23, and >20.23 months), the HRs were 1.044 (95% CI, 0.741-1.471), 0.975 (95% CI, 0.690-1.377) and 0.539 (95% CI, 0.374-0.778), respectively. For the tertiles of cumulative dose of <5,040, 5,040-15,330, and >15,330 mg, the HRs were 1.008 (95% CI, 0.710-1.429), 1.090 (95% CI, 0.785-1.515) and 0.484 (95% CI, 0.330-0.711), respectively. A significantly lower risk associated with pioglitazone use could only be seen in patients aged <65 years (HR, 0.578; 95% CI, 0.360-0.927) but not in patients aged ≥65 years. CONCLUSIONS: A significantly lower risk of prostate cancer is observed after a cumulative duration of pioglitazone therapy for >20.23 months or a cumulative dose of >15,330 mg. The risk reduction is mainly observed in patients aged <65 years.