Candidate for the Lee B. Lusted Student Prize Competition
Purpose: This systematic review and meta-analysis aimed to assess the associations of metformin intake with incidence and mortality of prostate cancer.
Methods: A comprehensive literature search was performed using EMBASE, MEDLINE and PubMed databases with their latest available date of May 8th 2013. The following keywords and/or corresponding MESH terms were used: (‘‘metformin'' OR ‘‘biguanides'' OR ‘‘diabetes mellitus, type 2/therapy'' OR ‘‘hypoglycemic agents'' OR ‘‘glucose-lowering therapy'') AND (‘‘prostate cancer'' OR ‘‘prostate carcinoma'' OR ‘‘prostatic carcinoma''). In addition, a hand-check was done to identify relevant publications from other meta-analyses or from reference lists. Inclusion Criteria included: (1) epidemiological studies conducted in humans (case-control or cohort); (2) exposure to metformin; (3) prostate cancer risk estimates including relative risk/odds ratio/hazard ratio (RR/OR/HR) with their 95% confidence intervals (CIs) were reported or data for these risk estimates can be calculated; (4) if there were multiple publications for the same study, the most recent and complete publication was considered. Epidemiological studies which reported prostate cancer incidence or mortality (RR/OR/HR with 95% CI) were selected for meta-analysis of prostate cancer outcomes associated with metformin use. Summery Risk Ratio (SRR) was calculated using random-effect models. Heterogeneity was tested using Chi2 statistics and measured with the I2statistic. Funnel plot and Egger's regression asymmetry test were used to evaluate publication bias.
Results: A total of 17 studies (12 reported prostate cancer incidence, 5 reported prostate cancer mortality) which met inclusion/exclusion criteria were evaluated for meta-analysis, after reviewing a total of 1266 articles. Among 1,071,130 participants (with or without type 2 diabetes), metformin use was associated with a lower risk of incident prostate cancer compared with no metformin use. The SRR for prostate cancer incidence was 0.86 (95% CI: 0.78-0.94). Heterogeneity and publication bias were also found. Among 117,164 patients with prostate cancer, metformin users were not found to reduce prostate cancer-specific mortality, as compared with those without exposure to metformin. The SRR for prostate cancer mortality was 0.92 (95% CI: 0.69-1.24).
Conclusions: Metformin use appears to reduce prostate cancer incidence in the meta-analysis. However, there was no significant association of metformin with reducing mortality among patients with prostate cancer. Associations of other anti-diabetic medications and health outcomes among patients with prostate cancer should be examined in future systematic reviews and meta-analyses.
