Metformin and breast cancer risk: a meta-analysis and critical literature review

Observational studies have suggested that metformin decreases the incidence of several common cancers. However, findings regarding breast cancer have been mixed. In order to explore this issue, a systematic literature review and meta-analysis were performed with a focus on potential biases. We conducted a comprehensive literature search for all pertinent studies addressing metformin use and breast cancer risk by searching PubMed, Cochrane Library, and Scopus (which includes Embase, ISI Web of Science) using the Mesh terms: "metformin" or "biguanides" or "diabetes mellitus, type 2/therapy" and "cancer" or "neoplasms." When multiple hazard ratios (HR) or odds ratio (OR) were reported, the most adjusted estimate was used in the base-case analysis. We pooled the adjusted HR and performed sensitivity analyses on duration of metformin use (> or a parts per thousand currency sign3 years use), study quality (assessed using the GRADE system), and initial observation year of the cohort (before vs after 1997). From a total of 443 citations, 18 full-text articles were considered, and seven independent studies were included. All were observational (four cohort and three case control). Our combined OR of all seven studies was 0.83 (0.71-0.97). Stronger associations were found when analyses were limited to studies estimating the impact of longer metformin use (OR = 0.75. 95 % CI 0.62, 0.91) or among studies that began observing their cohort before 1997 (OR = 0.68. 95 % CI 0.55-0.084). Stratification according to study quality did not affect the combined OR but higher quality studies had smaller CI and achieved statistical significance. Interpretation is limited by the observational nature of reports and different comparison groups. Our analyses support a protective effect of metformin on breast cancer risk among postmenopausal women with diabetes. Clinical trials are needed for definitive determination of the role of metformin in breast cancer risk reduction.