The Influence of Hormone Therapy on secondary diabetes mellitus in Breast Cancer: A Meta-analysis.

Abstract

BACKGROUND\nGrowing evidences have implied that patients with primary breast cancer (BC) were at increased risks of developing diabetes mellitus (DM). However, as a major adjuvant treatment, the influence of hormone therapy (HT) on secondary DM in primary BC remains controversial; we conducted a meta-analysis of existing studies to evaluate the association of hormone therapy and secondary DM.\n\n\nMETHODS\nWe searched online databases (PubMed, EMBASE, the Cochrane library, Scopus, and Google Scholar) for studies exploring the influence of hormone therapy on secondary DM in BC. The summarized effect sizes (ES) and 95% confidence interval (95% CI) are calculated by STATA software utilizing fixed-effect or random-effect models, depending on the heterogeneity of the eligible studies.\n\n\nRESULTS\nUltimately, 7 retrospective publications including a total of 44,524 primary BC patients are eligible in present meta-analysis. HT use significantly increased the risk of developing DM in primary BC patients, whenever compared with NON-HT BC patients (pooled adjusted HR 1.30, 95% CI: 1.19-1.43) or NORMAL participants (HR 1.19, 95% CI: 1.14-1.25). As to specific HT medications, our sub-analysis demonstrates the risk for DM in tamoxifen (TAM) users elevates by 30% than NON-TAM use BC patients (pooled HR 1.30, 95% CI: 1.20-1.40) and by 18% than NORMAL participants (pooled HR 1.18, 95% CI: 1.12-1.24). However, for aromatase inhibitors (AIs) users, the risks for DM do not elevate significantly. Funnel plots and Egger s tests are used to evaluate publication bias and no apparent bias is detected in all analysis.\n\n\nCONCLUSION\nThe present study is the first meta-analysis which thoroughly reveals that adjuvant HT is a risk factor of secondary DM in primary female BC patients. As to specific HT medications, TAM use significantly enhances the incidence of secondary DM, while AIs use does not influence the DM incidence significantly. Our results can help clinicians to tailor more appropriate strategies for the therapy and follow-up of primary BC patients.