Repurposing existing medications offers a practical and economical way to tackle the demand for new breast cancer prevention and treatment options. Our goal was to pinpoint druggable targets with potential clinical relevance through Mendelian randomization (MR) and confirm promising drug candidates via real-world population data. We selected genetic variants as instruments for 1406 actionable targets of licensed non-cancer drugs, drawing from gene expression (eQTL), DNA methylation (mQTL), and protein expression (pQTL) quantitative trait loci. Summary statistics came from a large breast cancer genome-wide association study (GWAS) by the Breast Cancer Association Consortium (122,977 cases and 105,974 controls). We also performed a nested case-control analysis using Swedish national register data to evaluate the drugs highlighted by the MR results. MR analyses revealed six notable links at the gene expression level (TUBB, MDM2, CSK, ULK3, MC1R, and KCNN4) and two at the DNA methylation level involving 21 CpG sites (RPS23 and MAPT). In the nested case-control study, raloxifene use (which targets MAPT) correlated with a 35% lower breast cancer risk (odds ratio [OR]: 0.65; 95% confidence interval [CI]: 0.51–0.83). Conversely, estradiol, tolterodine, and nitrofurantoin (also linked to MAPT) showed associations with elevated risk, with adjusted ORs (95% CIs) of 1.10 (1.07–1.13), 1.16 (1.09–1.24), and 1.09 (1.05–1.13), respectively. The associations for raloxifene and nitrofurantoin no longer reached significance in sensitivity analyses employing active-comparator and new-user designs. This extensive MR investigation, supported by population-level confirmation, uncovered eight druggable genes linked to breast cancer and highlighted raloxifene as a valuable agent for chemoprevention.
