Polygenic risk scores in breast cancer: A systematic review of predictive utility, clinical applications, and future directions

Abstract: BACKGROUND: The most common cancer among women in the world is breast cancer (BC), and the need for accurate risk prediction tools to enhance early detection and personalized prevention is great. Polygenic risk scores (PRSs), which aggregate the effects of many genetic variants, have been promising to enhance BC risk stratification beyond traditional risk factors. METHODS: Systematic review was conducted in PubMed, Scopus, Web of Science, and Google Scholar between January 2012 and May 2025. Randomized controlled trials, cohort, case-control, and case reports reporting PRS construction, validation, or clinical application to predict risk of BC were included. Two reviewers screened articles independently and extracted data on study design, population, PRS methodology, predictive accuracy, and clinical integration. Joanna Briggs Institute critical appraisal tools were used to assess the quality. RESULTS: Several studies show that PRS improve BC risk prediction, particularly in BRCA1/2 mutation carriers and high-risk individuals. Inclusion of epidemiological and clinical variables improves prediction accuracy. Most PRS models were constructed using European ancestry cohorts and lack generalizability in multi-ethnic populations. There are constraints in translating relative risk to absolute measures and communicating this in the clinic. There is emerging evidence to support PRS in personalized screening and prevention, but further validation and clinical integration are needed. CONCLUSION: PRSs are an exciting new technology in BC risk prediction that has the potential to enhance personalized medicine. Heterogeneity in the population, model improvement, and clinical application are still the top priorities for future research.