Precision Breast Cancer Screening with a Polygenic Risk Score

Breast cancer (BC) is the leading cause of cancer deaths in women in the world. Genome-wide association studies have identified numerous genetic variants (SNPs) independently associated with BC. The effects of such SNPs can be combined into a single polygenic risk score (PRS). Stratification of women according to PRS could be introduced to primary and secondary prevention. Our aim was to revalidate a PRS model and to develop a pipeline for individualizing breast cancer screening. Previously published PRS models for predicting the risk of breast cancer were collected from the literature. Models were validated on the Estonian Biobank (EGC) dataset consisting of 32,548 quality-controlled genotypes with 315 prevalent and 365 incident BC cases and on 249,062 samples in the UK Biobank dataset consisting of 8637 prevalent and 6825 incident cases. The best performing model was selected based on the AUC in prevalent data and independently validated in both incident datasets. Using Estonian BC background information, we performed absolute risk simulations and developed individual risk-based recommendations for prevention. The best-performing PRS included 2803 SNPs. The C-index of the Cox regression model associating BC status with PRS was 0.656 (SE = 0.05) with a hazard ratio of 1.66 (95% confidence interval 1.5 - 1.84) on the incident EGC dataset. The PRS is able to stratify individuals with more than a 3-fold risk increase. The observed 10-year risks of individuals in the 99th percentile exceeded the 1st percentile more than 10-fold. PRS is a powerful predictor of breast cancer risk. Currently, PRS scores are not implemented in routine BC screening. We have developed PRS-based recommendations for personalized primary and secondary prevention and our approach is easily adaptable to other nationalities by using population-specific background data of other genetically similar populations.