Clinical applicability of the Polygenic Risk Score for breast cancer risk prediction in familial cases

Abstract

BackgroundCommon low-risk variants are presently not used to guide clinical management of familial breast cancer (BC). We explored the additive impact of a 313-variant-based Polygenic Risk Score (PRS313) relative to standard gene testing in non-BRCA1/2Dutch BC families.MethodsWe included 3918 BC cases from 3492 Dutch non-BRCA1/2BC families and 3474 Dutch population controls. The association of the standardised PRS313with BC was estimated using a logistic regression model, adjusted for pedigree-based family history. Family history of the controls was imputed for this analysis. SEs were corrected to account for relatedness of individuals. Using the BOADICEA (Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm) V.5 model, lifetime risks were retrospectively calculated with and without individual PRS313. For 2586 cases and 2584 controls, the carrier status of pathogenic variants (PVs) inATM,CHEK2andPALB2was known.ResultsThe family history-adjusted PRS313was significantly associated with BC (per SD OR=1.97, 95% CI 1.84 to 2.11). Including the PRS313in BOADICEA family-based risk prediction would have changed screening recommendations in up to 27%, 36% and 34% of cases according to BC screening guidelines from the USA, UK and the Netherlands (National Comprehensive Cancer Network, National Institute for Health and Care Excellence, and Netherlands Comprehensive Cancer Organisation), respectively. For the population controls, without information on family history, this was up to 39%, 44% and 58%, respectively. Among carriers of PVs in known moderate BC susceptibility genes, the PRS313had the largest impact forCHEK2andATM.ConclusionsOur results support the application of the PRS313in risk prediction for genetically uninformative BC families and families with a PV in moderate BC risk genes.