Abstract The past two decades have seen dramatic progress in our understanding of the inherited basis of breast cancer, driven by large international collaborative epidemiological studies. Genetic factors identified include rare coding variants in susceptibility genes, and more than 200 common polymorphisms, identified through genome-wide association studies (GWAS) and largely in non-coding regions, that confer more moderate risks. The majority of the common variants are more strongly associated with ER-positive disease, but a subset is specifically associated with ER-negative disease. Variants are strongly enriched for regulatory regions and transcription factor binding sites, and the putative target genes are strongly enriched for known breast cancer drivers: however, the underlying target genes and mechanisms remain largely obscure. The associations for common variants appear to be largely independent of each other and can be summarised by a polygenic risk score (PRS) that is strongly predictive of risk. The most extensively validated PRS, comprising 313 SNPs, is associated with a relative risk of approximately 1.6 per standard deviation; women in the top 1% of the risk distribution have an approximately 4-fold risk of ER-positive disease and 3-fold risk of ER-negative disease. The PRS was developed in women of European ancestry but is also predictive of risk in Asian ancestry women, though less so in African ancestry women. In contrast, despite their widespread use, the evidence for association with cancer for coding variants in genes on genetic testing panels is often weak. Recent very large-scale targeted sequencing studies, however, have clarified the position, providing strong evidence for an association between protein truncating or missense variants and breast cancer risk for nine genes: BRCA1, BRCA2, ATM, CHEK2, PALB2, BARD1, RAD51C, RAD51D and TP53. Recent large-scale exome-wide sequencing analysis has shown that rare variants in several additional genes are also risk associated, with the majority being tumour suppressor genes: these include MAP3K1, ATRIP and SAMHD1. These analyses also demonstrate, however, that the contribution of coding variation, beyond the known genes, is likely to be small. The currently identified variants explain about half of the familial risk of breast cancer. Part of the “missing” heritability is due to weaker common variants, many of which should be identified in ongoing larger GWAS. The causes of the remainder are obscure and may include rare non-coding variants, repeat polymorphisms, and epistatic interactions. Risk discrimination is most powerful when all risk factors are considered. The BOADICEA model, implemented in the CanRisk tool, provides a well validated model that incorporates the effects of gene variants, polygenic risk scores, lifestyle factors and breast density to predict breast and ovarian cancer risk. Multiple translational studies are ongoing to evaluate the feasibility and utility of such personalised risk prediction for stratifying screening and prevention. Citation Format: Douglas F Easton. Genetic susceptibility to breast cancer: Recent advances and application to personalised prevention [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Breast Cancer Research; 2023 Oct 19-22; San Diego, California. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_1):Abstract nr IA18.
Read full abstract