Unexplained familial breast cancer risk – how can we find the missing genes?


Year:

Session type:

Julian Peto
London School of Hygiene and Tropical Medicine, UK

Abstract

The familial relative risk (FRR) for cancer is the risk in patients' parents, siblings and children relative to the risk in the general population. BRCA1, BRCA2 and other high-risk susceptibility genes, together with the common SNPs identified in GWASs, account for only one third of the variation in individual breast cancer risk implied by an FRR of 2, which is the observed risk in relatives averaged across all ages.

The estimated effects of SNPs that were included in GWASs but did not achieve statistical significance increases the proportion of variation explained to almost 50%. The search for these and other genetic variants that account for the unexplained FRR (the "missing heritability") will involve very large GWASs including less common SNPs, whole-genome sequencing studies to identify rarer variants, and detailed studies of the combined effects of variants in genes and regulatory elements in functional pathways. Genes discovered through their correlation with relevant phenotypes, particularly breast density, which is the strongest predictor of breast cancer risk, may make a substantial additional contribution.

The FRR in relatives of young cases is more than 5 below age 40, falling to about 1.4 above age 60, while in relatives of older cases aged the FRR is less at younger ages but similar (about 1.4) above age 60. The effects of most known SNPs show little or no trend with age, however, and a major weakness in the conventional polygenic model of genetic susceptibility is its failure to account for this strong age-dependence in FRR. This pattern, with the risk appearing at an earlier age in some families, suggests (1) that a substantial proportion of the FRR above age 60 is already explained by known genes, and (2) that the undiscovered genes may include many that accelerate age at onset but have little effect on risk at older ages. Such genes may be missed by conventional analysis, but they might be identified from GWASs by analysing the trend in risk with age at diagnosis, which is statistically independent of their overall effect.

Share this abstractTweet about this on TwitterPrint this pageShare on FacebookEmail this to someoneShare on LinkedIn