Abstract
Background: While protein truncating variants in RAD51C have been shown to predispose to triple negative breast cancer (TNBC) and ovarian cancer, little is known about the pathogenicity of missense (MS) variants. Methods: The frequency of rare RAD51C MS variants were assessed in the BEACCON study of 5,734 familial breast cancer cases and 14,382 population controls, and integrated with tumour sequencing data from 21 cases carrying a candidate variant to assess bi-allelic inactivation. Results: Collectively, a significant enrichment of rare missense variants was detected in cases (MAF < 0.001, OR 1.57, 95%CI 1.00 - 2.44, p = 0.05), particularly for variants with a REVEL score > 0.5 (OR 3.95, 95%CI 1.40 - 12.01, p = 0.006). Despite the large sample size, the majority of variants detected were very rare, precluding definitive conclusions about pathogenicity based solely on the case-control data. Sequencing of 21 tumours from carriers of one of eight candidate MS variants, identified four cases with bi-allelic inactivation through loss of the wild-type allele, while six lost the variant allele and ten remained heterozygous. Loss of the wild-type alleles corresponded strongly with ER- and triple-negative breast tumours and a high homologous recombination deficiency score. Conclusions: Using this approach, the p.Gly264Ser variant, which was previously suspected to be pathogenic based on small case-control analyses and loss of activity in in vitro functional assays, was shown to be benign with similar prevalence in cases and controls, and eight out of nine tumours showing loss of the variant allele or retention of heterozygosity. Conversely, the combined case-control and tumour sequencing data identified p.Ile144Thr, p.Arg212His, p.Gln143Arg and p.Gly114Arg as variants warranting further investigation.
Spatial analysis of lung, colorectal, and breast cancer on Cape Cod: An application of generalized additive models to case-control data