Mammographic density, pathogenic breast cancer susceptibility gene variants, and breast cancer risk: a pooled case–control analysis

Abstract

Background Major determinants of breast cancer risk include mammographic density and pathogenic variants in breast cancer susceptibility genes, but the association of these two factors and their joint effects on breast cancer risk are unclear. We aimed to investigate the association between the presence or absence of pathogenic variants in breast cancer susceptibility genes and mammographic density measures, and their joint effects on breast cancer risk. Methods This pooled case–control analysis comprised 6654 cases and 18 036 controls from 15 studies participating in the Breast Cancer Association Consortium (BCAC). Women were aged 19–92 years and had mammograms taken at least 1 year before diagnosis and had germline sequencing data available. Measures of mammographic density—dense area, non-dense area, percentage density, and absolute difference in percentage density between left-side and right-side breast—were scored with either Cumulus or Stratus software tools for computing dense and non-dense areas from processed mammogram images. We investigated the association between measures of mammographic density and the presence or absence of pathogenic variants in eight genes (ATM, BARD1, BRCA1, BRCA2, CHEK2, PALB2, RAD51C, and RAD51D) using linear regression and their joint effects on breast cancer risk using logistic regression. Analyses of CHEK2 c.1100delC were extended to an additional 12 244 participants for whom we had reliable array genotyping of the variant. Mendelian randomisation analyses were done to investigate causality between measures of mammographic density and breast cancer outcomes. Findings No association was found between the overall burden of pathogenic variants and any measures of mammographic density. Some evidence was noted for a negative interaction between the burden of pathogenic variants in the eight genes and percentage density (odds ratio [OR] 0·79, 95% CI 0·62–1·00; p=0·046), which appeared to be largely driven by a positive interaction with non-dense area. Mendelian randomisation analyses indicated attenuated effects for BRCA1 pathogenic variant carriers (for percentage density, OR per SD 1·02, 95% CI 0·78–1·34) but not for carriers of BRCA2 pathogenic variants (1·55, 1·08–2·24). Interpretation Our analysis found no evidence of an association between pathogenic variants in eight breast cancer susceptibility genes and measures of mammographic density. However, a weaker association between percentage density and breast cancer risk was noted in pathogenic variant carriers than in non-carriers. Replication of these findings in further large datasets is required. Funding Cancer Research UK, NIHR Cambridge Biomedical Research Centre, and the European Union.