Common variants in breast cancer risk loci predispose to distinct tumor subtypes

Abstract

Background: genome-wide association studies have identified over 170 common breast cancer susceptibility loci, many of them with differential associations by estrogen receptor (ER). How these variants are related to other tumor features is unclear. 

Methods: analyses included 106,571 invasive breast cancer cases and 95,762 controls of European ancestry with data on 178 genotyped or imputed single nucleotide polymorphisms (SNPs). We used two-stage polytomous logistic regression models to evaluate SNPs in relation to multiple tumor features (ER, progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2) and grade) adjusting for each other, and in relation to molecular subtypes. 

Results: nearly half of the SNPs (85 out of 178) were associated with at least one tumor feature (false discovery rate <5%). Case-case comparisons identified ER and grade as the most common heterogeneity sources, followed by PR and HER2. Case-control comparisons among these 85 SNPs with molecular subtypes identified four main clusters of SNPs: two clusters most strongly or exclusively (P<0.05) associated with luminal-like subtypes (65 SNPs), one cluster associated with all subtypes at various strengths (5 SNPs), and one cluster primarily associated with non-luminal tumors, especially triple-negative (TN) disease (15 SNPs). A CHEK2 variant (rs17879961) showed significant risk associations with luminal A-like (P=9.26x10-14) and TN (P=2.55x10-4) subtypes in opposite directions.

Conclusion: breast cancer susceptibility loci have complex associations with multiple tumor features. ER and tumor grade are the most common sources of heterogeneity. These findings provide insights into the genetic predisposition of breast cancer subtypes and can inform subtype-specific risk predictions.

More information

Identifiers

ORCID for Diana Eccles: orcid.org/0000-0002-9935-3169

Catalogue record

Export record

Altmetrics