Evaluation of a polygenic risk score as a predictor of early onset triple-negative breast cancer in Black women.

Abstract

10501 Background: Black women in the U.S. often develop early, biologically aggressive breast cancer (BC). Triple-negative breast cancer (TNBC) is a particularly aggressive type that occurs more frequently in Black than white women and often develops before recommended regular screening. More accurate risk prediction methods are urgently needed to identify young Black women with elevated risk of TNBC. Incorporating polygenic risk scores (PRS) into clinical models can substantially improve risk assessment, but most PRS have demonstrated poor performance among non-European ancestries. We previously described a multiple-ancestry PRS (MA-PRS), based on 56 ancestry-informative and 329 BC-associated single-nucleotide polymorphisms (SNPs), that predicts overall BC risk for diverse populations by characterizing genetic ancestry at each BC SNP and applying ancestry-specific SNP risks and frequencies. Here, we evaluated the extent to which MA-PRS improves upon clinical factors for the prediction of TNBC and early onset ( < 50 years) TNBC in a large, independent cohort of self-reported Black women. Methods: We examined clinical and genetic records from self-reported Black women referred for hereditary cancer testing from 8/22 - 9/23 and negative for pathogenic variants in BC-associated genes. The association of MA-PRS with TNBC was analyzed using multivariable logistic regression adjusted for personal and family cancer history, age and genetic ancestry. Analyses were conducted within the full cohort and the subpopulation of patients < 50 years old. Odds ratios (OR) are reported per standard deviation (SD) with 95% confidence intervals (CI). Results: We identified 14728 Black women who met study eligibility criteria. Nearly 70% (9895/14728) of patients were under the age of 50. In the full cohort, 629 (4.3%) patients were diagnosed with TNBC, and 173 (1.7%) were younger than 50 years old. MA-PRS significantly improved TNBC risk prediction over clinical factors in the full cohort and in the subpopulation of women under 50 (Table). The effect of MA-PRS on risk stratification was comparable to the 1.4 OR per SD reported in current literature for mammographic density, which is widely recognized as an important risk factor. Women in the top 5% of the MA-PRS distribution were at roughly 2-fold increased risk of TNBC (Table). Conclusions: MA-PRS substantially improved upon clinical factors for the prediction of TNBC and early-onset TNBC in a large cohort of Black women. MA-PRS appears to be equally as important for managing risk of early-onset TNBC as mammographic density for overall BC. Incorporation of MA-PRS into BC risk assessment has the potential to improve TNBC survival through more accurate identification of women at high risk. [Table: see text]