Comprehensive Breast Cancer Risk Assessment for and Pathogenic Variant Carriers Incorporating a Polygenic Risk Score and the Tyrer-Cuzick Model.

PURPOSE

Breast cancer risks for and pathogenic variant (PV) carriers are modified by an 86-single nucleotide polymorphism polygenic risk score (PRS) and individual clinical factors. Here, we describe comprehensive risk prediction models for women of European ancestry combining PV status, PRS, and individual clinical variables.

MATERIALS AND METHODS

This study included deidentified clinical records from 358,095 women of European ancestry who received testing with a multigene panel (September 2013 to November 2019). Model development included PV carriers (n = 4,286), PV carriers (n = 2,666), and women negative for other breast cancer risk gene PVs (n = 351,143). Odds ratios (ORs) were calculated using multivariable logistic regression with adjustment for familial cancer history. Risk estimates incorporating PV status, PRS, and Tyrer-Cuzick v7.02 were calculated using a Fixed-Stratified method that accounts for correlations between risk factors. Stratification of PV carriers into risk categories on the basis of remaining lifetime risk (RLR) was assessed in independent cohorts of PV carriers.

RESULTS

ORs for association of PV status with breast cancer were 2.01 (95% CI, 1.88 to 2.16) and 1.83 (95% CI, 1.68 to 2.00) for and PV carriers, respectively. ORs for PRS per one standard deviation were 1.51 (95% CI, 1.37 to 1.66) and 1.45 (95% CI, 1.30 to 1.64) in and PV carriers, respectively. Using the combined model (PRS plus Tyrer-Cuzick plus PV status), RLR was low (≤ 20%) for 24.2% of PV carriers, medium (20%-50%) for 63.8%, and high (> 50%) for 12.0%. Among PV carriers, RLR was low for 31.5% of patients, medium for 58.5%, and high for 9.7%.

CONCLUSION

In and PV carriers, risk assessment including PRS, Tyrer-Cuzick, and PV status has the potential for more precise direction of screening and prevention strategies.