Development and validation of a next-generation sequencing-based method for calculating the breast cancer polygenic risk score PRS313

Abstract

Polygenic risk scores (PRS) are genetic tools that quantify an individual's predisposition to certain diseases by combining the effects of many genetic variants. The PRS₃₁₃ includes 313 genomic variants and has been incorporated into the BOADICEA prediction model and in the CanRisk software to refine breast cancer risk. However, its current implementation relies on SNP microarrays, limiting its use in sequencing-based clinical workflows.

In this study, we directly compared SNP microarray technology and targeted NGS sequencing to determine the PRS₃₁₃. The two methods were tested for 154 patients. To replace PRS₃₁₃ SNPs with low sequencing coverage and/or in regions of low complexity, 27 proxy SNPs in high linkage disequilibrium were integrated into the panel. After this optimization, the NGS-derived PRS₃₁₃ demonstrated strong concordance with the microarray reference (R² = 0.95), with sensitivity and specificity reaching 96 % and 97 %, respectively. Moreover, the clinical risk category, as defined by CanRisk, remained consistent in 97 % of cases across both methods.

These findings validate the use of targeted NGS for PRS₃₁₃ calculation, demonstrating its feasibility, accuracy, and potential for easy integration into routine oncogenetic workflows. By enabling PRS calculation from the same sequencing data used for gene panel testing, this approach eliminates the need for separate genotyping platforms, offering a cost-effective and clinically practical solution to support the broader implementation of personalized breast cancer risk prediction.