Polygenic Risk Scores for Prediction of Breast Cancer and Breast Cancer Subtypes.

Area of Science:

• Genetics and Genomics
• Oncology
• Epidemiology

Background:

• Stratifying breast cancer risk using polygenic risk scores (PRSs) can enhance screening and prevention.
• Estrogen receptor (ER)-specific risk prediction is crucial for personalized strategies.

Purpose of the Study:

• Develop and validate PRSs optimized for predicting ER-specific breast cancer.
• Utilize the largest available genome-wide association dataset for PRS development.

Main Methods:

• Developed PRSs using stepwise regression or lasso penalized regression on a large dataset (94,075 cases, 75,017 controls).
• Validated PRSs in independent prospective studies (11,428 cases, 18,323 controls) and UK Biobank (190,040 women).
• Selected 313 single-nucleotide polymorphisms (SNPs) for the final PRS models.

Main Results:

• The best PRS demonstrated an odds ratio of 1.61 per SD for overall breast cancer (AUC=0.630).
• Top centile PRS indicated a 32.6% lifetime risk for overall breast cancer.
• Highest PRS group showed 4.37-fold (ER-positive) and 2.78-fold (ER-negative) risks; lowest PRS group showed reduced risks (0.16- and 0.27-fold).

Conclusions:

• The developed PRS is a powerful and reliable predictor of breast cancer risk.
• PRS demonstrates good calibration and accurate prediction, especially in the tails of the risk distribution.
• This PRS holds potential for improving breast cancer prevention programs and personalized risk assessment.