Abstract PS3-01-11: Breast Cancer Risk Prediction for Racially Diverse Women With Benign Breast Disease

Abstract Purpose: Existing breast cancer risk prediction models are ineffective for women with benign breast disease (BBD), especially in racially and ethnically diverse populations. We aimed to identify risk predictors for accurately estimating breast cancer development in a contemporary cohort of women from diverse racial backgrounds diagnosed with BBD. Patients and Methods: We identified 9,202 women diagnosed with histologically-confirmed benign lesions at the Joanne Knight Breast Health Siteman Cancer Center in St. Louis from 2010-2023. Risk modeling was performed with Cox regression using 95% confidence intervals (95%CI) and internally validated through a bootstrap resampling method with 100 replacements (i.e., corrected for overfitting). Discriminatory model performance was estimated by Harrell's C-index. Calibration was assessed by estimating observed and predicted ratio (O/E) at 5-year BC predicted risk. Results: Among the 9,202 women with BBD, 64.3% were White, 32.5% were Black, and 3.2% were Asian women, 3.3% had a subsequent breast cancer at least six months following BBD. The final model predictors included: age at BBD diagnosis, family history of breast cancer, atypical hyperplasia (AH), proliferative disease without atypia (PDWA), breast density, and race. The model discrimination in the original sample had a Harrell's C-index of 0.68 (95%CI 0.65, 0.71) and in the bootstrap sample of 0.67 (95%CI 0.63, 0.70). At the 5-year, the overall model calibration was good (O/E=0.98, 95%CI 0.86, 1.12). The model calibration was satisfactory across most risk-factor-defined subgroups; however, it tended to underestimate risk in the low-risk group and overestimate risk in the high-risk group (as defined by 20% quantiles). Conclusion: The final model outperformed existing breast cancer risk models among BBD women and future research will externally validate the model evaluating the risk among a subset of women with atypia. Citation Format: A. Koric, Y. Park, J. Jiang, F. Boul, G. Colditz. Breast Cancer Risk Prediction for Racially Diverse Women With Benign Breast Disease [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS3-01-11.