Development and validation of the Snehita BRISK model: A breast cancer risk assessment tool for risk stratification in women of the Indian subcontinent

<h2>Abstract</h2><h3>Background</h3> Breast cancer is the most prevalent cancer among women globally and in India, with rising incidence rates, particularly among younger women. Early detection remains crucial for reducing mortality. Risk assessment may motivate women to undergo examinations and utilize the calculator for risk stratification in early detection programs. However, existing breast cancer risk prediction models, primarily developed in Western populations, may not be fully applicable to the Indian context due to demographic, lifestyle, and genetic differences. This study aims to develop and validate the Snehita BRISK model, a logistic regression-based risk prediction tool tailored to the Indian subcontinent. <h3>Methods</h3> A case-control study involving 660 cases and 910 controls was conducted. Data were collected through structured interviews covering socio-demographic details, reproductive history, prior breast biopsies, and family history of breast cancer. Logistic regression analysis was employed to identify significant risk factors of breast cancer, leading to the formulation of the Snehita BRISK model. The model's performance was evaluated using Receiver Operating Characteristic (ROC) analysis, with additional validation tests. <h3>Results</h3> Significant risk factors included advancing age, early age at menarche, late age at first childbirth, nulliparity, irregular menstrual cycles, family history of breast cancer, and history of previous breast biopsies. The Snehita BRISK model demonstrated strong predictive accuracy (67.3 %) with an overall Area Under the Curve (AUC) of 0.699, sensitivity of 45.6 %, and specificity of 83 %. The model showed particularly high predictive accuracy (71.1 %) in women aged 50 years or older (AUC = 0.750). The overall probability cut-off for identifying high-risk women was 0.4338. <h3>Conclusion</h3> The Snehita BRISK model is a promising tool for breast cancer risk assessment, specifically tailored for women in the Indian subcontinent. By identifying high-risk women and motivating them to undergo timely examinations, this model supports effective risk stratification in screening programs and enhances early detection efforts. Presented as the ‘Snehita Breast Cancer Risk Calculator,' this tool aims to elevate early detection rates, particularly among older women. However, further validation in independent populations is necessary to confirm its broader applicability and effectiveness across diverse groups.