Breast cancer risk prediction model specific to sub-Saharan African women outperforms existing models

Recent research in Nigeria developed an absolute breast cancer risk prediction model specific to sub-Saharan African women to identify high risk individuals for cancer screening as well as to assess its performance against existing models. The research used data from the Nigerian Breast Cancer Study (NBCS) to compare women with confirmed invasive breast cancer to control groups without the disease, totalling about 4000 individuals. The risk prediction model was constructed using 9 breast cancer risk factors which included age, family history of breast cancer, age at menarche, and benign breast illnesses.

The resulting NBCS absolute risk prediction model was found to have a significantly higher discriminating accuracy (Area Under Curve (AUC) = 0.703) of 5-year risk when compared to three existing breast cancer risk prediction models: the Black Womens Health Study Model (AUC = 0.605), the Gail model for white population (AUC = 0.551), and the Gail model for black population (AUC = 0.545). The NBCS was also notably better at distinguishing between breast cancer cases and controls at thresholds used to determine eligibility for breast cancer prevention trials (5 year risk exceeding 1.66%). For detailed information about this research, click here.