Abstract
Mammographic breast density is an established risk factor for breast cancer. However, results are inconclusive regarding its use in risk prediction models. The current study evaluated 13,409 postmenopausal participants in the NSABP Study of Tamoxifen and Raloxifene. A measure of breast density as reported on the entry mammogram report was extracted and categorized according to The American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) classifications. An increased risk of invasive breast cancer was associated with higher mammographic breast density (P < 0.001). The association remained significant after adjusting for age, treatment, and smoking history [HR 1.35, 95% confidence interval (CI): 1.16-1.58], as well as when added to a model including the Gail score (HR 1.33, 95% CI: 1.14-1.55). At five years after random assignment, time-dependent area under the curve (AUC) improved from 0.63 for a model with Gail score alone to 0.64 when considering breast density and Gail score. Breast density was also significant when added to an abbreviated model tailored for estrogen receptor-positive breast cancers (P = 0.02). In this study, high BI-RADS breast density was significantly associated with increased breast cancer risk when considered in conjunction with Gail score but provided only slight improvement to the Gail score for predicting the incidence of invasive breast cancer. The BI-RADS breast composition classification system is a quick and readily available method for assessing breast density for risk prediction evaluations; however, its addition to the Gail model does not seem to provide substantial predictability improvements in this population of postmenopausal healthy women at increased risk for breast cancer.