Cumulative Advanced Breast Cancer Risk Prediction Model Developed in a Screening Mammography Population

Abstract

Background: Estimating advanced breast cancer risk in women undergoing annual or biennial mammography could identify women who may benefit from less or more intensive screening. We developed an actionable model to predict cumulative 6-year advanced cancer (prognostic pathologic stage II or higher) risk according to screening interval.

Methods: We included 931 186 women aged 40-74 years in the Breast Cancer Surveillance Consortium undergoing 2 542 382 annual (prior mammogram within 11-18 months) or 752 049 biennial (prior within 19-30 months) screening mammograms. The prediction model includes age, race and ethnicity, body mass index, breast density, family history of breast cancer, and prior breast biopsy subdivided by menopausal status and screening interval. We used fivefold cross-validation to internally validate model performance. We defined higher than 95th percentile as high risk (>0.658%), higher than 75th percentile to 95th or less percentile as intermediate risk (0.380%-0.658%), and 75th or less percentile as low to average risk (<0.380%).

Results: Obesity, high breast density, and proliferative disease with atypia were strongly associated with advanced cancer. The model is well calibrated and has an area under the receiver operating characteristics curve of 0.682 (95% confidence interval = 0.670 to 0.694). Based on women's predicted advanced cancer risk under annual and biennial screening, 69.1% had low or average risk regardless of screening interval, 12.4% intermediate risk with biennial screening and average risk with annual screening, and 17.4% intermediate or high risk regardless of screening interval.

Conclusion: Most women have low or average advanced cancer risk and can undergo biennial screening. Intermediate-risk women may consider annual screening, and high-risk women may consider supplemental imaging in addition to annual screening.

Figure 1.

Cumulative percentage of predicted risk of advanced cancers (stage II or higher) among annual screeners stratified by cumulative percentage of 6-year advanced cancer risk sorted from highest to lowest risk. The shaded regions indicate 4 risk groups: high risk, >95th percentile; intermediate risk, >75th and ≤95th percentile; average, >25th and ≤75th percentile; low and very low risk, ≤25th percentile.

Figure 2.

Cumulative percentage of predicted risk of advanced cancers (stage II or higher) among biennial screeners stratified by cumulative percentage of 6-year advanced cancer risk sorted from highest to lowest risk. The shaded regions indicate 4 risk groups: high risk, >95th percentile; intermediate risk, >75th and ≤95th percentile; average, >25th and ≤75th percentile; low and very low risk; ≤25th percentile.

Figure 3.

Mean cumulative risk of advanced breast cancer (stage II or higher) after 6 years of annual vs biennial screening and mean absolute risk difference from biennial to annual screening. Absolute risk difference not shown for proliferative disease with atypia because differences are so large estimates are off the graph. Estimates are based on a model including age, breast cancer family history, breast density, breast biopsy, body mass index, and race and ethnicity. Adjusted by US population weights and standardized to same population for annual and biennial. ^aIntermediate or high risk with biennial and annual screening. ^bIntermediate risk with biennial screening and average risk with annual screening. — = Absolute risk difference not applicable for ages 60-74 for premenopausal women; BI-RADS = Breast Imaging Reporting and Data System; IQR = interquartile range.