Validation Study of Several Models of Breast Cancer Risk
In a validation study reported in The Lancet Oncology, Terry et al found that the Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm model (BOADICEA) and the International Breast Cancer Intervention Study model (IBIS) were the best of 4 models tested for predicting 10-year breast cancer risk.
The study involved data from 15,732 women from Australia, Canada, and the United States in the Breast Cancer Prospective Family Study Cohort who did not have breast cancer at recruitment between March 1992 and June 2011. The cohort was used to calculate 10-year risk scores and compare risk prediction by the BOADICEA, BRCAPRO, the Breast Cancer Risk Assessment Tool (BCRAT), and IBIS models.
Risk Prediction Accuracy
After a median follow-up of 11.1 years, 619 (4%) of 15,732 women were prospectively diagnosed with breast cancer after recruitment, with 519 of these (84%) having histologically confirmed disease. In predicting risk at 10 years, BOADICEA and IBIS were well calibrated in the overall validation cohort, whereas BRCAPRO and BCRAT underpredicted risk; the ratio of expected cases to observed cases was 1.05 for BOADICEA, 1.03 for IBIS, 0.59 for BRCAPRO, and 0.79 for BRCAT. Estimated C-statistics for the validation cohort were 0.70 for BOADICEA, 0.71 for IBIS, 0.68 for BRCAPRO, and 0.60 for BCRAT.
In subgroup analyses by BRCA mutation status, the ratio of expected to observed cases among BRCA-negative women was 1.02 for BOADICEA, 1.00 for IBIS, 0.53 for BRCAPRO, and 0.97 for BCRAT. Among BRCA-positive participants, the three models that include information specifically for BRCA carriers were compared; the ratio of expected to observed cases was 1.17 for BOADICEA, 1.14 for IBIS, and 0.80 for BRCAPRO. Similar calibration patterns for BOADICEA and IBIS were observed among women aged < 50 years at recruitment. BOADICEA and IBIS predictive scores were not markedly affected by limiting input data to family history for first-degree and second-degree relatives.
The investigators concluded, “Our results suggest that models that include multigenerational family history, such as BOADICEA and IBIS, have better ability to predict breast cancer risk, even for women at average or below-average risk of breast cancer. Although BOADICEA and IBIS performed similarly, further improvements in the accuracy of predictions could be possible with hybrid models that incorporate the polygenic risk component of BOADICEA and the nonfamily-history risk factors included in IBIS.”
Mary Beth Terry, PhD, of the Department of Epidemiology, Columbia University Mailman School of Public Health, is the corresponding author for The Lancet Oncology article.
Disclosure: The study was funded by the National Cancer Institute, Breast Cancer Research Foundation, Australian National Health and Medical Research Council, Victorian Health Promotion Foundation, Victorian Breast Cancer Research Consortium, Cancer Australia, and others. The study authors' full disclosures can be found at thelancet.com.
The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.
