Abstract 2761: Evaluation of the BOADICEA model for predicting BRCA1 and BRCA2 mutation carrier probabilities in high-risk US Hispanic and Mexican families: A report from the Clinical Cancer Genetics Community Research Network

Background: BRCA1 and BRCA2 (BRCA) mutation prediction models are low cost tools that enable clinicians to target individuals with a higher probability to carry mutations in the BRCA genes. Most of the data used to construct these models were derived from a predominantly non-Hispanic White population. The predictive accuracy of these models in diverse populations is unclear. The Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm (BOADICEA) was developed using population-based families with multiple affected individuals that were primarily from the UK. The purpose of this study was to investigate and compare the accuracy of BOADICEA for prediction of BRCA mutation status among US Hispanics and a cohort of Mexicans in Guadalajara and Mexico City. Methods: Hispanic patients undergoing genetic cancer risk assessment (GCRA) were enrolled in an IRB-approved prospective registry from sites within the Clinical Cancer Genetics Community Research Network (CCGCRN), comprised of the City of Hope and 47 collaborating sites representing primarily community-based oncogenetic practices across the US and Latin America, including two cancer institutes in Mexico (University of Guadalajara; Instituto Nacional de Cancerología, Mexico City). Pedigrees were created in Progeny, exported to text files, and then uploaded for probablity calculations according to BOADICEA version 3 guidelines. Accuracy was assessed using an observed versus expected and an area under the receiver operating characteristic (ROC) curve analysis. Results: Overall, BOADICEA performed better in US Hispanics (area under the curve [AUC] = 0.75) for predicting a BRCA mutation than in the Mexican cohort (AUC = 0.64; p-value = 0.0015). Although BRCA mutations were documented in 238 of 1,247 (19.1%) US Hispanics and 23 of 234 (9.8%) Mexicans, BOADICEA predicted a mean mutation probability of 10.3% and 5.6%, respectively. Summary: Though generally the best performing model, BOADICEA's ability to predict BRCA mutations has been investigated in previous studies with different populations with varying results. The lower performance of BOADICEA in the Mexican cohort could be a result of the limitations in pedigree data. It's possible that families don't share information about cancer due to cultural norms or fear of stigmitation. It is also possible that the model assumptions about population mutation frequency are inaccurate. The results of this study indicate the need to document the c