Performance of BOADICEA and BRCAPRO genetic models and of empirical criteria based on cancer family history for predicting BRCA mutation carrier probabilities: A retrospective study in a sample of Italian cancer genetics clinics

PURPOSE: To evaluate in current practice the performance of BOADICEA and BRCAPRO risk models and empirical criteria based on cancer family history for the selection of individuals for BRCA genetic testing. PATIENTS AND METHODS: The probability of BRCA mutation according to the three tools was retrospectively estimated in 918 index cases consecutively undergone BRCA testing at 15 Italian cancer genetics clinics between 2006 and 2008. RESULTS: 179 of 918 cases (19.5%) carried BRCA mutations. With the strict use of the criteria based on cancer family history 173 BRCA (21.9%) mutations would have been detected in 789 individuals. At the commonly used 10% threshold of BRCA mutation carrier probability, the genetic models showed a similar performance [PPV (38% and 37%), sensitivity (76% and 77%) and specificity (70% and 69%)]. Their strict use would have avoided around 60% of the tests but would have missed approximately 1 every 4 carriers. CONCLUSION: Our data highlight the complexity of BRCA testing referral in routine practice and question the strict use of genetic models for BRCA risk assessment.