A comparison of internet-based models assessing the likelihood of a BRCA mutation.

22051 Background: Mutations in the BRCA1 and BRCA2 genes account for the majority of hereditary breast and ovarian cancer. Various risk assessment models are available to calculate the risk of a BRCA mutation, and a greater than 10% risk is often used to identify an eligible candidate for genetic testing. Of these models, the Penn II ( http://www.afcri.upenn.edu/itacc/penn2/ ) and BOADICEA ( https://pluto.srl.cam.ac.uk/cgi-bin/bd1/v1/bd.cgi ) are the only models available on the internet. Here, we compare these models to determine their utility in patients from a large breast cancer program in our tertiary referral center.,Risk assessments were done using the Penn II and BOADICEA models for 200 consecutively tested non-BRCA carriers, and 100 consecutively tested BRCA carriers from Mount Sinai Hospital, Toronto (Canada) from August 1995 and March 2006. Sensitivity and specificity were calculated at the 10% testing threshold. Receiver operating curves were constructed by plotting sensitivity over the false positive rate (1-specificity) and area under the curves (AUCs) were calculated. Testing thresholds were established at 90% sensitivity.,100 carriers (91% female, 58% BRCA1 mutation positive, 39% Ashkenazi Jewish, mean age: 50.7 years, 79% affected with breast cancer) and 200 non-carriers (97.5% female, 39.5% Ashkenazi Jewish, mean age: 52 years, 73% affected with breast cancer) were included. The AUC was 0.74 for both models. The sensitivity and specificity at the 10% testing threshold were: Penn II (0.93, 0.32); BOADICEA (0.70, 0.65). The testing threshold to achieve 90% sensitivity was 10.9% and 2.8% respectively.,Both the Penn II and BOADICEA risk assessment models identify those at highest risk of carrying a BRCA mutation, and have been used to determine eligibility for publicly and privately funded genetic testing. A major advantage of both models is that they are available on the internet and thus easily accessible. In our patient population, substantially higher sensitivity was achieved with the Penn II model at the 10% testing threshold. A much lower testing threshold (2.8%) would be required to achieve a similar sensitivity with the BOADICEA model. An additional benefit of the Penn II model is the ease of data entry when compared to the BOADICEA model. No significant financial relationships to disclose.