Tissue-based predictors of germ-line BRCA1 mutations: implications for triaging of genetic testing

SummaryBRCA testing of patients with breast cancer considered at high risk for having a germ-line BRCA mutation usually consists of comprehensive mutational analysis of both BRCA1 and BRCA2. A more cost-effective strategy of triaging patients for analysis of a single gene could be adopted if tissue-based predictors indicated a high risk specifically for either BRCA1 or BRCA2. To identify potentially useful tissue-based predictors of BRCA mutation status in breast cancer, we evaluated multiple histopathologic features of invasive breast carcinoma on archival tissue sections from 196 high-risk patients who had undergone BRCA testing, and we analyzed which individual or combination of features was most associated with BRCA mutations. Of the 196 patients with invasive breast carcinoma, there were 44 (22%) with a deleterious BRCA1 mutation and 27 (14%) with a deleterious BRCA2 mutation. For patients with available untreated surgical resection specimens for evaluation (n = 172), estrogen receptor-positive phenotype was inversely associated with the presence of a BRCA1 mutation (odds ratio, 0.243; 95% confidence interval, 0.070-0.840; P = .025), and high mitotic activity (≥25 mitotic figures per 10 high-power fields) was directly associated with the presence of a BRCA1 mutation (odds ratio, 4.222; 95% confidence interval, 1.353-13.18; P = .013). The combination of estrogen receptor-negative phenotype and high mitotic rate had high specificity (99%; 95% confidence interval, 95%-100%) but low sensitivity (43%; 95% confidence interval, 26%-61%) for identifying a deleterious BRCA1 mutation. In patients with breast cancer at high risk for carrying a BRCA mutation, those with estrogen receptor-negative phenotype and high mitotic rate could be triaged specifically for BRCA1 testing instead of initially performing mutational analysis for both BRCA1 and BRCA2.