Adenocarcinoma, Clear Cell/genetics* ; Adult ; Aged ; Aged, 80 and over ; Asian Continental Ancestry Group/genetics* ; Breast Neoplasms/genetics ; Carcinoma, Endometrioid/genetics* ; Decision Support Techniques ; Female ; Genes, BRCA1* ; Genes, BRCA2* ; Genetic Predisposition to Disease ; Genetic Testing ; Hereditary Breast and Ovarian Cancer Syndrome/diagnosis ; Hereditary Breast and Ovarian Cancer Syndrome/epidemiology ; Hereditary Breast and Ovarian Cancer Syndrome/genetics* ; Humans ; Middle Aged ; Models, Theoretical ; Mutation* ; Neoplasms, Cystic, Mucinous, and Serous/genetics* ; Ovarian Neoplasms/genetics* ; Probability ; ROC Curve ; Republic of Korea ; Risk Assessment
OBJECTIVE: To evaluate the predictive efficacies including sensitivity and positive predictive value of the genetic risk prediction model BRCAPRO and the Myriad BRCA risk calculator in Korean ovarian cancer patients.
METHODS: Individuals undergoing genetic testing for BRCA mutations from November 2010-August 2016 were recruited from the Department of Obstetrics and Gynecology at a single institute in Korea. The observed BRCA1 and BRCA2 mutation statuses were compared with the predicted carrier probabilities using BRCAPRO and the Myriad BRCA risk calculator.
RESULTS: Two hundred thirty-two patients were recruited, of whom 99.1% (230/232) were of Korean ethnicity. Of the 232 individuals, 206 and 26 had ovarian and double primary breast/ovarian cancer, respectively. Thirty-six individuals had a family history of breast/ovarian cancer in first-degree relatives. Fifty-seven patients (24.6%) tested positive for BRCA mutation (41 BRCA1, 16 BRCA2). The mean BRCAPRO and Myriad scores for all patients were 6.4% and 7.7%, respectively. The scores were significantly higher for patients with positive BRCA mutation status (29.0% vs. 6.1%, P<0.001, 12.1% vs. 7.7%, P<0.001, respectively). For all patients, the respective areas under the receiver operating characteristics curves were 0.720 and 0.747 for the BRCAPRO and Myriad models to predict the risk of carrying a BRCA mutation. Both models overestimated the mutation probability in patients with a family history of breast/ovarian cancer (1.55-fold and 1.50-fold, respectively) and underestimated the probability in patients without a family history (both, 0.54-fold).
CONCLUSION: BRCAPRO and Myriad seem to be acceptable risk assessment tools for determining the risk of carrying BRCA mutations in Korean ovarian cancer patients.