Limited value of currently used germline brca mutations predictive tools in prostate cancer

Abstract

Background: Germline BRCA1 and BRCA2 mutations have been associated to poor prostate cancer (PrCa) outcomes and may have implications for cancer treatment. Identification of these carriers would also serve for the early identification of other family members at increased risk of breast and ovarian cancer. Several tools have been developed to estimate the probability of a gBRCA mutations in the context of a family history of breast and/or ovarian cancer but its performance has not been evaluated in PrCa patients. Methods: This is single-centre study aimed to: 1) compare the contribution of PrCa to identify families known to harbour a germline BRCA1 or BRCA2 mutation; and 2) estimate the ability of predicting a BRCA mutation in these families based on the cancer history at the time of PrCa diagnosis. A comprehensive reassessment of families attending our Familial Cancer screening program atMÁlaga Univ. Hospitals between 2012-16 identified 104 families known to harbour a gBRCA mutations. gBRCA mutation risk estimations were calculated with 2 commonly used risk assessment models: BRCAPRO 6.0 and Manchester Score (MS). Results: Finally, a total of 98 families (42 BRCA1, 56 BRCA2) were included in the study, after exclusion for further analyses of families with PrCa cases in non-carriers (phenocopies). As expected, PrCa was more common in BRCA2 carriers (2 vs 19, p=0.002). Median age of PrCa diagnosis was 70 yrs (48-83). Male breast cancer was more common in families with PrCa (24% vs 4% p=0.003), particularly in BRCA2 families (26.3% vs 5.4%, p=0.023), but no other differences in family history of cancer were observed between families with or without PrCa cases and therefore their scores using BRACAPRO and MS did not differ. A≥10% probability of finding a BRCA2 mutation was identified in 47% of families using BRCAPRO, decreasing to 21% when the proband was the PrCa patient (p=0.002). Similar results were observed when the probability was calculated usingMS (42% vs 21%, p=0.011) Conclusions: The currently available predictive tools underestimate the probability of a BRCA mutation when the proband is a prostate cancer patient and should not be used as unique tools to decide which PrCa patients should undergo genetic testing.