321 Background: The BLM gene encodes a RecQ DNA helicase that is involved in homologous recombination. Biallelic BLM inactivation leads to Bloom syndrome, an inherited genetic disorder marked by chromosomal instability and multiple cancer susceptibilities. Conflicting studies have suggested that heterozygous BLM mutation carriers may have an increased risk of various cancers. Here we explored the role of germline pathogenic BLM mutations in prostate cancer. Methods: Prostate cancer patients with heterozygous BLM mutations were assembled from Tulane Cancer Center (TCC), Johns Hopkins Hospital (JHH) and University of Washington (UW). BLM germline mutations were identified either through commercial germline testing (Invitae), the UW-BROCA panel, or whole-exome sequencing. Corresponding tumor tissue was analyzed by DNA sequencing for somatic alterations. Population level control data were obtained from the Genome Aggregation Database (gnomAD). Results: 6 BLM germline carriers were identified among 985 advanced prostate cancer case; 2/295 TCC patients, 2/172 JHH patients, and 2/518 UW patients. Overall, pathogenic BLM mutations were detected in 0.609% (6/985) of prostate cancer cases. All mutations were loss-of-function truncating lesions (splicing or nonsense alterations). No Ashkenazi BLM mutations were observed. The population frequency of pathogenic or likely pathogenic BLM alterations detected in gnomAD was 0.025% (31/124,589). Compared to gnomAD controls, the relative risk of BLM mutations in prostate cancer patients was 24.3 (95% CI 10.2 to 58.2; P < 0.0001). One family had a pathogenic splice variant in BLM that cosegregated with disease in three of three cases with lethal/high risk prostate cancer. Tumor DNA sequencing was possible in 5 of 6 BLM carriers; no case demonstrated LOH or additional somatic BLM mutations. Interestingly, 2/5 cases on tumor sequencing also had bi-allelic BRCA2 inactivation. Conclusions: Germline BLM mutations may play a role in prostate cancer risk. Given the role of BLM in chromosomal stability and evidence of concurrent BRCA2 inactivation in a subset of cases, larger cohorts and functional analyses will be critical for better understanding the role of BLM in prostate cancer.
Individual and cumulative effect of prostate cancer risk-associated variants on clinicopathologic variables in 5,895 prostate cancer patients
