Abstract
PURPOSE: In cancers with both high microsatellite instability (MSI-H) and BRCA1/2 mutations, BRCA1/2 mutations may be incidentally caused by MSI and represent passenger mutations versus drivers of those cancers. Reporting of these mutations without additional clarification may result in poly (ADP-ribose) polymerase (PARP) inhibitor therapy, where there is not likely true benefit. The purpose of this work was to identify BRCA1/2 passenger mutation hotspots that are secondary to MSI-H status rather than truly independent driving mutations. METHODS: We analyzed over 100,000 pancancer patient cases in cBioPortal with both BRCA1/2 mutations and MSI-H status to identify passenger mutation hotspots that recur in microsatellite sites. We validated these hotspots as likely reflective of MSI-H status using a data set of nearly 20,000 patient cases from the University of Washington. RESULTS: We identified six recurrent frameshift passenger mutation hotspots in BRCA1 (K339fs and K654fs) and BRCA2 (I605fs, W1692fs, N1784fs, and T3033fs). These hotspots represented 17% and 21% of all truncating BRCA1/2 mutations detected in the cBioPortal and University of Washington data sets, respectively, and were seen almost exclusively in MSI-H tumors. These hotspots had a mean variant allele fraction of 17%, supporting their occurrence as passenger mutations. All hotspots are annotated in ClinVar as pathogenic variants, and all but one in catalogue of somatic mutations in cancer as somatic hotspots. Current annotations do not mention MSI. CONCLUSION: Our findings emphasize the need for molecular pathology laboratories and clinical variant databases to annotate BRCA1/2 passenger mutation hotspots with more context to interpret their pathogenicity in the setting of concurrent MSI. Identification and annotation of such hotspots will improve how oncology providers guide patients regarding therapeutic options.