A prioritization framework for BRCA1/2 variants of uncertain significance identified by comprehensive genomic profiling

Comprehensive genomic profiling (CGP) has significantly advanced cancer genomics by enabling broad detection of clinically relevant genomic alterations across diverse cancers. In the context of BRCA1/2, CGP has expanded analysis beyond conventional testing for hereditary breast and ovarian cancer (HBOC), thereby identifying otherwise unrecognized variants. Nevertheless, the high prevalence of variants of uncertain significance (VUS) remains a major obstacle to clinical implementation. To address this challenge, we analyzed 2172 CGP tests performed at Hiroshima University Hospital and affiliated institutions in Japan. BRCA1/2 VUS identified through CGP were systematically prioritized using an integrative framework combining in silico prediction and functional evidence. From 526 BRCA1/2 variants, 153 were classified as VUS. Our variant prioritization filter based on ten in silico predictors narrowed these to 10 candidates, including two splice-site and eight missense variants, most of which were concordant with prior functional studies. Among these, the significance of BRCA2:c.67 G > C (p.D23H, NM_000059.4) had remained unclear. Functional analysis demonstrated exon 2 skipping consistent with loss of function, and clinical observations from two patients carrying this variant showed that therapeutic responses aligned with the biology of homologous recombination deficiency. These findings present a proof-of-concept framework for prioritizing and interpreting BRCA1/2 VUS detected in real-world CGP testing. By integrating multiple in silico predictors with functional evidence, this approach enables systematic prioritization and interpretation of BRCA1/2 VUS and may be broadly applicable to variant assessment in hereditary cancer predisposition genes.