Answers are grounded in this paper and cite their sources.
Retrospective analysis reveals low prevalence of CGP-undetectable germline BRCA variants in advanced cancer, suggesting clinical implications.
161 Background: BRACAnaylsis in Japan is widely used as a companion diagnostic for PARP inhibitors in multiple cancer types, including pancreatic cancer. Comprehensive genomic profiling (CGP) has also become increasingly utilized in clinical practice, and some platforms evaluate germline variants using matched tumor–normal analysis. However, CGP does not assess germline structural variants, and therefore cannot fully substitute for BRACAnalysis. The prevalence and clinical impact of CGP-undetectable germline BRCA pathogenic variants (GPVs) in patients with advanced cancer remain unclear. Methods: We retrospectively analyzed patients with pancreatic, breast, ovarian, and prostate cancers who underwent BRACAnalysis at our institution. Among patients with pathogenic BRCA variants, we evaluated variant characteristics, the frequency of CGP-undetectable GPVs, and clinical features. CGP-undetectable GPVs were defined as variants considered technically or procedurally undetectable by CGP, such as copy number alterations, rearrangements, or large insertions/deletions. Results: Between September 2018 and September 2025, 2,869 patients underwent BRACAnalysis, and 257 patients (9.0%) were identified as BRCA GPV carriers. Among them, 7 patients (2.7%) had CGP-undetectable GPVs, including 5 BRCA1 and 2 BRCA2 variants. Cancer types included breast cancer (n = 3), ovarian cancer (n = 3), and pancreatic cancer (n = 1). The median age was 56 years, and all patients had a family history of cancer within second-degree relatives. No patients with CGP-undetectable GPVs underwent CGP testing. Conclusions: The prevalence of CGP-undetectable germline BRCA GPVs was low and predominantly observed in BRCA1 . When matched pair CGP panel is used as a substitute for BRACAnalysis in germline BRCA evaluation, approximately of pathogenic variants may be missed.
No takes yet. Share an insight, caveat, or question.
Hakui et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: