5584 Background: Homologous recombination deficiency (HRD) is a pivotal molecular phenotype in ovarian cancer (OC) and serves as a key biomarker for guiding PARP inhibitor (PARPi) therapy. Current clinical algorithms assess HRD status by integrating pathogenic BRCA1/2 mutations and genomic scar scores. Nevertheless, a paradoxical phenotype exists: BRCA mutations with a low HRD score. The incidence and genomic characteristics of this subgroup remain poorly understood. Methods: We retrospectively analyzed a cohort of OC patients tested at a centralized laboratory (2024-2025). A composite genomic instability score was calculated as the sum of three biomarkers: loss of heterozygosity (LOH), telomeric allelic imbalance (TAI), and large-scale state transitions (LST). HRD positivity was defined as the presence of a BRCA1/2 pathogenic/likely pathogenic variant and/or a HRD score (LOH+TAI+LST) ≥ 42. Results: Among 441 ovarian cancer samples, 151 of 310 (34.24%, 151/441) BRCA wild-type patients had an HRD score ≥ 42, primarily driven by other homologous recombination repair (HRR) gene alterations. Conversely, among 131 BRCA-mutant patients, 30 (6.80%, 30/441) had an HRD score < 42. Notably, 13 patients (2.95%, 13/441) with pathogenic BRCA variants exhibited remarkably low HRD scores (< 10). This subgroup comprised 10 BRCA1 and 3 BRCA2 alterations, including: 4 somatic truncating mutations Exon11 S1389* (VAF=4.41%; HRD=0), Exon11 D1868Vfs*5 (VAF=4.80%; HRD=4), Exon10 E597* (VAF=17.37%; HRD=3), Exon10 K339Rfs*2 (VAF=30.65%; HRD=6); 6 germline truncating mutations Exon21 Q1785* (VAF=50.29%; HRD=9), Exon13 G1492* (VAF=47.61%; HRD=7), Exon10 L1086* (VAF=46.52%; HRD=2), Exon11 S1882* (VAF=50.83%; HRD=0), Exon10 Q1200* (VAF=49.37%; HRD=0), Exon11 L1908Rfs*2 (VAF=44.44%; HRD=0); and 3 germline copy number deletions (HRD=2, 2, 0). We hypothesize potential mechanisms for this discordance: 1) Monoallelic inactivation: Single-allele inactivation (first hit) without a complete loss-of-function second hit, leading to partial HR function retention. 2) Subclonal variants: Low VAF suggests BRCA mutations may reside in tumor subclones, not dominating the pan-tumor HRD phenotype, with signals diluted by HR-proficient cells. 3) Time-dependency: BRCA inactivation might occur late in tumor evolution, insufficient for accrual of HRD-associated genomic scars at testing. Conclusions: This study quantifies a rare subgroup (~3%) of OC with pathogenic BRCA mutations but low HRD scores. This suggests monoallelic BRCA mutation alone may be insufficient to induce high genomic instability, potentially contributing to primary resistance to PARPis, though validation with clinical outcome data is required. Future studies integrating multi-omics data are needed to precisely evaluate HR pathway status and guide individualized treatment decisions.
Liu et al. (Wed,) studied this question.