Abstract
BACKGROUND: Germline mutations in cancer-predisposition genes are critical for clinical risk assessment and therapeutic decisions in breast cancer, yet large-scale genomic studies and population-specific tools remain limited for Asian populations. METHODS: In this prospective clinic-based cohort study, 2700 unselected Chinese breast cancer patients underwent germline sequencing for single nucleotide variations, insertion/deletions, and large genomic rearrangements using a clinically validated 32-gene panel. Multiplex ligation-dependent probe amplification was used for confirmation of copy number variations. A multivariate logistic regression model (PEEKABOO) was developed to predict mutation probability. Therapeutic impact was assessed in 632 patients receiving neoadjuvant therapy. RESULTS: The overall prevalence of deleterious germline variants was 11.4%, predominantly BRCA2 (3.7%) and BRCA1 (3.1%) with protein-truncating variants accounting for 81.8% of alterations. Mutation prevalence progressively increased across hereditary risk tiers: 4.9% for the NeoFHS-zero group, 13.1% for the NeoFHS-low group, and 19.9% for the NeoFHS-high group (p < 0.001). In HER2-negative breast cancer, germline homologous recombination repair gene mutations (gHRRm) independently predicted higher pathological complete response (pCR) rates (OR, 2.24; 95% CI, 1.09-4.66; p = 0.028). A numerically higher pCR rate was observed in gHRR-mutant TNBC patients receiving neoadjuvant immunotherapy combined with chemotherapy (80.0% vs 55.6%, p = 0.6). The PEEKABOO model exhibited strong performance in predicting mutation probability in both panel genes (area under curve [AUC], 0.73; accuracy, 57%; sensitivity/specificity, 76%/54%; PPV/NPV, 17%/95%) and BRCA1/2 (AUC, 0.80; accuracy, 62%; sensitivity/specificity, 81%/61%; PPV/NPV, 13%/98%). CONCLUSIONS: Our study establishes a unique germline mutation profile of Chinese breast cancer in a large-scale targeted sequencing cohort. Germline HRR gene mutation status is a potential biomarker for response to neoadjuvant treatment with DNA-damaging chemotherapeutics for HER2-negative breast cancer. The population-specific PEEKABOO model improves the predictive efficiency of germline mutations and represents a clinically applicable tool for risk stratification in Chinese patients.