Integrated Genomic Profiling Identifies Predictive Biomarkers for Neoadjuvant Therapy Response in Chinese Breast Cancers

Abstract Background Neoadjuvant therapy (NAT) has emerged as a standard treatment strategy for locally advanced breast cancer (BC), yet effective predictive markers are still lacking. Methods We constructed a large-scale NAT cohort of 1,161 primary BC patients, including 1,145 cases with paired clinicopathological data and targeted sequencing, to identify biomarkers of NAT efficacy. Results Systematic analysis identified pan-subtype predictors of NAT efficacy (e.g., PIK3CA mutations associated with resistance in HR+/HER2 − and triple-negative BC) and subtype-specific markers (e.g., ERBB2 and GRIN2A alterations predicting resistance in HER2-enriched tumors). In non-pathological complete response (non-pCR) patients, multiple genomic alterations (e.g., TP53 and TOP2A) were identified as independent predictors of metastatic recurrence. Furthermore, a machine learning model integrating somatic mutations and clinical features demonstrated consistent NAT-response prediction (training AUC = 0.82; validation AUC = 0.81). Conclusions Overall, our study presents a comprehensive genomic atlas of NAT responsiveness in Asian populations, providing molecular guidance for personalized treatment regimens that may enhance clinical outcomes.