Early Ki67 change predicts prognosis and supports response-adapted therapy in breast cancer treated with neoadjuvant chemotherapy.

BACKGROUND: Early prediction of response to neoadjuvant chemotherapy (NACT) in breast cancer is critical for optimizing treatment strategies and improving outcomes. This study assessed the prognostic value of early Ki67 change (ΔKi67 %) via on-treatment core needle biopsy (CNB) in stratifying event-free survival (EFS) and informing potential treatment escalation. METHODS: In this prospective cohort study, 1388 breast cancer patients treated from 2013 to 2021 were randomly divided into training and validation sets (7:3 ratio). ΔKi67 % was calculated as the percentage change from baseline to on-treatment CNB after a median of two NACT cycles. K-means clustering determined an optimal 40 % cutoff classifying patients as poor (≤40 %) or good responders (>40 %). EFS was analyzed using Kaplan-Meier estimates, multivariable Cox models, and restricted mean survival time (RMST). RESULTS: Good responders had significantly superior 5-year EFS compared to poor responders in both training (78.8 % versus 62.4 %, p < 0.001) and validation (78.6 % versus 60.9 %, p = 0.001) sets. ΔKi67 % showed stronger stratification than imaging-based metrics in RMST analysis and remained an independent predictor after adjustment. Subgroup analyses suggested poor responders in the ER-negative/HER2-negative subgroup derived a 32.0 % 3-year EFS benefit from chemotherapy intensification. The 3-year survival benefit was 14.1 % in poor responders in the HER2-positive subtype with dual HER2 blockade, though these findings require further validation. CONCLUSION: Early ΔKi67 % change using a 40 % cutoff via on-treatment CNB is a reliable prognostic predictor supporting response-adapted treatment tailoring, particularly in ER-negative/HER2-negative and HER2-positive populations.