Abstract
Background: Triple-negative breast cancer (TNBC) is an aggressive subtype with limited targeted therapies. Tumor-infiltrating lymphocytes (TILs) are promising prognostic and predictive biomarkers, but dynamic changes during neoadjuvant chemotherapy (NACT) and immunotherapy remain underexplored. We investigated whether longitudinal TIL dynamics improve prediction of pathological complete response (pCR) and survival outcomes in TNBC patients receiving NACT with or without immune checkpoint inhibitors.Methods: We retrospectively analyzed 248 TNBC patients treated with NACT (platinum-based) at two academic centers (2015–2023). Stromal TILs (sTILs) were assessed at baseline, after 2 cycles, and at surgery using standardized guidelines. CD8+ TIL density was quantified by immunohistochemistry. Neutrophil-to-lymphocyte ratio (NLR) was obtained from blood counts. Logistic regression and Cox models evaluated associations with pCR and event-free survival (EFS). A composite score integrating TIL dynamics and NLR was developed.Results: Baseline sTILs ≥20% were associated with higher pCR rates (OR 2.45, p=0.003). An increase in sTILs ≥10% during NACT independently predicted pCR (OR 3.12, pConclusions: Longitudinal TIL dynamics, combined with NLR, enhance prediction of immunotherapy response in TNBC. Incorporating dynamic immune biomarkers into clinical decision-making may optimize treatment selection and monitoring.