Abstract
PURPOSE: This study aims to investigate the predictive value and optimal threshold of tumor-infiltrating lymphocytes (TILs) for neoadjuvant treatment response and long-term prognosis in breast cancer patients. METHODS: Patients with primary breast cancer who were diagnosed and received neoadjuvant chemotherapy (NAC) at Ruijin Hospital affiliated to Shanghai Jiao Tong University School of Medicine and Quanzhou First Hospital affiliated to Fujian Medical University between January 2013 and December 2023 were included. The assessment of TILs was performed on pre-NAC breast tumor tissue according to the guidelines of the International Immuno-Oncology Biomarker Working Group. Restricted cubic spline (RCS) regression was used to explore the potential nonlinear relationships between the continuous variable TILs and pathological complete response (pCR) as well as prognosis. Logistic regression was used to examine the association between TILs and pCR, and Cox proportional hazards regression assessed the effects on breast cancer-free interval (BCFI) and overall survival (OS). RESULTS: In this study, a total of 424 patients were included in the analysis. Median follow-up time was 95 months. RCS analysis indicated that a TILs threshold of 10% is the optimal cutoff for predicting pCR in this cohort. Among the participants, 147 patients (34.7%) exhibited high TIL expression, defined as > 10%. Notably, the pCR rate was significantly higher in patients with Elevated TIL levels, achieving 29.3% compared to only 8.7% in those with lower TIL levels (p < 0.001). The odds ratio (OR) for achieving pCR in patients with high TILs was 0.29, with a 95% confidence interval (CI) of 0.16 to 0.52 (p < 0.001). Furthermore, multivariate analysis revealed that patients with low TIL levels are at a substantially increased risk of breast cancer recurrence, with a hazard ratio (HR) of 2.36 (95% CI: 1.47–3.80, p < 0.001). Univariate Cox regression analysis showed that low TIL expression significantly compromised OS (HR: 2.22, 95% CI: 1.17–4.19, p = 0.014), while a trend towards worse in multivariate analysis. In addition, high TIL levels were associated with improved BCFI and OS in triple-negative breast cancer (TNBC) patients; however, no significant relationship was observed in hormone receptor-positive, HER2-negative tumors. CONCLUSIONS: TILs could serve as an independent predictor of pCR. Our results also support TILs as long-term prognostic predictors for TNBC and HER2-positive breast cancers, but not for HR + HER2- subtype.