Integrating traditional biomarkers and emerging predictors to assess neoadjuvant chemotherapy efficacy in breast cancer: a multifactorial analysis of Ki-67, CDK4, EGFR, TILs and ctDNA.

Abstract

Objective: This study aimed to analyse the correlation between the expression of cell proliferation-associated antigen (Ki-67), cell cycle protein-dependent kinase 4 (CDK4), epidermal growth factor receptor (EGFR), tumour-infiltrating lymphocytes (TILs) and circulating tumour DNA (ctDNA) with the outcome and prognosis of patients with breast cancer (BC) undergoing neoadjuvant chemotherapy (NACT).

Methods: We retrospectively analysed the clinicopathological data of 231 patients with BC who underwent preoperative NACT at XX Hospital between 1 January 2018 and 31 December 2021. Logistic regression models were used to analyse factors influencing NACT efficacy. The Cox risk regression model was used to analyse prognostic factors. The TILs were assessed on pre-treatment biopsies, and ctDNA levels were monitored during NACT. Propensity score matching and subgroup analyses were performed.

Results: After 4-6 cycles of chemotherapy, the response rate was 77.92% (180/231), with 58.87% (136/231) achieving pathological complete response (pCR). Multifactorial analysis showed that tumour, node and metastasis (TNM) stage II, EGFR positivity, low Ki-67 expression, CDK4 negativity, non-triple-negative subtypes and effective NACT results were associated with higher pCR rates. Higher TIL levels correlated with increased pCR rates (72.4% for high TILs vs 39.1% for low TILs, p < 0.001). The ctDNA levels decreased significantly in patients with pCR compared with patients without pCR during NACT (p < 0.001). After propensity score matching, the 3-year disease-free survival rate was significantly higher in the pCR group (88.9% vs 71.1%, p = 0.003). Subgroup analysis revealed varying pCR rates and predictive biomarkers across BC subtypes.

Conclusion: The TNM classification, EGFR, Ki-67, CDK4 expression, BC subtype and NACT results have predictive value for pCR in patients with BC. Lower TNM classification, lower Ki-67 expression and EGFR positivity are associated with better outcomes. High TIL levels and significant decreases in ctDNA during NACT correlate with improved response and prognosis. These findings highlight the potential for integrating traditional clinicopathological factors with novel biomarkers for personalised treatment strategies in BC.