Abstract 7646: Evaluation of tumor immune microenvironment after neoadjuvant immunotherapy in patients with esophageal squamous cell carcinoma (ESCC): Towards a prediction for recurrence risk

Abstract Background: Despite the promising potential of neoadjuvant immune checkpoint blockade (ICB) in ESCC, postoperative recurrence remains a critical challenge. Besides pathological response, there remains no biomarkers for recurrence risk prediction after neoadjuvant ICB. Recent studies have highlighted a strong association between ICB efficacy and tumor immune microenvironment (TIME). Thus, we sought to investigate the TIME features in surgically resected tumors following neoadjuvant ICB from patients with ESCC to develop a potential way for recurrence risk prediction. Methods: This is a side study of a phase 2 NICE trial (ChiCTR1900026240) of 51 patients with resectable locally advanced ESCC who received neoadjuvant ICB plus chemotherapy followed by surgery from 2019 to 2020. TIME features of surgically resected tumors after neoadjuvant therapy were assessed by multiplex immunohistochemistry (mIHC) for immune markers (CD8, PD-L1, CD45RO, PD-1, Pan-CK, CD4, FoxP3, CD68, CD163). We first analyzed TIME features in tumors evaluated as pathological complete response (pCR) and major pathological response (MPR). Then, we divided 51 patients into a training and validation set; and machine learning was applied to identify a list of TIME features that predict recurrence-free survival (RFS) and to develop a TIME-based signature (TIS) on neoadjuvant ICB prognosis. Results: Fifty-one patients from NICE trial were used for primary analysis. A higher positive rate of CD8+ cells, CD163−CD68+ macrophages and CD8+CD45RO+ cells was observed in patients with pCR. Similarly, MPR patients exhibited a higher positive rate of CD8+ cells and CD8+CD45RO+ cells. Regarding RFS prediction, machine learning selection of TIS were identified: CD8+PD1−(cytotoxic T cells), CD4+Foxp3+ (Treg), CD8+CD45RO+ (memory T cells), and M1 macrophages (CD163−CD68+) exhibited significant predictive value. Stratifying patients into low- and high-risk groups using TIS for RFS revealed a prolonged prolonged RFS in the low-risk group (p<0.001; HR: 0.08, 95% CI [0.03-0.25]). Notably, when applying TIS to pCR patients, we identified a subset of high-risk pCR patients (median RFS = 8.9 months, 1- year RFS=32%) relapsed earlier than low-risk pCR patients (median RFS = not reached, 1-year RFS=100%, p<0.001). Multivariable Cox regression analysis demonstrated that TIS was an independent predictor for RFS as well as overall survival (both p<0.001). Further independent validation (ICB and non-ICB cohort) of TIS is ongoing. Conclusion: The establishment of a risk prediction model based on TIME demonstrated the potential to enhance postoperative monitoring management for ESCC patients who underwent surgery after neoadjuvant ICB. Our study provides insights into development of optimal post-immunotherapy management strategy. Citation Format: Zhichao Liu, Boyao Yu, Yuxin Yang, Cong Qi, Chang Yuan, Yang Yang, Jinzhi Wei, Sidong Chen, Yang Wang, Zhigang Li. Evaluation of tumor immune microenvironment after neoadjuvant immunotherapy in patients with esophageal squamous cell carcinoma (ESCC): Towards a prediction for recurrence risk [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7646.