CD45RO(+)记忆T细胞作为非小细胞肺癌患者预后标志物的研究

BACKGROUND AND OBJECTIVE: Lung cancer is the most common malignancy world-wide. There are a variety of immune infiltrating cells in tumor microenvironment, which is an important component of tumor immunity and has clinical significance for the prognosis of patients. CD45RO is a surface marker of memory T cells. The expression of CD45RO(+) tumor infiltrating lymphocytes (TILs) is associated with the prognosis of many tumors. The purpose of this study was to evaluate the relationship between the density of CD45RO(+) TILs in tumor and stromal area and the clinical characteristics of patients with non-small cell lung cancer (NSCLC) and its impact on the prognosis of patients. We aimed to explore the clinical value of CD45RO(+) TILs and programmed cell death ligand 1 (PD-L1) as prognostic markers. METHODS: Multiple fluorescent immunohistochemical staining was used to stain the tissue microarray chips of 167 patients with NSCLC, marking CD45RO, cytokeratin (CK) and PD-L1. Using artificial intelligence image recognition technology and tumor cell-specific CK staining, divide the tumor and stromal area in the tissue, evaluate the density of CD45RO(+) TILs in the tumor and stromal area, and the expression level of PD-L1 in tumor cells. The non-parametric test was used to analyze the relationship between CD45RO(+) TILs and the clinical characteristics of patients, and the Kaplan-Meier method and Cox risk ratio model were used to analyze the relationship between CD45RO(+) TILs independently or in combination with PD-L1 and tumor prognosis. RESULTS: The density of CD45RO(+) TILs was significantly associated with patient age, smoking, tumor stage, and pathological type. Single-factor survival analysis showed that NSCLC (P=0.007) stromal region and lung adenocarcinoma (LUAD) (P < 0.001) with CD45RO(+) TILs high density had better OS. Multivariate survival analysis showed that the high density of CD45RO(+) TILs in the stromal region of NSCLC (HR=0.559, 95%CI: 0.377-0.829, P=0.004) and lung adenocarcinoma (HR=0.352, 95%CI: 0.193-0.641, P=0.001) were independent prognostic factors for overall survival time (OS). Combined with PD-L1 score of tumor cells in tumor tissues and infiltration score of CD45RO(+) TILs in all tumor tissues, the patients were divided into 4 groups: patients with PD-L1(+)/CD45RO(+) had the longest disease-free survival (DFS) time, and patients with PD-L1(+)/CD45RO(-) had the shortest DFS time. Multivariate Cox regression analysis showed that PD-L1(+)/CD45RO(-) was an independent prognostic factor for DFS and had a higher risk of poor prognosis compared to the other three groups (HR=2.221, 95%CI: 1.258-3.919, P=0.006). CONCLUSION: In tumor tissues, the density of CD45RO(+) TILs, as well as the combination of CD45RO(+) TILs and PD-L1 in tumor areas, significantly correlated with clinicopathological features and prognosis of NSCLC, which can be used as a new prognosis marker.