Multiplex immunohistochemistry defines the tumor immune microenvironment and immunotherapeutic outcome in CLDN18.2-positive gastric cancer

BACKGROUND: The FAST study identified claudin-18 (CLDN18.2) as a promising novel therapeutic target for gastric cancer (GC). However, the tumor immune microenvironment and clinicopathological features of CLDN18.2-positive GC are unclear, making it difficult to develop and optimize CLDN18.2-targeted treatments. METHODS: This study included 80 GC patients, 60 of whom received anti-PD-1/PD-L1 treatment. CD4/CD8/CD20/CD66b/CD68/CD163/PD-1/PD-L1/TIM-3/LAG-3/FoxP3/CTLA-4/HLA-DR/STING, and CLDN18.2 were labeled using multiplex immunohistochemistry (m-IHC) to decipher the rate and spatial distribution of T cells, B cells, macrophages, and neutrophils in formalin-fixed, paraffin-embedded tumor tissues isolated from these patients. Tumor immune-microenvironmental features and patient survival stratified by CLDN18.2 expression were analyzed using two independent-sample t-tests and log-rank tests, respectively. RESULTS: We considered moderate-to-strong CLDN18.2 expression ≥ 40% of tumor cells as the cut-off for positivity. The proportion of CD8(+)PD-1(−), CD8(+)LAG-3(−), and CD8(+)TIM-3(−) T cells was significantly higher in CLDN18.2-positive tumors than in negative tumors (0.039 vs. 0.026, P = 0.009; 0.050 vs.0.035, P = 0.024; 0.045 vs. 0.032, P = 0.038, respectively). In addition, the number of neutrophils (CD66b(+)) was higher in the CLDN18.2-positive group than in the negative group (0.081 vs. 0.055, P = 0.031, respectively), while the rates of M1 (CD68(+)CD163(−)HLA-DR(+)), M2 macrophages (CD68(+)CD163(+)HLA-DR(−)), and B cells (CD20(+)) were comparable between the CLDN18.2-positive and negative groups. The average numbers of CD8(+)PD-1(−), CD8(+)LAG-3(−), and CD8(+)TIM-3(−)T cells surrounding tumor cells within a 20-μm range were higher in CLDN18.2-positive tumors than in the CLDN18.2-negative tumors (0.16 vs. 0.09, P = 0.011; 0.20 vs. 0.12, P = 0.029; 0.18 vs. 0.12, P = 0.047, respectively). In addition, in the CLDN18.2-positive group, tumor cells surrounded by CD8(+)PD-1(−), CD8(+)LAG-3(−) T cells, or M1 macrophages within a 20-μm range accounted for a higher proportion of all tumor cells than those in the CLDN18.2-negative group (10.79% vs. 6.60%, P = 0.015; 12.68% vs. 8.70%, P = 0.049; 9.08% vs. 6.56%, P = 0.033, respectively). These findings suggest that CLDN18.2-positive GC harbors complex immune-microenvironmental features. Additionally, CLDN18.2-positive group had shorter OS and irOS than CLDN18.2-negative group (median OS: 23.33 vs.36.6 months, P < 0.001; median irOS: 10.03 vs. 20.13 months, P = 0.044, respectively). CONCLUSIONS: CLDN18.2-positive GC displayed unique immune-microenvironmental characteristics, which is of great significance for the development of CLDN18.2-targeted therapies. However, the impact of CLDN18.2-related microenvironmental features on prognosis requires further investigation. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12916-022-02421-1.