Comprehensive Analysis of Immune Infiltration Identifies Predictors for Immunotherapy Efficacy in Hepatocellular Carcinoma

Abstract

Background The aim of this study was to determine whether differences in the cellular composition of the immune infiltrate in HCC influence survival and identify predictors for immunotherapy efficacy in hepatocellular carcinoma. Methods A total of 362 patients from TCGA cohort and 204 patients from ICGC with HCC were included in the study. Two immune features were selected out of 24 immune features to construct immunotypes based on the Cox regression model. Hub genes of DEGs were identified by STRING and Cyto-scape. The role of hub genes on immunotherapy efficacy prediction was evaluated by Kaplan–Meier survival analysis in immunotherapy cohorts. The effects of LCK on HCC cell proliferation and migration were evaluated by CCK8, trans-well and wound healing assays. Results Eight immune cell subsets were associated with HCC prognosis. Two immune cells (MAIT and central-memory) were selected to construct 3 immunotypes which could predict overall survival in the TCGA cohort (X² = 24.13, P < 0.0001) and ICGC cohort (validation cohort, X² = 10.51, P = 0.005). GO and GSEA analysis showed up-regulated immune-related pathway in Cluster3, and Cluster3 showed significantly higher immune checkpoint molecules (PD-L1, PD-1, CTLA-4, PD-L2, LAG3 and TIM3) expression. Three hub genes (CCR5, CCR7 and LCK) were identified based on the differential expression genes between Cluster3 and Cluster1. CCR5, CCR7 and LCK were efficient predictors for immune infiltration, especially CTL, and immunotherapy efficacy. We also verified that LCK conferred proliferation and metastasis of HCC cells and immunotherapy resistance of HCC patients. Conclusion Immune cell abundance and immunotypes could effectively predict prognosis of HCC. Furthermore, CCR5, CCR7 and LCK were identified as predictors for immunotherapy efficacy in hepatocellular carcinoma.