The Landscape of Immune Cells Indicates Prognosis and Applicability of Checkpoint Therapy in Hepatocellular Carcinoma

Background: Tumor-infiltrating immune cells are important components of tumor microenvironment (TME), and their composition reflects the confrontation between host immune system and tumor cells. However, the relationship between the composition of infiltrating immune cells, prognosis and the applicability of anti-PD-1/PD-L1 therapy in hepatocellular carcinoma (HCC) needs systematic examination. Methods: Cell-type Identification By Estimating Relative Subsets Of RNA Transcripts (CIBERSORT) was applied to evaluate the infiltration of immune cells based on The Cancer Genome Atlas (TCGA) liver hepatocellular carcinoma (LIHC) cohort. Diagnostic and prognostic models were constructed based on immune cells, and the models were validated by two external cohorts. The relationship between immune cells and PD-L1 was evaluated by Spearman correlation, and the finding was validated in our in-house HCC sample. Result: Patients in TCGA LIHC cohort were classified into 6 subtypes with different prognosis based on the proportion of tumor-infiltrating immune cells simulated via CIBERSORT. Among 22 types of immune cells, intratumoral PD-L1 mRNA level exhibited linear relationship with the fraction of 5 types of immune cells (M1 macrophages, plasma cells, CD8+ T cells, resting mast cells, and regulatory T cells), and M1 macrophages showed the strongest relevance (R = 0.26, p < 0.001). Immunohistochemistry of our in-house HCC specimens verified this conclusion. Moreover, intratumoral mRNA levels of M1 macrophage-associated cytokines were positively correlated with PD-L1 level. Conclusions: Our study demonstrated that the prognosis of HCC patients was associated with the pattern of infiltrating immune cells in TME, and macrophage-associated cytokines might be a potential non-invasive marker for predicting the PD-L1 level for HCC patients.