Abstract
BACKGROUND: The prognostic value of m6A-related genes in hepatocellular carcinoma (HCC) and its correlation with the immune microenvironment still requires further investigation. METHODS: Consensus clustering by m6A related genes was used to classify 374 patients with HCC from The Cancer Genome Atlas (TCGA) database. Then we performed the least absolute shrinkage and selection operator (LASSO) to construct the m6A related genes model. The International Cancer Genome Consortium (ICGC) and Gene Expression Omnibus (GEO) datasets were used to verify and evaluate the model. ESTIMATE, CIBERSORTx, the expression levels of immune checkpoint genes, and TIDE were used to investigate the tumor microenvironment (TME) and the response to immunotherapy. Gene set enrichment analyses (GSEA), tumor-associated macrophages (TAMs), and gene-drug sensitivity were also analyzed. RESULTS: By expression value and regression coefficient of five m6A related genes, we constructed the risk score of each patient. The patients with a higher risk score had a considerably poorer prognosis in the primary and validated cohort. For further discussing TME and the response to immunotherapy, we divided the entire set into two groups based on the risk score. Our findings implied that the tumor-infiltrating lymphocytes (TILs) were proportional to the risk scores, which seemed to contradict that patients with higher scores had a poor prognosis. Further, we found that the high-risk group had higher expression of PD-L1, CTLA-4, and PDCD1, indicating immune dysfunction, which may be a fundamental reason for poor prognosis. This was further reinforced by the fact that the low-risk group responded better than the high-risk group to monotherapy and combination therapy. CONCLUSION: The m6A related risk score is a new independent prognostic factor that correlates with immunotherapy response. It can provide a new therapeutic strategy for improving individual immunotherapy in HCC.