Abstract
BACKGROUND: Available treatments for hepatocellular carcinoma (HCC), a common human malignancy with a low survival rate, remain unsatisfactory. Macropinocytosis (MPC), a type of endocytosis that involves the non-specific uptake of dissolved molecules, has been shown to contribute to HCC pathology; however, its biological mechanism remains unknown. METHODS: The current study identified 27 macropinocytosis-related genes (MRGs) from 71 candidate genes using bioinformatics. The R software was used to create a prognostic signature model by filtering standardized mRNA expression data from HCC patients and using various methods to verify the reliability of the model and indicate immune activity. RESULTS: The prognostic signature was constructed using seven MPC-related differentially expressed genes, GSK3B, AXIN1, RAC1, KEAP1, EHD1, GRB2, and SNX5, through LASSO Cox regression. The risk score was acquired from the expression of these genes and their corresponding coefficients. HCC patients in the discovery and validation cohorts were stratified, and the survival of low-risk score patients was improved in both cohorts. Time-dependent ROC analysis indicated that the model's prediction reliability was the highest in the short term. Subsequent immunologic analysis, including KEGG, located the immune action pathway of the differentially expressed genes in the direction of the cancer pathway, etc. Immune infiltration and immune checkpoint tests provided valuable guidance for future follow-up experiments. CONCLUSION: A risk model with MRGs was constructed to effectively predict HCC patient prognoses and suggest changes in the immune microenvironment during the disease process. The findings should benefit the development of a prognostic stratification and treatment strategy for HCC.