APM-Related gene signature model to predict prognosis and immunotherapy response in hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is a primary liver malignancy with a dismal prognosis. This study established and validated a prognostic model based on antigen-processing and presenting machinery (APM)-related genes through Mendelian randomization and publicly available datasets. Systematic analysis revealed CXCL5, SGPP2, and GLP1R as critical prognostic biomarkers, which were subsequently integrated into a risk model. The model demonstrated significant associations with pathways linked to bile acid, fatty acid, and amino acid metabolism, alongside variations in immune cell infiltration and genomic mutations, including TTN, TP53, and MUC16. Patient stratification into high- and low-risk groups indicated that low-risk individuals exhibited reduced immune infiltration, potentially correlating with enhanced immunotherapy sensitivity. These findings offer a robust gene signature for HCC prognosis and a framework for evaluating responses to immunotherapy.