Construction of a liver hepatocellular carcinoma risk model based on fumarate metabolism-associated genes to predict prognosis and immunological landscape.

Liver hepatocellular carcinoma (LIHC) continues to present a serious global health burden, with high mortality driven by late-stage detection and poor therapeutic responses. While fumarate has emerged as a key oncometabolite driving tumor progression through metabolic reprogramming and epigenetic regulation, its specific role in LIHC prognosis warrants further investigation to develop novel therapeutic strategies. We constructed and validated a prognostic signature for LIHC based on fumarate metabolism-associated genes (FAMRGs) using data from TCGA and GEO. After screening prognosis-related FAMRGs using differential expression and univariate Cox regression, we built a four-gene model via LASSO and multivariate Cox regression. Model performance was assessed by survival analysis and ROC curves. Immune infiltration was characterized by ssGSEA, CIBERSORT, and ESTIMATE; immunotherapy response was predicted using TIDE scores and the IMvigor210 cohort. Drug sensitivity was estimated with pRRophetic and CellMiner, and molecular subtypes were identified by consensus clustering. This study constructed and validated a prognostic model for LIHC comprising four FAMRGs. The model effectively distinguished between high- and low-risk patients, with the low-risk group demonstrating longer overall survival, more active CD8⁺ T cell infiltration, and lower TIDE scores, suggesting potential heightened sensitivity to immunotherapy. Moreover, distinct risk-group patterns were found in immune features, mutational load, drug responses, and molecular classifications, thereby advancing our understanding of precision medicine for LIHC. The FAMRGs-based prognostic model presented here not only deciphers immune heterogeneity and immunotherapy sensitivity in LIHC but also furnishes a theoretical scaffold for precision medicine approaches.