BACKGROUND: Amino acid metabolism (AAM) reprogramming plays a crucial role in hepatocellular carcinoma (HCC), but its genetic pathophysiology was not fully elucidated. Therefore, we employed a summary data-based Mendelian randomization (SMR) approach to identify putative causal effects of the AAM-related genes on hepatitis B virus (HBV)-HCC survival via integrating multi-omics data. METHODS: Multivariate Cox proportional hazards regression models were used to evaluate associations between genetic variants of AAM-related genes and overall survival (OS) of HBV-HCC patients (n = 866). Next, we developed a pathway-specific genetic risk score (GRS) comprising variants in the AAM pathway. Subsequently, putative causal SNPs were prioritized using SMR by integrating HBV-HCC OS data with expression quantitative trait loci (eQTLs) and DNA methylation QTLs (mQTLs) from the blood, as well a eQTLs of liver tissues. RESULTS: We identified 23 independent variants associated with HBV-HCC OS, and the pathway-specific GRS derived from the identified variants was a significant predictor of HBV-HCC OS. The addition of the GRS significantly improved the predictive performance of the 5-year survival model (AUC increased from 72.04% to 84.67%, P <
0.001). By integrating HBV-HCC OS associated with the eQTLs and mQTLs from the blood, we identified a putative causal variant rs2074038 across HBV-HCC OS, ACCS expression, and DNA methylation. Furthermore, the integration of liver eQTL data revealed that increased expression levels of ACCS by rs2074038 were associated with a worse HBV-HCC OS. Mechanistically, bioinformatics annotation and luciferase reporter assays further demonstrated that rs2074038 contributes to HBV-HCC progression by allele-specific regulation of the ACCS expression. CONCLUSIONS: This study identified rs2074038 as a novel functional SNP associated with poor HBV-HCC survival, likely mediated genetic regulation of ACCS expression. These findings suggest that ACCS is a potential therapeutic target and highlight the need for further validation in clinical settings.