Hepatocellular carcinoma (HCC) remains a leading cause of cancer-related mortality globally, with current diagnostic methods like AFP and imaging lacking sensitivity and specificity. Despite advances in proteomics, validated biomarkers reflecting HCC’s metabolic heterogeneity are limited. This study addresses this gap by identifying novel protein biomarkers and constructing a diagnostic model to improve early HCC detection.

Quantitative proteomics using data-independent acquisition (DIA) was performed on 31 pairs of HCC and adjacent tissues (Cohort 1). Differentially expressed proteins (DEPs) were analyzed via bioinformatics (GO, KEGG) and validated through Western blot, RT-PCR, and immunohistochemistry in two independent cohorts (18 and 34 patients). Partial least squares discriminant analysis (PLS-DA) and logistic regression were employed to develop a diagnostic model.

Quantitative proteomics identified 8908 proteins, with 800 differentially expressed proteins (DEPs; 306 upregulated, 494 downregulated) significantly enriched in metabolic pathways. ADH1C and BHMT, metabolically linked to alcohol and homocysteine regulation, emerged as core biomarkers, showing consistent downregulation in HCC and correlations with the Edmondson-Steiner classification ( P < 0.05). A diagnostic nomogram integrating these biomarkers demonstrated superior accuracy (AUC=0.946, 95% confidence interval (CI), 0.897–0.994, sensitivity = 91.18%, specificity = 88.24%) compared to single-marker assessments. Independent validation confirmed concordant suppression of ADH1C and BHMT at both protein and transcriptional levels.

This study identifies metabolic dysregulation as a hallmark of HCC and proposes a robust diagnostic model with tissue-validated ADH1C and BHMT. Based on tissue-based validation, our findings pave the way for future development of liquid biopsy exploration and new insights into HCC metabolic mechanisms.