Development and Verification of a Combined Immune- and Metabolism-Related Prognostic Signature for Hepatocellular Carcinoma

Immune escape and metabolic reprogramming are becoming important characteristics of tumor biology, which play critical roles in tumor initiation and progression. However, the integrative analysis of immune and metabolic characteristics for the tumor microenvironment in hepatocellular carcinoma (HCC) remains unclear. Herein, by univariate and least absolute shrinkage and selection operator (LASSO) Cox regression analyses, a prognostic signature associated with tumor microenvironment was established based on five immune- and metabolism-related genes (IMRGs), which was fully verified and evaluated in both internal and external cohorts. The C-index was superior to previously published HCC signatures, indicating the robustness and reliability of IMRGs prognostic signature. A nomogram was built based on IMRGs prognostic signature and various clinical parameters, such as age and T stage. The AUCs of nomogram at 1-, 3-, and 5-year (AUC = 0.829, 0.749, 0.749) were slightly better than that of IMRGs signature (AUC = 0.809, 0.734, 0.711). The relationship of risk score (RS) with immune checkpoint expressions, immunophenoscore (IPS), as well as microsatellite instability (MSI) together accurately predicted the treatment efficacy. Collectively, the IMRGs signature might have the potential to better predict prognostic risk, evaluate immunotherapy efficacy, and help personalize immunotherapy for HCC patients.

Automated summary

Immune escape and metabolic reprogramming are important characteristics of tumor biology in hepatocellular carcinoma (HCC). This study establishes a prognostic signature based on immune- and metabolism-related genes (IMRGs) to predict prognosis, evaluate immunotherapy efficacy, and personalize immunotherapy for HCC patients. The signature shows superior performance compared to previously published HCC signatures and is validated in internal and external cohorts. A nomogram combining the IMRGs signature with clinical parameters is also developed to predict prognostic risk.