• Integrative Immunometabolic Profiling for Predicting Immune Checkpoint Inhibitor Response in Hepatocellular Carcinoma: A Retrospective–Prospective Multi-Omics Study
  • Pouriya Katouzi,1,* Mohammadreza Akbarian Khorasgani,2 Melika Khalifeh Hadi,3
    1. Qilu Hospital of Shandong University
    2. Qilu Hospital of Shandong University
    3. Qilu Hospital of Shandong University


  • Introduction: Immune checkpoint inhibitors (ICIs) have transformed the therapeutic landscape of hepatocellular carcinoma (HCC), yet treatment responses remain highly heterogeneous, and currently available biomarkers such as PD-L1 expression provide limited predictive value. Emerging evidence underscores the pivotal role of immunometabolism—including T-cell exhaustion, metabolic reprogramming, and cytokine signaling—in shaping ICI efficacy within the tumor microenvironment. A multi-dimensional approach that integrates immune and metabolic parameters holds promise for enhancing predictive precision. To address this need, the present study aims to develop and validate a predictive model for ICI response in HCC by combining retrospective clinical and immunometabolic parameters with prospective profiling of immunometabolism-related gene expression.
  • Methods: A hybrid retrospective–prospective study design was implemented. In the retrospective arm (n = 60), clinical variables such as TNM stage, HBV status, and BMI were evaluated alongside immune markers, including PD-L1 and CD8 expression assessed by immunohistochemistry (IHC), and cytokines such as IFN-γ and IL-6 measured by ELISA. Metabolic indicators, including lactate dehydrogenase (LDH), glucose, and triglycerides, were also analyzed using archived patient records and serum samples obtained at Qilu Hospital. In the prospective arm (n = 15), fresh or frozen tumor biopsies from patients treated with immune checkpoint inhibitors (ICIs) were processed for RNA extraction, followed by cDNA synthesis and SYBR Green–based quantitative PCR to assess expression of key immunometabolic genes, including PD-L1, GLUT1, IDO1, CPT1A, TIGIT, and IFN-γ. Data integration and analysis involved the application of multivariate logistic regression, principal component analysis (PCA), and machine learning models to construct a predictive tool, with model robustness and accuracy assessed using receiver operating characteristic (ROC) curve analysis and k-fold cross-validation.
  • Results: The study aims to establish a clinically deployable biomarker panel with greater than 85% sensitivity and specificity for stratifying immune checkpoint inhibitor (ICI) responders. In addition, it is expected to generate molecular signatures that link gene expression profiles with treatment outcomes, thereby providing deeper insights into therapeutic response. Building on these findings, a decision-support tool in the form of a Stratified Immunotherapy Guidance Flowchart will be developed to facilitate personalized treatment selection for hepatocellular carcinoma (HCC). Finally, the project is designed to provide foundational data that can support the extension of immunometabolic predictive strategies to other solid tumors.
  • Conclusion: This study proposes a novel, integrative immunometabolic model for predicting ICI response in HCC, with strong translational potential in precision oncology. The dual-arm design and multi-omics integration not only enhance predictive accuracy but also bridge clinical and molecular domains, offering a data-driven approach to personalize immunotherapy decisions.
  • Keywords: Hepatocellular carcinoma; Immune checkpoint inhibitors; Immunometabolism; Multi-omics; Predictive mo