Zeyu Zhang,1 Hongxi Zhao,1 Pengyu Wang,2 Xueyan Geng,1 Maopeng Yin,1 Yingjie Liu,1 Shoucai Zhang,1 Yongyuan Liang,1 Jian Ji,1 Guixi Zheng1,3 

1Department of Clinical Laboratory, Qilu Hospital of Shandong University, Jinan, Shandong, 250012, People’s Republic of China; 2Faculty of Science, University of Alberta, Edmonton, Alberta, T6G 2E9, Canada; 3Shandong Engineering Research Center of Biomarker and Artificial Intelligence Application, Jinan, Shandong, 250012, People’s Republic of China

Correspondence: Guixi Zheng, Email zhengg@sdu.edu.cn

Background: This study aimed to establish an immune-glycolysis-related prognostic signature (IGRPS) to predict hepatocellular carcinoma (HCC) outcomes. Additionally, it explored the role of this signature in the tumor immune microenvironment (TIME), glycolytic pathways, and immunotherapy.
Methods: We analyzed RNA-seq, single-cell sequencing, and immune- and glycolysis-related gene datasets from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO). Using weighted gene co-expression network analysis (WGCNA), F-test, and Cox regression, we identified key survival-related immune and glycolytic genes (SRIGRGs) and developed an IGRPS through multivariate Cox regression. The IGRPS’s predictive performance was validated in training and validation cohorts using Kaplan-Meier survival analysis, receiver operating characteristic (ROC) curves, and a prognostic nomogram. Its correlation with TIME and its ability to predict immunotherapy outcomes were also assessed. In vitro experiments were conducted to analyze the expression and function of IGRPS genes in HCC.
Results: Thirteen SRIGRGs were identified for constructing the IGRPS. Patients with low-risk scores had significantly longer survival times. The area under the curve (AUC) for ROC curves was over 0.73 for training and 0.7 for validation cohorts, with C-indices of 0.721 and 0.79, respectively. IGRPS was confirmed as an independent prognostic indicator. Patients in the low-risk group showed better responses to combined anti-CTLA4 and anti-PD-1 therapies. In vitro experiments indicated that PRKAG1 and B3GAT3 were upregulated, enhancing glycolysis and promoting HCC cell proliferation and migration.
Conclusion: The IGRPS, based on immune- and glycolysis-related genes, effectively predicted prognosis and immunotherapy responses in HCC patients.

Keywords: HCC, glycolysis, tumor immune microenvironment, prognostic, immunotherapy