Identification of a Five-Immune Gene Model as an Independent Prognostic Factor in Hepatocellular Carcinoma

Abstract

Background: Hepatocellular carcinoma (HCC) is a common cancer with a poor prognosis. We purposed to identify a prognostic risk model of HCC according to the differentially expressed (DE) immune genes.Methods: The DE immune genes were identified based on 374 HCC and 50 adjacent normal samples from the Cancer Genome Atlas database. Univariate Cox analysis, Lasso regression, and multivariate Cox analysis were used to determine the immune genes used to construct the model based on the training group. The testing group and the entire group were applied for the validation of the model.Results: A five-immune gene model comprising HSPA4, ISG20L2, NDRG1, EGF, and IL17D was identified. Based on the model, overall survival was significantly different between the high-risk and low-risk groups (P = 7.953e-06). The AUC for the model at 1- and 3-year was 0.849 and 0.74, respectively. The validating groups confirmed the reliability of the model. The risk score was identified as an independent prognostic factor and was closely related to the content of immune cells from HCC samples.Conclusion: We identified a five-immune gene model, which could be treated as an independent prognostic factor of HCC.