A Novel Epithelial-Mesenchymal Transition Gene Signature Correlated With Prognosis, and Immune Infiltration in Hepatocellular Carcinoma

Abstract

Background: Although many genes related to epithelial-mesenchymal transition (EMT) have been explored in hepatocellular carcinoma (HCC), their prognostic significance still needs further analysis.Methods: Differentially expressed EMT-related genes were obtained through the integrated analysis of 4 Gene expression omnibus (GEO) datasets. The univariate Cox regression and Lasso Cox regression models are utilized to determine the EMT-related gene signature. Based on the results of multivariate Cox regression, a predictive nomogram is established. Time-dependent ROC curve and calibration curve are used to show the distinguishing ability and consistency of the nomogram. Finally, we explored the correlation between EMT risk score and immune immunity.Results: We identified a nine EMT-related gene signature to predict the survival outcome of HCC patients. Based on the EMT risk score’s median, HCC patients in each dataset were divided into high and low-risk groups. The survival outcomes of HCC patients in the high-risk group were significantly worse than those in the low-risk group. The prediction nomogram based on the EMT risk score has better distinguishing ability and consistency. High EMT risk score was related to immune infiltration.Conclusion: The nomogram based on the EMT risk score can reliably predict the survival outcome of HCC patients, thereby providing benefits for medical decisions.