Six Genes Involved in Prognosis of Hepatocellular Carcinoma Identified by Cox Hazard Regression

Qinghong Dai,Tao Liu,Yongchao Gao,Honghao Zhou,Xiong Li,Wei Zhang
DOI: https://doi.org/10.1186/s12859-021-04095-7
2021-01-01
Abstract:Background Hepatocellular carcinoma (HCC), derived from hepatocytes, is the main histological subtype of primary liver cancer and poses a serious threat to human health due to the high incidence and poor prognosis. This study aimed to establish a multigene prognostic model to predict the prognosis of patients with HCC. Results Gene expression datasets (GSE121248, GSE40873, GSE62232) were used to identify differentially expressed genes (DEGs) between tumor and adjacent or normal tissues, and then hub genes were screened by protein-protein interaction (PPI) network and Cytoscape software. Seventeen genes among hub genes were significantly associated with prognosis and used to construct a prognostic model through COX hazard regression analysis. The predictive performance of this model was evaluated with TCGA data and was further validated with independent dataset GSE14520. Six genes (CDKN3, ZWINT, KIF20A, NUSAP1, HMMR, DLGAP5) were involved in the prognostic model, which separated HCC patients from TCGA dataset into high- and low-risk groups. Kaplan-Meier (KM) survival analysis and risk score analysis demonstrated that low-risk group represented a survival advantage. Univariate and multivariate regression analysis showed risk score could be an independent prognostic factor. The receiver operating characteristic (ROC) curve showed there was a better predictive power of the risk score than that of other clinical indicators. At last, the results from GSE14520 demonstrated the reliability of this prognostic model in some extent. Conclusion This prognostic model represented significance for prognosis of HCC, and the risk score according to this model may be a better prognostic factor than other traditional clinical indicators.
What problem does this paper attempt to address?