A novel four-gene signature for predicting the prognosis of hepatocellular carcinoma.

ABSTRACT

OBJECTIVE: To identify and utilize gene signatures for the prognostic evaluation of postoperative patients with hepatocellular carcinoma (HCC). METHODS: The gene mRNA expression profiles and corresponding clinicopathological data of postoperative patients with HCC were downloaded from The Cancer Genome Atlas (TCGA) database. Highly differentially expressed genes (DEGs) in tumor tissues compared to adjacent tissues were identified, and their associations with the overall survival (OS) of HCC patients were analyzed. The strongly associated genes were used to develop a prognostic score for the survival stratification of HCC, and the underlying mechanisms were analyzed using bioinformatics. RESULTS: A total of 376 DEGs were identified and four DEGs (ADH4, COL15A1, RET and KCNJ16) were independently associated with OS. A prognostic score derived from the four genes could effectively stratify HCC patients with different OS outcomes, independent of clinical parameters. Patients with high scores exhibited poorer OS than patients with low scores (HR 5.526, 95% CI 2.451-12.461, p < .001). The four genes were involved in cancer-related biological processes and were independent of each other in bioinformatics analyses. CONCLUSION: Four genes strongly associated with the prognosis of postoperative patients with HCC were identified, and the derived prognostic score was simple and valuable for overall survival prediction.