A New Gene Signature Based on Pyroptosis for Predicting the Overall Survival of Hepatocellular Carcinoma

Abstract

Background: Pyroptosis is currently being explored for anti-tumor activity. However, few studies focus on the prognostic value of pyroptosis in hepatocellular carcinoma (HCC). Objective: To develop and characterize an effective gene signature capable of anticipating the prognosis of HCC, based on pyroptosis. Methods: Initially, the RNA (ribonucleic acid) sequencing data for HCC, along with its reciprocal clinical data, was taken from Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA). Following this, pyroptosis-related genes (PRGs) were selected from the data of the TCGA cohort. Those PRGs having a differential expression between normal and HCC samples were selected and identified, using consensus clustering. Subsequently, differentially expressed genes (DEGs) in different clusters associated with pyroptosis were identified. At last, a gene signature was constructed, and its independent predicting power was illustrated using the Cox regression analysis, followed by functional analysis and single sample gene set enrichment analysis (ssGSEA). External validation of the gene signature was done using patient data obtained from the GEO dataset. Results: Forty-two differentially expressed PRGs were identified and utilized to obtain two pyroptosis-related clusters from the patient data of the TCGA cohort. A total of 2667 DEGs were distinguished between two clusters (C1 and C2) (p < 0.05). Fourteen DEGs were utilized for constructing a prognostic gene signature capable of quantifying each patient. The median risk score (3.5) of the gene signature was used as a basis to classify the patients into two groups: high- (81 patients) and low-risk groups (81 patients). A comparison of survival rate between the two risk groups showed a significantly lower overall survival rate for the high-risk group (p < 0.001). The area under the ROC (receiver operating characteristic) curve values of the gene signature were 0.843, 0.798, and 0.752 over one, three, and five years, respectively. A functional study exhibited that the 14-DEGs presented diverse complement and coagulation cascades. The ssGSEA demonstrated that different risk groups presented diverse immunity conditions. All the above results have been externally verified. Conclusions: A new gene signature based on pyroptosis can serve as an effective prognostic tool for predicting the survival of HCC patients. Pyroptosis may be related to both complement and immunity in HCC.