Identification and validation of a five-gene prognostic signature for hepatocellular carcinoma

Abstract

BACKGROUND: ARID1A is a commonly mutated tumor suppressor gene found in all human cancer types, but its clinical significance, oncogenic functions, and relevant mechanisms in hepatocellular carcinoma (HCC) are not well understood. OBJECTIVE: We aimed to improving the prognosis risk classification of HCC from the perspective of ARID1A mutations. MATERIALS AND METHODS: We examined the interaction between ARID1A mutations and the overall survival via Kaplan-Meier survival analysis. We used gene set enrichment analysis (GSEA) to elucidate the influence of ARID1A mutations on signaling pathways. A prognostic model was constructed using LASSO and multivariate Cox regression analyses. A receiver operating characteristic (ROC) curve was used to estimate the performance and accuracy of the model. RESULTS: HCC patients with ARID1A mutations presented poor prognosis. By GSEA, we showed that genes upregulated by reactive oxygen species (ROS) and regulated by MYC were positively correlated with ARID1A mutations. A prognostic signature consisting of 5 genes (SRXN1, LDHA, TFDP1, PPM1G, and EIF2S1) was constructed in our research. The signature showed good performance in predicting overall survival (OS) for HCC patients by internal and external validation. CONCLUSION: Our research proposed a novel and robust approach for the prognostic risk classification of HCC patients, and this approach may provide new insights to improve the treatment strategy of HCC.