Development and External Validation of a Novel Immune Checkpoint-Related Gene Signature for Prediction of Overall Survival in Hepatocellular Carcinoma.

Enfa Zhao,Ying Dang,Shimin Chen

doi:10.3389/fmolb.2020.620765

Abstract

Objective: The purpose of this study was to develop and validate a novel immune checkpoint–related gene signature for prediction of overall survival (OS) in hepatocellular carcinoma (HCC).Methods: mRNA expression profiles and clinical follow-up information were obtained in the International Cancer Genome Consortium database. An external dataset from The Cancer Genome Atlas (TCGA) Liver Hepatocellular Carcinoma database was used to validate the results. The univariate and multivariate Cox regression analyses were performed based on the differentially expressed genes. We generated a four-mRNA signature to predict patient survival. Furthermore, the reliability and validity were validated in TCGA cohort. An integrated bioinformatics approach was performed to evaluate its diagnostic and prognostic value.Results: A four-gene (epidermal growth factor, mutated in colorectal cancer, mitogen-activated protein kinase kinase 2, and NRAS proto-oncogene, GTPase) signature was built to classify patients into two risk groups using a risk score with different OS in two cohorts (all P < 0.0001). Multivariate regression analysis demonstrated the signature was an independent predictor of HCC. Furthermore, the signature presented an excellent diagnostic power in differentiating HCC and adjacent tissues. Immune cell infiltration analysis revealed that the signature was associated with a number of immune cell subtypes.Conclusion: We identified a four–immune checkpoint–related gene signature as a robust biomarker with great potential for clinical application in risk stratification and OS prediction in HCC patients and could be a potential indicator of immunotherapy in HCC. The diagnostic signature had been validated to accurately distinguish HCC from adjacent tissues.

Highlights

Hepatocellular carcinoma (HCC), which is characterized by a low survival rate, aggressive nature, and high metastasis potential, is the most common subtype of hepatic malignancies worldwide, accounting for ∼90% of primary liver cancers (Bray et al, 2018)
Samples with clinicopathological and follow-up information were included for survival analysis in this study, consisting of 232 HCC samples in the International Cancer Genome Consortium (ICGC) cohort and 370 HCC samples in the The Cancer Genome Atlas (TCGA) cohort, respectively
Univariate Cox analysis identified seven genes associated with survival (Figure 2B), and five genes retained after LASSO Cox regression (Figure 2C)

Summary

Introduction

Hepatocellular carcinoma (HCC), which is characterized by a low survival rate, aggressive nature, and high metastasis potential, is the most common subtype of hepatic malignancies worldwide, accounting for ∼90% of primary liver cancers (Bray et al, 2018). The great developments in radiotherapy, chemotherapy, liver transplantation, and other potentially curative treatment have revolutionized the treatment of HCC, the long-term prognosis remains poor because most HCC patients are at the late stage at the time of diagnosis and have lost the opportunity of surgical removal of their lesion (Bruix et al, 2014). High-risk HCC patients with potentially poor prognosis must be monitored, and timely and effective treatment should be taken to prolong the survival and improve the quality of life (Llovet et al, 2015). Traditional methods utilizing clinical tumor-node-metastasis (TNM) staging, vascular invasion, and other clinicopathologic parameters contribute to predict HCC prognosis (Bruix et al, 2016). Considering the great complexity and heterogeneity of HCC, the predictive ability of such models is still far from satisfying

Methods

Results

Discussion

Conclusion