Telomere-Related Gene Risk Model for Prognosis and Immune Landscape in Hepatocellular Carcinoma

Liver hepatocellular carcinoma (LIHC) is a very aggressive kind of cancer that has a dramatic impact on the quality of life and mean survival of the patient. Consequently, a specific requirement emerges to predict the prognosis of individual patients as well as to guide the individualized therapeutic strategy in clinic. Telomere- related genes (TRGs) have recently been unraveled as key players in tumor biology and a constituent of the tumor immune microenvironment. Thus, the authors constructed a risk prediction model rooted in TRGs for the purpose of improving the predictive value of prognosis in LIHC patients.

The data in different datasets such as The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus were collected in TCGA-LIHC as well as GSE116174 and GSE14520. The differential expression analysis was performed to identify telomere location-related differential expression genes (TRGs), and the gene ontology (GO) and KEGG enrichment analyses were performed to investigate the function of TRGs in bioprocess, metabolism, and signaling pathways. Prognostic risk prediction model correlated with outcome was constructed by the LASSO Cox regression model and the key genes associated with the prognosis of LIHC. The predictive capacity of the risk signature based on TRG was further confirmed in two external cohorts. The predictive ability of risk model was assessed, and a series of clinical factors associated with the prognosis of liver cancer were determined. Univariate and multivariate analyses were used to identify independent prognostic factors of LIHC.

The authors discovered a set of TRG-associated DGEs with telomere states compared between LIHC and normal. Functional enrichment analysis of these DGEs indicated that they might participate in fundamental biological processes, such as genome maintenance and replication as well as multiple metabolic and signaling pathways. A risk prediction model and signature genes associated with patient prognosis were established by the LASSO Cox regression analysis for LIHC. The prognostic accuracy of the TRG-based risk model was also verified in two independent datasets. Furthermore, the prediction accuracy of the model was analyzed, and clinical indicators associated with the prognosis of liver cancer patients were enumerated. Univariate and multivariate analyses were conducted to investigate the association of clinical variables and prognosis in patients with LIHC.

In conclusion, the authors validate that diagnostic, therapeutic, and prognostic accuracy would be enhanced through the study of gene expression data, construction of risk prediction models, and identification of risk-associated clinical factors of LIHC patients. The findings provide new biomarkers and risk prediction models for clinicians to better estimate the risk of patients for the purpose of treatment decisions.