A novel prognostic model based on telomere-related lncRNAs in gastric cancer.

Abstract

Telomeres are specialized structures at the ends of chromosomes that are important for their protection. Over time, long non-coding RNAs (lncRNAs) have gradually come into the spotlight as essential biomarkers of proliferation, migration, and invasion of human malignant tumors. Nevertheless, the impact of telomere-related lncRNAs (TRLs) in gastric cancer is currently unknown. In the present study, we screen the TRLs and identify a prognostic TRLs signature in gastric cancer. First, telomere-related genes (TRGs) were retrieved from the website, and RNA sequencing (RNA-seq) data and clinical data of stomach adenocarcinoma (STAD) patients were gathered from The Cancer Genome Atlas (TCGA) database. Gastric cancer patients' lncRNAs and overall survival (OS) were found to be related using univariate Cox regression analysis. Next, least absolute shrinkage and selection operator (LASSO) regression analysis and multifactorial Cox regression analysis were used to further screen telomere-related differentially expressed lncRNAs (TRDELs), and finally six lncRNAs were obtained, including LINC01537, CFAP61-AS1, DIRC1, RABGAP1L-IT1, DBH-AS1, and REPIN1-AS1. According to these six TRDELs, a prognostic model for gastric cancer was constructed. The samples were divided into the training group and the testing group at random, and the reliability of prognostic model was validated in both groups and overall samples. In addition, we performed Kaplan-Meier (K-M) survival curve analysis, independent prognostic analysis, and functional enrichment analysis to validate the predictive value and independence of the model, as well as immune cell correlation analysis, clustering analysis, and principal component analysis (PCA) to further explore the relationship between this model and the tumor cells. Finally, we performed the drug sensitivity analysis to identify a few small molecules that may have a therapeutic effect on gastric cancer. Finally, we constructed a prognostic model for gastric cancer consisting of six TRDELs. According to the K-M curve, the prognosis of the low-risk group was noticeably superior than that of the high-risk group. Multivariate Cox regression analysis suggested that risk score was an independent prognostic element. Receiver operating characteristic (ROC) curves, nomogram, and calibration curve indicated that the prognostic model had good predictive ability. Functional enrichment analysis demonstrated major pathways with high- and low-risk groups. Next, both tumor microenvironment (TME) and immune correlation analysis showed discrepancy in the high- and low-risk groups. Through drug sensitivity analysis, we screened four small molecules that might be beneficial for gastric cancer treatment. A prognostic model consisting of these six TRDELs was capable to predict the prognosis of gastric cancer patients.