Big data analysis identified a telomere-related signature predicting the prognosis and drug sensitivity in lung adenocarcinoma

Abstract

Telomeres exert a critical role in chromosome stability and aberrant regulation of telomerase may result in telomeres dysfunction and genomic instability, which are involved in the occurrence of cancers. However, limited studies have been performed to fully clarify the immune infiltration and clinical significance of telomeres-related genes (TRGs) in lung adenocarcinoma (LUAD). The number of clusters of LUAD was determined by consensus clustering analysis. The prognostic signature was constructed and verified using TCGA and GSE42127 dataset with Least Absolute Shrinkage and Selection Operator cox regression analysis. The correlation between different clusters and risk-score and drug therapy response was analyzed using TIDE and IMvigor210 dataset. Using several miRNA and lncRNA related databases, we constructed a lncRNA-miRNA-mRNA regulatory axis. We identified 2 telomeres-related clusters in LUAD, which had distinct differences in prognostic stratification, TMB score, TIDE score, immune characteristics and signal pathways and biological effects. A prognostic model was developed based on 21 TRGs, which had a better performance in risk stratification and prognosis prediction compared with other established models. TRGs-based risk score could serve as an independent risk factor for LUAD. Survival prediction nomogram was also developed to promote the clinical use of TRGs risk score. Moreover, LUAD patients with high risk score had a high TMB score, low TIDE score and IC50 value of common drugs, suggesting that high risk score group might benefit from receiving immunotherapy, chemotherapy and target therapy. We also developed a lncRNA KCNQ1QT1/miR-296-5p/PLK1 regulatory axis. Our study identified 2 telomeres-related clusters and a prognostic model in LUAD, which could be helpful for risk stratification, prognosis prediction and treatment approach selection.