A Novel Model Based on Genomic Instability-Associated Long Non-Coding RNAs for Predicting Prognosis and Response to Immunotherapy in Patients With Lung Adenocarcinoma

Abstract

Background: Emerging scientific evidence has shown that long non-coding RNAs (lncRNAs) exert critical roles in genomic instability (GI), which is considered a hallmark of cancer. To date, the prognostic value of GI-associated lncRNAs (GI-lncRNAs) remains largely unexplored in lung adenocarcinoma (LUAC). The aims of this study were to identify GI-lncRNAs associated with the survival of LUAC patients, and to develop a novel GI-lncRNA-based prognostic model (GI-lncRNA model) for LUAC.Methods: Clinicopathological data of LUAC patients, and their expression profiles of lncRNAs and somatic mutations were obtained from The Cancer Genome Atlas database. Pearson correlation analysis was conducted to identify the co-expressed mRNAs of GI-lncRNAs. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses were conducted to determine the main biological function and molecular pathways of the differentially expressed GI-lncRNAs. Univariate and multivariate Cox proportional hazard regression analyses were performed to identify GI-lncRNAs significantly related to overall survival (OS) for construction of the GI-lncRNA model. Kaplan–Meier survival analysis and receiver operating characteristic curve analysis were performed to evaluate the predictive accuracy. The performance of the newly developed GI-lncRNA model was compared with the recently published lncRNA-based prognostic index models.Results: A total of 19 GI-lncRNAs were found to be significantly associated with OS, of which 9 were identified by multivariate analysis to construct the GI-lncRNA model. Notably, the GI-lncRNA model showed a prognostic value independent of key clinical characteristics. Further performance evaluation indicated that the area under the curve (AUC) of the GI-lncRNA model was 0.771, which was greater than that of the TP53 mutation status and three existing lncRNA-based models in predicting the prognosis of patients with LUAC. In addition, the GI-lncRNA model was highly correlated with programed death ligand 1 (PD-L1) expression and tumor mutational burden in immunotherapy for LUAC.Conclusion: The GI-lncRNA model was established and its performance was found to be superior to existing lncRNA-based models. As such, the GI-lncRNA model holds promise as a more accurate prognostic tool for the prediction of prognosis and response to immunotherapy in patients with LUAC.