An Effective Hypoxia-Related Long Non-Coding RNA Assessment Model for Prognosis of Lung Adenocarcinoma

Abstract

Background: Lung adenocarcinoma (LUAD) represents one of the highest incidence rates worldwide. Hypoxia is a significant biomarker associated with poor prognosis of LUAD. However, there are no definitive markers of hypoxia-related long non-coding RNAs (lncRNAs) in LUAD. Methods: From The Cancer Genome Atlas (TCGA) and the Molecular Signatures Database (MSigDB), we acquired the expression of hypoxia-related lncRNAs and corresponding clinical information of LUAD patients. The hypoxia-related prognostic model was constructed by univariable COX regression analysis, least absolute shrinkage and selection operator (LASSO), and multivariable Cox regression analysis. To assess the performance of the model, the Kaplan–Meier (KM) survival and receiver operating characteristic (ROC) curve analyses were performed. Results: We found seven lncRNAs, AC022613.1, AC026355.1, GSEC, LINC00941, NKILA, HSPC324, and MYO16-AS1, as biomarkers of the potential hypoxia-related prognostic signature. In the low-risk group, patients had a better overall survival (OS). In addition, the results of ROC analysis indicated that the risk score predicted LUAD prognosis exactly. Furthermore, combining the expression of lncRNAs with clinical features, two predictive nomograms were constructed, which could accurately predict OS and had high clinical application value. Conclusion: In summary, the seven-lncRNA prognostic signature related to hypoxia might be useful in predicting clinical outcomes and provided new molecular targets for the research of LUAD patients.