Background: Long non-coding RNAs (lncRNAs) were demonstrated to be key to cancer progression and highly associated with the tumor immune microenvironment. Oxidative stress and immune may modulate the biological behaviors of tumors. Therefore, biomarkers that combined oxidative stress, immune, and lncRNA can be a promising candidate bioindicator in clinical therapy of cancers.
Methods: Immune-related genes (IRGs) and oxidative stress-related genes (ORGs) were identified based on a detailed review of published literatures. The transcriptome data and clinical information of lung adenocarcinoma (LUAD) patients were obtained from TCGA database. Lasso and Cox regression analyses were conducted to develop a prognostic model. Additionally, the link between immune checkpoints, immune cells, and the prognostic model was investigated, and predict the sensitivity of immunotherapy.
Results: 2498 IRGs and 809 ORGs were extracted from previous studies, and 190 immune- and oxidative stress-related genes (IOGs) were acquired by overlapping the above genes. 658 immune- and oxidative stress-related lncRNAs (IOLs) were screened by Pearson correlation analysis. A total of 25 prognosis-related IOLs were screened by univariate regression analysis. Finally, LASSO Cox regression analysis was adopted for determining a 12-IOLs prognostic risk signature. The signature performance was confirmed in the training cohort and the testing cohort, and cases were classified into low- and high-risk groups by the risk score calculated from the signature. Patients in the high-risk group had poor prognoses and immunosuppression, while the risk score was significantly associated with tumor-infiltrating immune cells, immune checkpoint expression, and immunotherapy responses. In vitro experiments further confirmed the expression of key signature gene.
Conclusion: Our new IOLs-related prognostic signature can be reliable prognostic tools and therapeutic targets for LUAD patients.

Keywords: immune, oxidative stress, lung adenocarcinoma, prognostic model