Abstract:BACKGROUNDThe incidence of lung cancer ranks first among malignant tumors all over the world. Based on the histological features, lung cancer could be divided into small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC), of which non-small cell lung cancer accounts for about 80%. NSCLC mainly includes lung squamous cell carcinoma (LUSC), lung adenocarcinoma (LUAD), and large cell lung cancer (LLC). Most of the patients were diagnosed at a late stage, which means the 5-year overall survival rate for patients was shallow. According to the situation, new biomarkers are needed to recognize patients at high risk, which would help give them appropriate treatment to improve their outcomes.Ferroptosis, an important iron-dependent cell death driven by oxidative phospholipid damage and characterized by lipid peroxidation, was recently found to play a novel role in several cancers. Existing research have identified many genes related to ferroptosis in tumor tissues. However, research on ferroptosis-genes-related long non-coding RNA (lncRNA) in lung cancer is still insufficient. METHODSWe acquired the statistics from the public database TCGA Lung Squamous Cell Carcinoma (LUSC). Then a multi-lncRNA signature was constructed to recognize patients at high risk based on differentially expressed ferroptosis-genes-related lncRNAs in lung squamous cell carcinoma. RESULTSWe finally identified eight differently expressed ferroptosis-genes-related lncRNAs are predictive of outcomes in LUSC patients. Kaplan-Meier analyses revealed that the high-risk-related lncRNAs signature has strong predictive power for poor LUSC prognosis (AUC=0.686). Our risk-predictive model was superordinate to the traditional predictive method based on clinicopathological characteristics (Stage, AUC=0.563). Gene set enrichment analysis (GSEA) revealed signaling pathways that ferroptosis-genes-related lncRNAs may participate in. Further study of immune function-related gene sets showed that parainflammation, APC co-stimulation, inflammation-promoting, T cell co-stimulation, Type I and type II INF response were significantly associated with risk-related lncRNAs signature. Immune checkpoint-related genes such as PDCD-1, CD70, CD28, and CD27, etc., also expressed differently between the two risk groups.CONCLUSIONThe specific differentially expressed ferroptosis-related lncRNAs have a strong predictive effect on the prognosis in patients with lung squamous cell carcinoma.

Identification of a Ferroptosis-related lncRNA Signature for Prognosis Prediction in Lung Squamous Cell Carcinoma