A N6-methyladenosine-related long noncoding RNAs model for predicting prognosis in oral squamous cell carcinoma: Association with immune cell infiltration and tumor metastasis

Abstract

ObjectivesWe aim to investigate the prognostic value of potential prognostic m6A-related lncRNAs in oral squamous cell carcinoma (OSCC) samples and construct a m6A-related lncRNAs prognostic model of OSCC survival outcomes, find new clues for OSCC prognosis improvement. Materials and methodsData of m6A-related lncRNAs were obtained from OSCC samples in TCGA database. Prognostic m6A-related lncRNAs were selected by univariate Cox regression analysis. M6A-related lncRNAs prognostic signature was analyzed by least absolute shrinkage and selection operator (LASSO) regression. Results271 m6A-related lncRNAs were identified in OSCC lncRNAs, 16 of which were highly correlated with OSCC survival outcomes. Two clusters were created based on 16 prognostic valuable m6A-related lncRNAs. Characteristics of tumor immune and metastasis were identified in cluster 2 and the overall survival (OS) was worse of cluster 2 than that of cluster 1. Eleven m6A-related lncRNAs were selected from prognostic lncRNAs to establish a risk prediction signature via LASSO regression. Results showed that the OS of 1 year was lower in the high-risk group than low-risk group. The area under the curve in the training cohort was 0.704 at 1 year and 0.880 at 5 years. Multivariate Cox regression results suggested that the risk score calculated based on the m6A-related lncRNAs differential expressions was an independent predictor of OS in both two cohorts. ConclusionWe identified prognostic valuable m6A-related lncRNAs and demonstrated a reliable m6A-related lncRNA prognostic model for OSCC patients.