Development of a lncRNA-based prognostic signature for oral squamous cell carcinoma

Background We aimed to establish a long noncoding RNA (lncRNA)-based signature for accurately predicting prognosis and guiding the personalized clinical management of oral squamous cell carcinoma (OSCC). Methods OSCC RNA sequencing profiles were acquired from The Cancer Genome Atlas and Gene Expression Omnibus. Univariate Cox regression, least absolute shrinkage and selection operator (LASSO) and multivariate Cox regression analyses were performed to construct a lncRNA-based prognostic signature. Kaplan-Meier survival analysis, receiver operating characteristic (ROC) curves and calibration curves were used to assess the effectiveness and accuracy of the signature. Additionally, we conducted single-sample gene-set enrichment analysis to infer the different degrees of immunocyte infiltration. Weighted correlation network analysis, enrichment analysis and Spearman's correlation analysis were implemented to screen immune-related genes that interact with the lncRNA signature. Results In total, 14 lncRNAs were defined as potential prognostic biomarkers. Based on these lncRNAs, patients were divided into low- and high-risk subgroups with different survival times (p < 0.001). In addition, the reliability of the prognostic signature was verified by Kaplan-Meier analysis, ROC analysis and calibration curves. Patients in the low-risk group exhibited more significant immune cell infiltration. Simultaneously, a potential regulatory network consisting of eight lncRNAs and 159 protein-coding genes in the top 10 immune-related biological process terms was constructed. Conclusions Our findings suggested that the 14-lncRNA signature has satisfactory performance in predicting the prognosis of OSCC, thereby providing new insights to the pathogenesis, clinical patient management and therapeutic intervention. The different immune cell infiltration statuses of OSCC patients may encourage immunotherapy.