A PROGNOSTIC FIVE-LNCRNA EXPRESSION SIGNATURE FOR SURVIVAL PREDICTION IN PATIENTS WITH TONGUE SQUAMOUS CELL CARCINOMA

Zijing Yu*, Lingyun Xia, Weidong Leng, Jun Shi, Wei He, Ming Mao, Hongbo Chai, Yan Yan 

Department of stomatology, Taihe Hospital, Hubei University of Medicine, Shiyan 442000, Hubei, China

Introduction: Tongue squamous cell carcinoma (TSCC) is the most common and a highly aggressive malignancy of oral cancer. Deregulated tumorigenic long non-coding RNA (lncRNA) has been reported in several malignancies. However, few studies have investigated systematically the prognostic value of lncRNAs in TSCC patients till date. This study aimed to construct a lncRNA-based prognostic signature to improve the survival prediction for TSCC patients.

Materials and methods: We performed a comprehensive analysis for lncRNA expression profiles and corresponding clinical information in a cohort of 122 TSCC patients from The Cancer Genome Atlas (TCGA). Prognostic lncRNAs were identified by employing univariate and multivariate Cox proportional hazards regression analysis, the least absolute shrinkage and selection operator analysis and the Kaplan-Meier curve method. A multi-lncRNA signature and risk score were constructed according to the lncRNA expression and corresponding Cox coefficient. Functional enrichment analysis and drug prediction were also performed using online tools.

Results: We established a five-lncRNA expression signature associated with overall survival (OS) based on risk scoring strategy in TSCC. Risk score was dichotomized to result in high and low risk patients. Receiver operating characteristic analysis indicated that this signature exhibited excellent performance for 1-, 3- and 5-year OS event prediction (AUC= 0.856, 0.883 and 0.879 for 1-, 3- and 5-year, respectively). Moreover, multivariate Cox regression analysis revealed that this five-lncRNA signature was an independent predictor for OS when after adjusting T stage, tumor stage and cigarettes history. Functional analysis suggested that the prognostic lncRNAs were mainly involved in apoptosis and sphingolipid signaling pathways. Drug prediction indicated that panobinostat was potentially a promising agent for the treatment of TSCC.

Conclusion: Our study demonstrated that the five-lncRNA expression signature may serve as a novel independent biomarker for predicting survival in patients with TSCC.