A multi-lncRNA signature for suvival prediction in esophageal squamous cell carcinoma (ESCC) patients

Abstract Background: The aim of this study was to identify prognostic long non-coding RNAs (lncRNAs) and develop a multi-lncRNA signature for suvival prediction in esophageal squamous cell carcinoma (ESCC) patients.Methods: The clinical and gene expression data from 301 ESCC patients were downloaded, including a corhort used as training set from Gene Expression Omnibus database (GSE53624, n=119), another cohort of 98 paired ESCC tumor and normal tissues as test set and an independent validation cohort including 84 ESCC tissues. Survival analyses, Cox regression and Kaplan–Meier analysis were used.Results: we screened a prognostic marker of ESCC from the GSE53624 dataset and named it as the five-lncRNA signature including AC007179.1, MORF4L2-AS1, RP11-488I20.9, RP13-30A9.2, RP4-735C1.6, which could classify patients into high- and low-risk groups with significantly different survival(median survival: 1.75 years vs. 4.01 years, log rank P<0.05). Then test dataset and validation dataset confirmed that the five-lncRNA signature can determine the prognosis of ESCC patients. Predictive independence of the prognostic marker was proved by multivariable Cox regression analyses in the three datasets (P<0.05). In addition, the signature was found to be better than TNM stage in terms of prognosis.Conclusion: The five-lncRNA signature could be a good prognostic biomarker for ESCC patients and has important clinical value.