Abstract
BACKGROUND: Long non-coding RNAs (lncRNAs) are highlighted as novel cancer biomarkers with great promise. Herein, we focused on summarizing the overall diagnostic performance of lncRNAs for gastric cancer (GC). METHODS: Publications fulfilling the search criteria were selected from the online databases. Study quality was assessed according to the Quality Assessment for Studies of Diagnostic Accuracy (QUADAS) checklist. The summary receiver operator characteristic (SROC) curve was plotted using a bivariate meta-analysis model. Statistical analysis was performed based on the platforms of STATA 12.0 and Meta-Disc 1.4 software. RESULTS: Fifteen studies with 1252 patients and 1283 matched controls were included. The pooled sensitivity and specificity for lncRNA expression profile in differentiating GC patients from cancer-free individuals were 0.68 (95%CI: 0.61-0.74) and 0.79 (95%CI: 0.72-0.84), respectively, corresponding to an area under curve (AUC) of 0.80. Moreover, the stratified analyses demonstrated that plasma-based lncRNA profiling harbored higher accuracy than that tissue-based assay (specificity: 0.80 versus 0.75; AUC: 0.84 versus 0.77). CONCLUSIONS: LncRNA profiling hallmarks a moderate diagnostic value in the management of GC and that lncRNA expression patterns may potentially be utilized as auxiliary biomarkers in confirming GC.