Gastric cancer risk-scoring system based on analysis of a competing endogenous RNA network.

Background: Long noncoding RNAs (lncRNAs) can play vital roles in tumor initiation, progression, invasion, and metastasis. However, the functional role of the lncRNA-based competing endogenous RNA (ceRNA) networks in gastric cancer (GC) is still unclear. We aimed to identify novel lncRNAs and their association with GC prognosis.

Methods: The lncRNA, miRNA, and mRNA expression profiles of GC patients data were obtained from The Cancer Genome Atlas (TCGA) database. Differentially expressed genes (DEGs) were identified using the edge-R package. Then, the relationship among lncRNAs-miRNAs-mRNAs was integrated into a constructed ceRNA network with Cytoscape software. Using Cox regression analysis, a risk score system based on DEGs associated with patient prognosis in GC was established. Finally, a nomogram was founded to predict the prognosis of GC patients.

Results: A total of 971 differentially expressed lncRNAs (DElncRNAs), 144 differentially expressed miRNAs (DEmiRNAs) and 2,789 differentially expressed mRNAs (DEmRNAs) were identified and found to be associated with GC risk. Using the bioinformatics method, a ceRNA network involving 62 DElncRNAs, 21 DEmiRNAs and 59 DEmRNAs was constructed. Based on the results of the Cox regression analysis, a risk-scoring system involving 3 lncRNAs (i.e., ADAMTS9-AS1, C15orf54, and AL391152.1) was set up for the survival analysis of GC patients. The area under the receiver operating characteristic (ROC) curve for the risk-scoring system was 0.674, with a C-index of 0.64 [95% confidence interval (CI): 0.59-0.69, P=2.806485e-08]. Univariate and multivariate Cox regression analyses demonstrated that the risk-scoring system was an independent prognostic factor for GC. The risk-scoring system is positively associated with advanced tumor grade. The expression of these 3 lncRNAs were validated in GEPIA database. A nomogram based on these 3 lncRNAs was created to predict the prognosis of GC patients.

Conclusions: Our study established a novel lncRNA-expression-based ceRNA network and an ADAMTS9-AS1-C15orf54-AL391152.1-based risk-scoring system, which can be used to predict the prognosis of GC patients.