Analysis of microRNA expression profiles reveals a 5-microRNA prognostic signature for predicting overall survival time in patients with gastric adenocarcinoma

There is growing evidence supporting dysregulated microRNAs (miRNAs) as potential prognostic biomarkers in cancer. The present study aimed to identify an miRNA model set with prognostic power for patients with gastric adenocarcinoma. miRNA-seq data from 155 patients and 37 controls were downloaded from The Cancer Genome Atlas (TCGA) database for a comprehensive analysis of miRNA expression profiles and were used as training data. A total of 5 prognostic miRNAs, which have not been previously reported, were identified using univariate and multivariate Cox regression analyses. A separate 155-patient TCGA cohort was used as a validation set for evaluation of the risk model. Patients in the training set were assigned into high- and low-risk groups according to the 5-miRNA signature risk scores. Kaplan-Meier survival analyses demonstrated that patients with high risk scores had significantly shorter survival times than those with low risk scores. The risk model validation confirmed the prognostic ability of this 5-miRNA signature in predicting the risk status of patients. Stratification analysis for clinical prognostic variables demonstrated recurrence and age were significant prognostic factors in the low- and high-risk groups, respectively. In conclusion, the present 5-miRNA signature is a potential independent risk factor for patient outcomes. The risk model based on the 5-miRNA signature performed well in predicting overall survival time in patients with gastric adenocarcinoma.