Establishment of the Prognosis Predicting Signature for Endometrial Cancer Patient

Background: Novel biomarkers provide clinicians more critical information on tumor genetic features and patients' prognosis. Here, we aimed to establish prognosis-predicting signatures for endometrial carcinoma (EC) patients based on the miRNA information. Material/Methods: The Cancer Genome Atlas (TCGA) website was available for dataset extraction. Prognosis-associated miRNAs were generated by univariate Cox regression test. Online websites were used to predict the targeted genes of these enrolled miRNAs. The miRNA-mRNA network was described by Cytoscape software, while the relevant signaling pathways of these targeted genes were enriched by Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses. Results: The miRNA-based overall survival (OS) and recurrence-free survival (RFS) predicting signatures were constructed by LASSO Cox regression analyses, respectively, by which, the endometrial carcinoma patients were separated into high- and low-risk groups in both the discovery and validation sets. Univariate Cox regression analyses suggested that these high-risk patients had elevated death and recurrence risk compared to low-risk patients. In addition, multivariate Cox regression analysis confirmed that our signatures were independent prognosticate factors with or without clinicopathological features for endometrial carcinoma patients. Moreover, the miRNA-mRNA network was displayed by Cytoscape software, and the pathway enrichment analyses found that the targeted genes of these enrolled miRNAs were enriched in tumor progression and drug resistance-related pathways. Conclusions: The OS and RFS predicting classifiers serve as independent prognosis-associated determiners for EC patients.