Identification of EMT-associated LncRNA Signature for Predicting the Prognosis of Patients with Endometrial Cancer

Background: Endometrial cancer (EC) is one of the most normal malignancies globally. Growing evidence suggests epithelial–mesenchymal transition (EMT) related markers are closely correlated with poor prognosis of EC. However, the relationship between multiple EMT-associated long non-coding RNAs (lncRNAs) and the prognosis of EC has not yet been studied. Methods: The transcriptome data and clinical information of EC cases were obtained from The Cancer Genome Atlas (TCGA), respectively. Then, we identified differentially expressed EMT-associated lncRNAs between tumor and normal tissue. Univariate cox regression analysis and multivariate stepwise Cox regression analysis was applied to identify EMT-associated lncRNAs that related to overall survival (OS). Kaplan-Meier curve, receiver operating characteristic (ROC), nomograms and multi-index ROC curves were further established to evaluate the performance of the prognostic signature. In addition, we also investigated the distribution of immune cell characteristics, sensitivity to immune checkpoint inhibitor (ICI) and chemotherapeutics, and tumor mutation burden (TMB) between high- and low-risk score predicated on a prognostic model. Results: We established nine EMT-associated lncRNA signature to predict the OS of EC, the area under the ROC curve (AUC) of the risk score has better values compared with other clinical characteristics, indicating the accuracy of the prognostic signature. As revealed by multivariate Cox regression, the prognosis model independently predicted EC prognosis. Moreover, the signature and the EMT-associated lncRNAs showed significant correlations with other clinical characteristics,including . Multi-index ROC curves for estimating 1-, 3- and 5-year overall survival (OS) of EC patients showed good predictive accuracy with AUCs of 0.731, 0.791, and 0.782, respectively. The high-risk group had specific tumor immune infiltration, insensitive to ICI, higher chemotherapeutics sensitivity and higher expression of TP53 mutation. Finally, the five lncRNAs of signature was further verified by qRT-PCR. Conclusion: We constructed an EMT-associated lncRNA signature that can predict the prognosis of EC effectively, and the prognostic signature also played an essential role in the TME; thus, the establishment of EMT-associated lncRNA signature may provide new perspectives for the treatment of EC.