Diagnostic and prognostic value of genomic instability-derived long non-coding RNA signature of endometrial cancer

Abstract

Objective: To investigate whether genomic instability (GI)-derived long non-coding RNAs (lncRNAs) have a prognostic impact on the patients with endometrial cancer. Material and methods: Patients with Uterine Corpus Endometrial Carcinoma (UCEC) were selected from The Cancer Genome Atlas (TCGA) database. Systematic bioinformatics analyses were performed, including Pearson correlations, GO and KEGG enrichment analysis, bivariate and multiple logistic regression analysis, and Kaplan–Meier (KM) method. Results: A total of 552 UCEC samples were included in the study. The differentially expressed lncRNAs (DELs) were identified, including 79 down-regulated lncRNAs and 31 up-regulated lncRNAs. Bivariate logistic regression analysis showed that 19 GI-derived lncRNAs were prognostic factors. By further multivariate logistic regression analysis, AC005256.1 (estimated coefficient = −0.474), AC026336.3 (estimated coefficient = −0.030), AL161618.1 (estimated coefficient = −1.661), and BX322234.1 (estimated coefficient = 1.511) were used to construct a prognostic risk model. In the train set and test set, the risk model was shown to have both a high prognostic and a diagnostic value. Conclusion: We developed a novel GI-derived 4-lncRNA signature for the diagnosis and prognosis of patients with endometrial cancer. These findings offered a novel perspective in the clinical management of endometrial cancer.

Keywords

• Genomic instability
• Endometrial cancer
• The cancer genome atlas
• Prognosis