Identification And Bioinformatics Analysis Of Prognostic Biomarkers For Endometrial Carcinoma

Endometrial carcinoma(EC)is one of the most common gynecological malignant tumors in world.In recent years,although the treatment of EC has made rapid progress,the high recurrence rate and the high metastasis rate remain a clinical challenge.For early stage EC patients,the prognosis is good,and the 5-year survival rate is nearly90%.However,the prognosis of advanced EC patients is poor,and the 5-year overall survival rate is only 7.7%.Therefore,there is an urgent need to explore biomarkers related to the occurrence,development and prognosis of EC to distinguish high-risk or low-risk patients and guide the schedule of individualized treatments.This paper aims to study the molecular pathogenesis of endometrial cancer and identify more reliable biomarkers.We collected the mRNA sequencing and clinical data of 583 EC patients and miRNA sequencing and clinical data of 575 EC patients from the Cancer Gene Atlas(TCGA)database.First,we constructed the co-expression network of mRNA,lncRNA and miRNA by WGCNA,and identified the key modules related to the important clinical traits(clinical stages)of EC patients from three co-expression networks.Through differential expression analysis,712 mRNAs(456 up-regulated mRNAs,256 down-regulated mRNAs),378 lncRNAs(188 up-regulated lncRNAs,190down-regulated lncRNAs)and 44 miRNAs(15 up-regulated miRNAs and 29 downregulated miRNAs)were screened out from the key modules.To further explore the molecular function of the selected differentially expressed mRNA,we carried out Gene Ontology(GO)and Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway enrichment analysis.The results showed that they may participate in biological processes such as cell differentiation,transcription fromRNA polymerase II promoter.Based on the differentially expressed RNAs,we constructed a ceRNA regulatory network containing 11 mRNAs,4 miRNAs and 18 lncRNAs to study the molecular regulatory mechanism of endometrial carcinoma.To identify the RNAs associated with the overall survival of EC patients,we carried out the Kaplan-Meier survival analysis of each RNA in the ceRNA network and found that 7 mRNAs,4 lncRNAs and 1 miRNA were closely related to the overall survival of EC patients.In addition,we conducted regression analysis on the mRNA-lncRNA pairs in the ceRNA network and found that 35 of them were positively correlated.Finally,five mRNAs(CBX6,PIM1,RIMS3,SOX11 and XKR7),three lncRNAs(WT1-AS,LINC00494 and LINC00501)and one miRNA(miR-195)were identified as potential prognostic biomarkers of EC.The Gene Set Enrichment Analysis(GSEA)of these five mRNAs showed that they were highly enriched in cell cycle and tumor-related signaling pathways such as Nikolsky breast cancer 7p15 amplification,rosty cancer promotion cluster.Lnc RNA controls the upstream part of ceRNA network and acts as the main effector of miRNA and mRNA.In addition,the expression and distribution of lncRNA were highly specific,which made it the best biomarker for the diagnosis and prognosis of EC.The samples were randomly divided into the training set and test set according to the ratio of 7:3.Then,based on the training set,the Cox regression model related to the overall survival of EC patients was established by using the maximum likelihood method,and the regression coefficient(β)of the multivariate Cox regression model was obtained.Finally,a prognostic risk scoring system based on four lncRNAs was established.Prognostic Index(PI)=(0.1695×LINC00501)+(0.0337×WT1-AS)+(0.0855 ×ADAMTS9-AS1)+(0.0092 ×LINC00494).Kaplan-Meier curve showed that the risk score model could distinguish the high-risk patients and lowrisk patients.The area under the curve(AUC)was 0.789 of the five years receiver operating characteristic(ROC)curve of the test set,which showed that the risk scoring system had high survival prediction performance.The univariate and multivariate Cox regression analysis showed that risk score,clinical stage and histological grade were independent prognostic factors.Therefore,our risk scoring system is a valuable prognostic model,which can help clinicians effectively distinguish high-risk or low-risk patients,and then guide the schedule of individualized treatments of EC patients.We had identified five mRNAs(CBX6,PIM1,RIMS3,SOX11 and XKR7),three lncRNAs(WT1-AS,LINC00494 and LINC00501)and one miRNA(miR-195)as potential prognostic biomarkers for EC patients,which would contribute to the early diagnosis,prognosis judgment and development of new treatment strategies of EC patients.In addition,a lncRNA-related risk scoring system was constructed to help clinical workers effectively distinguish high-risk or low-risk patients and guide personalized treatment of EC patients.