Table_4_Identification of Potential Crucial Genes Associated With the Pathogenesis and Prognosis of Endometrial Cancer.XLSX

Background and ObjectiveEndometrial cancer (EC) is a common gynecological malignancy worldwide. Despite advances in the development of strategies for treating EC, prognosis of the disease remains unsatisfactory, especially for advanced EC. The aim of this study was to identify novel genes that can be used as potential biomarkers for identifying the prognosis of EC and to construct a novel risk stratification using these genes.
Methods and ResultsAn mRNA sequencing dataset, corresponding survival data and expression profiling of an array of EC patients were obtained from The Cancer Genome Atlas and Gene Expression Omnibus, respectively. Common differentially expressed genes (DEGs) were identified based on sequencing and expression as given in the profiling dataset. Pathway enrichment analysis of the DEGs was performed using the Database for Annotation, Visualization, and Integrated Discovery. The protein–protein interaction network was established using the string online database in order to identify hub genes. Univariate and multivariable Cox regression analyses were used to screen prognostic DEGs and to construct a prognostic signature. Survival analysis based on the prognostic signature was performed on TCGA EC dataset. A total of 255 common DEGs were found and 11 hub genes (TOP2A, CDK1, CCNB1, CCNB2, AURKA, PCNA, CCNA2, BIRC5, NDC80, CDC20, and BUB1BA) that may be closely related to the pathogenesis of EC were identified. A panel of 7 DEG signatures consisting of PHLDA2, GGH, ESPL1, FAM184A, KIAA1644, ESPL1, and TRPM4 were constructed. The signature performed well for prognosis prediction (p < 0.001) and time-dependent receiver–operating characteristic (ROC) analysis displayed an area under the curve (AUC) of 0.797, 0.734, 0.729, and 0.647 for 1, 3, 5, and 10-year overall survival (OS) prediction, respectively.
ConclusionThis study identified potential genes that may be involved in the pathophysiology of EC and constructed a novel gene expression signature for EC risk stratification and prognosis prediction.

背景与目的：子宫内膜癌（Endometrial Cancer, EC）是全球范围内常见的妇科恶性肿瘤。尽管子宫内膜癌的治疗策略研发已取得进展，但该病的预后仍不尽如人意，尤其对于晚期子宫内膜癌患者。本研究旨在筛选可作为子宫内膜癌预后潜在生物标志物的新型基因，并基于这些基因构建全新的风险分层模型。
方法与结果：本研究分别从癌症基因组图谱（The Cancer Genome Atlas, TCGA）与基因表达综合数据库（Gene Expression Omnibus, GEO）获取了子宫内膜癌患者的mRNA测序数据集、对应生存数据以及表达谱数据。基于测序数据与表达谱数据集的信息，筛选得到共同差异表达基因（Differentially Expressed Genes, DEGs）。利用注释、可视化与综合发现数据库（Database for Annotation, Visualization, and Integrated Discovery, DAVID）对共同差异表达基因进行通路富集分析。通过STRING在线数据库构建蛋白质相互作用网络，以识别核心基因（hub genes）。采用单因素与多因素Cox回归分析筛选预后相关差异表达基因，并构建预后特征模型。基于该预后特征模型，对TCGA子宫内膜癌数据集进行生存分析。本研究共筛选得到255个共同差异表达基因，并鉴定出11个与子宫内膜癌发病机制密切相关的核心基因：TOP2A、CDK1、CCNB1、CCNB2、AURKA、PCNA、CCNA2、BIRC5、NDC80、CDC20及BUB1BA。随后构建了由7个差异表达基因组成的预后特征模型，涉及基因包括PHLDA2、GGH、ESPL1、FAM184A、KIAA1644、ESPL1、TRPM4。该特征模型在预后预测中表现优异（p < 0.001），时间依赖性受试者工作特征（Receiver-operating Characteristic, ROC）分析显示，其1年、3年、5年及10年总生存（Overall Survival, OS）预测的曲线下面积（Area Under the Curve, AUC）分别为0.797、0.734、0.729及0.647。
结论：本研究筛选得到了可能参与子宫内膜癌病理生理过程的潜在基因，并构建了全新的基因表达特征模型，可用于子宫内膜癌的风险分层与预后预测。