Establishment of a nine‐gene prognostic model for predicting overall survival of patients with endometrial carcinoma

Abstract

Endometrial carcinoma (EC) is the most common malignant tumor of the female genital tract in developed countries. The prognosis of early stage EC is favorable, but a subset faces high risk of cancer progression or recurrence. EC has a poor prognosis upon progression to advanced or metastatic stages. Therefore, our goal is to build a robust prognostic model for predicting overall survival (OS) in EC patients. In this study, 1571 genes were identified as being associated with OS based on genomewide expression profiles using a training dataset. Kyoto Encyclopedia of Genes and Genomes enrichment analysis revealed that these genes were involved in various cancer‐related signaling pathways. Nine signature genes were further selected using stepwise selection, and their potential role in the development of EC was demonstrated by performing differential expression analysis between EC and normal uterine tissues. A prognostic model that aggregated these nine signature genes was ultimately established and effectively divided EC patients into two risk groups. OS for patients in the high‐risk group was significantly poorer compared with that of the low‐risk group. This nine‐gene model was subsequently validated and evaluated using the TCGA dataset and shown to have a high discriminating power to distinguish EC patients with an elevated risk of mortality based on the FIGO staging system and other prognostic factors. This study provides a novel prognostic model for the identification of EC patients with elevated risk of mortality and will help to improve our understanding of the underlying mechanisms involved in prognostic EC factors.

Novel miRNA markers for the diagnosis and prognosis of endometrial cancer

“…Prior to this study, several prognostic models have been developed for EC based on gene signatures. Previously, we have proposed a nine‐gene model for survival prediction in EC, whose AUC values reached 0.82 and 0.676 in the training and validation datasets, respectively . A six‐gene signature devised by Wang et al also achieved good performance with AUC values of 0.841 and 0.722.…”

“…Subsequently, multivariate Cox regression was applied to construct a prognostic model and remove any miRNAs that might not be independent factors in the model. For the gene model devised by our previous work, the risk score for each patient was computed using the list of nine genes ( SLC16A1‐AS1 , KDM4B , MAP2K5 , SYP , MPP1 , DLX4 , BOLA3‐AS1 , HOMEZ and STAP2 ) . For the prognostic model proposed by Wang et al, six genes ( PCSK4 , IHH , CTSW , LRRC8D , TNFRSF18 and CDKN2A ) were used to calculate the risk score for EC patient.…”

Biomarker Signatures of Quality for Engineering Nasal Chondrocyte-Derived Cartilage

“…Here we show how multiple logistic regression can be used to model purity percentages with the advantage of being able to provide biologically relevant estimates, i.e., between 0 and 100% (Zhao et al, 2001). When further screening the most significant genes that contributed to the purity model with stepwise selection (Ying et al, 2018;Liu et al, 2019), we found the combination of collagen II and versican expression to be predictive, a relatively uncommon gene pair compared to the often studied gene expression ratios of collagen II:I and aggrecan:versican. The selection appears reasonable, with one chondrocyte marker, collagen II, and one perichondrium marker, versican, used in the model and being inversely related to each other.…”

Identification of an immune gene signature for predicting the prognosis of patients with uterine corpus endometrial carcinoma

“…Numerous reports have described the relationship between differentially expressed genes and various aspects of tumors, including tumorigenesis and prognosis [18][19][20]. However, a vast majority of genes implicated in playing a central role in predicting tumor prognosis are limited by certain factors, such as insu cient sample sizes.…”