Abstract
Endometrial carcinoma (EC) remains an ambiguous pathogenesis. This study aimed to investigate the potential of causal genes in predicting EC prognosis. The prognostic biomarkers of EC were identified using univariate Cox regression analyses based on data from The Cancer Genome Atlas Uterine Corpus Endometrial Carcinoma (TCGA-UCEC). Mendelian randomization (MR) analyses were conducted to infer causal relationships, utilizing expression quantitative trait loci (eQTLs) derived from prognostic genes as exposures, and a dataset from European populations with EC as outcomes. Single nucleotide polymorphisms (SNPs) that significantly influenced gene expression (eQTLs) were selected as instrumental variables. The inverse variance weighted (IVW) method was employed as the primary analytical approach. Sensitivity analyses were performed to ensure robustness of the findings. Causal genes with potential prognostic significance were further evaluated using multivariate Cox regression analysis, Kaplan-Meier (KM) overall survival curves, and receiver operating characteristic (ROC) curve analysis. Additionally, results from gene ontology (GO) and gene set enrichment analysis (GSEA) of differentially expressed genes (DEGs), along with immune infiltration analyses in the high- and low-risk groups, are presented. 18 genes exhibiting a negative correlation with EC demonstrated a protective effect, whereas 9 genes identified as risk factors for EC exerted an adverse effect on the disease. A prognostic model was developed consisting of 8 genes selected from 27 genes. According to the KM overall survival curve data, ECs classified with high-risk ratings exhibited significantly poor prognoses (P < .0001). The ROC curve analysis indicated that the area under the curve (AUC) for this risk model in predicting the 1-, 3-, and 5-year EC survival rates were 0.704, 0.735, and 0.766, respectively. Furthermore, GO and GSEA results of DEGs in both the high- and low-risk groups revealed strong associations with pathways related to cell motility and immune response, among others. In addition, an analysis of immune cell infiltration demonstrated significant differences between the high- and low-risk groups. A prognostic model for EC using causal genes identified using MR has good sensitivity and specificity. These findings provide new insights into ECs pathogenesis and suggest promising strategies for the diagnosis and treatment of ECs.