Endometrial cancer (EC) exhibits significant heterogeneity in clinical outcomes, and neutrophils are increasingly implicated in its progression and the tumor microenvironment. The objective of this research was to discover neutrophil-associated genes in EC and construct a prognostic model along with biologically distinct molecular subtypes. Data on clinical features and gene expression obtained from the Cancer Genome Atlas (TCGA) were subjected to analysis. To identify prognostic markers, we applied univariate Cox regression combined with several machine learning approaches, including least absolute shrinkage and selection operator (LASSO) regression, Random Forest (RF), and extreme gradient boosting (XGBoost), to screen for prognostic neutrophil-related genes. To classify molecular subtypes, we applied non-negative matrix factorization (NMF) on the gene expression data. Using the selected genes, a risk score model was formulated. We established a nomogram for predicting the clinical outcome of EC. Validation of the model was carried out on datasets from internal and external sources. We characterized the risk groups and molecular subtypes using CIBERSORT, ESTIMATE, and Single Sample Gene Set Enrichment Analysis (ssGSEA) algorithms to investigate immune cell infiltration along with immune-related functional pathways. Additionally, we investigated the potential responsiveness to drugs linked to the identified genes and risk categories. We identified four key neutrophil-related genes, including LEF1, GHH, CCL22, and PLA2G2A, which constitute a robust prognostic biomarker set for EC. Based on these genes, the risk score distinguished patients into distinct high- and low-risk categories with markedly different overall survival outcomes. Furthermore, NMF analysis revealed two distinct molecular subtypes based on these four genes, which displayed significant differences in prognosis and were characterized by unique infiltration of immune cells and expression levels of immune checkpoints. We also observed associations between the risk groups and potential drug sensitivities. In this research, we discovered a novel prognostic signature comprising four genes and classified EC into two distinct molecular subtypes driven by neutrophil-associated genes. These results enhance our understanding of EC's prognostic profile and its immune microenvironment, which may facilitate improved risk stratification and guide the design of personalized treatment approaches.
Liu et al. (Sun,) studied this question.