# Diabetes-associated differentially expressed genes as prognostic biomarkers and therapeutic targets in endometrial cancer: a comprehensive molecular analysis

## TL;DR

This study identifies genes linked to diabetes that affect endometrial cancer prognosis and could guide new treatments.

## Contribution

A novel prognostic model using diabetes-related genes in endometrial cancer is developed and validated.

## Key findings

- 186 diabetes-related differentially expressed genes (DM-DEGs) were identified in endometrial cancer.
- A five-gene prognostic model (TRPC1, SELENOP, CDKN2A, GSN, PGR) effectively stratifies patient risk with strong predictive accuracy.
- The model highlights pathways related to glucose and lipid metabolism, linking diabetes to cancer progression.

## Abstract

Uterine corpus endometrial carcinoma (UCEC) is a prevalent malignancy increasingly observed in patients with diabetes mellitus. A comprehensive understanding of the intricate molecular interplay between diabetes and UCEC is crucial to develop effective prognostic and therapeutic strategies. This study aims to elucidate the relationship between diabetes and UCEC by identifying diabetes-related differentially expressed genes (DM-DEGs) and to establish a prognostic model to enhance clinical outcomes.

Transcriptomic data sourced from The Cancer Genome Atlas (TCGA) was analyzed alongside diabetes-associated genes from GeneCards. Differential expression analysis revealed 931 differentially expressed genes (DEGs) in the training cohort and 1,206 DEGs in the validation cohort. By intersecting these DEGs with diabetes-related genes, we pinpointed 186 DM-DEGs, which were further subjected to Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses.

The univariate Cox analysis identified 17 DM-DEGs that demonstrated significant prognostic relevance. Through protein-protein interaction assessments, a LASSO regression model discerning five pivotal genes (TRPC1, SELENOP, CDKN2A, GSN, PGR) for prognostic modeling was constructed. This model successfully stratified patients into high- and low-risk cohorts, with Kaplan-Meier survival analysis and Receiver Operating Characteristic (ROC) curve assessment confirming notable survival differentiations. A personalized nomogram, integrating clinical parameters and risk scores, exhibited robust predictive capability, yielding a C-index of 0.781. Gene set enrichment analysis (GSEA) suggested significant involvement in pathways related to glucose and lipid metabolism.

In conclusion, our study establishes and validates a robust prognostic signature based on diabetes-related genes (DM-DEGs) for UCEC. This signature not only effectively stratifies patient risk but also delineates specific molecular pathways, such as those involving SELENOP, CDKN2A, and PGR, through which the diabetic milieu may drive tumor aggressiveness. These findings provide a mechanistic rationale for the diabetes-UCEC link and pave the way for developing personalized treatment strategies. Future work should focus on translating this signature into clinical practice and elucidating the precise biological roles of these DM-DEGs.