# Discovery of a Ferroptosis-Related lncRNA–miRNA–mRNA Gene Signature in Endometrial Cancer Through a Comprehensive Co-Expression Network Analysis

## TL;DR

This study identifies a new RNA-based model for predicting endometrial cancer prognosis by analyzing ferroptosis-related gene interactions.

## Contribution

The first comprehensive lncRNA–miRNA–mRNA regulatory network for ferroptosis in endometrial cancer.

## Key findings

- A 16-RNA signature (10 lncRNAs, 2 miRNAs, 4 mRNAs) was developed for EC prognosis.
- The model outperformed traditional clinical factors with AUC values up to 0.768.
- High-risk patients had significantly shorter survival times (p < 0.001).

## Abstract

Ferroptosis, a newly discovered form of cell death related to cancer, is regulated by mRNAs as well as non-coding RNA molecules, including long non-coding RNAs (lncRNAs) and microRNAs (miRNAs). However, examination of ferroptosis within endometrial cancer (EC) remains limited. In the current investigation, we used publicly available data for 521 patients with EC and identified 16 signal ferroptosis-related RNA molecules containing 10 lncRNAs, 2 miRNAs, and 4 mRNAs, associated with prognosis. A predictive system for patient prognosis on the basis of these ferroptosis-associated RNAs was developed. Notably, this prognostic framework exhibited superior predictive accuracy for EC prognosis, with a higher AUC value than that of conventional clinical factors, containing diagnostic age, tumor differentiation, and stage classification. This study is the first to propose a comprehensive ferroptosis-associated lncRNA–miRNA–mRNA regulatory circuitry in EC. The present findings offer a robust basis for future analyses of ferroptosis in EC.

Background: As a newly recognized type of cell death implicated in cancer, ferroptosis plays multiple roles in tumor biology. Here, we sought to construct a prognostic framework for EC on the basis of ferroptosis-related long non-coding RNAs (FerlncRNAs), microRNAs (FermiRNAs), and mRNAs (FRGs) for endometrial cancer (EC). Methods: Transcriptomic profiles of tumors and matched clinical data for 544 EC patients were retrieved from TCGA-UCEC. A prognostic framework was generated through Cox regression, integrating ferroptosis-linked lncRNAs, miRNAs, and mRNAs. EC cases were stratified into groups with high or low predicted risk based on ferroptosis-related gene expression. The model’s prognostic utility was examined through Kaplan–Meier (K–M) analysis and receiver operating characteristic curves. Results: A prognostic model based on 16 RNAs, including 10 FerlncRNAs, 2 FermiRNAs, and 4 FRGs, was developed. Analysis using K–M plots showed that high-risk patients experienced shorter overall survival than their low-risk counterparts (p < 0.001). The model’s area under curve (AUC) values were 0.731, 0.749, and 0.768 at 1-, 3-, and 5-year time points, surpassing those of standard clinical parameters. Furthermore, in an external validation cohort, these signature RNAs were associated with EC prognosis. Conclusions: The novel ferroptosis-related lncRNA–miRNA–mRNA prognostic model provides a basis to assess clinical prognosis in EC patients.