Identification and Validation of Prognostic Genes Related to Lipid Intake in Endometrial Cancer from a Transcriptomic Perspective

Background Endometrial cancer (EC) is a common gynecological malignancy with high invasiveness and metastatic potential. Although abnormal lipid intake has been recognized as a key etiological factor, the role of lipid intake-related genes (LIRGs) in EC remains poorly understood. This study aimed to elucidate the clinical significance of LIRGs in EC. Methods Candidate EC-related LIRGs were identified through differential expression analysis and weighted gene co-expression network analysis using transcriptomic data. Prognostic models were developed using univariate Cox regression, least absolute shrinkage and selection operator, and multivariate Cox regression, followed by nomogram construction. Associations with immune microenvironment, genetic mutations, and therapeutic responses were investigated. Reverse transcription quantitative polymerase chain reaction was employed to validate prognostic gene expression in EC tissues. Results SLC47A1 and ATP5F1E were identified as prognostic genes for EC. Patients were stratified into high- and low-risk groups based on these genes, with the high-risk group showing significantly elevated mortality (P < 0.05). Both age and risk score were independent prognostic factors, and the nomogram demonstrated strong predictive accuracy (C-index = 0.78). Significant differences between risk groups were observed in immune infiltration and immunotherapy response. PTEN and TP53 exhibited high mutation frequencies. Validation experiments confirmed decreased SLC47A1 expression and increased ATP5F1E expression in EC tissues (both P < 0.01). Conclusions This study establishes SLC47A1 and ATP5F1E as novel prognostic biomarkers for EC. The prognostic model based on these genes provides valuable insights for risk stratification and may facilitate personalized therapeutic strategies in EC.