Background: Dysregulation of lipid metabolism is common in cancer. However, the molecular mechanism underlying lipid metabolism in esophageal squamous cell carcinoma (ESCC) and its effect on patient prognosis are not well understood. The objective of our study was to construct a lipid metabolism-related prognostic model to improve prognosis prediction in ESCC.
Methods: We downloaded the mRNA expression profiles and corresponding survival data of patients with ESCC from the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases. We performed differential expression analysis to identify differentially expressed lipid metabolism-related genes (DELMGs). We used Univariate Cox regression and least absolute shrinkage and selection operator (LASSO) analyses to establish a risk model in the GEO cohort and used data of patients with ESCC from the TCGA cohort for validation. We also explored the relationship between the risk model and the immune microenvironment via infiltrated immune cells and immune checkpoints.
Results: The result showed that 132 unique DELMGs distinguished patients with ESCC from the controls. We identified four genes (ACAA1, ACOT11, B4GALNT1, and DDHD1) as prognostic gene expression signatures to construct a risk model. Patients were classified into high- and low-risk groups as per the signature-based risk score. We used the receiver operating characteristic (ROC) curve and the Kaplan-Meier (KM) survival analysis to validate the predictive performance of the 4-gene signature in both the training and validation sets. Infiltrated immune cells and immune checkpoints indicated a difference in the immune status between the two risk groups.
Conclusion: The results of our study indicated that a prognostic model based on the 4-gene signature related to lipid metabolism was useful for the prediction of prognosis in patients with ESCC.
Keywords: biomarkers, esophageal squamous cell carcinoma, immune microenvironment, lipid metabolism, prognosis