A Glutathione Metabolism-Related Transcriptomic Signature for Prognostic Assessment and Biological Characterization of Lung Adenocarcinoma
TL;DR
The proposed glutathione metabolism-related signature demonstrated prognostic value in both training and validation cohorts and was associated with immune characteristics, pathway enrichment patterns, genomic alterations, and tumor mutation burden in LUAD.
Abstract
Background Glutathione metabolism plays an important role in redox homeostasis, oxidative stress responses, and metabolic adaptation in cancer. However, its prognostic significance in lung adenocarcinoma (LUAD) remains incompletely understood. This study aimed to develop a glutathione metabolism-related prognostic signature and investigate its associations with immune characteristics, genomic alterations, and biological pathways in LUAD. Methods Transcriptomic, clinical, and somatic mutation data from TCGA-LUAD were analyzed, and GSE50081 was used as an independent validation cohort. Glutathione metabolism-related genes were identified through differential expression and Cox regression analyses, followed by least absolute shrinkage and selection operator (LASSO) Cox regression to construct a prognostic signature. Survival analysis, time-dependent receiver operating characteristic (ROC) analysis, Cox regression, immune infiltration analysis, Gene Set Enrichment Analysis (GSEA), mutation profiling, tumor mutation burden (TMB) analysis, and drug-sensitivity prediction were subsequently performed. Results A glutathione metabolism-related signature stratified patients into high- and low-risk groups with significantly different overall survival in both the TCGA training cohort and the GSE50081 validation cohort. The risk score remained an independent prognostic factor in multivariable Cox regression analysis. High-risk tumors exhibited reduced B-cell and dendritic-cell infiltration, enrichment of cell-cycle- and metabolism-related pathways, higher frequencies of TP53 and KEAP1 mutations, and elevated tumor mutation burden. Computational drug-sensitivity analysis identified differences in predicted responses to several therapeutic agents between risk groups. Conclusions The proposed glutathione metabolism-related signature demonstrated prognostic value in both training and validation cohorts and was associated with immune characteristics, pathway enrichment patterns, genomic alterations, and tumor mutation burden in LUAD. These findings provide additional insights into glutathione metabolism-related heterogeneity in LUAD and warrant further biological and clinical validation.