Comprehensive analysis of multiomics data for the identification of a cuproptosis-related gene signature predicting prognostic outcomes and drug responses in gastric cancer

Abstract

Abstract Background: Cuproptosis, a recently elucidated copper-dependent mechanism of cell death associated with the tricarboxylic acid cycle, lacks a comprehensive understanding of its relation to clinical prognosis and drug response in gastric cancer (GC). This study aims to discern potential prognostic signatures of cuproptosis-related genes (CRGs) and evaluate drug response. Methods: Using publicly available datasets from TCGA and GEO, we initially obtained transcriptomic and clinical data of GC patients. We employed consensus clustering approach to delineate molecular subtypes based on the expression of CRGs. Utilizing least absolute shrinkage and selection operator (LASSO) regression analysis, we formulated a prognostic signature derived from the differentially expressed genes among these molecular subtypes. We constructed a nomogram that amalgamates both clinical characteristics and the prognostic model to provide a comprehensive prognosis prediction. Rigorous assessment of prognostic performance was carried out through Kaplan‒Meier curve analysis, the log-rank test, univariate and multivariate Cox regression, and time-dependent ROC curve analysis. Tumor Immune Dysfunction and Exclusion (TIDE) andthepRRophetic package in R were used to assess the potential response to chemotherapy and immunotherapy. Seurat was utilized to analyze the general characterization of the single-cell dataset. Additionally, the validation of hub gene expression in both cells and clinical samples was undertaken via qRT‒PCR. Results: Upon conducting an exhaustive investigation into the distinct differential expression and prognostic implications of each CRG, we delineated two distinct cuproptosis-associated molecular subtypes. Following Lasso regression analyses, we formulated a prognostic model comprising six specific genes. Patients were effectively stratified into either high-risk or low-risk categories by utilizing this model. Patients classified as high-risk experienced poorer prognosis and were associated with higher TNM stages compared to those with low risk. Furthermore, patients belonging to the low-risk group exhibited enhanced benefits from chemotherapeutic drugs and demonstrated better susceptibility to immunotherapy. The validation of our prognostic model's efficacy was established through ROC analysis, affirming its commendable sensitivity and specificity. Conclusions: Our study illuminates the significance of cuproptosis in drug response and clinical prognosis in Asian GC patients, underscoring its clinical significance and providing a reliable tool for predicting overall survival in this patient population.