Abstract
BACKGROUND: Laryngeal squamous cell carcinoma (LSCC) is an aggressive malignant tumor, characterized by high incidence and mortality. Metabolic pathways within cancer cells are frequently dysregulated; thus, exploring fumaric acid metabolism-related genes (FAMRGs) appears interesting. We aimed to identify a signature prognostic genetic profile to develop tailored management strategies for patients with LSCC. METHODS: Data from The Cancer Genome Atlas (TCGA), Gene Expression Omnibus (GEO), and GeneCards databases were used to identify differentially expressed genes related to fumaric acid (FA) metabolism in LSCC. To explore the underlying mechanisms, we conducted analyses using the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). Additionally, we employed Cox regression and the least absolute shrinkage and selection operator (LASSO) to develop a risk signature based on FAMRGs. This signature was validated in TCGA and GEO cohorts. The association of the risk score with clinical characteristics, microenvironmental characteristics, and drug sensitivity was explored by correlation analyses. Finally, expression of FAMRGs was validated using datasets from the Gene Expression Profiling Interactive Analysis (GEPIA) and the Human Protein Atlas (HPA) databases. Moreover, the robustness of our findings was further confirmed through molecular docking and single-cell sequencing. RESULTS: A FA metabolism-associated model for laryngeal cancer was constructed using seven genes (ABCC2, ADH7, AQP9, CXCL11, GPT, PAEP, and PLCG1). Functional analysis suggested that FAMRGs were strongly associated with the chemotaxis and cytokine-cytokine receptor interaction. High-risk score subgroups, as indicated by the Kaplan-Meier curves, demonstrated poorer outcomes in both TCGA and GEO cohorts. A predictive nomogram was developed for LSCC survival probability; FAMRGs were significantly associated with the immune checkpoints. Additionally, six small molecule drugs that appeared promising as therapeutic agents in combating LSCC were identified. Besides, CXCL11 and AQP9 exhibited significantly high expression in tumor tissues, while GPT showed low expression, as confirmed by the HPA and GEPIA databases. Molecular docking confirmed the interaction between the seven core genes and FA. This finding was corroborated by single-cell sequencing, which revealed significant expression differences across various cell clusters in LSCC. CONCLUSIONS: A prognostic model associated with FA metabolism was established for LSCC based on seven genes. This model can effectively predict LSCC prognosis. Additionally, six small molecule drugs with potential therapeutic value for LSCC were identified.