Folate metabolism–based risk stratification identifies CYP27B1 as a determinant of tumor progression in HNSCC

Head and neck squamous cell carcinoma (HNSCC) is characterized by frequent recurrence and poor survival, highlighting the need for reliable biomarkers for risk stratification and therapeutic guidance. Folate-mediated one-carbon metabolism has been linked to tumor development; however, its prognostic and immunological relevance in HNSCC remains unclear. Using The Cancer Genome Atlas (TCGA) transcriptomic data, we developed and validated a folate metabolism–related gene signature (FMRG_score). An 11-gene model constructed by least absolute shrinkage and selection operator (LASSO) and multivariate Cox regression effectively stratified patients into distinct survival groups and served as an independent prognostic factor. A nomogram integrating FMRG_score with clinical variables demonstrated favorable predictive performance. Functional enrichment analyses revealed activation of oncogenic pathways in high-risk tumors. Immune deconvolution indicated that elevated FMRG_score was associated with reduced anti-tumor immune infiltration, increased macrophage-related signals, and an immunosuppressive tumor microenvironment. Drug response prediction suggested higher estimated IC50 values in the high-risk group, showing potential chemoresistance. Mechanistically, CYP27B1 knockdown suppressed malignant phenotypes and enhanced cisplatin sensitivity. Collectively, these findings establish folate metabolism as a metabolic–immune axis in HNSCC and suggest that the FMRG_score may serve as a prognostic and therapeutic stratification tool.