A multi-omics approach to identify prognostic subtypes and therapeutic targets in head and neck squamous cell carcinoma

Background. Head and neck squamous cell carcinoma (HNSCC) remains a significant clinical challenge due to high recurrence rates and limited treatment efficacy. Improved diagnostic and therapeutic strategies are urgently needed. Methods . We integrated multi-omics data (transcriptomics, DNA methylation, and clinical information) from TCGA-HNSCC and GEO datasets to identify novel biomarkers and therapeutic targets. This involved subtype classification, pathway enrichment analysis (ssGSEA and GSEA), development of a consensus machine learning-based survival (CMLS) model, tumor microenvironment (TME) analysis (IOBR package), and molecular docking/cell experiments. Results . Our analysis revealed two distinct HNSCC subtypes with significantly different prognoses (p<0.001). The CMLS model, demonstrating a high C-index (0.85), identified six key genes (ZNF557, HPRT1, DSCAM, ZNF652, DUSP3, ANKRD44) associated with subtype and survival. Pathway enrichment analyses revealed associations with radiotherapy response (hypoxia) and DNA replication in the identified subtypes. Low CMLS scores correlated with increased immune cell infiltration (T cells, CD4+ T cells, B cells), suggesting immunotherapy potential. Molecular docking and cell experiment implicated HPRT1 as a potential dasatinib off-target, enhancing its anti-cancer activity. High CMLS scores were associated with pathways linked to poor prognosis (Myogenesis, EMT, hypoxia, coagulation). Conclusions . Our multi-omics approach identified novel HNSCC subtypes, prognostic biomarkers, and therapeutic targets, including HPRT1 as a potential modifier of dasatinib efficacy. These findings provide actionable insights for personalized treatment strategies and warrant further clinical validation.