Abstract: Objective Treatment resistance in head and neck squamous cell carcinoma(HNSCC)is closely associated with apoptotic defects. Necroptosis is an alternative regulated cell death modality that can bypass apoptotic deficiencies; however, its role in tumor microenvironment(TME)remodeling and prognosis in HNSCC remains insufficiently characterized at the single-cell level level. This study aimed to integrate single-cell and bulk transcriptomic data to elucidate the cellular heterogeneity of necroptosis-related features in HNSCC, and to construct a prognostic prediction model. Methods We obtained single-cell RNA sequencing(scRNA-seq)and bulk transcriptomic data of HNSCC were obtained from the GEO and TCGA databases. Single-cell analysis was performed using the Seurat package and the AUCell algorithm to quantify necroptosis activity across distinct cell subpopulations. Key genes were identified through differential expression analysis, as well as Cox and LASSO regression, in order to construct a prognostic nomogram model. The associations of the model genes with immune cell infiltration and drug sensitivity were then evaluated using CIBERSORT and the CellMiner database, respectively. Results Single-cell profiling revealed that malignant cells, macrophages, and mast cells exhibited the highest levels of necroptosis activity. Keratin family genes were found to be highly correlated with necroptosis activity in malignant cells. An integrative analysis was performed to construct a four-gene prognostic model comprising LAT, KRT6A, APP, and DUSP5. Patients in the high-risk group had significantly worse prognosis and exhibited an “immune desert” phenotype characterized by reduced infiltration of CD8⁺ T cells and activated CD4⁺ memory T cells, accompanied by increased resting dendritic cells. Furthermore, the gene expression in the model correlates with sensitivity to multiple drugs, including nelarabine and Bcl-2 inhibitors. This offers preliminary insights for subsequent experiments. It should be clearly noted that, due to limitations of public databases, HPV status information was not included in this study due to limitations of public databases. Consequently, the generalizability of the current conclusions remains to be validated by HPV-stratified analyses. Conclusion This study systematically elucidates the molecular features of necroptosis in HNSCC at a single-cell level. The established four-gene model enables effective risk stratification and provides novel molecular targets for understanding the formation of necroptosis-driven immunosuppressive TME driven by necroptosis, as well as for guiding individualized therapy.

Key words: Head and neck squamous cell carcinoma, Necroptosis, Prognostic model, Tumor microenvironment, Immunotherapy