Ferroptosis resistance is increasingly appreciated as an indispensable factor for tumor initiation, progression, and therapeutic resistance in various human malignancies including head and neck squamous cell carcinoma (HNSCC). Herein, we sought to develop a novel signature utilizing ferroptosis-related genes (FRGs) for prognosis and therapeutic prediction in HNSCC.

A prognostic signature specific to HNSCC was developed using univariate Cox regression and LASSO-penalized multivariate Cox regression analyses. A nomogram incorporating this signature and selected clinicopathological factors was created through multivariate Cox regression. The effectiveness of the FRG signature in predicting tumor mutation burden (TMB), immune status, and responses to chemotherapy was also evaluated.

The FRG signature based on eight genes (AURKA, LPIN1, MIOX, CDKN2A, PRKAA2, CISD2, TRIB3, and ASNS) successfully classified patients into subgroups with distinct outcomes across multiple cohorts. A FRG nomogram was constructed with good-prognostic performance. Additionally, higher FRG signature scores were positively associated with TMB and negatively correlated with tumor-infiltrating immune cells, which were linked to sensitivity to several chemotherapeutic drugs.

Our findings provide strong evidence that the FRG-derived signature/nomogram can effectively predict both prognosis and therapeutic response in HNSCC.