Abstract:Objectives Intricate associations between programmed cell death (PCD) and cancer development and treatment outcomes have been increasingly appreciated. Here, we integrated 12 PCD patterns to construct a novel biomarker, cell death index (CDI), for oral squamous cell carcinoma (OSCC) prognostication and therapeutic prediction. Materials and Methods Univariate Cox regression, Kaplan–Meier survival, and LASSO analyses were performed to construct the CDI. A nomogram combining CDI and selected clinicopathological parameters was established by multivariate Cox regression. The associations between CDI and immune landscape and therapeutic sensitivity were estimated. Single‐cell RNA‐seq data of OSCC was used to infer CDI genes in selected cell types and determine their expression along cell differentiation trajectory. Results Ten selected PCD genes derived a novel prognostic signature for OSCC. The predictive prognostic performance of CDI and nomogram was robust and superior across multiple independent patient cohorts. CDI was negatively associated with tumor‐infiltrating immune cell abundance and immunotherapeutic outcomes. Moreover, scRNA‐seq data reanalysis revealed that GSDMB, IL‐1A, PRKAA2, and SFRP1 from this signature were primarily expressed in cancer cells and involved in cell differentiation. Conclusions Our findings established CDI as a novel powerful predictor for prognosis and therapeutic response for OSCC and suggested its potential involvement in cancer cell differentiation.

Leveraging programmed cell death patterns to predict prognosis and therapeutic sensitivity in OSCC