Abstract
BACKGROUND: Alternative splicing (AS) plays crucial roles in tumorigenesis by regulating cancer development, progression, and patient outcomes. However, the prognostic value of AS in HPV-negative head and neck squamous cell carcinoma (HNSCC), a subtype with distinct clinical features and poor prognosis compared to HPV-positive HNSCC, remains underexplored. This study aimed to develop a robust predictive signature based on AS events in HPV-negative HNSCC. METHODS: We retrieved RNA-sequencing (RNA-seq) data and clinical profiles of the HNSCC cohort from The Cancer Genome Atlas (TCGA) via its data portal. Percentage Spliced In (PSI) values were derived from TCGA SpliceSeq. To identify prognostic AS events in HPV-negative HNSCC, we first performed univariate Cox analysis. Subsequently, we used Least Absolute Shrinkage and Selection Operator (LASSO) regression and stepwise multivariate Cox regression to develop a prognostic signature. The efficacy of this signature in predicting outcomes was assessed using Kaplan-Meier survival curves and receiver operating characteristic (ROC) analyses. Additionally, gene set enrichment analysis (GSEA) was applied to the prognostic signatures. The levels of tumor-infiltrating immune cells were quantified using established deconvolution algorithms (e.g., CIBERSORT, xCell, MCP-counter). RESULTS: We identified 42,849 AS events across 388 HPV-negative HNSCC cases, of which 1,062 events were significantly associated with patient prognosis. We developed a prognostic signature based on the 21 most significant AS events, demonstrating high accuracy in predicting survival in HPV-negative HNSCC. This signature emerged as an independent prognostic factor. GSEA and immune microenvironment profiling revealed that the low-risk group exhibited enrichment in immune activation and antigen presentation pathways, in contrast to the high-risk group, which was linked to cell proliferation processes. CONCLUSIONS: We successfully constructed and validated a novel AS-related prognostic signature that enhances personalized survival prediction in patients with HPV-negative HNSCC.