Splicing variability and oncogenic isoforms: Transforming our understanding of HNSCC biology

Head and neck squamous cell carcinoma (HNSCC) remains a cancer type with limited effective targeted therapies, currently restricted to anti-EGFR (Cetuximab) and anti-PD1 agents (Nivolumab and Pembrolizumab), and only one clinically established biomarker (HPV or its surrogate p16, applicable to ~30% of patients). Despite extensive genomic sequencing efforts, no new targeted therapy or biomarker has been approved for HNSCC in the past two decades. Our research addresses this critical gap by focusing on alternative splicing events (ASEs), which can profoundly alter gene function independently of mutations and represent an underexplored mechanism of oncogenesis. We hypothesize that ASEs in HNSCC contribute to tumor progression and can serve as biomarkers and therapeutic targets. To test this, we analyzed RNA-Seq data from The Cancer Genome Atlas (TCGA) and institutional cohorts, identifying 110 recurrent ASEs, including those in AKT3, PIK3R1, and HOXC6. These isoforms frequently originate from alternative transcription start sites (TSS) and display a strong inverse correlation between promoter DNA methylation and ASE expression, coupled with a positive correlation between H3K27ac enrichment and ASE expression. Together, these findings implicate chromatin architecture as a critical regulator of cancer-specific splicing programs. Functional validation revealed that these isoforms are frequently upregulated in tumors and drive oncogenic phenotypes such as increased proliferation and migration through activation of PI3K/AKT and other signaling pathways. To systematically characterize splicing variability and its immunogenic potential, we developed Splice Expression Variability Analysis (SEVA), enabling detection of differential isoform usage across tumor subtypes, and SpliceMutr, a computational pipeline for predicting splicing-derived neoantigens. These tools have uncovered splicing-driven vulnerabilities that may inform immunotherapy strategies and precision oncology. In conclusion, our findings created a map of ASE in HNSCC and established ASEs as a major contributor to HNSCC oncogenesis, independent of coding mutations, and highlighted the interplay between epigenetic regulation and RNA processing. Detection of oncogenic isoforms offers a path toward novel biomarkers, while targeting isoform-specific mechanisms could lead to innovative therapeutic strategies. This work introduces a paradigm shift in understanding HNSCC biology and opens new avenues for biomarker discovery and treatment development.