PROJECT SUMMARY This proposal aims to optimize therapy for human papilloma virus-related (HPV+) squamous cell carcinomas of the head and neck (HNSCCs), which are rapidly increasing in incidence. The relatively favorable prognosis for the HPV+ subtype of HNSCC has justified ongoing efforts to de-intensify their treatment with high dose radiation and cisplatin, whose toxicities can leave lifelong disabilities in survivors. A recently popularized approach to therapy de-escalation is transoral robotic surgery (TORS), which has markedly altered practice patterns for this disease and allowed for reduction in cisplatin use and radiation dosing relative to nonsurgical therapy. However, poor ability to stratify recurrence risk after TORS is a key barrier to both safely de-escalating adjuvant therapy for typical HPV+ HNSCCs and intensifying treatment for certain cases that are predisposed to lethal outcome. This proposal seeks to fill this knowledge gap by leveraging a unique set of archival HPV+ HNSCC specimens from patients treated by TORS to pursue novel molecular biomarkers of treatment response and prognosis. Our studies of older cohorts with widely variable treatment suggest worse outcomes for HPV+ HNSCCs with high oxidative metabolic gene expression, reduced E2F target gene upregulation, and lower viral E6 oncogene expression. Thus, our overall hypothesis is that genetic variants and expression profiles of both host and viral genes will allow prospective discrimination of HPV+ HNSCCs at risk of lethal outcome after TORS-based therapy. To test this hypothesis, Aim 1 will identify host and viral expression profiles distinguishing HPV+ HNSCCs that recur after TORS. Case-control analyses will be applied to a cohort of 634 TORS-treated HPV+ HNSCCs with uniquely long-term follow-up in order to identify viral and host genes differentially expressed in tumors that later recurred. Recurrent cases will be matched to nonrecurrent controls based on stage, adjuvant therapy, and follow-up. Transcriptomic analysis will be followed by protein level validation for select differentially expressed genes. Aim 2 will define genetic traits of HPV+ HNSCCs that recur after TORS and pursue a multi-marker stratifier of recurrence risk. Whole exome sequencing will be used to identify somatic mutations and copy number alterations that distinguish tumors that recurred from nonrecurrent controls. In addition, viral genome sequencing will be used to assess for viral subtypes, sub-lineages, and nonsynonymous SNPs that are over-represented in recurrence-prone tumors. Genetic traits associated with treatment failure will be integrated with transcriptomic data to develop a multi-marker signature that stratifies HPV+ HNSCCs for lethal recurrence risk. This assessment of molecular traits that distinguish tumors with high recurrence risk after TORS-based therapy will create a discrete panel of molecular features that can be tested in large cohorts, leading to creation of strong prognost...