Abstract Oral squamous cell carcinomas (OSCCs) account for 35-38% of head and neck SCC, the 6th most common cancer worldwide, and have a disappointing 5-year survival rate of ~50%. Oral epithelial dysplasia (OED) are precursors to OSCC, but many go undiagnosed. Studies of OED allow characterization of the aberrant molecular changes that lead to cancer. Yet due to inconsistencies in standard care and a lack of sufficient longitudinal data, few studies have examined the molecular characteristics that drive the transformation of OED to OSCC across time in the same patients to identify features unique to recurrence with progression. There are very few longitudinal studies of OED progression, and those that exist have little epidemiologic characterization and lack data integration from multi-omics. Additionally, existing reports concentrate on a few common candidate genes and/or are limited to a small sample size. We will use our uniquely detailed longitudinal cohort of OED to integrate clinical, histopathologic, epidemiologic and multiple molecular characteristics of OED lesions to identify those hallmarks of OED lesions that lead to OSCC. This will benefit patients by leading to improved risk prediction, more effective preventive measures, and insight into OSCC biomarkers that indicate carcinogenesis. Use of a large, highly-characterized longitudinal cohort for this type of multi-omics study, while using a second, independent population to validate generalizability, is highly innovative. Viewed in light of the molecular profiles of OSCCs identified from our current set of tumors and from TCGA HNSC data, this study will enable us to focus on the most relevant, recurrent changes, and understand how OSCCs that proceed OED fit into the heterogeneity of overall OSCCs. Identification of biomarkers predictive of progression will enable clinicians to provide more targeted preventive measures, and vary the extensiveness of surgery and frequency of monitoring to increase quality of life. In Aim 1, we will use existing samples identify genomic, transcriptomic, and epigenetic profile signatures distinguishing progressors from non-progressors at diagnosis for each OED stage. Our approach allows us to identify a set of candidate biomarkers that distinguish high from low risk lesions, accounting for important biologic variables. The goal of Aim 2 is to chart the timing of key genomic, transcriptomic, and epigenetic changes associated with progression to OSCC, using extensive longitudinal cases, and comparing observed changes to genes and pathways known to be aberrant in OSCC. In Aim 3, we will validate the characterization of our unique longitudinal clinical cohort in OED with and without a subsequent diagnosis of OSCC from the University of Michigan to identify clinicopathologic and epidemiologic predictors of recurrence with progression.

Project Narrative Despite the fact that a substantial percent of oral dysplastic lesions progress to oral cancer, little is known regarding the molecular characteristics of the lesions that progress to cancer versus those that do not. Using advanced DNA sequencing technologies and bioinformatics methods, we will study oral dysplasias and cancers to identify biomarkers that will help to predict an individual’s risk of oral cancer. This will allow clinicians to adjust preventive measures according to risk.