ABSTRACT
Oral squamous cell carcinoma (OSCC) is on the rise, increasing by two-thirds in 20 years. Each year 30,000 Americans are diagnosed with OSCC, and 50% of these cases are early stage I/II. Despite the early stage at diagnosis, these patients suffer from significant morbidity, and a 5-year mortality rate of 40%. Treatment for early stage OSCC is highly variable, ranging from just cancer resection, to the addition of adjuvant treatments like elective neck dissection (END), radiotherapy (RT), or chemoradiation (chemoRT). While stage is primarily used to assess risk and assign adjuvant treatment, its prognostic value is low. There is currently no reliable clinical, histologic or molecular marker to determine individual risk in patients within the same cancer stage. There is a need to develop a robust prognostic biomarker to guide treatment and improve survival. We recently defined a mortality risk score for early stage OSCC patients, composed of methylation and clinicopathologic features, using a discovery cohort and The Cancer Genome Atlas (TCGA) data, which has strong predictive performance to identify patients at high risk of death in 5 years. In this application, we propose to validate this methylation biomarker in early stage OSCC patients with known 5-year survival from a multi-institutional retrospective cohort of formalin-fixed, paraffin embedded (FFPE) tissues. We will combine our validated methylation (molecular) biomarker with clinicopathologic (non-molecular) markers to construct the high-Risk Epigenetic And clinicopathologic Score for Oral caNcer (REASON) score. We hypothesize that this personalized score will accurately predict the risk of 5-year cancer-specific mortality. The study will proceed with three aims. Firstly, we will perform an epigenome wide association study (EWAS) using the EPIC array, to construct and validate the REASON score with a retrospective cohort (cohort 1, n=400) of early stage OSCC patients with known 5-year survival outcome, who underwent cancer resection only. Secondly, we will apply the REASON score to a separate retrospective cohort (cohort 2, n=400) of early stage OSCC patients who underwent adjuvant treatments (i.e., END, RT, chemoRT) in addition to cancer resection. We will determine whether these adjuvant treatments confer a survival advantage in high risk (high REASON score) patients over cancer resection alone. We will also determine whether these adjuvant treatments could be spared in low risk (low REASON score) patients. We will also perform technical validation of the methylation features discovered in the EWAS with MethylCap-Seq (MC-Seq), a robust, Clinical Laboratory Improvement Amendments (CLIA) certifiable platform. Lastly, in an exploratory aim, we will prospectively enroll early stage OSCC patients and collect noninvasive brush swabs and cancer tissues. We will determine the concordance of methylation signatures between paired brush swabs and cancer tissues in these patients using MC-Seq, and determine the prognostic performance of the REASON score in this prospective cohort (cohort 3, n=200). This study assembles the largest cohort (n=1000) early stage OSCC patients to date, and is expected to produce a clinically robust mortality risk score.
