Head and neck squamous cell carcinoma is the 6th most common cancer worldwide with an annual incidence of 12000 cases in the UK alone. More than 60% of cases are diagnosed at the locally advanced stage. These patients are treated with radical intent, using a combination of surgery, radiotherapy and/or chemotherapy. Unfortunately 5 in 10 patients relapse within 2 years, with most relapses occurring within the first year since treatment. Unlike many other solid tumours, 80% of relapses occur locoregionally. Salvage surgery offers the best chance of long-term survival for patients with loco-regional recurrence, but this is only possible if the recurrence is amenable to resection. Salvage surgery has been estimated to improve survival outcomes in relapsed cancer by up to 73%. For salvage surgery to be feasible, relapses need to be detected early. Current surveillance strategies have little evidence base, with imaging often driven by clinical symptoms - often when the recurrence is no longer amenable to salvage surgery. With this study, we will address the unmet clinical need to develop a risk-stratified surveillance pathway to enhance detection of early relapse of radically treated head and neck cancer. At present, tumour grade and biomarkers such as HPV status have offered important but insufficient information to guide surveillance strategies.