Selection of patients for surveillance imaging after radiotherapy for squamous cell carcinoma of oral cavity and oropharynx.

Abstract

6533 Background: NCCN guidelines do not recommend routine surveillance imaging for distant failure (DF) after definitive treatment of head & neck squamous cell carcinoma (SCC). We hypothesized that there exists a subset of patients with sufficiently high enough risk for DF to benefit from surveillance imaging. This study attempts to define high risk cohorts of oropharynx (OP) and oral cavity (OC) patients. Methods: A retrospective review was conducted of patients with SCC of the OP or OC at a single tertiary care institution from 1994-2019. Patients were staged according to AJCC 7th edition and included in this study if they completed definitive-intent treatment and received 60 Gray or higher of radiotherapy (RT). Local, regional, and distant failure were estimated with cumulative incidence. Univariable & multivariable risk factors for DF were identified with Fine & Gray competing risk regression. Significant variables were compiled to calculate a risk score. Results: 863 patients were included (676 OP/187 OC). OC patients were 60.4% male, median age 61, with median follow up of 77.5 months. Smoking status was 27.3% current, 44.4% former, 28.3% never, with 30 median pack years. Disease was 57.3% T1-2, 42.7% T3-4, 55.6% N0-2a, 44.4% N2b-3. 94.1% had surgery & 34.3% had concurrent systemic therapy. OP patients were 87.9% male, median age 58, 96.3% HPV+, with median follow up of 60.8 months. Smoking status was 20.9% current, 44.5% former, 34.6% never, with 20 median pack years. Disease was 67.9% T1-2, 32.1% T3-4, 29.9% N0-2a, 70.1% N2b-3. 11.5% had surgery & 87.3% had concurrent systemic therapy. Specifically, 52.2% of OP patients received concurrent cisplatin, 10.6% concurrent cetuximab, and 24.5% other systemic therapies. 11.7% of patients experienced DF, of which 77% failed in the lung. Within the OC cohort, nodal stage 2b or higher was the only predictive factor (HR 3.26, p < 0.001), conferring a 3 year risk of DF of 34% vs 10%. Within the OP cohort, a high risk cohort of 87 patients (12.9%) was identified with a 3 year incidence DF of 22%, compared to 10% or less in lower risk cohorts. This high risk cohort consisted of active smokers treated with definitive RT and either concurrent cisplatin or no concurrent therapy, with at least T3 and N2b disease, as well as any patients treated with definitive RT and concurrent cetuximab. Conclusions: We identified groups of OC & OP patients with greater than 20% risk of developing DF at 3 years, the majority of which occurred in the lung. Surveillance imaging of the chest should be considered for patients meeting these criteria.