Positron Emission Tomography (PET)–Based Adaptive Radiation Therapy for Locally Advanced Squamous Carcinoma of the Head and Neck: Initial Results of a Prospective Trial

To report initial clinical outcomes of a prospective, non-randomized protocol of a PET-based adaptive radiation therapy (RT) approach for patients with locally advanced head and neck squamous carcinoma (HNSCC). Patients with a diagnosis of clinically node-positive nasopharyngeal, oropharyngeal, laryngeal, or hypopharyngeal squamous carcinoma were eligible and treated on an IRB-approved protocol. Patients underwent concurrent cisplatin-based chemotherapy with RT. Prescription doses were 70 Gy to sites of gross disease (both primary and nodal) and 60 Gy to at-risk but uninvolved nodal basins. Treatment was delivered in a concomitant boost technique over 35 fractions. Simulation was performed using a thermoplastic mask on a PET-CT simulator and was repeated following treatment fractions 10 and 22. Plans were modified at those points taking into account reduction in disease burden and anatomical changes secondary to weight loss. Additional weekly PET-CT scans were considered optional and acquired on all agreeable patients to monitor response. The primary study endpoint was locoregional failure (LRF). Secondary endpoints included treatment toxicity, distant metastases, and both overall (OS) and disease-free survival (DFS). Toxicity rates were scored using CTCAE v4.0. A total of 25 patients have enrolled, 20 of whom with > 6 months’ follow up form the basis for this analysis. Median follow up time is 15 months (range 6-39). Median age at diagnosis was 61 years (range 48-76). Primary tumor site was oropharynx in 13 patients, oral cavity in 1, supraglottic larynx in 3, hypopharynx in 1, and unknown in 3. 11 patients were p16 positive, 8 negative, and 1 unknown. Primary tumor stage was Tx in 3 patients, T1 in 2, T2 in 7, T3 in 7, and T4 in 1. Nodal stage was N1 in 2 patients, N2b in 13, N2c in 4, and N3 in 1. All patients completed cisplatin based chemotherapy concurrently, with 1 switched to cetuximab mid-treatment due to toxicity. 3 LRF events occurred, yielding a 2-year local control rate of 83%. All LRF events occurred in p16-negative patients. 4 patients developed distant metastases (2-yr rate = 22.5%), 2 synchronous with LRF. 2-year OS and DFS estimates were 100% and 72.3%, respectively. The most common chronic grade 2 toxicities were dysgeusia (54%), xerostomia (31%), and dysphagia (23%). No chronic grade 3 toxicities have occurred to date. Target volume reduction likely may be safely performed based on radiographic and PET responses occurring during a course of concurrent chemoradiation therapy for node-positive HNSCC. Toxicity rates are favorable, and additional investigation will continue to define the optimal integration of such approaches into disease management paradigms.