Institutional review board (IRB) approval was obtained for this phase I/ trial in which eligible patients gave written informed consent and were subsequently enrolled. Eligibility criteria included patients with intermediate- or high-risk SCC of the oropharynx, as defined by Ang and colleagues.4 Patients considered to have intermediate-risk features included: (1) HPV-positive, N2b to N3 tumors and smoking history of at least 10 pack-years; or (2) HPV-negative, T2 or T3 tumors, and smoking history of less than 10 pack-years. Patients at high-risk had: (1) HPV-negative tumors and at least 10 pack-years of smoking history, or (2) T4 disease regardless of HPV status. High-risk HPV subtype positivity was assessed by fluorescence in-situ hybridization analysis, and the expression of EGFR, p16, and p53 was assessed by immunohistochemistry.
Each patient in the study received fractionated radiation therapy concurrent with intravenous platinum-based chemotherapy either weekly or every three weeks. For safety purposes, and to clearly define SRS-specific toxicities, no systemic therapy was allowed after the final week of fractionated radiotherapy. Radiotherapy was delivered Monday through Friday in 30 daily fractions as follows: areas of gross disease and high risk received 60 Gy or 66 Gy in 2 Gy or 2.2 Gy fractions and areas of low risk received 54 Gy or 52.5 Gy in 1.8 Gy or 1.75 Gy fractions, respectively. An IMRT technique was utilized to spare swallowing organs at risk.1,16,17 One week after IMRT completion, patients received a single fraction SRS boost of 8 Gy corresponding to a total biological effective dose (BED) of 79.1 Gy in 2 Gy equivalent dose fractions assuming an alpha-beta ratio for tumor of 10 Gy (EQD10/2). Over time, the boost dose was escalated to a single fraction SRS boost of 10 Gy (corresponding to 83.8 Gy or 76.7 Gy EQD10/2) if no dose limiting toxicities developed with the 8Gy arm according to study design. Lastly, the final seven patients on the protocol received SRS boost 10 Gy in 2 fractions (corresponding to 72.5 Gy EQD10/2) after DLTs were identified in the single-fraction arm. Figure 1a diagrams patient flow through this protocol.
Radiation Planning and Technique
Patients enrolled in the protocol underwent contrast-enhanced magnetic resonance imaging (MRI) of the neck and non-contrast computed tomography (CT) simulation before IMRT. Figure 1b displays example IMRT and SRS plans for an enrolled patient. For the IMRT portion, the clinical target volume (CTV) receiving 60 Gy or 66 Gy included the primary tumor plus region(s) of grossly involved lymphadenopathy and high-risk subclinical regions with 7 mm expansion. Initial 16 patients (11 of the 8 Gy group and 5 in the 10 Gy in single fraction group) out of the 34 patients were treated to 66 Gy followed by the boost. The rest of the 18 patients were treated with 60 Gy followed by the boost based on the departmental protocol. The CTV receiving 54 Gy or 52.5 Gy included all other regions at risk for harboring microscopic disease. The planning target volume (PTV) was defined as a 3 mm expansion of the CTV. The boost target volume was determined by repeating CT simulation and MRI after the last day of IMRT, and fusing the T1 post-contrast fat suppression MRI sequence to the CT images. Boost gross tumor volume (GTV) included areas of primary and nodal gross tumor delineated on the CT-MRI fusion, and was uniformly expanded by 2 mm to create a boost PTV. In patients whose tumors demonstrated complete response on post-IMRT MRI, the pre-IMRT gross tumor was delineated to form the boost GTV. Two-dimensional kilovoltage imaging was used for set-up and intra-fraction verification prior to delivery of each field. The boost was delivered five working days after completion of IMRT in for all enrolled patients. Patients were evaluated 6 weeks after treatment completion, and followed by imaging every 3 to 4 months for two years and then annually up to 5 years.
Dose constraints for the IMRT portion were applied to the involved pharyngeal constrictor muscles (V60Gy ≤ 0.5 cc), uninvolved pharyngeal constrictor muscles (D33 < 45 Gy; D15 < 60Gy; mean dose < 35 Gy), involved supraglottic larynx (SGL) (mean dose < 30 Gy), and uninvolved SGL (mean dose < 18 Gy). Dose constraints for the boost portion included coverage of the boost PTV by the 80% isodose line (ranging between 60-90% isodose lines), with “hot spot” kept within the PTV. Organs at risk were specified to receive no more than 30% of the boost dose. The cumulative maximum spine dose for the combined SRS and IMRT treatment was limited to no more than 50 Gy (EQD2). No carotid artery constraints were applied.
Adverse event reporting
Acute and late adverse events including mucositis, dysphagia, xerostomia, neurologic toxicity, skin toxicities, vascular effects, and constitutional symptoms related to treatment were reported and scored for severity using the NCI Common Terminology Criteria for Adverse Events (CTCAE) version 4.03.18 A dose-limiting toxicity (DLT) was defined as any treatment-related CTCAE grade 3, 4, or 5 toxicity that was not present prior to boost. All reported DLTs were verified by the study chair and Data Safety Monitoring Committee before final determination that a DLT did in fact occur. Initial SRS boost dose was 8 Gy with a plan to escalate to the maximum tolerated dose (MTD), defined as the highest dose level at which ≤ 33% of patients experience DLT. The study was designed to be considered complete when either of the following events occurred: (1) the MTD for a cohort was reached, or (2) the highest protocol dose level was treated and tolerated where therapy was likely to be tumorcidal per the determination of investigators.
To assess adverse functional and subjective symptom effects from treatment, patients completed the PSS-HN questionnaire and the M. D. Anderson Symptom Inventory Head and Neck Module (MDASI) core and head-and-neck surveys at baseline, treatment completion, and at routine follow-up appointments over the subsequent 24 months.19,20
Percutaneous endoscopic gastrostomy (PEG) tubes were prophylactically placed only in patients who met departmental criteria, which included initial abnormal swallowing pattern as assessed by a speech pathology team and multiple comorbidities. Patients who lost more than 10% body weight during IMRT typically received reactive PEG placement. Overall duration of PEG dependence was calculated from date of boost completion. Radiologic outcomes were assessed by head-and-neck radiologists. Objective response was calculated from assessments performed at baseline, 6-8 weeks after completion of treatment, and during the follow-up period using RECIST version 1.1 criteria.
The study was a “7+2+3+3” dose escalation design. For a given SRS boost doselevel, a total of seven to fifteen patients will be enrolled. For a given dose, there will be up to four cohorts of subjects enrolled (Cohorts 1, 2, 3, 4), with maximum sample sizes of 7, 2, 3, and 3, respectively. Patient enrollment in the study, as well as the decision to escalate, add a new cohort, or stop the trial, were restricted by the number of subjects who developed DLTs during their 90-day follow-up. The sample size of this study was not to exceed 45 patients, to allow for a maximum of 3 SRS dose cohorts containing 15 patients each.
The primary objective of the study was safety, as measured by development of DLTs during the follow-up period. The secondary objectives, for which we used statistical analysis, was to estimate the failure rates (overall, local, regional, and distant) using standard methods for proportions with their corresponding exact binomial 95% confidence intervals, as well as local, regional, and distant control rates (calculated as 100 minus failure rates). The Fisher’s exact test was used to compare those subjects who received 8 Gy versus 10 Gy for each of the failure rates. Kaplan-Meier curves were used to portray the outcomes. A result was considered statistically significant at the p<0.05 level of significance. All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC).