The application of first-line therapy for low-tointermediate risk oropharynx squamous cell carcinoma (OPSCC) remains highly variable. There is recent high-level evidence suggesting a nonclinically meaningful difference in favor of primary radiation for this population and additional studies are underway or in planning. Here, our objective is to describe swallowing-related outcomes after primary singlevs multi-modality intensity-modulated proton therapy (IMPT) or transoral robotic surgery (TORS) using patient-reported and clinician-graded measures. Materials/Methods: 215 patients with HPV/P16+ T1-T3, N0-2b (AJCC 7) OPSCC were included from a prospective longitudinal registry. Patients were grouped by single modality (IMPT 20/82, 24%; TORS 62/82, 76%) vs multi-modality treatment (IMPT 67/133, 50%; TORS 66/133, 50%). The primary endpoint, MD Anderson Dysphagia Symptom Inventory (MDADI) was collected in addition to DIGEST (Dynamic Imaging Grade of Swallowing Toxicity; per MBS) at baseline, 3-6 months and 12 months after treatment, respectively. MDADI was compared between groups using two-way ANOVAs and DIGEST grade was compared between groups using Kruskal-Wallis test with post-hoc Dunn's test and sidak correction. Results: Concurrent chemoradiation (IMPT 67/87, 77% vs TORS 41/128, 32%, p<0.001) and age (mean, SD: IMPT 62.4 AE 8 vs TORS 58.8 AE 9, p Z 0.003) differed by group, while T-classification did not, p Z 0.48. Baseline and 3-6 months MDADI scores did not differ between treatment groups (p Z 0.74). After adjusting for covariates, MDADI scores at 12 months were significantly different (mean, SD: IMPT 81.1 AE 14; TORS 88.8 AE 12 ,p Z 0.001), with a significant interaction found between single (IMPT 75.1 AE 20; TORS 91.6 AE 11) and multi-modality treatment (IMPT 82.6 AE 12; TORS 86.2 AE 12, p Z 0.02). Dysphagia grade per DIGEST differed by primary treatment strategy favoring TORS at 3-6 months (DIGEST >0, IMPT 31/46, 67%; TORS 36/66, 55%, p Z 0.02). Conclusion: Prospective longitudinal evaluation of swallowing function, using patient-reported and clinician-graded measures, favor the first-line application of TORS in selected low-to-intermediate risk OPSCC. The differences were most pronounced for single modality therapy and should inform future trial designs.
