Osteoradionecrosis as a complication following intensity-modulated radiation therapy or proton therapy in the treatment of oropharyngeal carcinoma.

6066 Background: Osteoradionecrosis (ORN) is a late complication of head and neck radiotherapy (RT) that negatively impacts survivorship. Although there is an abundance of literature reporting on ORN for photon-based RT, few studies have investigated the correlation between proton therapy and ORN; also, the current literature on ORN incorporates a broad mix of head and neck subsites. Therefore, we report our 10-year institutional experience to assess rates of ORN in a homogenous and consecutive cohort of patients with oropharyngeal squamous cell carcinoma (OPSCC) treated with curative-intent radiotherapy, representing the largest available institutional series. Methods: A consecutive cohort of 1564 OPSCC patients (1344 definitive, 220 post-operative) who received at least 50Gy were treated at our institution between 2013 and 2023 were included in this study. Patients were treated with either IMRT or proton therapy. CTCAE version 5 was used for ORN grading. Results: Overall ORN rate was 4.35%. Of 1389 patients who underwent IMRT, 56 (4.03%) developed ORN, vs 12/175 (6.86%) treated with proton therapy (hazard ratio [HR] 2.62, 95%CI 1.39–4.93, P=0.003). Median time to ORN in the IMRT arm was 25mo (range, 2mo–91mo). Median time to ORN in the proton arm was 23.5mo (2mo–45mo). Post-operative vs definitive treatment setting was not associated with the rate of ORN (univariate Cox HR 1.00, 95% CI 0.51–1.95, P=0.99). On subset analysis of the 1344 patients treated in the definitive setting, 47/1210 (3.88%) patients treated definitively with IMRT developed ORN as compared to 11/134 (8.21%) patients treated definitively with protons (univariable Cox HR 3.62, 95% CI 1.85–7.09, P<0.001). On multivariable analysis including treatment modality and use of chemotherapy, proton therapy was associated with increased hazard of ORN (HR 2.75, 95%CI 1.46–5.19, P=0.002). Concurrent chemotherapy was also independently associated with increased hazard of ORN (HR 3.34, 95%CI 1.05–10.65, P=0.041). A total of 10 out of 1564 (0.64%) patients developed CTCAE grade 3 ORN. The rates of grade 3 ORN were 2/175 (1.14%) in the proton cohort and 8/1389 (0.58%) in the IMRT cohort. This difference was not statistically significant on univariable Cox analysis (HR 2.44, 95%CI 0.51–11.60, P=0.26). Conclusions: The overall prevalence of ORN was 4.35%; the prevalence of >/= grade 3 ORNs was 0.64% in this consecutive cohort of patients with OPSCC treated with either IMRT or proton therapy. The overall prevalence of ORN of any grade was statistically higher for protons versus IMRT, a difference that was more pronounced in the definitive setting. Given the uncertainties with relative biological effectiveness calculations in proton therapy, avoidance of hot-spots, frequent replanning, and use of empirical proton-specific normal tissue constraints may help to reduce rates of ORN. Future work should explore the role of combination proton and photon treatment, especially in the definitive setting.