3506 - Incidence and Dosimetric Characteristics Associated with Mandible Osteoradionecrosis in Pediatric Patients Treated with Proton Therapy for Head and Neck Malignancies

A. Grippin¹, S. A. Copling², A. Kim², Y. Liu³, A. Nalin¹, H. Cheng⁴, Z. Zhang⁵, M. F. McAleer¹, D. R. Grosshans¹, S. L. McGovern¹, and A. C. Paulino¹; ¹Department of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, ²UTHealth Houston McGovern Medical School, Houston, TX, ³The University of Texas MD Anderson Cancer Center, Houston, TX, ⁴UT Southwestern Medical Center, Dallas, TX, ⁵Department of Sociology, Rice University, Houston, TX

Purpose/Objective(s): Osteoradionecrosis (ORN) is a feared complication in patients receiving head and neck radiation, but little is known regarding the incidence and causes of ORN in children. In this retrospective analysis of a prospectively collected dataset, we tested the hypothesis that ORN would be uncommon in children treated with proton therapy for head and neck malignancies and associated with maximum doses to the mandible greater than 60Gy. Materials/

Methods: We utilized a prospective database including all pediatric patients treated with protons for head and neck malignancies at a single large cancer center between December 2006 and February 2020. We gathered baseline demographic data, disease characteristics, treatment plan, and incidence of ORN from the electronic medical record. We evaluated treatment plans for each patient including prescribed dose, fractionation, target, technique, and actual dosimetry including calculated doses to the mandible for each patient. Comparisons of medians were completed with Mann-Whitney U tests. Effect size of each parameter was estimated with simple linear regression. Fischer’s exact tests were used to identify differences in incidence of ORN above threshold values.

Results: We identified 117 pediatric patients in our cohort. The most common histology was rhabdomyosarcoma (46%), and the most common involved site was the orbit (30%). The majority received passive scatter (n=81, 69%) and the remainder IMPT (n=36, 31%). In our cohort, only two patients (1.7%) developed ORN. Age, mean mandible doses, and radiation technique (3D vs. IMPT) did not differ between groups. Patients with ORN had higher doses to the mandible than those without ORN, including Dmax (p=0.029), D0.5cc (p=0.039), D1cc (p=0.036), V60Gy (p=0.01), V55Gy (p=0.02), and V50Gy (p=0.048). The most significant predictors of ORN by linear regression were mandible V50Gy (r²=0.053, p<0.001,), V55Gy (r²=0.094, p<0.01), and V60Gy (r²=0.17, p<0.001).

Dosimetric thresholds associated with development of ORN included Dmax>60Gy (9% vs. 0%, p=0.037), D0.5cc>60Gy (9.5% vs. 0%, p=0.031), D1cc>59Gy (10% vs. 0%, p=0.028), V55Gy>2cc (9% vs. 0%, p=0.034).


Conclusion: ORN is uncommon in pediatric patients treated with proton therapy for head and neck malignancies. Patients with mandible Dmax>60Gy, D0.5cc>60Gy, D1cc>59Gy, and V55Gy>2cc are at greatest risk for ORN.