ESTRO 2021 Abstract Book

S1529

ESTRO 2021

PO-1801 Regional variation in relative biological effectiveness and pediatric brainstem toxicity L.F. Fjæra 1 , D.J. Indelicato 2 , K.S. Ytre-Hauge 3 , A.H. Handeland 4 , Y. Lassen-Ramshad 5 , L.P. Muren 6 , C.H. Stokkevåg 7 1 University of Bergen, Department of Physics and Technology, Bergen, Norway; 2 University of Florida, Department of Radiation Oncology, Jacksonville, USA; 3 University of Bergen, Department of Physics and Technology , Bergen, Norway; 4 Unversity of Bergen, Department of Physics and Technology, Bergen, Norway; 5 Aarhus University Hospital, Danish Centre for Particle Therapy, Aarhus, Denmark; 6 Aarhus University, Department of Medical Physics, Aarhus, Denmark; 7 Haukeland University Hospital, Department of Oncology and Medical Physics, Bergen, Norway Purpose or Objective Pediatric patients with brain tumors are at risk of brainstem toxicity following radiotherapy. Clinically, a fixed relative biological effectiveness (RBE) of 1.1 is used for children treated with proton therapy. However, studies have shown that the RBE varies with properties such as cell type, dose, the linear energy transfer (LET), and biological endpoint. Furthermore, the brainstem is often regarded as a single organ with uniform tolerance levels. However, there may be regional differences in radiation sensitivity within brainstem substructures, which have not yet been scrutinized with respect to the effects of variable RBE. To investigate if brainstem toxicity could be related to increased RBE, we compared variable RBE- weighted doses and LET within the brainstem and substructures for pediatric brain tumor patients with and without toxicity. Materials and Methods A cohort of 36 anonymized patients were selected from 954 pediatric brain tumor patients treated with double-scattering proton therapy between 2006-2017. Nine children with symptomatic brainstem toxicity of CTCAE grade 2+ were each matched to three controls based on age (± 1.5 years), diagnosis, adjuvant therapy, and brainstem dose parameters (median; D 50% ± 2 Gy, and maximum; D 0.1cc ± 2 Gy). Using T1/T2-weighted MRI scans fused with CT images, the brainstem substructures pons, midbrain, and medulla oblongata were segmented. To approximate physiologic fiber tracts, the pons was further divided into transversal zones (posterior, middle posterior, middle anterior, and anterior). To obtain LET and variable RBE-weighted doses, all 36 patients were imported into the FLUKA Monte Carlo code for recalculation. Dose- averaged LET (LET d ) and doses using the phenomenological models of Rørvik (ROR) and McNamara (MCN) were scored in FLUKA. RBE-weighted doses and LET d to the brainstem and brainstem substructures were compared between cases and controls. Results The median variable RBE-weighted doses from ROR and MCN were on average higher for the cases vs. controls for the entire brainstem, pons, and midbrain. The median LET d was on average higher for the cases in all substructures, with the largest difference of 16% in the midbrain (Figure 1). For the maximum RBE-weighted dose, the differences between cases and controls were overall small with narrow confidence intervals (Figure 2).

Conclusion When considering the entire brainstem, both variable RBE doses and LET d

values were on average higher for cases compared