ESTRO 2021 Abstract Book

S278

ESTRO 2021

Conclusion Our results demonstrate that in treatment of large brain tumors, a treatment planning approach using Bremstrahlung x-ray based Linacs can achieve superior dose fall-off for intracranial hSRS compared to GKI. We dispel the myth that physical characteristics of Cobalt gamma-rays are needed to create the sharpest dose fall-off. Given the prevalence of Linac-based treatments compared to GKI, we present a novel optimized planning technique that achieves a new benchmark of sharpest dose fall-off. PH-0379 Influence of tumor site on neurovascular structure doses in proton therapy of pediatric brain tumors L. Toussaint 1 , S. Peters 2 , R. Mikkelsen 3 , S. Karabegovic 3 , C. Bäumer 2 , L.P. Muren 1 , L. Tram-Henriksen 4 , M. Høyer 1 , Y. Lassen-Ramshad 1 , B. Timmermann 2 1 Danish Centre for Particle Therapy, Aarhus University Hospital, Aarhus, Denmark; 2 West German Proton Therapy Centre Essen (WPE), West German Cancer Center (WTZ), Essen, Germany; 3 Aarhus University Hospital, Department of Neuroradiology, Aarhus, Denmark; 4 Aarhus University Hospital, Department of Pediatrics, Aarhus, Denmark Purpose or Objective Long-term survivors of pediatric brain tumors are at increased risk of developing neurovascular disease. Radiation-induced neurovascular late effects are currently being investigated, with radiation dose to the Circle of Willis (CW) being proposed as a risk factor. The aims of this study were to develop a delineation atlas of neurovascular structures, to investigate if the suprasellar cistern (SC) dose could be used as surrogate for the CW dose, and to study the influence of tumor location on the dose to neurovascular structures. Materials and Methods Within the frame of the HARMONIC prospective cohort study (https://harmonicproject.eu/), an atlas of the CW, the first segment of the Middle Cerebral Artery (M1) and the SC was developed by neuroradiologists. This contouring atlas was subsequently applied in treatment plans of 30 pediatric brain tumor patients previously treated with proton therapy (Fig. 1). The treatment plans were divided into three subcohorts, depending on tumor location: 10 central, 10 lateral hemispheric, and 10 posterior fossa tumors. The agreement between dose metrics (mean dose, Dmean; maximum dose, Dmax) to the CW vs. the SC was investigated. When target volume and SC were not overlapping, the shortest distance between the structures was calculated as the minimal Hausdorff distance (HDmin), and further related to the Dmean to the SC as a potential guide for dose sparing of neurovascular structures.

Results Across all 30 patients, the median Dmean/Dmax to the CW were 35 Gy(RBE)/52 Gy(RBE). In the three subcohorts, the median Dmean/Dmax to the CW were 53 Gy(RBE)/55 Gy(RBE) in the central tumors, 18 Gy(RBE)/25 Gy(RBE) in the lateralized tumors and 30 Gy(RBE)/49 Gy(RBE) in the posterior fossa tumors. There was a good agreement between the Dmax/Dmean to the CW and the SC for all three tumor locations (Fig. 2; left and middle). Across all 30 patients, cases with HDmin lower than 10 mm between the target and the SC received the highest Dmean to the SC (Fig. 2; right). For all central tumors, the SC and target were overlapping and for all but two cases, SC Dmean was equal to the prescription dose. In the lateralized tumors, cases with a HDmin larger than 20 mm between the target and SC had a SC Dmean of 0 Gy(RBE). In the posterior fossa tumors, all cases had a HDmin lower than 14 mm, and half of these cases had a SC Dmean of at least half the prescription dose (for HDmin < 1mm).