ESTRO 37 Abstract book

S986

ESTRO 37

Research, Maastricht, The Netherlands 2 Mevion, Medical Systems Inc, Littleton- MA, USA

planning system and the CT-to-ED table used for all the clinical images (figure1).

Purpose or Objective To model the proton beam-shaping component of a pencil beam scanning system using a Monte Carlo (MC) toolkit for commissioning and as an independent validation of The proton beam nozzle was modelled using TOPAS, a MC particle simulator. The nozzle includes the Adaptive Aperture dynamic collimator, consisting of moving leaves which travel from spot to spot, to shape the beam downstream of the nozzle’s exit window. Its unique energy selection system, slabs of polycarbonate interposed in the beam following a given pattern, was also modelled. First, simulations with an open aperture were made to assure good agreement with experimental dose distributions for a set 42 energies ranging between 51 MeV to 230 MeV created by range shifters. Then, the Adaptive Aperture was modelled and simulated for different positions and energies. Finally, a plan to cover a 1L water sphere with a uniform dose was simulated and compared with the experimental data. the commercial TPS. Material and Methods

Figure 1. CT to relative electron density graph. We scanned the Imaging and Radiation Oncology Core (IROC) Lung Phantom, that contains an insert, which is part of the lung with a centrally located water equivalent density CTV (3cmx5cm) and we defined a PTV with an expansion of 0.5 cm in axial plane and 1 cm in longitudinal plane. We passed the credentialing IROC process for the phantom irradiation component. We used the same calculations settings than for tumor dose verification with two VMAT arcs to deliver 6 Gy to the PTV in 1 fraction and a 6MV Elekta accelerator equipped with Agility multileaf collimator. The isocentre was set at the centre of the PTV. The dose distribution was calculated using the Monte Carlo algorithm with 0.3 cm volumetric grid size and mean variance 2%. The plan was calculated with modified lung ED with the same monitor units and segments. The ED values were modified to simulate the different densities obtained from different systems, kVp and FOV. Results Dose errors were evaluated by comparing changes in mean dose for the PTV relative to the baseline. The baseline was set to real electron density relative to water for one lung insert; inhale lung ED 0.190. For 6 MV photons to produce 1.6% PTV mean dose error required a variation in relative ED of 0.06 (from 0.19 to 0.25) but the dose volume histogram was quite different (figure 2) .

Figure 1. Adaptive aperture modelled in TOPAS, placed downstream to the range shifter (energy selection system). Results With the open aperture, the 80% distal dose fall-off (R80) for both the experimental and simulated integrated depth dose curves showed an agreement under 1 mm, which was within the experimental resolution. The use of the adaptive aperture showed a decrease from 15% to 97% in the lateral penumbrae (80-20%), depending on its opening.

Figure 2: Dose volume histogram for PTV. Lung relative electron density (ED) was forced to 0.25 and 0.19 for plan 6Gy1ssdensidades-QA19 and plan 6Gy1ssdensidades-QA2 respectively. The plan 6Gy1ssdensidades corresponds to the original plan with a mean lung ED of 0.346. Conclusion Although small PTV mean dose errors where obtained due to small lung ED variations in lung SBRT, the dose volume histogram has shown significant differences in PTV coverage. Different CT to ED tables must be used for different kVp, FOV and image system in lung SBRT VMAT with Monaco treatment planning system. EP-1828 Monte Carlo simulation of the dynamic collimator in a pencil beam scanning system I.P. Almeida 1 , M. Sure 1 , G. Vilches-Freixas 1 , J. Cooley 2 , T. Zwart 2 , A. Langenegger 2 , F. Verhaegen 1 1 Maastricht Radiation Oncology MAASTRO clinic, Physics

Figure 2. Lateral penumbra (measured in the x-direction)