ESTRO 2020 Abstract book

S170 ESTRO 2020

(difference of 0.04 ± 1.01 Gy, p=0.85) and the bladder (differences of 0.38 ± 2.52 Gy, p=0.68) in the training cohort. For the manual treatment plan validation cohort (Figure 2 middle) this difference was larger: 0.80 ± 1.30 Gy, p<0.001, and 5.43 ± 4.71 Gy, p<0.001, for the bowel bag and the bladder, respectively. If treatment plans were optimized with RapidPlan TM this difference decreased again to levels similar to the model fit with differences of 0.46 ± 1.54 Gy (p=0.16) and 0.44 ± 1.55 Gy (p=0.18) for the bowel bag and the bladder, respectively (Figure 2 right).

cm3 being located at different areas in the brain. Additionally, several organs at risk (OARs) structures were contoured accordingly. A dedicated framework utilizing the Eclipse Scripting Research API was used to determine gantry-table and gantry-collimator paths based on contoured structures by means of an A* path searching algorithm. These paths served as input for the multi-leaf sequence optimization using a research version of the VMAT optimization algorithm in Eclipse leading to a DTRT treatment plan for each brain case. Additionally, a HyperArc treatment plan was generated for each of the five cases. Resulting dose distributions for DTRT and HyperArc plans were compared based on DVH parameters. The deliverability of DTRT and the agreement between calculated dose distributions and measurements were assessed by gafchromic film measurements. Results The comparison of DVHs for the target shows a similar dose coverage for all brain cases. On average over all five cases mean doses for parallel OARs and near maximum doses for serial OARs improved for DTRT treatment plans by 5.5% and 5.0% relative to the prescribed dose, respectively. Measured and calculated doses show a gamma passing rate >99.5% for global 2%/2 mm criteria and a threshold of 10%. Conclusion The DTRT treatment plan generally performs better or similar compared to the HyperArc plan, depending on the case and on the OAR considered. The results demonstrate that DTRT has a great potential to reduce dose to OARs, while target coverage is preserved compared to HyperArc. This work was supported by Varian Medical Systems. PD-0305 SABR re-irradiation of pelvic cancer recurrences: photon vs proton beam therapy E. Glassborow 1 , J. Richardson 1 , M. Clarke 1 , L. Murray 2 , R. Speight 2 , L. Aspin 2 , S. Gregory 2 , J. Handley 1 , R. Chuter 1 1 The Christie NHS Foundation Trust, Christie Medical Physics and Engineering, Manchester, United Kingdom ; 2 Leeds Teaching Hospitals NHS Trust, Leeds Cancer Centre, Leeds, United Kingdom Purpose or Objective Patients who experience a pelvic cancer recurrence in or near the region that received initial radiotherapy for the primary disease, typically have few treatment options. OARs have often reached their dose constraint limits leaving standard re-irradiation (reRT) unavailable. Alternatives include surgery, which may be extensive, and chemotherapy, which is typically reserved for wide-spread or symptomatic disease. Despite OAR dose concerns, photon SABR reRT has been utilised with promising initial results but can still struggle to meet constraints. Proton beam therapy (PBT) could offer an improvement through characteristic distal dose fall off. This project aims to establish whether PBT can achieve reduced OAR doses for equal or improved target coverage, compared to photon SABR. Material and Methods PBT plans have been retrospectively created in Eclipse v13.7 on CT data sets for 15 patients treated with photon SABR reRT for pelvic recurrence reRT under NHS England’s Commissioning through Evaluation (CtE) programme between 2016-2017. The photon SABR plans were created in Monaco v5.1, using a VMAT technique and had been prescribed to a 5mm CTV-PTV expansion with an ALARP OAR approach. For the PBT plans a spot scanning technique was used, and plans were created to match the SABR prescription of 30Gy in 5 fractions and the CTV coverage

Conclusion An independent anatomy-based DVH prediction tool for rectal cancer patients was developed and showed to predict mean bowel bag and bladder doses accurately. Compared to manual treatment optimization techniques, RapidPlan TM treatment plans presented lower variation and a closer agreement between predicted and achieved results than manual planning. This prediction tool could also be used for general plan QA purposes and to investigate differences between planning techniques, e.g. manual and automatic. [1] Petit SF, van Elmpt W. Accurate prediction of target dose-escalation and organ-at-risk dose levels for non-small cell lung cancer patients. Radiother and Oncol 2015; 117:453–8. PD-0304 Treatment plan comparisons between dynamic trajectory radiotherapy and HyperArc M. Fix 1 , S. Mueller 1 , D. Frei 1 , W. Volken 1 , D. Terribilini 1 , D. Frauchiger 1 , A. Joosten 1 , A. Henzen 1 , E. Herrmann 1 , D.M. Aebersold 1 , P. Manser 1 1 Division of Medical Radiation Physics and Department of Radiation Oncology, Inselspital-Bern University Hospital, and University of Bern, Switzerland Purpose or Objective Treatments of brain tumors benefit from non-coplanar beam arrangements. One approach in this context is HyperArc applying a pre-defined set of non-coplanar arcs. Recently, we proposed a new technique called dynamic trajectory radiotherapy (DTRT), which additionally includes dynamic table and collimator rotations during beam on to further increase degrees of freedom for the selection of beam directions. In this work DTRT and HyperArc were compared in order to investigate the potential benefit of DTRT for brain tumors. Material and Methods In this study five clinically motivated brain cases were included. Thereby the target volumes ranged from 3 to 270