ESTRO 2021 Abstract Book

S1560

ESTRO 2021

on Figure 1. A mean dose difference of 0.5% ± 0.9% between calculated and measured dose was obtained. A homogenous dose distribution was obtained for phantom thickness ≤25cm. Above a thickness of 25cm, it was not possible to keep the maximal dose <110% while keeping the midline dose between 95% and 105%. The plan robustness was assessed my moving the isocenter 3cm in the sup-inf and lateral direction. The dose difference with the original plan was less than 2%.

Conclusion Dynamic arc therapy for TBI provides a homogeneous dose distribution, is robust in respect to patient motion and is applicability in small treatment rooms. It can be used for patient with thickness ≤ 25cm. For larger patient, higher photons energy should be used in order to avoid high hot spots. PO-1830 Investigating the efficacy of robust optimisation and evaluation in clinical lung SABR planning M. de la Bastide 1 1 Imperial College Healthcare NHS Trust, Radiation Physics and Radiobiology, London, United Kingdom Purpose or Objective To evaluate the clinical impact on plan dosimetry of incorporating motion and setup uncertainties into lung SABR plan optimisation and evaluation, in comparison to conventional PTV-based planning. Materials and Methods A total of 8 patients were selected retrospectively for lung SABR planning, covering a range of tumour locations and dose prescriptions in accordance with the UK SABR Consortium guidelines [1] . For each patient, 3 plans were generated in Raystation V9B using: (i) conventional PTV volume-based optimisation with 5mm ITV-PTV margin, (ii) ITV-based robust optimisation with a 5mm setup uncertainty and (iii) GTV-based robust optimisation including setup uncertainty and internal motion uncertainty using 4DCT image phases. Plan quality was compared using conventional SABR dosimetric parameters. Plan robustness was assessed by (i) re-calculating the dose distribution on each CT phase of the 4DCT dataset and (ii) using robust evaluation to determine the plan pass rate and worst-case-scenario (WCS) dose distribution under specified setup uncertainty. The dosimetric effect of 4D motion on the different plan types is being investigated using the CIRS respiratory phantom. [1] Stereotactic ablative body radiation therapy (sabr): A resource. SABR UK Consortium 6.1 (2019). Faculty of Clinical Oncology of The Royal College of Radiologists. Results All plans resulted in full D99 coverage of the prescription dose (PD) to the GTV in the nominal planning scenario, with the PTV, ITV-robust and GTV-robust plans giving a mean D99 GTV coverage of 113(±2)%, 112(±2)% and 110(±3)% respectively as seen in figure 1. Mean ITV coverage for each planning technique was 111(±2)%, 109(±3)% and 108(±4)% of the PD respectively. OAR doses showed no significant correlation to target coverage or optimisation technique used. The most significant influence on OAR dose was relative position to target. Plan robustness was also found to be comparable across planning methods, with mean D99 GTV coverage pass rates of 97(±7)%, 98(±4)% and 96(±5)% obtained for non-robust, ITV- based robust and GTV-based robust plans respectively. Using robust evaluation, WCS D99 coverage was found to be marginally greater in non-robust plans, with a mean GTV D99 coverage of 103(±5)% of the PD for non-robust plans and 101(±2)% using either robust optimisation method (see figure 1). For the 4DCT phase re-calculations, GTV coverage by the PD was achieved on each phase for all plans, indicating that all plans were robust against respiratory motion.