ESTRO 2020 Abstract book

S918 ESTRO 2020

ischemic cardiac events). ACE cumulative risk was calculated twice for each patient (once for each pretreatment risk factors, 0 and 1) simulating two cohorts grouped on the basis of the two risk factors. Results The use of ABC mDIBH resulted in similar target coverage with no significant statistical differences with respect to FB plans. Median MHD was 3.7 Gy (range:2.8-6.2 Gy) in FB and 2.5 Gy (range:1.6-4.1 Gy) in mDIBH (p<0.05), resulting in absolute and relative reduction of 1.3 Gy (range: 0.2- 2.5 Gy) and 35.0% (range: 4.7-53.6%), respectively. The use of ABC mDIBH reduced MHD by 20% or greater in 90% of patients. Using ABC mDIBH, the dosimetric reduction in MHD translated in an average decrease of ACE 9-year excess cumulative risk from 0.5% to 0.3% for risk0 group and from 2.7% to 1.8% for risk1 group. The impact of ABC mDIBH on ACE cumulative risk becomes increasingly remarkable with patients’ age. For patients age >60 years, average ACE 9- year excess cumulative risk decreased from 1.4% to 1.0% for risk0 group and from 7.0% to 4.9% for risk1 group, respectively. Conclusion ABC mDIBH technique resulted in a significant reduction of MHD for left sided breast radiotherapy. Excess risk of ACEs can be remarkably reduced, in particular for patients age >60 years. PO-1595 MR guided tumour tracking on a high field MR Linac: feasibility and first experimental results S. Nill 1 , I. Hanson 1 , F. Costa 1 , M.J. Menten 1 , A. Wetscherek 1 , U. Oelfke 1 1 The Institute of Cancer Research and The Royal Marsden NHS Foundation Trust, Joint Department of Physics, London, United Kingdom Purpose or Objective The Elekta Unity MR Linac (MRL) consists of a high field 1.5 T MR scanner and a 7 MV linear accelerator equipped with an 80 leaf pair multi-leaf collimator, with the leaf motion being parallel to the patient caudal-cranial direction. The MRL allows 2D MR images to be continuously acquired during treatment to detect tumour motion in real-time. MLC based tumour tracking using these MR images is one promising technology that will allow further reduction of PTV margins and consequently improve healthy tissue dose sparing. Material and Methods Our in-house MLC tracking system DynaTrack developed for conventional linacs was adapted to support the MRL running non-clinical research firmware. Geometric tracking errors (root mean square of the difference between the location of a radiopaque marker and the centre of the leaf aperture opening as measured on the on-board EPID) were measured for an insert moving on a sine curve (sup-inf direction) with two breathing periods (T=4 s, 5 s) and three amplitude settings (A=5 mm, 10 mm, 15 mm) inside the Quasar MRI4D motion phantom. The motion was determined using a template matching algorithm on coronal MR images (coherent FFE sequence, update rate: 3 Hz, image latency: 200 ms) as input. A linear regression motion predictor with either a fixed look ahead time of 410 ms or for a variable time point between 243 ms and 486 ms depending on the time of the last received data point was applied within the tracking tool during the delivery. A four field treatment plan was delivered for three different arrangements using the fixed predictor setting: 1) no phantom motion – no tracking 2) phantom motion (T=5 s, A=10 mm) - no tracking 3) phantom motion – with tracking. GAFChromic EBT3 films were irradiated for each case to quantify the dosimetric impact. A comparison using dose profiles and gamma criterion pass rate between the non- tracked and tracked delivery with the results from the no phantom motion case as reference were performed.

Results Table 1 summarizes the results for the geometric tracking accuracy. Breaths per Minute Amplitude [mm] Geometric tracking error [mm] Predictor setting 12 5 0.40 Variable 12 10 0.45 Variable 12 10 0.80 Fixed 12 15 0.57 Variable 15 5 0.41 Variable 15 10 0.53 Variable 15 15 0.68 Variable Table 1: Results for the geometric tracking accuracy using MR images as input. The tracking error for all measurements is well below 1mm. The gamma pass rates for 2 %/1 mm / 10 % dose cutoff are 28.3 % and 98.7 % for the non-tracked and tracked delivery. Figure 1 shows a comparison for one exemplarily measured profile.

Conclusion We have demonstrated the feasibility of using MR images to track tumour motion using an MLC on the Elekta Unity system. Our framework improves the geometric accuracy compared to previously reported work (Glitzner PMB 64 15NT02) and the dosimetric evaluation does encourage further experiments for the delivery of complex IMRT treatment plans exposed to more realistic tumour movements. PO-1596 End-to-end test in MRgRT: 3D dose verification of a gated treatment with a porcine lung phantom P. Mann 1,2 , D. Stefan 1,2 , S. Katharina 2,3 , K. Christian P 1,2 1 German Cancer Research Center DKFZ, Medical Physics in Radiation Oncology, Heidelberg, Germany ; 2 National Center for Radiation Research in Oncology NCRO, Heidelberg Institute for Radiation Oncology HIRO, Heidelberg, Germany ; 3 University Hospital Heidelberg, Department of Radiation Oncology, Heidelberg, Germany Purpose or Objective To test and validate the entire treatment chain in MRgRT for a gated treatment in a realistic treatment setup combined with 3D dose verification Material and Methods An anthropomorphic porcine lung phantom (PLP) including a heart was combined with the PAGAT polymer gel (PG) serving as a target. The PG was stored inside a BAREX container (length=40mm, diameter=27mm) and sewed onto the mediastinum of the lung. In this study, the PLP was used to verify a gated treatment on a 0.35 T MR-LINAC (MRIdian, Viewray). For treatment planning, a CT (SOMATOM, Siemens) was acquired 1h prior to irradiation. The target volume delineation and plan calculation was carried out using the Viewray TPS. A total of 7 gantry angles with 39 segments and a D 50 = 4Gy was planned homogeneously to the PTV. The PTV contained the