ESTRO 2020 Abstract Book

S179 ESTRO 2020

still occurs. Applying noninvasive mechanical hyperventilation (causing hypocapnia) with pre- oxygenation enables prolonged breath-holds (PBHs; ~5min). A complete radiation fraction could be delivered during a single PBH. However, information on motion during PBHs is lacking.The aim of this pilot study is to assess in- and exhale PBH durations and to quantify diaphragm motion during PBHs. Material and Methods Six healthy volunteers were trained to be non-invasively hyperventilated through a face mask causing safe hypocapnia to perform PBHs of at least 3min. In each of two MR sessions volunteers were asked to perform one inhale and one exhale PBH. During each PBH, a 3D cine- MRI was acquired (balanced turbo field echo: TE, 1.51ms; TR, 3.0ms; resolution, 1x1x1.6mm3; field-of-view, 450x400x220mm3) yielding a 3D dynamic image every 8s. In- and exhale PBH durations were recorded. Dynamics were registered to the first dynamic (i.e. the reference) using the right diaphragm dome as region of interest, to obtain diaphragm motion in craniocaudal direction. Three dynamics over the course of PBHs and three dynamics towards the end of PBHs were used to linearly fit the velocity of diaphragm motion, with the diaphragm dome position at the start of the MRI acquisition defined as (0mm), and velocity defined as slope of the fit (in mm/min). We tested for differences in median velocity during in- and exhale PBHs (Wilcoxon signed-rank test, α=0.05). To investigate the reproducibility of PBHs recorded in respectively MR sessions 1 and 2, differences of the velocity per volunteer were calculated. Results Overall, in- and exhale PBH durations ranged from 2min2s to 8min2s, and from 1min45s to 6min21s, respectively (Figure 1), with diaphragm motion varying from 10–45mm. V02 was excluded from analysis due to PBH duration <3min. For the remaining volunteers who performed PBH of >3min, velocity of diaphragm motion during inhale PBHs (n=8; median 3.1 mm/min; range 1.0–4.5 mm/min) was significantly slower than during exhale PBHs (n=8; median 5.9 mm/min; range 1.8–7.9 mm/min) (Table 1). Per volunteer, the difference in velocity of diaphragm motion between the two MR sessions varied from 0.1–1.6 mm/min during inhale PBHs, and from 0.4–2.5 mm/min during exhale PBHs.

Conclusion Performing in- and exhale PBHs of 3min and more during MRI is feasible. The velocity of the right diaphragm dome motion was slower during inhale than during exhale PBHs. To make PBHs effective in clinical radiation therapy and to improve reproducibility, we need to compensate for the gradual lung deflation that causes motion during PBHs. OC-0340 Intrafraction motion during SBRT in deep inspiration breath-hold for liver metastases L.B. Stick 1 , I.R. Vogelius 1 , S.N. Risum 1 , M. Josipovic 1 1 Rigshospitalet, Department of Oncology, Copenhagen, Denmark Purpose or Objective To assess intrafraction motion in patients with liver metastases treated with stereotactic radiotherapy (SBRT) in deep inspiration breath-hold (DIBH). Material and Methods Ten patients treated with liver SBRT in DIBH in a pilot study between July 2018 and August 2019 were analyzed. Prior to imaging for treatment planning, three gold markers were percutaneously implanted in the tumour vicinity using ultrasound guidance. Patients were trained in voluntary DIBH with visual guidance (RPM system). Gating window of 2.5-3.0 mm was used for all DIBH imaging and treatment delivery. Three DIBH CTs were acquired during imaging for radiotherapy and the DIBH CT with median position of the fiducial markers was used for treatment planning. DIBH CBCTs were performed before (pre- treatment CBCT) and after (post-treatment CBCT) each of the three treatment fractions. Daily patient position correction was performed online using rigid registration (x, y, z and yaw) on fiducials. During the volumetric modulated arc treatment, planar 2D kV images were acquired for every 10° of the gantry rotation for 5 patients. It was only possible to evaluate the CC position on these images due to 2D geometry. Results All patients had one liver metastasis. In one patient, two fiducials were placed close together and appeared as one on CT and CBCT. In another patient, one marker was split in two parts during implementation and appeared as two markers on CT and CBCT. Median 3D distance from marker to center of GTV on the planning CT was 3.1 cm (range 0.4 to 5.0 cm). The median 3D difference in marker position between the three DIBH CTs was 0.3 cm (range 0.0 to 0.9 cm) for all patients. The median 3D difference in marker position between the pre-treatment CBCT and the post-