ESTRO 2020 Abstract book

S70 ESTRO 2020

strategy) or a CBCT with additional kV imaging of fiducial markers (FM) to correct for motion occurring during the CBCT analysis (CBCT+kV strategy). Material and Methods Treatment plans and motion data were gathered from 22 patients treated with 5 x 7-7.25Gy fractions using two 10 MV flattening filter free VMAT-beams. CTV-to-PTV margins were 3 mm posteriorly and 5 mm elsewhere. Patient setup included initial Calypso localization, CBCT check of rectum and bladder filling and additional kV or Calypso based position corrections prior to irradiation. Beam gate-off and couch corrections were applied during treatment if the Calypso-measured motion exceeded 2 mm posteriorly or 3 mm elsewhere. After treatment motion trajectories for the CBCT and CBCT+kV correction strategies were simulated by correcting the motion data for Calypso- guided couch adjustments. The kV-based couch adjustments were maintained in the simulated CBCT+kV strategy and removed in the CBCT strategy. The dosimetric effect of the motion was assessed by motion-including dose reconstruction. The dose reconstruction involved emulation of the 3D prostate motion as multiple isocenter shifts and recalculation of the motion-encoded plans by the treatment planning system. Motion inclusive DVH parameters for target and OAR structures (bladder and rectum) were compared to planned values. Results The total number of analyzed fractions with dose reconstruction was 103 for the Calypso- and CBCT+kV correction strategies and 102 for the CBCT strategy. Motion simulation is illustrated in fig.1. Table 1 presents differences in target and OAR DVH parameters between motion inclusive and planned dose distributions. In general, the PTV margins covered most of the motion and CTV dose deficits were small in each correction strategy. For individual fractions, larger dose deviations were seen, especially in the CBCT strategy. The volume of the OARs that received high doses increased due to motion, whereas lower dose volumes were less affected.

Conclusion Single CBCT-guided pre-treatment setup without further correction may lead to inaccurate treatment delivery with clinically relevant dose deficits in prostate SBRT. Additional pre-treatment position correction with kV imaging increases the accuracy and is adequate for most of the fractions. Position corrections based on continuous monitoring ensure high target dose coverage and minimizes the OAR doses best, although the dosimetric benefit of gating during the irradiation is small. PH-0125 Intra-fractional stability of MR-guided online adaptive SBRT for prostate cancer J. Heitmann 1 , M. Chamberlain 1 , L. Wilke 1 , M. Baumgartl 1 , J. Krayenbühl 1 , M. Zamburlini 1 , M. Mayinger 1 , N. Andratschke 1 , S. Tanadini-Lang 1 , M. Guckenberger 1 1 University Hospital Zürich, Radiation Oncology, Zurich, Switzerland Purpose or Objective MR-guided online adaptation for prostate stereotactic body radiation therapy (SBRT) aims to decrease safety margins and consequently reduce toxicity by compensation of inter-fractional non-rigid target and organ-at-risk variations. However, the process of online adaptation currently takes approximately 45 minutes during which internal movements remain unaccounted for. The aim of this study was to analyze the dosimetric benefit of adaptation and evaluate the robustness of the adapted plan over the time period of one treatment fraction. Material and Methods Baseline MR-scans at the MR-linear accelerator (MRIdian Linac System Version 5, ViewRay Inc.) were acquired for ten healthy male volunteers. A mock-prostate SBRT plan was generated with 5mm CTV-to-PTV safety margins and dose prescription of 5 x 7.25Gy. On a separate day, the volunteers were positioned on the couch for one hour and repetitive MR images were acquired every 15 minutes to assess the stability of the adapted plan of the day. PTV- and CTV-coverage as well as dose to organs at risk were analyzed. Results We observed a dosimetric benefit in 90% of volunteers in the adapted plan compared to the baseline plan (D 95 CTV increased 0.41-9.4 Gy, average 2.8Gy). The median D 95 CTV- and D 95 PTV coverage was improved from 34.8Gy to 35.5Gy and 30.7Gy to 34.6Gy, respectively. This benefit was not associated with higher dose to the organs at risk, most importantly the rectum. The median D 1cc rectum was lower in the adapted plans compared to the baseline plans copied onto the MR of the day with 33.33Gy vs. 32.76Gy. The benefit of online adaptation remained stable over 45 minutes for 90% of volunteers. The median D 95 CTV remained sufficient (timepoint 0: 35.5Gy, timepoint 60mins: 35.3Gy; p = 0.44). However, at 60 minutes, CTV- coverage in 30% of volunteers was below a threshold of 32.5Gy (29.3Gy, 30.6Gy, 32.0Gy). Importantly, in those three cases D 95 CTV of the adapted plan still exceeded the