ESTRO 2023 - Abstract Book

S787

Monday 15 May 2023

ESTRO 2023

Materials and Methods Four specialist therapeutic radiographers reviewed 40 CBCTs from the first 10 patients treated in the UK. The 4 radiographers all had cardiac anatomy training from a radiologist and had experience in delivering cardiac SABR treatment. Anonymised images were imported into a Varian training system from 3 UK centres using CBCTs from both Elekta and Varian linacs. A series of image matches were conducted by each radiographer: a manual match (manual), an automatic match to the heart structure (auto) and the auto match followed by manual adjustment to the PTV (PTV), all using three degrees of freedom (DoF) only. The auto and PTV matches were also repeated using 6 DoF. Inter-observer variability was quantified using 95% limits of agreement from a modified Bland-Altman analysis.

Results Table 1. Limits of agreement (mm) (mm)

Lateral Vertical Longitudinal Pitch Yaw Roll

3DoF Manual 1.69 3DoF Auto 0.94

1.87 1.05 2.06 0.99 1.24

2.03 1.40 2.11 1.06 1.68

- - -

- - -

- - -

3DoF PTV 6Dof Auto 6DoF PTV

1.57 0.85 1.06

0.58 0.38 0.52

0.65 0.44 0.51 The table shows the limits of agreement were smallest in the automatic matches which suggests the algorithm matching to the heart structure is reliable. A manual adjustment from the auto match to the PTV is clinically appropriate to optimise target coverage. The mean difference between the 3DoF PTV match and the 6DoF PTV match is 0.6 mm, which could be clinically significant in a target with small margins and OAR in close proximity to high dose gradients. The limits of agreement were smaller in the 6DoF auto and PTV matches than the 3DoF matches. This demonstrates that when the rotations are corrected for as part of the match, matching to the PTV is more consistent. Conclusion This data shows that a 6DoF CBCT image match has less variability and therefore more accurate treatment delivery. Therapeutic radiographers have no clinical experience of matching to a target in the heart, this multi-centre study addresses the gap in the knowledge base. OC-0940 Assessment of cardiorespiratory motion for cardiac radioablation using a digital 17-segment model R. Stevens 1 , C. Hazelaar 1 , M. Bogowicz 1 , R.M. ter Bekke 2 , P.G. Volders 2 , K. Verhoeven 1 , D. de Ruysscher 1 , J.J. Verhoeff 3 , M.F. Fast 3 , S. Mandija 3 , J. Cvek 4 , L. Knybel 4 , P. Dvorak 4 , O. Blanck 5 , W. van Elmpt 1 1 GROW - School for Oncology, Maastricht University Medical Center+, Radiation Oncology (Maastro), Maastricht, The Netherlands; 2 CARIM - School for Cardiovascular Disease, Maastricht University Medical Center+, Cardiology, Maastricht, The Netherlands; 3 University Medical Center Utrecht, Radiotherapy, Utrecht, The Netherlands; 4 University Hospital and Faculty of Medicine, Oncology, Ostrava, Czech Republic; 5 University Medical Center Schleswig-Holstein, Radiation Oncology, Kiel, Germany Purpose or Objective The impact of various motion management strategies, such as the internal target volume (ITV) approach or treatment in breath-hold, for stereotactic arrhythmia radioablation (STAR) is not fully understood. We developed a comprehensive framework using a digital CT phantom that allows simulation of cardiorespiratory motion in combination with different motion management strategies to gain more insight in the impact of cardiorespiratory motion on STAR. Materials and Methods The 4D extended cardio-torso XCAT CT phantom (Segars, 2010) was expanded with the 17-segment heart model, a standard in cardiology used to describe left ventricle (LV) anatomy. Segments were parametrized in XCAT to create heart lesions based on the anatomical region definitions from the 17-segment model. This allowed placement of targets within the phantom based on these standardized heart regions. Next, we simulated cardiac (5 phases) and respiratory-gated (10 phases) 4D-CTs applying cardiorespiratory motion during free-breathing and breath-hold (cardiac motion only). Translation of the heart due to breathing was set to population-averaged values measured in VT patients, i.e. 14, 8, and 0mm in the Sup-Inf, Ant-Pos, and Left-Right direction, respectively. Cardiac contraction was set to baseline XCAT values. Next, the targets were extracted from all phases and used to construct ITVs that encompass the complete cardiac and/or respiratory target motion to quantify cardiorespiratory motion (Figure 1). In addition, we investigated the clinical imaging scenario when cardiac motion is not fully captured during a 4D respiratory CT scan for radiotherapy planning. Hereto, ITVs that