ESTRO 38 Abstract book

S1067 ESTRO 38

EP-1957 An MR-based simulation procedure for individual assessment of tumor motion in MR-guided lung SABR M. Palacios 1 , J.R. Van Sornsen de Koste 1 , F.O.B. Spoelstra 1 , C.J.A. Haasbeek 1 , A.M.E. Bruynzeel 1 , B.J. Slotman 1 , F.J. Lagerwaard 1 , S. Senan 1 1 VU University Medical Center, Radiation Oncology Department, Amsterdam, The Netherlands Purpose or Objective To evaluate a simulation protocol to account for irregular breathing patterns in MR-guided SABR for lung tumors, and to report on the reproducibility of GTV position and duty- cycle efficiency during treatment. Material and Methods A workflow for initial simulation for MR-guided SABR was developed using MR-planar acquisitions (TR/TE = 2.41ms/1.09ms, FA=60˚, 4 frames-per-second) in sagittal, coronal and axial planes, allowing for an individualized assessment of tumor motion in three directions for each patient. MR-planar acquisitions allowed the choice of the most suitable breathing-phase for SABR delivery and evaluation of the tracking performance during simulation. A 4D-CT scan was omitted. Ten patients who exhibited variable tumor motion and were treated using two dose schemes (8x7.5Gy and 5x11.0Gy) were identified for this study. Treatment was delivered with GTV contour tracking during repeated breath-holds based on real-time deformable image registration using MR-planar acquisitions in the sagittal plane. PTV (GTV+3mm) was used as gating window. Treatment delivery performance was analyzed by quantifying the fraction of GTV inside the PTV in each frame during beam-on and corresponding duty-cycle efficiency for a total of 67 fractions. Results Substantial and variable tumor motion during respiration was observed: range 0.3 – 6.5 mm (AP), 0.2 – 4.7 mm (LR) and 1.5 – 27.1 mm (CC). Five patients were treated during shallow inspiration and five during expiration breath-hold. Reasons for expiration breath-hold were: lateral tumor movements, irregular tumor trajectories, and fast and substantial tumor motion in CC-direction. Patients treated in expiration breath-hold with MR-guided SABR exhibited better reproducibility of the GTV inside the PTV and higher duty-cycle efficiencies, than patients treated during shallow inspiration. On average across all fractions, 96.9% (SD: 2.1%) and 94.3% (SD: 2.2%) (p<0.001) of the GTV area displayed on the MR-cine during treatment beam-on was inside the PTV for expiration and inspiration patients, respectively. Corresponding duty-cycle efficiencies were 71.9% (SD: 10.5%) and 64.5% (SD: 16.0%) (p=0.016), respectively.

on the delivery time and includes a scatter component arising from treatment using MV beams. These factors may affect the image quality of in-treatment 4D-CBCT. This study aimed to quantitatively evaluate the image quality of in-treatment 4D-CBCT with various prescription doses (PDs) for accurately assessing tumor location in VMAT for stereotactic body radiation therapy (SBRT) of lung tumor. Material and Methods Spherical targets of diameters 10, 20, and 30 mm inserted in a dynamic thorax phantom moved sinusoidally with respiratory cycles of 4 s and amplitudes (A) of 5 and 10 mm along the superior-inferior direction. The target volumes were contoured from 4D-CT and merged into internal target volume (ITV). The PTV was defined by adding a uniform 5-mm margin to the ITV. The treatment plan was created with a D95 prescription of 2, 6, 7.5, 10, and 12 Gy for PTV using a single-arc VMAT with 6 MV. Pre- treatment 4D-CBCT (Elekta Symmetry™) scans were performed with an acquisition time of 3 min for the phantom setup, and those images were defined as reference images. The in-treatment 4D-CBCT images with various PDs were compared with the reference images using image quality metric. The image quality was evaluated by using signal-to-noise ratio (SNR), contrast-to- noise ratio (CNR), and dice similarity coefficient (DSC; to evaluate the spatial overlapping of target volumes with pre- and in-treatment 4D-CBCT). A Kruskal–Wallis test was performed for the statistical analysis of the results obtained with various PDs, and a P value < 0.05 was regarded as a significant difference. Results The figure shows representative images of pre- and in- treatment 4D-CBCT for the 10 mm target with A of 10 mm. The streak artifacts on the in-treatment 4D-CBCT images decreased as the PDs increased from 2 Gy to 12 Gy. For the 10 mm target with A of 10 mm, the mean values (± SD) of the SNR and CNR increased from 9.8 ± 0.8 to 17.4 ± 1.5 and from 6.4 ± 1.1 to 8.3 ± 1.4, respectively, as the PDs increased from 2 to 12 Gy. The mean values (± SD) of the DSC were not statistically significant from 0.75 ± 0.15 to 0.7 ± 0.11 (P = 0.669) as the PDs increased from 2 to 12 Gy. A similar tendency was observed for target sizes of 20 and 30 mm with A of 5 and 10 mm, respectively.

Conclusion PD is an important factor that affects the image quality in in-treatment 4D-CBCT image acquisition. Moreover, in- treatment 4D-CBCT obtained with PDs > 2 Gy is acceptable for assessing tumor location in VMAT for SBRT of lung tumor.

Conclusion An MR-based workflow for simulation prior to SABR for lung patients allowed individualized assessment of tumor motion and irregular breathing trajectories. Complex motion patterns observed led to use of SABR in expiration