S2273
Physics - Intra-fraction motion management and real-time adaptive radiotherapy
ESTRO 2026
images were acquired at 5Hz perpendicular to the treatment beam throughout 6-11 fractions per patient and used to reconstruct the 3D marker motion during treatment using a probability-based 2D-to-3D estimation method. Motion-including dose reconstruction was performed by incorporating target motion in a motion-encoded plan as multiple isocenter shifts and calculating the dose of this plan in the treatment planning system (Eclipse). For each patient, the motion-including CTV of involved LNs (CTV-LN) V95% and D98% were compared with the planned values for each fraction and averaged over all imaged fractions of the same treatment plans. Results: Results: Motion substantially distorted the dose at individual fractions (Figure 1,2) such that the CTV-LN coverage criterion V95% ≥ 99% was not fulfilled at 41 out of 141 investigated fractions (29%). For some patients, replacing initial motion-sensitive treatment plans with adaptive plans during the treatment course greatly improved the motion robustness (Patients 4,5a,6,9 in Figure 2). Motion on average reduced V95% by 1.6±4.5% compared to the plan for individual fractions and 0.5 ± 0.9% when averaged over all imaged fractions of a plan. The motion-induced reduction in D98% was on average 1.2 ± 1.8% for individual fractions and 0.4 ± 1.0% when averaged over all imaged fractions.
from 95% to 100%, and 2% (1/53) adjusted to 98%, while 68% maintained 100%. Mean absolute motion magnitude was 1.2 ± 1.0mm (left-right), 2.6 ± 2.6mm (superior-inferior), and 1.4 ± 1.1mm (anterior- posterior), with mean 95th percentile values of 1.5(-2.0 to 7.9), 4.0(-0.7 to 14.8) , and 1.9(-2.0 to 5.8)mm, respectively. Mean PTV D ₉₅ % was 100.2%; mean organ doses were kidneys 7.3Gy, liver 4.8Gy; D ₀ . ₅ cc for duodenum and stomach were 27.4Gy and 32.9Gy; D ₀ . ₁ cc for great vessel was 42.5Gy. Conclusion: CMM integrating automatic gating and real-time tumor tracking was successfully implemented for pancreatic MRgSBRT, achieving robust motion control with minimal baseline shift corrections. The low frequency of corrective interventions and well-characterized motion profiles support margin-reduced, motion- adaptive delivery. These findings establish the feasibility and safety of CMM-integrated MRgSBRT for pancreatic cancer. Keywords: MRgSBRT, Pancreatic Cancer, CMM Digital Poster 4631 Mediastinal lymph node target doses in DIBH lung cancer radiotherapy reconstructed from the motion monitored during treatment Sara Shahzadeh 1 , Lone Hoffmann 2 , Mai Lykkegaard Ehmsen 1 , Marianne Knap 2 , Esben Worm 2 , Ditte Sloth Møller 2 , Per Rugaard Poulsen 1,2 1 Danish Centre for Particle Therapy, Aarhus University Hospital, Aarhus, Denmark. 2 Department of Oncology, Aarhus University Hospital, Aarhus, Denmark Purpose/Objective: Purpose: Deep-inspiration breath-hold (DIBH) during lung cancer radiotherapy may reduce lung and heart irradiation. However, geometrical position errors and intra-fraction motion can lead to doses that deviate from the planned dose distribution. In this study, we monitored internal target motion during DIBH lung cancer treatments and investigated its impact on the mediastinal lymph node (LN) target dose. Material/Methods: Methods: Fifteen locally advanced lung cancer patients treated with 5-8 field intensity modulated radiotherapy (IMRT) for primary tumors and involved LNs were included. Treatment plans were created to achieve CTV V95% ≥ 99%, D98% ≥ 95%, and Dmax < 107% of the prescribed dose. Each patient had 1-3 fiducial markers implanted in the LNs. Daily DIBH cone-beam CT (CBCT) was acquired for patient setup with soft tissue matching on the primary tumor. Each field was delivered during DIBH with a maximum duration of 20 seconds, guided by an external gating block with 2mm gating window. Fluoroscopic kV
Made with FlippingBook - Share PDF online