S2171
Physics - Inter-fraction motion management and daily adaptive radiotherapy
ESTRO 2026
Life, Amsterdam, Netherlands. 3 department of radiation oncology, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, Netherlands Purpose/Objective: Online adaptive radiotherapy (oART) enables daily plan adaptation to account for anatomical changes, improving target coverage and sparing organs-of- interest (OOI).1 Clinical oART workflows rely on manual review and correction of automatically generated delineations, prolonging treatment and limiting routine use.2 While fully automated oART workflow could improve efficiency, its dosimetric validity remains uncertain. This study reports dosimetric outcomes of a clinical oART workflow in esophageal cancer patients, compares them with a fully automated workflow, and assesses patient- reported experience of oART. Material/Methods: In a prospective single-center study, 10 patients with esophageal cancer received daily oART (41.4Gy/23 fractions). Each fraction, a Cone-Beam CT was acquired on Ethos 2.0 with Hypersight (Varian Medical Systems, Palo Alto, CA) and used to automatically propagate planning-CT delineations to the daily anatomy. These delineations were reviewed and, if necessary, manually corrected by a radiation oncologist or trained RTT. To assess a fully automated workflow, plans based on the uncorrected delineations were used, with dosimetric parameters calculated on corrected delineations. Fraction- equivalent (FE) internal clinical target volume (ICTV) and OOI doses of corrected and automated workflows were compared with a Wilcoxon signed-rank test; p<0.05, and against ICTV coverage goals (D98%>95%, V95%>98%). Patient-reported experience was scored on a 5-point Likert scale (1, strongly disagree; 5, strongly agree) on perceived treatment duration and willingness to receive oART again. Results: All 230 clinically delivered oART plans achieved excellent target coverage (median ICTVFE- D98% 1.79Gy, ICTVFE-V95% 99.99%), with median lung doses of Dmean=5.5Gy, VFE20Gy=3.4%, VFE10Gy=16.7%, and median heart Dmean=9.9Gy and VFE30Gy=6.2%. Automatically propagated delineations were adapted in 117/230 fractions. Of these, 74 fractions had both corrected and uncorrected delineations with corresponding plans available, enabling comparison of automated and clinical workflows. Results were largely comparable, showing similar mean target coverage (ICTVFE-D98%=1.78Gy, ICTVFE-V95%=99.98%, p<0.01), though four fractions fell below clinical thresholds (Figure 1). A significant decrease in LungsFE-V20Gy was found (3.7% vs 3.24%, p=0.01), but only minimal differences in Lungs DFE- mean (0.24 vs 0.23Gy, p=0.03). Notably, the automated
Figure 1 Signal intensity change vs control point for spinal cord and PTV70 structures projected onto the EPID images for one patient for original CT scan (ORG) and rescan CT (Rescan). The original patient plan is used in both casesThe rescan CT showed an average 2-7% increase in EPID signal intensity compared with the original CT plans for both spinal cord and PTV structures across all beams. Analysis of maximum signal intensity differences across control points and gantry angles revealed distinct correspondence between high-difference regions and gantry angles where the largest patient weight loss occurred along treatment beam’s-eye view.
Figure 2 Comparison of the patient axial slice (left) ORG CT and (right) Rescan CT scans showing the regions of non-uniform weight-loss Conclusion: This study evaluated EPID transit imaging by analysing control point-level signal variations in HN radiotherapy to identify clinically relevant anatomical changes. The control-point resolved data is more sensitive than single image analysis and yields directional weight- change information. Future work will focus on correlating EPID signal differences with corresponding in-patient dosimetric differences to support adaptive decision-making. References: 1. Nuyts, S., et al., Adaptive radiotherapy for head and neck cancer: Pitfalls and possibilities from the radiation oncologist's point of view. Cancer Med, 2024. 13(8): p. e7192. Keywords: Anatomical Changes, Real-time EPID Transit Images Digital Poster Highlight 3886 Online adaptive radiotherapy in esophageal cancer: dosimetric findings, implications of a fully automated workflow, and patient experience Thomas Weststrate 1,2 , Leigh A.P. Bruijs 1,2 , Irma W.E.M. van Dijk 1,2 , Karin N. Goudschaal 1,2 , Zdenko van Kesteren 1,2 , Tezontl S. Rosario 3,2 , Joost J.C. Verhoeff 3,2 , Elisabeth D. Geijsen 1,2 , Jorrit Visser 1,2 , Arjan Bel 1,2 , Peter S.N. van Rossum 3,2 1 department of radiation oncology, Amsterdam UMC, University of Amsterdam, Amsterdam, Netherlands. 2 -, Cancer Center Amsterdam, Treatment and Quality of
Made with FlippingBook - Share PDF online