S1860
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
toxicity occurring within 6 months after SBRT. This analysis reports the impact of a protocol amendment in connective tissue delineation on target coverage. Material/Methods: Eleven consecutive patients (41–93 years) from two centers were included. All treated on a 0.35 Tesla MR- linac (MRidian, Viewray) following the STAR-LUNG protocol. One patient was excluded from dosimetric comparison due to absence of a connective tissue structure, leaving 10 for analysis. The prescribed dose was 56 Gy in eight fractions to the PTV (GTV + 5 mm), aiming for V95% > 95%. An inhomogeneous distribution allowed mean GTV doses up to 85 Gy while limiting the maximum to 110 Gy. Organ-at-risk (OAR) constraints were prioritized over target coverage.The protocol includes delineation of the aorta, cutis, main and lobar bronchi, esophagus, heart, trachea, lungs, spinal cord, thoracic wall, and GTV. In the original approach (+Conn.Tissue), connective tissue was defined as all mediastinal tissue not defined as OAR or GTV. Following a protocol amendment, this approach (-Conn.Tissue) was revised to exclude a 3 mm rim of connective tissue around the GTV, as the D0.3 cc < 45 Gy constraint in this region was dose-limiting without anticipated clinical justification.To assess the amendment’s impact, two plans per patient (baseline and fraction 1) were re- optimized using the two approaches, yielding 40 plans. Optimization aimed to maximize PTV coverage and GTV mean dose while adhering to OAR constraints. Differences in target coverage (PTV V95%, GTV mean dose) between the two approaches were evaluated using a right-tailed Wilcoxon signed-rank test (testing for higher values using -Conn.Tissue). Results: With the -Conn.Tissue approach, PTV coverage increased in all cases, with a significant median V95% difference of +6.7% (p < 0.001, range 0.2–19.2%), c.f. Figure 1. The mean GTV dose increased significantly by a median of +1.5 Gy (p < 0.001, range –0.4 to 18.2 Gy). In 16 of 20 plans, the mean GTV dose increased, while in the remaining four plans dose escalation was limited by cutis constraint (1 plan) or an already high GTV mean dose (>84 Gy, 3 plans).
Conclusion: This study integrates probabilistic modelling of target and contour propagation uncertainties with robust optimization for setup errors. Using voxel probability weighting, scenario selection, and statistical evaluation, it enables uncertainty-aware planning, with future work extending to other tumours and clinical
integration. References:
1) Gordon JJ et al. Med Phys 37, 550–63 (2010)2) Buti G et al. Phys Med Biol 66 (2021) 3) Unkelbach J, Paganetti H. Semin Radiat Oncol 28, 88–96 (2018)4) Wuyckens S et al. (2023)5) Sterpin E et al. Radiother Oncol 197 (2024) 6) Sterpin E et al. Phys Med Biol 66 (2021)7) Rivetti L et al. Phys Med Biol 69 (2024) Keywords: Infiltration-mapping, Robustness, Optimisation.
Digital Poster 1878
Impact of connective tissue delineation on target coverage in daily adaptive MR-guided SBRT for ultra-central lung tumors (NCT05354596) Susanne N Bekke 1 , Laura A Rechner 1 , Susan BN Biancardo 1 , Cécile Peucelle 2 , Kristian Boye 2 , Anders S Bertelsen 3 , Lone Hoffmann 4,5 , Tine Schytte 3,6 , Filippa Sundbye 1 , Gitte Persson 1 , Mette Pøhl 2 1 Department of Oncology, Copenhagen University Hospital – Herlev and Gentofte, Herlev, Denmark. 2 Department of Oncology, Copenhagen University Hospital - Rigshospitalet, Copenhagen, Denmark. 3 Department of Oncology, Odense University Hospital, Odense, Denmark. 4 Department of Oncology, Aarhus University Hospital, Aarhus, Denmark. 5 Department of Clinical Medicine, Aarhus University, Aarhus, Denmark. 6 Institute of Clinical Research, University of Southern Denmark, Odense, Denmark Purpose/Objective: STAR-LUNG (NCT05354596) is an ongoing prospective multicenter phase II trial evaluating the safety and efficacy of MR-guided adaptive stereotactic radiotherapy (SBRT) for ultra-central lung tumors. The primary endpoint is grade ≥ 4 treatment-related
Made with FlippingBook - Share PDF online