S1746
Physics - Dose prediction/calculation, optimisation and applications for particle therapy planning
ESTRO 2026
19 Particle Therapy Interuniversity Centre Leuven (PARTICLE) Proton Therapy Centre University Hospital Leuven, Department of Radiation Oncology, Leuven, Belgium. 20 Department of Radiation Oncology (Maastro), GROW School for Oncology, Maastricht University Medical Centre, Maastricht, Netherlands. 21 Center for Proton Therapy, Paul Scherrer Institute, Villigen, Switzerland. 22 Danish Center for Particle Therapy, Aarhus University Hospital, Aarhus, Denmark. 23 Medical Physical Unit, National Center for Oncological Hadrontherapy, Pavia, Italy. 24 Department of Medical Physics, Proton Therapy Center Czech, Prague, Czech Republic. 25 Westdeutsches Protonentherapiezentrum Essen, Universitätsklinkum Essen, Essen, Germany. 26 Karolinska University Hospital, Section of Radiotherapy Physics and Engineering, Stockholm, Sweden. 27 Department of Radiation Oncology, University of Washington & Fred Hutch Cancer Center, Seattle, WA, USA Purpose/Objective: A previous study, conducted in 2018-2019 within the European Particle Therapy Network (EPTN), revealed substantial inter-center variations in CT-based stopping-power ratio (SPR) prediction, translating into clinically relevant proton range deviations [1], highlighting the need for standardization. As a consequence, several participating particle therapy (PT) centers updated their clinical SPR prediction procedure and an ESTRO consensus guide for the creation of a Hounsfield look-up table was established [2]. The present work reassesses inter-center variation in SPR prediction following these developments. Material/Methods: Twenty PT centers and three affiliated hospitals participated, including seventeen centers form the first study run. The methodology, including phantom and evaluation procedure, remained unchanged (Fig.1). Each participant scanned the same phantom containing 17 tissue-mimicking surrogates with experimentally determined ground-truth SPR in head and body phantom configurations using their clinical scan protocols. SPR values were derived via each institution’s clinical CT-number-to-SPR conversion. Inter-center SPR variation and deviation from ground- truth were quantified and summarized for lung, soft tissue, and bone tissues. Subsequently, the impact on water-equivalent range prediction was assessed for representative beam paths corresponding to brain, lung and prostate cancer treatments traversing different tissue types.
Results: The 2 σ inter-center variation in SPR relative to water was 3.2%, 0.9% and 4.7% for lung soft-tissue and bone surrogates in the head setup, respectively (Fig.2). The body setup yielded comparable results with 2.9%, 0.9% and 3.7%. This corresponds to 2 σ inter-center variation in range prediction of 1.4%, 1.5% and 1.6% for brain, lung and prostate beam paths, respectively. For brain and prostate cases, this corresponds to a 40% relative reduction in inter-center variation compared to the first run, whereas variation for lung treatments increased slightly by 16%. Centers that participated in both runs, outperformed the overall cohort, yielding absolute range deviations consistently below 2%, with the largest improvement observed for brain cancer (62% reduction in inter-center variation compared to first run).
Made with FlippingBook - Share PDF online