Author: Caroline Esposito, Keith T Griffin, Jae Won Jung, Choonik Lee, Choonsik Lee, Matthew Mille, Harald Paganetti, Sergio Morato Rafet, Jan PO Schuemann, Jungwook Shin, Torunn I Yock 👨🔬
Affiliation: East Carolina University, University of Michigan, Massachusetts General Hospital, National Cancer Institute, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Massachusetts General Hospital and Harvard Medical School 🌍
Purpose: The National Cancer Institute’s Pediatric Proton and Photon Therapy Comparison Cohort aims to collect and analyze data from cancer centers across the United States and Canada to quantify differences in the risk of developing second malignancies among patients treated with proton versus photon radiotherapy. In this cohort, Monte Carlo (MC) simulations will be used to retroactively calculate dose estimates for passive scattering proton therapy patients. However, such high-energy simulations rely on physics models to estimate interaction probabilities, acting as a source of dose uncertainty. The purpose of this study is to investigate the impact of physics model selection on normal tissue dose reconstruction.
Methods: Data from thirty patients undergoing passive scattering proton radiotherapy were incorporated into an established TOPAS MC workflow to individually reconstruct patient-specific tissue dose estimates. These simulations were performed using extended whole-body patient models and both manual and automated tissue segmentation. Dose estimates were calculated using three distinct physics models to manage particle interactions within TOPAS: Binary Cascade, Bertini, and INCL++. Data on particle yield, energy, and angular distributions were collected at the point of treatment nozzle exit.
Results: Differences in organ dose estimates across various patient treatment types will be presented with supportive evidence captured on the quality of the radiation field exiting the nozzle. For example, the neutron dose estimates were more conservatively calculated by the Bertini Model, whereas differences between the Binary and Intranuclear cascade models exhibited a dependency on the distance from the target volume.
Conclusion: Dose characterization of the current work facilitates a more informed selection of the appropriate physics model for normal tissue dose reconstruction. Findings on the impact of physics model selection will be integrated into the uncertainty analyses applicable to all passive scattering patient dose estimates within the cohort.