Author: David Ayala Alvarez, Facundo Ballester, Luc Beaulieu, Francisco Berumen-Murillo, Jean-Simon Cote, Ernesto Mainegra-Hing, Iymad Mansour, Gaël Ndoutoume-Paquet, Jan P. Seuntjens, Rowan M. Thomson, Christian Valdes, Javier Vijande, Peter G. Watson 👨🔬
Affiliation: Département de physique, de génie physique et d'optique, Université Laval, Princess Margaret Cancer Centre & University of Toronto, McGill University, Department of Physics and Medical Physics Unit, McGill University, IFIC-UV, University of Valencia, National Research Council Canada, Carleton University, 5Service de Physique Médicale et de Radioprotection, Centre Intégré de Cancérologie, CHU de Québec- Université Laval et Centre de recherche du CHU de Québec, Nuclear Medicine Department, Hospital Regional de Antofagasta, Université Laval, Carleton Laboratory for Radiotherapy Physics, Department of Physics, Carleton University 🌍
Purpose: Low-energy X-ray beams used in electronic brachytherapy (eBT) pose unique dosimetric challenges due to high depth-dose gradients, material-dependent detector responses, and the absence of standardized dose-to-water references. This study evaluates the dosimetric impact of different physics approaches in three advanced Monte Carlo (MC) systems for eBT, focusing on discrepancies in bremsstrahlung interactions that influence dosimetric consistency.
Methods:
MC simulations of the Axxent S700, Esteya, and INTRABEAM eBT systems were conducted using EGSnrc (egs_brachy/egs_kerma), TOPAS, and PENELOPE-2018 (PEN18). Simplified X-ray tube models preserved key characteristics such as target mode (reflection or transmission), material, thickness, and applicator geometry. Depth doses and fluence were compared across systems for transmission and reflection target modes. Additional simulations evaluated the predictive capability of simplified models for detailed applicator geometries. All MC systems use Seltzer and Berger's scaled function for bremsstrahlung but vary in shape function approximations. KQP (relativistic partial-wave approach) is considered the theoretical benchmark, being used by PEN18, 2BN (Born approximations without screening correction) and 2BS (with screening correction, extreme-relativistic and small angles approach) by TOPAS, and KM (2BS corrected by 2BN's angular term) by EGSnrc.
Results: Significant differences were observed in transmission target mode. Bremsstrahlung fluence spectra varied by 15% between PEN18, EGSnrc, and TOPASliv, with PEN18 consistently lower. PEN18 and EGSnrc agreed within 3% for characteristic spectra, while PEN18’s lines exceeded TOPASliv’s by 40%, causing depth dose discrepancies of ~9%. With TOPASpen, compensating spectral differences yielded depth dose deviations within 1%. In reflection mode, PEN18 and EGSnrc aligned within 4%, but TOPASpen showed 12% lower bremsstrahlung and 6% higher characteristic lines, reducing dose differences to ~3%.
Conclusion: Significant discrepancies exist among MC systems in bremsstrahlung simulation for eBT, driven by differences in angular distribution and atomic relaxation modeling. Further theoretical and experimental studies are essential to understand these impacts on clinical dosimetry.