Double scattering and pencil beam scanning Monte Carlo workflows for proton therapy retrospective studies on radiation-induced toxicities.

PURPOSE: Monte Carlo (MC) simulations can be used to accurately simulate dose and linear energy transfers (LET) distributions, thereby allowing for the calculation of the relative biological effectiveness (RBE) of protons. We present hereby the validation and implementation of a workflow for the Monte Carlo modelling of the double scattered and pencil beam scanning proton beamlines at our institution. METHODS: The TOPAS/Geant4 MC model of the clinical nozzle has been comprehensively validated against measurements. The validation also included a comparison between simulated clinical treatment plans for four representative patients and the clinical treatment planning system (TPS). Moreover, an in-house tool implemented in Python was tested to assess the variable RBE-weighted dose in proton plans, which was illustrated for a patient case with a developing radiation-induced toxicity. RESULTS: The simulated range and modulation width closely matches the measurements. Gamma-indexes (3%/3mm 3D), which compare the TPS and MC computations, showed a passing rate superior to 98%. The calculated RBE-weighted dose presented a slight increase at the necrosis location, within the PTV margins. This indicates the need for reporting on the physical and biological effects of irradiation in high dose regions, especially at the healthy tissues and increased LET distributions location. CONCLUSION: The results demonstrate that the Monte Carlo method can be used to independently validate a TPS calculation, and to estimate LET distributions. The features of the in-house tool can be used to correlate LET and RBE-weighted dose distributions with the incidence of radiation-induced toxicities following proton therapy treatments.

Influence of beam incidence and irradiation parameters on stray neutron doses to healthy organs of pediatric patients treated for an intracranial tumor with passive scattering proton therapy.

PURPOSE: In scattering proton therapy, the beam incidence, i.e. the patient's orientation with respect to the beam axis, can significantly influence stray neutron doses although it is almost not documented in the literature. METHODS: MCNPX calculations were carried out to estimate stray neutron doses to 25 healthy organs of a 10-year-old female phantom treated for an intracranial tumor. Two beam incidences were considered in this article, namely a superior (SUP) field and a right lateral (RLAT) field. For both fields, a parametric study was performed varying proton beam energy, modulation width, collimator aperture and thickness, compensator thickness and air gap size. RESULTS: Using a standard beam line configuration for a craniopharyngioma treatment, neutron absorbed doses per therapeutic dose of 63µGyGy(-1) and 149µGyGy(-1) were found at the heart for the SUP and the RLAT fields, respectively. This dose discrepancy was explained by the different patient's orientations leading to changes in the distance between organs and the final collimator where external neutrons are mainly produced. Moreover, investigations on neutron spectral fluence at the heart showed that the number of neutrons was 2.5times higher for the RLAT field compared against the SUP field. Finally, the influence of some irradiation parameters on neutron doses was found to be different according to the beam incidence. CONCLUSION: Beam incidence was thus found to induce large variations in stray neutron doses, proving that this parameter could be optimized to enhance the radiation protection of the patient.

Configuration and validation of an analytical model predicting secondary neutron radiation in proton therapy using Monte Carlo simulations and experimental measurements.

PURPOSE: This study focuses on the configuration and validation of an analytical model predicting leakage neutron doses in proton therapy. METHODS: Using Monte Carlo (MC) calculations, a facility-specific analytical model was built to reproduce out-of-field neutron doses while separately accounting for the contribution of intra-nuclear cascade, evaporation, epithermal and thermal neutrons. This model was first trained to reproduce in-water neutron absorbed doses and in-air neutron ambient dose equivalents, H*(10), calculated using MCNPX. Its capacity in predicting out-of-field doses at any position not involved in the training phase was also checked. The model was next expanded to enable a full 3D mapping of H*(10) inside the treatment room, tested in a clinically relevant configuration and finally consolidated with experimental measurements. RESULTS: Following the literature approach, the work first proved that it is possible to build a facility-specific analytical model that efficiently reproduces in-water neutron doses and in-air H*(10) values with a maximum difference less than 25%. In addition, the analytical model succeeded in predicting out-of-field neutron doses in the lateral and vertical direction. Testing the analytical model in clinical configurations proved the need to separate the contribution of internal and external neutrons. The impact of modulation width on stray neutrons was found to be easily adjustable while beam collimation remains a challenging issue. Finally, the model performance agreed with experimental measurements with satisfactory results considering measurement and simulation uncertainties. CONCLUSION: Analytical models represent a promising solution that substitutes for time-consuming MC calculations when assessing doses to healthy organs.