Bridging the gaps in Particle Therapy using Monte Carlo Codes

Charged particle therapy is well known for precise dose delivery to the tumour, while simultaneously increasing sparing of surrounding healthy tissues, compared to high energy photon beams. As a constantly evolving technology, new research challenges must be addressed in order to further optimize treatment outcomes. These challenges include improvement of dose calculation accuracy for treatment planning, development of independent quality assurance methods, as well as evaluation of the biological effectiveness of a treatment. A significant part of such research can be carried out using particle interaction and transport simulations. Monte Carlo (MC) codes, which are considered as the ’gold standard’ in radiation therapy, can help bridge research gaps. As such an improvement, the development and application of MC codes for charged particle therapy are presented in this thesis. Specifically, this thesis is focused on the development and application of a Patient Specific Quality Assurance (PSQA) tool for scanned particle beam delivery, as well as benchmarking and application of a new radiobiological model for proton and ion beam therapy - BIANCA. The general purpose MC code FLUKA was selected to pursue this research. FLUKA profits from its recent developments in charged hadron interactions an transport models, at the therapeutic energy range. During the course of this work, the code was upgraded and several new features were subsequently implemented. As a result, the code was able to simulate a full set of treatment beamlets for an active beam scanning technique for proton and ion beam therapies. The performed study was finalized with development of the FLUKA Particle Therapy Tool (FPTT), which enables simulation of realistic clinical treatment planning scenarios using a voxelized patient geometry. A next step for PSQA development involved an application and a validation of the FPTT, i.e., benchmarking studies were performed using data sets from two particle therapy facilities - Trento Proton Facility and CNAO. The beam parameters were modelled in FPTT and compared with commissioning data of the respective clinical Treatment Planning Systems (TPS). Further evaluation included simulations of four clinical treatment planning scenarios in FPTT: two proton chordoma cases (one from each facility), where small differences between FPTT calculations of dose deposition and TPS calculations were noticed; one proton head-and-neck case (using a range shifter), where significant differences between TPS and MC calculations were highlighted and discussed; and one chordoma carbon ion case, where biological dose calculations were performed using the clinical radiobiological model - LEM I. In addition, a research application related to the dose-averaged Linear Energy Transfer (LET) was presented on a carbon ion case. Overall, the FPTT was proved to be reliable and well-integrated with the FLUKA MC code. It supported an import of particle therapy treatment planning data, adjustment to the user’s requirements, and translation into FLUKA input file, without superior experience in the MC simulations. Finally, it proved its usability for a complex treatment plan scenarios in a PSQA simulations and research applications. The second part of this thesis focuses on the radiobiological effectiveness of proton and ion beam therapies, presenting an upgrade and a benchmarking of a research biophysical model, BIANCA. An accurate description of underlying phenomena is crucial for evaluation of tumour control probability and normal tissue complication. First, the BIANCA model was expanded to new ion species, and various cells of interests. This was followed by an integration with the FLUKA MC code and the FPTT. Subsequently, BIANCA cell survival predictions were benchmarked against experimental data of Chinese Hamster Ovary Cells (CHO) irradiated by two opposed fields of proton or carbon ions. The results showed that the prediction of BIANCA and LEM I models were comparable, and the BIANCA model was in good agreement with the experimental data. The second benchmarking study compared the outcome of BIANCA modelling with cell survival of the rat spinal cord, following proton or carbon ion irradiation. In this case, the BIANCA model was again in a very good agreement with the data, well reproducing the trend of the RBE-dose and the RBE-LET dependence. This work provided a baseline for the first application of the BIANCA model to carbon-ion patient cases. Three cases from CNAO facility were chosen: a chordoma, a head-and-neck, and a prostate case. BIANCA Relative Biological Effectiveness (RBE) predictions were calculated, based on chordoma cell survival (RBEsurv), or on dicentric aberrations in peripheral blood lymphocytes (RBEab), which are indicators of late normal tissue damage. Simulation outcomes were compared with those provided by LEM I. Overall, results suggested that, if RBEsurv is used to evaluate the beam effectiveness at killing tumour cells, and RBEab is used to estimate (late) normal tissue damage, BIANCA provides lower RBE-weighted doses (with respect to LEM I) in the tumour and in the entrance channel; whereas in the Organs at Risk, BIANCA and LEM I provide very similar values. The presented work, and previous benchmarking studies, are encouraging, suggesting that BIANCA can be applied for radiobiological optimization of carbon ion treatment planning, using RBE database based on cell-specific tumour cell survival and RBE database based on normal tissue aberrations. In summary, this thesis resulted in implementation of a MC-based tool for particle therapy, that can be used for clinical and research applications, supporting dose distribution calculation for PSQA purposes, and evaluation of the biological-weighted dose for ion beam therapy. In addition, the final chapter showed that MC codes can support the development of prototype accelerators, and its commissioning in a commercial TPS. As such, the research presented in this thesis bridges certain gaps in particle therapy by providing solutions using general-purpose MC code.