Dosimetric feasibility of MRI-guided proton therapy

Abstract

The goal of radiotherapy is to deliver a high conformal radiation dose to a target, while sparing healthy surrounding tissue. Proton therapy, where protons are used to deliver the dose, promises higher dose conformality in comparison with photon-based radiotherapy. This is due to the existence of the Bragg Peak, a point where a large amount of the dose is delivered and after which the protons stop in the body. This leads to a sharp dose fall-off, potentially sparing tissue beyond the target. Because of these properties, up-to-date knowledge of the location of the target and the surrounding anatomy is crucial. Similar to the development in photon therapy, image-guided radiotherapy (IGRT) is emerging as well in proton therapy and in the recent years, an increasing number of studies have been performed on the development of image-guided proton therapy (IGPT). Various imaging modalities are available to deliver the necessary imaging in IGPT. One of those modalities is Magnetic Resonance Imaging (MRI). MRI has several benefits, such as the fact that no ionizing radiation is used and the superior soft-tissue contrast, which can lead to better tissue classification. Therefore an ideal solution for IGPT would be a hybrid MRI-proton therapy system, similar to the already existing MR-linac in photon therapy. As a step towards the development of such a system, the dosimetric feasibility of proton therapy inside the strong magnetic fields of an MRI is addressed in this thesis. For this purpose, Monte Carlo (MC) simulations are used. First, an MC model of the MD Anderson Cancer Center Proton Therapy Center clinical scanning proton beam is created using the TOol for PArticle Simulations (TOPAS), an MC toolkit based on Geant4 and specifically tailored for medical particle simulations. This beam model is then used for the simulation of a quality assurance phantom measurement and is the basis for the further simulation studies in this thesis. To account for inter-fraction and intrafraction motion during treatment, an adaptive planning workflow is presented. As in proton therapy anatomical changes can introduce profound dose changes, adaptive planning could significantly improve proton dose delivery. It is shown that for IMPT, the deterioration in target coverage is mostly restored with the adaptive plan. Next, a study on the dosimetric feasibility of Intensity Modulated Proton Therapy (IMPT) in a transverse magnetic field of 1.5T is presented. It Is shown that the impact of the magnetic field is small and, when taken into account into the planning, the resulting dose distributions are equivalent for 0T and 1.5T, concluding that IMPT in a 1.5T transverse magnetic field is dosimetrically feasible. Finally, the implementation of proton transport inside a magnetic field in the commercial treatment planning system RayStation is validated. This validation paves the way for broad clinical planning studies, which is necessary to build the clinical rationale for the development of MRI-guided proton therapy. In conclusion, this thesis lays the foundation for future dosimetric and clinical studies for the further development and, finally, the implementation of MRI-guided proton therapy.