Dottorato in FISICA DEGLI ACCELERATORI

Titolo della tesi: Development of a Treatment Control System for IOeRT FLASH beam

Four decades ago, Intra-Operative electron Radiation Therapy (IOeRT) was developed to improve precision in local cancer treatment by combining real-time surgical exploration and resection with electron irradiation at 4-12 MeV. A clear delimitation of the volume to treat, the possibility of moving temporarily normal tissues to spare them from radiation, the reduced interaction with skin or other organs, thanks to temporary beam modifiers, and a greater antitumoral efficacy make IOeRT a very effective technique. Despite that, planning technology has not evolved since its conception, being outdated in comparison to the current state of the art in other radiotherapy techniques and therefore slowing down the adoption of IOeRT in clinical practice. As of today, in fact, a dedicated Treatment Planning System (TPS) for IOeRT treatments does not exist, meanly due to the very limited amount of time available during the surgery to obtain both a new imaging of the patient, needed to account for the morphological changes brought by the surgery, and perform dose optimization through the TPS computation. Therefore, the required dose delivery planning is actually done using non-accurate models that rely on the knowledge and predictions of the surgeon and the radiotherapist that visually identify the target to be treated and decide the beam energy and applicator dimension and position. Today, the availability of accurate dose calculation algorithms for clinical use, such as real-time Monte Carlo (MC) simulation, the possibility to use in-room imaging, and the prospected coming of radically new irradiation scheme, e.g. the so-called FLASH effect, pushed the scientific community to explore possible new simulation and treatment planning system tools for IOeRT to improve the technique and expand its field of applications. The goal of my Ph.D. research activity was to address the technology gap between the IOeRT and other radiotherapy techniques, by developing the first-ever complete TPS for IOeRT treatments in collaboration with the S.I.T. Sordina IORT Technologies S.p.A. (Aprilia, Italy) company. In particular, while the S.I.T. has solved the problem of providing an online intra-operative image through the ultrasound imaging system, my contribution was mainly focused on the development of treatment simulation and optimization tools. To this aim, an accurate and fast dose engine was needed. The Applied Radiation Physics Group (ARPG) at “La Sapienza" University of Rome, in which I work, has developed the FRED (Fast paRticle thErapy Dose evaluator) dose engine based on MC calculation. FRED has been developed to allow a fast optimization of TPS in Particle Therapy while keeping the dose release accuracy typical of an MC tool. For the exceptional speed of the protons and recently also of the carbon ions tracking algorithms implemented in FRED and for the excellent results achieved in terms of dose accuracy, we decided to develop the electromagnetic FRED model to extend the use of this MC-on-GPU-based software to other radiotherapy techniques where the time factor is crucial, i.e. the IOeRT. This thesis work includes both the complete modeling of the interaction between the electromagnetic beam and patient, balancing the speed of calculation with the accuracy of implemented physical models, and its first clinical application as a treatment planning and optimization tool for IOeRT. The obtained FRED performance allows for computing electron and photons dose distributions in few minutes with an accuracy of ~1%, tested against state-of-art full MCs tools, such as FLUKA and GEANT4, and experimental data, in the energy range from 1 to 200 MeV (clinical application from IOeRT to the future Very High Energy Electron therapy). The developed IOeRT treatment planning and optimization tool, based on the FRED MC and the S.I.T. intra-operative imaging system, was tested in the context of breast cancer treatment obtaining encouraging results, confirming the FRED potential as a fast and accurate planning tool.