Development of a Clinical Prototype for Range Verification in Proton Therapy based on Prompt Gamma-Ray Spectroscopy

Abstract

Range uncertainties in proton therapy pose a limitation on its benefits for cancer treatment. Robust planning and conservative safety margins of up to 1 cm are necessary for guaranteeing full tumor coverage, at the price of a larger dose to normal tissue. At the Massachusetts General Hospital, a full-scale prototype system for in vivo proton range verification is under development, based on spectroscopy of prompt gamma-rays emitted from proton-nuclear reactions with tissue. The aim is to verify the proton range during patient treatments with 1 mm precision. The system consists of eight LaBr3 detectors and a tungsten collimator that are mounted on a rotating frame. The electronics and data processing algorithms are designed to cope with the highly variable count rates that occur during pencil-beam scanning. The prototype was tested by irradiating a water phantom with a clinical dose of 0.9 Gy, a beam current of 2 nA and a field size of 10 cm x 10 cm. Energy- and time-resolved gamma-ray spectra were acquired during the irradiation and further analyzed to subtract the neutron background and to determine the prompt gamma-ray line magnitudes. A GPU-accelerated Monte Carlo model of the gamma-ray emissions was developed, which relies on measured nuclear reaction cross sections and the treatment plan CT. We reconstruct both the absolute proton range and the elemental concentrations of the irradiated tissue by minimizing the deviation between the measurement and the parameterized model. Range shifters and changes in the elemental composition were introduced in different parts of the water phantom to verify the precision and robustness of the range verification method. A statistical precision of 1.1 mm at 95% confidence level and a mean systematic error of 0.5 mm were obtained, when merging protons delivered to the two distal pencil-beam layers within a 10 mm radius. In an upcoming clinical study, the prototype will be tested during the treatment of brain cancer patients.