Loading the tumor with P-31, Cu-63 and Y-89 provides an in vivo prompt gamma-based range verification for therapeutic protons

Introduction: The main rationale for using protons in cancer treatment is based on the highly conformal dose distribution and normal tissue spearing compared to conventional radiotherapy. The main limit of proton therapy is the particle range uncertainty due to patient setup, dose calculation and imaging. To account for this, a safety margin is added to the tumor to ensure the prescribed dose to the target. Reducing range uncertainties would result in the reduction of irradiation volume and would allow full exploitation of the proton therapy benefits. In this work, we presented a feasibility study for a strategy to achieve in vivo proton range verification based on prompt gammas (PG). This approach relies on the detection of signature prompt gammas, generated by the interaction of primary protons with a non-radioactive element, that is selectively loaded into a tumor with a drug carrier. The number of characteristic gammas is directly related to the proton range, and its measurement provides an estimate of the position at which the primary beam stops with respect to the tumor location. Method: We identified the criteria for selecting potential candidate materials and combined them with TALYS predictions to make the selection. We carried out an experimental campaign to characterize the PG spectra generated by the chosen materials when irradiated with therapeutic protons and compared them with TOPAS Monte Carlo toolkit predictions. Results: We identified 31-Phosphorous, 63-Copper and 89-Yttrium as potential candidates for this application based on TALYS calculations. The experimental data confirmed that all candidates emit signature prompt gammas different from water (here used as a proxy for normal tissue), and that the gamma yield is directly proportional to the element concentration in the solution. Four specific gamma lines were detected for both P-31 (1.14, 1.26, 1.78, and 2.23 MeV) and Cu-63 (0.96, 1.17, 1.24, 1.326 MeV), while only one for Y-89 (1.06 MeV). The simulations indicate that the count of characteristic gammas is directly proportional to the proton range, reaching in some cases a saturation value around the tumor's far edge. The results also indicate that to achieve a range accuracy below the current value of 2-3 mm, the uncertainty on the prompt gammas count has to be below 5% for 31-Phosphorous and 63-Copper, or 10% for 89-Yttrium. Discussion: We demonstrated that loading the tumor with a label element prior to proton treatment generates signature gammas that can be used to verify the beam range in vivo, reaching a potential range accuracy below the current limitations. This approach can be either used stand-alone or combined with other existing methodologies to further improve range resolution.