Radiotherapy Beamline Design for Laser-driven Proton Beams

Motivation: Radiotherapy is an important modality in cancer treatment commonly using photon beams from compact electron linear accelerators. However, due to the inverse depth dose profile (Bragg peak) with maximum dose deposition at the end of their path, proton beams allow a dose escalation within the target volume and reduction in surrounding normal tissue. Up to 20% of all radiotherapy patients could benefit from proton therapy (PT). Conventional accelerators are utilized to obtain proton beams with therapeutic energies of 70 – 250 MeV. These beams are then transported to the patient via magnetic transferlines and a rotatable beamline, called gantry, which are large and bulky. PT requires huge capex, limiting it to only a few big centres worldwide treating much less than 1% of radiotherapy patients. The new particle acceleration by ultra-intense laser pulses occurs on micrometer scales, potentially enabling more compact PT facilities and increasing their widespread. These laser-accelerated proton (LAP) bunches have been observed recently with energies of up to 90 MeV and scaling models predict LAP with therapeutic energies with the next generation petawatt laser systems. Challenges: Intense pulses with maximum 10 Hz repetition rate, broad energy spectrum, large divergence and short duration characterize LAP beams. In contrast, conventional accelerators generate mono-energetic, narrow, quasi-continuous beams. A new multifunctional gantry is needed for LAP beams with a capture and collimation system to control initial divergence, an energy selection system (ESS) to filter variable energy widths and a large acceptance beam shaping and scanning system. An advanced magnetic technology is also required for a compact and light gantry design. Furthermore, new dose deposition models and treatment planning systems (TPS) are needed for high quality, efficient dose delivery. Materials and Methods: In conventional dose modelling, mono-energetic beams with decreasing energies are superimposed to deliver uniform spread-out Bragg peak (SOBP). The low repetition rate of LAP pulses puts a critical constraint on treatment time and it is highly inefficient to utilize conventional dose models. It is imperative to utilize unique LAP beam properties to reduce total treatment times. A new 1D Broad Energy Assorted depth dose Deposition (BEAD) model was developed. It could deliver similar SOBP by superimposing several LAP pulses with variable broad energy widths. The BEAD model sets the primary criteria for the gantry, i.e. to filter and transport pulses with up to 20 times larger energy widths than conventional beams for efficient dose delivery. Air-core pulsed magnets can reach up to 6 times higher peak magnetic fields than conventional iron-core magnets and the pulsed nature of laser-driven sources allowed their use to reduce the size and weight of the gantry. An isocentric gantry was designed with integrated laser-target assembly, beam capture and collimation, variable ESS and large acceptance achromatic beam transport. An advanced clinical gantry was designed later with a novel active beam shaping and scanning system, called ELPIS. The filtered beam outputs via the advanced gantry simulations were implemented in an advanced 3D TPS, called LAPCERR. A LAP beam gantry and TPS were brought together for the first time, and clinical feasibility was studied for the advanced gantry via tumour conformal dose calculations on real patient data. Furthermore, for realization of pulsed gantry systems, a first pulsed beamline section consisting of prototypes of a capturing solenoid and a sector magnet was designed and tested at tandem accelerator with 10MeV pulsed proton beams. A first air-core pulsed quadrupole was also designed. Results: An advanced gantry with the new ELPIS system was designed and simulated. Simulated results show that achromatic beams with actively selectable beam sizes in the range of 1 – 20 cm diameter with selectable energy widths ranging from 19 – 3% can be delivered via the advanced gantry. ELPIS can also scan these large beams to a 20 × 10 cm2 irradiation field. This gantry is about 2.5 m in height and about 3.5 m in length, which is about 4 times smaller in volume than the conventional PT gantries. The clinical feasibility study on a head and neck tumour patient shows that these filtered beams can deliver state-of-the-art 3D intensity modulated treatment plans. Experimental characterization of a prototype pulsed beamline section was performed successfully and the synchronization of proton pulse with peak magnetic field in the individual magnets was established. This showed the practical applicability and feasibility of pulsed beamlines. The newly designed pulsed quadrupole with three times higher field gradients than iron-core quadrupoles is already manufactured and will be tested in near future. Conclusion: The main hurdle towards laser-driven PT is a laser accelerator providing beams of therapeutic quality, i.e. energy, intensity, stability, reliability. Nevertheless, the presented advanced clinical gantry design presents a complete beam transport solution for future laser-driven sources and shows the prospect and limitations of a compact laser-driven PT facility. Further development in the LAP-CERR is needed as it has the potential to utilize advanced beam controls from the ELPIS system and optimize doses on the basis of advanced dose schemes, like partial volume irradiation, to bring treatment times further down. To realize the gantry concept, further research, development and testing in higher field and higher (up to 10 Hz) repetition rate pulsed magnets to cater therapeutic proton beams is crucial