ESTRO 2020 Abstract book

S1031 ESTRO 2020

This work has demonstrated that the presented method was applicable for characterisation of the MR linac geometrically performance and stability. A comparison showed that the MR linac is performing better than the conventional linac which likely is due to the ring gantry construction of the MR linac. In addition, the method is very cheap to implement having access to a 3D-printer. PO-1759 A dose validation tool prior to proton and photon treatment in spine SRS G. Liu 1 , A. Qin 1 , J. Zhou 2,3 , D. Yan 1 , L. Zhao 1 , C. Stevens 1 , I. Grills 1 , P. Kabolizadeh 1 , X. Li 1 , X. Ding 2 1 Beaumont Health Proton Therapy Center, Radiation Oncology, Royal Oak, USA ; 2 Beaumont Health Proton Therapy Center, Radiation Oncology, Bloomfield Hills, USA ; 3 Emory University, Radiation Oncology, Atlanta, USA Purpose or Objective In addition to the setup accuracy based on the cone beam computed tomography (CBCT) alignment, a comprehensive dose validation information prior to the treatment delivery could be important and sometimes critical to the hypofractionated regimen such as single fraction spine Stereotactic Radiosurgery (SRS). The study proposed a practical and clinical feasible synthetic CT method based on the CBCT to assess the dose delivery accuracy. Material and Methods Seven patients with spine metastases between T2-L5 who were previously treated using VMAT SRS in our institution were retrospectively evaluated in this study. In addition to the clinical VMAT plans, Intensity Modulated Proton Therapy (IMPT) and a new treatment technique, Spot- scanning Proton Arc therapy (SPArc) plans, were generated. Robust optimization parameter with ±3.5% range uncertainty was used in proton planning optimization. According to RTOG-0631, the radiation dose prescription (RX) to PTV was 1800 cGy in a single fraction. To accurately reconstruct treatment dose for patients on the CBCT, the initial planning CT and the corresponding CBCT was deformable registered with each other. The synthesis CT was generated by warped CT number in initial planning CT to the paired CBCT, which was utilized to recalculate dose to represent the dose distribution during treatment. The algorithm first performs a block-wise non- linear registration to get a robust initial alignment, followed by a dense local-correlation-coefficient (LCC) based deformable registration to get the final Deformable Vector Field (DVF). Then, PTV was mapped to synthesis CT. PTV coverage index V90(percentage of target volume covered by 90% of prescription dose) was used to assess the dose accuracy compared to the initial plan. Results Six patients’ synthetic CTs result showed a good agreement with the initial plan. The V90 of PTV coverage on the treatment day maintained 99±0% (VMAT),98±2%(IMPT) and 98%±1% (SPArc) respectively (Fig. 1). However, patient #5’s geometry was found changed. The reconstructed dose showed significant dose degradation (V90 drops to 90%) in the IMPT plan while the plan quality of VMAT and SPArc remained solid (V90=99% in both VMAT and SPArc plan). The following Water- Equivalent-Thickness (WET) calculation and analysis showed that the proton range changed up to 2cm in the beam directions from IMPT plan (red dot in Fig. 2c). Based on such information, a clinical decision on the plan adaptive could have been made if IMPT were used at the time of treatment.

Conclusion This synthetic CT method could be used as an effective tool to identify the dose deviation based on the CBCT prior to the proton and photon treatment. Such tool is especially important to the proton treatment where the patient’s geometry consistency in the beam path is critical. On the other hand, SPArc could mitigate such dosimetric impact by using more degrees of freedom via arc trajectory. PO-1760 Magnetically focused minibeams for proton therapy G. McAuley 1 , C. Lim 2 , A. Teran 1 , J. Slater 1 , A. WROE 1 1 Loma Linda University, Department of Radiation Medicine, Loma Linda, USA ; 2 Loma Linda University, School of Medicine, Loma Linda, USA Purpose or Objective Proton minibeam radiation therapy (pMBRT) allows normal tissue sparing by utilizing an array of beamlets to deliver a spatially fractionated proximal dose that blends into a homogeneous dose at the target. These beamlets are typically generated via the use of a collimator, however in the present work we investigate the potential to deliver planar proton minibeams using a single quadrupole magnet. Material and Methods Geant4 based Monte Carlo simulations of unmodulated proton beams with a 10,12 and 15 mm initial diameter and 9.8 cm range in water were focused with a single quadrupole magnet of length 8.0 cm and a field gradient of 225 T/m or 250 T/m. The resultant dose distributions of single minibeams were evaluated in a water phantom that was comprised of 0.25 mm x 0.25 mm x 0.25 mm voxels. In addition, composite minibeam dose distributions were created by shifting laterally copies of the single beamlet distribution while modulated beams with a SOBP width of 15 mm were created using appropriate range shifts and weighting. Results The combined dose distribution from three beamlets (both modulated and unmodulated) showed high proximal spatial fractionation. For the various unmodulated beam cases tested, the peak-to-valley dose ranged from as high as 27 to 1.0 over 90% of particle range.