Purpose: To benchmark three techniques for Spatial Fractionated Radiation Therapy (SFRT) and evaluate their dosimetric performance in multiple treatment sites.
Methods: Five patients (multiple sites) who underwent conventional single-field GridBlock therapy were replanned using GridBlock, MLC-based, and Lattice-based SFRT (15Gy/1fraction, 6 MV). The GridBlock consisted of a hexagonal pattern of divergent holes (hole diameter=1.43cm, center-to-center spacing=2.11cm). The MLC-based grid was generated by converting a static beam to a step and shoot beam delivering grid pattern (square aperture=1cm, open-to-closed ratio=0.375/0.625). The Lattice-based grid was generated by placing cylinders with 1.5cm diameter and 3cm center-to-center spacing within the GTV to mimic conventional GridBlock therapy. The quality of the plans was evaluated using different dosimetric parameters (Valley-to-Peak Ratio (VPR=D10valley/D90peak), mean GTV dose, equivalent uniform doses (EUD) for both GTV and Normal Tissue, delivery time).
Results: The VPR is the major postulated contributor of the bystander and abscopal effects in SFRT. The lattice technique achieved a better mean VPR (54.03%±15%) when compared to those of the Grid-Block (19.09%±4%) and MLC (19.40%±16%) techniques. For dose coverage of the GTV, the lattice technique also achieved a better dose coverage (D=9Gy±3Gy) compared to Grid-Block (D=5.8Gy±1Gy) and MLC (D=4Gy±1Gy). The EUD predicts the effectiveness of tumor control and normal tissue toxicity and the lattice technique predicted better tumor control and normal tissue toxicity (mean EUD_tumor =7.97Gy, mean EUD_normal=5.87Gy) than the GridBlock (mean EUD_tumor=1.08Gy, mean EUD_normal=6.91Gy) and MLC (mean EUD_tumor=1.67Gy, mean EUD_normal=5.45Gy). The beam-delivery time was clinical acceptable for all the techniques (mean T_lattice=8.1min, mean T_GridBlock=2.83min, mean T_MLC=7.3min).
Conclusion: When available, the lattice technique should be preferred above the GridBlock and MLC techniques because it produces better quality plans. It achieves better VPRs and coverage of the GTV. It gives better predictions of the tumor control and normal tissue toxicity, and its delivery times are clinical acceptable.