2393 - Dosimetric Verification of a TPS-Guided Novel Arc-TBI Technique

S. Burgess¹, D. Delaney¹, M. A. Stevenson², and S. Peca²; ¹Suffolk University, Boston, MA, ²Beth Israel Deaconess Medical Center, Boston, MA

Purpose/Objective(s): Total Body Irradiation (TBI) for patients with hematological malignancies aims to deliver a prescribed dose ±10% to the whole body. Methods for TBI vary in complexity, but are generally very resource intensive. We propose to develop a robust yet simple TBI method that does not require patient-specific CT planning. The patient lies supine, then prone, on a floor mat in the gantry plane of rotation and the linac delivers jaw-defined arcs. The gantry speed is varied to account for: inverse-square law, changes in scatter with distance, attenuation through spoiler plate, differing patient thickness, lung sparing. This goal requires three objectives. First, to quantify the dose calculation accuracy of the treatment planning system (TPS, Eclipse) at 1-1.5 m from isocenter for different energies and field sizes. Second, to determine the energy and field size combination which gives best agreement with measurements and best dose homogeneity. Third, to build a plan using two arcs (AP and PA) and modulate the weight every 5° to obtain adequate dose coverage. Materials/

Methods: Using an ion chamber in solid water we measured dose at 190 cm SSD and 90 cm laterally under many conditions: jaws 20x40cm², 30x40cm², 40x40cm²; 6 MV, 10 MV; depths 1 cm, 3 cm, 10 cm. Further measurements were performed with a 1.27cm acrylic sheet placed above the phantom (the ‘beam spoiler’ used to provide surface dose). We replicated the setups in Eclipse and made a total of 30 plans (6 & 10 MV, three field sizes, five phantom thicknesses of 5, 10, 20, 30, 40 cm) using two isocenters and two spoilers to simulate the patient turning over in a AP/PA arrangement. Arc plans were created on phantoms 20 and 30 cm thick, 40 cm wide, 200 cm long (to encompass patient positioning) with 24 arcs of 5° each anteriorly (spanning 120°) and another 24 posteriorly (jaws: 20x40 cm²). Weights of each 5° arc were changed iteratively to produce the flattest dose profile in the center of the phantom.

Results: For the static fields, TPS calculations and ion chamber measurements agreed within 2.5% for all setups with beam spoiler. Regarding dose homogeneity: 10 MV produced the more homogenous dose, while field size had a small effect. For the arc plans, DVH metrics showed very good dose coverage: 96% of the 20 cm phantom and 85% of the 30 cm phantom received the prescription dose ±10%.

Conclusion: The high agreement between ion chamber measurement and calculation indicates Eclipse can accurately model doses for very extended distances, justifying the use of TPS calculations to calculate arc weights for TBI plans. The arc plans results suggest that 10 MV arc-TBI with TPS-guided weights can produce adequate total body dose with high homogeneity, likely higher than most non-CT-based TBI on linacs. These results will need to be validated on anthropomorphic phantoms to verify dose homogeneity to thicker and thinner sections of the patient, and lung sparing. We believe our method will allow robust and high quality TBI while maintaining simplicity and efficiency for the clinic.