ESTRO 37 Abstract book

S1075

ESTRO 37

specification differs significantly from conventional arm- gantry linacs. The department has also three beam- matched Clinacs on-site with 120-leaf Millennium MLC. Halcyon requires Aria v15.1. Its beam model in Eclipse is reconfigured and cannot be changed by the user; the Halcyon is tuned at the factory to conform the beam model and the machine can be tuned on site if necessary. Halcyon has a single 6MV FFF beam energy, with a double-stacked MLC giving a 28cmx28cm maximum field size. Each stack’s leaves project 1cm at the isocentre, and the stacks are offset by 0.5cm giving a potential resolution of 0.5cm: in Eclipse v15.1, only the lower layer of leaves is used for optimisation/field shaping. The rotational gantry speed is up to 2 revs/min to deliver VMAT treatments, and up to 4 revs/min between static- gantry IMRT fields. Halcyon v1.0 has MVCBCT and orthogonal pairs for MV imaging, and daily imaging is mandatory. Eclipse calculates the dose from this imaging and takes it into consideration when optimising, i.e. it is part of the treatment dose. Twenty pelvic cases were planned for intended treatment on Halcyon, with Clinac 10MV plans as back-ups. Eclipse v15.1 AAA algorithm was used to calculate dose for all plans. VMAT plans had two or three full arcs; IMRT plans had five or seven fields, using sliding-window technique. All plans were normalised with prescription dose to median volume. MV CBCTs were obtained on Halcyon and for these the imaging dose was calculated in the plan. Plans were analysed according to department protocols, using DVHs to ensure that OARs and target coverage achieved clinical criteria. Results All plans passed clinical criteria specified in our protocols and were appropriate for treatment. Halcyon plans had up to 25% more MUs than the 10MV Clinac plans. This is to be expected for 6MV compared with 10MV, but also the 1cm-wide Halcyon leaves may not be optimising as efficiently as the Clinac’s central 0.5cm leaves. In the patient scan, the total dose was up to 29% higher in Halcyon plans, but note that this includes imaging dose, which accounts for up to 11.3% of the dose and is not currently calculated in Clinac plans. Conclusion The Halcyon can be used to treat pelvic sites that are currently treated on a conventional 10MV linac, with comparable and clinically-acceptable plan quality. EP-1974 Multi-scenario robustness evaluation; transition to a ‘proton proof’ alternative to PTV evaluation E.W. Korevaar 1 , D. Scandurra 1 , M. Gelderman 1 , R.G. Kierkels 1 , A.C. Knopf 1 , S. Both 1 , M. Unipan 2 , M.G.C. Eenink 3 , J.A. Langendijk 1 1 University of Groningen- University Medical Center Groningen, Department of Radiation Oncology, Groningen, The Netherlands 2 Zuid-Oost Nederland Protonen Therapie Centrum, Maastricht, The Netherlands 3 Holland Proton Therapy Center, Delft, The Netherlands Purpose or Objective When dosimetrists, physicians and physicists review the dose distribution of an intended treatment, PTV coverage is checked to ensure that the CTV receives adequate dose despite patient setup errors. The underlying assumption that the dose distribution is invariant to errors is not generally true in photon treatments and becomes problematic in intensity modulated proton treatments where range errors should be taken into account as well. This limitation is solved in a multi-scenario simulation by explicitly calculating doses for various (typically > 10) error scenarios. Clinical introduction of this method to replace PTV based plan evaluation, however, introduces challenges of how to evaluate multiple dose distributions and what criteria to apply for consistency with (historic)

PTV planned treatments. The purpose if this work was to determine which robustness evaluation methods and which criteria are suitable for transition to a PTV-less treatment plan evaluation procedure that is generally applicable to photon treatments as well as to proton treatments. Material and Methods To obtain a clinically feasible workflow for evaluation of multiple scenario doses, including visual inspections, the relatively high number of scenario doses needed to be replaced by just a few evaluation dose distributions from which risk of under-dosage of the CTV and over-dosage of critical structures could be determined. Included were worst scenario dose, voxel-wise worst dose and voxel- wise mean dose. Determined metrics for the CTV were the V95 of these doses, as well as the average V95 of the scenario doses (). Acceptance criteria for these metrics were then determined in a calibration procedure by comparison to clinical PTV evaluations for 18 HN IMRT/VMAT treatment plans, to obtain consistency in the pass ratio with the PTV evaluations. Results The PTV V95>98% criterion corresponded 1:1 to the CTV V95>98% for the voxel-wise minimum dose, while for the average V95 of the scenario doses it corresponded to > 99.8%. For the high-dose CTV the same number of plans passed for PTV as well as for evaluation and (except for one plan) for the voxel-wise minimum evaluation. The low-dose target showed lower pass rates for and voxel-wise minimum dose compared to the PTV evaluation (Table 1). This indicates that the criteria were more strict for the low-dose CTV compared to the PTV evaluation and could be slightly relaxed to accept the same plans. Found differences, however, were caused by regions where dose was reduced by only about 1Gy, and would only need minor adjustments to bring the pass rate in agreement with the PTV evaluation. Visual inspection showed generally good agreements between PTV evaluation and voxel-wise minimum dose to the CTV (Figure 1).