Abstract Book

S269

ESTRO 37

Results Average dosimetric parameters, late complication probabilities and plan parameters are compiled in Table 1. Figure 1 shows the average dose-volume histogram for a selection of organs-at-risk. Mimicking TB2 to HA2 degraded the dose homogeneity in PTV High and increased the dose to the ipsilateral parotid. HA2 plans however consistently were clinically acceptable. HA3 improved the dose distribution quality in the high-dose target. Though still inferior to TB2, the achievable target homogeneity with HA3 was well within our clinical criteria. Also, the glottic area and swallowing muscles were better spared with HA3 compared to TB2, resulting in a minor (-1.3%) but statistically significant reduction in late dysphagia risk. The number of MUs to be delivered per arc was significantly lower for HA3 compared to TB2 and HA2. The reduction in imaging and plan delivery time compared to TB2 was 1.5 minute and 1.0 minute for HA2 and HA3, respectively, which was verified on the actual machines. Conclusion VMAT of HNC with HA was described for the first time. Compared to our current dual arc class solution with TB, equal or better plan quality has been achieved using 3 arcs with HA in a considerably reduced treatment time.

344minutes, respectively. The optimization time was similar for both RP versions with an optimization time of 27min on average; AIO needed on average 4min longer to perform the optimization and AP and RS needed on average 56 and 190 additional minutes. Table 1

Conclusion All ATPS were able to fulfill all the required planning DVH constraints and the effective working time was kept bellow 20’ for each ATPS except for OC-0514 VMAT treatment planning for head-and-neck cancer with the novel fast-rotating linac Halcyon S. Michiels 1 , K. Poels 2 , W. Crijns 2 , B. Vanstraelen 2 , K. Haustermans 1,2 , S. Nuyts 1,2 , T. Depuydt 1,2 1 University of Leuven, Department of Oncology, Leuven, Belgium 2 University Hospitals Leuven, Department of Radiation Oncology, Leuven, Belgium Purpose or Objective The novel Varian Halcyon linear accelerator can offer a next step forward in the efficient image-guided delivery of intensity-modulated treatment plans. Distinctive characteristics of this unit are the increased leaf speed (5 cm/s), the dual-layered low leakage (<0.1%) MLC, the absence of leaf over-travel limitation, the 1 cm leaf width and the O-ring encapsulated gantry design, allowing increased gantry speeds (up to 4 rpm). Since clinical experience with Halcyon treatment planning is lacking, data is presented acquired during the transition in our clinic from volumetric modulated arc therapy (VMAT) on a Varian TrueBeam (TB) to Halcyon (HA) in head-and-neck cancer (HNC). Material and Methods Ten patients with locally advanced HNC were planned for parotid- and swallowing-sparing VMAT with TB and HA. Prescription was 70 Gy to the high-risk planning target volume (PTV High ) and 54.25 Gy to the low-risk PTV, to be delivered simultaneously in 35 fractions. TB plans were generated according to our class solution TB2, dual arc VMAT, with collimator angles 10°/350° due to the 15 cm leaf over-travel limitation. For HA, the proposed solution consisted of two arcs (HA2) with collimators set at 10°/80°. Given the broader leafs of HA, triple arc VMAT (HA3) with collimators 10°/45°/80° was studied as well. Plans were normalized to the prescribed D Mean to PTV High. The Photon Optimizer algorithm and Anisotropic Analytical Algorithm (all v15.1) were used in Varian Eclipse for each plan. Within a patient, the same optimization weights were used for each plan. Temporarily only MV-CBCT is available for HA. Therefore this daily imaging dose was included in the plan optimization and calculation both on TB and HA, for the sake of correct comparison focusing on plan quality differences only.