ESTRO 2021 Abstract Book

S511

ESTRO 2021

1 Amsterdam UMC, Radiotherapy, Amsterdam, The Netherlands

Purpose or Objective Due to more advanced therapies for patients with multiple brain metastases (MBM), overall survival increased and high rates of local control can be achieved in patients with good performance scores. Complex cases of MBM can be treated rapidly with highly conformal mono-isocentric VMAT techniques. In this study, 2 of such (new) non-coplanar techniques are compared: the HyperArc (äVarian ) and our in-house developed VMAT technique. Materials and Methods This study included 25 patients (108 lesions) with 2-8 MBM. Ten patients presented 2-3 MBM, 10 patients 4-5 MBM, and five patients 6-8 MBM. GTV was delineated on a co-registered T1 weighted MR-scan, using 1 mm slice thickness and 3D distortion correction. A 1 mm PTV-margin was used from GTV to PTV. PTV volumes ranged from 0.01-33cc. All patients were treated on a standard TrueBeam accelerator with 6 degrees-of- freedom couch with 10X-FFF using a stereotactic mask (Brainlab AG) and using a RapidArc technique (RA- VMAT) using 4 arcs (1 dual coplanar, 1 dual non-coplanar, fig 1). The clinical plans were re-planned with Hyperarc-VMAT (HA-VMAT) using 4 arcs (3 non-coplanar and 1 coplanar, fig 1). For both techniques high- resolution structures, a grid size of 1.25 mm, and the ACUROS algorithm were used. Specific tools were used for the HyperArc optimization process: automatic isocenter positioning and collimator angle selection, automated lower dose objective for multiple targets, and the SRS normal tissue objective. For the RA-VMAT plans isocenter position and collimator angle were manually optimized. Structures used for optimization were 2 ring structures outside the PTV and a brain minus PTV (Brain-PTV) structure. For both RA-VMAT and HA-VMAT techniques the Paddick Conformity index (CI) and Gradient Index (GI), as well as the Gradient Measure (GM, Eclipse) and V12Gy of Brain-PTV were compared.

Results Values of GI and V12Gy of Brain-PTV were significantly higher (using a paired sample t-test, p <0.05) for HA- VMAT although mean differences were small, 5.6 and 7.7 for HA-VMAT versus 5.3 and 7.4 for RA-VMAT, respectively. CI and GM were not significantly different with mean values of 0.79 and 0.45 cm for RA-VMAT and 0.80 and 0.46 for HA-VMAT (fig 2). Mean values of Dmax of PTV were 142% of the prescribed dose for RA- VMAT and 141% for HA-VMAT. For HA-VMAT the amount of MU/Gy was reduced considerably compared to RA- VMAT, mean values (SD) were 381 (76) and 475 (155) respectively. Larger differences between RA-VMAT and HA-VMAT were found in individual cases, due to differences in volume and distance between lesions.