ESTRO 2020 Abstract Book

S393 ESTRO 2020

the repeated CTs, the WDO plan performed better, but still failed the criteria.

OC-0701 Can we use the same planning strategies for diaphragm motion in IMPT as we use in VMAT? S. Visser 1 , H. Neh 2 , C. O. Ribeiro 1 , E.W. Korevaar 1 , A. Meijers 1 , B. Poppe 2 , S. Both 1 , J.A. Langendijk 1 , C.T. Muijs 1 , A. Knopf 1 1 University Medical Centre Groningen, Department of Radiation Oncology, Groningen, The Netherlands ; 2 University of Oldenburg, Medical Radiation Physics Working Group, Oldenburg, Germany Purpose or Objective Since oesophageal cancers are in close proximity to critical structures, a clinical gain is anticipated using intensity- modulated proton therapy (IMPT) instead of volumetric- modulated arc therapy (VMAT). However, IMPT is known to be sensitive to motion and range uncertainties. Especially for the treatment of distal oesophageal tumours, plan robustness remains a concern due to the diaphragm moving in and out of the beam. Applying a density override to the region where the diaphragm is moving is a method to ensure target coverage in VMAT. In this study, we evaluate if this override approach is also useful in IMPT planning, and its potential benefit towards robustness over the Weekly 4DCTs and reconstructed averaged CTs (aCT) of ten oesophageal cancer patients were used in this study. For each patient, two robust optimised IMPT plans were created on the initial aCT: one with (WDO) and one without (NDO) a diaphragm density override of 1.05 g/cm 3 . The override region was delineated using the extreme phases (inhale and exhale) of the planning 4DCT (Figure 1A). The override was removed before final Monte Carlo dose calculation. A right-posterior-oblique and a posterior- anterior beam were used for all plans (Figure 1B). Dose was prescribed to the internal target volume (ITV), which was delineated on all weekly aCTs, based on all phases of the respective 4DCT. When all clinical criteria were met, robustness evaluation was performed, simulating errors in patient setup (8 mm) and CT densities (±3%). For acceptance, the resulting voxel-wise minimum (VW min ) and voxel-wise maximum (VW max ) dose distributions had to be within pre-defined thresholds regarding target coverage and hotspots. Target coverage over the treatment course was robustly evaluated on all repeated aCTs, considering setup and range errors (2 mm and ±3% respectively). Results Plan outcomes were very similar for both WDO and NDO plans. However, visual inspection of the VW max showed consistent lateral dose shifts between the WDO and the NDO plans. For one patient, this shift made the WDO plan superior (Figure 1B–Patient 2). For other patients, hotspots were created behind the override region (Figure 1B- Patient 1). We found no differences in target coverage in the weekly evaluations (Table 1). For patient 9, the coverage of the ITV was compromised in all weeks. In this patient, where the diaphragm position was more cranial in course of treatment. Material and Methods

Conclusion Since protons are more sensitive to density variations than photons, the translation of override methods from VMAT to IMPT planning requires some caution. The override of regions to higher densities in the proton beam path can create hotspots. There was no benefit of the diaphragm override approach in terms of inter-fractional robustness. We therefore do not recommend this approach as a default option in IMPT robust optimisation. OC-0702 Is there a correlation between robustness and tumor control for skull base proton PBS treatments? C. De Angelis 1 , F. Albertini 1 , D.C. Weber 1 , M. Walser 1 , A. Lomax 1 , A. Bolsi 1 1 PSI, Protontherapy, Villigen, Switzerland Purpose or Objective Chordomas and Chondrosarcomas are Skull base bone tumors typically treated with proton therapy. Due to their proximity to critical organs, the surgical approach for a total resection, as well as the delivery of high dose radiation with photons, is often not possible. Due to the well-defined range of protons, highly conformal dose distribution to the target can be delivered while sparing neighboring critical structures. However, there are fears that plans might be sensitive (unrobust) to range and set- up uncertainties, potentially affecting patient’s clinical outcome. In this study, we investigated potential correlations between the plan robustness (to both range and set-up uncertainty) and tumor recurrence for skull base