ESTRO 38 Abstract book

S983 ESTRO 38

effect of these features on dosimetry and delivery metrics for radiotherapy to locally advanced lung tumours. Material and Methods 12 randomly selected locally advanced NSCLC patients planned to 60 Gy in 30 fractions were prospectively re- optimised. All geometry and optimisation settings were constant except for ACC (‘On’ vs ‘off’) and ASC (‘off’, ‘moderate’, ‘very ‘high’). Two ipsilateral VMAT arcs were used and plans were calculated with AcurosXB (AXB, D m ) and AAA (both version 15.5.11). To facilitate comparison of near minimum and near maximum doses the plans were normalised to median PTV dose received prescription dose. Intermediate dose was calculated during the optimisation. The following parameters were extracted: MU/Gy, edge metric (MLC open leaf edges/area), D98% for GTV and PTV, mean heart dose, D0.33cc for the oesophagus, and D5Gy and D20Gy for the (lungs-CTV) structure. Within each algorithm, ASC ‘moderate’ and ‘very high’ was compared with ASC ‘off’, and ACC ‘off’ was compared with ACC ‘on’, using paired two tailed t- test. Results Table 1 shows the dosimetry variation for target and OARs, and plan complexity metrics. For AXB and AAA, as ASC was changed from ‘off’ to ‘moderate’ and ‘very high’ there was no statistical difference in GTV D98%, PTVD98%, PTVD2%, mean heart or oesophagus D0.03cc. Lung dose however increased slightly. Use of aperture shape controller decreased plan complexity (edge metric and MU/Gy). Turning ACC ‘off’ resulted in no difference in treatment plans for AXB, but worse GTV D98% for AAA. ACC had no impact on plan complexity. Table 1: Mean ± st. dev. dosimetry and plan efficiency metrics for AXB and AAA plans optimised with varying ASC and ACC settings. Blue cells represent statistical significance at p < 0.0005, green at p < 0.005 and red at p < 0.05, relative to ASC ‘Off’ for ASC variation and comparing ACC ‘Off’ against ‘On’ Conclusion The Air Cavity Correction is useful for improving target coverage with AAA but not with AXB. Use of Aperture Shape Controller results in no difference in target coverage, but slightly worse lung dose. ASC reduces treatment plan complexity and monitor units. EP-1815 MCO in VMAT treatment planning for locally advanced head and neck cancer B. Farnault 1 , V. Favrel 1 , L. Moureau-Zabotto 1 , J. Rolland 2 , A. Tallet 1 , P. Fau 1 1 Institut Paoli Calmettes, Radiotherapy Oncology, Marseille, France ; 2 Centre Hospitalier Intercommunal des Alpes du Sud, Radiotherapy Oncology, Gap, France Purpose or Objective Efficacy of inverse planning is becoming increasingly important for advanced radiotherapy techniques, and especially for head and neck cancer to decrease radiation therapy toxicity. However, the inverse planning process can create a suboptimal plan despite meeting all constraints. Multicriteria optimization (MCO) may improve doses at organs at risk (OARs) and provides better treatment planning without being time consuming.The aim of this study was therefore to evaluate the benefit of VMAT with multi-criteria optimization (MCO) in RayStation (v6.1.1.2, RaySearch Laboratories, Sweden) for head and neck cancer patients and compare the DVH difference

Purpose or Objective To evaluate the dose calculation accuracy of the Varian Eclipse anisotropic analytical algorithm (AAA) for stereotactic ablative body radiotherapy (SABR) in comparison with Monte Carlo (MC) in order to investigate the dosimetric consequences to organs at risk (OAR) and coverage of planning target volume (PTV) in lung SABR plans. Material and Methods 25 cases of non–small-cell lung cancer (NSCLC) that were previously treated with SABR to 48 Gy in 4 fractions at our center were selected for this study. These cases were treated from March 2016 to February 2018 and were selected such that the PTV size covers a wide range, from 8.9 cc to 163.2 cc. The internal gross target volume (IGTV) has been contoured from the 4DCT and a 5 mm isotropic expansion was applied to form PTV. The original treatment plans were calculated with 6 MV flattening filter free (FFF) beams using AAA in Eclipse treatment planning system (TPS). The same plans were recalculated using MC for the purpose of this study. Dose volume histogram (DVH) data has been exported for all cases and later processed using in-house code developed in the R programming language. The following dose-volume parameters were used for the comparison: V 100% , V 90% and D min , to the PTV; conformality index (V 100% /V PTV ), low dose conformality (V 50% /V PTV ) and D 2cm ; V 100% to IGTV; V 20Gy for the lung; dose parameters to OARs including chest wall, esophagus, great vessels, brachial plexus, trachea, heart, bronchial tree, skin and cord. The statistical comparison has been done by paired t-test analysis. Results Comparable results were obtained for AAA and MC calculations except for V 100% to the PTV (p < 0.001), conformity index (p = 0.008), low dose conformality (p < 0.001) and lung V 20Gy (p < 0.001). The largest difference was observed for V 100% to PTV which in turn impacts the conformality index too. AAA calculations underestimated dose to lungs and PTV compared to MC. The dose differences in PTV D min and PTV V 90% as well as D 2cm were not statistically significant. No correlation has been observed between the PTV size and the dose differences between AAA and MC. Occasionally, MC revealed hot and cold spots which were not present in the AAA calculations. Conclusion Our results demonstrate good agreement between AAA and MC doses to OARs for the most part. Lung is an exception as AAA underestimates V 100% to PTV. AAA emerges a good choice for routine planning, but occasionally MC reveals hot or cold spots in sensitive places, e.g. the PTV center and therefore any plan- specific QA strategy should employ better algorithms to detect such instances. Further investigations are necessary in larger patient cohorts to determine whether AAA is still appropriate for dose calculations in cases of very small field size lung SABR treatment plans. EP-1814 On the aperture shape controller and the air cavity correction for lung plans using AcurosXB and AAA L. Fog 1 , K. Offer 1 , N. Hardcastle 1 1 The Peter MacCallum Cancer Centre, Physical Sciences, Melbourne, Australia Purpose or Objective Volumetric modulated arc therapy (VMAT) plan optimisation for radiotherapy in the thorax is subject to uncertainties in dose calculation algorithms during optimisation and to variations in plan complexity. The Eclipse TM v15.5 (Varian Medical Systems) Photon Optimiser incorporates two features: Air Cavity Correction (ACC) to improve the accuracy of scattered dose in the optimiser, and Aperture Shape Controller (ASC) to increase contiguity of apertures during VMAT delivery. We investigate the