ESTRO 2023 - Abstract Book

S1413

Digital Posters

ESTRO 2023

addressed to all STOPSTORM project participants and was mandatory for prospective patient inclusion into the STOPSTORM database. Results 23 centers with 28 treatment units (C-arm LINAC, CyberKnife, MRidian and Synchrotron) using mostly 6-10 FFF beams collimated by MLCs with leaves from 2.5 to 10 mm or cones from 5-60mm participated in the audit survey. Eight different reference and four different small field dosimetry protocols with 14 different types of detectors are used for system commissioning and absolute dose calibration. Absolute dosimetry for reference fields is verified with an accuracy of 1% - 2%, with a frequency from daily to yearly. Furthermore, most participants perform external dosimetry audits. Planning CT quality parameters are tested regularly, and the hounds field curve stability is mostly verified on a yearly basis. Treatment unit geometric and isocenter tests (mechanical and Winston-Lutz tests) are mostly performed with different frequencies. The various imaging systems (kV, MV, CBCT, ExacTrac, Catalyst, IsoCrystal, MRI) are tested with static and reposition tests, however, image quality is mostly validated during commissioning or maintenance and less frequently on a regular basis performed by the participants itself. A system specific end-2-end test is implemented by only 65.2% of the participants, performed from monthly to yearly, while even fewer (34.7%) centers included motion simulation in this test. For patient-specific QA, 20 centers use 2D or 3D evaluation methods, however, with varying phantoms, detectors, and criteria. Conclusion This STOPSTORM multi-center treatment unit audit survey provided data concerning current quality assurance practices in major European centers performing SBRT and STAR. Notable differences and even deficiencies were identified in small field dosimetry, external audits, image guidance verification, particularly in repositioning tests and image quality assessment both on the treatment unit and on the CT scanner, end-2-end tests and patient specific-plan verification. Based on the audit, recommendations were issued to harmonize quality assurance procedures for STAR within the STOPSTORM project. 1 Azienda USL-IRCSS di Reggio Emilia, Medical Physics unit, Reggio Emilia, Italy; 2 Azienda USL - IRCSS di Reggio Emilia, Medical Physics unit, Reggio Emilia, Italy; 3 AOU di Modena, Medical Physics unit, Modena, Italy Purpose or Objective Verification procedures for patient-specific quality assurance (QA) in advanced radiotherapy are laborious and time consuming. Moreover, it has been shown that these procedures cannot detect some inaccuracies for some particular complex cases involving inaccuracies due to tissue inhomogeneity, such as in lung cases, and overmodulated plans for brain case. Aim of the study is to perform dosimetric patient-specific QA of lung and brain cancer plans through an independent Monte Carlo (MC) software based, the RadCalc system (LAP, Germany), opportunely tuned with volumetric measurements on uniform phantom (Octavius 4D). Materials and Methods The Radcalc system was used to evaluate intensity-modulated lung plans and brain dose-painting plans (Eclipse v3.7 TPS, Varian, USA) with different complexity (Modulation Complexity Score [MCS] range: 0.1-0.4). 70 VMAT treatment plans performed with AAA and Acuros XB dose calculation algorithms were selected. 20 plans (tuning-set) were used to tune the 4 MC models (5 plans for each energy: 6MV, 10MV, 6MV [FFF], 10MV [FFF]), while the other plans (test-set) were used to validate the goodness of the tuned MC models. MC models were built using specific commissioning measurements and then the additional radiation to light field-offset parameter (dosimetric leaf gap parameter) was tuned to achieve the best dose comparison agreement with phantom patient-specific QA measurements of the tuning set plans, an example for one energy is shown in Figure 1. Gamma Passing Rate (GPR) and Dose Volume Histograms (DVHs) agreements determined by MC simulations on the test-set plans were compared and evaluated to GPR dosimetric QA measurements and to Eclipse TPS DVHs, respectively. PO-1702 Radcalc 3D Monte Carlo: tuning and validation of a new independent calculation-based tool G. Sceni 1 , A. Botti 2 , O. Matteo 2 , R. Sghedoni 2 , V. Trojani 2 , D. Finocchiaro 3 , D. Lambertini 2 , A. Barani 2 , V. Bertolini 2 , M. Iori 2