Type of publication:
Conference abstract
Author(s):
*Patel A.; Albaladejo M.M.; Puchades V.P.; Amores D.R.; Arteaga J.S.; Gonzalez A.O.; Berna A.S.
Citation:
Radiotherapy and Oncology. Conference: ESTRO 2024. Glasgow United Kingdom. 194(Supplement 1) (pp S4770-S4773), 2024. Date of Publication: 01 May 2024.
Abstract:
Purpose/Objective: The purpose of this study is to demonstrate the experience in commissioning and optimising Varian's Mobius3D secondary dosimetry software for IMRT/VMAT patient specific QA using Varian TrueBeam HD-MLC and Ethos linacs, performed in the radiotherapy department at Complejo Hospitalario Universitario de Cartagena (CHUC). Material/Methods: Mobius3D provides an independent plan check against the TPS using a separate beam model and dose algorithm. This can be quantified with a 3D gamma pass rate (3%, 3mm threshold at CHUC), as well as point dose differences of seven positions within a Mobius Verification Phantom (MVP), which can be practically verified using a Semiflex 3D ionisation chamber (PTW 31021). For every plan at CHUC, this is initially done in the phantom's central position. Mobius3D was commissioned following Varian's guide for Ethos (energy 6FFF) and TrueBeam (energies 6X and 6FFF) linacs, which included a reference point dose calculation, and verification of the PDD curves, output factors, wedge factors, off-axis ratios, and the CT electron density table. The system was then evaluated against simple conformal plans, followed by more complex clinical VMAT/SBRT/SRS plans. As per the Mobius3D commissioning guidance provided by Varian, if plans are systematically returned with target volumes too hot or too cold with respect to the TPS then it is recommended that the model's dosimetric leaf gap (DLG) correction value is adjusted, which may be different for VMAT and IMRT techniques. For optimisation, Varian recommends using a small ionization chamber within an MVP insert to measure the delivered point dose from 5-10 clinical plans and comparing against the values returned by Mobius3D. This was performed on 8 IMRT plans over a range of DLG correction values. However, as the 3D gamma pass rate metric is generally more often used for comparing dose distributions, it may be more beneficial to optimise against this rather than the point dose difference. This was therefore also performed following the point dose optimisation. Result(s): The results following commissioning from plans on the Ethos linac were promising; for the default DLG values the target volume doses agreed to a sufficient degree, the 3D gamma pass rate (3%, 3mm) had a median of 99.5%, and the point dose difference had a median of -0.1%, as shown in Figure 1 (left and centre) for approximately 200 plans. This was also similar for VMAT plans on the TrueBeam linac. However, the Mobius3D results for IMRT treatments on the TrueBeam model gave target dose distributions which were consistently lower than those provided by the Eclipse TPS (AcurosXB v16). Additionally, the 3D gamma pass rates (3%, 3mm) were below the tolerance of 95%, with a median of 83.1% (n = 21), also shown in Figure 1 (right side). This therefore required optimisation of the DLG values. In this last scenario, the average point-dose difference between the plans was found for each DLG value. A trendline was plotted using linear regression, as depicted in Figure 2, and the DLG corresponding to a 0% point-dose difference was found to be 1.48 +/- 0.05mm. Similarly, the 3D gamma (3%, 3mm) pass rates were also found for each DLG value. Polynomial regression was performed to fit a cubic function to this data, also depicted in Figure 2, which gave a maximum corresponding to a DLG of 1.25 +/- 0.4mm. Considering the results from both methods, the DLG correction on the Mobius3D system for this TrueBeam and both energies was set as the average 1.4mm for IMRT, and 0mm for both VMAT and Ethos. This first value aligns closely with the value used for TrueBeam plans on the Eclipse TPS. From analysing plans following this optimisation, it was observed that the gamma3D (3%, 3mm) pass rates significantly improved, with a new higher median of 96.5% (n = 28, p < 0.001), as shown in Figure 1 (right). Conclusion(s): There is the need to optimize the DLG value for IMRT treatment plans on TrueBeam HD-MLC. Following this adjustment, the Mobius3D software gave satisfactory agreements to the TPS dose distribution for TrueBeam IMRT plans, with a substantial increase in 3D gamma (3%, 3mm) median pass rates. Ethos plans gave strong agreements without the need for optimisation, as did TrueBeam VMAT plans for the software default DLG values. It can therefore be concluded that the Mobius3D software offers a rigorous independent dose check against the TPS and is suitable for clinical use on Ethos and TrueBeam platforms, provided that a proper verification and optimization process has been previously performed.
DOI: 10.1016/S0167-8140%2824%2901289-1
