Tuning a secondary dose verification software for a CT-guided online adaptive delivery system

Background: Quality assurance (QA) remains unstandardized for CT-guided online adaptive radiotherapy (CTgART) platforms (Ethos, Varian Medical Systems, Inc., Palo Alto, CA), as they become more clinically prevalent. A secondary dose calculation software (Mobius3D, Varian Medical Systems, Inc., Palo Alto, CA) is provided for this closed CTgART system. However, the clinical impact of tuning dosimetric leaf gap (DLG) correction values for specific delivery techniques for CTgART secondary dose calculations remains uninvestigated.

Purpose: Tuning the DLG correction value for different delivery techniques of the independent secondary dose verification software for adaptive online QA.

Methods: A total of 31 volumetric arc therapy (VMAT) and 13 fixed-gantry intensity modulated radiation therapy (IMRT) plans were selected from representative anatomical sites treated in our clinic. All plans were evaluated on a patient CT dataset and a QA dataset of a solid water phantom with an embedded ion chamber placed at the center in both primary treatment planning systems (TPS) and secondary dose verification software. Primary TPS plan doses were compared with secondary calculation doses on patient CT by calculating 3D gamma passing criteria under different DLG correction values ranging from -2 to 2 mm to determine the optimal DLG correction range. Ion chamber verification measurements were then compared to secondary calculation dose to determine the optimal DLG correction value by minimizing the difference for IMRT and VMAT plans, separately.

Results: The optimal DLG correction values for VMAT and IMRT techniques were -0.3 and 0.4 mm respectively for the selected clinical cohort of patients. The mean gamma passing rate between primary and secondary doses for VMAT and IMRT were 99.0% ± 1.0% and 99.9% ± 0.1% with 3%/2 mm and excluding 10% low dose criteria. The mean ion chamber agreement for VMAT and IMRT were 0.0% ± 2.1% and 0.2% ± 1.4%.

Conclusion: DLG correction value should be tuned for each delivery technique (VMAT and IMRT) separately to maximize the robustness of CTgART online QA programs.