Development of a New Independent Monte Carlo Dose Calculation Quality Assurance Audit Tool for Clinical Trials

Introduction: Commercially available treatment planning systems (TPS) may use a number of different radiation dose calculation algorithms during the planning process. The Radiological Physics Center (RPC), tasked with ensuring clinically comparable and consistent dose delivery amongst institutions participating in NCI funded multi-institutional clinical trials, has traditionally relied upon measurements to achieve this objective. As a supplement to the tools used by the RPC, an independent dose calculation tool is needed to determine patient dose distributions in three dimensions so as to act as a quality assurance tool for the dose calculations. Methods: Multiple source models representing the output of Elekta 6MV and 10MV and Varian TrueBeam Flattening Filter Free (FFF) 6MV and FFF 10MV therapeutic x-ray beams were developed. The Monte Carlo technique, using the Dose Planning Method (DPM) algorithm, was used in radiation dose calculations. During validation calculations were compared to open field measurements in a water phantom. Benchmarking was a measurement based comparison of mock treatment plans in anthropomorphic phantoms. Treatment plans included intensity modulated radiation therapy and stereotactic body radiation therapy techniques. Past phantom treatment plans submitted through a remote auditing program were recalculated using the tool and compared to submitted measurement data as a test of the models’ robustness. Results: The average percentage of data passing a ±2%/2mm gamma criterion during validation testing was 99.5%, 99.6%, 98.1%, and 98.1% for Elekta 6MV, 10MV, Varian TrueBeam FFF 6MV, and FFF 10MV beams, respectively. The percentage of data passing the benchmarking evaluation criterion of ±3%/2mm was 87.4%, 89.9%, 90.1%, and 90.8% for Elekta 6MV, Elekta 10MV, Varian TrueBeam FFF 6MV, and Varian TrueBeam FFF 10MV beams, respectively. Conclusions: Elekta 6MV and 10MV and Varian TrueBeam FFF 6MV and FFF 10MV multiple source models based on dose calculations using the DPM Monte Carlo code were successfully developed, validated, and benchmarked against measurements. A recalculation of TPS dose from archived phantom credentialing audits was performed as a proof of concept for the models’ utility as a quality assurance tool for use in clinical trial audits