Enhancing Delivery Efficiency on the Magnetic Resonance-Linac: A Comprehensive Evaluation of Prostate Stereotactic Body Radiation Therapy Using Volumetric Modulated Arc Therapy

Purpose: Long treatment sessions are a limitation within magnetic resonance imaging guided adaptive radiation therapy (MRIgART). This work aims for significantly enhancing the delivery efficiency on the magnetic resonance linear accelerator (MR-linac) by introducing dedicated optimization and delivery techniques for volumetric modulated arc therapy (VMAT). VMAT plan and delivery quality during MRIgART is compared with step-and-shoot intensity-modulated radiation therapy (IMRT) for prostate stereotactic body radiation therapy.

Methods and materials: Ten patients with prostate cancer previously treated on a 1.5T MR-linac were retrospectively replanned to 36.25 Gy in 5 fractions using step-and-shoot IMRT and the clinical Hyperion optimizer within Monaco (Hyp-IMRT), the same optimizer with a VMAT technique (Hyp-VMAT), and a research-based optimizer called optimal fluence levels and pseudo gradient descent with VMAT (OFL+PGD-VMAT). The plans were then adapted onto each daily magnetic resonance imaging data set using 2 different optimization strategies to evaluate the adapt-to-position workflow: "optimize weights" (IMRT-Weights and VMAT-Weights) and "optimize shapes" (IMRT-Shapes and VMAT-Shapes). Treatment efficiency was evaluated by measuring optimization time, delivery time, and total time (optimization+delivery). Plan quality was assessed by evaluating organ at risk sparing. Ten patient plans were measured using a modified linac control system to assess delivery accuracy via a gamma analysis (2%/2 mm). Delivery efficiency was calculated as average dose rate divided by maximum dose rate.

Results: For Hyp-VMAT and OFL+PGD-VMAT, the total time was reduced by 124 ± 140 seconds (P = .020) and 459 ± 110 seconds (P < .001), respectively, as compared with the clinical Hyp-IMRT group. Speed enhancements were also measured for adapt-to-position with reductions in total time of 404 ± 55 (P < .001) for VMAT-Weights as compared with the clinical IMRT-Shapes group. Bladder and rectum dosimetric volume histogram (DVH) points were within 1.3% or 0.8 cc for each group. All VMAT plans had gamma passing rates greater than 96%. The delivery efficiency of VMAT plans was 89.7 ± 2.7 % compared with 50.0 ± 2.2 % for clinical IMRT.

Conclusions: Incorporating VMAT into MRIgART will significantly reduce treatment session times while maintaining equivalent plan quality.