G. G. Y. Kim, and X. Ray; Department of Radiation Medicine and Applied Sciences, University of California San Diego, La Jolla, CA
Purpose/Objective(s): This study examines the enhancements in plan quality achieved through a prototype plan optimization algorithm (VMATp) that employs dynamic collimator rotation and integrates static angle ports in VMAT treatments. We present a preliminary evaluation of its effectiveness for complex recurrent glioblastoma (GBM) cases. Materials/
Methods: 10 complex recurrent GBM cases treated at our institution with VMAT were replanned with three sets of parameters to evaluate the novel optimizer: (i) dynamic collimator rotation during gantry rotation, (ii) dynamic collimator rotation and 5 integrated static ports with user-selected collimator angles, and dynamic collimator rotation and 5 integrated static ports with optimizer selected collimator angles, (iii) one VMAT arc and 5 integrated static ports with manual and full optimized dynamic collimator rotation for a total of 60 plans. All plans used the prescriptions (median dose 35Gy in 10 fx +/- 5 fx, 3 cases with simultaneous integrated boost), optimization objectives, and plan normalization from the clinically approved plan. Plan quality was assessed by comparing target coverage, target point max, organ-at-risk (OAR) sparing, dose gradient, conformity, and delivery efficiency against the clinical plan using paired t-tests. Results: VMATp demonstrated similar target dose coverage to traditional VMAT plans. Using dynamic collimator rotation, the Brain-PTV Dmean was reduced by a median of 3.1% (p=0.017) and the dose fall-off distance (Gradient Measure) was reduced from a median of 1.55 cm to 1.52 cm (p=0.006). Doses to other structures, including the brainstem, optic chiasm, optic nerves, and cochlea, were not statistically different from those in the clinical VMAT plan. 5 manually selected integrated ports fields on top of the original VMAT arcs had equivalent plan quality, while including 5 static ports with full optimization for the collimator rotation, led to a median of 1.6% (p=0.032) dose reduction in the Brain-PTV Dmean and slight decrease in MUs. From test (iii) using a single VMAT arc with 5 fields using the full optimization algorithm, the Monitor Units decreased significantly by 2.9% (p=0.018). Conclusion: A preliminary evaluation was conducted for a novel VMATp optimizer, featuring dynamic collimator rotation and integrated static angle ports, specifically for complex recurrent GBM cases. The increased degrees of freedom showed modest enhancements in dose gradient and reduced exposure to the normal brain. Future research will explore whether adjusting the optimization objectives could better utilize the extra flexibility offered by this sophisticated treatment approach.
