Comparison of anatomically informed class solution template trajectories with patient‐specific trajectories for stereotactic radiosurgery and radiotherapy

Abstract

Class solution template trajectories are used clinically for efficiency, safety, and reproducibility. The aim was to develop class solutions for single cranial metastases radiotherapy/radiosurgery based on intracranial target positioning and compare to patient‐specific trajectories in the context of 4π optimization. Template trajectories were constructed based on the open‐source Montreal Neurological Institute (MNI) average brain. The MNI brain was populated with evenly spaced spherical target volumes (2 cm diameter, N = 243) and organs‐at‐risk (OARs) were identified. Template trajectories were generated for six anatomical regions (frontal, medial, and posterior, each with laterality dependence) based on previously published 4π optimization methods. Volumetric modulated arc therapy (VMAT) treatment plans generated using anatomically informed template 4π trajectories and patientspecific 4π trajectories were compared against VMAT plans from a standard four‐arc template. Four‐arc optimization techniques were compared to the standard VMAT template by placing three spherical targets in each of six anatomical regions of a test patient. This yielded 54 plans to compare various plan quality metrics. Increasing plan technique complexity, the total number of OAR maximum dose reductions compared to the standard arc template for the 6 anatomical classes was 4+/−2 (OFIXEDc) and 7+/−2 (OFIXEDi). In 65.6% of all cases, optimized fixed‐couch positions outperformed the standard‐arc template. Of the three comparisons, the most complex (OFIXEDi) showed the greatest statistical significance compared to the least complex (VMATi) across 12 plan quality metrics of maximum dose to each OAR, V12Gy, total plan Monitor Units, conformity index, and gradient index (p < 0.00417). In approximately 70% of all cases, 4π optimization methods outperformed the standard‐arc template in terms of maximum dose reduction to OAR, by exclusively changing the arc geometry. We conclude that a tradeoff exists between complexity of a class solution methodology compared to patient‐specific methods for arc selection, in the context of plan quality improvement.

Static couch non-coplanar arc selection optimization for lung SBRT treatment planning

“…As discussed in the methodology, OAR constraints were not applied to the treatment planning in general to ensure that plan quality metric statistics were biased as little as possible by the VMAT optimizer. Previous research where OAR objectives were included yielded dose reductions for optimized geometries in realistic conditions (MacDonald and Thomas 2015, MacDonald et al 2020, QM Reis et al 2021, Lincoln et al 2022). In this research, we aim to demonstrate the potential gains made due to arc geometry alone for the entire patient cohort.…”

“…OAR sparing has been correlated with choosing optimized couch and gantry combinations based on geometric overlap scores (Dong et al 2013a, 2013b, MacDonald and Thomas 2015 . When 4π methods are applied, OAR sparing has been realized with IMRT (Llacer et al 2009, Dong et al 2013a, 2013b, static couch VMAT (MacDonald et al 2020, Lincoln et al 2022), and dynamic couch and gantry VMAT (Smyth et al 2013, MacDonald and Thomas 2015 for a variety of anatomical sites and fractionation schemes. The principal efficiency improvements come from the nature of simultaneously rotating the couch and gantry together, thus covering more 4π sampling in less time (Smyth et al 2013, MacDonald and Thomas 2015, MacDonald et al 2018.…”