Purpose: The purpose of this study was to evaluate the differences between optimized and final dose for lung SBRT patients using the Eclipse to create volume modulated arc therapy (VMAT) plans. The resulting database of plans was evaluated to develop insights to further assist in the planning.
Methods: Lung SBRT plans were created with Eclipse using the Photon Optimizer, the Analytical Anisotropic Algorithm, and the Acuros Algorithm. The minimum dose to the ITV, minimum dose to the PTV, PTV-D98, PTV-D80, PTV-D20, PTVmax, Cordmax, and maximum dose to the chest wall were recorded for each algorithm. The following characteristics of the ITV and PTV were also measured: volume, mean CT numbers, standard deviation of CT numbers, and closest distance to the surrounding non-lung tissue (e.g., chest wall, mediastinum, etc…). A Pearson’s coefficient test was used to evaluate possible correlation with dose differences.
Results: Treatment plans were created for 27 lung SBRT patients with PTV volumes ranging from 6.8 to 127.2 cm³. The average density of the PTV for the cohort of plans was -552 ± 266 HUs. The dose difference between optimized dose and final dose was calculated for each algorithm. Weak correlation was found between ITV location and dose difference (r = .0.5987), and strong correlation was found between the mean HU’s of the PTV minus the ITV and dose (r = 0.7757 and p < 0.0001). A relationship was developed that can be used to adjust the prescribed dose and normal tolerances during optimization to account for the anticipated changes in dose after final dose calculation.
Conclusion: The differences between optimized and final dose were evaluated for the Eclipse planning system. A relationship between the HU’s of the PTV minus the ITV was identified and a technique for compensating for dose differences was developed.
Pencil Beam Algorithms, Treatment Planning, Optimization
TH- External Beam- Photons: IMRT/VMAT dose optimization algorithms