Advanced Collapsed cone Engine dose calculations in tissue media for COMS eye plaques loaded with I-125 seeds

作者：Morrison Hali; Menon Geetha; Larocque Matthew P; van Veelen Bob; Niatsetski Yury; Weis Ezekiel; Sloboda Ron S

来源：Medical Physics, 2018, 45(7): 3349-3360.

DOI：10.1002/mp.12946

• ACE
• dosimetry
• eye plaque brachytherapy
• model-based dose calculations
• Monte Carlo

PurposeTo investigate the dose calculation accuracy of the Advanced Collapsed cone Engine (ACE) algorithm for ocular brachytherapy using a COMS plaque loaded with I-125 seeds for two heterogeneous patient tissue scenarios.
MethodsThe Oncura model 6711 I-125 seed and 16 mm COMS plaque were added to a research version (v4.6) of the Oncentra((R)) Brachy (OcB) treatment planning system (TPS) for dose calculations using ACE. Treatment plans were created for two heterogeneous cases: (a) a voxelized eye phantom comprising realistic eye materials and densities and (b) a patient CT dataset with variable densities throughout the dataset. ACE dose calculations were performed using a high accuracy mode, high-resolution calculation grid matching the imported CT datasets (0.5 x 0.5 x 0.5 mm(3)), and a user-defined CT calibration curve. The accuracy of ACE was evaluated by replicating the plan geometries and comparing to Monte Carlo (MC) calculated doses obtained using MCNP6. The effects of the heterogeneous patient tissues on the dose distributions were also evaluated by performing the ACE and MCNP6 calculations for the same scenarios but setting all tissues and air to water.
ResultsAverage local percent dose differences between ACE and MC within contoured structures and at points of interest for both scenarios ranged from 1.2% to 20.9%, and along the plaque central axis (CAX) from 0.7% to 7.8%. The largest differences occurred in the plaque penumbra (up to 17%), and at contoured structure interfaces (up to 20%). Other regions in the eye agreed more closely, within the uncertainties of ACE dose calculations (similar to 5%). Compared to that, dose differences between water-based and fully heterogeneous tissue simulations were up to 27%.
ConclusionsOverall, ACE dosimetry agreed well with MC in the tumor volume and along the plaque CAX for the two heterogeneous tissue scenarios, indicating that ACE could potentially be used for clinical ocular brachytherapy dosimetry. In general, ACE data matched the fully heterogeneous MC data more closely than water-based data, even in regions where the ACE accuracy was relatively low. However, depending on the plaque position, doses to critical structures near the plaque penumbra or at tissue interfaces were less accurate, indicating that improvements may be necessary. More extensive knowledge of eye tissue compositions is still required.

• 出版日期2018-7