Monte Carlo Evaluation of Magnetic-Field Effects In Prostate Step-and-Shoot IMRT on a 0.5 T Inline Linac-MR
Abstract
Purpose
To simulate prostate treatment plans with Monte Carlo (MC) and compare with dose calculations from RayStation 11B (RS-11B) treatment planning system (TPS) for a 0.5 T bi-planar inline Linac-MR (LMR).
Methods
A validated TOPAS MC model of the LMR, including a custom multileaf collimator and a 3D magnetic field that rotates with gantry, is further developed to simulate step-and-shoot IMRT (sIMRT) plans. Several 5-field sIMRT prostate plans are generated in a clinically validated RS-11B (no magnetic field (Bo)), and MC simulated (with and without Bo). The MC-calculated dose is normalized to PTV60 (Rx) and compared with TPS calculations for composite plans and individual fields, using 3D gamma analysis (3% | 2 mm). Skin dose metrics, including mean dose within the hottest contiguous 1cm3 region (D̄1cc) and mean dose to skin receiving >5% of prescription Rx (D̄>5%) are evaluated for a 2-mm skin structure.
Results
Composite plan 3D gamma pass-rates ranged from 96.1–96.9% with Bo and 95.7–96.4% without Bo. Individual pass-rates are on average ~96% with and without the Bo; 3 of 15 fields fell below 95%, with minimum pass-rates of 93.7% (with Bo) and 92.6% (without Bo). For the 2-mm skin structure, the D̄1cc increased in the presence of Bo, reaching 34–40% Rx, compared to 29–33% Rx without Bo. Consistent with this trend, the D̄>5% ranged from 15.5–16.3% Rx with Bo and 14.3–15.0% Rx without Bo.
Conclusion
A patient-specific framework is established to simulate sIMRT plans using the LMR MC model. The primary magnetic-field–induced effect observed in the simulated prostate plan is elevated skin dose, with D̄1cc increased on average by 6.2% in simulations with Bo relative to without Bo scenarios, and D̄>5% increased by 1.23%. Overall, the MC-calculated plan with Bo shows no significant differences compared to RS-11B calculations, which does not account for Bo.