SU-F-T-363: Monte Carlo Dose Verification for Volumetric Modulated Arc Therapy in Multiple Brain Metastases

Abstract

Purpose: To verify the calculation accuracy of treatment planning system (TPS)-VMAT plans in multiple brain metastases, by EGSnrc Monte Carlo (MC)-based VMAT calculations. Methods: First, six VMAT plans for four patients with multiple brain metastases were created by analytical anisotropic algorithm in an Eclipse Ver. 8.6.17 (AAA8). The accuracy of TPS-VMAT dose distributions for three algorithms of AAA8 and AAA10 and Acuros XB (AXB10) in the Eclipse Ver. 10.0.24 was verified by MCbased VMAT calculations. Similarly, a set of TPS-VMAT dose distributions were also verified by MC under the identical calculation parameters as patients in a water-equivalent phantom. Dose distributions for PTVs and OARs were quantitatively verified from dose volume histograms (DVHs) and passing rates by three dimensional (3-D) gamma analysis. Results: The dose differences between each algorithm and MC for six VMAT plans were 12.2% at a maximum and 3.5 ± 2.4% on average for D9 5 in PTVs, and 1.8% or less for mean doses in OARs in patients. Similarly, the dose differences in a water-equivalent phantom were improved up to 5.1% and 0.6 ± 1.9% for PTVs and 1.8% or less for OARs. The passing rates with 3%/3 mm criteria in PTVs were up to 57.3% in patients and 80.0% in a waterequivalent phantom. Dose differences of DVHs and the reduction of passing rates for AAA in patients may be attributed to the calculation accuracy for inhomogeneous tissues like bone around target. Similarly, for AXB10 they may result from differences in CT numbers and mass densities of materials assigned to CT images in MC and AXB10, and cut-off energy (AXB10: 0.5 MeV, MC: 0.189 MeV). Conclusion: DVHs and passing rates by 3-D gamma analysis for every structure based on MC-based VMAT dose distributions are useful for dose verification of VMAT plans with multiple brain metastases.