Quantitative evaluation of 3D dosimetry for stereotactic volumetric-modulated arc delivery using COMPASS.

The purpose of this study was to evaluate quantitatively the patient-specific 3D dosimetry tool COMPASS with 2D array MatriXX detector for stereotactic volumetric-modulated arc delivery. Twenty-five patients CT images and RT structures from different sites (brain, head & neck, thorax, abdomen, and spine) were taken from CyberKnife Multiplan planning system for this study. All these patients underwent radical stereotactic treatment in CyberKnife. For each patient, linac based volumetric-modulated arc therapy (VMAT) stereotactic plans were generated in Monaco TPS v3.1 using Elekta Beam Modulator MLC. Dose prescription was in the range of 5-20 Gy per fraction. Target prescription and critical organ constraints were tried to match the delivered treatment plans. Each plan quality was analyzed using conformity index (CI), conformity number (CN), gradient Index (GI), target coverage (TC), and dose to 95% of volume (D95). Monaco Monte Carlo (MC)-calculated treatment plan delivery accuracy was quantitatively evaluated with COMPASS-calculated (CCA) dose and COMPASS indirectly measured (CME) dose based on dose-volume histogram metrics. In order to ascertain the potential of COMPASS 3D dosimetry for stereotactic plan delivery, 2D fluence verification was performed with MatriXX using MultiCube phantom. Routine quality assurance of absolute point dose verification was performed to check the overall delivery accuracy. Quantitative analyses of dose delivery verification were compared with pass and fail criteria of 3 mm and 3% distance to agreement and dose differences. Gamma passing rate was compared with 2D fluence verification from MatriXX with MultiCube. Comparison of COMPASS reconstructed dose from measured fluence and COMPASS computed dose has shown a very good agreement with TPS calculated dose. Each plan was evaluated based on dose volume parameters for target volumes such as dose at 95% of volume (D95) and average dose. For critical organs dose at 20% of volume (D20), dose at 50% of volume (D50), and maximum point doses were evaluated. Comparison was carried out using gamma analysis with passing criteria of 3 mm and 3%. Mean deviation of 1.9% ± 1% was observed for dose at 95% of volume (D95) of target volumes, whereas much less difference was noticed for critical organs. However, significant dose difference was noticed in two cases due to the smaller tumor size. Evaluation of this study revealed that the COMPASS 3D dosimetry is efficient and easy to use for patient-specific QA of VMAT stereotactic delivery. 3D dosimetric QA with COMPASS provides additional degrees of freedom to check the high-dose modulated stereotactic delivery with very high precision on patient CT images.