Volumetric modulated arc therapy or step-shoot IMRT? A 4D dosimetry study of motion effect in lung SBRT using a dynamic virtual patient model

Abstract: Purpose: To investigate the impact of delivery techniques and planning parameters on interplay effect in lung SBRT. Methods: A dynamic virtual patient model containing normal structures and a tumor with adjustable sizes, locations, and 3D breathing motion was utilized. SBRT plans were developed using both step-and-shoot IMRT and VMAT with different planning parameters (energy, isocenter location, PTV margin, and PTV dose heterogeneity). 4D doses were calculated by simulating synchronized delivery of SBRT to the virtual patient model with random initial positions of tumor motion. The expected dose (average) and the standard deviation of the 4D doses were obtained. The relative difference between the expected GTV minimal/mean (GTVMin/GTVMean) dose and the planned ITVMin/ITVMean dose (denoted by %E/P), and between the GTVMin and the prescription dose (DRx) were computed. Results: The %E/P for GTVMean was significantly lower for IMRT than VMAT (0.5% +/- 7.7% v.s. 3.5% +/- 5.0%, p=0.04). The expected GTVMin was lower than DRx in 9.4% of all IMRT plans versus 3.1% in VMAT. The worst-case scenario, 4D GTVMin was 14.1% lower than the ITVMin. Choices of PTV margin or dose heterogeneity to be achieved in PTV can result in significant difference (p<0.05) in motion interplay depending on delivery techniques. Conclusion: Motion interplay may cause the expected GTVMin to be less than the planned ITV minimal dose and DRx for both IMRT and VMAT plans. The differences between the expected GTV dose and the ITV dose depended on the delivery technique and planning parameters. Overall, VMAT is less prone to motion interplay than IMRT.