Moving Phantom Study of Stereotactic Body Radiation Therapy for Lung Cancer
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Content Notes
Abstract
To assess the accuracy of current stereotactic body radiation therapy (SBRT) lung treatment methodologies, we performed a systematic evaluation using phantoms that simulated motions from real patients (irregular motions) as well as sinusoidal motions (regular motions). The irregular patterns investigated in this study were of two types: small range irregular breathing motions (=10mm) and large range irregular breathing motions (=20mm). Four-dimensional computed tomography (4DCT) and cone beam computed tomography (CBCT) are important methodologies for SBRT, but previously have only been used to evaluate regular patterns. For targets moving regularly or irregularly within a small range (7.0 ± 1.8 mm, n = 6), we observed good agreement between the measured and computed dose distributions. However, for targets moving irregularly with a larger range (20.8 ± 2.6 mm, n = 4), the measured isodose lines were found to be shifted relative to the planned distribution, resulting in an underdosing (over 20%) in a portion of the PTV. In this underdosed volume, 1-2% of the PTV is underdosed by over 18 Gy, causing a 35-40% drop in the local control rate. We further observed that the discrepancy between the planned and measured dose distribution was due to the inaccurate representation of the irregular target motion in the maximum intensity projection (MIP) images generated from 4DCT, which could not be corrected by CBCT. A method of Extended Distance Virtual Isocenter (EDVI) was developed to lower the toxicity of healthy tissues. In all, caution should be used when planning from 4DCT images in the presence of large and irregular target motion. The inaccuracy inherent in 4DCT MIP and CBCT images can be mitigated through the application of methodologies to reduce respiratory motion, such as abdominal compression, and through the use of volumetric image guidance to assure precise targeting with minimal shifts.