Dosimetric Effect of Calculation Grid Size on Epidural Space in Spine Stereotactic Body Radiotherapy Using Volumetric Modulated Arc Therapy

Abstract

To perform spine stereotactic body radiotherapy (SBRT), it is important to spare the dose to the spinal cord to prevent radiation myelopathy. Conversely, we should keep sufficient dose to the target volume in the epidural space neighboring the spinal cord because the epidural space is reported as the most common site of recurrence after spine SBRT. Especially for re-irradiation, the dose constraint of the spinal cord is severer than first time radiation and it is more difficult to keep the dose in epidural space high. With the development of computer technology in recent year, a finer calculation grid size (CGS) has come to be used in dose optimization of volumetric modulated arc therapy (VMAT). However, in the spinal SBRT using VMAT, the effect of CGS on the dose to the epidural space has not been elucidated. The aim of this study was to determine the dosimetric effect on the target in the epidural space and spinal cord of the CGS in dose optimization of VMAT. This study was a single center retrospective dosimetric study of 9 patients previously treated for the spinal metastasis at our institution. Planning target volume (PTV) and planning organ at risk volume (PRV) for spinal cord were defined by adding isotropic margins of 2 mm for clinical target volume (CTV) and of 1.5 mm for the spinal cord, respectively. In addition, kidneys, trachea, and esophagus around the PTV were defined as the organ at risks. The prescribed dose was 24 Gy in 2 fractions for the volume obtained by subtracting the PRV of the spinal cord from the PTV. The dose tolerance for PRV of the spinal cord was 12.2 Gy. The following dose parameters were compared between the CGSs of 1 mm and 2 mm in dose optimization of VMAT: D99%, D98%, D95%, D50% for the PTV in the epidural space, and D0.035cc for the PRV of the spinal cord. The dose parameters for the CGS of 2 mm were evaluated after resampling to 1 mm without the change of fluence per delivery angle. Regarding the CGSs of 1 mm and 2 mm, D99%, D98%, D95%, D50% for the PTV in the epidural space were 10.1±1.3 Gy, 10.7±1.3 Gy, 11.7±1.5 Gy, 18.6±4.6 Gy and 9.6±1.1 Gy, 10.1±1.1 Gy, 11.0±1.3 Gy, 17.8±4.8 Gy, respectively. The CGS of 2 mm showed dose decrease of approximately 0.7 Gy for all parameters (p ˂0.01). The values of D0.035cc for the PRV of the spinal cord were 12.1±0.0 Gy and 11.4±0.2 Gy, respectively. The value obtained by the CGS of 2 mm decreased approximately 0.7 Gy (p <0.001) compared to that of 1 mm. In addition, visual evaluation of the difference image obtained by subtracting the dose distribution of the CGS of 1 mm from that of 2 mm showed the decrease of the PTV dose in the epidural space. For the epidural space where the most frequent recurrence occurs in the spinal SBRT, the dosimetric effect of the CGS in dose optimization of VMAT were investigated. In our analysis, the CGS of 2 mm underestimates the dose to the target in the epidural space, while the CGS of 1 mm would be more accurate for calculation of the target in the epidural space.