Volumetric Dosimetry Analysis of Vertebral Compression Fracture Risk Following Spine Stereotactic Body Radiation Therapy

Abstract

Spine stereotactic body radiotherapy (SBRT) has offered improved tumor response and palliation in treating spinal metastases. However, risk of vertebral compression fracture (VCF) following spine SBRT has been reported in 15-20% of patients. Dose per fraction has been found to be an important risk factor for the development of VCF. However, vertebral body dosimetry and its effect on risk for VCF is unknown. We performed detailed volumetric analysis of vertebral body dosimetry and its association with VCF. We utilized our institutional database to identify eligible patients for this study. We included patients who underwent single fraction spine SBRT who had at least a one month follow-up. Exclusion criteria included patients with sacral involvement, prior in-field spine SBRT treatment as well hardware at the site of spine SBRT. We identified 44 consecutive treatments that met criteria for inclusion. The primary outcome was de novo or progression of VCF. Each treated vertebral body was contoured using the simulation CT. Dosimetric endpoints included mean dose, maximum dose, minimum dose, percentage of vertebral body receiving >18 Gy, > 20 Gy, and >22 Gy. Dosimetric endpoints were correlated with the occurrence of VCF using the Wilcoxon rank-sum test. Eighty-four spinal segments (10 cervical, 49 thoracic, 25 lumbar) from 44 treatments were included in the analysis. Ten segments (12%) had either a new VCF or progression of pre-existing VCF. Median treatment dose was 16 Gy (range, 14-18 Gy). Median follow-up was 13 months (range, 1-23). The average mean vertebral body dose for spinal segments that developed VCF was 19.0 Gy vs. 17.5 Gy for spinal segments without VCF (p=0.21). The average maximum/minimum vertebral body dose was 21.6 Gy/12.0 Gy for spinal segments with VCF and 20.6 Gy/10.6 Gy for spinal segments without VCF (p=0.33). The average proportion of vertebral body receiving >18 Gy, >20 Gy and >22 Gy for spinal segments with VCF/without VCF was 70.5%/56.7% (p=0.26), 29.7%/22.5% (p=0.62), 8.7%/6.7% (p=0.27), respectively. None of the patients in this series required kyphoplasty and there were no grade ≥2 toxicities. This study is suggestive that spinal segments that developed VCF had higher maximum and minimum dose as well as a greater proportion of the vertebral body receiving >18 Gy in a single fraction. Although not significant in our small series, our findings corroborate previous studies that show that doses >18 Gy per fraction to the vertebral body increase the risk for developing a VCF. Our data further adds that the proportion of vertebral body receiving >18 Gy may also be an important predictor for the development of VCF. Given small patient numbers, additional data is required to further establish this relationship.