Rates of Radiosurgical Decompression for High Grade Epidural Spinal Disease Due to Solid Tumor Metastases

Abstract

A recent phase 2 study reported a significant difference in Bilsky grading of treated lesions before and after SBRT. This study reports rates of epidural downgrading following spine SBRT in a larger cohort of patients with high grade epidural disease. Patients with high grade epidural disease from solid spinal metastases treated with SBRT from 2009-2021 were retrospectively reviewed. High grade epidural disease was defined as Bilsky grade 1c-3 for lesions above the conus. A modified Bilsky grading was developed and used for lesions below the conus; grade 0 is spinal bone involvement only; grade 1 is epidural involvement without contact of cauda equina; grade 2 is contact of cauda equina with visible cerebrospinal fluid (CSF); grade 3 is compression of cauda equina with no visible CSF. High grade epidural disease below the conus was defined as a modified Bilsky grade 1-3. Patients who received upfront surgery were included in the current study if post-surgical grading met inclusion criteria. The first post-SBRT MRI was compared to the SBRT simulation imaging to evaluate the extent of epidural downgrading. Epidural disease grading pre- and post-SBRT were compared using the Wilcoxon signed-rank test. A total of 460 vertebral segments from 201 patients were included. The most common histologies were GI (19.3%), breast (16.0%), renal cell carcinoma (13.9%), and prostate (12.6%). Metastases were most often in the T-spine (45.8%), L-spine (34.8%), C-spine (11.8%), and Sacrum (7.6%). Most cases were treated with 3 (37.4%) or 5 (44.5%) fractions with a median prescription dose of 27 Gy or 30 Gy, respectively. 43.3% of cases were treated in the postoperative setting, while 18.5% of cases had prior radiation therapy at the same vertebral level. Median time to first post-SBRT MRI was 3.0 months (IQR 1.8-3.6). Among those with disease above the conus, 98 of the 139 cases (70.5%) had a pre-SBRT Bilsky grade of 2. For those with disease below the conus, 70 of the 99 cases (70.7%) had a pre-SBRT modified Bilsky grade of 1. Epidural downgrading was seen in 32.3% of all cases, 24.3% of postoperative cases, and 25.0% of reirradiation cases. The change in epidural grading was statistically different between pre- and post-SBRT evaluation for all cases (p<0.001), postoperative cases (p<0.001), and reirradiation cases (p = 0.01). We report the largest study to date exploring epidural downgrading following spine SBRT. Although surgery is typically deemed essential for decompression of neural elements, this study suggests that SBRT results in epidural downgrading in as many as one quarter to one third of patients, depending on the clinical scenario.