A Detailed Dosimetric Analysis of Spinal Cord Tolerance in High-Dose Spine Radiosurgery

Objective: Dose-volume tolerance of the spinal cord (SC) in spinal stereotactic radiosurgery (SRS) is difficult to define because radiation myelitis rates are low, and published reports document cases of myelopathy but do not account for the total number of patients treated at given dose-volume combinations who do not have myelitis. This study reports SC toxicity from single-fraction spinal SRS and presents a comprehensive atlas of the incidence of adverse events to examine dose-volume predictors.

Methods and materials: A prospective database of all patients undergoing single-fraction spinal SRS at our institution between 2004 and 2011 was reviewed. SC toxicity was defined by clinical myelitis with accompanying magnetic resonance imaging (MRI) signal changes that were not attributable to tumor progression. Dose-volume histogram (DVH) atlases were created for these endpoints. Rates of adverse events with 95% confidence limits and probabilities that rates of adverse events were <2% and <5% for myelitis were determined as functions of dose and absolute volume.

Results: Information about DVH and myelitis was available for 228 patients treated at 259 sites. The median follow-up time was 14.6 months (range, 0.1-138.3 months). The median prescribed dose to the planning treatment volume was 24 Gy (range, 18-24 Gy). There were 2 cases of radiation myelitis (rate r=0.7%) with accompanying MRI signal changes. Myelitis occurred in 2 patients, with Dmax >13.33 Gy, and minimum doses to the hottest 0.1, 0.2, 0.5, and 1 cc were >10.66, 10.9, and 8 Gy, respectively; however, both myelitis cases occurred below the 34th percentile for Dmax and there were 194 DVHs in total with Dmax >13.33 Gy.

Conclusions: A median SC Dmax of 13.85 Gy is safe and supports that a Dmax limit of 14 Gy carries a low <1% rate of myelopathy. No dose-volume thresholds or relationships between SC dose and myelitis were apparent. This is the largest study examining dosimetric data and radiation-induced myelitis in de novo spine SRS.