Reirradiation Stereotactic Body Radiation Therapy (SBRT) for Spinal Metastases: A Multi-Institutional Outcome Analysis

We report the largest multi-institutional pooled analysis specific to imaging-based local control in patients previously radiated with conventional external beam radiation (cEBRT) and salvaged with re-irradiation spine stereotactic body radiotherapy (SBRT). A retrospective review of 247 patients from seven different institutions was performed. Overall survival (OS) was based on per patient treated with SBRT, while local control was based on the spinal target volume treated, using the Kaplan Meier method. Local control was defined as imaging based progression within the SBRT target volume. Equivalent dose in 2 Gy fractions (EQD2) were calculated for the cEBRT and SBRT course and summed to yield the cumulative dose with an α/ß of 10 for tumor and 2 for both spinal cord and cauda equina. The median total dose and number of fractions of the initial cEBRT was 30 Gy and 10, respectively. The median SBRT total dose and number of fractions was 18 Gy and 1, respectively. Sixty percent were treated with single fraction SBRT (median, 18 Gy and EQD2/10 = 36.8 Gy), and 40% with multiple fraction SBRT (median, 24 Gy in 3 fractions and EQD2/10 = 36 Gy). The median time interval from cEBRT to re-irradiation SBRT was 13.5 months and median patient follow up was 8.1 months. OS at 6 and 12 months was 64% and 48%, respectively. Thirteen percent of patients suffered a local failure, and the local control rate at 6 months and 12 months was 93% and 83%, respectively. Multivariate analysis identified (Karnofsky performance Status) KPS <70 as a significant prognostic factor for worse OS, and single fractionated SBRT as a significant predictive factor for better local control. There were no cases of radiation myelopathy, and the vertebral compression fracture rate was 4.5%. Re-irradiation spine SBRT is effective in yielding imaging-based local control with a clinically acceptable safety profile. A randomized trial would be required to determine the optimal fractionation.