Abstract

OBJECTIVE Variation exists in approaches to delivery of spine stereotactic radiosurgery (SSRS). Here, the authors describe outcomes following single-fraction SSRS performed using a simultaneous integrated boost for the treatment of prostate cancer spine metastases. METHODS Health records of patients with prostate cancer spine metastases treated with single-fraction SSRS at the authors’ institution were reviewed. Treatment was uniform, with 16 Gy to the clinical tumor volume and 18 Gy to the gross tumor volume. The primary endpoint was local recurrence, with secondary endpoints including vertebral fracture and overall survival. Univariate and multivariate competing risk regression models made using the Fine and Gray method were used to identify factors predictive of local recurrence, considering death to be a competing event for local recurrence. RESULTS A total of 87 targets involving 108 vertebrae in 68 patients were included, with a median follow-up of 22.5 months per treated target. The 1-, 2-, and 4-year cumulative incidence rates of local failure for all targets were 4.6%, 8.4%, and 19%, respectively. The presence of epidural disease (subdistribution hazard ratio [sHR] 5.43, p = 0.04) and SSRS as reirradiation (sHR 16.5, p = 0.02) emerged as significant predictors of local failure in a multivariate model. Hormone sensitivity did not predict local control. Vertebral fracture incidence rates leading to symptoms or requiring intervention at 1, 2, and 4 years were 1.1%, 3.7%, and 8.4%, respectively. In an exploratory analysis of patterns of failure, 3 (25%) failures occurred in the epidural space and only 1 (8%) occurred clearly in the clinical tumor volume. There were several lesions for which the precise location of failure with regard to target volumes was unclear. CONCLUSIONS High rates of local control were observed, particularly for radiotherapy-naïve lesions without epidural disease. Hormone sensitivity was not predictive of local control in this cohort and fracture risk was low. Further research is needed to better predict which patients are at high risk of recurrence and who might benefit from treatment escalation.