Oligometastatic Disease State Is Associated with Improved Local Control in Patients Undergoing 3 or 5 Fraction Spine Stereotactic Body Radiotherapy

Abstract

Stereotactic body radiotherapy (SBRT) is an increasingly utilized treatment modality for patients with spine metastases. The objective of our study was to describe the local control (LC) outcomes with 3 or 5 fraction SBRT to the spine and identify predictors for improved LC. We retrospectively reviewed outcomes of patients treated with SBRT for vertebral body metastases. No patients had prior surgical intervention or spinal cord compression. All patients were treated with 3 (median dose 27 Gy; range 24-30 Gy) or 5 fraction (median dose 35 Gy, range 25-40 Gy) SBRT. The Kaplan-Meier method and SPINO criteria were used to determine LC. Patients were classified as having an oligometastatic disease state (< 5 systemic metastases) or polymetastatic disease state (≥ 5 metastases). Univariate analysis (UVA) was performed using the log-rank test to predict for clinical factors associated with LC. Multivariable analysis (MVA) was conducted using the cox proportional hazards model and included all covariates with p < 0.1 on UVA. Predictors of compression fracture were also evaluated using logistic regression. We included 56 patients with a total of 64 distinct SBRT targets. Median follow-up time was 9 months. Fifteen targets and 49 targets were treated with 3 and 5 fractions, respectively. Thirteen patients (20%) had prior radiotherapy to the SBRT area. Twenty-one patients (33%) had an oligometastatic disease state. One year LC was 82% (100% for 3 fraction, 77% for 5 fraction, p = 0.11). On UVA, polymetastases (1 year LC 76% vs. 100%; p = 0.038), prior radiotherapy to the SBRT target (65% vs. 87%; p = 0.041), and inclusion of greater than 2 vertebral levels in the target (63% vs. 86%; p < 0.01) were associated with worse LC. Male sex (84% vs. 87%; p = 0.091) had a trend toward worse LC at one year. Pseudoprogression occurred in 3 patients (5%) and was not associated with LC on UVA (p = 0.470). On MVA, patients with a polymetastatic state (hazard ratio 21.3, 95% confidence interval 1.21-373.73; p = 0.036) and prior radiotherapy to the SBRT target (HR 7.01, 95% confidence interval 1.46-33.6; p = 0.015) had worse LC. Sex and number of treated vertebral levels were not statistically significant (p = 0.395 and p = 0.233, respectively). Eight patients had grade 1 toxicity, and 1 patient had grade 2 toxicity. Vertebral body compression fractures occurred in 11 (17%) patients. Five fractions was associated with a higher risk of compression fracture (22% versus 0%, p=0.044), while increasing biologically effective dose (p = 0.110) and prior radiotherapy (p = 0.328) were not associated with an increased risk of compression fracture. LC of vertebral body metastases is encouraging with 3 or 5 fraction SBRT. Our study is hypothesis-generating in that patients with an oligometastatic disease state appear to have improved LC following SBRT, suggesting there may be a biologic advantage with local therapy in this group of patients not seen in patients with polymetastatic disease.