Bone density and fracture risk following SBRT for non-spine bone metastases

Purpose/methods: This retrospective study evaluated local recurrence (LR) and fracture risk in non-spine bone metastases treated with SBRT. Results: 181 lesions in 116 patients are reported. The median dose was 27 Gy (range 15-40) in 3 fractions (range 1-6). The cumulative incidence of LR was 2.8%, 7.2% and 12.5% at 6 mo, 1 yr and 2 yrs. Fractures occurred in 11 lesions (6%). Radioresistant histology and increasing PTV predicted for LR on univariate analysis, while rib location was associated with control. Increasing PTV remained a significant predictor for LR on multivariate analysis. Univariate predictors of fracture risk included female gender, lytic lesions and poorer KPS. Average CT-approximated L1 trabecular attenuation in patients with fracture was significantly lower than in patients without fracture (112.2 vs. 142.6 Hounsfield units). Conclusion: In the largest series to date, we report excellent local control for SBRT to non-spine bone metastases and a novel relationship between CT-based bone quality assessment and fracture risk.