Changes of Bone Strength Evaluated by CT-based Finite Element Methods in Radiotherapy for Bone Metastases of the Spine in Comparison With Bone Density.

Purpose/Objective(s) Radiotherapy (RT) is performed not only to reduce pain but also to prevent pathological fractures due to bone metastases of the spine. However, it is unclear to what extent radiotherapy increases bone strength. Bone density is sometimes used as a marker for assessment of local response to RT in spinal bone metastases. The purpose of this study was to evaluate the changes of bone strength using CT-based finite element (FE) methods in radiotherapy for metastases at the lumber vertebra in comparison with bone density. Materials/Methods Five patients with metastases to the lumber vertebra who could be followed for more than 6 months were included. Primary lesions were lung cancer in all patients. All patients had osteolytic changes of the bone cortex in the vertebral body. Median radiation dose was 37.8 Gy in 14 fractions. CT images were taken before and at 1, 3, 6, 9 and 12 months after irradiation. Bone density was measured in Hounsfield units (HU) in each CT scan. The regions of interest were set to include the whole vertebral body of the metastatic spine and that of a non-irradiated unaffected neighboring spine used for reference. The 3-dimensional FE models of the same vertebrae were constructed from each CT image. The models consisted of 1- or 2-mm linear tetrahedral elements and the cortical shell with a thickness of 0.1 mm. A uniaxial compressive load with uniform distribution was applied on the upper surface of the vertebral body with the lower surface completely restrained. The predicted fracture load was defined to be present when at least one element or cortical shell failed. The ratios of the values on the follow-up CT to the values on the pre-irradiation CT were calculated, and single regression equations were determined for these ratios. Results At one month after irradiation, the bone density decreased in all except one case of the affected vertebral bodies, and the predicted fracture load decreased in all except one case of the unaffected vertebral bodies. In the affected vertebral bodies, the slopes of the single regression equations of bone density were positive in 4 out of 5 cases. The slopes of those of predicted fracture loads were negative in 3 out of that 4 cases. In the unaffected vertebral bodies, the slopes of the single regression equations of bone density were positive in 3 out of 5 cases. The slopes of predicted fracture loads were positive in those 3 cases. Conclusion The bone strength might not improve in one month after irradiation. The predicted fracture load did not necessarily increase even if the bone density increased in the irradiated affected vertebral bodies. Bone density alone might not be sufficient as an indicator of bone strength. The FE method could evaluate the changes in bone strength in radiotherapy for bone metastases based on predicted fracture loads.