The relationship between Computed Tomography and DXA results: A potential bias in Bone Mineral Density assessment
Andrey Bokov, Sergey Mlyavykh, Alexander Aleynik, Marina Rasteryaeva, Tatiana Malysheva
Abstract
Background
It has been reported that the results of the lumbar spine bone mineral density assessment can be strongly biased by degenerative changes. However, the reported data remains controversial and a potential bias has not been assessed.
Aims
To evaluate the relationships between the results of DXA and CT with the assessment of potential bias related to the influence of different structures.
Methods
This is a cross-sectional study and 25 patients were enrolled. Using DXA scan, Bone Mineral density (BMD) (g/cm2) was calculated from 100 vertebra from the lumbar spine. Out of all the CT measurements, a mean radiodensity in HU (Hounsfield Units) for cancellous bone and total vertebra body, a mean square and radiodensity of vertebra pedicles and facet were calculated.
Results
Linear regression analysis demonstrated a strong correlation between BMD measured by DXA and CT data. Multiple correlation coefficient of model accounts for 0.8093, r2=0.6550, p < 0.0001. Parameters that have significant relationships with the results of DXA were: a product of facet joints radiodensity and mean square on axial images (B= 0.000003379, p < 0.0001) and total vertebral body radiodensity B=0.0016395253, p=0.0201. Beta coefficients for those variables accounted for 0.6729 and 0.3037 respectively.
Conclusion
The results of the bone mineral density assessment of the lumbar spine using DXA, can be strongly influenced by facet joints condition, especially in cases of degenerative changes. The results of BMD, provided by DXA, are partly relevant to vertebral body bone quality assessment and irrelevant to the characteristics of bone in pedicles. This means that the prognostic value of the DXA measurement results might be limited in relation to predicting low energy vertebra fracture and implant stability.
Full Text:
It has been reported that the results of the lumbar spine bone mineral density assessment can be strongly biased by degenerative changes. However, the reported data remains controversial and a potential bias has not been assessed.
Aims
To evaluate the relationships between the results of DXA and CT with the assessment of potential bias related to the influence of different structures.
Methods
This is a cross-sectional study and 25 patients were enrolled. Using DXA scan, Bone Mineral density (BMD) (g/cm2) was calculated from 100 vertebra from the lumbar spine. Out of all the CT measurements, a mean radiodensity in HU (Hounsfield Units) for cancellous bone and total vertebra body, a mean square and radiodensity of vertebra pedicles and facet were calculated.
Results
Linear regression analysis demonstrated a strong correlation between BMD measured by DXA and CT data. Multiple correlation coefficient of model accounts for 0.8093, r2=0.6550, p < 0.0001. Parameters that have significant relationships with the results of DXA were: a product of facet joints radiodensity and mean square on axial images (B= 0.000003379, p < 0.0001) and total vertebral body radiodensity B=0.0016395253, p=0.0201. Beta coefficients for those variables accounted for 0.6729 and 0.3037 respectively.
Conclusion
The results of the bone mineral density assessment of the lumbar spine using DXA, can be strongly influenced by facet joints condition, especially in cases of degenerative changes. The results of BMD, provided by DXA, are partly relevant to vertebral body bone quality assessment and irrelevant to the characteristics of bone in pedicles. This means that the prognostic value of the DXA measurement results might be limited in relation to predicting low energy vertebra fracture and implant stability.
PDF