Dual-energy X-ray Absorptiometry Does Not Represent Bone Structure in Patients with Osteoporosis: A Comparison of Lumbar Dual-Energy X-Ray Absorptiometry with Vertebral Biopsies
Spine. 2021 Jul 1;46(13):861-866.
Study design: Prospective cross-sectional exploratory study.
Objective: To evaluate the correlation between in vivo lumbar dual-energy x-ray absorptiometry (DXA) and parameters of bone architecture in micro-computed tomography (micro-CT) in patients with osteoporosis.
Summary of background data: DXA is the current diagnostic standard for evaluating osteoporosis. However, there are various concerns regarding its validity, especially in the spine. No study has so far investigated whether in vivo DXA correlates with the actual lumbar bone architecture.
Methods: Lumbar DXA scans were compared with micro-CT analysis of vertebral biopsies in patients with osteoporotic vertebral fractures (fracture group) and those without (control group). Preoperatively, all patients underwent a DXA scan (L1-L4). Intraoperative biopsies from nonfractured vertebrae (preferably L3) were analyzed by micro-CT regarding bone quantity and quality. The groups were compared regarding differences in DXA and micro-CT results. In each group, a correlation analysis was performed between DXA and micro-CT.
Results: The study included 66 patients (33 per group). Preoperative DXA results were worse in the fracture group than the control group (areal bone mineral density [aBMD] 0.95 vs. 1.31, T-score -1.97 vs. 0.92, each P < 0.001). Micro-CT analysis confirmed differences regarding quantitative parameters (bone/total volume: 0.09 vs. 0.12, P < 0.001) and qualitative parameters (connectivity index: 15.73 vs. 26.67, P < 0.001; structure model index: 2.66 vs. 2.27, P < 0.001; trabecular number: 2.11 vs. 2.28, P = 0.014) of bone architecture between both groups. The DXA results did not correlate with micro-CT parameters in the fracture group. In the control group, correlations were found for some parameters (bone/total volume vs. aBMD: r = 0.51, P = 0.005; trabecular number vs. aBMD: r = 0.56, P = 0.001).
Conclusion: These data constitute the first comparison of DXA measurements with microstructural analysis of vertebral biopsies in patients with osteoporosis. Our results indicate that lumbar DXA neither qualitatively nor quantitatively represents microstructural bone architecture and is therefore not a reliable tool for the evaluation of bone quality in the spine.