Category: Central DXA: (DXA, TBS)

29 - TBS association with biomechanical properties of human vertebrae, ex-vivo

Objective: The clinical utility of Trabecular Bone Score (TBS) to evaluate the risk for osteoporotic fracture has been widely recognized by the scientific community. To date only one study presented data on the relation between TBS and bone mechanical properties, with a relatively small sample size. The aim of this study was to evaluate the relationship between TBS, BMD and mechanical properties of on a larger sample of human vertebra ex vivo.


Method:
52 vertebrae (L1 to L4) were harvested from 13 post-mortem human subjects (mean age 74.9 ± 8.2 yrs). The isolated vertebrae (including posterior arches) were scanned with a DXA device (Prodigy, GE-Lunar) using water to simulate the soft tissues around the bone. Bone texture was evaluated using TBS (TBS iNsight software, Medimaps). BMD and TBS were evaluated over the same ROI, the largest rectangle included within the vertebral body. Biomechanical testing was performed on each whole vertebra to measure the failure load and the stiffness. The quasi-static compression tests were realized with a traction-compression device (INSTRON 55000, INSTRON). Relationships between TBS and mechanical properties were evaluated using a Pearson correlation test. A multivariate analysis, that includes both TBS and BMD, has also been performed.


Results:
Over the 13 subjects, 2 were excluded due to their abnormal L1-L4 BMD when looking at the global distribution (0.409 and 1.401 g/cm²). No outliers were detected for L1-L4 TBS. Among the 44 remaining vertebrae, 9 were excluded due to presence of severe arthrosis (presence of osteophytes) or established vertebral fracture. Mean BMD, TBS, failure load and stiffness were 0.854±0.161 g/cm², 1.561±0.098, 2351±1064 N, 4170±1260 N.mm-1 respectively. Moderate but significant correlations (figure 1) were observed with failure load for BMD (r=0.65; p < 0.0001) and for TBS (r=0.63; p < 0.0001). Stronger significant correlations (figure 1) were observed with stiffness for BMD (r=0.65; p < 0.0001) and for TBS (r=0.73; p < 0.0001). In the multivariate analysis, both BMD and TBS remained significantly associated with the failure load (r=0.50, p=0.002 and r=0.47, p=0.005 respectively). The combined model explained 52.3% (r²-adjusted) of the failure load. The association with the stiffness also remained significant for BMD and TBS (r=0.49, p=0.004 and r=0.62, p=0.0001 respectively). The combined model explained 64.6% (r²-adjusted) of the stiffness.


Conclusion:
The present study confirmed on a larger sample the correlations between TBS and biomechanical properties of the lumbar spine and those results were independent of spine BMD. Both BMD and TBS characterize a significant part of the vertebral bone strength which may explain the ability of TBS in conjunction with BMD to improve the risk of fracture in clinical practice.


Didier Hans

Professor
Center of Bone diseases, Bone & Joint Department, Lausanne University Hospital, Lausanne, Switzerland
Lausanne, Vaud, Switzerland

Christophe Lelong

CTO
Medimaps SASU, Merignac, France
Merignac, Aquitaine, France

Franck Michelet

Medical imaging scientist
Medimaps SASU, Merignac, France
Merignac, Aquitaine, France

Doris Tran

Clinical Scientist
Medimaps SASU, Merignac, France
Merignac, Aquitaine, France