Original Article
Novel Assessment of Subregional Bone Mineral Density Using DXA and pQCT and Subregional Microarchitecture Using Micro-CT in Whole Human Vertebrae: Applications, Methods, and Correspondence Between Technologies

https://doi.org/10.1016/j.jocd.2010.01.120Get rights and content

Abstract

In the clinical environment dual-energy X-ray absorptiometry (DXA) is the current tool of first choice for assessing and monitoring skeletal integrity. A major drawback of standard DXA is that the bone mineral density (BMD) data cannot be used with certainty to predict who will sustain a vertebral fracture. However, measurement of BMD within vertebral subregions, instead of relying on a gross estimate of vertebral BMD, may improve diagnostic sensitivity. The aim of this article was to describe a validation study for subregional BMD measurement using lateral-projection DXA and to present preliminary data. Concurrent validity of measuring subregional BMD with DXA was established against measures of volumetric subregional BMD from peripheral quantitative computed tomography (pQCT) and subregional bone volume fraction from μCT at the L2 vertebral body in 8 cadaver spine specimens. The novel approaches for measuring subregional parameters with each imaging modality are described. Significant differences in bone parameters between vertebral subregions were observed for each imaging modality (p < 0.05). Correspondence ranged from R2 = 0.01–0.79 and R2 = 0.06–0.80 between “DXA vs. pQCT” and “DXA vs. micro-CT,” respectively. For both imaging modalities, correspondence with DXA was high for centrally and anteriorly positioned subregions. These data provide a basis for larger studies to examine the biological significance of heterogeneity in vertebral BMD.

Introduction

The strong relationship between bone mineral density (BMD) and vertebral bone strength underlies the rationale for the use of bone densitometry in assessing and monitoring skeletal integrity and making clinical decisions concerning vertebral fragility (1). Although considered somewhat crude relative to other technologies, dual-energy X-ray absorptiometry (DXA) remains the clinical tool of first choice for this purpose owing to its high precision, accuracy, efficiency, low radiation dose, accessible measurement sites, and low cost relative to other densitometry technologies 2, 3, 4. The strong association between BMD and bone strength (1) explains vertebral BMD as a good predictor of vertebral fracture risk, analogous to blood pressure for stroke, yet BMD cannot be used reliably to determine who will sustain a fracture (5). This creates some uncertainty for clinicians in individual patient care.

Several studies have demonstrated marked differences in the prevalence rate of vertebral fractures among individuals with a comparable BMD as measured by DXA 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18. There are several possible explanations for the poor predictive value of standard DXA parameters when used in isolation, including the influence of clinical risk factors other than areal BMD, inability to measure bone quality, inadequate measurement specificity, and the generally stochastic nature of vertebral fractures (19). Among these, measurement specificity is the only factor that may be improved to optimize the predictive value of DXA, and this remains a focus for our research group.

Both ex vivo and in vivo research has established that the distribution of bone varies throughout the vertebral body 19, 20, yet standard DXA estimation of whole vertebral BMD cannot capture the heterogeneous distribution of intra-vertebral (subregional) BMD. The distribution pattern of bone density is known to influence the bone strength characteristics of the vertebra 21, 22. It follows therefore, that the distribution of intravertebral BMD may be a defining characteristic between individuals with and without osteoporotic vertebral fractures.

We developed a method to measure subregional vertebral BMD using lateral-projection DXA scans, acquired using a Hologic QDR4500A densitometer (23). Our pilot data demonstrate that measurement of areal BMD within vertebral subregions using lateral-projection DXA can better differentiate between individuals with and without osteoporotic vertebral fractures compared with standard AP-projection vertebral DXA parameters (24). Ultimately, a subregional BMD approach could be applied to standard clinical bone densitometry using DXA to potentially provide more reliable and extensive information concerning vertebral fragility. Although compelling, results acquired in the pilot study need replication and verification in a larger clinical study and the DXA protocol requires further validation.

Validation of the subregional BMD approach using DXA can be established against other densitometry modalities, such as peripheral quantitative computed tomography (pQCT) and micro computed tomography (μCT), thereby establishing concurrent validity.

Vertebral bone has trabecular bone structures with thicknesses as small as100 μm (25). Thus, imaging methods with high resolution are essential for their accurate description. The first studies characterizing vertebral microarchitecture with μCT were based on excised bone core samples from which parameters such as bone volume fraction, trabecular thickness, trabecular separation and trabecular number were derived (25). It is now possible to scan the whole vertebral body, rather than an excised core 26, 27, 28, providing data for finite element models (27). Although these studies demonstrate the capacity to examine subregional microarchitecture in the whole vertebral body using μCT, correspondence with subregional analysis of the same vertebral body obtained with clinical instruments, such as DXA, has not yet been examined. In this manuscript we describe the novel approaches used with each modality (DXA, pQCT, and μCT) to derive a vertebral subregional densitometric or micro-architectural value. The aim of this article is to investigate, within the same vertebral bodies, (1) the correspondence in subregional BMD measurements obtained by DXA with those obtained by pQCT, that is, the clinical imaging device for 3D spine measurements in-vivo having higher spatial resolution compared with DXA and (2) with those obtained by μCT, the non-destructive 3D imaging device having the highest resolution.

Section snippets

Study Design

This multi-centre project is currently run through three Australian centres: Curtin University of Technology, Western Australia; SA Pathology, South Australia; and University of Melbourne (Royal Melbourne Hospital), Victoria. Correspondence between subregional measures of vertebral areal BMD, volumetric BMD and microarchitecture are examined using DXA, pQCT, and μCT, respectively, using cadaveric material. Initially, DXA and pQCT scanning are performed in Melbourne, followed by μCT scanning in

Dual Energy X-ray Absorptiometry (DXA)

All scanning was performed using a Hologic (Hologic Inc., Waltham, MA; USA) QDR4500A fan beam densitometer, running operating software version 9.10D. The 12-month precision of the densitometer for the Hologic spine phantom was 0.39% for BMD and 0.58% for BMC. Spines samples were placed supine in a water bath (270 × 180 × 150 mm) of tap water to a depth of 18 cm to simulate soft tissue composition. Specimens were wrapped in water-tight plastic wrap free of air and were secured to the base of the water

DXA Parameters: Areal BMD (mg/cm3), ap.vBMD1 (mg/cm3), ap.vBMD2 (mg/cm3)

DXA-derived data included areal BMD in each subregion (ROI 1–7) and the standard Hologic DXA parameters (total vertebral BMD for the PA and lateral projections, and mid-lateral BMD) for L2. Subregional areal BMD was transformed to apparent volumetric BMD (ap.vBMD) on the assumption that each ROI represented a cylinder. Throughout the remainder of the manuscript, “transformed DXA data” refers to ap.vBMD calculated from areal BMD and subregional geometry. It was not possible to calculate a

Statistical Analysis

Precision of the pQCT method was expressed using a percent coefficient of variation (%CV). A repeated measures ANOVA was performed to evaluate differences in quantitative bone parameters between subregions for each imaging modality, in keeping with earlier work 24, 30. Although there were seven regions of interest, the within subject factor (ROI) was set a priori at k = 4 to ensure that overlapping subregions were not compared post hoc. The first ANOVA model included ROIs 1–4 and the second ANOVA

pQCT Precision

Moderate to high intra-rater precision was observed for the entire pQCT protocol (scan rescan), reflected by a mean (range) %CV of 3.08% (1.02–6.61%) (Table 1). Similarly, moderate to high precision was observed for the Matlab analysis process with a mean (range) %CV of 1.65% (0.34–3.36%) for rater 1, 1.90% (0.56–4.92%) for rater 2, and 3.25% (0.82–9.12%) for inter-rater precision (Table 1).

Subregional Densitometric Characteristics

The mean (SD) areal BMD of the whole L2 vertebrae was 833.625 (144.790) mg/cm2 in the DXA-PA projection

Discussion

We report on the development and application of quantifying subregional bone properties in lumbar vertebrae using DXA, pQCT, and μCT. The data described here support the contention that measurement of subregional vertebral bone properties is feasible using lateral projection DXA by virtue of the correspondence between DXA-derived subregional parameters with pQCT and μCT-derived parameters. Moreover, the patterns of correspondence between DXA and the two criterion modalities (pQCT and μCT) are

Acknowledgments

Funding for these studies was provided by the National Health and Medical Research Council (NHMRC) of Australia, Scoliosis Research Society (USA), and Arthritis Australia. Dr Andrew Briggs is supported by a fellowship awarded by the NHMRC. In kind support was provided by the University of Melbourne Department of Medicine (Royal Melbourne Hospital), and SA Pathology.

References (48)

  • B. Kaymakci et al.

    Precise accurate mineral measurements of excised sheep bones using X-Ray densitometry

    Bone Miner

    (1994)
  • R. Muller et al.

    Morphometric analysis of human bone biopsies: A quantitative structural comparison of histological sections and micro-computed tomography

    Bone

    (1998)
  • D.L. Kopperdahl et al.

    Quantitative computed tomography estimates of the mechanical properties of human vertebral trabecular bone

    J Orthop Res

    (2002)
  • E.N. Ebbesen et al.

    Vertebral bone density evaluated by DXA and QCT in vitro

    Bone

    (1998)
  • M. Kleerekoper et al.

    Which bone density measurement?

    J Bone Miner Res

    (1997)
  • J.A. Kanis et al.

    European guidance for the diagnosis and management of osteoporosis in postmenopausal women

    Osteoporos Int

    (2008)
  • D. Marshall et al.

    Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures

    BMJ

    (1996)
  • M. Bhambhani et al.

    Differential involvement of the dorsal and lumbar spine in osteoporosis

    Ann Rheum Dis

    (1992)
  • R. Eastell et al.

    Classification of vertebral fractures

    J Bone Miner Res

    (1991)
  • C. Grey et al.

    Vertebral deformity in the thoracic spine in post-menopausal women: value of lumbar spine bone density

    Brit J Radiol

    (1996)
  • M. Jergas et al.

    Which vertebrae should be assessed using lateral dual-energy X-ray absorptiometry of the lumbar spine

    Osteoporos Int

    (1995)
  • E. Legrand et al.

    Trabecular bone microarchitecture, bone mineral density, and vertebral fractures in male osteoporosis

    J Bone Miner Res

    (2000)
  • D. Mitra et al.

    The prevalence of vertebral fractures in mild ankylosing spondylitis and their relationship to bone mineral density

    Rheumatology

    (2000)
  • N.F.A. Peel et al.

    Risk of vertebral fracture and relationship to bone mineral density in steroid-treated rheumatoid arthritis

    Ann Rheum Dis

    (1995)
  • Cited by (23)

    • Subregional areal bone mineral density (aBMD) is a better predictor of heterogeneity in trabecular microstructure of vertebrae in young and aged women than subregional trabecular bone score (TBS)

      2019, Bone
      Citation Excerpt :

      This might explain the lack of a significant R2 for the anterior region of the aged group with BV/TV as dependent variable and aBMD as predictor. This effect could be accounted for by scaling the anterior values to a theoretical similar volume [14]. In this study, we refrained from correcting the values since we aimed to investigate if easily applied changes to the ROI could already yield meaningful results.

    • Impact of bone quality on the performance of integrated fixation cage screws

      2018, Spine Journal
      Citation Excerpt :

      Furthermore, the heterogeneous nature of the vertebrae cannot be assessed with a single BMD value as local bone quality varies spatially, particularly within trabecular bone because of its complex three-dimensional (3D) microstructure. In fact, a research study confirmed this by performing DEXA measurements on whole vertebra as well as specific regions of vertebra and reported that the BMD of the whole vertebra is significantly different from the BMD within subregions of the vertebra [8]. Cadaver biomechanical studies have also used BMD measurements to classify vertebral bone quality as normal, osteopenic, or osteoporotic.

    • Functional Morphology and Medical Imaging

      2013, Research Methods in Human Skeletal Biology
    • Failure strength of human vertebrae: Prediction using bone mineral density measured by DXA and bone volume by micro-CT

      2012, Bone
      Citation Excerpt :

      In order to then study mechanical properties, the cores were subjected to mechanical testing [22–24]. It is now possible to scan the whole vertebral body, rather than an excised core [25–29]. This provides high-resolution structural data of the whole vertebra in 3D, which can be used in combination with finite element models [26], or with mechanical test data of the whole vertebral body [20,30,31], to examine the resistance to fracture of the vertebra from which vertebral fragility can be inferred.

    View all citing articles on Scopus
    View full text