Elsevier

Clinical Biomechanics

Volume 17, Issue 2, February 2002, Pages 81-88
Clinical Biomechanics

Clinical Biomechanics Award 2001
High-resolution MRI and micro-FE for the evaluation of changes in bone mechanical properties during longitudinal clinical trials: application to calcaneal bone in postmenopausal women after one year of idoxifene treatment

https://doi.org/10.1016/S0268-0033(01)00110-3Get rights and content

Abstract

Objective. To investigate whether recently developed in vivo high-resolution magnetic resonance-imaging and micro-finite element techniques can monitor changes in bone mechanical properties during long-term clinical trials aiming at evaluating the efficacy of new drugs for the treatment of osteoporosis.

Design. Comparison of baseline and follow-up mechanical parameters calculated using micro-finite element analysis of the calcaneus for subjects participating in a study investigating the effect of idoxifene.

Background. Contemporary measurements for the evaluation of bone mechanical properties, based on dual-energy X-ray absorptiometry measurements, are not very accurate and require large trial populations.

Methods. A total of 56 postmenopausal subjects received either a placebo, 5 mg or 10 mg per day of idoxifene. Magnetic resonance-images of the calcaneus were made at baseline and after one year. Mechanical parameters of a trabecular volume of interest in the calcaneus were calculated using micro-finite element analysis.

Results. Although there were no significant differences between the mean changes in the treated groups and the placebo group, there were significant changes from baseline within groups after one year of treatment. Significant changes, however, were found only for mechanical parameters and only in the treated groups.

Conclusions. The present study is the first demonstration that longitudinal changes in bone mechanical properties due to trabecular micro-architectural changes may be quantified in long-term clinical studies. Since significant changes in mechanical parameters were obtained for the treated groups whereas no significant change in bone mass was found we conclude that the application of these techniques may increase the clinical significance of these trials.
Relevance

A precise diagnosis of in vivo bone mechanical properties that accounts for (changes in) trabecular bone architecture is of particular importance for longitudinal clinical trials aiming at evaluating the efficacy of new drugs since it can lead to clinically relevant results from shorter follow-up intervals and may enable a reduction of the number of patients involved in the trial.

Introduction

The mechanical integrity of the skeleton can be seriously affected by bone diseases, of which osteoporosis is the most common one. An accurate assessment of the mechanical integrity of bones in vivo is essential for the diagnosis of such diseases. Most quantitative diagnostic methods for bone mechanical integrity used nowadays are based on estimates of bone density measured by dual-energy X-ray absorptiometry (DEXA); this is the only material property that can now be clinically measured non-invasively. Although such measurements quantify the amount of bone, they do not sufficiently reflect (changes in) bone mechanical integrity [1], [2], in particular since they do not account for the mechanical consequences of bone micro-architecture degradation typically seen with osteoporosis. Delmas [3] stated that the prediction of osteoporotic fractures from bone mass measurements is “… as good as that of using blood pressure to predict stroke”. Hence, although bone mass correlates significantly with fracture risk, its predictive value is poor, and cannot be used as a reliable diagnostic tool by itself for the purposes implied above. From a mechanical point of view this result is not surprising. The mechanical properties of trabecular bone (e.g., stiffness and strength) depend not only on the amount of bone mass (as quantified by DEXA) but also on the spatial arrangement of the trabeculae. Consequently, changes in bone mechanical properties due to changes in its architecture, in particular those that do not go together with changes in bone mass, might not be detected with these diagnostic methods.

A precise diagnosis of bone mechanical properties that accounts for changes in bone architecture is of particular importance during longitudinal clinical trials aiming at evaluating the efficacy of new drugs for the treatment of osteoporosis. Such drugs could induce subtle changes in bone architecture that can lead to significant changes in its mechanical properties without significant changes in bone mass. An evaluation of the mechanical effects of these structural changes potentially could lead to the detection of significant changes in bone mechanical properties before significant changes in bone mass can be detected or in response to agents in which the relationship between changes in bone mechanical properties and bone mineral density (BMD) was different from currently approved therapies. This, in turn, could lead to clinically relevant results from shorter follow-up intervals. Accounting for changes in bone architecture may also enable a reduction of the number of patients involved in the trial because information on bone mechanical properties may be obtained from every participant instead of the small proportion who may sustain a fracture.

An accurate determination of bone mechanical properties that fully accounts for the structural arrangement of trabeculae is possible with a recently developed micro-finite element (FE) technique [4], [5]. With this technique, the morphology of the trabecular bone is measured by a large number of sequential cross-sectional images. These digitized images are stacked in a computer in which the 3-D trabecular structure is reconstructed as a rectangular voxel grid, with voxels representing bone tissue or marrow. By converting voxels representing bone tissue to equally shaped brick elements in a micro-FE model a micro-FE model is generated that can represent the trabecular structure in great detail. By simulating compression tests, this technique enables a complete evaluation of trabecular bone anisotropic elastic properties [6], [7], tissue loading [8] and multi-axial strength [9]. The micro-FE approach, however, was developed for use in combination with images obtained from in vitro imaging techniques, such as micro-computed tomography (CT) and serial sectioning, that enable a resolution of 50 μm or better.

Results of several recent studies demonstrated that the micro-FE approach can provide an adequate evaluation of structure-related bone mechanical properties also for bone in vivo. Presently, two imaging techniques exist that can be applied to bone in vivo and provide a resolution adequate to visualize individual trabeculae. With the first of these techniques, a peripheral quantitative computed tomography (pQCT) device is used with an isotropic spatial resolution of 165 μm [10], [11]. With the second, a whole-body magnetic resonance (MR) scanner is used, providing a resolution of approximately 150–200 μm in plane at a plane thickness of 300–500 μm [12], [13]. The resolution of these imaging techniques thus is considerably less than that of the in vitro imaging techniques. In addition, these imaging techniques cannot provide images with the same level of signal-to-noise-ratio and reproducibility. Nevertheless, when applying proper image processing techniques, adequate 3-D reconstructions of the trabecular architecture can be generated from images at this resolution [14], [15]. In a number of earlier validation studies, the feasibility of this micro-FE approach in combination with such pQCT and MR images was investigated [16], [17], [18]. In these studies it was found that micro-FE analyses can provide accurate results, in particular when using a special `mass preservative' segmentation technique [16], but that correction factors might be needed to predict accurate values. In a recent study, it was found that micro-FE analyses based on pQCT images of whole bones in situ can provide a prediction of bone strength that is much better than predictions based on DEXA scans [19].

In the present study, we aim to find a rigorous answer to the question whether these recently developed high-resolution MR-imaging and micro-FE techniques can monitor changes in bone mechanical properties during long-term clinical trials aiming at evaluating the efficacy of new drugs for the treatment of osteoporosis. For this study, high-resolution MR-images of a large number of patients were available from a large study investigating the effect of idoxifene, a selective estrogen receptor modulator, on change in bone mineral density in a cohort of postmenopausal women. In the present study we use MR-images of the calcaneus that were available at baseline and after one year of the treatment. A specific purpose of this study was to investigate whether the use of micro-FE can reveal changes in bone mechanical properties over the one-year interval.

This study is part of a larger project aimed at establishing the feasibility and potential of MR imaging and micro-FE analyses for bone in vivo in longitudinal trials. In an earlier study, we investigated the reproducibility of the MR-imaging procedure and the parameters obtained from these [20]. In that study it was found that a reproducibility of 2–4% can be expected for 2-D structural parameters and from 4–9% for 3-D parameters (including micro-FE calculated elastic moduli). In another earlier study [21], we used MR images from the same idoxifene trial to quantify micro-FE calculated moduli, morphological parameters DEXA measurements, biochemical markers and their interrelationships for subsets of the population at baseline. The present study is the first to investigate changes in bone mechanical properties after one year of treatment.

Section snippets

Patients and subjects

A subset of 56 subjects participating in a large randomized double blind multicenter study investigating the effect of idoxifene on lumbar spine BMD were recruited for MR imaging of the calcaneus at two institutions, namely University of California, San Francisco (UCSF) and the University of Washington (UW) Seattle, USA, in accordance with the regulations of the Committee of Human Research at the respective institutions. Subjects did not have a history of substance abuse (including alcohol)

Descriptive statistics

An overview of the mean and standard deviations for the investigated parameters at baseline and after one-year follow-up is given in Table 1 for the placebo, the 5 mg and the 10 mg group. The percentage change from baseline for the parameters is represented in a box plot (Fig. 3). This plot revealed that the median value of changes in BV/TV and elastic constants are close to zero for the placebo group and increase with increasing idoxifene dose. The plots also show a number of outliers,

Discussion

In the present study, we asked the question whether recently developed high-resolution MR-imaging and micro-FE techniques can monitor changes in bone mechanical properties during long-term clinical trials. Based on the results obtained here we conclude that, indeed, the application of these techniques can be justified. Significant changes in mechanical parameters were obtained for the treated groups whereas no significant change in bone mass was found. Consequently, the application of these

Acknowledgements

The following persons made significant contributions to part of this work: Gabi von Ingersleben, Steve Harris, Harry K. Genant, Charles Chestnut, and Patrick Garnero. This study was supported by SmithKline Beecham Pharmaceuticals, Collegeville, PA.

References (27)

  • Rüegsegger P. Bone density measurements. In: Bröll and Dambacher, editors. Osteoporosis: A guide to diagnosis and...
  • P. Delmas

    Editorial. Proceedings of the third international conference on osteoporosis, Paris, October 1995

    Bone

    (1996)
  • A.J. Ladd et al.

    Finite-element modeling of trabecular bone: comparison with mechanical testing and determination of tissue modulus

    J Orthop Res

    (1998)
  • Cited by (85)

    • MicroCT-based finite element models as a tool for virtual testing of cortical bone

      2017, Medical Engineering and Physics
      Citation Excerpt :

      However, recent advances in medical image resolution and computer processor speed provide alternatives, reducing the need for invasive mechanical testing and replacing it with computational biomechanics to simulate in-vivo bone-loading conditions. This advancement provides the opportunity to assess bone strength through non-invasive methods, thus reducing the cost, time, and number of experiments [2]. The greatest challenge in the µFE modeling of bone is assigning the tissue-level material properties.

    • Computer modelling integrated with micro-CT and material testing provides additional insight to evaluate bone treatments: Application to a beta-glycan derived whey protein mice model

      2016, Computers in Biology and Medicine
      Citation Excerpt :

      Additionally, destructive compressive tests on the radius have shown strong correlations with equivalent FE models [4]. There have been studies that have used micro-FE models to assess bone strength in response to treatments [5–7]. Those studies revealed that modelling isolates strength changes in cortical and trabecular bone separately, that subject-specific variations within groups are important and that modelling reveals whole strength measures that may be different from surrogate measures of strength (like micro-CT indices).

    View all citing articles on Scopus

    Presented at the International Soceity of Biomechanics XVIII Congress, Zurich, Switzerland, July 2001.

    View full text