SEARCH

SEARCH BY CITATION

Keywords:

  • true three-dimensional microarchitecture;
  • DXA;
  • bone loss;
  • bone growth;
  • histomorphometry

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We tested a novel microcomputed tomograph designed to longitudinally and noninvasively monitor bone alterations in hindlimb-unloaded female rats at a resolution of 26 μm over a period of 3 weeks. This prototype has a potential to detect three-dimensional trabecular microarchitectural changes induced by growth and unloading.

Introduction: Until now, data concerning structural changes of cancellous bone have only been available after necropsy of animals. In this study, we tested a novel microcomputed tomography (μCT) technique designed to monitor such changes repeatedly at a resolution of 26 μm with an acquisition time of about 10 minutes to map the entire proximal tibial metaphysis.

Materials and Methods: Four-month-old female Wistar rats were randomized to seven groups of 10 animals to be either tail-suspended or to act as controls. μCT and DXA measurements were performed at 0, 7, 14, and 23 days in suspended and control rats. One group was killed at each of these time points, and bone samples were processed for histomorphometry and ex vivo μCT.

Results: We verified that a good correlation was obtained between two-dimensional bone parameters evaluated in longitudinal tibial sections either by histomorphometry or μCT and μCT parameters obtained from either in vivo or ex vivo tibias. The longitudinal survey allowed earlier detection of both growth and unloading-related bone changes than the transverse survey. In controls, aging induced denser bones, reorganization of the trabecular network toward a more oriented plate-like structure, and an isotropic pattern. Unloading first inhibited cortical and cancellous bone growth and then induced bone loss characterized by fewer trabeculae, reduced connectivity density, and enhanced structure model index (SMI), revealing a lighter cancellous structure with development of rod-like characteristics.

Conclusion: We show for the first time that this μCT prototype has a great potential to accurately, repeatedly, reliably, and rapidly investigate alterations of three-dimensional trabecular microarchitecture.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

There has been a growing interest in microstructural changes of cancellous bone, because bone densitometry cannot entirely account for the observed decrease in bone quality.(1–3) There are current medications that increase bone mass in osteoporotic patients.(4) However, bone strength could be increased while bone mass is not, suggesting that bone quality is the prime factor.(5,6) Palle et al.,(7) in a bed rest study, found that bone microarchitecture was altered, despite no modification of bone mass.

However, assessment of bone microarchitecture by conventional histomorphometry leads to typical problems such as destruction of the bone sample (embedding and cutting). Moreover, the distribution of cancellous bone is heterogeneous, and bone histomorphometric analysis is limited to a few fields of view. The standard procedure leads to evaluation of three-dimensional (3D) morphometric indices, derived from two-dimensional (2D) images using stereological methods, and a range of important indices can be indirectly derived from a structural model assumed to be fixed. The assumption of a fixed model type is critical, because the structure of trabecular bone may change continuously because of aging and disease effects.

Over recent years, microtomography has become an increasingly popular method to measure bone samples, because of its relative rapidity compared with conventional histology and its potential as a nondestructive method. The progress from 2D to 3D analysis and direct evaluation of 3D parameters without assuming a fixed model structure have contributed to the growth of this technique. In rodent bone studies, higher spatial resolution can be achieved at the expense of invasiveness, which is the prime factor limiting current computed tomography (CT) scanner developments. The method applied today currently consists of killing the animal and examining the bone by in vitro microtomography. The disadvantage of this technique is that the animal cannot be examined at various stages of disease or more than once during the course of treatment. Other techniques have been recently developed to accurately image trabecular bone structure in vivo at a lower resolution.(8) Spatial resolution and signal/noise ratio are fundamentally limited by the radiation dose that can be applied: the dose increases with the fourth power of linear resolution. For this reason, resolution is generally lower in in vivo measurements, and the dose becomes an important issue for in vivo measurements. Because the bones of rodents, such as rats and mice, are considerably smaller than human bones and because the acceptable radiation dose might be proportionally relatively higher than in humans, spatial resolution achieved in laboratory animals is clearly better than in humans.

The primary purpose of this study was to evaluate the accuracy of a newly developed noninvasive experimental CT scanner. This scanner, a rotating gantry version of the previously developed 3D microcomputed tomography (3D μCT), offers high scanning speed and high resolution, at a moderately low radiation dose. This instrument was codeveloped by the Institute for Biomedical Engineering (Institute for Biomedical Engineering, University of Zürich and Swiss Federal Institute of Technology [ETH], Zürich, Switzerland) and Scanco Medical (Scanco Medical AG, Bassersdorf, Switzerland) and is designed to image small laboratory animals in vivo. To validate this tool, we characterized the bone changes of tail-suspended rats by simultaneously assessing bone histomorphometry, in vivo DXA, and in vivo 3D μCT. We compared the bone dynamics investigated by these three methods to determine the kinetics of bone loss in this model of osteoporosis.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Animals

Seventy female rats (Charles River, Iffa Credo, L'arbresle, France), 12 weeks old and weighing 250 ± 10 g, were randomly assigned to seven groups of 10 animals each. Rats were tail suspended (S) or were maintained in suspension cages but not suspended (control groups; Ctr), for 7, 14, and 23 days. A baseline group was killed on the first day of the experiment.

The rats were acclimatized for 1 week with standard temperature (23 ± 1°C) and light:dark (12:12) conditions. Animals were individually housed, provided with food (standard diet) and water ad libitum. The suspension procedure was performed according to the Wronski and Morey-Holton(9) recommendations. Fluorochrome double bone labeling was performed 4 days and 1 day before death. The 7-, 14-, and 23-day rats received an intraperitoneal injection of 30 mg/kg of tetracycline (Sigma Aldrich). Baseline rats were not labeled.

μCT and DXA measurements were performed on days 0, 7, 14, and 23. Before measurement, rats were anesthetized by an intraperitoneal dose of 5 mg/kg of ketamine/xylazine solution, and the animals were killed with a high dose of Nesdonal (Specia, Paris, France).

The left tibia of each animal was measured in vivo and ex vivo by μCT and DXA, and bone slices from the same sample were processed for conventional histomorphometry.

Histomorphometric analysis

The proximal tibial metaphysis was fixed in 4% formaldehyde solution, dehydrated in acetone, and embedded in methylmethacrylate. Longitudinal frontal slices were cut from the embedded bones with the Jung Model K microtome (Carl Zeiss, Heidelberg, Germany). Six nonserial sections, 8 μm thick, were used for modified Goldner staining. Fourteen-micrometer-thick sections were used to determine the dynamic indices of bone formation (double-labeled surface [dLS]/bone surface [BS], mineral apposition rate [MAR], bone formation rate [BFR]/BS). MAR was derived from fluorochrome interlabel distances. BFR were subsequently obtained from the product of dLS/BS and MAR. Six-micrometer-thick sections were used for TRACP staining, allowing determination of osteoclastic parameters. For each section, the data were collected in 2.6-mm-large region of interest (ROI) within the secondary spongiosa, according to the areas measured in 2D and 3D μCT (see below). Bone volume and parameters reflecting trabecular structure were measured using an automatic image analysis system (Biocom). Bone cellular and macroscopic parameters were measured with a semiautomatic system:digitizing tablet (Summasketch; Summagraphics, Paris, France) connected to a Macintosh personal computer with software designed in our laboratory.

High resolution μCT

The left tibias of 7-, 14-, and 23-day groups were scanned with a high resolution μCT prototype (VivaCT20) from Scanco Medical, recently described by Kohlbrenner et al.(10) (Fig. 1). This apparatus is able to noninvasively examine, in vivo, the bones of small laboratory animals (with a diameter not exceeding 20 mm) with a high resolution. The VivaCT 20 has a rotating gantry, and the X-ray source and detector rotate around the object. The spatial resolution of the system is 20 μm in the x and y axes and 40 μm in the z axis. The following CT settings were used: voxel matrix 20 × 20 × 26 μm3.

thumbnail image

Figure FIG. 1.. VivaCT 20 image acquisition and cancellous bone measurement. A 6.58-mm height zone is acquired (A) during 10 minutes, and then (B) a ROI is chosen that represents (C) a 2.60-mm height zone. (D) Finally the cancellous envelope is extracted from the chosen ROI.

Download figure to PowerPoint

The anesthetized animals were placed inside the VivaCT20, and the left tibia was fixed inside a carbon tube and measured. The scanned region corresponded to a zone of 253 transverse slices (6.58 mm) under the left proximal tibial growth plate, each slice consisting of 10242 pixels (Fig. 1). The net scanning time was about 10 minutes.

From the acquired data, the ROI in the axial direction was delimited anatomically from the bottom of the primary spongiosa (ISP) up to a height of 2.6 mm (100 slices, Fig. 1). Then we adjusted the ROI top at the primary-secondary frontier susceptible to vary according to growth or bone loss processes. For each transverse slice, the ROI was established manually in an area of trabecular bone, as large as possible.

In a preliminary experiment, 10 rats, 5 of which were tail-suspended, were killed and their tibias were processed for μCT measurements (both in vivo and ex vivo). The samples were then processed for conventional histomorphometry. The scanned acquired images were reprocessed to test the threshold giving the best fit with histomorphometric BV/TV data. We therefore defined the optimal threshold that was used for the rest of the study. All gray-scale images were segmented using a Gaussian filter and a fixed threshold (15% of maximal grayscale, corresponding to a value of 150) for all data. The CVs under these conditions are shown in Table 1.

Table Table 1. CV Expressed as Percentage as Obtained on the Different 3D Parameters Measured With the VivaCT 20, When Measuring the Same Animal 10 Times
Thumbnail image of

2D trabecular parameters

Six serial 2D frontal longitudinal slices (104 μm) were extracted from the acquired images for further evaluation. The axial and transverse ROI (Fig. 1) corresponded to similar regions measured by conventional histomorphometry: 2.6 mm from the bottom of the ISP and 104 μm longitudinally (the volume explored by conventional histomorphometry was 96 μm). 2D structural indices were calculated according to Parfitt et al.(11) for 14 samples selected at random from all acquired data. Similar areas were measured by histomorphometry and μCT to compare the two methods.

3D trabecular parameters

The bone volume fraction was calculated directly by plotting gray voxels representing bone fraction against black voxels (non-bone objects; VOX BV/TV). Bone volume (BV) and bone surface (BS) were calculated using a tetrahedron meshing technique generated by the “marching cubes method,”(12) and total volume (TV) was calculated as the volume of the volume of interest (VOI). The normalized indices (BV/TV, BS/TV, and BS/BV) were used.

Mean trabecular number (Tb.N), mean trabecular thickness (Tb.Th), and mean trabecular separation (Tb.Sp) were calculated according to Parfitt et al.,(11) assuming a constant structure model and applying stereological techniques. Tb.Th is defined as twice BV/BS, Tb.Sp as 2(TV − BV)/BS, and Tb.N as 0.5BS/TV.

Three-dimensional metric indices were also calculated using direct techniques based on the distance transformation,(13,14) without assuming a constant model. Direct indices, indicated by an asterisk to differentiate them from model-dependent indices, were calculated as follows:Tb.Thwas calculated by filling maximal spheres into a structure, and the average thickness of all voxels corresponded to Tb.Th.

The same procedure was used to determine Tb.Sp, but in this case the non-bone voxels were filled with maximal spheres, and the mean thickness of the marrow cavities represented Tb.Sp.

Tb.Nwas the inverse of the mean distance between the midaxes of the observed structure. The midaxes of the structure were assessed from the binary 3D image using the 3D distance transformation and by extracting the center points of non-redundant spheres that fill the structure completely. The mean distance between the midaxes was then determined by analogy with the Tb.Spcalculation, and the distance between the midaxes was assessed.

The plate-rod characteristic of the structure was estimated by the structure model index (SMI),(15) calculated by differential analysis of a triangulated surface of a structure: SMI = 6{[BV(dBS/dr)]/BS2}. dBS/dr is the surface area derivative with respect to a linear r (one-half thickness or the radius assumed to be constant over the entire structure). The SMI value is 0 for an ideal plate and 3 for an ideal rod structure. Values between 0 and 3 correspond to a structure with both plates and rods, depending on the volume ratio between rods and plates.

The geometric degree of anisotropy (D.A.) is defined as the ratio between the maximal and minimal radius of the mean intercept length (MIL) ellipsoid.(16,17)

Connectivity density (Conn.D.) was calculated using the Euler method of Odgaard and Gundersen.(18)

The linear X-ray attenuation coefficient (Lin.Att.), likened to apparent bone mineral density (BMD), was also evaluated.

To analyze the cortex, we choose a cross-sectional slice on the original images for which the distance between the tibia and fibula was approximately 4 mm. We assumed that the individual distance was constant, and for each measurement point, cortical area (Ct.Ar.), total area (T.Ar.), and marrow cavity Area (Ma.Ar.) were evaluated, with the same Gaussian filter and the same fixed threshold as for the trabecular structure.

DXA

A DXA PIXImus densitometer (Lunar Corp., Madison, WI, USA) with small animal software was used to measure BMD and bone mineral content (BMC). It is a rapid (5-minute image acquisition) and precise small animal densitometer with a precision of 1% CV for total skeletal BMD and 1.5% CV for femur ROI BMD. After completing the measurement, the ROI rectangle was moved and reshaped to cover a portion of the left forearm. The PIXI software automatically calculated bone density and recorded the data as Microsoft Excel files. Entire left tibias and left femora were analyzed for the 7-, 14-, and 23-day experiments. Left humeri were also measured for the 14- and 23-day animals.

Statistical analysis

Statistical analysis was performed using commercially available statistical software (STATISTICA; StatSoft Inc., Tulsa, OK, USA). For densitometric and tomographic data, differences between groups were initially analyzed by two-way ANOVA, with a between-groups factor (tail-suspended or control) and a repeated measures factor (within subjects factor). When F values for a given variable were found to be significant, the sequentially rejecting Bonferroni-Holm test(19) was subsequently performed using the Holm's adjusted p values. Results were considered to be significantly different at p < 0.05.

Two-way ANOVA was performed on histomorphometric data to determine the influence of both suspension and time period factors on structural and cellular parameters. When F values for a given variable were found to be significant, a post hoc Scheffé test were performed, and the results were considered to be significantly different at p < 0.05.

Pearson correlation analysis was performed to show potential relationships between the various morphometric methods, and p values were considered to be significant at p < 0.05.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Body weight

A slow but steady increase in body weight was observed in control groups throughout the observation period, as shown in Ctr 23 (Fig. 2). In S 23 rats, an acute decrease in body weight was observed during the first 3 days followed by a return to baseline values after 1 week. Then S rats grew in parallel to Ctr rats. The body weight in Ctr 7, Ctr 14, S 7, and S 14 groups changed in a similar way as in Ctr 23 and S 23 groups (data not shown).

thumbnail image

Figure FIG. 2.. Body weight evolution as observed during a 23-day experiment. Results are expressed as mean ± SD in control animals (filled squares and solid line) and in suspended animals (open squares and dashed line). (a) Significant difference vs. baseline. *Significant difference vs. Ctr; p < 0.05; N = 10 rats per group.

Download figure to PowerPoint

Correlations

Our first aim was to verify that 2D bone parameters evaluated in frontal longitudinal sections either by histomorphometry or μCT were well correlated. A strong correlation was found for BV/TV (r = 0.98; p < 0.0001; Fig. 3A), and a significant positive correlation was observed for the other parameters, such as Tb.N, Tb.Sp, and BS/BV (r = 0.781; r = 0.883; r = 0.699, respectively, where p ≤ 0.01).

thumbnail image

Figure FIG. 3.. (A) Correlation between bone volume fraction as measured by 2D μCT (BV/TV1) vs. volume fraction measured by histomorphometry (BV/TV). (B) Correlation between bone volume fraction as measured by 3D μCT (BV/TV*3D) vs. volume fraction measured by histomorphometry (BV/TV 2D).

Download figure to PowerPoint

BV/TV evaluated by histomorphometry was related to 3D μCT BV/TV, although to a lesser extent than in 2D measurements, as shown in Fig. 3A (Fig. 3B). Correlation coefficients between Tb.N, Tb.Sp, and BS/BV measured by histomorphometry and similar parameters evaluated with or without models by 3D μCT ranged between 0.555 and 0.732 (p ≤ 0.001). No significant correlation was found for Tb.Th.

A significant correlation was also observed between 2D and 3D parameters measured by VivaCT20 with the same thresholding (0.357 < r < 0.802; p ≤ 0.01).

Significant correlations were observed between all the 3D μCT parameters evaluated in bones scanned either in vivo or ex vivo (excised bones after death); the experiment was carried out on 20 animals (Table 2).

Table Table 2. Correlations Coefficients and p Values Between the 3D μCT Parameters Evaluated in Bones Scanned Either In Vivo or Ex Vivo (Excised Bones After Death, N = 20)
Thumbnail image of

Histomorphometric assessment of bone alterations

A two-way ANOVA analysis demonstrated that “immobilization effects” was the essential factor having a greater influence on architectural modifications than the “growth effect” factor or the interaction of both factors.

BV/TV and Tb.N increased significantly throughout the experiment in Ctr rats, revealing a growth effect (Table 3). Bone loss was revealed after the second week of immobilization, because BV/TV and Tb.N were significantly lower in S rats than in Ctr rats. Moreover, a significant increase in Tb.Sp was observed in S 23 compared with S 7 rats. Tb.Th remained constant throughout the experiment in both Ctr and S rats, and no significant difference was observed between these two groups at any time point. The longitudinal growth rate (LGR) did not change significantly, although a trend toward decreases was seen in S 7 and S 14 groups compared with the respective Ctr groups.

Table Table 3. Suspension and Aging-Induced Changes in Longitudinal Growth Rate (L.G.R) and in Bone Microarchitectural Parameters in the Secondary Spongiosa of the Proximal Tibial Metaphysis
Thumbnail image of

Dynamic bone formation parameters (BFR/BS) were reduced in S 14 rats compared with S 7 rats because of a decrease in dLS/BS (Table 4), while MAR remained constant throughout the observation period. A group difference between S and Ctr rats was observed on day 23. These parameters remained unchanged in Ctr rats. Bone resorption as assessed by Oc.S/BS was not significantly altered by the model.

Table Table 4. Suspension and Aging-Induced Changes in Bone Cellular Activities, Evaluated in the Secondary Spongiosa of the Tibial Proximal Metaphysis
Thumbnail image of

Microtomographic evaluation of tibial alteration

Bone growth was accentuated in Ctr rats (Tables 5 and 6), as an increase in bone mass as well as microarchitectural changes were demonstrated. Trabeculae became more oriented toward isotropic bone (reduction of D.A. on day 14) and organization into plates was observed (reduction in SMI on day 23), reflecting progression of trabecular microarchitecture toward a more compact structure. In S rats, a decrease in BV/TV was observed as early as 7 days (Fig. 4), which further declined with time and was significantly lower than in Ctr rats on day 14. Tb.N decreased significantly on day 7 in S groups compared with baseline and became significantly lower than in Ctr rats on day 14 and remained so until day 23. Tb.Sp increased in S rats on day 7 and became significantly higher than in Ctr rats after 14 days, and then increased slowly until day 23. Tb.Th remained constant throughout the experiment in both Ctr and S rats, and no significant difference was observed between the groups. Variations in 3D metric parameters were fairly similar for both model-dependent and model-independent parameters (Table 5). In S rats, the decrease in Lin.Att. from 14 days paralleled the decrease in BV/TV. Moreover, a decrease in Conn. Den. was observed from day 7 and subsequently decreased until day 23. The difference between S rats and Ctr only became statistically significant at day 14. In suspended rats, SMI was more indicative of a rod structure. This pattern remained stable throughout the experimental period. However, a significant difference was demonstrated between S and Ctr rats at days 7 and 14, because SMI was higher in S rats. No significant variation of DA was demonstrated over the same time period. This suggests that the growth-related progression toward plates was inhibited in S rats.

Table Table 5. μCT Metric Parameters in the Secondary Spongiosa of the Proximal Tibial Metaphysis
Thumbnail image of
Table Table 6. μCT Non Metric Parameters in the Secondary Spongiosa of the Proximal Tibial Metaphysis
Thumbnail image of
thumbnail image

Figure FIG. 4.. Bone volume fraction evolution as measured by 3D μCT during 7 (triangles), 14 (circles), and 23 days (squares). Results are expressed as mean ± SE in control animals (filled objects and solid lines) and in suspended animals (open objects and dashed lines). (a) Significant difference vs. baseline. (b) Significant difference vs. 7 days; *Significant difference vs. aged-matched Ctr; p < 0.05; N = 10 rats per group.

Download figure to PowerPoint

Ct.Ar. significantly increased in Ctr rats over the experimental period. Suspension stopped cortical growth (Fig. 5) until day 14, and then induced a decrease of Ct.Ar. Neither T.Ar. nor Ma.Ar. was significantly altered by aging or suspension in either S or Ctr rats, although T.Ar decreased in S rats.

thumbnail image

Figure FIG. 5.. Cortical area (Ct.Ar.) evolution as measured by μCT during a 23-day experiment. (a) Significant difference vs. baseline. (b) Significant difference vs. 7 days. (c) Significant difference vs. 14 days. *Significant difference vs. aged-matched Ctr; p < 0.05. Results are expressed as mean ± SD; N = 8 rats per group.

Download figure to PowerPoint

DXA assessment of bone alteration

Left tibia BMD and BMC increased with aging in Ctr rats on day 14. Suspension did not seem to induce any real bone loss in the whole tibia (Table 7), but rather inhibition of age-related bone growth, because no time-related changes in BMD or BMC were observed in suspended animals. It is noteworthy that, despite body weight adjustment, baseline BMD or BMC values differed significantly between the groups. This was also true for the other 3D μCT bone parameters.

Table Table 7. Suspension and Aging-Induced Changes in Bone Densitometric Parameters of the Entire Left Tibias
Thumbnail image of

In control animals, femoral BMC increased by 12% after 23 days, whereas BMD increased by only 8%. Bone area therefore increased to a lesser degree than mineral accumulation (Table 8). In suspended groups, only an increase of BMC was observed (3.6% after 1 week, reaching 6.5% at 23 days) with no significant variation in BMD. Group differences for baseline BMC values were observed, as for the tibia.

Table Table 8. Suspension and Aging-Induced Changes in Bone Densitometric Parameters of the Entire Left Femora
Thumbnail image of

Humeral growth was not altered by suspension, because BMC and BMD increased at a similar rate in both S and Ctr rats.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Our main objective was to determine to what extent 3D μCT parameters acquired in vivo were comparable with those obtained from optical images of the corresponding histological sections. For this purpose, frontal longitudinal tomographic sections of the proximal tibial metaphyses were extracted from the reconstructed excised tibia to match the 2D histomorphometric ROI. The 2D μCT BV/TV ratio was strongly correlated (r = 0.98) with the 2D histomorphometric BV/TV ratio, despite the lower μCT resolution (nominal resolution 26 μm versus 1 μm for microscopic resolution). Significant correlations were also observed for the structural parameters (Tb.Th, Tb.N, Tb.Sp) derived from the Parfitt-based parallel plate model (r values ranging from 0.699 to 0.883), although lower than for BV/TV. Müller et al.,(20) in a study on human iliac crest, also found a high correlation between these parameters on histology and μCT (nominal resolution of 14 μm). We also demonstrated that the absolute value of BV/TV was similar regardless of the technique used. The values of Tb.Sp and Tb.N were fairly similar (20% percent difference). However, Tb.Th, as estimated by 3DμCT, was 2-fold greater than histological values. An even greater difference has also been previously found in human cancellous bone either by comparing computed tomography (150 μm resolution, 330 μm section thickness) and histology(21) or by comparing MR-derived measurements of trabecular structures obtained from slices of the radius (resolution 156 μm, slice thickness 300 μm) with those obtained from high-resolution X-tomography microscopy images at an isotropic resolution of 18 μm.(22) In the present study, no correlation was demonstrated between 2D histomorphometric Tb.Th and 3D μCT Tb.Th (evaluated by either direct or model-dependent methods), although such a correlation was observed for the other parameters. This lack of correlation does not seem to be caused by insufficient image resolution, which would have led to significant underestimation of Tb.N, not observed in this study. It is also not caused by heterogeneity of the size of the regions measured (this parameter is very sensitive in metaphyseal long bone), because ROI were determined very carefully with the two methods (see Material and Methods section). In addition to this, we observed that the direct Tb.Thmeasurement was systematically higher than Tb.Th assessed with the plate model assumption by approximatively 10%. The reason of this difference might reside in the deviation of the trabecular structure from the ideal plate-model. In our study, the SMI changed during the experimental period in growing rats, but it was globally higher than 2 (1 corresponds to a plate-like structure and 3 corresponds to a rod-like structure), indicating the presence of more rods than plates. Rods result in a greater surface-to-volume ratio for a given thickness, leading to a smaller apparent thickness derived from BV/BS. Hildebrand et al.(14) found that, even in the human femoral head, which has a pronounced plate-like structure, thickness is underestimated by model-dependent evaluation. Various studies, both in humans(20,21,23) and in the ovariectomized rat model,(24,25) have demonstrated the need to use 3D measuring techniques that are able to visualize the real architecture of cancellous bone without assumptions concerning the type of structure. Model-based algorithms may potentially introduce biases affecting the parameters determined, and these biases may modify the impact of age- or treatment-related changes.

The results of this study show that tail-suspension induces metaphyseal bone loss. The longitudinal survey (DXA, μCT) allowed earlier detection of bone changes than the transverse histomorphometric, tomographic, or densitometric surveys. Initial measurements by DXA and μCT both revealed differences in baseline values between the groups, indicating that adjustment based on body weight data did not ensure homogeneity of bone parameters. Adjustment should therefore be performed based on densitometric data. This group heterogeneity might also be detrimental to detect early changes in a transverse survey. One of the greatest advantage of carrying out longitudinal studies is to compare changes against the animal own baseline value.

3D μCT quantification of bone microarchitecture based on model-independent methods reveals structural modifications earlier than 2D histomorphometric measurements. With a nominal resolution of 20 μm in the x and y axes and 26 μm in the z axis, the prototype used for microtomographic acquisitions (VivaCT 20) showed a sufficient discriminating capacity to allow precise qualitative and quantitative study of bone microstructure in vivo.

Because it is difficult and not frequent to use skeletally mature rats (age 12 months or older), we built an experiment with growing animals. It is necessary to be able to distinguish between changes related to inhibition of bone growth and changes attributable to real bone loss. Comparison of the kinetics between control and suspended animals and between the “baseline control group” (or baseline values for longitudinal survey) and experimental groups should help us to make this distinction. No detectable growth effect was observed during the first week, but bone loss was already demonstrated after 7 days in the S groups, together with a reduction of Tb.N, and an increase in Tb.Sp. Tb.N was further decreased after 7 days. This is also reflected by the decrease in connectivity density observed after only 7 days of suspension, which continued to decrease thereafter. Using ex vivo μCT (μCT 20; Scanco Medical AG), we previously reported a pronounced decline in connectivity density in tail-suspended male rats.(26) Complete loss of trabeculae has been demonstrated in postmenopausal women(1) and ovariectomized rats.(24,27) Ito et al.(24) found a dramatic decrease in Tb.N and a modest decrease in Tb.Th in female rats after neurectomy. A decrease in Tb.N with a similar or less pronounced decrease in Tb.Th has been reported in tail-suspended rats.(26,28) Analysis of bone cellular activities in the suspended groups showed that BFR/BS was unchanged at 7 days, suggesting that the reduction in BV/TV at this time is caused by increased bone resorption, as demonstrated in other models of immobilization.(29) Trabecular bone loss was accentuated between 7 and 14 days of suspension, while the bone formation decreased during this second week, as shown by the decrease in dLS/BS.

Analysis of 3D μCT nonmetric parameters showed that tail suspension abolishes the tibial growth-induced decrease in both DA and SMI, as seen in control animals. SMI is even increased by tail suspension. Unloading therefore counteracts the effects of growth on the bone microarchitecture; instead of having a trabecular pattern evolving toward a more isotropic and plate-like structure, unloading preserves anisotropy and accentuates the rod-like structure. The studies by Ito et al.(24) in neurectomized rats and by Laib et al.(26) in male suspended rats also demonstrated an increase in SMI. These results emphasize the importance of assessing real 3D parameters not based on a model assumption.

Suppression of growth in the femur shows that the metabolism of cortical bone is altered by immobilization, because we observed a decrease in BMC in the S groups, while BMD was not altered. These results suggest that bone area is altered by tail suspension. Inhibition of femoral growth, as observed after 14 days of suspension, is caused by metaphyseal bone loss associated with inhibition of cortical growth.

In conclusion, we have validated in vivo μCT by comparing the same 2D parameters measured on histological sections and on the same areas reconstructed by tomography. We then confirmed that the plate-model assumption is not appropriate to evaluate the kinetics of bone changes during growth and hindlimb unloading. In our model, the tibial cortical and cancellous bone density and volume increased in growing female rats, and their trabecular network become more isotropic and developed a more plate-like appearance. These effects are counteracted by tail suspension, which initially inhibits the bone growth process and then induces bone loss accompanied by less numerous and less well-connected trabeculae presenting more rod-like characteristics.

We showed for the first time that this μCT prototype has great potential to accurately, repeatedly, reliably, and rapidly investigate changes in the 3D trabecular microarchitecture in a rat model of osteoporosis.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We thank Andres Laib, Bruno Koller, and Stephan Haemmerle from Scanco Medical (Scanco Medical AG, Bassersdorf, Switzerland) for technical support and image analysis. This study was supported by the ERISTO (European Research In Space and Terrestrial Osteoporosis) group with the partners LBBTO, SCANCO, and Institute for Biomedical Engineering (contract number 14232/00/NL/SH) and INSERM (Institut National de la Santé et de la Recherche Médicale).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  • 1
    Kleerekoper M, Villanueva AR, Stanciu J, Rao DS, Parfitt AM 1985 The role of three-dimensional trabecular microstructure in the pathogenesis of vertebral compression fractures. Calcif Tissue Int 37:594597.
  • 2
    Parfitt AM 1992 Implications of architecture for the pathogenesis and prevention of vertebral fracture. Bone 13(Suppl 2):S41S47.
  • 3
    Goldstein SA, Hollister SJ, Kuhn JL, Kikuchi N 1990 The mechanical and remodeling properties of trabecular bone. In: MowVC, RatcliffeA, WooSLY (eds.) Biomechanics of Diarthrodial Joints, Vol. 2. Springer Verlag, New York, NY, USA, pp. 6181.
  • 4
    Sosa M, Hernandez D, Segarra MC, Gomez A, de laPena E, Betancor P 2002 Effect of two forms of alendronate administration upon bone mass after two years of treatment. J Clin Densitom 5:2734.
  • 5
    Boivin GY, Chavassieux PM, Santora AC, Yates J, Meunier PJ 2000 Alendronate increases bone strength by increasing the mean degree of mineralization of bone tissue in osteoporotic women. Bone 27:687694.
  • 6
    Riggs 1990 A new option for treating osteoporosis. N Engl J Med 323:124125.
  • 7
    Palle S, Vico L, Bourrin S, Alexandre C 1992 Bone tissue response to four month anti-orthostatic bedrest: A bone histomorphometric study. Calcif Tissue Int 51:189194.
  • 8
    Laib A, Ruegsegger P 1999 Calibration of trabecular bone structure measurements of in vivo three-dimensional peripheral quantitative computed tomography with 28-micron-resolution microcomputed tomography. Bone 24:3539.
  • 9
    Wronski TJ, Morey-Holton ER 1987 Skeletal response to simulated weightlessness: A comparison of suspension techniques. Aviat Space Environ Med 58:6368.
  • 10
    Kohlbrenner A, Koller B, Hammerle S, Ruegsegger P 2001 In vivo micro tomography. Adv Exp Med Biol 496:213224.
  • 11
    Parfitt AM, Mathews CHE, Villanueva AR, Kleerekoper M, Frame B, Rao DS 1983 Relationships between surface, volume, and thickness of iliac trabecular bone in aging and in osteoporosis. Calcif Tissue Int 72:13961409.
  • 12
    Lorensen WE, Cline HE 1987 Marching cubes: A high resolution 3D surface construction algorithm. Comput Graphics 21:163169.
  • 13
    Hildebrand T, Ruegsegger P 1997 A new method for the model independent assessment of thickness in three-dimensional images. J Microsc 185:6775.
  • 14
    Hildebrand T, Laib A, Müller R, Dequeker J, Rüegsegger P 1999 Direct three-dimensional morphometric analysis of human cancellous bone: Microstructural data from spine, femur, iliac crest and calcaneus. J Bone Miner Res 14:11671174.
  • 15
    Hildebrand T, Rüegsegger P 1997 Quantification of bone microarchitecture with the structure model index. Comput Methods Biomech Biomed Engin 1:1523.
  • 16
    Harrigan TP, Mann RW 1984 Characterization of microstructural anisotropy in orthoptic materials using a second rank tensor. J Mater Sci 19:761767.
  • 17
    Whitehouse WJ 1974 The quantitative morphology of anisotropic trabecular bone. J Microsc 101:153168.
  • 18
    Odgaard A, Gundersen HJG 1993 Quantification of connectivity in cancellous bone, with special emphasis on 3D reconstructions. Bone 14:173182.
  • 19
    Holm S 1979 A simple sequentially rejective multiple test procedure. Scand J Stat 6:6570.
  • 20
    Müller R, Van Campenhout H, Van Damme B, Van Der Perre G, Dequeker J, Hildebrand T, Rüegsegger P 1998 Morphometric analysis of human bone biopsies: A quantitative structural comparison of histological sections and micro-computed tomography. Bone 23:5966.
  • 21
    Cendre E, Mitton D, Roux JP, Arlot ME, Duboeuf F, Burt-Pichat B, Rumelhart C, Peix G, Meunier PJ 1999 High-resolution computed tomography for architectural characterization of human lumbar cancellous bone: Relationships with histomorphometry and biomechanics. Osteoporos Int 10:353360.
  • 22
    Majumdar S, Newitt D, Mathur A, Osman D, Gies A, Chiu E, Lotz J, Kinney J, Genant H 1996 Magnetic resonance imaging of trabecular bone structure in the distal radius: Relationship with X-ray tomographic microscopy and biomechanics. Osteoporos Int 6:376385.
  • 23
    Ding M, Odgaard A, Linde F, Hvid I 2002 Age-related variations in the microstructure of human tibial cancellous bone. J Orthop Res 20:615621.
  • 24
    Ito M, Nishida A, Nakamura T, Uetani M, Hayashi K 2002 Differences of three-dimensional trabecular microstructure in osteopenic rat models caused by ovariectomy and neurectomy. Bone 30:594598.
  • 25
    Laib A, Kumer JL, Majumdar S, Lane NE 2001 The temporal changes of trabecular architecture in ovariectomized rats assessed by MicroCT. Osteoporos Int 12:936941.
  • 26
    Laib A, Barou O, Vico L, Lafage-Proust MH, Alexandre C, Rugsegger P 2000 3D micro-computed tomography of trabecular and cortical bone architecture with application to a rat model of immobilization osteoporosis. Med Biol Eng Comput 38:326332.
  • 27
    Yoshida S, Yamamuro T, Okumura H, Takahashi H 1991 Microstructural changes of osteopenic trabeculae in the rat. Bone 12:185194.
  • 28
    Barou O, David V, Vico L, Tirode C, Barbier A, Alexandre C, Lafage-Proust MH 2002 High-resolution three-dimensional micro-computed tomography detects bone loss and changes in trabecular architecture early: Comparison with DEXA and bone histomorphometry in a rat model of disuse osteoporosis. Invest Radiol 37:4046.
  • 29
    Weinreb M, Rodan GA, Thompson DD 1989 Osteopenia in the immobilized rat hindlimb is associated with an increased bone resorption and decrease bone formation. Bone 10:187195.