Bone mineral density and computer tomographic measurements in correlation with failure strength of equine metacarpal bones

Information regarding bone mineral density and fracture characteristics of the equine metacarpus are lacking. The aim of this study was to characterize the relationship between mechanical properties of the equine metacarpal bone and its biomechanical and morphometric properties. Third metacarpal bones were extracted from horses euthanized unrelated to musculoskeletal conditions. In total, bone specimens from 26 front limbs of 13 horses (7.8 ± 5.8 years old) including Lipizzaner (n = 5), Hungarian Warmblood (n = 2), Holsteiner (n = 2), Thoroughbred (n = 1), Hungarian Sporthorse (n = 1), Friesian (n = 1), and Shagya Arabian (n = 1) were collected. The horses included 7 mares, 4 stallions and 2 geldings. Assessment of the bone mineral density of the whole bone across four specific regions of interest was performed using dual-energy X-ray absorptiometry. The bones were scanned using a computer tomographic scanner to measure cross-sectional morphometric properties such as bone mineral density and cross-sectional dimensions including cortical area and cortical width. Mechanical properties (breaking force, bending strength, elastic modulus) were determined by a 3-point bending test. Significant positive linear correlations were found between the breaking force and bone mineral density of the entire third metacarpal bones (P < 0.001, r = 0.72), the medial cortex region of interest (P < 0.001, r = 0.68) and the transverse region of interest (P < 0.001, r = 0.61). The correlation between the breaking force and bone mineral density of the equine third metacarpal bone found in this study warrants in vivo investigations. Third metacarpal bone, densitometry, biomechanical properties, CT, horse Although bone mineral densitometry is an essential diagnostic method to determine increased fracture risk in human patients, application in the veterinary medicine is negligible. In horses, the current works are in experimental stages (Shryver 1978; Hanson and Markel 1995; McClure et al. 2001). Evaluation of equine skeletal structures with dual energy x-ray absorptiometry (DXA) was first described by Lawrence and Ott (1985). Since then, DXA has mainly been applied for the measurements of bone mineral density of the metacarpal bones (van Harreveld et al. 2002; Walker et al. 2004). In vitro studies evaluated the bone density of different equine bones (Hanson and Markel 1995; Firth et al. 1999; Tóth et al. 2010). The DXA method could be valuable to assess fracture risk of the metacarpus in horses. However, data describing correlation between the failure strength and the different bone mineral density (BMD) values are missing. Therefore, the aim of this in vitro study was to describe the relationship between mechanical properties of the equine metacarpal bone, such as breaking force, elastic moduli and bending strength, and its biomechanical and morphometric structural properties such as BMD and computer tomographic (CT) cross-sectional dimensions. We hypothesized that there is a linear correlation between BMD, CT morphometric indices and the bending strength, elastic modulus and breaking force of the equine third metacarpal bone. ACTA VET. BRNO 2014, 83: 045–050; doi:10.2754/avb201483010045 Address for correspondence: Dr. Péter Tóth Clinic for Large Animals, Faculty of Veterinary Science Szent István University 2225 Üllő, Dóra-major, Hungary E-mail: Toth.Peter@aotk.szie.hu http://actavet.vfu.cz/ Materials and Methods


Bone dissection and preparation
After dissection and manual removal of all soft tissue, the bones were stored in 70% ethyl-alcohol at room temperature until measurements, as previously suggested by Beaupied et al. (2006).

Bone mineral densitometry
Post mortem examinations were done by using a Norland XR-26 densitometer (Norland Corporation, Fort Atkinson, WI, USA) at the First Department of Medicine, Semmelweis University of Medicine, Budapest, Hungary.During the measurements the bones were placed on a 20 mm wide plexiglass plate to imitate soft tissue density as described in detail elsewhere (T ó t h et al. 2010).The bones were measured × 3 from the dorsopalmar (DP) direction and averaged.The regions of interest (ROI) were the entire bone, the medial cortex region, the lateral cortex, the transverse area of the longitudinal centre and the perpendicular medial cortex as shown in (Plate III, Fig. 1).Positions of the ROIs were chosen in the loaded area of the assessed metacarpal bones because the question was the correlation between the BMD and failure strength.

CT scan
The bones were scanned with a Siemens Somaton Emotion 6 Multislice CT (130 kV, 20 mAs, slides: 2 mm) (Siemens AG, Erlangen, Germany) at the Institute of Diagnostic Imaging and Radiation Oncology, Kaposvár University, Hungary.The cortical width was measured × 3 and averaged at each bone quadrant at the longitudinal centre and the cross-sectional area was calculated using Siemens SIENET software (Siemens AG, Erlangen, Germany).

Load testing
The bones were tested with an INSTRON 8872 servohydraulic Universal Testing Machine /UTM/ (Instron, Norwood, MA, USA) at the Laboratory of Biomechanics, University of Technology and Economics, Budapest, Hungary.The specimens were supported by proximal and distal metal rods placed 180 mm apart.A third rod fixated to the actuator was used to transmit the load to the palmar mid-diaphyseal cortex, 90 mm from the proximal and distal rod in a palmaro-dorsal direction at a speed of 25 mm/sec until breaking.

Data analysis
Statistical analysis of the data was performed using commercially available software (Minitab 16: Minitab Inc., PA, USA).Descriptive statistical analyses were performed to calculate the mean, standard deviation, median and range of each individual variable.Distribution of the data was tested with the Shapiro-Wilk method.Pearson's linear regression analyses were performed to reveal possible correlations between the bone length, BMD indicators of the selected ROIs and mechanical properties (bending strength, elastic moduli and breaking force).Pearson's linear regression analyses were also performed to reveal possible correlations between morphometric CT measurements (lateral, medial, dorsal, palmar width and area) and the above mentioned mechanical properties.A P < 0.05 was considered significant in all tests and a correlation coefficient (r) value greater than 0.6 was considered to be sufficient to assume a linear relationship between the two variables in a given model.

Results
Descriptive statistical values for each variable across the equine third metacarpal bone specimens are summarised in Table 1.
The breaking force was found to correlate with the BMD of the entire bone (P < 0.001, r = 0.72; Fig. 2), the medial cortex ROI (P < 0.001, r = 0.68) (Fig. 3), the transverse ROI (P < 0.001, r = 0.61), the lateral cortex ROI (P < 0.001, r = 0.59), and the medial perpendicular ROI (P < 0.001, r = 0.6).Breaking force was not found to correlate well with the bone dimensions such as length.Cortical width measured by CT did not correlate with breaking force.
There were no significant correlations between bending strength and any of the bone mineral density measurements.Furthermore, bending strength did not significantly correlate with cortical widths in each quadrant or bone length.
No significant correlations were found between the elastic modulus and BMD measurements.Furthermore, the elastic moduli were not found to adequately correlate with the dorsal cortex width and with the width of the remaining quadrants.analysing the entire bone, and that there were no significant differences between the whole bone measurements between dissected and intact metacarpal bones (Carter et al. 1992).
The load testing method used in this study involved the artificial application of force in the palmaro-dorsal direction at the longitudinal centre of the bone.This approach was chosen due to practical reasons, as we aimed to focus on the midshaft of the third metacarpal bone during the loading test.In this position the proximal and the distal rod of the testing machine lie under the cannon bone and not under the splintbones.In this positioning, the confounding effects of the splintbones are minimal on the breaking strength values of the cannon bones.It should be reiterated that the most prevalent fracture site in racehorses is the lateral condyle (Riggs et al. 1999); however, the loading test of that single region or the proximodistal compression described by Davies (2009) was not feasible in our settings.
In conclusion, the most determinant factors were those representing the entire bone rather than specific regions.Bone strength is determined by a complex set of structural variables that need to be collectively considered.Therefore, despite differences among cortical regions, the strong relationship between whole bone BMD and breaking force suggests that further studies should focus on the entire bone rather than specific regions.Complete understanding of fracture risk requires combining these factors into a single cohesive model.Considering the fact that portable DXA devices are a feasible, accurate and precise tool in standing horses (Donabedian et al. 2005), correlation between breaking force and whole bone mineral density identified in this study warrants in vivo investigations.

Fig. 2 .
Fig. 2. Correlation between the whole bone mineral density and the breaking force of third metacarpal bone of horses McIII -third metacarpal bone, BMD -bone mineral density, S -skewness, R-Sq -coefficient of determination, R-Sq(adj) -adjusted coefficient of determination

Table 1 .
Descriptive statistical values for each variable across the third metacarpal bone specimens in horses.