Articular stress fracture arising from the distal end of the third metacarpal bone (MC3) is a common serious injury in Thoroughbred racehorses. Currently, there is no method for predicting fracture risk clinically. We describe an ex-vivo biomechanical model in which we measured subchondral crack micromotion under compressive loading that modeled high speed running. Using this model, we determined the relationship between subchondral crack dimensions measured using computed tomography (CT) and crack micromotion. Thoracic limbs from 40 Thoroughbred racehorses that had sustained a catastrophic injury were studied. Limbs were radiographed and examined using CT. Parasagittal subchondral fatigue crack dimensions were measured on CT images using image analysis software. MC3 bones with fatigue cracks were tested using five cycles of compressive loading at -7,500N (38 condyles, 18 horses). Crack motion was recorded using an extensometer. Mechanical testing was validated using bones with 3 mm and 5 mm deep parasagittal subchondral slots that modeled naturally occurring fatigue cracks. After testing, subchondral crack density was determined histologically. Creation of parasagittal subchondral slots induced significant micromotion during loading (p
Functional adaptation is readily detectable by 4 months of race training . Contact stresses on the palmar or plantar regions of the distal end of the MC3/MT3 bones from the proximal sesamoid bones are more than twice the stresses imposed on the dorsal region at the canter; this is a result of more load being shifted to the suspensory apparatus during increased fetlock joint extension . As training increases, adaptation in the subchondral plate leads to sclerosis of the trabecular bone in the palmar/plantar aspect of the condyles, endochondral ossification of the joint surface, and advancement of the tidemark to the articular surface , . These changes are associated with site-specific microdamage accumulation in calcified cartilage and the underlying subchondral bone of the parasagittal condylar grooves , , . Microcrack initiation occurs in the calcified cartilage layer  and stimulates a targeted remodeling response that results in the formation of resorption spaces containing activated osteoclasts in the damaged bone . This reparative response is associated with an increase in bone porosity, and may make horses more vulnerable to stress fracture if athletic activity is ongoing . Accumulation and coalescence of these microcracks leads to development of macroscopic crack arrays in the subchondral bone of the condylar grooves , , . Crack propagation through porous bone compromises the overlying cortical shell at the distal end of the MC3/MT3 bone , . Once the cortical shell of the distal end of the MC3/MT3 bone is mechanically compromised, crack propagation proximally along trabecular planes can easily develop, thereby rendering the horse at high risk of developing condylar stress fracture , .
Currently, there are differing opinions in the field regarding the etiology of fetlock breakdown injury in the racing Thoroughbred. The causative mechanism is likely complex and involves all of the joint structures, the third metacarpal or metatarsal, the proximal sesamoids, and the proximal phalanx. Very little is known about the mechanical stability of the fatigue cracks in the subchondral bone plate that precede development of a condylar stress fracture. It is likely that condylar stress fracture is preceded by cracks in the subchondral plate that become sufficiently severe to permit development of micromotion at the high joint loads associated with training or racing. Therefore, knowledge of the relationship between crack dimensions determined by cross-sectional imaging and crack micromotion at high load may facilitate early identification of horses that are at high risk of condylar stress fracture. Ultimately, this knowledge could help reduce the incidence of serious or catastrophic injuries at the racetrack.
Multiple imaging methods relevant to the distal limb are available to the equine clinician. Radiography is inferior to computed tomography (CT) and magnetic resonance imaging (MRI) for identification of pathologic features in subchondral bone and articular cartilage, such as fatigue cracks in the subchondral bone plate , . In addition to increased sensitivity for identifying pathologic changes in the articular surface of the fetlock joint, CT allows for multi-planar analysis of cross-sectional images of bone, facilitating a more accurate determination of size or extent of specific structures, such as subchondral fatigue cracks.
The objective of the present study was to develop an ex-vivo biomechanical model in which subchondral crack micromotion at high loads could be measured in distal MC3. We then used this model to determine the relationship between subchondral crack dimensions measured using CT and crack micromotion. We hypothesized that subchondral crack micromotion would be commonly detectable in the distal end of the MC3 bone in Thoroughbred racehorses at joint loads that model racing activity. Such a result would suggest that athletic Thoroughbred horses commonly train and race with incipient condylar stress fracture and are vulnerable to fracture propagation during athletic activity. A secondary objective was to study detection of fetlock joint abnormalities by CT imaging.
Thirty-six pairs of entire distal forelimbs and 4 individual limbs from 40 Thoroughbred racehorses that died or were euthanatized for reasons unrelated to the present study were given for use in this work (Fig. 1). Horses were euthanatized humanely by a veterinarian at the racetrack using an intravenous anesthetic overdose. Euthanasia was performed for clinical reasons because of serious injury during racing. Only thoracic limbs were used for this study. Condylar fractures most commonly affect the thoracic limbs , , , . Limbs were transected at the level of the carpus, sealed in plastic bags, and stored at -20C until needed. The age, gender, and racing history were collected from the Jockey Club Information Systems, Inc. Thoroughbred racehorse database (www.equineline.com) for horses with parasagittal subchondral cracks that were tested mechanically.
CT imaging was performed at high-resolution using 0.625 mm contiguous slices (64 slices GE Discovery CT750 HD, GE Healthcare Technologies, Waukesha, WI, USA). CT images were reconstructed in the sagittal, transverse, and frontal planes using a 64-bit Dicom viewer (OsiriX, Pixmeo SARL, Bernex, Switzerland). CT images were examined for evidence of pathological changes, including the presence of subchondral sclerosis, fractures, defects in the joint surface, and parasagittal fatigue cracks within the condylar grooves. The presence of palmar osteochondral disease (POD), initially referred to in the literature as traumatic osteochondrosis , , was defined by the presence of subchondral bone lysis in the palmar condyle area . The dimensions of subchondral cracks were measured radiographically on transverse and reconstructed frontal plane CT images (Fig. 2). In addition, crack area was measured in the parasagittal plane using image analysis (Image J, NIH) (Fig. 2). Using commercially available software (Mimics 13.1 Materialise, Ann Arbor, Michigan), the distal end of each MC3 bone with a parasagittal condylar groove lesion was segmented manually. A 3D volumetric model of the subchondral crack lesion was then created from the segmented images on a voxel-by-voxel basis to yield measurements of subchondral crack volume and the surface area of the subchondral crack volume.
Subchondral fatigue crack dimensions were measured in transverse (A) and reconstructed frontal plane (B) computed tomography (CT) images. In addition, the area of the crack array was measured in the parasagittal plane (C).
To study subchondral crack micromotion at high loads associated with athletic performance, we developed a non-destructive MC3 bone-loading model. An isolated single bone model was chosen for this work because of the high compressive load used for mechanical testing. Each bone was potted in an aluminum cylinder that was 10 cm long using epoxy (Evercoat 100156 Lite Weight Autobody Filler, Fiber Glass-Evercoat, Cincinnati, OH, USA). An 8 mm hole was drilled in the proximal end of the potted bone and the surrounding cylinder approximately 2 cm from the end of the cylinder, perpendicular to the long axis of the bone and the joint surface, and a stainless steel pin was inserted for additional stability.
Bone specimens without detectable CT lesions in the condylar grooves were used to validate the biomechanical model. Holes, 1.25 mm in diameter, were drilled at the distal palmar aspect of the sagittal ridge and 1 mm abaxial to the condylar groove in the same oblique frontal plate to span the region-of-interest. The potted bone specimen was then positioned in a custom jig (Fig. 3). The bones were oriented obliquely in the jig with the palmarodistal aspect facing up in order to mimic compressive loading by the proximal sesamoid bones from weight-bearing. Hypodermic needles (18 gauge) were placed in the drill holes and connected to an extensometer (MTS Systems Corp., Model 632-120-20, Eden Prairie, MN). Lateral and medial condyles of each bone were tested separately. As the bones were not tested destructively each specimen acted as it's own control. Each intact condyle was tested once, in order to obtain a baseline extensometer measurement before creation of an artificial subchondral slot to model an in-vivo fatigue crack.
(A) Drill holes 1.25 mm in diameter were made on the lateral and medial side of the artificial slot or crack array in the parasagittal condylar groove. (B) Hypodermic needles (18 g) were then placed in the drill holes and attached to an extensometer to measure motion across the slot or crack. The actuator consisted of a metal rod that was contoured to conform to the curved surface of the condyle. 2b1af7f3a8