The MCPJ is the most common site of musculoskeletal disease that results in reduced performance, premature retirement, catastrophic failure, and euthanasia of Thoroughbred racehorses in the United States, Canada, United Kingdom, and Hong Kong.1–12,a Review of results of a 15-year summary7 of Thoroughbred racehorses submitted to the California postmortem examination program revealed that 57% of submissions had catastrophic failure of the MCPJ, of which the affected structures included the PSBs and suspensory apparatus in 34% (651/1,915), MC3 in 19% (364/1,915), and the first phalanx in 4% (77/1,915). Results of necropsies performed in other studies indicate that biaxial (medial and lateral) PSB fracture is the most common cause for failure of the MCPJ for racehorses in California,7,13 Kentucky,2 Florida,4,b,c and Hong Konga and an uncommon cause for failure of the MCPJ for racehorses in the United Kingdom.10,14,15
Histologic evaluation has identified structural changes in the PSBs of trained versus untrained Thoroughbreds.16,17 Results of a study17 in which histomorphometric analysis was used to evaluate the PSBs indicate that bony material is more compacted in horses with biaxial PSB fracture, compared with that of horses without biaxial PSB fracture, and the investigators concluded that early identification of that structural change could provide an opportunity for prevention of PSB fractures.
The articulating surface of long bones is composed of many layers of tissues that are collectively referred to as osteochondral tissues.18 The superficial layers are composed of hyaline articular cartilage and subjacent to this is calcified cartilage followed by SCB layers that consist of the SCB plate and trabecular bone.18 Overload arthrosis can develop in the MCPJ18,19 at the distopalmar aspect of the MC3 condyles that is contacted by the PSBs during the stance phase of the gallop.20,21 The SCB damage to MC3 that results from overload arthrosis is a painful condition that is recognized in racehorses of all breeds20–28 and is believed to be a biomechanical disorder created by repetitive overload trauma from cyclic, high-intensity exercise.18–21 The distal aspect of the MC3 condyles is the most common site for abnormally increased radiopharmaceutical uptake.29 In a report24 of the postmortem findings for 64 Thoroughbred racehorses that were trained at the Hong Kong Jockey Club, 43 (67%) horses had SCB lesions in at least 1 MC3 condyle. In another study30 in which sMRI was used to evaluate the MCPJs and metatarsophalangeal joints of 131 Thoroughbred racehorses in training, 72 (55%) horses had SCB lesions in at least 1 MC3 condyle.
Radiography, ultrasonography, and gamma scintigraphy are the imaging modalities most frequently used to evaluate the MCPJs of Thoroughbred racehorses29,31,32; however, the number of racehorses euthanized because of MCPJ injuries continues to increase.7 The sensitivity of plain radiography for identification of stress fractures is poor with detection rates as low as 15%.33 Triple-phase bone scintigraphy is an extremely sensitive method for identification of stress injuries, but it lacks specificity because it cannot differentiate between a normal bone response to race training and a pathological bone response to mechanical stress.33 Although CT is accurate for characterizing fractures, it is less sensitive than bone scintigraphy or MRI for the early detection of stress injuries.33 The preferred imaging modality for the identification of stress fractures is MRI because bone abnormalities can be identified weeks before radiographic abnormalities develop, it has comparable sensitivity and superior specificity to bone scintigraphy, and it is the only imaging modality that can be used to identify bone marrow edema, which is pathognomonic for a stress response.32–35,d Results of multiple studies30,a–d indicate that MRI is valuable for the identification of subclinical pathological lesions and bone marrow edema in Thoroughbred racehorses, which suggests that MRI evaluation might have prognostic value and aid in the prevention of catastrophic MCPJ injuries in racehorses.
The purpose of the study reported here was to compare bony changes in the MCPJ of Thoroughbred racehorses with and without biaxial PSB fracture as determined by 2 grading scales applied to images of cadaveric forelimbs obtained by sMRI. The specific areas of interest for the grading scales were the SCB of the distopalmar aspect of the MC3 condyles and the associated PSBs. Our hypothesis was that bony changes would be more severe (ie, grades would be higher) in the PSBs of the affected limb and in the SCB of the MC3 of the contralateral limb of horses with biaxial PSB fracture, compared with those of horses without biaxial PSB fracture.
Materials and Methods
Samples—The forelimbs of Thoroughbred racehorses that were euthanized at racetracks in south Florida between September 15, 2011, and March 1, 2013 were harvested. All horses were in training or race condition at the time of euthanasia. A case was defined as a horse that was euthanized subsequent to a biaxial PSB fracture. A control was defined as a horse that was euthanized for a condition that was not associated with the MCPJ (eg, enteritis, pneumonitis, neurologic disease, laminitis, forelimb fractures that did not involve the MCPJ, or hind limb fractures).
Both forelimbs were harvested from cases and controls. For cases, the limb with the biaxial PSB fracture was referred to as the ipsilateral limb and the nonfractured limb was referred to as the contralateral limb. Although an sMRI examination was conducted on both forelimbs from each control, a single forelimb was used for statistical analysis and was referred to as a control limb.
It was assumed that the forelimb in which the biaxial PSB fracture occurred in each case was random. Therefore, the side of the ipsilateral limb was recorded as cases were identified, which created a random pattern list that was used to select which forelimb was used for statistical analysis from each of the controls. For example, if the order of the first 6 ipsilateral limbs enrolled was RF, RF, LF, RF, LF, and LF, that was the order of the first 6 control limbs enrolled. The pattern was repeated as necessary to select a control limb from each of the 53 controls.
For all cases, 1 investigator (PM) disarticulated both forelimbs through the middle carpal joint immediately after euthanasia. The selected forelimbs of the control horses were similarly disarticulated by the same investigator as soon as possible after euthanasia. The distal portion of all study forelimbs was packaged and sent cooled but unfrozen to the Equine Medical Center of Ocala.
sMRI evaluation—All images were obtained within 48 to 72 hours after euthanasia with an sMRI systeme by the same investigator (LH). During image acquisition, each limb was supported by a rope and pulley system. The rope was positioned at the proximal aspect of the MC3 to support the limb in a vertical position with the foot resting on the floor. The exact center of the MCPJ was identified in 3 orthogonal planes and positioned within the isocenter of the sMRI system. The radio-frequency coil was positioned so that images of the MCPJ were obtained from approximately 1 cm proximal to the apex of the PSB distally to the proximal 1 cm of the first phalanx. Images were obtained in the sagittal, dorsal, and transverse planes. The specific MRI sequence parameters and technical data were summarized (Appendix).
One investigator (JGP) assessed all sMRI images. The presence or absence of a biaxial PSB fracture precluded the investigator from being unaware (ie, blinded) of the status of the limbs. The sagittal image of the bone marrow from the unaffected first phalanx was used as the reference for the signal intensity of clinically normal PSBs. Hypointense was used to describe areas that were dark or produced a decrease in signal intensity, and hyperintense was used to describe areas that were light or produced an increase in signal intensity. Sclerosis is a radiographic term used to describe areas with increased bone mineral density, which on MRI images is represented by areas that are hypointense when compared with normal bone on non-fat suppressed images (ie, T1W GRE and T2W FSE).
A grading system with a scale of 0 to 3 was developed (Figure 1) to characterize the densification of the medial and lateral PSBs. Sagittal, dorsal, and transverse T1W images were assessed concurrently to subjectively estimate the percentage of the volume of each PSB with a hypointense signal on all pulse sequences, which provided a measure of the increase in the PSB density.
Representative T1W sMRI images of 3 equine forelimb MCPJs in the transverse (A, C, and E; dorsal is to the top) and sagittal (B, D, and F; dorsal is to the left) planes provided to help describe the PSB grading system developed for a study of bony changes in the forelimbs of Thoroughbred racehorses with and without catastrophic biaxial PSB fracture. For each pair of transverse and sagittal images that was obtained from the same forelimb (A and B, C and D, and E and F), the signal intensity of the bone marrow of the unaffected first phalanx in the sagittal image (black arrow) was used as the reference for the signal intensity of clinically normal bone. Hypointense was used to describe dark areas that had lower signal intensity, compared with the signal intensity of the reference. Hyperintense was used to describe light areas that had higher signal intensity, compared with the signal intensity of the reference. Sclerosis is a radiographic term that describes areas with an increase in bone mineral density and is represented as hypointense areas on MRI images. A and B—PSB grade, 0 (no increase in sclerosis or densification of the PSB [ie, clinically normal]). The PSB is composed primarily of non-sclerotic medullary bone (red arrows). The signal intensity of the medullary bone is derived from a combination of porous bone filled with fatty marrow. Notice that the signal intensity of the medullary bone of the PSB is hyperintense and has a similar intensity as the medullary bone of the metaphysis of the first phalanx in panel B. The signal intensity of the compact bone forming the SCB plate of MC3 and the cortex at the periphery of the PSB is hypointense (blue arrows). Reinforced trabecular bone along the palmar aspect of the PSB in the sagittal image (B) is also hypointense (white arrow), which reflects clinically normal sclerosis or densification at this location. A PSB grade of 1 was used to describe PSBs with < 50% densification at the palmar aspect (images not provided). C and D—PSB grade, 2 (≥ 50% to < 75% densification). Notice most of the PSB has a hypointense signal (blue arrows), compared with the reference area. In the transverse image (C), the remaining fatty marrow in the PSB created a hyperintense Y signal (white arrows), whereas densification and reinforcement with trabecular bone created a hypointense signal at the dorsal and palmar axial and abaxial aspects of the PSB. E and F—PSB grade, 3 (≥ 75% densification or continuous dorsal to palmar densification identified in sagittal slices). Notice that the PSB contains only a small amount of clinically normal bone (red arrow) because of densification and trabecular reinforcement in the marrow space (blue arrows). In panel E, only 1 PSB is visible because the other PSB was fractured and retracted from view (black arrow).
Citation: Journal of the American Veterinary Medical Association 246, 6; 10.2460/javma.246.6.661
Representative T1W sMRI images of 3 equine forelimb MCPJs in the transverse (A, C, and E; dorsal is to the top) and sagittal (B, D, and F; dorsal is to the left) planes provided to help describe the PSB grading system developed for a study of bony changes in the forelimbs of Thoroughbred racehorses with and without catastrophic biaxial PSB fracture. For each pair of transverse and sagittal images that was obtained from the same forelimb (A and B, C and D, and E and F), the signal intensity of the bone marrow of the unaffected first phalanx in the sagittal image (black arrow) was used as the reference for the signal intensity of clinically normal bone. Hypointense was used to describe dark areas that had lower signal intensity, compared with the signal intensity of the reference. Hyperintense was used to describe light areas that had higher signal intensity, compared with the signal intensity of the reference. Sclerosis is a radiographic term that describes areas with an increase in bone mineral density and is represented as hypointense areas on MRI images. A and B—PSB grade, 0 (no increase in sclerosis or densification of the PSB [ie, clinically normal]). The PSB is composed primarily of non-sclerotic medullary bone (red arrows). The signal intensity of the medullary bone is derived from a combination of porous bone filled with fatty marrow. Notice that the signal intensity of the medullary bone of the PSB is hyperintense and has a similar intensity as the medullary bone of the metaphysis of the first phalanx in panel B. The signal intensity of the compact bone forming the SCB plate of MC3 and the cortex at the periphery of the PSB is hypointense (blue arrows). Reinforced trabecular bone along the palmar aspect of the PSB in the sagittal image (B) is also hypointense (white arrow), which reflects clinically normal sclerosis or densification at this location. A PSB grade of 1 was used to describe PSBs with < 50% densification at the palmar aspect (images not provided). C and D—PSB grade, 2 (≥ 50% to < 75% densification). Notice most of the PSB has a hypointense signal (blue arrows), compared with the reference area. In the transverse image (C), the remaining fatty marrow in the PSB created a hyperintense Y signal (white arrows), whereas densification and reinforcement with trabecular bone created a hypointense signal at the dorsal and palmar axial and abaxial aspects of the PSB. E and F—PSB grade, 3 (≥ 75% densification or continuous dorsal to palmar densification identified in sagittal slices). Notice that the PSB contains only a small amount of clinically normal bone (red arrow) because of densification and trabecular reinforcement in the marrow space (blue arrows). In panel E, only 1 PSB is visible because the other PSB was fractured and retracted from view (black arrow).
Citation: Journal of the American Veterinary Medical Association 246, 6; 10.2460/javma.246.6.661
Representative T1W sMRI images of 3 equine forelimb MCPJs in the transverse (A, C, and E; dorsal is to the top) and sagittal (B, D, and F; dorsal is to the left) planes provided to help describe the PSB grading system developed for a study of bony changes in the forelimbs of Thoroughbred racehorses with and without catastrophic biaxial PSB fracture. For each pair of transverse and sagittal images that was obtained from the same forelimb (A and B, C and D, and E and F), the signal intensity of the bone marrow of the unaffected first phalanx in the sagittal image (black arrow) was used as the reference for the signal intensity of clinically normal bone. Hypointense was used to describe dark areas that had lower signal intensity, compared with the signal intensity of the reference. Hyperintense was used to describe light areas that had higher signal intensity, compared with the signal intensity of the reference. Sclerosis is a radiographic term that describes areas with an increase in bone mineral density and is represented as hypointense areas on MRI images. A and B—PSB grade, 0 (no increase in sclerosis or densification of the PSB [ie, clinically normal]). The PSB is composed primarily of non-sclerotic medullary bone (red arrows). The signal intensity of the medullary bone is derived from a combination of porous bone filled with fatty marrow. Notice that the signal intensity of the medullary bone of the PSB is hyperintense and has a similar intensity as the medullary bone of the metaphysis of the first phalanx in panel B. The signal intensity of the compact bone forming the SCB plate of MC3 and the cortex at the periphery of the PSB is hypointense (blue arrows). Reinforced trabecular bone along the palmar aspect of the PSB in the sagittal image (B) is also hypointense (white arrow), which reflects clinically normal sclerosis or densification at this location. A PSB grade of 1 was used to describe PSBs with < 50% densification at the palmar aspect (images not provided). C and D—PSB grade, 2 (≥ 50% to < 75% densification). Notice most of the PSB has a hypointense signal (blue arrows), compared with the reference area. In the transverse image (C), the remaining fatty marrow in the PSB created a hyperintense Y signal (white arrows), whereas densification and reinforcement with trabecular bone created a hypointense signal at the dorsal and palmar axial and abaxial aspects of the PSB. E and F—PSB grade, 3 (≥ 75% densification or continuous dorsal to palmar densification identified in sagittal slices). Notice that the PSB contains only a small amount of clinically normal bone (red arrow) because of densification and trabecular reinforcement in the marrow space (blue arrows). In panel E, only 1 PSB is visible because the other PSB was fractured and retracted from view (black arrow).
Citation: Journal of the American Veterinary Medical Association 246, 6; 10.2460/javma.246.6.661
A classification system36,37 developed for osteochondral lesions on the dome of the talus in human patients was modified by 2 investigators (JBV and JGP) to develop a grading system with a scale of 0 to 5 to characterize changes in the SCB of the distopalmar aspect of the medial and lateral condyles of MC3 (Figure 2). A separate grade was assigned to each condyle. The central feature for this system was the SCB plate of MC3. Sagittal, dorsal, and transverse T1W images were assessed concurrently to subjectively assign a grade to the SCB of MC3, although the sagittal images of the SCB of the condyles provided the best perspective for evaluation of the MC3 SCB plate. The area of interest was the contact area between the distopalmar aspect of the MC3 condyles and the PSBs during the stance phase of the gallop, specifically the SCB of MC3 located 5 to 8 mm palmar and proximal to the transverse ridge and 3 to 15 mm abaxial to the sagittal ridge on either the medial or lateral condyle.20,21
Representative T1W sMRI images obtained in the sagittal plane for 6 equine forelimb MCPJs provided to help describe the MC3 SCB grading system developed for a study of bony changes in the forelimbs of Thoroughbred racehorses with and without catastrophic biaxial PSB fracture. This grading system focused on the contact area of the PSB with the palmar aspect of the condyles of the MC3 during the stance phase of the gallop (red arrows in panel A). In each image, dorsal is to the left, MC3 is at the top, and the first phalanx is at the bottom. A—MC3 SCB grade, 0. No abnormalities were identified in the SCB plate or the signal in the SCB subjacent to the SCB plate. The SCB plate (blue arrows) curves from the dorsal to palmar aspect of MC3 and is a continuation of the dense cortical bone of the distal portion of MC3. It creates a hypointense signal that is sandwiched between the extremely hyperintense signal of the superficial articular cartilage of MC3 and the subjacent medullary trabecular bone and fatty marrow. B—MC3 SCB grade, 1. The SCB plate of MC3 has normal signal intensity and morphology, but there is an increase in densification of the subjacent medullary bone marrow (red arrows), compared with the subjacent medullary SCB marrow of MC3 in panel A, and a somewhat linear area of hyperintensity immediately subjacent to the SCB plate (blue arrow). Joint fluid does not extend into the SCB plate, which implies that the overlying articular cartilage of MC3 is intact. C—MC3 SCB grade, 2. Notice that the SCB plate of MC3 has an abnormal hyperintensity and a contour deformity (red arrow) that is suggestive of focal compression or fracturing of the SCB plate or resorption or fracturing of the subjacent trabecular bone. Joint fluid does not extend into the SCB plate, which implies that the overlying articular cartilage is intact. Also, there is an increase in the densification of the subjacent SCB marrow. D—MC3 SCB grade, 3. Notice that there is a full-thickness defect in the SCB plate (red arrow) with a signal intensity similar to that of joint fluid, which implies that the overlying articular cartilage is not intact, and an increase in the densification of the subjacent SCB marrow. Joint fluid that extends into the subjacent bone marrow may cause cyst formation. E—MC3 SCB grade, 4. Notice that there is a partial or complete detachment of an osteochondral fragment that is not displaced (red arrows) and an increase in the densification of the subjacent SCB marrow. In this particular image, the signal intensity around the fragment is not as hyperintense as that of joint fluid, which implies that the overlying articular cartilage is still intact. F—MC3 SCB grade, 5. Notice that an osteochondral fragment is displaced from the original defect, which has left a large area of the SCB of MC3 exposed (red arrows), and there is an increase in the densification of the subjacent SCB marrow.
Citation: Journal of the American Veterinary Medical Association 246, 6; 10.2460/javma.246.6.661
Representative T1W sMRI images obtained in the sagittal plane for 6 equine forelimb MCPJs provided to help describe the MC3 SCB grading system developed for a study of bony changes in the forelimbs of Thoroughbred racehorses with and without catastrophic biaxial PSB fracture. This grading system focused on the contact area of the PSB with the palmar aspect of the condyles of the MC3 during the stance phase of the gallop (red arrows in panel A). In each image, dorsal is to the left, MC3 is at the top, and the first phalanx is at the bottom. A—MC3 SCB grade, 0. No abnormalities were identified in the SCB plate or the signal in the SCB subjacent to the SCB plate. The SCB plate (blue arrows) curves from the dorsal to palmar aspect of MC3 and is a continuation of the dense cortical bone of the distal portion of MC3. It creates a hypointense signal that is sandwiched between the extremely hyperintense signal of the superficial articular cartilage of MC3 and the subjacent medullary trabecular bone and fatty marrow. B—MC3 SCB grade, 1. The SCB plate of MC3 has normal signal intensity and morphology, but there is an increase in densification of the subjacent medullary bone marrow (red arrows), compared with the subjacent medullary SCB marrow of MC3 in panel A, and a somewhat linear area of hyperintensity immediately subjacent to the SCB plate (blue arrow). Joint fluid does not extend into the SCB plate, which implies that the overlying articular cartilage of MC3 is intact. C—MC3 SCB grade, 2. Notice that the SCB plate of MC3 has an abnormal hyperintensity and a contour deformity (red arrow) that is suggestive of focal compression or fracturing of the SCB plate or resorption or fracturing of the subjacent trabecular bone. Joint fluid does not extend into the SCB plate, which implies that the overlying articular cartilage is intact. Also, there is an increase in the densification of the subjacent SCB marrow. D—MC3 SCB grade, 3. Notice that there is a full-thickness defect in the SCB plate (red arrow) with a signal intensity similar to that of joint fluid, which implies that the overlying articular cartilage is not intact, and an increase in the densification of the subjacent SCB marrow. Joint fluid that extends into the subjacent bone marrow may cause cyst formation. E—MC3 SCB grade, 4. Notice that there is a partial or complete detachment of an osteochondral fragment that is not displaced (red arrows) and an increase in the densification of the subjacent SCB marrow. In this particular image, the signal intensity around the fragment is not as hyperintense as that of joint fluid, which implies that the overlying articular cartilage is still intact. F—MC3 SCB grade, 5. Notice that an osteochondral fragment is displaced from the original defect, which has left a large area of the SCB of MC3 exposed (red arrows), and there is an increase in the densification of the subjacent SCB marrow.
Citation: Journal of the American Veterinary Medical Association 246, 6; 10.2460/javma.246.6.661
Representative T1W sMRI images obtained in the sagittal plane for 6 equine forelimb MCPJs provided to help describe the MC3 SCB grading system developed for a study of bony changes in the forelimbs of Thoroughbred racehorses with and without catastrophic biaxial PSB fracture. This grading system focused on the contact area of the PSB with the palmar aspect of the condyles of the MC3 during the stance phase of the gallop (red arrows in panel A). In each image, dorsal is to the left, MC3 is at the top, and the first phalanx is at the bottom. A—MC3 SCB grade, 0. No abnormalities were identified in the SCB plate or the signal in the SCB subjacent to the SCB plate. The SCB plate (blue arrows) curves from the dorsal to palmar aspect of MC3 and is a continuation of the dense cortical bone of the distal portion of MC3. It creates a hypointense signal that is sandwiched between the extremely hyperintense signal of the superficial articular cartilage of MC3 and the subjacent medullary trabecular bone and fatty marrow. B—MC3 SCB grade, 1. The SCB plate of MC3 has normal signal intensity and morphology, but there is an increase in densification of the subjacent medullary bone marrow (red arrows), compared with the subjacent medullary SCB marrow of MC3 in panel A, and a somewhat linear area of hyperintensity immediately subjacent to the SCB plate (blue arrow). Joint fluid does not extend into the SCB plate, which implies that the overlying articular cartilage of MC3 is intact. C—MC3 SCB grade, 2. Notice that the SCB plate of MC3 has an abnormal hyperintensity and a contour deformity (red arrow) that is suggestive of focal compression or fracturing of the SCB plate or resorption or fracturing of the subjacent trabecular bone. Joint fluid does not extend into the SCB plate, which implies that the overlying articular cartilage is intact. Also, there is an increase in the densification of the subjacent SCB marrow. D—MC3 SCB grade, 3. Notice that there is a full-thickness defect in the SCB plate (red arrow) with a signal intensity similar to that of joint fluid, which implies that the overlying articular cartilage is not intact, and an increase in the densification of the subjacent SCB marrow. Joint fluid that extends into the subjacent bone marrow may cause cyst formation. E—MC3 SCB grade, 4. Notice that there is a partial or complete detachment of an osteochondral fragment that is not displaced (red arrows) and an increase in the densification of the subjacent SCB marrow. In this particular image, the signal intensity around the fragment is not as hyperintense as that of joint fluid, which implies that the overlying articular cartilage is still intact. F—MC3 SCB grade, 5. Notice that an osteochondral fragment is displaced from the original defect, which has left a large area of the SCB of MC3 exposed (red arrows), and there is an increase in the densification of the subjacent SCB marrow.
Citation: Journal of the American Veterinary Medical Association 246, 6; 10.2460/javma.246.6.661
Data collection—All information about past performance was collected from an equine database of racing information and statisticsf The following variables were recorded for each horse: name, confidential study identification number, age, sex (gelding, colt, or mare), date of race, case or control status, study limb (LF or RF) and whether it was fractured (yes or no), whether the horse had experienced a change in class from its previous race (higher, lower, or no change), number of career starts and wins, lifetime earnings, and the reason for euthanasia.
Statistical analysis—Three additional variables were created for analysis purposes. A variable called total PSB grade was created because PSB fracture in Thoroughbred racehorses frequently occurs biaxially. This variable represented the sum of the grades assigned to the medial and lateral PSBs of a given limb. We decided a priori to dichotomize the data such that limbs with a PSB grade 3 were compared with those with a grade < 3, and limbs with a total PSB grade ≥ 5 were compared with those with a total PSB grade < 5. We chose ≥ 5 as the cutoff for the total PSB grade because we wanted at least one of the PSBs in each of the limbs in the ≥ 5 group to have a PSB grade of 3.
Similar to the total PSB grade, a variable called total MC3 SCB grade was created that represented the sum of the SCB grades for the medial and lateral condyles of the MC3 of a given limb. We decided a priori to dichotomize the data such that individual condyles with an SCB grade ≥ 3 were compared with those with an SCB grade < 3, and limbs with a total MC3 SCB grade ≥ 5 were compared with those with a total MC3 SCB grade < 5. We chose ≥ 5 as the cutoff for the total MC3 SCB grade because we wanted at least 1 condyle in each of the limbs in the ≥ 5 group to have a SCB grade of 3.
A variable called contralateral MC3 orthopedic disease was also created because 3 cases had substantial pathological lesions in the contralateral MC3 that were not adequately described or characterized by the SCB grading system. Horses classified as having contralateral MC3 orthopedic disease included those 3 cases and all horses with a total MC3 SCB grade ≥ 5.
Data were analyzed with descriptive and inferential methods. For descriptive purposes, frequency counts were provided for categorical variables and medians and ranges were provided for continuous variables. Comparisons between cases and controls were made by the use of Wilcoxon rank sum tests for continuous variables and χ2 tests for categorical variables. Age and the number of starts were evaluated both as continuous and dichotomous variables. Age was dichotomized as < 3 years old and ≥ 3 years old, and the number of starts was dichotomized as < 10 starts and ≥ 10 starts. Logistic regression was used to identify variables associated with biaxial PSB fracture. Independent variables included in the regression models included sex, age, number of starts, PSB grade, and MC3 SCB grade. The magnitude of the associations between the respective independent variables and biaxial PSB fracture was summarized with ORs and 95% CIs that were calculated with maximum likelihood methods. All analyses were performed with statistical software,g and values of P ≤ 0.05 were considered significant.
Results
Animals—During the period between September 15, 2011, and March 1, 2013, 74 horses were euthanized at the Florida racetrack that met the criteria for study enrollment. Twenty-one horses were classified as cases, of which 20 had biaxial PSB fracture and 1 had a uniaxial fracture of the lateral PSB and a uniaxial fracture of the base of the medial PSB with rupture of the distal sesamoidean ligament. The affected, or ipsilateral, limb was the LF in 11 of the cases and the RF in the remaining 10 cases. Fifty-three control horses were enrolled in the study, from which the LF was evaluated for 28 and the RF was evaluated for the remaining 25.
The cases included 16 (76%) males (8 geldings and 8 colts) and 5 (24%) mares, and the controls included 25 (47%) males (11 geldings and 14 colts) and 28 (53%) mares. The proportion of males in the case group was significantly (P = 0.045) greater than the proportion of males in the control group. The odds of horses being male were 3.6 times as high for cases as for controls (OR, 3.6, 95% CI, 1.1 to 11.2; Table 1).
Results of univariable logistic regression analyses to identify variables associated with biaxial PSB fracture in the forelimbs of Thoroughbred racehorses that were euthanized at Florida racetracks between September 15, 2011, and March 1, 2013.
| Variable | OR | 95% CI | P value |
|---|---|---|---|
| Sex | |||
| Male | 3.6 | 1.1–11.2 | 0.031 |
| Age | |||
| Per year | 1.4 | 1.1–2.0 | 0.031 |
| < 3 y | Referent | — | — |
| ≥ 3 y | 2.4 | 0.8–6.9 | 0.106 |
| Starts | |||
| Per start | 1.05 | 1.01–1.09 | 0.010 |
| < 10 | Referent | — | — |
| ≥ 10 | 2.6 | 0.9–7.5 | 0.080 |
| PSB grade for ipsilateral and control limbs | |||
| Lateral PSB | |||
| < 3 | Referent | — | — |
| 3 | 12.5 | 2.9–53.2 | 0.001 |
| Medial PSB | |||
| < 3 | Referent | — | — |
| 3 | 7.5 | 2.0–29.0 | 0.004 |
| Total PSB grade for ipsilateral and control limbs | |||
| < 5 | Referent | — | — |
| ≥ 5 | 10.4 | 3.1–35.1 | < 0.001 |
| MC3 SCB grade for contralateral and control limbs | |||
| Lateral condyle | |||
| < 3 | Referent | — | — |
| ≥ 3 | Inestimable | Inestimable | Inestimable |
| Medial condyle | |||
| < 3 | Referent | — | — |
| ≥ 3 | 8.7 | 0.8–88.5 | 0.073 |
| Total MC3 SCB grade for contralateral and control limbs | |||
| < 5 | Referent | — | — |
| ≥ 5 | 20.8 | 2.3–186.3 | 0.008 |
| Presence of orthopedic disease in contralateral MC3* | 39.0 | 4.6–331.8 | 0.001 |
The study was a case-control study; cases (n = 21) were horses with biaxial PSB fracture of the forelimb, and controls (53) were horses that were euthanized for a condition that was not associated with the MCpJ (eg, enteritis, pneumonitis, neurologic disease, laminitis, forelimb fractures that did not involve the MCPJ, or hind limb fractures). Both forelimbs were harvested from cases and controls, but only 1 randomly selected forelimb from each control was used for statistical analysis. The limb with the biaxial PSB fracture was referred to as the ipsilateral limb, the nonfractured limb was referred to as the contralateral limb, and the forelimbs from controls were referred to as control limbs. The PSB grade was subjectively assigned to each PSB within a given limb on a scale of 0 to 3, with 0 indicative of no abnormalities and 3 indicative of ≥ 75% densification or continuous dorsal to palmar densification identified in sagittal sMRI images. The MC3 SCB grade was subjectively assigned to the distopalmar aspect of the lateral and medial condyles of each MC3 on a scale of 0 to 5, with 0 indicative of no abnormalities and 5 indicative of complete detachment and displacement of an osteochondral fragment, which may or may not be observed as a loose body within the MCPJ, along with an increase in the densification of the subjacent SCB marrow.
Includes all horses with a total MC3 SCB grade ≥ 5 and 3 cases with pathological lesions of the contralateral MC3 that were not adequately characterized by the MC3 SCB grade.
— = Not applicable.
The median age of the cases (5 years; range, 2 to 8 years) was significantly (P = 0.036) greater than that of the controls (3 years; range, 2 to 8 years). However, the proportion of cases ≥ 3 years old (14/21 [67%]) did not differ significantly (P = 0.161) from the proportion of controls ≥ 3 years old (24/53 [45%]). Although the odds of a horse having a biaxial PSB fracture increased significantly (P = 0.031) with each year of age (OR, 1.4; 95% CI, 1.1 to 2.0), the odds of a horse ≥ 3 years old having a biaxial PSB fracture did not differ significantly (P = 0.106) from those for a horse < 3 years old (OR, 2.4, 95% CI, 0.8 to 6.9; Table 1).
The median number of race starts for the cases (16 starts; range, 0 to 46 starts) was significantly (P = 0.012) greater than the median number of race starts for the controls (6 starts; range, 0 to 53 starts). However, the proportion of cases with ≥ 10 starts (14/21 [67%]) did not differ significantly (P = 0.122) from the proportion of controls with ≥ 10 starts (23/53 [43%]). Although the odds of a horse having a biaxial PSB fracture increased significantly (P = 0.010) with each start (OR, 1.01; 95% CI, 1.01 to 1.09), the odds for a horse with ≥ 10 race starts having a biaxial PSB fracture did not differ significantly (P = 0.080) from those for a horse with < 10 race starts (OR, 2.6; 95% CI, 0.9 to 7.5; Table 1).
sMRI evaluation—The presence of STIR or T2*W signal increases did not differ significantly between cases and controls. Three cases had pronounced pathological lesions of the distal aspect of the contralateral MC3 that could not be characterized by the MC3 SCB grading system. One horse had a prominent fracture of the medial condyle that was considered chronic because of the absence of an abnormally hyperintense signal within the bone surrounding the fracture line on the STIR images. Evaluation of the STIR images for each of the other 2 horses revealed large, localized areas of hyperintense signal in the distal aspect of the contralateral MC3 that were dissimilar to the signal produced by bone marrow edema associated with condylar fractures previously described.30,b,c
PSB grades for the ipsilateral limb—The frequency distributions of the PSB grades for the ipsilateral and control limbs were summarized (Table 2). The distribution of the grades for the lateral PSB differed significantly (P = 0.002) between the cases and controls. The proportion of cases with a lateral PSB grade of 3 (9/21 [43%]) was significantly (P < 0.001) greater than the proportion of controls with a lateral PSB grade of 3 (3/53 [6%]). Horses with a lateral PSB grade of 3 were 12.5 times as likely to have a biaxial PSB fracture as were horses with a lateral PSB grade < 3 (OR, 12.5; 95% CI, 2.9 to 53.2; P = 0.001; Table 1).
Frequency distribution of PSB grades for the lateral and medial PSBs and total PSB grade (sum of the PSB grades for the lateral and medial PSBs) for the horses of Table 1.
| Lateral PSB | Medial PSB | Total PSB grade | ||||
|---|---|---|---|---|---|---|
| Grade | Case | Control | Case | Control | Case | Control |
| 0 | 0 (0) | 4 (8) | 1 (5) | 5 (9) | 0 (0) | 4 (8) |
| 1 | 4 (19) | 28 (53) | 4 (19) | 26 (49) | 0 (0) | 1 (2) |
| 2 | 8 (38) | 18 (34) | 8 (38) | 18 (34) | 3 (14) | 25(47) |
| 3 | 9 (43) | 3 (6) | 8 (38) | 4 (8) | 2 (10) | 3 (6) |
| 4 | — | — | — | — | 4 (19) | 14 (26) |
| 5 | — | — | — | — | 9 (43) | 5 (9) |
| 6 | — | — | 3 (14) | 1 (2) | ||
Values represent the number (%) of horses. Grades reported for the cases are for the ipsilateral limb. Within a column, the percentages may not sum to 100 because of rounding.
See Table 1 for remainder of key.
The distribution of the grades for the medial PSB differed significantly (P = 0.002) between cases and controls. The proportion of cases with a medial PSB grade of 3 (8/21 [38%]) was significantly (P = 0.004) greater than the proportion of controls with a medial PSB grade of 3 (4/53 [8%]). Horses with a medial PSB grade of 3 were 7.5 times (95% CI, 2.0 to 29.0; P = 0.004) as likely to have a biaxial PSB fracture as were horses with a medial PSB grade < 3 (Table 1).
The distribution of the total PSB grades for the ipsilateral limb differed significantly (P = 0.032) from the total PSB grades for control limbs. The proportion of cases with a total PSB grade for the ipsilateral limb ≥ 5 (12/21 [57%]) was significantly (P < 0.001) greater than the proportion of controls with a total PSB grade ≥ 5 (6/53 [11%]). Horses with a total PSB grade for the ipsilateral limb ≥ 5 were 10.4 times (95% CI, 3.1 to 35.1; P < 0.001) as likely to have a biaxial PSB fracture as were horses with a total PSB grade < 5 (Table 1).
PSB grades for the contralateral limb—The PSB grades for the contralateral limb of the cases were generally higher than those for the controls (data not shown). However, the lateral (P = 0.238) and medial (P = 0.202) PSB grades and total PSB grades (P = 0.268) for the contralateral and control limbs did not differ significantly.
MC3 SCB grades for the ipsilateral limb—The MC3 SCB grades for the ipsilateral limbs were generally higher, compared with the MC3 SCB grades for the control limbs. However, the SCB grades for the lateral (P = 0.646) and medial (P = 0.629) condyles and the total MC3 SCB grades (P = 0.914) did not differ significantly between the ipsilateral and control limbs.
MC3 SCB grades for the contralateral limb—The frequency distributions for the MC3 SCB grades for the contralateral and control limbs were summarized (Table 3). The distribution of the SCB grades for the lateral condyle of MC3 differed significantly (P = 0.003) between the cases and controls. The proportion of cases with SCB grades for the lateral condyle ≥ 3 (6/21[29%]) was significantly (P < 0.001) greater than the proportion of controls with SCB grades for the lateral condyle ≥ 3 (0/53 [0%]).
Frequency distribution of the SCB grades for the distopalmar aspect of the lateral and medial condyles of MC3 and the total MC3 SCB grade (sum of the SCB grades for the lateral and medial condyles of MC3) for the horses of Table 1.
| Lateral condyle | Medial condyle | Total MC3 SCB grade | ||||
|---|---|---|---|---|---|---|
| Grade | Case | Control | Case | Control | Case | Control |
| 0 | 6 (29) | 41 (77) | 6 (29) | 40 (75) | 5 (24) | 37 (70) |
| 1 | 5 (24) | 9 (17) | 5 (24) | 10 (19) | 2 (10) | 7 (13) |
| 2 | 4 (19) | 3 (6) | 7 (33) | 2 (4) | 2 (10) | 5 (9) |
| 3 | 5 (24) | 0 (0) | 3 (14) | 1 (2) | 4 (19) | 2 (4) |
| 4 | 1 (5) | 0 (0) | 0 (0) | 0 (0) | 2 (10) | 1 (2) |
| 5 | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 2 (10) | 1 (2) |
| 6 | — | — | — | — | 4 (19) | 0 (0) |
| 7 | — | — | — | — | 0 (0) | 0 (0) |
| 8 | — | — | — | — | 0 (0) | 0 (0) |
| 9 | — | — | — | — | 0 (0) | 0 (0) |
| 10 | — | — | — | 0 (0) | 0 (0) | |
Grades reported for the cases are for the contralateral limb.
See Tables 1 and 2 for remainder of key.
The distribution of the SCB grades for the medial condyle differed significantly (P = 0.002) between the cases and controls. However, the proportion of cases with an SCB grade ≥ 3 for the medial condyle (3/21 [14%]) did not differ significantly (P = 0.120) from the proportion of controls with an SCB grade ≥ 3 for the medial condyle (1/53 [2%]). Horses with an SCB grade ≥ 3 for the medial condyle were 8.7 times (95% CI, 0.8 to 88.5) as likely to have a biaxial PSB fracture as were horses with a SCB grade for the medial condyle < 3, although that association was not significant (P = 0.073; Table 1).
The distribution of the total MC3 SCB grades differed significantly (P = 0.014) between contralateral and control limbs. The proportion of contralateral limbs with total MC3 SCB grades ≥ 5 (6/21 [29%]) was significantly (P = 0.002) greater than the proportion of control limbs with a total MC3 SCB grade ≥ 5 (1/53 [2%]). Horses with a total MC3 SCB grade ≥ 5 for the contralateral limb were 20.8 times (95% CI, 2.3 to 186.3; P = 0.008) as likely to have a biaxial PSB fracture as were horses with a total MC3 SCB grade < 5 for the contralateral limb (Table 1). The odds of horses having a total ipsilateral PSB grade ≥ 5 were significantly (P < 0.001) greater for horses with a total contralateral MC3 SCB grade ≥ 5, compared with the odds for horses with MCS SCB grade < 5 (OR, 3.7; 95% CI, 1.9 to 7.2).
Contralateral MC3 orthopedic disease—The proportion of cases with contralateral MC3 orthopedic disease (9/21 [43%]) was significantly (P < 0.001) greater than the proportion of controls with contralateral MC3 orthopedic disease (1/53 [2%]). Horses with contralateral MC3 orthopedic disease were 39.0 times (95% CI, 4.6 to 331.8; P = 0.001) as likely to have a biaxial PSB fracture as were horses without contralateral MC3 orthopedic disease (Table 1).
Multivariable logistic regression analyses—Sex was not significantly associated with biaxial PSB fracture in any of the multivariable logistic regression models assessed. A strong positive correlation existed between age and the number of starts; therefore, the multivariable logistic regression models included either age or the number of starts, but not both variables. The results of the multivariable models that included age were similar to the results of the multivariable models that included number of starts. We chose to report the results for the models that included the number of starts because we believed that variable was more clinically relevant than age. Similarly, there was a strong correlation between total MC3 SCB grade for the contralateral limb and the presence of contralateral MC3 orthopedic disease (the horses with contralateral MC3 orthopedic disease included all horses with a total MC3 SCB grade ≥ 5 for the contralateral limb plus 3 cases with substantial pathological lesions of the contralateral MC3 that were not adequately characterized by the MC3 SCB grading system); therefore, the multivariable logistic regression models included either total MC3 SCB grade or the presence of contralateral MC3 orthopedic disease, but not both variables.
Results of the various multivariable logistic regression models assessed were summarized (Table 4). Total PSB grade and the presence of contralateral MC3 orthopedic disease were significantly associated with biaxial PSB fracture in all multivariable models in which each of those respective variables were assessed. Total MC3 SCB grade was significantly associated with biaxial PSB fracture in all multivariable models in which it was assessed except the model that included total MC3 SCB grade for the contralateral limb, total PSB grade, and number of starts modeled as a dichotomized variable. Number of starts was not significantly associated with biaxial PSB fracture in any of the multivariable models assessed except the model that included total PSB grade and number of starts modeled as a continuous variable. Inclusion of number of starts in a model had a negligible effect on the magnitude of the OR for the other variables assessed in that model regardless of whether it was modeled as a continuous or dichotomous variable; therefore, number of starts was not considered a confounder. For each multivariable model assessed, all possible 2- and 3-way interaction terms were assessed, but none were found to be significantly associated with biaxial PSB fracture.
Results of multivariable logistic regression analyses to identify variables associated with biaxial PSB fracture in the forelimbs of the horses of Table 1.
| Model | Variable | OR | 95% CI | P value |
|---|---|---|---|---|
| Association of biaxial PSB fracture with total PSB grade and total MC3 SCB grade | Total PSB grade | |||
| < 5 | Referent | — | — | |
| ≥ 5 | 7.8 | 2.2–27.9 | 0.003 | |
| Total MC3 SCB grade | ||||
| < 5 | Referent | — | — | |
| ≥ 5 | 11.6 | 1.1–119.6 | 0.044 | |
| Association of biaxial PSB fracture with total PSB grade and contralateral MC3 orthopedic disease | Total PSB grade | |||
| < 5 | Referent | — | — | |
| ≥ 5 | 8.4 | 2.1–33.3 | 0.003 | |
| Contralateral MC3 orthopedic disease present* | 30.7 | 3.2–297.3 | 0.004 | |
| Association of biaxial PSB fracture with total PSB grade and number of starts (dichotomized) | Total PSB grade | |||
| < 5 | Referent | — | — | |
| ≥ 5 | 11.1 | 3.1–39.5 | < 0.001 | |
| No. of starts | ||||
| < 10 | Referent | — | — | |
| ≥ 10 | 2.9 | 0.9–9.9 | 0.088 | |
| Association of biaxial PSB fracture with total MC3 SCB grade and number of starts (dichotomized) | Total MC3 SCB grade | |||
| < 5 | Referent | — | — | |
| ≥ 5 | 16.5 | 1.7–159.0 | 0.018 | |
| No. of starts | ||||
| < 10 | Referent | — | — | |
| ≥ 10 | 1.6 | 0.5–4.9 | 0.453 | |
| Association of biaxial PSB fracture with total PSB grade, total MC3 SCB grade, and number of starts (dichotomized) | Total PSB grade | |||
| < 5 | Referent | — | — | |
| ≥ 5 | 8.5 | 2.3–32.0 | 0.002 | |
| Total MC3 SCB grade | ||||
| < 5 | Referent | — | — | |
| ≥ 5 | 8.0 | 0.7–91.5 | 0.098 | |
| No. of starts | ||||
| < 10 | Referent | — | — | |
| ≥ 10 | 2.0 | 0.5–7.2 | 0.299 | |
| Association of biaxial PSB fracture with total PSB grade, contralateral MC3 orthopedic disease, and number of starts (dichotomized) | Total PSB grade | |||
| < 5 | Referent | — | — | |
| ≥ 5 | 9.1 | 2.2–37.2 | 0.003 | |
| Contralateral MC3 orthopedic disease present* | 26.3 | 2.6–267.2 | 0.007 | |
| No. of starts | ||||
| < 10 | Referent | — | — | |
| ≥ 10 | 1.6 | 0.4–6.4 | 0.473 | |
| Association of biaxial PSB fracture with total PSB grade and number of starts (continuous) | Total PSB grade | |||
| < 5 | Referent | — | — | |
| ≥ 5 | 10.3 | 2.9–37.1 | < 0.001 | |
| No. of starts (per start) | 1.05 | 1.01–1.09 | 0.024 | |
| Association of biaxial PSB fracture with total MC3 SCB grade and number of starts (continuous) | Total MC3 SCB grade | |||
| < 5 | Referent | — | — | |
| ≥ 5 | 11.4 | 1.0–126.2 | 0.050 | |
| No. of starts (per start) | 1.02 | 0.98–1.07 | 0.256 | |
| Association of biaxial PSB fracture with contralateral MC3 orthopedic disease and number of starts (continuous) | Contralateral MC3 orthopedic disease present* | 27.8 | 3.1–249.8 | 0.004 |
| No. of starts (per start) | 1.02 | 0.98–1.07 | 0.264 | |
For all models, the total PSB grade is for ipsilateral or control limbs and the total MC3 SCB grade is for contralateral or control limbs.
See Table 1 for remainder of key.
Discussion
In the present study, new grading systems to assess the PSBs and the SCB of the MC3 of horses on MRI images were developed, described, and used to assess changes in the MCPJ of Thoroughbred racehorses with (cases) and without (controls) catastrophic biaxial PSB fracture. Results supported the hypothesis that cases had more severe bony changes associated with the MCPJ as determined by sMRI than did controls. For the cases, the ipsilateral (affected) limb had more extensive PSB densification and the contralateral limb had more severe MC3 SCB or other orthopedic disease, compared with the limbs of the controls. Bony changes in the MCPJ observed on images obtained by sMRI may be useful for the identification of Thoroughbred racehorses at risk of developing a catastrophic biaxial PSB fracture.
Disease of the SCB on the distopalmar aspect of MC3 or distoplantar aspect of the third metatarsal bone is common in Thoroughbred racehorses.18–21,24,29 The area most commonly affected is the contact area between the distopalmar aspect of MC3 or distoplantar aspect of the third metatarsal bone and the PSBs during the stance phase of the gallop.20,21 In the present study, MC3 SCB lesions were identified more frequently in both the lateral and medial condyles of cases, compared with the lateral and medial condyles of controls. These findings are in agreement with those of other studies that involved clinical,30 scintigraphic,29 and postmortem24 examination of the MCPJ of horses. Results of the study24 in which the MCPJs were examined during necropsy indicate a relationship between the severity of MC3 SCB disease and other pathological changes associated with generalized osteoarthritis.
For the present study, the human tibiotalar joint was used as the template for the equine MCPJ because those joints have 3 distinct similarities that create analogous limitations when imaged with MRI. The articular cartilage layers of both the human tibiotalar joint (mean thickness, 1.1 mm)38 and equine MCPJ (mean thickness, 0.79 mm)39 are thin. The tibial plafond and the talar dome of the human tibiotalar joint and the distal aspect of the equine MC3 and first phalanx have similar curved surfaces.36–38 Finally, the lack of physical separation between adjacent cartilage surfaces in both the human tibiotalar joint and equine MCPJ makes it difficult to identify the opposing cartilage surfaces as distinct structures.38–41 It is recommended that suspected cartilage lesions in the human tibiotalar joint38 and equine MCPJ39–42 be evaluated with a 3-T MRI system because the aforementioned functional and structural characteristics of the cartilage layers and the imaging difficulties they produce make imaging of those joints fundamentally susceptible to volume-averaging artifacts. Results of 1 study40 indicate that identification of cartilage lesions in equine MCPJs by evaluation of images obtained by high-field (1.5-T) MRI or sMRI systems was much less accurate than histologic examination. In another study,41 the sensitivity and specificity of 3-T MRI for identification of articular cartilage defects in the MCPJ of non-Thoroughbreds were 41% and 93%, respectively, compared with gross anatomic evaluation. Investigators of those studies,39–42 recognizing the limitations of the use of 1.5-T and sMRI systems for identification of cartilage lesions in the equine MCPJ, did not describe a grading scale for articular cartilage changes.
Although grading scales have been developed to assess the severity of osteochondral disease during evaluation of MRI images for most major joints of human patients,35,43,a an MRI classification scheme for lesions of the SCB on the distopalmar aspect of the MC3 in horses has not been described. In the present study, we developed and described a grading scale to assess SCB changes in the distal portion of the MC3 that focused on the SCB plate. The SCB plate of the MC3 is a dense, discrete, and uniform structure with a low signal intensity (ie, hypointense) that is sandwiched between the hyperintense signals of articular cartilage and subjacent trabecular bone and is easily identified in sagittal plane T1W GRE, T2W FSE, and T2*W GRE images. For the MC3 SCB grading system developed for the present study, a grade ≥ 3 indicated that there was a breach in the integrity of the SCB plate and damage to the overlying articular cartilage. In the present study, we used a total MC3 SCB grade ≥ 5 as the cutoff for analysis purposes because that would mean that at least one of the condyles of MC3 had a grade ≥ 3, which is suggestive of exposed SCB, which in turn causes signs of pain, lameness, and arthritis that affect the entire MCPJ. The proposed MC3 SCB grading system described for the present study could be used to assess MRI images of the MCPJ of horses with lameness localized to that joint to help make recommendations regarding training intensity. This grading system may also be used to assess MRI images of the MCPJs of racehorses prior to injury to provide information regarding the risk for catastrophic PSB fracture.
Images obtained by low-field sMRI are not as detailed as images produced by high-field MRI,44 which has led to concerns that assessment of images obtained by sMRI may fail to identify lesions and underestimate pathological lesions. However, high-field MRI requires that horses be under general anesthesia.44 Although the performance of either high-field MRI or sMRI on a motionless anesthetized patient has benefits, many Thoroughbred trainers are reluctant to anesthetize an otherwise healthy racehorse to diagnose lameness.h Results of another study40 indicate that high-field (1.5-T) MRI and sMRI have similar sensitivity and specificity for detection of disease of SCB or trabecular bone when compared with histologic evaluation. Given that most lesions of the MCPJ of Thoroughbred racehorses involve bone7,8 and sMRI is highly sensitive for identification of bony lesions,40 it seems reasonable that sMRI should be the preferred imaging modality following radiographic evaluation for lameness involving the MCPJ in Thoroughbred racehorses to avoid the anesthetic risks associated with high-field MRI. Horses that require additional imaging after sMRI can still be evaluated with high-field MRI, and this proposed algorithm for diagnostic imaging should reduce the number of horses that require anesthesia.
On MRI images, densification is identified as a hypointense area on all MRI pulse sequences owing to a lack of mobile protons.30,45 The signal is abnormally hypointense during evaluation of bone on non–fat suppressed (ie, T1W GRE, T2W GRE, and T2*W FSE) images. In the present study, cases were significantly more likely to have PSBs with > 75% densification (PSB grade, 3) and a total PSB grade ≥ 5 than were controls, which suggested that determination of the extent of PSB densification might be an effective screening method for identification of horses at risk of biaxial PSB fracture. Results of 2 other studies16,17 also support the role of PSB densification in biaxial PSB fracture. In 1 study,16 the porosity (ie, density) of the trabecular bone of the PSBs of 2-year-old Thoroughbreds was positively associated with the extent of training and the firmness of the training surface. In the other study,17 results of histologic and histomorphometric evaluation of the PSBs of Thoroughbred racehorses indicated that PSBs with porous medullary trabeculae were less likely to fracture than were PSBs with compacted medullary trabeculae, which led the investigators to conclude that knowledge of structural changes in the PSBs of racehorses would provide an opportunity for prevention of PSB fractures. Results of another study46 indicate that MC3s with high bone density are brittle and at risk of fracturing when loading forces exceed the bone's yield strength. It is possible that high bone density may have an analogous effect on PSBs.
Preexisting disease frequently plays a role in the catastrophic failure of most major bones in Thoroughbred racehorses6,7; however, elucidation of a cause-and-effect relationship between preexisting disease and biaxial PSB fracture has been challenging.6,7,14–17,47,48 In the present study, ipsilateral limbs had significantly higher PSB grades (ie, PSB densification) and contralateral limbs had significantly higher total MC3 SCB grades and were more likely to have orthopedic disease of MC3, compared with control limbs. We propose that horses with biaxial PSB fracture of a forelimb have preexisting pathological lesions in the contralateral MC3 that cause those horses to favor the contralateral limb during training and racing. Further research of a large population of racehorses is necessary to explore this proposition. Results of multiple studies suggest that the forelimbs of both Thoroughbred22,23 and Standardbred25,26 racehorses are not symmetrically loaded while racing. It is possible that this asymmetric loading contributes to biaxial PSB fracture. Human patients with end-stage osteoarthritis of a hip joint often undergo replacement of the contralateral knee joint preferentially before replacement of the ipsilateral knee joint.49 In human patients with unilateral osteoarthritis of a hip joint, the contralateral knee joint is at greater risk of developing progressive symptomatic osteoarthritis than is the ipsilateral knee joint, and asymmetric loading forces lead to structural asymmetries in osteoarthritic knee joints well before the disease becomes symptomatic.50
In the present study, the number of starts was significantly associated with biaxial PSB fracture when it was modeled as a continuous variable but not when it was modeled as a dichotomous variable (Table 1). Results of multiple studies14,15,17,47 suggest that biaxial PSB fracture is associated with chronic use and the risk of biaxial PSB fracture increases as the number of race starts increases. In 1 study,17 intensive training and racing caused microstructural changes at the tissue level that predisposed Thoroughbred racehorses to acute complete PSB fractures. The investigators of a study14 conducted in the United Kingdom concluded that chronic use causes biaxial PSB fractures in Thoroughbred racehorses because the median number of starts for horses with biaxial PSB fracture was much higher, compared with the median number of starts for horses with all other types of fractures of the distal portion of the limbs.
Results of a study30 in which the MCPJs of 131 Thoroughbreds were evaluated with sMRI indicate that 47 (36%) had bone marrow edema suggestive of a condylar fracture. Bone marrow edema generally causes an abnormally hyperintense signal on STIR, T2*W, and fat-suppressed T2W MRI images.30,51,b,c However in the present study, the bone marrow signal on STIR and T2*W images did not differ significantly between cases and controls. This finding was consistent with the results of another study17 in which gross examination of parasagittal sections of the PSBs of Thoroughbred racehorses with and without PSB fracture failed to detect the heme pigment in any location along the PSB. In a review35 of bone marrow edema in sports medicine, a distinction is made between the biomechanics of a fracture and the amount of edema that subsequently forms; compressive forces produce more extensive bone marrow edema than do tensile forces. For the cases of the present study, it is possible that failure to detect bone marrow edema on the sMRI images was because the biaxial PSB fracture was caused by tensile forces or the marrow space had become fibrotic from repetitive injury and repair, which diminished its blood supply.45
The present study had several important limitations. Histopathologic data were unavailable for the limbs evaluated in this study. However, histopathologic information has been reported for California racehorses that were euthanized because of biaxial PSB fracture,17 and necropsy results for Thoroughbred racehorses with MC3 SCB lesions have been described in multiple studies.20,21,24 The described MC3 SCB grading scale used in the present study has not been systematically validated for interobserver and intraobserver agreement. Also, because the sMRI images evaluated in this study were obtained from motionless cadaveric limbs, it is likely that the quality of the images was higher than that of images obtained from standing and sedated clinical patients, which might have some movement artifact. Furthermore, the presence of a biaxial PSB fracture in the ipsilateral limbs prevented the investigator assessing the bony changes of the MCPJ from being blinded, which might have introduced bias in the assignment of the PSB and MC3 SCB grades. A prospective cohort study with a large sample size is necessary to better understand the basic pathogenesis of PSB fracture, MC3 SCB disease, and their potential roles in the development of lameness in the contralateral limb, and the progression or improvement of these conditions following MRI evaluation and altered training and racing regimens. Finally, the present study was a case-control study; therefore, we could not definitively identify causative associations. Catastrophic biaxial PSB fracture is an acute injury, and it is unlikely that chronic changes such as sclerosis occurred simultaneously or as a result of the fracture. It is possible that the sMRI findings associated with biaxial PSB fracture in this study were confounded by other variables that were or were not measured. The frequency distributions of sex, age, and number of starts differed between cases and controls. Although we used multivariable methods to control for the effects of age and the number of starts, we cannot exclude the possibility that the results were confounded by other unidentified variables.
The MCPJ is the most common site of musculoskeletal disease that results in reduced performance, premature retirement, and catastrophic failure and euthanasia in Thoroughbred racehorses throughout the world. Given this fact and the recent intense criticism of the Thoroughbred racing industry in North America,52 legitimate, viable strategies to prevent catastrophic failure of the MCPJ, specifically biaxial PSB fracture, are needed. Development of a 3-D diagnostic modality capable of precise and accurate imaging of bone architecture and density is urgently needed to monitor the evolution of bony changes in the PSBs and SCB of the MC3 of Thoroughbred racehorses in response to changes in training or exercise intensity.53 Such a tool could be used to establish a surveillance system to distinguish horses with reversible disease from horses with permanent injury and at risk for MCPJ failure.53 On the basis of the results of the present study, we propose that sMRI is such a tool and expedient evaluation of this approach is warranted.
ABBREVIATIONS
| CI | Confidence interval |
| LF | Left forelimb |
| MC3 | Third metacarpal bone |
| MCPJ | Metacarpophalangeal joint |
| PSB | Proximal sesamoid bone |
| RF | Right forelimb |
| SCB | Subchondral bone |
| sMRI | Standing magnetic resonance imaging |
| STIR | Short tau inversion recovery |
| T1W GRE | T1-weighted gradient echo |
| T2W FSE | T2-weighted fast spin echo |
Riggs CM. The fetlock joint in the racing Thoroughbred: what are the clinical problems, what do we know about them and what do we need to find out? (abstr), in Proceedings. 48th Br Equine Vet Assoc Cong 2009;217.
Peloso JG, Cohen ND, Vogler JD, et al. Standing magnetic resonance imaging identifies bone changes in catastrophic fracture formation of Thoroughbred racehorse fetlock joints (abstr), in Proceedings. 41st Annu Conv Am Assoc Equine Pract 2012; 539.
Peloso JG, Cohen ND, Vogler JD, et al. Using low-field standing MRI to compare bone changes in the fetlock joint of 21 Thoroughbred racehorses that sustained a catastrophic fracture with 16 control horses (abstr), in Proceedings. 42nd Annu Meet Am Col Vet Surg 2013;E-23.
Murray RC, Tranquille CA, Parkin TD. Using MRI for recognition/monitoring of fetlock pathology, and detection of fracture warning signs (abstr), in Proceedings. 48th Br Equine Vet Assoc Cong 2009;131.
Hallmarq EQ2 MR imaging system, Hallmarq Veterinary Imaging, Guilford, Surrey, England.
Equibase Co LLC, Lexington, Ky.
Spotfire, version 8.2, TIBCO Software Inc, Seattle, Wash.
Blea J, President, American Association of Equine Practitioners, Sierra Madre, Calif: Personal communication, 2013.
References
1. Krook L, Maylin GA. Fractures in Thoroughbred racehorses. Cornell Vet 1988; 78(suppl):1–133.
2. Peloso JG, Mundy GD, Cohen ND. Prevalence of, and factors associated with musculoskeletal racing injuries of Thoroughbreds. J Am Vet Med Assoc 1994; 204: 620–626.
3. Estberg L, Stover SM, Gardner IA, et al. Fatal musculoskeletal injuries incurred during racing and training in Thoroughbreds. J Am Vet Med Assoc 1996; 208: 92–96.
4. Hernandez J, Hawkins DL, Scollay MC. Race-start characteristics and risk of catastrophic musculoskeletal injury in Thoroughbred racehorses. J Am Vet Med Assoc 2001; 218: 83–86.
5. Johnson BJ, Stover SM, Daft BM, et al. Causes of death in racehorses over a 2-year period. Equine Vet J 1994; 26: 327–330.
6. Stover SM, Murray A. The California postmortem program: leading the way. Vet Clin North Am Equine Pract 2008; 24: 21–36.
7. Stover SM. The epidemiology of Thoroughbred racehorse injuries. Clin Tech Equine Pract 2003; 2: 312–322.
8. Cruz AM, Poljak Z, Filejski C, et al. Epidemiologic characteristics of catastrophic musculoskeletal injuries in Thoroughbred racehorses. Am J Vet Res 2007; 68: 1370–1375.
9. Parkin TD. Epidemiology of racetrack injuries in racehorses. Vet Clin North Am Equine Pract 2008; 24: 1–19.
10. Parkin TD, Clegg PD, French NP, et al. Risk of fatal distal limb fractures among Thoroughbreds involved in the five types of racing in the United Kingdom. Vet Rec 2004; 154: 493–497.
11. Reed SR, Jackson BF, Mcilwraith CW, et al. Descriptive epidemiology of joint injuries in Thoroughbred racehorses in training. Equine Vet J 2012; 44: 13–19.
12. Riggs CM. Osteochondral injury and joint disease in the athletic horse. Equine Vet Educ 2006; 18: 128–144.
13. Anthenill LA, Stover SM, Gardner IA, et al. Association between findings on palmarodorsal radiographic images and detection of a fracture in the proximal sesamoid bones of forelimbs obtained from cadavers of racing Thoroughbreds. Am J Vet Res 2006; 67: 858–868.
14. Kristoffersen M, Parkin TD, Singer AE. Catastrophic biaxial proximal sesamoid bone fractures in UK Thoroughbred races (1999–2004): horse characteristics and racing history. Equine Vet J 2010; 42: 420–424.
15. Kristoffersen M, Hetzel U, Parkin TD, et al. Are bi-axial proximal sesamoid bone fractures in the British Thoroughbred racehorse a bone fatigue related fracture? A histological study. Vet Comp Orthop Traumatol 2010; 23: 336–342.
16. Young DR, Nunamaker DM, Markel MD. Quantitative evaluation of the remodeling response of the proximal sesamoid bones to training-related stimuli in Thoroughbreds. Am J Vet Res 1991; 52: 1350–1356.
17. Anthenill LA, Gardner IA, Pool RR, et al. Comparison of macrostructural and microstructural bone features in Thoroughbred racehorses with and without midbody fracture of the proximal sesamoid bone. Am J Vet Res 2010; 71: 755–765.
18. Kawcak CE. The role of subchondral bone in joint disease: a review. Equine Vet J 2001; 33: 120–126.
19. Norrdin RW, Kawcak CE, Capwell BA, et al. Subchondral bone failure in an equine model of overload arthrosis. Bone 1998; 22: 133–139.
20. Pool RR, Meagher DM. Pathologic findings and pathogenesis of racetrack injuries. Vet Clin North Am Equine Pract 1990; 6: 1–30.
21. Pool RR. Pathologic manifestations of joint disease in the athletic horse. In: McIlwraith CW, Trotter GW, eds. Joint disease in the horse. Philadelphia: WB Saunders Co, 1996; 87–104.
22. Arthur RM, Ross MW, Moloney PJ, et al. North American Thoroughbred. In: Ross MW, Dyson SJ, eds. Diagnosis and management of lameness in the horse. 3rd ed. Philadelphia: WB Saunders Co, 2003; 868–879.
23. Pilsworth RC. The European Thoroughbred. In: Ross MW, Dyson SJ, eds. Diagnosis and management of lameness in the horse. 3rd ed. Philadelphia: WB Saunders Co, 2003; 879–894.
24. Barr ED, Pinchbeck GL, Clegg PD, et al. Post mortem evaluation of palmar osteochondral disease (traumatic osteochondrosis) of the metacarpo/metatarsophalangeal joint in Thoroughbred racehorses. Equine Vet J 2009; 41: 366–371.
25. Mitchell JB, Mitchell JS, Nolan PM, et al. The North American Standardbred. In: Ross MW, Dyson SJ, eds. Diagnosis and management of lameness in the horse. 3rd ed. Philadelphia: WB Saunders Co, 2003; 895–912.
26. Torre F. The European and Australasian Standardbreds. In: Ross MW, Dyson SJ, eds. Diagnosis and management of lameness in the horse. 3rd ed. Philadelphia: WB Saunders Co, 2003; 913–927.
27. Sarrafian TL, Case JT, Kinde H, et al. Fatal musculoskeletal injuries of Quarter Horse racehorses: 314 cases (1990–2007). J Am Vet Med Assoc 2012; 241: 935–942.
28. Lewis RD. The racing Quarter Horse. In: Ross MW, Dyson SJ, eds. Diagnosis and management of lameness in the horse. 3rd ed. Philadelphia: WB Saunders Co, 2003; 927–930.
29. Arthur RM, Constantinide D. Results of 428 nuclear scintigraphic examinations of the musculoskeletal system at a Thoroughbred racetrack, in Proceedings. 41st Annu Conv Am Assoc Equine Pract 1995; 84–87.
30. Powell SE. Low-field standing magnetic resonance imaging findings of the metacarpo/metatarsophalangeal joint of racing Thoroughbreds with lameness localized to the region: a retrospective study of 131 horses. Equine Vet J 2012; 44: 169–177.
31. Davidson EJ, Ross MW. Lameness in racehorses: clinical recognition of stress-related bone injury in racehorses. Clin Tech Equine Pract 2003; 2: 296–311.
32. Datir AP, Saini A, Connell A, et al. Stress-related bone injuries with emphasis on MRI. Clin Radiol 2007; 62: 828–836.
33. Spitz DJ, Newberg AH. Imaging of stress fractures in the athlete. Radiol Clin North Am 2002; 40: 313–331.
34. Koh ES, Lee JC, Healy JC. MRI of overuse injury in elite athletes. Clin Radiol 2007; 62: 1036–1043.
35. Vanhoenacker FM, Snoeckx A. Bone marrow edema in sports: general concepts. Eur J Radiol 2007; 62: 6–15.
36. Anderson IF, Crichton KJ, Grattan-Smith T, et al. Osteochondral fractures of the dome of the talus. J Bone Joint Surg Am 1989; 71: 1143–1152.
37. Mintz DN, Tashjian GS, Connell DA, et al. Osteochondral lesions of the talus: a new magnetic resonance grading system with arthroscopic correlation. Arthroscopy 2003; 19: 353–359.
38. Forney M, Subhas N, Donley B, et al. MR imaging of the articular cartilage of the knee and ankle. Magn Reson Imaging Clin N Am 2011; 19: 379–405.
39. Olive J, d'Anjou MA, Girard C, et al. Fat suppressed spoiled gradient-recalled imaging of equine metacarpophalangeal articular cartilage. Vet Radiol Ultrasound 2010; 51: 107–115.
40. Smith MA, Dyson SJ, Murray RC. Reliability of high- and low-field magnetic resonance imaging systems for detection of cartilage and bone lesions in the equine cadaver fetlock. Equine Vet J 2012; 44: 684–691.
41. Hontoir F, Nisolle JF, Meurisse H, et al. A comparison of 3-T magnetic resonance imaging and computed tomography arthrography to identify structural cartilage defects of the fetlock joint in the horse. Vet J 2014; 199: 115–122.
42. O'Brien T, Baker TA, Brounts SH, et al. Detection of articular pathology of the distal aspect of the third metacarpal bone in Thoroughbred racehorses: comparison of radiography, computed tomography and magnetic resonance imaging. Vet Surg 2011; 40: 942–951.
43. Sherlock CE, Mair TS, Braake FT. Osseous lesions in the metacarpo(tarso) phalangeal joint diagnosed using low-field magnetic resonance imaging in standing horses. Vet Radiol Ultrasound 2009; 50: 13–20.
44. Werpy NM, Ho CP, Pease A, et al. Preliminary study on detection of osteochondral defects in the fetlock joint using low and high field strength magnetic resonance imaging, in Proceedings. 54th Annu Conv Am Assoc Equine Pract 2008; 447–451.
45. Winter MD. The basics of musculoskeletal magnetic resonance imaging terminology, imaging sequences, image planes, and descriptions of basic pathologic change. Vet Clin North Am Equine Pract 2012; 28: 599–616.
46. Les CM, Stover SM, Key JH, et al. Stiff and strong compressive properties are associated with brittle post-yield behavior in equine compact bone material. J Orthop Res 2002; 20: 607–614.
47. Anthenill LA, Stover SM, Gardner IA, et al. Risk factors for proximal sesamoid bone fractures associated with exercise history and horseshoe characteristics in Thoroughbred racehorses. Am J Vet Res 2007; 68: 760–771.
48. Bukowiecki CF, Bramlage LR, Gabel AA. In vitro strength of the suspensory apparatus in training and resting horses. Vet Surg 1987; 16: 126–130.
49. Shakoor N, Hurwitz DE, Block JA, et al. Asymmetric knee loading in advanced unilateral hip osteoarthritis. Arthritis Rheum 2003; 48: 1556–1561.
50. Shakoor N, Dua A, Thorp LE, et al. Asymmetric loading and bone mineral density at the asymptomatic knees of patients with unilateral hip osteoarthritis. Arthritis Rheum 2011; 63: 3853–3858.
51. Tranquille CA, Parkin TD, Murray RC. Magnetic resonance imaging-detected adaptation and pathology in the distal condyles of the third metacarpus, associated with lateral condylar fracture in Thoroughbred racehorses. Equine Vet J 2012; 44: 699–706.
52. Bogdanich W, Drape J, Miles DL, et al. Death and disarray at Americas racetracks: mangled horses, maimed jockeys. The New York Times 2012;Mar 25:A1.
53. Cruz AM, Hurtig MB. Multiple pathways to osteoarthritis and articular fractures: is subchondral bone the culprit? Vet Clin North Am Equine Pract 2008; 24: 101–116.
Appendix
Sequences and parameters used to obtain sMRI images of the MCPJ of equine forelimbs in the sagittal, dorsal, and transverse planes.
| Pulse | TE (ms) | TR (ms) | Flip angle (°) | Spectral width (kHz) | Matrix size* | FOV (mm) | Slices obtained (No.) | Slice thickness (mm) | Gap between slices (mm) |
|---|---|---|---|---|---|---|---|---|---|
| T1W GRE HR | 8 | 100 | 75 | 33.3 | 384 × 192 | 170 | 8 | 3.5 | 0.7 |
| T2*W GRE HR | 13 | 140 | 32 | 16.7 | 384 × 192 | 170 | 8 | 3.5 | 0.7 |
| T2W FSE HR | 87 | 1,970 | 90 | 25 | 384 × 195 | 170 | 12 | 3.5 | 0.7 |
| STIR FSE | 27 | 2,910 | 90 | 25 | 336 × 165 | 170 | 12 | 5 | 1 |
Matrix size represents the number of data points collected in each dimension; it has no units.
FOV = Field of view. HR = High resolution. TE = Echo time. TR = Repetition time.