Although osteoarthritic lesions of the femorotibial joint are a common cause of lameness in horses, they are poorly described in the scientific literature.1 The femorotibial joint compartments (medial and lateral) are formed by the femoral and tibial condyles, with an interposed meniscus that confers stability to the joint.2 Degenerative lesions of articular cartilage are more commonly identified in the medial versus lateral femorotibial compartment,3 similar to the situation in humans.4 However, noninvasive detection of cartilage degenerative lesions in femorotibial joints of horses remains a clinical challenge.
Lameness is suspected to be attributable to the femorotibial joint when a horse has a positive response to flexion testing of the upper aspect of the limb and when synovial effusion is detected by palpation5 or ultrasonographic examination.6 Confirmation that a lesion in this region is responsible for causing signs of pain is provided when lameness is ameliorated after intra-articular administration of anesthetic, but such a test is not specific for detection of cartilage lesions.
For many years, radiographic examination alone was used to image the femorotibial joint,7 allowing diagnosis of osteoarthritis in horses with advanced disease. Characteristic radiographic features of femorotibial osteoarthritis include periarticular osteophytes, flattening of articular surfaces, sclerosis, and lucent zones in subchondral bone.8 However, radiography has many shortcomings that reduce its diagnostic sensitivity. These shortcomings are particularly problematic during the early stages of disease when lesions are small.
Unlike the distal aspect of the limb, the stifle joint is difficult to image because of its size and proximity to the trunk of the horse. This joint is one of the most voluminous joints in horses, and rendering it to a 2-D image reduces the sensitivity for detection of small lesions.9 Second, there is a large muscle mass in the stifle region that contributes to superposition and limits the ability to obtain oblique views. Finally, unlike bone, articular cartilage, menisci, or soft tissues are not visible or only partially visible on radiographs.
The advent of ultrasonographic examination has enhanced our ability to assess equine femorotibial joint structures.10 A comprehensive examination now allows detection of some meniscal, collateral, and cruciate ligament lesions; osteophytes; and some cartilage or subchondral bone lesions.11 However, because of anatomic constraints, accessibility to all joint structures is limited with ultrasonography of the femorotibial joint. The combination of radiographic and ultrasonographic examination has expanded our ability to detect lesions in this joint and now provides an extensive, although still incomplete, examination of the femorotibial joint.12,13
Magnetic resonance imaging of the equine femorotibial joint provides information on all tissue structures and in many planes.14 Major limitations include high equipment costs and limited availability of magnets of appropriate diameter to accommodate the stifle joint; these magnets are available in only a few referral centers worldwide. Furthermore, the costs associated with an examination that requires that horses be anesthetized are sometimes prohibitive. A novel technique, CT arthrography, although slightly more invasive, allows a comprehensive but indirect assessment of articular cartilage of the femorotibial joint.15,16 One of its chief advantages, compared with MRI, is rapidity of image acquisition. However, CT arthrography also has many of the disadvantages of MRI.
Despite its limitations, radiographic examination of femorotibial joints remains a mainstay examination for decision making in prepurchase examinations and femorotibial lameness evaluation for most equine practitioners as well as a first-line imaging option in referral practice. Radiographic assessment of the mJSW of femorotibial joints is considered an indirect means of measuring cartilage (femoral and tibial combined) and meniscal thickness in humans17 and, when decreased relative to the mJSW of healthy joints, is considered an indicator of cartilage degeneration due to osteoarthritis.1,14 Despite the advent of new imaging technologies, mJSW measurements remain a viable outcome variable for clinical trials designed to evaluate the effects of disease-modifying osteoarthritis drugs in human medicine.18 Lessons from human medicine suggest that the lack of standardization of femorotibial radiographic techniques leads to errors in radiographic interpretation of mJSW.19
We hypothesized that results of measurement of mJSW in femorotibial joints of horses would vary with radiographic projection angle. The purpose of the study reported here was to measure mJSW in the medial and lateral compartments of femorotibial joints of standing, orthopedically normal horses by use of caudocranial radiographic views, to compare the accuracy of those measurements with that of a software program designed for use in human medicine, and to identify the ideal caudocranial projection angle for measurement of mJSW.
Materials and Methods
Animals
The study consisted of 3 phases: a preliminary phase to determine which radiographic projection angles to use, a principal phase in which the effect of projection angle on mJSW measurements was determined, and a validation phase to validate JSW measurements ex vivo. For the preliminary phase, 2 Standardbred females (10 and 12 years old) from the institutional research herd were used. For the principal phase, 12 skeletally mature horses (mean ± SD age, 14.4 ± 2.9 years) from the institutional research herd were enrolled. These included 11 Standardbred females and 1 crossbred female with a mean body weight of 485 ± 45.5 kg. Routine lameness examination of each horse enrolled for the principal phase revealed no clinical signs of lameness attributable to stifle joint disease. For the validation phase, 2 Standardbred females and 1 crossbred female (mean age, 11.0 ± 3.5 years) euthanized for reasons unrelated to joint disease were used. The experimental protocol was approved by the Institutional Animal Care and Use Committee of the Université de Montréal Faculté de Médecine Vétérinaire.
Radiographic examination
In preparation for radiographic examination, horses were initially sedated with detomidine hydrochloride (0.01 mg/kg, IV) and butorphanol tartrate (0.01 mg/kg, IV). Images were then acquired with each horse standing squarely on all 4 feet. A long bubble level (constructed with a wooden rod and a bubble level at 360°) was used to verify alignment between the point of the tuber ischiae and the tuber calcanei (Figure 1).
Schematic illustration of radiographic beam angles in the caudocranial view selected for proximodistal and lateromedial planes to measure the mJSW in both femorotibial compartments of orthopedically normal standing horses. A—In the proximodistal plane, the central x-ray was angled relative to the ground (0°; representing the horizontal). The cassette was held orthogonal to the beam. B—In the lateromedial plane, a caudoproximal-craniodistal oblique projection was obtained at 10° from the horizontal (Cd10Pr-CrDiO), at 10° from the horizontal with 5° of lateral angulation (Cd10Pr5L), and at 10° from the horizontal with 5° of medial angulation (Cd10Pr5M). Cd = Caudal. Cr = Cranial. Di = Distal. L = Lateral. M = Medial. O = Oblique. Pr = Proximal.
Citation: American Journal of Veterinary Research 77, 2; 10.2460/ajvr.77.2.127
For caudocranial images of the femorotibial joint, cassettesa were held orthogonally to the x-ray beam. Source-to-film distance was 100 cm. The primary beam angle of the radiographic machineb was controlled with the positioning light (relative to horizontal) to obtain images 10° from the horizontal with lateral or medial angulation (Figure 1). The x-ray beam was centered on the femorotibial joint (detected by palpation). Exposures of 75 kVp and 3.2 mAs were used.
Lateromedial femorotibial images were acquired to assess for any abnormalities elsewhere in the femorotibial joint. The cassette was placed high into the inguinal region, and the central x-ray beam was directed horizontally to focus on the lateral collateral ligament (detected by palpation). Exposures of 75 kV and 1.2 mAs were used. Radiography was repeated when the femoral condyles were not superimposed on radiographs.
Establishment of radiographic projection angles
A series of caudocranial femorotibial radiographs were acquired from 2 horses at various angles in proximodistal and lateromedial planes. First, radiographs were acquired with a horizontal beam in a sagittal plane. The first caudocranial view was obtained with the x-ray beam centered just above the proximal aspect of the tibia midline on the axis of the limb at the level of the femorotibial joint. To determine the effects of angulation, radiographs were then obtained at objective 5° increments in a plane proximodistal or lateromedial to that of the initial radiograph by use of the graduated scale on the generator. In the proximodistal plane, the angle of the beam was increased in increments of 5° until the medial or lateral tibial condyle overlapped with the corresponding femoral condyle, precluding joint space measurements. In the lateromedial plane, the rotation of the beam was also increased in increments of 5° until overlapping occurred between the medial or lateral tibial intercondylar tubercule and the corresponding femoral condyle, thereby precluding measurement.
Measurement of mJSW
Angles identified in the preliminary phase of the study as yielding no overlap of anatomic structures were used for assessment of mJSW. Radiographs were acquired (without diagnostic evaluation) and then screened for evidence of osteoarthritis lesions, on both caudocranial and lateromedial views, by a third independent evaluator (PT). Only radiographs of orthopedically normal joints were included for subsequent assessment.
The mJSW19 of the medial and lateral aspect of the femorotibial joint was independently measured by a board-certified veterinary radiologist (KA) and an imaging resident (JDL). These 2 evaluators selected a site at which the JSW appeared smallest and measured the distance between a point on the femoral condyle and the tibial condyle, on all caudocranial views, on digital images in dedicated imaging softwarec with electronic calipers that measured to the nearest 0.01 mm and a computer monitor with a resolution of 2,880 × 1,800 pixels (86.6 pixels/cm).
The mJSW was also measured with the aid of a digital image analysis system.d This system was originally designed for measurement of radiographic hip JSW in humans20 and reportedly increases both the reproducibility of radiographic JSW measurements in osteoarthritic knees and the sensitivity to change in JSW in serially obtained radiographs.21 Briefly, all radiographs were oriented so that the medial femorotibial compartment was situated on the left of the computer screen. Landmarks were defined, and the joint space contour was automatically detected by the computer with the help of an edge-based algorithm. Landmarks were the medial or lateral edge of the femorotibial compartment and the estimated base of the intercondylar tubercles (Figure 2). When the algorithm failed to delineate the contours, the evaluator (PT) could correct the computer-drawn contours as appropriate.22 After the bone edges were delineated by the software, the computer automatically calculated the measurements. The mJSW result that provided the largest mJSW measurement was considered most representative of the actual value.
Representative caudocranial radiographic view of the femorotibial compartments of a horse chosen to illustrate a technique for automated digital mJSW measurement. The computer automatically detected the edges of the femoral condyles and the tibial condyles (red contours) in each compartment. The mJSW was calculated as the minimum distance (blue lines) between the delineated femoral and tibial contours.
Citation: American Journal of Veterinary Research 77, 2; 10.2460/ajvr.77.2.127
Magnification effects were determined in a group of horses on the view obtained from a standardized caudoproximal-craniodistal oblique projection made 10° from the horizontal (n = 6 joints). A coin (diameter, 21.5 mm) was placed at the level of the lateral femorotibial joint prior to acquisition of the image. Degree of magnification was then calculated by measuring the coin diameter on the radiograph by use of imaging softwarec and electronic calipers.23
Validation of JSW measurements ex vivo
Radiographs of femorotibial joints were acquired, as described previously, from 3 horses prior to euthanasia for reasons unrelated to joint disease. The standing femorotibial angle was identified as the angle between the long axis of the femur and tibia (lines drawn on the cranial aspect of the cortex of femur and the tibia) on the lateromedial view. Femorotibial joints (2/horse) were then harvested at postmortem examination and frozen at this standing femorotibial angle, which was retained by placement of a metal rod between the caudal aspect of the femoral and tibial diaphyses. The medial femorotibial joint compartment was later sectioned in a sagittal plane with a band saw to reveal the site where the mJSW was measured with the automated technique. Osteochondral samples were harvested from the medial femoral condyle and medial tibial condyle at these sites.
Harvested samples were fixed in 10% formalin solution overnight, decalcified in 50% formic acid, embedded in paraffin, and sectioned at 4 mm. Sections were stained with safranin O fast green as described elsewhere.24 Representative sections from each site were examined microscopically at a magnification of 80×. A graduated scale was used to measure thickness of HAC and HAC plus ACC at 5 sites within each section (Figure 3). Means of 5 measurements for the femoral condyle and tibial condyle sites were summed and then compared with values obtained through automated measurements made prior to euthanasia.
Representative photographs obtained for validation of radiographic femorotibial mJSW measurements in horses by use of histologic specimens from the same joints. A—Photograph of a sagittal section of the medial femorotibial joint compartment. Joints were harvested at postmortem examination and frozen at the standing femorotibial angle. The angle was retained by the placement of a metal rod between the caudal aspects of the femoral and tibial diaphyses. The medial femorotibial joint compartment was sectioned in a sagittal plane to reveal the site where the mJSW was measured in radiographs obtained from standing horses prior to euthanasia. Note there is no meniscus between the femur and tibia within the black square. Consequently, measurements made at this angle of flexion and site should reflect only cartilage thickness and not meniscal protrusion or pathological lesions. B—Photograph of an osteochondral section corresponding to site of radiographic mJSW measurement. C—Photomicrograph of a corresponding histologic section showing measurement of noncalcified cartilage. The red stain represents proteoglycan in HAC, and the blue stain represents bone. The paler area between both represents the ACC. Black lines represent HAC measurements. Safranin O fast green stain; bar = 1 mm.
Citation: American Journal of Veterinary Research 77, 2; 10.2460/ajvr.77.2.127
Statistical analysis
Minimal JSW measurements of the medial and lateral femorotibial joint compartments on caudocranial views were assessed for agreement (ICC) between the 2 evaluators and also between 1 evaluator (board-certified radiologist) and the automated digital measurements. The following ICC interpretation scale was used25: poor to fair (< 0.40), moderate (0.40 to < 0.60), excellent (0.60 to < 0.80), and almost perfect (0.80 to 1). Summary data for mJSW are reported as mean ± SD.
The effect of the radiographic projection angle on mJSW in both femorotibial compartments was analyzed with a mixed linear model for heterogeneous variances, with horse as a random factor (to control for several measurements for each individual) and projection angle as a fixed factor.e The Anderson-Darling test was performed, revealing no deviation from normality in each projection angle class. The Levene test was also performed, revealing that variances were statistically heterogeneous among projection angle classes. When the main effect of projection angle was significant, Tukey post hoc tests were used to compare pairs of means. Values of P < 0.05 were considered significant.
Results
In the preliminary phase of the study involving 2 horses, overlap of the tibial condyle and the corresponding femoral condyle (medial or lateral) occurred at angles < 0° and > 20° relative to horizontal in the proximodistal plane images. In the lateromedial plane images, only a projection 10° from the horizontal with 5° lateral or medial angulation avoided overlap of the tibial intercondylar tubercule and femoral condyles. Projection angles of 5°, 10°, and 15° (all caudoproximal-craniodistal oblique projections) in the proximodistal plane (Figure 4) and projection angles 10° from the horizontal with 5° lateral or medial angulation (Figure 5) were subsequently selected for additional evaluation.
Representative caudocranial radiographic views of the femorotibial compartments of orthopedically normal horses in the proximodistal plane at a projection angle of 20° (A; obtained during preliminary evaluation), 15° (B), 10° (C), or 5° (D). The primary beam of the radiography machine was angled relative to horizontal. Arrows in panel A represent the overlap of the tibial condyle and the corresponding femoral condyle.
Citation: American Journal of Veterinary Research 77, 2; 10.2460/ajvr.77.2.127
Representative caudocranial radiographic views of the femorotibial compartments of orthopedically normal horses obtained with the primary beam angle of the radiography machine angled at 10° relative to horizontal (A) and with 5° of lateral (B) or medial (C) angulation in the lateromedial plane.
Citation: American Journal of Veterinary Research 77, 2; 10.2460/ajvr.77.2.127
In the principal phase of the study, radiographic evaluation of the 12 horses revealed moderate signs of osteoarthritis (small osteophytes on the medial tibial condyle) in the stifle joints of 2 horses. Radiographs of those 2 joints were excluded from the study, leaving 22 orthopedically normal joints for interpretation. Almost perfect agreement was achieved between the 2 evaluators who manually measured mJSW for the medial (ICC, 0.94) and lateral (ICC, 0.87) femorotibial compartments by use of radiographs (Figure 6). Mean ± SD differences between the 2 evaluators were 0.21 ± 0.17 mm for the medial femorotibial compartment and 0.31 ± 0.34 mm for the lateral femorotibial compartment. However, the maximal difference between the 2 evaluators was 0.79 mm for the medial femorotibial compartment but 2.21 mm (almost 3 times as high as the medial value) for the lateral compartment.
Agreement between the board-certified radiologist's manual assessment and the automated digital measurements was almost perfect for the medial (ICC, 0.94) and lateral (ICC, 0.89) femorotibial compartments (Figure 6). Mean differences between the 2 types of assessment were 0.27 ± 0.19 mm for the medial femorotibial compartment and 0.39 ± 0.33 mm for the lateral femorotibial compartment. The maximal difference between assessment types was 0.88 mm for the medial femorotibial compartment and 1.54 mm (almost double the medial value) for the lateral femorotibial compartment.
Box-and-whisker plots of mJSW measurements for the medial (left) and lateral (right) compartments of 22 femorotibial joints of orthopedically normal horses as calculated manually by a board-certified veterinary radiologist (R1) and an imaging resident (R2) and with automated digital analysis software. Measurements were obtained from caudocranial radiographic views at caudo-5°-proximal-craniodistal oblique (A), at caudo-10°-proximal-craniodistal oblique (B), at caudo-15°-proximal-craniodistal oblique (C), at caudo-10°-proximo-5°-lateral-craniodistomedial oblique (D), and at caudo-10°-proximo-5°-medial-craniodistolateral oblique (E). Boxes represent 95% of the values of mJSW, the horizontal line within each box represents the mean, and whiskers represent minimum and maximum values. The solid horizontal line outside the boxes represents the maximum value, and the dashed horizontal line represents the maximum mean value.
Citation: American Journal of Veterinary Research 77, 2; 10.2460/ajvr.77.2.127
Effects of projection angle on mJSW
For the medial femorotibial joint compartment, mJSW measurements varied significantly (P < 0.001) with the angle of the caudocranial radiographic projection. Post hoc Tukey tests revealed that the mean of mJSW measurements for the medial femorotibial joint compartment was significantly (P < 0.05) higher when projections were obtained at 10° from the horizontal, at 10° from the horizontal with 5° of lateral angulation, and at 10° from the horizontal with 5° of medial angulation, compared with when they were obtained at 5° or 15° from the horizontal.
For the lateral femorotibial joint compartment, the automated mJSW measurement required more manual adjustment to specify the location of the landmarks for joint space contour detection by the edge-based computer algorithm, compared with the amount of adjustment required for the medial compartment. The lateral tibial intercondylar tubercule was difficult to differentiate from the lateral tibial condyle because of negligible differences in orientation and its proximity to the lateral femoral condyle axially, making measurement of the lateral mJSW difficult. No significant (P = 0.95) difference in mean mJSW measurements was identified in relation to the projection angle for this compartment.
Magnification effects
Mean degree of magnification was 11.46 ± 0.96%. When this value was applied to the measured mJSW (automated measurements), it did not alter the ICC.
Validation of JSW measurements ex vivo
Mean values of mJSW were 5.06 ± 0.55 mm (range, 4.37 to 5.73 mm) when the automated digital analysis software was used to measure mJSW of the medial femorotibial joint compartment in 3 standing horses prior to euthanasia (2 joints/horse) by use of radiographs obtained in a caudocranial plane at 10° from the horizontal. Mean thickness of the HAC in tissue specimens obtained from the medial femorotibial joint compartments of the same horses after euthanasia was 5.23 ± 0.79 mm (range, 4.27 to 6.23 mm). Mean thickness of the HAC and ACC combined was 5.72 ± 0.73 mm (range, 4.78 to 6.66 mm). Agreement between the software measurements and ex vivo measurements was excellent but higher for HAC (ICC, 0.77) than for HAC and ACC combined (ICC, 0.71). When values were adjusted for magnification effects, no differences in ICC were identified.
Discussion
The present study involving measurement of mJSW in caudocranial radiographs of the femorotibial joint of orthopedically normal horses provided information that can be used to improve standardization of radiographic views to permit more accurate measurements. First, mJSW measurements of the medial compartment of the femorotibial joint were highly reproducible between evaluators and slightly less so for the lateral compartment. Second, the angle of the radiographic projection significantly affected mJSW measurements of the medial joint compartment, and the highest values (most representative of cartilage thickness) were obtained on the projections angled 10° from the horizontal. Third, the maximal angle in the lateromedial plane, to avoid overlap of anatomic structures, was identified as between a medial or lateral rotation of 5°. Finally, mJSW values calculated by use of the optimal radiographic views accurately represented the thickness of HAC within the medial joint compartment.
Measurement of JSW on standardized radiographs of the knee joint is the simplest tool by which to evaluate progression of cartilage destruction of the joint in humans22,26 and is still used to assess articular cartilage loss or joint space narrowing in clinical trials to assess effects of disease-modifying osteoarthritis drugs.18 The mJSW represents the articular cartilage thickness of the femoral condyle and tibial condyle combined. Although other variables, including a mean of various measurements at different sites (mean JSW) or the calculation of the joint area, have been investigated for their usefulness in assessment of articular cartilage thickness in femorotibial joints, the most sensitive and reproducible variable is mJSW.21,27 Nonetheless, the automated methods used in the present study reportedly have better reproducibility than even manual measurements of mJSW in humans.28
The mJSW of the femorotibial joint in humans is influenced by rotation of the knee joint, degree of knee joint flexion, alignment of the tibial condyle with the center of the x-ray beam, and weight bearing.29 Additional shortcomings of the technique have been extensively addressed elsewhere.30 Briefly, these include the fact that joint space narrowing is determined from the distance between the femur and tibia at only 1 point of contact during standing and, therefore, cartilage loss along other areas of the femur and tibia that come into contact during the flexion and extension excursion of the knee joint would not be captured. Additionally, all hardware and radiographic settings must be kept constant between baseline and followup measurements to avoid systematic offsets biasing assessment of longitudinal change. Finally, radiography allows only indirect assessment of cartilage, whereas advanced imaging, such as MRI or quantitative MRI, permits assessment of the composition of cartilage and quantitative characteristics of bone.30 Because of this, standardized techniques for limb positioning were developed for greater accuracy and precision in the measurement of mJSW.19 In human medicine, the recommended position for best accuracy for clinical trials of disease-modifying osteoarthritis drugs is the Lyon schuss radiographic technique31: standardized caudocranial radiography with placement of the anterior aspect of the hip joint, patella, and tip of the hallux (big toe) against the radiographic cassette or surface of the vertical radiographic table.
The association between the radiographic progression of narrowing of the JSW and loss of cartilage detected by MRI has been evaluated in humans with clinical osteoarthritis of the knee joint, revealing that the radiographic progression of narrowing of the JSW was highly specific (91%) for detecting cartilage loss.29,32 However, a proportion of patients in those studies29,32 had evidence of cartilage loss on MRI without evidence of radiographic progression. Radiographic progression is specific but less sensitive for detection of cartilage loss than is MRI. Furthermore, a highly significant correlation has been identified between arthroscopic and radiographic variables of the same joint in an investigation33 of chondropathy in osteoarthritic knee joints in humans.
The authors were unaware of any reported studies regarding measurement of the femorotibial joint space or factors affecting that joint space in horses. However, a reduction in JSW is considered a radiographic sign of femorotibial osteoarthritis in horses.34 Accurate assessment of femorotibial JSW is particularly important because equine veterinarians still frequently rely on radiographic assessment of joint lesions owing to the previously mentioned challenges associated with other noninvasive imaging technologies, such as MRI and CT.
In the present study, excellent correlation was achieved between 2 evaluators for mJSW measurements of the medial femorotibial joint compartment, despite differences in their expertise, suggesting that this type of measurement can be reliable even for less experienced evaluators. However, agreement was less for the lateral femorotibial joint compartment, probably owing to anatomic variation among the study horses. That particular finding is of lesser clinical impact than the finding for the medial compartment, considering that most cartilage lesions reportedly occur in the medial compartment.3 Interestingly, the automated digital analysis software used also had more difficulty detecting the landmarks for the joint space contour tracings in the lateral compartment of the femorotibial joint than in the medial compartment. The radiographic projection of the lateral tibial intercondylar tubercule was difficult to differentiate from that of the lateral tibial condyle because of their small difference in orientation and their proximity to the lateral femoral condyle axially, making measurement of the lateral mJSW difficult.
The angle of the radiographic beam affected the mJSW measurements obtained for the medial femorotibial joint compartment. In the proximodistal plane, the optimal view for assessment of the mJSW in the medial compartment was identified as a caudoproximal-craniodistal oblique projection made 10° from the horizontal. In the lateromedial plane, overlap of the lateral intercondylar tubercule on the axial aspect of the lateral condyle of the distal aspect of the femur occurred with large angle views. To avoid overlap, the best lateromedial angles were ≤ 5°. With regard to measurements of the mJSW of the lateral compartment relative to x-ray beam angle, angle had no significant impact on mJSW measurements in both planes. These results, combined with difficulties in automatic localization of the mJSW, call into question the accuracy of mJSW measurements and the thickness of articular cartilage in the lateral femorotibial joint compartment.
An important objective of the present study was to validate the mJSW radiographic measurements by use of histologic measurements performed on tissue specimens. We measured the HAC and the HAC and ACC combined to determine whether the radiographic mJSW also delineated the ACC. The intraclass correlation with radiographic mJSW was excellent for both types of ex vivo measurements but was greater for HAC than for HAC and ACC combined, revealing that the radiographic mJSW in healthy joints accurately reflected the combined thickness of HAC of the femoral and tibial condyles, validating the measurement. It is also important to note that the meniscus is not present between the femur and tibia at this central site; consequently, measurements of JSW made at this angle of flexion and specific site should always reflect cartilage thickness and not meniscal protrusion or pathological lesions. However, in advanced osteoarthritis with bone lysis and loss at this site, the JSW could also potentially indirectly measure meniscal lesions. Given that this would also be an indicator of more advanced disease, the interpretation would only add to the diagnosis.
The present study had several limitations. The stifle joint in a standing horse is partially flexed, which makes standardization of the angle of flexion difficult. However, when the horse stands square, major differences in angle of flexion are unlikely to exist among horses.35 In some exceptional circumstances, the angle of flexion may be atypical because of extreme conformational changes or pathological changes in other parts of the limb. The mJSW measurements provided by the present study apply to adult horses and are likely to vary with age and size. Additional studies are warranted to assess mJSW in femorotibial compartments of horses with early-stage osteoarthritis and to compare those measurements with arthroscopic or histologic measurements of mJSW. One additional limitation is that we did not correct for magnification of mJSW values, which would have increased the precision and accuracy of our measurements.33 However, because of the high degree of standardization of the radiographic technique used, the magnification was likely consistent among measurements and should therefore have no important impact on results of comparisons. Indeed, we found that manual measurements of mJSW in the medial femorotibial joint compartment were reliable, reproducible, and optimal with caudoproximal-craniodistal oblique projections made 10° from the horizontal. Most cartilage lesions in horses with clinical osteoarthritis arise in the medial versus lateral femorotibial joint compartment,3 and veterinarians may reliably assess mJSW on these views.
On the basis of our study findings, we recommend that equine veterinarians standardize their caudocranial femorotibial radiographic views to achieve an accurate assessment of mJSW. This is particularly important for horses in the early stages of osteoarthritis, when cartilage lesions may exist without concurrent bone changes in the joint. The radiographic beam should be centered just above the proximal region of the tibial midline on the axis of the limb at the level of the femorotibial joint. A bubble level could be used in a clinical setting with compliant horses to improve view positioning.
Acknowledgments
Supported by Zoetis, Fonds du Centenaire and the Association des Veterinaires Équins du Quebec. Sheila Laverty's laboratory is funded by the National Sciences and Engineering Council of Canada, Fonds de recherche du Québec Réseau ThéCell, and a donation from John Magnier and family, Coomore Stud, Ireland.
Presented as a poster at the Annual Symposium of the American College of Veterinary Surgeons, San Diego, October 2014.
The authors thank Guillot Christian for advice related to the automated mJSW measurements.
ABBREVIATIONS
| ACC | Articular calcified cartilage |
| HAC | Hyaline articular cartilage |
| ICC | Intraclass correlation coefficient |
| JSW | Joint space width |
| mJSW | Minimal joint space width |
Footnotes
Agfa CR DX System, Toronto, ON, Canada.
Polydoros SX 65/80, Siemens, Oakville, ON, Canada.
OsiriX, version 5.8.5, Pixmeo, Geneva, Switzerland.
Holy's software-β13, Actibase, Lyon, France.
SAS, version 9.2, SAS Institute Inc, Cary, NC.
References
1. Cohen JM, Richardson DW, McKnight AL, et al. Long-term outcome in 44 horses with stifle lameness after arthroscopic exploration and debridement. Vet Surg 2009; 38: 543–551.
2. Barone R. Anatomie comparée des mammifères domestiques: arthrologie et myologie. Paris: Vigot, 2000; 263–285.
3. Bolam CJ, Hurtig MB, Cruz A, et al. Characterization of experimentally induced post-traumatic osteoarthritis in the medial femorotibial joint of horses. Am J Vet Res 2006; 67: 433–447.
4. Duncan RC, Hay EM, Saklatvala J, et al. Prevalence of radiographic osteoarthritis—it all depends on your point of view. Rheumatology (Oxford) 2006; 45: 757–760.
5. Jeffcott LB, Kold SE. Stifle lameness in the horse: a survey of 86 referred cases. Equine Vet J 1982; 14: 31–39.
6. Cauvin ER, Munroe GA, Boyd JS, et al. Ultrasonographic examination of the femorotibial articulation in horses: imaging of the cranial and caudal aspects. Equine Vet J 1996; 28: 285–296.
7. Jeffcott LB, Kold SE. Radiographic examination of the equine stifle. Equine Vet J 1982; 14: 25–30.
8. Fowlie JG, Stick JA, Nickels FA. Stifle. In: Auer JA, Stick JA, eds. Equine surgery. 4th ed. St Louis: WB Saunders Co, 2012; 1419–1442.
9. Bourzac C, Alexander K, Rossier Y, et al. Comparison of radiography and ultrasonography for the diagnosis of osteochondritis dissecans in the equine femoropatellar joint. Equine Vet J 2009; 41: 685–692.
10. Penninck DG, Nyland TG, O'Brien TR, et al. Ultrasonography of the equine stifle. Vet Radiol 1990; 31: 293–298.
11. Jacquet S, Audigie F, Denoix J-M. Ultrasonographic diagnosis of subchondral bone cysts in the medial femoral condyle in horses. Equine Vet Educ 2007; 19: 47–50.
12. Barrett MF, Frisbie DD, McIlwraith CW, et al. The arthroscopic and ultrasonographic boundaries of the equine femorotibial joints. Equine Vet J 2012; 44: 57–63.
13. Hoegaerts M, Nicaise M, Van Bree H, et al. Cross-sectional anatomy and comparative ultrasonography of the equine medial femorotibial joint and its related structures. Equine Vet J 2005; 37: 520–529.
14. Judy CE. Magnetic resonance imaging of the equine stifle in a clinical setting, in Proceedings. Am Coll Vet Surg Vet Symp 2008; 163–166.
15. Van der Vekens E, Bergman EHJ, Vanderperren K, et al. Computed tomographic anatomy of the equine stifle joint. Am J Vet Res 2011; 72: 512–521.
16. Crijns CP, Gielen IM, van Bree HJ, et al. The use of CT and CT arthrography in diagnosing equine stifle injury in a Rheinlander gelding. Equine Vet J 2010; 42: 367–371.
17. Spector TD, Hart DJ, Byrne J, et al. Definition of osteoarthritis of the knee for epidemiological studies. Ann Rheum Dis 1993; 52: 790–794.
18. Hellio Le Graverand MP, Clemmer RS, Brunell RM, et al. Considerations when designing a disease-modifying osteoarthritis drug (DMOAD) trial using radiography. Semin Arthritis Rheum 2013; 43: 1–8.
19. Hellio Le Graverand MP, Mazzuca S, Duryea J, et al. Radiographic grading and measurement of joint space width in osteoarthritis. Rheum Dis Clin North Am 2009; 35: 485–502.
20. Conrozier T, Lequesne M, Favret H, et al. Measurement of the radiological hip joint space width. An evaluation of various methods of measurement. Osteoarthritis Cartilage 2001; 9: 281–286.
21. Vignon E, Piperno M, Le Graverand MP, et al. Measurement of radiographic joint space width in the tibiofemoral compartment of the osteoarthritic knee: comparison of standing anteroposterior and Lyon schuss views. Arthritis Rheum 2003; 48: 378–384.
22. Conrozier T, Favret H, Mathieu P, et al. Influence of the quality of tibial plateau alignment on the reproducibility of computer joint space measurement from Lyon schuss radiographic views of the knee in patients with knee osteoarthritis. Osteoarthritis Cartilage 2004; 12: 765–770.
23. Conn KS, Clarke MT, Hallett JP. A simple guide to determine the magnification of radiographs and to improve the accuracy of preoperative templating. J Bone Joint Surg Br 2002; 84: 269–272.
24. Lacourt M, Gao C, Li A, et al. Relationship between cartilage and subchondral bone lesions in repetitive impact trauma-induced equine osteoarthritis. Osteoarthritis Cartilage 2012; 20: 572–583.
25. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 159–174.
26. Mazzuca SA, Brandt KD, Katz BP. Is conventional radiography suitable for evaluation of a disease-modifying drug in patients with knee osteoarthritis? Osteoarthritis Cartilage 1997; 5: 217–226.
27. Lequesne M. Chondrometry. Quantitative evaluation of joint space width and rate of joint space loss in osteoarthritis of the hip (lett). Rev Rhum Engl Ed 1995; 62: 155–158.
28. Duryea J, Li J, Peterfy CG, et al. Trainable rule-based algorithm for the measurement of joint space width in digital radiographic images of the knee. Med Phys 2000; 27: 580–591.
29. Amin S, LaValley MP, Guermazi A, et al. The relationship between cartilage loss on magnetic resonance imaging and radiographic progression in men and women with knee osteoarthritis. Arthritis Rheum 2005; 52: 3152–3159.
30. Eckstein F, Guermazi A, Gold G, et al. Imaging of cartilage and bone: promises and pitfalls in clinical trials of osteoarthritis. Osteoarthritis Cartilage 2014; 22: 1516–1532.
31. Le Graverand MP, Vignon EP, Brandt KD, et al. Head-to-head comparison of the Lyon schuss and fixed flexion radiographic techniques. Long-term reproducibility in normal knees and sensitivity to change in osteoarthritic knees. Ann Rheum Dis 2008; 67: 1562–1566.
32. Wirth W, Duryea J, Hellio Le Graverand MP, et al. Direct comparison of fixed flexion, radiography and MRI in knee osteoarthritis: responsiveness data from the Osteoarthritis Initiative. Osteoarthritis Cartilage 2013; 21: 117–125.
33. Ayral X, Dougados M, Listrat V, et al. Arthroscopic evaluation of chondropathy in osteoarthritis of the knee. J Rheumatol 1996; 23: 698–706.
34. Butler J, Colles C, Dyson S, et al. Clinical radiology of the horse. 3rd ed. Oxford, England: Blackwell Publishing, 2008.
35. Holmström M, Magnusson LE, Philipsson J. Variation in conformation of Swedish Warmblood horses and conformational characteristics of élite sport horses. Equine Vet J 1990; 22: 186–193.
