Development of an experimental model of Palmar osteochondral disease and repetitive stress injury in the equine metacarpophalangeal joint

Abstract

Palmar osteochondral disease (POD) is a traumatic fatigue injury of the subchondral bone of the palmaro/plantarodistal third metacarpal/metatarsal (MC3/MT3) condyles that has a wide spectrum of disease severity and is highly prominent amongst the racehorse population. The development and progression of this disease tend to be unpredictable, with the time to lesion progression, the degree and timing of clinical effects, and the severity of disease varying between individual horses. A prominent feature of early POD is trabecular and subchondral bone sclerosis, microcrack formation, and osteocyte necrosis, often visible on advanced diagnostic imaging as a bone marrow lesion (BML). BMLs have been shown to precede and be associated with fatigue injury and structural deterioration in the subchondral bone. A successful model for inducing POD, or repetitive stress injury in general, does not yet exist. Most experimental research involves the use of impact devices that diffusely injure the articular cartilage, serving as a catalyst for disease progression without persistent BMLs. Since there is little to no overlying cartilage damage in the early stages of POD, a technique that produces subchondral bone changes, in a minimally-invasive manner, without diffusely affecting cartilage at the time of lesion induction is critical. An experimental model that would allow researchers to study the progression in a controlled, prospective manner would open the doors for future studies on variables that may affect lesion severity, as well as an individual horse's risk of injury, and facilitate development of an optimized treatment plan. This dissertation describes a pilot study focused on the initial development of an experimental model of POD, or maladaptive stress remodeling, in horses. This research evaluated the potential of a previously validated pin penetration model developed by Stewart et al. to create BMLs in the distal third metacarpal condyles of horses. By combining this model with a high-speed treadmill exercise protocol previously shown to induce subchondral bone remodeling, the goal was to induce lesions similar to those in the earlier stages of POD, and to evaluate lesion progression and host response. The results of this pilot study confirmed that the pin penetration model – which involved using a 1.1 mm Steinmann pin to create an osteochondral defect that extended approximately 8 mm from the articular surface into the trabecular bone – combined with 6 months of high-speed treadmill exercise, consistently resulted in BMLs in the palmarodistal MC3 condyles of horses that persisted throughout the study. The first portion of this work (chapter 3) describes the clinical effects of this experimental model including lameness evaluation and synovial fluid analysis. Similar to POD, this model induced a dynamic, mild to moderate, bilateral forelimb lameness that persisted, though to variable degrees. This corresponded to an increased response to flexion, increased joint effusion and a reduction in range of motion. When evaluating the response using synovial fluid analysis, the total nucleated cell counts and total protein were near or below the upper limits of the reference interval (1,000 cells/uL and 2.5g/dL, respectively). Findings from this experimental model were consistent with a primarily mononuclear response, indicative of a non-inflammatory arthropathy such as that from degenerative or traumatic joint diseases. These results demonstrated why it is critical to pair synovial fluid analysis with clinical examination, as degenerative processes affecting joints can have minimal to no inflammation reflected in the synovial fluid, with overlap between cell counts from normal and osteoarthritic joints. The next section (chapter 4) describes the appearance and progression of the experimentally-induced BMLs using digital radiography, magnetic resonance imaging (MRI), and computed tomography (CT). Modifications to both the subchondral and trabecular bones were observed without evidence of significant articular cartilage degeneration on advanced diagnostic imaging. Importantly, the BMLs induced using this model changed over time. While histopathology was not performed until the end of the study period, the combined information obtained from MRI, CT, and the initial study by Stewart et al. allow some assumptions to be made about lesion composition; this would need to be confirmed in future research with earlier histologic evaluation. The initial composition when imaged after 4 weeks was consistent with hemorrhage and edema within the marrow spaces, at which point the BML occupied the greatest percentage of the condylar area. This fluid signal decreased significantly by week 12 and was no longer significantly different compared to baseline. Bone sclerosis was present by week 4 and remained unchanged in size throughout the remainder of the study period; however, these lesions subjectively became more defined by weeks 12 and 24. Resorptive lesions were also observed on CT and MRI near the tip of the pin tract. While the study by Stewart et al., in which the model was developed in the medial femoral condyles of sheep, reported that the trabecular bone must be damaged in addition to the subchondral plate for BMLs to persist, it is possible that a shallower pin tract would allow resorption to occur closer to the subchondral bone, as occurs in clinical POD, with the addition of high-speed treadmill exercise allowing for lesion persistence. The final portion of this research (chapter 5) describes the characteristics of the BMLs on post-mortem examination at the end of the 24-week study. This was performed via gross evaluation, microcomputed tomography (microCT), and histopathology. Gross abnormalities were consistent with pathologies of the MC3 condyles shown to be associated with POD including wear lines, subchondral bone bruising visible through the articular cartilage, cartilage loss and ulceration, marginal remodeling of the proximal sesamoid bones, dorsal impact injury, and linear fissures. In support of the diagnostic imaging findings, microCT showed marked and extensive subchondral and trabecular bone sclerosis throughout the palmarodistal condyles in all horses, with density patterns similar to previous reports. Histopathology confirmed the presence of a larger volume of less mineralized bone closer to the articular surface, characterized by regionally extensive thickening of the cancellous trabeculae with variable immature, woven bone in the region surrounding the pin tracts. Additional changes included cavitation of the articular cartilage and subchondral bone plate, extensive myelofibrosis, cellular debris within marrow spaces, and fibroplasia within the tract. While the osseous changes were more extensive, degenerative changes were also observed in the articular cartilage in the regions overlying and immediately adjacent to the pin tracts; however, distant articular cartilage was within normal limits. Based on the results by Stewart et al., this degeneration is suspected to be a result of the extensive subchondral bone damage, similar to clinical disease, and prolonged treadmill exercise. The results of this pilot study have provided in vivo proof-of-concept that the pin penetration technique combined with high-speed treadmill exercise will produce significant BMLs in the palmar third metacarpal condyles, similar to those seen in the early stages of POD. While further research is still required, this pilot study will provide a foundation for future studies bridging the gap between descriptive, associative, and predictive factors for the progression of POD-type lesions, as well as studies on how to better optimize treatment.