Introduction
Osteoarthritis (OA) is an articular joint disorder that leads to mechanical failure within the knee, causing pain and deformity. Although this degenerative joint disease is common throughout North America [
1‐
3], there is an unmet need for pharmacological therapies that modify or reverse the structural damage and alleviate symptoms [
4]. The erosion of cartilage (that is, full thickness lesions or loss of surface integrity) is commonly associated with OA [
5,
6]. However, changes to the underlying subchondral bone, through variations in regional volumetric bone mineral density (vBMD) or the rapid unorganized remodeling seen after injury, are likely to cause greater structural damage to joints by inhibiting local blood flow [
7] and the formation of sclerosis [
8,
9]. Moreover, it has been shown that alterations in the biomechanical integrity of subchondral bone can modify the ability of the overlying cartilage to function normally [
4,
10,
11]. Identifying pathological features within patients exhibiting aggressive disease progression is paramount to ensuring the proper treatment regimen is used, whether that includes physical therapy, pharmacological or surgical interventions [
12].
Intra-osseous lesions, commonly known as subchondral bone cysts (SBC), 'pseudo-cysts' or 'geodes' [
13,
14], within OA knees have recently been associated with greater pain and disease progression. SBC were first identified by Ondrouch [
15] and Landells [
16] in the load-bearing regions of the femur, patella, and shoulder of arthritic patients, although the exact cause is not well known. Currently, there are two conflicting theories proposed for the origin of SBC in OA: 'synovial fluid intrusion', via a breach in the subchondral plate caused by the diminished local cartilage, leading to a rapid inflammatory response; or 'bony contusion', where stresses in the bone below the joint surface (due to trauma or thinned cartilage) exceed the functional strength of the trabecular bone, causing micro-fracture, edema, and focal bone resorption [
14,
15,
17]. However, those earlier studies only utilized histological techniques, and were limited to analyzing tissue that was removed during a surgical procedure, such as total knee replacement [
18], where the diseased or lesion-occupied area is severely degraded. Concurrently, the ability to monitor the progression of these cystic lesions
in vivo has been limited to retrospective analyses of patients' MRI scans [
18‐
21] which provide limited information regarding the onset of SBC formation. Given recent evidence that associates the presence of SBC, found in 25% to 47% of painful OA knees, with greater cartilage loss and increased risk of joint replacement among OA patients [
19,
21], it is apparent that this subchondral bone feature warrants further study.
Pre-clinical models can provide valuable insights into the cause of SBC by producing patterns of cartilage and bone degeneration that are similar to those seen in human post-traumatic or secondary OA. The surgical induction of a partial medial meniscectomy, combined with a severed anterior cruciate ligament (ACLX), in the rodent knee can emulate several aspects of secondary OA (including SBC) in a predictable and reproducible manner [
22,
23]. The use of preclinical imaging, specifically micro-computed tomography (micro-CT), has become routine in the past several years for longitudinal,
in vivo monitoring of disease progression [
24]. When combined with the superior soft-tissue contrast provided by high-field magnetic resonance imaging (MRI) [
25,
26], dual-modality imaging has the potential to characterize joint degradation comprehensively [
23] and differentiate the bone, fluid, and tissue components found within SBC [
16]. However, few studies have capitalized on the ability to incorporate image registration, which ensures any observations are conducted at the same three-dimensional (3D) location within each volume.
The purpose of this study was to use dual-modality, in vivo micro-imaging techniques to monitor the development of intra-osseous lesions occurring in an established pre-clinical model of OA. By combining our quantitative measurements with end-stage histology and immunohistochemistry, we intend to accurately determine the contents within these SBC and identify the possible existence of an open communication between the subchondral bone and joint space. Thus, these advanced analyses will provide valuable insight into the origin of OA SBC through the possible confirmation of one of the prevailing theories while simultaneously capturing the progressive remodeling of osteoarthritic bone. Furthermore, we hope to define more accurately this increasingly prevalent condition that is associated with more actively degenerative OA.
Discussion
To the best of our knowledge, this is the first study to indentify and monitor SBC as they occur in a preclinical model of OA. The induction of OA through surgical destabilization in the rat knee produced SBC in a consistent manner within the medial tibial plateau, which could be measured with medical imaging techniques similar to those used in humans. High-field MRI aided in the depiction of SBC having both fluid and fibrous tissue components as they increased in size and number and the rapid acquisition time enabled us to obtain an
in vivo, dual-modality images of the rat knee in under an hour. Using histological analyses and a grading system of OA severity designed for humans [
33], we found that SBC occur at the site of greatest disease severity, which agrees with previous findings in humans [
21]. This study further validates the rat ACLX model of secondary OA with SBC development.
The presence of a breached subchondral plate was identified by Landells and proposed to be the source of SBC formation [
16]. As synovial fluid entered the cavity, it was thought to increase the intra-osseous pressure and lead to subsequent cyst expansion. However, this is unlikely for the following reasons: the defect in the subchondral plate is small relative to the size of the SBC (Figure
5c); the course, if it existed, from joint space to SBC is tortuous or oblique to the surface, limiting the ability of easy fluid exchange (Figure
5c); the defect is filled with fibro-cartilage (Figure
5c,
7a), which is less stiff than bone, but may not allow fluid flow in the same manner as hyaline cartilage behaves as a viscoelastic solid under weight-bearing. In addition, as the SBC were still expanding by the final time point, there was evidence of immature bone deposition in some knees that would further inhibit fluid flow (Figure
3f). Thus, Ondrouch proposed that due to a thinned cartilage layer, normal weight-bearing forces within the adjacent bone are increased beyond its physiological limits, causing focal damage, subsequent resorption and SBC formation [
15]. Recent mechanical simulation studies have confirmed higher intra-osseous stress values beneath the joint surface when the amount of overlying cartilage is diminished [
17], which could also lead to SBC formation in the opposing bone in the joint. Our results indicate that the SBC are developing and expanding due to the mechanical instability imposed by the ACLX. Increased shear stress on the surface of the cartilage, due to the loss of the ligament, combined with a medial compartment-weight distribution bias, due to the meniscectomy, caused cartilage reduction or fibrillation and increased intra-osseous stress simultaneously. Therefore, the defect in the subchondral bone is likely a micro-fracture in response to the overloaded region in the joint. Repeated loading of this area during gait prevents sufficient healing and the space remains occupied with fibrous tissue (Figures
6,
7). Consequently, the bone below undergoes further damage leading to focal osteonecrosis, which is defined as a lack of osteocytes in the damaged area followed by the formation of fibrotic, avascular marrow [
34,
35].
The expansion of these SBC in OA joints occurs due to stress-induced bone resorption of necrotic tissue. In the hip, this can continue until the majority of the epiphyseal trabecular bone in the femoral head or acetabulum is resorbed, causing great pain and deformity [
14,
16,
17]. The SBC in knee OA tend to be smaller and contain an outer rim of sclerotic bone that makes them distinguishable even when imaged with planar radiographs [
13,
36]. This may indicate a form of self-limitation within the bone, where the SBC become more spherical and are surrounded by a necrotic shell to help dissipate the increased loading. This may explain how some SBC can resolve or reduce in size over time [
21] if the fibro-cartilage tissue within the cavity is provided the opportunity to mineralize. Sclerotic bone formation is emulated in our rat model as vBMD increased by 13% (at 12 weeks post-ACLX) on the superficial surface of the SBC after a significant decrease in vBMD at 8 weeks post-ACLX versus presurgical values. Furthermore, simulations of SBC within the human knee using finite-element analysis indicated that intra-osseous stress concentrations were higher near the joint surface [
28], which could explain the sclerotic bone formation seen in the proximal tibia (Figure
2f).
One of the features lacking in our SBC models was the detectable presence of bone marrow edema, which is commonly associated with human SBC development [
18‐
20]. A diffuse, and high-intensity fluid-like, signal upon MRI examination of the trabecular bone is the proposed evidence of edema in the joint compartment, typically as a result of injury and the subsequent initiation of the bone remodeling process [
20]. Edema has been found to precede SBC formation in most cases [
20,
21], but its presence does not guarantee trabecular remodeling when compared with histological examination [
37]. The inability to detect bone edema may have resulted from the MRI sequence used, as it did not incorporate fat suppression techniques that enhance edema visualization against the fatty bone marrow [
38]. However, we were concerned about masking any marrow components that may be included within the SBC and chose not to use fat suppression in our MRI protocol. In addition, allowing a shorter duration between imaging sessions (that is, two weeks) may improve our detection of edema in the early stages of OA progression and SBC formation. With the advancement of custom designed rat knee coils [
39] that incorporate longer scan times for increased signal-to-noise ratio, it will be feasible to further quantify the tissue composition of SBC over time, without reliance on tissue-destructive, end-stage histological analysis. This is further evident in the recent work using the rat model of secondary OA, combined with dual-modality imaging that demonstrated the utility of longitudinal imaging when evaluating the effect of anti-inflammatory drugs on disease progression [
23].
Conclusions
In summary, the rat ACLX model of post-traumatic knee OA provides valuable insights into the cause and composition of a degenerative feature within the subchondral bone that is associated with greater pain and disease severity. Subchondral cysts, as they are commonly referred to, appeared in a predictable manner within the overloaded joint compartment of the medial tibial plateau. These SBC do not contain any evidence of an epithelial lining and therefore are often referred to as 'pseudo-cysts' or geodes, indicating the formation of a lesion in response to stress [
13,
14]. Our evidence suggests that these spherical or pyriformal intra-osseous lesions contain fibrous tissue, which may initially contain fluid and can ossify in later stages, surrounded by regions that lack osteocytes and can possibly form a sclerotic margin in response to repeated, pathological loads. Therefore, we feel further studies are warranted to elucidate the unique pathogenesis of SBC that focus on the presence of focal, OA-induced osteonecrosis. A breach of the subchondral bone plate was found overlying all SBC and we believe the expansion of these SBC is more likely due to micro-fracture and stress-induced bone resorption than the influx of fluid through the breach. It remains unclear whether SBC first initiate due to a loss of cartilage integrity or as a result of a subchondral bone failure, but this debate is common in OA research and future studies should strive to characterize the response of both tissues to pathological stress conditions.
In addition, it should be noted that SBC tended to appear in combination with greater cartilage severity scores. Finally,
in vivo dual-modality imaging proved an invaluable method for quantifying the efficacy of therapeutic interventions developed to treat SBC in humans [
40] or animals [
23].
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
DDM conceived of the study, developed the imaging protocols, performed all imaging experiments, statistical analyses, and drafted the manuscript. VU performed histological and imaging analyses, and drafted portions of the manuscript. MD performed histological analyses. VP was involved in the development of the animal model, assisted in imaging experiments and histological analysis. FB was involved in coordination of the study. DWH was involved in the conception and coordination of the study and design of the imaging protocols. ll authors read, edited, and approved the final manuscript.