Osteoarthritis (OA) is often a progressive and disabling disease, which occurs in the setting of a variety of risk factors – such as advancing age, obesity and trauma – that collude to incite a cascade of pathophysiological events within joint tissues. An important emerging theme in OA is a broadening of focus from a disease of cartilage to one of the ‘whole joint.’ The synovium, bone and cartilage are each involved in pathological processes that lead to progressive joint degeneration. Additional themes that have emerged over the past decade are novel mechanisms of cartilage degradation and repair, the relationship between biomechanics and biochemical pathways, the importance of inflammation and the role of genetics. In this article, we review the molecular, clinical and imaging evidence that synovitis is not an ‘incidental finding of OA’, but plays a significant role in disease pathogenesis, and could therefore represent a target for future treatments.
Pathological evidence for synovitis in osteoarthritis
Synovial proliferation and inflammation
It is increasingly appreciated that some degree of synovitis may be observed even in early osteoarthritis (OA) . Synovial histological changes include synovial hypertrophy and hyperplasia with an increased number of lining cells, often accompanied by infiltration of the sublining tissue with scattered foci of lymphocytes . Synovitis is often localised and may be asymptomatic: arthroscopic studies suggest that localised proliferative and inflammatory changes of the synovium occur in up to 50% of OA patients, and the activated synovium may produce proteases and cytokines that accelerate the progression of the disease . Cartilage breakdown products, derived from the articular surface as a result of mechanical or enzymatic destruction of cartilage, can provoke the release of collagenase and other hydrolytic enzymes from synovial cells and macrophages. Cartilage breakdown products are also believed to result in mononuclear cell infiltration and vascular hyperplasia in the synovial membrane in OA. A consequence of these low-grade inflammatory processes is the induction of synovial interleukin-1β (IL-1β) and tumour necrosis factor α (TNF-α), which are likely contributors to the degradative cascade. There are also reports of increased numbers of immune cells in synovial tissue, such as activated B cells and T lymphocytes, including evidence for a clonally expanded, response of antigen-driven B cells that may contribute to the development or progression of the disease .
Characterisation of pannus-like tissue in osteoarthritis
Although OA is classified as non-inflammatory arthritis, mechanical stress and low-grade inflammatory process in the joint micro-environment may lead to synovial inflammation and growth in OA . Meachim and Osborne were the first to report that fibrous-to-loose-textured soft tissue were observed in all the OA femoral-head specimens studied . As OA progresses, the proliferating synovial tissue may resemble the pannus tissue that characterises rheumatoid arthritis, as reported by Nakamura and co-workers . The pannus-like tissue in OA overlies and invades cartilage often in focal areas; the tissue has less cellular density and is devoid of lymphocyte follicles and CD68+ cells as compared with rheumatoid arthritis (RA) pannus tissue. The causes of synovitis in osteoarthritis are poorly understood and likely multifactorial. Prevalent theories assert that the cells of the synovial lining are activated by the release of cartilage fragments, causing synovitis . Other possibilities include potential contributions by crystal-initiated (BCP or CPPD) low-grade inflammatory changes in chronic OA . In addition, synovitis may involve an antigen-driven immune response; cartilage-derived neo-antigens have been implicated in inducing infiltration of oligoclonal B cells in OA joints . OA pannus cells have been shown to express vimentin, a mesenchymal marker, supporting the possibility that its origin is subarticular or synovial mesenchymal cells . Furuzawa-Carballeda et al. have reported that OA pannus-like tissue, similar to that in RA, produces metalloproteinases as well as pro-inflammatory (IL-1β, TNF α, IL-8 and IL-6) cytokines; IL-10, IL-12 and IFN- γ were undetectable in the cultures . OA synovial tissues have also been shown to produce vascular endothelial growth factor (VEGF) and other angiogenic factors that promote neovascularisation of subchondral bone .
Activation markers of synovial tissue in osteoarthritis
Histologic changes in OA synovium also show mild-to-moderate synovitis characterised by an increase in the number of inflammatory mononuclear cells in the sublining tissue including activated B cells and T lymphocytes . Activated T-cell markers and T-cell cytokine transcripts are observed in the synovial membrane of OA patients . Two other T-cell-specific common γ-chain cytokines, namely IL-21 and IL-2, are substantially detected in the synovial tissues of both early- and end-stage OA patients. IL-15, like IL-21, is also elevated in early knee OA and its expression has also been correlated with CD8 transcripts and matrix metalloproteinase (MMP-1 and -3), implicating a role in OA pathogenesis . Synovial membrane and fluids from advanced OA patients were also positively stained for CD5+ lymphocytes cells. These observations support the view that activated lymphocytic populations are present in the OA synovial tissue, reflecting local inflammation . However, it is still controversial whether these cells contribute substantially to the inflammatory response in OA.
Inflammation and angiogenesis in osteoarthritis synovium
Synovitis has now become recognised as a common and important feature of osteoarthritis; vascular growth is enhanced in the osteoarthritic synovium when infiltrating cells generate angiogenic factors. A modest but statistically significant increase in vascular density was detected in a subset of OA synovium, indicative of angiogenesis . These findings suggest that inflammation or hypoxia resulting from mechanical joint lesions may lead to increased vascularity, likely by neovascularisation . In addition, increased mononuclear cell infiltration and overexpression of mediators of inflammation were observed in synovium isolated from early OA, compared with late OA patients . Synovial fibroblast-like cells isolated from these tissues also had elevated expression of levels of cyclooxygenase-2 (COX-2), VEGF, IL-1 and TNF α. These molecules are known to activate angiogenesis in other cell types. A greater percentage of OA synovial-lining cells were positive for activated focal adhesion kinase (pFAK) . The FAK family kinases, FAK and proline-rich kinase 2 (Pyk 2) are the predominant mediators of integrin-signalling events that play an important role in angiogenesis. Walsh and associates have shown in a meniscal transection model of OA both osteochondral and synovial angiogenesis in rat, which suggests that changes in vascularisation may occur early during the development of OA and may also contribute to the pathogenesis of OA. They also further suggested that osteochondral and synovial angiogenesis appear to be independent processes and synovial angiogenesis is associated with histological synovitis in human OA patients .
Synovial biomarkers and animal models
Serum hyaluronan levels are reported to reflect primarily synovial activity. Increased risk of radiographic progression at the knee has been associated with elevations of serum hyaluronan and serum C-reactive protein (CRP) . Richardot et al. have recently developed a neo-epitope-based antibody raised against type III collagen N-telopeptide (nitrated QY*DSY*DVKSG sequence, IIINys), which is the major collagen molecule in the synovial membrane. They further showed that serum IIINys levels were on average 1.5-fold higher in patients with knee OA than in healthy controls and significantly correlated with CRP value . Thrombin is another protein in the synovial fluid, which is also a potential biomarker of synovial inflammation and is correlated with angiogenic factor VEGF expression .
The other potential synovial fluid protein is lubricin or PREG4 . Lubricin is produced by synoviocytes and superficial zone chondrocytes in articular cartilage. Mutation or absence of lubricin results in the disappearance of superficial zone chondrocytes and hyperplastic synoviocytes in mouse joints and also develops early wear and higher friction than in normal joints. Furthermore, intra-articular treatment of rat meniscal tear model of OA with lubricin had significant disease-modifying activity, inhibited synovial cell overgrowth and demonstrated chondroprotective effects .
There is direct evidence for the presence of synovitis and is also correlated with radiographic progression . However, currently, there are no animal models to study the role of synovium in the development and progression of osteoarthritis. Smith et al. recently reported significant synovial pathology after meniscectomy in an ovine OA model . They further observed that treatment of these animals with hyaluronic acid (HA) reduced synovial pathological changes and improved joint mobility.
Intra-articular injection of transforming growth factor (TGF- β) induces synovial hyperplasia and osteophyte formation in mice. Van den Berg et al. have shown the crucial role of synovial macrophages in TGF- β-induced osteophyte formation. They also further established that inhibitors of TGF β and BMP signalling in papain-induced OA model have reduced synovial thickening and osteophyte formation . Similar to MMPs, cysteine proteases such as cathepsins are also overexpressed in OA cartilage and synovium. IL-1 and TNF-α induced expression of cathepsin K in chondrocytes and synovial fibroblasts. Cathepsin K-overexpressing transgenic mice spontaneously developed synovitis and subsequently led to cartilage and bone degeneration .
Mesenchymal stem cells of synovium
Synovial tissue had been thought to be comprised primarily of macrophages, fibroblast-like cells, mast cells and endothelial cells. Recently, De Bari et al. successfully demonstrated derivation of mesenchymal stem cells (MSCs) from synovial membrane tissue . However, the concepts of synovium containing MSCs have been supported by literature for a long time. For example, human synovial chondromatosis is a disorder of the joint where normal joint lining synovium forms cartilage tissue . Furthermore, synovial pannus obtained from RA patients has been shown to contain MSCs. MSC derived from synovium (SMSCs) are considered to have a high chondrogenic and osteogenic differentiation potential similar to bone-marrow-derived MSC. SMSCs have potential therapeutic use for cartilage and joint-tissue regeneration and tissue-engineering clinical applications . However, many challenges still remain to understand the such as the phenotype of SMSCs cells phenotype of SMSCs cells, requirement of growth factors and molecular mechanism underlying chondrogenic and osteogenic differentiation .
Clinical assessment of synovial inflammation and proliferation in osteoarthritis
Arthroscopy and MRI
Arthroscopic samples of early OA synovial tissue have been found to have increased mononuclear cell infiltration, blood vessel formation and overexpression of inflammatory mediators (TNF-α, IL-1 β, COX-2 and nuclear factor kappa light chain-enhancer of activated B cells (NF- κB)) compared with arthroplastic samples of late OA synovial tissue . Synovial membrane specimens taken from late-stage OA patients reveal the expression of cytokines similar to those seen in RA synovium, the only differences being quantitative . These mediators are well known to foster an environment of cartilage destruction by activating chondrocytes to produce catabolic factors and to further stimulate their own production and that of other cytokines, by promoting angiogenesis .
Regarding synovitis as a predictor of progression, some have reported that synovitis does not seem to be associated with cartilage loss as assessed by non-enhanced magnetic resonance imaging (MRI) ; yet others have demonstrated that the presence of synovitis is associated with progression when evaluated by arthroscopic assessments . Several studies have also indirectly linked synovial inflammation with the progression of the disease . Sub-clinical chronic synovitis may exist from the early stages of knee OA, and it may play an active role in the perpetuation of the disease rather than being an epiphenomenon. Cartilage defects on arthroscopy are often seen directly abutting areas of inflammatory synovitis, implicating that the synovium has direct effects on adjacent cartilage, reminiscent of the advancing synovial pannus seen in RA joints . Furthermore, MRI studies have shown that synovitis in OA has a predilection for the cartilage–pannus junction rather than being randomly distributed , suggesting that synovitis in OA is geographically positioned to make conditions favourable for disease progression.
Much of the increasing appreciation of the events in all three joint compartments, and their relationships to disease onset and disease progression is due to recent advances in MR imaging of cartilaginous and non-cartilaginous compartments . These advances have facilitated the understanding that OA is a failure of the whole joint organ. Imaging of non-cartilaginous structures in knee OA has highlighted the presence of synovitis in OA ( Fig. 1 ). Contrary to early OA-imaging studies, more recent work has shown that synovitis is not restricted to those patients with end-stage OA undergoing joint-replacement surgery, and that it can be seen in patients with all stages of established disease . MRI has demonstrated synovitis in early OA in joints where synovitis was not clinically detected . Even if using MRI, non-contrast techniques using semi-quantitative scoring methods often underestimate the degree of synovial disease. Quantitative measurement of synovial thickness using gadolinium-enhanced MRI is the gold standard for assessment of synovitis on MRI .
The developing field of musculoskeletal ultrasound has provided further evidence that synovitis is a characteristic feature of OA. Ultrasound images depict synovial hypertrophy as abnormal hypoechoic (low contrast, white) intra-articular and periarticular tissue that is non-displaceable and poorly compressible from pushing the probe against the skin, and which may exhibit a power Doppler signal. By contrast, a synovial effusion, seen as anechoic (black), is both displaceable and compressible, without any Doppler signal (although a lack of Doppler flow does not exclude synovitis) ( Fig. 2 ).
Ultrasound offers the ability to (1) obtain dynamic images while moving the patient to gain better insight to the pathology and (2) evaluate microvascular flow in inflamed tissues such as synovium with the power Doppler settings . In OA, which, until recently, was viewed as more of a chronic degenerative disease without a significant inflammatory component, ultrasound efforts had focussed accordingly on identification of osteophytes, articular cartilage damage, Baker’s cysts and bursitis, in joints such as the hip, knee and hand . As we are learning that OA, in fact, has inflammatory components, including synovitis, evidence has emerged showing that synovitis is detectable by ultrasound ( Fig. 2 ) – with both grey scale and Doppler images . Such diagnostic ability is paramount in the OA research arena, as plain radiographs do not show soft-tissue changes or synovial inflammation. MRIs are not as helpful in detecting synovitis without the use of gadolinium, which has become controversial in a non-life-threatening disease such as OA due to the possible, albeit low, risk of nephrogenic fibrosing dermopathy (NFD) or ‘nephrogenic systemic fibrosis’ (NSF) . Thus, the field could potentially rely on ultrasound to score or diagnose an early OA patient with a discrepancy between radiographic grading and symptoms (pain and functional limitations), which are secondary to synovitis unlike minimal osteophytes or cartilage damage. However, while advances in technology have increased the sensitivity of ultrasound to identify synovial effusions and hypertrophy in OA, ultrasound remains less sensitive than MRI in detecting effusions (70% vs. 85%) and synovial thickening (34 vs. 50%) .
A European League against Rheumatism (EULAR) study of 600 patients with knee OA identified 44% of OA patients with demonstrable effusions and 17% with synovial thickening – but with stricter definitions (4 mm or more for each) and without Doppler imaging because the Doppler data at that time were ‘highly machine dependent’ . Even so, these authors concluded that they found a combined effusion/synovitis prevalence of 47% by ultrasound in their OA cohort, further bolstering the argument that OA is, in fact, an inflammatory disease. Their data further suggested that inflammation identified by ultrasound correlated statistically with clinical signs and symptoms of increased pain and swelling or ‘flares’.
Two recent systematic reviews by Joshua and colleagues focussed on the role of ultrasound as an outcome measure for synovitis in musculoskeletal disease – and expectedly found that OA was infrequently included. The first found that only 10 of 54 articles studied synovitis in OA, while the second reported that of the 53 articles specifically involving Doppler signal, only six included OA patients. In part as an update to Joshua’s data but with a narrowed scope, Keen et al. subsequently published the first systematic literature review of ultrasound as a tool for assessing structural pathology in OA. They found that, from 1982 to 2008, only 10 of the 47 OA studies used Doppler function or assessed synovitis – and that until only recently, such reports lacked consistency and reliability (with regard to both synovitis and other structural changes in OA). Jung and colleagues recently reported that higher amounts of synovial thickening (here >3.1 mm) and effusions in knee OA correlate with elevations in biochemical serum markers of synovium and cartilage metabolism. Their study of 51 patients with painful knee OA found that serum HA and cartilage oligomeric protein (COMP) were higher in patients with more severe sonographic findings, although this was only statistically significant with HA. Still, such data further emphasise that ultrasound can serve as an important marker of active synovitis inflammation. Novel techniques to increase the utility of ultrasound in OA also include the ongoing development of three-dimensional (3-D) images . One such study of 22 patients with knee OA helped to better identify the shape, pattern and thickness of the synovium, while also locating synovial fluid collections.
Ultrasound findings in OA may also provide important prognostic information to help guide the physician’s treatment plan and the advising of patients. Conaghan and colleagues recently reported an ultrasound-identified knee effusion of 4 mm or more, as well as pain and severity of radiographic damage, as key predictors of subsequent joint-replacement surgery. This multicentred European study of 600 patients with knee OA found that those with a significant knee effusion were 2.6 times as likely to have a total knee replacement within 4 years (hazard ratio (HR) = 2.63, p < 0.0001).
A number of groups are evaluating hand OA by ultrasound and are targeting synovitis by grey scale and power Doppler as key components of proposed scoring systems . While Keen and colleagues reported that ultrasound is more sensitive than radiography in finding osteophytes and joint space narrowing, their semi-quantitative methods for identifying synovitis were statistically significant . By evaluating all of the metacarpophalangeal (MCP), proximal interphalangeal (PIP) and distal interphalangeal (DIP) joints in 33 patients with hand OA (1077 joints as one patient had an amputated finger) and 19 controls (570 joints), the authors found 46% of the patients’ joints with grey-scale synovitis and 7% with Doppler signal ( p < 0.001 for both when compared with controls). While the extent of synovitis did not correlate with the degree of pain, the study indicated that ultrasound could identify OA joints with active inflammation. In addition, their studies also demonstrated ‘moderately good intra- and inter-reader reliability’ by calculating kappa values .
Taken together, in summary, there is pathologic, arthroscopic and imaging evidence that synovitis is an integral component of progressive osteoarthritis. The data suggest that synovial proliferation and inflammation is not merely an incidental finding, but a feature of the disease, evidence that progressive OA is a failure of the ‘whole joint’. What remains to be determined is whether synovitis contributes to the cartilage destruction and, therefore, whether drugs targeted at synovial inflammation can slow disease progression.