Synovial pathology in rheumatoid arthritis

INTRODUCTION — No single histologic feature or group of features in synovium is diagnostic of rheumatoid arthritis (RA). Many of the histologic changes that are seen can also occur in other inflammatory joint diseases and even in osteoarthritis. Thus, the diagnosis of RA is made by history and physical examination of the patient, and is supported by the presence of characteristic pathologic findings in synovial tissue. (See "Diagnosis and differential diagnosis of rheumatoid arthritis".)

There has been a resurgence of interest in the study of synovial tissue in RA over the past several years. Synovial tissue obtained at joint replacement surgery as well as needle and arthroscopic biopsy samples from patients with RA have been examined by molecular techniques, including single-cell ribonucleic acid (RNA) sequencing, cellular indexing of transcriptomes and epitopes (CITE) sequencing, and immunohistochemistry, in order to gain a better understanding of the pathogenic events in this disease. In addition, synovial biopsies are being used to assess the effects of medical interventions on the production of cytokines, joint-damaging enzymes, adhesion molecules, and other inflammatory mediators [1].

An overview of the histologic features of normal synovium and the changes that are characteristic of RA is presented first, followed by a discussion of the synovial response to treatment.

NORMAL SYNOVIUM — The synovial tissue that lines the fibrous capsule of a joint is derived from cells of the embryonic mesenchyme [2]. Three types of subintima have been described: fibrous, adipose, and areolar [3].

In the normal state, the synovium is a thin membrane that attaches to skeletal tissues at the bone-cartilage interface and does not encroach upon the surface of articular cartilage. The synovial membrane is divided into two layers:

●The synovial intima (also termed synovial lining cell layer) is the most superficial layer of the normal synovium and is in contact with the joint cavity. It is normally composed of synovial lining cells (SLCs) arranged one to three cells thick. Although these cells form an epithelium-like unit, images obtained by transmission electron microscopy demonstrate that this is a discontinuous cellular layer without a basement membrane.

●The SLCs are loosely attached to the subintima, which contains blood vessels, lymphatics, and nerves. Capillaries and arterioles are generally located directly deep to the SLCs; venules are located closer to the joint capsule. There is a continuous transition from loose to dense connective tissue as one goes from the joint cavity toward the capsule. Most cells in normal subintimal tissues are fibroblasts and some macrophages, although adipocytes and occasional lymphocytes and mast cells are present [4].

Synovial lining cells — SLCs are approximately 6 to 12 micrometers (microns) in diameter, excluding the numerous finger-like processes that extend from the main body. Classic electron microscopy studies have identified two populations of SLCs [5].

●Macrophage-like cells, which contain vacuoles, a prominent Golgi apparatus, and filopodia but little rough endoplasmic reticulum. These cells express numerous cell surface markers of the monocyte-macrophage lineage, including CD11b, CD68, and CD14 [6]. Macrophage-like cells are phagocytic; in the normal joint, they provide a mechanism by which particulate matter can be cleared from the joint cavity.

●Fibroblast-like synovial cells, which contain fewer vacuoles and filopodia, as well as abundant protein synthetic organelles. These cells are responsible for the synthesis of extracellular matrix proteins of the synovium, including collagen. They also produce fibronectin [7], synovial fluid hyaluronic acid, and lubricin necessary for the normal motion of cartilage surfaces in the diarthrodial joint [3]. Finally, fibroblast-like lining cells synthesize plasminogen activator which may be involved in both the degradation of fibrinogen and the activation of metalloproteinase enzymes that have been implicated in joint destruction [8].

Direct interaction between these cell types in the synovial lining later is suggested by the presence of ligand-receptor pairs on the surface of each cell type. Lining macrophage-like cells express CD97, a member of the secretin superfamily of transmembrane receptors; and lining fibroblast-like cells express CD55, or decay accelerating factor, a specific ligand for CD97 [9]. Although the functional consequences of this ligand-receptor pair are unknown in synovium, it is noteworthy that CD55 protects cells from complement-mediated damage and it is likely that CD97 transduces signals in the macrophage. (See "Regulators and receptors of the complement system".)

SYNOVIAL PATHOLOGY IN RHEUMATOID ARTHRITIS — As previously noted, the diagnosis of rheumatoid arthritis (RA) is clinical and cannot be made by examination of synovial tissue alone. However, characteristic features of RA synovitis include (picture 1):

●Hypertrophy of the lining layer.

●Neo-angiogenesis.

●Infiltration of immune cells of both the innate and adaptive immune system; immune cells can be found either randomly distributed within the synovial sublining or spatially grouped into follicle-like structures.

●Fibrin is often deposited on synovial surfaces, especially in clinically active disease.

In synovial biopsy studies, a comparable number of macrophages, T cells, and plasma cells were demonstrated in the sublining region of the rheumatoid synovium of different joints [10]. However, macrophage- and fibroblast-like synovial lining layer cells in these samples did not show a significant correlation between different joints in the same patient. Furthermore, fibroblasts show different DNA methylation and transcriptome patterns from one joint to another, suggesting that differences exist in synovium from joint to joint within the same patient with RA [11]. These differences need to be taken into account when assessing data from synovial biopsy studies.

RA synovitis is also highly heterogeneous from patient to patient, with diverse cellular and molecular signatures (pathotypes) emerging as potential taxonomic classifiers of disease phenotypes [1]. Diverse histomorphological features of RA synovitis are described as follicular, diffuse, and pauci-immune synovitis after staining with hematoxylin and eosin (H&E) and by immunohistochemistry (IHC) for B cells (CD20), T cells (CD3), macrophages (CD68), and plasma cells (CD138) [1,12].

Synovial lining cell hyperplasia — Hyperplasia of the synovial lining cell layer is a hallmark of RA. The lining cell layer may become 6 to 10 cells thick (versus the normal of one to three cells) and there is hypertrophy of individual cells.

Hyperplasia in established RA is primarily due to an increase in the number of macrophage-like cells [6]. Evidence for the recruitment of macrophage-like cells from the bone marrow comes from the study of radiation chimeras of normal and beige mice [13]. Beige mice carry giant secondary lysosomes in several cells, including those of the monocyte-macrophage lineage. Histocompatible normal mice were irradiated and bone marrow cells were grafted from beige mice. Giant granules were subsequently noted in 1 to 7 percent of synovial lining cells (SLCs), all of which had ultrastructural features of macrophage-like cells.

Macrophage-like SLCs do not appear to be actively proliferating in the synovial lining. However, the degree to which in situ proliferation of fibroblast-like cells contributes to the hyperplasia seen in the rheumatoid joint is a matter of debate. Proliferating cell nuclear antigen (PCNA), an auxiliary factor for replication polymerases, plays a critical role in DNA synthesis and its presence correlates with other indices of cellular proliferation [14]. The presence of PCNA in RA SLCs has been used as evidence that in situ proliferation of fibroblast-like synoviocytes contributes to synovial lining hyperplasia [15]. On the other hand, immunostaining for Ki-67, a marker of cell proliferation, reveals absence of this marker in the synovial tissue of patients with RA [16].

Cellular infiltrate — Mononuclear cellular infiltrates are typically present in RA deep to the synovial lining layer. These infiltrates are comprised predominantly of macrophages and lymphocytes, which can form aggregates.

●CD4+ T cells are present diffusely, in aggregates, and in perivascular sites.

●CD4+, CD45RO+ differentiated memory T lymphocytes are common, especially in areas between aggregates [17]. This phenotype is consistent with cells previously exposed to antigen (see "The adaptive cellular immune response: T cells and cytokines", section on 'Memory T cells'). Conditions in the synovial tissues may allow for an altered pool of peripheral blood T cells with limited diversity to act as proinflammatory cells [18].

●CD8+ cells are present diffusely and many express granzyme K, an enzyme that may be important in the activation of synovial proteins. These cells are a major source of interferon-gamma and tumor necrosis factor (TNF) production in synovial tissues [19].

●B lymphocytes and immunoglobulin producing plasma cells are present within and between aggregates and in foci; many plasma cells produce rheumatoid factor as well as antibodies to citrullinated proteins.

●Citrullination is a post-translational protein modification performed by the enzymes peptidyl-arginine deiminases (PADs), which convert the amino acid arginine into the noncoded amino acid citrulline. Protein citrullination can be detected in the inflamed joints in RA-affected synovial tissue. (See "Pathogenesis of rheumatoid arthritis", section on 'Citrullinated proteins and peptides'.)

●T cells play a key role in pathogenesis of RA. In a study to evaluate activated T cells, cell types were isolated from RA synovial tissues using mass cytometry, demonstrating an expanded population of programmed cell death (PD)-1^hiCXCR5^-CD4⁺ T cells [20]. Despite the high expression of PD-1, these cells do not have a phenotype of exhaustion; rather, they appeared to be a peripheral T-helper subset (ie, T peripheral helper [TpH] cells) that are located within clusters of synovial B cells, as well as in the peripheral circulation. These cells promote production of interleukin 21 to support B-cell proliferation and differentiation into plasma cells [20].

However, it is still not clear which other of the T-cell phenotypes in RA synovium are responsible for disease, which are present as a response to the synovial microenvironment, and which are merely spectators [21].

Multinucleated cells are sometimes present in the SLC layer and occasionally deep to the lining. These cells are of two types: macrophage polykaryons, and multinucleated cells with a fibroblast-like morphology. They are presumably formed by fusion of SLCs [21]. Although common in RA, these cells can also be seen in other forms of arthritis. The biological significance of these multinucleated cells and the factors involved in their generation are unknown. True multinucleated osteoclasts, however, are localized to the pannus-bone interface.

Other cells may also be seen in the synovium of patients with RA. Polymorphonuclear leukocytes are sometimes present, especially in acute disease flares. Natural killer (NK) cells and mast cells have also been identified; the latter may play a role in inflammation and in articular damage at the cartilage-pannus junction [22]. Prolonged morning stiffness may be related to impaired fibrinolysis of neutrophil-enmeshed fibrin deposits along the synovial membrane [23].

Dendritic cells have the ability to activate resting T lymphocytes. These cells are present in RA synovial tissues and fluid. The surface molecules CD80 (B7-1) and CD86 (B7-2) are costimulatory molecules required for effective antigen presentation. The presence of CD80/86 on dendritic cells within T cell clusters in RA implicates these cells as antigen-presenting cells within the rheumatoid joint [24].

Adhesion molecules have well-defined functions in inflammation (see "Leukocyte-endothelial adhesion in the pathogenesis of inflammation"). These molecules are expressed on endothelium and in other synovial sites in patients with RA. Adhesion molecules play a role in the activation and recruitment to the joint of circulating mononuclear cells, consistent with the observed accumulation of mononuclear cells around blood vessels. Specialized venules with a columnar endothelium, similar to the "high endothelial venules" described in lymph nodes, are apparent morphologically and have been specifically implicated in mononuclear cell influx into synovium.

Relevant adhesion molecules include selectins, vascular cell adhesion molecule 1 (VCAM-1), intercellular adhesion molecules 1 and 2 (ICAM-1 and ICAM-2), integrins, and others [25,26]. Expression of VCAM-1 mRNA and protein is present in and on type B SLCs, and to a lesser extent on endothelium of small vessels beneath the lining layer [27].

Angiogenesis — The hypertrophied synovium is nourished by a plethora of blood vessels [28,29]. This growth of new blood vessels is thought to be important in the influx of inflammatory cells into the RA synovial tissue. It is also important in the growth of the RA pannus.

Pannus — Destruction of cartilage and bone is a major source of morbidity in patients with established RA. The destructive process is mediated by a synovial membrane-derived tissue termed "pannus" (Latin for "cloth"). Pannus tissue adheres to the surface of articular cartilage, and cells within pannus produce proteinases, which degrade and destroy the extracellular matrix of the cartilage (picture 2) [30]. Electron microscopic studies suggest that pannus results from synovial derived fibroblast-like cells migrating over the surface of cartilage. Deep to this fibroblast-like cell layer, ongoing destruction occurs via the invasion of cells with the electron microscopic appearance of macrophages [31].

The relationship of these macrophage-like cells to the fibroblast-like cells on the cartilage surface is unclear. Over time, blood vessels as well as other cell types migrate into the pannus. Eventually, the pannus is transformed into a fibrous tissue termed "inactive" pannus. One limitation of sampling synovial tissue via needle biopsy or arthroscopy is that these samples usually do not evaluate tissue at the cartilage-pannus junction.

Invasion of bone by pannus is mediated by osteoclasts leading to the characteristic marginal erosions observed radiographically in patients with RA. Destruction of cartilage and bone lead to joint instability, disability, and eventually to fibrous ankylosis.

Multinucleated cells with the full phenotype of osteoclasts have been identified at the pannus-bone interface in bone resorption lacunae and are required for bone resorption in RA [32,33]. Synovial T-cells play a role in osteoclastogenesis. A subset of T-cells expressing CD4 and receptor activator of nuclear factor (NF)- kappa B ligand (RANKL) cause peripheral blood mononuclear cells to differentiate into osteoclasts, an effect that is abrogated by osteoprotegerin (OPG), a soluble protein that binds to RANKL to prevent its biologic activity [34] (see "Normal skeletal development and regulation of bone formation and resorption", section on 'Osteoclasts'). However, conditional deletion studies have demonstrated that synovial fibroblasts are a more important source of RANKL for inducing osteoclastogenesis in RA tissues and subsequent articular bone erosion [35].

Changes in early RA — Histologic changes in early RA have been studied in an attempt to better understand the pathogenetic events in this disease. Histologic examination of RA synovium in cases of up to six weeks' disease duration has shown the following [36]:

●Hyperplasia of the synovial lining cells; these cells are predominantly fibroblast-like synoviocytes by electron microscopy (versus a predominance of macrophage-like cells in established RA).

●Inflammatory infiltrates are perivascular and are localized to the superficial areas within the sublining. No lymphocyte aggregates are present in these early tissue samples.

●There is slight predominance of CD4+ T cells over CD8+ cells. B cells are present in the majority of cases, but in fewer numbers [37].

●Microvascular changes are prominent in early cases, with vascular congestion, high endothelial cells, and microvascular occlusion by platelets and/or thrombus.

One study compared synovial tissue samples from patients with RA of less than one year's duration with samples from patients with RA of more than five years' duration [38]. Although synovial lining hyperplasia was present in both groups, it was more pronounced in patients with established disease. Measurements of synovial inflammation did not differ significantly between the two groups.

In another study, 55 individuals who were IgM rheumatoid factor-positive and/or anti–citrullinated protein antibody (ACPA)-positive and who were without any evidence of arthritis upon physical examination were evaluated. All individuals underwent magnetic resonance imaging (MRI) and mini-arthroscopic synovial biopsy sampling of a knee joint at inclusion and were prospectively followed. Results showed that subtle infiltration by synovial T cells may precede the signs and symptoms of arthritis in preclinical RA [39].

Biopsies of clinically uninvolved knee joints in patients with RA have demonstrated histological evidence of synovitis [40,41]. These observations suggest that the synovial changes typical of RA can predate clinical signs of disease activity.

MEDIATORS OF INFLAMMATION IN THE SYNOVIUM — The analysis of synovial tissues in rheumatoid arthritis (RA) has provided information on the localization and regulation of mediators of inflammation including cytokines, enzymes, adhesion molecules, and transcription factors.

Cytokines — Cytokines are proteins that act as soluble mediators of inflammation (see "The adaptive cellular immune response: T cells and cytokines", section on 'Cytokines'). Much of what is known about the role of cytokines in RA comes from extensive studies of these proteins in serum and synovial fluid. The study of synovial tissue has provided complementary information, including the identification of cells that produce cytokines. There is great interest in the role of macrophages and macrophage-like synovial lining cells (SLCs) in the initiation and perpetuation of rheumatoid synovitis (see "Pathogenesis of rheumatoid arthritis"). The dramatic clinical improvement in RA with cytokine blockade has further heightened interest in synovial cytokines [42,43]. (See "Overview of biologic agents in the rheumatic diseases".)

Dissection of the cytokine network in RA has localized the production of several cytokines to synovial tissue and led to the notion that positive feedback loops contribute to the perpetuation of synovial inflammation. Many cytokines have been immunolocalized to the RA joint, the following are illustrative [44-49]:

●Interleukin 1 (IL-1) and tumor necrosis factor (TNF), both of which are produced by synovial macrophages, stimulate fibroblast proliferation and increase production of IL-6, IL-8 (and other chemoattractant molecules known as chemokines), granulocyte-macrophage colony-stimulating factor (GM-CSF), and destructive enzymes such as stromelysin and collagenase (see 'Chemokines' below and 'Enzymes' below).

●GM-CSF, which is produced by synovial macrophages and activated fibroblasts, is a potent inducer of IL-1, TNF, and IL-8 secretion. GM-CSF, especially in combination with TNF, also activates macrophages and induces human leukocyte antigen (HLA)-DR expression.

●TNF, IL-1-beta, IL-6, GM-CSF, and transforming growth factor (TGF)-beta 1 have been identified by immunohistochemistry at the cartilage-pannus junction, mostly localized to cells expressing the macrophage marker CD68 [49].

●Fibroblast-like cells produce autocrine stimulation and contribute to their own dysregulation through the elaboration of growth factors like fibroblast growth factor, as well as anti-apoptotic molecules [50].

●IL-18, which shares properties with IL-1, has other detrimental effects including induction of interferon gamma, and upregulation of inducible nitric oxide synthetase and cyclooxygenase (COX)-2 [51].

Chemokines — The recruitment of cells to sites of inflammation is largely directed by chemokines and their receptors. The chemokine receptors, CXCR3 and CCR5, are expressed by most T cells in the synovial fluid compared with a minority of T cells in the peripheral blood of subjects with RA, and virtually all T cells in rheumatoid synovial tissue express CXCR5. This phenotype suggests that lymphocytes are recruited to the synovium [52]. Interactions between a stromal cell derived factor (SDF)-1 and its receptor (CXCR4) may reduce the egress of T cells from the synovial compartment [53,54]. A number of chemokines, such as IL-8/CXCL8, epithelial neutrophil derived peptide-78 (ENA-78/CXCL5), and growth-regulated oncogene (GRO)-alpha/CXCL1 have been localized to the RA synovium and may mediate neutrophil recruitment and angiogenesis in the RA joint [55-57]. Likewise, chemokines such as macrophage inhibitory protein-1 alpha (MIP-1 alpha/CCL3) or macrophage chemoattractant protein (MCP-1/CCL2) may attract macrophages [48,58].

Enzymes — Several classes of enzymes have been implicated in the joint destruction that occurs in established RA. Perhaps the best studied are the family of matrix metalloproteinases (MMPs). Cytokines known to be expressed in RA synovial tissues, including IL-1-beta, TNF, IL-6, and IL-17 are potent activators of these MMP genes [59].

MMP family members are divided into three categories based upon substrate specificity.

●The collagenases cleave native collagen types I, II and III, the major collagens present in bone and cartilage, respectively.

●The stromelysin group of enzymes has a broad substrate specificity, and stromelysin-1 (MMP-3) can degrade proteoglycans, gelatin, fibronectin, and laminin. Enzymes in this group may be of particular importance because of their ability to cleave MMP pro-enzymes to yield active enzymes.

●The gelatinases A (MMP-2) and B (MMP-9) degrade denatured collagen as well as basement membrane collagen and type V collagen in bone.

In situ hybridization studies have demonstrated that interstitial collagenase (MMP-1) and stromelysin-1 (MMP-3) mRNA are expressed in rheumatoid SLCs (image 1) [60-62]. Synovial cells producing mRNA for MMP-1 and higher levels of active metalloproteinases in synovial tissue are associated with erosive disease [63,64]. Gelatinases have also been identified in RA synovial tissues.

The tissue inhibitors of metalloproteinases are naturally occurring proteins which bind in a 1:1 stoichiometry to various MMP family members, rendering them inactive. Tissue inhibitor of metalloproteases (TIMP)-1 and TIMP-2 have been identified in RA synovial tissues and are expressed in the synovial lining layer and in cells in the sublining [61,62,64]. The balance between metalloproteinases and TIMPs is expected to determine the rate of extracellular matrix degradation. However, it is difficult to make a functional interpretation of the relative intensities of their detection in tissue samples.

Granzymes A and B are serine proteases present in activated natural killer (NK) cells and cytotoxic T cells. These enzymes are present in some synovial lymphocytes, particularly in NK cells in patients with RA and to a lesser extent in those with osteoarthritis [65,66]. Synovial tissue staining for granzyme B correlates with elevated serum levels of acute phase reactants [66].

Cathepsins B, D, K, L, and S are cysteine proteinases that have been identified in human rheumatoid and osteoarthritis synovium [67-69]. Cathepsin K can degrade collagen of both type I (bone) and type II (articular cartilage). An increased number of cathepsin K-producing cells are present in the synovium of patients with RA compared with those with osteoarthritis [69]. Cathepsin K is highly expressed in osteoclasts; patients with the osteosclerotic disorder, pycnodysostosis, lack activity of this enzyme [70].

Transcription factors and cell signaling molecules — The primary control point for the production of the proteins that mediate inflammation and joint destruction (eg, cytokines, enzymes, and adhesion molecules) is at the level of gene transcription. This process is regulated by transcription factors which are proteins in the nucleus of the cell that bind to specific promoter regions of DNA. There is considerable interest in understanding the role of transcription factors in the pathogenesis of RA, especially since they are potential targets for therapeutic intervention. (See "Basic genetics concepts: DNA regulation and gene expression", section on 'Transcription'.)

The AP-1 family of transcription factors (c-Jun, c-Fos, JunB, JunD, FosB, and Fra1, Fra2) activate the transcription of several genes relevant to synovitis, including the metalloproteinases collagenase and stromelysin.

The transcription factor nuclear factor kappa B (NF-kB) is a dimer, typically of p65 (RelA) and p50 (NF-kB1); like AP-1, these may be substituted with other family members (c-Rel, RelB). NF-kB is likely to be important for the expression of a number of cytokines found in rheumatoid synovium. Immunohistochemistry (IHC) has identified activated p65 and p50 in the nuclei of many deep and superficial cells in rheumatoid synovium, including endothelial cells (picture 3 and picture 4) [71,72]. Using double IHC NF-kB within the synovial lining is localized predominantly, although not exclusively, in CD14-positive (macrophage-like) cells [71]. This finding is in contrast to the distribution of AP-1 in fibroblast-like cells. The immunolocalization of TNF to CD14-positive SLCs in patients with RA may be attributed, at least in part, to the activity of NF-kB in these cells [49].

Other molecules involved in the signaling pathways have been identified in synovium [73], including those of the mitogen-activated protein kinase (MAPK) pathway. Three major members of this pathway include ERK, Jun kinase (JNK), and p38 MAPK, all of which are expressed in active form in RA synovial tissue [74,75]. It may be that in the context of RA, shear, heat, cytokine, and oxidative stress activate these signaling pathways [76-78].

SYNOVIAL TISSUE RESPONSE TO THERAPY AND CORRELATION WITH DISEASE ACTIVITY — The advantage of synovial biopsy over measurements of peripheral blood to assess the response to therapeutic interventions was highlighted in a report of two patients with rheumatoid arthritis (RA) treated with the monoclonal antibody alemtuzumab (Campath-1H) [79]. The study subjects developed a prolonged depletion of circulating lymphocytes in response to the drug despite persistence of clinical synovitis. Synovial tissue analysis revealed significant T lymphocytic infiltrates at a time when circulating T lymphocytes were markedly depleted, suggesting that peripheral blood analysis may not accurately reflect the synovial tissue response to therapy. In contrast, synovial T cell depletion and reconstitution did correlate with changes in clinical disease activity in a series of seven patients who had synovial biopsies before and after autologous stem cell transplantation [80].

Quantitative analysis of factors within synovial tissue and peripheral blood enables the response to a therapeutic intervention to be measured at the target organ.

●Preinflammatory mesenchymal (PRIME) cells, which bear hallmarks of fibroblasts, have been identified in peripheral blood and are more common in patients with RA than in healthy controls. An increase in these cells is seen in blood just before flares. In a recent study, serial home fingerstick collection of blood was combined with RNA sequencing techniques and weekly home monitoring of patient-reported disease activity with the use of a questionnaire [81].

●Quantitative immunohistochemical analysis of sequential synovial biopsy specimens has been used to measure the response at four weeks to intravenous anti-tumor necrosis factor (TNF) monoclonal antibodies. One study of 14 patients with RA found significant reductions in staining for adhesion molecules E-selectin, vascular cell adhesion molecule-1 (VCAM-1) and T lymphocytes after such treatment compared with controls [82].

●Treatment of 16 patients with a chemokine receptor CCR1 antagonist resulted in a decrease in the number of synovial tissue macrophages and CCR1+ cells after therapy, as assessed by synovial biopsy [83].

●In situ hybridization for metalloproteinase mRNA has been used to quantitatively assess the synovial response to drugs. Intraarticular glucocorticoids decreased the synovial lining layer expression of collagenase mRNA in three patients with RA in one report [61]. In a pilot clinical trial of N-[4-hydroxyphenyl] retinamide, however, there was no decrease in mRNA for metalloproteinases collagenase or stromelysin [84].

●In a biopsy-driven, randomized clinical trial in RA (R4RA), patients who were TNF-inhibitor-inadequate responders were randomized to receive either rituximab or tocilizumab after stratification by their synovial B cell signatures. Only 12 percent of patients with a low synovial B cell molecular signature responded to rituximab, while 50 percent responded to tocilizumab. In patients with a high synovial B cell signature, the two drugs appeared equally effective [85].

Variability in synovial tissue from different joint sites and the representative nature of small synovial biopsies are important issues dictating that sampling be done in a controlled manner. The availability of small arthroscopes has facilitated the selection of biopsy sites under direct visualization. One report found that multiple synovial samples taken from within a rheumatoid joint are histologically very similar, even when specimens are taken from maximally and minimally inflamed areas [86]. Another group has taken sequential arthroscopic biopsies from distinct areas, in an attempt to reduce potential variability [87,88].

In a study comparing early disease (RA of less than one year) and longstanding disease (RA of greater than five years), there was no change in immunohistological features with disease duration, although knee pain in the subjects correlated with macrophage numbers, interleukin 6 (IL-6), and TNF expression in the synovial biopsies [89]. Knee pain did not correlate with scores for CD4+ T cells.

The following studies illustrate the use of synovial biopsy as a tool to better understand RA pathogenesis and the response to therapy:

●A study of 86 pre/post-TNF-inhibitor arthroscopic biopsies (24 posttreatment repeats) demonstrated that ectopic lymphoid-like structures (ELS) were an independent negative predictor of response, whereas their regression was associated with good response [90]. In line with this, synovial transcripts associated with poor responses to TNF-inhibitors were found in samples expressing high levels by immunohistochemistry (IHC) of IL-7-R, normally associated with ELS [91]. However, in another study of 97 pre/post-TNF-inhibitor arthroscopic biopsies (15 posttreatment repeats), lymphocyte aggregates were reported as positive predictors of response at 16 weeks [92]. There are a number of explanations for these discrepancies including: variable disease duration and follow-up times, different past-TNF-I exposure, and, importantly, different ELS categorization.

●With regard to rituximab use, in a study of 24 patients, biopsies were performed pretreatment and 4 and 16 weeks posttreatment. Basal levels of synovial B cells were not found to be predictors of a response, whereas a reduction of plasma cell numbers was a predictor, presumably as a consequence of depletion of their immediate memory B-cell precursors by rituximab [93]. In another study of 13 patients (biopsies pretreatment and eight weeks posttreatment), patients with higher levels of responses may have had more consistent depletion of synovial B cells and of synovial immunoglobulin synthesis [94].

●A study with abatacept (a CD80/86-CD28 inhibitor) involving 15 patients (biopsies pre/12 weeks posttreatment) showed a small reduction in synovial cells post therapy, which was significant only for CD20-positive B cells, and favorable changes in interferon (IFN)-gamma, osteoprotegerin, and receptor activator of nuclear factor (NF)- kappa B ligand (RANKL) in the responder group [95].

●A biopsy study comparing tocilizumab plus methotrexate-treated (n = 10) patients with methotrexate-treated patients (n = 10) demonstrated a complete blockade of synovial IL-6 expression and a significant reduction in CD20-expressing B cells in the tocilizumab-treated group compared with the group treated with methotrexate alone, whereas TNF-alpha, matrix metalloproteinase (MMP)-3, and CD68 were similarly expressed in both groups. Thus, synovially derived IL-6 and CD20 cells appear to be differentially affected by tocilizumab treatment compared with treatment with methotrexate alone [96].

With further research to integrate synovial tissue analysis with clinical prediction models, the analysis of synovial biopsy specimens might provide valuable information regarding cellular and molecular signatures that could help identify patients who will respond to specific traditional and biologic disease-modifying antirheumatic drugs (DMARDs) [1].

SUMMARY

●Normal synovium – The synovial tissue that lines the fibrous capsule of a joint is derived from cells of the embryonic mesenchyme. In the normal state, the synovium is a thin membrane that attaches to skeletal tissues at the bone-cartilage interface and does not encroach upon the surface of articular cartilage. The synovial membrane is divided into two layers: the intima or synovial lining cell (SLC) layer, which is the more superficial layer; and the subintima, which contains primarily fibroblasts, macrophages, blood vessels, lymphatics, and nerves and which is adjacent to the joint capsule. (See 'Normal synovium' above.)

●Role of synovial lining cells – There are two types of SLC: macrophage-like cells and fibroblast-like cells. These cells appear to directly interact and are major effector cells in the joint damage that occurs in rheumatoid arthritis (RA). Mononuclear cellular infiltrates are typically present in RA deep to the synovial lining layer. These infiltrates are comprised predominantly of macrophages and lymphocytes, which can form aggregates. The hypertrophied synovium is nourished by a plethora of blood vessels, which is thought to be important in the influx of inflammatory cells into the RA synovial tissue and in the growth of the RA pannus, a synovial membrane-derived tissue which adheres to the surface of articular cartilage. (See 'Synovial lining cells' above and 'Synovial pathology in rheumatoid arthritis' above and 'Synovial lining cell hyperplasia' above and 'Cellular infiltrate' above and 'Angiogenesis' above.)

●Pathogenesis – The destruction of cartilage and bone in RA is mediated at the interface between pannus and cartilage or bone; cells within pannus produce proteinases, which degrade and destroy cartilage. Invasion of bone by pannus is mediated by osteoclasts and results in the development of the characteristic marginal erosions observed radiographically in patients with RA. Destruction of cartilage and bone can lead to joint instability, disability, and, eventually, to fibrous ankylosis. Histologic changes in early RA include SLC hyperplasia, lymphocytic cellular infiltrates, and some polymorphonuclear leukocytes, as well as microvascular changes. (See 'Pannus' above and 'Changes in early RA' above.)

●Mediators of inflammation – The analysis of synovial tissues in RA has provided information on the localization and regulation of mediators of inflammation, including cytokines, chemokines, enzymes, adhesion molecules, and transcription factors. (See 'Mediators of inflammation in the synovium' above.)

●Monitoring response to therapy – Peripheral blood analysis may not accurately reflect the synovial tissue response to therapy, as indicated by studies designed to assess the response to therapeutic interventions. However, the synovial tissue response to therapeutic interventions may yield information regarding the pathogenesis of RA which cannot be obtained by other means. (See 'Synovial tissue response to therapy and correlation with disease activity' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Alisa Koch, MD, who contributed to earlier versions of this topic review.