Search

Background:Rheumatoid arthritis (RA) is a chronic inflammatory disease of immune dysregulation affecting the joints. While T cells play a recognised role in disease pathogenesis, the presence of autoantibodies years before the clinical onset of disease, and the efficacy of the B cell-depleting therapy rituximab, highlight a pathogenic role for these functionally diverse lymphocytes. A novel subset, termed age-associated B cells (ABCs), are described as CD19high CD21- CD11c+, with a high proportion expressing the transcription factor T-bet; they are elevated in murine models of autoimmunity and produce autoantibodies characteristic of autoimmune disease. A detailed characterisation of ABCs in early, drug naïve RA has not yet been conducted.Objectives:We aimed to characterise peripheral blood (PB) and synovial fluid (SF) ABCs in patients suffering from early drug naïve RA. As a secondary objective we sought to determine whether this population differs between RA patients and age-matched early psoriatic arthritis (PsA; disease controls) and healthy controls.Methods:Newly presenting early RA and other inflammatory arthritis patients, naïve to immunomodulatory treatment, were recruited from the Newcastle Early Arthritis Clinic. Patients with established RA (≥1 year duration on treatment and age-matched healthy controls were recruited in parallel. B cell subsets (naïve (CD19+IgD+CD27-), switched memory (CD19+IgD-CD27+) and ABC (CD19+CD21-CD11c+)) in PB and SF were characterised by flow cytometry. FACS sorted PB B cell subset gene expression profiles were assessed using a customised NanoString nCounter Human Immunology v2 Panel.Results:Transcriptionally, ABC differed from both naïve and switched memory B cells. In keeping with published studies ABCs had an activated memory phenotype, exemplified by elevated expression of CD69, CD80, and CD86, as well as Ki67, HLA-DR, IgG and T-bet. Interestingly, ABC had high expression of the Fc Receptor Like (FcRL) family (FcRL1-5), and an inflammatory homing profile (high levels of CXCR3 and low levels of CXCR4 and CXCR5). We found no difference in the proportion of PB ABC between RA patients and control groups, and no association of ABC frequencies with age. Focussing on RA ABCs specifically, we observed elevated ABC frequencies in females, but no association with disease activity. In keeping with their chemokine receptor profile, cross-sectional analysis also demonstrated ABC enrichment in SF compared to PB, with SF frequencies higher in established than early disease. The transcriptome of ABCs from early RA patients differed from both age-matched disease controls and healthy donors (Figure 1 next page).Conclusion:These data demonstrate that ABCs have a unique, activated, class-switched proliferative phenotype that is transcriptionally distinct from switched memory and naïve B cells. Interestingly, at the transcriptome level early RA ABCs differ to their counterparts in health and other inflammatory arthritides, suggesting they may differ functionally and contribute to pathogenesis. Of note, their high expression of MHC class II (HLA-DR) and co-stimulatory (CD80 and CD86) molecules suggests an important antigen-presentation function of ABC, which together with their unique FcRL family expression pattern, warrants further functional characterisation.Figure 1.Differential gene expression in ABCs from early RA patients compared to early PsA patients (A) and healthy controls (B). A NanoString nCounter Technologies chip was used to assess gene expression. Raw counts were normalised to the housekeeping genes. Sample quality was then assessed using the arrayQualityMetrics package. Gene expression profiles between different donor groups was assessed using the DESeq2 R package. In both hierarchical clustering heatmaps gene expression intensities were log2 transformed and their z-scores are displayed as colours ranging from yellow (low expression) to red (high expression) as shown in the key.Disclosure of Interests:Gemma Vidal-Pedrola: None declared, Najib Naamane: None declared, Dagmar Scheel-Toellner: None declared, Arthur Pratt Grant/research support from: GSK, Pfizer and Janssen, Andrew Mellor: None declared, John D Isaacs Speakers bureau: Abbvie, Gilead, Roche, UC, Consultant of: Abbvie, Gilead, Roche, UC, Grant/research support from: Pfizer, GSK and Janssen, Amy E. Anderson Grant/research support from: Pfizer, GSK and Janssen

ObjectivesTo examine how rituximab may result in the inhibition of joint destruction in rheumatoid arthritis (RA) patients.MethodsTwenty-eight patients with active RA were treated with rituximab. Radiographs of hands and feet before and 1 year after therapy were assessed using the Sharp–van der Heijde score (SHS). Expression of bone destruction markers was evaluated by immunohistochemistry and immunofluorescence of synovial biopsies obtained before and 16 weeks after the initiation of treatment. Serum levels of osteoprotegerin, receptor activator of nuclear factor κB ligand (RANKL), osteocalcin and cross-linked N-telopeptides of type I collagen (NTx) were measured by ELISA before and 16 weeks post-treatment.ResultsAfter 1 year, the mean (SD) change in total SHS was 1.4 (10.0). Sixteen weeks after treatment there was a decrease of 99% in receptor activator of nuclear factor κB-positive osteoclast precursors (p=0.02) and a decrease of 37% (p=0.016) in RANKL expression in the synovium and a trend towards reduced synovial osteoprotegerin expression (25%, p=0.07). In serum, both osteoprotegerin (20%, p=0.001) and RANKL (40%, pConclusionsRituximab treatment is associated with a decrease in synovial osteoclast precursors and RANKL expression and an increase in the osteoprotegerin/RANKL ratio in serum. These observations may partly explain the protective effect of rituximab on the progression of joint destruction in RA.

For successful translational research it is of key importance that communication between researchers mainly based in the clinic and those working in a laboratory setting functions effectively. This presentation will highlight some characteristic problems the speaker has encountered at the interface of clinical and non-clinical staff and will offer a few suggestions to avoid misunderstandings and misconceptions Disclosure of Interest None declared

Background We have recently identified a subset of pro-inflammatory memory B cells in RA synovial fluid and tissue, characterised by the expression of Fc receptor-like 4 (FcRL4)1-3. These cells produce RANKL, and express high levels of TNFa mRNA, indicating a destructive, pro-inflammatory role. Their immunophenoptype suggest that they are activated B cells; however, they show no signs of differentiation towards plasmablasts, suggesting that their functional role does not involve antibody production. Objectives The aim of this project was to investigate the characteristics of the immunoglobulin genes expressed in FcRL4+ B cells isolated from RA synovial fluid and tissue and to determine whether their B cell receptors recognise known autoantigens. Methods Single FcRL4-positive and -negative B cells were sorted from synovial tissue (n=2) and synovial fluid (n=5) of patients with active RA. Their Ig variable region genes were sequenced and subsequently expressed to generate recombinant monoclonal antibodies. Antigen specificities of the generated monoclonal antibodies were determined by ELISA. Results In total, we have cloned and sequenced B cell receptors from 332 individual B-cells (171 from FcRL4+ and 160 from FcRL4- cells). Ig gene sequence analysis demonstrated that the Ig repertoire was highly diverse with no major differences in the IGH and IGK or IGL gene segment usage or IGH CDR3 features between FcRL4+ and FcRL4- memory B cells. The prevalence of mutations, suggesting somatic hypermutation and selection in germinal centres, was equivalent in the FcRL4+ and FcRL4- populations. From the sequenced clones we have generated recombinant monoclonal antibodies from both FcRL4+ (n=29) and FcRL4– B cells (n=26) and have determined their specificity for citrullinated candidate autoantigens. A similar proportion of recombinant antibodies derived from the FcRL4+ and FcRL4- B cell populations (23% and 24% respectively) reacted with citrullinated antigens (including a-enolase, fibrinogen, vimentin and histones) Conclusions We conclude that memory B cells from both the FcRL4+ and FcRL4- populations are post-germinal centre hypermutated B cell subsets with similar Ig gene features. Their antigen specificity suggests that these functionally distinct B subpopulations both emerge from the autoantigen driven immune response in the inflamed joint. References 1. Yeo L, Toellner KM, Salmon M, et al. Cytokine mRNA profiling identifies B cells as a major source of RANKL in rheumatoid arthritis. Ann Rheum Dis. Nov 2011;70(11):2022-2028. 2. Yeo L, Lom H, Juarez M, et al. Expression of FcRL4 defines a pro-inflammatory, RANKL-producing B cell subset in rheumatoid arthritis. Ann Rheum Dis. Jan 15 2014. 3. Ehrhardt GR, Hsu JT, Gartland L, et al. Expression of the immunoregulatory molecule FcRH4 defines a distinctive tissue-based population of memory B cells. J Exp Med. Sep 19 2005;202(6):783-791. Disclosure of Interest None declared

Background and Objectives The importance of B cells in the pathogenesis of rheumatoid arthritis (RA) is highlighted by the clinical effectiveness of rituximab. Potential mechanisms for B cells in driving disease processes include autoantibody production, antigen presentation and cytokine secretion. Our recent findings suggest a pro-inflammatory and destructive role for B cells via production of RANKL, TNF-α and IL-6. A memory B cell subset in tonsils which expresses Fc receptor-like 4 (FcRL4) has previously been found to produce RANKL. We sought to determine whether the RANKL-expressing B cells present in the RA synovium belong to this subset. Methods: Synovial fluid and peripheral blood cells from RA patients were analysed for B cell differentiation markers by flow cytometry. Synovial tissue samples from RA patients were assessed using immunofluorescence and immunohistochemistry. FcRL4 mRNA expression was determined in synovial samples from RA patients and uninflamed control subjects by real-time PCR. FcRL4+ and FcRL4- CD19 + B cells were sorted from synovial fluid using a MoFlo cell sorter. mRNA expression of 48 cytokines was determined using Taqman microfluidic cards. Results FcRL4 expressing B cells were detected in the synovial fluid (median 14% of B cells) at a significantly higher frequency compared to peripheral blood (median 0.04%; p = 0.004). RANKL was significantly enriched in the FcRL4+ B cell population compared to the FcRL4- population. FcRL4+ B cells were predominantly switched memory B cells, expressing significantly higher levels of CD80, CD86, CD95, CD20 and CD11c than FcRL4- cells. Sorted FcRL4+ B cells expressed significantly higher mRNA levels of RANKL and TNF-α than FcRL4- B cells. FcRL4+ B cells expressing RANKL were present in synovial tissue, and were distributed under the synovial lining layer and in perivascular cuffs. FcRL4 mRNA was detected in synovium from RA patients but not uninflamed controls, and FcRL4 expression in RA synovium correlated positively with the degree of leukocyte infiltration. Conclusions We have identified a novel pro-inflammatory subset of B cells in the RA synovium which expresses RANKL and TNF-α and shows considerable expression of co-stimulatory molecules indicating a potential pathogenic role in RA. As these cells express high levels of CD20 it is conceivable that their removal contributes to the anti-inflammatory clinical effect of rituximab.

Background and Objectives Variation in endogenous glucocorticoid (GC) activity during inflammation has been linked to susceptibility to developing rheumatoid arthritis (RA). In RA patients, inflamed synovial tissue can generate active GCs through the expression of the 11β-hydroxysteroid dehydrogenase type 1 enzyme (11β-HSD1), which converts cortisone to cortisol. We examined whether the total body activity of 11β-HSD1 or its expression within synovium was associated with the risk of developing RA in patients with newly presenting joint inflammation. Materials and Methods Blood and urine were collected from 42 patients with inflammatory arthritis and a clinical symptom duration of less than 12 weeks; of these 17 developed RA and 25 resolved spontaneously. Total body 11β-HSD1 activity was determined by urinary gas chromatography/mass spectrometry and calculated as the tetrahydrocortisol + allotetrahydrocortisol/tetrahydrocortisone ((THF + alloTHF)/THE) ratio and the cortols/cortolones ratio. Urinary 11β-HSD2 activity was measured as the urinary free cortisol/ urinary free cortisone (UFF/UFE) ratio. Synovial tissue expression of 11β-HSD1 and 11β-HSD2 was assessed by qPCR. Inflammatory disease activity was assessed by ESR, CRP and DAS28. Results Total body 11β-HSD1 activity was significantly lower in patients with arthritis that subsequently resolved compared with patients whose arthritis developed into RA ( RA 1.34 ± 0.013, Resolving arthritis , 0.96 ± 0.07, P = 0.012). Similar changes were seen in the cortols/cortolones ratio ( RA 0.64 ± 0.05, Resolving arthritis , 0.43 ± 0.03, P = 0.0002). There was no significant difference in renal 11β-HSD2 activity between patients with resolving disease and those that developed RA. Despite the difference in total body 11β-HSD1 activity there was no significant difference between groups in synovial tissue 11β-HSD1 expression. There was no difference in ESR or DAS28 between patients with resolving arthritis or RA although the former had a significantly lower level of CRP ( RA, 49 ± 16.7 mg/dL, Resolving arthritis 19 ± 5.7 mg/dL, P = 0.018). All values are represented as mean ± SEM. Conclusions These studies demonstrate that a high total body 11β-HSD1 activity during early arthritis is associated with a reduced probability of resolution. The excess 11β-HSD1 may have an articular and/or extra-articular origin. This work raises the possibility that targeting 11β-HSD1 activity in early arthritis could impact on the development of RA.

Background and Objectives Within the context of a comprehensive study of the pathology of the early stages of rheumatoid arthritis (RA) we compared cytokine mRNA expression in the synovium of patients with early inflammatory arthritis who later progressed to rheumatoid arthritis with that of patients with a resolving disease course. Interestingly, we found a trend towards higher expression of platelet related chemokines CXCL4 and CXCL7 mRNA in early RA synovium. We therefore investigated CXCL4 and CXCL7 expression at the protein level and its co-localisation with platelets, macrophages and blood vessels. Methods Synovial tissue biopsies were obtained from treatment naive patients presenting with at least one clinically swollen joint within the first 12 weeks of symptom onset. Patients who went on to develop RA (according to the 1987 ACR criteria) at an 18 month follow-up (n = 8), as well as patients whose arthritis spontaneously resolved (n = 9) were included. In addition, biopsies collected from longer duration (> 12 weeks) treatment naive RA patients (n = 10) and patients with mechanical symptoms undergoing knee arthroscopy without obvious signs of inflammation were included as controls (n = 7). Synovial tissue sections were stained with antibodies specific to CXCL4 or CXCL7, CD41, CD68 and vWF using immunofluorescence and staining was quantified using Zeiss imaging software. Specificity of CXCL7 staining was confirmed by blocking with recombinant cytokine. Results We observed a statistically significant increase of CXCL4 and CXCL7 protein expression in patients with early RA when compared to early resolvers (CXCL4 p = 0.036, CXCL7 p = 0.011). This increase reflected a transient stage of early disease, as in treatment-naive patients with more than 12 weeks disease duration the expression level of these chemokines was found at levels comparable to non-inflamed synovium. Both CXCL4 and CXCL7 co-localised with blood vessels, platelets and CD68 + macrophages within the synovial tissue. However, only staining found outside blood vessels, which co-localised largely with CD68 as a marker for macrophages, differed between synovium from patients with resolving, early and established arthritis (CXCL4 p = 0.063, CXCL7 p = 0.028). Conclusions We have identified two chemokines, CXCL4 and CXCL7, that are expressed at higher levels on macrophages during a transient phase in early RA. Future work will investigate whether these chemokines play a role in disease progression and/or may be used as biomarkers for prediction of disease outcome.

Upon activation, in an active process known as netosis, neutrophils can release their chromatin into the extracellular space. The resulting meshwork of DNA is called Neutrophil Extracellular Traps or NETS (1). These NETS are thought to retain bacteria at a site rich in neutrophil derived proteases and thus contribute to the anti-bacterial action of neutrophils. This process, therefore, is regarded as beneficial in the context of bacterial infection. However, recent studies in SLE patients suggest that in other situations, when neutrophils are activated in the context of autoimmunity, NETS can add to disease pathology (2). The release of the normally tightly packed chromatin depends on the citrullination of histone proteins (3), by an enzyme known as peptidylarginine deiminase 4 (PAD4). PAD4 and related other PAD family members have caused great interest in the research of the pathogenesis of rheumatoid arthritis. Proteins citrullinated by this group of enzymes are recognized by acpa, or anti-citrullinated protein specific antibodies. These antibodies are highly specific for rheumatoid arthritis, although their contribution to disease pathology is currently unclear. Here, we have been investigating the role of NETS in the inflamed joints of patients with rheumatoid arthritis. Most sites of chronic inflammation contain few neutrophils. An exception here is the synovial fluid (SF) of patients with chronic arthritides. The majority of the infiltrating immune cells found here are activated neutrophils. We initially quantified extracellular DNA in SF from patients with RA, OA and crystal arthritis, confirming previous reports of the presence of DNA in SF (4). Further, we applied a range of imaging technologies to generate images of NETS and attached proteins such as neutrophil elastase, PAD4 as well as citrullinated proteins. NETS were enriched from in vivo and in vitro activated neutrophils and probed for the presence of citrullinated proteins and PAD4 by western blotting. In functional experiments, we are investigating the consequences of NET generation by synovial fluid neutrophils. Taken together, our data suggest that generation of NETS by locally activated neutrophils contributes to the production of autoantigens in the inflamed joints of patients with rheumatoid arthritis. References Brinkmann V, Reichard U, Goosmann C, Fauler B, Uhlemann Y, Weiss DS, et al. Neutrophil extracellular traps kill bacteria. Science. 2004 Mar 5;303(5663):1532-5. Garcia-Romo GS, Caielli S, Vega B, Connolly J, Allantaz F, Xu Z, et al. Netting neutrophils are major inducers of type I IFN production in pediatric systemic lupus erythematosus. Science translational medicine. 2011 Mar 9;3(73):73ra20. Wang Y, Li M, Stadler S, Correll S, Li P, Wang D, et al. Histone hypercitrullination mediates chromatin decondensation and neutrophil extracellular trap formation. J Cell Biol. 2009 Jan 26;184(2):205-13. Leon SA, Revach M, Ehrlich GE, Adler R, Petersen V, Shapiro B. DNA in synovial fluid and the circulation of patients with arthritis. Arthritis Rheum. 1981 Sep;24(9):1142-50. Disclosure of Interest None Declared

Background The importance of the B cell lineage in RA pathogenesis is highlighted by the clinical effectiveness of rituximab. Potential mechanisms by which B cells drive disease processes in RA include their ability to produce autoantibodies, act as antigen-presenting cells, or secrete cytokines. Our recent findings suggest a pro-inflammatory and destructive role for this cell type via production of cytokines including RANKL, TNF-α and IL-6 [1]. A memory B cell subset normally resident in epithelial niches which is characterised by expression of the inhibitory receptor FcRL4 has previously been described to express RANKL [2]. Objectives As we previously made the observation that B cells in the RA synovium produce RANKL, we sought to determine whether RANKL-expressing cells belonged to the FcRL4+ B cell subset. Methods Synovial fluid, peripheral blood and synovial tissue samples were obtained from patients fulfilling 1987 ACR criteria for RA. Mononuclear cells were isolated from synovial fluid and peripheral blood by density gradient centrifugation, stained with antibodies against CD19, CD20, CD27, IgD, CD11c, RANKL, FcRL4, CD95 and CD21, and analysed by flow cytometry. Mononuclear synovial fluid cells were stained with antibodies against CD19 and FcRL4 and sorted using a Mo-Flo™ cell sorter. Taqman™ microfluidic cards were used to detect mRNA expression of 48 cytokines by real-time PCR in these samples. Immunofluorescence staining was performed on 5 µm frozen tissue sections with antibodies against CD20, RANKL and FcRL4. Results FcRL4-expressing B cells, comprising 4-22% of the B cell population, were detected in the synovial fluid at a significantly higher frequency compared to matched peripheral blood (n=12, p=0.0004). RANKL was significantly enriched in the FcRL4+ B cell population compared to the FcRL4- population (p=0.006). Both FcRL4+ B cells and RANKL+ B cells were predominantly switched IgD-CD27+/- memory B cells, expressing significantly higher levels of CD11c and CD95, and lower CD21 levels than the FcRL4- or the RANKL- population. FcRL4+ B cells expressing RANKL were also present in synovial tissue (n=4). At the mRNA level, sorted FcRL4+ B cells were found to express considerably higher levels of both RANKL and TNF compared to the FcRL4- fraction (n=5). Conclusions We have identified a novel subset of B cells in the RA synovium which is characterised by expression of FcRL4 and has not previously been described in RA or any other chronic inflammatory joint condition. FcRL4+ B cells produce RANKL, and express high levels of TNF mRNA, indicating a destructive, pro-inflammatory role for this B cell subpopulation in RA pathogenesis. References Yeo L, Toellner KM, Salmon M, Filer A, Buckley CD, Raza K, Scheel-Toellner D: Cytokine mRNA profiling identifies B cells as a major source of RANKL in rheumatoid arthritis . Ann Rheum Dis 2011, 70 (11):2022-2028. Ehrhardt GR, Hsu JT, Gartland L, Leu CM, Zhang S, Davis RS, Cooper MD: Expression of the immunoregulatory molecule FcRH4 defines a distinctive tissue-based population of memory B cells . J Exp Med 2005, 202 (6):783-791. Acknowledgements The University of Birmingham has filed a patent on the subject of this work. Disclosure of Interest None Declared

Objectives Anti-citrullinated protein antibodies (ACPA) are characteristically detected in patients with rheumatoid arthritis (RA) and growing evidence suggests that they are involved in disease pathogenesis. During disease flares large numbers of neutrophils enter the joint space of RA patients. These cells can extrude genomic DNA in an active process termed NETosis. NETosis critically depends on histone citrullination by peptidyl deiminase 4 (PAD4). Here we tested the hypothesis that activation and release of PAD4 during NETosis contributes to the production of autoantigens in the inflamed joint. Materials and Methods Levels of extracellular DNA in the synovial fluid (SF) of patients with RA (n = 23) and osteoarthritis (OA) (n = 15) or crystal arthritis (n = 6) were quantified using the non cell-permeable DNA dye SYTOX Green. PAD4, neutrophil elastase, citrullinated proteins and bound human IgG were labelled in association with DNA on smear preparations of RA SF and detected by immunofluorescence. NETs from SF and in vitro stimulated neutrophils were isolated and western blotting and mass spectrometry were used to identify proteins released during NET formation and proteins co-fractionated with NETs. PAD enzymatic activity was determined after NETosis in vitro and in the SF of patients with RA (n = 9) and OA (n = 9). Results Extracellular DNA was detected in SF from patients with RA and crystal arthritis at significantly higher levels than in OA SF (RA versus OA p Conclusions These findings suggest that neutrophil NETosis, as a source of extracellular activated PAD enzymes, is involved in the production of auto-antigens in the joints of RA patients. In ACPA positive patients the resulting immune complexes can contribute to the perpetuation of the inflammatory response.

Background Dickkopf related protein 1 (DKK1) is an inhibitor of the Wnt signalling pathway involved in osteoblast growth and differentiation and cell adhesion and invasion in health and disease. In cancer, increased plasma levels of DKK1 have been associated with the osteolytic lesion of multiple myeloma and with tumour invasion and progression in solid tumours. In rheumatoid arthritis (RA) increased DKK1 plasma levels have been shown to correlate with inflammation and bone erosions. Our aim was to study expression of DKK1 in synovial fibroblasts from patients from three different clinical outcome groups: early RA, established RA and resolving arthritis. Materials and methods Synovial tissue was obtained from ultrasound guided biopsies done at the time of first presentation to clinic. Resolving (n=8) and early RA (n=12) patients presented with arthritis of less than 3/12 duration and were followed up for 18/12 at which point a diagnostic category was assigned (resolving vs early RA). Established RA (n=9) was defined as RA of >3/12 duration. Synovial fibroblasts from tissue were cultured and mRNA quantified using real-time PCR. Values were expressed as 2 −dCt relative to glyceraldehyde 3-phosphate dehydrogenase. Results DKK1 expression was higher in synovial fibroblasts from patients with early RA than in those from patients with resolving arthritis (0.024 vs 0.009, p Conclusion The question of why arthritis resolves in some patients but persists as RA in others is still not resolved. Mechanisms allowing perpetuation of inflammation and joint damage must be in place to allow this. DKK1 is an inhibitor of the Wnt signalling pathway that has been shown to promote cell invasion and imbalance of oestoblast/ostoeclast function in a way that favours bone destruction in cancer. In RA it has been proposed as a biomarker of disease activity and bone erosions. RA synovial fibroblasts have an invasive phenotype and attach to and degrade cartilage causing permanent damage in disease. Our data suggests that DKK1 may be involved in mediation of these actions.

Backgroundand objectives In the earliest clinically apparent phases of arthritis, it is difficult to distinguish patients who will develop RA from those whose disease will resolve. The authors hypothesised that synovial tissue cytokine expression would differ between early patients with these distinct outcomes. As control groups the authors included patients with newly presenting established RA and individuals with no evidence of inflammatory arthritis. Materials and methods Synovial tissue biopsies were obtained from 18 early synovitis patients who eventually developed RA (according to 1987 American College of Rheumatology (ACR) criteria within 18 month follow-up) and 13 early synovitis patients with self-limiting disease; all samples were collected within 12 weeks of the onset of any musculoskeletal symptom attributed to inflammatory arthritis. In addition, 12 patients with established RA (according to 1987 ACR criteria with a symptom duration of >3 months at sample collection), and 12 control subjects undergoing arthroscopy with no macroscopic or radiological evidence of inflammatory pathology were studied. No individuals were taking disease-modifying antirheumatic drug therapy at the time of sample collection. mRNA expression of 135 cytokines and related molecules (including interleukins, tumour necrosis factor family members, chemokines, growth factors, interferons and adipokines) was determined by real-time PCR using microfluidic cards. The Kruskal-Wallis test and Dunn9s post-test were used for comparison of four groups, and the Mann–Whitney test was used for paired analysis. Results Of the 135 cytokines analysed, 20 genes were differentially expressed between the four groups (early resolving synovitis, early RA, established RA and controls), 16 genes were not detected and 99 genes did not show statistically significant differences between groups. Significant differences included higher levels of IFNβ in established RA compared to early RA. Levels of IL-1α, IL-2 and IL-21 were significantly higher in established RA than early resolving synovitis patients. Conversely, XCL1 and VEGFα were higher in early resolving synovitis patients than established RA. The levels of many cytokines and chemokines were elevated in the synovium of early and established arthritis patients compared to control subjects. 11 genes were significantly different between anticitrullinated protein antibodies (ACPA)-positive and ACPA-negative RA patients. Conclusions The profile of cytokine mRNA expression in patients with synovitis is consistently different to that of uninflamed synovial tissue. Although significant differences exist between cytokine expression levels in different outcome groups and time points in RA disease, for the majority of cytokines studied, expression is similar between groups and may reflect the extent of inflammation rather than showing specificity for its outcome in early arthritis.