Background
Early treatment, the availability of biologic therapies and treat-to-target strategies has made remission a realistic goal in rheumatoid arthritis (RA) [
1‐
5]. However, significant numbers of RA patients classified as being in clinical remission have detectable synovitis on ultrasound (US) or magnetic resonance imaging (MRI) [
6‐
10]. These patients have an increased probability of reactivation of RA and/or radiological progression during follow up. The persistence of subclinical synovitis, as evaluated by power Doppler (PD) US has been associated with a high risk of flares and joint damage [
11‐
14]. Understanding the immunopathology underlying US-defined synovitis in RA patients in remission may increase knowledge of the physiopathology of RA and identify potential biological biomarkers of further reactivation and the progression of structural damage.
Studies suggest that PD scores correlate with vascular and inflammatory cell infiltration [
15,
16]. However, in RA patients in clinical remission, only retrospective, limited semiquantitative pathological data obtained from the joints of patients undergoing replacement surgery have been reported [
17]. To identify the pathological basis of these findings, we aimed to analyze the immunopathologic characteristics of synovial tissue (ST) obtained from RA patients in clinical remission with US- defined synovitis (presence of PD signal) and to determine whether immunopathologic changes predict the relapse from clinical remission during a 12 months of follow up.
Methods
Patient recruitment
Consecutive patients with RA were selected, who had been in clinical remission for ≥6 months and had PD signal. Patients had a 28-joint Disease Activity Score (DAS28) <2.6 and no swollen or tender joints as evaluated by two independent rheumatologists in the RA outpatient clinic. Patients aged <18 years, with allergy to local anesthetics, anticoagulant treatment, or unhealthy skin at the site of interest, were excluded. Clinical and US examinations were performed at the outpatient clinic of the rheumatology service and the US-guided synovial biopsy was carried out within 24 h of the clinical and US examination. We obtained biopsies from 24 patients, but 4 of the biopsies were non-evaluable due to a lack of a well-defined synovial lining. Finally, 20 patients with suitable synovial samples were clinically followed for 12 months to detect relapse from remission. All 20 patients completed the follow up.
We also included ST from two control groups: (1) synovial biopsies from unselected patients with clinically active RA (n = 22), and (2) non-inflammatory ST (control) obtained on arthroscopic surgery from 10 selected patients (60 % male; mean age (± SD) 42 (±8) years), who had meniscal lesions, but had no osteoarthritic or inflammatory lesions on MRI and had histologically normal ST. The study was approved by the Ethics Committee of the Hospital Clinic of Barcelona (Comité Ético de Investigación Científica del Hospital Clínic de Barcelona, Spain (2011/6490)) and signed informed consent was obtained from each patient.
US assessment and US-guided synovial biopsy
All US assessments were performed using high-sensitivity US equipment (Acuson Antares®, Siemens AG, Erlangen, Germany), using a frequency range from 10 to 12 MHz and pulse repetition frequency between 500 and 800 Hz. Receiver gain settings were controlled to eliminate the appearance of artifacts. Joint US findings were characterized according to published Outcome Measures in Rheumatology (OMERACT) definitions [
18]. An experienced sonographer (JR), blinded to the results of the clinical joint examination, evaluated both knees and 11 joints in each hand (including proximal interphalangeal joints, metacarpophalangeal joints (MCP) and wrists) for synovial hypertrophy (SH) and intra-articular PD signal according to European League Against Rheumatism (EULAR) guidelines [
19]. SH and PD scores were independently quantified (grades 0–3). US-defined synovitis was based on the presence of PD signal. Intra-rater agreement was calculated as previously described [
20], and was 0.81 for SH and 0.92 for PD.
US-guided synovial biopsies were performed using the same US equipment. Two operators (JDC and JR) performed all biopsies. Only joints with PD signal were selected for biopsy. Biopsies were carried out in an operating theatre and six to eight synovial biopsies were taken per procedure. Each US biopsy was made according to the technique described by Kelly et al. [
21]. There were no complications after the procedures.
Histological and immunohistochemical assessments
ST were routinely fixed and embedded in paraffin. Deparaffinized sections were cooked to perform antigen retrieval when required. Slides were subsequently stained with an automated immunostainer (TechMate 500 Plus; Dako, Cambridge, UK) using the following monoclonal antibodies: anti-CD3 (clone PS1; Novocastra, Newcastle, UK) for T-lymphocytes, anti-CD20 (clone L26; Dako) for B-lymphocytes, anti-CD68 (clone KP-1; Dako) for CD68 + macrophages, anti-CD117 (rabbit anti-human polyclonal antibody; Dako) for mast cells, anti-Hsp47 monoclonal antibodies (IgG2b M16.10A1 clone; Assay Designs) for synovial fibroblasts, anti-CD31 (clone JC70A; Dako) for endothelial cells, and anti-basic fibroblast growth factor (bFGF) (polyclonal SC-79, Santa Cruz Biotechnology, Santa Cruz, CA, USA) and anti-CXCL12 (clone K15C [
22]) for angiogenic markers, which were significantly increased in serum from these patients in our previous study [
20]. As a negative control, the primary antibodies were substituted by isotype-matched and concentration-matched control antibodies. The primary antibodies were subsequently detected by an avidin-biotin-peroxidase-based method (Envision System; Dako) and an aminoethylcarbazole color reaction (Sigma-Aldrich, St. Louis, MO, USA) as previously described [
23]. Finally, the slides were counterstained with hematoxylin.
Digital image analysis
Stained slides were scored on digital image analysis by an independent observer (RC), who was blinded to the diagnosis and the clinical data. Only slides with well-defined lining and sub-lining areas were included. Each stained slide was scored in its entirety by dividing it into different regions. Within each region, the number of stained cells per area and the percentage of stained cells were measured in at least 20 high-power fields using the AnalySIS®Imaging processing program (Olympus®) as previously described [
24].
Statistical analysis
Quantitative immunohistologic data were analyzed using the nonparametric Mann–Whitney test or the Kruskal-Wallis test with the post-hoc Dunn test where appropriate. Correlation between variables was analyzed using Spearman’s test. Values of p < 0.05 were considered significant. The statistical analyses were carried out using SPSS V.18 software.
Discussion
To our knowledge, this is the first study to quantitatively analyze changes in the infiltrating and resident cell components that characterize RA patients in remission with US-defined synovitis, based on the presence of PD signal. Previous studies in active or end-stage joints in patients with RA and osteoarthritis have shown correlation between US findings, mainly PD scores, and vascularity [
15]. In a recent study of patients with clinically active RA there was significant correlation between inflammatory cell infiltration (CD68 macrophages and CD3 T lymphocytes) and vascularity with a PD color fraction [
16].
Our results show that RA patients in clinical remission, who have a persistent PD signal, have high macrophage infiltration, comparable to patients with clinically active RA, and increased vascularity. In contrast, lymphocytic and mast cell infiltration and fibroblastic hyperplasia were significantly reduced in these patients compared with patients with clinically active disease. The comparison with ST from individuals without inflammatory disease reinforces these conclusions as there was increased vascularity but similar fibroblastic density in RA patients in remission compared with synovium from patients without inflammatory disease. Furthermore, the expression of angiogenic factors, particularly bFGF, which was found to be the best systemic biomarker of US synovitis in our previous study [
20], was also significantly elevated in the ST of these patients.
These results have implications for the mechanistic understanding of US-defined synovitis in RA patients in clinical remission. Synovial macrophage infiltration is the best marker of active disease, and the most sensitive to change after effective therapy [
25,
26]. Therefore, the finding of persistent macrophage infiltration suggests that US synovitis does not differ physiopathologically from clinically active synovitis. Despite the absence of clinical signs (i.e., non-tender and non-swollen joints), RA patients in remission, who have PD signal, would appear to have a pathologic status whereby macrophage depletion has not been achieved by therapy.
Our results also confirm the link between PD and increased vascularity in these patients, as was previously suggested to be characteristic of joints in clinically active disease [
15,
16]. The persistent expression of angiogenic bFGF and macrophage infiltration provides a mechanistic explanation for this pathologic observation. Interestingly, bFGF-induced angiogenesis has been linked to macrophage infiltration, underlining the pathogenic role of this factor in RA [
27,
28].
The main pathologic difference between clinical and US-defined synovitis was in lymphocytes, mast cells and fibroblasts, which were significantly-reduced in ST from RA patients in remission who had PD signal. These cells might, therefore, play a relevant role in the clinical expression of synovitis, but identification of the potential mechanisms remains speculative. It is also interesting that the levels of mast cell and B lymphocyte infiltration were associated with clinical progression to symptomatic, clinically active RA during the follow up.
Mast cells play an essential role in some animal models of arthritis [
29] and are an important synovial reservoir of pro-inflammatory cytokines, including TNFα and IL-17, in RA [
30,
31]. However, few studies have involved analysis of their changes or their prognostic implications in patients on treatment [
32]. Despite the significant reduction in asymptomatic US synovitis compared with clinical synovitis, the levels in patients in remission correlated with further clinical reactivation during follow up, supporting a role for mast-cells in disease activity and in the transition from subclinical to clinically active synovitis.
A higher density of synovial B lymphocytes was also associated with RA reactivation. The efficacy of anti-CD20 therapy in RA has shown the pathogenic relevance of B cells, but the specific role of local synovial B-cells is less clear [
33]. An increased basal density of B cells has been associated with a worse therapeutic response to anti-CD20 but not to other therapies [
34,
35].
Our study was limited by the relatively small sample size. However, we were able to take synovial biopsies from >80 % of eligible patients (meeting the PD signal criteria). However, ST from the group with clinically active RA was taken from joints with clinically active disease (swollen joints with inflammatory synovial fluid) and was included retrospectively to specifically compare their immunopathologic characteristics with the RA remission group. Therefore, the association between the immunopathologic findings and disease reactivation are only exploratory and require further confirmation. Finally, 80 % of patients with RA in remission, who had PD signal, also had synovial hypertrophy grade ≥2, thus, meeting a more stringent criterion for US synovitis [
20]. These patients had a high risk of coming out of remission, as 50 % had arthritic flares during follow up.
Acknowledgements
The authors thank Dr. Beatriz Bravo (Servicio de Traumatología y Cirugía Ortopédica, Hospital 12 de Octubre, Madrid, Spain) for providing non-inflammatory ST. All the data from this study are included in the manuscript. We have no more data to share. This study was supported by grants PI11/1890 (JDC), PI I12/439 (JLP), Rio Hortega Program (RF), and RIER RD12/09, from the Instituto de Salud Carlos III, Ministerio de Economía y Competitividad, Spain (co-financed by FEDER, European Union “Una manera de hacer Europa”).
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
JR helped design the study, performed data acquisition and data analysis, and helped write the paper. RC, AU, and RF performed immunohistochemical analysis, data acquisition and data analysis, and drafted the manuscript. VR-E and AC contributed to data acquisition and analysis and revised the manuscript. RS helped design the study and interpret the data, and helped to draft the manuscript. JDC conceived the study, and JDC and JLP designed the study, participated in the research and data analysis, and wrote the paper. All authors read and approved the final manuscript.