Introduction
Rheumatoid arthritis (RA) is the most common chronic inflammatory joint disease, with a worldwide prevalence of 0.5% to 1%. RA is characterized by synovial joint inflammation, which often leads to progressive joint destruction and disability [
1]. Early treatment improves the outcome and therefore early diagnosis is crucial. Rheumatoid factors (RFs) were the first biological markers discovered for RA and remain the only laboratory criterion included in the American College of Rheumatology criteria for RA classification [
2]. Two major disadvantages of RF are low specificity and possible absence in the first year of the disease [
3]. Several other auto-antibodies specific to RA have been found. Among them, anti-filaggrin antibodies, anti-keratin antibodies (AKAs), and anti-perinuclear factor (APF) exhibit a high specificity and sometimes present early in the disease. However, AKA is not sufficiently sensitive to be used for diagnosing RA. APF detection is available only at specialized laboratories, as obtaining and standardizing the substrate is technically challenging, and interpreting the immunofluorescence results is largely subjective [
4].
Anti-filaggrin antibodies recognize citrulline residues formed by post-transcriptional modification of arginine by peptidylarginine deiminase [
5]. Enzyme immunoassays (EIAs) based on synthetic cyclic citrullinated peptides (CCP2) are available for detecting anti-CCP. We and others previously showed that the sensitivity of these antibodies was about 80% in established RA [
6,
7] compared with 55% in early RA [
6,
8] and 40% in very early RA [
9,
10]. The usefulness of anti-CCP for monitoring RA patients, particularly during treatment, is controversial. A significant decrease was found after 6 months of tumor necrosis factor (TNF) antagonist therapy in one study [
11], whereas decreases were slow and inconsistent during infliximab therapy in two other studies [
12,
13].
Antibodies to other citrullinated peptides or proteins have been suggested as good candidates for diagnosing RA. Vimentin is an intermediate filament that is widely expressed by mesenchymal cells and macrophages and easy to detect in the synovium. Modification of the protein occurs in macrophages undergoing apoptosis, and antibodies to citrullinated vimentin may emerge if the apoptotic material is inadequately cleared [
14]. The first antibodies to citrullinated vimentin described in the literature were anti-Sa antibodies detected by Western blot, which were as specific as anti-CCP but not sufficiently sensitive (20% to 45%) to serve as diagnostic tools [
15,
16]. Recombinant mutated citrullinated vimentin (MCV) was recently produced, and an enzyme-linked immunosorbent assay (ELISA) was developed for detecting anti-MCV. Few data are available on the performance of anti-MCV for diagnosing RA. In most studies, anti-MCV and anti-CCP testing produced similar results [
17‐
20]. Two studies, however, suggested that anti-MCV might be more sensitive than anti-CCP [
21,
22].
The aims of this study were to evaluate the usefulness of anti-MCV for diagnosing RA in anti-CCP-negative patients and to monitor anti-MCV titres during infliximab therapy for RA. First, we compared the results of anti-MCV tests in RA patients with and without positive tests for anti-CCP. Then, we obtained serial anti-MCV assays in RA patients receiving infliximab therapy.
Materials and methods
Patients
We studied 156 patients seen at the Rheumatology Department of the Bichat-Claude Bernard Teaching Hospital (Paris, France) for RA meeting American College of Rheumatology (ACR) criteria. Among them, 20 had early RA (<12 months). The control group was composed of 50 healthy controls and 158 patients seen at the same department for other diseases, including psoriatic arthritis (n = 51), primary Sjögren syndrome (n = 58), and ankylosing spondylitis (n = 49).
Among the RA patients, 23 were evaluated after 6, 12, 18, and 24 months of infliximab therapy. All 23 patients had a history of inadequate response or tolerance to at least one conventional disease-modifying antirheumatic drug (DMARD). Previous treatment with DMARDs (methotrexate, hydroxychloroquine, azathioprine, or sulfasalazine), steroids, and nonsteroidal anti-inflammatory drugs could be continued provided that the dosage was kept stable for at least 4 weeks before infliximab initiation and throughout infliximab therapy. Infliximab was administered (3 mg/kg) at baseline, 2 and 6 weeks later, and every 8 weeks thereafter in combination with a DMARD. None of the patients had active or latent tuberculosis, other infections, or severe comorbidities.
The following variables were collected at baseline and then after 3, 6, 12, 18, and 24 months: tender and swollen joint counts of a total of 28 joints, score on a visual analog scale for global health completed by the patient, erythrocyte sedimentation rate, and C-reactive protein level. The disease activity score using 28 joint counts (DAS28) was computed at 0, 6, and 12, months; the amount of missing data was too large for reliable DAS28 determination at the 3-month time point. Responders were defined at different time points as having a DAS28 decrease of greater than 1.2 with a DAS28 value of less than 3.2, and nonresponders were defined as having a DAS28 decrease of less than 0.6 or a decrease in the 0.6 to 1.2 range with a score value of greater than 5.1 [
23]. Data for responders and nonresponders were compared only at 6 and 12 months as the number of nonresponder patients available at 18 and 24 months was too low for a statistical analysis.
The patients were informed of the purpose of the study and gave their informed consent. The trial was approved by the ethics committee of our hospital. All procedures were conducted in accordance with the hospital's ethics rules. All the sera were stored at -20°C until the assays.
Methods
IgM-RF was determined by nephelometry using a BNII analyser (N RF Latex; Dade Behring, Paris La Défense, France). Levels greater than 20 IU/mL were considered positive. Anti-CCP2 was assayed using an EIA (Immunoscan RA; Euro-Diagnostica, Arnhem, The Netherlands) in accordance with the instructions of the manufacturer. Titres lower than 25 units were considered negative. Anti-MCV levels were determined using a commercial ELISA (Orgentec Diagnostika, Mainz, Germany) with MCV as the antigen. Briefly, sera diluted 1:100 were incubated for 30 minutes on the coated plate, which was then washed before the addition of horseradish peroxidase-labelled goat anti-human IgG for 15 minutes. The reaction was revealed by the addition of TMB (3,3', 5,5'-tetramethylbenzidine) substrate, and color intensity was measured at 450/620 nm. Values greater than 20 U/mL were considered positive, which was in accordance with the instructions of the manufacturer.
Statistical analysis
The chi-square test was used to compare the specificities of anti-MCV, anti-CCP, and IgM-RF. The nonparametric Wilcoxon signed rank test was used for paired comparisons of changes in anti-CCP and anti-MCV titres during infliximab treatment for all of the RA patients as well as after a separation between responders and nonresponders. The Mann-Whitney test was used to compare anti-CCP, anti-MCV, and IgM-RF titres at baseline in responders and nonresponders. The Spearman rank correlation test was used to study the correlation between the DAS28 and anti-CCP or anti-MCV titres at baseline. A P value of 0.05 or less was considered statistically significant.
Discussion
We evaluated the interest of antibodies to the citrullinated protein MCV in anti-CCP negative patients who met ACR criteria for RA. At comparable specificity, of the 63 patients with negative tests for both anti-CCP and RF, 6 (7.9%) were positive for anti-MCV. Moreover, anti-MCV titres decreased faster than did anti-CCP during infliximab therapy and showed a closer association with the DAS28 response.
Anti-CCP assays are effective and widely used for diagnosing RA. However, their sensitivity is limited to 40% in patients with early RA [
9,
10]. CCP is not expressed in the synovium, and citrullinated proteins expressed in the rheumatoid joint would probably be more relevant as targets of auto-antibodies used to diagnose RA. Citrullinated vimentin is present in synovial membranes and is released in increased amounts in response to growth factors and proinflammatory cytokines, suggesting involvement in the pathophysiology of RA [
24]. The few studies of anti-MCV usually found performance characteristics similar to those of anti-CCP for diagnosing RA [
17‐
20]. For Bang and colleagues [
21], anti-MCV antibodies were even more sensitive than anti-CCP.
The most useful characteristic of anti-CCP in clinical practice is high specificity for RA. Therefore, we assessed the specificity of anti-MCV in controls and a large group of patients with established inflammatory rheumatic diseases. At the cutoff values recommended by the manufacturer, the specificity was lower than that of anti-CCP, as previously reported [
20,
25], and we adjusted the cutoff to obtain the same specificity as that of anti-CCP. Longer follow-up will determine whether positive anti-MCV assays, observed in several non-RA patients, predict subsequent RA as shown for anti-CCP [
26,
27].
An important finding from our study is that 11.8% of anti-CCP-negative RA patients were positive for anti-MCV. Furthermore, in 7.9% of our patients meeting ACR criteria for RA but having negative tests for anti-CCP, anti-MCV antibodies were positive although RF was negative. Similarly, 13.2% of anti-CCP-negative patients who also tested negative for anti-MCV had positive RF titres. In contrast, our results showed that an equivalent number of anti-CCP-positive RA patients (11.2%) tested negative for anti-MCV and among them 5% were also RF-negative. Therefore, the best diagnostic strategy may be to assay both anti-MCV and RF in anti-CCP-negative RA patients without replacing them. Of the 13 patients who had early RA and negative tests for anti-CCP, 3 had anti-MCV antibodies, which is in keeping with previous evidence that anti-MCV may help to improve the early diagnosis of RA [
22].
We found that 9 anti-CCP-negative patients tested positive for anti-MCV (among them, 3 also tested positive for RF). Follow-up of this subgroup will determine whether anti-CCP antibodies emerge later on and whether the anti-CCP-negative anti-MCV-positive profile is of prognostic significance as suggested by Mathsson and colleagues [
22]. Indeed, a recent report suggests that anti-MCV antibodies are better predictors of disease activity than anti-CCP [
28]. Conceivably, the anti-CCP-negative anti-MCV-positive profile may be associated with specific gene polymorphisms. An association between the presence of RF and the mutation of the protein tyrosine phosphatase nonreceptor type 22 (
PTPN22) gene has been evidenced [
29]. Anti-CCP antibodies were associated with the
PTPN22 1858 C/T polymorphism [
30] and the HLADRB1 allele shared epitope [
31].
Monitoring auto-antibody titres has been suggested as a means of assessing treatment efficacy. Several studies evaluated the effect of TNF antagonist treatment on anti-CCP titres in RA patients. A significant decrease was found after 6 months of TNF antagonist therapy in one study [
11], whereas decreases were slow and inconsistent during infliximab therapy in two other studies [
12,
13]. Thus, anti-CCP seems to be of limited value for monitoring the effect of TNF antagonist therapy. Our study provides the first data on changes in anti-MCV titres during infliximab therapy for RA. We found a significant decrease in anti-MCV titres, which antedated the anti-CCP decrease; the former was first significant after 18 weeks and the latter after 24 weeks. Moreover, the anti-MCV decrease was related to the DAS28 response after 12 months, which is not the case for IgM-RF as their decrease is significant in responder as well as nonresponder RA patients. Previous reports showed a decrease of IgM-RF during anti-TNF treatment [
11‐
13]. Association to clinical improvement was reported [
12,
32] at 6 months based essentially on the ACR response and not on the DAS28 response.
Most of the patients treated with infliximab in our study had long-standing RA. A disease duration of less than 1 year predicted a greater reduction in anti-CCP antibody titres with conventional treatment [
33], suggesting that a study specifically designed to monitor anti-MCV titres in patients with early RA treated with TNF antagonists might be of interest. However, a recent report of Ursum and colleagues [
34] found a very low correlation between the DAS28 and anti-MCV levels during a 2-year follow-up of early RA, suggesting that other parameters of the response are involved and should be studied.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
PNR contributed to the conception of the study and to the analysis and interpretation of data and was the main contributor to the preparation of the manuscript. SGM contributed to the acquisition and analysis of the data and to the preparation of the manuscript. AB contributed to the analysis of the data from infliximab-treated patients. MH contributed to the acquisition and analysis of the data. EP, GH, and PD were clinical investigators who contributed to the recruitment and treatment of the patients. OM was the main clinical investigator and contributed to the recruitment of the patients and to the critical review of the manuscript. SCM contributed to the interpretation of data and to the critical review of the manuscript. All authors read and approved the final manuscript.