Synovial fluid levels of anti–cyclic citrullinated peptide antibodies and IgA rheumatoid factor in rheumatoid arthritis, psoriatic arthritis, and osteoarthritis

Authors


Abstract

Objective

To assess the levels of anti–cyclic citrullinated peptide (anti-CCP) and IgA rheumatoid factor (IgA-RF) in synovial fluids of patients with rheumatoid arthritis (RA), psoriatic arthritis (PsA), and osteoarthritis (OA).

Methods

Knee effusions of 29 patients with RA (23 women, 6 men; mean ± SD age 60 ± 15 years), 20 with PsA (6 women, 14 men; mean age 51 ± 12 years), and 19 with OA (9 women, 10 men; mean age 73 ± 11.8 years) were aspirated, tested for white blood cell (WBC) counts, centrifuged, and stored at −20°. Sera of 22, 11, and 12 of these patients with RA, PsA, and OA, respectively, were similarly stored. IgG anti-CCP and IgA-RF were detected by enzyme-linked immunosorbent assay. Erythrocyte sedimentation rate and C-reactive protein levels were used as measures of disease activity.

Results

Mean levels of synovial fluid anti-CCP and IgA-RF were significantly increased in RA joint effusions compared with PsA and OA (anti-CCP: 150 ± 134, 34 ± 29, and 24 ± 26 units, respectively [P < 0.003]; IgA-RF: 76 ± 77, 15.7 ± 10, and 18 ± 20 units, respectively). No significant difference was noted between OA and PsA. A significant correlation was found between synovial fluid anti-CCP and serum anti-CCP and IgA-RF. In patients with RA, a significant correlation was found between synovial fluid WBC counts and IgA-RF (P = 0.03) and serum IgA-RF (P = 0.008), but not between synovial fluid and serum anti-CCP levels. In RA patients, C-reactive protein correlated with serum IgA-RF.

Conclusion

Anti-CCP and IgA-RF were significantly increased in synovial fluid of RA in comparison with PsA and OA patients.

INTRODUCTION

Rheumatoid arthritis (RA) is a systemic autoimmune disease characterized by chronic inflammation of the joints, eventually resulting in erosive changes and joint deformities. The diagnosis of RA is primarily based on clinical manifestations of the disease, supported in many cases by serologic findings. Rheumatoid factor is present in up to 70% of patients but is also detected in a variety of autoimmune disorders, other nonrheumatic conditions, and in healthy individuals (1). According to the specific test used, antibodies directed toward citrullinated proteins have been described in 60–80% of patients with RA with a specificity of 85–95% (2, 3).

Citrullinated proteins seem to originate in the synovium (4), and anti–cyclic citrullinated proteins (anti-CCP) appear to be produced in the inflamed synovium by local plasma cells (5). Furthermore, anti-CCP–producing B cells have been detected in the synovial fluid of anti-CCP–positive patients with RA (6).

It has recently been shown that the presence of citrullinated peptides is not specific to the rheumatoid synovium and may be locally detected in other inflammatory arthropathies (7). However, although citrullinated proteins are present in the synovium of mice with collagen-induced arthritis, the induction of autoantibodies directed to these proteins is a more specific phenomenon, detectable only in human patients with RA (4). Therefore, the presence of CCPs in the synovium of different arthropathies does not preclude a specific local humoral response of the inflamed synovium to CCP, which may vary in different arthritides or be correlated with the degree of inflammation. The aim of this study was to measure the levels of anti-CCP and IgA rheumatoid factor (IgA-RF) in joint effusions of patients with RA, psoriatic arthritis (PsA), and osteoarthritis (OA) to evaluate the specificity and possible diagnostic value of anti-CCP and IgA-RF, as well as their correlations with parameters of disease activity.

PATIENTS AND METHODS

Twenty-nine, 20, and 19 consecutive patients with RA, PsA, and OA, respectively, participated in this study. All the patients underwent an arthrocentesis of the knee due to acute synovitis. The synovial fluid sample was tested for white blood cell (WBC) counts and culture; the rest of it was centrifuged and the supernatant was frozen at −20°. Sera of 22, 11, and 12 of the same patients with RA, PsA, OA, respectively, were frozen at −20°.

Anti-CCP enzyme-linked immunosorbent assay.

IgG anti-CCP of the synovial fluid and serum was performed using a second-generation commercial kit (Quanta lite; Inova Diagnostics, San Diego, CA), according to the manufacturer's instructions. Serial dilution of the synovial fluids was performed (1:10, 1:100, and 1:1,000). Synovial fluid samples were diluted 1:10 and serum samples 1:100. Samples were considered weakly positive if the antibody titer was between 20 and 39 IU, moderately positive between 40 and 59 IU, and strongly positive ≥60 IU.

IgA rheumatoid factor.

IgA-RF of the synovial fluid and serum was performed using a commercial kit (Quanta lite; Inova Diagnostics), according to the manufacturer's instructions. Synovial fluid samples were diluted 1:10 and serum samples 1:100. Samples were considered positive if the antibody titer was ≥6 IU. Erythrocyte sedimentation rate (ESR; Westergren) and C-reactive protein (CRP) were used as measures of disease activity.

Statistical analysis.

Comparisons between the 3 groups of patients (PsA, OA, and RA) with regard to demographic and clinical factors were carried out using Kruskal-Wallis nonparametric analysis of variance and chi-square tests as applicable. Gabriel and Games-Howell post-hoc multiple comparisons between the 3 groups were performed on the ranks of the original levels of the parameters. Spearman's correlation coefficients were calculated for all continuous parameters for each group separately. Receiving operating characteristic curves were compared using the area under the curve criteria. The statistical significance level was set to 0.05 and the SPSS for Windows software, version 12.0 (SPSS, Chicago, IL) was used for the analysis.

RESULTS

The demographic and clinical characteristics of the patients are summarized in Table 1. Most of the patients with RA were women, whereas the patients with PsA and OA were predominantly men. As expected, the patients with OA were older, with the mean ± SD age of the RA, PsA, and OA groups being 60 ± 15 years, 51 ± 12 years, and 73 ± 11.8 years, respectively. The 3 groups had similar mean disease duration. Sixty-three percent of the patients with RA were seropositive for IgM-RF, whereas none of the patients with OA and PsA were found to be positive (Table 2).

Table 1. Demographic and clinical characteristics of the patients
CharacteristicRheumatoid arthritisPsoriatic arthritisOsteoarthritis
Female/male23/66/149/10
Age, mean ± SD years60 ± 1551 ± 1273 ± 11.8
Disease duration, mean ± SD years11.7 ± 68.9 ± 58.7 ± 8
Table 2. Values of anti-CCP and IgA-RF in synovial fluids and serum of patients with rheumatoid arthritis, psoriatic arthritis, and osteoarthritis*
CharacteristicRheumatoid arthritisPsoriatic arthritisOsteoarthritis
  • *

    Values are the mean ± SD. Normal values of anti-CCP: negative <20 IU, weak positive 20–39 IU, moderate positive 40–59 IU, strong positive >60 IU. Normal values of IgA-RF: <6 IU. Anti-CCP = anti–cyclic citrullinated peptide; IgA-RF = IgA rheumatoid factor; CRP = C-reactive protein; ESR = erythrocyte sedimentation rate.

  • P < 0.0001.

Synovial fluid anti-CCP150 ± 13434 ± 2924 ± 26
Serum anti-CCP115 ± 12062 ± 9413 ± 12
Synovial fluid IgA-RF76 ± 7715.7 ± 1018 ± 20
Serum IgA-RF33.6 ± 446.5 ± 34.9 ± 2
CRP, mg/liter14 ± 373 ± 30.2 ± 0.08
ESR, mm/hour51 ± 2330 ± 1749 ± 37

Levels of anti-CCP and IgA-RF in the serum and synovial fluids of the patients.

The mean ± SD levels of synovial fluid anti-CCP and IgA-RF were significantly increased in RA joint effusions compared with PsA and OA (anti-CCP: 150 ± 134 units, 34 ± 29 units, and 24 ± 26 units, respectively; P < 0.003). Thus, the mean anti-CCP for patients with RA was above the upper normal limit of 60 units, whereas it was below this level for patients with PsA and OA. The mean ± SD level of IgA-RF was 76 ± 77 units in the RA group, 15.7 ± 10 units in the PsA group, and 18 ± 20 units in the OA group (P < 0.008). Although IgA-RF was significantly higher in patients with RA in comparison with those with PsA and OA, the mean level of IgA-RF was above the normal limit (6 units) in the 3 groups, suggesting that synovial fluid IgA-RF may not discriminate between the 3 groups. No significant difference was noted between patients with OA and PsA (Table 2).

Serum levels of anti-CCP and IgA-RF were significantly increased in patients with RA in comparison with those with PsA and OA, although the serum levels of anti-CCP in patients with PsA were also elevated. A significant correlation was found between anti-CCP in serum and synovial fluids (P < 0.0001) (Figure 1), IgA-RF in serum and synovial fluids (P < 0.0001) (Figure 2), and between anti-CCP and IgA-RF in serum and synovial fluids.

Figure 1.

Correlation between serum and synovial fluid anti–cyclic citrullinated peptide (anti-CCP).

Figure 2.

Correlation between serum and synovial fluid IgA rheumatoid factor (IgA-RF).

ESR, CRP, and synovial fluid WBC.

The levels of ESR, CRP, and synovial WBC were significantly lower in patients with OA, whereas no difference was found between patients with RA and those with PsA (Table 2). A significant correlation was found between CRP levels and serum IgA-RF levels in patients with RA. Likewise, WBC counts in the synovial fluid clearly correlated with serum IgA-RF (P = 0.008) and synovial fluid IgA-RF (P = 0.03) levels, but not with serum and synovial fluid anti-CCP levels. Serum levels of CRP correlated with serum IgA-RF.

Sensitivity and specificity of serum and synovial fluid levels of anti-CCP and IgA-RF.

Serum anti-CCP levels ≥40 IU (moderately and strongly positive according to the recommendations of the manufacturer [Inova Diagnostics]) had a sensitivity of 58% with a specificity of 84% for the diagnosis of RA. The combination of serum and synovial fluid anti-CCP levels ≥40 IU increased the sensitivity of the test up to 74%, mildly lowering the specificity to 74%.

A serum level of IgA-RF >6 IU had a sensitivity of 63% with a specificity of 78% for the diagnosis of RA. The combination of serum and synovial fluid IgA-RF >6 IU provided a sensitivity of 84% on the count of the specificity (50%).

Receiver operating curves for anti-CCP and IgA-RF.

Using the area under the curve, we calculated cutoff values with the best profile of specificity and sensitivity for synovial fluid anti-CCP and IgA-RF. Using this method, a cutoff of 21 for IgA-RF had a sensitivity of 80% with a specificity of 70%. A cutoff of 31 for anti-CCP was associated with a sensitivity of 80% and a specificity of 78%.

DISCUSSION

Antibodies directed toward citrullinated proteins are highly specific for RA, but their sensitivity is around 65–75% (2, 3). In our study, we have shown that levels of anti-CCP and IgA-RF are significantly increased in the serum and synovial fluid of patients with RA in comparison with another inflammatory joint effusion condition (PsA) and a noninflammatory synovial fluid condition (OA). Interestingly, mean serum levels of anti-CCP in patients with PsA were above the normal limits, confirming recent studies on the possible presence of anti-CCP in the serum of patients with PsA (8). We could clearly show that the combined positivity of anti-CCP in both serum and synovial fluids increased the sensitivity of the test for the diagnosis of RA from 58% to 74%.

Although it was previously thought that the presence of tissue citrullinated proteins in the synovium is characteristic of RA (9), it has recently be shown that these proteins are also found in the synovium of other arthritides such as OA and reactive arthritis (7). The significant finding of increased anti-CCP levels in the synovial fluid suggests that it is the humoral immunologic response to these proteins that characterizes RA rather than the presence of the triggering antigen in the synovium. This capacity to develop antibodies to CCP seems to be confined to humans because murine synovium does not have the capacity to produce anti-CCP (4).

RF constitutes one of the diagnostic criteria of RA proposed by the American College of Rheumatology (formerly the American Rheumatism Association) (10). However, RF positivity shows low diagnostic specificity (1). Among the various RFs, IgA-RF has a similar sensitivity (65%) to IgM-RF (11) but has a higher correlation with extraarticular manifestations of RA, including sicca (12), and with erosive disease (13). The specificity of serum anti-CCP is significantly increased when combined with IgA-RF in the early diagnosis and prognosis of RA (14, 15). Data on the presence and significance of RFs in the synovial fluid is scarce. IgM-RF and IgG-RF have been detected in the synovial fluid of patients with RA (16) but there are no data concerning IgA-RF, although it has been shown that synovial cells have the capacity to produce IgA-RF in vitro (17). Similarly, there are no clinical data on anti-CCP in the synovial fluid of patients with RA or other joint diseases, although experimental data on this subject are scarce and concern mainly the synovial tissue (7). In our present study, we have demonstrated that anti-CCP is present in the synovial fluid of patients with RA, and that similar to IgA-RF, it is significantly increased in comparison with that of patients with PsA and OA, and clearly correlates with serum IgA-RF and serum and synovial fluid anti-CCP. This first study of synovial fluid anti-CCP and IgA-RF in the diagnosis of RA, although limited by the relatively small number of patients, documented significantly increased levels of anti-CCP and IgA-RF in patients with RA and may suggest a previously untested discrimination between different arthritides.

Ancillary