Although the precise aetiology of rheumatoid arthritis (RA) remains elusive, evidence for autoimmunity is strong because several autoantibodies are associated with the disease. Besides the rheumatoid factor (RF), another group of autoantibodies was detected in the serum of patients with RA: the anti-cyclic citrullinated peptide (anti-CCP) antibodies. Recently, we compared the diagnostic value of the RF and anti-CCP antibodies in a consecutive cohort of patients with inflammatory joint symptoms (patients with and without RA) at a specificity of at least 98.5%, and we concluded that the anti-CCP antibodies were more sensitive than the RF (sensitivity anti-CCP antibodies 54.2–73.7% v sensitivity RF 12.8%).¹ Although there is an important overlap between the presence of RF and anti-CCP antibodies in the serum of patients with RA, evidence has suggested that the RF and anti-CCP antibodies are two separate autoantibody systems: (a) anti-CCP antibodies are associated with the shared epitope,^1–3 whereas the RF is not associated or less associated with the presence of the shared epitope¹; and (b) extra-articular manifestations are related to the RF¹ and not to anti-CCP antibodies.^1,2

A correlation between RF titres and clinical disease activity has been reported,^4,5 because RF titres decrease with successful treatment, particularly methotrexate or parenteral gold, this suggests an indirect link with disease activity.^4,6 Because infliximab treatment (tumour necrosis factor α (TNFα) blockade) has proved to be effective in the treatment of patients with refractory RA,⁷ and because we previously demonstrated that infliximab treatment induces new antibody reactivities, leading to antinuclear antibodies and anti-dsDNA antibodies,⁸ this study aimed at analysing the effect of infliximab treatment on pre-existing autoimmune profiles (IgM RF and anti-CCP antibodies). Furthermore, we investigated whether baseline IgM RF and anti-CCP antibodies are associated with changes in acute phase reactants (C reactive protein (CRP) and erythrocyte sedimentation rate (ESR)) during infliximab treatment.

PATIENTS AND METHODS

Patients and samples

Sixty two patients (24 men and 38 women; mean age at baseline 53.9 years (range 32–76)) with refractory RA, treated with infliximab at one centre (Department of Rheumatology, Ghent University Hospital, Ghent, Belgium) as part of an expanded access programme, were included in this study. All patients fulfilled the American College of Rheumatology classification criteria for RA.⁹ They received 3 mg/kg infliximab intravenously at weeks 0, 2, 6, and every 8 weeks thereafter in combination with methotrexate. After 30 weeks of infliximab treatment, all patients showed a clinical improvement of at least 20% according to American College of Rheumatology response criteria.¹⁰

Serum samples were collected before infusion at baseline and at week 30, and stored at –20°C until further analysis. Serum samples were collected after informed consent from the patient and approval by the local ethics committee had been obtained.

Detection of IgM RF

IgM RF assay using particles sensitised with rabbit IgG

This agglutination assay uses gelatin particles sensitised with denatured rabbit IgG (SERODIA-RA, Fuijrebio Inc, Tokyo, Japan). After incubation with progressively diluted human serum samples, the plates were inspected for observable agglutination.

IgM RF assay using particles sensitised with human IgG

This assay is based on latex particles sensitised with human IgG (RFscan Latex assay, BD Diagnostic Systems, Sparks, Maryland, USA). After incubation with progressively diluted human serum samples, the plates were inspected for observable agglutination.

Detection of anti-CCP antibodies

The anti-CCP1 enzyme linked immunosorbent assay (ELISA; Immunoscan RA, Eurodiagnostica, Arnhem, The Netherlands) uses plates coated with highly purified synthetic peptides containing citrulline residues. The test was performed according to the manufacturer’s instructions.

Statistical analysis

Statistical analysis was performed using SPSS 10.0 software (SPSS, Chicago, IL). We used the Wilcoxon signed rank test for paired analysis and the Mann-Whitney U test for comparison of patient groups with and without a marked decrease of the acute phase reactants. Correlations were sought using Spearman’s correlation coefficients (r_s). Values of p<0.05 were considered significant.

RESULTS

IgM RF, but not anti-CCP antibodies, is modulated by infliximab treatment

IgM RF assay using particles sensitised with rabbit IgG

As shown in table 1 for the total cohort and for different cut off points, the RF titre was significantly reduced during infliximab treatment. Furthermore, the majority of patients with RA showed a decrease in RF titre (n = 32), including a reduction of at least two titre steps in 9/32 patients, whereas in nine patients the RF titre increased, including an increase of at least two titre steps in only one patient, and in 21 patients the RF titres did not change.

IgM RF assay using particles sensitised with human IgG

For this alternative RF assay, a similar, significant reduction in RF titre was observed (as illustrated for the total cohort and for different cut off points in table 1). The majority of patients had a decrease in RF titre (n = 33), including a reduction of at least two titre steps in 19/33 patients, whereas the RF titre increased in 15 patients, including an increase of at least two titre steps in 8/15 patients, and remained unchanged in 14 other patients.

Anti-CCP antibodies

A comparison of the concentrations of anti-CCP antibodies at baseline and after 30 weeks of infliximab treatment showed no significant differences, indicating that anti-CCP antibodies are not modulated by infliximab treatment (as given for the total cohort and for the manufacturer’s cut off point in table 1). Moreover, 38/62 patients with RA showed a decrease in concentrations of anti-CCP antibodies, whereas 24/62 patients had an increase after infliximab treatment. Only 23/38 patients had a marked reduction (at least 20% decrease) in anti-CCP antibody levels, whereas in 13/24 patients the concentrations of anti-CCP antibodies increased by at least 20%.

Furthermore, we observed no correlations between changes in RF and changes in anti-CCP antibodies during infliximab treatment (for RF assay using particles sensitised with rabbit IgG: r_s = 0.015, p = 0.908; for RF assay using particles sensitised with human IgG: r_s = −0.083, p = 0.529), supporting the hypothesis that RF and anti-CCP antibodies are two independent autoantibody systems in RA.

IgM RF, but not anti-CCP antibodies, is associated with changes in acute phase reactants during infliximab treatment

We further investigated whether IgM RF and anti-CCP antibodies at baseline are predictive of the biological response during infliximab treatment. In patients with increased CRP (>10 mg/l, n = 44) or ESR (>20 mm/1st h, n = 37), we analysed the correlations between titres of autoantibodies at baseline and the percentage change in acute phase reactants during infliximab treatment. Table 2 shows that a significant inverse correlation was found between the baseline IgM RF titres measured by an agglutination assay using particles sensitised with human IgG and the percentage change in CRP and ESR, indicating that patients with high baseline IgM RF titres have a less pronounced decrease in acute phase reactants. This was further confirmed by a similar trend using the other RF assay (table 2). In contrast, baseline concentrations of anti-CCP antibodies did not correlate significantly with changes in CRP or ESR during infliximab treatment (table 2).

Furthermore, we analysed the baseline autoantibody titres in the patients with and without a marked decrease of at least 20% in CRP and ESR during infliximab treatment. We observed significantly lower IgM RF titres in the patients with a marked decrease in acute phase reactants than in those with a less pronounced decrease in CRP and ESR (table 3). Again, no statistically significant differences were found in baseline concentrations of anti-CCP antibodies (table 3).

DISCUSSION

The importance of the different RA associated antibodies as diagnostic markers for RA has been extensively analysed, and a valid comparison clearly indicates that anti-CCP antibodies show a better diagnostic performance than RF.^1,11–13 However, few data are available on the relationship between antibody titres and response to treatment. Some studies describe a decrease in RF titres during successful treatment with methotrexate or parenteral gold,^4,6 but until now, changes in RF titres during anti-TNFα treatment have only been analysed in two studies (by Maini et al and Charles et al), who reported a decrease in RF titres in infliximab treated patients with RA.^14,15

Our study describes the effect of TNFα blockade on both the RF and anti-CCP antibodies in the same cohort of patients with RA. Interestingly, a clearly different effect of infliximab treatment on the IgM RF and anti-CCP antibodies was found: RF titres decreased significantly (confirming the results reported by Maini et al and Charles et al), but concentrations of anti-CCP antibodies did not change. These remarkable findings indicate that anti-CCP antibodies could act as a disease-specific marker for RA and are not modulated by infliximab treatment, whereas IgM RF titres could be related to disease activity as suggested by the decrease in IgM RF titres during infliximab treatment.

Another important topic in view of these new biological treatments such as TNFα blockade is the identification of biomarkers that predict response to treatment. When analysing the RA associated autoantibodies in relation to changes in acute phase reactants, we observed an inverse correlation between the baseline IgM RF titres and the changes in CRP and ESR. In contrast, the baseline concentrations of anti-CCP antibodies did not correlate significantly with the changes in acute phase reactants during infliximab treatment. In other words, patients with RA with high baseline IgM RF titres have a less pronounced decrease of CRP and ESR. These findings were confirmed by the lower baseline RF titres in patients with a marked decrease in acute phase reactants compared with those with a less pronounced decrease in CRP and ESR.

In conclusion, both the differential effect of infliximab treatment on IgM RF and the anti-CCP antibodies, and the different predictive value on the changes in acute phase reactants during infliximab treatment add support to the existing evidence that RF and anti-CCP antibodies are two different, independent autoantibody systems in RA. Our data indicate that the RF and anti-CCP antibodies may provide different and, eventually, complementary biological information on the disease process in RA.