Objective: To investigate time courses of autoantibody profiles in patients with early arthritis.
Patients and methods: Two hundred patients with very early arthritis (<3 months duration), among them 102 patients with a final diagnosis of rheumatoid arthritis (RA) and 98 with other rheumatic diseases, were followed up for several years. First follow-up testing was done in all patients (mean 5 months from baseline), and 82 RA and 35 non-RA patients were available for last follow-up testing (mean 32 months from baseline). IgM-rheumatoid factor (RF) was measured by nephelometry, IgA-RF, IgG-RF and anti-citrullinated peptide antibodies (ACPA) by ELISA, and anti-RA33 antibodies were determined by immunoblotting.
Results: At baseline, IgA-RF was detectable in 29% and IgG-RF in 14% of RA patients while IgM-RF>50 IU/ml (RF50) was positive in 45% of the patients; specificities were 97%, 99% and 96%, respectively. However, the vast majority of IgA- or IgG-RF positive patients were also positive for RF50 or ACPA. During follow-up prevalence of ACPA slightly increased while prevalence of all RF subtypes and anti-RA33 decreased. Remarkably, the number of patients positive for RF50 and/or ACPA remained constant; and these patients had a highly increased risk for developing erosive disease, in contrast to patients solely positive for anti-RA33.
Conclusions: Testing for RF subtypes did not provide additional diagnostic information. Patients positive for RF50 and/or ACPA had an unfavourable prognosis, irrespectively of changes in the antibody profile during follow-up, whereas anti-RA33 positivity was inversely associated with erosiveness both at baseline and at later time points.