Anti-cyclic citrullinated peptide (CCP) antibodies are very specific for the development of rheumatoid arthritis (RA) in patients with very early synovitis. 1 Their predictive value is underscored by their high weighting in an algorithm validated in patients with undifferentiated synovitis of≤ 3 months’ duration. 2 However, using this algorithm, one cannot accurately predict the outcome in 25% of patients and additional predictive markers are needed. We have reported that antibodies against type II collagen were unhelpful in this phase of the disease. 3 Several groups have measured anti-modified citrullinated vimentin (MCV) antibodies in patients with RA. In patients with symptoms of< 12 months’ duration, the specificity and sensitivity of anti-MCV were 95% and 71%, respectively. 4 The enhanced sensitivity compared with anti-CCP2 raised the possibility of its clinical utility. Subsequently, in patients with synovitis of< 2 years’ duration, a specificity and sensitivity of 83% and 62%(vs 93% and 57%, respectively, for anti-CCP2) have been reported. 5 We assessed the utility of anti-MCV in patients with clinically apparent synovial swelling and inflammatory joint symptoms of≤ 3 months’ duration. Ethical permission and informed consent were obtained. Patients were followed for 18 months and assigned to their final diagnostic groups as previously described. 6 Anti-MCV antibodies were measured using a commercial ELISA kit (Orgentec Diagnostika, Mainz, Germany) according to the manufacturer’s instructions in serum which had been stored at− 80 C. A level of≥ 20 U/ml was considered positive. 4 5 Anti-CCP2 was also measured (ELiA, Phadia, Germany); anti-CCP2 data were not available for two patients (both non-RA).

Samples were available from 175 patients: 63 patients developed RA7 and 112 did not (70 unclassified, 10 reactive arthritis, 10 psoriatic arthritis, 10 crystal arthritis, 4 inflammatory bowel disease-related arthritis, 8 other). Anti-MCV antibodies were positive in 34 patients who developed RA (54%)(figure 1). Patients with anti-MCV antibody-positive early RA had a median level of 377 U/ml. Of these 34 patients, 29 were positive for anti-CCP2 antibodies. Anti-MCV levels in the 5 patients negative for anti-CCP2 antibodies were towards the lower end of the positive range (23, 26, 36, 41 and 114 U/ml). Anti-MCV antibodies were present in an additional 10 non-RA patients (1 crystal arthritis, 2 reactive arthritis, 3 psoriatic arthritis, 4 unclassified arthritis, figure 1). Two of these 10 (with the highest anti-MCV levels), but no other non-RA patients, were positive for anti-CCP2 antibodies; both were classified as psoriatic arthritis, one of them positive for rheumatoid factor. The anti-MCV levels in the patients with antibody-positive non-RA were lower than in the patients with antibody-positive early RA (median 34 vs 377 U/ml; p= 0.002, Mann–Whitney U test). Significantly more patients with early RA than patients without RA were positive for anti-MCV antibodies (p< 0.001, χ2 test). The diagnostic performance of the anti-MCV and anti-CCP2 antibody tests for the development of RA is shown in table 1. There was no relationship between erythrocyte sedimentation rate, C-reactive protein, tender or swollen joint counts and anti-MCV antibody levels in patients who developed RA (Spearman test; data not shown).