Background Early diagnosis and early treatment with DMARDs lead to better outcomes in rheumatoid arthritis (RA) (1,2). The 2010 ACR/EULAR criteria for RA were developed for early classification (3) and have good sensitivity, lower specificity and overall moderate accuracy. Their usefulness to lead treatment selection has not been investigated yet.
Objectives To compare clinical remission (CR) rates in patients with early polyarthritis in which the decision to start methotrexate (MTX) was based on the 1987 ACR vs 2010 ACR/EULAR criteria, over the first 12 months follow up.
Methods This is an observational non concurrent cohort study. Patients classified as RA or undifferentiated arthritis (UA) attending for the first time our early arthritis clinic (2005-2013) were eligible for inclusion. At baseline, before October 2010, patients classified as RA according to the 1987 criteria were treated with MTX (from 10 mg/wk up to 20 mg/wk), while patients with UA received hydroxychloroquine (HCQ) (1987 cohort). After October 2010, patients fulfilling the 2010 criteria received MTX (from 15 mg/wk up to 25mg/wk), while UA received HCQ (2010 cohort). Low-dose prednisone could be given according to clinician's decision. Patients were seen every 2 months in the first six months and every 3 afterwards; treatment was increased in order to achieve low disease activity (DAS28<3.2). CR (DAS28<2.6) was evaluated at every visit. Analyses were performed with a Cox proportional hazard regression analysis, and results presented as hazard ratios (HR) and 95% confidence intervals (CI).
Results Out of 676 patients, 467 were included in the 1987 cohort and 209 in the 2010 cohort. There were no significant differences between the two cohorts in terms of age, gender, VAS pain, RF and ACPA positivity. Patients in the 2010 cohort had significantly fewer median (IQR) tender (4 (2-8) vs 5 (2-10), p=0.018) and swollen joints (4 (2-7) vs 6 (3-10), p<0.0001) over 28 joints, ESR (19 (10-34) vs 22 (13-39), p=0.007) and CRP (0.4 (0.3-1.2) vs 0.7 (0.31-2.09), p=0.001), mean (SD) DAS28 (4.44 (1.14) vs 4.74 (1.25), p=0.005) and median (IQR) HAQ (0.75 (0.375-1.25) vs 1 (0.5-1.625), p=0.0001). Comparing the two cohorts, the 2010 cohort was more likely to achieve CR even when the analysis was limited to patients who strictly followed the protocol or actually received MTX, both in crude and adjusted analyses (Tab1).
Conclusions Patients with early arthritis in which the decision to start MTX is driven by the 2010 criteria achieve more often CR compared to those treated according to the 1987 criteria. Beside the limited diagnostic accuracy of the 2010 criteria, these results support their usefulness as treatment selection guidance in practice.
Sakellariou G et al, Ann Rheum Dis 2013;72:245-9
Breedveld FC et al, Arthritis Res Ther 2011;13:S3
Aletaha et al, Ann Rheum Dis 2010;69:1580-8
Disclosure of Interest None declared