Objective: To evaluate the predictive value of clinical, biological and radiological parameters for the prognosis of rheumatoid arthritis (RA) in a community-recruited cohort.
Methods: Ninety-one patients (mean age 49 yr, female/male ratio 2.9) with RA of limited duration (median 2 yr), 80% recruited from the community, were prospectively enrolled in 1996 (T1) and followed until 1999 (T2). Data collected at T1 were demographic characteristics, Ritchie articular index (RAI), extra-articular manifestations, Health Assessment Questionnaire (HAQ) score, C-reactive protein (CRP) and autoantibodies (autoAbs) [rheumatoid factors (RF), detected by latex fixation test and ELISA (IgM, IgA and IgG isotypes), anti-filaggrin, detected by immunofluorescence (anti-keratin antibodies, AKA; anti-perinuclear factor antibodies, APF) and ELISA (anti-citrullinated rat filaggrin antibodies, ACRFA), anti-Sa, anti-calpastatin recognizing the 27 C-terminal fragment (ACAST-C27) and domain I (ACAST-DI), anti-cardiolipin (ACL), antineutrophil cytoplasmic antibodies (ANCA), anti-annexin V (aANX V) and anti-Ro]. Hands were radiographed at T1 and T2, and read using the Sharp method as modified by van der Heijde. The main assessment criterion was progression of radiologically detected damage between T1 and T2.
Results: At T1, RA activity was mild (RAI 11/78; mean CRP 14 mg/ml), with minor functional disability (HAQ 0.8/3) and mild X-ray destruction (mean total Sharp score 9.2/280). At T1, 96% of the patients were on treatment (prednisone 72%, DMARDs 95%). The latex test detected autoAb in 46% of patients, RF-IgM was detected in 51%, RF-IgA in 36%, RF-IgG in 32%, AKA in 33%, APF in 45%, ACRFA in 45%, ACAST-C27 in 14%, ACAST-DI in 5%, anti-Sa in 22%, ACL in 3%, ANCA in 28%, aANX V in 9% and anti-Ro in 2%. At T2, the mean total Sharp score was 22.9. According to univariate analysis, T1 parameters associated with the independent variable were RAI, HAQ, CRP, latex test positivity and T1 Sharp scores. Multivariate analysis retained only latex test positivity and, to a lesser degree, joint-space narrowing score as independent predictors of radiological progression.
Conclusion: RF is the main factor that can predict radiological progression in community cases of RA of limited duration.