Reductions in Radiographic Progression in Early Rheumatoid Arthritis Over Twenty-Five Years: : Changing Contribution From Rheumatoid Factor in Two Multicenter UK Inception Cohorts
McWilliams, Daniel F.
Rennie, Kirsten L.
Walsh, David Andrew
on behalf of the Early Rheumatoid Arthritis Study and the Early Rheumatoid Arthritis Network
Objective: To assess the 5-year progression of erosions and joint space narrowing (JSN) and their associations with rheumatoid factor (RF) status in 2 large, multicenter, early rheumatoid arthritis cohorts, spanning 25 years. Methods: Radiographic joint damage was recorded using the Sharp/van der Heijde (SHS) method in the Early Rheumatoid Arthritis Study (ERAS), 1986–2001, and the Early Rheumatoid Arthritis Network (ERAN), 2002–2013. Mixed-effects negative binomial regression estimated changes in radiographic damage over 5 years, including erosions and JSN, separately. RF, along with age, sex, and baseline markers of disease activity were controlled for. Results: A total of 1,216 patients from ERAS and 446 from ERAN had radiographic data. Compared to ERAS, ERAN patients had a lower mean total SHS score at baseline (ERAN 6.2 versus ERAS 10.5; P < 0.001) and mean annual rate of change (ERAN 2.5 per year versus ERAS 6.9 per year; P < 0.001). Seventy-four percent of ERAS and 27% of ERAN patients progressed ≥5 units. Lower scores at baseline in ERAN were largely driven by reductions in JSN (ERAS 3.9 versus ERAN 1.2; P < 0.001), along with erosions (ERAS 1.9 versus ERAN 0.8; P < 0.001). RF was associated with greater progression in each cohort, but the absolute difference in mean annual rate of change for RF-positive patients was substantially higher for ERAS (RF positive 8.6 versus RF negative 5.1; P < 0.001), relative to ERAN (RF positive 2.0 versus RF negative 1.9; P = 0.855). Conclusion: Radiographic progression was shown to be significantly reduced between the 2 cohorts, and was associated with lower baseline damage and other factors, including changes in early disease-modifying antirheumatic drug use. The impact of RF status as a prognostic marker of clinically meaningful change in radiographic progression has markedly diminished in the context of more modern treatment.