Background Anemia in rheumatoid arthritis is often neglected, but a significant relation between this extra-articular manifestation and radiographic progression of erosions was recently shown.

Objectives To study the relation between time in anemia and radiographic progression in addition to time integrating variables for different levels of clinically assessed disease activity, in patients with and without anti-TNF therapy.

Methods This study is nested in a prospective cohort of 4377 RA patients. Data were collected between 1996 and 2007. Anemia was defined according to WHO (♀ Hb<12, ♂: Hb<13 g/dl) or to less stringent criteria (♀ Hb<12.2 g/dl, ♂ Hb<13.7 g/dl or 13.2 g/dl, dependent on age 20-59 or 60+ years [1]. Erosions were assessed with the validated Ratingen erosion score of 38 joints at hands and feet. Levels of clinical disease activity were defined by dichotomizing the population at different thresholds of high (DAS28_ESR>5.1), moderate (DAS28_ESR>3.2) or low disease activity (DAS28_ESR>=2.6). In the absence of CRP data, we also used a modified remission criterion (SJC<=1 & TJC<=1 & patient global <=1), which is reported to perform similar well to predict radiographic and functional patient outcome as the official ACR/EULAR 2010 remission criteria [2]. Damage progression was analyzed in mixed effects models for longitudinal data, including time-integrated variables for clinical disease activity on the different indicated levels, anemia and TNF blockade, and adjustments for the potential confounding factors age, gender, rheumatoid factor, disease duration and erosion scores at inclusion, DAS28_ESR at baseline, NSAID, glucocorticoid, MTX and other DMARD therapy.

Results Progression of erosions was studied in 2681 patients with at least two (mean 3.7) radiographic assessments over mean 2.8 years. Regardless of the applied threshold for clinically active disease in the different models, mean annual erosion progression rates were always significantly higher in patients with anemia. Anemic patients according to the WHO definition experienced by mean 0.65-0.76% of the total score (p<0.001) faster, and patients fulfilling the alternative anemia definition by 0.28-0.38% (p<0.05) faster erosion progression than the non-anemic comparators. Mean joint damage progression was significantly retarded by anti-TNF therapy, irrespective of the different applied cut-offs for anemia or clinical disease activity (Δ mean -0.41% to -0.49%, p<0.001). Effect-sizes for annual damage progression in WHO-defined anemia were at least comparable to the protective effects of anti-TNF therapy, clinical disease activity thresholds of ACR/EULAR remission (Δmean 0.19%), DAS28_ESR remission (Δmean 0.33%), or low disease activity (Δmean 0.57%), but lower than the effect-size of high disease activity (Δ mean 0.77%).

Conclusions Progression of erosions was accelerated in anemic patients and varied with anemia severity. The anemia effect was complementary to the effects observed in relation to clinical disease activity and in opposite direction to anti-TNF therapy. Thus, anemia may capture unmeasured inflammatory processes at any level of clinical disease activity in patients with and without TNF blockade.

References

1. Beutler E, Waalen J. Blood 2006;107:1747-50.

2. Felson DT, Smolen JS, Wells G, et al. ACR/EULAR provisional definition of remission. Ann Rheum Dis 2011;70:404-13.

Disclosure of Interest None Declared