Anemia may provide clinically relevant information beyond conventional disease activity assessment to predict radiographic damage progression in Rheumatoid Arthritis

Burkhard Möller, Frauke Förger, Peter M. Villiger, Axel Finckh, on behalf of the Swiss Clinical Quality Management Program for Rheumatic Diseases.

Background/Purpose: Anemia in rheumatoid arthritis (RA) is prototypic for anemia of chronic disease (ACD) and often neglected in clinical practice. Previous studies have shown a relation between this extra-articular manifestation of RA and disease severity. Purpose: To study the relation between anemia and radiographic progression in RA.

Methods: This study is nested in a large prospective RA patient cohort. Data were collected between 1996 and 2007. Anemia was defined according to the WHO criteria (♀ Hb<12, ♂: Hb<13 g/dl). We also used one less (♀ Hb<12.2, ♂ 20-59y: Hb<13.7 g/dl, ♂ 60+y: Hb<13.2 g/dl) and one more severe definition of anemia (♀ Hb<11, ♂: Hb<12 g/dl) in sensitivity analyses. Erosions were assessed with a validated erosion score (Ratingen score) of 38 joints at hands and feet in 11255 radiograph sets from 4005 RA patients, over a mean follow-up of 2.2 years. Radiographic progression was analyzed using longitudinal regression models for repeated data and adjusted for potential confounding factors, such as age, gender, rheumatoid factor, disease duration and erosion score at inclusion, baseline and time dependent DAS28_ESR, DMARD, glucocorticoid and anti-TNF therapy. Subgroup analyses were performed in patients starting on methotrexate (n=728) or anti-TNF-therapy (n=938), in patients achieving different stages of severe, moderate, mild disease activity or remission (DAS28_ESR<2.6, n=1253) at the final visit, and in 508 patients achieving remission after TNF blockade.

Results: Erosions progressed significantly (p<0.001) faster in patients with than without anemia, despite adjusting for clinical disease activity and other important predictors of joint damage. As an example, erosion had progressed by mean 1.36% (95% CI 1.21-1.51) in non-anemic versus 2.66% (2.17-3.14) in anemic patients after one year. Patients with less severe anemia had a less rapid and patients with more severe anemia displayed a more rapid joint damage progression, suggesting a ‘dose response effect'. Effect modification by anemia went insignificant in subgroups of patients with persistent clinically detectable disease activity, but remained a significant factor for more rapid erosive progression in patients achieving remission upon any treatment (p<0.001) or upon anti-TNF-therapy (p<0.001).

Conclusion: Anemia appears to capture structural disease processes that remain unmeasured by conventional disease activity measures in patients with or without TNF blockade. Thus, anemia may help to identify patients with more rapid erosive disease.