BIOMARKERS AND THEIR ASSOCIATION TO INFLAMMATION AND JOINT DESTRUCTION MEASURED BY MAGNETIC RESONANCE IMAGING IN EARLY RHEUMATOID ARTHRITIS PATIENTS

Background: Strategies for treating patients with rheumatoid arthritis (RA) aim to both suppress inflammation and to prevent future joint destruction. As current predictors of joint destruction have low specificity and ongoing inflammation are commonly seen despite low CRP and DAS-scores, serological biomarkers reflecting both inflammation and joint destruction may potentially improve disease monitoring. Objectives: The aim of this study was to examine whether a panel of biomarkers in serum and urine were longitudinally associated to inflammation measured by MRI, and to assess whether the biomarker levels at baseline were predictive of progression in erosions measured by MRI and conventional radiographs (CR). Methods: A cohort of 84 early RA patients (disease duration < 1 yr, mean age 58.4 yrs, 73.9% females, 55% ACPA pos) was followed longitudinally with assessment at baseline, 3, 6 and 12 months including clinical examination, CR of hands and MRI scans of the dominant wrist (1.5 T scanner). MRI images were evaluated according to the RAMRIS score (erosions range 0-150, synovitis range 0-9 and bone marrow oedema range 0-45). The CRs were scored according the van der Heijde modified Sharp score (vdHSS) with 1U as the cut-off for progression. With the exception of CRP (Nefelometric, Dade Behring), all measurements were performed by ELISA: sOPG and sRANKL (Biomedica), sCTX I and uCTX II (Nordic Bioscience), sYKL40 (Quidel), sCOMP (Anamar) and sMMP 3 (Bender MedSystems). Longitudinal associations between the biomarker levels and MRI inflammation were explored by multiple linear mixed model analyses for repeated measures, and multiple linear and logistic regression analyses were use to identify predictors of progression of damage assessed by MRI and CR, respectively. Results: Levels of sYKL-40 and sMMP 3 were longitudinally associated (p<0.05) to MRI measures of inflammation (RAMRIS synovitis score and RAMRIS bone marrow oedema score) and also to DAS28 in linear mixed model analyses corrected for age, gender, CRP and treatment. sCTX I levels were predictive of progression in RAMRIS erosion score from baseline to 12 months follow-up (beta 2.42 (95%CI 0.48-4.36)). The model was corrected for age, gender and known predictors of erosive progression. As the RAMRIS do not include evaluation of joint space narrowing, predictors of progression measured by CR were also examined. The levels of CTX I (0.38 ng/ml (0.32-0.46) vs 0.29 (0.23-0.46)) and CTX II (233 ng/mmol (172-387) vs 154 (94-261) were sign. higher in progressors than in non-progressors and both biomarkers predicted radiographic progression assessed by the total vdHSS (p<0.05) in multivariate logistic regression analyses corrected for other known predictors of radiographic progression. Conclusion: Measurements of MMP 3 and YKL 40 were consistently associated with both MRI - and clinical scores of inflammation during the study. This association was independent of CRP, indicating an additional value to this well-established marker of inflammation. The bone marker sCTX I was a predictor of erosive progression measured by MRI and CR, and the cartilage marker uCTXII predicted progression measured by CR. This study indicates that biomarkers might contribute to the monitoring of disease activity and joint destruction in the longitudinal assessment of patients with RA. Disclosure of Interest: None declaredCitation: Annals of the Rheumatic Diseases, volume 68, supplement 3, year 2009, page 550Session: Rheumatoid arthritis Prognosis, predictorsand outcome (Poster Presentations )