Background Rheumatoid arthritis (RA) treatment has shifted from reducing symptoms to earlier inhibition of disease progression, thus preventing irreversible structural damage and disability. Detection of synovitis (SYN), osteitis (OST), and erosions (ERO), utilizing magnetic resonance imaging (MRI) may facilitate an earlier decision to progress from traditional disease modifying anti-rheumatic drugs (DMARDs) to the addition of biologic therapy. Arresting joint damage within 3 to 6 months of RA treatment determined by MRI, a more sensitive imaging technique compared to conventional radiography, should lead to low disease activity or remission, and subsequent better patient outcome with less disability. Abatacept (ABA) is a selective co-stimulation modulator that prevents T cell co-stimulation through CD28 pathway by binding avidly to CD80/86 and is clinically effective in reducing joint inflammation and structural damage.

Objectives Assess the impact of inhibition and progression of structural joint damage by ABA in patients (pts) with RA using MRI and x-rays. Evaluate the relationship between RA MRI scoring system (RAMRIS) mean changes for SYN, OST, ERO and clinical outcomes.

Methods 12 moderate to severe RA pts, methotrexate (MTX)-inadequate responders and biologic naive, were dosed to continue stable MTX combined with ABA 125 mg subcutaneous (sc) weekly. The dominant hand (metacarpophalangeal joints 1-5)/wrist were imaged with low field (0.2T), extremity MRI at baseline (BL), weeks (wks) 12, and 24 and x-rays at BL and wk 24. MR images were assessed according to the Outcome Measures in Rheumatology (OMERACT) RAMRIS and x-rays by Genant score. Change in scores from BL through wk 24 was compared to evaluate early effects of ABA on joint inflammation and damage.

Results BL mean RAMRIS scores showed low (OST) to moderate (SYN, ERO) disease severity in contrast to high disease activity by DAS28 (ESR). Despite the small number of patients included, decreased SYN was observed at wk 12 with continued significant suppression by wk 24. No significant changes from BL to wk 24 were seen in mean OST scores. Although reductions in ERO were observed at wk 12, a small non- statistically significant mean change (worsening) from BL to wk 24 was observed. BL SYN correlated strongly and significantly with swollen joint count (SJC) and physician global assessment (PGA). BL OST correlated with PGA, and ERO correlated with SJC and PGA.

Conclusions Low-field, extremity MRI of only 12 pts was able to demonstrate suppression of synovitis by ABA at wk 12 with further improvement through wk 24 and suppression of erosion progression through wk 24. Low BL levels of osteitis may explain the lack of significant change in mean RAMRIS osteitis.

References

1. American College of Rheumatology Rheumatoid Arthritis Clinical Trials Task Force Imaging Group and Outcome Measures in Rheumatology Magnetic Resonance Imaging Inflammatory Arthritis Working Group. Arthritis Rheum. 2013;65(10):2513-23

Disclosure of Interest O. Troum Grant/research support: Bristol-Myers Squibb, Consultant for: Bristol-Myers Squibb, O. Pimienta: None declared, L. Duan: None declared, J. Crues: None declared, C. Peterfy Consultant for: Bristol-Myers Squibb

DOI 10.1136/annrheumdis-2014-eular.5001