SAT0011   INHIBITION OF RADIOGRAPHIC PROGRESSION WITH RITUXIMAB IS NOT DEPENDENT ON CLINICAL EFFICACY: RESULTS FROM A STUDY IN RHEUMATOID ARTHRITIS PATIENTS WITH AN INADEQUATE RESPONSE TO ONE OR MORE TNF INHIBITORS (REFLEX)

E. Keystone1, P. Emery2, C. G. Peterfy3, P. P. Tak4, S. B. Cohen5, M. C. Genovese6, S. Williams7, D. Hagerty8, M. W. Cravets8, T. M. Shaw7
1Rheumatology, University of Toronto, Toronto, Canada, 2Rheumatology, Leeds General Infirmary, Leeds, United Kingdom, 3Synarc Inc., San Francisco, CA, United States, 4Division of Clinical Immunology and Rheumatology FOCIS Center of Excellence, Academic Medical Center/University of Amsterdam, Amsterdam, Netherlands, 5Rheumatology and Musculoskeletal, Radiant Research, Dallas, TX, 6Department of Immunology and Rheumatology, Stanford University, Palo Alto, CA, United States, 7Medical Sciences, Roche Products Ltd, Welwyn Garden City, United Kingdom, 8Biogen Idec, San Diego, CA, United States

Objectives: To investigate the correlation of clinical efficacy at Week 24 with inhibition of structural joint damage by rituximab (RTX) plus methotrexate (MTX) compared with MTX alone at 56 weeks in rheumatoid arthritis (RA) patients (pts) with an inadequate response (IR) to ≥1 tumour necrosis factor (TNF) inhibitors.
Methods: Pts receiving MTX and with an IR to ≥1 TNF inhibitors were randomised to RTX (two infusions of 1 g, 2 weeks apart) or placebo as previously described (1). From Week 24, eligible pts in both placebo and RTX arms could receive a repeat course of RTX. Radiographs of hands and feet were obtained at baseline and at Weeks 24 and 56. Pts who withdrew for any reason were included within their randomised treatment arm, irrespective of any subsequent RA therapy. Radiographs were assessed using the Genant–Sharp method, and were read blinded to sequence and treatment by an independent central reading site. Analysis at Week 56 using linear extrapolation for missing radiographic data is presented. Subgroup analysis of pts who were ACR20 non-responders at Week 24 was also conducted. Observed data were used to define ACR20 response.
Results: Pts had long-standing active disease that had been previously treated with multiple disease-modifying anti-rheumatic drugs (DMARDs) as well as ≥1 TNF inhibitors. At Week 56, RTX pts exhibited significant differences in radiographic endpoints compared with placebo pts (2). Analysis of pts who were ACR20 non-responders at Week 24 also showed that RTX-treated pts had significantly less total progression compared to non-responding placebo pts (p=0.004) (Table). Pts who were non-responders by EULAR criteria and who received RTX tended to exhibit less progression relative to the corresponding placebo pts (p=0.07). When pts were segmented according to the number of swollen joints at 24 weeks, pts receiving RTX had less progression relative to placebo pts with the same number of swollen joints. For pts with >13 swollen joints, change in mean total Genant–Sharp scores was 3.01 for pts in the placebo arm compared with 1.24 for pts in the RTX arm.
Table:
 
Placebo + MTX (56 weeks)RTX + MTX (56 weeks)
ACR20 non-responders
(at Week 24)		n=104n=87
Mean change in total
Genant–Sharp score (SD)		2.82 (6.36)0.66 (2.05) p=0.004
Mean change in erosion score (SD)1.72 (3.75)0.33 (1.14) p=0.0056
Mean change in joint space narrowing (SD)1.10 (3.11)0.33 (1.25) p=0.0076
EULAR non-responders
(at Week 24)		n=134n=77
Mean change in total
Genant–Sharp score (SD)		2.67 (5.83)1.19 (3.11) p=0.07
SD = standard deviation

Conclusion: These data represent the first evidence that the inhibition of structural joint damage exhibited by RTX, a selective CD20+ B cell targeted therapy, is independent of achieving a clinical response with RTX. They are consistent with the concept of an uncoupling of clinical efficacy and structural damage inhibition with RTX.
References: 1. Cohen, et al. Arthritis Rheum 2006;54:2793–2806.
2. Keystone, et al. EULAR 2006 (Abstract No. OP0016).