A multicenter, double-blind, dose-ranging, randomized, placebo-controlled study of recombinant human interleukin-1 receptor antagonist in patients with rheumatoid arthritis - Radiologic progression and correlation of Genant and Larsen scores

Objective. To evaluate radiographic progression and the relationship of rad iologic scores obtained by the Genant and Larsen methods in a clinical tria l of recombinant human interleukin-1 receptor antagonist (IL-1Ra), Methods. Patients with rheumatoid arthritis (RA) were randomized into 4 gro ups: placebo (n = 121) or IL-1Ra at a daily dosage of 30 mg (n = 119), 75 m g (n = 116), or 150 mg (n = 116), Hand radiographs obtained at baseline, 24 weeks, and 48 weeks were scored using both methods. Results. At 24 weeks, by the Genant method, there was significant reduction in the score for progression of joint space narrowing (JSN) and the total score (a combination of erosion and JSN) in all treatment groups, Least-squ ares mean changes in the Genant erosion score from baseline to 24 weeks wer e significantly reduced after treatment with IL-1Ra at 30 mg/day and for al l IL-1Ra treatment groups combined. The changes corresponded to a reduction of 38% in erosion, 58% in JSN, and 47% in total score, Patients treated wi th IL-1Ra at 75 mg/day had a significant reduction in the Larsen erosive jo int count (LEJC), and all IL-1RA-treated groups combined showed a 45% reduc tion. Correlations (r) between the Genant total and Larsen scores were 0.84 at baseline, 0.83 at week 24, and 0.83 at week 48 (P < 0.0001); correlatio ns between the Genant erosion score and the LEJC were 0.83 (P < 0.0001) at all visits; correlations between the Genant total and the Larsen scores wer e 0.32 and 0.49 (P < 0.0001) for progression from baseline to week 24 and f rom baseline to week 48, respectively; correlations between the Genant eros ion score and the LEJC were 0.36 and 0.41 (P < 0.0001) for progression to w eeks 24 and 48, respectively. Conclusion. IL-1Ra reduced radiologic progression of RA, Scores by the 2 me thods correlated strongly for each individual time point, but much less str ongly for assessments of disease progression.