Tabalumab | p Value | |||||||
---|---|---|---|---|---|---|---|---|
Placebo (N=36) | 1 mg (N=30) | 3 mg (N=20) | 10 mg (N=15) | 30 mg (N=18) | 60 mg (N=13) | 120 mg (N=24) | ||
ACR50 obs, n (%) (p)*† | 8 (22.2) | 3 (10.0) (0.896) | 2 (10.0) (0.940) | 5 (33.3) (0.311) | 2 (11.1) (0.919) | 1 (7.7) (0.954) | 9 (37.5) (0.255) | NA |
ACR50 model, % (p) *† | 18.1 | 17.7 (0.827) | 17.0 (0.828) | 15.0 (0.833) | 11.8 (0.834) | 11.8 (0.767) | 37.0 (0.037) | (0.059)‡ |
ACR20 obs, n (%) (p) *† | 16 (44.4) | 12 (40.0) (0.730) | 9 (45.0) (0.594) | 7 (46.7) (0.563) | 11 (61.1) (0.193) | 7 (53.8) (0.397) | 17 (70.8) (0.039) | NA |
ACR20 model, % (p) *† | 43.2 | 43.6 (0.171) | 44.3 (0.170) | 46.9 (0.162) | 53.5 (0.132) | 61.5 (0.070) | 70.1 (0.005) | (0.044)‡ |
DAS28, mean±SD change from baseline (p)§¶ | −1.5±1.3 | −1.5±1.3 (0.457) | −1.0±1.1 (0.874) | −1.7±1.0 (0.278) | −1.5±1.3 (0.357) | −1.6±1.2 (0.271) | −1.9±1.2 (0.048) | (0.071)** |
*1-sided Fisher exact test that the tabalumab group has more responders than placebo.
†Imputed by non-responder imputation.
‡Likelihood ratio test of quadratic logistic regression model, testing the existence of a dose-response with a prespecified 2-sided type 1 error rate of 0.10.
§1-sided pairwise comparison using contrast statements with an analysis of covariance (ANCOVA) model with treatment as the fixed factor and baseline as the covariate.
¶Imputed by last-observation-carried-forward.
**Dose-response relationship from 1-sided Spearman non-parametric correlation analysis.
ACR50, American College of Rheumatology 50 responder index; ACR20, American College of Rheumatology 20 responder index; DAS28, Disease Activity Score based on 28-joint count; obs, actual observations; model, assessed using a statistical model; NA, not available; p, p Value.