Exposure-Response Analysis of Tofacitinib in Active Psoriatic Arthritis: Results from Two Phase 3 Studies

Abstract

Tofacitinib is an oral Janus kinase inhibitor for the treatment of psoriatic arthritis (PsA). These post hoc exposure-response (E-R) analyses of pooled data from two Phase 3 studies (NCT01877668 and NCT01882439) characterized the relationships between tofacitinib exposure and efficacy (American College of Rheumatology [ACR] criteria), and changes in hemoglobin (Hgb) in patients with PsA. Efficacy data for the proportion of patients receiving tofacitinib 5 or 10 mg twice daily, or placebo, achieving ACR ≥20%, ≥50%, or ≥70% response criteria (ACR20, ACR50, and ACR70, respectively) at Month 3, were modeled jointly using a four-category ordered categorical exposure-response model (ACR20 non-responder, ACR20 responder but not ACR50 responder, ACR50 responder but not ACR70 responder, and ACR70 responder). A maximum drug effect (E_max) model (using average concentrations of tofacitinib at steady state [C_avg]) adequately described the exposure-ACR response rate relationship. Model-predicted response rates for tofacitinib 5 and 10 mg twice daily were 51% and 58%, respectively, for ACR20; 29% and 36% for ACR50; and 15% and 20% for ACR70. The E-R relationship between tofacitinib exposure and changes in Hgb was assessed using an indirect response model, which generally predicted Hgb concentration-time profiles across treatments well. The proportions of patients experiencing a decrease in Hgb of >2 g/dL were similar with tofacitinib 5 mg twice daily or placebo. These results were generally consistent with previous analyses in rheumatoid arthritis and psoriasis, and support the use of tofacitinib 5 mg twice daily for active PsA.

Keywords: efficacy; exposure–response; hemoglobin; psoriatic arthritis; tofacitinib.

Figure 1

Observed and predicted ACR20 (a), ACR50 (b), and ACR70 (c) responder rate with 90% CIs of the prediction (predictive check)—pooled data from OPAL Broaden and OPAL Beyond. ACR20/50/70, proportion of patients achieving American College of Rheumatology ≥20%, ≥50%, or ≥70% response criteria improvement from baseline; BID, twice daily; C_avg, average concentration at steady state; CI, confidence interval; Obs, observed responder values (0 or 1, shown as vertical tick marks along the x‐axis). Observed ACR20, ACR50, and ACR70 responder rate are represented as tick marks along the x‐axis at either 0 (non‐responder) or 1 (responder); diagonal numbers and intervals in black represent the C_avg tertile intervals used for binning the observed responder rate (green circles) with 90% CIs (error bars); green circles are plotted at the midpoint of each bin. Solid line and grey shaded area (90% CIs) represent model‐predicted responder rate. Vertical dotted blue lines represent dose‐standardized median of the exposure values scaled to tofacitinib 5 or 10 mg BID.

Figure 2

Observed and predicted incidence of hemoglobin decrease of >2 g/dL versus C_avg (a), and C_avg distribution for tofacitinib 5 and 10 mg BID (b). BID, twice daily; C_avg, average concentration at steady state; CI, confidence interval. Panel a: The green circles (error bars) represent observed incidence (90% CI) at each bin of C_avg, and the blue line (grey shaded area) represents the incidence (90% CI) predicted from a logistic regression model. The dashed lines represent 25th, 50th, and 75th percentiles for C_avg. Panel b: The bold lines in each box indicate the median C_avg (approximately 17.6 and 36.1 ng/mL for tofacitinib 5 and 10 mg BID, respectively). The lower and upper hinges of each box correspond to the first and third quartiles (25th and 75th percentiles), respectively. The left/right whisker extends from the hinge to the lowest/highest value no further than 1.5 times of the inter‐quartile range. The black circles represent data points beyond the end of the whiskers.