Background/Purpose: Most biologic trials in juvenile idiopathic arthritis (JIA) treat all participants with the biologic under study for 12 to 16 weeks before randomizing responders into a withdrawal phase. This implies that patients are expected to respond quickly to biologics. If treatment response can be predicted from early response patterns (< 12-16 weeks), this information may be applied to identify likely responders from likely non-responders early during clinical care, allowing earlier termination of ineffective treatment. We aimed to delineate the early treatment response trajectory of JIA patients to biologics.

Methods: Longitudinal data from early treatment phase (12-16 weeks) from 3 JIA trials: 1) Tocilizumab (Roche), 2) Etanercept (Amgen) and 3) the Trial of Early Aggressive Therapy TREAT (Etanercept), were obtained and combined. Primary outcome is the ACR response states: ACR50, 70, 90. Clinically significant minimal improvement was defined as attaining at least an ACR50 state. Secondary outcome was the clinical juvenile arthritis disease activity score (cJADAS): inactive, low, moderate and high disease states. Longitudinal responses (ACR states and cJADAS) were modelled using the Markov multistate model, corrected for interval censoring and allowing heterogeneous transition rates.

Results: 342 patients (188 Tocilizumab, 69 Etanercept, 85 TREAT) with polyarticular course JIA were studied. 27% (92/341) were males, median age at baseline (25 th -75 th percentile P) was 10.0 (5.7- 15.0) years and median (25 th – 75 th P) duration of disease was 3.6 (1.0- 8.7) years. Baseline median (25 th -75 th P) active joint counts was 15 (7-24). Pateints were on methotrexate (46%) and corticosteroids (34%) at baseline. 87% patients were on biologics (Tocilizumab 188, Etanercept 111). At week 4, the probabilities of transitioning from active disease to ACR50, 70 and 90, was 0.29, 0.13, 0.03. Probabilities of transition continued to increase from week 4 to 16, without plateau. At week 16, the probabilities of transition to ACR50, 70, 90 were: 0.19, 0.32, 0.25. By week 16, the mean durations of time spent in ACR50, 70, 90 were 0.91, 0.76 and 0.43 months. Even if no clinically significant minimal improvement was achieved by week 4, the probabilities of attaining this was substantial by weeks 12 (0.55) and 16 (0.67).  At week 4, the probabilities of transition from a high cJADAS state to a moderate state was 0.20, to low state was 0.01, and to inactive disease < 0.01. At week 16, the probabilities for a moderate cJADAS state was 0.36, for low state was 0.10 and for inactive disease was 0.11. By week 16, the mean durations of time spent in low and inactive disease states were 0.15 and 2.37 months.

Conclusion: Patients treated with biologics continued to improve over the first 12-16 weeks, without plateau. Lack of early (4 weeks) clinically significant minimal response did not preclude attaining such response by 16 weeks. Our work suggests that response to biologics (Tocilizumab, Etanercept) continues to improve over the first 16 weeks of treatment. Therefore, patients should not be switched out of biologics before 16 weeks for a lack of effectiveness.

« Back to 2019 ACR/ARP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/delineating-early-response-trajectories-to-biologics-in-polyarticular-course-juvenile-idiopathic-arthritis/