Background/Purpose:
In patients (pts) with active non-radiographic axial spondyloarthritis (nr-axSpA) in the ABILITY-3 study who achieved sustained remission with adalimumab (ADA), continued ADA therapy was associated with significantly fewer pts flaring than treatment withdrawal. Sustained remission is an important treatment goal in pts with nr-axSpA, but factors predicting sustained remission and absence of flare are unknown. The purpose of this analysis was to determine whether an ADA concentration threshold is predictive of achievement of sustained remission and absence of flare in pts with nr-axSpA.

Methods:
ABILITY-3 included a 28-wk open-label (OL) ADA (40 mg every other wk) lead-in in which pts who achieved sustained remission (ASDAS <1.3 at wks 16, 20, 24, and 28) were randomized to double-blind (DB) placebo (PBO) or continued ADA for 40 wks (68 wks total). Pts not achieving remission were discontinued at wks 20, 24, or 28. The primary efficacy variable was the proportion of pts who did not experience a flare (ASDAS ≥2.1 at 2 consecutive visits) by wk 68. Pts who flared received OL ADA rescue. ADA trough concentrations at wks 12 or 28 were used to predict achieving sustained remission by wk 28 (all OL patients, N=673) and absence of flare at wk 68 (patients randomized to DB ADA, n=152). Threshold analysis was conducted by constructing receiver operating characteristic (ROC) curves (R Ver 3.3.3) to assess predictive ADA trough concentration thresholds based on specificity and sensitivity values.

Results:
Of 673 pts enrolled in the OL phase, 305 met criteria for remission at wk 28 and were randomized to DB treatment. Mean ± SD ADA trough concentrations were 6.68±5.23 μg/mL at wk 12 and 8.36±5.27 μg/mL at wk 28. During the DB phase, 81 and 45 pts flared and 68 and 36 pts received ≥12 wks of OL ADA rescue in PBO (n=153) and ADA (n=152) arms, respectively. ADA mean ± SD trough concentration at wk 68 for pts randomized to DB ADA with flare (7.64±5.22 μg/mL) was slightly lower than for DB ADA pts without flare (8.12±4.35 μg/mL). ROC curves showed AUC values ≤0.6 for achieving sustained remission by wk 28 and absence of flare at wk 68, with no concentration threshold meeting sensitivity and specificity criteria for reliable prediction of such endpoints (Figure).

Conclusion:
ROC analyses did not identify an ADA trough concentration threshold that reliably predicted whether a pt with nr-axSpA would achieve sustained remission (by wk 28) or absence of flare (at wk 68).

Figure.  Receiver Operating Characteristic Curves to Predict Sustained Remission by Week 28 and Absence of Flare at Week 68 by Adalimumab Serum Concentration