Background/Purpose: Abatacept (ABA), a selective T-cell co-stimulation modulator, showed promise for the treatment of PsA in a Phase II trial.¹ This prompted the conduct of the Phase III Active pSoriaTic aRthritis rAndomizEd triAl (ASTRAEA; NCT01860976); key results are presented here.

Methods: In this international, double-blind, multicenter study, patients (pts) with PsA were randomized (1:1) to SC ABA 125 mg wkly or placebo for 24 wks, and then treated with open-label SC ABA up to 24 mths. Pts had active disease (≥3 tender and ≥3 swollen joints), ≥2 cm target lesion of plaque psoriasis and inadequate response or intolerance to ≥1 non-biologic DMARD. Randomization was stratified by MTX use, prior TNFi use and skin involvement ≥3% of body surface area (BSA). Pts not achieving ≥20% improvement in SJC and TJC at Day 113 were switched to open-label ABA (early escape). Primary endpoint: ACR20 response at Wk 24. Key secondary endpoints at Wk 24: HAQ response (change from baseline ≥0.35); ACR20 response in the TNFi-naïve and -exposed subgroups; radiographic non-progression (PsA-modified total Sharp/van der Heijde score; change from baseline ≤0). Other secondary/exploratory endpoints included: ≥50% improvement in Psoriasis Area and Severity Index score (PASI50) in pts with ≥3% BSA; change in HAQ score; safety. Comparisons were performed using a 2-sided Cochran–Mantel–Haenszel chi-square test, adjusted for stratification criteria. Pts designated as early escape or with missing data were imputed as non-responders/radiographic progressors. Change in HAQ score was analyzed using a longitudinal repeated measures model (early escape pts set to missing at Days 141 and 169). Primary and key secondary endpoints were tested using a hierarchical approach.

Results: Of 424 pts enrolled, 213 received ABA and 211 placebo; 76 were early escape pts in ABA and 89 in placebo; 12 pts discontinued in ABA and 24 in placebo. Table 1 shows baseline characteristics. Most (>60%) pts had prior exposure to TNFis. ABA significantly improved the proportion of pts achieving an ACR20 response at Wk 24 (p<0.001) (Table 2). The proportion of HAQ responses was numerically higher with ABA vs placebo (p=0.097). Higher proportions of pts receiving ABA vs placebo had an ACR20 response in the TNFi-naïve and -exposed subgroups, and radiographic non-progression (nominal p<0.05), with modest numerical improvement in PASI50. Efficacy was maintained at 1 year. The safety profile of ABA was similar to placebo, with no new safety signals.

Conclusion: Abatacept improved disease and was well tolerated in pts with active PsA, regardless of prior exposure to TNFis.   1.    Mease P, et al. Arthritis Rheum 2011;63:939–48.