Op0261 treat-to-target dose reduction and withdrawal strategy of tnf inhibitors in psoriatic arthritis and axial spondyloarthritis: a randomized controlled non-inferiority trial

BackgroundTumour Necrosis Factor inhibitors (TNFi) are effective in Psoriatic Arthritis (PsA) and axial SpondyloArthritis (axSpA), but are associated with a somewhat increased infection risk, patient burden, and high costs. Treat-to-target (T2T) tapering might ameliorate these drawbacks, but no high quality T2T tapering studies have been done in PsA and axSpA.ObjectivesTo investigate whether a T2T strategy with tapering is non-inferior to a T2T strategy without tapering.MethodsWe performed a pragmatic open-label, monocenter, randomized, controlled non-inferiority (NI) trial on T2T tapering of TNFi. PsA and axSpA patients using a TNFi with ≥6 months stable low disease activity (LDA) were randomized to a T2T tapering or no-tapering strategy, in a ratio of 2:1 and followed-up for 12 months. LDA was defined as a Psoriatic Arthritis Disease Activity Score (PASDAS) ≤3.2 for PsA and/or Ankylosing Spondylitis Disease Activity Score (ASDAS) <2.1 and/or judgement of physician and patient. Tapering consisted of 3-monthly tapering steps (66%, 50%, 0%), with re-intensification in case of flare. Primary endpoint was the difference in proportion of patients having LDA at 12 months, compared to a prespecified NI margin of 20%. The primary Bayesian analysis was adjusted for stratification factors (diagnosis and csDMARD use). Secondary endpoints included the mean percentage Daily Defined Dose (%DDD) at month 12 and mean 3-monthly disease activity and %DDD. Other endpoints included the proportion of patients discontinuing their TNFi, the cumulative incidence of flares (change from baseline PASDAS≥0.8 or ASDAS≥0.9), the use of concomitant medication and (serious) adverse events.Results122 patients were included (N=81 tapering (PsA, N=42, axSpA, N=39); N=41 no-tapering (PsA, N=22, axSpA, N=19)) (Table 1). Proportion of patients in LDA at 12 months for the tapering and no-tapering group was 69% and 73%: adjusted difference 5% (Bayesian 95% credible interval: -10% to 19%), confirming NI. The mean percentage %DDD was respectively 53% and 91% at month 12 (Figure 1). At 12 months, 58 (72%) patients of the tapering group were successfully tapered, of whom 23 (28%) discontinued their TNFi. The cumulative incidence of flares was 85% in the tapering and 78% in the no-tapering group (p=0.32). Start or escalation of concomitant medication was more frequent in the tapering group, significantly so for NSAID use: csDMARDs (only for PsA): 1 (2%) vs. 1 (5%) (p=0.64); NSAIDs: 44 (54%) vs. 10 (24%) (p=0.002); glucocorticoids: 24 (30%) vs. 7 (17%) (p=0.13). For serious adverse events, similar results were seen. The risks of grade 3/4 infections and injection site reactions were respectively 46% and 23% lower in the tapering group than the no-tapering group.Table 1.Baseline characteristics of T2T strategy treated PsA and axSpA patients with or without tapering.CharacteristicT2T with tapering (N=81)T2T without tapering (N=41)Diagnosis, n (%)-Psoriatic arthritis42(52%)22(54%)-Axial spondyloarthritis39(48%)19(46%)Female, n (%)28(35%)20(49%)Age in years at inclusion, mean (SD)50(14)52(15)Disease duration at inclusion, years, median (IQR)11(5-21)12(5-21)CASPAR Criteria, n (%)34(81%)17(77%)ASAS Criteria, n (%)35(90%)17(89%)Disease activity, mean (SD)-PASDAS - (64/64 PsA)-ASDAS - (57/58 axSpA)1.60(1.26) 1.34(0.87)1.63(0.98) 1.21(0.61)Duration of current bDMARD use, years, median (IQR)2(1-6)2(2-7)Current bDMARD use, n (%)-Adalimumab62(77%)28(68%)-Etanercept10(12%)6(15%)-Certolizumab Pegol2(2%)1(2%)-Golimumab2(2%)1(2%)-Infliximab5(6%)5(12%)Figure 1.Mean disease activity and %DDD of T2T strategy treated PsA (A. and C.) and axSpA (B. and D.) patients with or without tapering at baseline, 3, 6, 9 and 12.ConclusionA T2T TNFi tapering strategy is non-inferior to a T2T strategy without tapering with regard to the proportion of patients still in LDA at 12 months and results in a substantial reduction of TNFi use, albeit with slightly more use of other medication.AcknowledgementsWe thank all the patients who were willing to participate in this study and the rheumatologists in the Sint Maartenskliniek Nijmegen and Woerden for participation in patient recruitment and data collection; S.R. van de Plassche, A.H. Verkerk and M. den Broeder for data collection and entering; M. Roelofs, I. Cillessen, C. Kleinveld, I. van Neste for aiding with study coordination; D. van Aggelen, D. Rotteveel, L. Schiersbergen for aiding with laboratory implementation and procedures; B.J.F. van de Bemt and M. Flendrie for being part of the data safety monitoring board.Disclosure of InterestsCelia Michielsens: None declared, Nathan den Broeder: None declared, Frank van den Hoogen: None declared, Elien Mahler: None declared, Steven Teerenstra: None declared, Désirée van der Heijde Consultant of: AbbVie, Bayer, BMS, Cyxone, Eisai, Galapagos, Gilead, Glaxo-Smith-Kline, Janssen, Lilly, Novartis, Pfizer, UCB Pharma, L.M. Verhoef: None declared, Alfons den Broeder Grant/research support from: Abbvie, Galapagos, Pfizer, Novartis, Lilly, Sanofi, Gilead