P413 Withdrawal versus continuation of thiopurine in vedolizumab-treated patients with Ulcerative Colitis (VIEWS): a multi-centre randomised controlled trial

Abstract

Abstract Background The benefit of continuing thiopurine in ulcerative colitis (UC) responders to vedolizumab (VED) is unclear. We aimed to determine the effect of thiopurine withdrawal in VED-treated UC patients on combination therapy. Methods This prospective multi-centre, single-blind, randomised controlled trial recruited UC patients on VED 300mg IV every 8 weeks and a thiopurine for ≥6 months. Patients in steroid-free clinical remission (partial Mayo score[pMS]≤2, no subscore>1) and endoscopic remission/improvement (Mayo endoscopic subscore[MES]≤1) were randomised 2:1 to withdraw or continue thiopurine. Primary outcome was comparing week 48 VED trough levels. Secondary outcomes were disease relapse (pMS≥3 and faecal calprotectin>150μg/g or increase in MES≥1 from baseline), centrally-read endoscopic remission (MES=0), histologic remission (Nancy index=0), histo-endoscopic remission, faecal calprotectin (remission<150μg/g) and adverse events. Results Sixty-two consecutive patients (58% males, median follow-up 23.0 months) were randomised to withdraw (n=42) or continue (n=20) thiopurine. Randomisation was balanced. Complete data was available for 58 patients. There was a non-significant trend toward increased disease relapse in the withdrawal group (P=0.12, Fig. 1). In the withdrawal group, baseline histologic activity significantly increased disease relapse (HR:5.8, 95%CI:1.6-20.8, P=0.007, Fig. 2). There was a trend towards higher relapse rates in the withdrawal versus continue groups for bio-exposed patients (50.0% [6/12] vs 0.0% [0/4], P=0.07) but not bio-naïve patients (20.0% [6/30] vs 12.5% [2/16], P=0.52). At week 48, the continue group had significantly higher histologic remission (OR:4.6, 95%CI:1.1-18.9, P=0.03) and histo-endoscopic remission rates (OR:4.6, 95%CI:1.3-16.1, P=0.01, Fig. 3) versus the withdrawal group, but similar endoscopic remission rates (P=0.09, Fig. 3). Faecal calprotectin remission was significantly higher in the continue (94% [16/17]) versus withdrawal group (70% [28/40], P=0.047) with mean calprotectin 44.5μg/g±SD56.2 vs 209.6μg/g±SD344.2 (P=0.003) respectively. On multivariate analysis, histologic activity at baseline (HR:5.3, 95%CI:1.1-26.2,P=0.04) predicted disease relapse. Week 48 median VED levels were 15.9µg/mL (IQR:10.1-22.7) in the withdrawal group versus 14.7µg/mL (IQR:12.1-18.7) in the continue group (P=0.43). No patients had anti-VED antibodies. There was no significant difference in adverse events between groups. Conclusion Although thiopurine withdrawal did not affect VED trough level, it increased calprotectin, histologic and histo-endoscopic activity in UC. In patients with histological activity despite deep remission, thiopurine withdrawal significantly increased the risk of disease relapse.