P344 Real-world effectiveness of tofacitinib in ulcerative colitis: a multi-centre study

Abstract

We aimed to describe the real-world effectiveness of tofacitinib in ulcerative colitis (UC).We analysed a retrospective, multi-centre cohort from six centres in the USA. UC patients started on tofacitinib (10 mg BID) for active disease were included. Primary outcome was clinical response (>50% reduction in symptoms) at Week 8 as determined by physician global assessment. Secondary outcomes included clinical remission (no symptoms) at Week 8, clinical response/remission at Week 16 and endoscopic healing (defined as Mayo endoscopic score ≤1 or absence of erosions/ulcerations) within 6 months of initiating tofacitinib. Descriptive statistics and Fisher exact tests were performed. Logistic regression assessed predictors of Week 8 response. A multi-variable model was created using backward elimination.A total of 123 UC patients were included with a median age of 38 years (IQR 27–46) and 5 years disease duration (IQR 2–9). 56.1% were men and 60.2% had pancolitis. 28.5% were bio-naïve while 40.7% had been exposed to both anti-tumour necrosis factor (anti-TNF) biologics and vedolizumab (VDZ). Ninety-six patients completed 8 weeks of tofacitinib. 60.8% had clinical response and 13.5% clinical remission at Week 8. At Week 16 (total n = 74), 55.4% had clinical response and 48.6% clinical remission. 64.9% (total n = 57) had endoscopic healing. A larger proportion of bio-naïve patients achieved clinical response with no difference between those exposed to both anti-TNF and VDZ or either alone (Table 1). Patients with prior exposure to 2 biologic classes (anti-TNF and VDZ) had lower rates of endoscopic healing compared with bio-naïve and 1 biologic class exposure (Table 1). Bio-naïve status and higher albumin were associated with greater chance of Week 8 response while pancolitis, baseline endoscopic Mayo score 3, concomitant steroids at start of tofacitinib, and male gender were associated with lower chance of response (Table 2). In multi-variable analysis, bio-naïve status (aOR 5.50, 95% CI 1.71–17.65), concomitant steroids (aOR 0.25, 95% CI 0.07–0.83), and male gender (aHR 0.25, 95% CI 0.08–0.83) were associated with Week 8 response.Tofacitinib is effective at inducing clinical response in a real-world clinical setting. Prior exposure to biologics is associated with reduced chance of clinical response and endoscopic healing.Table 1. Tofacitinib response rates by prior biologic exposure. (*p = 0.007, **p < 0.001, *** p < 0.001. Patients who discontinued tofacitinib before Week 8 or 16 were considered non-responders.)Prior biologic exposure statusClinical response Week 8*Clinical response Week 16**Endoscopic healing by 6 months***Bio-Naïve81.8% (total n = 33)81.3% (total n = 32)87.1% (total n = 31)Prior exposure to 1 class (Anti-TNF or VDZ)44% (total n = 25)36.4% (total n = 22)57.1% (total n = 14)Prior exposure to 2 classes (Anti-TNF and VDZ)55.3% (total n = 38)35% (total n = 20)16.7% (total n = 12)Table 2. Baseline variables significantly associated with tofacitinib clinical response at Week 8 OR: odds ratio; CI: confidence interval.Univariable logistic regression variable OR (95% CI)p-value Bio-Naïve4.50 (1.64–12.37)0.004Pancolitis (ref = limited colitis)0.34 (0.14–0.86)0.02Albumin (g/dl)2.63 (1.02–6.80)0.046Mayo endoscopic score 3 (ref=score 2)0.27 (0.10–0.72)0.01Male (ref=female)0.28 (0.11–0.70)0.007Concurrent steroids at start of tofacitinib0.22 (0.08–0.58)0.002