Evaluation of the Efficacy of Tofacitinib as Maintenance Therapy in Patients With Ulcerative Colitis Stratified by OCTAVE Sustain Baseline Endoscopic Subscore

Abstract Background Tofacitinib is an oral, small molecule JAK inhibitor for the treatment of ulcerative colitis (UC). Efficacy and safety of tofacitinib in patients (pts) with UC were evaluated in two 8-week, Phase 3 induction studies (OCTAVE Induction 1&2), a 52-week, Phase 3 maintenance study (OCTAVE Sustain), and an open-label, long-term extension study. Here, we evaluate tofacitinib efficacy in OCTAVE Sustain (NCT01458574) by baseline Mayo endoscopic subscore (ES). Methods Pts with an OCTAVE Sustain baseline ES of 0 or 1 were included. Proportion of pts achieving efficacy endpoints at Week 52 of OCTAVE Sustain (responders from OCTAVE Induction 1&2; tofacitinib 5 or 10 mg twice daily [BID] or placebo [PBO]) was evaluated in relation to OCTAVE Sustain baseline ES. Using logistic regression, the difference in treatment effect (tofacitinib vs PBO) between baseline ES (0 vs 1) for each efficacy endpoint was assessed. Cox proportional hazards regression was used to model the difference in treatment effect between baseline ES (0 vs 1) for time to treatment failure and loss of response. Results At Week 52 of OCTAVE Sustain, a numerically higher proportion of pts with a baseline ES of 0 achieved remission and endoscopic improvement vs pts with a baseline ES of 1, regardless of tofacitinib dose (Table). Logistic regression analyses showed a larger treatment effect of tofacitinib 5 mg BID at Week 52 in pts with a baseline ES of 0 vs 1 (p=0.0306) for clinical response (Table). Treatment effect of tofacitinib 10 mg BID was not significantly different between baseline ES 0 vs 1 for any endpoint (Table). In tofacitinib 5 mg BID-treated pts, Cox proportional hazards regression showed difference in risk vs PBO to be larger in pts with a baseline ES of 0 vs 1 for treatment failure (p=0.0231) and loss of response (p=0.0209); corresponding differences for 10 mg BID-treated pts were not significant (p=0.9239 and p=0.9613, respectively). Conclusion In general, at Week 52 of OCTAVE Sustain, a higher proportion of tofacitinib-treated pts achieved endpoints with a baseline ES of 0 vs 1. Based on proportions, the treatment effect of tofacitinib 5 mg BID was more evident in pts with a baseline ES of 0 vs 1. This difference in treatment effect related to baseline ES was not observed with tofacitinib 10 mg BID. Baseline ES may be an important factor in predicting outcomes, and suggests that aiming for a deeper remission after induction may allow successful maintenance with tofacitinib 5 mg BID. These analyses were post hoc and limited by the small sample sizes.