Direct and Indirect Effects of Tofacitinib on Treatment Satisfaction in Patients with Ulcerative Colitis

Background and aims: This mediation modelling analysis evaluated direct and indirect effects of tofacitinib, an oral, small molecule Janus kinase inhibitor under investigation for ulcerative colitis, on patient treatment satisfaction.

Methods: Data from an 8-week randomized Phase 2 trial [NCT00787202] in adults with moderate-to-severe, active ulcerative colitis receiving twice-daily tofacitinib 0.5-15mg [n=146] or placebo [n=48] were analysed in patient-reported [n=149] and clinician-reported [n=170] outcomes-based mediation models. Binary predictor variable: Treatment [pooled active treatment vs placebo]. Eventual dependent variable: Week 8 patient treatment satisfaction [measured on a five-point Likert scale]. Mediators of treatment effect on satisfaction: Week 8 Inflammatory Bowel Disease Questionnaire domains [Bowel Symptoms, Emotional Health, Social Function and Systemic Symptoms] and Mayo scale domains [Stool Frequency, Rectal Bleeding, Physician's Global Assessment and Endoscopic Disease Activity] for patient-reported and clinician-reported models, respectively.

Results: Overall tofacitinib indirect effect on satisfaction via Inflammatory Bowel Disease Questionnaire domains was 40.5% [p<0.05] and via Mayo scale domains was 84.0% [p<0.01] for patient-reported and clinician-reported models, respectively. Bowel function had the most important indirect effect: of the total tofacitinib effect on satisfaction, 32.4% [p=0.05] was indirectly mediated via Bowel Symptoms; and 30.0% [p=0.04] via Stool Frequency. In total, 59.5% [p<0.01] and only 16.0% [p=0.56] of tofacitinib's effect on satisfaction was unrelated to Inflammatory Bowel Disease Questionnaire and Mayo scale domains in the patient-reported and clinician-reported models, respectively.

Conclusions: Bowel function is an important factor for patient treatment satisfaction with tofacitinib. Treatment effect on patient satisfaction was almost completely mediated via improvement in Mayo scale domains.