Introduction Anti-TNFs are effective agents in inducing and maintaining remission in patients with inflammatory bowel disease (IBD). However, the optimal duration of treatment is still a matter of debate. It has been described that approximately 50% of patients who discontinue biologics while in remission will relapse and this proportion may be increasing with time. Risk factors for relapse have been identified but these do not seem repeatable in all studies^1-2. The aim of this study was to capture relapse rates in IBD patients in remission post anti-TNF withdrawal in a real life cohort and to identify predictors of relapse or ‘early’ relapse (≤ 1 year).

Methods All patients who achieved clinical remission (defined as a Harvey-Bradshaw score <5) on anti-TNFs for ≥1 year and discontinued biologics were included in the study. Retrospective data regarding patients’ demographics, medical history and course of disease were retrieved and patients were prospectively followed up post anti-TNF discontinuation. Potential risk factors for relapse as well as all laboratory, endoscopic and imaging data were recorded. Univariate analysis used t-test for parametric and Mann-Whitney for non-parametric continuous variables, while logistic regression was used for multivariate analysis.

Results 42 patients discontinued anti-TNFs on clinical remission between 2002–2014. All but two had Crohn’s disease. Mean duration of anti-TNF treatment prior to discontinuation was 3.5±2 years (range 1–11). Follow up period of the study was 36±28 months. Prior to withdrawal 24 patients (57%) had a colonoscopy (confirming endoscopic and histologic remission), 22 (52%) had imaging of their small bowel and 10 (24%) had a faecal calprotectin (FC, mean value 33±21, range 0–70). 28 patients (67%) relapsed post anti-TNF discontinuation, most of them within the first year post withdrawal (n = 16). Out of several factors that were univariately associated with higher rates of relapse (no calprotectin measurement prior withdrawal p = 0.042, younger age at diagnosis p = 0.026, male gender p = 0.023, absence of stricturing/ penetrating disease p = 0.048, concomitant immunomodulators p=NS) only not requesting a calprotectin prior withdrawal predicted higher rates of relapse in multivariate analysis (p = 0.036, OR 8.839, 95% CI 1.159–67.438). 23/28 (82%) re-started anti-TNF on relapse, 90% of whom optimally responded on re-initiation. 24 (60%) patients were in clinical remission at three years post withdrawal.

Conclusion A normal FC in IBD remission can guide decision on stopping anti-TNF therapy. Rates of relapse in our cohort were higher than in published literature but were not associated with worse overall outcomes as majority of patients responded well on drug re-initiation.

References 1 Molander, et al. Inflamm Bowel Dis 2014 Jun;20(6):1021–8.

2 Louis, et al. Gastroenterology 2012;142:63–70.

Disclosure of Interest None Declared