Search in the Abstract Database

Search Abstracts 2016

* = Presenting author

P584 Relapse risk and predictors for relapse in a real-life cohort of inflammatory bowel disease patients after discontinuation of anti-tumour necrosis factor therapy

S. Bots*, S. Kuin, M. Löwenberg, C. Ponsioen, G. van den Brink, G. D’Haens

Academic Medical Centre (AMC), Department of Gastroenterology and Hepatology, Amsterdam, Netherlands


The objective of this study was to investigate the success of planned anti-tumour necrosis factor (TNF) withdrawal in a real-life cohort of Crohn’s disease (CD) and ulcerative colitis (UC) patients in sustained clinical remission and to identify predictors of relapse.


In this mono-centre observational study, 79 patients (60 CD and 19 UC) who were in steroid-free remission for ≥ 1 year at the time of withdrawal of infliximab (IFX) or adalimumab (ADA), were followed-up for a median of 14 months. The frequency of clinical relapse after anti-TNF withdrawal was analysed according to clinical, biochemical, and endoscopic variables. Disease relapse was defined as relapsing symptoms and the need to (re)start treatment with anti-TNF agents, immunomodulators or corticosteroids. Predictors for relapse were studied using cox proportional hazard analysis.


Median duration of anti-tumour necrosis factor (TNF) therapy at stop was 52 months and the median follow-up after anti-TNF cessation was 14 months. IFX was discontinued in 44 (56%) and anti-drug antibodies (ADA) in 35 patients (44%). In total, 27 patients (34%) experienced a relapse (CD 21; UC 6) with a median time to relapse of 5 and 3 months in CD and UC, respectively. Fifteen (83%) patients responded to retreatment with anti-TNF therapy. Continuing immunomodulatory treatment was associated with a lower relapse rate in UC patients (HR 0.1, 95% CI 0.0–10.9). We observed a trend that IFX or ADA trough levels ≥ 2 (irrespective of anti-TNF therapy) in the year before anti-TNF discontinuation was associated with lower relapse rates, but this did not reach statistical significance in CD patients (HR 3.9, 95% CI 0.9–18.5). Endoscopic remission in the previous year, bowel-related surgery, previous anti-TNF use, smoking, perianal disease, disease duration, and duration of anti-TNF use were not associated with higher or lower relapse rates in this study. Factors such as CRP and faecal calprotectin as predictors for relapse were not addressed because they were within the normal range in most patients.


Our results confirm that anti-TNF withdrawal is associated with a high relapse risk. Continuing immunomodulatory treatment was associated with a lower relapse rate in UC patients. Although not statistically significant, trough levels ≥ 2 seem to be associated with higher relapse rates. Retreatment with anti-TNF agents was successful in 83% of patients. This study is still ongoing, with longer follow-up time and more inclusions we expect to find higher relapse rates and more predictors associated with relapse. Discontinuation of anti-TNF does not appear to be beneficial in inflammatory bowel disease patients.