Search in the Abstract Database

Search Abstracts 2016

* = Presenting author

P456 Efficacy and safety of TNF antagonist for the treatment of internal fistulising Crohn’s disease

A. Huguet1, A.-L. Mallet1, V. Desfourneaux1, L. Siproudhis1, J.-F. Bretagne2, G. Bouguen*2

1Université Rennes 1 & CHU Pontchaillou Rennes, Service de chirurgie digestive, Rennes, France, 2Université Rennes 1 & CHU Pontchaillou Rennes, Service des Maladies de l’Appareil Digestif, Rennes, France


Despite numerous studies on entoerocutaneous fistula (mainly anoperineal fistula), data on the use of TNF antagonist for internal fistulising disease remain scarce, and their use debated related to the risk of abscess. The aim of the study was to assess efficacy and safety of anti-TNF for internal fistula (exclusion of entoerocutaneous and anoperineal fistula).


All chart of patients with Crohn’s disease and treated with TNF antagonist for internal fistula were assessed. Primary endpoint was the onset of a major abdominal surgery. Secondary analysis included disappearance of fistula tract during the follow-up, and safety. Kaplan–Meier method was used for statistical analysis.


In total, 66 patients were included with a median follow-up of 43.2 months between anti-TNF initiation and last news. Fistula tract was ileal in 86% of patients. The mean length of ileal involvement was 22.2 cm, and 47% (n = 32) of patients had associated abdominal abscess. Eight abdominal abscesses required percutaneous drainage. The fistula tract was complex for 44% of patients. Infliximab was use for a majority of cases. Half of patients (n = 33) had concomitant antibiotics, and 23 (35%) patients had concomitant immunosuppressant. The median duration between the diagnostic of internal fistula and the initiation of TNF antagonist was 2 weeks. During the follow-up, abdominal surgery was required for 43.9% of patients (n = 30). The surgery was decided because of primary non-response or abscess recurrence for 18 patients, in 2 patients for adenocarcinoma, and in 13 patients for stricturing disease. The cumulative probabilities of surgery were 19%, 29%, and 50% at 6 months, 1 year and 3 years, respectively. One operated patient had post-surgical complication with and abscess. Complex fistula tract with more than 2 tracts (HR = 2.5), haemoglobin level < 12.5 g/dL (HR = 4.1), and low albumin level < 35 g/L (HR = 3.28) at infliximab initiation predicted the need for surgery. Amongst the 37 surgery-free patients, disappearance of fistula tract was observed for 62% of patients (magnetic resonance imaging [MRI] or computed tomography [CT] scan or ultrasound).The mean time to fistula healing was 14.1 +/- 1.8 months.

During the follow-up, an abscess occurred in 11 patients (16.6%), and 9 of these 11 patients were operated on. Further, 3 patients (4%) died during the follow-up period: 2 patients developed an adenocarcinoma (ileal and colic) 5 and 7 years, respectively, following infliximab, and a 78-year-old woman died from a septic shock 3 months after the start of infliximab.


TNF antagonist is useful for one-third of patients with internal fistula. However, it increases the risk of sepsis and maybe cancer and therefore its use must be assess by a long-term prospective cohort. Use of TNF antagonists should be carefully discussed, patient by patient.