Surgical Closure vs Anti-TNF in the Treatment of Perianal Fistulas in Crohn's Disease (PISA-II): a Comprehensive Cohort Design

Currently, the treatment of Crohn's patients with perianal fistulas predominantly exists of anti-TNF medication. However, its efficiency has never been directly compared to surgical closure of the perianal fistula. The aim of this study is to compare radiological fistula healing at MRI after 18 months follow-up in Crohn's patients undergoing surgical closure to anti-TNF medication as treatment of perianal Crohn's fistulas. Study design: In this multicenter comprehensive cohort design (CCD) Crohn's patients with a (re)active high perianal fistula will be allocated to anti-TNF for 1 year or surgical closure after 2 months under a short course of anti-TNF. Patients with a distinct preference will be treated accordingly, whereas only indifferent patients will be randomised in the usual way. Main study parameters/endpoints: The primary outcome parameter is the number of patients with radiologically closed fistulas based on an evaluated MRI-score after 18 months. Secondary outcomes are clinical closure, number of patients undergoing surgical re-interventions and number of re-interventions, recurrences and quality of life based on the Perianal Disease Activity Index (PDAI). Nature and extent of the burden and risks associated with participation, benefit and group relatedness: All patients will receive one of the two standard treatment approaches that are currently used for Crohn's fistulas. All effort has been performed to ensure most optimal treatment, according to best available evidence and current guidelines. Since there is no experimental study-arm, there are no additional risks associated with participation. During the study, the medical staff and trial nurses will monitor the necessity of surgical interventions and hospitalizations. At baseline and after 18 months all patients will undergo a MRI to score the fistula. Secondary outcome parameters will be assessed during visits to the outpatient clinic or telephone consultations at baseline and at intervals of 3 months for the duration of the study period. Every 6 months patients were asked to fill out the PDAI questionnaire with their physician. Based on the available literature, radiological closure of fistulas is expected in 40% of patients in the surgical closure group compared to 15% in the anti-TNF group. The increase in closure rate from 15% to 40% is considered clinically relevant. Due to the combination of a preference and randomized cohort, the appropriate sample size to detect this 25% difference is flexible and is adjusted for a skewed distribution. The minimal sample size, in case of a 1:1 treatment allocation, needed to detect this difference with a Chi-square test equals 86 patients (alpha 0.05, power 80%). The maximal allowed skewed distribution is set at 1:4, which will result in a maximal sample size of 116 patients.