Objectives: To determine the cumulative incidence and the prognostic factors of ileorectal anastomosis (IRA) failure after colectomy for ulcerative colitis (UC).
Background: Although ileal pouch-anal anastomosis is recommended after colectomy for UC, IRA is still performed.
Methods: This was a multicenter retrospective cohort study, which included patients with IRA for UC performed between 1960 and 2014. IRA failure was defined as secondary proctectomy and/or rectal cancer occurrence. Uni- and multivariate survival analyses were performed using Cox-proportional hazards models.
Results: A total of 343 patients from 13 French centers were included. Median follow up after IRA was 10.6 years. IRA failure rates were estimated at 27.0% (95% confidence interval, CI, 22-32) and 40.0% (95% CI 33-47) at 10 and 20 years, respectively. Median survival time without IRA failure was estimated at 26.8 years. Two thirds of secondary proctectomies were performed for refractory proctitis, and 20% for rectal neoplasia. Univariate analysis identified factors associated with IRA failure: IRA performed after 2005, a longer duration of disease at the time of IRA, indication for colectomy and having received immunomodulative agents before IRA. In multivariate analysis, treatment with both immunosuppressant (IS) and anti-TNF before colectomy was independently associated with IRA failure (HR=2.9, 95% CI 1.2-7.1). Conversely, colectomy for severe acute colitis was associated with decreased risk of IRA failure (HR=0.6, 95% CI 0.4-0.97).
Discussion: Patients with UC have a high risk of IRA failure, particularly when colectomy is performed for refractory disease. However, IRA could be discussed after colectomy for severe acute colitis, or in patients naive to IS and anti-TNF.