DOP046. A multicenter evaluation of clinical and surgical risk factors for anastomotic leak after restorative procto-colectomy with ileal pouch-anal anastomosis
S. Sahami1, C. Buskens1, R. Lindeboom2, T. Young-Fadok3, A. de Buck van Overstraeten4, A. D'Hoore4, W. Bemelman1, 1Academic Medical Center, Surgery, Amsterdam, Netherlands, 2Academic Medical Centre, Divisions of Clinical Methods and Public Health, Amsterdam, Netherlands, 3Mayo Clinic, Surgery, Phoenix, United States, 4University of Leuven, Surgery, Leuven, Belgium
Anastomotic leakage (AL) is one of the most feared complications after restorative proctocolectomy with ileal pouch-anal anastomosis (IPAA) that could negatively impact long-term patient outcome in pouch function and quality of life. Although previous studies have identified several risk factors for AL, predictive factors in the specific current IBD patient remain subject of debate. Since timely identification of high-risk patients could influence surgical decision-making and diminish the risk for complications, the aim of our study is to identify clinical and surgical parameters associated with AL.
Between September 1990 and April 2013, a total of 691 patients who underwent IPAA for IBD, dysplasia, or FAP were identified from prospectively maintained databases of 3 colorectal tertiary referral centres. Retrospective chart review identified data on demographic and surgical variables. AL was defined as any leak confirmed by either contrast extravasation on imaging or during re-laparotomy (leak grades B; drainage and C; re-laparotomy). Multivariate regression models were developed to identify risk factors for AL.
A total of 691 patients (55.7% male) were included with a median age of 39 years (17–77). One hundred and two (14.8%) patients developed postoperative AL. Univariate analysis identified, age at surgery (>55 years), long-term disease course (>5 years), overweight (BMI >25), high ASA classification (>3), steroids (>20 mg), anti-TNF (<3 months preoperatively) and the combination of both therapies as risk factors. Surgical factors were multistaged procedures (primary IPAA vs subtotal colectomy with completion proctectomy and IPAA at a later stage), J-pouch and perioperative blood transfusion. Multivariate regression models demonstrated, long-term disease course (OR 2.01, 95% CI 1.27–3.19), high ASA score (OR 1.94, 95% CI 1.09–3.47) and a combination of anti-TNF and steroid treatment (OR 5.61, 95% CI 1.71–18.48) as independent preoperative risk factors for AL. The only surgical risk factor that was independently associated with decreased leak rate was subtotal colectomy with IPAA at a later stage (OR 0.53, 95% CI 0.33–0.846). Since a staged procedure was therefore considered as a confounding variable, subgroup analysis of patients with primary IPAA demonstrated that long-term disease course (OR 1.79, 95% CI 1.03–3.14) and a combination of anti-TNF and steroids (OR 3.96, 95% CI 1.15–13.77) remained independent preoperative risk factors.
Long-term disease course, high ASA score, and a combination of anti-TNF and steroid treatment within 3 months before IPAA surgery were all independent preoperative risk factors for AL. A staged procedure seems an appropriate strategy when these risk factors are identified.