An endoscopic mucosal grading system is predictive of leak in stapled rectal anastomoses.

Abstract

Anastomotic leak is a devastating postoperative complication following rectal anastomoses associated with significant clinical and oncological implications. As a result, there is a need for novel intraoperative methods that will help predict anastomotic leak. From 2011 to 2014, patient undergoing rectal anastomoses by colorectal surgeons at our institution underwent prospective application of intraoperative flexible endoscopy with mucosal grading. Retrospective review of patient medical records was performed. After creation of the colorectal anastomosis, application of a three-tier endoscopic mucosal grading system occurred. Grade 1 was defined as circumferentially normal appearing peri-anastomotic mucosa. Grade 2 was defined as ischemia or congestion involving <30% of either the colon or rectal mucosa. Grade 3 was defined as ischemia or congestion involving >30% of the colon or rectal mucosa or ischemia/congestion involving both sides of the staple line. From 2011 to 2014, a total of 106 patients were reviewed. Grade 1 anastomoses were created in 92 (86.7%) patients and Grade 2 anastomoses were created in 10 (9.4%) patients. All 4 (3.8%) Grade 3 patients underwent immediate intraoperative anastomosis takedown and re-creation, with subsequent re-classification as Grade 1. Demographic and comorbidity data were similar between Grade 1 and Grade 2 patients. Anastomotic leak rate for the entire cohort was 12.2%. Grade 1 patients demonstrated a leak rate of 9.4% (9/96) and Grade 2 patients demonstrated a leak rate of 40% (4/10). Multivariate logistic regression associated Grade 2 classification with an increased risk of anastomotic leak (OR 4.09, 95% CI 1.21-13.63, P=0.023). Endoscopic mucosal grading is a feasible intraoperative technique that has a role following creation of a rectal anastomosis. Identification of a Grade 2 or Grade 3 anastomosis should provoke strong consideration for immediate intraoperative revision.