Back to 2011 Program

Objective: An anastomotic leak (AL) can refer to a broad range of clinical entities from an asymptomatic radiologic finding to a devastatingly morbid complication. We sought to define the spectrum of leaks that occur after low anterior resection (LAR) for rectal cancer and the associated consequences.
Design: Retrospective cohort study
Setting: Tertiary care academic medical center
Patients: 210 consecutive patients who underwent LAR for rectal cancer from 1996-2010.
Main Outcome Measures: Anastomotic leak, need for fecal diversion
Results: 17 pts had some form of AL. There were 6 intraperitoneal leaks (3%). 11 had contained presacral contrast extravasation; only 6 of these were diagnosed during the index admission. 3 were diagnosed more than 45 days postop. As such, the actual overall leak rate was 8.6%. Male gender, diabetes and neoadjuvant chemo/XRT were associated with a leak (relative risk of 6, 3.1 and 1.7 respectively). The RR for permanent stoma was 8.6 for free leak vs 4.5 for contained presacral collection. 10 additional pts had free pelvic fluid noted on a postop CT scan without a defined leak and three had pelvic abscesses. Both of these subgroups had an inferior functional result. There was one death from a leak.
Conclusions: AL after LAR actually takes several different forms with varied consequences for both functional outcome and the need for a permanent stoma. This has important consequences for quality reporting and assessing outcomes.