Relevant studies published between January 2000 and December 2016 were identified from the search of the Medline databases and Cochrane databases. The following search terms were used: rectal, rectum, proctectomy and leakage, failure, integrity, insufficiency, breakdown, defect, separation, dehiscence. Further articles were then selected based upon a review of the citations found in selected papers from the first search. All English language publications which primarily focused on the management of anastomotic leaks following low anterior resections were selected. Exclusion criteria included: (1) studies primarily focusing on risk factors, prevention, recurrence, or the treatment of other types of complications; (2) studies in which the majority of leaks were not involving a rectal anastomosis (ileo-colic, colo-colonic, ileo-anal); (3) non-English papers; (4) animal or laboratory studies. To avoid redundant studies, all of the authors and organizations, community of patients and study dates were routinely checked. When a study reporting the same patient cohort was included in several publications, only the most recent or complete study was selected.

The patients that are the most in need of infectious source control are those with generalized peritonitis and/or sepsis. Most surgeons would agree on the need for fluid resuscitation, antibiotics and operative intervention to drain and divert. However, aside from gross ischemia or complete dehiscence, there continues to be controversy over whether the anastomosis should be taken down or salvaged. We specifically wanted to know whether or not anastomotic salvage leads to inferior source control and therefore higher mortality rates as compared to anastomotic takedown. Also, as has been suggested in prior studies, does anastomotic salvage provide any benefits over takedown in terms of the ability to re-establish intestinal continuity and prevent the number of permanent stomas?

There is very little consensus regarding the best methods for preserving or reestablishing intestinal continuity. Despite the high rates of permanent stomas and poor rectal function in patients with anastomotic leak, many surgeons continue to rely on a wait and see approach. Definitive treatment to allow for complete closure is therefore delayed in the hopes of spontaneous healing. Ultimately, this approach has been and will continue to be challenged by the emergence of active therapies to treat anastomotic leaks.

At our institution we have used endoluminal vacuum (E-Vac) therapy in the treatment of anastomotic leaks. While this therapy is not commonly utilized in the US, it has been used primarily in Germany since around 2002 [8]. Without any established gold standard for comparison, we decided to create PICO tables that compared E-Vac therapy to redo surgery, and conservative management (including the “wait and see approach,”). Other methods such as endoscopic stent placement will be less formally reviewed as it was anticipated that available studies concerning these methods would be limited.

A manual search was also performed focusing on the search terms endoluminal vacuum therapy, endoscopic vacuum therapy, endo-SPONGE, endosponge, endo sponge, and endoluminal negative pressure therapy. Additional articles were found using a Google Scholar search using the same search terms as well as from a review of the citations of selected articles. All studies evaluating the use of E-Vac therapy were reviewed and assessed for treatment related complications. Only studies including ten or more patients treated with E-Vac therapy were used in comparing outcomes between interventions. One final manual search was performed for a better evaluation of the baseline risk of permanent stoma in patients with and without anastomotic leak. Keywords used were “permanent stoma,” “definitive stoma,” “permanent ostomy”, “definitive ostomy.”

Traditionally the treatment of choice for a leaking colorectal anastomosis has been resection with end colostomy. This is despite the limited evidence to support this practice (Table 49.1a, b). In fact, one of the commonly referenced studies which emphasized the need for anastomotic takedown contained only three patients so treated [13]. More recently, the need for anastomotic takedown has been questioned and the trend continues to be moving away from this approach and towards performing anastomotic salvage.

Comparisons between anastomotic takedown and salvage were limited to four studies reporting on two of the four important outcomes. Patients treated with anastomotic salvage had statistically significant fewer postoperative deaths [9] and permanent stomas [9, 12] compared to patients treated with anastomotic takedown. Patients treated with anastomotic takedown as compared to anastomotic salvage also had more episodes of recurrent sepsis [22.7 % (5/22) vs. 0 % (0/10)] [12] and underwent an additional laparotomy more often [18.5 % (10/54) vs. 7.7 % (3/39)] [9], respectively.

These differences must be analyzed with caution based upon the overall quality of the studies (low to very low). Treatment bias may result in severe leakage (larger defects, colon necrosis etc.) being treated with anastomotic takedown, but remains unlikely account for differences in outcomes.

The authors of all four included studies reported favoring the use of anastomotic salvage [9, 10, 12, 11]. In three of the four studies, anastomotic takedown with creation of an end stoma was only favored in the management of anastomoses with ≥ 50–100 % dehiscence or in the presence of bowel ischemia or necrosis [9, 11, 12]. In the absence of the above criteria, diverting ostomy and salvage of the anastomosis is an effective method of controlling peritoneal sepsis resulting from leakage of both intraperitoneal and extraperitoneal rectal anastomoses. Anastomotic salvage and diversion is also the favored approach when anastomoses are inaccessible or poorly visualized as a result of significant inflammation, exudate, and/or adhesions.