Our study found that laparoscopic reintervention with anastomotic preservation for the management of colorectal anastomotic leakage was a viable option in 49 patients, thanks to early revisional surgery. The overall morbidity rate was 30.6% and one patient died (mortality rate 2%). Combined laparoscopy and transanal endoluminal repair of anastomotic leakage was associated with a lower overall complication rate (13.6% vs. 44.4%, p = 0.03) and fewer deep/organ space intraabdominal infections (4.5% vs. 29.6%, p = 0.03) compared with the control group. Also, a lower permanent stoma rate was noted in the repair group compared to the no repair group (18.2% vs. 25.9%, p = 0.73).
In the absence of consensus on standardized management of colorectal anastomosis leakage, some authors would argue that patients with stoma or localized abscess should been managed conservatively or undergo percutaneous drainage without reoperation [15, 16]. However, definitive closure of leaks with these methods can take several weeks, or be complicated with sinus formation or fistula [17, 18]. Laparoscopic reintervention for anastomotic leakage after laparoscopic colorectal surgery has gained widespread acceptance and several reports have shown that it is safe [13, 15, 19,20,21,22]. Traditionally, management for anastomotic leakage was via laparotomy because of the fear of bowel injury due to distended bowel and/or inadequate exposure through the laparoscopic approach . In our series, two patients had serosal injury which were repaired immediately; both patients had undergone late reintervention (> 5 days), that is recognized as a cause of bowel distension, dense adhesions and inadequate visualization . This is an argument in favor of early reintervention. Compared to relaparotomy, laparoscopic reintervention has the advantage of reduced wound complications (infection and incisional hernia) [15, 19, 20] and maintains the advantages of laparoscopic surgery . Indeed, as in our series, the previous trocar wounds can be reused for laparoscopic reintervention and the minilaparotomy wound does not have to be re-opened. There were three patients who had surgical site infection and three patients with ventral hernia in our series. All six patients with wound complications were patients who had undergone conversion.
Another concern of laparoscopic reintervention for leaks has been that pneumoperitoneum would cause fecal ascites and then exacerbate intra-abdomen infection . However, several studies have shown that laparoscopy does not increase intra-abdomen infection when compared with laparotomy [16, 20]. Lee et al. reported a 6.6% intra-abdominal infection rate after laparoscopic reintervention compared to 31.3% in the open group . However, there was no clear definition of intra-abdomen infection or the additional interventions related to intra-abdomen infection. Although intraabdominal infection was the major complication after laparoscopic reintervention in our series, all patients had an uneventful recovery after either conservative therapy or additional intervention.
Most studies on laparoscopic reintervention for colorectal anastomosis leakage have recommended “divert and drain” with anastomosis preservation [13, 14, 19, 20]. Reports of colorectal anastomosis repair are rare [13, 20, 22]. Lee et al. reported 61 laparoscopic reinterventions for colorectal anastomosis leakage: 12 patients had trans anal repair . However, the authors did not describe the outcome after anastomotic repair. Brunner et al. reported two patients who underwent trans anal repair with a single-port device (SILSTM Port CovidienTM); no stoma was performed  similar to our previously reported technique .
Our overall permanent stoma rate was 20.4% (10/49), similar to other studies [14, 16, 20]. This includes both the patients with stoma formation at the original operation (created according to patient status and surgeon preferences) and those who received a stoma at the reoperation. Of note, although there were fewer patients with permanent stoma in the repair group (18.2% vs. 25.9%), the difference was not statistically significant. When complicated by sinus formation and/or fistula, diverting stoma and simple drainage without repair may lead to delayed or no stoma closure . None of our patients had sinus formation or late fistula. Furthermore, intestinal healing after anastomotic leak can be associated with intense fibrosis and eventually some degree of anastomotic stricture [24, 25]. We believe that early repair could possibly avoid these complications or at least reduce the inflammatory response associated with their persistence. The improved post-operative course with fewer intra-operative complications may explain why more patients in our series were scheduled for stoma reversal. However, our numbers are small and long term follow-up is needed.
Our morbidity rate (30.6%) is at the lower limit of the range reported in the literature [19, 25, 26]. However, none of these studies concerned the outcome of transanal endoluminal repair. Morbidity in the control (no-repair) group was high (13.6% vs. 44.4%, p = 0.03) although there was no difference in APACHE II score (7.9 vs. 10.3, p = 0.12), anastomotic defect characteristics, or diffuse peritoneal contamination (27.3% vs. 37.1%, p = 0.47) between the two groups. Possible explanations include the higher deep O/S SSI rate in the control group (29.6% vs. 4.5%, p = 0.03). When the anastomotic site is not repaired, it is possible that the leaking site could still be an active source of infection, even though the leak will eventually heal. Endoscopic vacuum-assisted closure has been reported as viable option to deal with anastomotic leak after low anterior resection with effective control of septic focus . However, the duration of treatment can be long (34.4 ± 19.4 days) . Although we have no formal proof, we believe that the ease with which the leak can be assessed and repaired endo-luminally (vs. trans abdominal laparotomy or laparoscopy) are strong arguments in favor of the transanal endoluminal route for extraperitoneal colorectal anastomosis leakage repair and could decrease the intraabdominal infection rate associated with laparoscopic reintervention. Indeed, intraabdominal infection is the main complication after laparoscopic reintervention and it is also the main reason for additional interventions after reoperation [19, 20]. In our study, all patients who had O/S SSI (including four patients after additional intervention) could be managed safely by antibiotics, radiological drainage or re-laparoscopy. There was no statistically significant difference in additional intervention rates, hospital stay or outcomes between two groups.
Our study has several limitations. This study was retrospective, and the sample size was small (n = 49). Although there were variations in the stage of disease stage and the distance of the anastomosis from the anal verge, these differences were not statistically significant. As well, there was no statistically significant difference in outcome between the two groups in spite of these variations as well as to whether leakage occurred early or late, or the technique of anastomosis at primary surgery. In addition, the indication for transanal endoluminal repair and stoma creation was made at the surgeon’s discretion. Although it may be argued that many patients with anastomotic leak after anterior resection may be suitable for conservative treatment, this may prolong hospital stay, delay reintervention or compromise stoma reversal. According to our experience, re-laparoscopy is an early diagnostic tool for anastomosis leakage and transanal endoluminal repair is easier when re-operation is early [12, 13] because bowel tissues become more inflamed, and adhesions are firmer when the intervention is performed at a later stage.