Our experience with endoscopic repair of large colonoscopic perforations and review of the literature

Feasibility of endoscopic closure of an iatrogenic colon perforation occurring during colonoscopy

BACKGROUND

Colon perforation is one of the most dreaded complications of colonoscopy. Traditionally, patients with a colon perforation have been treated surgically. Although there are several case reports documenting the usefulness of endoscopic closure of colon perforations, there are few current data evaluating the feasibility of endoscopic closure for an iatrogenic perforation on consecutive patients undergoing colonoscopy.

OBJECTIVE

To assess the incidence of colon perforations and the utility of immediate endoscopic closure during colonoscopy.

DESIGN

Retrospective, observational study.

SETTING

Tertiary-care academic medical center.

PATIENTS

All patients who underwent colonoscopy at 1 institution from June 2002 to December 2008 were identified.

INTERVENTION

An attempt at immediate colon perforation closure by endoscopic means.

MAIN OUTCOME MEASUREMENTS

Successful endoscopic closure of colon perforation.

RESULTS

During the study period, a total of 8601 colonoscopies were performed (2472 therapeutic interventions, 28.7%). A total of 12 iatrogenic colon perforations occurred, yielding a rate of 1.4/1000. Five (41.7%) occurred during a diagnostic colonoscopy (0.8/1000), and 7 perforations (58.3%) occurred as the result of a therapeutic intervention (2.8/1000). Endoscopic closure of the perforation site was possible in 5 patients (42%). Seven patients were treated surgically (large defects [n = 3], including 1 failed endoscopic closure, difficult endoscope position [n = 2], stool contamination [n = 1], and endoscopist's inexperience with closure of mucosal defects [n = 1]).

LIMITATION

Retrospective design.

CONCLUSIONS

In this study, the incidence of colon perforations was 1.4/1000. Endoscopic closure of iatrogenic colon perforations was attempted in 50% of patients and was successful in 83%. All patients with successful endoscopic closure had lesions smaller than 10 mm.

Tulip bundle technique and fibrin glue injection: Unusual treatment of colonic perforation

We report a case of a 63-year-old male who experienced an iatrogenic sigmoid perforation repaired combining three endoscopic techniques. The lesion was large and irregular with three discrete perforations, therefore, we decided to close it by placing one clip per perforation, and then connecting all the clips with two endoloops. Finally we chose to use a fibrin glue injection to obtain a complete sealing. Four days after the colonoscopy the patient underwent a laparoscopic right hemicolectomy due to evidence of a large polyp of the caecum with high grade dysplasia and focal carcinoma in situ. Inspection of the sigma showed complete repair of the perforation. This report underlines how a conservative approach, together with a combination of various endoscopic techniques, can resolve complicated iatrogenic perforations of the colon.

Endoscopic repair of small symptomatic gastrogastric fistulas after gastric bypass surgery: a single center experience

Gastrogastric fistula (GGF) is a known complication of gastric bypass surgery. Revisional surgery for GGF repair can be technically challenging. We describe our experience with endoscopic repair of small GGFs. A retrospective review was performed to identify patients in whom symptomatic GGF was repaired endoscopically at our institution between September 2004 and September 2008. At endoscopy, the fistulous margins were debrided using cold biopsy forceps or ablated using Argon Plasma Coagulation (APC). The fistula was then repaired with endoclips. Status of GGF repair was assessed intra-operatively, at 2 weeks by upper gastrointestinal (UGI) series, and at regular follow-up thereafter. GGF repair was attempted in eight female patients (mean age = 47 years). The average time interval between gastric bypass surgery and GGF presentation was 81 months. The presenting symptoms included nausea, vomiting, abdominal pain, and weight regain. The average duration of endoscopic procedure was 55 min. All GGFs were small (<20 mm). Endoscopic repair of GGF was successful intra-operatively in all patients. Two patients had failure of GGF repair at 2 weeks. Other two patients experienced recurrent symptoms after several weeks and had a delayed failure of GGF repair diagnosed by UGI series. Endoscopic repair has remained successful in four patients at 8-46 months follow-up. Endoscopic repair of small GGFs using endoclips is feasible. It must be considered as an option for management of small GGFs, given its safety, and ease of performance compared to revisional surgery.

Enterovesical fistulas: diagnosis and management

Enterovesical fistula (EVF) is an abnormal communication between the intestine and the bladder. It represents a rare complication of inflammatory or neoplastic disease, and traumatic or iatrogenic injuries. The most common aetiologies are diverticular disease and colorectal carcinoma. Over 75% of affected patients describe pathognomonic features of pneumaturia, faecaluria and recurrent urinary tract infections. The diagnosis of EVF can be challenging, and frequently patients are monitored for months before the condition is recognised and treated effectively. Diagnostic tools include laboratory tests, imaging studies and endoscopic procedures. Although conservative management can be attempted in selected patients, in most cases, the treatment is mainly based on surgical interventions. Recently, the laparoscopic approach to EVF has been shown to be safe and effective. Although it is a rare condition in a general surgery setting, EVF is a challenging condition leading to high morbidity and mortality rates.

[Large iatrogenic colonic perforation treated by endoscopic suturing]

Perforation of the colon during colonoscopy is still one of the most severe complications of this technique and occurs with a frequency of between 0.12 % and 0.2 % of cases after diagnostic colonoscopy and in up to 3 % of patients after therapeutic colonoscopy. The site of perforation is usually the sigmoid colon. The gold standard for treatment of this complication is surgery to be performed as rapidly as possible: a simple suture and peritoneal cleaning, with limited resection and anastomosis or colostomy only in case of confirmed fecal peritonitis. However, interventional endoscopy has made progress, in particular endoscopic suturing and Natural Orifice Transluminal Endocopic Surgery (NOTES) has been developed. There are several reports of endoscopically sutured perforated colons, most less than 10mm. We report our experience of two colonic perforations which were at least 10mm treated by endoscopic suturing with hemoclips: a perforated sigmoid diverticulum during simple colonoscopy in the first case and a large polypectomy by endoscopic mucosal resection of the ascending colon in the second.

Natural orifice translumenal endoscopic surgery used for perforated viscus repair is feasible using lower peritoneal pressures than laparoscopy in a porcine model

BACKGROUND

Procedure-related complications contribute to 1-year mortality in patients with perforated ulcers. Natural orifice translumenal endoscopic surgery (NOTES) might offer a new repair approach.

STUDY DESIGN

Swine were randomized to laparoscopic or NOTES repair. Laparoscopic gastrotomy creation (1 cm) was followed by 4 hours soilage time. After peritoneal cavity irrigation (per group assignment), repair proceeded with a laparoscopic or NOTES approach. For NOTES repair, omentum was endoscopically grasped, pulled into the gastric lumen, and fixed with metallic clips. Feasibility; time to complete procedures; pneumoperitoneal pressures; and clinical parameters, including necropsy and peritoneal culture at 2 weeks, were recorded.

RESULTS

NOTES repair failed in 1 animal (technical); repair was completed laparoscopically, and data were analyzed as intention to treat. Specific NOTES repair time (minutes) was comparable with laparoscopy (36 versus 46; p = 0.2). Mean abdominal pressure (mmHg) required to complete NOTES repair was lower than in laparoscopy (4 versus 12; p < 0.001). Nineteen of 23 animals thrived until necropsy at 2 weeks. Three animals succumbed to airway compromise in recovery; 1 NOTES animal failed to thrive on postoperative day 7. No intra-abdominal cause for these deaths was found. At necropsy all repairs were intact, and peritoneal cultures revealed a small and equivalent amount of colony-forming units in each group.

CONCLUSIONS

Endoscopic ulcer repair appears technically feasible with similar clinical and infectious outcomes to laparoscopy. The lower required pneumoperitoneal pressures used in these NOTES techniques are recognizable different outcomes from laparoscopy and can be advantageous in critically ill patients.

Colonoscopic perforation: Incidence, risk factors, management and outcome

This review discusses the incidence, risk factors, management and outcome of colonoscopic perforation (CP). The incidence of CP ranges from 0.016% to 0.2% following diagnostic colonoscopies and could be up to 5% following some colonoscopic interventions. The perforations are frequently related to therapeutic colonoscopies and are associated with patients of advanced age or with multiple comorbidities. Management of CP is mainly based on patients’ clinical grounds and their underlying colorectal diseases. Current therapeutic approaches include conservative management (bowel rest plus the administration of broad-spectrum antibiotics), endoscopic management, and operative management (open or laparoscopic approach). The applications of each treatment are discussed. Overall outcomes of patients with CP are also addressed.