Endoscopic treatment of postoperative enterocutaneous fistulas after bariatric surgery with the use of a fistula plug: report of five cases

• Introduction
• Case report
• Discussion
• References

Anastomotic leaks frequently occur after bariatric surgery and their management includes different options. The present study describes the management of enterocutaneous fistulas in patients in whom surgical or endoscopic treatments have failed, by insertion of a biomaterial (Surgisis fistula plug) to facilitate healing of the gastrocutaneous fistula. Five patients with leaks after bariatric surgery were treated. All patients had undergone previous failed surgical or endoscopic attempt(s) at closure. Our technique entailed insertion of the Surgisis fistula plug into the fistula tract by a ”rendezvous” procedure, via both percutaneous and endoscopic routes. The data were collected retrospectively. Initially, two patients were treated by fistula plug alone and three received fistula plug plus a self-expanding stent. In two patients, cutaneous fistula outflow ceased within a few days. The other three patients required one additional endoscopic procedure. At the end we observed healed leaks in four of the five patients (80 %). The median follow-up duration was 18 months. In conclusion, the combined therapy consisting of fistula plug implantation with optional stenting helps closure in these difficult refractory cases of gastrocutaneous fistula.

Introduction

Bariatric surgery has increased in popularity due to an epidemic of obesity. In sleeve gastrectomy and Roux-en-Y gastric bypass (RYGBP) procedures, gastrocutaneous fistulas occurring at the level of the gastrojejunal anastomoses or at the level of the gastric stapling are among the most frequent technical complications. Leaks are responsible for a high morbidity rate [1], and mortality associated with reinterventions for such leaks arising after RYGBP is as high as 10 %, with a morbidity of 50 % [2]. The use of endoscopic techniques is thus appealing, and various procedures have been described over the last 10 years for the closure of fistulas. Recently, management with the insertion of self-expanding metal stents (SEMS), followed – once the self-expanding stents were made of plastic (SEPS) – by stent removal, was shown to be successful in 81 % of patients presenting with leaks after bariatric surgery [3]. Nevertheless, these cases still represent management failures.

Surgisis (Cook Biotech Inc., West Lafayette, Indiana, USA) ([Fig. 1]) is an acellular matrix biomaterial derived from the porcine small intestine submucosa that stimulates proliferation of fibroblasts in the region of wounds and is incorporated into the scar without stimulating a foreign-body inflammatory reaction [4]. It was previously used for repair of abdominal hernias [5], treatment of varicose ulcers, and treatment of bladder fistulas [6]. A cone-shaped fistula plug made of Surgisis (Surgisis AFP anal fistula plug; Cook Biotech Inc.) was developed for the repair of anorectal fistulas and was suggested to be superior to fibrin glue injection for this indication [7] [8].

The present study describes the technique of implantation of these fistula plugs in patients with refractory enterocutaneous fistula after bariatric surgery, and the long-term outcome of treatment.

Case report

Between April 2006 and March 2007, five patients presenting with enterocutaneous fistulas refractory to other treatments were treated with placement of a Surgisis anal fistula plug. The majority of patients were referred from other centers for endoscopic management of fistulas after previous surgical (n = 5) or endoscopic (n = 4) attempts at closure had failed. The fistulas were defined by an enterocutaneous tract, an outflow of at least 50 ml/day for more than 2 weeks, and occurrence following bariatric surgery. Three patients had undergone sleeve gastrectomy and two RYGBP. Two patients were treated by fistula plug alone and three by fistula plug plus self-expanding stent placement (due to a huge internal orifice). There were four women and one man, and the mean age was 40 years (range 34 – 45 years).

The technique consisted in insertion of the fistula plug into the fistula tract by a ”rendezvous” procedure, working via both percutaneous and endoscopic routes. All procedures were performed with the patient under general anesthesia. The technique involves two endoscopists, or one endoscopist and one physician assistant. A therapeutic gastroscope (GIF-1TQ160; Olympus Corp., Tokyo, Japan) is placed at the level of the internal opening of the fistula. First, the fistula tract is opacified by the injection of contrast medium. To perform the ”rendezvous”, a guide wire (Jagwire; Boston Scientific, Natick, Massachusetts, USA) is introduced with the help of an endoscopic retrograde cholangiopancreatography (ERCP) catheter (Cook Endoscopy, Winston-Salem, North Carolina, USA) under fluoroscopic control, from either the cutaneous orifice or the endoluminal orifice. The guide wire is then grasped either internally, with a snare through the operative channel of the scope, or externally, allowing an oral-to-cutaneous access through the fistula tract ([Fig. 2]). Under endoscopic and fluoroscopic guidance, the fistula tract is abraded using a large-caliber (11-Fr) modified ”stent pusher” with multiple barbs ([Fig. 3]) passed over the guide wire. Several push-and-pull movements are performed in order to induce abrasion and bleeding to promote better tissue healing. The pusher is then removed and a snare is passed through the scope and the fistula, attached to the guide wire. The snare becomes visible at the skin orifice and is used to grasp the thinner end of the fistula plug and pull it back inside the fistula tract ([Fig. 4]). Under endoscopic and fluoroscopic guidance, the fistula plug(s) (from one to three, according to the diameter of the fistula) is (are) left inside the fistula tract with less than 1 cm of the thinner part inside the digestive lumen ([Fig. 5]). Optionally, an endoloop or endoscopic clips may be used to fix this part to the gastrointestinal mucosal layer. SEMS were placed in some patients to cover large fistulas in order to increase the chance of healing of the fistula tract. Patients were given nothing by mouth for 2 – 3 days after which they underwent a barium X-ray. If no leakage was visible, they were then fed orally. They received one prophylactic dose of 2 g cefazolin before the procedure if they had not already been treated with antibiotics.

Patients were followed after the procedure. Primary outcomes taken to show definite healing were fistula closure, absence of intra-abdominal collections, and disappearance of or blind fistula tract at barium swallow without any stent in place. For the patients who had additional SEMS placement, healing was considered to have taken place only after stent removal according to the previously reported technique [3].

Overall, there were two patients in whom fistula outflow ceased within a few days after a single procedure. The other three patients required additional endoscopic treatment, which allowed definitive closure in two more cases. The individual results of the procedures are shown in [Tables 1] and [2]. The median duration of follow-up among patients with healed leaks was 17.8 months (8 – 29 months).

Fistula plug insertion, in some cases repeated and/or associated with stent placement, allowed four of these five difficult refractory enterocutaneous fistulas to heal. No complication related to fistula plug placement was observed, and no mortality was associated with the procedure.

Discussion

Our results indicate that application of a fistula plug (cone-shaped Surgisis AFP) offers a chance for curative treatment of enterocutaneous fistulas where previous attempts at surgical or endoscopic repair have failed. We also report a technique of implantation, using a rendezvous procedure, which allows safe and accurate placement of these plugs in the fistula tract. Pross et al. [9] and Truong et al. [10] reported closure rates for postoperative fistulas of 100 % (2/2) and 78 % (7/9), respectively, by combining a Vicryl plug and fibrin glue injection. The major difference between these studies and the present one is that our patients had already undergone stenting, fibrin glue injection, or reoperation before being referred to our center for rescue treatment, and that all of our patients had large fistulas demonstrated by barium X-rays. The patients to whom this treatment was offered were in all cases in the third line of treatment, after failed repair, and the median interval between fistula onset and treatment was greater than 20 weeks (range 4 – 44 weeks). We were therefore facing a highly selected group of patients with refractory fistula.

Our limited experience does not allow us to define in what conditions fistula plug alone (between one and three in the same tract) or in addition to stenting allows the best healing rate in these difficult situations. However, we can say that our one failure occurred in a patient with a short fistula tract which did not allow implantation of the whole plug. The simultaneous placement of a stent has the major advantage of avoiding the necessity for the patient to fast for another few weeks, thus improving the patient’s comfort and reducing social cost.

Our classical treatment for fistulas is the transient placement of metallic stents removed after a few months with the help of a plastic stent [3]. This has allowed an improved prognosis for those patients whose mortality rate after redo surgery may reach 10 % [2]. However, there still remain patients whose condition is difficult to manage, mainly those with large fistulas such as those with a fistula tract arising in a sleeve gastrectomy.

Although it is difficult at this stage to draw firm conclusions, due to the differences in the treatments (SEMS, SEPS, fistula plug) and the relative heterogeneity of the patient population, it is clear that fistula plug implantation, combined or not combined with stenting, helps closure in these difficult cases. This treatment should be tested as a first approach at least for the group of patients who have enterocutaneous fistulas in gastric sleeves and those in whom previous endoscopic stenting has failed. Technical refinements are of course needed to make the procedure more standardized in terms of implantation devices and technique.

References

• 1 Colquitt J, Clegg A, Loveman E. et al . Surgery for morbid obesity.  Cochrane Database Syst Rev. 2005;  4 CD003641
  PubMedSearch in Google Scholar
• 2 Gonzalez R, Sarr M G, Smith C D. et al . Diagnosis and contemporary management of anastomotic leaks after gastric bypass for obesity.  J Am Coll Surg. 2007;  204 47-55
  CrossrefPubMedSearch in Google Scholar
• 3 Eisendrath P, Cremer M, Himpens J. et al . Endotherapy including temporary stenting of fistulae of the upper gastrointestinal tract after laparoscopic bariatric surgery.  Endoscopy. 2007;  39 625-630
  Thieme ConnectPubMedSearch in Google Scholar
• 4 Ansaloni L, Cambrini P, Catena F. et al . Immune response to small intestinal submucosa (surgisis) implant in humans: preliminary observations.  J Invest Surg. 2007;  20 237-241
  CrossrefPubMedSearch in Google Scholar
• 5 Franklin Jr M E, Treviño J M, Portillo G. et al . The use of porcine small intestinal submucosa as a prosthetic material for laparoscopic hernia repair in infected and potentially contaminated fields: long-term follow-up.  Surg Endosc. 2008;  22 1941-1946
  CrossrefPubMedSearch in Google Scholar
• 6 Alpert S A, Cheng E Y, Kaplan W E. et al . Bladder neck fistula after the complete primary repair of exstrophy: a multi-institutional experience.  J Urol. 2005;  174 1687-1689; discussion 1689 – 1690
  CrossrefPubMedSearch in Google Scholar
• 7 Champagne B J, O’Connor L M, Ferguson M. et al . Efficacy of anal fistula plug in closure of cryptoglandular fistulas: long-term follow-up.  Dis Colon Rectum. 2006;  49 1817-1821
  CrossrefPubMedSearch in Google Scholar
• 8 Johnson E K, Gaw J U, Armstrong D N. Efficacy of anal fistula plug vs. fibrin glue in closure of anorectal fistulas.  Dis Colon Rectum. 2006;  49 371-376
  CrossrefPubMedSearch in Google Scholar
• 9 Pross M, Manger T, Reinheckel T. et al . Endoscopic treatment of clinically symptomatic leaks of thoracic esophageal anastomoses.  Gastrointest Endosc. 2000;  51 73-76
  CrossrefPubMedSearch in Google Scholar
• 10 Truong S, Böhm G, Klinge U. et al . Results after endoscopic treatment of postoperative upper gastrointestinal fistulas and leaks using combined Vicryl plug and fibrin glue.  Surg Endosc. 2004;  18 1105-1108
  PubMedSearch in Google Scholar

E. Toussaint, MD

Department of Gastroenterology, Hepatopancreatology and Digestive Oncology
Erasme Hospital

808 Route de Lennik
1070 Brussels
Belgium

Fax: +32–2-5554697

Email: emmanuel.toussaint@bordet.be

References

• 1 Colquitt J, Clegg A, Loveman E. et al . Surgery for morbid obesity.  Cochrane Database Syst Rev. 2005;  4 CD003641
  PubMedSearch in Google Scholar
• 2 Gonzalez R, Sarr M G, Smith C D. et al . Diagnosis and contemporary management of anastomotic leaks after gastric bypass for obesity.  J Am Coll Surg. 2007;  204 47-55
  CrossrefPubMedSearch in Google Scholar
• 3 Eisendrath P, Cremer M, Himpens J. et al . Endotherapy including temporary stenting of fistulae of the upper gastrointestinal tract after laparoscopic bariatric surgery.  Endoscopy. 2007;  39 625-630
  Thieme ConnectPubMedSearch in Google Scholar
• 4 Ansaloni L, Cambrini P, Catena F. et al . Immune response to small intestinal submucosa (surgisis) implant in humans: preliminary observations.  J Invest Surg. 2007;  20 237-241
  CrossrefPubMedSearch in Google Scholar
• 5 Franklin Jr M E, Treviño J M, Portillo G. et al . The use of porcine small intestinal submucosa as a prosthetic material for laparoscopic hernia repair in infected and potentially contaminated fields: long-term follow-up.  Surg Endosc. 2008;  22 1941-1946
  CrossrefPubMedSearch in Google Scholar
• 6 Alpert S A, Cheng E Y, Kaplan W E. et al . Bladder neck fistula after the complete primary repair of exstrophy: a multi-institutional experience.  J Urol. 2005;  174 1687-1689; discussion 1689 – 1690
  CrossrefPubMedSearch in Google Scholar
• 7 Champagne B J, O’Connor L M, Ferguson M. et al . Efficacy of anal fistula plug in closure of cryptoglandular fistulas: long-term follow-up.  Dis Colon Rectum. 2006;  49 1817-1821
  CrossrefPubMedSearch in Google Scholar
• 8 Johnson E K, Gaw J U, Armstrong D N. Efficacy of anal fistula plug vs. fibrin glue in closure of anorectal fistulas.  Dis Colon Rectum. 2006;  49 371-376
  CrossrefPubMedSearch in Google Scholar
• 9 Pross M, Manger T, Reinheckel T. et al . Endoscopic treatment of clinically symptomatic leaks of thoracic esophageal anastomoses.  Gastrointest Endosc. 2000;  51 73-76
  CrossrefPubMedSearch in Google Scholar
• 10 Truong S, Böhm G, Klinge U. et al . Results after endoscopic treatment of postoperative upper gastrointestinal fistulas and leaks using combined Vicryl plug and fibrin glue.  Surg Endosc. 2004;  18 1105-1108
  PubMedSearch in Google Scholar

E. Toussaint, MD

Department of Gastroenterology, Hepatopancreatology and Digestive Oncology
Erasme Hospital

808 Route de Lennik
1070 Brussels
Belgium

Fax: +32–2-5554697

Email: emmanuel.toussaint@bordet.be