866 Obesity Surgery, 17, 2007 © Springer Science + Business Media, Inc.

Obesity Surgery, 17, 866-872

Background: Duodenal switch (DS) is one of the mosteffective techniques for the treatment of morbid obesi-ty and its co-morbidities, with mortality rate <1%, butwith 9.4% morbidity rates (6.5% due to leaks). In ourexperience, leaks of the staple-line after sleeve gastrec-tomy (SG) are the most frequent sites of fistula forma-tion and conservative treatment usually takes a longtime.We present our experience in the treatment of gas-tric leaks with coated self-expandable stents (CSES).

Methods: 6 patients had gastric leaks at the gastro-esophageal (GE) junction after SG or DS. One patient hada symptomatic gastro-bronchial fistula. Stents wereplaced by the interventional radiologist under fluoro-scopic control and removed endoscopically. In one case,we used an uncoated Wallstent. In two patients, percuta-neous microcoil embolization of the fistula was added.

Results: The patient treated with the Wallstentrequired a total gastrectomy 6 months after place-ment of the uncovered stent. In the other 5 patients,coated stents were successfully removed and thegastric leaks completely sealed.

Conclusions: CSES are proposed as an alternativetherapeutic option for the management of GE junctionleaks in bariatric surgery with good results in termsof morbidity and survival.

Key words: Gastric leaks, morbid obesity, bariatric sur-gery, sleeve gastrectomy, coated stents

Introduction

Open and laparoscopic bariatric operations are veryeffective for morbidly obese patients, not only in

terms of weight loss but also for the management ofassociated co-morbidities.1 Laparoscopic sleevegastrectomy (LSG) and open duodenal switch(ODS) or laparoscopic DS (LDS) are complexbariatric restrictive and malabsorptive techniquesthat, in our hands, have shown the best results interms of weight loss, quality of life and resolution ofco-morbidities with acceptable complication rates.2

Leaks after bariatric operations are usually life-threatening complications, traditionally treated withsurgery, distal enteral feeding or total parenteralnutrition (TPN).3 The use of flexible stents has beenrecently proposed as an alternative for the treatmentof enteric leaks after esophago-gastric surgery4-9

and also for leaks after bariatric surgery.4,10 The aimof the current study is to present our experience withcoated self-expandable stents (CSES) for the treat-ment of gastric leaks after ODS, LDS and LSG.

Materials and Methods

DS as represented in Figure 1A, consists on a SG withresection of 80% of the stomach at the greater curva-ture, biliopancreatic diversion with duodenal switch,with a common limb of 65 cm, alimentary limb of 235cm, and biliopancreatic limb of 300 cm.11-13 IsolatedLSG is indicated for the treatment of morbidly obesepatients with low BMI or as the first stage of the LDSin patients with BMI>6014 (Figure 1B).

From 1994 to March 2007, we have performed993 SG (519 ODS, 349 hand-sewn LDS and 125isolated LSG). We have managed one patient withGE junction leak with an uncovered stent

Treatment of Gastric Leaks with Coated Self-Expanding Stents after Sleeve Gastrectomy

Carlos Serra, MD, PhD; Aniceto Baltasar, MD; Luis Andreo, MD1; NievesPérez, MD; Rafael Bou, MD; Marcelo Bengochea, MD; Juan José Chisbert

General Surgery Service and 1Interventional Radiology, Virgen de los Lirios Hospital, Alcoy,Alicante, Spain

Correspondence to: Carlos Serra MD, PhD, Servicio de CirugíaGeneral, Hospital Virgen de los Lirios, Polígono de Caramanxels/n, 03804, Alcoy, Alicante, Spain. E-mail: serra_car@gva.es

(Wallstent) and five patients with CSES(Hanarostent®, diameter 24 mm, length 9 cm, MITEC, Seoul, Korea). Percutaneous embolization ofthe fistulas was made in two patients with 2-5 mmplatinum microcoils (Target Vascular, BostonScientific, USA).

Patient 1

A 33-year-old male with BMI 35 kg/m2, with hyper-tension, in 1987 underwent a vertical banded gastro-plasty (VBG). In January 2000, for inadequate weightloss, an ODS was performed. On the 2nd postopera-tive day, leakage at the GE junction was detected.Conservative management with naso-enteral nutritionand drainage was initially performed, but deteriora-tion of the patient due to inadequate drainage of theleak prompted us to reoperate. At surgery, a huge leakwas seen and managed with a drain near the fistula(Figure 2A). An uncovered self-expandable Wallstent

was placed at the level of the GE junction in order tocontrol the leak (Figure 2B). The stent and the pig-tailsubphrenic drain controlled the leak and the patientwas discharged (Figure 2C). Six months later, the leakrecurred due to mucosal hypertrophy at the level of thestent, and finally, 7 months after the DS, a total gas-trectomy was indicated (Figure 2D), Reconstructionwas performed with an antecolic Roux-en-Y, and thefull intestinal anatomy was restored. Currently, BMI is22 and % excess BMI loss (%EBL) is 138%.

Patient 2

A 44-year-old woman with BMI 51 who had under-gone a Roux-en-Y gastric bypass (RYGBP) 13years previously with poor weight loss, was con-verted at our institution to an ODS. RYGBP conver-sion to a ODS is a very complex operation withseven sites of possible leaks (Figure 3A). She had aleak at the gastro-gastric anastomosis of the SG(Figure 3B) and was treated with the placement oftwo sequential 9-cm CSES (Figure 3C). After 2months, the stents were endoscopically removed(Figure 3D) and the leak sealed (Figure 3E).

Patient 3

A 46-year-old woman with BMI 41 underwent aLSG. One week after the operation, a gastric leak atthe angle of His was detected and managed conserv-atively with naso-enteral nutrition and drainage. Onemonth later, the patient had a severe stenosis of thegastric sleeve that produced a gastro-bronchial fistu-la (Figure 4A). The stricture of the stomach wastreated with balloon dilatation, and two 9-cm CSES

Treatment of Leaks with Coated Self-Expanding Stents

Obesity Surgery, 17, 2007 867

Figure 1. A. Duodenal switch. B. Sleeve gastrectomy.

A B

Figure 2. A. Massive leak at the GE junction. B. Wallstent placed to control the fistula. C. Obstruction of the stentedstomach due to mucosal hypertrophy. D. Gastrografin® image after total gastrectomy.

A C DB

RemovedStomach

B.P.L.BiliopancreaticLoop

C.C.CommonChannel

D.L.DigestiveLoop

were placed at the GE junction (Figure 4B). The gas-tro-bronchial fistula was effectively managed withpercutaneous microcoil embolization. Three monthslater, the stents were endoscopically removed after afluoroscopic control. One week later, the gastro-bronchial fistula was patent again (Figure 4C), and apair of 9-cm CSES were placed. Two months later,the stents were removed (Figure 4D) and the leak

had successfully closed (Figure 4E).

Patient 4

A 45-year-old male with BMI 41 underwent a LSG.On the 1st postoperative day, a gastric leak at the GEjunction was noted (Figure 5A) and was initiallytreated with distal enteral nutrition and a drain. One

Serra et al

868 Obesity Surgery, 17, 2007

Figure 4. A. Stenosis of the gastric sleeve and fistula at GE junction. B. CSES placed at the gastric sleeve and micro-coils embolization. C. Persistent leak at GE junction. D. Stents after removal. E. Gastrografin® image after stent removal.

A C D EB

Figure 3. A. Schematic conversion of gastric bypass to DS. B. Leak at the gastro-gastric anastomosis. C. CSES placedat the gastric sleeve. D. Stents after removal. E. Gastrografin® image after stent removal.

A

C

D

E

B

month later, the leak persisted and a CSES wasplaced by the interventional radiology personnel(Figure 5B). Two weeks later, the leak was stillpatent and a second CSES was placed distally. Thedistal stent migrated upwards and extended to theproximal CSES but did not affect the control of thefistula (Figure 5C). Two months later, the stents wereendoscopically removed, and fluoroscopic controlshowed successful closure of the leak (Figure 5D).

Patient 5

A 35-year-old woman with BMI 48 underwent aODS and had a gastric leak at the GE junction (Figure6A). Conservative management was initiated withdrainage and naso-enteral nutrition. Three monthsafter the operation, she had a persistent leak andempyema. An anti-reflux valve CSES was placed andpercutaneous microcoil embolization of the gastro-cutaneous fistula was performed. Inadequate controlof the leak was mainly due to the effect of the antire-

flux valve (Figure 6B) which maintained the leak.The stent was removed and was replaced by a newCSES (Figure 6C). She rapidly progressed to an oraldiet, and 2 months later the stent was removed after afluoroscopic study. Two days later, the leak waspatent on fluoroscopy and another SECS was placed.One month later, the CSES was removed endoscopi-cally and the leak had sealed (Figure 6D).

Patient 6

A 38-year-old female with BMI 50 underwent anuneventful LSG. Three days later, she had fever anda leak at the GE junction (Figure 7A). One day later,she was reoperated to drain a subphrenic abscess.The patient remained septic, and a CECS was placed(Figure 7B). Four weeks later, she had fever and sep-sis. The CSES was removed and an open Roux-en-Yanastomosis to the GE junction fistula was done.Figure 7C shows a patent and functioning Roux-en-Y loop and a non-functioning gastric sleeve 1 monthafter the operation.

Treatment of Leaks with Coated Self-Expanding Stents

Obesity Surgery, 17, 2007 869

Figure 5. A. Leak at GE junction. B. CSES placed at the gastric sleeve. C. Proximal migration of the distal stent.D. Gastrografin® image after stent removal.

A C DB

A C DB

Figure 6. A. Leak at the GE junction. B. Antireflux CSES placed at the gastric sleeve with persistent leak. Microcoilsembolization C. CSES placed at the gastric sleeve. D. Gastrografin® image after the stent removal.

Results

We have thus used an uncovered Wallstent and fiveCSES in six patients to treat leaks after LSG and ODS(Table 1). In all cases, the point of leakage was at theGE junction of the LSG and one at the gastro-gastricanatomosis in a conversion from RYGBP to DS. Ofthe six patients, two had previous bariatric operations.One had had a VBG and the other a RYGBP. Threehad undergone a LSG and one had an ODS.

All patients had good tolerance to the stents, withmild and transient nausea or retrosternal discomfort.The uncoated Wallstent controlled the leak initiallyand stabilized the patient, but stenosis due to mucos-al hypertrophy made a total gastrectomy necessary.CSES contributed to the leak resolution in four of thesix cases. The CSES failed in patient 6. In onepatient with an anti-reflux valve CSES, the stent hadto be removed due to persistent leakage and wasreplaced by a regular CSES. The duration for leak

resolution was 4-7 months. During that period, allbut one patient were discharged, eating regular foodat home and asymptomatic.

Discussion

Complication rates after primary bariatric operationsare today within acceptable limits, although in thehands of the most experienced surgeons, both mor-bidity and mortality are inevitable.15,16 The high inci-dence of complications (18-50%) and mortality (2.5-12.5%) associated with conversional or revisionalbariatric operations are well known.17-20 Conversionsare always a challenge for the bariatric surgeon,interventions are technically difficult, blood supplyfor the gastric sleeve may be altered, and reconstruc-tion implies more anastomoses and consequentlyhigher risk for developing serious complications andmortality. Two of our patients had previous bariatricoperations – one a VBG and the other a RYGBP.

LDS is our standard operation today. We use lin-ear staplers to divide stomach and bowel, and allanastomosis are always hand-made. As can be seenin Figure 1, in our technique there are several possi-ble sites for developing leaks: the gastric sleeve, theduodenal stump, the duodeno-ileal anastomosis, andthe Roux-anastomosis. In our experience, the GEjunction has been the most frequent origin of leaks,followed by the duodeno-ileal anastomosis. Toreduce the incidence of leaks, we cover all staple-lines with running sutures and the duodeno-ilealanastomosis is performed in two running layers.

Serra et al

870 Obesity Surgery, 17, 2007

Figure 7. A. Leak at the GE junction. B. CSES placed at the gastric sleeve. C. Functioning Roux-en-Y loop and naso-gastric tube into the gastric sleeve.

A B C

Table 1. Patient summary

Patient Surgery Leak Stent LeakNo. site duration resolution

(months)

1 VBG ; DS GEJ 7 No2 GBP ; DS Gastro-gastric 4 Yes3 LSG GEJ 7 Yes4 LSG GEJ 4 Yes5 ODS GEJ 7 Yes6 LDS GEJ 1 No

GEJ = gastroesophageal junction

Thus, our incidence of leaks has been reduced sig-nificantly. Hess has reported his experience in thesame direction.21 We routinely test the incidence ofpossible leaks intraoperatively with methylene blue,the 1st postoperative day with upper GIGastrografin® studies, and afterwards with oralmethylene blue daily until the drains are removed.

Anastomotic leakage and formation of fistulasrepresent one of the most dangerous complicationsafter bariatric surgery that can lead to life-threaten-ing situations and death.22-27 If they are found dur-ing the early postoperative course, reoperation andsurgical management with sutures, and eventualresection and reconstruction of an anastomosis, maybe effective. However, in many cases, inflammatorychanges are present in the surrounding tissues andleakage may recur. Operative treatment is the main-stay for patients with signs of sepsis and hemody-namic instability.20,28-31

On the other hand, stable patients with minimal clin-ical signs or patients who develop fistulas during thelate postoperative course, may be managed conserva-tively with proper drainage, TPN or distal enteral feed-ing, and antibiotics. Enteral feeding must be initiatedearly during the treatment, because adequate nutritionis important for leak closure. Others have suggestedthe use of fibrin sealant applied endoscopically to closethe fistula10,32 and also the use of percutaneous micro-coil embolization via the fistula tract.33 Nevertheless,not all fistulas will close with this type of therapy.

The placement of CSES can temporarily bypassthe site of leakage and allow the patients to maintainenteral nutrition until complete closure of the leak.Any intra-abdominal fluid collections or abscessesmust be properly drained, besides placement of thestent. We prefer fluoroscopy for the placement ofstents and we only use endoscopy if there are anyproblems in positioning. Stents must be tested withoral Gastrografin® after the implantation and alsoimmediately after removal. Stent withdrawal is theonly way to know if the fistula will persist. Almostall authors have reported that the optimal time forremoval of the stent is around 6-8 weeks.4,10,20 Wehave maintained stents longer, and, in some cases,leaks have reappeared. Then, another stent has beenplaced and removed 6-8 weeks later. Thus, we thinkthat stents must be in place for 6-8 weeks, thenremoved, and if the fistula persists, another stentshould be placed until final closure.

Tolerance of the stent is variable. Nausea, hyper-sialysis, early satiety and mild retrosternal discomfortare the most common symptoms, and usually disap-pear in the first few days.4 Patients can be dischargedfrom hospital tolerating soft food and normal diets.

Removal of the stent is not always easy. The pros-theses are silicone-coated, but intense adherence ofthe stent is frequent, and mucosal laceration andbleeding are frequent after removal. Migration isanother concern with the use of CSES. The incidenceof migration has been reported to be as high as one-third of cases.4 In our experience, one stent migratedupwards and became included into the proximal onebut did not affect the control of the leak.

Mejía34 reported a modification of the stent byleaving the upper flare uncoated and with twopolypropylene (Prolene®) threads at 180˚ brought outand fixed to prevent migration. Mejía also recom-mends, as we do, that the removal of the stent shouldbe done by pulling from the distal end, thus invagi-nating the stent.

Stents with an anti-reflux valve, in our opinion,are not adequate for the treatment of fistulas,because the valve may increase the resistance to out-flow and produce an overflow around the stent thatfinally maintains the leak.

Modifications of stents and the delivery systemmay improve anchorage of the stent to the supportingtissues. Future developments of easy removablestents or absorbable biodegradable materials will bemore attractive to treat leakage after bariatric surgery.

Conclusions

The use of CSES in the treatment of leakage afterbariatric operations appears to be practical and reli-able. The dehiscence of an anastomosis or a staple-line in a bariatric patient may lead to septic situa-tions if optimal treatment is not initiated early. Inour opinion, “very early” implantation of stents as aminimally invasive technique avoids re-operation,complications and stabilizes the patient. Thus, prop-er oral nutrition, a mainstay in the management ofgastrointestinal leaks, can be initiated early duringthe course of treatment. Further studies are neededto find better materials and delivery systems tomake this approach even more effective.

Treatment of Leaks with Coated Self-Expanding Stents

Obesity Surgery, 17, 2007 871

Serra et al

872 Obesity Surgery, 17, 2007

References

1. Buchwald H. Consensus Conference Statement.Bariatric surgery for morbid obesity: Health implicationsfor patients, health professionals, and third-party payers.Surg Obes Relat Dis 2005; 1: 371-83.

2. Baltasar A, Bou R, Bengochea M et al. Mil operacionesbariátricas. Cir Esp 2006; 79: 349-55.

3. Baltasar A, Bou R, Miro J et al. Laparoscopic biliopan-creatic diversion with duodenal switch: technique andinitial experience. Obes Surg 2002; 12: 245-8.

4. Fukumoto R, Orlina J, McGinty J et al. Use of Polyflexstents in treatment of acute esophageal leaks afterbariatric surgery. Surg Obes Relat Dis 2006; 2: 570-2.

5. Gelbmann CM, Ratiu NL, Rath HC et al. Use of self-expandable plastic stents for the treatment of esophagealperforations and symptomatic anastomotic leaks.Endoscopy 2004; 36: 695-9.

6. Hunerbein M, Stroszczynski C, Moesta KT et al.Treatment of thoracic anastomotic leaks after eophagec-tomy with self-expanding plastic stents. Ann Surg 2004;240: 801-7.

7. Langer FB, Wenzl E, Prager G et al. Management ofpostoperative esophageal leaks with the Polyflex self-expanding covered plastic stent. Ann Thorac Surg 2005;79: 398-403.

8. Schubert D, Scheidbach H, Kuhn R et al. Endoscopictreatment of thoracic esophageal anastomotic leaks byusing silicone-covered self-expanding polyester stents.Gastrointest Endosc 2005; 61: 891-6.

9. Scileppi T, Li JJ, Iswara K et al. The use of Plyflex coat-ed esophageal stent to assist in the closure of a colonicanastomotic leak. Gastrointest Endosc 2005; 62: 643-5.

10. Salinas A, Baptista A, Santiago E et al. Self-expandablemetal stents to treat gastric leaks. Surg Obes Relat Dis2006; 2: 570-2.

11. Baltasar A, Bou R, Cipagauta LA et al. “Hybrid”bariatric surgery: biliopancreatic diversion and duodenalswitch. Preliminary Experience. Obes Surg 1995; 5:419-23.

12. Baltasar A, Del Río J, Escrivá C et al. Preliminary resultsof the duodenal switch. Obes Surg 1997; 7: 500-4.

13. Baltasar A, Bou R, Bengochea M et al. Duodenal switch:an effective therapy for morbid obesity; intermediateresults. Obes Surg 2001; 11: 54-8.

14. Baltasar A, Serra C, Pérez N et al. Laparoscopic sleevegastrectomy. A multi-purpose operation. Obes Surg2005; 15: 1124-8.

15. Livingston EH. Procedure incidence and in-hospitalcomplication rates of bariatric surgery in the UnitedStates. Am J Surg 2004; 188: 105-10.

16. Birkmeyer JD, Stukel TA, Siewers AE et al. Surgeonvolume and operative mortality in the United States. NEngl J Med 2003; 349: 2117-27.

17. Cates JA, Drenick EJ, Abedin MZ et al. Reoperative sur-gery for the morbidly obese. Arch Surg 1990; 125: 1400-4.

18. Forse RA, Deitel M, MacLean LD. The conversion of ahorizontal to vertical banded gastroplasty: a hazardousprocedure. Gastroenterol Clin North Am 1987; 16: 533-5.

19. Schwartz RW, Strodel WE, Simpson WS et al. Gastricbypass revision: lessons learned from 920 cases. Surgery1988; 104: 806-11.

20. Farahmand M, Deveney CW, Deveney KE et al.Gastrectomy for complications of bariatric procedures.Obes Surg 1996; 6: 351-5.

21. Hess DS, Hess DW. Biliopancreatic diversion with aduodenal switch. Obes Surg 1998; 8: 267-82.

22. Kriwanek S, Ott N, Ali-Abdullah S et al. Treatment ofgastro-jejunal leakage and fistulization after GastricBypass with coated self-expanding stents. Obes Surg2006; 16: 1669-74.

23. Mason E, Ito C. Gastric bypass. Ann Surg 1969; 170:329-36.

24. Mason E, Printen K, Hartford C et al. Optimizing resultsof gastric bypass. Ann Surg 1975; 182: 405-13.

25. Halverson J, Zuckerman G, Koelher R et al. Gastric bypassfor morbid obesity. Ann Surg 1981; 194: 152-60.

26. Thompson W, Amaral J, Caldwell M et al. Complicationsand weight loss in 150 consecutive gastric exclusionpatients. Am J Surg 1983; 146: 602-12.

27. Podnos Y, Jimenez J, Wilson S et al. Complications aftergastric bypass. Arch Surg 2003; 138: 957-61.

28. Baker R, Foote J, Kemmeter P et al. The science of sta-pling and leaks. Obes Surg 2004; 14: 1290-8.

29. Gonzalez R, Nelson L, Gallagher S et al. Anastomoticleaks after laparoscopic gastric bypass. Obes Surg 2004;14: 1299-307.

30. Thodiyil P, Rogula T, Mattar S et al. Managenet of com-plications after laparoscopic gastric bypass. In: InabnetW, DeMaria E, Ikramuddin S, eds. LaparoscopicBariatric Surgery. Philadelphia: Lippincott Williams &Wilkins 2005: 225-37.

31. Marshal J, Srivastava a, Gupta S et al. Roux-en Y gastricbypass leak complications. Arch Surg 2003; 138: 520-4.

32. Papavramidis ST, Eleftheriodis EE, Papavramidis TS etal. Endoscopic management of gastrocutaneous fistulaafter batriatric surgery by using a fibrin sealant.Gastrointest Endosc 2004; 59: 296-300.

33. Hunt JA, Gallagher PJ, Heintze SW et al. Percutaneousmicrocoil embolization of intraperitoneal intrahepaticand extrahepatic biliary fistulas. Aust N Z J Surg 1997;67: 424-7.

34. Mejía AF, Bolaño E, Chaux CF et al. Endoscopic treat-ment of gastrocutaneous fistula following gastric bypassfor obesity. Obes Surg 2007; 17: 544-6.

(Received March 26, 2007; accepted April 30, 2007)