Langenbeck's Archives of Surgery

, 396:981

Gastric leakage after sleeve gastrectomy—clinical presentation and therapeutic options


    • Klinik und Poliklinik für Allgemein-, Viszeral-, Gefäß- und Kinderchirurgie
  • Andreas Thalheimer
    • Klinik und Poliklinik für Allgemein-, Viszeral-, Gefäß- und Kinderchirurgie
  • Florian Seyfried
    • Klinik und Poliklinik für Allgemein-, Viszeral-, Gefäß- und Kinderchirurgie
  • Martin Fein
    • Klinik und Poliklinik für Allgemein-, Viszeral-, Gefäß- und Kinderchirurgie
  • Gwendolyn Bender
    • Medizinische Klinik, Abteilung für Endokrinologie
  • Christoph-Thomas Germer
    • Klinik und Poliklinik für Allgemein-, Viszeral-, Gefäß- und Kinderchirurgie
  • Christian Wichelmann
    • Klinik und Poliklinik für Allgemein-, Viszeral-, Gefäß- und Kinderchirurgie
Original Article

DOI: 10.1007/s00423-011-0800-0

Cite this article as:
Jurowich, C., Thalheimer, A., Seyfried, F. et al. Langenbecks Arch Surg (2011) 396: 981. doi:10.1007/s00423-011-0800-0



To analyze gastric leakage following sleeve gastrectomy depending on its point of detection and localization in order to evaluate therapeutic strategies.


From Dec 2006 until June 2010, data of all patients undergoing bariatric surgery were entered into a prospectively documented database. Evaluation contained patient′s gender, age, body mass index (BMI), type of surgery, clinical symptoms, diagnostics, onset and localization of leakage, type of therapy, length of stay (LOS), and clinical outcome.


Forty-five of 196 bariatric patients underwent sleeve gastrectomy, 22 male and 23 female with mean age 43 ± 9.7 years and mean BMI 54.9 ± 10 kg/m2. Four patients developed a gastric leak (8.9%)—three proximal leaks and one distal leak. Leakage was detected by upper gastrointestinal (UGI) radiography in two cases, by gastroscopy in one case, and by abdominal computed tomographic (CT) scan in another case. In two cases, CT scan was not feasible because of patient′s conditions. Three patients underwent relaparoscopy with re-suture of staple line, abdominal lavage, and placement of an intraabdominal drain. Both patients with proximal leaks required stent graft application as leakage reoccurred within 5 days after relaparoscopy. LOS varied between 30 and 120 days. None of the patients died.


The location of leakage, and the presence or absence of an intraabdominal drain are determining factors for its treatment. UGI radiography with contrast media and gastroscopy are comparable and superior to standard CT scan. Stent graft application is a promising therapy in case of proximal leakage; re-suture or resection of the staple line are possible solutions in case of a distal leak.


Sleeve gastrectomyLeakage


Sleeve gastrectomy has become an important therapeutic option in the treatment of morbid obesity during recent years. Being a relative “new” procedure—first described in 2003 by Gagner et al [1]—there are striking influences on weight loss and associated comorbidities [2, 3]. Initial findings after sleeve gastrectomy are comparable to those following RnYGB [46]; therefore sleeve gastrectomy may be thought to be far more than just a “restrictive” bariatric procedure as supposed primarily. Sleeve gastrectomy can be performed laparoscopically and is also feasible in patients suffering from “super obesity” or high-risk patients [1, 79].

Although being initially performed as a typical “conditioning-procedure” or as the first operative step of a two-step-procedure [1, 7], sleeve gastrectomy is carried out more and more as a single operation [6, 10]. It is gaining importance, for example, in those cases in which a dumping syndrome has to be strictly avoided or when an extensive long-term medication has to remain unaffected (level of active pharmaceutical ingredient).

In spite of its encouraging and convincing results, the pretended feasibility of this operative procedure can be associated with a remarkable operative morbidity [11]. The main reason is the leakage of the long staple line.

The aim of this study was to analyze gastric leakage following sleeve gastrectomy depending on its timepoint of detection and on its localization in order to evaluate therapeutic strategies.

Patients and methods

From December 2006 till June 2010, 196 patients underwent bariatric surgery for morbid obesity in our department. Data were entered into a prospectively documented database. Sleeve gastrectomy was performed in 45 patients; in 40 of these cases, operation was carried out laparoscopically. Conversion of a laparoscopic procedure into open surgery occurred in four cases; one patient underwent primary open surgery. Among the 45 patients, 22 were male and 23 female; mean age was 43 ± 9.7 years and mean body mass index (BMI) 54.9 ± 10 kg/m2.

Operative procedure

Laparoscopic sleeve gastrectomy (LSG) was performed using a classical five-trocar-technique: a 5-mm subxiphoid trocar for the liver retractor; one 12-mm right upper quadrant trocar and two 12-mm left upper quadrant trocars served as working channels, and one of the left trocars was used to remove the specimen; a 12-mm supraumbilical trocar serves as optic system. LSG is performed by dividing the greater curvature vessels using an Ultracision device (Ethicon, Johnson&Johnson and Medical GmbH), beginning 5 cm proximal to the pylorus close to the Angle of His, cutting the short gastric and posterior fundic vessels. Then, a 36F bougie is introduced into the stomach and further into the pyloric channel. An ENDO GIA stapler with 3.8–4.1-mm staples (golden/green cartridge) is introduced through one of the 12-mm ports for the resection which begins 5 cm proximal to the pylorus. Gastric tubulization is completed by dividing the gastric corpus straight to the Angle of His applying three to four golden cartridges of the ENDO GIA stapler. Reinforcement with absorbable sutures (Vicryl 2–0 or 3–0) is performed over the mechanical closure only the lower third of the staple line followed by administration of methylene blue. Routine staple line reinforcement (Peri-stripes Dry® Synovis) was performed from January 2009. The resected specimen is removed through the port site at the left upper abdominal quadrant. Routine volumetry of the specimen is performed to assure adequate extent of resection. No leaks were observed during surgery in any patient.

Clinical evaluation

Clinical evaluation for pain and discomfort, fever, tachycardia, and tachypnea were recorded for all patients three times daily. Laboratory testing was performed routinely on the second postoperative day and in case of suspicious clinical symptoms.

Diagnosis of leakage

In case of clinically suspected suture leakage, upper gastrointestinal (UGI) radiography with liquid contrast media was performed, although it is inferior to barium sulfate [12]. Computed tomographic (CT) scan was performed when feasible in a standard multi-slice technique with liquid contrast media. Additionally, gastroscopy was indicated in those cases where endoluminal stent graft application was aspired.

Definition of leakage

According to the UK Surgical Infection Study Group, a gastric leak was defined as “the leak of luminal contents from a surgical joint between two hollow viscera”. In the current series, gastrointestinal leak in the staple line led to distribution of luminal content around the organ. Determining the time of appearance, leaks were classified according to Csendes et al. [13] in leaks being detected 1 to 3 days after surgery (postoperative days (POD) −3), those being detected 4 to 7 days after surgery (POD 4–7), and those appearing more than 7 days after surgery (POD ≥ 8).


Four of 45 patients who underwent laparoscopic sleeve gastrectomy for morbid obesity developed a gastric leak (8.9%) within the first 30 postoperative days. Data of these patients are presented in Tables 1 and 2. None of the patients presented with abdominal pain. Leading symptoms of patients with gastric leak were tachycardia, increased WBC, and elevated C-reactive protein levels. Concerning the clinical presentation, no difference was found between the group POD −3 and POD 4–7. Two of four patients had fever (>38°C), one in group POD −3 and one in group POD 4–7. Only in one patient (group POD 4–7) did the intraabdominal drain show abnormal secretion. There were three proximal leaks near the Angle of His and one distal leak in the lower third of the staple line. None of the leakages occurred at junction areas of different staple lines. The area of leakage in the presented cases was additionally secured by oversewing only in the lower third of the staple line. Leakage was detected by UGI radiography with liquid contrast media in two cases, in one case by gastroscopy and in one case by abdominal CT scan with liquid contrast media. In two cases, CT scan was not feasible because of patient′s too large waist circumference or absolute weight. Three patients underwent relaparoscopy. In all of those sutures of staple line, invagination of the leakage area was realized in addition to abdominal lavage and placement of an intraabdominal drain. In both patients with proximal staple line leakage, it reoccurred within 5 days after relaparoscopy. All patients with proximal leaks required stent graft application (Niti-S Esophageal Covered Stents® by TaeWoong Medical Co. Ltd., Corea) in the course of the treatment (Fig. 1). Two patients developed further complications as septicemia, multiple organ failure, or intraabdominal abscess. The length of stay varied between 30 and 120 days. None of the patients died.
Table 1

Patients suffering from staple line leakage—overview



Suture line reinforcement

Initial drain

Clinical presentationi


Localization of leckage




Further complication


Lap.-conv. sleeve

Yes (Peri-stripes dry®)


Tachycardia WBC ↑ CRP ↑



Gastroscopy, UGI radiography

Stent, jejunal feeding tube



Lap. sleeve

Yes (Peri-stripes dry®)


Tachycardia WBC ↑ CRP ↑



Gastroscopy, CT laparoscopy

Relaparoscopy/suture, gastroscopy and stent, jejunal feeding tube


Gastrocutane fistula, renal failure, septicemia and MOF


Lap. sleeve



Tachycardia WBC ↑ CRP↑



UGI radiography



Perisplenic abscess (CT drainage)


Lap. sleeve



Tachycardia WBC ↑ CRP ↑




Relaparoscopy/suture/drainage jejunal feeding tube, gastroscopy and stent


Legend: MOF multiple organ failure, POD postoperative day, LOS length of stay, suture line reinforcement Peri-stripes Dry® Synovis

Table 2

Patient characteristics and comorbidities














Chronic renal failure









Sleeping apnea










Reflux surgery and gastric banding in history


Leakage after sleeve gastrectomy occurs in about 0.5–5.3% of the cases [4, 9, 1418]; therefore its incidence is comparable to the leakage rate after RnYGB [19]. The presence of leakage determines a serious and life-threatening complication of this bariatric procedure. The relative high rate of gastric leakage among patients included in this study might be due to the learning curve, but no technical mistakes were detected. To our opinion, complication and especially leakage rate after sleeve gastrectomy is underestimated. A detailed review by the ASMBS (American Society for Bariatric and Metabolic Surgery) showed the overall complication rate after sleeve gastrectomy to range between 0% and 24% with an overall mortality rate of 0.39% [20]. The incidence of leakage rates ranges between 0% and 20% [2123]. Stroh et al. [24] observed a leakage rate of 7% among 144 patients in which a laparoscopic sleeve gastrectomy was performed in a large German multicenter observational study.

Table 3 shows the latest studies dealing with staple line leakage after laparoscopic sleeve gastrectomy for morbid obesity, its clinical appearance, and parameters and gives a short overview of the proceeded management of this operative complication [11, 21, 22, 2426].
Table 3

Leakage after sleeve gastrectomy in literature 2009/2010: incidence, clinical parameters, and therapeutic management


n (SG)

n (leak)/rate (%)

Mortality rate (%)

Time of onset

Localization GEJ/upper third

Therapeutic strategy

Casella [22]/2009




3 early, 3 late


6/6: TPN, PPI, AB

5/6: Percut. drain

3/6: Stentgraft

Frezza [11]/2009




Reoperation, drainage

Burgos [21]/2009




2 early, 2 interm., 3 late


4/7: Reoperation (lavage, drainage, re-suture in 2 cases with 50% failure)

3/7: TPN, drainage, percutaneous CT intervention

Stroh [24]/2009




4/10: Relaparotomy

Ser [25]/2010





3/4: Reoperation (fistula exclusion (2), abscess drainage/jejunostomy (1)

Lacy [26]/2010





10/11: Relaparoscopy (lavage, drainage (+omental flap in 2 cases))

1/11: Conservative

6/11: Stent graft (6/6 leak persistence)

11/11: TPN + AB

Own data/2011




2 early, 2 interm.


3/4: Relaparoscopy + re-suture (stentgraft 2/3)

3/4: Stent graft

3/4: Jejunal feeding tube

SG sleeve gastrectomy; early <3 postoperative day (POD), intermediate 3–7 POD, late >8 POD, GEJ gastroesophageal junction, TPN total parenteral nutrition, PPI proton-pump inhibitors, AB broad-spectrum antibiotics

Corresponding to Chen et al. [27] and Kasalicky et al. [28], usage of routine staple line reinforcement or oversewing did not influence the occurrence of staple line leakage in our patients. In the study by Casella et al. [22] in which, routinely, a running suture of the complete stomach wall as a reinforcement of the staple line was performed in 100 patients undergoing sleeve gastrectomy, leak presented in two cases. Though oversewing prevented bleeding, it did not prevent leakage completely. Furthermore, oversewing of the proximal stomach wall means a technically demanding procedure to the surgeon combined with prolonged operation time. For these reasons, we do not perform routinely oversewing of the proximal staple line during proceeding sleeve gastrectomy for morbid obesity.

As explained by other authors so far, staple line leakage could be observed at different times [21]. No differences between gastric leak POD −3 and gastric leak POD 4–7 could be found as far as clinical parameters were concerned. As reported by others, tachycardia was the leading clinical sign [16, 29].

When leakage was clinically suspected, different diagnostic procedures led to the correct diagnosis. “Diagnostic tools” should be used by means of patient`s clinical appearance and technical practicability. In our series, only in one case did abdominal computed tomography with administration of contrast media lead to the correct diagnosis. This underlines the experience that sensitivity of CT scan declines with the degree of obesity. The abdominal girth and total weight of the patient may cause difficulties at the CT gantry. In addition, artifacts related to large body dimensions may reduce the quality of the images [30]. In contrast to Triantafyllidis et al. [31], in our opinion, CT scan in patients with BMI 50+ is usually not helpful. In the other cases, gastroscopy and UGI radiography were beneficial. We therefore recommend the diagnostic and therapeutic algorithm presented in Fig. 2.
Fig. 1

Endoluminal stent graft therapy. Proximal leak after sleeve gastrectomy (A), placement of stent (B) and control of stent closure with contrast media (C) and control of passage after stent removement (D)
Fig. 2

Diagnostic and therapeutic algorithm in case of suspected gastric leak

According to the literature [11, 14, 22], in three of our own four patients, gastric leakage was located in the proximal stomach near the Angle of His. Relaparoscopy with hand-suture of the proximal leak was performed in two cases without primary insertion of a drain but did not meet with success, although the suture appeared to be leak-proof after administration of methylene blue dye via nasogastric tube intraoperatively. This might be due to an inflammatory component around the leak and is seen by others as well [17].

According to Casella et al. [22], exclusive application of a stent graft in case of proximal gastric leakage may suffice depending on the existence of an intraabdominal drain which is also known from bypass procedures [29]. In case of an intraabdominal drain having been inserted primarily, no further measures might be necessary. From our point of view, relaparoscopy, lavage, and insertion of an intraabdominal drain are necessary in case of removal or absence of a drain. This is in accordance with reports of others [17]. Only when distal leakage is present, re-suture with application of an intraabdominal drain was successful. The insufficient staple line could be tucked in by inverting sutures.

Furthermore, enteral nutrition by alimentary fistulas or naso-jejunal tubes should be provided and secured [21].


The determining factor for the treatment of gastric leakage within the first week after laparascopic sleeve gastrectomy is the location of the leakage and the presence or absence of an intraabdominal drain. For detection of leakage, UGI radiography with contrast media and gastroscopy are comparable. Standard CT scan even with oral application of contrast media is inferior in patients with BMI 50+.

In case of proximal leakage, endoluminal stent graft application is a promising therapy in the early postoperative course, irrespective of the exact postoperative day. In case of absence of an intraabdominal drain, relaparoscopy with abdominal lavage and insertion of a drain is necessary. Therefore, we recommend primary insertion of an intraabdominal drain close to the staple line for six postoperative days on a routine base.

Re-suture or resection of the staple line may be a possible solution to the problem only in case of distal leakage. In our experience, re-suture of proximal leakages is not helpful.

The complicated multi-disciplinary management of gastric leakage by stent graft or relaparoscopy with drainage application should be provided by specialized centers.

Conflicts of interest


Copyright information

© Springer-Verlag 2011