Types of surgical procedures: (1) gastrojejunal anastomosis, (2) Roux-en-Y gastric bypass, (3) gastrectomy with esojejunal anastomosis. 

Contexts in source publication

Context 1

... gastric resection, or some other unidentified factors and is achieving 60%–70% excess weight loss by 3 years [1]. A dreaded complication after sleeve gastrectomy is staple line leak. The reported gastric leak rates from the sleeve staple line are 1.4%–2.5% for primary sleeve gastrectomies and 16%–20% for reoperative surgery for which a previous gastric operation has been performed [2–5]. Numerous articles have been written about the different approaches for the gastric fistula, but few exist concerning the reconstructive surgical treatment of chronic gastric leak [6–10]. The purpose of the present study was to report our experience with this approach in 8 patients. To our knowledge, in the literature, this is the largest series reported for chronic gastric fistula (GF) treated by aggressive reconstructive surgery. Eight patients (7 women and 1 man; mean age 43.62 years) with an average body mass index (BMI) of 46.2 kg/m 2 who underwent LSG complicated by persistent high-output GF were enrolled in the study. Five patients had their original LSG surgery performed at another hospital and were referred to us with persistent gastric leaks, despite intensive measures, such as antibiotherapy, nutritional support, and treatment of the abdominal sepsis by reintervention, endoscopy, or computed tomography–guided drainage, at the original institution. To this group, we added 3 patients who had their initial surgery performed in our institution. The demographic data, the initial intervention, the diagnosis time, and the initial management of the fistula are summarized in Table 1. The technique of LSG for the 3 patients operated in our institution was antrum preserving sleeve gastrectomy cali- brated with a 36F bougie. We used 5–7 gold cartridges with no additional reinforcement by oversewing or buttress materials. For the other 5 patients who had their original LSG surgery performed in another hospital, the operative records were not found. The first patient had a history of gastric banding for 5 years and underwent sleeve gastrectomy and concomitant gastric band removal for weight loss failure (BMI 1⁄4 50 kg/ m 2 ) in another center. After 3.5 months, she was diagnosed with GF treated by peritoneal lavage and jejunostomy. The management of the GF was completed by a 28-mm endoscopic prosthesis-type Ultraflex (UltraflexTM Esophageal NG Stent System, Boston Scientific Corporation, Natick, MA) 10 days after. The removal of the prosthesis after 4 months was impossible because of the intense process of fibrosis. The endoscopic decision was to place in a second plastic prosthesis type Rusch (Willy Rusch GmbH, Teleflex Medical Company, Research Triangle Park, NC) for 15 days and to realize a complementary fulguration of the granulomatous tissue. After several failed attempts to remove the endoscopic prosthesis, another fistula was diagnosed in the lower part of the gastric tube. Facing multiple GFs and the impossibility of removing the prosthesis, we decided to perform a total gastrectomy with a Roux-en-Y esophagojejunal anastomosis. The second patient had an LSG in our institution complicated on postoperative day (POD) 17 by consecutive episodes of severe, intractable vomiting. The Gastrografin swallow and the upper endoscopy revealed the persistence of a residual gastric fundus resulting from incomplete dissection (Fig. 1). After 4 weeks of conservative treatment, we performed a resleeve gastrectomy complicated by fistula on POD 3, which necessitated drainage of the collection by laparoscopy. The patient was included in a protocol of serial endoscopic treatments by biologic glue. Six months later, because of the persistence of the fistula with no sign of healing despite multiple endoscopic sessions of biologic glue, the decision for a surgical approach was made. The third patient had undergone band removal and concomitant sleeve gastrectomy for weight loss failure in a different center (BMI 1⁄4 42 kg/m 2 ) after an 8-year history of inefficient gastric banding. The postoperative course was complicated 3 days later by a fistula, drained initially by laparotomy and secondarily by radiology on POD 8. After 15 days, he was transferred to our unit because of uncontrollable sepsis. Another laparotomy for drainage was performed with simultaneous feeding jejunostomy. One month afterward, an endoscopic prosthesis was inserted, which was complicated by the migration of the prosthesis. The patient moved to another region, and during the next 2.5 years, he had more than 40 endoscopic sessions and 6 prosthesis insertions, 2 of them complicated by migration. Because of the persistence of a high-output fistula and the inefficiency of the endoscopic treatment, a surgical approach was chosen. In all cases, the fistula was diagnosed by abdominal computed tomography with oral contrast Fig. 1. There are 3 possible surgical procedures: gastrojejunal lateral anastomosis, Roux-en-Y gastric bypass, and gastrectomy with esojejunal anastomosis (Fig. 2). For types I and II, the anastomosis was performed on the fistulous orifice. The choice of the specific surgical procedure was based on the following variables: intraoperative findings, presence of perigastric abscess or multiple fistulas, and the anatomy of the gastric tube and fistula. After the initial management of the fistula, with good control of the sepsis by drainage and antibiotherapy, the endoscopic therapy was attempted for all 8 patients. Two patients had an initial treatment by endoscopic clips at 2 and 7 months, with recurrence of the leakage, followed by endoscopic stents. The first patient presented with an intraabdominal collection with the prosthesis in place; therefore, the stent was removed. The second patient presented with a migration of the prosthesis; the second prosthesis was changed after 3 months. Two other patients had the endoscopic stent as the first choice treatment (Fig. 3): in the first case at the same time with the peritoneal lavage and in the second case 10 days postoperatively. In the first case, the prosthesis was impossible to remove after 4 months, and a second fistulous orifice was diagnosed at the lower part of the stent. A second plastic stent was placed inside of the first one. Both stents were removed 7 months after the diagnosis of the fistula during the reconstructive surgery by gastrectomy with esophagojejunal anastomosis. In the second case, the stent was changed after 3 months. After the second stent’s removal, because of the persistent fistula, it was decided to use sequential endoscopic sessions with sealing glue. After 14 months, the output of the fistula was constantly high and reconstructive surgery was performed. In 2 other patients the fibrin sealant Beriplast (Beriplast, CSL Behring GmbH, Marburg, Lahn) was used as a tissue adhesive. In both patients, we had signs that the defect continued to exist, so multiple applications were performed. For both patients, the output of the fistula remained 5 months after diagnosis, and the decision for reconstructive surgery was made. The last 2 patients had a sequential treatment with glue and prosthesis. The overall average diagnosis time of the fistula was 14.4 months (range 5–44 months). The surgical procedures (Fig. 2) performed for chronic fistula after sleeve gastrectomy were (1) gastrojejunal lateral anastomosis in 4 cases, (2) Roux-en-Y gastric bypass in 2 cases, (3) and gastrectomy with esojejunal anastomosis in 2 cases. The intraoperative time for all 8 patients is illustrated in Table 2. The mean operative time was 193 minutes (range 164–238 minutes). All procedures were performed by laparotomy. For the last patient, the laparoscopic approach was attempted, but conversion was necessary because of the intense adhesions between the posterior part of the gastric tube and the anterior part of the pancreas. The postoperative fistula was recorded for 4 patients: 3 patients had ‘‘early’’ leaks, and only 1 patient had a ‘‘late’’ leak 4 1 week after surgery. The mean healing time of the fistula was 32 days (range 22–63 days). None of these patients needed another surgical procedure for drainage, 3 patients had an ‘‘early’’ leak with the drain already in place, and the fourth patient had radiological drainage. Two patients required admittance to the intensive care unit for respiratory problems. No mortality was recorded. LSG has increasingly gained worldwide acceptance among bariatric surgeons during the past 7 years. Initially, LSG was accepted as a first-stage procedure in high-risk or super-obese patients, but the popularity of the procedure increased as it started to be used as a single-stage procedure. Despite the above, the procedure has been associated with specific and life-threatening complications, the most feared being GF. A new problem occurs for high-activity bariatric centers — the management of the chronic gastric leak. The literature does not offer a clear definition for the chronic gastric leak or its treatment. In our experience, we start considering chronic gastric leak after 6 weeks of diagnosis with remission of major inflammatory signs with constant output. We propose reconstructive surgical treatment when all endoscopic approaches have failed to close the fistula. When faced with a GF, the surgeon must consider all options to confront it adequately. Upon the establishment of the diagnosis of a GF, blood and electrolyte imbalance restoration, alimentary tract resting, optimal nutrition launch- ing, and sepsis management must be an absolute priority. Then we should proceed to the endoscopic exploration of the gastric area to assess for different methods of endoscopic treatment. The use of coated self-expandable stents in the treatment of leakage after bariatric operations appears to be practical and reliable, as was shown in several studies [11– 19], but this is often complicated by migration of the prosthesis, or if not, it has to be removed or changed 2–3 months after its ...

Context 2

... management of the GF was completed by a 28-mm endoscopic prosthesis-type Ultraflex (UltraflexTM Esophageal NG Stent System, Boston Scientific Corporation, Natick, MA) 10 days after. The removal of the prosthesis after 4 months was impossible because of the intense process of fibrosis. The endoscopic decision was to place in a second plastic prosthesis type Rusch (Willy Rusch GmbH, Teleflex Medical Company, Research Triangle Park, NC) for 15 days and to realize a complementary fulguration of the granulomatous tissue. After several failed attempts to remove the endoscopic prosthesis, another fistula was diagnosed in the lower part of the gastric tube. Facing multiple GFs and the impossibility of removing the prosthesis, we decided to perform a total gastrectomy with a Roux-en-Y esophagojejunal anastomosis. The second patient had an LSG in our institution complicated on postoperative day (POD) 17 by consecutive episodes of severe, intractable vomiting. The Gastrografin swallow and the upper endoscopy revealed the persistence of a residual gastric fundus resulting from incomplete dissection (Fig. 1). After 4 weeks of conservative treatment, we performed a resleeve gastrectomy complicated by fistula on POD 3, which necessitated drainage of the collection by laparoscopy. The patient was included in a protocol of serial endoscopic treatments by biologic glue. Six months later, because of the persistence of the fistula with no sign of healing despite multiple endoscopic sessions of biologic glue, the decision for a surgical approach was made. The third patient had undergone band removal and concomitant sleeve gastrectomy for weight loss failure in a different center (BMI 1⁄4 42 kg/m 2 ) after an 8-year history of inefficient gastric banding. The postoperative course was complicated 3 days later by a fistula, drained initially by laparotomy and secondarily by radiology on POD 8. After 15 days, he was transferred to our unit because of uncontrollable sepsis. Another laparotomy for drainage was performed with simultaneous feeding jejunostomy. One month afterward, an endoscopic prosthesis was inserted, which was complicated by the migration of the prosthesis. The patient moved to another region, and during the next 2.5 years, he had more than 40 endoscopic sessions and 6 prosthesis insertions, 2 of them complicated by migration. Because of the persistence of a high-output fistula and the inefficiency of the endoscopic treatment, a surgical approach was chosen. In all cases, the fistula was diagnosed by abdominal computed tomography with oral contrast Fig. 1. There are 3 possible surgical procedures: gastrojejunal lateral anastomosis, Roux-en-Y gastric bypass, and gastrectomy with esojejunal anastomosis (Fig. 2). For types I and II, the anastomosis was performed on the fistulous orifice. The choice of the specific surgical procedure was based on the following variables: intraoperative findings, presence of perigastric abscess or multiple fistulas, and the anatomy of the gastric tube and fistula. After the initial management of the fistula, with good control of the sepsis by drainage and antibiotherapy, the endoscopic therapy was attempted for all 8 patients. Two patients had an initial treatment by endoscopic clips at 2 and 7 months, with recurrence of the leakage, followed by endoscopic stents. The first patient presented with an intraabdominal collection with the prosthesis in place; therefore, the stent was removed. The second patient presented with a migration of the prosthesis; the second prosthesis was changed after 3 months. Two other patients had the endoscopic stent as the first choice treatment (Fig. 3): in the first case at the same time with the peritoneal lavage and in the second case 10 days postoperatively. In the first case, the prosthesis was impossible to remove after 4 months, and a second fistulous orifice was diagnosed at the lower part of the stent. A second plastic stent was placed inside of the first one. Both stents were removed 7 months after the diagnosis of the fistula during the reconstructive surgery by gastrectomy with esophagojejunal anastomosis. In the second case, the stent was changed after 3 months. After the second stent’s removal, because of the persistent fistula, it was decided to use sequential endoscopic sessions with sealing glue. After 14 months, the output of the fistula was constantly high and reconstructive surgery was performed. In 2 other patients the fibrin sealant Beriplast (Beriplast, CSL Behring GmbH, Marburg, Lahn) was used as a tissue adhesive. In both patients, we had signs that the defect continued to exist, so multiple applications were performed. For both patients, the output of the fistula remained 5 months after diagnosis, and the decision for reconstructive surgery was made. The last 2 patients had a sequential treatment with glue and prosthesis. The overall average diagnosis time of the fistula was 14.4 months (range 5–44 months). The surgical procedures (Fig. 2) performed for chronic fistula after sleeve gastrectomy were (1) gastrojejunal lateral anastomosis in 4 cases, (2) Roux-en-Y gastric bypass in 2 cases, (3) and gastrectomy with esojejunal anastomosis in 2 cases. The intraoperative time for all 8 patients is illustrated in Table 2. The mean operative time was 193 minutes (range 164–238 minutes). All procedures were performed by laparotomy. For the last patient, the laparoscopic approach was attempted, but conversion was necessary because of the intense adhesions between the posterior part of the gastric tube and the anterior part of the pancreas. The postoperative fistula was recorded for 4 patients: 3 patients had ‘‘early’’ leaks, and only 1 patient had a ‘‘late’’ leak 4 1 week after surgery. The mean healing time of the fistula was 32 days (range 22–63 days). None of these patients needed another surgical procedure for drainage, 3 patients had an ‘‘early’’ leak with the drain already in place, and the fourth patient had radiological drainage. Two patients required admittance to the intensive care unit for respiratory problems. No mortality was recorded. LSG has increasingly gained worldwide acceptance among bariatric surgeons during the past 7 years. Initially, LSG was accepted as a first-stage procedure in high-risk or super-obese patients, but the popularity of the procedure increased as it started to be used as a single-stage procedure. Despite the above, the procedure has been associated with specific and life-threatening complications, the most feared being GF. A new problem occurs for high-activity bariatric centers — the management of the chronic gastric leak. The literature does not offer a clear definition for the chronic gastric leak or its treatment. In our experience, we start considering chronic gastric leak after 6 weeks of diagnosis with remission of major inflammatory signs with constant output. We propose reconstructive surgical treatment when all endoscopic approaches have failed to close the fistula. When faced with a GF, the surgeon must consider all options to confront it adequately. Upon the establishment of the diagnosis of a GF, blood and electrolyte imbalance restoration, alimentary tract resting, optimal nutrition launch- ing, and sepsis management must be an absolute priority. Then we should proceed to the endoscopic exploration of the gastric area to assess for different methods of endoscopic treatment. The use of coated self-expandable stents in the treatment of leakage after bariatric operations appears to be practical and reliable, as was shown in several studies [11– 19], but this is often complicated by migration of the prosthesis, or if not, it has to be removed or changed 2–3 months after its placement. Also endoscopic sealing glue should be considered as an option of treatment as proven in several studies. Alternatively, the applications of endoscopic clips are suggested. Endoscopic clip application is considered successful when no leakage occurs for a minimum of 3 days. After initial successful management of the sepsis, patients with gastric leak are referred to the endoscopic unit. In this way, 40 of 54 (74.04%) patients with fistula as a complication after sleeve gastrectomy have healed by different endoscopic approaches, including endoprothese, endoscopic clips, endoscopic sealing glue, or balloon dilation. These patients were included in a multicenter study. Calculating an exact rate for each of these approaches was impossible because of the use of multiple approaches for each patient. In our experience, the successful use of the endoprothese is very limited when the leak diagnosis is done 4 1 month postoperatively. This explains the nonuse of stents in some of the patients. Of 54 patients, 6 achieved the healing of the leak by medical treatment with no therapeutic endoscopic procedure. The other 8 patients, with persistent leak despite aggressive endoscopic approach, were referred for surgical treatment, and they represent the subject of this report. In patients in whom all the aforementioned conservative techniques fail, aggressive management with surgical reconstructive procedures may be performed. Only a few reports on the surgical treatment of chronic gastric leak were found [6–10]. The first case in the literature was described by Baltasar et al. in 2007 [7]. They presented the case of a 48-year-old woman who developed a GF at POD 3 after LSG surgery and was treated by conservative measures. Six weeks after the original surgery, a Roux-en-Y limb was brought to the esogastric junction and anastomosed side-to-end to the fistula. At the beginning, the Roux limb was the only functioning outlet, and finally, 2 months later, both path- ways (the gastric sleeve and the Roux-en-Y) were patent. No fistula occurred in that patient. The same team reported 1 year previously, in another article [10], a series of 9 cases of total gastrectomies out of 846 patients who underwent the duodenal switch for different ...