Role of Flexible Endoscopy in the Practice of Bariatric Surgery

Source

Year

No. of patients

Mean BMI (kg/m²)

Prevalence of pathologic findings on endoscopy (%)

Madan et al. [10]

2004

102

48.2

91.0

Loewen et al. [8]

2008

448

48.6

29.2

de Moura Almeida et al. [9]

2008

162

44.1

77.2

Munoz et al. [12]

2009

626

42.0

46.0

Kuper et al. [6]

2010

47.6

79.7

Dietz et al. [11]

2012

126

51.2

57.9

BMI body mass index

Identification of anatomic defects or upper GI pathology in both asymptomatic and symptomatic patients can alter the course of treatment and also the surgical operation. Recommendations for upper endoscopy in all patients prior to bariatric surgery, regardless of the presence of symptoms, have already been suggested by guidelines outside of the United States [13]. A study conducted at the University of Virginia examined 667 patients with routine endoscopy prior to surgery; as a result, 4.6 % of patients had their operations altered by the findings from endoscopy [14]. The most common alteration in this cohort was the addition of a remnant gastrostomy tube based on preoperative endoscopic findings of the distal stomach. Another study of 447 patients demonstrated that preoperative endoscopy changed medical management prior to surgery in 18 % of patients but only altered or postponed surgery in <1 % [8]. Common findings by upper endoscopy that may impact the choice of operation or preoperative management include: esophagitis, GERD, ulcers, Helicobacter pylori, hiatal hernia, cancer, and polyps. Sharaf and colleagues retrospectively reviewed records of patients that were endoscopically evaluated prior to surgery and demonstrated that out of 195 patients, 89.7 % had one or more lesions, and of these, 61.5 % were clinically significant [15]. Biopsies for H. pylori should be obtained during endoscopy, especially in the presence of gastritis, and treated if positive. Studies have demonstrated a lower rate of marginal ulcers and foregut symptoms after bariatric surgery in patients that were either H. pylori negative or had been tested and treated [14, 16].

The utilization of preoperative flexible endoscopy is essential in the evaluation of patients prior to revisional bariatric surgery. Operative records should be reviewed before intervention, but specific anatomic constructs may not be accurately described in the report or the original operative reports may not be available. The indications for reoperation vary, and they often involve poor or failed weight loss, acute symptoms, or complications from the primary operation. This type of patient can present with complex problems and variable anatomy. Determination of the anatomy, assessment of the anastomoses and staple lines, and identification of any upper GI pathology will dictate the appropriate surgical intervention.

Flexible upper endoscopy traditionally examines the mucosa of the upper GI tract and allows biopsies for histologic diagnosis. Supplemental techniques, such as the use of a pH probe and manometry, can provide further objective measures to evaluate a patient and determine treatment. In addition, radiographic studies may enhance medical decisions when used in combination with endoscopy. An upper GI series and a barium esophagram are two such examples. In a retrospective series of patients presenting for weight regain after previous bariatric surgery, Brethauer et al. demonstrated that the use of both endoscopy and upper GI series allowed for the detection of abnormalities in 90 % of patients [17].

Intraoperative Management and Techniques

Applications of endoscopy during bariatric surgery vary depending on the operation and technique. It has been described in the literature for many intraoperative applications, including identification of anatomy and inspection of surgical technique, and is also widely practiced.

For RYGB, different techniques have been described for the creation of the gastrojejunostomy, some utilizing endoscopy. Wittgrove and Clark described the use of endoscopy to pass the anvil of an end-to-end anastomosis (EEA) stapler into the gastric pouch during an RYGB [18]. Initial concerns for perforation with transoral passage of the anvil led to modifications in this technique, including manipulating the anvil to facilitate passage and also the development of a pre-tilted anvil [19]. Endoscopy also has a role in SG, and although sizing of the gastric tube is often done by means of a bougie, Diamantis and colleagues reported their experience of using an endoscope to perform a laparoscopic SG safely and effectively [20].

Anastomotic integrity is vital for a successful outcome and is often evaluated endoscopically at the time of operation. Failure to identify anastomotic leaks imposes significant morbidity [21]. Methylene blue dye injected near the gastrojejunal anastomosis via a nasogastric tube can identify leaks intraoperatively, and it has been used successfully in clinical practice. Unfortunately, positive leak tests stain the operative field, prohibit repeat exams, and may not precisely identify the area of leakage. Intraoperative endoscopy may circumvent these pitfalls. Endoscopy allows for visualization of the GI tract, for direct placement of the endoscope near the anastomosis or staple line, and for multiple leak tests since it does not utilize methylene blue dye. It also provides additional information about the pouch size and mucosal perfusion and allows for treatment of bleeding that may occur at staple lines. Intraoperative endoscopy can be used to evaluate both the gastrojejunostomy as well as the jejunojejunostomy. Schauer and colleagues described the use of endoscopy to evaluate the staple line of all gastrojejunostomies during RYGB as well as to perform a leak test [22]. A leak test is performed by clamping the Roux limb and submerging the gastrojejunal anastomosis and pouch in saline, while the lumen is inflated endoscopically. Leaks may be evidenced as bubbles emanating from a staple line. If a leak is present, it can be directly repaired at that time. The utilization of intraoperative endoscopy identified a 3.7 % anastomotic leak rate in 290 patients in one study [23] and a 4.1 % incidence of intraoperative technical errors in another, including 29 suture and staple line leaks, 2 bougie perforations, 2 inadvertent stoma closures secondary to the suture line, and 1 mucosal perforation in a gastric pacemaker, during 825 bariatric procedures [24]. Alaedeen and colleagues performed a retrospective review of 400 bariatric cases that included intraoperative endoscopy or methylene blue to identify leaks [25]. Postoperatively, all the patients underwent an upper GI series to evaluate for missed leaks, and the reported anastomotic leak rate was significantly lower after the use of endoscopy instead of methylene blue, at 0.4 % vs. 4 %, respectively. Similar techniques can be applied to SG to assess the staple line construction with a leak test, to check for bleeding, and to identify any technical errors such as narrowing at the esophageal–hiatal junction or incisura or endoluminal twisting of the sleeve. A leak test in an SG is performed by clamping proximally to the pylorus and submerging the stomach and staple line in saline while the lumen is inflated.

Endoscopy also serves as a valuable technique when facing challenging surgical cases or when treating complicated patients. Intraoperative endoscopy can be used to control bleeding and to deliver instruments. It can also be used for direct, percutaneous placement of feeding or decompression tubes into the Roux limb if needed.

Intraoperative endoscopy during revisional surgery is often helpful as an adjunct to external, surgical views. Even after extensive preoperative evaluation, intraoperative endoscopy is important to verify the distorted anatomy as well to determine the placement of new anastomoses. Similarly to an initial bariatric operation, endoscopy can be used to assess the integrity of anastomoses and to inspect for bleeding [26]. In reoperative bariatric surgery, a leak test may be performed on the newly revised or created pouch prior to anastomosis creation, evaluating the staple line integrity before committing to an anastomosis, thus allowing for immediate revision if needed. Intraoperative endoscopy during revisional bariatric surgery can also decrease operating times by helping to locate and identify gastrogastric fistulas, stenotic lesions, and gastrojejunal stoma locations [27].

Intraoperative endoscopy can also advance the training of future gastrointestinal and bariatric surgeons. It allows for the development and mastery of endoscopic skills under a structured and supervised setting [28, 29].

Postoperative Management and Techniques

Experience and skill in flexible endoscopy are important to bariatric surgeons in the postoperative period as well. Patients may present with varying symptoms suggestive of postoperative GI pathology or complications that require evaluation. Flexible endoscopy allows visualization of anatomy, identification of pathology, and potential intervention. Furthermore, the utilization of endoscopy by the surgeon who performed the primary operation establishes a unique firsthand knowledge when treating a patient; however, this is not always possible. Physicians unfamiliar with the anatomic changes as a result of bariatric surgery may potentially misinterpret findings on upper endoscopy.

Indications for Endoscopy

Symptoms frequently direct the clinical evaluation of postoperative bariatric surgery patients. Common symptoms include nausea, emesis, abdominal or retrosternal pain, dysphagia, and inadequate weight loss or weight regain [30, 31]. In general, the two most common indications for endoscopy in these patients are the evaluation of symptoms and the treatment of complications. The etiology of symptoms is often multifactorial; however, symptoms are frequently associated with dietary noncompliance and insufficient mastication. Patients with persistent symptoms should be further evaluated as these symptoms may indicate the development of complications after surgery. Patient history may be helpful in differentiating the etiology of pain and may guide patient work-up. Endoscopy is often the preferred diagnostic strategy and can effectively assess mucosal integrity, detect stenosis, and/or exclude other pathologic abnormalities in the surgically altered GI tract. Nausea, emesis, and bloating, with or without abdominal pain, can suggest an obstructive cause (strictures, internal hernias, or bezoars), a marginal ulcer, and band erosion or slippage or be indicative of dumping syndrome [26]. Dysphagia can be caused by esophageal dysmotility or anastomotic stenosis. In a study by Wilson and colleagues, 62 % of patients who were seen with persistent nausea and emesis after RYGB had significant findings of upper endoscopy (ulcers, stomal stenosis, staple line dehiscence) [32]. Retrosternal or abdominal pain may be caused by acid reflux, bile reflux, ulceration, or band erosion and should be evaluated by endoscopy. In addition to endoscopic evaluation, an upper GI contrast series or computed tomography (CT) scan with oral contrast should be considered.

Endoscopic Findings in Normal Postsurgical Anatomy

Roux–en–Y Gastric Bypass—The esophagus and esophagogastric junction should appear normal after an RYGB (Fig. 1a). It is important to limit the amount of air insufflation when evaluating the gastric pouch as it is of variable size. Further, special care should be made to examine the pouch and suture line for fistulas and ulcerations. The gastrojejunostomy will normally have a stoma measuring 10–20 mm in diameter. When an endoscopy is not performed by the surgeon who operated on the patient, variation in surgical technique must be determined and thoroughly identified. There can be variations in the anastomosis of the gastrojejunostomy and also in the length of the Roux limb. The gastrojejunostomy anastomosis is dependent on surgical technique, hand-sewn or stapled, and also the type of stapler used, circular or linear. Distal to the gastrojejunostomy, a short, blind limb is often seen alongside the efferent jejunal limb, often referred to as a candy cane. The Roux limb length typically ranges from 75 to 150 cm. The jejunojejunostomy may or may not be able to be reached with a standard upper endoscope.

Fig. 1.

Endoscopic appearance of (a) normal gastrojejunostomy and gastric pouch after gastric bypass, (b) normal compression from band seen on retroflexed view after adjustable gastric banding, and (c) normal staple line after sleeve gastrectomy.

Adjustable Gastric Banding and Vertical Banded Gastroplasty—Endoscopy is relatively straightforward after AGB and VBG. Dependent on the fluid volume in the band, AGB can produce a variable amount of extrinsic circumferential compression on the stomach that can be seen with the endoscope (Fig. 1b). At the time of endoscopy, it is important to determine the length of the pouch above the compression of the band to the gastroesophageal junction in order to assess for pouch dilatation or band slippage. The endoscopist should also evaluate for possible band erosion into the gastric wall. This may best be seen on retroflexion. In a VBG, the lesser curvature channel allows endoscopic visualization of the pouch, and the stoma is typically located 7–8 cm distal to the gastroesophageal junction. The banded portion in a VBG is variable in diameter, and once this area is traversed, the distal stomach and duodenum are accessible. In both procedures, retroflexion of the endoscope within the antrum will reveal the greater curvature and gastric fundus.

Sleeve Gastrectomy—The SG creates a long tubular stomach limited in expansion by a staple line that parallels the lesser curvature. During endoscopy, the staple line should be examined for defects and ulcerations (Fig. 1c). Specific attention should also be paid to patency at the incisura, located approximately midway to 2/3 of the distance to the pylorus from the esophagogastric junction, as well as to twisting of the lumen of the sleeve.

Biliopancreatic Diversion/Duodenal Switch—This procedure is often performed in conjunction with a partial gastrectomy, but it also involves a duodeno-ileal anastomosis that can be visible just distal to an intact gastric pylorus. The ampulla is not available for visualization or for endoscopic retrograde cholangiopancreatography (ERCP) in a standard fashion.

Bariatric Surgery Complications

Complications after bariatric surgery may present early or late in the postoperative course. Studies have reported varying rates of postoperative complications (Table 2) [22, 33–36]; however, the rate of major adverse postoperative complications has been demonstrated between 4 and 6 % [35, 37]. Early complications such as bleeding, infection, and anastomotic leaks often require surgical intervention [37–39]. Furthermore, if a leak is suspected shortly after surgery, a contrast radiologic study can serve as the initial diagnostic test and is helpful to delineate anatomy. Complications can also develop later in the postoperative period. Late complications, such as ulcers, stenosis, gastrogastric fistulas, obstruction, band slippage or erosion, pouch dilation, and primary weight loss failure or weight regain, can occur with varying rates after any operation; however, some are more procedure specific (e.g., erosion after adjustable gastric banding). Several of these complications can be managed successfully endoscopically.

Table 2

Complication rates after bariatric surgery

Source	Year	No. of patients	Late complication rate (%)	Early complication rate (≤30 days) (%)
Schauer et al. [22]	2000	275	47.3	30.5
Weller et al. [34]	2008	19,156	NR^a	5.0
Encinosa et al. [33]^b	2009	2,522/7,060	41.7/32.8	33.7/25.5
Flum et al. [35]	2009	4,610	NR	4.1
Masoomi et al. [36]	2012	226,043	NR	4.9

^a NR not reported

^bTwo time frames: 2001–2002 and 2005–2006

Endoscopic Management of Postoperative Complications

Marginal Ulcers—Defined as ulcers that occur at the gastrojejunal anastomosis, they can occur in 1–16 % of patients following RYGB [40–43]. Marginal ulcers typically present with epigastric or abdominal pain, bleeding, or nausea, although they may be asymptomatic [32]. Ulceration can occur at any time postoperatively, but most ulcers occur in the first several months following surgery [44]. In a study conducted on all patients who underwent bariatric surgery, postoperative endoscopic examination 1 month after surgery demonstrated ulcers in 4.1 % of patients after open RYGB and in 12.3 % of patients after laparoscopic RYGB. 28 % of the demonstrated ulcers occurred in the absence of symptoms [45]. Marginal ulcers are frequently located on the jejunal side of the anastomosis, so careful attention should be paid to this area during endoscopy (Fig. 2). When possible, a retroflexed view can identify a potentially missed location of an ulcer. Even though the exact etiology is unknown, ulcers may result from gastric acidity (due to staple line dehiscence or gastrogastric fistula), pouch orientation and size (that may incorporate a greater parietal cell mass), H. pylori infection, the presence of staples and suture material (inciting a localized inflammatory reaction), and local ischemia and tension at the anastomosis [41, 44, 46]. Smoking and nonsteroidal anti-inflammatory drug use increase the risk of marginal ulcers, and the use of proton pump inhibitors appears to decrease risk [1].

Fig. 2.

(a) Endoscopic view of large marginal ulcer and (b) retroflexed view of marginal ulcer.

The role of endoscopy for marginal ulcers is primarily diagnostic with limited therapeutic use. When identified during endoscopy, the pouch must be carefully examined for a fistula. Staple line dehiscence and the formation of a gastrogastric fistula can result in an increase in acid exposure in the pouch, stoma, and jejunum and make the mucosa more vulnerable to damage [26]. One study reported that stomal ulcers were associated with gastrogastric fistula in as many as 65 % of cases [46]. Ulcers may also represent foreign body reactions to sutures or staples, and judicious removal of foreign material with endoscopic tools may cause ulcer resolution [30]. If marginal ulcerations are not associated with staple line dehiscence or foreign body reaction, the management includes evaluation of the pouch for H. pylori status, proton pump inhibitor therapy, and liquid sucralfate and the elimination of ulcerogenic medication [38]. If marginal ulcers are severe and persist despite these measures, surgical revision may be required to prevent complications such as perforation, recurrent bleeding, and anastomotic strictures.

Stenosis—Luminal stenosis or stricture is an important complication of bariatric surgery. After RYGB, postoperative stricture formation is around 3 % [28]. The gastrojejunal anastomosis is the most common site of stenosis after bariatric surgery and has been reported in 5.1–6.8 % of patients after laparoscopic RYGB, typically within the first year [30]. Other locations where stenosis can develop include the gastric band, site of passage through the mesocolon, jejunojejunal anastomosis, and at adhesions. Anastomotic strictures are defined as anastomoses that are smaller than 10 mm in diameter [1]. Stenosis may arise from ischemia or ulceration, but the rates of stenosis are also somewhat technique dependent; the use of circular staplers has a higher rate of stricture than hand-sewn or linear staplers. Also, the use of 25 mm circular stapler reduced the rates of stricture when compared to the use of a 21 mm circular stapler [47]. Patients may present postoperatively with nausea, emesis, dysphagia, malnourishment, or unhealthy weight loss. Stenosis can be diagnosed by contrast radiography, but direct endoscopic visualization is preferable because it has high sensitivity and therapeutic measures can be performed (Fig. 3) [48]. Typical findings on endoscopy include a narrowed orifice precluding the passage of the endoscope; however, other potential findings include gastric pouch dilatation, undigested food, or foreign material [49].

Fig. 3.

(a) Radiographic evidence of a stricture and (b) endoscopic appearance of stenosis after gastric bypass.

Endoscopic treatment of strictures can be safely and effectively performed by using through-the-scope (TTS) balloon dilators or wire-guided bougie dilators [30, 50, 51]. Although initial success rates of up to 93 % have been reported, management may require multiple dilations. The length of time from surgery to stricture formation and the diameter achieved with the first dilation procedure are significant predictors of the need for further dilations [52]. Repeat dilation with progressively larger balloons may also be required to achieve more durable results, and gradual dilations over multiple sessions may reduce the risk of perforation [53, 54]. Even still, some stenoses cannot be sufficiently dilated, and these patients will require surgical revision. Overaggressive dilation should be avoided, not only to reduce perforation risk, but also because dumping symptoms and weight regain can occur [49]. There have been reports of successful dilation up to 20 mm [55] without weight regain, but data is controversial on dilation greater than 15 mm, and many authors recommend against it [49]. When standard dilation is unsuccessful, additional strategies can be utilized such as the removal of exposed sutures with endoscopic scissors, injection at the anastomosis with saline or steroids after dilation, needle-knife electrocautery of scar tissue, or even argon plasma coagulation combined with diathermy [26, 30, 56].

Stenosis following AGB may be due to fibrosis of gastric tissue in the region of the band, formation of adhesions, or band angulation or slippage. Endoscopic dilation may be effective when the cause is fibrosis or adhesions but is rarely useful in the setting of band angulation or rotation [26]. Such patients should not have repeated dilations but instead treated surgically with either band removal, removal and replacement, or conversion to another procedure. After a VBG, stenosis is usually the result of stricturing and scarring of the outflow tract in the proximal stomach and thus the creation of a hypertrophic scar or frank erosion. The incidence of stricture has been reported at 13 % [57]. Endoscopic balloon dilation is not durable in the long term but may transiently alleviate symptoms [58]. Operative revision is typically required due to the fixed nature of the mesh band or for erosion into the lumen.

Stenosis can occur after SG with an incidence that ranges from 0.2 to 4 % [59]. Strictures have a higher occurrence with the use of a smaller bougie size and a tighter sleeve. They are generally seen in the proximal to mid stomach, at the incisura, or at the esophagogastric junction. The incisura is a common spot of narrowing resulting from stapling too close to the lesser curvature. Management options of strictures and stenosis after SG include: observation, endoscopic dilation with or without stent placement, seromyotomy, and conversion to RYGB. If endoscopic dilation has failed for 6 weeks, reoperation is typically recommended [60]. After SG, torsion or rotation of the remnant sleeve may present similarly to a stenosis with obstructive symptomatology and can be managed with dilation, myotomy, or revisional surgery.

Gastrointestinal Bleeding—Bleeding in patients after bariatric surgery may be acute or chronic or present as an iron deficiency anemia [61]. Bleeding can occur anywhere in the GI tract, including in the biliopancreatic limb and remnant stomach after an RYGB. Significant upper GI bleeding occurs in about 1–4 % of patients after RYGB [62], about 0.1 % after AGB [63, 64], and between 1 and 2 % after a sleeve gastrectomy [65, 66]. Patients with signs or symptoms of acute or chronic bleeding should be evaluated with endoscopy, preferably in close consultation with a surgeon, should complications arise or endoscopic interventions fail [62, 67]. The benefit of endoscopy is the ability to provide diagnosis and treatment simultaneously. However, endoscopy in the early postoperative period may be challenging, especially after RYGB and BPD because of the inaccessibility of the biliopancreatic limb, remnant stomach, and the jejunojejunostomy and the potential risks associated with early postoperative endoscopy such as perforation at the surgical anastomoses [68]. If more advanced endoscopic techniques, such as double-balloon enteroscopy, are unsuccessful at accessing the bypassed anatomy, access may be gained through a surgically created gastrostomy [69]. Endoscopy after an SG, VBG, and AGB is relatively straightforward, and standard flexible upper endoscopy is usually sufficient for the management of endoluminal bleeding in this situation.

Numerous approaches for treating active upper GI bleeds have been described in the literature. Techniques that involve the use of thermal energy (electrocoagulation, heater probe, and argon plasma coagulation), mechanical application of clips, and local injections with epinephrine, sclerosants, and thrombin/fibrin glue have all been successfully reported [70]. A retrospective review of 933 patients after RYGB reported a 3.2 % incidence of postoperative hemorrhage and an 80 % rate of successful endoscopic intervention [71]. Bleeding after an SG tends to occur at the staple line and is usually self-limited. Rarely, the use of endoscopy to suction out or push out a blood clot may be necessary [66].

Leaks and Fistulas—Gastric leaks and fistulas are potentially serious complications of bariatric surgery and cause significant morbidity. Overall, the occurrence of leaks is between 0.4 and 6 % in gastric bypass patients [24, 50, 72] and is 2.4 % in SG patients [73]. High-volume centers tend to report anastomotic leak rates of less than 2 % [24, 74, 75]. Identification of a leak in the immediate postoperative period suggests that either an intraoperative leak test with endoscopy missed the defect or that the leak developed after the completion of the operation. Staple line disruption can result in extraluminal leaks or eventually gastrogastric fistulas (most common type). Extraluminal leaks tend to present early in the postoperative period and can result in peritonitis, abscess, sepsis, organ failure, and even death [30, 76]. After RYGB, most leaks occur at the gastrojejunal anastomosis followed by the remnant stomach; leaks at the jejunojejunal anastomosis are rare but do occur and usually need reoperation (Fig. 4a). After SG, leaks are typically found in the proximal third of the stomach and specifically at the areas of the esophagogastric junction [73]. Upper GI studies are typically used to diagnose extraluminal leaks (Fig. 4b). CT scans are another common imaging modality used to examine the anatomy of the anastomoses and staple lines. Typically, leaks with clinical signs of sepsis require operative repair, drainage of infection, and establishment of enteral access. Endoscopy has an adjunct role in the operating room, for example, to define the precise location of the leak and, increasingly, to be used as a therapeutic measure. There have been small series that have described endoscopic management of leaks using partially covered self-expanding metal stents, Polyflex stents, argon plasma coagulation, endoscopic clips, and fibrin glue [77, 78]. Merrifield and colleagues reported successfully treating three patients with leaks endoscopically and concluded that endoscopy may be a feasible, less invasive alternative to surgical repair [77]. A study at the Cleveland Clinic, which used three different types of stents for the management of anastomotic complications after bariatric surgery (a prototype salivary stent, a partially or fully covered self-expanding metal stent, and a silicone-coated polyester stent), demonstrated that endoscopic stent placement successfully resolved anastomotic leaks in 85 % of patients (Fig. 5) [79]. Such novel methods are still currently investigational, and further research is needed to define the role of endoscopy to treat postoperative anastomotic leaks.

Fig. 4.

(a) Illustration of common leak locations after Roux-en-Y gastric bypass (RYGB). (b) Upper gastrointestinal study showing leak at gastrojejunal anastomosis after RYGB (white arrow pointing at leak). Reprinted with permission, Cleveland Clinic Center for Medical Art & Photography © 2010–2013. All Rights Reserved.

Fig. 5.

(a) Sleeve gastrectomy with stent and radiographic appearance of stent after placement. (b) Roux-en-Y gastric bypass with stent and radiographic appearance of stent. Reprinted with permission, Cleveland Clinic Center for Medical Art & Photography © 2010–2013. All Rights Reserved.

Chronic fistulas may be found in the presence of marginal ulcers or as a result of staple line disruption. Staple line dehiscence after SG has been reported at rates ranging from 0.3 to 5 % [80]. Patients with a chronic fistula may present with nausea, emesis, epigastric pain, and weight gain. However, many fistulas may remain subclinical, and the true incidence is not entirely known [41]. Fistulas, similar to acute leaks, are typically diagnosed with upper GI series. A large fistula may also be visualized by endoscopy. A gastrogastric fistula is the most common type and is depicted in Fig. 6. Surgery is the mainstay of treatment, but techniques of endoscopic management are being actively investigated [81]. For example, endoscopic treatment of postoperative fistulas has been accomplished with self-expanding stents and endoscopic suturing and clipping [82–85]. Successful closure of gastrocutaneous fistulas after VBG and BPD using endoscopic fibrin sealant injection has also been reported [86]. Although these techniques are feasible, long-term durability is dependent on fistula size, with large fistulas yielding suboptimal results [87].

Fig. 6.

Band Erosion and Slippage—Band erosion into the gastric lumen can occur after AGB and VBG. Band slippage is another complication that can occur after AGB. The incidence of band erosion after VBG and AGB is uncommon and has been reported at 1–3 % after VBG and 0.9–3.8 % after AGB [88–90]. Band erosion can be asymptomatic or can cause abdominal pain, nausea, emesis, access port site infection, fistula, increased food intake, and GI bleeding. Band slippage may present with weight gain, worsening reflux symptoms, or obstruction. Erosion is best diagnosed endoscopically. Endoscopy can allow for direct visualization of a band eroding through gastric mucosa (Fig. 7), and endoscopic removal techniques for near completely eroded bands have been described [89, 91]. However, surgical repair is usually recommended with excision and replacement or conversion. While band slippage can be demonstrated with endoscopy, it may be best diagnosed with contrast radiography, since findings on upper endoscopy are variable and dependent on the degree and type of slippage encountered. Findings may include an enlarged pouch size, reflux esophagitis, gastritis, or ulcers. Severe cases are potentially life-threatening as they can lead to gastric necrosis [92, 93]; this may be demonstrated endoscopically as mucosal ischemia.