Approach to Marginal Ulceration Following RYGB Surgery: Laparoscopic Excision of the Marginal Ulcer and Retrocolic, Retrogastric Rerouting of the Roux Limb with Truncal Vagotomy and Hiatal Hernia Repair

Gastrogastric fistula is a rare complication following a Roux-en-Y gastric bypass procedure wherein there is a communication between the proximal gastric pouch and the distal gastric remnant. Patients typically present with nausea and vomiting, abdominal pain, intractable marginal ulcer, bleeding, reflux, poor weight loss, and weight regain. Etiologies include postoperative Roux-en-Y gastric bypass leaks, incomplete gastric division, marginal ulcers, distal obstruction, and erosion of a foreign body. Diagnosis is made through upper gastrointestinal contrast radiography or CT scan and endoscopy. Barium contrast radiography is particularly useful and is the preferred initial study method for the detection of staple-line dehiscence, which may be small and overlooked during endoscopy. Once identified, a gastrogastric fistula may be treated surgically with remnant gastrectomy or gastrojejunostomy revision. Here, we present a case of a female patient status post Roux-en-Y gastric bypass surgery who presented with abdominal pain. Upon endoscopy, she was noted to have an inflamed gastric pouch and a gastogastric fistula. A laparoscopic gastric bypass revision was done to divide the gastrogastric fistula and to separate the gastric pouch from the gastric remnant in order to alleviate the inflamed gastric pouch and prevent further ulcer formation.

Obesity is one of the fastest-growing public health concerns in the United States. Of the currently available management options, it is evident that metabolic and bariatric surgery (MBS) is among the most effective at inducing long-term weight loss and resolution of obesity-related co-morbidities such as type 2 diabetes, obstructive sleep apnea, and non-alcoholic fatty liver disease. The Roux-en-Y gastric bypass (RYGB) remains one of the most effective and durable metabolic and bariatric procedures producing consistently excellent weight loss and metabolic outcomes.^1,2  Although effective, MBS is still a major surgery that carries the risk of various complications. Marginal ulcers (MU) are a relatively common complication that occurs following RYGB. MU typically form at the gastrojejunal anastomosis in the proximal jejunum.³ The reported incidence of marginal ulcers varies widely, typically in a range between 1–16% of all RYGB cases.^4-6 This variability is likely due to the retrospective nature of many studies, and the inconsistencies in diagnosing MU—some only include those diagnosed by endoscopy, while others include those with a probably clinical history. Furthermore, because endoscopy is typically performed on symptomatic patients only, some studies miss asymptomatic MU cases.⁷ In a prospective study using endoscopy to screen all RYGB patients at one month following surgery, MU was found in 5.6% of patients.⁷  

Patients with marginal ulcers most commonly present with epigastric pain (50–60%) and/or bleeding (15–25%) which may manifest as melena or hematemesis.^3,8 About 20% of patients also experience nausea and vomiting. If the MU is accompanied by other complications, additional symptoms may appear in history. Gastrogastric fistulas, which occur in approximately 1.18% of patients undergoing RYGB, present with weight regain, insulin resistance, and pain.⁹ MU perforations may present with pain and an acute abdomen.¹⁰ Stenosis, a complication of untreated marginal ulcers, presents with dysphagia, solid food intolerance, and delayed vomiting of undigested food.¹¹ Critically, up to 28% of patients with MU may be asymptomatic, and some patients may present only with painless upper GI bleeding.⁷

The patient’s history may also reveal modifiable risk factors for developing MU. Smoking history is one of the strongest independent predictors of MU developing following RYGB, conferring nearly 5-fold higher risk compared to non-smokers regardless of how heavily they smoke or even whether the patient has quit smoking.^12,13 Similar to regular peptic ulcers, NSAID use also contributes to increased incidence and impaired healing of MU after RYGB, although some studies dispute this.^14-16 Daily low-dose aspirin has not been found to increase MU risk.¹⁷ Some studies have found a significant association between diabetes and MU incidence, while others have not.^3,18 BMI and alcohol use do not appear to predict MU occurrence.^19,20 

Marginal ulcers are primarily diagnosed by symptoms and imaging, but physicians may note physical exam findings secondary to the primary ulcer. If they have a bleeding ulcer, the patient may be anemic and show pale skin, tachycardia, and postural hypotension. Patients may be malnourished and dehydrated, caused by food avoidance due to pain, chronic vomiting, or gastric stricture. Dehydration may present with ketosis, tachycardia, sunken eyes, and decreased skin turgor, and suspicion of malnutrition warrants blood labs for micronutrients and serum proteins.

Upon presentation with symptoms consistent with a marginal ulcer, patients should undergo imaging for definitive diagnosis. An upper GI series or a CT scan with oral contrast are both rapid and effective methods to detect fistulas following gastric bypass however they are not sensitive for detecting MU.⁹ Although an upper GI series may be insensitive to detecting small fistulas, more subtle imaging findings, such as contrast in the gastric remnant, can help diagnose fistulas. If the patient presents with signs of bleeding, a CT scan with IV contrast may also help identify the bleeding source. Upper GI endoscopy is the gold standard for diagnosing MU, enabling providers to note the size of the ulcer and the presence of any removable foreign bodies such as sutures or staples.20 Endoscopy can also offer therapeutic intervention such as bleeding control, dilatation of strictures, or closure of fistulas. In one study of patients who received endoscopy for upper GI symptoms following RYGB, 15.8% were diagnosed with marginal ulceration.²¹ In the same study, patients who developed symptoms three months or earlier in the postoperative period were more likely to have an abnormal endoscopy result.²¹

The natural history of marginal ulcers can be divided into the etiologies that cause them, and the potential complications that follow them.

Etiology: Acid

Similar to regular peptic ulcers, acid is also related to the pathogenesis of marginal ulcers. Current surgical standards for RYGB creates a small, proximal pouch that excludes most of the parietal cells in the antrum, significantly reducing the risk of developing MU.⁸ In cases of dilated pouches or gastrogastric fistulas, the pouch or the gastric remnant may produce excessive acid that is passed to the jejunum, which lacks the buffering capacity of the duodenum. These patients are at high risk of developing marginal ulcers. Although it is clear that acid impairs ulcer healing, it cannot explain the etiology of all marginal ulcers because patients still form ulcers early in their recovery course when pouches are not yet dilated, and some marginal ulcers don’t heal even with high-dose proton-pump inhibitors (PPI).¹⁵

Etiology: Ischemia

Local ischemia significantly increases the risk of marginal ulceration.⁸ Smoking, diabetes, and coronary artery disease all cause microvascular insufficiency and ischemia and are all independent risk factors for developing MU.^3,15 Furthermore, the blood supply in the mesentery is often disrupted during surgery, and the resultant anatomy requires that blood must flow anti-gravity to reach the gastrojejunostomy. The anastomosis, most distal to the blood supply, is most likely to contain an ischemic ulcer.³ Furthermore, ulcers biopsied and excised routinely show ischemic pathology.

Etiology: Foreign body

During diagnostic endoscopy, sutures or staples are found in approximately one-third of marginal ulcers.¹⁸ The foreign body causes irritation and mucosal erosion, significantly increasing ulcer risk. For this reason, many surgeons now opt for absorbable sutures, which significantly reduces the risk of MU and gastrogastric fistula development.²²

Etiology: H. Pylori

The exact role of H. pylori in the pathogenesis of MU is unclear. Some studies have not found an increased complication rate in patients infected with H. pylori, while others have identified H. pylori infection as an independent predictor of MU formation following gastric bypass.^23,24 Compared to peptic ulcers in the general population, H. pylori is not an important risk factor for marginal ulcers, although many bariatric centers still screen for H. pylori and preoperative H. pylori eradication is widely practiced.^25,26 

Complication: Fistula

One study found that nearly one-fifth of patients diagnosed with MU also had a concomitant gastrogastric fistula (GGF). GGFs are much more common in patients who have undergone gastric partitioning used in open surgery instead of the divided gastric pouch created in laparoscopic and robotic surgeries; however, the incidence of MU is no different.²⁷ In some cases the fistula may be a result of a prior leak or staple line disruption that occurs in the pouch and is not directly associated with the ulcer. In other cases, the ulcer creates the fistula by eroding into the gastric remnant or adjacent organs. Of patients with GGF, 60% have a history of marginal ulcers.⁹  Although gastrogastric fistulas are most common, fistulas related to MU can also involve the small intestine, colon, pancreas, liver, and major blood vessels including in one reported case, the aorta.²⁸

Complication: Perforation

Perforated marginal ulcers are potentially lethal and may be surgical emergencies. About 1% of all LRYGB patients present with perforated MU.¹⁰ Importantly, patients who perforate may present several years following surgery with symptoms that are atypical for most hollow organ perforations.²⁹ Serum inflammation markers may be normal, and signs of free air and peritonitis may not be present on imaging. However, of all patients with perforated MU, 80% had an identifiable risk factor such as smoking history, NSAID use, steroid use, or history of prior MU.¹⁰ Although perforation is relatively rare after RYGB, 25% of patients who experience perforation have a recurrence of MU, so this patient population must be carefully followed.³⁰

About two-thirds of MU cases respond to medical treatment alone, but in cases of recalcitrant or complicated ulcers, surgical intervention is necessary.^3,8 Conservative medical treatment involves smoking cessation, NSAID discontinuation, H. pylori eradication, PPIs, and sucralfate.⁵ Unlike PPI therapy, H₂ antagonists have not proven to be effective against MU secondary to the gastrogastric fistula.²² The administration of PPI prophylaxis to prevent MU formation is not standardized in the bariatric community. Although some studies have found no protective effect, others have shown that PPI prophylaxis is effective in reducing the risk of MU, particularly if the patient is already taking NSAIDs.^8,20,31 A study of 2830 RYGB cases found that postoperative PPI reduced MU risk by half, and a 90-day course has been shown to be more effective than a 30-day course.^15,32

Untreated ulcers can lead to strictures and gastric outlet obstruction, so repeat endoscopies to determine healing are critical. Furthermore, endoscopy has emerged as a less invasive option to treat medically refractory marginal ulcers. Balloon dilatation of strictures caused by marginal ulcers are more safely performed after the ulcers have healed. If there is severe stenosis that fails 3 attempts at dilatation and/or significant ulceration, a covered stent could be used to treat the stricture.^33,34 These techniques range from simple interventions, such as removing sutures from the site of a non-healing ulcer, to more complex procedures, such as placing sutures and stents to treat perforated ulcers.^21,35-38

Despite the excellent options for medical treatment, about 17% of MU cases require surgical intervention within 8 years of diagnosis.³⁹ Patients with perforations, dilated pouch, gastrogastric fistula, or no modifiable ulcer risk factors are more likely to require revisional surgery. Of MU patients who require revisional surgery, about 72% have a GG fistula.²⁷ Surgery typically involves excising the ulcer and reconstructing a new GJ anastomosis. The success of surgical interventions is variable depending on the center and the patient population. In one study, 87% of patients remain symptom-free after revisional surgery, while in another study 57% of patients had MU recurrence 1-year after revisional surgery.^15,27

To avoid complications that require revisional surgery, patients at significant risk for MU following RYGB may consider an alternative form of bariatric surgery, such as laparoscopic sleeve gastrectomy (LSG). This includes patients who smoke or are exposed to significant second-hand smoke, patients dependent on steroids, or patients who rely on pre-operative NSAIDs.¹⁵ In the management of MU, patients who have no modifiable risk factors and/or live in remote areas without ready access to a bariatric center may consider revision surgery as a first-line definitive treatment for MU. This limits the risk of an untreated ulcer causing a surgical emergency, such as perforation or massive hemorrhage.

This case describes a laparoscopic surgical revision of a gastric bypass following the development of a marginal ulcer with a gastrogastric fistula. It includes a partial gastrectomy and revision of the gastrojejunostomy to excise the ulcer, vagotomy, and rerouting of the roux limb to the retrocolic, retrogastric location.

The development and subsequent management of MU following RYGB merits several technical considerations. First, gastric pouch size and separation from the remnant affect MU risk. Larger, more distal pouches (such as in biliopancreatic diversions) have a higher risk of MU, whereas a smaller pouch (5–6 cm) significantly reduces the risk of MU.^8,40,41 In some RYGB cases, the pouch and the gastric remnant are partitioned by a staple line, but not transected and separated. Complete transection, or even removal of the gastric remnant, significantly reduce the risk of GGF and subsequent MU.^7,22

Second, absorbable sutures should be used for the GJ anastomosis to prevent mucosal erosion caused by permanent sutures. On endoscopy, foreign materials such as sutures or staples are found in one-third of marginal ulcers.⁸ In one study, the incidence of gastrogastric fistulas decreased from 5.1% to 0% when the anastomosis was created with absorbable sutures rather than staples.²²

Third, truncal vagotomy is often performed to reduce acid production by eliminating acetylcholine stimulation on parietal cells. Historically used to treat peptic ulcer disease, truncal vagotomy may also decrease parietal cell sensitivity to gastrin and histamine, and reduce gastrin production from the antrum.⁴² Recent advances in robotic or minimally-invasive thoracoscopic vagotomy may be useful to treat recalcitrant marginal ulcers.^43,44 Truncal vagotomy may lead to development of post-vagotomy syndrome, including diarrhea and gastroparesis. While post-vagotomy gastroparesis is not functionally significant in patients with gastric bypass, post-vagotomy diarrhea should still be considered a possible complication.^45-47 Patients who undergo vagotomy must remain on PPI for at least three months post surgery to allow time for the parietal cells to regress.

Finally, some surgeons prefer to place the Roux limb in the retrocolic position to reduce tension on the mesentery and avoid anastomotic ischemia. There is no data to suggest that the positioning of the Roux limb affects MU risk, so the decision of whether to use antecolic or retrocolic placement depends on the individual surgeon.

The time of MU onset is an important consideration in identifying the underlying etiology. Early (<30 day post-op) ulcers are rare, occurring in less than 1% of RYGB patients.⁴⁸ Because these ulcers develop even when absorbable sutures are used, and because they develop before pouch dilation may increase the number of parietal cells, it is unlikely that foreign bodies or acid production are the cause.⁷ A more plausible explanation is that MU occurring less than a month following surgery is likely the result of inflammation, ischemia, electrocautery, and general tissue damage associated with the surgery.^7,48 Patients who underwent anticoagulation therapy prior to surgery were also much more likely to be diagnosed with an early MU.⁴⁸ The majority of MU occur later, due to gastrogastric fistulae, pouch dilation, and foreign bodies. Ulcers caused by pouch dilation are typically more aggressive and may present with perforation or severe bleeding.⁷ Later MU are also more likely to be related to modifiable risk factors, such as smoking or NSAID use.^12,49

Improvements in surgical techniques and less-invasive management will greatly improve MU outcomes in the future. During the initial gastric bypass, determining adequate blood flow to the Roux limb is critical to preventing ischemia – a major cause of marginal ulcers. Advancements in fluorescence-based intraoperative angiography can provide real-time information on Roux limb perfusion, allowing surgeons to better visualize anatomy and plan their operative course.^50,51 The fluorescent molecule indocyanine green (ICG) is a useful tool, as it binds to plasma lipoproteins and is cleared by the liver in a first-pass effect.⁵² It has been used in LSG, and a recent case report demonstrates its use in revisional surgery for MU following RYGB.^51,53

The One-Anastomosis gastric bypass (OAGB) is a recently developed bariatric surgery that creates a longer pouch and better preserves the mesentery, allowing better Roux limb perfusion.⁵⁴ OAGB produces excellent weight-loss outcomes, and the ulcer rate appears to be less than with standard RYGB, most likely owing to the presence of pancreatic fluid at the gastrojejunal anastomosis buffering the acid from the larger pouch size.^55,56 Robotic RYGB using the double loop technique also preserves the mesentery and therefore may decrease the risk of marginal ulcers.⁵⁷ While there is no large study showing decreased risk of ulceration, current studies show only a small number of postoperative ulcers.⁵⁸

Finally, endoscopy is becoming more widely utilized in the management of MU, not merely in the diagnosis. From basic interventions, like endoscopic removal of foreign bodies, to more advanced procedures, endoscopy may be a less invasive method to treat ulcers. One study reports suturing of a mucosal advancement flap endoscopically for the treatment of a nonhealing MU.^59,60

GGF after RYGB can also be managed endoscopically. The usage of endoscopic self-expanding stents, clips and fibrin glue have shown promising results in management of leaks and fistulas after bariatric surgery, with success rates reported between 70–93%. This minimally-invasive approach effectively reduces symptoms and promotes healing of the fistula, offering a viable option for patients experiencing complications post-surgery in properly selected patients.⁶¹

No special equipment was used for this operation. 

Nothing to disclose.

The patient referred to in this video article has given their informed consent to be filmed and is aware that information and images will be published online.