Ulcer Disease in the Excluded Segments after Roux-en-Y Gastric Bypass: a Current Review of the Literature

After RYGB, there is continued, although diminished, basal, and stimulated acid excretion in the excluded stomach. Gastrin levels are decreased after RYGB, yet the gastric mucosa maintains its ability to respond to vagal and hormonal stimuli [30], thus preserving an acidic environment [12, 30‐32]. The amount of acid production is influenced by the proportion of the parietal cell mass, partitioned by surgery to the pouch and the distal stomach, a factor certainly influencing the development of MU [33‐35]. In theory, high transection increases acid production in the excluded stomach and contributes to acid-related mucosal injury in the antrum and duodenum [12, 31]. Furthermore, post-RYGB, cellular hypertrophy of the gastric mucosa occurs in the presence of a reduction in G cells, again lowering gastrin production [36]. Additionally, in proton-pump inhibitor (PPI)–treated mice with normal anatomy, the resulting hypochlorhydria causes bacterial overgrowth resulting in gastric inflammation [37].

After RYGB, there is no counteracting effect on the acid in the antrum and duodenum as the ingested food bypasses the excluded segments. However, the exclusion of the duodenum also leads to a reduced and desynchronized pancreatic secretion as the strongest stimulus of bicarbonate secretion is the duodenal presentation of nutrients. Furthermore, bile refluxing into the gastric remnant also contributes to mucosal injury [12, 21, 38, 31].

In large database-driven studies, up to a quarter of post-RYGB patients are on PPIs [39, 40]. There is an interplay between PPI intake and gastrin levels, and long-term use leads to hypergastrinemia stimulated by a reduction in gastric acid secretion via somatostatin feedback [41]. RYGB results in altered gastric emptying, reduction in intestinal absorption surface, and a change in pH. This leads to pharmacokinetic changes where PPIs, as lipophilic drugs, are absorbed to a lesser degree [42]. As PPIs are degraded by luminal acid, they are usually protected by a coating to prohibit premature activation and enable later absorption in the small bowel which again is altered. Likewise, opened PPI capsules have a significantly better effect on MU healing than intact capsules [43]. In line with these pharmacokinetic changes, plasma concentrations of omeprazole metabolites in patients after RYGB are significantly reduced compared to the corresponding levels in pre-RYGB and controls [30].

Gastroparesis can occur following planned or inadvertent vagotomy [59]. Its incidence after partial gastrectomy is up to 5% [60]. Changes in ghrelin secretion influence gastric motility [60, 61]. Bile reflux into the gastric remnant is observed in 36% of post-RYGB patients, and its deleterious effects on the mucosal barrier lead to chronic gastritis and intestinal metaplasia and may contribute to the formation of ulcers [62, 63].

Furthermore, delayed emptying and stasis in the biliopancreatic limb may predispose patients to bacterial overgrowth followed by inflammatory changes in the intestinal wall potentially leading to ulceration [64].

Studies evaluating the gastric remnant after RYGB are rare. Among 53 patients with RYGB, remnant gastritis was found in 87% of patients with a normal mucosa in the pouch, indicating a harmful effect of unbuffered acid on the gastric remnant [31]. Elsewhere, in more than half of patients taking PPI, endoscopy of the gastric remnant revealed peptic changes [30].

Considering symptoms as surrogates for gastritis and PUD, upper abdominal pain by far is the most frequent symptom leading to readmissions after RYGB [65]. Of the reported patients, over two-thirds reported epigastric and/or upper abdominal pain.

With a sensitivity of approximately 86% in detecting gastrointestinal perforation, CT is widely accepted as accurate in the evaluation of such patients [66]. However, in post-RYGB patients, the presence of free intra-abdominal air due to perforation of the excluded segments is much less common if not absent. Patients with perforation had plain X-rays in 20 of 41 cases (49%), and free air was detected in two patients. When performed, 44% of CT scans showed free intra-abdominal air. Thus, in patients after RYGB, negative CT findings do not exclude a viscus perforation, and operative exploration should be performed with a low threshold. Indeed, the presence of free air and even more fluid should raise suspicions about a pouch-gastric fistula or irregularities at the level of the jejuno-jejunal anastomosis [6, 15, 20, 67].

The sensitivity and specificity of CT angiography for active gastrointestinal bleeding are over 90%. However, its sensitivity for obscure bleeding is only approximately 45% [68]. Even though it lacks therapeutic options, CT angiography provides essential information regarding the bleeding origin and allows for subsequent mesenteric angiography to be targeted to an area of interest leading to an improved angiographic detection rate. The technical success rate of transarterial embolization in gastrointestinal bleeding is reported to be up to 93% with a rate of bleeding cessation up to 81% [69].

Among the reported cases of ulcer bleeding following RYGB, only indirect signs but no active bleeding were detected. Mesenteric angiography without previous CT angiography was performed in 25% of patients, and angiography showed active bleeding in 67% of patients. One patient had successful embolization, whereas embolization failed in the other patient, and surgical intervention was needed to control bleeding [14, 25].

Transarterial angiography with subsequent embolization is an option to address bleeding in those patients. This procedure is non-invasive and has a high success rate [69]. However, in a general context, it might be more purposeful to address such patients surgically, or rather endoscopically via temporary gastrostomy.

There are various ways to access the excluded segments post-RYGB. Upper endoscopy is considered a first-line approach for the diagnosis and treatment of upper gastrointestinal bleeding and provides a valid alternative in the treatment of postbariatric complications such as leaks or anastomotic bleeding [70]. Whereas gastro-jejunal anastomoses can easily be evaluated, the combined length of alimentary and biliopancreatic limbs complicates access to the gastric remnants and duodenums. With the use of a pediatric colonoscope, the remnant can be reached in up to 68% of cases [31]. This rate can be raised to 88% by applying a double-balloon technique albeit with a higher perforation rate of 10% [71].

Eleven (20%) of the reported patients underwent a standard upper endoscopy, which was non-diagnostic in every case. In 5 patients, retrograde endoscopy was attempted; the remnant was reached in three patients, and treatment of the bleeding was successful in two patients [12, 14, 22, 28]. Thus, in time-critical situations, conventional upper endoscopy can rule out pathologies of the gastric pouch, anastomosis, and alimentary limb, but its diagnostic value in assessing the excluded segments is limited.

Access to the excluded segments can further be achieved with percutaneous transgastric endoscopy. CT or ultrasonographic guidance to introduce a guidewire is followed by placement of a gastrostomy tube into the remnant stomach. Serial dilations of the gastrostomy over several weeks are then required until an endoscope can be introduced. This technique is safe and effective but obviously not suitable for emergency situations [4, 5, 62].

Intraoperative transgastric endoscopy allows for addressing or excluding other pathologies at the same time [72]. Usually, access to the gastric remnant is established during laparoscopy via temporary remnant gastrostomy [72]. Among the reviewed patients with ulcer bleeding or perforation after RYGB, 8 of 54 had (15%) intraoperative transgastric endoscopy [13, 18, 25, 28]. Endoscopy revealed duodenal ulcers in 3 cases and gastritis and benign ulcers in 2 cases.

As patients with PUD after RYGB are more likely to be diagnosed with complicated PUD, the treatment is generally operative. There are several controversial issues, such as whether to perform remnant gastrectomy, ulcer repair with concomitant vagotomy, or mandatory intraoperative gastroscopy. In the case of free air or fluid, assessing the patency of the jejuno-jejunostomy is mandatory [67].

The recurrence rate of PUD determines the extent of surgery needed. So far, the paucity of data and the rather short follow-up overall do not allow for general recommendations. Although remnant gastrectomy seems drastic, it certainly facilitates the follow-up and most effectively counteracts PUD unless there is ectopic gastric tissue elsewhere. In the case of perforated PUD, oversewing after biopsies is an option. It is certainly advisable to enable easier future access to the remnant stomach, especially when intraoperative gastroscopy is not performed at the time of the procedure. (Super)selective vagotomy may be added to reduce acid production.

Of 54 described cases, 50 (93%) underwent surgery. In 48% of these patients, remnant gastrectomy was performed, combined with resection of the first portion of the duodenum in 6% and pancreas-preserving duodenal resection in 2%. Oversewing or Graham patches were performed in 37% of patients. Other procedures included ulcer excision with bypass reversal, duodenostomy with drainage, coagulation by intraoperative endoscopy, and angiographic coiling (tbl 2).

In a short-term follow-up, 94% of patients were cured, one needed reoperation because of early postoperative bleeding following remnant gastrectomy with resection of the first part of the duodenum, and two patients died, one from pulmonary complications and the other one intraoperatively due to multiorgan failure.

There is a wide variety in the duration of follow-up, ranging from 5 days to 4 years and only limited data are available overall.

Postoperative eradication of HP was routine for positive patients. Secondary prophylaxis with PPIs was initiated in 23 of 54 patients (of whom one had a remnant gastrectomy), with varying durations, from 2 weeks to lifelong. Meta-analytic data suggest a significant benefit of prophylactic PPI in reducing MU after RYGB [73]. Despite lacking evidence, PPI may therefore be considered to counteract recurrent PUD in the excluded segments.

Cancer in the remnant stomach is rare; a meta-analysis found 17 patients who reported with pain, abdominal distension, and weight loss as main symptoms [74].