Marginal ulcer rates in primary Roux-en-Y gastric bypass versus conversion of sleeve gastrectomy to Roux-en-Y gastric bypass

Abstract

Background

Sleeve gastrectomy (SG) and Roux-en-Y gastric bypass (RYGB) are well-established surgical interventions for morbid obesity, offering significant weight loss and improvement or resolution of many obesity-related comorbidities. Some patients who initially undergo SG later elect to convert to RYGB due to inadequate weight loss. Marginal ulcers (MUs), characterized by ulceration at the gastrojejunal anastomosis, are a known and prevalent complication following RYGB.

Methods

This retrospective study was conducted at a single community hospital. Records were reviewed to identify patients who underwent either primary or conversion RYGB between 2015 and 2023. Data collected included patient demographics, perioperative variables, relevant risk factors, and incidence of gastrojejunal anastomotic ulcers, diagnosed via endoscopy or computed tomography scans.

Results

A total of 113 patients were included in this study, with 49.6% of them undergoing primary RYGB and 50.4% undergoing conversion RYGB. No significant differences were found between groups in preoperative weight, BMI, ASA status, or other preoperative risk factors. However, there were significant differences in racial distribution between the groups—the primary RYGB group included 69.6% Black or African American patients while the conversion RYGB group consisted of 86% Black or African American patients (p = 0.037). Comparatively, the primary RYGB group consisted of 28.6% White patients while only 8.8% of the patients in the conversion RYGB group were White.(p = 0.007). The incidence of marginal ulceration was not significantly different between the groups (21.4% in the primary group vs. 24.6% in the conversion group; p = 0.69).

Conclusion

No differences were observed in the rates of marginal ulceration in patients in the group who underwent primary Roux-en-Y gastric bypass surgery versus those who underwent conversion RYGB from a previous SG. This study’s results contradict the results of some similar studies and challenge the hypothesis that the gastric anatomical changes after SG could possibly result in significantly compromised perfusion, thus increasing the risk of marginal ulceration. The discrepancy between our results and previous reports highlights the need for further research to clarify the underlying mechanisms and risk factors contributing to marginal ulcer development in various surgical contexts.

Introduction

Sleeve gastrectomy (SG) and Roux-en-Y gastric bypass (RYGB) are well-established surgical interventions for morbid obesity, providing substantial weight loss and resolution of many obesity-related comorbidities. Some patients who initially undergo SG later elect for conversion to RYGB due to inadequate weight loss, weight regain, and, severe or refractory gastroesophageal reflux disease. Marginal ulcers (MUs), characterized by ulceration at the gastrojejunal anastomosis, are a relatively common complication following RYGB surgery. These ulcers may present with abdominal pain, bleeding, nausea, and/or vomiting. Nevertheless, they also may be asymptomatic [1].

The pathophysiology of MUs is multifactorial. Causes of MUs can be a result of local ischemia, staple line disruption, acid exposure to the intestinal mucosa, or the presence of luminally exposed staples or suture material [1, 2]. Reported incident occurrences of MUs after RYGB vary widely, with estimates ranging from 0.6% to 36% [1, 3]. Identified risk factors include the following: Helicobacter pylori (H. pylori) infection, smoking, diabetes, excessive gastric pouch length, and the use of nonsteroidal anti-inflammatory drugs (NSAIDs) [1, 3]. The American Society for Gastrointestinal Endoscopy notes that proton pump inhibitors (PPIs) are associated with a reduced risk of MUs [1]. However, prophylactic acid suppression alone does not fully prevent MUs, indicating that acid exposure is not the sole causative factor [4].

Management of MUs typically involves PPI therapy and risk factor modification. The latter generally involves smoking cessation and H. pylori eradication [2]. For more complicated ulcers, surgical intervention may be required to manage perforations or resect strictures [2]. In recurrent or recalcitrant cases, treatment options include endoscopic coverage of the ulcer bed, anastomotic resection, and truncal vagotomy [2].

Given the increasing number of conversion RYGB (c-RYGB) procedures following SG performed globally, it is important to better understand the incidence and risk factors for MUs in this population [4]. Recent studies suggest that MU incidence may be higher after conversion RYGB compared to primary RYGB (p-RYGB). This highlights the need for more active research into prevention and underlying pathophysiology of MUs [5]. This study aims to evaluate the incidence and risk factors for MU development in patients undergoing p-RYGB versus c-RYGB.

Methods

Data acquisition

This study was approved by the Institutional Review Board of our institution. Informed consent was not required, as this was a retrospective cohort analysis. Electronic medical records and the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) Patient Use Files (PUF) were queried to identify all patients who underwent primary or conversion RYGB procedures between 2015 and 2023. All surgeries were performed by four bariatric surgeons using a standardized technique.

Data collection

Data collected included patient demographics, perioperative variables, relevant risk factors, and the incidence of gastrojejunal anastomotic ulcers, diagnosed via endoscopy or CT scan. The primary outcome was the incidence of MUs. Secondary outcomes included analysis of risk factors associated with MU development, such as NSAID use and smoking.

Statistical analysis

A two-sample t-test was used to compare continuous variables between groups. Categorical variables were compared using a two-sample proportion test or Fisher’s exact test, with statistical significance defined as p < 0.05. All statistical analyses were performed using GraphPad Prism.

Sleeve gastrectomy surgical technique

The laparoscopic sleeve gastrectomy was performed with the patient in the supine position. After establishing pneumoperitoneum and trocar placement, the greater curvature of the stomach was mobilized beginning approximately 4–5 cm proximal to the pylorus and extending to the angle of His. Dissection was carried out along the greater curvature, with careful division of the gastrocolic and gastrosplenic ligaments, while preserving the phrenoesophageal fat pad and the gastric vessels along the lesser curvature throughout the procedure in order to minimize the risk of postoperative complications, including reflux and sleeve migration. A 36–40 French bougie was placed along the lesser curvature to calibrate the gastric sleeve. The stomach was divided vertically using sequential applications of a linear stapler, starting 2–5 cm from the pylorus and proceeding toward the angle of His, maintaining a safe distance from the esophagus. The staple line was inspected for hemostasis and integrity, and reinforcement was performed at the surgeon’s discretion. The gastric antral posterior ligament and the vascular supply along the lesser curvature were preserved to maintain anatomical integrity. The procedure concluded with leak testing and confirmation of hemostasis.

Roux-en-Y gastric bypass surgical technique

Laparoscopic Roux-en-Y gastric bypass was performed with the patient in the supine position under general anesthesia. After sterile preparation, a stab incision was made at Palmer’s point and a Veress needle was inserted to establish pneumoperitoneum. A 5 mm incision was made in the left upper quadrant (LUQ), and the abdomen was entered with a 5 mm 30-degree laparoscope. Additional ports were placed: 5 mm supra-umbilical, right upper quadrant, 15 mm right lateral, 5 mm epigastric (for liver retraction), and an extra 5 mm LUQ port, all under direct visualization. Attention was first directed to the stomach. The gastrohepatic ligament was divided using a harmonic energy device, and the stomach was lifted anteriorly with dissection carried into the lesser sac. A 30 cc balloon was inflated just below the gastroesophageal junction to calibrate the pouch. A 60 mm GIA stapler with green loads and Seamguard reinforcement was used to fashion a small gastric pouch by firing transversely across the stomach and then vertically up to the angle of His, resulting in a pouch of approximately 30 mL volume. The omentum and transverse colon were elevated, and the ligament of Treitz was identified. The small bowel was measured approximately 80 cm distal to the ligament of Treitz and transected with a GIA stapler (white load); the mesentery was divided with a harmonic scalpel. The biliopancreatic limb was marked with a clip. The distal bowel was then measured an additional 100 cm, and a jejunojejunostomy was fashioned at this point. Enterotomies were created in the antimesenteric borders of both the Roux limb and the biliopancreatic limb. A side-to-side anastomosis was performed using a 60 mm GIA stapler (white load), and a second white load was fired between the proximal Roux limb and distal common channel limb to create a wide, “W”-shaped anastomosis. The enterotomy was closed with an additional 60 mm GIA white load. The mesenteric defect was closed with a running 2 − 0 silk suture, in accordance with best practices to reduce internal hernia risk. The proximal Roux limb was brought over the transverse colon in an antecolic fashion. An enterotomy was made at the end of the Roux limb, and a small opening was created at the end of the gastric pouch. An orogastric tube was placed and the anvil was passed through the pouch opening. The 15 mm port was extended, and an EEA stapler was introduced into the abdomen. The stapler was passed through the enterotomy into the bowel, and the spike was brought out through the antimesenteric wall. The spike and anvil were approximated, and the EEA stapler was fired to create an end-to-side gastrojejunostomy. The roux limb enterotomy was closed using a GIA 60 mm white load. The procedure concluded with an upper endoscopy to evaluate the GJ anastomosis. The scope was easily passed into the pouch, through the anastomosis, and into the roux limb. A leak test was performed to ensure no leak, after which the endoscope was withdrawn.

Results

A total of 113 patients who underwent either p-RYGB or c-RYGB between 2015 and 2023 were included in the analysis. Of these, 49.6% (n = 56) underwent p-RYGB and 50.4% (n = 57) underwent c-RYGB. In the p-RYGB group, 76.8% (n = 43) were female while 77.2% (n = 44) of the c-RYGB group were female. Mean age was 44.57 years in the p-RYGB group and 44.7 years in the c-RYGB group (Table 1).

Table 1.

Baseline patient demographics and risk factors at the time of surgery as well as incidence of MU

		Total	Primary RYGB	Conversion RYGB	Statistical Significance
		N = 113	N = 56	N = 57
Age (years)
	Mean	44.64	44.57	44.7	p = 0.93
	Range	18.01 - 67	18.01 - 66.00	21.4 - 67.0
Sex
	Female	87	43	44	p = 0.96
	Male	26	13	13	p = 0.96
Race
	White	21	16	5	p = 0.007
	Black or African American	88	39	49	p = 0.037
	Asian	0	0	0
	Hispanic	1	1	0	p = 0.31
	Other	3	0	3	p = 0.082
Pre-Op Weight (lbs)
	Mean	291.57	286.49	296.57	p = 0.94
	Range	195 - 500	195 - 500	212 - 457
Pre-Op BMI (kg/m2)
	Mean	48.22	47.23	49.19	p = 0.94
	Range	35.11 - 83.11	35.11 - 83.11	42.15 - 75.55
ASA Class
	I	0	0	0
	II	6	4	2	p = 0.39
	III	101	50	51	p = 0.97
	IV	6	2	4	p = 0.41
Pre-op Risk Factors
	GERD	75	34	41	p = 0.21
	NSAID Use	18	10	8	p = 0.58
	Smoking	17	8	9	p = 0.82
	Diabetes Mellitus	32	15	17	p = 0.72
Incidence of MU		26	12	14	p = 0.69

No significant differences were observed between groups in terms of preoperative weight, BMI, ASA classification, or preoperative risk factors. However, there were significant differences in racial distribution: 69.6% of p-RYGB patients identified as Black or African American compared to 86% of the patients in the c-RYGB group (p = 0.037). Furthermore, 28.6% of p-RYGB patients identified as White while 8.8% of the c-RYGB group’s patients identified as White (p = 0.007).

The mean preoperative weight was 291.57 pounds in the p-RYGB group and 286.50 pounds in the c-RYGB group (p = 0.94). The corresponding mean BMIs were 48.22 and 47.23, respectively. Marginal ulceration was observed in 21.4% of patients in the p-RYGB group and 24.6% in the c-RYGB group (p = 0.69), thus demonstrating no statistically significant difference.

Discussion

Patients may elect for conversion from SG to RYGB for various reasons. The most common of such reasons is suboptimal weight loss which accounts for nearly 70% of these revision cases. Subsequently, GERD and dysphagia account for the majority of additional reasons patients opt for SG conversion to RYGB [7–9]. Hany M. et al. found c-RYGB showed a similar safety profile to p-RYGB, with no significant differences in rates of complications, mortality, and reoperation [7]. Their study reported a slightly lower, though not statistically significant, incidence of marginal ulcers in c-RYGB (2.7%) compared to p-RYGB (3.9%).

In contrast, a multi-institutional study by Anderson et al. demonstrated a significantly higher incidence of MUs in the c-RYGB group (13.6%) compared to the p-RYGB group (5.8%) [5]. They noted a shorter time to ulcer development and a twofold higher MU incidence in c-RYGB patients, despite similar risk factor profiles. They proposed several factors contributing to increased MU development including the following: refractory GERD [10], larger sized gastric pouches with increased acid production [11], crossed staple lines inhibiting adequate tissue perfusion, higher smoking rates, and “retained antrum syndrome”—referring to the situation when the remnant stomach retains a nearly intact antrum potentially resulting in increased acid secretion due to ongoing exposure to an alkaline environment.

In this single-institution retrospective cohort study, no significant difference in MU incidence was observed between p-RYGB and c-RYGB following SG. Our findings align more closely with those of Hany et al. and differ from the results reported by Anderson et al. We speculate that our outcomes may be influenced by consistent surgical technique, as all procedures were performed by one bariatric group using standardized methods. RYGBs were performed with a bougie or 30 cc balloon to fashion a small gastric pouch, an 80–120 cm alimentary limb, and a 100–150 cm biliopancreatic limb. SGs were also performed in a standardized fashion, preserving the phrenoesophageal fat pad and gastric vessels along the lesser curvature—factors which may contribute to maintaining adequate perfusion of the future gastric pouch. All patients received aggressive postoperative acid suppression protocols. Additionally, differences in the patient population may have contributed to our observed MU rates, though both groups had similar rates of smoking, GERD, NSAID use, and diabetes. Variations in postoperative surveillance and ulcer prophylaxis practices between institutions may also affect the timing and reporting of MU incidence.

Race may be an additional factor influencing MU risk. A retrospective Swedish study by Wennerlund et al. reported an increased risk of MUs in patients with immigrant backgrounds, although specific ethnic groups were not detailed [6]. Two MBSAQIP-based studies suggested that African American race may be associated with increased MU risk after RYGB [11, 12]. Our patient population, which included a high percentage of Black or African American patients, may differ from those in comparable studies, potentially contributing to variation in outcomes. These findings underscore the need for further investigation into racial and ethnic disparities in MU development.

The discrepancy between our findings and previous literature highlights the multifactorial nature of MU pathogenesis which involves mechanical, ischemic, and acid-related components. Continued research is essential to better understand and prevent this complication, especially in the context of bariatric revision surgery. Identifying patient-specific and technical risk factors will be critical for refining surgical strategies and improving long-term outcomes.

This study has several important limitations. First, the retrospective, single-center design and modest sample size (n = 113) limit statistical power and generalizability, consistent with prior bariatric surgery studies that have noted similar constraints in real-world and single-institution cohorts. The absence of a formal sample size calculation or power analysis raises the possibility that the study may be underpowered to detect clinically meaningful differences in marginal ulcer incidence. In addition, the absence of multivariable adjustment is a limitation, and unmeasured confounding may influence our results. Second, the potential for selection bias and unmeasured confounding is inherent to observational research. Although demographic and clinical variables were compared between groups, residual confounding may persist due to incomplete data on important risk factors such as prior NSAID use, PPI adherence and H. pylori status. Third, the study relied on electronic health record data, which may be subject to missing or misclassified outcomes, as has been described in other large cohort studies. The follow-up period and completeness of outcome ascertainment may be limited, potentially underestimating the true incidence of marginal ulcers and related complications. Fourth, the single-center setting and predominance of a specific patient population may restrict the external validity of the findings, as outcomes may differ in other geographic regions, practice settings, or among more diverse populations. Finally, the study did not capture patient behavioral factors, such as medication adherence, dietary habits, or physical activity, which are known to influence postoperative outcomes but are difficult to measure retrospectively. Despite these limitations, the study provides valuable comparative data on marginal ulcer rates after primary versus conversion Roux-en-Y gastric bypass in a real-world clinical setting. Future multicenter, prospective studies with larger sample sizes and more comprehensive confounder control are needed to confirm these findings and further elucidate risk factors for marginal ulceration.

Future research should focus on multicenter prospective studies with standardized definitions, surgical protocols, and postoperative management strategies to better elucidate MU risk factors across diverse populations and clinical settings. Greater understanding in this area will support the development of targeted perioperative strategies aimed at reducing MU incidence and enhancing long-term success in bariatric surgery.

Conclusion

In this single-institution retrospective cohort study, we observed no significant difference in the rates of marginal ulcers between patients who underwent primary Roux-en-Y gastric bypass and those who underwent conversion Roux-en-Y gastric bypass from a sleeve gastrectomy. These findings contrast with recent studies that have reported an increased rate of marginal ulcers in patients undergoing conversion RYGB procedures [5].

It is possible that even with the anatomical changes after sleeve gastrectomy, the perfusion of the gastric pouch isn’t significantly impacted enough to cause a higher rate of marginal ulceration. The discrepancy between our findings and those of previous studies underscores the multifactorial nature of marginal ulcer risk and the importance of individualized patient assessment and management. In addition, this study highlights the need for further research to better understand the underlying mechanisms and risk factors for marginal ulcer development in different surgical contexts. Understanding these differences is essential for optimizing patient outcomes and tailoring perioperative management strategies.

Authors’ contributions

All authors contributed equally to the manuscript and approved the manuscript as it is written.

Funding

This research received no specific grant from any funding agency in the public or commercial sectors.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

Our study adhered to the Declaration of Helsinki and was approved by our institutional IRB, the Mercy Fitzgerald Medical Center IRB. Informed consent was not required, as this was a retrospective cohort analysis.

Consent for publication

Not applicable.