The One Anastomosis Gastric Bypass Is a Suitable Alternative to Roux-en-Y Gastric Bypass in Patients with Body Mass Index ≥50 kg/m²: A Propensity Score-Matched Analysis

Body mass index (BMI) ≥50 kg/m² is more challenging for the metabolic bariatric surgeon because of a thicker abdominal wall, more visceral fat, and hepatomegaly by liver steatosis [1]. These factors can lead to a longer operation time and more postoperative complications [2].

Sleeve gastrectomy (SG), Roux-en-Y gastric bypass (RYGB), and one anastomosis gastric bypass (OAGB) are the most commonly performed metabolic bariatric procedures worldwide [3]. These different procedures all have their specific indications, contraindications, and procedural advantages.

In patients with BMI ≥50 kg/m², the OAGB may be more preferable compared to RYGB as less lifting of the intestines is necessary because OAGB has a longer pouch. In addition, OAGB has only one anastomosis, requiring less maneuvering space. Only a limited number of studies compared OAGB and RYGB in patients with BMI ≥50 kg/m² [4‒7]. Limitations of these previous studies are poor matching, a small number of patients, a limited follow-up duration, or no report of mid- or long-term complications. Therefore, this study aimed to investigate midterm outcomes in a matched population after OAGB and RYGB in patients with BMI ≥50 kg/m² during 5 years of follow-up in terms of weight loss, remission of comorbidities, and complications.

This is a retrospective cohort study in a nonacademic teaching hospital in the North of the Netherlands. All adult patients with BMI ≥50 kg/m² undergoing a primary OAGB or RYGB between 2015 and 2017 were included.

This study was approved by the Medical Ethical Committee (RTPO Leeuwarden, nWMO 2022 0007). Data were extracted from electronic patient records. All patients gave written informed consent for the use of their data.

Briefly, the creation of the pouch in OAGB started at the crow’s foot, resulting in a length of approximately 12 cm, with a biliopancreatic limb (BPL) between 150 and 250 cm. An antireflux suture was placed [8]. The pouch in RYGB was approximately 6 cm long, with a BPL between 120 and 200 cm and an alimentary limb (AL) of 50–150 cm. The length of the limbs was not uniform between the different surgeons and was primarily based on BMI, traction on the bowel, adhesions, and space to maneuver. In both procedures, the gastrojejunal anastomosis was made with a linear stapler and measured approximately 3 cm in length. The remaining defect was closed using either V-Loc absorbable sutures or interrupted Vicryl sutures. All patients received intravenous antibiotics perioperatively and the integrity of the gastrojejunal anastomosis was tested by methylene blue and air test. Mesenteric defects were not routinely closed during both procedures.

Remission of comorbidities was evaluated by patient-reported changes in treatment: no change, less treatment (partial remission), or no treatment anymore (total remission). Total remission of diabetes mellitus was considered with an HbA1c <48 mmol/mol after more than 6 months without medication.

Weight loss was expressed in kilograms, BMI, the percentage excess weight loss, and total weight loss (%TWL). TWL <20% was considered as a nonresponder on metabolic bariatric surgery (MBS). The cutoff point of 20% is based on current literature about successful weight loss after MBS [9].

Complications were categorized as short-term (≤30 days postoperative) and midterm (>30 days postoperative and <5 years). The Clavien-Dindo classification was used to classify all complications, with grades I and II considered minor and III as major complications [10]. While the Clavien-Dindo classification traditionally applies to postoperative complications, we have extended its application to grade midterm complications as well. Specifically, grade I denotes complications managed with conservative treatment without medication, grade II involves conservative treatment with medication, grade III entails invasive treatment (endoscopic, radiological intervention, or operation), grade IV designates life-threatening complications requiring invasive intervention, and grade V signifies complications that, if left untreated, would lead to death within 24 h.

Reflux disease was defined as one or more episodes of reflux. Treatment of reflux was initiated with lifestyle recommendations to which medication was added in case of insufficient relief. Patients who experienced reflux of fluid in their mouth or lungs during the night for at least two nights a week and/or experienced reflux in their mouth/lungs multiple times during the day when stooping with no effect of conservative treatment could opt for conversion to RYGB after being informed by their attending surgeon. Gastroscopy was conducted in cases where patients reported epigastric pain/discomfort or exhibited other signs indicative of an ulcer. Gastroduodenoscopy was not routinely performed to demonstrate the presence of biliary reflux.

In total, 158 patients underwent OAGB and 32 patients RYGB. Baseline characteristics differed slightly between both populations; therefore, a 1:1 propensity score matching was performed which resulted in two nearly identical cohorts of 28 patients, matched by age, gender, BMI before surgery, and comorbidities (hypertension, diabetes mellitus, obstructive sleep apnea syndrome [OSAS], reflux, and asthma/chronic obstructive pulmonary disease [COPD]).

Analysis of endpoints was performed as follows. All patients were analyzed for short-term complications. Patients with at least one annual follow-up visit were analyzed for midterm complications. Patients who did not attend their follow-up visit after 5 years were interviewed by telephone to complete the data. For analysis of remission of comorbidities, only patients with a follow-up visit after 5 years were included.

Continuous data are presented as mean ± standard deviation or median (interquartile range) in case of skewed distribution. Categorical data are presented as total numbers (percentages). The t-test was used for normally distributed data and the Mann-Whitney U test for skewed distributed data. Categorical data were analyzed with a chi-square test. A two-sided p value of ≤0.05 was considered statistically significant. All statistical analyzes were performed using SPSS Statistics version 28. GraphPad (Boston, MA, USA) was used to visually depict the data.

The baseline characteristics of the matched and unmatched populations are shown in Table 1. BMI at the preoperative screening was slightly higher in patients undergoing RYGB compared to OAGB (52.9 kg/m² versus 51.1 kg/m²; p = 0.012) (Table 1).

Two patients underwent a RYGB instead of an OAGB because of a complicated procedure: one bleeding of the pouch requiring resection of the lower part of the pouch and one stenosis at the anastomosis.

Adherence to follow-up visits in patients after OAGB and RYGB were respectively 96% and 82% in year 1 (p = 0.084), 89% and 89% in year 2 (p = 1.000), 86% and 86% in year 3 (p = 1.000), and 64% and 82% in year 4 (p = 0.131). Patients with at least one follow-up visit who missed the follow-up visit after 5 years were interviewed by phone. Therefore, 5-year data were available from 79% of the patients with an OAGB and 82% of the patients with a RYGB (p = 0.737).

No significant differences in TWL were seen between OAGB and RYGB during all 5 years of follow-up (Fig. 1). Also, no differences in the proportion of responders (>20% TWL) to MBS were found after 5 years (91% after OAGB versus 87% after RYGB; p = 0.578) (Table 2).

No significant differences were found in the partial or total remission of hypertension, diabetes mellitus, OSAS, and asthma/COPD with both procedures (Table 3).

No differences were seen in short-term complications between both procedures. However, minor midterm complications were more present after OAGB (50% versus 18%; p = 0.011) due to the high incidence of reflux (50% versus 7%; p < 0.001). This resulted in conversion to RYGB in 7% of the patients after OAGB (major complication).

The number of patients with major midterm complications did not differ between both procedures. Conversion to RYGB was the only major complication following OAGB. In contrast, patients who underwent RYGB experienced a variety of major complications, including internal herniation, anastomotic stenosis, diagnostic laparoscopies, perforation at the entero-enterostomy, intra-abdominal abscess, and laparoscopic gastrostomy (Table 4). Data from the unmatched groups about weight loss, remission of comorbidities, and complications can be found in Supplementary Tables 5–7 (for all online suppl. material, see https://doi.org/10.1159/000542681).

This study compared midterm outcomes after OAGB and RYGB in patients with BMI ≥50 kg/m² in a well-matched population. Both procedures resulted in similar weight loss, remission of comorbidities, short-term complications, and major midterm complications; making OAGB a suitable alternative to RYGB for patients with a BMI ≥50 kg/m².

No significant differences in weight loss were seen between OAGB and RYGB during 5 years of follow-up. Compared to the literature, only a limited number of small studies compared OAGB and RYGB in patients with BMI ≥50 kg/m² [4‒7]. Two studies found more weight loss after 2 and 3 years in patients with OAGB [5, 6]. One study found no significant differences in weight loss over 10 years [4].

After 5 years, TWL was 32% in our patients with OAGB; this is less than 37.7% and 40.8% reported in previous studies [4, 11]. Most of our patients (61%) had a BPL of 180 cm, compared to BPL of 200 cm and 250–350 cm in these previous studies [4, 11]. A recent systematic review found more weight loss after OAGB with a BPL of 200 cm, compared to 150 cm [12]; however, this was accompanied by more nutritional deficiencies.

After 5 years, TWL was 30% in our patients with RYGB. Salte et al. [13] found a comparable percentage of 28.9%. Also, higher percentages around 33.1% and 37.2% are reported in the literature [4, 14]. This can potentially be explained by differences in BPL and AL lengths and thus in lengths of the common channel. Yet, up till now, no consensus on the ideal limb lengths of the RYGB has been reached [15‒17].

For both procedures, there is a lack of consensus on the optimal limb lengths. Additionally, the role of gastric pouch length in influencing surgical outcomes lacks sufficient data and warrants further research to determine if and how variations in pouch length can optimize patient outcomes. While limb length is crucial, the success of metabolic bariatric surgeries also depends on other factors such as hormonal changes, damage to the vagus nerve, microbiome alterations, and more [18, 19]. Given these complexities, future studies should focus on determining the optimal lengths for BP and AL limbs, investigating the impact of gastric pouch lengths, and exploring the combined effects of various physiological changes. In the meantime, comparing traditional RYGB with standard OAGB remains a prudent approach to establish a foundational understanding while awaiting more detailed research insights.

We found no significant differences in the remission of hypertension, diabetes mellitus, OSAS, and asthma/COPD 5 years after OAGB or RYGB; however, the numbers were small. Soong et al. [4] also found no differences in the remission of hypertension and diabetes mellitus after 5 years, but the proportion of patients with remission of OSAS was higher after OAGB (70.3% versus 39.7%; p = 0.001). Asthma/COPD was not investigated in this previous study.

Our study found no differences in short-term complications after both procedures. However, the sample size in our study is small. Other small studies in patients with BMI ≥50 kg/m² varied from no significant differences to fewer or even more major short-term complications after OAGB [4‒6].

Minor midterm complications were more present after OAGB compared to RYGB due to the high incidence of reflux, contributed by our definition of reflux. With only one episode of reflux, a patient already met our definition of reflux disease. Our conversion rate of 7% is high compared to the literature, varying from 0% to 1.6% [4, 20, 21]. During the inclusion period of our study, preoperative reflux was not an exclusion criterion for OAGB. Based on a previous study of our group we now refrain from offering OAGB to patients who report reflux symptoms or are on proton pump inhibitors, as the revision rate in this group tends to be significantly higher [8]. Consequently, these percentages may be lower in contemporary practice.

Based on our findings, OAGB emerges as a good alternative to RYGB for patients with BMI ≥50 kg/m². Notably, OAGB yields comparable outcomes regarding weight loss, remission of comorbidities, short-term complications, and major midterm complications. In a recent Delphi consensus, ≥90% of the experts agreed that the OAGB is an acceptable and suitable one-stage procedure in patients with BMI ≥50 kg/m² [22]. As mentioned in the introduction, metabolic bariatric procedures have their own technical (dis)advantages in patients with a higher BMI. Hence, it is imperative to consider the surgeon’s preference as a crucial factor in the decision-making process.

To provide an impartial and comprehensive assessment of the most suitable metabolic bariatric procedure for patients with BMI ≥50 kg/m², the SG should also be considered. However, in our cohort, only 15 patients underwent SG between 2015 and 2017. Therefore we compared existing literature about results of SG.

Soong et al. [4] found more major postoperative complications after RYGB (4.8%) compared to SG (0.5%) and OAGB (0.8%). OAGB resulted in better weight loss with TWL of 40.8% after 5 years compared to SG (35.1%) and RYGB (37.2%). Remission of dyslipidemia was the best after SG, but remission of diabetes did not differ between the procedures. The overall revision rate was the lowest after SG (2.6%) compared to RYGB (8.1%) and OAGB (6.9%).

Two systematic reviews compared SG and RYGB in patients with BMI ≥50 kg/m² [4, 23]. RYGB showed better results regarding weight loss and control of dyslipidemia after 12 months [4, 23]. However, weight loss after 2 years was equal for both procedures [4, 23]. Maroun et al. [24] retrospectively analyzed long-term outcomes of SG, RYGB, and biliopancreatic diversion with duodenal switch (BPD-DS). BPD-DS achieved the highest TWL after 5 years (38.4%), compared to RYGB (26.3%) and SG (13.8%) [24]. However, this success was associated with more surgical and nutritional complications [24].

Three studies compared SG with OAGB in patients with BMI ≥50 kg/m² [25‒27]. All studies found more weight loss after OAGB after 1, 2, and 3 years [25‒27]. However, data about major short-term complications were different in all studies: no difference in fewer or even more complications after OAGB [25‒27].

The strength of this study is the 5-year follow-up duration including data on weight loss, remission of comorbidities, and short- and midterm complications. In addition, not many previous studies compared the midterm outcomes after OAGB and RYGB in patients with BMI ≥50 kg/m².

Our study also had some limitations: first, the retrospective design. Second, compared to our current clinical practices, there have been some notable changes. Nowadays, OAGB is not conducted in patients with reflux, and mesodefects are perioperatively closed in both procedures. Third, gastroduodenoscopy was not routinely performed to confirm the presence of biliary reflux. As a result, aligning our data with existing literature and clinical practices, where gastroduodenoscopy is standard in cases of reflux, poses a challenge. Nevertheless, our conversion rate does offer insight into the prevalence of therapy-resistant reflux complaints. In future investigations, it might be interesting to perform a randomized control trial between OAGB and RYGB in patients with BMI ≥50 kg/m² because this patient category is excluded or poorly represented in previous studies.

For clinical decision-making, it is important to know that both procedures give an excellent result after 5 years of follow-up, thus leaving the choice for one of the procedures to the preference of the patient or the surgeon.

We would like to thank Maaike K. Veenstra, BSc, and Gaatsche K. Visser-De Witte, BSc, for their help with the collection of data.

This study was approved by the Medical Ethical Committee (RTPO Leeuwarden, nWMO 2022 0007). Data were extracted from electronic patient records. All patients gave written informed consent for the use of their data.

This study was not supported by any sponsor or funder.

The salary of L.v.d.L. is funded by Medical Center Leeuwarden. The funder had no role in the design, data collection, data analysis, and reporting of this study.

L.v.d.L. participated in data acquisition, analyzed data, and drafted the manuscript. D.S. assisted with the data analysis and the data interpretation. L.J.M.d.H., A.P.v.B., and M.E. assisted with data interpretation, revised the manuscript, and gave the final approval for the version to be published.

All data generated or analyzed during this study are included in this article and its supplementary material files. Further inquiries can be directed to the corresponding author.