Gastric Bipartition with Functional Duodenum Exclusion (GBp-FDE) as a Possible Surgical Conversion of Sleeve Gastrectomy in Patients with De Novo GERD and Obesity Recidivism: Preliminary Results of a Case Series

Introduction

Obesity has reached alarming global prevalence, and metabolic bariatric surgery (MBS) is recognized as the most effective and durable treatment [1]. Several techniques are available, each with distinct mechanisms of action [2,3]. Sleeve gastrectomy is the most widely performed procedure worldwide [4,5], but it is frequently associated with new-onset or worsening gastroesophageal reflux disease. This occurs mainly due to the longitudinal gastrectomy, which reduces gastric compliance, increases intragastric pressure, and alters the esophagogastric junction anatomy [6–8]. A strong correlation has been established between sleeve pouch dilation, obesity recurrence, hiatal hernia (HH), and GERD after SG [9,10]. Conversely, SG preserves gastrointestinal continuity, allowing complete endoscopic evaluation and reducing nutritional deficiencies [11]. Long-term metabolic outcomes and weight loss are satisfactory [12], particularly in patients with body mass index (BMI) <50 kg/m2, although results are slightly inferior to those of RYGB [13].

RYGB was the most performed bariatric procedure worldwide until 2013, when sleeve gastrectomy (SG) started surpassing it in the United States [14]. Despite this shift, RYGB remains widely recommended, demonstrating effective weight loss and metabolic improvements in patients with morbid obesity [15]. It is considered the most effective surgical option for obese individuals with GERD [16,17]. The procedure excludes most of the stomach, duodenum, and proximal jejunum, limiting transoral endoscopic access to these segments [18–23]. For this reason, in regions with high gastric cancer prevalence, resection of the remnant stomach has been proposed as a preventive strategy [24].

Managing patients with GERD and recurrence of weight gain after SG remains challenging, as neither re-sleeve gastrectomy (Re-SG) nor RYGB adequately addresses both conditions. Re-SG may promote further weight loss (WL) in cases of gastric pouch dilation but is suboptimal for GERD control [25]. Conversely, while RYGB effectively treats GERD, it often fails to achieve sustained additional weight loss [26]. Ileal surgeries such as duodenal switch (DS), single anastomosis duodenum-ileostomy with sleeve (SADI-S) and one anastomosis gastric bypass (OAGB) promote further WL, but have questionable results in gastroesophageal reflux control [27]. Transit bipartition (TB) has emerged as a feasible surgical option to deal with obesity recidivism and GERD after SG [28], although studies addressing this operation in severe gastroesophageal reflux disease are lacking.

This case series examines patients presenting with moderate to severe de novo GERD and recurrent weight gain following SG, with associated HH and sleeve pouch dilation in most cases, for whom revision surgery was indicated. Inclusion criteria comprised obesity recurrence and confirmed endoscopic GERD diagnosis per the Lyon consensus [29], using the Los Angeles Endoscopic Esophagitis Classification (LA Classification) [30]. All patients declined operations that included gastrointestinal exclusions for varied reasons: 4 had a family history of digestive tract cancer, 2 had relatives with biliary tree gallstones, 3 knew individuals with malnutrition after other techniques associated with anatomical exclusions, and 1 was concerned about blind gastrointestinal segments. Considering these factors and after thorough discussion about surgical options, GBp-FDE was proposed as a revisional approach to address both GERD and obesity recidivism without anatomical exclusions [31,32]. In cases associated with gastric pouch dilation and/or HH, concomitant re-sleeve and HH repair were performed. The surgery consisted of adding a wide Roux-en-Y antroileostomy on the juxta pre-pyloric anterior gastric wall of the sleeved stomach, aiming to reduce intragastric pressure and alleviate reflux. This anastomosis was also intended to promote preferential gastric flow to the ileum, functionally excluding the duodenum. Duodenal diversion and ileal stimulation – enhanced by a 40% proportion biliopancreatic limb – are expected to support further weight loss via metabolic pathways [33,34].

Preoperative and 1-year postoperative assessments included clinical evaluation of gastroesophageal reflux symptoms using a validated GERD questionnaire [35], which was an adaptation to Portuguese language by Velanovich et al. The GERD questionnaire (Table 1) [36], translated and validated by the authors, aimed at easier comprehension by Brazilians, with the inclusion of regurgitation assessment. Upper endoscopy was performed at both time points to assess hiatal hernia – defined by the distance between the proximal gastric folds and crura constriction, during endoscopic withdrawal – erosive esophagitis per Los Angeles classification [30], and sleeve pouch morphology, including tortuosity, stenosis, or dilation (the latter inferred when endoscopic retroflexion was feasible). Radiological contrast studies were conducted pre- and postoperatively to evaluate HH, pouch tortuosity, stenosis, and pouch dilation (defined as total gastric volume >250 mL) [37]. Postoperative radiological series (300 mL of barium solution) also assessed the preferential contrast flow between the antroileostomy and the pylorus. Three patterns were considered: (1) unidirectional flow to the ileum; (2) preferential flow to the ileum, with minimal contrast passage to the duodenum; (3) split flow between the ileum and the duodenum. The first 2 patterns were interpreted as functional duodenum exclusion. Anthropometric parameters and comorbidity status were monitored throughout the study.

Case Reports

STATISTICAL ANALYSIS:

Continuous variables are summarized as mean±SD, median (P25–P75), minimum–maximum, and the 95% confidence interval of the mean. Categorical variables are presented as counts and percentages, with exact Clopper-Pearson 95% confidence intervals when informative. The number of missing observations is reported per variable.

Longitudinal comparisons (4 time points). Weight and BMI measured at initial, nadir, pre-revision, and 1-year follow-up were analyzed using one-factor repeated-measures ANOVA (within-subjects). Sphericity was evaluated with Mauchly’s test; when violated, degrees of freedom and P values were reported with Greenhouse-Geisser correction (and Huynh-Feldt as a sensitivity check). Effect size is reported as partial eta squared (ηp²). Planned pairwise comparisons between time points used paired t tests with Holm adjustment for multiple testing. Figures display the mean with 95% confidence intervals across time.

Before-after outcomes such as gastroesophageal reflux disease (GERD) and hiatal hernia (HH) were tested with McNemar’s test; in addition to the continuity-corrected chi-square, the exact binomial P value for discordant pairs was reported. A paired odds ratio with 95% confidence interval was computed using the Haldane-Anscombe 0.5 correction when needed. For GERD defined from endoscopy, Los Angeles grades B/C/D were treated as positive and grade A as non-confirmatory; sensitivity analyses considered suspected cases as positive (conservative) or negative (strict).

The distribution of contrast destination at postoperative seriography (ileum vs duodenum) was compared using an exact binomial test of the ileum proportion against a null of 0.50, with the corresponding exact 95% confidence interval. Results are illustrated with monochrome bar charts in percentages and reported in figure captions.

Descriptive summaries used available data (pairwise). Repeated-measures ANOVA and paired tests were performed on complete cases for the respective variables. No imputation was performed given the sample size and design. Significance and reporting. All tests were two-sided with α=0.05. Exact P values are reported. Where applicable, adjusted P values (Holm) and effect sizes (ηp² for repeated-measures ANOVA; paired odds ratio for binary outcomes; proportions with exact 95% confidence intervals) are provided. Analyses were conducted in R (R Foundation for Statistical Computing) using base stats (t tests, binom.test, mcnemar.test), ez (ezANOVA), and tidyverse/ggplot2 for data handling and visualization.

Results

Out of a total of 22 individuals, 12 were excluded from this study: 6 for presenting grade A or no esophagitis according to LA classification, 5 due to absence of obesity recidivism, and 1 for not complying with the FU protocol. In the remaining 10 individuals enrolled in this series, during 1-year FU there was a universal reflux symptoms improvement and decrease in validated GERD questionnaire scores (P=0.0059) (Figure 8). Endoscopic findings showed complete healing of erosive esophagitis in 8 patients and improvement in 2 who presented grade A esophageal erosions (P=0.008) (Figure 9), using the conservative rule that counts “suspected” as positive (Grade A esophagitis). These 2 individuals were asymptomatic. Hiatal hernia resolved in all 9 patients who presented HH preoperatively, decreasing from 100% before surgery to 0% at 1 year (P=0.002) (Figure 10). Table 2 shows the individual evolution of hiatal hernia and esophagitis after revision surgery. No marginal ulcer was observed, and the anastomoses remained widely patent (2.5–3 cm in diameter). In 90% of cases, the pylorus was spastic during endoscopic assessment, while in 10% it was partially spastic. The whole endoscopic view of postoperative phase is illustrated in Figure 11.

Radiological oral contrast studies revealed functional duodenal exclusion in all individuals, and unidirectional flow to the ileum was observed in 9 of them (P=0.021) (Figure 12A–12C). No case of split contrast flow was observed (Figure 13).

During 1-year follow-up weight and BMI dropped from 99.7±9.9kg (97.5 [94.5–108.8]) and 37.2±3.2 kg/m2 (37.0 [34.7–38.0]) to 64.5±8.4 kg (62.9 [58.2–70.2]) and 24.0±1.7 kg/m2 (23.8 [23.2–24.9]), respectively (Figures 14 and 15 shows mean weight and BMI evolution since primary operation until 1-year FU of revision). Mean%EWL was 108.0±11.4 (109.9 [102.1–115.0]) and%TWL was 35.2±6.6 (37.8 [33.7–38.4]), 1 year after revision (Figure 16).

All individuals improved their comorbidities, and reduced drugs need, 1 year after revision, summarized in Table 3. Surgical details are available in Video 1 (link).

Discussion

Although sleeve gastrectomy (SG) is the most commonly performed bariatric procedure worldwide [4,5], obesity recurrence and gastroesophageal reflux disease (GERD) remain major concerns postoperatively [38,39]. These complications may limit its indication and often require surgical revision. In this series, all patients met the updated Lyon consensus criteria for GERD [29] and exhibited recidivism of weight gain after primary SG.

Multiple factors contribute to GERD onset or its exacerbation following SG [40]. The marked reduction in gastric volume from longitudinal gastrectomy decreases gastric compliance and elevates intraluminal pressure [41,42]. Additionally, inadvertent injury to the crura and sling fibers of the lower esophageal sphincter may impair the region’s anti-reflux barrier [43,44]. The increased incidence of HH post-SG is well documented and may further drive GERD pathophysiology [10]. In this cohort, HH was present in 90% of patients, potentially contributing to de novo GERD.

These findings underscore the importance of actively assessing for sleeve pouch proximal migration and repairing HH when identified. Furthermore, abdominal fixation of the gastric pouch may help prevent HH formation, gastric rotation, or tortuosity – mechanisms implicated in GERD development [40,45–47].

Gastric dilation typically occurs 2–3 years after SG and, depending on its volume, may contribute to GERD development and recurrent weight gain [25]. A sleeve pouch volume >250 mL within the first 3 years after SG has been strongly linked to GERD and obesity recurrence [48]. This threshold was adopted in the present study, where 80% of patients demonstrated varying degrees of gastric chamber dilation on preoperative radiological assessment. These cases required re-sleeve gastrectomy as part of revisional surgery.

A large percentage of patients with prior SG and pathological GERD are asymptomatic [25]. In this cohort, 90% of individuals reported symptoms during the preoperative phase of revisional surgery, as assessed by a validated GERD questionnaire [35] – a finding that contrasts with the existing literature.

The interval between primary and revisional surgery in this series was longer than typically reported in the literature [49,50], yet did not appear to impact outcomes during the 1-year follow-up.

Although some authors report a 15–25% incidence of de novo GERD following primary RYGB [51,52], conversion to RYGB remains a common recommendation when GERD arises after SG, given its well-established efficacy [53]. However, in cases where GERD after SG is accompanied by obesity recurrence, RYGB may not be the optimal revisional approach due to its limited effectiveness regarding further weight loss [54]. In the clinical context of this study, it is essential to identify the main concern, as other bariatric procedures (eg, OAGB and SADI-S) may offer better obesity and comorbidities control, compared to Roux-en-Y gastric bypass (RYGB), despite their inferiority in GERD mitigation [49,55].

Transit bipartition has emerged as a viable surgical alternative for recurrent weight gain and mild GERD after SG [28]. Nonetheless, its efficacy in patients with severe GERD remains unsupported by consistent evidence. In this case series, the coexistence of obesity and moderate/severe GERD presented a complex challenge for surgical revision. Additionally, patients declined anatomical gastrointestinal exclusions for previously mentioned reasons.

Given this scenario, revision to GBp-FDE was selected, an operation described in the literature [31,32,56]. Anatomically, this technique resembles transit bipartition with some modifications, preserving the safety of antroenterostomies, already well-documented [57–60]. Considering the preliminary report of this conversion, there are no comparative studies with other revisional strategies.

This procedure was recently renamed by the authors as “gastric bipartition with functional duodenum exclusion” (GBp-FDE) to better describe its anatomical similarity with TB (ie, Roux-en-Y antroileostomy over a sleeve pouch) and its functional implications. The preferential flow through the anastomosis, functionally bypassing the duodenum [56], mimics the alimentary route of a long-pouch distal RYGB, which anatomically diverts the duodenum [61–63].

Long-pouch RYGB proved effective in managing gastroesophageal reflux and yielded superior mid-term weight loss compared to short-pouch RYGB [64,65]. The extended gastric pouch in GBp-FDE not only can support sustained weight reduction but also enhance nutritional safety by preserving gastric cellularity. This factor may optimize the gastric phase of digestion and reduce the risk of malabsorption and malnutrition, even when paired with an elongated biliopancreatic limb (BPL) [33,61,66]. Accordingly, in the proposed technique, the biliopancreatic limb was extended to 40% of the total small intestine [56], reaching the ileum and its metabolic and weight loss benefits [67,68], notably enhanced incretin release [33,69]. The additional weight loss also contributes to reduced intra-abdominal pressure, offering further mechanisms for GERD mitigation [70].

The prevailing functional duodenal exclusion observed in GBp-FDE is a novel concept in metabolic surgery, likely resulting from the proximity of a wide antroileostomy to a spastic pylorus [56]. The observed partial or complete pyloric spasm may stem from injury to the anterior vagus nerve branches – intramural within the antrum – during wedge resection of the anterior wall in this region, as part of anastomosis [71]. This hypothesis is an assumption that needs further investigation to be confirmed.

Moreover, disruption of the interstitial cells of Cajal (ICC) network during antral wedge resection can further contribute to pyloric dysmotility. A reduction in ICC density within the antrum has been associated with impaired pyloric function and spasm [72–77]. Further investigation is necessary to confirm this hypothesis. Although pyloric spasm has been observed in GBp-FDE, either complete or partial, it is not possible to infer its role in functional duodenum exclusion, considering the wide antrum-ileal anastomosis near the pylorus could solely promote it. Further investigation is demanded to clarify this.

Functionally, GBp-FDE differs from TB due to the anatomical configuration of the latter, which allows near-equal food flow through both the pylorus and gastroileostomy [78]. This enables the use of a shorter, fixed common channel while maintaining nutritional safety. However, preservation of the pylorus in food pathway in TB, combined with the ileal brake mechanism typical of ileal procedures [79], can impede effective gastric decompression in severe GERD cases, which is the scenario in this series.

In contrast, SADI-S and DS preserve the pyloric sphincter as the only path for ingested food. Consequently, intragastric pressure within the sleeve pouch remains elevated and possibly aggravated by the ileal break effect [80], potentially limiting GERD resolution in more advanced cases, which completely differs from the revision proposed in this case series, which prompts reduced intragastric pressure due to the wide gastrointestinal anastomosis.

Notably, 80% of patients in this series required a re-sleeve procedure due to gastric pouch dilation [25]. A 32-Fr bougie was used and the stapling was performed 2–3 mm lateral to it, followed by staple line invagination. This technique aimed to create a narrow sleeve pouch, that potentially accelerates gastric emptying via the gastroenterostomy, leading to gastric decompression – mechanisms that may contribute to improved GERD control and enhanced metabolic outcomes [81,82]. In this study, complete resolution of erosive esophagitis was achieved in 80% of patients, while all patients had symptoms improvement, reinforcing this hypothesis.

Duodenal exclusion may be a pivotal metabolic mechanism in bariatric surgery [83]. Elevated serum levels of fibroblast growth factor 19 (FGF19) and bile acids, along with partial restoration of postprandial GLP-1 peaks and reductions in GIP and glucagon responses, have been observed following duodenal isolation. This hormonal profile collectively enhances lipid and glycemic regulation [34,84–87]. However, anatomical duodenal exclusion via gastroduodenal partitioning – as seen in RYGB, OAGB, SADI-S, and DS – presents a limitation by obstructing routine endoscopic access to the biliopancreatic papillae [68,88,89]. The potential to achieve functional duodenal isolation, as proposed in this study, could be clinically significant. It preserves standard endoscopic access to the duodenum, as the spastic pylorus seems to yield to the forward pressure of the endoscope, successfully achieved in all patients in this series. Moreover, the absence of a bypassed stomach in GBp-FDE may be particularly advantageous in populations or ethnic groups at increased risk for gastric cancer [90].

In this study, during 1-year FU, the WL achieved resembled the one from other ileal surgeries when used as a rescue procedure after sleeve gastrectomy in individuals with obesity recidivism [27,28,49,55,91–94]. No distinct WL pattern was observed when minimal contrast flow to the duodenum occurs, suggesting the importance of a very preferential flow to the ileum as the goal to be reached in this operation. Again, the role of pyloric spasm in this physiology is not established, and the wide gastrointestinal anastomosis itself could “steal” all the alimentary flow.

Complete comorbidities control was achieved in 80% and partial in 20%, in this series. These results are consistent with the expected outcomes of other ileal metabolic surgeries [33,61,69] and mirrored the revisions of SG when converted to potent ileal operations [27,28,49,91,92].

Marginal ulceration was a concern in GBp-FDE due to its elevated incidence in gastrointestinal anastomoses involving large gastric reservoirs [95]. However, procedures utilizing a long, narrow gastric pouch (eg, OAGB [96], long-pouch-RYGB [61], and TB [60,97) have negligible rates of marginal ulcers. The usual extensive longitudinal gastrectomy in these operations, which removes or isolates most of gastrin- and acid-producing cells [98], may account for the low incidence of marginal ulceration and justify its absence in this series.

This case series is constrained by its retrospective design, limited sample size, and absence of a control group, reducing both clinical and statistical robustness. Furthermore, the lack of esophageal manometry hindered assessment of esophageal function, limiting insight into GERD pathophysiology following primary sleeve gastrectomy (SG) and its effect on motility. The absence of manometric data also precluded evaluation of the revisional procedure’s impact on esophageal motility and lower esophageal sphincter competence. Although postoperative grade A esophagitis has been considered as pathological reflux in this series (sensitivity analyses considered suspected cases as positive), lack of esophageal pHmetry compromised the accurate detection of pathological reflux, particularly in patients with this grade of erosive esophagitis or no esophagitis, after revisional surgery.

The observation of duodenal functional exclusion was inferred from indirect findings obtained via endoscopic and contrast radiological evaluations. Direct pyloric pressure measurements were not performed, representing a limitation in the assessment of pyloric function, warranting further investigations to confirm its occurrence and to clearly elucidate its underlying mechanisms and influence on duodenal diversion. Also, the role of pylorus spasm in functional duodenum exclusion cannot be inferred, as the exclusion may be provided solely by a huge gastric flow diversion towards the ileum, promoted by the wide gastroileal anastomosis.

The 1-year follow-up period is limited, hindering strong conclusions and compromising the nutritional safety of the proposed surgical revision; thus, the results of this case series should be regarded as preliminary. Validation through larger, long-term controlled studies, incorporating comprehensive gastroesophageal physiological assessment, is recommended.

Conclusions

The proposed GBp-FDE, as a revisional approach to SG in patients with de novo GERD and recurrent obesity, aims to decompress the SG pouch via wide gastroileostomy, leading to GERD control; to redirect most of the gastric flow to the ileum, enhancing weight loss and metabolic outcome; and to promote duodenal exclusion without anatomical partitioning, preserving thorough gastrointestinal endoscopic access and increasing metabolic benefits. Controlled studies are necessary to validate these findings.