Introduction

Roux-en-Y gastric bypass (RYGBP) is considered the gold standard operation for severe obesity, with a low mortality rate [1, 2]. However, there is no consensus on whether the placement of a ring is beneficial or not. The ring seems to be related to long-term weight loss maintenance [3], but vomiting may occur as a consequence of ring slippage, which can lead to gastric pouch outlet stenosis (GPOS) [4]. Vomiting may also occur even when there is no gastric stenosis, due to the presence of the prosthesis [5, 6].

Non-banded gastric bypass seems to be a global trend [7], but there are many patients that already have a ring implanted and may present with vomiting, requiring prosthesis removal. Despite being an invasive method that promotes the risk of intra- and postoperative complications, postoperative discomfort, hospitalization, and late return to regular activities, surgical ring removal has been the standard approach (Table 1).

Table 1 Author’s table
Full size table

Endoscopic dilation with achalasia balloon has emerged as a less invasive option. This procedure promotes either rupture or stretching of the thread running inside the ring, resolving patient’s symptoms [4].

We aimed to evaluate the safety and efficacy of aggressive dilation as an outpatient procedure for the treatment of solid food intolerance after banded RYGBP without GPOS. Long-term weight regain after this procedure was also analyzed.

Methods

Sixty-three patients [45 women (71.4 %); mean age 42.4 years] with an average preoperative body mass index (BMI) of 42.4 kg/m2 who underwent either open (n = 40) or laparoscopic (n = 23) banded (silastic ring) RYGBP at various Brazilian bariatric surgery institutions took part in this study. Balloon dilation was performed at two institutions of gastrointestinal endoscopy and surgery in Brazil, with extensive experience in the management of bariatric surgery postoperative complications (Fig. 1, Table 2).

Fig. 1
figure 1

Flow chart of food intolerance treatment

Full size image
Table 2 Demographic data and descriptive analysis
Full size table

In this longitudinal prospective study, we evaluated the endoscopic management of food intolerance. From 2002 to 2011, the authors enrolled all patients who underwent banded RYGBP and were referred to us with food intolerance symptoms of greater than four vomiting episodes per week due to solid food. All patients had good dental health.

Exclusion criteria were gastrojejunal anastomotic stenosis, gastric stenosis due to ring slippage, or presence of a ring other than a silastic ring. The local institutional review board approved this study and all patients provided informed consent. Patients who refused endoscopic procedure or presented with any contraindications were referred for surgical ring removal.

Food intolerance was considered when patient presented with vomiting episodes and was confirmed by upper digestive endoscopy to be without GPOS, allowing standard 9.8 mm endoscope passage. The endoscopic procedure was then indicated based on the clinical symptoms (Fig. 2). Four patients were not suitable to undergo endoscopic dilation based on exclusion criteria: two cases of food intolerance due to ring slippage, one case of gastrojejunal anastomotic stenosis, and one had a ring other than a silastic ring which was removed by laparoscopy. The other three underwent a different endoscopic technique.

Fig. 2
figure 2

Gastric pouch endoscopic image evidencing presence of food due to delayed gastric emptying caused by the ring, despite endoscopic free passage

Full size image

Patients who had some degree of malnutrition received nutritional counseling with specific replacement of deficient nutrients, which usually involved dense liquid nutrition. After stability of minimal nutritional status, endoscopic procedure was performed.

The procedures were carried out in outpatient endoscopic units under deep sedation, with the presence of a consultant anesthesiologist. A standard flexible endoscope was used in all procedures. The ring area in the gastric pouch was dilated up to 30 mm, using a Rigiflex® balloon (Boston Scientific, Natick, MA), which was gradually inflated (maximum 20 psi), according to endoscopic technique as previously described [5].

The aim was to promote rupture or stretch of the thread running inside the ring, widening the luminal diameter (Fig. 3). The first 16 (25.4 %) procedures were performed under fluoroscopic guidance. After that, according to early experience, sessions were performed exclusively under endoscopic control. Each dilation session lasted 5 to 30 min, per protocol.

Fig. 3
figure 3

X-ray and endoscopic image of gastric pouch evidencing Rigiflex® balloon inflated revealing ring compression. X-ray and endoscopic image evidencing ring opened ring after few minute dilation

Full size image

When thread rupture or appropriate pouch diameter for symptoms resolution were not achieved after the first session, the procedure was subsequently performed every other week up to four sessions, until symptoms resolved. Patients who still showed four or more vomiting episodes after four sessions were considered treatment failures and referred to surgery. Patients presenting with up to two vomiting episodes per week after four sessions were considered to have partial symptom improvement. No vomiting episodes corresponded to complete symptom improvement. Symptom recurrence, procedure complications, and long-term weight regain were also evaluated.

Data analysis was carried out using one-way ANOVA, chi-square, and Tukey test with SPSS software, version 13. An alpha risk of 5 % was assumed for statistical significance.

Results

Endoscopic dilation for vomiting treatment was performed an average time of 35.5 months (range, 11 to 156 months) after banded RYGBP. At the time of the first session, mean BMI was 25.3 kg/m2 (range, 17.9 to 40 kg/m2; Table 2).

Number of dilations was indicated according to each patient’s symptom improvement, ranging from one to four. Four sessions were needed in 12 (19 %), three in 14 (22.2 %), two in 24 (38 %), and one in 13 (20.6 %) patients.

Symptom improvement was reached in 61 cases: 59 (93.6 %) demonstrated complete improvement and two (3.2 %) had partial improvement (Table 3). These two patients with partial improvement reported that symptom relief was adequate and opted to quit further dilations. As the prior surgery was a laparotomic RYGBP, the two patients who failed to achieve improvement were referred for surgical ring removal by laparotomy, considering the existent adherences.

Table 3 Endoscopic treatment outcomes
Full size table

There was no symptom recurrence in the group of patients showing improvement (complete or partial) at a mean of 46.1 months follow-up. At this time, mean BMI increased from 25.3 to 27.8 kg/m2 (Table 4).

Table 4 BMI analysis before and after endoscopic dilation
Full size table

In terms of complications related to the procedure, three (4.7 %) cases of bleeding were found and treated with adrenaline solution injection, hospitalization (ranging from two to four days) for observation, liquid diet, sucralfate, double dose proton pump inhibitor administration, endoscopic control, and hematologic tests. There was no need for blood transfusion in any case. Two (3.2 %) intragastric ring erosions were identified and treated by endoscopic removal using scissors and foreign body forceps [8], and one (1.6 %) case of pneumoperitoneum managed with hospitalization for clinical treatment, which included antibiotic therapy and nothing per oral for 2 days. After clinical improvement, liquid diet was initiated, being the patient discharged after two more days (Table 5).

Table 5 Mild complications due to aggressive balloon dilation
Full size table

Discussion

Although silastic ring placement is associated with long-term weight control and maintenance [3, 7, 9], its placement is still a matter of controversy since it may cause complications such as intraluminal erosion, slippage, gastric stenosis, and food intolerance without GPOS.

Currently, non-banded gastric bypass seems to be a global trend [7], but there are a large number of patients with an implanted ring that may present with gastric stenosis or the newly defined food intolerance without GPOS. Both conditions may lead to vomiting, excessive weight loss, and malnutrition, which are unfavorable factors for surgical ring removal. Despite this, surgery has been the standard approach at many institutions worldwide.

In 2006, Stubbs et al. presented a prospective study in which 5 % of 415 banded gastric bypass patients underwent surgical ring removal due to major restriction to eating [10]. Surgical intervention in these cases is more challenging once ring adherence to the gastric pouch is considered, with a greater risk of perforation and bleeding [11].

These ring complications have been treated by endoscopy at our medical service with high success rates, low morbidity, and no mortality [4]. We also described the use of endoscopic dilation with a 30-mm Rigiflex® balloon as treatment of GPOS due to ring slippage [5].

In the current study, food intolerance was not related to GPOS but was also treated with endoscopic dilation. This was a safe outpatient approach that avoided surgical procedures and their associated complications in 96.8 % of patients.

Despite the safety of the endoscopic procedure, the occurrence of balloon distal slippage in three of the initial cases was responsible for the one case of pneumoperitoneum and three digestive bleeding during balloon dilation. However, such events can be avoided through gradual balloon inflation using a manometer and by firmly holding the balloon catheter [4].

Although discussed in the literature (Table 1) [6, 10–13], patients presenting with food intolerance are a poorly defined group as there is no consensus to characterize the clinical presentation and give a proper diagnosis. We defined food intolerance as absence of GPOS, allowing 9.8 mm endoscope passage with no difficulty through the ring area. Nevertheless, this group may present with solid food intolerance, major dysphagia, and more than four vomiting episodes per week [5, 8].

Balloon dilation consists of attempting to rupture or stretch the internal ring thread, leading to symptom improvement. However, if the ring is closed, as in an adjustable gastric band with a locker and no internal tying, dilation should not be performed. Gastrointestinal ischemia at the ring area is another contraindication due to the risk of perforation [14].

Concerning the authors’ experience, balloon dilation may cause weight regain and, in some cases, should not be performed [4]. In cases like VBG, where the gastric pouch is involved by a Marlex mesh, endoscopic dilation has resulted in mild food intolerance improvement but with early recurrence. If the ring has a lock with no internal thread, as in adjustable gastric bands, balloon dilation is not indicated [15]. Thus, the authors restricted this procedure indication only for cases involving a silastic ring with internal thread.

If the thread is not ruptured, the balloon will at least cause stretching, leading to a wider gastric pouch diameter at the ring site. It is not possible to determine precisely whether the thread is going to rupture or stretch. Silk thread is more likely to rupture than polyester thread. As it is more elastic, polypropylene thread tends to stretch during balloon dilation [4].

Procedures were performed under deep sedation because 30 mm balloon inflation usually causes abdominal pain, which may persist for several minutes after balloon deflation. When patients presents with shoulder and back pain, pneumoperitoneum can be in the differential diagnosis. After performing 16 procedures under fluoroscopy, endoscopic guidance was then elected. This technique proved to be safe, with no cases of pneumoperitoneum in those performed without fluoroscopic guidance. Additionally, this procedure avoids radiation exposure and allows the dilation session to be an outpatient procedure [16].

Endoscopic guidance reveals effective dilation during the procedure, once the extrinsic compression is no longer observed. Immediately after balloon deflation, an increase in the diameter of the gastric pouch is seen. As a collateral effect, the gastrojejunal anastomosis is also dilated because the balloon has a 10 cm length and the gastric pouch a 5 to 7 cm length. This could lead to dumping syndrome in some cases, which can be improved with clinical management.

Weight regain of patients presenting with intraluminal ring erosion is described in the literature [8]. Similar weight regain was expected after balloon dilation and surgical ring removal [8, 11], due to interruption of the restrictive function of the ring. However, in this study, patient’s BMI only increased from 25.3 to 27.8 kg/m2. This indicates that non-banded RYGBP also achieves good results and that silastic ring restriction is not the exclusive determinant factor in terms of weight loss maintenance.

As an outpatient procedure, the technique described here has advantages when compared to inpatient surgery, including reduced cost, less stressful environment, and more rapid patient discharge. In addition, this minimally invasive procedure has a high success rate with more than 96 % of patients reporting complete or partial symptom resolution, low morbidity, and no mortality. Despite the two patients in this study who failed to achieve improvement and had the ring removed by laparotomy, currently at our service, the laparoscopic procedure has been our first choice when surgical ring removal is needed.

Complications related to the procedure are infrequent and can be managed without the need for reoperation.

In conclusion, food intolerance is a condition that may occur even when GPOS is absent. Breaking the paradigm of surgical resolution, aggressive endoscopic dilation for the treatment of food intolerance is as a viable, safe, and minimally invasive alternative, with a low rate of weight regain.