Laparoscopic Gastric Band Erosion: A Retrospective Analysis and Review of Management

Introduction

In the United States, obesity continues to be an epidemic. Over the past two decades, obesity rates have drastically increased, and now nearly 33% of adults are obese. In 2010, not a single state had a prevalence of obesity <20%, and 36 states had a prevalence of at least 25%. ¹ Surgical options for obesity remain a mainstay of treatment. These options include gastric bypass, sleeve gastrectomy, and gastric banding.

Gastric banding, first introduced in the early 1990s by Belachew, has been frequently used in the treatment of morbidly obese patients. ² Gastric banding involves placement of an implantable device around the proximal stomach, which can be performed laparoscopically. The appeal of this restrictive procedure is weight reduction without the need for anatomical manipulation or resection.

Despite the weight loss success of patients undergoing laparoscopic gastric banding, there are some long-term complications associated with this procedure. Some of the more common causes of revision, or at times, reoperation include band erosion, slippage or band intolerance, and esophageal and pouch dilation. Some of the less common causes of reoperation are band or port leaks, intolerable gastroesophangeal reflux disease (GERD) or dysphagia, bowel perforation, and bowel obstruction. In addition, patient dissatisfaction from either inadequate weight loss or food intolerance can prompt revision to another restrictive/malabsorptive procedure.^3,4

The incidence of gastric band erosion ranges from 0.5% to 11%. ⁵ The definitive cause of gastric band erosion is unclear, but there are several hypotheses, including damage to the gastric wall during implantation, infection of the band or port site, and overfilling of the band. ⁶ When reviewing the literature regarding management of this complication, there is no consensus, and many different management options exist. Egberts et al. performed a large review of 25 studies in which 231 erosions were identified. The most common treatment was removal of the band with repair of the stomach and later band replacement. ⁷ Here, we describe the gastric band erosions that have occurred at our bariatric center of excellence, and the management of these patients and their outcomes.

Methods

All patients who had undergone laparoscopic gastric banding between 2006 and 2009 were reviewed following approval from the institutional review board. The pars flaccida technique was the method of insertion in each case. We identified the total number of gastric bands placed during those 3 years at our institution, which includes an outpatient surgery center. We then identified all patients who underwent reoperation for gastric band erosion, all of which were within our main institution. No band removals took place at the outpatient surgery center. One additional patient, whose initial band was placed at an outside hospital, was managed in our hospital and is included in this review. We believe this review has identified all gastric band erosions, but any patient who may have been evaluated elsewhere may have been missed.

The operative records for each patient with gastric erosion were reviewed, and the technique of band removal, whether closure of the defect and/or placement of a drain, was noted. Other demographics collected were age, sex, presenting symptoms, the manufacturer of the band type removed, length of hospital stay at band removal, initial body mass index (BMI), BMI at band removal, and the total number of days the band was in place before removal.

Results

A total of 2,570 patients underwent laparoscopic gastric banding between the years 2006 and 2009, and reoperation occurred for 17 gastric band erosions, giving an erosion rate of 0.6%. All patients in whom the erosion occurred had their band placed laparoscopically. Removal was also completed using a laparoscopic technique. Of the bands placed at our institution, the mean number of days the band was in place prior to removal due to erosion was 484 (range 12–1,054); the average BMI at placement of the band was 44.5 (range 30.2–60), and at removal was 35.7 (range 23–46.9). A total of eight REALIZE bands and nine lap-bands were removed at our institution from 11 females and six males with an average age of 52 years.

The most common presenting symptoms included abdominal pain, nausea, vomiting, port-site infection, and dysphagia. Of these, the most common was port-site infection, affecting 8 of 17 (47%) patients. Of the patients with a port-site infection, the mean number of days when they presented with erosion was 494 (range 295–828). Seven (41%) had abdominal pain, and six (35%) had nausea and/or vomiting. Only two (13%) patients had symptoms of dysphagia at the time of diagnosis (Table 1).

All patients underwent operative management of the gastric band erosions, which consisted of removal of the band and port device, and either primary closure of the defect with drain placement or intraluminal placement of a drain alone. Of the 17 patients managed at our institution, nine were managed with intraluminal drain placement (53%) and eight were managed with primary closure of the defect and drain placement (47%). Postoperatively the patients were all fed on postoperative day 1 or 2, and the drains were removed as an outpatient during follow-up. All of the patients had follow-up as an outpatient, and no postoperative complications to date have been noted (see Table 2).

Discussion

When thinking about gastric band erosion, it is important to be vigilant for early clinical signs. When a patient first presents with complaints of nausea, abdominal pain, food intolerance, or other symptoms of band intolerance, we first start by removing some of the fill from the band. If symptoms persist, next we proceed with either an upper gastrointestinal series (UGI) or a computed tomography (CT) scan, looking for the possibility of slippage, pouch dilation, esophageal dilation, or erosion. If this still fails to yield a diagnosis, we then perform an esophagogastroduodenoscopy (EGD). If concern is high for erosion, we bypass the imaging and proceed straight to endoscopy, as this is the gold standard, and CT or UGI can miss erosions.

The published rates of gastric band erosion in the literature range from 0.5% to 11%. Our institution's rate of gastric band erosion is comparable. In reviewing the literature, we noted many different management techniques for gastric band erosions, which include open, laparoscopic, and endoscopic techniques, with and without placement of drains, or conversion to gastric bypass or sleeve gastrectomy at time of band removal. Techniques described include removing the band through a separate gastrotomy site and primary repair, the theory being that one can allow closure without placement of sutures through friable tissue, as well as provide a safer removal without manipulation of the densely inflamed tissue. Also, direct removal of the band with placement of omental plug has been described. ⁵ In Europe, an endoscopic gastric band cutter seems to be the treatment of choice for band removal. ⁶ Although this system is not available in the United States, endoscopic scissors and piecemeal removal can still be employed to remove the eroded band endoscopically. It is recommended that this procedure be performed in the operating suite in the event that the need for conversion to a laparoscopic or open approach arises. For those patients that have the buckle of the band visible during EGD, endoscopic removal is a good alternative to laparoscopy and may prevent further adhesions and potential injury during laparoscopic removal. ⁸ Whatever technique is chosen, it should be based on surgeon comfort level with laparoscopy/endoscopy. At our institution, it is surgeon preference to perform laparoscopic removal.

We found that simply removing the band and port device with placement of an intraluminal drain was sufficient in the series of 17 patients that were managed at our institution. There was no difference in outcome when comparing whether or not the defect was primarily closed.

One patient presented within 1 week of having the gastric band placed and had symptoms of abdominal pain, fevers, and chills, which were consistent with peritonitis. When this patient was reexamined operatively, there was a notable gastric defect with free gastric fluid within the abdomen. A laparoscopic repair with band removal, Graham patch, and drain placement was performed in this case. This erosion, because of its occurrence early in the postoperative period, is almost certainly related to a technical complication.

The remainder of the patients presented with late band erosion (>30 days), and when removing the band, we noted that a significant inflammatory reaction develops an inflammatory capsule around the band itself, and believe that this reaction is what prevents patients from presenting with peritonitis. In addition, it is this same intense inflammatory reaction and capsule that prevents patients from developing peritonitis following the removal of the band, which would hold true for any patient whose band was removed endoscopically as well. This is what we believe allows for safe band removal without closure of the gastric defect.

The exact cause of gastric band erosions is uncertain, and many hypotheses exist. We attempted to determine whether overfilling of the band was a causative factor, but all of the patients at the time of band removal had no fluid within the band. This is most likely related to the fact that these patients frequently present with abdominal pain, nausea, or vomiting to the office first, and the initial management is to remove some amount of fluid from the band. As their symptoms continue, the entire volume of fluid had been removed prior to the definitive diagnosis of erosion and removal of the band.

Port-site infections in our series of patients was the most frequent finding. Similar studies have shown that these infections may in some way contribute to the erosion of the band. ⁶ It is uncertain whether the port-site infection occurs first, which then allows the bacteria to travel toward the band and stomach along the port catheter, or if the erosion occurs first due to some visceral damage at the initial time of placement and the infection travels toward the port. ⁹ In the future, cultures of the port site may be of some advantage in determining which of these occurs first.

Conclusion

Management of gastric band erosion can be performed successfully laparoscopically when occurring either early or late. In addition, when erosion has occurred, our patients have done well with removal of the band and port system and drain placement in the area of the erosion. Our experience has also shown that when patients present with port-site infections, a work-up for erosion is indicated.