EUS-guided liver palpation as a screening tool for advanced fibrosis and cirrhosis in patients with suspected metabolic dysfunction-associated steatotic liver disease: a pilot study.

BACKGROUND AND AIMS: Endoscopic liver "palpation" can be performed by indenting the liver surface under EUS. Indentation depth is measured with the use of sonographic calipers. We hypothesized that fibrotic livers are more difficult to indent, and that indentation can accurately predict liver fibrosis staging. We compared EUS-guided liver palpation and conventional screening modalities in patients with suspected metabolic dysfunction-associated steatotic liver disease. METHODS: This was a cross-sectional pilot study. Consecutive patients at 3 hospitals from 2021 to 2023 underwent EUS-guided palpation with liver biopsy. Liver palpation was compared with fibrosis-4 index (FIB-4), aspartate transaminase to platelet ratio index (APRI), nonalcoholic fatty liver disease fibrosis score (NFS), and transient elastography in predicting fibrosis staging on histology. Area under the receiver operating characteristic curve analysis was performed. RESULTS: Seventy-three patients were included. Mean age was 49.1 years, and 71.2% were female. Mean body mass index was 41.1 kg/m.2 Indentation depth was negatively correlated with fibrosis stage (Kruskal-Willis test, P < .0001). EUS palpation demonstrated c-statistics of 0.79 and 0.95 in discriminating advanced fibrosis and cirrhosis, respectively. EUS liver palpation was superior to NFS in predicting advanced fibrosis (P = .0057) and superior to APRI and NFS in predicting cirrhosis (P = .0099 and P = .045, respectively). EUS palpation was not significantly different from FIB-4. EUS palpation was superior to transient elastography in predicting cirrhosis (P = .045). When optimal cutoffs were used, indentation measurement ≤3.5 mm yielded 100% predictive value for ruling in advanced fibrosis, and ≥4.0 mm yielded 100% predictive value for ruling out cirrhosis. CONCLUSIONS: EUS liver palpation is a novel, accurate, and easy-to-use screening tool for advanced fibrosis and cirrhosis in patients with metabolic dysfunction-associated steatotic liver disease.

Magnetic cholecysto-duodenal anastomosis.

INTRODUCTION: There is a growing interest in minimally invasive endoscopic gallbladder drainage therapies. Unlike stenting, a technology based on magnetic compression could theoretically produce a durable drainage anastomosis without residual foreign material. This study aimed to evaluate the safety and technical feasibility of a cholecysto-duodenal magnetic compression anastomosis. MATERIAL AND METHODS: We performed a survival study of two Yorkshire pigs. Duodenal magnets were deployed endoscopically; reciprocal gallbladder magnets were placed laparoscopically, and the magnets were coupled. Pigs underwent serial endoscopy documenting magnet expulsion and evolution of cholecysto-duodenal anastomosis creation. Necropsies and histological evaluation were performed. Primary endpoints were technical success and safety. Secondary endpoints included anastomosis integrity, patency, and histologic characteristics. RESULTS: Magnets were successfully delivered and coupled. Patent, leak-free anastomoses formed by day 4. Magnets were expelled by day 10. All anastomoses were widely patent at one month (mean diameter 15 mm). Necropsy showed the absence of adhesions affecting the anastomosis. Histology showed complete re-epithelialization without inflammation or foreign body reaction. CONCLUSIONS: Magnetic cholecysto-duodenal anastomosis for gallbladder drainage appears safe and feasible in the animal model. The anastomoses are patent, leak-free, and without inflammation from the presence of foreign material. Technical modifications for magnet delivery under endoscopic ultrasound (EUS) guidance are currently underway.

Partial jejunal diversion using an incisionless magnetic anastomosis system: 1-year interim results in patients with obesity and diabetes.

BACKGROUND AND AIMS: Most patients with type 2 diabetes mellitus have obesity. Studies show that bariatric surgery is superior to medical treatment for remission of type 2 diabetes mellitus. Nevertheless, very few patients undergo surgery, and a less-invasive endoscopic alternative is desirable. METHODS: This was a single-arm first-in-human pilot study designed to evaluate the technical feasibility, safety, and clinical performance of the incisionless magnetic anastomosis system (IMAS) to create a partial jejunal diversion (PJD). Ten patients with obesity and type 2 diabetes mellitus, prediabetes, or no diabetes were enrolled. A PJD to the ileum was attempted in all patients under general anesthesia. The IMAS was delivered through the working channel of a colonoscope, with laparoscopic supervision. The patients were not required to participate in an intensive lifestyle/diet management program. Endoscopic visualization of the anastomosis was obtained at 2, 6, and 12 months. Patient weight, glycemic profile, and metabolic panels were acquired at 0.5, 1, 2, 3, 6, 9, and 12 months. RESULTS: A PJD was created in all patients with no device-related serious adverse events. The anastomosis remained widely patent in all patients at 1 year. Average total weight loss was 14.6% (40.2% excess weight loss at 12 months). A significant reduction in glycated hemoglobin level was observed in all diabetic (1.9%) and prediabetic (1.0%) patients, while reducing or eliminating the use of diabetes medications. CONCLUSIONS: Permanent anastomosis for PJD was created in all patients with the IMAS. This resulted in improvement in measures of hyperglycemia and progressive weight loss. (Clinical trial registration number: NCT02839512.).

Effect of vagotomy during Roux-en-Y gastric bypass surgery on weight loss outcomes.

BACKGROUND: During Roux-en-Y gastric bypasses (RYGB), some surgeons elect to perform a vagotomy to reduce symptoms of gastro-oesophageal reflux (GER). Routine vagotomy during RYGB may independently affect weight loss and metabolic outcomes following bariatric surgery. We aimed to determine whether vagotomy augments percent excess weight loss in obese patients after RYGB. METHODS: We examined the effect of vagotomy in 1278 patients undergoing RYGB at our institution from 2003 to 2009. Weight and percent excess weight loss (%EWL) were modelled at three months and annually up to five years using a longitudinal linear mixed model controlling for differences in age, gender, initial body mass index (BMI), ideal body weight, and presence of vagotomy. RESULTS: Vagotomy was performed on 40.3% of our cohort. Vagotomy patients had significantly lower initial BMI (46.4±6.2 vs. 48.3±7.7kg/m(2), p<0.001), but there were no other significant differences at baseline. The strongest predictor of %EWL over time was initial BMI, with lower BMI patients exhibiting greater %EWL (p<0.001). Age and gender effects were also significant, with younger patients (p<0.04) and males (p<0.002) attaining greater %EWL. Vagotomy had no effect on %EWL in either simple or multiple regression models. CONCLUSION: Our series suggest that vagotomy does not augment %EWL when performed with RYGB.

Laparoscopic adjustable gastric banding and hypoglycemia.

Obesity is commonplace, and surgical treatment usually includes Roux-en-Y gastric bypasses (RYGBs). RYGBs have the most documented side effects including vitamin deficiencies, rebound weight gain, and symptomatic hypoglycemia; fewer series exist describing hypoglycemia following other bariatric operations. We reviewed all patients undergoing laparoscopic adjustable gastric banding (LAGB) at our institution between 2008 and 2012. Three patients were identified to have symptomatic hypoglycemia following LAGB. Mean time from surgery was 33 months (range 14-45 months), and mean weight loss was 32.7 kg (range 15.9-43.1 kg). None of the patients had preexisting diabetes. Therefore, symptomatic hypoglycemia should be investigated irrespective of bariatric operation.

Endoscopic suturing for transoral outlet reduction increases weight loss after Roux-en-Y gastric bypass surgery.

BACKGROUND & AIMS: Weight regain or insufficient loss after Roux-en-Y gastric bypass (RYGB) is common. This is partially attributable to dilatation of the gastrojejunostomy (GJ), which diminishes the restrictive capacity of RYGB. Endoluminal interventions for GJ reduction are being explored as alternatives to revision surgery. We performed a randomized, blinded, sham-controlled trial to evaluate weight loss after sutured transoral outlet reduction (TORe). METHODS: Patients with weight regain or inadequate loss after RYGB and GJ diameter greater than 2 cm were assigned randomly to groups that underwent TORe (n = 50) or a sham procedure (controls, n = 27). Intraoperative performance, safety, weight loss, and clinical outcomes were assessed. RESULTS: Subjects who received TORe had a significantly greater mean percentage weight loss from baseline (3.5%; 95% confidence interval, 1.8%-5.3%) than controls (0.4%; 95% confidence interval, 2.3% weight gain to 3.0% weight loss) (P = .021), using a last observation carried forward intent-to-treat analysis. As-treated analysis also showed greater mean percentage weight loss in the TORe group than controls (3.9% and 0.2%, respectively; P = .014). Weight loss or stabilization was achieved in 96% subjects receiving TORe and 78% of controls (P = .019). The TORe group had reduced systolic and diastolic blood pressure (P < .001) and a trend toward improved metabolic indices. In addition, 85% of the TORe group reported compliance with the healthy lifestyle eating program, compared with 53.8% of controls; 83% of TORe subjects said they would undergo the procedure again, and 78% said they would recommend the procedure to a friend. The groups had similar frequencies of adverse events. CONCLUSIONS: A multicenter randomized trial provides Level I evidence that TORe reduces weight regain after RYGB. These results were achieved using a superficial suction-based device; greater levels of weight loss could be achieved with newer, full-thickness suturing devices. TORe is one approach to avoid weight regain; a longitudinal multidisciplinary approach with dietary counseling and behavioral changes are required for long-term results. ClinicalTrials.gov identifier: NCT00394212.

Implementing a bariatric surgery program.

The number of people who are obese or morbidly obese is increasing in the United States. Currently, the most effective means of losing a substantial amount of weight and maintaining the weight loss is bariatric surgery, and health care providers, especially those in surgical services, must be able to safely care for patients undergoing these surgeries. Financial implications of starting a bariatric surgery program and the ongoing costs must be fully understood and supported by both administrators and employees. Special equipment and supplies are needed to handle the higher weight of bariatric patients, and careful planning is required for adequate medical and nursing expertise, staffing, equipment, supplies, facility resources, and patient support services.

Does laparoscopic gastric banding create hiatal hernias?

BACKGROUND: We hypothesized that laparoscopic adjustable gastric band (LAGB) placement might result in the development of a hiatal hernia (HH) over time. The objective of our study was to determine whether HHs develop after LAGB in the setting of a university hospital. METHODS: We retrospectively reviewed all outcomes for consecutive LAGB patients in our institutional, longitudinal prospective bariatric surgical database to identify those patients without evidence of a HH at LAGB placement, who subsequently underwent delayed HH repair. RESULTS: From 2005 to 2009, 695 gastric bands were implanted. Twelve patients (1.72%) were identified who had no radiographic or intraoperative evidence of a HH at LAGB placement and who subsequently underwent HH repair at re-exploration. Patients presented 18 ± 10 months after band placement. Of these patients, 75% presented with gastroesophageal reflux disease or food intolerance (50% with gastroesophageal reflux disease alone). Also, 2 presented with acute pain due to band slippage and 1 with chronic pain and vomiting. In 50% of the patients, revision procedures detected the HH at operation despite negative preoperative studies. CONCLUSION: In our series, a significant HH developed in 1.7% of LAGB patients who had no clinically identifiable HH at LAGB placement. Persistent dysphagia after band deflation requires careful inspection of the hiatus during surgical revision, even in the absence of radiologic depiction of HH, and might represent an underlying etiology of LAGB dysfunction. This complication, along with esophageal dilation and annular pouch dilation, might represent a constellation of conditions with a common etiology. From the results of our small series, we raise the question of the existence of chronic backpressure created by LAGB restriction and accounting for these complications.

Resection or reduction? The dilemma of managing retrograde intussusception after Roux-en-Y gastric bypass.

BACKGROUND: Retrograde intussusception (RI) at the jejunojejunostomy can occur after Roux-en-Y gastric bypass (RYGB). Although this complication is rare, it has been encountered more frequently as the number of bariatric procedures have increased. Little data is available to assist surgeons with the optimal management of this condition. Our objectives were to identify the risk factors for RI after RYGB and report on outcomes after surgical intervention at a tertiary academic surgical unit. METHODS: We used our prospective longitudinal institutional bariatric surgical database to identify patients with post-RYGB RI from 1996 to 2011. RESULTS: We identified 28 post-RYGB RI cases. The median interval between RYGB and RI was 52 months, and the median percentage of excess weight loss was 75%. Patients presented with acute symptoms in 36% of the cases. All patients underwent surgical exploration, including resection and revision of the jejunojejunostomy (46%) or operative reduction with or without enteropexy (54%). Those undergoing resection had a longer hospital stay but similar 30-day complication rates. At a median follow-up of 9 months, only 1 recurrence was documented. CONCLUSIONS: RI is a rare and late complication of RYGB and typically occurs after significant weight loss. In the presence of ischemia or nonreducible RI, resection and revision of the jejunojejunostomy is recommended. In less acute patients, laparoscopic management with reduction and/or enteropexy offers a reduced hospital length of stay while maintaining equivalent morbidity and low recurrence compared with resection.

Endoscopic sclerotherapy for the treatment of weight regain after Roux-en-Y gastric bypass: outcomes, complications, and predictors of response in 575 procedures.

BACKGROUND: Weight regain after Roux-en-Y gastric bypass (RYGB) is common. Endoscopic sclerotherapy is increasingly used to treat this weight regain. OBJECTIVES: To report safety, outcomes, durability, and predictors of response to sclerotherapy in a large prospective cohort. DESIGN: Retrospective analysis of a prospective cohort study of patients with weight regain after RYGB. PATIENTS: A total of 231 consecutive patients undergoing 575 sclerotherapy procedures between September 2008 and March 2011. INTERVENTIONS: Single or multiple sclerotherapy procedures to inject sodium morrhuate into the rim of the gastrojejunal anastomosis. MAIN OUTCOME MEASUREMENTS: We report weight loss, complications, and predictors of response. We also used Kaplan-Meier survival analysis and log-rank test to compare time to continuation of weight regain after sclerotherapy in patients undergoing a single versus multiple sclerotherapy procedures. RESULTS: At 6 and 12 months from the last sclerotherapy procedure, weight regain stabilized in 92% and 78% of the cohort, respectively. Those who underwent 2 or 3 sclerotherapy sessions had significantly higher rates of weight regain stabilization than those who underwent a single session (90% vs 60% at 12 months; P = .003). The average weight loss at 6 months from the last sclerotherapy session for the entire cohort was 10 lb (standard deviation 16), representing 18% of the weight regained after RYGB. A subset of 73 patients (32% of the cohort) had greater weight loss at 6 months (26 lb, standard deviation 12), representing 61% of the weight regained. Predictors of a favorable outcome included greater weight regain and the number of sclerotherapy procedures. Bleeding was reported in 2.4% of procedures and transient diastolic blood pressure increases in 15%, without adverse health outcomes. No GI perforations were reported. CONCLUSIONS: Endoscopic sclerotherapy appears to be a safe and effective tool for the management of weight regain after RYGB.

Impact of lap-band size on weight loss: does gender matter?

BACKGROUND: Laparoscopic adjustable gastric band (LAGB) has gone through major design modifications to improve clinical endpoints and reduce complications. Little is known, however, about the effects of LAGB size on clinical outcomes, or whether outcomes differ based on gender. We set out to examine the impact of band size on surgical weight loss, reoperations, comorbidity resolution, and compare outcomes within gender. METHODS: We reviewed our prospectively collected longitudinal bariatric database between 2008 and 2010, and compared patients with BMI 35-50 kg/m(2) who had undergone LAGB with the LAP-BAND® APS to those who had the larger APL. Those patients with initial BMI > 50 kg/m(2) were excluded to reduce any possible selection bias which favors larger band use in such subjects. RESULTS: Three hundred ninety-four patients met our inclusion criteria; 230 (58 %) in the APS group and 164 (42 %) in the APL group. Female patients in APS group experienced significantly higher percentage excess body weight loss at 6 months, 1 year, and 2 years in comparison to female patients in APL group (p < 0.001 for all time points). In contrast, a reverse pattern was observed for male patients. No significant differences were observed between the groups regarding frequency of band adjustments, complications, or comorbidity resolution. CONCLUSIONS: Male patients might benefit from APL bands, in contrast to female patients who appear to experience superior weight loss with the smaller APS bands. This study provides the first set of evidence to facilitate surgical decision making for band size selection and highlights differences between genders.

Establishing a reproducible large animal survival model of laparoscopic Roux-en-Y gastric bypass.

BACKGROUND: The advent of metabolic surgery and the increasing focus on the substantial resolution rate of type 2 diabetes after laparoscopic Roux-en-Y gastric bypass (LRYGB) call for additional fundamental investigations as to the mechanisms behind this effect. These investigations require an adequate animal model. Our objective was to develop a reproducible survival model of LRYGB performed in a large animal at a tertiary university hospital. METHODS: LRYGB was performed on 11 Yorkshire pigs that where then followed for 6 weeks. The operative time, morbidity, and mortality were recorded for each case. Necropsy was performed, and the anastomoses were harvested and inspected for leaks. RESULTS: The surgical technique and difficulties are carefully described. Of the 11 pigs, 10 survived to the end of the study period. The 1 death was from intraoperative cardiac dysrhythmia. The postoperative complications consisted of a postoperative febrile episode in 2 pigs. The mean initial weight was 31.5 ± 3.4 kg. The mean operative time was 214 ± 71 minutes. No anastomotic leaks were identified at necropsy or on histologic examination of anastomoses. The mean weight gain at the end of the study period was .8 ± 1.4 kg compared with an expected 17.5 kg weight gain. CONCLUSION: We have described an effective survival porcine model of LRYGB that can be consistently reproduced. This will enable additional investigation into the complex physiologic mechanisms that control hunger, weight loss, and the development, as well as resolution, of type 2 diabetes, potentially leading to the development of novel, targeted bariatric procedures and diabetic treatments.

Obesity overview: epidemiology, health and financial impact, and guidelines for qualification for surgical therapy.

The aim of this article is to describe the context in which this issue of Gastrointestinal Endoscopy Clinics of North America is established. The authors review the current worldwide dimensions and trends of the obesity epidemic; associated mortality and comorbid diseases including diabetes, cancer, cardiovascular disease and obstructive sleep apnea; the financial impact of obesity; and current national and international guidelines for referral and qualification for surgical treatment of obesity.

Endoscopic management of gastrogastric fistulae does not increase complications at bariatric revision surgery.

BACKGROUND: Gastrogastric fistula (GGF) is a challenging complication of primary obesity surgery that often leads to revision surgery. The impact of prior endoscopic intervention on subsequent surgical revisional outcomes remains unknown. We present the largest series of Roux-en-Y gastric bypass GGF with subsequent surgical revision of fistulae to date. METHODS: A database of bariatric surgical revisions performed at a single institution was collected. The cohort was divided between patients with and without attempted endoscopic fistula closure prior to surgical revision. Thirty-day morbidity and mortality was the primary outcome. RESULTS: Thirty-five cases of revision were performed for GGF. Of the 35 cases, 22 patients had attempted endoscopic closure prior to surgical revision while 13 patients went directly to surgical revision. In the endoscopy group, two minor complications and seven major complications occurred (total 9 of 22; 40.9%). In the surgery only group, three minor complications and three major complications occurred (total 6 of 13; 46.1%). No deaths occurred. CONCLUSION: Prior attempts at endoscopic fistula closure do not lead to increased surgical complications at the time of surgical revision.

Band revision versus Roux-en-Y gastric bypass conversion as salvage operation after laparoscopic adjustable gastric banding.

BACKGROUND: With the increasing popularity of laparoscopic adjustable banding (LAGB), reoperations to treat band complications have become increasingly more common. Few studies, however, have documented the outcomes of such revisional procedures. The purpose of the present study was to compare the different reoperative approaches after LAGB placement. The present study was conducted at a tertiary referral academic medical center in the United States. METHODS: We reviewed our bariatric database to identify patients who had undergone LAGB from 2004 to 2007 and had undergone either revision of their original band or conversion to Roux-en-Y gastric bypass. RESULTS: Of the 66 patients who met our inclusion criteria, 47 (71.2%) had undergone revision and 19 (28.8%) conversion. The patient demographics were similar between the 2 groups. The interval to revision or conversion was 22 and 28 months, respectively. Band slippage was the most common surgical indication in the revision group, and inadequate weight loss was the most common in the conversion group. Although patients in the revision group had experienced good weight loss before their reoperation, the conversion group had not. However, the conversion group experienced good weight loss after their reoperation, with an average excess body weight loss of 48%. CONCLUSION: Patients who have experienced successful weight loss with LAGB and experience band complications will have satisfactory outcomes with band revision, maintaining their excess body weight loss. These data suggest that patients with inadequate weight loss after LAGB can do well after conversion to Roux-en-Y gastric bypass.

Gastrojejunal stoma diameter predicts weight regain after Roux-en-Y gastric bypass.

BACKGROUND & AIMS: Weight regain after Roux-en-Y gastric bypass (RYGB) is associated with reductions in health status and quality of life. We evaluated whether gastrojejunal stoma diameter is a risk factor for weight regain after RYGB. METHODS: We examined data collected over 4 years from consecutive patients referred to a tertiary care bariatric center for upper endoscopy after RYGB. We used linear regression analysis to determine the association between the gastrojejunal stoma diameter and weight regain. We applied a logistic regression model using clinical and endoscopic parameters to develop a prediction rule for weight gain after RYGB. RESULTS: Among 165 patients included in our study, 59% had significant weight regain (≥ 20% of maximum weight lost after the RYGB) and 41% did not. The mean percentage of maximal weight lost after RYGB that was regained in the entire cohort was 30% ± 22%. Gastrojejunal stoma diameter was associated significantly with weight regain after RYGB surgery in univariate analysis (ß = .31, P < .0001). This association remained significant after adjusting for several known or purported risk factors for weight regain (ß = .19, P = .003). We developed a simple prediction rule for weight regain after RYGB using a 7-point scoring system that includes the gastrojejunal stoma diameter, race, and percentage of maximal body weight lost after RYGB; a cut-off score of 4 or more points had an area under receiver operating characteristic curve of 0.76 and a positive predictive value of 75%. CONCLUSIONS: Increased gastrojejunal stoma diameter is a risk factor for weight regain after RYGB and can be incorporated in a novel prediction rule.

Endoscopic foreign body removal for treatment of chronic abdominal pain in patients after Roux-en-Y gastric bypass.

BACKGROUND: Common endoscopic findings in patients who have undergone Roux-en-Y gastric bypass (RYGB) with chronic abdominal pain have included marginal ulceration, gastrogastric fistula, and jejunal erosion. However, suture or staples eroding into the gastric pouch can also contribute to abdominal pain. Redundant suture is typically regarded as a normal part of the postoperative anatomy. The objectives of the present study were to assess the effects of endoscopic foreign body removal of partially exposed sutures and staples in post-RYGB patients with chronic abdominal pain at a university hospital in the United States. METHODS: We performed a retrospective study of consecutive patients from January 2006 to July 2007. Post-RYGB patients with chronic abdominal pain underwent endoscopic foreign body removal of exposed sutures/staples. Pain scores were obtained before the procedure, immediately after the procedure, and at the telephone follow-up (median 7.2 months). RESULTS: Of 21 patients, 15 (71%) reported immediate symptomatic improvement. Specific endoscopic accessories were found to be more useful than others in managing the various foreign materials. Of the 21 patients, 15 (71%) were available for telephone follow-up. Of these 15 patients, 13 (87%) reported continued symptomatic improvement, with 9 (60%) reporting complete pain resolution and 4 (27%) reporting partial improvement. Eroded foreign material was seen in association with marginal ulcers in 3 patients (14%), gastritis in 7 patients (33%), and an inflammatory polyp in 1 patient (5%). CONCLUSIONS: Eroded suture and staples can cause chronic abdominal pain in post-RYGB patients. In symptomatic patients, visible suture or staples should be considered a potential etiology of chronic pain, instead of normal postoperative findings. Endoscopic foreign body removal might be of therapeutic benefit in these patients.

Endoscopic repair of gastrogastric fistula after Roux-en-Y gastric bypass: a less-invasive approach.

BACKGROUND: Gastrogastric fistulas (GGFs) are a well-known complication of Roux-en-Y gastric bypass. Surgical repair of such fistulas is technically difficult, with significant associated morbidity. The aim of the present study was to evaluate the efficacy of endoscopic GGF closure at a university hospital in the United States. METHODS: Patients with Roux-en-Y gastric bypass and confirmed GGFs on esophagogastroduodenoscopy or barium study. Endoscopic repair was performed with the EndoCinch suturing system (group 1) or clips (group 2). All patients were followed up in the outpatient clinic or interviewed by telephone at 1, 6, and 18 months after the procedure, then as indicated by symptoms. RESULTS: A total of 95 patients were included in the present series (group 1, n = 71, 75%; group 2, n = 24, 25%). The mean GGF size was significantly larger in group 1 than in group 2 (14.5 +/- 8.7 versus 7.7 +/- 6, P = .01). An average of 2.2 sutures or 3 clips (range 2-7) was used. Complete initial GGF closure was achieved in 90 patients (95%), with reopening in 59 (65%) an average of 177 +/- 202 days. The average follow-up was 395 +/- 49 days, with 22 patients lost to follow-up. Two significant complications were reported (bleeding and an esophageal tear). None of the GGFs with an initial size >20 mm remained closed during the follow-up period compared with 10 (32%) of the 31 fistulas