Correlation of radiographic and endoscopic evaluation of gastrojejunal anastomosis after Roux-en-Y gastric bypass

Surgery for Obesity and Related Diseases 2 (2006) 617– 621

Original article

Correlation of radiographic and endoscopic evaluation of gastrojejunal anastomosis after Roux-en-Y gastric bypass Samuel Szomstein, M.D.a,*, Orit Kaidar-Person, M.D.a, Kristoff Naberezny, M.D.a, Marcia Cruz-Correa, M.D.b, Raul Rosenthal, M.D.a a

Bariatric Institute, Section of Minimally Invasive Surgery, Cleveland Clinic Florida, Weston, Florida b Department of Gastroenterology, Cleveland Clinic Florida, Weston, Florida Received January 18, 2006; revised May 28, 2006; accepted September 12, 2006

Abstract

Background: Anastomotic stenosis presents as one of the most common late complications in the postoperative period after bariatric surgery. It is often diagnosed by upper gastrointestinal series (UGIS) and/or upper endoscopy (UE). The aim of this study was to determine whether a correlation exists between the Gastrografin UGIS and UE findings in the determination of gastrojejunal anastomotic strictures after Roux-en-Y gastric bypass (RYGB). Methods: Between July 2001 and October 2003, all medical records of patients who underwent RYGB at our institution were retrospectively reviewed. The medical records of patients who underwent UE because of symptoms suggestive of gastric outlet obstruction and those of patients who were initially evaluated by Gastrografin UGIS before UE were evaluated further. Results: Of 535 morbidly obese patients who underwent RYGB, 52 (9.7%) had UE and were included in this study. The mean number of UEs performed per patient was 2.67. Of these 52 patients, 30 underwent Gastrografin UGIS before UE. The mean diameter of the anastomosis on the first UE was 5.97 mm and on Gastrografin UGIS was 6.83 mm. A good correlation was found between the Gastrografin UGIS and UE findings using Pearson’s correlation coefficient (0.44, P ⫽ .02) and single linear regression analysis using the endoscopic diameter as the outcome and radiographic findings as the predictor (␤ ⫽ 0.27, P ⫽ .025, 95% confidence interval 0.30 – 0.49). Conclusion: In our study, the Gastrografin UGIS findings correlated positively with the endoscopic gastrojejunal anastomosis findings in patients with anastomotic stricture who had undergone RYGB. © 2006 American Society for Bariatric Surgery. All rights reserved.

Keywords:

Gastrografin; Endoscopy; Obesity; Gastrojejunal anastomosis strictures; Roux-en-Y gastric bypass

Anastomotic stenosis presents as one of the most common late postoperative complications of Roux-en-Y gastric bypass (RYGB). An average of 12% of all bariatric patients develop this complication, with estimates ranging from 4% to 19% [1– 4]. Early detection of related complications can substantially reduce the morbidity and mortality [4]. TherePresented as a podium presentation in the South Florida Chapter of the American College of Surgeons, 2005, and the Annual Meeting of the American Society for Bariatric Surgery, 2005 *Reprint requests: Samuel Szomstein, M.D., Bariatric Institute, Section of Minimally Invasive Surgery, Cleveland Clinic Florida, 2950 Cleveland Clinic Boulevard, Weston, FL 33331. E-mail: [email protected]

fore, patients who present with symptoms suggestive of anastomotic strictures, such as nausea, vomiting, and abdominal pain, undergo costly investigations that usually consist of a Gastrografin upper gastrointestinal series (UGIS) and/or upper endoscopy (UE). Endoscopy is an invasive procedure that, although not frequently, may be associated with serious complications. Morbid obesity, in addition to imposing technical difficulties when performing the procedure, may be associated with a greater risk of developing these complications. Because of various considerations, symptomatic morbidly obese patients after RYGB often undergo various initial diagnostic studies. The aim of this study was to determine whether Gastro-

1550-7289/06/$ – see front matter © 2006 American Society for Bariatric Surgery. All rights reserved. doi:10.1016/j.soard.2006.09.004

618

S. Szomstein et al. / Surgery for Obesity and Related Diseases 2 (2006) 617– 621

the Gastrografin protocol implemented at Cleveland Clinic Florida. According to the protocol, all the UGIS were performed using the same distance and same magnification, allowing a more accurate comparison between films of the same patient at different times and standardization among different patients (Figs. 1 and 2). The protocol consisted of oral administration of 20 mL of Gastrografin solution followed by anteroposterior films at 5 and 10 seconds, pending emptying of the pouch. The same process was repeated for the right and left oblique projections, followed by an anteroposterior projection demonstrating contrast in the distal bowel. The radiologist selected and measured a single projection that, according to his judgment, best represented the anastomotic diameter. UE Examinations

Fig. 1. Gastrografin UGIS demonstrating patent anastomosis.

grafin UGIS could evaluate the diameter of an anastomotic stricture in correlation with the UE findings. Methods After approval by the institutional review board, all medical records of patients who had undergone RYGB between July 2001 and October 2003 at our institution were retrospectively reviewed. The medical records of patients who underwent UE because of symptoms suggestive of gastric outlet obstruction, including nausea, vomiting, and upper abdominal pain, and the records of patients who were initially evaluated with Gastrografin UGIS before UE were evaluated further.

The UE approximations of anastomotic diameters, which were performed by 3 endoscopists, involved comparing the diameter of the anastomosis to the diameter of a 9-mm Pentax EG endoscope (Figs. 3 and 4). Statistical Analysis Pearson’s correlation coefficient and linear regression analysis were used to evaluate the relationship between the radiographic and endoscopic findings in patients who underwent both studies. STATA, version 8.0, software (STATA 2003) was used for statistical analysis, with the endoscopic diameter as the outcome and the radiographic findings as the predictor of significant stricture.

Surgical Procedure All procedures were performed as laparoscopic antecolic, antegastric gastric bypass. The gastrojejunostomy was created using a single firing of a 45-mm linear cutter and 2.5-mm stapler between the posterior wall of the pouch and jejunum. The anterior wall of the anastomosis was hand sewn with a double row of running 2-0 polyglactin sutures. The omentum and small bowel mesentery were divided only in cases of anastomotic tension. If this step was required, the omentum was divided first, followed by division of the mesentery, if necessary, to further relieve the tension. The anastomosis was then tested for leakage using air insufflation and methylene blue. All patients received proton pump inhibitors postoperatively for 2 months. Gastrografin UGIS Gastrografin UGIS consisted of anteroposterior and lateral-posterior oblique projections performed according to

Fig. 2. Gastrografin UGIS demonstrating complete anastomotic obstruction.

S. Szomstein et al. / Surgery for Obesity and Related Diseases 2 (2006) 617– 621

619

Table 1 Symptom of anastomotic stricture after RYGB Symptom

%

Nausea and vomiting Dysphagia Salivation Upper abdominal pain

56 42 12 4

RYGB ⫽ Roux-en-Y gastric bypass.

Fig. 3. UE demonstrating patent anastomosis.

Results A total of 535 patients underwent RYGB during the study period. Of these 535 patients, 52 (36 women [69%]) with an overall mean age of 44.5 ⫾ 10.21 years underwent UE because of symptoms suggestive of anastomotic stenosis. The symptoms commonly described included nausea,

vomiting, dysphagia to solids, and upper abdominal pain. Overall, nausea and vomiting combined were documented in 29 patients (56%), and upper abdominal pain was noted only in 2 patients (4%; Table 1). Of these 52 symptomatic patients, 30 were evaluated with Gastrografin UGIS before UE. In the 52 patients who underwent UE, an average of 2.67 ⫾ 1.34 (range 1– 6) postoperative UEs was performed. The initial UE was performed at an average of 79 days (range 16 –502) postoperatively. The mean anastomotic diameter was 5.98 ⫾ 2.51 mm (range 0 –10). Strictures were the most common finding, observed in 41 patients (81%), followed by the presence of ulcers and strictures in 10 patients (19%), and only ulcers in 1 patient. For the 30 patients who underwent Gastrografin UGIS before UE, the mean anastomotic diameter was 6.8 ⫾ 3.43 mm (range 0 –14). The correlation between the radiographic and endoscopic findings was good using both Pearson’s correlation coefficient (0.44, P ⫽ .02) and simple linear regression analysis (␤ ⫽ .27, P ⫽ .02, 95% confidence interval 0.30 – 0.49) in an estimation of anastomotic stricture diameter, with sensitivity of 77% (Fig. 5). Discussion GI symptoms after RYGB are not uncommon. RYGB induces a combination of both malabsorption and restric-

10

Endoscopy (mm) 5

0 0

5

10

15

X-Ray (mm)

Fig. 4. UE demonstrating complete anastomotic obstruction.

Fig. 5. Regression line demonstrating correlation between endoscopic and radiographic findings.

620

S. Szomstein et al. / Surgery for Obesity and Related Diseases 2 (2006) 617– 621

tion. Restriction is achieved by creating a small gastric pouch and a relatively narrow gastrojejunal anastomosis approximately 12 mm in diameter [5]. Most patients with dysphagia and other GI symptoms after RYGB usually respond to dietary manipulation and do not require surgical or endoscopic intervention [6]. UE may offer both diagnostic and therapeutic measures for patients with GI symptoms. Although, millions of UEs are performed annually in the United States, they are performed by gastroenterologists or surgeons and are not as available as UGIS [7]. Currently, no standard protocol is available for evaluating patients after bariatric surgery with GI complaints. Also, even though stricture development after RYGB is a common postoperative complication, no clinical or radiologic criteria have been established that define the diagnosis in terms of pathognomonic symptoms or minimal diameter. A stricture is diagnosed when the patients’ symptoms correlate with the radiographic and endoscopic findings. However, symptoms in this patient population can be misleading. Huang et al. [8] evaluated the endoscopic findings in patients with upper GI symptoms after RYGB and correlated the clinical features with the endoscopic findings. Fortynine patients who underwent 69 UEs were included in that study. The most common endoscopic findings were normal postoperative anatomy (21 patients, 43%), and abdominal pain was the most common symptom (26 patients, 53%) and was more frequent among patients with normal endoscopic findings than among those with abnormal endoscopic findings (P ⫽ .04). Anastomotic stenosis was present in 39% of the patients who presented with nausea, vomiting, or dysphagia and was not present in any patient who did not have these symptoms (P ⫽ .001). The investigators concluded that among patients with symptoms after RYGB presenting for UE, normal postoperative anatomy was the most common finding. In our patient population, the most common symptoms were nausea and vomiting; abdominal pain was noted in only 2 patients. Vomiting usually occurred 10 –30 minutes after a meal. Another interesting symptom was increased salivation (or foaming of the mouth), which was present in ⱖ12% of the patients with no association with the timing after a meal (Table 1). The overall complication rate of UE is relatively low. Also, UE can serve not only as a diagnostic measure, but also as a therapeutic tool [9]. However, it is important to remember that the severely obese population is already considered high risk because of the co-morbidities. Performing UE in obese patients is difficult and is often associated with technical problems such as positioning [5]. In patients with suspected strictures or anastomotic leaks or high-risk patients, UE is usually preformed under general anesthesia [9]. General anesthesia or heavy sedation imposes a great risk to obese patients who are at an increased risk of pulmonary gastric aspiration and difficult airways, especially in patients with obstructive sleep apnea or those with symptomatic gastroesophageal reflux and other predis-

posing conditions. Thus, severely obese patients necessitate a multidisciplinary evaluation before induction of general anesthesia or heavy sedation by a well-trained team [10]. Currently, managed care dictates that cost-effectiveness and outcomes are paramount in patient treatment. Gastrografin UGIS is significantly less invasive and less expensive than UE, and, although it is not therapeutic, it is often used as the initial diagnostic tool in patients with GI symptoms after bariatric surgery and may eliminate the need for UE. Although this technique is commonly used in the immediate postoperative period to evaluate the newly created bypass anatomy and for leaks and possible strictures, its benefit in the evaluation of late strictures is less recognized. Moreover, GI complaints are common in patients after bariatric surgery, and these patients are often referred to various tests to evaluate the GI system. In our study, all 52 patients were evaluated by esophagogastroduodenoscopy (EGD), because we reviewed only the medical records of patients who underwent UE owing to symptoms suggestive of gastric outlet obstruction. Of these patients, we further identified those who underwent Gastrografin UGIS before UE. This study was specifically undertaken only to determine whether a correlation existed between Gastrografin UGIS and UE in determination of gastrojejunal anastomotic strictures after RYGB due to the paucity in the literature regarding this issue. Although we did not review all the medical records of patients who underwent Gastrografin UGIS at our institution, in our experience we often perform this test as the initial evaluation of patients for varying indications. In cases in which the patient was initially referred to the gastroenterology department, the initial evaluation was EGD because of the availability and ease of performing this diagnostic test. We used only Gastrografin UGIS and not thin barium for safety considerations, because the possibility of anastomotic leak in these patients should always be considered. Despite the main weakness of our investigation, its small sample size, a good correlation was found between the radiographic and endoscopic findings.

Conclusion Gastrografin UGIS correlated positively with the endoscopic gastrojejunal anastomosis findings in patients with anastomotic stricture who had undergone RYGB.

References [1] Go MR, Muscarella P II, Needleman BJ, Cook CH, Melvin WS. Endoscopic management of stomal stenosis after Roux-en-Y. gastric bypass. Surg Endosc 2004;18:56 –9. [2] Wittgrove AC, Clark GW. Laparoscopic gastric bypass, Roux-enY.—500 patients: technique and results, with 3– 60 month follow up. Obes Surg 2000;10:233–9.

S. Szomstein et al. / Surgery for Obesity and Related Diseases 2 (2006) 617– 621 [3] Sims TL, Mullican MA, Hamilton EC, Provost DA, Jones DB. Routine upper gastrointestinal Gastrografin swallow after laparoscopic Roux-en-Y. gastric bypass. Obes Surg 2003;13:66 –72. [4] Hamilton EC, Sims TL, Hamilton TT, Mullican MA, Jones DB, Provost DA. Clinical predictors of leak after laparoscopic Rouxen-Y. gastric bypass for morbid obesity. Surg Endosc 2003;17: 679 – 84. [5] Merkle EM, Hallowell PT, Crouse C, Nakamoto DA, Stellato TA. Roux-en-Y. gastric bypass for clinically severe obesity: normal appearance and spectrum of complications at imaging. Radiology 2005; 234:674 – 83. [6] Bell R, Reinhardt KE, Flowers JL. Surgeon-performed endoscopic dilatation of symptomatic gastrojejunal anastomotic strictures follow-

[7]

[8]

[9]

[10]

621

ing laparoscopic Roux-en-Y. gastric bypass. Obes Surg 2003; 13:728 –33. Cappell MS, Friedel D. The role of esophagogastroduodenoscopy in the diagnosis and management of upper gastrointestinal disorders. Med Clin North Am 2002;86:1165–216. Huang CS, Forse RA, Jacobson BC, Farraye FA. Endoscopic findings and their clinical correlations in patients with symptoms after gastric bypass surgery. Gastrointest Endosc 2003;58:859 – 66. Reed WP, Kilkenny JW, Dias CE, Wexner SD, for the SAGES EGD Outcome Study Group. A prospective analysis of 3525 esophagogastroduodenoscopies performed by surgeons. Surg Endosc 2004;18:11–21. Freid EB. The rapid sequence induction revisited: obesity and sleep apnea syndrome. Anesthesiol Clin North Am 2005;23:551– 64.