Effect of staple height on gastrojejunostomy during laparoscopic gastric bypass: a multicenter prospective randomized trial

Surgery for Obesity and Related Diseases 6 (2010) 477– 484

Original article

Effect of staple height on gastrojejunostomy during laparoscopic gastric bypass: a multicenter prospective randomized trial Ninh T. Nguyen, M.D.a,*, Gregory Dakin, M.D.b, Brad Needleman, M.D.c, Alfons Pomp, M.D.b, Dean Mikami, M.D.c, David A. Provost, M.D.d, Daniel J. Scott, M.D.d, Daniel B. Jones, M.D.e, Scott Gallagher, M.D.f, Michel Gagner, M.D.b, Michel Murr, M.D.f a

Department of Surgery, University of California Irvine Medical Center, Orange, California b Department of Surgery, Weil Cornell Medical College, New York, NY c Department of Surgery, Ohio State University, Columbus, Ohio d Department of Surgery, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas e Department of Surgery, Beth Israel Deaconess Medical Center, Boston, Massachusetts f Department of Surgery, University of South Florida, USF Health, Tampa, Florida Received June 10, 2009; revised October 6, 2009; accepted March 29, 2010

Abstract

Background: Gastrointestinal (GI) bleeding and anastomotic stricture are frequent complications associated with the construction of the gastrojejunostomy during laparoscopic gastric bypass. Staplers with shorter staple height can reduce the rate of postoperative GI hemorrhage. The aim of the present study was to assess the outcomes of patients who had undergone gastric bypass with construction of the gastrojejunostomy using a 25-mm circular stapler with a 3.5- versus 4.8-mm staple height. Methods: From January 2007 to February 2009, 357 patients underwent laparoscopic gastric bypass using a circular stapler for construction of the gastrojejunostomy were randomly assigned to either the 3.5-mm (n ⫽ 180) or 4.8-mm (n ⫽ 177) group. Two patients randomized to the 4.8-mm group did not undergo the operative procedure and were excluded from the analysis. The primary outcome measures included the rate of GI hemorrhage, anastomotic stricture, and wound infection. Results: The 2 groups were similar with regard to the demographics and baseline body mass index (47 versus 48 kg/m2). The operative time, blood loss, and postoperative hematocrit on day 2 were similar between the 2 groups. No significant differences were seen in the overall rate of intraoperative GI bleeding or postoperative GI bleeding from all sources (3.3% for 3.5 mm versus 6.3% for 4.8 mm, P ⬎.05); however, a trend was seen toward a lower rate of postoperative GI bleeding from the gastric pouch or gastrojejunostomy (.5% for 3.5 mm versus 3.4% for 4.8 mm, P ⫽ .06). The rate of anastomotic stricture was significantly lower in the 3.5-mm group (3.9% versus 16.0%, P ⬍.01). No significant differences were seen in rate of wound infection between the 2 groups. Other morbidities for the entire study cohort included leaks (1.1%), pulmonary embolism (.6%), gastrointestinal obstruction (1.4%), and reoperation (3.4%). The overall in-hospital mortality rate was .3%, and the 30-day mortality rate was .8%. Conclusions: In the present prospective, randomized trial, using a circular stapler with a shorter staple height (3.5 mm) during construction of the gastrojejunostomy, significantly reduced the rate of postoperative anastomotic stricture, with a trend toward a lower rate of GI bleeding from the gastrojejunostomy. (Surg Obes Relat Dis 2010;6:477– 484.) © 2010 American Society for Metabolic and Bariatric Surgery. All rights reserved.

Keywords:

Obesity; Gastric bypass; Gastrojejunostomy; Gastrointestinal bleeding; Laparoscopy

Supported by Covidien, North Haven, Connecticut. Presented at the American Society for Metabolic and Bariatric Surgery, June 25, 2009, Grapevine, Texas. *Correspondence: Ninh T. Nguyen, M.D., Department of Surgery, University of California Irvine Medical Center, 333 City Boulevard West, Suite 850, Orange, CA 92868. E-mail: [email protected] 1550-7289/10/$ – see front matter © 2010 American Society for Metabolic and Bariatric Surgery. All rights reserved. doi:10.1016/j.soard.2010.03.294

478

N. T. Nguyen et al. / Surgery for Obesity and Related Diseases 6 (2010) 477– 484

Roux-en-Y gastric bypass has long been the bariatric procedure of choice for the treatment of morbid obesity. Concerning perioperative complications after gastric bypass have included anastomotic leak, gastrointestinal bleeding, and anastomotic stricture. These complications might be related to the construction of the gastrojejunostomy. The initial description for construction of the gastrojejunostomy during Roux-en-Y gastric bypass used the circular stapler technique [1]. Subsequently, some surgeons switched to using the linear stapler or a hand-sewn technique; however, the circular stapler technique continues to be the most commonly performed method for construction of the gastrojejunostomy [2]. Gastrointestinal (GI) bleeding is not an uncommon complication during and after gastric bypass. In a review of 3464 gastric bypass cases, Podnos et al. [3] reported a 1.9% (range 1.1–3.8%) incidence of postoperative GI bleeding. The most likely source of GI bleeding is from the multiple staple lines within the GI tract, including the gastrojejunostomy. One of the methods to prevent GI bleeding after gastric bypass is to use a stapler with a shorter staple height. For example, the gastric pouch is now commonly constructed using the blue load (3.5-mm staple height) instead of the green load (4.8-mm staple height), and the jejunojejunostomy is constructed using a white load (2.5 mm staple height) instead of a blue load. Until now, the commercially available circular stapler used for construction of the gastrojejunostomy had a 4.8-mm staple height. Recently, a circular stapler with a shorter staple height (3.5 mm) was developed in an effort to enhance tissue compression and possibly minimize the GI bleeding arising from the gastrojejunostomy. The aim of the present prospective, randomized trial was to compare the outcomes of patients who had undergone laparoscopic Roux-en-Y gastric bypass using the 25-mm circular stapler with either a 3.5- or 4.8-mm staple height. Methods Study design Six centers participated in the present trial. The patients were recruited to participate at the bariatric surgery clinic of each respective institution. The institutional review board of each center approved the present study. The study was also registered on the ClinicalTrials.gov website (identifier NCT00254072). All patients who agreed to undergo randomization provided written informed consent. The patients were eligible for the study if they had a body mass index of 35– 60 kg/m2, had an acceptable operative risk, and were aged 18 – 60 years. The exclusion criteria were the presence of a large ventral hernia, hiatal hernia, or liver cirrhosis and history of venous thrombosis or pulmonary embolism. After the patients gave written informed consent, they were randomly assigned to undergo laparoscopic gastric bypass us-

ing a 25-mm circular stapler with either 3.5- or 4.8-mm staples. The patients were unaware of the randomized assignment. Surgical technique All patients within both groups received venous thrombosis prophylaxis, consisting of both sequential compression devices and preoperative and perioperative anticoagulation. Laparoscopic gastric bypass was performed using 5–7 abdominal ports. A 15–30-cm3 gastric pouch was constructed by stapling the stomach, starting at the lesser curvature and exiting through the angle of His. An anvil from the circular stapler was placed, either transabdominally or transorally, within the gastric pouch. With the transoral technique, a gastric pouch was constructed, and the Orvil device (Covidien, North Haven, CT) was passed transorally to a position at the tip of the gastric pouch. With the transabdominal technique, the anvil was placed intraperitoneally; passed through a gastrotomy, and positioned 1–2 cm distal to the gastroesophageal junction. A gastric pouch was then constructed around the anvil. For anvil placement within the 3.5-mm group, the Orvil device was used for the transoral technique, and either the Orvil device or a conventional anvil was used for the transabdominal technique. For anvil placement within the 4.8-mm group, the conventional anvil was used for both the transoral and the transabdominal techniques. The ligament of Treitz was identified, and the jejunum was divided 30 –100 cm distal to the ligament of Treitz. The Roux limb was measured to 75–150 cm, and a side-to-side jejunojejunostomy was constructed using a 60-mm linear stapler with a 2.5-mm staple height. The mesenteric defect was closed with running or interrupted sutures. The remaining jejunal enterotomy was sutured or stapled closed. The Roux limb was placed in a retrocolic or antecolic pathway. According to the preoperative randomized assignment, a 25-mm circular stapler with either 3.5- or 4.8-mm staples (Covidien) was used for construction of an end-to-side gastrojejunostomy. In addition to having a shorter staple height, the circular stapler with 3.5-mm staples has advanced technology termed “DST series,” which includes a new cutting system that maximizes the anastomotic patency and directional stapling technology that might improve staple formation. On completion of the gastrojejunostomy, the jejunal enterotomy was closed with a linear stapler. The gastrojejunal anastomosis was reinforced with 2–3 interrupted sutures. The gastrojejunostomy was tested for air leaks using either endoscopy with air insufflation or intraluminal injection of methylene blue dye. Postoperative care and follow-up Patient-controlled analgesia was used for pain control. A Gastrografin contrast study was performed at the discretion of the surgeon. Clear liquid was started after the contrast study had demonstrated no obstruction or leaks. Patients

N. T. Nguyen et al. / Surgery for Obesity and Related Diseases 6 (2010) 477– 484

were discharged as soon as oral intake was tolerated. Follow-up was on an outpatient basis at 1 week, 1 month, and 2 months postoperatively. Outcomes The endpoints of the present study were the intraoperative and perioperative morbidities. The outcome measures included demographics, operative time, estimated blood loss, length of hospital stay, morbidity (the rate of intraoperative and postoperative GI bleeding, anastomotic stricture, anastomotic leaks), and mortality. Intraoperative GI bleeding was defined as the presence of intraluminal bleeding observed through the Roux limb enterotomy after construction of the gastrojejunostomy and requiring the need for therapeutic intervention such as oversewing of the anastomosis for hemostasis. Postoperative GI bleeding was defined as the presence of signs or symptoms consistent with GI hemorrhage such as hematemesis, bright red blood per rectum, melena, or the presence of hypotension with a decreasing hematocrit. Postoperative GI bleeding specifically attributed to the gastrojejunostomy or the gastric pouch included bleeding that manifested as hematemesis or objective confirmation of staple line bleeding as determined by endoscopy. Anastomotic stricture was defined as symptoms of frequent vomiting after meals, with endoscopy showing narrowing of the gastrojejunostomy, and the inability to pass the 9.8-mm endoscope through the anastomosis. The rate of anastomotic stricture was reported at the 2-month follow-up period. All surgical complications were recorded, and a single patient could have multiple unrelated complications. Each complication was categorized according to its severity. The severity of the complication was graded using the classification system proposed by Clavien et al. [4]. Grade I complications were alterations from the ideal postoperative course, not life-threatening, and with no lasting disability. Grade II complications were potentially life-threatening but without residual disability and were subdivided into 2 groups: grade IIa (requiring the use of blood transfusions, total parenteral nutrition, drug therapy, or a hospital stay more than twice the median stay) and grade IIb (requiring therapeutic procedures such as endoscopy or reoperation). Grade III complications were those with residual disability or requiring organ resection. Finally, a grade IV complication was death as a result of complications. Statistical analysis All data are expressed as the mean ⫾ standard deviation. Analyses of the differences between the 2groups for demographic and operative data were performed using 2-sample t tests or Fisher’s exact test for categorical data. A comparison of the complications was made using the chi-square test for independence. All statistical analyses were conducted using Statistical Analysis Systems, version 9.1 (SAS Insti-

479

tute, Cary, NC). A P value of ⬍.05 was considered significant. During the study design, a sample size calculation was performed with the assumption that the rate of intraoperative and postoperative clinical GI hemorrhage using the circular stapler with 4.8-mm staples would be 10% (6% for intraoperative bleeding and 4% for clinical postoperative bleeding) and the rate of hemorrhage using the circular stapler with 3.5-mm staples would be 2%. A difference of 8% in the rate of hemorrhage between the 2 groups was considered statistically significant. A minimum of 162 patients in each group was necessary to detect this difference using a 2-tailed t test with a probability of a type I error (␣) of .5 and a probability of a type II error (␤) of .2 (power of 80%). Results Demographics From January 2007 to February 2009, 357 patients at 6 centers were randomly assigned to undergo laparoscopic gastric bypass using a 25-mm circular stapler with either 3.5- or 4.8-mm staples. Center 1 enrolled 41 patients, center 2 enrolled 80 patients, center 3 enrolled 104 patients, center 4 enrolled 32 patients, center 5 enrolled 90 patients, and center 6 enrolled 10 patients. Two patients randomized to the 4.8-mm group did not undergo the operation and were excluded from the analysis. The characteristics of the 2 groups are listed in Table 1. The mean body mass index, gender, and age were similar between the 2 groups. The preoperative hematocrit and co-morbidities were also similar between the 2 groups. The technique for transoral placement of the anvil was performed more frequently in the Table 1 Patient characteristics Characteristic

3.5-mm Group (n ⫽ 180)

4.8-mm Group (n ⫽ 175)

P value

Female gender Age (yr) Mean body mass index (kg/m2) White race Transoral placement of anvil Baseline hematocrit (g/dL) Co-morbidity Sleep apnea Gastroesophageal reflux Hypertension Diabetes Dyslipidemia Coronary artery disease

135 (75.0) 43.1 ⫾ 10.8 47.8 ⫾ 5.6 144 (80.0) 78 (22.0) 40.3 ⫾ 3.5

138 (78.8) 43.1 ⫾ 10.1 46.9 ⫾ 5.4 131 (74.9) 51 (14.4) 39.9 ⫾ 3.5

NS* NS† NS† NS* ⬍.01* NS†

68 (37.8) 74 (41.1) 109 (60.5) 70 (38.9) 78 (43.3) 11 (6.1)

72 (41.1) 79 (45.1) 106 (60.6) 66 (37.7) 71 (40.1) 10 (5.7)

NS* NS* NS* NS* NS* NS*

NS ⫽ not significant. Data presented as numbers, with percentages in parentheses, or mean ⫾ standard deviation. * Fisher’s exact test. † Two sample t test.

480

N. T. Nguyen et al. / Surgery for Obesity and Related Diseases 6 (2010) 477– 484

Table 2 Perioperative outcomes Outcome

3.5-mm Group (n ⫽ 180)

4.8-mm Group (n ⫽ 175)

P value

Operative time (hr) Estimated blood loss (mL) Patients requiring conversion to laparotomy Patients requiring intensive care unit stay Length of hospital stay (d) Intraoperative bleeding at gastrojejunostomy Intraoperative leak at gastrojejunostomy Postoperative GI bleeding Overall Specific to gastrojejunostomy 30-Day reoperation Hematocrit on first postoperative day (g/dL) Hematocrit on second postoperative day (g/dL)

2.2 ⫾ 0.9 88.1 ⫾ 101.8 2 (1.1)

2.2 ⫾ 0.7 93.5 ⫾ 96.5 0 (0)

NS* NS* NS†

9 (5.0)

10 (5.7)

NS†

3.5 ⫾ 5.7 2 (1.1)

3.0 ⫾ 2.3 3 (1.7)

NS* NS†

4 (2.2)

2 (1.1)

NS†

6 (3.3) 1 (0.5) 4 (2.2) 35.9 ⫾ 3.9

11 (6.3) 6 (3.4) 8 (4.6) 35.5 ⫾ 3.7

NS† .06† NS† NS*

34.3 ⫾ 4.0

34.2 ⫾ 4.0

NS*

NS ⫽ not significant; GI ⫽ gastrointestinal. Data presented as numbers, with percentages in parentheses, or mean ⫾ standard deviation. * Two-sample t test. † Fisher’s exact test.

3.5-mm group (22% versus 14% of cases). Complete 2-month follow-up data were available for all patients in both groups. Perioperative data The perioperative data of the 2 groups are listed in Table 2. Conversion to open laparotomy was required in 1.1% of patients in the 3.5-mm group and 0% in the 4.8-mm group. The mean operative time and blood loss were similar between the 2 groups. The length of hospital stay was also similar between the 2 groups. An intensive care unit stay was required for 5.0% of patients in the 3.5-mm group and 5.7% of patients in the 4.8-mm group. No significant difference was seen in the overall rate of GI bleeding between the 2 groups (3.3% for the 3.5-mm group and 6.3% for the 4.8-mm group). However, a trend was seen toward a lower rate of GI bleeding arising specifically within the gastrojejunostomy or gastric pouch (.5% versus 3.4%, P ⫽ .06). Although the rate of GI bleeding was similar between the 2 groups, none of the 6 patients with GI bleeding within the 3.5-mm group required a therapeutic intervention compared with 5 of 11 patients within the 4.8-mm group (P ⫽ .16). The hematocrit levels on postoperative day 1 and 2 were similar between the 2 groups. No significant difference was seen in the 30-day reoperation rates between the 2 groups. The 30-day mortality for the entire study cohort was .8%. Transoral placement of the anvil was successful in all attempted cases. No esophageal laceration or injuries oc-

curred. No significant difference was seen in the rate of wound infection between the transoral and transabdominal techniques for placement of the anvil (8.5% versus 7.5%, respectively). Additionally, no significant differences were seen in the rate of stenosis or GI bleeding between those with transoral versus transabdominal placement of the anvil. Complications The major (grade IIb, III, and IV) and minor (grade I and IIa) intraoperative and postoperative complications between the 2 groups are listed in Table 3. No significant difference was seen in the rate of intraoperative bleeding from the gastrojejunostomy between the 2 groups (1.1% for the 3.5-mm group versus 1.7% for the 4.8-mm group). The rates of major and minor complications were significantly greater in the 4.8-mm group than in the 3.5-mm group. Within the 3.5-mm group, 5 major complications (2.8%) occurred; 1 patient had a leak secondary to proximal jejunal perforation, 1 patient developed an intra-abdominal abscess secondary

Table 3 Early complications according to severity* Complication

Intraoperative complications Gastrojejunal anastomosis bleeding Positive gastrojejunal anastomotic leak test Splenic injury Total Major complications (grades IIb, III, IV) Pneumonia Acute renal failure GI hemorrhage requiring therapeutic treatment Anastomotic leak/intra-abdominal abscess Myocardial infarction GI obstruction Pulmonary embolism Total Minor complications (grades I, IIa) GI hemorrhage (expectant management) Anastomotic stricture Cardiac arrhythmia Superficial wound infection Prolonged diarrhea Peripheral neuropathy/paresthesia Marginal ulcer Dehydration Total

3.5-mm Group (n ⫽ 180)

4.8-mm Group (n ⫽ 175)

2 4

3 2

0 6 (3.3)

3 8 (4.6)

0 1 0

1 0 5

2

2

1 1 0 5 (2.8) 6 7 0 16 1 1 1 1 33 (18.3)

0 4 2 14 (8.0)

P value

NS†

.03†

6 28 1 12 0 1 3 2 52 (29.7)

.02†

Abbreviations as in Table 2. Data presented as numbers, with percentages in parentheses. * Included all complications; a single patient could have multiple complications. † Fisher’s exact test.

N. T. Nguyen et al. / Surgery for Obesity and Related Diseases 6 (2010) 477– 484

to anastomotic leak, 1 patient developed early postoperative bowel obstruction, 1 developed acute renal failure, and 1 experienced a myocardial infarction. Within the 4.8-mm group, 14 major complications (6.6%) occurred; 4 patients had GI obstruction, 1 developed pneumonia, 5 developed GI bleeding requiring therapeutic intervention, 2 developed leaks, and 2 developed pulmonary embolism. The most common minor complications in both groups were GI hemorrhage with expectant management, anastomotic stricture requiring endoscopic balloon dilation, and superficial wound infection. The rate of anastomotic stricture at 2 months of follow-up was significantly greater in the 4.8-mm group compared with the 3.5-mm group (16.0% versus 3.9%, respectively, P ⬍.01). Discussion Laparoscopic gastric bypass is an effective bariatric operation but can be associated with perioperative complications such as leaks, anastomotic stricture, and GI hemorrhage. These complications can occur in part from the construction of the gastrojejunostomy. In an effort to reduce the morbidity associated with gastric bypass, the present randomized trial compared the outcomes of patients who had undergone laparoscopic gastric bypass using a 25-mm circular stapler and either a 3.5- or a 4.8-mm staple height. The results of our study have demonstrated that compared with the 4.8-mm group, the patients in the 3.5-mm group had a significantly lower rate of anastomotic stricture, with a trend toward a reduction in the rate of GI hemorrhage arising within the gastrojejunostomy. Anastomotic stricture is the most common perioperative complication after gastric bypass. Patients normally present with intolerance for food and/or liquids 3– 6 weeks postoperatively. Although stricture can be easily treated with endoscopic balloon dilation, it often requires multiple sessions and a small risk of perforation exists. Multiple studies have reported a greater frequency of stricture with the use of the 21-mm circular stapler compared with a 25-mm circular stapler [5– 8]. In clinical practice, most surgeons have preferred the 25-mm circular stapler for construction of the gastrojejunostomy. Despite the use of the 25-mm circular stapler, the rate of anastomotic stricture from multiple published reports has been 7.0 –10.2% [5–7]. In the present trial, the 3.5-mm staple height significantly reduced the rate of anastomotic stricture compared with the rate using the 4.8-mm staple height (3.9% versus 16.0%, respectively). The low rate of anastomotic stricture within the 3.5-mm group in the present study was similar to the reported rate of stricture after construction of the gastrojejunostomy using the linear stapler with a shorter staple height (3.5 or 2.5 mm). In a prospective study of 1012 patients who had undergone laparoscopic gastric bypass using a linear stapler with a 2.5-mm staple height, Ukleja et al. [9] reported a stricture rate of 6%. In a comparative study of the linear (3.5-mm staple height) versus circular stapler technique,

481

Sczepaniak et al. [10] reported a 12.9% anastomotic stricture rate using the 25-mm circular stapler and a 0% rate using the linear stapler technique. The pathophysiologic mechanisms for the lower rate of anastomotic stricture among patients who underwent gastric bypass with a circular stapler with a shorter staple height could have been contributed to by the improved tissue compression associated with the shorter staple height, which might minimize wound contraction, and the improved function of the DST series circular stapler that maximizes the anastomotic diameter. GI bleeding can occur intraoperatively or postoperatively, with life-threatening consequences. Intraoperative bleeding from the gastrojejunostomy can often be detected as evidenced by the outpouring of fresh blood through the jejunal enterotomy. Treatment consists of oversewing the anastomosis or intraluminal packing of the anastomosis with hemostatic agents. In the present study, no significant differences were seen in the rate of intraoperative bleeding at the gastrojejunostomy between the 2 groups (1.1% for the 3.5-mm group and 1.7% for the 4.8-mm group). Regarding postoperative GI bleeding, Spaw and Husted [11] reported a 3.1% rate of GI bleeding in 2895 patients who had undergone laparoscopic gastric bypass. In the present study, we only evaluated differences in outcomes between circular staplers with 3.5- and 4.8-mm staples for construction of the gastrojejunostomy. Therefore, we were not expecting to find a significant difference in the overall rate of GI bleeding. However, examining the rate of GI bleeding specifically from the gastrojejunostomy or gastric pouch, we found a trend toward a lower rate of GI bleeding within the 3.5-mm group (.5% for the 3.5-mm group versus 3.4% for the 4.8-mm group, P ⫽ .06). Additionally, none of the patients with GI bleeding within the 3.5-mm group required therapeutic intervention to control the bleeding, but 5 of 11 patients within the 4.8-mm group did. Finally, despite a reduction in the staple height to 3.5 mm, no significant difference was seen in the rate of anastomotic leaks between the 2 groups. One of the major benefits of the circular stapler with 3.5-mm staples is the improved option for transoral placement of the anvil. The original description of the construction of the gastrojejunostomy used a circular stapler with transoral placement of the anvil [1]. However, complications, such as esophageal perforation and laceration, have been reported owing to the technical difficulty in passing the mushroom-like anvil through the esophagus [12]. Subsequently, some surgeons have mechanically tilted the head of the anvil to facilitate transoral passage. In the present study, the 3.5-mm circular stapler included an anvil pretilted for transoral introduction, and 22% of the cases in the 3.5-mm group were performed with transoral placement of the anvil. In the 4.8-mm group, 14% of cases were performed with transoral placement of the anvil, which was manually tilted by the surgeon. No intraoperative or postoperative esopha-

482

N. T. Nguyen et al. / Surgery for Obesity and Related Diseases 6 (2010) 477– 484

geal complications related to the use of the transoral technique occurred in either group. A major advantage of the transoral technique is the elimination of constructing a gastrotomy to introduce the anvil and hence the elimination of the need to close the gastrotomy. A disadvantage of the transoral technique is the risk of wound infection, because the oral gastric tube is retrieved laparoscopically and removed through 1 of the trocar sites. However, our study did not show a significant difference in the rate of wound infection between the transoral and transabdominal techniques for placement of the anvil (8.5% versus 7.5%, respectively). Our study had limitations. The main limitation was the difficulty in discerning GI bleeding specifically attributed to the gastrojejunostomy staple line. GI bleeding can occur at the gastrojejunostomy, gastric pouch, jejunojejunostomy, or gastric remnant staple lines. GI bleeding from the gastric remnant or the jejunojejunostomy can be identified by the signs and symptoms of presentation such as the presence of bright red blood per rectum or the passing of melena. In contrast, it is difficult to clinically differentiate between GI bleeding from the staple lines of the gastrojejunostomy and that from the staple lines of the gastric pouch, because bleeding from both of these areas can result in hematemesis. Additionally, although the main difference between the circular staplers with 3.5- versus 4.8-mm staples is the staple height, other technological advances were present with the 3.5-mm circular stapler, including the presence of a new cutting system that provides a maximal luminal diameter and the directional stapling technology, which could explain the differences in outcomes. Also, some variation was present in the operative techniques among the study sites, such as differences in the route (antecolic versus retrocolic) and length (75 versus 150 cm) of the Roux limb. However, these differences should not have affected the clinical rate of GI bleeding within the gastrojejunostomy because the construction of the gastrojejunostomy was standardized among the centers. Finally, our study sample size might have been underpowered, such as was seen by the trend for a lower rate of GI bleeding arising specifically from the gastrojejunostomy in the 3.5-mm group, with the P value close to significant (P ⫽ .06). A study with a larger sample size might have been able to demonstrate statistically significant differences between the 2 groups. Despite these limitations, this is the first clinical trial to date to compare the

outcomes for the construction of the gastrojejunal anastomosis using a circular stapler with a varying staple height. In the present prospective, randomized trial, we found that the use of a circular stapler with a shorter staple height (3.5 mm) during construction of the gastrojejunostomy significantly reduced the incidence of anastomotic stricture, with a trend toward a lower rate of GI bleeding arising from the gastrojejunostomy. Disclosures The authors have no commercial associations that might be a conflict of interest in relation to this article. References [1] Wittgrove AC, Clark GW, Tremblay LJ. Laparoscopic gastric bypass, Roux-en-Y: preliminary report of five cases. Obes Surg 1994;4: 353–7. [2] Madan AK, Harper JL, Tichansky DS. Techniques of laparoscopic gastric bypass: on-line survey of American Society for Bariatric Surgery Practicing Surgeons. Surg Obes Relat Dis 2008;4:166 –72. [3] Podnos YD, Jimenez JC, Wilson SE, et al. Complications after laparoscopic gastric bypass: a review of 3464 cases. Arch Surg 2003;138:957– 61. [4] Clavien PA, Sanabria JR, Strasberg SM. Proposed classification of complications of surgery with examples of utility in cholecystectomy. Surgery 1992;111:518 –26. [5] Fisher BL, Atkinson JD, Cottam D. Incidence of gastroenterostomy stenosis in laparoscopic Roux-en-Y gastric bypass using 21- or 25-mm circular stapler: a randomized prospective blinded study. Surg Obes Relat Dis 2007;3:176 –9. [6] Nguyen NT, Stevens CM, Wolfe B. Incidence and outcome of anastomotic stricture after laparoscopic gastric bypass. J Gastrointest Surg 2003;7:997–1003. [7] Suggs WJ, Kouli W, Lupovici M, et al. Complications at gastrojejunostomy after laparoscopic gastric bypass: comparison between 21and 25-mm circular staplers. Surg Obes Relat Dis 2007;3:508 –14. [8] Takata MC, Ciovica R, Cello JP, et al. Predictors, treatment, and outcomes of gastrojejunal stricture after gastric bypass for morbid obesity. Obes Surg 2007;17:878 – 84. [9] Ukleja A, Afonso BB, Pimentel R, et al. Outcome of endoscopic balloon dilation of strictures after laparoscopic gastric bypass. Surg Endosc 2008;22:1746 –50. [10] Sczepaniak JP, Owens ML. Results of gastrojejunal anastomotic technique designed to reduce stricture. Surg Obes Relat Dis 2009;5:77– 80. [11] Spaw AT, Husted JD. Bleeding after laparoscopic gastric bypass: case report and literature review. Surg Obes Relat Dis 2005;1:99 – 103. [12] Nguyen NT, Wolfe BM: Hypopharyngeal perforation during laparoscopic Roux-en-Y gastric bypass. Obes Surg 2000;10:64 –7.

Editorial comment

Comment on: Effect of staple height on gastrojejunostomy during laparoscopic gastric bypass: a multicenter prospective randomized trial In this issue of the Journal, Nguyen et al. [1] have published a report on the first prospective trial evaluating the effect of staple size on the construction of a circular

stapled (EEA) gastrojejunal anastomosis during Roux-en-Y gastric bypass (RYGB). This well-conducted, multi-institution study compares the clinical outcomes of using a