Original article

Closure of mesenteric defect can lead to decrease in internal herniasafter Roux-en-Y gastric bypass

Nestor de la Cruz-Muñoz, M.D., F.A.C.S., F.A.S.M.B.S.*, Juan C. Cabrera, M.D.,Melissa Cuesta, L.P.N., Scott Hartnett, D.O., Renan Rojas, M.D.

Department of Surgery, University of Miami Miller School of Medicine, Miami, Florida

Received May 7, 2010; accepted October 8, 2010

Abstract Background: Although most surgeons believe that internal hernia (IH) defects should be closed,no general consensus has yet been reached. The published studies have reported conflictingobstruction incidence estimates and variations in location. Some have also argued that closed herniasites recur.Methods: A retrospective review of 2079 Roux-en-Y gastric bypass procedures performed by1 surgeon from 2001 to 2009 was conducted. Chi-square analysis compared those performedfrom 2001 to August 2003 (group 1) with those from September 2003 to 2009 (group 2) for theincidence of IH. Since September 2003, the jejunojejunal anastomotic mesenteric defect hasbeen closed with a running, permanent suture. In contrast, before September 2003, the defectwas not closed.Results: A total of 37 patients presented with symptomatic IH from 2001 to 2009. A significantdifference was found in the proportion of those patients who experienced IH in group 1 (n � 36)compared with those who did so in group 2 (n � 1; P �.001). Group 1 developed 35 IHs (10%) atthe jejunojejunal anastomotic mesenteric defect and 6 (1.7%) at Petersen’s defect. The medianinterval to presentation was 20 months (range 5–34). Of the 35 patients, 5 (1.4%) presented acutelywith pain or bowel obstruction. Group 2 developed no IHs at the mesenteric defect (0%) and 1 atPetersen’s defect (.1%). The interval to presentation was 9 months.Conclusion: The incidence of IH after laparoscopic Roux-en-Y gastric bypass can be reducedwith closure of mesenteric defects. IHs can present either acutely with pain and obstructivesymptoms or chronically with vague, intermittent postprandial pain. (Surg Obes Relat Dis 2011;7:176 –180.) © 2011 American Society for Metabolic and Bariatric Surgery. All rights reserved.

Surgery for Obesity and Related Diseases 7 (2011) 176–180

Keywords: Bariatric surgery; Complications; Gastric bypass; Internal hernia

Laparoscopic Roux-en-Y gastric bypass (RYGB) has be-come a standard technique for the treatment of patients withmorbid obesity. The advantages of laparoscopic comparedwith the open technique have been widely demonstrated[1–3], but the development of internal hernias (IHs) afterlaparoscopic RYGB has been widely reported as a majorpostoperative complication [3–5]. The incidence and sitesof occurrence of IHs have varied among antecolic RYGB

*Correspondence: Nestor de la Cruz-Munoz, M.D., F.A.C.S.,F.A.S.M.B.S., Department of Surgery, University of Miami Miller Schoolof Medicine, 3650 Northwest 82 Avenue, Suite 302, Doral, FL 33166.

E-mail: ndelacruz@med.miami.edu

1550-7289/11/$ – see front matter © 2011 American Society for Metabolic anddoi:10.1016/j.soard.2010.10.003

series, perhaps related to the differences in the surgicaltechnique used by each bariatric surgeon. Although mostsurgeons believe that IH defects should be closed, no gen-eral consensus has yet been reached [5,6]. This has oc-curred in part because of the conflicting results publishedof the bowel obstruction rates and IH locations. Somehave also argued that IHs still occur at sites that havebeen closed [2].

The development of an IH is a potential complicationafter RYGB and carries the risk of bowel ischemia andperforation. Its diagnosis can be difficult because the pre-senting signs, symptoms, and physical and radiologic ex-

amination findings can be vague, nonspecific, and/or non-

Bariatric Surgery. All rights reserved.

177N. de la Cruz-Muñoz et al. / Surgery for Obesity and Related Diseases 7 (2011) 176–180

diagnostic [7]. The clinical presentation of an IH can varyfrom very unclear and intermittent symptoms to a dramaticacute abdomen secondary to small bowel necrosis or per-foration.

Three potential hernia sites exist after RYGB. The first isat the mesentery defect of the jejunojejunostomy. The sec-ond, often referred to as Petersen’s defect, is the defectcreated by the space posterior to the mesentery of the Rouxlimb as it is brought up to the stomach. The third is themesocolonic defect that occurs if the RYGB is done in theretrocolic fashion.

The purpose of the present analysis was to examine asingle-surgeon series of laparoscopic RYGB and describethe technique and incidence of IH between 2 groups ofpatients who had undergone same operative technique butdiffered in the closure or nonclosure of the mesentericdefect. It was hypothesized that those patients who hadundergone jejunojejunal anastomotic mesenteric defect(JJMD) closure with a running, permanent suture wouldhave a lower incidence of IH than those who did not un-dergo such closure. To our knowledge, this is the first studyto evaluate the effects of closing the mesenteric defect in asingle series.

Methods

A retrospective database review of 2079 patients whohad undergone antecolic laparoscopic RGYB performed by1 surgeon at multiple private institutions in South Florida(Mercy Hospital, Cedars Medical Center, and Hialeah Hos-pital) from 2001 to 2009 was conducted. The University ofMiami institutional review board approved the presentstudy. The patients were selected for surgery according tothe National Institutes of Health surgical criteria for severeobesity [8]. A change in the surgical technique occurred inSeptember 2003, dividing the series into 2 groups: thosewho had undergone RYGB from 2001 to 2003 when theJJMD was not closed (group 1) and those who had under-gone RYGB from 2003 to 2009 with complete closure ofthe JJMD using nonabsorbable continuous suture (group 2).

The patients who developed an IH postoperatively wereselected, and their medical records were reviewed. Thediagnosis of an IH was made clinically in most cases. It wassuspected in patients who presented with chronic intermit-tent colicky postprandial pain or bowel obstruction symp-toms. These patients were taken to the operating room fordiagnostic laparoscopy. The diagnosis was confirmed by thefinding of patent hernia defects on exploration and theresolution of symptoms after hernia repair. All patientsunderwent laparoscopic IH repair.

Other variables of interest included whether the JJMDhad been closed originally, interval from the initial surgeryto hernia repair, recurrence of the hernia after repair, symp-toms at presentation, radiologic findings at presentation, and

place of presentation (office or emergency room).

RYGB surgical technique

The small bowel was transected 50 cm from the liga-ment of Treitz (the mesentery was not entered �1 cm).The jejunum was run an additional 150 cm. The smallbowel and its associated mesentery were laid out in acircular fashion to give full visualization of the entiredefect. A jejunojejunostomy was performed using a bi-directional staple technique. The closure of the mesen-teric defect was performed with permanent running syn-thetic suture starting from the small bowel and continuingdown until the mesenteric defect had been completelyclosed near the base of the mesentery. It required 8 –12needle passes on each side of the mesenteric defect toclose it completely. Before September 2003, the smallbowel anastomosis was done similarly, but the mesenterydefect was not closed.

Petersen’s defect was not actively closed in any of thecases. During the procedure, the omentum was split downthe middle to decrease the tension on the Roux limb duringits antecolic course. The right half of the omentum wastucked against the mesocolon and mesentery of the Rouxlimb right at the defect.

Hernia reduction description

The small bowel (Roux limb) was run from the gastricpouch distally. In cases in which herniated bowel was pres-ent, it was best reduced by running the bowel retrograde,starting at the cecum. In patients with more mild, intermit-tent, postprandial pain, we often found a patent herniadefect without no bowel within in it.

Petersen’s space was also inspected. In most cases, it hadbeen sealed closed by the raw edge of the split omentum. Ifa defect was found, it was closed in similar fashion byflipping the Roux limb over the patient’s left site and su-turing the mesocolon to the mesentery of the Roux limbwith permanent synthetic running suture.

Statistical analysis

Fisher’s exact chi-square test analysis was conducted toexamine the statistical differences in the proportion of IHsbetween groups 1 and 2. The level of statistical significancewas set at P �.05. Data management and all statisticalanalyses were performed using Statistical Analysis Sys-tems, version 9.1 (SAS Institute, Cary, NC).

Results

The percentage of patients continuing follow-up at 1 and5 years after surgery was 62% and 60% and 37% and 30%of groups 1 and 2, respectively.

Of the 2079 patients who had undergone RYGB, 37patients presented with a symptomatic IH. A significant

difference was found in the number of patients in group 1

178 N. de la Cruz-Muñoz et al. / Surgery for Obesity and Related Diseases 7 (2011) 176–180

(2001–2003) who developed an IH (n � 36) compared withthose in group 2 (2003–2009) who did so (n � 1; P �.001).

Of the 36 group 1 patients with an IH, 35 (97.2%)developed the IH at the JJMD and 6 (16.7%) at Petersen’sdefect (between the Roux limb and transverse mesocolon orRoux limb mesentery). Of the 6 Peterson’s hernias found, 5were found in association with an IH at the JJMD. Themedian interval to presentation was 20 months (range5–34).

Of the 36 patients in group 1 with an IH, 5 (13.8%)presented acutely with pain or bowel obstruction. Of those5 patients, 4 had bowel incarcerated within the JJMD and 1had bowel within the Petersen’s defect. The remaining 31patients (86.2%) presented with chronic, postprandial, col-icky upper abdominal pain. In the chronic group, 100% hadpatent JJMDs, 0% had bowel within Petersen’s defect, and10 (27.7%) had bowel herniated within the JJMD.

Group 2 developed no hernias at the JJMD (0%) and 1at Petersen’s defect (.1%). This patient presented acutelywith severe pain. The interval to presentation was 9months.

The mortality rate from IH repair was 0%. No patientdeveloped irreversible bowel ischemia or perforation,and all patients had resolution of their preoperativesymptoms.

Discussion

The present analysis has shown that patients who havehad the JJMD closed with running, permanent sutureexperience significantly lower incidence of IH comparedwith those who did not undergo closure of the defect.This is one of the first studies to evaluate the effects ofclosing the mesenteric defect in a single series. Further-more, among those who did have the permanent suture,no deaths were reported, indicating this is a safe andeffective procedure.

The change from retrocolic to antecolic laparoscopicbypass has been shown in multiple studies to reduce theincidence of IH [9,10], perhaps by eliminating the mesoco-lic defect, which was the main site of IH [3]. In 1 analysisof postoperative IH, Higa et al. [2] presented a series of2000 laparoscopic RYGB patients. They were all performedin a retrocolic antegastric fashion with a hand-sewn gastro-jejunal anastomosis [2]. A total of 66 IHs (3.1%) werereported, and 52.3% of the patients presented with smallbowel obstruction and two thirds of the IHs were at themesocolic defect. The investigators concluded that IH sitesshould be closed with permanent suture and that cliniciansneed to be vigilant for patients presenting with symptoms ofIH that were not necessarily obstructed. It was also reportedthat patients would present to the office with postprandialcolicky type pain in the upper abdomen.

The true rate of IH might have been underestimated in

published studies. Several studies have calculated the IH

rates by the percentage of patients who presented emer-gently with small bowel obstruction due to IH [11–13].Many have not considered those patients with IH whopresented with chronic pain. The IH rates quoted by thisgroup of studies ranged from .2% to 1.5%, consistent withour findings of bowel obstruction rates [11–13].

Patients can also present to the office with occasionalmild abdominal discomfort that can become intense and isepisodic and most often postprandial, usually about 30 min-utes after completing the meal, regardless of the type offood ingested. Nausea has been a frequent associated symp-tom, as well as some upper abdominal distension [14]. Higaet al. [2] discovered 48% of IHs because of intermittentpain, and, in our series, nonobstructed IHs accounted for86% of our cases. We found that many patients did notmention these symptoms during the office visit. Most as-sumed that pain after bariatric surgery was normal. It wasonly after direct questioning that patients relayed relativelyconsistent symptoms.

Suturing of the mesenteric defect as a part of laparo-scopic RYGB has been a point of debate in some studies(Table 1). No consensus has been reached regardingwhether closure of the mesenteric defect will reduce theincidence of IH. Some investigators have suggested thatpartial closure of a defect might be more dangerous thanleaving the defect wide open [12], stating that a smaller holemight increase the risk of incarceration. Our clinical expe-rience would suggest these are the same patients most likelyto develop intermittent symptoms.

Furthermore, some investigators have suggested thatclosure of the defect can increase the costs, cause tensionon the anastomoses, and result in hematomas or injury tothe mesenteric vessels or adhesions that can cause smallbowel obstruction [5]. In our experience, closure of themesenteric defect eliminated the incidence of IH at thatsite and did not show add any adverse effects. Addition-ally, our experience has shown that adequate exposure ofthe defect is key to adequate closure. If the bowel andmesentery have been appropriately aligned, the entirespace can be easily defined. If the anatomy has not beenwell exposed, it would be easy to underestimate the sizeof the defect. In our experience, it has required 8 –12needle passes of permanent running suture to close theentire defect.

The present study had several limitations. First, it wasa retrospective review with a historical control group.The patients were not randomized. Also, we did not havehigh follow-up percentages for either group, and somepatients who developed IH might have been treated else-where. Also, some patients who had no preoperativeradiographic confirmation of an IH nor any bowel withinthe defect were included in the IH group because theyhad complete resolution of symptoms after closure of the

defect.

cggetrtefl

C

cswsdua

o

sp

D

b

R

TI

H

C

H

GP

B

MG

CQ

.

179N. de la Cruz-Muñoz et al. / Surgery for Obesity and Related Diseases 7 (2011) 176–180

In our series, a change in the surgical technique com-pletely changed our IH incidence. Other studies have alsoshown that seemingly minor changes in technique can havesignificant effects. Quebbemann and Dallal found [10] thathanging the orientation of the antecolic Roux limb at theastrojejunostomy altered the IH incidence between 2roups of patients who had undergone antecolic left-ori-nted (antimesenteric side of the roux limb pointed towardhe patient’s left side, with the cut end toward the patient’sight side) or antecolic right-oriented (antimesenteric end ofhe Roux limb toward the patient’s right side, with the cutnd toward the patient’s left side). The incidence of IHormation was 9.0% in the former group versus .5% in theatter [10].

onclusion

In our series, the patients who had undergone JJMDlosure with a running, permanent suture experienced aignificantly lower incidence of IH compared with thoseho had not undergone such closure. This is one of the first

tudies to evaluate the effects of closing the mesentericefect in a single series. Furthermore, of those who didndergo closure, no deaths were reported, indicating this issafe and effective procedure.Patients can present either acutely with small bowel

able 1nternal hernia rates in various studies

Investigator Year Totalcases (n)

Incidenceof IH (%)

Roux limb typ

wang et al. [12] 20041310 .3 (4) Antecolic405 1.48 (6) Retrocolic

hampion andWilliams [11]

2003

246 .4 (1) Antecolic465 1.5 (7) Retrocolic

iga et al. [2] 2003 2000 3.1 (66) Retrocolicantegastric

arza et al. [13] 2004 1000 4.5 (45) Retrocolicresent study —

352 11.7 (42) Antecolic1727 .05 (1) Antegastric

auman andPirrello [14]

2009 1047 6.80 (72) Antecolic ante

adan et al. [5] 2009 387 0 Antecolic anteunabushanamet al. [1]

2009 835 1.50 (13) Antecolic ante

ho et al. [6] 2006 1400 .20 (3) Antecolic anteuebbemann andDallal [10]

2005 Antecolic ante

200 9.00 (18) Left oriented200 .50 (1) Right oriented

IH � internal hernia; JJ � jejunojejunal; NS � not specified.Data in parentheses are numbers of patients, unless otherwise specified

bstruction or more subtly with chronic pain. Thus, IH

hould be considered in a gastric bypass patient who com-lains of intermittent abdominal pain.

isclosures

The authors have no commercial associations that mighte a conflict of interest in relation to this article.

eferences

[1] Gunabushanam G, Shankar S, Czerniach DR, Kelly JJ, Perugini RA.Small-bowel obstruction after laparoscopic Roux-en-Y gastric bypasssurgery. J Comput Assist Tomogr 2009;33:369–75.

[2] Higa KD, Ho T, Boone KB. Internal hernias after laparoscopic roux-en-Y gastric bypass: incidence, treatment and prevention. Obes Surg2003;13:350–4.

[3] Iannelli A, Facchiano E, Gugenheim J. Internal hernia after laparo-scopic Roux-en-Y gastric bypass for morbid obesity. Obes Surg2006;16:1265–71.

[4] Paroz A, Calmes JM, Giusti V, Suter M. Internal hernia afterlaparoscopic Roux-en-Y gastric bypass for morbid obesity: a con-tinuous challenge in bariatric surgery. Obes Surg 2006;16:1482–7.

[5] Madan AK, Lo Menzo E, Dhawan N, Tichansky DS. Internal herniasand nonclosure of mesenteric defects during laparoscopic Roux-en-Ygastric bypass. Obes Surg 2009;19:549–52.

[6] Cho M, Pinto D, Carrodeguas L, et al. Frequency and management ofinternal hernias after laparoscopic antecolic antegastric Roux-en-Ygastric bypass without division of the small bowel mesentery orclosure of mesenteric defects: review of 1400 consecutive cases. Surg

Mesentericdivision

Mesentericclosure at JJ

Peterson’sdefect closure

Follow-up (mo)

Yes Yes NS 21

Yes Variable Variable 43

Yes Variable Variable NS

NS NS NS NS

Minimal No No 94–108Yes No 12–94

Minimal Yes No NS

Minimal No No 23.5 (range 1–60)No Yes No NS

No No No 41.2%, �12NS Yes No

25.219.2

e

gastric

gastricgastric

gastricgastric

Obes Relat Dis 2006;2:87–91.

[

[

[

[

[

180 N. de la Cruz-Muñoz et al. / Surgery for Obesity and Related Diseases 7 (2011) 176–180

[7] Morgan H, Chastanet R, Lucha PA Jr. Internal hernia after laparo-scopic gastric bypass surgery: a case report and literature review.Postgrad Med 2008;120:E01–5.

[8] Hubbard VS, Hall WH. Gastrointestinal surgery for obesity. ObesSurg 1991;1:257–65.

[9] Ahmed AR, Rickards G, Husain S, Johnson J, Boss T, O’Malley W.Trends in internal hernia incidence after laparoscopic Roux-en-Ygastric bypass. Obes Surg 2007;17:1563–6.

10] Quebbemann BB, Dallal RM. The orientation of the antecolic Rouxlimb markedly affects the incidence of internal hernias after laparo-

scopic gastric bypass. Obes Surg 2005;15:766–70.

11] Champion JK, Williams M. Small bowel obstruction and internalhernias after laparoscopic Roux-en-Y gastric bypass. Obes Surg 2003;13:596–600.

12] Hwang RF, Swartz DE, Felix EL. Causes of small bowel obstructionafter laparoscopic gastric bypass. Surg Endosc 2004;18:1631–5.

13] Garza E Jr, Kuhn J, Arnold D, Nicholson W, Reddy S, McCarty T.Internal hernias after laparoscopic Roux-en-Y gastric bypass. Am JSurg 2004;188:796–800.

14] Bauman RW, Pirrello JR. Internal hernia at Petersen’s space after lapa-roscopic Roux-en-Y gastric bypass: 6.2% incidence without closure—a

single surgeon series of 1047 cases. Surg Obes Relat Dis 2009;5:565–70.