| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, University of Alabama at Birmingham, 619 19th St. S,
Birmingham, AL 35249-6830.
2 Department of Biostatistics, University of Alabama at Birmingham, Birmingham,
AL 35294-0022.
3 Department of Surgery, University of Alabama at Birmingham, Birmingham, AL
35249-0016.
4 Present address: Department of Surgery, WakeMed Hospital, Raleigh, NC
27610.
Received April 19, 2006;
accepted after revision July 31, 2006.
Address correspondence to M. E. Lockhart
(mlockhart{at}uabmc.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. With institutional review board approval, the CT scans of 18 patients (17 women, one man) with surgically proven internal hernia after laparoscopic Roux-en-Y gastric bypass were retrieved, as were CT studies of a control group of 18 women who had undergone gastric bypass but did not have internal hernia at reoperation. The scans were reviewed by three radiologists for the presence of seven CT signs of internal hernia: swirled appearance of mesenteric fat or vessels, mushroom shape of hernia, tubular distal mesenteric fat surrounded by bowel loops, small-bowel obstruction, clustered loops of small bowel, small bowel other than duodenum posterior to the superior mesenteric artery, and right-sided location of the distal jejunal anastomosis. Sensitivity and specificity were calculated for each sign. Stepwise logistic regression was performed to ascertain an independent set of variables predictive of the presence of internal hernia.
RESULTS. Mesenteric swirl was the best single predictor of hernia; sensitivity was 61%, 78%, and 83%, and specificity was 94%, 89%, and 67% for the three reviewers. The combination of swirled mesentery and mushroom shape of the mesentery was better than swirled mesentery alone, sensitivity being 78%, 83%, and 83%, and specificity being 83%, 89%, and 67%, but the difference was not statistically significant.
CONCLUSION. Mesenteric swirl is the best indicator of internal hernia after laparoscopic Roux-en-Y gastric bypass, and even minor degrees of swirl should be considered suspicious.
Keywords: CT • gastrointestinal radiology • hernia • small bowel
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recognized gastrointestinal complications of gastric bypass surgery include anastomotic leak, stenosis, and small-bowel obstruction [5-9]. Internal hernia, the movement of bowel into abnormal spaces, is more common after laparoscopic gastric bypass than after the open procedure. It is a particularly sinister complication with a variable, nonspecific clinical presentation [10, 11]. Although several CT findings have been described, including clustering of bowel loops against the anterior abdominal wall and swirled appearance of the mesentery, studies have been limited by small numbers (the largest series to our knowledge included only five patients who had undergone laparoscopic Roux-en-Y gastric bypass, and individual CT signs were not evaluated), by the inclusion of patients who did not undergo gastric bypass, and by the lack of a control group of bypass patients without internal hernia [12-15]. We undertook this study to evaluate the ability of radiologists using seven CT signs to diagnose internal hernia after laparoscopic Roux-en-Y gastric bypass.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Image Analysis
Seven signs of internal hernia were chosen for study. The first four signs,
all previously described in the literature
[12,
16], were as follows: swirled
appearance of mesenteric fat or vessels at the root of the mesentery
(Fig. 1), small-bowel
obstruction (Fig. 2), clustered
loops of small bowel (Fig. 3),
and mushroom shape of the herniated mesenteric root with crowding and
stretching of the mesenteric vessels (Fig.
4). The following three signs, which were based on our own
experience, had not been described previously, to our knowledge: tubular or
round shape of distal mesenteric fat closely surrounded by bowel loops
(Fig. 5), small bowel other
than duodenum passing posterior to the superior mesenteric artery (SMA)
(Fig. 6), and right-sided
location of distal jejunal anastomosis
(Fig. 7).
|
|
|
|
|
The CT studies were randomly presented to two experienced abdominal radiologists and a third-year radiology resident. These observers were asked to evaluate the images for presence or absence of the seven signs of internal hernia. The three reviewers were not the clinical reviewers of the images obtained before the study. To ensure consistency, the reviewers were first shown line drawings depicting the signs. For patients judged to have mesenteric swirl, the amount of swirl was estimated as < 45°, 45-90°, 91-180°, 181-270°, 271-360°, or > 360° relative to the origin of the SMA. The entire transverse CT series for each patient was available for review on a PACS workstation. The reviewers scrolled through the data set and adjusted window level and width as needed. Images were interpreted in a double-blind manner, and an observer monitored interpretation of the images. The CT findings were correlated with clinical data and surgical outcome.
Statistical Analysis
A biostatistician performed the statistical analysis using SPSS 12.0
software (issued September 2003, SPSS). The demographics of the two patient
cohorts were compared with use of Student's t test for continuous
variables and Fisher's exact test for discrete variables. Sensitivity for each
sign was calculated as the percentage of patients with hernia who had the
sign, and specificity was the percentage without hernia who did not have the
sign. Stepwise logistic regression was performed on the data to determine an
independent set of variables predictive of the presence of internal hernia.
Computations were performed separately for the three reviewers. Statistical
significance was considered p <0.05.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Accuracy of Individual CT Signs
The sensitivity and specificity of the seven individual signs of hernia and
the interpreters' overall impressions are shown in
Table 2. Mesenteric swirl was
the best single predictor of hernia with a sensitivity of 61%, 78%, and 83%
and a specificity of 94%, 89%, and 67% for the two experienced abdominal
imagers and one resident reviewer. The combination of swirled mesentery and
mushroom shape of the mesentery had higher sensitivity than swirled mesentery
alone with a sensitivity of 78%, 83%, and 83%, but specificity was not better
(83%, 89%, and 67%). The difference was not statistically significant. In
patients found to have mesenteric swirl but no hernia, the median amount of
swirl was less than 90°. The median amount of swirl in patients with
hernia was 180-270°. In all cases of at least 270° swirl, hernia was
found at surgery.
|
Distal tubular mesenteric fat surrounded by bowel loops, a small-bowel loop behind the SMA, and distal anastomosis to the right of midline were highly specific (89-100%) but had low sensitivity (0-44%). Clustered loops of bowel was the only CT finding in two patients with hernia, and one hernia patient had none of the seven signs.
Logistical Regression
Stepwise logistical regression showed the best indicator of internal hernia
after bypass was the presence of any degree of mesenteric swirl. Addition of
mushroom shape of the mesentery or the interpreter's overall impression did
not significantly improve the predictive value. One of the abdominal
radiologists successfully identified a single case in which a hernia had less
than 90° of swirl without other positive signs; the resident identified
two cases. Exclusion of swirl less than 180° would have caused one
abdominal radiologist to miss an additional hernia, the other abdominal
radiologist to miss two hernias, and the resident to miss two hernias.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Laparoscopic Roux-en-Y gastric bypass surgery produces three potential sites for internal hernia formation: at the defect in the transverse mesocolon through which the Roux loop passes (if it is placed in the retrocolic position), at the mesenteric defect at the enteroenterostomy, and behind the Roux limb mesentery placed in a retrocolic or antecolic position (retrocolic Petersen and antecolic Petersen type). Petersen hernia is a specific type in which intestine moves into the potential space between the caudal surface of the transverse mesocolon and the edge of the Roux limb. It has been suggested [6] that patients are more prone to internal hernia after laparoscopic Roux-en-Y gastric bypass than after an open operation because there are fewer adhesions to tether small-bowel loops and prevent them from herniating. In addition, patients who have greater degrees of weight loss after laparoscopic Roux-en-Y gastric bypass may be more prone to internal hernia because of loss of the protective, space-occupying effect of mesenteric fat [5]. In a review of 1,000 laparoscopic Roux-en-Y gastric bypass procedures, Garza and coworkers [18] identified 45 internal hernias, almost all of which were transmesocolic, in 43 patients. Another review [11] revealed 66 internal hernias in 63 of 2,000 patients who underwent laparoscopic Roux-en-Y gastric bypass. Forty-four of the hernias were transmesocolic, 14 were through the small-bowel mesentery, and five were of the Petersen type.
The symptoms of internal hernia are nonspecific, making the diagnosis challenging. Most patients report a combination of postprandial abdominal pain, nausea, and emesis [18]. We have found that patients typically do not have clinical signs of abdominal inflammation (elevated white blood cell count, peritoneal irritation, lactic acidosis) until there is infarction or perforation of the intestine, which increases the morbidity and mortality of remedial operations.
CT has been shown valuable in the detection of leaks, abscesses, and other perioperative complications of laparoscopic Roux-en-Y gastric bypass, but CT has been found less useful in detection of internal hernia. Blachar et al. [12] prospectively evaluated the ability of radiologists to detect paraduodenal and transmesenteric internal hernias on CT, but only one patient had undergone laparoscopic Roux-en-Y gastric bypass. Those investigators concluded that no sign was pathognomonic and suggested that the diagnosis could be made by applying a combination of signs with variable sensitivity and specificity.
In our clinical practice, we had been frustrated by our inability to confidently diagnose internal hernia in patients who had undergone laparoscopic Roux-en-Y gastric bypass. We had observed anecdotally that certain signs, notably mesenteric swirl, seemed to be helpful in diagnosis. Moreover, published reports of studies of the utility of CT in detection of internal hernia in patients who had undergone laparoscopic Roux-en-Y gastric bypass were limited by various factors, including small populations, inclusion of patients without gastric bypass, and lack of a control group. We therefore designed a study for retrospective evaluation of the sensitivity and specificity of seven CT signs.
Swirling of the mesentery was the best-performing single sign in our series with sensitivity and specificity of 61% and 94%, 78% and 89%, and 83% and 67% for the three reviewers. The value of the mesenteric swirl sign probably reflects its increased conspicuity compared with the other findings, which may not be as obvious on CT. The CT anatomy in patients who have undergone laparoscopic Roux-en-Y gastric bypass is not straight forward. The undulating leaves of mesentery through which hernias occur are very thin and rarely conform to any imaging plane. Therefore, the CT signs of herniation are largely based on identification of the effects of the hernia on the herniating structures, such as mesenteric vessels, rather than on delineation of the defect itself. The amount of swirl sufficient to make the diagnosis is open to debate because mesenteric swirling occurs in healthy persons and patients who have undergone laparoscopic Roux-en-Y gastric bypass but do not have an internal hernia. We believe, however, that any amount of mesenteric swirl should be viewed with suspicion in a patient with a history of laparoscopic Roux-en-Y gastric bypass who presents with abdominal pain.
Four other CT signs (distal tubular mesenteric fat surrounded by bowel loops, mushroom-shaped mesentery, small bowel behind the SMA, and right-sided anastomosis) were more specific but far less sensitive than mesenteric swirling. Distal tubular fat surrounded by bowel loops, resembling the eye of a hurricane, is similar to mesenteric swirl but occurs in the distal mesentery with bowel loops closely surrounding the round mesenteric fat rather than at the mesenteric root. Three of these signs, which have not been described previously, to our knowledge, were prospectively identified before this study on the basis of our anecdotal experience and discussions with our bariatric surgeon. Although they overlap some of the previously described signs to some extent, their higher specificity may be valuable in some cases.
We found good interobserver agreement for all three reviewers. A third-year radiology resident, who received the same pretest training as the two abdominal radiologists, was included to help us determine whether the findings were useful to imagers with less experience. The abdominal imaging specialists and the radiology resident performed similarly. This finding suggests that the CT signs chosen for the study were robust.
We believe the results of this study emphasize the need for radiologists who interpret CT scans of patients who have undergone laparoscopic Roux-en-Y gastric bypass to make special effort to review and understand the signs of herniation. To some extent, this finding parallels the results of others [14], who have commented on the value of review of CT images by bariatric surgeons, who are intimately familiar with the complex anatomy and the expected CT findings in their own patients. However, patients with symptoms after laparoscopic Rouxen-Y gastric bypass often present to an emergency department or clinic, and the initial imaging assessment is usually performed by a radiologist. Because a history of laparoscopic Roux-en-Y gastric bypass may not be available at interpretation, it is critical for radiologists to familiarize themselves with the CT appearance after laparoscopic Roux-en-Y gastric bypass and the signs of internal hernia.
As always, however, clinical correlation of CT findings with clinical presentation is critical. If the presentation suggests that an internal hernia is present (intermittent postprandial abdominal pain, nausea with or without vomiting) the surgeon should consider laparoscopic exploration to definitively make the diagnosis. This point is extremely important because the consequences of missed internal hernia with strangulation of the bowel are dire.
One limitation of our study was the high proportion of women relative to men, but the numbers represent the referral pattern for bariatric surgery at our institution. Another limitation was the difficulty in selecting a control group. Although we could have easily identified a matched group of controls with normal CT findings, the lack of laparoscopic Roux-en-Y gastric bypass in these patients would have been obvious to the reviewers. Therefore, because there were ethical concerns about imaging of patients who had undergone laparoscopic Roux-en-Y gastric bypass and did not have symptoms, we identified a group of laparoscopic Roux-en-Y gastric bypass patients who underwent CT and reoperations for other reasons and in which internal hernia was not found. This control probably accounts for the difference in time between surgery and CT in the clinical and control groups, which may account for the greater weight loss in the group with hernia. Prospective confirmation of our findings would be valuable.
The CT finding of swirled mesentery is the best single predictor of internal hernia after laparoscopic Roux-en-Y gastric bypass. Radiologists should become familiar with this sign and the other CT signs of internal hernia in laparoscopic Roux-en-Y gastric bypass patients. Even minor degrees of swirl should be considered suspicious for internal hernia in this population. We emphasize that CT findings are helpful in the overall management of complications after laparoscopic Roux-en-Y gastric bypass, but careful clinical correlation is paramount to avoid missed diagnoses that may be detrimental to the patient. This point also emphasizes the importance of a close, collaborative relationship between the bariatric surgeon and the radiologist, who should be committed to the long-term care of the bariatric patient.
Acknowledgments
We thank Charles G. Wells for assistance in data collection and Trish
Thurman for assistance with manuscript preparation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  L. R. Carucci, M. A. Turner, and S. D. Shaylor Internal Hernia Following Roux-en-Y Gastric Bypass Surgery for Morbid Obesity: Evaluation of Radiographic Findings at Small-Bowel Examination Radiology, June 1, 2009; 251(3): 762 - 770. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
