| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, C. U. B. Hôpital
Erasme, 808, Route de Lennik, 1070 Bruxelles, Belgium.
2
Department of Gastrointestinal Surgery, C. U. B.
Hôpital Erasme, 1070 Bruxelles, Belgium.
Received February 17, 2000;
accepted after revision May 25, 2000.
Address correspondence to M. Zalcman.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. All patients seen over a 3-year period with a CT diagnosis of small-bowel obstruction were included. There were 144 examinations in 142 patients. Images were interpreted prospectively with consensus by a fellow and an experienced gastrointestinal radiologist. Attention was focused on the presence of the following signs of strangulation and ischemia: reduced enhancement of the small-bowel wall, mural thickening, mesenteric fluid, congestion of small mesenteric veins, and ascites. A diagnosis of ischemia was made if enhancement of the bowel wall was reduced or if at least two of the other signs were found. Results were correlated with surgical findings in 73 cases and clinical follow-up in 71 cases.
RESULTS. A diagnosis of ischemia was made at surgery in 24 patients. CT diagnosis was correct in 23 patients (96% sensitivity). There were nine false-positive diagnoses (93% specificity). The negative predictive value of CT was 99%. Reduced enhancement of the bowel wall had a sensitivity of 48% and specificity of 100%, mural thickening had a sensitivity of 38% and specificity of 78%, mesenteric fluid had a sensitivity of 88% and specificity of 90%, congestion of mesenteric veins had a sensitivity of 58% and specificity of 79%, and ascites had a sensitivity of 75% and specificity of 76%.
CONCLUSION. Helical CT is a highly sensitive method to diagnose or rule out intestinal ischemia in the context of acute small-bowel obstruction.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Patient Characteristics
Our series included 144 consecutive CT examinations of 142 patients with CT
evidence of small-bowel obstruction. Two patients underwent two examinations
with an interval of more than 6 months. For the sake of clarity of results,
these four cases can be considered as four different patients. There were 81
men and 61 women (age range, 24-88 years; mean age, 60.9 years). A history of
previous laparotomy was present in 106 patients (75%). Fifty-two patients
(37%) had a history of malignancy, including 10 patients (7%) who had
undergone abdominal radiotherapy. Ten patients (7%) had a history of Crohn's
disease, eight of whom had undergone an intestinal resection.
Examination Technique
All the examinations were performed on a CT scanner (Somatom Plus 4A or 4C;
Siemens, Erlangen, Germany). Oral contrast material was not administered. The
acquisition of information was helical in all cases. The examinations started
with a study of the abdomen before IV contrast material injection. A volume of
120-180 mL of a 35 mg/100 mL iodinated agent was then injected at a rate of 2
mL/sec for the 130 patients without medical contraindication. The delay
between the start of injection and imaging varied between 40 and 70 sec. The
slice thickness was 5 or 8 mm, the pitch was 1.4, and the reconstruction
interval was 4 or 6 mm.
Image Analysis
Reduced enhancement of the bowel wall.— This finding (Fig.
1A,1B)
was based on visual comparison of the attenuation of the bowel wall at the
site of the obstruction with the attenuation of bowel loops distant from this
site, after contrast material injection. No attempt was made to quantify the
enhancement abnormality or to differentiate reduced enhancement from absent or
delayed enhancement.
|
|
Thickening of the bowel wall.—A bowel-wall thickness of 2 mm or less was considered normal. Bowel-wall thickness was considered abnormal if it appeared to be thicker than 2 mm in a noncollapsed segment. Owing to the difficulty of accurately measuring the thickness of the small-bowel wall, this evaluation was partly subjective, and the values were therefore not recorded.
Mesenteric fluid.—The presence of localized fluid in the small-bowel mesentery attached to abnormal (dilated, thickened) small-bowel loops was considered abnormal. The amount of fluid was not evaluated, and the attenuation was not measured.
Congestion of small mesenteric veins.—Enlargement of the small mesenteric veins around the site of the obstruction as compared with mesenteric veins distant from this site was considered a positive finding.
Ascites.—This was defined as presence of fluid in the peritoneal cavity as opposed to fluid in the mesentery. The amount and attenuation value of ascites were not measured.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Only two of the nine patients with an unconfirmed CT diagnosis of ischemia underwent surgery. One had multiple ileal strictures caused by Crohn's disease, and one had an incarcerated hernia of the ventral abdominal wall without evidence of ischemia. The seven other patients did not undergo surgery because the discrepancy between clinical and radiologic diagnoses was considered too important by the patient's surgeon. These seven patients recovered with conservative treatment, and their diagnoses of ischemia were considered false-positive (Fig. 2A,2B). Eighteen of the 23 patients with a true-positive CT diagnosis of ischemia underwent surgery within 12 hr, three underwent surgery 24-36 hr after CT, and one underwent surgery 3 days later. Ischemia was caused by an adhesive volvulus in 11 patients, by adhesive bands in seven patients, by a hernia in four patients (3 external, 1 internal), by mesenteric infarction in one patient, and by mesenteric vein thrombosis in one patient.
|
|
Bowel resection was necessary in 14 of 23 patients with a true-positive CT diagnosis. The overall sensitivity of CT in the diagnosis of strangulation and ischemia in acute small-bowel obstruction was thus 96% (23/24) with a specificity of 93% (111/120), a positive predictive value of 72% (23/32), and a negative predictive value of 99% (111/112).
Findings on the 144 CT scans of patients with small-bowel obstruction are shown in Table 1. The diagnostic value of the various combinations of two signs of ischemia is shown in Table 2.
|
|
Reduced Enhancement of the Bowel Wall
Reduced enhancement of the ischemic bowel wall was observed in 10 of the 21
patients with confirmed ischemia who had a contrast material injection
(sensitivity, 47.6%) (Figs.
1A,1B,
3, and
4A,4B,4C).
It was never present in noncomplicated obstruction (specificity, 100%). The
positive and negative predictive values of this sign were 100% and 90.8%
(109/120), respectively.
|
|
|
|
Thickening of the Bowel Wall
This sign, usually considered a cardinal feature of ischemia
[16,17,18,19],
was observed in only nine of the 24 patients with proven ischemia
(sensitivity, 37.5%) (Fig. 4C).
When only the patients with contrast material injection were considered, the
sensitivity was 42.9% (9/21). Bowel-wall thickening was also present in 26 of
120 patients with noncomplicated obstruction (specificity, 78.3%). As a
result, the positive (25.7%) and negative (86.2%) predictive values of this
sign were insufficient. In four patients with ischemia, the thickened wall had
a target configuration, constituted by the presence of rings of different
attenuation, after contrast material injection. This form of wall thickening
was more specific of ischemia; it was visible in only four of 109 patients
with contrast material injection in the nonischemia group (specificity, 96%).
Positive and negative predictive values for the target appearance were 50% and
86%.
Twelve of 21 patients had an abnormal enhancement of the bowel wall (2 patients had both a reduced enhancement and target appearance) in the ischemia group (sensitivity, 57%).
Mesenteric Fluid
Mesenteric fluid was present in 21 of the 24 patients with ischemia
(sensitivity, 87.5%) and was also visible in 12 of 120 patients in the
non-complicated group (specificity, 90%). Positive and negative predictive
values were 63.6% (21/33) and 97.3% (108/111), respectively.
Mesenteric Congestion
Focal congestion of the small mesenteric veins around the site of
obstruction was seen in 14 of 24 patients in the ischemia group (sensitivity,
58.3%) and in 25 of 120 patients in the nonischemia group (specificity,
79.2%). In the ischemia group, one patient only had this feature without
evidence of mesenteric fluid. Positive and negative predictive values were
35.9% and 90.5%, respectively.
Ascites
Ascites was present in 18 of 24 patients in the ischemia group
(sensitivity, 75%) and 29 of 120 patients in the nonischemia group
(specificity, 75.8%). Positive and negative predictive values were 38.3%
(18/47) and 93.8% (91/97), respectively.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The objective of our study was to determine prospectively the value of CT in obstruction-associated ischemia and the value of the CT signs we thought were the most reliable for the diagnosis. Unlike three other groups of researchers [12,13,14], we avoided opacification of the small bowel for several reasons. Obstruction prevents the progression of oral contrast material; ingestion of fluid by a patient with intestinal obstruction will invariably lead to vomiting; and, above all, the natural contrast offered by intestinal fluid allows an optimal study of the bowel wall after contrast material injection. Furthermore, it is illogical and potentially hazardous to delay CT examination up to 2 hr [12, 13] while waiting for the opacification of bowel loops in a patient with abdominal emergency. All illustrations accompanying the latest reports on strangulation [12,13,14] show nonopacified ischemic bowel loops, which confirm that there is no indication for intestinal contrast medium in CT performed in the context of small-bowel obstruction.
The overall sensitivity and negative predictive value of CT in the diagnosis of strangulation and ischemia were high, 96% and 99%, respectively. The discrepancy between our results and those of the latest extensive study by Balthazar et al. [14] is probably because more than two thirds of their patients underwent a sequential study of the abdomen instead of a helical study. As reported, reduced enhancement of the bowel wall can be a transient feature and can thus be inconspicuous on a nonhelical scan of the abdomen [20]. Furthermore, angiographic studies of the small-bowel volvulus [21] have shown the possibility of intense bowel-wall opacification in a later phase, which may contribute to the inadequate conspicuity of this sign on scans obtained during an intermediate phase. Reduced enhancement of the bowel wall was the most frequent mural abnormality in our series and had a 100% specificity. It is the only direct sign of vascular impairment of the small bowel and has been reported in predominantly arterial diseases such as mesenteric infarction [22, 23] as well as in predominantly venous diseases such as strangulation [20]. Reduced enhancement of the bowel wall was not associated with necrosis of the small bowel in three of our patients. The first had a small-bowel volvulus with discolored bowel at surgery, which returned to normal after devolvulation (Fig. 3). The second had a strangulated hernia with a cyanosed loop that returned to normal after the cure of the hernia. The third had a superior mesenteric vein thrombosis and did not undergo surgery (Fig. 4A,4B,4C). The explanation for this sign is thus not totally clear, but it could at least partly be the result of an arterial spasm caused by the obstruction to the venous outflow; in one patient, it was visible in a nonischemic loop proximal to the diseased segment (Fig. 4B).
The most unexpected finding of our study was the limited value of mural thickening as an individual sign of ischemia. The target configuration of wall thickening had a high specificity (96%) but was visible in few patients (19%). This discrepancy between our experience and previously reported results is mainly related to the fact that there is no wide consensus regarding the normal thickness and appearance of the small-bowel wall in variable states of distention so that "the initial observation of bowel wall thickening on CT is usually made subjectively" [24]. A major cause of error, which has not been reported yet in the literature, is the confusion between true wall thickening and pseudowall thickening due to transverse scanning of Kerckring's folds in a distended fluid-filled bowel. The presence of a gas bubble in this pseudothickening can lead to a false-positive diagnosis of ischemia-induced pneumatosis (Fig. 2B) and unnecessary surgery. The high number of patients with a thickened bowel wall in the non-complicated group was related to two factors: the difficulty of assessing accurately the bowel-wall thickness and the fact that virtually half of our patients had medical conditions that can be associated with thickening of the bowel wall (cancer and Crohn's disease). The low sensitivity of this sign is probably multifactorial. Most of our examinations were performed early in the evolution of the obstruction (i.e., at admission), dilatation of the bowel can reduce the conspicuity of moderate mural thickening and, as already mentioned, subtle CT findings are open to subjective interpretations.
Fluid in the mesentery adjacent to abnormal (thickened, dilatated, or both) bowel loops was the individual sign most frequently associated with strangulation (88%) and had a good specificity (90%), despite the large proportion of patients with malignant or inflammatory disease in our series. The value of mesenteric changes as signs of small-bowel ischemia has already been reported in previous articles [12, 13, 21] and must be underscored for two reasons. These abnormalities are visible on unenhanced scans, and they have an extremely high negative predictive value. Only two of the 24 patients with ischemia had no mesenteric changes. One patient had two bowel wall abnormalities, including reduced enhancement, and the other one had the false-negative examination with artifacts.
We did not find in our previous experience that bowel strangulation was associated with an increase in the attenuation value of ascites. Although the volume of peritoneal fluid can be larger in patients with strangulation, it does not appear reasonable to consider ascites as a good predictor of ischemia. However, the absence of peritoneal fluid has a high negative predictive value, and thus the search for ascites is not without importance.
The benefit of early surgery is underscored by the fact that all patients with small-bowel ischemia who did not undergo surgery the day of the CT diagnosis (n = 4) had a bowel resection and the eight patients with proven ischemia who had no bowel resection underwent surgery the day of the CT diagnosis. It must be added that seven of the nine patients with a false-positive diagnosis of ischemia did not undergo surgery and recovered fully with conservative treatment alone. It is impossible to tell if these cases represent in reality false-positive diagnoses of ischemia or cases of reversible ischemia. It will be interesting in future studies to evaluate whether CT can enable radiologists to differentiate cases of mild reversible ischemia from potentially life-threatening ischemia, when seen at an early stage. In the meantime it seems preferable to accept a somewhat imperfect specificity rather than decrease the high sensitivity of CT.
Another issue that needs further evaluation is that of IV contrast material: Is it always necessary and what is the optimal delay between injection and imaging to visualize enhancement abnormalities? In two of 24 patients with confirmed ischemia, the diagnosis was made on an unenhanced examination (Fig. 5A,5B) in the presence of mesenteric changes and ascites. Ideally, contrast material injection should be limited to patients who have no clear evidence of ischemia on the initial CT scan of the abdomen. This would reduce the potential hazard of injecting a dehydrated patient, suppress the risk of an adverse reaction to iodine contrast material, and reduce the cost of the diagnostic procedure. If contrast material is necessary, choosing the adequate time-window to study the enhancement of the small bowel is another variable that needs further study. At the beginning of our series, we chose a relatively short delay of 40 sec that was gradually increased to 60-70 sec to avoid potential diagnostic problems related to the variability of blood flow velocity between patients. A 60- to 70-sec delay between the start of contrast material injection and the start of imaging seems a reasonable compromise between a purely arterial phase with the risk of overdiagnosing enhancement abnormalities and a delayed phase that can render inconspicuous the abnormal enhancement.
|
|
In summary, we believe that CT should be the initial radiologic examination each time high-grade bowel obstruction is suspected. It should be performed early after admission with a helical acquisition. If the examination is correctly performed and of acceptable technical value, it can show signs of ischemia with a high sensitivity or can safely exclude the presence of ischemia.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  T Hamada, M Yamauchi, M Tanaka, Y Hashimoto, K Nakai, and K Suenaga Prospective evaluation of contrast-enhanced ultrasonography with advanced dynamic flow for the diagnosis of intestinal ischaemia Br. J. Radiol., August 1, 2007; 80(956): 603 - 608. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. M. Thompson, R. K. Kilani, B. B. Smith, J. Thomas, T. A. Jaffe, D. M. Delong, and E. K. Paulson Accuracy of Abdominal Radiography in Acute Small-Bowel Obstruction: Does Reviewer Experience Matter? Am. J. Roentgenol., March 1, 2007; 188(3): W233 - W238. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. P. Sheedy, F. Earnest IV, J. G. Fletcher, J. L. Fidler, and T. L. Hoskin CT of Small-Bowel Ischemia Associated with Obstruction in Emergency Department Patients: Diagnostic Performance Evaluation Radiology, December 1, 2006; 241(3): 729 - 736. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Aguirre, A. C. Santosa, G. Casola, and C. B. Sirlin Abdominal Wall Hernias: Imaging Features, Complications, and Diagnostic Pitfalls at Multi-Detector Row CT RadioGraphics, November 1, 2005; 25(6): 1501 - 1520. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Vernhet, J. M. Serfaty, M. Serhal, E. McFadden, E. Bonnefoy, P. Adeleine, D. Revel, and P. Douek Abdominal CT Angiography Before Surgery as a Predictor of Postoperative Death in Acute Aortic Dissection Am. J. Roentgenol., April 1, 2004; 182(4): 875 - 879. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Wiesner, B. Khurana, H. Ji, and P. R. Ros CT of Acute Bowel Ischemia Radiology, March 1, 2003; 226(3): 635 - 650. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Macari, H. Chandarana, E. Balthazar, and J. Babb Intestinal Ischemia Versus Intramural Hemorrhage: CT Evaluation Am. J. Roentgenol., January 1, 2003; 180(1): 177 - 184. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. K. Chou CT Manifestations of Bowel Ischemia Am. J. Roentgenol., January 1, 2002; 178(1): 87 - 91. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
