| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, Nagasaki University School of Medicine, 1-7-1
Sakamoto, Nagasaki 852-8501, Japan.
2
Department of First Pathology, Nagasaki University School of Medicine,
Nagasaki 852-8501, Japan.
3
Department of Radiology, Yamaguchi University School of Medicine, 1-1-1
Minamikogushi, Ube 755-8505 Japan.
4
Department of Radiology, Oita Medical University, 1-1 Idaigaoka, Hasama-machi,
Oita 879-5593, Japan.
Received June 19, 2000;
accepted after revision October 10, 2000.
Address correspondence to T. Fujimoto.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. The imaging studies, clinical records, and surgical and pathologic findings in 25 patients with intussusception were reviewed retrospectively. We evaluated six CT findings based on the abnormalities of each component of intussusception. Presence or absence of these findings was compared with the degree of vascular compromise as observed on pathologic examination, such as edema, ischemia, or necrosis.
RESULTS. The hypodense layer was observed in 16 of 18 intussusceptions with various degrees of vascular compromise. A fluid collection surrounded by the returning wall, which was revealed to correspond to trapped peritoneal fluid, was observed in eight of nine intussusceptions with ischemia or necrosis. A gas collection surrounded by the returning wall was observed in two of four intussusceptions with necrosis. Free peritoneal fluid coexisted with a fluid collection surrounded by the returning wall in all the intussusceptions except one. Bowel obstruction was observed in six of nine intussusceptions with ischemia or necrosis. The maximum wall thickness was not related to the degree of vascular compromise.
CONCLUSION. The CT findings of a hypodense layer in the returning wall, fluid collection in the space surrounded by the returning wall, and gas collection in the space surrounded by the returning wall can be useful in predicting the degree of vascular compromise in intussusception.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this study, we describe the fundamental components of intussusception and evaluate the CT findings based on the abnormalities of these components. Six CT findings were compared with the degree of vascular compromise, which was classified in four stages: minimal compromise, edema, ischemia, and necrosis. CT staging criteria were devised based on the evaluation of these findings. The final purpose of the study was to clarify which CT findings predict the degree of vascular compromise associated with intussusception.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The presence of intussusception was confirmed at surgery in 20 patients (21 intussusceptions); one patient had two separate intussusceptions. In the remaining five patients, no intussusception was found at surgery. We believe that these five intussusceptions were spontaneously reduced because typical findings of intussusception had been revealed on CT and sonography performed before surgery.
The primary symptoms of the 24 patients were abdominal pain, obstructive symptoms (vomiting, abdominal fullness, constipation, and diarrhea), and bloody stool. These symptoms were chronic (n = 12), acute (n = 10), or intermittent (n = 2). One patient had no abdominal symptoms; this patient had Peutz-Jeghers syndrome and was found to have an intussusception on screening CT.
The initial examination leading to the diagnosis was CT in 16 intussusceptions. In the remaining patients, the initial diagnostic examination was sonography (n = 8) or barium enema (n = 1); CT was performed to confirm the diagnosis. The interval from the onset of symptoms to CT ranged from 3 hr to several months, and the interval from CT to surgery was from 3 hr to 25 days (median, 2 days). The patients with an interval of more than 48 hr from CT to surgery (n = 12) had chronic or intermittent symptoms. The sites of intussusception were enteric (n = 7), ileocolic (n = 3), ileocecal (n = 12), and colocolic (n = 4). Intussusception was caused by malignant tumors (n = 13) and benign tumors (n = 11); no neoplasm was found in the remaining two intussusceptions.
Of 21 intussusceptions that were observed at surgery, en bloc resection was performed in four because of the presence of marked vascular compromise (n = 3) or a possibility of dissemination of cancer (n = 1). In the remaining 17 intussusceptions, manual reduction with resection was performed. Of these, the entire length was resected in 13 intussusceptions because of ischemia or necrosis (n = 6), the presence of a malignant tumor without ischemia or necrosis (n = 5), or the presence of multiple benign tumors without ischemia or necrosis (two separate intussusceptions in one patient). Limited resection comprising removal of the causative tumor with the adjacent bowel was performed in the remaining four. In our series no intussusception was treated by manual reduction only.
CT scans were obtained with a 10-mm collimation, a 10- to 15-mm interval, a scan time of 1-4 sec, 120 kV or 137 kV, and 140-360 mA.
Image Analysis
The unenhanced CT images were analyzed independently by two radiologists
who were unaware of the clinical, surgical, and pathologic findings but who
were aware of the diagnosis of intussusception. If the CT interpretations of
the two radiologists differed, a third radiologist evaluated the images, and
the majority opinion was used. The third observer guided the two to measure
the wall thickness at the same level when the two observers disagreed.
Diagnostic Criteria of Intussusception and Identification of Its
Components
In this article, the following terms are used to express the components of
intussusception (Fig. 1). These
terms and their definitions agree with those used in previous publications
[1,
2,
5,
9,10,11,12],
except the term of "thick bowel wall complex." The diagnosis of
all intussusceptions was made by identification of a fatty component
(invaginated mesentery or mesocolon) surrounded by a ringlike bowel wall
(thick bowel wall complex) on CT. The continuity of the fatty component to the
mesentery or mesocolon was visualized in all.
|
Thick bowel wall complex.—The thick bowel wall complex was defined as a ringlike bowel wall structure surrounding a fatty component. The structure consists of two bowel walls that are the outer intussuscipiens and the inner returning wall of the intussusceptum and includes the intraluminal space between them. However, the complex frequently appears as a single wall because the intraluminal space cannot always be identified.
Invaginated mesentery (or mesocolon).—Invaginated mesentery was defined as a fatty component surrounded by the thick bowel wall complex. The fat continues to the mesentery or mesocolon outside the intussusception and frequently makes a crescent shape in the space surrounded by the thick bowel wall complex.
Intussuscipiens.—Intussuscipiens refers to the outer part of the thick bowel wall complex separated from the returning wall of the intussusceptum by intraluminal gas or fluid. The wall continues to the distal bowel wall at the distal end of the intussusception.
Returning wall of the intussusceptum.—The returning wall of the intussuseptum is defined as the inner part of the thick bowel wall complex, which is separated from the intussuscipiens by intraluminal gas or fluid.
Entering wall of the intussusceptum.—The entering wall of the intussuseptum is defined as the bowel surrounded by invaginated mesentery. It may be located eccentrically in the space surrounded by the thick bowel wall complex.
Intraluminal space in intussusception.—The intraluminal space in intussuseption is defined as the space containing gas or fluid in either the entering wall of the intussusceptum or the thick bowel wall complex.
Analysis of CT Findings Related to Vascular Compromise
After identifying these components of intussusception on CT, we evaluated
whether the presence of the following five findings was related to the degree
of vascular compromise observed at surgery
(Fig. 2): hypodense layer in
the thick bowel wall complex or returning wall of the intussusceptum
(hypodense layer), fluid collection in the space surrounded by the returning
wall of the intussusceptum (fluid collection), gas collection in the space
surrounded by the returning wall (gas collection), free peritoneal fluid, and
bowel dilatation proximal to the intussusception (bowel obstruction). In
addition, the maximum thickness of the thick bowel wall complex was measured,
excluding the site of the local mass lesion that was suggested to be the
cause. The thickness of each of the intussuscipiens and the returning wall was
measured to reveal which was responsible for the bowel wall thickening, if we
could identify both.
|
Of these six CT criteria, the first three were proposed by us, and the latter three have been suggested in previous reports to be related to vascular compromise [6, 8]. We did not use any CT patterns, such as target, layered, sausage-shaped, or reniform, for evaluating CT findings of intussusception because these patterns have not been clearly defined.
Assessment of Degree of Vascular Compromise
To assess the degree of vascular compromise, we classified it into four
stages according to the pathologic findings. Pathologic stage 1 (minimal
compromise) meant that few changes due to vascular compromise were noted.
Pathologic stage 2 (edema) meant that edema, congestion, or both were observed
but were judged to be reversible. Pathologic stage 3 (ischemia) meant that
apparent edema or congestion, which necessitated total or near-total resection
of the intussusception, was present. Pathologic stage 4 (necrosis) meant that
necrosis, gangrene, or both were obvious. The histopathologic findings of
resected specimens were evaluated for the presence of edema, ischemia,
necrosis, and gangrene by a pathologist who was unaware of the CT
findings.
Statistical Analysis
Interobserver disagreement was evaluated with Cohen's kappa statistics. The
correlation between the stage of vascular compromise based on CT and
pathologic findings was analyzed with Spearman's rank correlation
coefficient.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Of the 26 intussusceptions, 10 showed uniform soft-tissue density of the intussusceptum or thick bowel wall complex (Fig. 3). The remaining 16 intussusceptions showed a hypodense layer, which appeared to be present in the returning wall of the intussusceptum (Fig. 4A,4B). A fluid collection was seen in nine of these 16 intussusceptions (Figs. 5A,5B and 6A,6B). This fluid collection was confirmed by the surgical en bloc specimen in one case (Fig. 5A,5B) and by sonography in four cases (Fig. 6A,6B). A gas collection was seen as air bubbles or air-fluid levels in two of nine (Fig. 7). Free peritoneal fluid, all instances of which were accompanied by a fluid collection, was present in eight patients. Bowel obstruction was seen in six of the eight patients with a fluid collection and in one without the fluid collection or any other findings (Fig. 3). A fluid collection was only seen in intussusceptions that showed a hypodense layer, and gas collections were always accompanied by a fluid collection.
|
|
|
|
|
|
|
|
The maximum thickness of thick bowel wall complex ranged from 9 to 23 mm (mean ± SD, 15.2 ± 3.4 mm). In 10 cases, the thickness of the intussuscipiens identified on CT was less than 3 mm. In the remaining 16 cases, the intussuscipiens could not be separately identified.
Interobserver agreement in the interpretation of the CT findings was complete for the presence or absence of free peritoneal fluid and bowel obstruction. Interobserver disagreement was seen in one case each for identifying the hypodense layer, a fluid collection, and a gas collection; the kappa coefficient was 92% for each of these three findings.
Pathologic Evaluation and Assessment of Degree of Vascular
Compromise
The numbers and frequencies for each stage of vascular compromise based on
pathologic findings were as follows: pathologic stage 1 was found in eight
intussusceptions (31%); pathologic stage 2, in nine intussusceptions (35%);
pathologic stage 3, in five intussusceptions (19%); and pathologic stage 4, in
four intussusceptions (15%). The presence or absence of free peritoneal fluid
and proximal bowel obstruction on CT was confirmed in all patients at surgery.
No significant discrepancy was found between the surgical findings and the
histopathologic findings in any of the cases.
Diagnostic Value of CT Findings for Detection of Vascular
Compromise
The hypodense layer was present in 16 (89%) of 18 intussusceptions of
pathologic stage 2 or greater and was absent in all eight (100%) of pathologic
stage 1 cases. A fluid collection was present in eight (89%) of nine
intussusceptions of pathologic stage 3 or greater and was absent in 16 (94%)
of 17 with pathologic stage 2 or less. The gas collection was present in two
(50%) of four intussusceptions of pathologic stage 4 and was absent in all 22
cases (100%) of pathologic stage 3 or less. Free peritoneal fluid was present
in eight (89%) of nine patients with pathologic stage 3 or greater and was
absent in all 16 (100%) with pathologic stage 2 or less. Proximal bowel
obstruction was present in one (11%) of nine intussusceptions of pathologic
stage 2, three (60%) of five with pathologic stage 3, and three (75%) of four
with pathologic stage 4. The mean maximum thickness (±SD) of the thick
bowel wall complex was 15 ± 3 mm in pathologic stage 1, 17 ± 3
mm in pathologic stage 2, 14 ± 4 mm in pathologic stage 3, and 15
± 4 mm in pathologic stage 4.
All 26 intussusceptions could be classified into four stages according to the presence or absence of the three CT findings: the hypodense layer, fluid collection, and gas collection. CT stage 1 meant that all three CT findings were absent; CT stage 2 meant that only the hypodense layer was present; CT stage 3 meant that both the hypodense layer and fluid collection were present; and CT stage 4 meant that all three CT findings were present. Four of the 10 CT stage 1 intussusceptions spontaneously reduced, whereas only one of the remaining 16 did. There was a close correlation between CT staging and pathologic staging (r = 0.90, p < 0.0001; Spearman's rank correlation coefficient) (Table 1).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The hypodense layer in the thick bowel wall complex is thought to indicate mural edema of the returning wall of the intussusceptum. A lack of this CT finding correlates with a stage of minimal vascular compromise. The intussuscipiens, when it was identified, was not included in this hypodense layer. The intussuscipiens is thought to be less damaged because its mesentery was not involved [13].
The fluid collection is considered to be extraluminal because the returning wall of the intussusceptum is everted. The order of the intraluminal and extraluminal spaces is reversed, and this reversed orientation should be kept in mind to realize that fluid or gas surrounded by the bowel wall is extraluminal. An en bloc resected specimen (Fig. 5A,5B) confirmed that the extraluminal fluid was trapped in the invaginated portion. In 1996, del-Pozo et al. [14] proposed the concept of trapped peritoneal fluid on sonography in cases of intussusception in infants. These authors speculated that the fluid was peritoneal transudate caused by vascular compromise. Our results agree with their speculation in the location of the fluid collection, its origin, and the association with vascular compromise. The trapped fluid in intussusception and the free fluid in the peritoneal cavity coexisted in all patients in our series except one. Therefore, in our series of adults, the significance of these factors, as reported by del-Pozo et al., appears to be almost the same in practice.
A gas collection in the space is considered to be extraluminal, like the fluid collection at the same site. Therefore, a gas collection should be regarded as a finding of perforation in intussusception. To our knowledge, gas in the space surrounded by the returning wall has not been described as a finding of perforation or necrosis. Because the gas with trapped peritoneal fluid closely resembles intraluminal contents, the finding might have been missed previously.
Bowel obstruction has been considered to indicate the presence of vascular compromise [6, 8]. However, our study showed that proximal bowel obstruction does not necessary indicate [ILL.] compromise. The compromised intussusception, even with gangrene, may not show [ILL.] of proximal bowel obstruction.
In intussusception, thickening of the returning wall has been explained to indicate mural edema associated with vascular compromise [5, 10, 12, 13, 15]. However, according to our observations, the thickening was not related to the degree of vascular compromise. One of the major factors associated with the thicke the returning wall is presumably sion force along the axis of intussusception. The compression may be caused by the wall contraction at the proximal end of intussuscipiens, which pushes the returning wall toward the apex against the restriction of its mesentery and connecting tissue.
Surgical intervention is thought to be unavoidable for the treatment of intussusception in adults, whereas the choice of surgical procedure and the indication for preoperative examination have been controversial [2,3,4, 15]. However, when the degree of vascular compromise is predicted before surgery, optimal treatment can be planned in each case on the basis of CT finding as follows.
CT stage 1 indicates that the returning wall of the intussusceptum is of uniform density without a hypodense layer. The degree of vascular compromise is minimal. Spontaneous reduction appears to be more common in this stage than the others. It seems to be important to distinguish whether a causative tumor is present or absent in intussusception of this stage.
CT stage 2 indicates that the hypodense layer of the returning wall of the intussusceptum is observed without a fluid collection in the space surrounded by the returning wall of the intussusception. The returning wall is edematous, but the change is still reversible.
CT stage 3 indicates that fluid without a gas collection in the space surrounded by the returning wall is observed. The vascular compromise is too severe to avoid resection of the entire length of the intussusception. Preoperative reduction can be skipped.
CT stage 4 indicates that a gas collection in the space surrounded by the returning wall is observed. Perforation or gangrene is strongly suggested. Prompt surgical treatment should be performed.
The limitations of our study, first, include the fact that the evaluations were made retrospectively on the basis of records from many different institutions. However, differences in CT equipment and techniques are negligible because the CT findings that we evaluated are considered to be detectable without the use of any special equipment or techniques. Second, the interval from CT examination to surgery varied among patients. The patients with a long interval before surgery were not excluded from this study, because such an interval is usual in adults with intussusception. Third, the number of patients was limited, especially of those with intussusception with high-grade vascular compromise. Further study should be undertaken in a larger number of patients.
In conclusion, six CT findings were evaluated to determine whether they can predict the vascular compromise in intussusception. The following three CT findings were important for predicting vascular compromise: the hypodense layer in the middle of the returning wall of intussusceptum, a fluid collection in the space surrounded by the returning wall, and a gas collection in the space surrounded by the returning wall. The degree of vascular compromise can be predicted by the classification of four stages based on these three findings. This classification is thought to be helpful in determining the optimal treatment of intussusception. Further study in a prospective series should be performed to prove the validity of CT staging in predicting the degree of vascular compromise.
Acknowledgments
We thank Aamer Aziz for his invaluable help in the revision and correction
of this manuscript.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S Yalamarthi and R C Smith Adult intussusception: case reports and review of literature Postgrad. Med. J., March 1, 2005; 81(953): 174 - 177. [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]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
