DOI:10.2214/AJR.05.2049
AJR 2007; 188:275-279
© American Roentgen Ray Society
Sonography of Pediatric Small-Bowel Intussusception: Differentiating Surgical from Nonsurgical Cases
Martha M. Munden1,
John F. Bruzzi2,
Brian D. Coley3 and
Reginald F. Munden2
1 Edward B. Singleton Diagnostic Imaging Services, Texas Children's Hospital,
Baylor College of Medicine, Houston, TX 77030-2399.
2 Department of Diagnostic Imaging, University of Texas M. D. Anderson Cancer
Center, 1515 Holcombe Blvd., Box 57, Houston, TX 77030.
3 Department of Radiology, Columbus Children's Hospital, Columbus, OH
43205.
Received November 23, 2005;
accepted after revision February 1, 2006.
Address correspondence to R. F. Munden.
Abstract
OBJECTIVE. The purpose of this study was to determine whether there
are clinical or sonographic findings that can be used to differentiate benign
self-limited small-bowel intussusception from pathologic small-bowel
intussusception that necessitates surgical intervention.
MATERIALS AND METHODS. A retrospective search was performed of
abdominal sonograms obtained at two institutions between January 1996 and June
2005. Sonographic findings were correlated with medical and surgical
records.
RESULTS. A total of 35 cases of isolated small-bowel intussusception
were found. Thirteen (37%) of these cases necessitated surgical intervention,
and 22 (63%) of the cases were benign and self-limiting. Patients with
self-limiting intussusception were younger than patients with intussusception
necessitating surgical intervention (mean, 4.2 vs 7.5 years; p =
0.0327). Abdominal sonograms depicted ascites and small-bowel obstruction
significantly more frequently in patients with small-bowel intussusception
necessitating surgery (n = 7 [54%] for each finding) than in patients
with self-limiting intussusception (n =2 [9%], n =0)
(p = 0.006 and p = 0.0003, respectively). At sonography,
patients who later underwent surgical intervention had small-bowel
intussusception of significantly greater length (mean, 7.3 cm) than those
treated conservatively (mean length, 1.9 cm) (p < 0.0001).
Intussusception length greater than 3.5 cm was considered a sensitive and
specific independent predictor of the need for surgery (sensitivity, 93%;
specificity, 100%).
CONCLUSION. When small-bowel intussusception is detected in infants
and children undergoing abdominal sonography, intussusception length greater
than 3.5 cm is a strong independent predictor of the need for surgical
intervention.
Keywords: abdomen • pediatric imaging • small-bowel intussusception • sonography
Introduction
Idiopathic ileocolic intussusception is a medical emergency that has the
classic presentation of acute onset of intermittent, cramping abdominal pain
occurring between periods of lethargy, often associated with hematochezia.
Early reduction by imaging-guided enema (air or water-soluble contrast
material) or by surgery is indicated to prevent irreversible bowel infarction.
Isolated small-bowel intussusception is less common than idiopathic ileocolic
intussusception and often is asymptomatic
[1]. It can, however, manifest
with vague symptoms of irritability, failure to thrive, diarrhea,
gastrointestinal bleeding, abdominal distention, or nonspecific abdominal pain
[2-4]
and may even be incidentally detected in the course of CT or sonography
performed for other reasons. Persistent small-bowel intussusception that
necessitates surgery often has an associated focal lead point, such as
Meckel's diverticulum, intraluminal polyp, duplication cyst, or lymphoma, or
is associated with an underlying condition such as cystic fibrosis, celiac
disease, or Henoch-Schönlein purpura
[5]. Not all cases of
small-bowel intussusception must be reduced by surgical means. Many cases of
intussusception encountered incidentally at sonography or CT are idiopathic
and reduce spontaneously within minutes of detection. For intussusception
lasting longer than the initial sonographic examination, however, it is
difficult to predict the need for further evaluation or intervention on
clinical grounds alone. Lvoff et al.
[6] found that small-bowel
intussusception detected incidentally on CT of adult patients was likely to be
self-limiting if less than 3.5 cm long. We undertook this study to ascertain
clinical and sonographic findings that can be used to differentiate transient,
self-limited small-bowel intussusception from nontransient small-bowel
intussusception, which may necessitate surgical intervention.
View larger version (93K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1 —14-year-old girl with intermittent abdominal pain of several
months' duration. Longitudinal sonogram shows small-bowel intussusception
(arrows) measuring 14 cm in longest dimension. At surgery, lead point
was found to be polyp. Diagnosis was Peutz-Jeghers syndrome. I =
intussusceptum.
|
|
View larger version (108K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2 —13-month-old girl with 5-day history of vomiting and
diarrhea. Transverse sonogram of left lower quadrant shows dilated,
fluid-filled loops of small bowel (black arrows) and ascites
(white arrow). At surgery, Meckel's diverticulum was found to be lead
point.
|
|
Materials and Methods
In a computer-aided search of the database records of Texas Children's
Hospital and Columbus Children's Hospital, all reports containing the word
"intussusception" were retrieved among records of patients
undergoing abdominal sonography between January 1996 and June 2005. Only
patients with isolated small-bowel intussusception were included. Patients
with ileocolic intussusception and intussusception associated with feeding
tubes were excluded. The following clinical data were obtained from medical
records: age, sex, reason for examination, nature of symptoms before
sonography, underlying medical or surgical conditions, subsequent clinical
course, and findings at surgical intervention. Symptoms were defined as acute
if they occurred within 48 hours before presentation and as chronic if they
had lasted longer than 48 hours. In addition, all sonograms from each
examination were reviewed at each institution by two pediatric radiologists
using PACS workstations (iSite, Stentor, GE Healthcare). The following
sonographic parameters were documented: presence of small-bowel
intussusception; length of intussusception in centimeters (measured with
electronic calipers along the greatest longitudinal axis of the
intussusception) (Fig. 1);
presence of sonographic features of small-bowel obstruction (dilated,
fluid-filled loops of small bowel greater than 3 cm in diameter proximal to
the level of intussusception with collapsed distal loops); and presence of
ascites (Fig. 2). Group 1 was
defined as all patients who later underwent surgical reduction of
intussusception. Group 2 was defined as all patients with transient
self-limiting intussusception that did not necessitate surgical intervention.
The study was approved by institutional review boards.
Statistical analysis included descriptive analyses of age, sex, time to
resolution (self-limiting or surgical), length of intussusception, and
frequency of ascites or small-bowel obstruction. The statistical significance
of differences between group 1 and group 2 were calculated with statistical
software (GraphPad Instat, GraphPad Software). Student's t test was
used for continuous variables and the Fisher's exact test for contingency
analysis. A significance level of 5% was used in all cases.
Results
Thirty-five cases of small-bowel intussusception were diagnosed with
sonography between January 1996 and June 2005. The patients with small-bowel
intussusception were 21 girls and 14 boys with a mean age of 5.4 years (range,
6 weeks-18 years). Indications for the examinations included symptoms
referable to the abdomen (abdominal pain [n = 17], vomiting
[n = 8], lethargy or failure to thrive [n = 7], lower
gastrointestinal bleeding [n = 6], weight loss [n = 3],
fever [n = 2], diarrhea [n = 2], hematemesis [n =
1], and constipation [n = 1]). Four patients had no symptoms, and
abdominal sonography was performed for other reasons (evaluation of an
inguinal hernia [n = 1], suspected patent urachus [n = 1],
surveillance after resection of Wilms' tumor [n = 1], and assessment
before liver transplantation [n = 1]). Three patients had
Henoch-Schönlein purpura and were undergoing assessment of abdominal pain
(n = 2) or lower gastrointestinal bleeding (n = 1).
The clinical characteristics of the two groups of patients are summarized
in Table 1. Thirteen (37%) of
the patients with small-bowel intussusception needed surgical intervention,
and 22 patients (63%) were treated conservatively. Patients in group 1 (eight
girls, five boys) were older (mean age, 7.5 years; range, 6 weeks-18 years)
than those in group 2 (13 girls, nine boys; mean age, 4.2 years; range, 18
months-10 years) (p = 0.0327). More patients in group 1 (n =
5 [38%]) than in group 2 (n = 3 [14%]) had acute symptoms, but this
difference was not statistically significant (p = 0.116). There were
no other distinguishing clinical features between the two groups.
View this table:
[in this window]
[in a new window]
|
TABLE 1: Characteristics and Clinical Presentation of Patients with Small-Bowel
Intussusception Detected with Abdominal Sonography
|
|
Among the 13 patients who underwent surgical reduction, presumed lead
points for the intussusception were found at laparotomy in nine (69%) of the
patients. The lead points included Meckel's diverticulum (n = 4),
intraluminal polyps (n = 3), and enlarged lymph nodes (n =
2). In the other four patients, no obvious cause was found. Necrotic bowel was
found at surgery in two patients (Fig.
3), and one of these patients died.
View larger version (109K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3 —18-year-old girl with midepigastric pain for 4 months who
presented with acute severe abdominal pain. Transverse midline sonogram shows
small-bowel intussusception (arrows). Maximum longitudinal
measurement was 14 cm. At surgery, 140 cm of necrotic small bowel was removed.
There was no identifiable lead point.
|
|
Twenty-two cases of small-bowel intussusception resolved without surgical
intervention. Most (n = 13 [59%]) of these cases reduced
spontaneously within seconds to minutes of being detected during the
sonographic examination (Fig.
4). In nine (41%) of the patients, however, intussusception lasted
longer, persisting after the initial sonographic examination had terminated.
Spontaneous reduction of intussusception in these nine patients was confirmed
at subsequent evaluation with sonography (n = 5), CT (n =
2), or small-bowel follow-through (n = 1). In one case follow-up
imaging was not performed. Time to follow-up imaging was as follows:
sonography, 39 minutes and 1, 4, 14, and 24 hours; CT, 7 and 60 days; and
small-bowel follow-through, 9 hours. The patient who did not undergo follow-up
imaging had chronic lethargy and no abdominal pain, and the intussusception
was presumed to be self-limiting. The patient was well in subsequent clinical
follow-up.
View larger version (92K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4 —18-month-old girl undergoing evaluation for patent urachus.
Transverse sonogram shows incidentally found small-bowel intussusception
(arrows) that resolved spontaneously by end of examination. Maximum
longitudinal measurement was 1.4 cm.
|
|
The sonographic findings for the two groups are detailed in
Table 2. Mesenteric
lymphadenopathy was detected in three (14%) of the patients in group 2. No
focal lead points were seen sonographically in any of the other patients in
group 1 or group 2. Both the presence of ascites and the finding of
small-bowel obstruction on sonography were significantly more frequent in
group 1 (n = 7 [54%] for each finding) than in group 2 (n =
2 [9%]; n = 0) (p = 0.006 and p = 0.0003,
respectively), but neither finding was consistently present in all patients
with intussusception that necessitated surgery. Length of small-bowel
intussusception, however, was consistently and significantly greater in group
2 (mean length, 7.3 cm; 95% CI, 4.8-9.7 cm) than in group 1 (mean length, 1.9
cm; 95% CI, 1.7-2.2 cm) (p < 0.0001) and was found to be the
strongest independent predictor of the need for surgical intervention. All
transient and self-limiting intussusceptions were 3.5 cm long or shorter. All
but one of the intussusceptions reduced surgically were longer than 3.5 cm.
The exception was a 2.5-cm intussusception in a 3.5-year-old patient who
underwent surgery because of clinical suspicion of sepsis. The intussusception
was easily reduced at surgery, and no lead point was found. When small-bowel
intussusception was detected with abdominal sonography, intussusception length
greater than 3.5 cm was a sensitive and specific predictor of the need for
surgical intervention, independent of other clinical and sonographic findings
(sensitivity, 93%; specificity, 100%).
Discussion
Transient, self-limiting small-bowel intussusception is not rare in the
pediatric population, accounting for as many as 17% of all cases of
intussusception [1]. More than
one half of cases of transient small-bowel intussusception are asymptomatic
and are often detected incidentally in the course of CT or sonography
performed for other reasons. When present, symptoms are often subacute and
nonspecific, consisting of a spectrum of clinical findings such as
irritability, abdominal distention, abdominal pain, vomiting, fever,
gastrointestinal bleeding, and palpable abdominal mass
[1,
7]. Underlying gastrointestinal
abnormalities or focal lead points causing small-bowel intussusception are
rarely detected sonographically, but these findings include Meckel's
diverticulum, duplication cyst, small-bowel polyp, lymphoma, and jejunal
feeding tubes. More diffuse pathologic findings are lymphoid hyperplasia
secondary to celiac disease, cystic fibrosis, Henoch-Schönlein purpura,
and gastroenteritis
[8-13].
Although rarely detected sonographically, lead points were found at surgery in
eight (42%) of 19 surgical cases in the series described by Ko et al.
[7] and in nine (69%) of our 13
cases. Most cases of small-bowel intussusception are idiopathic and are
thought to be due to benign lymph node hyperplasia, abnormal bowel-wall
motility, bowel-wall thickening, or impaction of secretions.
Our findings suggest that surgery is indicated in the care of patients who
have symptoms and small-bowel intussusception longer than 3.5 cm detected on
abdominal sonography. In the care of patients with incidentally detected
small-bowel intussusception, nonspecific symptoms, and no identifiable
predisposing condition, however, there is much diagnostic uncertainty over
correct management. Many of these patients can be treated conservatively. For
cases in which intussusception does not spontaneously resolve during the
sonographic examination, Kornecki et al.
[1] advocate repetition of
sonography within 45 minutes. On the other hand, to reduce the risk of
small-bowel ischemia and infarction, early surgical intervention is important
for patients with intussusception that is not self-limiting. It is therefore
important at the initial diagnostic examination to be able to predict which
cases of intussusception are benign and self-limiting and which necessitate
aggressive management.
In our patients, we found that length of intussusception was the most
useful independent predictor of the need for surgical intervention. None of
the intussusceptions longer than 3.5 cm had reduced spontaneously by the
follow-up examination. These patients had recurrent or intractable symptoms
and were ultimately treated with surgery. Lead points were found in nine (69%)
of the patients who underwent laparotomy. None of these points had been
detected with sonography. Conversely, all but one of the cases of
intussusception 3.5 cm or shorter at initial sonographic examination resolved
spontaneously. Most (n = 13 [59%]) of these cases reduced
spontaneously within minutes of being detected during the sonographic
examination. In nine (41%) of the patients, intussusception lasted longer than
the initial sonographic examination. The follow-up period after initial
imaging varied from 39 minutes to no follow-up. All of these cases of
intussusception were eventually determined to have resolved. This finding
suggests that intussusception length less than 3.5 cm is predictive of a
benign course and that these patients need only clinical follow-up for
confirmation of an asymptomatic condition.
Our results reflect those of a similar study
[6] of the predictive
usefulness of length of small-bowel intussusception in an adult population in
which length was measured on CT images. In that study, the authors found that
a length of 3.5 cm could be used to differentiate benign self-limiting
intussusception from intussusception necessitating surgery. Sonography may be
a more appropriate imaging technique for measuring the length of
intussusception in a pediatric population. Sonography is often the first
imaging examination to be performed, particularly for pediatric patients, and
can depict intussusception in real time in the longitudinal axis. Sonography
also is better suited to evaluation of changes in the bowel wall, such as
mucosal edema, bowel-wall motion, and the presence of blood flow
[2,
4].
The sonographic features of small-bowel intussusception are similar to
those of ileocolic intussusception, including a doughnut sign of alternating
hypoechoic and hyperechoic rings and a crescent-in-doughnut sign of a
hyperechoic crescent layered around a central mass
[2,
14]. Focal lead points such as
Meckel's diverticulum, duplication cysts, and jejunal feeding tubes are
sometimes well depicted on sonography
[15]. Sonography, however, has
low sensitivity for certain causes of intussusception, such as intraluminal
polyps, and is not always reliable for differentiation of benign lymphoid
hyperplasia and lymphoma [5,
15]. As far as we are aware,
there have been no prospective studies of the relative diagnostic accuracy
rates of CT and sonography in the detection of focal lead points causing
small-bowel intussusception.
In our study, demographic and clinical features were less useful than
length in prediction of the eventual outcome of small-bowel intussusception.
Patients who had small-bowel intussusception necessitating surgery tended to
be older than those with self-limiting conditions, but there was significant
overlap. Therefore, age was not found to be clinically useful in individual
cases. On sonographic examination, the presence ascites was detected more
frequently in the surgical group than in the self-limiting intussusception
group, but this finding was not consistent. Similarly, small-bowel obstruction
occurred only in patients who needed surgery, but this finding probably
reflected the fact that small-bowel obstruction is a more acute clinical
situation that necessitates intervention more immediately than conservative
management. When present, small-bowel obstruction was always associated with
intussusception longer than 3.5 cm, and the absence of small-bowel obstruction
did not preclude surgery. When all factors were considered, intussusception
length greater than 3.5 cm measured on abdominal sonography was the most
important independent predictor of the need for surgical reduction of
small-bowel intussusception.
Limitations of this study mainly arose from the retrospective nature of the
review. Factors such as varying surgical teams and changing surgical practices
over time were not fully accounted for. In addition, sonography is a
relatively operator-dependent tool, and foreshortening of small-bowel
intussusception on sonograms may have introduced a degree of error into the
calculations of length. Intussusception was not measured at surgery in
patients who underwent surgical reduction, and there was no other
corroborating evidence of the length of intussusception. However, the
sonograms represented cases seen routinely in two busy pediatric hospitals,
and the measurements of length are applicable to those commonly obtained in
routine clinical practice. In all cases, the sonograms were examined by an
experienced pediatric radiologist, and the most reliable image was chosen for
measurement. Any error in calculation due to foreshortening was unlikely to
have affected the overall conclusion that the presence of a longer small-bowel
intussusception is more likely to indicate the need for surgical
intervention.
In conclusion, when small-bowel intussusception is detected in infants and
children undergoing abdominal sonography, intussusception length greater than
3.5 cm is a strong independent predictor of the need for surgical
intervention.
References
- Kornecki A, Daneman A, Navarro O, et al. Spontaneous reduction of
intussusception: clinical spectrum, management and outcome. Pediatr
Radiol 2000;30:58
-63[CrossRef][Medline]
- Kim JH. US features of transient small bowel intussusception in
pediatric patients. Korean J Radiol2004; 5:178
-184[Medline]
- Doi O, Aoyama K, Hutson JM. Twenty-one cases of small bowel
intussusception: the pathophysiology of idiopathic intussusception and the
concept of benign small bowel intussusception. Pediatr Surg
Int 2004;20:140
-143[CrossRef][Medline]
- Tiao MM, Wan YL, Ng SH, et al. Sonographic features of small-bowel
intussusception in pediatric patients. Acad Emerg Med2001; 8:368
-373[Medline]
- Navarro O, Daneman A. Intussusception. Part 3. Diagnosis and
management of those with an identifiable or predisposing cause and those that
reduce spontaneously. Pediatr Radiol2004; 34:305
-312[CrossRef][Medline]
- Lvoff N, Breiman RS, Coakley FV, Lu Y, Warren RS. Distinguishing
features of self-limiting adult small-bowel intussusception identified at CT.
Radiology2003; 227:68
-72[Abstract/Free Full Text]
- Ko SF, Lee TY, Ng SH, et al. Small bowel intussusception in
symptomatic pediatric patients: experiences with 19 surgically proven cases.
World J Surg2002; 26:438
-443[CrossRef][Medline]
- Siaplaouras J, Moritz JD, Gortner L, et al. Small bowel
intussusception in childhood [in German]. Klin Padiatr2003; 215:53
-56[CrossRef][Medline]
- Blakelock RT, Beasley SW. The clinical implications of
non-idiopathic intussusception. Pediatr Surg Int1998; 14:163
-167[Medline]
- Carucci LR, Levine MS, Rubesin SE, Laufer I, Assad S, Herlinger H.
Evaluation of patients with jejunostomy tubes: imaging findings.
Radiology2002; 223:241
-247[Abstract/Free Full Text]
- Hui GC, Gerstle JT, Weinstein M, Connolly B. Small-bowel
intussusception around a gastrojejunostomy tube resulting in ischemic necrosis
of the intestine. Pediatr Radiol2004; 34:916
-918[CrossRef][Medline]
- Martinez G, Israel NR, White JJ. Celiac disease presenting as
entero-enteral intussusception. Pediatr Surg Int2001; 17:68
-70[CrossRef][Medline]
- Sonmez K, Turkyilmaz Z, Demirogullari B, et al. Conservative
treatment for small intestinal intussusception associated with
Henoch-Schönlein's purpura. Surg Today2002; 32:1031
-1034[CrossRef][Medline]
- Harris JP, Munden MM, Minifee PK. Sonographic diagnosis of multiple
small-bowel intussusceptions in Peutz-Jeghers syndrome: a case report.
Pediatr Radiol2002; 32:681
-683[CrossRef][Medline]
- Navarro O, Dugougeat F, Kornecki A, et al. The impact of imaging in
the management of intussusception owing to pathologic lead points in children:
a review of 43 cases. Pediatr Radiol2000; 30:594
-603[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
