Dr. Nagai
is a Clinical Instructor and
Dr. Feldstein
is an Associate Professor of Clinical Radiology in the Department
of Radiology, University of California, San Francisco, CA.
Acute abdominal pain is a frequent problem in childhood. The
list of differential diagnostic possibilities is long and varies
significantly with the patient's age and sex. Because it is often
difficult to reach a diagnosis on clinical grounds alone, and
several of the possible diagnoses require immediate surgical
attention, radiologists are often called upon to assist in the
evaluation of these patients.
Historically, standard abdominal roentgenography had been the
cornerstone in the evaluation of acute abdominal pain in children.
While plain film can be useful for finding free intraperitoneal
air, evidence of bowel obstruction, or abnormal abdominal
calcifications, results may be negative in up to 50% of patients.
1
Cross-sectional imaging techniques, including ultrasonography, have
significantly influenced the imaging approach in children with
certain abdominal findings. Three common, and important, clinical
diagnoses that may present with acute abdominal symptoms in
pediatric patients are appendicitis, intussusception, and
hypertrophic pyloric stenosis. For all three of these entities,
ultrasound (US) has been shown to be a reliable imaging tool for
rapid and accurate diagnosis.
Appendicitis
The most frequent surgical diagnosis in children who present
with right lower quadrant pain is acute appendicitis. Historically,
approximately 10% to 25% of children undergoing appendectomy did
not have acute appendicitis at surgery. This false-positive rate
has been considered acceptable, because of the relatively low
morbidity of laparotomy and the serious consequences of a missed
diagnosis of appendicitis. Early diagnosis and surgery are
essential because mortality and late complications increase
significantly if perforation occurs. Perforation with acute
appendicitis is more common in pediatric patients than in adults,
occurring in approximately 40% of children <4 years of age and
in up to 80% of infants. Also, the progression from onset of
symptoms to perforation is more rapid the younger the patient.
The clinical diagnosis of acute appendicitis is usually
suggested by the history of abdominal pain followed by anorexia and
vomiting. The pain is classically periumbilical initially and then
localized to the right lower quadrant. Fever and leukocytosis are
often present. The clinical diagnosis can be especially difficult
in certain pediatric subgroups, namely infants, adolescent girls,
and immunocompromised children.
Plain film of the abdomen can be normal in half of cases of
appendicitis. The presence of a calcified appendicolith is
considered strongly suggestive of acute appendicitis, although is
not specific enough to be the sole basis for diagnosis.
2
In 1986, Puylaert
3
described the use of ultrasonography in the diagnosis of acute
appendicitis. The technique provides direct visualization of the
appendix, using high-resolution sonography with graded abdominal
compression. A high-frequency linear array transducer is used to
evaluate the right lower quadrant, with particular attention to the
site of maximal tenderness. As normal overlying small bowel and
cecum are compressed and air is displaced, the appendix can be
visualized
3
(Figure 1).
A normal appendix appears as a blind-ending tubular structure
with a diameter measuring <6 mm. It can be distinguished from
normal small bowel by lack of peristalsis. A normal appendix
usually appears as a thin central echogenic region corresponding to
the submucosa, surrounded by a hypoechoic outer zone representing
the muscularis propria. Usually the lumen is collapsed, although in
a small percentage of normal appendices, a small amount of
intraluminal fluid may be seen.
4
Ultrasound diagnosis of acute appendicitis rests on finding a
tender, blind-ending, nonperistaltic tubular structure in the right
iliac fossa with a cross-sectional diameter >6 mm (from outer wall to outer wall) with
compression (Figure 2). Hyperemia, the presence of an
appendicolith, and periappendiceal inflammatory changes are
strongly suggestive of the diagnosis, but are nonspecific. Quillin
et al
5
found that increased blood flow in the appendiceal wall or in a
right lower quadrant mass on color Doppler suggests appendicitis.
However, absence of flow did not reliably distinguish a normal from
an abnormal appendix.
5
Other studies confirm the utility of color Doppler sonography in
the diagnosis of acute appendicitis,
6,7
including a study by Gutierrez et al
6
in which 10 of 12 patients with abnormal color Doppler signal in
the right lower quadrant had acute appendicitis, while 100 of 105
patients with no abnormal color flow did not (Figure 3).
Other findings include loss of the echogenic submucosal layer,
fluid-filled appendiceal lumen, increased echogenicity of pericecal
fat, or periappendiceal mass.
4
In focal appendicitis, inflammation may be localized to the distal
tip, highlighting the importance of imaging the entire length of
the appendix. In perforated appendicitis, the caliber of the burst
appendix may be normal. Thus, in these cases, there may be a higher
rate of false-
negative studies, and other indicators of appendicitis should be
sought, including periappendiceal fluid or fluid confined to the
right peritoneal reflection, free pelvic fluid, thickening of
adjacent bowel wall, right lower quadrant adenopathy, or a
well-defined abscess.
8
The reported sensitivity and specificity of the clinical
examination in children suspected to have acute appendicitis are
49% and 95%, respectively.
9
Numerous studies have shown that the sensitivity of US for acute
appendicitis in pediatric populations ranges from 74% to 100% with
specificities ranging from 74% to 95%.
10-15
One group found the sensitivity of US to be remarkably lower, at
44%.
16
Negative laparotomy rates ranging from 0.7% to 8.9% have been
reported when US was used.
9,12,17
Investigators have compared computed tomography (CT) scan and US
in the diagnosis of acute appendicitis in children. These studies
find higher sensitivities with CT (sensitivities 84% to 97%,
specificities 89% to 100%) versus US (sensitivities 44% to 95%,
specificities 88% to 100%), although in some studies the difference
is not statistically significant. Some authors have argued that US
is suitable as a first step in diagnosis, since US has relatively
high sensitivity and specificity compared with clinical
examination, and can triage most patients effectively. In those
cases that are equivocal, the use of limited CT scan with rectal
contrast (CTRC) may be useful for further evaluation. This
combination of US and CTRC in equivocal cases was found in one
study to increase overall sensitivity from 44% (US alone) to 94%
(US followed by limited CTRC).
16
The advantages of US over CT scanning include lack of ionizing
radiation, lack of need for contrast material, wide availability,
and lower cost. Garcia Pena et al
16
found that using US as the first-line imaging in the diagnosis of
acute appendicitis, followed by limited CTRC in equivocal cases,
there was an overall reduction in morbidity and acute care costs.
Some studies have found that US can aid in reducing unnecessary
admissions to the hospital for observation and decreasing the
number of unnecessary appendectomy procedures, without increasing
the risk of perforation.
13,18
There are some limitations of US in the evaluation of acute
appendicitis. In early appendicitis, the sonogram may have a
false-negative result. When clinical suspicion remains high despite
a negative US, serial US examinations can be used to follow the
patient in conjunction with serial clinical evaluation. In patients
with perforation, diagnosis can be made more difficult by
decompression of the appendix. Retrocecal location of the appendix
also makes diagnosis more difficult. Technical factors can also be
limiting. These include extreme abdominal tenderness and guarding,
obesity, and excessive bowel gas. As in other applications of US,
the graded compression technique is operator-dependent; the best
results are obtained when the examiner is experienced in this
procedure. However, despite these limitations, US has proven to be
a rapid and reliable tool for diagnosing acute appendicitis in the
pediatric population.
Intussusception
Intussusception is one of the common surgical emergencies in
infancy and early childhood. Most cases occur in the first 2 years
of life, with a peak occurrence between 3 to 9 months of age.
However, intussusception has become common in children older than 2
years of age, suggesting a changing pattern for this disease.
19
The classic clinical features include blood in the stool and a
palpable abdominal mass. Approximately 15% to 20% of children may
experience no pain. In up to 50% of children, the presenting sign
is listlessness.
In up to 95% of cases, there is no leading point for the
intussusception. The older the child is, the more likely there is a
pathologic leading point, the differential diagnosis of which
includes
lymphoma, Meckel's diverticulum, lymphadenopathy, and
Henoch-Schonlein purpura. In those cases without an identifiable
leading point, it is thought that viral infection leading to
hypertrophy of Peyer's patches may be the cause.
The most common location for intussusception is at the ileocecal
junction, with up to 85% of cases occurring here. Most of the
remaining cases occur in the colon, with rare cases in the small
bowel.
Plain films of the abdomen are often obtained in children
suspected of having intussusception. These may be normal in up to
25% of patients
20
; however, their main utility is in demonstrating free
intraperitoneal air indicative of perforation, for which the
patient would be immediately taken to surgery. Up to 50% of
patients may have an abnormal bowel-gas pattern and suggestion of a
soft tissue mass; 25% may have a pattern suggesting a small bowel
obstruction. The overall accuracy of plain radiography in
diagnosing or excluding intussusception ranges from 40% to 90%.
21
The gold standard for the diagnosis of intussusception is the
contrast enema, which is reported to have an accuracy of 100%.
22
In addition, performance of this study may be therapeutic. However,
for screening of patients suspected of having intussusception, many
of who will not have the condition, US has been proposed as an
alternate method. The main advantages of US include its speed and
accuracy, its noninvasiveness, and the absence of ionizing
radiation. Also, US is well suited for investigation of other
causes of abdominal complaints in children, many of which would not
be seen on contrast enema examination.
In addition to diagnosing intussusception, US can be used to
guide reduction. In one series, US-guided hydrostatic reduction was
performed successfully in 91% (42 of 46) of intussusceptions.
23
The typical sonographic finding is a mass, approximately 3 to 5
cm in diameter, with a hypoechoic intestinal wall surrounding an
echogenic central area, representing submucosa and intussuscepted
fat
24
(Figure 4). On transverse section, this has a target or doughnut
appearance. Multiple concentric rings or the bull's-eye sign of
alternating rings of hypo- and hyperechogenicity are even more
characteristic (Figure 5). On longitudinal section, the hypo-echoic
layers of swollen intestinal wall are seen on either side of the
centrally located intussusceptum (Figure 6). This has been called
the pseudokidney or sandwich sign.
Ultrasound has been shown to be highly accurate in the diagnosis
of intussusception, with sensitivities in the range of 98% to 100%
and specificities from 88% to 100%.
22,25-28
Some studies report the ability to offer alternate diagnoses based
on the US examinations, ranging from 6% to 27% of cases negative
for intussusception.
27,29
The use of color Doppler ultrasonography (CDU) has been
described as a useful tool to assess for bowel necrosis and to
predict reducibility of an intussusception by enema reduction. In a
series of 125 patients with a total of 134 intussusceptions who
underwent both CDU and air enema, the rate of successful reduction
was 91% (109 of 121) in those intussusceptions with blood flow, and
only 31% (4 of 13) in intussusceptions with no flow seen on CDU.
30
Similar correlation between the presence of color flow and
reducibility have been found by other authors.
31,32
In these smaller studies, bowel necrosis was seen only in those
patients with no color flow by CDU, suggesting a strong correlation
between lack of color flow and bowel gangrene.
Hypertrophic pyloric stenosis
The classic presentation of hypertrophic pyloric stenosis (HPS)
is non-bilious vomiting that occurs minutes after feeding, which
worsens over time and may become forceful and projectile. The
infant remains hungry and may feed immediately after vomiting. The
onset of symptoms typically occurs between 3 and 6 weeks of age.
Boys are affected more than girls by a ratio of 5:1.
33
Gastric outlet obstruction is caused by hypertrophy of the antral
and pyloric circular muscle, thought to be an acquired rather than
a congenital condition.
Clinical diagnosis can be made by palpation of the hypertrophied
muscle mass, likened to an olive. In skilled hands, palpation is a
fairly sensitive and highly specific method of detecting HPS. In
one series, physical examination of children with HPS revealed a
palpable abdominal mass in 80% to 87% of cases.
34
However, the art of clinical diagnosis has given way to imaging
diagnosis. There was a decline in the rate of clinical diagnosis
between the years 1974 to 1977 and the years 1988 to 1991, with a
threefold increase in imaging studies for HPS.
34
Plain films are not routinely obtained when the diagnosis of HPS
is suspected. Before the use of US for HPS, upper gastrointestinal
contrast study (UGI) had been the recommended method to establish
the diagnosis. Ultrasound was first introduced as an aid in the
diagnosis of HPS by Teele and Smith
35
in 1977, and is now widely considered the study of choice. The patient is examined in the supine and right lateral
decubitus positions. A 5.0 or 7.5 MHz transducer is applied to the
right upper quadrant, and the pylorus is imaged in its longest
dimension and in transverse plane (Figure 7).
A recent report suggests that placing the infant prone to obtain
posterior views may be useful in imaging the pylorus in cases in
which there is an abundance of overlying bowel gas.
36
The single most reliable measurement by US in the diagnosis of
HPS is muscle wall thickness, measured from the base of the
echogenic submucosa to the outer edge of the hypertrophic muscle
layer (Figure 8). Normal infants have a pyloric muscle wall
thickness ¾ 2 mm, while >4 mm is diagnostic of HPS.
37
The range between 2 and 4 mm is equivocal, and these infants can be
observed with follow-up examinations. Other quantitative criteria include pyloric channel length,
pyloric volume, pyloric volume to body weight ratio, pyloric muscle
index, or pyloric ratio. Recently, the pyloric ratio (wall
thickness to pyloric diameter) was assessed in a retrospective
review of 87 children evaluated for HPS, and a ratio >= 0.27
yielded a sensitivity and specificity of 96% and 94%, respectively.
38
In many cases, measurements and calculations are superfluous, as
an experienced sonographer can establish the diagnosis reliably
based on subjective evaluation. Some nonquantitative findings
include the antral nipple, increased gastric peristalsis,
prepyloric antral thickening, extension of fluid into the proximal
pylorus, and lack of stomach contents traversing the pyloric
channel.
38
In a series of 323 examinations by Blumhagen et al,
37
the sonographic diagnosis, based primarily on muscle wall
thickness, was correct in 321, yielding an accuracy of 99.4%. There
were two false-negative and no false-positive sonograms
(sensitivity 98.2%, specificity 100%). In this study, infants with
palpable pyloric masses were included, introducing a bias toward
higher predictive value of a positive sonogram. A more recent study
included 150 infants with normal or equivocal examinations.
39
Using muscle wall thickness >= 3 mm as diagnostic of HPS, the
sensitivity, specificity, and accuracy of sonography were found to
be 100%. A similar study of 142 infants without a palpable pyloric
mass or with equivocal findings at clinical examination, found an
overall sensitivity of 97% and specificity of 99%.
40
Other diagnoses
Ultrasound aids in the diagnosis of many other pediatric
conditions that may present with acute abdominal pain. Reports have
shown that many patients with acute abdominal symptoms evaluated
for one suspected diagnosis were found on US evaluation to have
other diagnoses. It is important to evaluate the entire abdomen and
pelvis in an effort to discover alternate causes of the clinical
signs and symptoms.
Conclusion
There is much evidence supporting the use of US as a key
diagnostic tool in acute pediatric abdominal pain. In some cases,
new observations have been shown to be useful in determination of
prognosis and therapy, as well. Given all the advantages of US in
the evaluation of pediatric patients, it is not surprising that
ultrasonography has become an increasingly important modality in
this clinical setting.
AR