Intestinal Imaging of Children with Acute

Rotavirus Gastroenteritis
*Dorsey Bass, Eileen Cordoba, *Cornelia Dekker, Anne Schuind, and Christopher Cassady
Department of *Pediatrics, Stanford-LPCH Vaccine Program, Stanford Center for Digestive Diseases, Stanford University,
Stanford, California; GlaxoSmithKline, Collegeville, Pennsylvania; and Department of Radiology, Stanford University,
Stanford, California
ABSTRACT
Objective: To examine the morphology and motility of the
distal small bowel of infants with rotavirus gastroenteritis using
non-invasive/non-ionizing imaging technology.
Methods: Prospective, non-randomized observational study of
five infants with symptomatic rotavirus infection. Infants were
imaged by real-time magnetic resonance imaging (MRI) and
ultrasound within 5 days of onset of gastroenteritis symptoms.
Imaging studies were repeated in the convalescent period 5 to
9 weeks later.
Results: Three of five infants had a significant increase in the
ileal wall thickness visualized by ultrasound during acute ro-
tavirus infection compared with convalescence. The number
and size of mesenteric lymph nodes visualized by ultrasound
appeared similar in the acute and convalescent phases, as did
peristaltic activity assessed by MRI.
Conclusion: Abdominal ultrasound can detect changes in ileal
wall thickness in infants with rotavirus infection. These
changes may reflect ileal inflammation elicited by viral infec-
tion. Such studies may prove useful in evaluating morphologic
response to attenuated rotavirus vaccines. JPGN 39:270274,
2004. Key Words: Rotavirus gastroenteritisMesenteric
lymph nodesIleal wall thickness. 2004 Lippincott Willi-
ams & Wilkins
INTRODUCTION
Rotavirus is the most important cause of pediatric de-
hydrating gastroenteritis in the world (13). Because of
the burden of rotavirus disease, effort has been directed
to the development of safe and effective vaccines. A
recently developed tetravalent rhesus-human reassortant
rotavirus vaccine was withdrawn from the market be-
cause of its association with intussusception (4). The
association with intussusception is a rare event, occur-
ring approximately once in 10,000 vaccine doses (5,6).
In prelicensure studies involving over 5000 children, no
significant association between the vaccine and intussus-
ception was observed despite active surveillance (7). To
test future vaccines for adverse events occurring at such
a frequency, large trials involving hundreds of thousands
of subjects will be required (8). Such large studies are
complex in their implementation and impose prohibitive
expense on vaccine development. A sensitive surrogate
marker of intussusception risk would facilitate the safety
assessment of new rotavirus vaccine candidates.
Intussusception is a common cause of intestinal ob-
struction in infants less than 1 year of age. Although the
vast majority of cases of intussusception are idiopathic,
intestinal masses such as polyps, lymphomas, Meckels
diverticulum, and lymphoid hyperplasia may serve as
lead points. Focal inflammation of the intestinal wall can
also predispose to intussusception, as it does in Henoch
Schnlein purpura. There is some evidence that changes
in gut motility may predispose to intussusception, as
shown by experiments in which intussusception is in-
duced in mice by the parenteral administration of lipo-
polysaccharide (9).
Wild-type rotavirus has been associated with intussus-
ceptions in a few studies (10,11), whereas others have
not found any such association (1013). The lack of sea-
sonality of intussusception and the strong winter season-
ality of rotavirus infections in temperate climates suggest
that rotavirus is probably not a major cause of intussus-
ception (14). Respiratory adenoviruses have been fre-
quently identified in the stool and/or intestinal mucosa of
children with intussusception (15,16).
Radiographic imaging of the intestine might reveal
morphologic changes that could act as markers of the
risk of intussusception. Such markers might include ileal
Supported by the Stanford Digestive Disease Center (NIH
DK56339), The Stanford-LPCH Vaccine Program, NIH DK7762, and
funds from GlaxoSmithKline.
Received October 3, 2003; accepted March 10, 2004.
Address correspondence and reprint requests to Dorsey Bass, MD,
Room G307, Department of Pediatrics, Stanford University Hospital,
Stanford CA 943055208 (e-mail: dorseybass@hotmail.com).
Journal of Pediatric Gastroenterology and Nutrition
39:270274 September 2004 Lippincott Williams & Wilkins, Philadelphia
270
wall thickening, mesenteric lymphadenopathy, and alter-
ations in small bowel motility. We undertook a pilot
study to identify and describe the changes occurring in
the intestine of infants during natural rotavirus infection
in infants using noninvasive radiologic methods.
SUBJECTS AND METHODS
All study procedures were approved by the Stanford Univer-
sity Institutional Review Board. Informed consent was obtained
from parents before entry.
Subjects were recruited from Lucile Packard Childrens Hos-
pital inpatient and outpatient units. Inclusion criteria included
age less than 2 years, onset of acute gastroenteritis less than 5
days before the initial imaging procedure, absence of prior
history of abdominal surgery, immunologic or hematologic dis-
ease. Rotavirus infection was confirmed by a commercial EIA
kit (Rotaclone, Meridian Laboratories, Cincinnati, OH) in all
patients at the time of study entry.
Parents maintained clinical symptom diaries to assess sever-
ity scores of the gastroenteritis disease as described by Ruuska
and Vesikari (17).
Initial imaging studies were obtained within 5 days of the
onset of gastroenteritis symptoms and follow-up studies were
performed 5 to 9 weeks later (convalescent data). Thirty min-
utes before imaging infants were offered 8 ounces of blueberry
juice, which was used as a contrast media for the MRI be-
cause of its natural metal content. For studies in the convales-
cent period, 8 ounces of juice offered contained 7.5 g of Mi-
ralax (Braintree Labs, Braintree MA), a mild osmotic laxative
that consists of polyethylene glycol 3350, to ensure presence of
fluid in the distal ileum for adequate ultrasonographic imaging.
Measurements and interpretation of radiologic images were
performed by a single pediatric radiologist.
Magnetic resonance imaging (MRI) was performed with an
open-bore configuration General Electric Signa 0.5 Tesla MR
imager scanner (GE, New York, NY). Parents and medical
personnel were allowed to attend the infants during imaging,
which was performed without sedation. The imager and pro-
prietary software had the unique capability to produce fluo-
roscopic rapid MR image acquisition and display. At an ac-
quisition rate of 3.5 frames/second and construction and display
at 12 frames/second, we were able to perform MRI in a real-
time mode that compares with conventional fluoroscopy with-
out the use of ionizing radiation. To investigate whether gross
peristaltic activity was affected by rotavirus, peristaltic waves
over a 10-minute period were quantitated and expressed as
waves per minute.
Ultrasonography was performed using a high-resolution (15
MHz) linear transducer. Images of the terminal ileum were
obtained by scanning the right lower quadrant. Wall measure-
ments were performed in a standardized fashion from maxi-
mally distended intestinal loops. Five independent measure-
ments were taken of the ileal wall thickness while the bowel
lumen was longitudinal to the transducer plane (Figure 1). In
addition, mesenteric lymph nodes were assessed quantitatively
for number and size.
Two-sample Students t test was used to compare means of
ileal wall thickness, mesenteric node size and number of mes-
enteric nodes.
RESULTS
Patient Characteristics
Five subjects (three female, two male) were studied
between February 1 and May 15, 2001. Ages ranged
from 8 to 20 months (mean, 14 2.2 months) (Table 1).
Three of the children were hospitalized for management
of moderate to severe dehydration. Vesikari scores av-
eraged 15.8 2.4 with all five subjects meeting criteria
for severe disease. All subjects had complete resolution
of symptoms by 7 days after the onset of gastroenteritis.
MRI Fluorography
Four of five subjects completed the MRI fluorography
portion of the analysis. Subject 5 was unable to complete
FIG. 1 A, ultrasound imaging in the acute phase of rotavirus
gastroenteritis. Adjacent ileal walls are measured. The bowel is
fluid filled. Images are obtained with a high-resolution linear
transducer in the right lower quadrant. Each measured distance
is 1.5 mm. B, convalescent repeat imaging in the same subject.
The ileal wall is measured 5 weeks after initial acute study. The
wall thickness is 1.0 mm. Short arrow indicates two adjacent thin
bowel walls as compared with the initial examination in panel A.
INTESTINAL IMAGING OF CHILDREN 271
J Pediatr Gastroenterol Nutr, Vol. 39, No. 3, September 2004
the MRI examinations for technical reasons. No particu-
lar anatomic abnormalities were observed (Fig. 2). Four
subjects completed both MRI examinations. Two sub-
jects had higher frequencies of peristalsis during acute
rotavirus infection and two had higher frequencies when
observed at the convalescent visit 4 to 9 weeks later.
Ultrasonigraphic Anatomy
High-resolution images of the ileal wall were obtained
in all subjects (Fig. 1). Figure 3A shows the number of
mesenteric lymph nodes identified for each subject, com-
paring the acute and convalescent examinations. Al-
though there was a trend toward an increased number of
nodes during the acute phase, there was no significant
difference between the time points. Similarly, when the
size of mesenteric lymph nodes was compared (Figure
3B), there was a trend, which did not reach statistical
significance, toward larger nodes during the acute phase.
No intussusception was noted in any examination.
The acute rotavirus examinations showed an increase
in ileal wall thickness of 20% to 70% compared with
measurements obtained in the convalescent phase (5 to 9
weeks later) in four of five subjects (Fig. 1). Another
subject had 5% change in the measured thickness (Figure
4 and Table 2). When the acute and convalescent ileal
wall thicknesses were compared for each subject, signifi-
cant differences were noted in three of five subjects.
DISCUSSION
Because of the recent association of a tetravalent
rhesus-human reassortant rotavirus vaccine with intus-
TABLE 1. Characteristics of subjects
Subject
number
Age,
months
(gender) Hospitalized Date entered
Vesikari/Ruska
GE illness
score (020)
1 13 (F) Y 02/08/01 18
2 11 (M) Y 02/09/01 17
3 20 (F) N 02/14/01 17
4 8 (M) N 02/28/01 12
5 18 (F) Y 05/11/01 15
FIG. 2. Frequency of peristaltic waves in children with acute ro-
tavirus gastroenteritis. Infants underwent MRI fluorography as
described under materials and methods. The mean number of
peristaltic waves per minute was calculated based on 10 minutes
of observation during the acute illness () and 5 to 9 weeks later
().
FIG. 3. Mean longitudinal diameter and number of mesenteric
lymph nodes. Subjects underwent ultrasound examinations of the
right lower quadrant as described in materials and methods dur-
ing acute rotavirus gastroenteritis and subsequently 5 to 9 weeks
later. Lymph nodes were identified, enumerated and measured in
their shortest axis. A, mean diameter; b, total nodes identified.
BASS ET AL. 272
J Pediatr Gastroenterol Nutr, Vol. 39, No. 3, September 2004
susception in infants, we were interested in examining
noninvasive methods for detecting changes in the mor-
phology and/or motility of the intestine that might pre-
dispose infants to intussusception. A further aim of this
study was to explore whether wild-type rotavirus infec-
tion might induce detectable changes in the distal ileal
wall and associated mesenteric lymph nodes. Knowledge
of such changes might help in assessing safety for future
rotavirus vaccines.
Intussusception is the most common cause of intesti-
nal obstruction in infants beyond the first month of life.
Despite this fact, very little is understood about the
pathophysiology of intussusception. It occurs most com-
monly in children 3 to 9 months of age. The majority of
cases are considered idiopathic in this age group. Most
cases are treated with hydrostatic or barostatic reduction.
In cases which require laparotomy for reduction or bowel
resection, irregularities of the bowel wall such as lym-
phoid hyperplasia and congenital abnormalities of the
bowel wall are reported as pathologic lead points. The
significance of mesenteric lymph nodes in producing in-
tussusception is not clear. Such nodes could conceivably
act as lead points, although this is rarely reported by
surgeons. Mesenteric nodes could alternatively be a re-
flection of simultaneous lymphonodular hyperplasia in
the bowel itself. A recent report suggests that larger mes-
enteric nodes (as measured by ultrasound) predict a
lower rate of successful hydrostatic reduction of intus-
susception (18).
There are few reports of intestinal imaging in children
with acute gastroenteritis. In one, ultrasound examina-
tions were obtained on children with either Salmonella
enterocolitis or rotavirus infection (19). No significant
findings were reported in the rotavirus infected children,
whereas the Salmonella infected subjects with the most
severe disease had significant colonic thickening and as-
cites. Unfortunately, the study did not include convales-
cent imaging to detect persistent changes and focused
mainly on the colon rather than the small intestine.
Several recent papers have examined the sensitivity
and specificity of abdominal ultrasound in evaluation of
the small intestine in other diseases. Pallota described the
use of isotonic nonabsorbable polyethylene solution to
facilitate detailed imaging of the small bowel (20). Fol-
vik found that this ultrasound technique compared quite
favorably with traditional barium contrast radiography in
56 patients evaluated for abdominal pain (21). Excellent
agreement has also been reported between ultrasound,
radiography and endoscopy in a group of 240 prospec-
tively studied adult subjects with inflammatory lesions
and tumors of the bowel (22).
To determine the ability of these radiographic tech-
niques to detect changes in the distal small bowel of
infants, we recruited subjects less than 2 years of age
with moderate to severe rotavirus disease. This young
group was selected to be most representative of the age
group that would be immunized with rotavirus vaccine.
Because there are no standards for ileal wall thickness in
this age group, we used the subjects as their own controls
after recovery from gastroenteritis. Our results support
this approach as the bowel wall thickness decreased in
four of five patients, suggesting that five to nine weeks
after resolution of diarrhea, the bowel wall had returned
to a normal state. The acute increase in ileal wall thick-
ness we observed was significant in three of the five
subjects. No evidence of intussusception was noted. Our
study also showed that ultrasound can be used to enu-
merate and measure mesenteric lymph nodes. The small
number of subjects enrolled precludes further interpreta-
tion of our data on mesenteric nodes. The physiologic
and pathologic correlates of the observed ileal wall thick-
ening are not known. Edema resulting from inflamma-
tion and lymphoid hyperplasia are possible hypotheses.
The significance of these findings with respect to the risk
for intussusception is unknown.
We also performed acute and convalescent MRI fluo-
rography in an effort to assess by non-invasive means
TABLE 2. Comparison of mean wall thickness between the
acute and the convalescent phases
Subject
number
Acute
mean
(mm)
Convalescent
mean (mm)
Difference
(mm)
Difference
(%) P value
1 1.36 1.10 0.26 24 0.005
2 1.32 0.82 0.50 61 0.002
3 1.46 0.86 0.60 70 0.001
4 0.84 0.80 0.04 5 0.72
5 1.06 0.88 0.18 20 0.17
FIG. 4. Mean ileal wall thickness. Thickness of the ileal wall was
obtained by ultrasound from longitudinal sections of maximally
fluid distended ileum. Measurements were made from five dis-
tinct points. Mean standard deviations for acute and convales-
cent measurements are depicted. *P < 0.05, acute versus con-
valescent measurements.
INTESTINAL IMAGING OF CHILDREN 273
J Pediatr Gastroenterol Nutr, Vol. 39, No. 3, September 2004
any intestinal motility changes that might result from the
acute rotavirus infection. Results of the real-time MRI
examinations did not provide evidence for viral induced
changes in intestinal motility in our subjects. Myoelec-
trical measurements of intestinal motility in animals with
viral gastroenteritis have suggested disordered propul-
sive activity (23). Real-time MRI may not be a sensitive
measurement of such changes.
This pilot study suggests that ultrasound is a reason-
able tool for investigating the morphology of the lower
small intestine in infants and that severe rotavirus infec-
tion may cause morphologic changes in this region. Im-
aging studies of children with routine gastroenteritis are,
of course, not clinically indicated but may provide im-
portant information in structured research protocols. This
pilot study is insufficiently powered to make conclusions
regarding the relationship between infection severity and
degree of ileal thickening or to rule out the possibility
that mesenteric nodes may vary in size and number dur-
ing rotavirus infection and recovery. Further studies are
underway to extend our preliminary observations. It will
be particularly interesting to discern if such changes oc-
cur in children with milder natural rotavirus infections,
which might more closely mimic immunization with at-
tenuated rotavirus vaccines.
Acknowledgment: The helpful advice and comments of
Bruce Innis of GlaxoSmithKline are gratefully acknowledged.
REFERENCES
1. Coffin S. Rotavirus vaccines: current controversies and future di-
rections. Curr Infect Dis Rep 2000;2:6872.
2. Bernstein D. Rotavirus vaccine: current status and future prospects.
Biodrugs 2000;14:27581.
3. Institute of Medicine. The prospects for immunizing against rota-
virus. In: New Vaccine Development: Establishing Priorities. Dis-
eases of Importance in Developing Countries. Washington, D.C.:
National Academy Press; 1986:30818.
4. Withdrawal of rotavirus vaccine recommendation. MMWR Morb
Mortal Wkly Rep 1999;48:1007.
5. Murphy TV, Gargiullo PM, Massoudi MS, Nelson DB, et al. In-
tussusception among infants given an oral rotavirus vaccine.
N Engl J Med 2001;344:56472.
6. Simonsen L, Morens D, Elixhauser A, Gerber M, Van Raden M.,
Blackwelder W. Effect of rotavirus vaccination programme on
trends in admission of infants to hospital for intussusception. Lan-
cet 2001;358(9289):12249.
7. Rennels MB, Parashar UD, Holman RC, Le CT, Chang HG, Glass
RI. Lack of an apparent association between intussusception and
wild or vaccine rotavirus infection. Pediatr Infect Dis J 1998;17:
9245.
8. Jacobson RM, Adegbenro A, Pankratz VS, Poland GA. Adverse
events and vaccination-the lack of power and predictability of
infrequent events in pre-licensure study. Vaccine 2001;19:
242833.
9. Lin Z. Bacterial wall lipopolysaccharide as a cause of intussuscep-
tion in mice. J Pediatr Gastroenterol Nutr 1998;27:3015.
10. Konno T, Suzuki H, Kutsuzawa T, et al. Human rotavirus and
intussusception. N Engl J Med 1977;297:945.
11. Konno T, Suzuki H, Kutsuzawa T, et al. Human rotavirus infection
in infants and young children with intussusception. J Med Virol
1978;2:2659.
12. Mulcahy DL, Kamath KR, de Silva LM, Hodges S, Carter IW,
Cloonan M. A two-part study of the aetiological role of rotavirus
in intussusception. J Med Virol 1982;9:515.
13. Nicolas JC, Ingrand D, Fortier B, Bricout F. A one-year virological
survey of acute intussusception in childhood. J Med Virol 1982;
9:26771.
14. Parashar UD, Holman RC, Cummings KC, et al. Trends in
intussusception-associated hospitalizations and deaths among US
infants. Pediatrics 2000;106:141321.
15. Porter HJ, Padfield CJ, Peres, LC, Hirschowitz L, Berry PJ. Ad-
enovirus and intranuclear inclusions in appendices in intussuscep-
tion. J Clin Pathol 1993;46:1548.
16. Hsu HY. Viral etiology of intussusception in Taiwanese childhood.
Pediatr Infect Dis J 1998;17:8938.
17. Ruuska T, Vesikari T. Rotavirus disease in Finnish children: use of
numerical scores for clinical severity of diarrhoeal episodes. Scand
J Infect Dis 1990;22:25967.
18. Koumanidou C, Vakaki M, Pitsoulakis G, Kakavakis K, Mirilas P.
Sonographic detection of lymph nodes in the intussusception of
infants and young children: clinical evaluation and hydrostatic re-
duction. AJR Am J Roentgenol 2002;178:44550.
19. Ueda D, Sato T, Yoshida M. Ultrasonographic assessment of Sal-
monella enterocolitis in children. Pediatr Radiol 1999;29:46971.
20. Pallotta N. Contrast ultrasonography of the normal small bowel.
Ultrasound Med Biol 1999;25:133540.
21. Folvik G. Hydrosonography of the small intestine: comparison
with radiologic barium study. Scand J Gastroenterol 1999;34:
124752.
22. Bozkurt T, Richter F, Lux G. Ultrasonography as a primary diag-
nostic tool in patients with inflammatory disease and tumors of the
small intestine and large bowel. J Clin Ultrasound 1994;22:8591.
23. Burrows CF, Merritt AM. Influence of coronavirus (transmissible
gastroenteritis) infection on jejunal myoelectrical activity of the
neonatal pig. Gastroenterology 1984;87:38691.
BASS ET AL. 274
J Pediatr Gastroenterol Nutr, Vol. 39, No. 3, September 2004