European Journal of Radiology 82 (2013) e1–e10

Contents lists available at SciVerse ScienceDirect

European Journal of Radiology

journal homepage: www.elsevier.com/locate/ejrad

Computed tomography of the bowel: A prospective comparison study between

four techniques
Francesco Paparo a , Alessandro Garlaschi b , Ennio Biscaldi a , Lorenzo Bacigalupo a ,
Luca Cevasco b , Gian Andrea Rollandi a,∗
a
Department of Radiology, E.O. Ospedali Galliera, Mura della Cappuccine 14, 16128 Genoa, Italy
b
School of Radiology, University of Genoa, Via Leon Battista Alberti 4, 16132 Genoa, Italy

a r t i c l e i n f o a b s t r a c t

Article history: Objectives: The major objective was to prospectively compare the grade of bowel distension obtained
Received 4 June 2012 with four different computed tomography (CT) techniques dedicated for the examination of the small
Received in revised form 28 July 2012 intestine (CT enteroclysis [CTE] and enterography [CTe]), of the colon (CT with water enema [CT-WE]), or
Accepted 26 August 2012
both (CTe with water enema [CTe-WE]). The secondary objective was to assess patients’ tolerance toward
each CT protocol.
Keywords:
Materials and methods: Recruitment was designed to obtain four groups of the same number of patients
Computed tomography enteroclysis
(30). Each group corresponded to a specific CT technique, for a total of 120 consecutive outpatients (65
Computed tomography enterography
Computed tomography with water enema
male and 55 female, mean age 51.09 ± 13.36 years).
Computed tomography enterography with CTE was performed after injection of methylcellulose through a nasojejunal tube, while in the CTe
water enema protocol a polyethylene glycol electrolyte solution was orally administered to patients prior to the
CT acquisition. In the CT-WE protocol intraluminal contrast (water) was administered only by a rec-
tal enema, while CTe-WE technique included both a rectal water enema and oral ingestion of neutral
contrast material to obtain a simultaneous distension of small and large bowel.
CT studies were reviewed in consensus by two gastrointestinal radiologists who performed a quan-
titative and qualitative analysis of bowel distension on a per segment basis. The presence and type of
adverse effects were recorded.
Results: CTE provided the best distension of jejunal loops (median diameter 27 mm, range 17–32 mm)
when compared to all the other techniques (p < 0.0001). The frequency of patients with an adequate
distension of the terminal ileum was not significantly different among the four groups (p = 0.0608). At
both quantitative and qualitative analysis CT-WE and CTe-WE determined a greater and more consistent
luminal filling of the large intestine than that provided by both CTE and CTe (p < 0.0001 for all colonic
segments). Adverse effects were more frequent in patients belonging to the CTE group (p < 0.0028).
Conclusions: CTE allows an optimal distension of jejunal loops, but it is the most uncomfortable CT proto-
col. When performing CT-WE, an adequate retrograde distension of the terminal ileum was provided in
a particularly high percentage of patients. CTe-WE provides a simultaneous optimal distension of both
small and large bowel.
© 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
A critical technical requirement for the CT study of the bowel
is the full distension of a clear lumen with complete separation of
the intestinal walls, and such a requisite is valid for both the small
and large bowel, because collapsed bowel loops can hide even large
Abbreviations: CTE, computed tomography enteroclysis; CTe, computed tomo-
graphy enterography; CT-WE, computed tomography with water enema; CTe-WE,
computed tomography enterography with water enema.
∗ Corresponding author. Tel.: +39 010 5634810.
E-mail address: gianandrea.rollandi@galliera.it (G.A. Rollandi).
lesions (i.e. tumors and polyps) and may falsely mimic wall thicken-
ing, thus simulating pathologic conditions (i.e. inflammatory bowel
diseases).
CT techniques designed for bowel imaging are all characterized
by the administration of intraluminal contrast material to provide
distension of the lumen and separation of the intestinal walls; the
way of contrast administration is the primary distinguishing fea-
ture between CT protocols dedicated to examine the small intestine
(i.e. CTE and CTe) and that optimized for the study of the colon
(CT-WE, also called hydrocolon-CT).
In both CTe and CTE small bowel distension is provided by
the anterograde administration of intraluminal contrast mate-
rial (through a nasojejunal tube in CTE and oral ingestion in

0720-048X/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ejrad.2012.08.021
e2 F. Paparo et al. / European Journal of Radiology 82 (2013) e1–e10
CTe), while mural enhancement is obtained with the injection of
intravenous iodinated contrast media, usually performing a single
CT acquisition in the portal venous phase (45 s after the arte-
rial peak in the upper abdominal aorta using a bolus-tracking
software) [1,2].
In CTE technique neutral contrast agents (usually water or
methylcellulose) are manually injected through the naso-jejunal
catheter or by means of a semi-automatic injector pump in a
variable amount (from 1500 up to 2500 mL), depending on the
tolerance of each patient [3].
In CTe negative contrast media that are most widely accepted
for oral use include water, oil emulsions, solutions containing sugar
alcohols, such as manitol or sorbitol, and polyethylene glycol solu-
tions [4–6].
When performing both CTE and CTe the enteric contrast mate-
rial may pass through the ileocecal valve, providing a variable grade
of large bowel distension, but the real capability of these techniques
to allow an adequate exploration of all colonic segments remains
largely undetermined.
In some clinical concerns the CT examination has to be focused
on the study of the colon. The most widely accepted CT technique
for large bowel evaluation is CT-colonography with air or carbon
dioxide [7–9], but the colonic distension may also be obtained using
water as intraluminal contrast media.
This technique, called CT with water enema – CT-WE – (or
hydrocolon-CT) is based on colonic distension with water fol-
lowed by intravenous administration of iodinated contrast media;
it offers an excellent visualization of the colonic wall due to parietal
enhancement by iodine contrast, as well as a good contrast between
wall, water-filled lumen and pericolic fat [10,11].
CT-WE may provide an optimal distension of the large bowel. In
CT-WE the neutral contrast is administered through a rectal enema
tube, and it may sometimes reflux through the ileocecal valve from
the cecum into the terminal ileum, revealing pathologic findings of
the last ileal loop. Despite the retrograde distension of the terminal
ileum represents a well known eventuality, its true consistency is
still unknown.
More recently a new CT technique, called CTe-WE (CT enterog-
raphy with water enema), has been proposed to obtain a
simultaneous distension of the small and large intestine. This orig-
inal technique has been employed to evaluate a cohort of 221
patients with Crohn’s disease, providing a complete assessment of
disease phenotype (localization [upper gastrointestinal, ileal, ileo-
colic, colonic, perianal] and behavior [inflammatory, stricturing,
penetrating]) [12].
The aim of the present study was to prospectively compare
the grade of distension of the different bowel segments (including
both small and large bowel) obtained using the above mentioned
four different CT techniques (CTE, CTe, CT-WE and CTe-WE), pro-
viding also data concerning patient’s tolerance toward each CT
protocol.
2. Patients and methods
2.1. Inclusion of patients
This was a prospective, institutional review board-approved
study of comparison between four CT techniques (CTE, CTe, CT-
WE, CTe-WE), which was performed in our Radiology Department
from October 2009 to December 2010.
A total of 120 consecutive outpatients (65 males and 55 females,
mean age 51.09 ± 13.36 years) with suspected or known patholo-
gies involving the gastrointestinal tract (including both small and
large bowel) were recruited. All patients enrolled in the study gave
written informed consent.
Inclusion of patients was designed to obtain four groups with the
same number (30 subjects each one), each group corresponding to
a specific, clinically indicated CT protocol.
Patients were assigned by our study coordinator (G.A.R.) to
undergo one of the four CT protocols on the basis of the clinical
question in their referral form.
The first consecutive 60 patients with a clinical indication for the
CT examination of the small bowel were randomly assigned by the
study coordinator to undergo either the CTE or the CTe protocol; so,
for both CTE and CTe, inclusion criteria were the same: suspected
or known Crohn’s disease, suspected carcinoid tumor, abdominal
pain of unknown origin, diarrhea, obscure gastrointestinal bleeding
in patients with negative upper and lower endoscopy. If a patient
refused the assigned CTE protocol (with placement of the naso-
jejunal tube), he was given the option of choosing the CTe protocol,
and it occurred in only three cases.
The first consecutive 30 patients who had to undergo a clini-
cally indicated CT investigation of the large bowel were assigned to
the CT-WE protocol; inclusion criteria were as follows: diverticuli-
tis, suspected or known ulcerative colitis, detection of colo-rectal
polyps, preoperative localization and staging of colorectal can-
cer detected at videocolonoscopy, positive fecal occult blood test,
chronic iron deficiency anemia, cramping or acute abdominal pain
of lower abdominal quadrants, and hematochezia.
Inclusion criteria for the CTe-WE group were comprehensive of
those of the other three groups, including patients with suspected
or known diseases affecting the small and/or the large bowel.
The mean age of patients enrolled in the CTE group was
48.63 ± 13.98 years (12 male and 18 female). In 13 CTE exami-
nations were present signs of active Crohn’s disease affecting the
ileum; 2 CTE demonstrated a small bowel tumor (1 adenocarci-
noma of the jejunum and 1 carcinoid tumor of the ileum); one
CTE revealed a single polyp of the ileum; in 14 CTE there were no
pathologic findings related to the small bowel.
The mean age of patients included in the CTe group was
49.47 ± 13.94 years (19 males and 11 females). In 17 CTe exami-
nations were present signs of active Crohn’s disease with an ileal
distribution; one CTe demonstrated an intestinal lymphoma of the
proximal jejunum, and one CTe a single polyp of the ileum; in 11
CTe there were no pathologic findings related to the small intestine.
The mean age of patients included in the CT-WE group was
56.13 ± 11.74 years (15 males and 15 females). In 8 CT-WE were
present signs of diverticulitis of the sigmoid colon; 2 CT-WE
allowed the detection of an adenocarcinoma of the sigmoid colon;
4 CT-WE were performed for staging an endoscopically detected
adenocarcinoma of the large bowel (2 adenocarcinomas of the
cecum and 2 adenocarcinomas of the sigmoid colon); 4 CT-WE
revealed polyps of the large intestine; one CT-WE revealed syn-
chronous multifocal carcinoid tumors (an appendicular carcinoid
and a carcinoid of the ileum); 2 CT-WE revealed signs of ulcerative
colitis; one CT-WE demonstrated an angiodysplasia of the cecum;
in 8 CT-WE there were no pathologic findings related to the large
intestine.
The mean age of patients included in the CTe-WE group was
51.13 ± 12.86 years (14 males and 16 females). In 14 CTe-WE were
present signs of active Crohn’s disease, including 7 cases of ileal
Crohn’s disease and other 7 cases of ileocolic Crohn’s disease; 3
CTe-WE revealed signs of diverticulitis of the sigmoid colon; one
CTe-WE demonstrated an appendicular carcinoid; 3 CTe-WE were
performed for staging of an endoscopically detected adenocarci-
noma of the large intestine (1 adenocarcinoma of the descending
colon and 2 adenocarcinomas of the sigmoid colon); 2 CTe-WE
revealed polyps of the large intestine (one polyp of the cecum and
one polyp of the descending colon); one CT-WE demonstrated an
angiodysplasia of the cecum; in the remaining 7 CTe-WE there were
no pathologic findings related to the intestine.
 F. Paparo et al. / European Journal of Radiology 82 (2013) e1–e10
Exclusion criteria included: the clinical suspicion and/or
conventional radiographic findings suggestive for bowel (sub-)
obstruction, age < 18 years, previous bowel resection, preg-
nancy, contraindication to intravenous injection of hyoscine-
N-butylbromide, and general contraindications to intravenous
administration of iodinated contrast media (i.e. eGFR < 45 mL/min
and documented previous reaction to iodinated contrast mate-
rial). The first 120 patients who agreed to participate and provide
informed consent were enrolled in the study.
2.2. CT protocols
All examinations were performed with a 64-slice multidetec-
tor CT scanner (Light Speed VCT, GE Medical Systems, Milwaukee,
WI) with the patient in supine position. Contrast-enhanced CT was
performed using the following scanning parameters: collimation
1.25 mm, table speed 13.75 mm per rotation, pitch 1.375, 120 kVp,
300 mAs. Bowel wall enhancement was produced by intravenous
injection of iodinated contrast medium with an iodine concen-
tration ranging between 350 and 370 mg/mL (iobitridol, xenetix
350, guerbet; or iopamidol, iopamiro 370, bracco). The iodine flow
injected per second per kilogram of body weight was maintained
constant for all examinations (1.11 g I/s). The flow rate was set at
3.2–3 mL/s with an automatic injector and acquisition was started
in the portal phase, 45 s after the arterial peak in the upper abdom-
inal aorta using a bolus-tracking software. Immediately before CT
acquisition, bowel hypotonia was obtained by i.v. injection of 2 mL
hyoscine-N-butylbromide 20 mg/mL (Buscopan, Boehringer Ingel-
heim). The estimated mean effective dose for each CT protocol was
16.4 mSv, except for the CTE protocol which includes an additional
radiation dose due to the nasojejunal tube placement under fluo-
roscopic guidance.
Bowel cleansing was standardized for all groups of patients:
a low fiber diet for 3 days before the CT examination, and oral
administration of 2 L of an isotonic non-absorbable electrolyte solu-
tion containing polyethylene glycol (Isocolan, Giuliani S.P.A, Milan,
Italy) the afternoon before examination.
2.2.1. CT-enteroclysis (CTE)
Patients enrolled in the conventional invasive CTE protocol
underwent nasojejunal intubation with a 13-F 155-cm nasoje-
junal tube with an anti-reflux balloon, which was followed by
hand injection of 1800–2000 mL of neutral contrast material
(0.5% methylcellulose) administered at a rate of approximately
60 mL/min before CT acquisition. The use of an anti-reflux balloon
may help to avoid the reflux of contrast material into the duo-
denum and gastric antrum. The mean time required to place the
nasojejunal tube was 10–15 min with a maximum time of pulsed
fluoroscopy of 3 min, and the average effective dose applied to the
patients was 5.33 mSv. The total mean effective dose applied during
the CTE protocol was 21.73 mSv.
2.2.2. CT-enterography (CTe)
In the CTe protocol small bowel lumen distension was achieved
by oral administration of 1500–2000 mL of a non-absorbable iso-
tonic solution containing polyethylene glycol (Isocolan) 45 min
prior to CTe-WE. All the volume of neutral enteral contrast mate-
rial had to be drunk in a time interval not superior to 15–20 min.
Patients were instructed about the importance of continuous intake
of the oral contrast and monitored by a nurse or technician as they
ingested the contrast. Discontinuous intake or poor timing of oral
contrast may almost certainly lead to a poor study.
2.2.3. CT-with water enema (CT-WE)
All CT-WE examinations were performed with the following
technique. A lubricated enema tube was inserted into the rectum.
e3
The tube was connected to a bag containing 2000 mL of lukewarm
tap water, which was gently infused through gravity in 3–4 min,
with the patient placed supine on the CT table.
2.2.4. CT-enterography with water enema (CTe-WE)
When performing CTe-WE both small and large bowel were dis-
tended with neutral enteral contrast material. Small bowel lumen
distension was achieved using the same technique of CTe (oral
ingestion of polyethylene glycol solution), while large bowel dis-
tension was obtained according to the CT-WE protocol, as described
in a previous work [12].
2.3. Image analysis
All CT examinations were reviewed on a dedicated workstation
(ADW4.5, General Electric Medical Systems), using the 1.25 mm
thickness reconstructed images to obtain multi planar recons-
tructions (MPR). Both quantitative and qualitative analyses were
performed in consensus by two radiologist (FP, AG) with a 5-year
and 10-year experience in abdominal imaging respectively. They
were blinded to the type of CT protocol of each examination, and
CTs were reviewed in a random order.
For purposes of analysis, the small bowel was divided into
jejunum, ileum, and terminal ileum according to Arslan et al. [13].
The anatomy of the intestine was primarily assessed by means of
coronal reformatted images. The small bowel occupies the inframe-
socolic space of the peritoneal cavity. The jejunum generally is
located in the left upper and mid quadrants, while the ileum occu-
pies the right mid- and lower portions of the abdominal cavity
including the area above the pelvic inlet. Jejunal loops may be dis-
tinguished from that of the proximal ileum due to their anatomical
localization and for the presence of more circular folds (valvu-
lae conniventes) per centimeter. It is known that, on average, the
jejunum presents four to seven folds every 2.5 cm, and the ileum,
three to five folds in the same length [14]. Distal jejunal loops cross
the spine to the right side and continue as the proximal ileum. The
last ileal loop has usually a right cephalic orientation and it may be
correctly visualized on the coronal plane.
The large intestine was divided into six segments: rectum, sig-
moid colon, descending colon, transverse colon, ascending colon,
and cecum, as proposed by Ajaj et al. [15].
In order to identify a potential source of bias in the assessment of
true luminal filling, CT examinations were preliminarily analyzed
to find out the presence of inflammatory or neoplastic strictures of
the small intestine that may have artificiously improved the disten-
sion proximally. A prestenotic dilatation was considered significant
when the small bowel lumen (proximal to a luminal narrowing)
exceeded a 2.5-cm diameter [12].
The quantitative analysis of intestinal distention was performed
measuring the largest cross-sectional diameter (from outer wall to
outer wall) of each bowel segment. The largest diameter (in mil-
limeters) could be chosen from either the coronal or the axial plane.
Distension measurements were carefully made in correspondence
to healthy, unaffected bowel segments.
The qualitative analysis was also performed on a per segment
basis using a continuous 3-point scale (0, poor; 1, good; 2, optimal),
as described by Megibow et al. [16]. An optimal score meant that
the segment was distended, the wall was uniformly visualized, and
a fold pattern could be recognized. A score of 0 meant that the seg-
ment was collapsed without any luminal separation, the walls could
not be seen, and a fold pattern could not be recognized. For this
purpose axial and multiplanar reformatted images were analyzed.
The last two degrees of distention (i.e. good and optimal) were
considered as a positive result. The percentage of bowel segments
with an adequate distension for diagnostic purposes was obtained
e4 F. Paparo et al. / European Journal of Radiology 82 (2013) e1–e10

Fig. 1. Volume-rendered CT reconstruction (A) and coronal reformatted CTE image (B) which show the position of the nasojejunal enteroclysis catheter (arrows) with the
tip at the duodeno-jejunal passage. The optimal distension of jejunal loops (J) is well appreciable in (B).

summing the relative percentages of segments that received both also, type I error was protected by using the Bonferroni adjustment,
good and excellent scores. setting the level of significance at 0.0083.

2.4. Adverse effects 3. Results

All the patients of each group were interviewed about their tol-
erance to the CT examination. Patients were asked whether they
experienced any adverse effect during, or immediately after the
procedure. The presence and type of adverse effects were recorded
on a standardized questionnaire. Nausea and abdominal discom-
fort were considered mild side effects, while vomiting, abdominal
pain and diarrhea were considered severe side effects.
2.5. Statistical analysis
Statistical analysis was performed to assess the presence of a sig-
nificant difference between the grade of bowel distension obtained
with the four CT protocols on a per segment basis.
The normality of data derived from quantitative per segment
analysis was assessed. It was observed that measurements of
the largest cross-sectional diameter of some bowel segments did
not follow the normal distribution, so all these quantitative data
were considered as non-parametric and expressed as medians and
range (minimum to maximum). The Kruskal–Wallis test for non
parametric data was used to assess the presence of a significant
difference among the medians of measurements obtained from the
four groups. After obtaining a significant Kruskal-Wallis test, the
Mann–Whitney U-test was applied for pairwise comparisons. The
type I error was protected by using the Bonferroni adjustment, and
the “a priori” alpha level (0.05) was divided by the number of pair-
wise comparisons (6), thus resulting in a level of significance of
0.0083.
Frequencies obtained from qualitative per segment analysis
were compared using the Chi-square test, while pairwise compar-
isons were conducted by means of the Fisher’s exact test. In this case
3.1. Patients’ characteristics
We did not find any statistical difference among the four
CT groups concerning the age (p = 0.09) and the sex of patients
(p = 0.8887).
3.2. Image analysis
In no CT examination there was an inflammatory or neoplastic
stricture of the small bowel associated to a significant prestenotic
dilatation (≥2.5 cm).
3.2.1. Quantitative analysis
Data retrieved from per segment quantitative analysis are
shown in Table 1.
Distension of the jejunum was significantly different among the
four CT protocols (p < 0.0001), and CTE provided the best distension
of jejunal loops among all the other techniques (p < 0.0001). There
was no significant difference between CTe and CTe-WE (p = 0.1453).
CT-WE provided the lowest median value of distension of jejunal
loops (10 mm, range 6–15 mm).
Distension of the ileum was significantly different among the
four CT protocols (p < 0.0001), and no significant difference was
found between the median values of ileal distension provided
by CTE, CTe and CTe-WE (CTE vs CTe p = 0.0811; CTE vs CTe-WE
p = 0.234; CTe vs CTe-WE p = 0.579). CT-WE provided the lowest
median value of distension of the ileum (10 mm, range 4–22 mm),
which was significantly inferior to that provided by the other CT
protocols (p < 0.001 for all pairwise comparisons).

Table 1
Results of per segment quantitative analysis. Median values (range: from minimum to maximum) of distension of each bowel segment are expressed in mm.

Technique Jejunum Ileum Terminal ileum Cecum Ascending colon Transverse colon Descending colon Sigma Rectum

CTE 27 (17–32) 18 (12–21) 17 (7–21) 22.5 (11–46) 32.5 (20–46) 25.5 (18–42) 21 (17–37) 19 (10–25) 20 (18–23)
CTe 13 (10–25) 17 (10–21) 14.5 (4–21) 25 (10–46) 33.5 (16–51) 23 (18–42) 25.5 (19–38) 21 (17–34) 23.5 (19–37)
CT-WE 10 (6–15) 10 (4–22) 15 (4–21) 57 (45–78) 58.5 (43–76) 46.5 (32–62) 40 (32–47) 35 (23–39) 54.5 (40–62)
CTe-WE 16 (8–30) 17 (10–21) 18 (8–21) 62 (47–77) 60.5 (37–77) 47 (33–54) 40.5 (34–47) 26 (16–39) 52 (40–60)
 F. Paparo et al. / European Journal of Radiology 82 (2013) e1–e10
Fig. 2. Coronal reformatted CTe-WE image shows simultaneous distension of ileum
(i), cecum (C) and ascending colon (ac). Note collapsed jejunal loops (J) in the left
upper abdominal quadrant.
No significant difference was found among the median values
of distension of the terminal ileum provided by the four CT pro-
tocols, despite the p-value tended to reach the level of significance
(p < 0.0608). In particular the comparison between CTe and CTe-WE
gave the lowest p-value (p = 0.0292), which was otherwise not sig-
nificant due to the Bonferroni adjustment for pairwise comparisons
(adjusted level of significance p = 0.0083).
The median values of distension of cecum, ascending, trans-
verse, descending colon and sigma provided by CT-WE and CTe-WE
were significantly higher than those provided by the other two
techniques. The p-value was < 0.0001 for all the following compar-
isons: CT-WE vs CTE, CT-WE vs CTe, CTe-WE vs CTE, CTe-WE vs CTe,
for all segments of the large bowel.
3.2.2. Qualitative analysis
Data concerning qualitative analysis are shown in Table 2.
3.2.2.1. Jejunum. CTE determined a significantly higher frequency
of patients with an optimal jejunal distension (90%; 27/30 patients)
when compared to CTe (3.3%; 1/30 patients; p < 0.001), CT-WE (0%),
and CTe-WE (20%; 6/30; p < 0.001) (Fig. 1). In this regard there was
no significant difference between CTe and CTe-WE protocols (3.3%
vs 20%, respectively; p = 0.1).
Fig. 3. Histogram representing the number of patients in each CT group with an adequate bowel distension on a per segment basis. CTE determines the highest frequency of
adequate distension of jejunum and ileum, while its ability to provide an adequate luminal filling of colonic segments decreases from proximal to distal. CTe-WE provides
very high frequencies of adequate distension of both small and large bowel segments with the only exception of jejunal loops.
e5
The frequency of patients with an adequate distension obtained
with CTE (30/30; 100%) was significantly higher than that achieved
by CTe (13/30 patients [43.3%]; p < 0.001) and CTe-WE (8/30
patients [26.7%]; p < 0.001) (Fig. 2). In no patient of the CT-WE group
the jejunum was adequately visualized. Concerning the percentage
of examinations with an adequate luminal filling, no significant
difference was found between CTe and CTe-WE groups (26.7% vs
43.3%; p = 0.28) (Fig. 3).
3.2.2.2. Ileum. The frequency of patients with an optimal disten-
sion of the ileum was not significantly different between CTE (27/30
patients; 90%), CTe (21/30 patients; 70%) and CTe-WE groups
(24/30 patients; 80%) (p = 0.15), but it was significantly lower in
the CT-WE group (4/30, 13.3%) when compared to other groups
(p < 0.001 at all pairwise comparisons). Ileal distension was ade-
quate in all patients who had undergone CTE (30/30; 100%), in
28/30 patients (93.3%) who were submitted to CTe, in 29/30 patient
(96.6%) of the CTe-WE group, and in 10 patients (33.3%) of the
CT-WE group; in this regard there was no significant difference
between CTE, CTe and CTe-WE protocols (p = 0.35), but the differ-
ence was significant when comparing CT-WE group with all the
other groups (p < 0.001).
3.2.2.3. Terminal Ileum. The frequency of patients with an opti-
mal distension of the terminal ileum was different between the
CTE (16/30 patients; 53.3%), CTe (9/30 patients; 30%), CT-WE (8/30
patients; 26.6%), and CTe-WE protocols (23/30 patients; 76.7%)
(p < 0.001); pairwise comparisons showed a significant difference
with regard to CTe-WE vs CTe and CTe-WE vs CT-WE (p < 0.001 in
both cases) (Fig. 4). No difference was found between the CTE and
CTe-WE protocols (p = 0.1). The frequency of examinations with an
adequate luminal filling of the terminal ileum was not significantly
different between CTE (27/30 patients; 90%), CTe (25/30 patients;
83.3%), CT-WE (21/30 patients; 70%) and CTe-WE groups (28/30;
93.3%) (p = 0.066) (Fig. 5).
3.2.2.4. Cecum and other colonic segments. The frequency of exam-
inations with an optimal visualization of the cecum was almost
the same for the CT-WE (29/30, 96.6%) and CTe-WE (28/30, 93.3%)
protocols, but it was largely inferior in the CTE (8/30, 26.6%) and
CTe (9/30, 30%) groups (p < 0.0001 at all pairwise comparisons [CT-
WE vs CTE, CT-WE vs CTe, CTe-WE vs CTE and CTe-WE vs CTe]).
There was also a significant difference among the four CT tech-
niques in the frequency of patients with an adequate luminal filling
of the cecum (p < 0.0001), with CT-WE and CTe-WE determining the
highest percentages (30/30, 100% in both cases). In 15/30 patients
(50%) CTE determined an adequate distension of the cecum, which
e6 F. Paparo et al. / European Journal of Radiology 82 (2013) e1–e10

Table 2
Results from qualitative per segment analysis. Number of patients with an optimal, good, poor and adequate distension of different bowel segments for each CT group (CTE,
CTe, CT-WE, CTe-WE). The number of examinations with adequate bowel distension was calculated summing those with an optimal and those with a good distension.

Technique Optimal Good Poor Adequate

CTE CTe CT-WE CTe-WE CTE CTe CT-WE CTe-WE CTE CTe CT-WE CTe-WE CTE CTe CT-WE CTe-WE

Jejunum 27 1 0 6 3 7 0 7 0 2 30 17 30 8 0 13
Ileum 27 21 4 24 3 7 6 5 0 2 20 1 30 28 10 29
Terminal ileum 16 9 10 23 11 16 11 5 3 5 9 2 27 25 21 28
Cecum 8 9 29 28 7 12 1 2 15 9 0 0 15 21 30 30
Ascending 7 10 27 28 10 9 3 2 13 11 0 0 17 19 30 30
Transverse 1 2 27 29 11 6 3 1 18 22 0 0 12 8 30 30
Descending 1 1 26 27 5 11 4 3 24 18 0 0 6 12 30 30
Sigma 0 4 26 28 1 7 3 2 29 19 1 0 1 11 29 30
Rectum 0 0 27 28 0 10 3 2 30 20 0 0 0 10 30 30

was obtained in 21/30 patients (70%) who had undergone CTe, thus and CT-WE techniques (p = 0.0012), but no difference was found
resulting in no significant difference between these two techniques among all the other CT protocols.
(p = 0.18).
All the other colonic segments distal to cecum (ascending, trans-
4. Discussion
verse, descending colon and sigma) were optimally distended by
both CT-WE and CTe-WE with very high frequencies (>86.6% for
As well as for conventional barium studies, luminal distention
each colonic segment with both techniques) (Fig. 6). For all colonic
is also required for CT imaging of the bowel, because collapsed
segments distal to cecum CTE and CTe provided lower frequencies
bowel loops can hide even large lesions and may falsely mimic
of patients with an adequate luminal filling than those obtained
wall thickening [2], and such a requisite is valid for both the small
with the other techniques. CTe provided a higher frequency of ade-
and large bowel. Macari et al. [14] have considered that appropri-
quate distension than that obtained by CTE for cecum (21/30 vs
ate luminal distension is achieved when the small bowel diameter
15/30, not significant), ascending colon (19/30 vs 17/30, not sig-
corresponds to at least 2 cm, and intraluminal contrast material
nificant), descending colon (12/30 vs 6/30, not significant), sigma
separates the intestinal walls and allows the identification of folds
(11/30 vs 1/30, p < 0.0024) and rectum (0/30 vs 10/30, p < 0.00032).
(valvulae conniventes), without collapsed loops.
In our work, in line with the results of Minordi et al. [6],
3.3. Adverse effects we found that CTE provides better distension of jejunal loops at
both qualitative and quantitative analysis (p < 0.001), and this may
The frequency of adverse effects were significantly different be considered an advantage of CTE over the other CT protocols.
among the four CT protocols (p = 0.0028) (Table 3). Adverse effects In both CTE and CTe the prominent mechanism responsible for
were more frequent in the CTE group than in the other groups, small bowel distension is the antegrade administration of contrast
and they were mostly represented by abdominal discomfort and material; when performing CTE, with the tip of the naso-jejunal
vomiting (7/18, 38.9% and 6/18, 33.3%, respectively). Pairwise catheter placed at the duodeno-jejunal junction, the direct infusion
comparisons using the Bonferroni adjustment demonstrated a sig-
nificant difference in the frequency of adverse effects between CTE

Fig. 4. Coronal reformatted CTe-WE image with optimal distension of the ileocecal Fig. 5. Coronal reformatted CT-WE image in a patient with Crohn’s disease showing
area, including both cecum and last ileal loop, which is characterized by the typical mild retrograde distension of the last ileal loop, which is characterized by mural
trilaminar mural stratification of Crohn’s disease (arrow). thickening and mucosal hyperemia (arrow).
 F. Paparo et al. / European Journal of Radiology 82 (2013) e1–e10 e7

Fig. 6. Coronal reformatted CTe-WE images (A and B) show optimal simultaneous distension of both small and large bowel. Legend: J, jejunum; i, ileum; ac, ascending colon;
tc, transverse colon.

Table 3
Frequencies and types of adverse reactions for each CT protocol.

No side effects Mild Severe

Nausea Abdominal discomfort Vomiting Diarrhea Abdominal pain

CTE 12 4 7 6 – 1
CTe 22 4 3 1 – –
CT-WE 25 2 2 – 1 –
CTe-WE 21 3 4 1 1 –

of methylcellulose in the jejunal loops results in a better distension
of this bowel tract than that provided by CTe. On the other hand, CTE
is characterized by an additional radiation dose given to the patient
(5.33 mSv in our material), due to the fluoroscopically guided place-
ment of the nasojejunal tube, and this technique is also associated
to the highest frequency of adverse effects (p < 0.0012). In addition
to exposing patients to considerable radiation and procedural dis-
comfort, CTE technique requires the use of two diagnostic rooms
and movement of the patient between them. However, the optimal
distension of jejunal loops, which is constantly provided by CTE,

Fig. 7. Axial CTE (A) and coronal reformatted CTE images (B) in a celiac patient with an histologically proven adenocarcinoma of the proximal jejunum. Image (A) demonstrates
the tumor as a focal circumferential mural thickening of the proximal jejunum with a nodular appearance. Image (B) shows the optimal distension of all jejunal loops in left
abdominal quadrants.
e8 F. Paparo et al. / European Journal of Radiology 82 (2013) e1–e10

Fig. 8. Two coronal reformatted CT-WE images (A and B) and a coronal oblique reformatted CT-WE image (C) in a patient with multifocal synchronous carcinoid tumors. In
image (A) the appendicular carcinoid appears as a thickened hyper-enhancing cecal appendix with some tiny calcifications (arrow), while the hyper-enhancing ileal carcinoid
(void arrow) is not easily detectable in image (B) due to its localization in a collapsed bowel loop. Image (C) shows both appendicular (arrow) and ileal (void arrow) carcinoid
tumors and typical hyper-enhancing mesenteric lymphadenopathies (L). All findings were confirmed by octreoscan scintigraphy.

may be particularly important to diagnose pathologies that selec- When performing CTe-WE, the terminal ileum is distended with
tively affect this segment of the small bowel (i.e. celiac disease) and a double mechanism: the anterograde passage of oral contrast and
their complications (i.e. lymphoma and adenocarcinoma in celiac the retrograde passage of trans-rectally injected water through
patients) [17] (Fig. 7). the ileocecal valve, which is incompetent in a large number of
A particular clinical indication to perform CTE is the detec- patients [12]. This second mechanism of distension of the termi-
tion and characterization of lower ileocolic inflammatory lesions nal ileum (reflux of water through the ileocecal valve) may be
in patients with stricturing Crohn’s disease involving the gastric quite consistently observed also when performing CT-WE (21/30,
antrum or the duodenal bulb and determining a gastric outlet syn- 70%). The incompetence of the ileocecal valve and the likelihood
drome [18]. ileocecal reflux are increased by the pharmacological effect of
The jejunum is not adequately distended by CTe (and CTe-WE) the smooth muscle relaxant (hyoscine-N-butylbromide), which is
because the duodenum and proximal jejunum distend earlier than administered before CT acquisition (Fig. 9).
terminal ileum (15–20 min vs 45–60 min after the ingestion of oral We found some similarities between CTe-WE and the peroral
contrast), and the large majority of CTe protocols, including that of pneumocolon technique, which has been recently reconsidered by
our study, are designed to explore the ileum and the terminal ileal P.J. Pickhardt as a way to obtain a good distension of both sides
loop [2,3]. of the ileocecal area also in CT studies [19]. The peroral pneumo-
So, the major limitations of CTe include the ingestion of a large colon was firstly employed as a complement of the standard SBFT
volume of fluid in a relatively small amount of time, the need for examination, and it is based on fluoroscopically guided gaseous
a precise timing of the CT scan in relation to the oral contrast insufflation per rectum following the arrival of barium at the cecum.
intake, and the lack of a simultaneous visualization of jejunum and It may provide, in the majority of cases, high-quality double con-
ileum. trast evaluation of the terminal ileum. Beyond conventional barium
With regard to the ileum, the median values of luminal disten- studies, a retrograde pneumocolon technique has been proposed
sion obtained with CTE, CTe and CTe-WE were not significantly as an attractive potential alternative to CTe and CTE techniques for
different between each other, and high frequencies of optimal ileocecal distension in CT studies, representing a new approach for
distension were found at qualitative analysis with all these CT investigating inflammatory bowel disease and other pathologies
techniques. The median values of ileal distension obtained by
CTE, CTe and CTe-WE (18 mm, 17 mm and 17 mm, respectively)
were comparable with those reported in a previous work based
on peroral CTe performed with lactulose solution (17.8 mm, range
15.1–23 mm) [13].
CT-WE technique allowed an adequate distension of the ileum
only in 33.3% of patients; as a result, in a patient affected by mul-
tifocal synchronous carcinoid tumors (an appendicular carcinoid
and an ileal carcinoid) the ileal tumor, which was localized in a col-
lapsed bowel loop, was not easy to detect on CT images, despite its
relatively large dimensions (Fig. 8).
The terminal ileum is an important, but sometimes difficult area
to adequately examine on both small bowel barium follow-through
(SBFT) and enteroclysis examinations. The degree of contrast opaci-
fication and luminal distension of the terminal ileum can vary
widely with both techniques [18].
In our study we found that there is no significant difference
among all CT techniques in regard to the frequencies of patients
with an adequate distension of the terminal ileum, with the great-
Fig. 9. CTe-WE coronal reformatted image which shows optimal distension of the
est median value of luminal distension provided by CTe-WE and
ileocecal area with patency of the ileocecal valve (arrow). Legend: c, cecum; ti,
the lowest one obtained with CT-WE. terminal ileum; asterisk, fat tissue of the ileocecal valve.
 F. Paparo et al. / European Journal of Radiology 82 (2013) e1–e10
which may simultaneously affect both the small and large intestine.
This hybrid CT technique derived from the peroral pneumocolon
has been called “CT colo-enterography”, and it mainly relies on the
reflux of gas through the ileocecal valve to improve distension of
the terminal ileum [19].
In the CTe-WE technique the reflux of water from the cecum
through the ileocecal valve into the terminal ileum combines with
the orally ingested contrast material, resulting in excellent luminal
distension of both sides of the ileocecal area [12]. This simultaneous
and combined distension of terminal ileum and cecum is not so
consistent with the other techniques, as we observed a significant
difference among the four CT protocols in the frequency of patients
with an adequate distension of cecum (p < 0.0001). In particular,
the anterograde passage of oral contrast through the ileocecal valve
was not sufficient for an adequate separation of cecal walls in 15/30
patients belonging to the CTE group and in 9/30 patients of the CTe
group. On the contrary, CTe-WE and CT-WE provided very high
frequencies of patients with adequate distension of the cecal lumen
(100% in both cases), thus underlining the advantage of techniques
that include a retrograde colonic distension for a comprehensive
assessment of the ileocecal area. The main advantage of CTe-WE
over CT-WE is that CTe-WE provides an adequate demonstration
of the entire ileum, and not only of the last ileal loop.
When evaluating the colon several requirements have to be ful-
filled. Of these, a full distension of a clean colon is needed for a
proper and valid analysis. Luminal collapse or incomplete colonic
filling can lead to false-negative findings because polyps and small
colorectal tumors can be obscured.
As for the small intestine, also in patients with large bowel
inflammatory diseases collapsed segments may mimic bowel wall
thickening and mucosal hyper-enhancement, thus resulting in a
over-estimation of inflammatory lesions and false-positives [20].
To the best of our knowledge only one work examined the
grade of distension and contrast opacification of large bowel loops
obtained with CTe technique.
In this work Johnson et al. [20] evaluated the grade of colonic
distension in 70 CTe of patients with Crohn’s granulomatous colitis
and ulcerative colitis performed either with (33/70, 47%) or with-
out (37/70, 53%) intravenous administration of 0.5 mg of glucagon.
They used a semiquantitative 4-point scale, from 0 = collapse to
4 = full distension, to assess the amount of colonic distension and
to assign a qualitative grade to each CT examination (0–1 = poor,
2–3 = good, 4 = excellent). Interestingly they found a relative high
frequency of patients with an adequate (good + excellent) colonic
distension for diagnostic purposes in both groups (66% of patients
who received glucagon and 63% of patients who did not receive
glucagon), but their assessment was not made on a per-segment
basis.
In our group of patients who had undergone CTe we found a
trend of progressive decrease in the frequency of colonic segments
with an adequate distension from the cecum (70%) to the rectum
(33.3%). In the CTE group results concerning colonic distension were
even more inconsistent.
On the other hand the trans-rectal introduction of a
1500–2000 mL water enema allowed to obtain a colonic disten-
sion suitable for image assessment in all patients belonging to the
CT-WE and CTe-WE groups.
Up to date CT-WE (which has also been called hydrocolon-CT or
WE-MDCT) has received little attention in the radiological litera-
ture. Two precedent works [10,11] were focused on the ability of
CT-WE to detect colorectal cancer and polyps using colonoscopy as
reference standard. Ridereau-Zins et al. [10] reported an overall CT-
WE sensitivity and specificity of 98.6 and 95.0%, respectively, while
Soyer et al. [11] provided a sensitivity of 95% for tumors ≥6 mm.
Soyer et al. [11] performed also a semi-quantitative analysis
of their CT-WE examinations using a tree-point scale (3 = optimal,
e9
2 = acceptable, and 1 = non diagnostic), depending on the grade of
colorectal distension. Image analysis revealed optimal colon dis-
tension in all but three cases of their study cohort (98/101, 97%),
and, in the remaining three cases (3/101, 3%), colonic distension
was acceptable but this did not hamper the diagnostic capabilities
of the CT examinations. According to Soyer et al. the CT-WE exam-
inations were well tolerated in all patients, without any vasovagal
reaction.
The lack of a simultaneous distension of both small and large
bowel loops may be seen as a great limitation in several clinical
concerns, as it may result in underestimation of multifocal inflam-
matory lesions in patients with ileocolic Crohn’s disease, or in
miss-detection of multiple synchronous tumors that may affect the
bowel along its entire length (i.e. carcinoids).
The simultaneous assessment of both terminal ileum and prox-
imal colon may be particularly important in patients with Crohn’s
disease, which affects the ileocecal area in up 50% of cases, and in
those with suspected multifocal synchronous intestinal tumors (i.e.
carcinoids).
In a previous work an original CT technique based on oral admin-
istration of neutral contrast material and trans-rectal introduction
of a water enema (CTe-WE) has been used to investigate a large
cohort of CD patients, allowing a complete staging of disease exten-
sion throughout the intestine [12].
A drawback of the quantitative analysis performed in our study
is that the widest wall-to-wall diameter is only a surrogate measure
of the overall distention within a segment of the gastrointestinal
tract. Currently there is no readily available method that would
allow us to measure the exact volume of the entire segment of
interest. The quantitative findings, however, were found to parallel
those of qualitative assessment, thus supporting reader’s qualita-
tive impression of the entire segment.
5. Conclusions
CT studies of the bowel should be tailored to answer the clinical
question. CTE is the only CT protocol that guarantees a constantly
good distension of the jejunum, but the positioning of the nasoje-
junal tube is labor-intensive and exposes patients to an additional
radiation dose. CT-WE may be considered a useful CT protocol to
explore the large intestine, including inflammatory and neoplastic
pathology, with the advantage of providing retrograde distension of
the last ileal loop in a high percentage of patients. CTe-WE allows a
simultaneous, combined and constant distension of both small and
large bowel.
Conflict of interest statement
All authors declare that they have no conflict of interest in the
publication of this work.
References
[1] Fletcher JG. CT enterography technique: theme and variations. Abdominal
Imaging 2009;34(May–June (3)):283–8 [Review].
[2] Dave-Verma H, Moore S, Singh A, et al. Computed tomographic enterography
and enteroclysis: pearls and pitfalls. Current Problems in Diagnostic Radiology
2008;37(November–December (6)):279–87.
[3] Maglinte DD, Sandrasegaran K, Lappas JC, Chiorean M. CT enteroclysis. Radiol-
ogy 2007;245(December (3)):661–71 [Review].
[4] Tochetto S, Yaghmai V. CT enterography: concept, technique, and interpreta-
tion. Radiologic Clinics of North America 2009;47(January (1)):117–32.
[5] Wold PB, Fletcher JG, Johnson CD, Sandborn WJ. Assessment of small
bowel Crohn’s disease: non invasive peroral CT enterography compared
with other imaging methods and endoscopy-feasibility study. Radiology
2003;229:275–81.
[6] Minordi LM, Vecchioli A, Mirk P, Bonomo L. CT enterography with polyethylene
glycol solution vs CT enteroclysis in small bowel disease. British Journal of
e10 F. Paparo et al. / European Journal of Radiology 82 (2013) e1–e10
Radiology 2011;84(February (998)):112–9 [Epub ahead of print; October 19,
2010].
[7] Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual colonoscopy
to screen for colorectal neoplasia in asymptomatic adults. The New England
Journal of Medicine 2003;349(December (23)):2191–200 [Epub ahead of print;
December 1].
[8] Mang T, Gryspeerdt S, Schima W, Lefere P. Evaluation of colonic lesions and
pitfalls in CT colonography: a systematic approach based on morphology, atten-
uation and mobility. European Journal of Radiology 2012 [Epub ahead of print,
July 18].
[9] Laghi A, Rengo M, Graser A, Iafrate F. Current status on performance of CT
colonography and clinical indications. European Journal of Radiology 2012
[Epub ahead of print, June 27].
[10] Ridereau-Zins C, Aubé C, Luet D, et al. Assessment of water enema com-
puted tomography: an effective imaging technique for the diagnosis of colon
cancer: colon cancer: computed tomography using a water enema. Abdom-
inal Imaging 2010;35(August (4)):407–13 [Epub ahead of print; May 22,
2009].
[11] Soyer P, Sirol M, Dray X, et al. Detection of colorectal tumors with water enema-
multidetector row computed tomography. Abdominal Imaging 2012 [Epub
ahead of print, January 21].
[12] Paparo F, Bacigalupo L, Garello I, et al. Crohn’s disease: prevalence of
intestinal and extraintestinal manifestations detected by computed tomo-
graphy enterography with water enema. Abdominal Imaging 2012;37(June
(3)):326–37 [Epub ahead of print, December 9].
[13] Arslan H, Etlik O, Kayan M, et al., Peroral CT. Enterography with lactu-
lose solution: preliminary observations. American Journal of Roentgenology
2005;185(November (5)):1173–9.
[14] Macari M, Megibow AJ, Balthazar EJ. A pattern approach to the abnormal
small bowel: observations at MDCT and CT enterography. American Journal
of Roentgenology 2007;188:1344–55.
[15] Ajaj W, Lauenstein TC, Langhorst J, et al. Small bowel hydro-MR imaging
for optimized ileocecal distension in Crohn’s disease: should an additional
rectal enema filling be performed? Journal of Magnetic Resonance Imaging
2005;22(July (1)):92–100.
[16] Megibow AJ, Babb JS, Hecht EM, et al. Evaluation of bowel distention and bowel
wall appearance by using neutral oral contrast agent for multidetector row CT.
Radiology 2006;238:87–95.
[17] Soyer P, Boudiaf M, Fargeaudou Y, et al. Celiac disease in adults: evaluation with
MDCT enteroclysis. American Journal of Roentgenology 2008;191(November
(5)):1483–92.
[18] Kohli MD, Maglinte DD. CT enteroclysis in small bowel Crohn’s disease. Euro-
pean Journal of Radiology 2009;69(March (3)):398–403 [Epub ahead of print,
January 3; Review].
[19] Pickhardt PJ. The peroral pneumocolon revisited: a valuable fluoroscopic
and CT technique for ileocecal evaluation. Abdominal Imaging 2012;37(June
(3)):313–25 [Epub ahead of print, June 18].
[20] Johnson KT, Hara AK, Johnson CD. Evaluation of colitis: usefulness of CT enterog-
raphy technique. Emergency Radiology 2009;16(July (4)):277–82 [Epub ahead
of print, February 13].