Download Multidetector CT Angiogra- phy for Acute Gastrointesti

Note: This copy is for your personal non-commercial use only. To order presentation-ready
copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.
TRAUMA/EMERGENCY RADIOLOGY
1453
Multidetector CT Angiography for Acute Gastrointestinal Bleeding: Technique and
Findings1
SA-CME
See www.rsna
.org/education
/search/RG
LEARNING
OBJECTIVES
FOR TEST 6
After completing this
journal-based SACME activity, participants will be able to:
■■Describe
the protocol for multidetector CT angiograpy
in patients with
acute gastrointestinal bleeding.
■■Recognize
multidetector CT angiographic findings of
acute gastrointestinal bleeding, including active bleeding
and signs of recent
bleeding, as well as
the most common
causes.
■■Discuss
the role
of CT angiography
in the clinical management of acute
gastrointestinal
bleeding.
José M. Artigas, MD, PhD • Milagros Martí, MD, PhD • Jorge A. Soto,
MD, PhD • Helena Esteban, MD • Inmaculada Pinilla, MD • Eugenia
Guillén, MD
Acute gastrointestinal bleeding is a common reason for emergency department admissions and an important cause of morbidity and mortality.
Factors that complicate its clinical management include patient debility
due to comorbidities; intermittence of hemorrhage; and multiple sites of
simultaneous bleeding. Its management, therefore, must be multidisciplinary and include emergency physicians, gastroenterologists, and surgeons, as well as radiologists for diagnostic imaging and interventional
therapy. Upper gastrointestinal tract bleeding is usually managed endoscopically, with radiologic intervention reserved as an alternative to be
used if endoscopic therapy fails. Endoscopy is often less successful in the
management of acute lower gastrointestinal tract bleeding, where colonoscopy may be more effective. The merits of performing bowel cleansing
before colonoscopy in such cases might be offset by the resultant increase
in response time and should be weighed carefully against the deficits in
visualization and diagnostic accuracy that would result from performing
colonoscopy without bowel preparation. In recent years, multidetector
computed tomographic (CT) angiography has gained acceptance as a
first-line option for the diagnosis and management of lower gastrointestinal tract bleeding. In selected cases of upper gastrointestinal tract bleeding, CT angiography also provides accurate information about the presence or absence of active bleeding, its source, and its cause. This information helps shorten the total diagnostic time and minimizes or eliminates
the need for more expensive and more invasive procedures.
©
RSNA, 2013 • radiographics.rsna.org
INVITED
COMMENTARY
See discussion on this
article by Paulson
(pp 1470–1471).
Abbreviation: MIP = maximum intensity projection
RadioGraphics 2013; 33:1453–1470 • Published online 10.1148/rg.335125072 • Content Codes:
From the Departments of Radiology, Miguel Servet University Hospital, Paseo de Isabel La Católica 1-3, 50009 Zaragoza, Spain (J.M.A., H.E., E.G.);
La Paz University Hospital, Madrid, Spain (M.M., I.P.); and Boston Medical Center, Boston, Mass (J.A.S.). Recipient of a Certificate of Merit award
for an education exhibit at the 2011 RSNA Annual Meeting. Received April 16, 2012; revision requested July 6 and received February 19, 2013; accepted February 26. For this SA-CME activity, the authors, editor, and reviewers have no financial relationships to disclose. Address correspondence
to J.M.A. (e-mail: [email protected]).
1
©
RSNA, 2013
1454 September-October 2013
Introduction
Acute gastrointestinal bleeding is a common, potentially life-threatening condition responsible for
1%–2% of all hospital admissions (1), with the estimated rate in the United States reaching 375 admissions per 100,000 persons per year in some series (2–4). Acute gastrointestinal bleeding is more
common in men than women (M/F ratio, 2:1).
The incidence increases with age, with 70% of patients older than 65 years (1,5). In addition, acute
gastrointestinal bleeding frequently occurs as a
complication in critically ill adults who are admitted with other primary diagnoses and complicates
their clinical course (1). Mortality ranges between
19% and 40% and increases with (a) older age,
(b) the presence of shock, and (c) associated underlying comorbidities (6–8). Variceal hemorrhage
is associated with a mortality rate of 15%–40%,
a rate higher than that from other causes of acute
gastrointestinal bleeding (1,9).
The framework for triage of patients with gastrointestinal bleeding varies on the basis of its
source—upper or lower gastrointestinal tract—
with important differences in terms of epidemiology, clinical management, and prognosis (10,11).
Compared with lower gastrointestinal tract bleeding, upper gastrointestinal tract bleeding tends
to affect younger people. Because the rate of occurrence of lower gastrointestinal tract bleeding
events increases with age, hospitalizations that are
due to gastrointestinal bleeding in the elderly are
more likely to be caused by lesions of the lower
gastrointestinal tract (12). Moreover, better control of peptic ulcers secondary to Helicobacter pylori infection, the increasing use of gastroprotectant treatment, and the progressive aging of Western populations, combined with an increased use
of antiplatelet therapy and anticoagulant therapy,
are changing the epidemiology of hospitalizations
for gastrointestinal bleeding. In the findings from
recent studies, investigators have pointed out that
the rates of hospitalization for patients with upper
and those with lower gastrointestinal tract bleeding are becoming similar, with poorer outcomes
and higher resource utilization for lower gastrointestinal tract bleeding events (3,13).
Identification of the source and cause of bleeding helps guide therapeutic decisions, and for
years, clinical management has been focused
mainly on endoscopic findings. Endoscopy represents a safe and effective method for the diagnosis
and, often, the treatment of patients with gastrointestinal bleeding, with sensitivity and specificity
approaching 98% and 100%, respectively. Nevertheless, implementation of endoscopic procedures
radiographics.rsna.org
in the emergency setting usually poses a variety of
challenges, such as variable availability of the service and an insufficient time window to perform
adequate bowel preparation in the most serious
cases. Mucosal visualization is poor owing to the
presence of intraluminal blood clots and other
intestinal contents, and the distal portion of the
duodenum and the small bowel are not routinely
accessed (14). Because of its ability to be used to
wash and aspirate the gastric contents, esophagogastroduodenoscopy is clearly the first-line tool for
diagnosis and treatment of upper gastrointestinal
tract bleeding. On the contrary, urgent colonoscopy remains challenging for assessment of acute
lower gastrointestinal tract bleeding because this
procedure can be used to identify the source of
bleeding in only 13% of cases in some series (15).
A specific cause cannot be identified with endoscopy or subsequent workup in as many as 20% of
patients with lower gastrointestinal tract bleeding
and as many as 14% of patients with upper gastrointestinal tract bleeding (3).
Other techniques, such as scintigraphy with
technetium 99m (99mTc)–labeled red blood cells
or 99mTc–sulfur colloid, enteroscopy, and video
capsule endoscopy, are not universally accessible
in the emergency setting (16). Catheter angiography and emergency surgery are options for patients with massive life-threatening bleeding and
should be used ideally as targeted therapeutic
procedures (5).
The purpose of this article is to summarize the
role of computed tomographic (CT) angiography
in the evaluation of acute gastrointestinal bleeding. First, the clinical context in which radiologists will encounter patients with acute gastrointestinal bleeding in the emergency setting is summarized. Then the initial diagnostic evaluation is
described, followed by the role of endoscopy. The
rationale for incorporating multidetector CT angiography as a useful alternative in the diagnostic
algorithm for these patients is explained, along
with the CT imaging technique and protocol.
The most common CT angiographic imaging
findings are reviewed, including active bleeding (blush) and recent bleeding (clots) and their
causes. Finally, imaging artifacts and potential
pitfalls encountered in the setting of acute gastrointestinal bleeding are identified.
Clinical Manifestations
Gastrointestinal bleeding is not a specific disease
but rather is a clinical manifestation of many
diseases affecting the digestive tract. Gastrointestinal tract bleeding is classified as upper or lower,
depending on whether its origin is proximal or
distal to the ligament of Treitz. The locations of
RG • Volume 33 Number 5
Artigas et al 1455
Figure 1. Drawing shows the various anatomic locations of upper (1–5) and lower
(6–12) gastrointestinal tract bleeding (not
in order of frequency). The main cause for
bleeding in each numbered location is 1,
esophagitis; 2, esophageal varices; 3, gastric
cancer; 4, peptic ulcer disease; 5, gastritis
or duodenitis; 6, colitis; 7, colonic angiodysplasia; 8, small bowel vascular lesions; 9,
diverticular disease; 10, colonic neoplasms;
11, rectal ulcer; and 12, hemorrhoids.
upper and lower gastrointestinal tract bleeding
and the main causes of bleeding in each location
are shown in Figure 1.
The clinical manifestations of gastrointestinal hemorrhage depend on the location of the
source, the rate of bleeding, and the effect of
blood loss on the general condition of the patient.
These factors dictate the priority and the order
in which diagnostic tests and therapeutic procedures should be performed (17). Severe bleeding can cause hypovolemic shock, but slow and
intermittent blood loss manifests most often with
iron deficiency anemia or stools with positive
findings on a fecal occult blood test (Hemoccult;
Beckman Coulter, Brea, Calif). Direct evidence
of bleeding can manifest with hematemesis (vomiting fresh blood), coffee-ground vomit (partially
digested dark blood), melena (black tarry stools),
or hematochezia (red blood per rectum). Patients
with massive upper gastrointestinal tract bleeding
may have hematochezia because of the cathartic
properties of blood (1,14,18).
Overall, one in four patients with gastrointestinal bleeding will need immediate medical attention and resuscitation, and the remainder will be
asymptomatic or will present with mild systemic
symptoms that can be treated conservatively (5).
Clinical scores can be used to estimate the risk
of death or recurrent hemorrhage and the need
for clinical intervention to control bleeding, and
these scores can be useful for initial patient triage (19). Regardless of the amount of blood lost,
approximately 75%–80% of episodes of gastrointestinal bleeding resolve spontaneously (15).
In 25% of patients, hemorrhage recurs, and this
recurrence increases the risk for in-hospital mor-
bidity and mortality; this group consumes a large
amount of hospital resources because these patients require a thorough clinical assessment and
proper diagnostic imaging methods (5).
Initial Diagnostic Evaluation
Hospitalization is usually required in elderly patients and those presenting with hemodynamic
instability or severe comorbidities (17). However,
only 20% of the patients with acute gastrointestinal bleeding will need urgent intervention. It is
therefore impractical and economically unjustified
to subject all patients with acute gastrointestinal
bleeding to a comprehensive emergency evaluation. Risk stratification criteria are used to assess
the severity of bleeding and to determine who may
potentially benefit the most from urgent resuscitation, endoscopy, and intensive clinical care (20).
The general goals of management include resuscitation, diagnosis, hemostasis, and prevention
of recurrent bleeding. Most complications from
acute gastrointestinal bleeding result from hypoperfusion; thus, stabilization of blood pressure and
restoration of volume precede any further diagnostic measures (9). Detection and correction of a
possible coagulopathy are also important (21).
Patients who show evidence of or are suspected of having an upper gastrointestinal tract
source of bleeding should undergo endoscopy
as the initial investigation. Endoscopy frequently
facilitates definitive treatment of lesions proximal
to the ligament of Treitz (18,22). If the patient
is not experiencing hematemesis or if endoscopy
is not immediately available, a nasogastric tube
1456 September-October 2013
radiographics.rsna.org
Figure 2. Lower gastrointestinal tract bleeding in
a 36-year-old man in whom the findings from initial
upper endoscopy had been unrevealing. (a) Axial unenhanced CT image shows intragastric hyperattenuating
material (*). (b) Axial arterial phase CT angiographic
image shows linear extravasation (a “jet”) (arrow) of
contrast material, as well as intragastric hyperattenuating material (*). (c) Axial CT image of the pelvis
shows hyperattenuating rectal contents (*) that are
due to hematochezia, owing to accelerated bowel transit
caused by intraluminal blood. Peptic ulcer was suggested as the probable cause of bleeding and was later
confirmed by the findings from repeat upper endoscopy.
should be inserted to assess the rate of bleeding
and to allow gastric lavage while the patient is
awaiting endoscopy. Although bleeding from the
duodenum may manifest with a heme-negative
nasogastric aspirate (5), the combination of the
absence of blood and the presence of bile in the
aspirate makes an upper gastrointestinal tract
source unlikely, and the focus of the workup then
turns to the lower gastrointestinal tract (17).
Lower gastrointestinal tract bleeding tends to
be self-limited and less dramatic in manifestation
(10). A digital rectal examination and proctoscopy
should be performed initially to exclude anorectal
sources of bleeding, such as hemorrhoids, and
to estimate the volume of blood lost (22). Lower
gastrointestinal tract bleeding typically manifests
with hematochezia, although bleeding from the
small intestine or the proximal portion of the
colon may appear as melena (1). However, an
episode of apparent lower gastrointestinal tract
bleeding may in fact originate from the upper gastrointestinal tract (1,17,21,22) (Fig 2).
Endoscopy
Because endoscopy is the initial diagnostic modality in patients suspected of having upper
gastrointestinal tract bleeding, the primary objectives of upper gastrointestinal tract endoscopy
are to confirm the source of the bleeding and to
determine whether the cause is variceal or not
(22). In addition, endoscopy allows the collection of histologic specimens and, frequently, the
performance of therapeutic interventions, such as
injection therapy, thermal coagulation, and laser
therapy, in patients with active bleeding. The benefits of urgent endoscopy in this setting are well
documented (23,24). The main limitations of upper gastrointestinal tract endoscopy include poor
visualization of the distal portion of the duodenum
and an inability to study the remaining portion of
the small bowel (8,9). When upper gastrointestinal
tract endoscopy is not feasible or its findings are
inconclusive, imaging is the next option to localize
and treat the source of bleeding (22).
In patients with lower gastrointestinal tract
bleeding, colonoscopy can be used to identify
the site of bleeding in more than 70% of patients (17); however, wide variability has been
reported in this percentage, ranging from 8% to
100% (10,12,14,15,24), a range that reflects inherent biases in the definition of end points and
differences in equipment, timing, and preparation. Colonoscopy can be performed urgently
or electively, depending on the patient’s hemodynamic status and risk stratification criteria.
For therapy, emergent colonoscopy offers no
RG • Volume 33 Number 5
Artigas et al 1457
Figure 3. Upper gastrointestinal tract bleeding and severe
epigastric pain in a 66-year-old woman in whom upper endoscopy had shown active bleeding from the papilla of Vater
(hemobilia). (a) Axial unenhanced CT image shows hyperattenuating bile in the gallbladder (g), a finding that represents
blood, as well as a large stone (s). (b) Axial portal venous
phase CT angiographic image clearly depicts a pseudoaneurysm (a) within the gallbladder lumen. There is a lack of mural enhancement in the gallbladder, secondary to gangrenous
cholecystitis (arrowheads). (c) Coronal arterial phase MIP
reformatted image shows a faintly hyperattenuating area that
represents the pseudoaneurysm (arrowheads), as well as the
neck of the pseudoaneurysm (arrow).
advantage compared with radiologic intervention; therefore, the choice of one or the other
approach is based on local expertise and available resources (24,25). In the emergency setting,
colonoscopy is hampered by incomplete colonic
cleansing and poor visualization of the mucosa in
the presence of massive hemorrhage. Although urgent colonoscopy can be performed without colonic
preparation (10), the available diagnostic information is limited, and expediting embolization or even
surgery may be preferable to delaying definitive
treatment (26). Thus, the clinical utility of urgent
colonoscopy has been questioned for patients with
acute lower gastrointestinal tract bleeding (15).
Furthermore, the complication rate of colonoscopy
is low but not negligible, ranging from 0.03% to
2.14%, with a mortality rate of 0.3% (15); these
rates may be even higher when colonoscopy is performed urgently.
Multidetector CT Angiography
Technical improvements in multidetector CT
technology, especially the higher temporal resolution that allows the ability to obtain high-resolution three-dimensional datasets with short acquisition times at different times during the administration and distribution of intravascular contrast
material (multiphasic studies), have expanded the
applications of CT angiography for evaluating
patients with vascular diseases, including acute
hemorrhage. Temporally resolved CT angiography allows the identification of active extravasation of contrast material and the accurate identification of the source of hemorrhage (22,27).
Other advantages of CT angiography include its
widespread availability in the emergency setting,
its minimal invasiveness, and its reproducible
results, which combine high sensitivity and accuracy for detecting or excluding active bleeding, as
well as detecting the potential source and cause
(7,9,19,28–35). CT angiography can be used
to evaluate the wall of the entire gastrointestinal
tract and other digestive structures that may occasionally be the source of bleeding, such as the
pancreas and biliary tract (Fig 3).
Imaging Technique
Although the specific technical parameters used
and the phase images acquired vary among institutions, most agree that three-phase examinations that include the unenhanced, arterial, and
portal venous phases provide the best and most
reproducible results for this clinical application
in patients with acute gastrointestinal bleeding
(6,7,14,19,29,32,33,36). For CT scanners with a
1458 September-October 2013
radiographics.rsna.org
Figure 4. Typical appearance of active gastrointestinal
bleeding in a 74-year-old man who had previously undergone a total colectomy and who presented with bleeding
from the ileostomy site, bleeding that was due to superficial erosions in the distal portion of the ileum. (a) Axial
unenhanced CT angiographic image shows intraluminal
hyperattenuation (arrows). (b, c) Axial arterial phase (b)
and portal venous phase (c) CT angiographic images
show an intraluminal jet (arrow) of contrast material
in the arterial phase, which changes in size and morphology in the portal venous phase.
detector configuration of 64 × 0.625 mm, the following acquisition parameters are recommended:
section thickness, 1 mm, with a reconstruction
interval of 0.8 mm; pitch factor, 0.828; rotation
time, 0.5 second; and tube voltage, 120 kV, with
automatic tube current modulation in the x-, y-,
and z-axis directions. The scan should include
the complete abdomen and pelvis, from the
diaphragm to the inferior pubic ramus. For CT
angiography performed to evaluate acute gastrointestinal bleeding, no oral contrast material is
routinely administered, because positive contrast
material may mask bleeding and because water or
other neutral contrast material may dilute the extravasated intravenous contrast material, thereby
decreasing the ability to identify the site of bleeding (28,32,33,37).
The CT angiographic examination starts with
a preliminary unenhanced imaging series to depict any preexisting intraluminal hyperattenuating material, such as foreign bodies, opaque pills,
hemostatic clips, suture material from previous
surgery, or residual barium in diverticula, that
might be misinterpreted later as active bleeding (7,33,38). A low-radiation-dose technique is
recommended for this unenhanced series. Intravenous contrast material is administered through
an antecubital vein with a power injector at a rate
of 4 mL/sec, followed by a chaser of 50 mL of
saline. The dose is adjusted for body weight and
iodine concentration, with the total volume of
contrast material typically varying between 100
and 125 mL of an agent high in iodine concentration (>300 mg of iodine per milliliter). The
arterial phase images are usually obtained by using automated bolus triggering, starting when the
attenuation coefficient in the proximal portion of
the abdominal aorta reaches 150 HU. Portal venous phase images are then obtained 40–60 seconds later (ie, 70–90 seconds after the beginning
of the injection of contrast material) (33,38).
Findings of Acute
Gastrointestinal Bleeding
Active Bleeding: Blush.—Intraluminal extrava-
sation of contrast material is the main criterion
used to define the location of the bleeding point
(7,33,36) (Fig 4). Identification of this finding
requires that active bleeding be present during the
period of time that the injected contrast material
circulates through the vascular system; CT angiography typically detects active bleeding with a rate
that exceeds 0.3–0.5 mL/min, a rate that is slightly
higher than the threshold assigned to catheter
angiography (0.5 mL/min) and is lower than that
of scintigraphy with 99mTc-labeled red blood cells,
RG • Volume 33 Number 5
Artigas et al 1459
Figure 5. Typical appearance of active intestinal
hemorrhage caused by a bleeding diverticulum in
a 72-year-old man who presented with hematochezia. (a) Sagittal unenhanced reformatted image
shows hyperattenuating contents (*) in the ascending colon and multiple diverticula. (b) Sagittal
arterial phase reformatted image shows contrast
material extravasation (arrowhead) within a diverticulum. (c) Sagittal portal venous phase reformatted image shows that the focus of extravasated
contrast material (arrowhead) grows and changes
in morphology.
which can be used to detect active bleeding at a
rate higher than 0.1 mL/min (7,22,34,36). Therefore, severe bleeding episodes, such as those manifesting with hemodynamic instability, increase the
pretest probability of a positive result for active
bleeding at CT angiography (28,32,33,39). Intermittent bleeding can cause false-negative findings
at CT angiography (22,28,31,40).
The usual appearance of active bleeding is that
of an intraluminal blush of contrast-enhanced
blood in the arterial phase or a hyperattenuating
focus of variable attenuation and morphology in
the portal venous phase (29,33,34) (Fig 5). For
detecting active bleeding with CT angiography,
the highest sensitivity is achieved by combining findings from the arterial and portal venous
phases (28,36); the changing appearance of the
focus of extravasated contrast material with time
(from the arterial phase to the portal venous
phase) unequivocally confirms the presence of
bleeding, especially when the unenhanced image
shows no hyperattenuating intraluminal material.
Although the attenuation coefficient thresholds of the luminal contents have been proposed
as a way to define extravasation of contrast material (starting at 90 HU) (31,41), a sequential
comparison of the unenhanced images with the
two-phase CT angiograms offers better performance (Figs 4–6) (30). Small foci of bleeding
and subtle blushes may not reach the suggested
threshold because of volume averaging (33,39).
Although the source axial dataset is usually sufficient to reach a diagnosis, postprocessing of
the entire acquired dataset with three-dimensional algorithms allows the radiologist to obtain
tailored reformatted images, particularly multiplanar and maximum intensity projection (MIP)
images. These reformatted images depict the
vascular anatomy (and variants) and facilitate
the identification and localization of the site of
bleeding (32,33,38).
1460 September-October 2013
radiographics.rsna.org
Figure 6. Rectal ulcer in a 33-year-old woman who was positive for the human immunodeficiency virus and who presented with brisk hematochezia. (a) Axial arterial phase CT angiographic image depicts a jet of extravasated contrast
material (arrow) arising from the left posterior wall of the rectum, which is distended with blood and clots. (b) Axial
portal venous phase CT angiographic image shows pooling of extravasated contrast material on the dependent surface
of the rectum, which creates a fluid–contrast material level (arrowheads). (c) Coronal MIP reformatted image depicts
bleeding from the left hemorrhoidal artery (arrow). (d) Catheter angiographic image confirms the finding of bleeding
from the left hemorrhoidal artery (arrow). Therapy with embolization (not shown) was successful. (Reprinted, with
permission, from reference 30.)
MIP reformatted images of the arterial phase
resemble conventional angiograms and can be
used to guide subsequent angiographic interventions (6,33). Identification of early-draining veins
and vascular tufts is suggestive of angiodysplasia.
Although extravasation of contrast material can
be detected in the arterial phase, the small blush
can be subtle and difficult to identify (36). Portal venous phase imaging depicts extravascular
blushes with higher sensitivity than arterial phase
imaging does because more time has elapsed,
which allows the focus of extravasated contrastenhanced blood to enlarge and increase in attenuation within the lumen of the bowel (36).
The morphology of the extravasated contrast
material varies, depending on the origin (arterial
or venous) and rate of the bleeding, such that an
arterial (linear) jet will be invariably present with
bleeding at a high rate (Fig 6a), whereas bleeding rates of less than 0.5 mL/min are associated
with a less-defined (or “cloudy”) morphology
(34) (Fig 7). The extravasated contrast material
may pool on the dependent surface of the bowel
(Fig 6b) or may be mobilized by peristalsis into a
more irregular shape (30).
In studies of experimental models, investigators have also suggested that the location of the
bleeding point in the circumference of the bowel
wall may create differences in the appearance of
extravasated contrast material; an anterior bleeding point can be associated with a typical “jet”
morphology, whereas if the bleeding point is located posteriorly, a “cloud” morphology will be
more likely (34). Thus, active bleeding can adopt
a variable morphology—linear and jetlike, swirled,
circular or ellipsoid, pooled or cloud-shaped—or
may even form a fluid–contrast material level
(32,33,38) (Figs 2, 3, 6–8). With brisk hemorrhage, extravasated contrast material can fill the
lumen completely, outlining bowel folds, and
can appear as a hyperattenuating loop (Figs 8, 9)
RG • Volume 33 Number 5
Artigas et al 1461
Figure 7. Lower gastrointestinal tract bleeding in a 74-year-old man. Sagittal portal venous
phase reformatted image shows a faint focus of
contrast material extravasation (arrow) arising
from the transverse colon, a finding that was
not depicted in the arterial phase (not shown).
Colonoscopy demonstrated findings of ischemic colitis and a bleeding ulcer.
Figure 8. Aortoduodenal fistulas in two patients. (a, b) Primary
aortoduodenal fistula (no previous aortic surgery or trauma) in
a 63-year-old man who presented with a severe episode of upper
gastrointestinal tract bleeding. (a) Coronal unenhanced reformatted
image shows hyperattenuating contents in the upper gastrointestinal
tract, a finding that indicates active or recent bleeding. Note intragastric clot (arrow) and stranding of the inferior margin of the periduodenal fat (arrowheads). (b) Sagittal multiplanar reconstruction
of the arterial phase dataset depicts an irregular aneurysmal aortic
wall with a direct communication with the duodenal lumen (arrows).
(c, d) Secondary aortoduodenal fistula in a 74-year-old woman who
presented with hematemesis 7 months after she had undergone an
aortobifemoral bypass graft procedure. (c) Coronal arterial phase
volume-rendered reformatted image shows irregularity of the aortic
wall and graft permeability, with faint opacification of the duodenal
lumen (h). (d) Coronal portal venous phase volume-rendered
reformatted image shows that the duodenal lumen becomes more
opacified, creating a typical hyperattenuating loop (h). The duodenal
folds (arrowheads) are outlined by the extravasation of intravenous
contrast material into the duodenal lumen.
1462 September-October 2013
radiographics.rsna.org
Figure 9. Right colonic angiodysplasia causing severe
lower gastrointestinal tract bleeding in a 44-year-old
woman with liver cirrhosis and portal hypertension.
Coronal (a), axial (b), and oblique (c) portal venous
phase MIP reformatted CT angiographic images show
brisk contrast material extravasation into the right colon
(*). Vascular tufts and a large draining vein (arrows) are
shown, which are typical findings of angiodysplasia as the
source of bleeding. Note that the cecum shows the hyperattenuating loop sign, owing to profuse bleeding, in c.
(32,33). Venous bleeding is depicted better in the
portal venous phase (36); upper gastrointestinal
tract bleeding of venous origin is usually due to
esophageal or gastric varices, although 16%–30%
of patients with portal hypertension and upper
gastrointestinal tract bleeding will have an arterial
bleeding source (5,18,22).
Recent Bleeding: Clots.—When bleeding has
stopped or does not reach the CT angiographic
detection threshold at the time of the injection
of contrast material, unenhanced CT can reveal
hyperattenuating material within the bowel lumen without additional findings in the contrastenhanced phases, a finding that indicates recent
hemorrhage (28,38). Attenuation values of the
usual intestinal contents approach those of water
(0–15 HU) (42). Extraluminal blood typically has
higher attenuation than other body fluids, mainly
owing to its high protein content. Unclotted extravascular blood usually has a measured attenuation of 30–45 HU (42), and the attenuation of
clotted blood ranges between 45 and 70 HU (43).
Accordingly, the mean attenuation level of blood
in the bowel lumen on unenhanced CT images
is approximately 47 HU (28), and the threshold
for diagnosing recent gastrointestinal bleeding has
been suggested to be more than 60 HU (29,44).
On CT images, the “sentinel clot” (hematoma
with the highest attenuation) has been described
as that closest to the site of bleeding, in contrast to
lower-attenuation unclotted blood, which is usually located farther from the source (45). Narrow
window settings facilitate recognition of hyperattenuating material on CT images, and comparison
with adjacent fluid-filled structures can be helpful.
When differences are subtle, the attenuation can
be measured directly (42).
Causes of Bleeding.—The detection and local-
ization of active hemorrhage and the concurrent
diagnosis of the specific underlying cause are the
optimal results of CT angiography. However, even
in the absence of active bleeding, CT angiography
may be used to identify the underlying lesion and
characterize it as diverticular, vascular, inflammatory, or neoplastic. The accuracy of CT angiography for detecting the cause of hemorrhage exceeds
80% (14,29). Abnormalities in the bowel wall or
surrounding fat provide diagnostic clues. These
clues include hyperattenuating perienteric fat (Figs
8a, 10), bowel wall thickening (Fig 10), abnormal
enhancement of the bowel wall (Figs 10, 11),
polypoid lesions, tumor masses (Figs 12, 13), and
focal vascular dilatation or early venous drainage
(features of angiodysplasia) (Fig 9).
RG • Volume 33 Number 5
Artigas et al 1463
Figure 10. Infectious colitis complicating non-Hodgkin lymphoma and causing lower gastrointestinal tract
bleeding in a 56-year-old man. Oblique portal venous
phase reformatted image, generated along the left
colon (c), shows mural thickening and pericolonic fat
stranding. No active bleeding was demonstrated with
CT angiography or colonoscopy. Retroperitoneal and
mesenteric lymphadenopathy was also present (not
shown).
Figure 11. Massive lower gastrointestinal
tract bleeding in a 17-year-old girl. (a, b) Coro­
nal (a) and sagittal (b) arterial phase reformatted images depict an abnormal commashaped saccular vascular structure (arrow)
located in the jejunal wall. (c) Axial portal
phase image shows extravasation of contrast
material (arrows). The findings at emergency
surgery confirmed a jejunal arteriovenous
malformation with active bleeding.
1464 September-October 2013
radiographics.rsna.org
Figures 12, 13. (12) Acute lower gastrointestinal tract bleeding in a 34-year-old woman in whom CT angiography
failed to show active bleeding but did show a 4-cm hypervascular partially exophytic mass originating in the wall
of the jejunum; the mass was resected, and the findings at histopathologic evaluation indicated that the mass was a
gastrointestinal stromal tumor (GIST). (a, b) Coronal arterial phase CT angiographic reformatted image (a) shows
the hyperenhancing mass (t), which is also depicted on the coronal MIP reformatted image (b). (c) Axial arterial
phase CT image shows a hyperenhancing hepatic metastasis (arrow). (13) Cecal adenocarcinoma causing lower gastrointestinal tract bleeding in a 79-year-old woman. (a) Axial portal venous phase CT angiographic image depicts a
6-cm cecal mass (t) with active extravasation of contrast material (arrows). Note the regional lymphadenopathy (arrowheads). (b) Coronal arterial phase MIP reformatted image shows the large area of extravasation originating from
a branch of the ileocolic artery (arrows).
RG • Volume 33 Number 5
Artigas et al 1465
Figure 14. Varices in a 67-year-old man with
cirrhosis. Upper endoscopy and CT angiography showed no active bleeding despite the
patient’s acute melena at presentation. Oblique
thin-slab portal venous phase MIP reformatted image shows esophageal varices (black arrows) and segmental bulbous dilatation (white
arrows) of intramural veins in the jejunum, a
finding that also represents varices. A likely
variceal source of bleeding was diagnosed.
Varices cause approximately 30% of all episodes of upper gastrointestinal tract bleeding. CT
angiography should be considered when endoscopy does not reveal a bleeding source and when
there is a suspicion of arterial bleeding or a source
in the lower gastrointestinal tract (Fig 14). Other
nonvariceal causes of bleeding include duodenal
and gastric ulcers, tumors, angiodysplasia, Mallory-Weiss tears, gastric erosions, and pancreatitis
(7). Aortoenteric fistulas are a rare cause of catastrophic bleeding. In this entity, there is a communication between the aorta and the gastrointestinal
tract, typically the distal portion of the duodenum,
most often after reconstructive surgery of the abdominal aorta. In 20%–100% of cases, previous
self-limited episodes of bleeding (“herald bleeds”)
are reported before the episode of exsanguinating
hemorrhage (46) (Fig 8). Upper gastrointestinal
tract bleeding may also occur in patients who are
hospitalized for other causes; stress ulcers can appear within 24 hours of admission in a patient who
has suffered severe trauma or burns. Hemobilia refers to blood originating in the liver, biliary tree, or
pancreas; blood passes to the duodenum through
the papilla of Vater (Fig 3).
The source of lower gastrointestinal tract bleeding is colorectal in approximately 90% of cases. In
the remaining 10%, the bleeding site is situated in
the small bowel. Among colorectal lesions, colonic
diverticula, angiodysplasia, inflammatory lesions,
and malignancies are the most common causes of
bleeding (7); and the accuracy of CT angiography
for determining the specific cause ranges from
80% to 85% in most series (14,28,29,40,47,48).
CT angiography can be used to easily diagnose
bleeding from colonic diverticula, the leading
cause of lower gastrointestinal tract bleeding (10)
(Fig 5). Colonic angiodysplasia is the second
most common cause of lower gastrointestinal tract
bleeding, accounting for as many as 40% of cases
in patients older than 60 years (31) (Fig 9). Investigators have reported values for the sensitivity,
specificity, and positive predictive value of 70%,
100%, and 100%, respectively, for CT angiography (49). When possible, it is important to distinguish between bleeding from a diverticulum and
angiodysplasia because the probability of rebleeding is as high as 85% in cases of angiodysplasia
and is only 25% in cases of diverticular bleeding
(33). In addition, CT angiography can be used to
accurately diagnose neoplasms and colitis, which
are the third and fourth most common causes of
lower gastrointestinal tract bleeding, respectively
(7,29,33) (Figs 10, 12, 13), and CT angiography
is also helpful for the diagnosis of benign anorectal
lesions (Fig 6). Postoperative bleeding can occur
at an anastomotic site or a more distant location
(18) (Fig 4).
Imaging Pitfalls
The ability to visualize active extravasation of
contrast material is influenced by many factors
related to the nature of the bleeding (severity,
intermittence), the patient (hemodynamic status,
body mass index, preexistent intestinal contents),
the CT technique (examination protocol, scanner
generation, iodine concentration of the contrast
material used), and the experience of the radiologist (29,50). In studies for the detection of active
bleeding with CT angiography, false-negative
1466 September-October 2013
radiographics.rsna.org
Figure 15. Initial false-negative interpretation in a
62-year-old man who presented to the emergency department with lower gastrointestinal tract bleeding and
malaise. (a) Coronal portal venous phase reformatted
image shows a tumor mass in the sigmoid colon (t1),
which was suggested as the likely cause of bleeding
despite an absence of active contrast material extravasation or hyperattenuating material in the vicinity of the
tumor (*). A mass in the left lung apex (t2) was also
depicted. (b) Oblique portal venous phase reformatted
image shows a duodenal ulcer (arrow) with a faint linear
focus of extravasation of contrast material (arrowheads)
corresponding to active bleeding. This finding was
discovered only on a second review of the results of
the examination.
results are most often due to intermittent or lowintensity bleeding, below the threshold of CT angiographic detection (32,34,36). Dilution of extravasated contrast material in fluid-filled dilated
bowel loops is another cause of nondepiction of
active bleeding with CT angiography (18,51).
Inadequate bowel distention precludes detection
of subtle tumors or inflammatory lesions that
may be better seen at CT enterography. Mucosal
enhancement of collapsed bowel loops can simulate extravasated contrast material (28,33). Errors
of perception and interpretation (such as satisfaction of search), which are usually related to the
inexperience of the radiologist, are other causes
of false-negative results (Fig 15).
Preexisting hyperattenuating material, such as
foreign bodies (including lines and tubes), metallic clips, suture material, or retained oral contrast
material in the bowel lumen, may mimic acute
gastrointestinal bleeding. These pitfalls can be
avoided by obtaining preliminary unenhanced images (18,51). Retention of previously administered
barium in colonic diverticula may be mistaken
for, or may obscure, acute extravasation of contrast material. A large amount of retained oral
contrast material in the digestive tract depicted
on the topogram may preclude completion of the
examination. Cone beam artifacts may produce
areas of hyperattenuation observed at the interface
between normal bowel contents and air in patients
with distended loops, a finding that can also lead
to false-positive results (32,51) (Fig 16). Hypervascular gastrointestinal tumors (stromal tumors,
carcinoid tumors, certain metastases) may appear
as hyperenhancing masses and may mimic acute
gastrointestinal bleeding. Importantly, these lesions can be the cause of true hemorrhage (Fig 12).
RG • Volume 33 Number 5
Artigas et al 1467
Figure 16. Potential pitfalls and image artifacts. (a) Coronal portal venous phase reformatted image shows cone
beam artifact (arrow) appearing as a focus of hyperattenuation in the gas-liquid interface, a finding that mimics extravasation of contrast material. (b) Axial portal venous phase image depicts hyperenhancing collapsed bowel wall
(arrows). (c, d) Axial portal venous phase image (c) shows hyperattenuating intraluminal material (arrow) that had
already been evident on the unenhanced image (d), thereby excluding active bleeding.
Conclusions
CT angiography is a noninvasive, robust, widely
available technique that can be easily implemented in patients with acute gastrointestinal
hemorrhage. The sensitivity of CT angiography
for detecting active bleeding and its cause is high.
The findings of CT angiography help guide optimal treatment (directed endoscopic intervention,
angiographic embolization, or surgery), as well
as determine the optimal timing to implement
these interventions. CT angiography is usually
performed when endoscopy is not feasible or is
nondiagnostic, but in most cases, CT angiogra-
phy can be completed while preparing for endoscopy. Even when active hemorrhage has ceased,
CT findings may assist in risk stratification and
may be helpful in selecting the best option for
definitive treatment, as well as its optimal timing.
Nevertheless, although this approach has been
successful in our hands, there is no consensus as
yet with regard to the primary role of CT angiography in the evaluation of gastrointestinal bleeding. A meticulous technique is necessary, and
radiologists should be aware of (a) the pertinent
1468 September-October 2013
radiographics.rsna.org
Figure 17. Proposed new algorithm for the initial evaluation of patients with acute gastrointestinal (GI) bleeding.
EGD = esophagogastroduodenoscopy, GIB = gastrointestinal bleeding, LGIB = lower gastrointestinal tract bleeding,
UGIB = upper gastrointestinal tract bleeding.
imaging findings that indicate active or recent
hemorrhage and (b) the most common sources of
potential pitfalls. Figure 17 summarizes our proposed new algorithm for the initial evaluation of
patients with acute gastrointestinal bleeding.
References
1.Zimmerman HM, Curfman K. Acute gastrointestinal bleeding. AACN Clin Issues 1997;8(3):
449–458.
2.Zhao Y, Encinosa W. Hospitalizations for gastrointestinal bleeding in 1998 and 2006. Statistical brief no.
65. Rockville, Md: Healthcare Cost and Utilization
Project (HCUP), Agency for Health Care Policy
and Research, 2008. HCUP Statistical Briefs Web
site. http://www.hcup-us.ahrq.gov/reports/statbriefs
/sb65.pdf. Published December 2008.
3.Whelan CT, Chen C, Kaboli P, Siddique J, Prochaska M, Meltzer DO. Upper versus lower gastrointestinal bleeding: a direct comparison of clinical
presentation, outcomes, and resource utilization. J
Hosp Med 2010;5(3):141–147.
4.Cappell MS, Friedel D. Initial management of acute
upper gastrointestinal bleeding: from initial evaluation up to gastrointestinal endoscopy. Med Clin
North Am 2008;92(3):491–509, xi.
5.Lee EW, Laberge JM. Differential diagnosis of gastrointestinal bleeding. Tech Vasc Interv Radiol 2004;
7(3):112–122.
6.Kennedy DW, Laing CJ, Tseng LH, Rosenblum DI,
Tamarkin SW. Detection of active gastrointestinal
hemorrhage with CT angiography: a 4½-year retrospective review. J Vasc Interv Radiol 2010;21(6):
848–855.
7.Friebe B, Wieners G. Radiographic techniques for
localization and treatment of gastrointestinal bleeding of obscure origin. Eur J Trauma Emerg Surg
2011;37(4):353–363.
8.Vreeburg EM, Snel P, de Bruijne JW, Bartelsman
JF, Rauws EA, Tytgat GN. Acute upper gastrointestinal bleeding in the Amsterdam area: incidence,
diagnosis, and clinical outcome. Am J Gastroenterol
1997;92(2):236–243.
9.Johnson JO. Diagnosis of acute gastrointestinal hemorrhage and acute mesenteric ischemia in the era
of multi-detector row CT. Radiol Clin North Am
2012;50(1):173–182.
10.Barnert J, Messmann H. Diagnosis and management of lower gastrointestinal bleeding. Nat Rev
Gastroenterol Hepatol 2009;6(11):637–646.
11.Wilcox CM, Clark WS. Causes and outcome of upper and lower gastrointestinal bleeding: the Grady
Hospital experience. South Med J 1999;92(1):
44–50.
12.Strate LL, Naumann CR. The role of colonoscopy
and radiological procedures in the management of
acute lower intestinal bleeding. Clin Gastroenterol
Hepatol 2010;8(4):333–343.
13.Lanas A, García-Rodríguez LA, Polo-Tomás M,
et al. The changing face of hospitalisation due to
gastrointestinal bleeding and perforation. Aliment
Pharmacol Ther 2011;33(5):585–591.
RG • Volume 33 Number 5
14.Frattaroli FM, Casciani E, Spoletini D, et al. Prospective study comparing multi-detector row CT
and endoscopy in acute gastrointestinal bleeding.
World J Surg 2009;33(10):2209–2217.
15.Lee SS, Oh TS, Kim HJ, et al. Obscure gastrointestinal bleeding: diagnostic performance of multidetector CT enterography. Radiology 2011;259(3):
739–748.
16.Manning-Dimmitt LL, Dimmitt SG, Wilson GR.
Diagnosis of gastrointestinal bleeding in adults. Am
Fam Physician 2005;71(7):1339–1346.
17.Steiner K, Gollub F, Stuart S, Papadopoulou A,
Woodward N. Acute gastrointestinal bleeding: CT
angiography with multi-planar reformatting. Abdom
Imaging 2011;36(2):115–125.
18.Chang WC, Tsai SH, Chang WK, et al. The value
of multidetector-row computed tomography for
localization of obscure acute gastrointestinal bleeding. Eur J Radiol 2011;80(2):229–235.
19.Angtuaco TL, Reddy SK, Drapkin S, Harrell LE,
Howden CW. The utility of urgent colonoscopy in
the evaluation of acute lower gastrointestinal tract
bleeding: a 2-year experience from a single center.
Am J Gastroenterol 2001;96(6):1782–1785.
20.Chu A, Ahn H, Halwan B, et al. A decision support
system to facilitate management of patients with
acute gastrointestinal bleeding. Artif Intell Med
2008;42(3):247–259.
21.Duchesne J, Jacome T, Serou M, et al. CT-angiography for the detection of a lower gastrointestinal
bleeding source. Am Surg 2005;71(5):392–397.
22.Kerr SF, Puppala S. Acute gastrointestinal haemorrhage: the role of the radiologist. Postgrad Med J
2011;87(1027):362–368.
23.Lee JG, Turnipseed S, Romano PS, et al. Endoscopy-based triage significantly reduces hospitalization rates and costs of treating upper GI bleeding:
a randomized controlled trial. Gastrointest Endosc
1999;50(6):755–761.
24.Green BT, Rockey DC, Portwood G, et al. Urgent
colonoscopy for evaluation and management of
acute lower gastrointestinal hemorrhage: a randomized controlled trial. Am J Gastroenterol 2005;100
(11):2395–2402.
25.Schuetz A, Jauch KW. Lower gastrointestinal bleeding: therapeutic strategies, surgical techniques and
results. Langenbecks Arch Surg 2001;386(1):17–25.
26.Copland A, Munroe CA, Friedland S, Triadafilopoulos G. Integrating urgent multidetector CT
scanning in the diagnostic algorithm of active lower
GI bleeding. Gastrointest Endosc 2010;72(2):
402–405.
27.Uyeda JW, Anderson SW, Sakai O, Soto JA. CT angiography in trauma. Radiol Clin North Am 2010;
48(2):423–438, ix–x.
28.Scheffel H, Pfammatter T, Wildi S, Bauerfeind P,
Marincek B, Alkadhi H. Acute gastrointestinal
bleeding: detection of source and etiology with
multi-detector-row CT. Eur Radiol 2007;17(6):
1555–1565.
Artigas et al 1469
29.Martí M, Artigas JM, Garzón G, Alvarez-Sala R,
Soto JA. Acute lower intestinal bleeding: feasibility
and diagnostic performance of CT angiography.
Radiology 2012;262(1):109–116.
30.Martí de Gracia M, Artigas Martín JM. Multidetector computed tomography as the first diagnostic
option for acute lower digestive tract bleeding in
the emergency department [in Spanish]. Radiologia
2011;53(suppl 1):43–50.
31.Yoon W, Jeong YY, Shin SS, et al. Acute massive
gastrointestinal bleeding: detection and localization
with arterial phase multi–detector row helical CT.
Radiology 2006;239(1):160–167.
32.Geffroy Y, Rodallec MH, Boulay-Coletta I, Jullès
MC, Ridereau-Zins C, Zins M. Multidetector CT
angiography in acute gastrointestinal bleeding: why,
when, and how. RadioGraphics 2011;31(3):E35–
E46. doi:10.1148/rg.313105206. Published May
2011. [Published corrections appear in RadioGraphics 2011;31(5):1496 and RadioGraphics
2011;31 (7):2114. Fullès, Marie-Christine (corrected to Jullès, Marie-Christine).]
33.Laing CJ, Tobias T, Rosenblum DI, Banker WL,
Tseng L, Tamarkin SW. Acute gastrointestinal
bleeding: emerging role of multidetector CT angiography and review of current imaging techniques.
RadioGraphics 2007;27(4):1055–1070.
34.Dobritz M, Engels HP, Schneider A, et al. Evaluation of dual-phase multi-detector-row CT for detection of intestinal bleeding using an experimental
bowel model. Eur Radiol 2009;19(4):875–881.
35.Wu LM, Xu JR, Yin Y, Qu XH. Usefulness of CT
angiography in diagnosing acute gastrointestinal
bleeding: a meta-analysis. World J Gastroenterol
2010;16(31):3957–3963.
36.Yoon W, Jeong YY, Kim JK. Acute gastrointestinal
bleeding: contrast-enhanced MDCT. Abdom Imaging 2006;31(1):1–8.
37.Lee S, Welman CJ, Ramsay D. Investigation of acute
lower gastrointestinal bleeding with 16- and 64-slice
multidetector CT. J Med Imaging Radiat Oncol
2009;53(1):56–63.
38.Dobritz M, Engels HP, Schneider A, Bauer J, Rummeny EJ. Detection of intestinal bleeding with
multi-detector row CT in an experimental setup:
how many acquisitions are necessary? Eur Radiol
2009;19(12):2862–2869.
39.Foley PT, Ganeshan A, Anthony S, Uberoi R.
Multi-detector CT angiography for lower gastrointestinal bleeding: can it select patients for endovascular intervention? J Med Imaging Radiat Oncol
2010;54(1):9–16.
40.Ernst O, Bulois P, Saint-Drenant S, Leroy C, Paris
JC, Sergent G. Helical CT in acute lower gastrointestinal bleeding. Eur Radiol 2003;13(1):114–117.
41.Tew K, Davies RP, Jadun CK, Kew J. MDCT of
acute lower gastrointestinal bleeding. AJR Am J
Roentgenol 2004;182(2):427–430.
1470 September-October 2013
42.Lubner M, Menias C, Rucker C, et al. Blood in
the belly: CT findings of hemoperitoneum. RadioGraphics 2007;27(1):109–125.
43.Kirchhof K, Welzel T, Mecke C, Zoubaa S, Sartor
K. Differentiation of white, mixed, and red thrombi:
value of CT in estimation of the prognosis of
thrombolysis—phantom study. Radiology 2003;228
(1):126–130.
44.Lane MJ, Katz DS, Shah RA, Rubin GD, Jeffrey RB
Jr. Active arterial contrast extravasation on helical
CT of the abdomen, pelvis, and chest. AJR Am J
Roentgenol 1998;171(3):679–685.
45.Hamilton JD, Kumaravel M, Censullo ML, Cohen
AM, Kievlan DS, West OC. Multidetector CT evaluation of active extravasation in blunt abdominal
and pelvic trauma patients. RadioGraphics 2008;28
(6):1603–1616.
46.Vu QD, Menias CO, Bhalla S, Peterson C, Wang
LL, Balfe DM. Aortoenteric fistulas: CT features
and potential mimics. RadioGraphics 2009;29(1):
197–209.
radiographics.rsna.org
47.Sabharwal R, Vladica P, Chou R, Law WP. Helical
CT in the diagnosis of acute lower gastrointestinal
haemorrhage. Eur J Radiol 2006;58(2):273–279.
48.Horton KM, Jeffrey RB Jr, Federle MP, Fishman
EK. Acute gastrointestinal bleeding: the potential
role of 64 MDCT and 3D imaging in the diagnosis.
Emerg Radiol 2009;16(5):349–356.
49.Junquera F, Quiroga S, Saperas E, et al. Accuracy
of helical computed topographic angiography for
the diagnosis of colonic angiodysplasia. Gastroenterology 2000;119(2):293–299.
50.Jaeckle T, Stuber G, Hoffmann MH, Jeltsch M,
Schmitz BL, Aschoff AJ. Detection and localization of acute upper and lower gastrointestinal (GI)
bleeding with arterial phase multi-detector row helical CT. Eur Radiol 2008;18(7):1406–1413.
51.Stuber T, Hoffmann MH, Stuber G, Klass O, Feuerlein S, Aschoff AJ. Pitfalls in detection of acute gastrointestinal bleeding with multi-detector row helical
CT. Abdom Imaging 2009;34(4):476–482.
TM
This journal-based SA-CME activity has been approved for AMA PRA Category 1 Credit . See www.rsna.org/education/search/RG.
Teaching Points
September-October Issue 2013
Multidetector CT Angiography for Acute Gastrointestinal Bleeding: Technique and Findings
José M. Artigas, MD, PhD • Milagros Martí, MD, PhD • Jorge A. Soto, MD, PhD • Helena Esteban, MD •
Inmaculada Pinilla, MD • Eugenia Guillén, MD
RadioGraphics 2013; 33:1453–1470 • Published online 10.1148/rg.335125072 • Content Codes:
Page 1457
Although the specific technical parameters used and the phase images acquired vary among institutions, most agree that three-phase examinations that include the unenhanced, arterial, and portal venous
phases provide the best and most reproducible results for this clinical application in patients with acute
gastrointestinal bleeding.
Page 1458
For CT angiography performed to evaluate acute gastrointestinal bleeding, no oral contrast material is
routinely administered.
Page 1459
Severe bleeding episodes, such as those manifesting with hemodynamic instability, increase the pretest
probability of a positive result for active bleeding at CT angiography.
Page 1459
For detecting active bleeding with CT angiography, the highest sensitivity is achieved by combining
findings from the arterial and portal venous phases (28,36); the changing appearance of the focus of extravasated contrast material with time (from the arterial phase to the portal venous phase) unequivocally
confirms the presence of bleeding, especially when the unenhanced image shows no hyperattenuating
intraluminal material.
Page 1460
Portal venous phase imaging depicts extravascular blushes with higher sensitivity than arterial phase
imaging does because more time has elapsed, which allows the focus of extravasated contrast-enhanced
blood to enlarge and increase in attenuation within the lumen of the bowel.