Management of Acute Lower

Gastrointestinal Bleeding
Ethan J. Speir, BS, R. Mitchell Ermentrout, MD, and Jonathan G. Martin, MD

Acute lower gastrointestinal bleeding (LGIB), defined as hemorrhage into the gastro-
intestinal tract distal to the ligament of Treitz, is a major cause of morbidity and mortality
among adults. Overall, mortality rates are estimated between 2.4% and 3.9%. The most
common etiology for LGIB is diverticulosis, implicated in approximately 30% of cases,
with other causes including hemorrhoids, ischemic colitis, and postpolypectomy
bleeding. Transcatheter visceral angiography has begun to play an increasingly
important role in both the diagnosis and treatment of LGIB. Historically, transcatheter
visceral angiography has been used to direct vasopressin infusion with embolization
reserved for treatment of upper gastrointestinal bleeding. However, advances in micro-
catheter technology and embolotherapy have enabled super-selective embolization to
emerge as the treatment of choice for many cases of LGIB.
Tech Vasc Interventional Rad ]:]]]-]]] C 2017 Elsevier Inc. All rights reserved.

KEYWORDS Acute lower gastrointestinal bleeding, hemorrhage, embolization

Clinical Evaluation/Indication 2 modalities are given equal appropriateness ratings by

Initial management of a patient presenting with acute
lower gastrointestinal bleeding (LGIB) should focus on
securing intravenous access and initiating fluid resuscita-
tion with crystalloids or blood products if necessary. Any
underlying coagulopathy or thrombocytopenia should be
promptly corrected.
Further evaluation of the patient’s clinical status will
direct management. A valuable method for categorizing
patient status is described in the American College of
Radiology (ACR) Appropriateness Criteria.1 Based upon
these criteria, patients presenting with hematochezia or
melena who are hemodynamically stable should first
undergo colonoscopy. However, endoscopic evaluation
may be limited by either inadequate bowel preparation or
due to active bleeding, which can obscure visualization of
colonic mucosa. In these patients, noninvasive radio-
graphic studies such as computed tomography angiogra-
phy (CTA) and 99mTechnetium-labeled red blood cell
(RBC) scintigraphy can be performed. Although these
Department of Radiology and Imaging Sciences, Emory University
Hospital, Emory University School of Medicine, Atlanta, GA.
Address reprint requests to Ethan J. Speir, Department of Radiology and
Imaging Sciences, Emory University Hospital, Emory University
School of Medicine, 1364 Clifton Rd NE, Atlanta, GA 30322. E-mail:
ethan.speir@emory.edu
1089-2516/14/$ - see front matter & 2017 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1053/j.tvir.2017.10.005
ACR Criteria, CTA is preferred at our institution as this
modality has greater accuracy in localizing the source of
the hemorrhage2 in addition to providing valuable ancil-
lary data regarding patient anatomy and underlying
pathology. Furthermore, recent data from within our
institution indicates that positive findings on CTA are
corroborated on transcatheter visceral angiography (TVA)
significantly more often compared to RBC scintigraphy.
For patients presenting with LGIB and hemodynamic
instability, defined as systolic blood pressure ≤90 mmHg,
or who require more than 5 units of blood, ACR Criteria
indicates that the most appropriate intervention is TVA.
Even in these patients, however, we will perform a pre-
procedural CTA if it can be completed without delay of
TVA.
Contraindications
There are no absolute contraindications for TVA, partic-
ularly when a patient is presenting with life-threatening
LGIB. For patients with a history of severe contrast allergy,
pre-treatment with corticosteroids can be considered, and
intubation for airway protection should be considered.
Alternatively, carbon dioxide (CO2) may be used as a
contrast agent.3
1
2 E.J. Speir et al.
Equipment Needed
The choice of equipment used for super-selection
of mesenteric arteries will often depend on product
availability, patient anatomy, and physician preference.
A wide range of guide and microcatheter systems are used
at our institution.
Available embolic agents for super-selective emboliza-
tion (SSE) include Gelfoam (Pharmacia and Upjohn Co,
Kalamazoo, MI), particles, microcoils, n-butyl cyanoacry-
late (NBCA) glue (Codman Neuro, Warren NJ), and
ethylene vinyl alcohol copolymer (Onyx) (ev3, Plymouth,
MN). In the absence of robust clinical trials comparing the
relative efficacy of embolic agents, operator preference
often dictates which agent is used. Microcoils are the most
commonly used embolic agent at our intuition. However,
there are particular clinical situations in which other
embolics may be preferred. If bleeding is nonfocal or if
the microcatheter cannot be advanced directly adjacent to
the bleed, flow-directed polyvinyl alcohol of a minimum
size of 300 μm may be used.4 For patients with uncorrect-
able coagulopathy, the use of a liquid casting agent such as
NBCA or Onyx may be considered as these agents do not
require thrombus formation in order to create vessel
occlusion.5,6
Overcoming Technical
Challenges
Distinguishing a true source of hemorrhage on TVA may be
difficult especially when patient motion or peristaltic bowel
activity results in misleading artifacts on digital subtraction
angiography. Reviewing suspicious findings on nonsub-
tracted imaging may alleviate uncertainty in this setting.
Additionally, administration of 1 mg intravenous glucagon
prior to TVA will often successfully arrest bowel peristalsis.
Occasionally, extravasation seen during evaluation of a
parent vessel may not be able to be reproduced after
further selection is performed. This may be due to
temporary vasospasm of the bleeding vessel. If vasospasm
is suspected, an aliquot of 100-200 μg of nitroglycerine
can be injected intra-arterially.
Given the intermittent nature of LGIB, a frequent
challenge encountered during TVA is inability to corrob-
orate positive findings observed on preprocedural imag-
ing. For example, given the high sensitivity of RBC
scintigraphy along with its prolonged acquisition time,
PPV for this modality has been reported to be as low as
24%.7 In the setting of nonrevealing TVA, one may
consider the use of CO2 contrast. The extremely low
viscosity of CO2 allows for easy delivery through even the
smallest diameter microcatheters and, as some have
argued, may even confer a higher sensitivity for active
LGIB.3 Alternatively, use of cone-beam CT with accom-
panying vessel-tracking software, if available, may improve
detection of bleeding and expedite selection of the culprit
vessel.8 Finally, one may consider provocative angiogra-
phy using systemic heparinization and selective injection
of a vasodilator along with tissue plasminogen activator
(tPA).9 At our institution, however, this approach is
typically reserved for patients with obscure LGIB who
have repeatedly negative TVAs.
Recognizing and Treating
Complications
Complications from TVA include those associated with
arteriography and include bleeding, hematoma, or
pseudoaneurysm at the at the access site. Although
nontarget embolization may occur, a more common
complication of SSE is bowel ischemia. This occurs
more often in the lower gastrointestinal tract due to the
lack of extensive collateral networks seen in the upper
tract. Minor ischemia manifesting as abdominal pain
with or without increase in serum lactic acid levels is
not uncommon. Major ischemia requiring surgical
resection occurs in 1.3%-5% of patients.10,11 Risk of
ischemia is directly proportional to the extent of
vascular territory embolized with animal models show-
ing a substantially increased risk when ≥4 vasa recta
are embolized.12,13
Clinical Follow-Up
In addition to monitoring for access site complications,
clinical observation should focus on hemodynamic
status and serial hemoglobin and hematocrit levels to
determine if there is continued resolution of bleeding.
Stool character is a less reliable predictor of persistent
bleeding as continued passage of blood may represent
residua of earlier bleeding. Routine monitoring of
serum lactic acid is not necessary following emboliza-
tion but is appropriate if the patient develops severe
and progressive abdominal pain. Temporary abdominal
pain or a transient rise in lactic acid warrants further
evaluation if either fails to resolve. Cross-sectional
imaging or endoscopy may be necessary to assess for
bowel infarction.
Expected Outcomes
TVA with SSE is highly effective in treating acute LGIB
with technical success rates as high as 96.4%-100%.14,15
Clinical success rates are somewhat lower, ranging from
63.0%-91.2%, often due to rebleeding following emboli-
zation.15,16 Notably, thrombocytopenia prior to SSE is
strongly associated with clinical failure.17
Case 1
A 91-year-old man with a history of bladder cancer,
prostate cancer, and diverticular bleeding presented from
an outside rehabilitation facility with hematochezia.
Upon presentation, the patient had a blood pressure of
115/65 and required 3 units of packed RBCs within a 24-
Acute LGIB 3

A B

C D

Figure 1 : (A) Axial abdominal CTA obtained during the arterial phase demonstrating pooling of contrast in the
mid-portion of the sigmoid colon (arrow) consistent with active bleeding. (B) DSA of the IMA without definite
evidence of active extravasation. (C) Super-selective DSA demonstrating active extravasation from distal branches
of the IMA supplying the rectosigmoid region (circle). (D) Super-selective embolization performed with a single
microcoil (arrow) placed as proximal to the source of bleeding as possible with resulting hemostasis.

hour time period. Given the patients hemodynamic
stability, he first underwent colonoscopy, which revealed
fresh blood along the mucosa of the sigmoid colon along
with multiple diverticula. However, no site of active
hemorrhage was able to be identified. A CTA was
subsequently performed showing pooling of contrast in
the midportion of the sigmoid colon on arterial phase
consistent with acute bleeding (Fig. 1A). The patient was
brought to interventional radiology for TVA with possible
embolization.
Under ultrasound guidance, the right common femoral
artery was accessed using a micropuncture needle.
A 0.018′′ Nitinol wire was advanced through the needle
into the artery and the needle was exchanged for a 5 Fr.
transitional catheter. The wire and inner dilator were
removed and a 0.035′′ Bentson wire was advanced into
the artery. The transitional catheter was exchanged for a
5 Fr. Pinnacle (Terumo, Tokyo, Japan) vascular introducer
sheath, and the sheath was attached to a heparinized,
pressurized bag of saline. A 5 Fr Mickelson (Cook Medical
Inc, Bloomington, IN) catheter was advanced over
the Bentson wire and formed in the abdominal aorta.
It was then used to select the inferior mesenteric
artery (IMA). Arteriography was performed through the
Mickelson catheter in the IMA with an injection rate of
3 mL/s without definite evidence of active extravasation
(Fig 1B).
Given the focally positive CTA findings, targeted inves-
tigation for a bleeding source was performed with micro-
catheter evaluation. A 2.4 Fr angled Renegade STC-18
(Boston Scientific, Marlborough, MA) microcatheter was
advanced over a 0.016′′ Fathom (Boston Scientific, Marl-
borough, MA) microwire into branches of the IMA. Super-
selective arteriography was performed demonstrating
active extravasation from distal branches supplying the
rectosigmoid region (Fig. 1C). Further selection of these
branches was performed with a 1.9 Fr. Echelon (Med-
tronic, Irvine, CA) microcatheter and a 0.018′′ double
angle GT (Terumo, Tokyo, Japan) microwire. Super-
selective coil embolization of a sixth order branch was
then performed with a single 2 mm × 4 cm Concerto
(Covidien, Dublin, Ireland) microcoil placed as proximate
to the source of bleeding as possible (Fig. 1D). Postembo-
lization arteriography of this vessel demonstrated
4 E.J. Speir et al.

A B

C D

Figure 2 : (A) RBC scintigraphy with intense radiotracer uptake within the cecal lumen (arrow) consistent with
active large bowel gastrointestinal bleed in this region. (B) DSA of the superior mesenteric artery (SMA) showing a
focal area of blush in the right colon (arrow). (C) DSA of branches of the ileocolic artery following super-selection
redemonstrating of the area of extravasation (arrow). (D) Postembolization DSA shows an NBCA cast filling the
culprit ileocolic branch (arrow) with resultant resolution of active hemorrahage (circle).

hemostasis. A right common femoral angiogram was
obtained through the sheath to evaluate the location of
the arterial access site. An Angio-Seal (Terumo, Tokyo,
Japan) device was deployed at the arteriotomy resulting in
immediate hemostasis.
Case 2
A 62-year-old man with a remote history of upper gastro-
intestinal bleeding due to hemorrhagic gastritis presented
to the emergency department following 2 episodes of
bright red blood per rectum. His blood pressure remained
stable at 120/80 following initiation of crystalloids and he
did not require blood transfusion. Despite his history,
LGIB was presumed given bright red blood per rectumin
the setting of hemodynamic stability. As such, a RBC
scintigraphy study was performed at the request of the
emergency medicine physician prior to interventional
radiology consultation. The study demonstrated radio-
tracer uptake within the cecum consistent with active
bleeding (Fig. 2A). Based upon these findings, the patient
was brought to interventional radiology for TVA with
possible embolization.
Under ultrasound guidance, the right common femoral
artery was accessed using a micropuncture needle. This
was upsized to a 5 Fr Pinnacle (Terumo, Tokyo, Japan)
vascular introducer sheath using the method described in
the preceding case. Through the sheath, a 5 Fr Sim-1
Glidecath (Terumo, Tokyo, Japan) was advanced over a
Bentson wire into the abdominal aorta. The reverse curve
was subsequently formed using a Tevdek (Teleflex,
Gurnee, IL) suture for traction, and the celiac artery
was engaged. Arteriography was performed showing
normal anatomy without evidence of extravasation. Next,
the superior mesenteric artery was selected and arteriog-
raphy performed. Active bleeding was identified with a
focal area of blush in the right colon supplied by ileocolic
arterial branches (Fig. 2B). Based on these findings, an
Echelon 14 (Medtronic, Irvine, CA) microcatheter was
advanced over a 0.014′′ Transend (Stryker Neurovascu-
lar, Fremont, CA) microwire into the ileocolic artery. This
was then used to super-select branches of the ileocolic
artery (Fig. 2C).
Acute LGIB
Once the culprit vessel was catheterized, there was
difficulty advancing the microcatheter across the origin
of the bleed. Due to the proximal microcatheter position-
ing, combined with a desire to embolize across the
bleeding origin and operator familiarity with the ability
to advance glue farther in to the target vessel, glue
embolization was performed with approximately
0.25 mL of NBCA glue diluted 1:3 in Lipiodol (Guerbert,
Villepinte, France). This mixture ratio was chosen based
upon evidence describing NBCA polymerization times.18
To prevent catheter adhesion, care was taken to gently pull
back the catheter during injection and, once the NBCA had
infiltrated the appropriate vascular distribution, the cath-
eter was quickly removed. Postembolization angiography
was performed demonstrating adequate hemostasis
(Fig. 2D). Angiography of the right common femoral
arterial access site was performed through the vascular
sheath to document access site. Finally, the arteriotomy
was closed using a VASCADE (Cardiva Medical, Inc, Santa
Clara, CA) closure device.
References
1. Darcy MD, Cash BD, Feig, BW, et al: American College of Radiology
ACR Appropriateness Criteria® Radiologic Management of Lower
Gastrointestinal Tract Bleeding. Available from: https://acsearch.acr.
org/docs/69457/Narrative/.
2. Feuerstein JD, Ketwaroo G, Tewani SK, et al: Localizing acute lower
gastrointestinal hemorrhage: CT angiography versus tagged RBC
scintigraphy. AJR Am J Roentgenol 3:578-584, 2016
3. Back MR, Caridi JG, Hawkins IF Jr, et al: Angiography with carbon
dioxide (CO2). Surg Clin North Am 4:575-591, 1998
4. Urbano J, Manuel Cabrera J, Franco A, et al: Selective arterial
embolization with ethylene-vinyl alcohol copolymer for control of
massive lower gastrointestinal bleeding: Feasibility and initial expe-
rience. J Vasc Interv Radiol 6:839-846, 2014
5. Guimaraes M, Wooster M: Onyx (ethylene-vinyl alcohol copolymer)
in peripheral applications. Semin Intervent Radiol 3:350-356, 2011
5
6. Jae HJ, Chung JW, Jung AY, et al: Transcatheter arterial embolization
of nonvariceal upper gastrointestinal bleeding with N-butyl cyanoa-
crylate. Korean J Radiol 1:48-56, 2007
7. Yi WS, Garg G, Sava JA: Localization and definitive control of lower
gastrointestinal bleeding with angiography and embolization. Am
Surg 4:375-380, 2013
8. Ierardi AM, Urbano J, De Marchi G, et al: New advances in lower
gastrointestinal bleeding management with embolotherapy. Br J
Radiol 1061:20150934, 2016
9. Kim CY, Suhocki PV, Miller MJ Jr., et al: Provocative mesenteric
angiography for lower gastrointestinal hemorrhage: Results from a
single-institution study. J Vasc Interv Radiol 4:477-483, 2010
10. Kickuth R, Rattunde H, Gschossmann J, et al: Acute lower gastro-
intestinal hemorrhage: Minimally invasive management with micro-
catheter embolization. J Vasc Interv Radiol 9 2008 1289-96.e2
11. Lipof T, Sardella WV, Bartus CM, et al: The efficacy and durability of
super-selective embolization in the treatment of lower gastrointes-
tinal bleeding. Dis Colon Rectum 3:301-305, 2008
12. Ikoma A, Kawai N, Sato M, et al: Ischemic effects of transcatheter
arterial embolization with N-butyl cyanoacrylate-lipiodol on the
colon in a Swine model. Cardiovasc Intervent Radiol 5:1009-1015,
2010
13. Jae HJ, Chung JW, Kim HC, et al: Experimental study on acute
ischemic small bowel changes induced by superselective emboliza-
tion of superior mesenteric artery branches with N-butyl cyanoa-
crylate. J Vasc Interv Radiol 5:755-763, 2008
14. Hur S, Jae HJ, Lee M, et al: Safety and efficacy of transcatheter
arterial embolization for lower gastrointestinal bleeding: A single-
center experience with 112 patients. J Vasc Interv Radiol 1:10-19,
2014
15. Koh DC, Luchtefeld MA, Kim DG, et al: Efficacy of transarterial
embolization as definitive treatment in lower gastrointestinal bleed-
ing. Colorectal Dis 1:53-59, 2009
16. Tan KK, Wong D, Sim R: Superselective embolization for lower
gastrointestinal hemorrhage: An institutional review over 7 years.
World J Surg 12:2707-2715, 2008
17. Chan DK, Soong J, Koh F, et al: Predictors for outcomes after super-
selective mesenteric embolization for lower gastrointestinal tract
bleeding. ANZ J Surg 6:459-463, 2016
18. Pollak JSWhite RI Jr: The use of cyanoacrylate adhesives in
peripheral embolization. J Vasc Interv Radiol 8:907-913, 2001