Clinical Update

American Society for Gastrointestinal Endoscopy

Editor: Ronnie Fass, MD

Vol. 14, No. 3 January 2007

Commentary: Upper GI bleeding is one of the most important therapeutic areas in gastroenterology. Many of the cases fall under the
category of an emergency situation and thus require urgent endoscopic intervention. In this review, Dr Loren Laine provides updated
information about the epidemiology, the clinical presentation, and therapeutic interventions in various GI bleeding scenarios. In
addition, Dr Laine provides a critical review of some of the diagnostic and therapeutic approaches that are commonly pursued
despite their very low yield. Lastly, Dr Laine emphasizes, in his review, the importance of combining endoscopic and medical
therapeutic interventions in patients who present with upper-GI bleeding.
Ronnie Fass, MD, Editor

UPPER GASTROINTESTINAL BLEEDING

Loren Laine, MD

University of Southern California School of Medicine, Los Angeles, California

EPIDEMIOLOGY
The annual incidence of hospitalizations for upper-GI bleeding (UGIB) is
approximately one per thousand adults.1 Currently, the average mortality is
likely to be in the 5% to 10% range. Independent predictors of recurrent
bleeding and death include the severity of the bleeding episode as manifested by tachycardia and hypotension, patient age, and comorbidities.2 Patients
with UGIB rarely die from exsanguination but, rather, death occurs from
decompensation of other illnesses. For example, in a large observational
study in the United Kingdom, mortality with UGIB in patients <60 years
of age with no concurrent illnesses was 0.1%.3

Melena is most commonly due to an upper GI source and can result from as
little as 50 to 100 mL of blood.4 The presence of melena indicates that
blood has been in the GI tract for at least 14 hours.5 Thus, the more proximal the bleeding lesion in the GI tract, the more likely melena will occur.
However, an upper-GI source may occasionally bleed so briskly that the
blood does not remain in the GI tract long enough for melena to occur. For
example, instillation of a liter of blood into the upper-GI tract leads, at least
initially, to hematochezia.4 If hematochezia is from an upper-GI site, it
reflects major bleeding, with hemodynamic instability and an increased
mortality.

PRESENTATION

INITIAL EVALUATION

UGIB, defined as bleeding from a source in the esophagus, the stomach, or

the duodenum, may be overt or occult. Overt bleeding may present as
hematemesis (vomiting blood or coffee-ground material), melena (black,
tarry stool), or hematochezia (red or maroon blood per rectum). Occult
bleeding may present as a positive fecal occult blood test (eg, guaiac testing),
laboratory evidence of anemia and iron deficiency, or symptoms of anemia
or blood loss (eg, fatigue, lightheadedness, syncope, dyspnea, angina). This
update will focus on patients with overt evidence of UGIB.

Heart rate and blood pressure provide the most important information in
the initial assessment of a patient with UGIB. Major bleeding leads to postural changes in heart rate or blood pressure, tachycardia, and, eventually
hypotension. The Hb level is less useful at the time of presentation. Even
with a major bleeding episode, the Hb level may be normal or only minimally
decreased at the time of initial presentation because people bleed whole
blood. As extravascular fluid enters the vascular space to restore volume, the
Hb level falls, but equilibration may take up to 72 hours.6

Hematemesis confirms an upper-GI source of bleeding, assuming that the

blood did not originate outside the GI tract (eg, nose, respiratory tract).

Nasogastric aspirates that are grossly bloody confirm an upper-GI source,

but a negative aspirate does not rule out an upper-GI source. Up to 18% of

Copyright 2007 American Society for Gastrointestinal Endoscopy

1520 Kensington Road, Suite 202, Oak Brook, IL 60523
Phone 630.573.0600, Fax 630.573.0691, www.asge.org, www.askasge.org
To purchase additional copies, please contact the ASGE ISSN 1070-7212
www.asge.org www.askasge.org

Sponsored by:

TABLE 1. OUTCOMES WITH ONLY SUPPORTIVE THERAPY10 AND RECOMMENDED MANAGEMENT OF PATIENTS WITH
BLEEDING ULCER BASED ON ENDOSCOPIC FINDINGS
Mean rate of surgery for bleeding
with only supportive therapy, %

Recommended therapy

Length of hospitalization, days*

0.5

Oral PPI

Flat pigmented spot

Oral PPI

Adherent clot

10

IV PPI infusion endoscopic therapy

Nonbleeding visible vessel

34

Endoscopic therapy + IV PPI infusion

Active bleeding

35

Endoscopic therapy + IV PPI infusion

Endoscopic finding
Clean base

*Assuming no other medical problems, no recurrent bleeding, normal vital signs, and stable Hb level.

patients with UGIB have a nonbloody nasogastric aspirate, most commonly

because of bleeding from the duodenum.7 Although some authors suggest
that a nonbloody bile-stained aspirate rules out a duodenal source, physicians
are incorrect about 50% of the time when they report bile in the aspirate.8
History and physical examination are not usually diagnostic of the source of
UGIB, and upper endoscopy is the diagnostic test of choice. Timing of an
endoscopy is controversial.9 The significant benefit of endoscopic therapy in
high-risk patients (discussed below) suggests that early endoscopy should be
beneficial in patients with high-risk clinical features. However, no randomized controlled trials have confirmed that early endoscopy improves clinical
outcomes. Nevertheless, I suggest that patients with clinical evidence of
major bleeding (eg, tachycardia, hypotension, orthostatic changes) undergo
an endoscopy soon after resuscitation and volume stabilization. An early
endoscopy decreases costs in patients with low-risk clinical features, because
patients with normal vital signs, a stable Hb level, and no other medical
problems who are found to have low-risk lesions (eg, clean-based ulcers,
nonbleeding Mallory-Weiss tears, erosive or hemorrhagic gastropathy) at
endoscopy can be discharged home.9

SOURCES OF UGIB
Peptic ulcer
A peptic ulcer is the most common cause of UGIB, although the incidence
of bleeding ulcers has decreased over the past few decades. Peptic ulcers are
responsible for up to approximately 50% of UGIB cases. The decrease in
bleeding ulcers is related to the decreasing Helicobacter pylori prevalence and
possibly the increased use of acid-suppressive medications. An increasing
proportion of bleeding ulcers is caused by nonsteroid anti-inflammatory
drug (NSAID) use.
In addition to the clinical predictors of outcome discussed above, the
appearance of an ulcer at endoscopy provides important prognostic information and is used to guide subsequent management (Table 1). Approximately a third of patients found to have an ulcer with active bleeding or a
nonbleeding visible vessel will have further bleeding that requires surgery if
treated expectantly.10 These patients should receive endoscopic therapy and
intravenous (IV) infusion of a proton pump inhibitor (PPI) (discussed
2

www.asge.org www.askasge.org

below). Patients with adherent clots also should receive IV infusion of a PPI,
and some investigators suggest endoscopic therapy as well.11,12 Patients with
flat pigmented spots and those with clean-based ulcers do not require endoscopic therapy or IV PPIs. Because patients with clean-based ulcers have
rates of recurrent bleeding that are close to zero, they can be discharged
home once vital signs are normal and the Hb level is stable. Patients with
other endoscopic stigmata generally are hospitalized for 3 days after the
bleeding episode (assuming no recurrence), because most recurrent bleeding
occurs within 3 days.
Randomized controlled trials show that endoscopic therapy significantly
improves outcomes, including further bleeding, transfusion, the need for surgery, the length of hospital stay, and costs in patients with high-risk bleeding ulcers (active bleeding, nonbleeding visible vessel, and perhaps adherent
clot). Furthermore, meta-analyses indicate that mortality is also significantly
decreased with endoscopic therapy.13 Standard forms of endoscopic therapy
include thermal contact devices that use tamponade and heat to provide
hemostasis (bipolar electrocoagulation or heat probe), injection therapy (in
which dilute epinephrine and/or sclerosant agents are injected into the base
of the bleeding ulcer), and mechanical treatment with clips.
Randomized controlled trials also indicate that, even after endoscopic therapy, a bolus of an IV PPI, followed by a constant infusion of PPI for 72
hours will significantly lessen recurrent bleeding (but not mortality).14 Most
studies used IV omeprazole, which is not available in the United States, and
the correct doses for the available PPIs have not been appropriately studied.
Nevertheless, the doses commonly used are a 80-mg bolus and a 8 mg/h
infusion for pantoprazole and esomeprazole and a 90-mg bolus and a 6 to
9 mg/h infusion for lansoprazole. The hypothesis behind the use of highdose constant infusion PPI therapy is based on older experimental studies
that suggest that maintaining intragastric pH >6 will enhance clot formation and clot stability.15 Studies from Kashmir and Iran report significantly
less recurrent bleeding with intermittent high-dose oral PPI therapy
(omeprazole 40 mg twice a day or 20 mg 4 times a day).16,17 However,
because PPI therapy also may have greater antisecretory potency in Asian
populations, we would not recommend the use of intermittent PPI therapy
in place of constant infusion IV therapy unless randomized controlled trials
in a Western population document efficacy with intermittent therapy.
Vol. 14, No. 3 January 2007

Prevention of recurrent ulcer bleeding in the long term should focus on the
3 main factors that cause ulcers: H pylori, NSAIDs, and acid. Eradication of
H pylori in patients with bleeding ulcers decreases the rate of recurrent bleeding to <5%. If a bleeding ulcer develops in a patient taking NSAIDs, the
NSAIDs should be discontinued if possible. If NSAIDs are required, the
most appropriate strategy to prevent recurrent bleeding ulcers is a cyclooxygenase-2 (COX-2) selective inhibitor plus a PPI. The annual incidence of
recurrent ulcer bleeding with either a COX-2 selective inhibitor or a PPI
plus traditional NSAID alone is approximately 10%, whereas the use of the
combination significantly reduces risk.18 Patients with bleeding ulcers unrelated to H pylori or NSAIDs should remain on full-dose antisecretory therapy
indefinitely.

Mallory-Weiss tears
Mallory-Weiss tears are diagnosed in approximately 5% to 15% of patients
with UGIB. Mallory Weiss tears occur at the gastroesophageal junction, primarily on the gastric side. Vomiting, retching, or coughing is classically
reported before hematemesis, especially in a patient who is alcoholic.
Bleeding stops spontaneously in most patients and rarely recurs (0% to 7%).
Endoscopic therapy is indicated only if the Mallory-Weiss tear is seen to be
actively bleeding at endoscopy.

Esophageal varices and portal hypertension

Esophageal varices account for approximately 5% to 30% of cases of UGIB,
and the proportion varies widely, depending on the population served.
Patients with variceal hemorrhage have higher rates of recurrent bleeding and
death than patients with the other common sources of UGIB. Endoscopic
therapy has been documented to significantly reduce further bleeding and
mortality in randomized controlled trials. Endoscopic therapy is used in the
acute setting, when patients present with bleeding and is also repeated at regular intervals (eg, 1 to 2 weeks) to eradicate varices. Once varices are eradicated, patients return periodically (eg, 3- to 12-month intervals) to identify
and treat recurrent varices. Rubber-band ligation has supplanted sclerotherapy for endoscopic treatment of esophageal varices, because it has less
recurrent bleeding, has a lower mortality rate, has fewer local complications,
and requires fewer treatment sessions to eradicate varices.19
Octreotide (50 g bolus and 50 g/h IV infusion for 2 to 5 days) improves
the control of acute bleeding when used in combination with endoscopic
therapy.20 Antibiotic therapy (eg, quinolones) is also recommended for
patients with cirrhosis with UGIB, because randomized controlled trials
demonstrate a significant decrease in bacterial infections and mortality.20
Long-term treatment with nonselective beta blockers is also recommended,
in combination with endoscopic ligation therapy, to decrease recurrent
bleeding from esophageal varices.20
Patients with persistent or recurrent bleeding despite endoscopic and medical therapy require more invasive therapy. Transjugular intrahepatic portosystemic shunt (TIPS) decreases recurrent bleeding more effectively than
endoscopic therapy, although hepatic encephalopathy is more common and
the mortality rates are comparable. Most patients with TIPS have shunt
stenosis within 1 to 2 years and require reintervention to maintain shunt
patency, although the use of coated stents appears to markedly decrease
shunt dysfunction, at least in the first year. A randomized comparison of
TIPS and distal splenorenal shunt in Child-Pugh class A or B cirrhosis with
www.asge.org www.askasge.org

refractory variceal bleeding revealed no significant difference in recurrent

bleeding, encephalopathy, or survival but a much higher rate of reintervention with TIPS (82% vs 11%).21 Thus, patients with milder, well-compensated cirrhosis should require fewer reinterventions with decompressive
surgery, although the higher risks of surgery at the time of the initial procedure must also be considered.
Patients with portal hypertension also may develop bleeding from gastric
varices, although this is less common than bleeding from esophageal varices.
Unfortunately, standard endoscopic therapy with ligation is relatively
ineffective at preventing further bleeding. Endoscopic therapy with tissue
adhesives is more effective, but this therapy is not approved for variceal treatment in the United States and is rarely used here. Therefore, in the United
States, patients with bleeding gastric varices typically are treated with TIPS.
The possibility of splenic vein thrombosis (most commonly from pancreatic
disease) should always be considered in patients with gastric varices but no
esophageal varices. If present, this can be cured with splenectomy. Portal
hypertensive gastropathy is commonly identified at endoscopy in patients
with cirrhosis but rarely causes major overt UGIB.

Hemorrhagic and erosive gastropathy

Hemorrhagic and erosive gastropathy, often incorrectly labeled as gastritis,
refers to subepithelial hemorrhages and erosions. These lesions are restricted
to the mucosa, where no large blood vessels are present and, therefore, do not
cause major bleeding.
The most common causes are NSAID use, alcohol intake, and stress. Half
of patients regularly taking traditional NSAIDs develop erosions, and up to
20% of patients who are alcoholic and are actively drinking and with symptoms of UGIB have subepithelial hemorrhages or erosions. The discontinuation of the offending agents and the use of antisecretory therapy are
recommended in patients with bleeding.
Stress-related injury occurs in extremely sick patients: eg, major trauma,
extensive burns, major intracranial disease, and severe medical illness (ie, ventilator dependence, coagulopathy). Major bleeding from stress injury probably requires progression of an erosion to an ulcer. Preventive therapy should
only be used in high-risk patients (eg, mechanical ventilation). Strong evidence supports the use of IV H2-receptor antagonist therapy. This therapy
is more effective than sucralfate but not superior to a PPI immediate-release
suspension via nasogastric tube.22,23

Other causes of UGIB

A great number of other lesions also may cause UGIB. These include erosive
esophagitis, neoplasms, vascular lesions (eg, hereditary hemorrhagic telangiectasias [Osler-Weber-Rendu syndrom], gastric antral vascular ectasia
[watermelon stomach], Dieulafoys lesion [in which an aberrant vessel in
the mucosa bleeds from a pinpoint mucosal defect]), prolapse gastropathy
(prolapse of the proximal stomach into the esophagus, with retching, especially
in patients with alcoholism), erosive duodenitis, aortoenteric fistulas, and
hemobilia and hemosuccus pancreaticus (bleeding into the duodenum from
the bile duct or pancreatic duct).
Vol. 14, No. 3 January 2007

DISCLOSURE STATEMENT
The author discloses receiving research support and/or consulting for Altana,
AstraZeneca, Eisai Inc., TAP Pharmaceuticals and Johnson & Johnson.

REFERENCES
1. Longstreth GF. Epidemiology of hospitalization for acute upper gastrointestinal
hemorrhage: a population-based study. Am J Gastroenterol 1995;90:206-10.
2. Rockall TA, Logan RF, Devlin HB, et al. Risk assessment after acute upper gastrointestinal haemorrhage. Gut 1996;38:316-21.
3. Rockall TA, Logan RFA, Devlin HB, et al. Incidence of and mortality from acute
upper gastrointestinal haemorrhage in the United Kingdom. BMJ
1995;311:222-6.
4. Schiff L, Stevens RJ, Shapiro N, et al. Observations on the oral administration
of citrated blood in man. II. The effect on the stools. Am J Med Sci
1942;203:409-12.
5. Hilsman, JH. The color of blood-containing feces following the instillation of
citrated blood at various levels of the small intestine. Gastroenterology
1950;15:131-4.
6. Ebert RV, Stead EA, Gibson JG. Response of normal subjects to acute blood
loss. Arch Intern Med 1941;68:578-90.
7. Ahmad A, Bruno JM, Boynton R, et al. Nasogastric aspirates frequently lead to
erroneous results and delay of therapy in patients with suspected UGI bleeding.
Gastrointest Endosc 2004;59:P163.
8. Cuellar RE, Gavaler JS, Alexander JA, et al. Gastrointestinal tract hemorrhage.
The value of a nasogastric aspirate. Arch Intern Med 1990;150:1381-4.
9. Spiegel BM, Vakil NB, Ofman JJ. Endoscopy for acute nonvariceal upper gastrointestinal tract hemorrhage: is sooner better? A systematic review. Arch Intern
Med 2001;161:1393-404.
10. Laine L, Peterson W. Bleeding peptic ulcers. N Engl J Med 1994;331:717-27.
11. Sung JJY, Chan FKL, Lau JYW , et al. The effect of endoscopic therapy in
patients receiving omeprazole for bleeding ulcers with nonbleeding visible vessels
or adherent clots. A randomized comparison. Ann Intern Med 2003;139:237-43.
12. Jensen DM, Kovacs TOG, Jutabha R, et al. Randomized, controlled trial of medical therapy compared to endoscopic therapy for prevention of recurrent ulcer
hemorrhage in patients with non-bleeding adherent clots. Gastroenterology
2002;123:407-13.

13. Cook DJ, Salena B, Guyatt GH, et al. Endoscopic therapy for acute non-variceal
upper gastrointestinal hemorrhage: a meta-analysis. Gastroenterology
1992;102:139-48.
14. Lau JYW , Sung JJ, Lee KK, et al. Effect of intravenous omeprazole on recurrent
bleeding after endoscopic treatment of bleeding peptic ulcers. N Engl J Med
2000;343:310-6.
15. Green FW , Kaplan MM, Curtis LE, et al. Effect of acid and pepsin on blood
coagulation and platelet aggregation. A possible contributor prolonged gastroduodenal mucosal hemorrhage. Gastroenterology 1978;74:38-43.
16. Javid G, Masoodi I, Zargar SA, et al. Omeprazole as adjuvant therapy to endoscopic combination injection sclerotherapy for treating bleeding peptic ulcer. Am
J Med 2001;111:280-4.
17. Kaviani MJ, Hashemi MR, Kazemifar AR, et al. Effect of oral omeprazole in
reducing re-bleeding in bleeding peptic ulcers: a prospective, double-blind, randomized, clinical trial. Aliment Pharmacol Ther 2003;17:211-6.
18. Chan FK, Wong VW , Suen BY, et al. Proton pump inhibitor plus a COX-2
inhibitor for the prevention of recurrent ulcer bleeding in patients with arthritis:
a double blinded, randomized trial. Gastroenterology 2006;130:A-105.
19. Laine L, Cook D. Endoscopic ligation compared with sclerotherapy for treatment of esophageal variceal bleeding: a meta-analysis. Ann Intern Med
1995;123:280-7.
20. DeFranchis R. Evolving consensus in portal hypertension. Report of the Baveno
IV consensus workshop on methodology of diagnosis and therapy in portal
hypertension. J Hepatol 2005;43:167-76.
21. Henderson JM, Boyer TD. Kutner MH, et al. Distal splenorenal shunt versus
transjugular intrahepatic portal systematic shunt for variceal bleeding: a randomized trial. Gastroenterology 2006;130:1643-51.
22. Cook DJ, Guyatt G, Marshall J, et al. A comparison of sucralfate and ranitidine
for the prevention of upper gastrointestinal bleeding in patients requiring
mechanical ventilation. N Engl J Med 1998;338:791-7.
23. Conrad SA, Gabrielli A, Margolis B, et al. Randomized, double-blind comparison of immediate-release omeprazole oral suspension versus intravenous cimetidine for the prevention of upper gastrointestinal bleeding in critically ill patients.
Crit Care Med 2005;33:760-5.

PRSTD STD
U.S. POSTAGE PAID
HANOVER, NH
PERMIT NO. 8

Elsevier Inc.
11830 Westline Industrial Drive
St. Louis, MO 63146-3318

www.asge.org www.askasge.org

CU-1403

Vol. 14, No. 3 January 2007