upper gastrointestinal hemorrhage

Introduction

Upper gastrointestinal (GI) hemorrhage refers to bleeding proximal to the jejunum (or ligament of Treitz). Mortality may be as high as 10%.

Etiology

• peptic ulcer* (34%)
  • duodenal ulcer
  • gastric ulcer
    • erosive gastritis (13%), NSAID use, H pylori infection
  • stress ulceration
  • anticoagulation
• esophageal varices*, gastric varices* (33%)
  • due to portal hypertension, cirrhosis
• erosive esophagitis (8%)
• Mallory-Weiss tear (6%)
  • hematemesis after heavy alcohol use, hematemesis after weightlifting, bulemia
• angiodysplasia (gastric & duodenal) (<6%)
• pyloric channel ulcer (2%)
• gastric antral vascular ectasia
• neoplasm (5%)
  • gastric neoplasm*
  • esophageal neoplasm*
  • gastrointestinal stromal tumor*
  • metastatic tumor to the duodenum
  • gastric polyps
• erosive duodenitis (2%)
• Dieulafoy's lesion* (1%)
• Cameron erosion
• anastomotic ulcer
• aortoenteric fistula* (patients with aortic graft surgery)
• hematobilia*
• Menetrier's disease
• Schonlein-Henoch purpura
• Osler-Weber-Rendu syndrome
• celiac disease
• hemosuccus pancreaticus
• idiopathic (13%)

* causes of severe upper GI bleed

Epidemiology

• more common than lower GI bleed

Clinical manifestations

• hematemesis
• coffee-ground emesis
• melena
• infrequently, bright-red-blood per rectum

Laboratory

• (see gastrointestinal hemorrhage)
• serum creatinine, serum urea nitrogen
  • BUN/creatinine ratio of > 30 suggests upper GI bleed
• H. pylori serology or other H. pylori testing if biopsy negative for H pylori

Diagnostic procedures

• Glasgow-Batchford score (range 0-23)
  • when score = 0, 100% negative predictive value for severe GI bleed & need for hospitalization
• upper GI endoscopy, esophagogastroduodenoscopy (EGD)
  • after hemodynamic stabilization
  • within 24 hours
  • within 12 hours if esophageal varices suspected
  • routine 2nd look not recommended
  • endoscopic coagulation or injection for
    • actively bleeding ulcers
    • visble vessels
    • sentinal clots
  • endoscopic ligation or sclerosis of esophageal varices
  • mechanical (clips) vs thermal hemostasis with similar outcomes
  • biopsy for Helicobacter pylori
• capsule endoscopy
  • repeat esophagogastroduodenoscopy with biopsies & colonoscopy or push enteroscopy indicated prior capsule endoscopy if initial studies were of low quality

Radiology

• CT angiography for hemodynamically unstable patients with small bowel bleeding
• technetium labeled nuclear scan (scintigraphy) provides the best sensitivity for actively bleeding lesion not detected by upper or lower GI endoscopy or capsule endoscopy
• arteriography
  • failure of EGD to visualize source of bleeding
  • bleeding of > 0.5 mL/min at the time of study
  • arterial angiotherapy
    • vasopressin 0.15-2.0 units/min in selectively catheterized bleeding artery
    • arterial embolization
      • selective arterial catheterization
      • absorbable gelatin powder (Gelfoam)
• upper GI with barium has no role in initial evaluation of active upper GI bleed

Complications

• higher mortality than lower GI bleed
  • mortality (up to 14%)
• tachycardia (> 100/min), hypotension (systolic BP < 100 mm Hg), age > 60 years, comorbidities associated with increased risk for rebleeding & increased mortality

Management

• general recommendations:
  • 2 large caliber IV catheters
  • IV crystalloids: target heart rate < 100/min, systolic BP > 100 mm Hg
  • transfuse patients with hemoglobin <7 g/dL in the absence of shock or symptomatic anemia; < 8 g/dL;
    • target 7-9 g/dL
  • perform endoscopy early (within 24 hours) for most patients
    • stabilize patients hemodynamically with fluids & erythrocytes prior to endoscopy
    • endoscopy within 12 hours if esophageal varices suspected
    • endoscopy within 6 hours no better thqn later endoscopy
  • do not delay endoscopy to correct coagulopathies or to administer proton-pump inhibitor (PPI) therapy to downsize lesions
  • if endoscopic treatment fails, percutaneous embolization can be considered as an alternative to surgery
  • bolus- & continuous-infusion PPI for high-risk patients for 3 days after successful endoscopy for high-risk peptic ulcer or adherent clots
    • proton pump inhibitor of no benefit for acute peptic ulcer bleeding or other cause of acute upper GI bleed before or after endoscopy
    • H2 receptor blockers are not beneficial
  • hospitalize high-risk patients at least 72 hours
    • Glasgow-Blatchford score <= 1 predicts low-risk patients
  • discharge low-risk patients soon after endoscopy
    • see peptic ulcer to "low risk" criteria
    • no benefit of observing for 24 hours in hospital
  • prescribe daily proton pump inhibitor at discharge
  • in patients with peptic ulcers, test for, treat, & eradicate Helicobacter pylori
  • nonspecific NSAIDs plus a proton pump inhibitor or a COX2 inhibitor alone associated with bleeding risk; COX-2 inhibitor plus a PPI recommended if NSAID required
  • restart low-dose aspirin for cardiovascular prophylaxis as soon as benefit outweighs bleeding risk
    • restarting low-dose aspirin 3-5 days after upper GI bleed reduces 30-day mortality 10-fold in patients with cardiovascular disease, while increasing rebleeding rates only 2-fold
    • clopidogrel confers higher risk aspirin plus a proton pump inhibitor
  • restart anticoagulation after resolution of bleeding
  • use prognostic scales to assess risk for rebleeding & death
• general measures (see gastrointestinal hemorrhage)
  • do not insert nasogastric tube for aspiration to confirm diagnosis
  • also see algorithm for management of GI bleed
• Minnesota Esophagogastric Tamponade Tube for acute bleeding refractory to other measures (see esophageal varices)
• esophagogastroduodenoscopy (EGD) see above
  • stabilize patients hemodynamically with fluids & erythrocytes prior to endoscopy
  • IV bolus of proton pump inhibitor before endoscopy, followed by infusion after EGD [2, 5]
  • octreotide 50 ug IV bolus, then 25-50 ug/hour IV drip (superior to vasopressin)
    • for use in conjunction with EGD
    • octreotide alone does not improve outcome of non- esophageal varices upper GI bleed
  • H2-receptor antagonists are NOT effective in stopping active UGI bleed
• therapy for specific lesions
  • peptic ulcer:
    • continue proton pump inhibitor only if peptic ulcer is the source of bleeding
  • esophageal varices
  • Mallory-Weiss tear
  • aortoenteric fistula
  • angiodysplasia
• empiric prophylactic therapy with broad-spectrum antibiotics is indicated in patients with cirrhosis
• prognosis
  • patients taking NSAIDs at the time of hemorrhage
    • switching to COX-2 inhibitor or adding omeprazole to NSAID results in similar rebleed rate = 5%
• prophylaxis
  • proton pump inhibitor reduces risk of GI bleed associated with anti-platelet agent

More general terms

• gastrointestinal hemorrhage

More specific terms

• Dieulafoy's lesion
• Esophageal Hemorrhage
• hematemesis
• hemosuccus pancreaticus
• Mallory-Weiss tear
• small intestinal hemorrhage

Additional terms

• aortoenteric fistula; aortoduodenal fistula
• comorbid conditions that increase mortality of upper GI bleed
• esophageal varices
• melena
• Minnesota esophagogastric tamponade tube
• octreotide (Sandostatin)
• peptic ulcer disease (PUD)
• stress ulceration
• upper gastrointestinal (GI) endoscopy; esophagogastroduodenoscopy (EGD)