I. What every physician needs to know.

Gastoesophageal varices occur in 30-40% of compensated cirrhotic patients and 60% of uncompensated cirrhotics. They cause 70% of upper gastrointestinal (GI) bleeding episodes in patients with liver cirrhosis. Mortality has improved to 15-20% for those that receive therapy.

Gastroesophageal varices form due to portal hypertension, usually when the hepatic venous pressure gradient is more than 10mmHg. The bleeding risk from varices is between 4-15% per year (depending on stage of cirrhosis) and approximately 30% during the course of the disease.

II. Diagnostic Confirmation: Are you sure your patient has gastroesophageal varices?

Esophagogastroduodenoscopy (EGD) is the gold standard for diagnosis of varices. Other modalities, such as ultrathin endoscopy, capsule endoscopy, endoscopic ultrasonography (EUS) and MRI may be of benefit to patients unable or unwilling to undergo EGD, but have little evidence supporting their capabilities in diagnosis of varices.

Continue Reading

A. History Part I: Pattern Recognition:

Patients who present with bleeding from gastroesophageal varices generally have a history of cirrhosis.

B. History Part 2: Prevalence:

Gastric varices are much less common than esophageal varices, occurring in 15-20% of cirrhotics. Gastric varices generally cause more severe bleeding but less often have bleeding episodes. They are classified as small (<10mm diameter), medium, or large (> 20mm diameter).

Esophageal varices are classified by size (small < 5mm, large > 5mm) and location (lower, mid or upper esophagus). There are several grading systems used for esophageal varices, and one of the most popular is the NIEC scale, which uses the Child-Pugh score, size, and findings on EGD (Table I).

Table 1

Secondary Prevention of variceal bleeding

C. History Part 3: Competing diagnoses that can mimic gastroesophageal varices.

Other common causes of Upper GI bleeding include peptic or duodenal ulcers, mucosal tears, erosive gastritis or esophagitis, gastric cancer, gastric leiomyoma, and vascular ectasias.

There are many possible causes for portal hypertension, including Hepatitis B or C cirrhosis, autoimmune hepatitis, hemochromatosis, Wilson’s disease, severe congestive heart failure, Budd-Chiari syndrome, portal or splenic vein thrombosis, primary biliary cirrhosis, primary sclerosing cholangitis, and sarcoidosis.

D. Physical Examination Findings.

Look for the stimata of chronic liver disease on physical exam, including spider nevi, jaundice, palmar erythema, Dupuytren’s contractures, testicular atrophy, loss of secondary sexual characteristics, splenomegaly, ascites, encephalopathy (asterixis), and caput medusae.

E. What diagnostic tests should be performed?

The most important test is an EGD, preferably as soon as possible in setting of suspected acute variceal hemorrhage.

1. What laboratory studies (if any) should be ordered to help establish the diagnosis? How should the results be interpreted?

Helpful labs include a hemoglobin/hematocrit, platelet count to assess if transfusions are necessary. Bilirubin, albumin and PT/INR may be helpful to classify the patient’s Child class and mortality risk.

2. What imaging studies (if any) should be ordered to help establish the diagnosis? How should the results be interpreted?

Imaging tests are generally not immediately necessary.

F. Over-utilized or “wasted” diagnostic tests associated with this diagnosis.

Factor VII levels are generally not helpful as Factor VII infusions to not improve morbidity or mortality in acute variceal hemorrhage.

III. Default Management.

Endoscopic variceal ligation (EVL) is the preferred therapy for most variceal bleeds.

The location of the varices on EGD help to guide the type of managament. Gastric variceal obturation (GVO) with a glue polymer has a high rate of hemostasis and is preferred for varices in the fundus of the stomach, whereas gastric variceal band ligation (GVL) or EVL is preferred for varices near or below the gastroesophageal junction.

ß-blockers should not be used in a setting of acute hemorrhage. However, ß-blockers do reduce mortality as secondary prophylaxis against gastroesophageal variceal hemorrhage.

Secondary prophylaxis

A reasonable starting doses for non-selective ß-blockers include propranolol 20mg orally twice daily or nadolol 40mg orally once daily. These drugs can then be titrated to maximum tolerated doses with the goal of a 25% reduction in heart rate from baseline (or decrease in hepatic venous pressure gradient to <12mmHg).

Although nitrates and spironolactone can further decrease portal pressures, the use of nitrates and/or spironolactone in combination with a non-selective ß-blocker is not recommended due to the high rate of side effects with these combinations.

Nitrates alone do not reduce hemorrhage from gastroesophageal varices and are not recommended as monotherapy even in those patients who cannot tolerate ß-blockers.

Table I. First line prevention of recurrent variceal hemorrhage.

Primary prophylaxis

Non-selective ß-blockers are recommended for small varices (<5mm) or medium/large varices without a high risk of bleeding (e.g. Child’s class A and no red wale markings). Non-selective ß-blockers or EVL are recommended for primary prevention of hemorrhage in medium/large varices at a high risk of bleeding (e.g. Child’s class B/C or red wale markings).

If a patient does not tolerate or is unable to reliably take a non-selective ß-blocker then EVL is a reasonable alternative even for small varices. Nitrates alone or in combination with ß-blockers, transjugular intrahepatic portosystemic shunt (TIPS), and sclerotherapy are not indicated for primary prevention of gastroesophageal varices hemorrhage.

A. Immediate management.

Consider the following management in persons with suspected hemorrhage from gastroesophageal varices:

  • Assess airway. Patient airway can become a problem with ongoing hemorrhage especially if hepatic encephalopathy is present. Consider intubation if patient is unable to protect airway or if intubation needed to protect airway during endoscopy.

  • Secure IV access (traditionally 2 large bore peripheral IVs).

  • Maintain perfusion with blood products and/or saline as needed, but avoid over resuscitation as this may increase portal pressures and increase risk of continued bleeding or early rebleeding. Goal hemoglobin should be about 8g/dL (Class 1B recommendation). Excess IV saline may contribute to accumulation of extravascular volume (e.g. ascites).

  • Check INR and platelet count. Fresh frozen plasma or platelets may be necessary to help stop hemorrhage in persons with significant coagulopathy or thrombocytopenia respectively. Recombinant factor VIIa is not routinely used.

  • Admit to intensive care unit (ICU) for monitoring and resuscitation.

  • Initiate octreotide 50 microgram IV bolus followed by 50 micrograms per hour continuous IV infusion to help decrease portal venous inflow through splanchnic vasodilitation. (Vasopressin is a more potent splanchnic vasoconstrictor, but it also causes significant systemic vasoconstriction as well with risk of unwanted systemic side effects.) Continue octreotide for 3-5 days if variceal hemorrhage confirmed on endoscopy (Class 1A recommendation).

  • Contact gastroenterology consultant for emergent endoscopy (usually within 12 hours of presentation) for diagnosis and endoscopic therapy (EVL).

  • Begin short-term prophylactic antibiotics such as norfloxacin 400mg orally twice a day, a similar oral or IV fluoroquinolone, or ceftriaxone 1g IV daily (especially in settings where there is significant gram-negative bacteria resistance to fluoroquinolones or in patients receiving prophylactic fluoroquinolones) to reduce bacterial infections and mortality in patients with cirrhosis and GI bleeding regardless of the presence or absence of ascites. Some of the reduction in mortality is due to decrease in rebleeding of gastroesophageal varices in patients who receive prophylactic antibiotics. Continue antibiotics for 5-7 days (Class 1A recommendation).

  • Proton pump inhibitors should be administered daily for at least 10 days.

Table II. First-line management of acute variceal bleed.

Table II

First line management of acute variceal bleed

Despite endoscopic and pharmacologic therapy, approximately 10-20% of patients with cirrhosis and variceal hemorrhage will continue to bleed or have early rebleeding. In these situations, options include TIPS, shunt surgery, or temporary balloon tamponade.

Temporary balloon tamponade can acutely stop bleeding in up to 80% of persons, but the complication rate (e.g. aspiration, esophageal rupture) is high with mortality of approximately 20%.

Balloon tamponade should be considered a bridge to more definitive therapy and should not be performed for more than 24 hours. TIPS and shunt surgery can also be risky and decisions to proceed with one of these two options will depend on local expertise and severity of patient underlying cirrhosis.

Table III. Management algorithm for presumed acute variceal bleed.

Table III

C. Laboratory Tests to Monitor Response To, and Adjustments in, Management.

Hemoglobin levels should be monitored closely in the setting of an acute gastroesophageal variceal bleed, and as there may be a lag time between the bleeding and severe anemia.

D. Long-term management.

If the patient was treated with EVL, he/she will need a repeat EGD in 1-4 weeks to assess for varices that require subsequent banding as there is a high rebleeding rate and repeat EVL has reduced the relative risk of rebleeding by 37%.

If the patient was treated with sclerotherapy, a repeat EGD should be performed in 6 months. If the patient is Child’s class A, a shunt should be considered. Beta blockers should also be administered (if tolerated) in patients along with periodic EVL (Class 1A recommendation).

E. Common Pitfalls and Side-Effects of Management

Risks of endoscopic treatment include ulcerations and perforation in all types of therapy, and venous thromboembolism in patients undergoing GVO.

A. Renal Insufficiency.

If acute renal insufficiency, these patients may require very large amounts of IV fluid resuscitation. If chronic renal insufficiency, the practitioner may choose fluids with higher osmotic pressure such as albumin or plasma.

B. Liver Insufficiency.

No change in standard management.

C. Systolic and Diastolic Heart Failure

If chronic systolic heart failure, the practitioner may choose fluids with higher osmotic pressure such as albumin or plasma. Octreotide therapy has a 10% risk of cardiac conduction abnormalities.

D. Coronary Artery Disease or Peripheral Vascular Disease

In patients on β-blockers for coronary artery disease, consider changing their selective β-blocker (e.g. atenolol, metoprolol) to a non-selective β-blocker. Octreotide therapy has a 10% risk of cardiac conduction abnormalities.

E. Diabetes or other Endocrine issues

May need to consider the possibility of worsening hypoglycemic awareness in patients with diabetes mellitus and hypoglycemia events who start non-selective β-blockers for gastroesophageal varices bleeding prophylaxis.

F. Malignancy

No change in standard management.

G. Immunosuppression (HIV, chronic steroids, etc).

No change in standard management.

H. Primary Lung Disease (COPD, Asthma, ILD)

May need to consider use of non-selective β-blocker for gastroesophageal varices bleeding prophylaxis in patients with reactive airway disease.

I. Gastrointestinal or Nutrition Issues

No change in standard managment.

J. Hematologic or Coagulation Issues

Consider more frequent monitoring of hemoglobin or platelet levels and different thresholds for transfusion, which should be discussed with a hematologist consultant.

K. Dementia or Psychiatric Illness/Treatment

No change in standard management.

V. Transitions of Care

A. Sign-out considerations While Hospitalized.

Follow up hemoglobin and platelet counts at regular intervals and transfuse if Hb is less than 8g/L or platelet is less than 50 or hemodynamic instability (especially systolic BP < 100 mmHg, Class 1B recommendation).

B. Anticipated Length of Stay.

Length of stay is typically at least 5 hospital days to receive octreotide infusion and close monitoring of laboratory values and vitals.

C. When is the Patient Ready for Discharge.

The patient may be discharged when bleeding has ceased, hemoglobin levels and vitals are stable, diet is tolerated, octreotide infusion has completed, and medications and follow-up has been arranged.

D. Arranging for Clinic Follow-up

A follow-up EGD should be arranged based on the type of endoscopic therapy received.

1. When should clinic follow up be arranged and with whom.

The patient should follow-up with a gastroenterologist for management of cirrhosis (and consideration for treatment if hepatitis C positive), and should also be considered for a liver transplant clinic follow-up (Level 1C recommendation).

2. What tests should be conducted prior to discharge to enable best clinic first visit.

If the patient has hepatitis C, the practitioner may consider a viral PCR and genotype prior to discharge as this lab may take several days to a week to return and may be helpful if considering treatment of hepatitis C.

3. What tests should be ordered as an outpatient prior to, or on the day of, the clinic visit.

A complete blood count (CBC), comprehensive metabolic panel, and protime/INR may be helpful to clinic physician for follow-up.

E. Placement Considerations.

Evaluation by physical therapy or occupation therapy prior to discharge may be warranted, especially if the patient had a prolonged course in the ICU.

F. Prognosis and Patient Counseling.

1-2 year mortality rate in patients with varices is 30% if patients are not treated with β-blockers and EVL.

New use of non-selective β-blockers may lead to fatigue, hypotension, and/or shortness of breath. Patients should be advised to contact their healthcare provider if they develop any of these symptoms.

VI. Patient Safety and Quality Measures

A. Core Indicator Standards and Documentation.


B. Appropriate Prophylaxis and Other Measures to Prevent Readmission.

No DVT prophylaxis indicated due to bleeding risk.

What's the evidence?

Cairdenas, A. “Management of acute variceal bleeding: emphasis on endoscopic therapy.”. Clin Liver Dis. vol. 14. 2010. pp. 251-62.

Coelho-Prabhu, N. “Current staging and diagnosis of gastroesophageal varices”. Clin Liver Dis. vol. 14. 2010. pp. 195-208.

D’Amico, M. “Refractory acute variceal bleeding; what to do next”. Clin Liver Dis. vol. 14. 2010. pp. 297-305.

Fontana, R. “Factors that determine the development and progression of gastroesophagel varices in patients with chronic hepatitis C”. Gastroenterology. vol. 138. 2010. pp. 2321-31.

Garcia-Tsao, G, Bosch, J. “Management of Varices and Variceal Hemorrhage in Cirrhosis.”. N Engl J Med.. vol. 362. 2010. pp. 823-32.

Garcia-Tsao, G, Sanyal, AJ, Grace, ND. “Prevention and management of gastroesophageal varices and variceal hemorrhage in cirrhosis”. Hepatology. vol. 46. 2007. pp. 922-38.

Lo, GH. “The role of endoscopy in secondary prophylaxis of esophageal varices”. Clin Liver Dis. vol. 14. 2010. pp. 307-23.

“Prediction of the first variceal hemorrhage in patient with cirrhosis of the liver and esophageal varices”. N Engl J Med. vol. 319. 1988. pp. 983-9.

Sarin, SK. “Endoscopic therapy for gastric varices”. Clin Liver Dis.. vol. 14. 2010. pp. 263-79.

Sass, DA. “Portal hypertension and variceal hemorrhage.”. Med Clin North Am.. vol. 93. 2009. pp. 837-53.

Jump to Section