Primary prophylaxis of esophageal variceal bleeding


  • Douglas A. Simonetto M.D.,

    1. Gastroenterology Research Unit, Advanced Liver Disease Study Group, Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN
    Search for more papers by this author
  • Vijay H. Shah M.D.

    Corresponding author
    1. Gastroenterology Research Unit, Advanced Liver Disease Study Group, Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.


Watch the interview with the author

Watch a video presentation of this article

Answer questions and earn CME




hepatic venous pressure gradient


North Italian Endoscopic Club


variceal band ligation

Portal hypertension is a common complication of cirrhosis and is the result of increased intrahepatic resistance and splanchnic and systemic hemodynamic changes. Elevated portal pressure leads to the formation of portosystemic collaterals, which comprise gastroesophageal varices (Fig. 1). The prevalence of esophageal varices in patients with compensated liver disease is approximately 30%, whereas their prevalence in decompensated patients is higher at 60%.[1] The incidence of new esophageal varices ranges from 5% to 10% per year in published series.[2-4] Once varices have developed, they tend to enlarge and may eventually bleed. The progression of variceal size occurs at a rate of 10% to 15% per year (related mostly to the severity of liver dysfunction),[5] and the estimated 2-year incidence of bleeding is approximately 24%.[6] The risk of death from a bleeding episode has declined considerably over the last 20 years, mostly as a result of new effective treatments. However, the reported mortality rate, ranging from 12% to 44%, is still appreciable.[7-9] Primary prevention of variceal hemorrhage remains an important and much-debated topic in the management of esophageal varices.[10]

Figure 1.

Schematic representation of the pathophysiology of portal hypertension in cirrhosis and the formation of esophageal varices.

Risk Stratification

Recommendations for primary prophylaxis for esophageal varices differ according to the associated risk of bleeding. The most important risk factor for variceal hemorrhage is size. Varices can be classified as small, medium, or large or simply as small or large (Fig. 2). The latter method of classification has been recommended most often, with a suggested cutoff diameter of 5 mm. The risk of bleeding from small varices has been reported to be 7%, whereas the risk from large varices increases to 30%.[6] The presence of “red wale” markings (i.e., longitudinal dilated venules that resemble whip marks) and the severity of liver disease have also been associated with an increased risk of bleeding. These parameters have been combined in the North Italian Endoscopic Club (NIEC) index.[11] The NIEC index has been validated in multiple independent studies; however, its prognostic yield is still limited (74% sensitivity and 64% specificity).[12]

Figure 2.

Endoscopic appearances of esophageal varices. (A) Upper gastrointestinal endoscopy demonstrates small varices in the lower esophagus (arrows). (B) Upper gastrointestinal endoscopy demonstrates large esophageal varices with a fibrin plug (arrow), representing the site of recent bleeding.

The hepatic venous pressure gradient (HVPG) is an indirect way of assessing portal pressure in patients with chronic liver disease, and it has been shown to be an independent predictor of variceal bleeding and death.[13] An HVPG > 12 mm Hg is associated with the development of esophageal varices and an increased risk of bleeding.[14]

Pharmacologic Prophylaxis

Nonselective beta-adrenergic blockers have been studied extensively in the prevention of variceal bleeding and are considered the cornerstone therapy for primary prophylaxis. A meta-analysis of 11 randomized controlled trials comparing beta-blockers with placebo revealed a 10% risk reduction of first variceal bleeding (from 24% in the placebo group to 15% with beta-blockers after 2 years).[6] The number of patients needed to treat to prevent one variceal bleed was 10. The mortality rate was also reduced from 28.4% in the control group to 23.9% in patients treated with beta-blockers (Fig. 3).

Figure 3.

Comparison of nonselective beta-blockers versus placebo in the prevention of first variceal bleeding and mortality according to a meta-analysis of 11 randomized controlled trials.[6] *P < 0.05

Nonselective beta-blockers decrease portal pressure via cardiac output reduction and unopposed splanchnic vasoconstriction, and this leads to diminished portal flow. In contrast, B1-selective agents lack the vasoconstrictive effect and therefore are not used for the prophylaxis of variceal bleeding. HVPG has been used to evaluate the hemodynamic response to beta-blockers with an overall response rate of approximately 49%[15] (HVPG reduction to <12 mm Hg or >20% from baseline). Because HVPG is not widely available, beta-blockers should be titrated to maximal tolerated doses. Propranolol and nadolol are the most widely used nonselective beta-blockers. The most common side effects include light-headedness, fatigue, and dyspnea. Only approximately 10% to 15% of patients require discontinuation of treatment due to side effects.[16] Starting beta-blockers at small doses and taking the medication at night may reduce side effects and improve compliance. In addition, approximately 15% of patients have contraindications to the use of beta-blockers, such as asthma, brittle diabetes mellitus, and atrioventricular block. More recently, carvedilol, a nonselective beta-blocker with mild anti–alpha-1 adrenergic activity, has been studied in primary prophylaxis of esophageal varices.[17] Through its alpha-1 blocking effect, carvedilol also decreases intrahepatic vascular resistance and results in a further decrease in portal pressure. Currently, carvedilol is most often being used in clinical trials ; however, its use could be considered in patients with concomitant systemic hypertension.[18]

The association of nitrates with beta-blockers is not recommended for primary prophylaxis of variceal bleeding.[19] Similarly, a trial using spironolactone in combination with nadolol showed no benefit over nadolol alone in primary prophylaxis.[20] Thus, nonselective beta-blockers are the only currently recommended pharmacologic treatment for primary prophylaxis of variceal bleeding.

Nonselective beta-blockers have been shown to be ineffective in preventing the development of varices in patients with cirrhosis; therefore, preprimary prophylaxis is not recommended.[21]

Endoscopic Prophylaxis

Variceal band ligation (VBL) has become the preferred method for endoscopic prevention of variceal bleeding because it has been associated with fewer side effects in comparison with endoscopic sclerotherapy. A meta-analysis of five randomized clinical trials comparing VBL with no treatment showed a decreased risk of first variceal bleeding and a lower mortality rate in the VBL group.[22] VBL has also been compared with nonselective beta-blockers in 19 trials, 12 of which have been published in full. A meta-analysis of these trials suggested that VBL is slightly superior to beta-blockers in reducing the risk of first variceal bleeding; however, this result was not confirmed when only high-quality trials were included (Fig. 4).[23] There was also no difference in bleeding-related and overall mortality in comparison with beta-blockers. Side effects from VBL are less frequent but more severe than those associated with beta-blockers and may lead to fatalities. Regarding its cost, a recent cost-effectiveness study comparing VBL with beta-blockers suggested that VBL is more cost-effective when quality of life is considered.[24]

Figure 4.

Comparison of nonselective beta-blockers versus VBL in the prevention of first variceal bleeding and mortality according to a meta-analysis of 19 randomized controlled trials.[23] *P < 0.05

Therefore, either nonselective beta-blockers or VBL may be an appropriate first-line choice for primary prophylaxis of variceal bleeding. The combination of both therapies has not been shown to be more effective and is associated with increased side effects.[25]


Currently, there are two main guidelines for the management of esophageal varices.[5, 26] Primary prophylaxis of variceal bleeding is recommended for all patients with medium or large esophageal varices independently of the severity of liver disease or the presence of risk factors. According to American Association for the Study of Liver Diseases guidelines, either nonselective beta-blockers or VBL could be used in patients with a high risk of hemorrhage (Child-Turcotte-Pugh class B or C or presence of red wales on varices; Fig. 5). In patients who are not at increased risk of bleeding, nonselective beta-blockers are preferred unless contraindicated.

Figure 5.

Schematic diagram of American Association for the Study of Liver Diseases recommendations for primary prophylaxis of esophageal variceal bleeding. High risk for variceal bleeding includes Child-Turcotte-Pugh class B and C or the presence of red wales on varices. Abbreviation: EGD, esophagogastroduodenoscopy.

Primary prophylaxis with beta-blockers for small varices is recommended only for patients with an increased risk of hemorrhage (Child-Turcotte-Pugh class B or C or presence of red wales on varices).