In patients with cirrhosis, the presence of esophageal varices on upper endoscopy may be an indication of the development of portal hypertension. Because variceal bleeding is a major cause of mortality in cirrhotic patients, prevention of such bleeding is a high priority of management. The risk of bleeding can be assessed by measuring the hepatic vein pressure gradient (HVPG) through catheterization of the hepatic vein using a transfemoral or transjugular entry. The calculation of HVPG involves subtracting the free hepatic vein pressure from the wedged hepatic venous pressure. An HVPG greater than 5 mm Hg demonstrates portal hypertension, and values greater than 10 mm Hg indicate clinically significant portal hypertension. De Franchis reviewed the management of incidental esophageal varices and the risk of rupture with increasing wall tension.
Because HVPG is not routinely available, assessment of esophageal size on esophagoscopy is the best available measure of risk. Varices eventually develop in the majority of patients with cirrhosis and tend to increase in size and to bleed. The mortality rate of first bleeds is approximately 20 percent.
Screening for esophageal varices may be worthwhile because useful prophylactic treatments exist. Unfortunately, good predictors of varices do not exist, indicating that all cirrhotic patients should have a screening endoscopy. Surveillance endoscopies are probably appropriate in patients who have no varices (at three-year intervals) or small varices (at one- to two-year intervals). Persons at highest risk of rapid variceal growth, including those with alcoholic cirrhosis or more severe liver impairment, and those with endoscopic risk signs, should be re-evaluated at shorter intervals. Prognostic indexes on postendoscopic patients who are at highest risk include the North Italian Endoscopic Club Index, which is based on the severity of liver disease (modified Child Class), the size of the varices, and the presence of red wale markings on the varices. Unfortunately, even the best risk score protocols miss a significant number of patients who eventually bleed.
|Beta blockers: significantly reduce bleeding-related mortality; most useful when HVPG falls by 20 percent from baseline or to a level < 12 mm Hg; usefulness limited by side effects|
|Nitrovasodilators: decrease hepatic resistance; lower efficacy than beta blockers|
|Sclerotherapy: not recommended because of variability of study results|
|Band ligation: significantly decreases incidence of first bleed and mortality; possibly more effective than beta blockers|
|Shunt surgery: abandoned because of high mortality|
|Combination medical treatment: beta blockers and nitrates may provide increased efficacy, but side effects and cost-effectiveness require further study|
|Combination medical and endoscopic therapy: inadequate information about efficacy and safety|
Prophylactic therapies include drug treatment, surgical or radiologic shunt treatments, and endoscopic variceal obliteration techniques (see accompanying table). Primary prophylaxis appears to have good results. Although the exact timing for initiation of prophylaxis is unclear, the risk of bleeding in patients with medium to large varices should be reduced. Universal prophylaxis in all cirrhotic patients with beta blockers has been recommended as the most cost-effective therapy.
The author concludes that patients with medium or large esophageal varices noted on screening endoscopy should be treated prophylactically with beta blockers. If medical therapy is contraindicated or not tolerated, the patient should undergo band ligation. The clinical value of monitoring the HVPG requires further study.