This article is more than 5 years old. Information may no longer be current.
Endoscopy Offers Reliable Detection of Varices
Q: Which Asymptomatic Patients With Cirrhosis Should Undergo an Upper Endoscopy to Determine Whether They Have Esophageal Varices?
A: In our practice, all patients with cirrhosis are offered endoscopy to assess for presence and size of varices. Patients with bridging fibrosis on biopsy may also merit screening endoscopy if they exhibit thrombocytopenia, splenomegaly, or other signs of portal hypertension. We do this for three key reasons: (1) esophageal variceal bleeding can occur at any point after development of portal hypertension and carries significant risk; (2) at the present time only endoscopy can reliably detect varices and predict the likelihood of variceal bleeding; and (3) available treatments can prevent variceal bleeding and reduce mortality in high-risk patients.
Preventing variceal hemorrhage is an important goal of management of patients with cirrhosis. Esophageal varices develop in up to two thirds of patients with cirrhosis, and 25% to 40% of all patients with varices will develop variceal hemorrhage. Each episode of active variceal hemorrhage is associated with at least a 20% mortality. Unlike other complications of cirrhosis, variceal bleeding is purely a complication of portal hypertension and can occur without warning in patients with otherwise well-compensated liver function. Also, unlike other complications of cirrhosis, variceal bleeding is abrupt and often catastrophic and can lead to sudden, unexpected out-of-hospital death.
Varices develop when increased pressure and flow in the portal vein cause expansion of collateral channels between the portal and systemic circulation. A critical threshold pressure gradient of about 12 mm Hg between these systems is required for varices to form. Intravariceal pressure, increased vessel size and thinning of the vessel wall all contribute to increasing variceal wall tension, according to LaPlace’s law:
Tension = Pressure × radius/wall thickness
Image: Toubia N, Habib A, Heuman DM
With increasing wall tension, areas of weakness develop in the vessel wall where superficial venules feed into the varix, forming endoscopically visible stigmata (red walls, hematocystic spots). Like blisters on a balloon tire, these areas of weakness can rupture abruptly, producing a “blowout.”
Numerous studies show that the risk of bleeding from esophageal varices can be predicted based on their endoscopic appearance. If varices are absent or small and stigmata are absent, risk of variceal hemorrhage is low; with increasing variceal size or stigmata, risk increases progressively. While assessment of variceal size is semiquantitative and somewhat subjective (Figure 1), clinical trials indicate that experienced endoscopists in the community setting can reliably distinguish between low-risk and high-risk varices based on their appearance, and this assessment is the single strongest predictor of bleeding risk. Other clinical parameters such as cirrhosis severity (Child-Turcotte-Pugh class), presence of tense ascites, and coagulopathy also affect bleeding risk but to a lesser degree.
In patients with high-risk varices, several treatments have been shown to reduce the risk of variceal hemorrhage and may prolong survival. The best established of these is use of nonselective beta-blockers: propranolol, nadolol, or timolol. These act mainly at the level of the mesenteric arterioles to block vasodilatory beta-1-adrenergic receptors. The resulting increase in mesenteric resistance causes a drop in pressure downstream in the portal vein. The response of portal pressure to beta-blockers is variable and is not clearly related to the systemic effects, so the usual practice is to push the dosage to the limits of tolerance, generally defined as a resting pulse less than 60. Isosorbide mononitrate may provide some synergy with beta-blockers. Band ligation of esophageal varices is also effective and has been advocated as primary prophylaxis by some groups. However, banding in a small percentage of cases may be complicated by massive delayed variceal hemorrhage, and fatalities from this complication have occurred. For this reason, we usually reserve banding for patients at high risk who cannot tolerate effective doses of beta-blockers or patients who have already experienced a variceal bleed, and current guidelines support this approach. Transjugular intrahepatic portosystemic shunting (TIPS) is very effective at lowering portal pressure and preventing variceal hemorrhage, but the cost and complication rate are too high to justify its use in primary prophylaxis.
Determination of hepatic venous pressure gradient (a measure of portal pressure in patients with cirrhosis) is favored by some groups as part of routine management of cirrhosis, but it is not universally practiced. The procedure is invasive, requiring jugular puncture with placement of a cannula in the hepatic vein to measure free and wedged pressures. Reliable interpretation of pressure tracings requires some experience and attention to details of technique. A gradient less than 12 mm Hg places the patient at low likelihood of variceal bleeding, but higher pressures correlate only weakly with variceal size and bleeding risk. The most important use of this test is to assess response of portal pressure to therapeutic interventions. Patients in whom hepatic venous pressure gradient falls by more than 20% or drops below 12 mm Hg with propranolol therapy have excellent results and low risk of bleeding, whereas patients in whom pressure fails to respond remain at risk and may require further prophylaxis with band ligation.
What about follow-up surveillance endoscopy? In asymptomatic cirrhotic patients without varices on initial evaluation, varices subsequently develop at an annual rate of 5% to 15%. Small varices progress to large varices at a rate of 4% to 10% each year. Prophylactic administration of beta-blockers to patients with small varices does not prevent progression to large varices and is associated with significant side effects. Our current practice is to repeat endoscopy at 2-year to 3-year intervals in cirrhotic patients who initially are free of varices, and at 1-year to 2-year intervals in patients with small varices. Large varices are most likely to be found in patients with clinical evidence of portal hypertension, including platelet count less than 100,000/mm3, elevated INR, or dilated portal vein greater than 13 mm in diameter on ultrasonography. In the absence of these indicators, screening endoscopy may not be cost-effective. The technique of wireless capsule endoscopy has recently been adapted for evaluation of esophageal varices and may eventually replace traditional endoscopy for this indication. Once large varices have been identified and prophylaxis has been initiated, further surveillance is not of proven value.
Excerpted from:
Shiffman ML, eds. Curbside Consultation of the Liver: 49 Clinical Questions, Second Edition (pp 183-186) ©2008 SLACK Incorporated.
- References:
- Chalasani N, Boyer TD. Am J Gastroenterol. 2005;100:805-807.
- D’Amico G, De FR. Hepatology. 2003;38:599-612.
- D’Amico G, et al. Gastroenterology. 2006;131:1611-1624.
- Eisen GM, et al. Endoscopy. 2006;38:31-35.
- Garcia-Tsao G, Wongcharatrawee S. Treatment of patients with portal hypertension and cirrhosis. Literature review and summary of treatment recommendations. Available at: http://www.hepatitis.va.gov/pdf/va01-pr/prtop-08/prtop08-01-gd-01.pdf.
- Groszmann RJ, et al. Hepatology. 2004;39:280-282.
- Groszmann RJ, et al. N Engl J Med. 2005;24:2254-2261.
- Merkel C, et al. Am J Gastroenterol. 2000;95:2915-2920.
- Poynard T, et al. N Engl J Med. 1991;324:1532-1538.
- Schepis F, et al. Hepatology. 2001;33:333-338.
- Spiegel BM, et al. Hepatology. 2003;37:366-377.
- The North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices. N Engl J Med. 1988;319:983-989.
- For more information on the Curbside Consultation in Gastroenterology Series:
- See Healio.com/Books/Gastroenterology.