Click a topic below for an index of articles:

New-Material

WGO-OMGE Practice Guideline:
Treatment of Esophageal Varices

May 2003

http://omge.org/globalguidelines/guide08/guideline8.htm

Dr. E Renner MD 

Esophageal varices are portosystemic collaterals, i.e. vascular channels that link the portal venous and the systemic venous circulation. They form with portal hypertension preferentially in the submucosa of the lower esophagus. Rupture and bleeding from esophageal varices is a major complication of portal hypertension and carries a high mortality.

Variceal bleeding accounts for 10-30% of all cases of upper GI-bleeding.

According to Ohm's law, portal venous pressure (P) is the product of vascular resistance (R) and blood flow (Q) in the portal bed (P = Q x R). Cirrhosis is the most common cause of portal hypertension. In cirrhosis, both, intrahepatic vascular resistance and portal flow are increased. The former due to a deranged hepatic (vascular) architecture, the latter due to the hyperdynamic circulation typically associated with cirrhosis which increases arterial blood inflow into the mesenteric arterial and hence in the portal venous bed. The hyperdynamic circulation of cirrhosis results from an imbalance between vasoconstrictor and vasodilator (e.g. nitric oxide) mechanisms leading to decreased resistance (i.e. net vasodilation) in the splanchnic and systemic circulation. As a consequence, the adrenergic (increased cardiac index) and the renin-angiotensin systems (renal Na+ and water retention) are activated as counterregulatory mechanisms.

Although varices form most often in the distal few cm of the esophagus and less often in the gastric fundus, they may form in any location along the tubular GI-tract.

A pressure difference between portal and systemic circulation (hepatic venous pressure gradient, HVPG) of 12mm Hg is necessary, but not sufficient for varices to form.

Varices rupture if the wall tension becomes too large. According to Laplace's law a vessel's wall tension is proportionate to the pressure within the vessel and to the fourth power of its radius. Thus, the likelihood of a varix to rupture and bleed increases with increasing varix size/diameter and with increasing variceal pressure which is again proportionate to the HVPG. Conversely, varices do not bleed, if HVPG is below 12 mmHg.

Approximately 30% of cirrhotics have esophageal varices at the time of diagnosis; this proportion increases with time and reaches 90% after approximately ten years.

Recently, an International Normalized Ratio INR) >1.5, a portal vein diameter of >13mm and thrombopenia have been found to predict the likelihood of varices being present in cirrhotics. Thus, <10%, 20-50%, 40-60%, >90% of patients were estimated to have varices if none, one, two or all three of these conditions were met, respectively. This may impact on the indication for gastroscopy in searching for varices with respect to primary prophylaxis of bleeding in cirrhotics.

About 30% of patients with esophageal varices will bleed within the first year of diagnosis. Despite all advances in intensive care, bleeding episodes still carry a high mortality which mainly depends on the severity of the underlying liver disease.

The mortality of any bleeding episode may range from <10% in well compensated Child-Pugh A to >70% in advanced Child-Pugh C cirrhotics, respectively.
Once a patient has bled, the risk of re-bleeding is high, reaching 80% within one year.

Clinical risk factors for an initial bleeding episode include poor liver function and continuing alcohol consumption. Endoscopic predictors of bleeding include the size of the varices and the presence of red color signs or red wale markings corresponding to areas of thinning of the varix wall due to high wall tension. Stopping beta-blockers acutely could also be considered a risk factor for major bleeding.

The differential diagnosis includes all etiologies of (upper) GI-bleeding. It is especially noteworthy that peptic ulcers are also more frequent in cirrhotics. Thus, diagnosis requires endoscopy.

Three different clinical situations have to be distinguished:

1. treatment of acute variceal hemorrhage
2. prevention of a first variceal bleed (primary prophylaxis)
3. prevention of re-bleeding after an initial bleeding episode (secondary prophylaxis)

5.1. Treatment of acute variceal hemorrhage

Initial measures are directed at general resuscitation, i.e. at securing the airway and at stabilization of the circulation.

Pharmacotherapy e.g. with octreotide, a synthetic somatostatin analogue, or terlipressin, a synthetic vasopressin analogue, has been shown to be effective in stopping hemorrhage, at least temporarily, in up to 80% of patients. Somatostatin may be superior to octeotride. It inhibits the release of vasodilator hormones, causing splanchnic vasoconstriction and decreased portal inflow. Both somatostatin and octeotride are safe medications with very few side effects.

Endoscopic sclerotherapy and variceal ligation are effective in stopping bleeding in up to 90% of patients. Endoscopic band ligation is as effective as sclerotherapy but prone to fewer side effects. In severe active bleeding, endoscopic band ligation may however be more difficult to apply than sclerotherapy.

Combining octreotide (or terlipressin) for 3-5 days with endoscopic therapy reduces re-bleeding and the requirement of blood tranfusions in the early phase after an index bleed compared to either measure alone.

The use of balloon tamponade is decreasing because of a high risk of rebleeding following deflation and a risk of major complications. A transjugular intrahepatic portosystemic shunt (TIPS) - if available - is now a good alternative when endoscopic treatment and pharmacotherapy fail. Nevertheless, in most cases balloon tamponade is effective in stopping hemorrhage at least temporarily and it can be used in regions of the world where TIPS is not readily available. It can help to stabilize the patient in order to gain time and access to TIPS later.

With the exception of endoscopic therapy, none of these measures, whilst being effective in stopping bleeding, have been shown to affect mortality.

Acute variceal hemorrhage is often associated with bacterial infection due to gut translocation and motility disturbances. Prophylactic antibiotic therapy has been proven to increase survival. In acute or massive variceal bleeding tracheal incubation can be of great help with avoiding blood bronchyo aspiration.

Thus, in summary, the following scheme is recommended if variceal hemorrhage is suspected:

1. general resuscitation measures
2. start pharmcotherapy, e.g. octreotide 50 ug iv bolus followed by 50ug/h iv for 3-5 days (or terlipressin)
3. administer antibiotics, e.g. chinolone or 3rd generation cepaholsporine
4. emergency endoscopy to verify diagnosis and to perform band ligation or sclerotherapy
5. in case of early (within 5 days of index-bleed) re-bleeding: repeat endoscopic therapy once, if possible
6. recurrent or uncontrolled bleeding or endoscopic treatment failure (early re-bleeding after two endsocopic attempts): place balloon tamponade, consider TIPS

5.2. Prevention of a first variceal bleed (primary prophylaxis)

Pharmacotherapy is aimed at reducing HVPG and thus collateral blood flow and pressure through/in the varices (if HVPG is £12mm Hg, varices will not bleed).

Non-cardioselective beta-blockers reduce cardiac output, splanchnic arterial inflow, hence portal venous flow and pressure and thus, variceal flow and pressure. In cirrhotics with esophageal varices, both, propranolol and nadolol have been shown to reduce the risk of an initial bleeding episode by 40-50%, while there was a trend only of reducing mortality. About 30% of patients will not respond to beta-blockers with a reduction in HVPG, despite adequate dosing. These non-responders can only be detected by invasive measurements of HPVG. Moreover, beta-blockers may cause side effects such as fatigue and impotence which may impair compliance especially in younger males, or they may be contraindicated for other reasons.

Isosorbide 5-mononitrate has been shown to lower portal pressure. The mechanism of action of nitrates in portal hypertension is not fully clarified, but may involve vasodilatation in the portal venous bed. Unfortunately, the use of nitrates in cirrhotics is limited by their systemic vasodilatory effects often leading to a further decrease in blood pressure and potentially in (prerenal) impairment of kidney function. Thus, nitrates alone are not recommended.

Combining isosorbide 5-mononitrate with non-cardioselective beta-blockers has been shown to have additive effects in lowering portal pressure and especially to be effective also in patients primarily not responding to beta-blockers alone. However, these beneficial effects may be outweighed by detrimental effects on kidney function and long-term mortality, especially in the those >50 years of age. Thus routine combination is not recommended.

While the side effects of endoscopic sclerotherapy outweigh its benefit in primary prophylaxis of esophageal variceal hemorrhage, endoscopic band ligation has recently been shown to be effective and well tolerated.

As with any primary prophylactic measures, overall cost-effectiveness strongly depends on the risk of the event to be prevented in the population prophylactically treated. Hence, primary prophylaxis of variceal bleeding is often performed in patients at a medium to high risk of bleeding only, their selection being based on endoscopic findings.

In summary therefore, the following scheme is recommended for primary prophylaxis of variceal hemorrhage:

1. selection of patients with at least medium sized esophageal varices and/or red color signs
2. non-cardioselective beta-blockers (propranolol or nadolol) starting at low dose, if necessary increasing dose step by step until reaching a reduction of resting heart rate by 25%, but not lower than 55/min
3. endoscopic band ligation is indicated in patients, who do not tolerate or have contraindications to beta-blockers

5.3. Prevention of re-bleeding after an initial bleeding episode (secondary prophylaxis)

Endoscopic eradication of varices after an index bleed is highly effective in reducing the risk of recurrent bleeding, the annual incidence decreasing from approx. 80% to 20-30%.

As mentioned earlier, variceal band ligation has a better efficacy/side-effects ratio than sclerotherapy and is the endoscopic treatment of choice for eradication of esophageal varices. It has to be emphasized that endoscopic eradication of esophageal varices requires several sessions (usually around 3), that varices may recur (seemingly at a higher rate after band ligation than after sclerotherapy), and that, therefore, regular endoscopic controls are mandatory longterm in order to detect and to enable eradication of recurrent varices prior to re-bleeding.

As in primary prophylaxis, pharmacotherapy is aimed at reducing HVPG and thus collateral blood flow and pressure through/in the varices. Reducing portal pressure by 20% of baseline, reduces the risk of rebleeding from over 60% to less than 10% at three years; if HVPG is lowered below 12mm Hg, varices will not bleed.

Non-selective beta-blockers such as propranolol and nadolol also block the adrenergic dilatory tone in the mesenteric arterioles resulting in unopposed alpha adrenergic mediated vasoconstriction and hence a decrease in portal inflow. In cirrhotics with esophageal varices, both, propranolol and nadolol, have been shown to reduce the risk of rebleeding by approx. 50% and to reduce mortality. About 30% of patients, however, will not respond to beta-blockers with a reduction in HVPG, despite adequate dosing. These non-responders can only be detected by invasive measurements of HPVG. Moreover, beta-blockers may cause side effects such as fatigue and impotence which may impair compliance especially in younger males, or they may be contraindicated for other reasons.

Isosorbide 5-mononitrate has been shown to lower portal pressure. The mechanism of action of nitrates in portal hypertension is not fully clarified, but may involve vasodilatation in the portal venous bed. Unfortunately, the use of nitrates in cirrhotics is limited by their systemic vasodilatory effects often leading to a further decrease in blood pressure and potentially in (prerenal) impairment of kidney function. Thus, nitrates alone are not recommended.

Combining isosorbide 5-mononitrate with non-cardioselective beta-blockers has been shown to have additive effects on lowering portal pressure and especially to be effective also in patients primarily not responding to beta-blockers alone. The combination of beta-blockers and isosorbide 5-mononitrate has been shown to more effectively decrease the risk of re-bleeding than sclerotherapy and even variceal ligation. This combination of beta-blockers and isosorbide results in side effects in 20% of patients.

Pharmacologic therapy seems, however, to be largely effective in Child Pugh A and B patients. As outlined above, there remain concerns about the potential longterm detrimental effects of the beta-blocker/nitrate combination, especially on renal function and mortality, in particular in the elderly and in advanced cirrhosis.

While it seems logical to combine pharmacological and endoscopic strategies for secondary prophylaxis of bleeding, at least until varices have been completely eradicated by endoscopy, this combination remains controversial.

The transjugular intrahepatic portosystemic shunt (TIPS) very effectively decreases portal pressure and hence the risk of re-bleeding. It is more effective than pharmacotherapy or endoscopic eradication of varices in preventing re-bleeding. However, a re-intervention for shunt dysfunction is necessary in about 50% of patients within a year and hepatic encephalopathy - albeit usually of mild to moderate grade and easily controllable by lactulose - is triggered or aggravated in at least 30% of patients. Thus, regular clinical and TIPS controls (Duplex sonography) are required. TIPS seems a good mid term option for prevention of re-bleeding, e.g. for bridging the waiting time until liver transplantation, whereas in the longterm, other options may be more cost-effective.

Surgical shunts (in particular the calibrated H graft but also the distal splenorenal shunt according to Warren) have a place in secondary prophylaxis, especially in patients presenting with portal hypertensive bleeding as their main clinical problem, a well preserved liver function, stable (or no) liver disease and only a small likelihood of requiring liver transplantation within the next 5-10 years. In these highly selected patients, a surgical shunt may be the best long term solution to the problem.

Patients who survived an index bleed from varices and are Child Pugh B or C should be considered for liver transplantation.

In summary therefore, the following scheme is recommended for secondary prophylaxis of esophageal variceal hemorrhage:

1. selection of patients with an index bleed that was stabilized
2. eradication of esophageal varices by endoscopic band ligation (every 7-14 days, until varices are eradicated); longterm endoscopic control and banding of recurrent varices every 3-6 months
3. if endoscopic band ligation is not available or contraindicated, non-cardioselective beta-blockers (propranolol or nadolol) starting at low dose, if necessary increasing dose step by step until reaching a reduction of resting heart rate by 25%, but not lower than 55/min; in younger patients with less advanced cirrhosis (Child Pugh A) addition of isosorbide 5-mononitrate (starting with 2x20 mg per day and increasing to 2x40 mg per day) may be considered
4. if sclerotherapy or pharmacotherapy failed, consider TIPS, especially in candidates for liver transplantation; in selected cases (well preserved liver function, stable liver disease) a calibrated H graft or a distal splenorenal shunt according to Warren may be considered
5. Always consider liver transplantation if patient is Child-Pugh B or C

As outlined above several therapeutic options are effective in most clinical situations of acute variceal hemorrhage, as well as in its secondary and primary prophylaxis. The optimal therapy in an individual setting very much depends on the relative ease of local availability of these methods/techniques. This is likely to vary widely in different parts of the world.

If endoscopy is not readily available, one has to resort to pharmacotherapy in any suspected variceal bleeding, i.e. in patients with hematemesis and signs of cirrhosis. Similarly, pharmacological therapy might be applied under such circumstances as primary prophylaxis in a cirrhotic with signs of portal hypertension (splenomegaly, thrombocytopenia) and/or impaired liver function and as secondary prophylaxis in a cirrhotic with a history of upper GI bleeding.

If pharmacotherapy is also not available and variceal bleeding is suspected, one must resort to general resuscitation measures and transport the patient as soon a possible to an institution where the necessary diagnostic/therapeutic means are available; balloon tamponade could be of great help in such a situation.

1. Gastroesophageal variceal hemorrhage Sharara AI and Rocky DC N Engl J Med 345: 669-681, 2001.  
2. UK guidelines on the management of variceal hemorrhage in cirrhotic patients. Jalan R and Hayes PC. Gut 46: (suppl III) 1-15, 2000  
3. Developing consensus in portal hypertension. De Franchis R. J Hepatol 25 : 390-394, 1996 [Editorial].  
4. Prediction of variceal hemorrhage in cirrhosis : a prospective folow-up study Kleber G, Sauerbruch T, Ansari H, Paumgartner G. Gastroenterology 100: 1332-1337, 1991.  
5. Prediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices - A prospective multicenter trial. The North Italian Endoscopic Club for the Study and Treatment of Esophagela Varices. New Engl. J Med 319: 983-989, 1988.  
6. Which patients with cirrhosis should undergo endoscopic screening for esophageal varices detection? Schepis F, Camma C, Niceforo D, Magnano A, Pallio S, Cinquegrani M, D'amico G, Pasta L, Craxi A, Saitta A, Raimondo G. Hepatology 33: 333-338, 2001.  
7. Treatment of oesophageal varices : a meta-analysis. De Franchis R. Scand J Gastroenterol 29 (suppl 207): 29-33, 1994.  
8. Endoscopic ligation compared with sclerotherapy for treatment of esophageal variceal bleeding - a meta-analysis. Laine L, Cook D. Ann Int Med 123: 280-287, 1995.  
9. Octreotide for acute esophageal variceal bleeding: a meta-analysis. Corley AC, Cello JP, Adkinson W, Ko WF, Kerlikowske K. Gastroenterology 120: 946-954, 2001.  
10. Endoscopic treatment versus endoscopic plus pharmacologic treatment for acute variceal bleeding: a meta-analysis. Banares R, Albillos A, Rincon D, Alonso S, Gonzalez M, Rtuiz-del-Arbol L, Slcedo M, Molinero L-M. Hepatology 35: 609-615, 2002.  
11. Early administration of terlipression plus glyceryl trinitrate to control active upper gastrointestinal bleeding in cirrhotic patients. Levacher S, Letoumelin P, Pateron D, Blaise M, Lapandry C, Pourriat J-L. Lancet 246: 865-868, 1995.  
12. Antibiotic prophylaxis for the prevention of bacterial infections in cirrhotic patients with gastrointestinal bleeding: a meta-analysis. Bernard B, Grange JD, Khac EN, Amiot X, Opolon P, Poynard T Hepatology. 1999 Jun;29(6):1655-61.  
13. Beta-adrenergic-antagonist drugs in the prevention of gastrointestinal bleeding in patients with cirrhosis and esophageal varices. Poynard T, Cales P, Pasta L, Ideo G, Pascal J-P, Pagliaro L, Lebrec D and the Franco-Italian Multicenter Study Group New Engl J Med 324: 1532-1538, 1991.  
14. Prevention of first bleeding in cirrhosis - a meta-analysis of randomized trials on nonsurgical treatment. Pagliaro L, D'Amico G, Thorkild I, Sörensen A, Lebrec D, Burroughs A, Morabito A, Tine F, Politi F, Traina M. Ann In Med 117: 59-70, 1991.  
15. A meta-analysis of endoscopic variceal ligation for primary prophylaxis of esophageal variceal bleeding. Imperiale TF, Chalasani N. Hepatology 33: 802-807, 2001.  
16. Endoscopic ligation compared with combined treatment with nadolol and isosorbide mononitrate to prevent recurrent variceal bleeding. Villanueva C, Ninana J, Ortiz J et al. N Engl J Med 345: 647-655, 2001.  
17. Beta-adrenergic antagonists in the prevention of gastrointestinal rebleeding in patients with cirrhosis: a meta-analysis. Bernard B, Lebrec D, Mathurin P, Opolon P, Poynard T. Hepatology 25: 63-70, 1997.  
18. Propranolol and sclerotherapy in the prevention of gastrointestinal rebleeding in patients with cirrhosis: a meta-analysis. Bernard B, Lebrec D, Mathurin P, Opolon P, Poynard T. J Hepatol 26: 312-324, 1997.  
19. Randomised trial of transjugular-intrahepatic-portosystemic shunt versus endoscopy plus propranolol for prevention of variceal rebleeding. Rössle M, Deibert P, Haag K, Ochs A, Olschewski M, Siegerstetter V, Hauenstein K-H, Geiger R, Stiepak C, Keller W, Blum HE. Lancet 349: 1043-1049, 1997.  
20. Distal spleno-renal shunt versus endoscopic sclerotherapy in the prevention of variceal rebleeding - a meta-analysis of 4 randomized clinical trials. Spina GP, Henderson JM, Rikkers LF, Teres J, Buroughs AK, Conn HO, Pagliaro L, Santambrogio R. J Hepatol 16: 338-345, 1992.  

1. American College of Gastroenterology: http://www.acg.gi.org
2. The American Gastroenterological Association: http://www.gastro.org/
3. The National Library of Medicine's PUBMED Medline is the best starting point for published research on Esophageal Varices: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi
4. MedlinePlus: The best source for consumer and patient information on Esophageal varices: http://www.nlm.nih.gov/medlineplus/
5. The US National Guidelines Clearing House is a good guideline source for Esophageal Varices: http://www.guideline.gov