Outlook Not Bright for Patients With Bleeding Oesophageal Varices[Drug & Ther Perspect 14(8):10-13, 1999. © 1999 Adis International Limited]
Bleeding from oesophageal varices secondary to portal hypertension is a medical emergency with a high mortality rate. Although optimal medical therapy in the emergency setting can improve the outlook, survival is often determined by the severity of the underlying liver disease. Survival rates can only be improved by:
- accurate diagnosis
- early resuscitation effective therapy to arrest haemorrhage
- follow-up therapy to prevent rebleeding.
Treatment to arrest haemorrhage often involves vasoactive drug therapy to reduce portal hypertension and stop bleeding (see Differential features table ) in combination with an endoscopic procedure (see table 1). However, the risk of rebleeding after the index bleed is high and eradication of the varices in conjunction with long-term propranolol therapy to reduce portal hypertension is recommended. In addition, propranolol prophylaxis may be considered in patients with oesophageal varices who are at high risk of bleeding as propranolol appears to be well tolerated, effective and inexpensive in the prevention of variceal haemorrhage.
Varices a Complication of Cirrhosis…
Oesophageal varices secondary to portal hypertension is one of the complications of alcoholic liver cirrhosis. In patients with portal hypertension, blood from the gastrointestinal tract can reach the systemic circulation by bypassing the liver through veins around the stomach. Varices form when these veins dilate.
In patients with cirrhosis, 30 and 60% of those with compensated and decompensated liver disease, respectively, will have varices.  Once varices occur, they tend to progressively increase in size, and the risk of bleeding has been shown to be related to varix size and appearance as well as the severity of the underlying liver disease.  About 30% of patients with varices will experience a bleed. 
…And a Leading Cause of Death
Variceal haemorrhage remains a leading cause of death in patients with liver cirrhosis, with mortality from the index bleed being reported as about 50%. [2,3] Without treatment, recurrent haemorrhage occurs in almost all patients who survive the initial bleeding episode, often within a year. In order to assess the risk of haemorrhage, an index combining Child’s classification (indicating severity of liver disease), size of varices and the presence/absence of red wale markings seen on endoscopic examination was developed. Patients were then stratified into groups with a risk of bleeding ranging from 6 to 63%. 
Portal hypertension is a crucial factor affecting the risk of recurrent haemorrhage. It has been observed that patients with variceal bleeding have a hepatic vein pressure gradient of >12mm Hg. 
Stabilise Patient and Confirm Diagnosis
Resuscitation and stabilisation of the patient is the primary objective, as aggressive early management may prevent complications such as hypotension-induced renal failure and deterioration in liver function. 
Haemodynamic stability is generally achieved using packed red blood cells to maintain the haematocrit at about 30% and fresh frozen plasma to manage coagulopathy. 
Drug Therapy to Reduce Portal Pressure
Immediately portal hypertensive haemorrhage is suspected, drug treatment to lower portal pressure should be commenced (see Differential features table ).  Nitroglycerin (glyceryl trinitrate) is often added to vasopressin or terlipressin therapy, because nitrates are powerful venodilators with a lesser arterial dilator effect. In addition, these agents further reduce portal pressure by venodilation. 
Once the patient is haemodynamically stable , urgent endoscopy should be performed. Endoscopy enables: 
- an accurate diagnosis to be made by excluding other causes of upper gastrointestinal tract blood loss
- localisation of the site of the bleeding varices.
Stop the Bleeding
Variceal haemorrhage may stop spontaneously, at least temporarily, in up to 40 to 60% of patients. In patients who continue to actively bleed, balloon tamponade may be used as a temporary measure until more definitive therapy can be undertaken. [2,3]I n balloon tamponade, the haemorrhage is controlled by direct compression of the bleeding vessels; h owever, further treatment is usually required as about 60% of patients rebleed after balloon removal. 
In clinical practice, most patients receive emergency sclerotherapy or band ligation, usually in conjunction with drug therapy (see Differential features table ), to control the haemorrhage. Vasoactive drugs contribute to control of bleeding and decrease portal pressure.
Sclerotherapy Successful for Many
The success rate of emergency sclerotherapy in stopping haemorrhage is about 90%. [2,3] The rate of early rebleeding may also be reduced following sclerotherapy. 
In sclerotherapy, a sclerosant is injected into the varix under direct vision. Injection techniques can be either intravariceal or paravariceal, or a combination of the two. Alternatively, adhesive agents can be used; however, as these agents harden quickly, they must be used quickly. Although, rebleeding does occur, it appears to be less common following sclerotherapy using adhesives than sclerosants.  Combined therapy with vasoactive drugs appears to be more effective than sclerotherapy alone in controlling haemorrhage. In addition, there appears to be a lower incidence of early rebleeding and reduced transfusion requirements using combined therapy. 
Band Ligation Better for Spurting Varices
Emergency band ligation is at least as effective as injection sclerotherapy in the control of bleeding from oozing varices and more effective in arresting haemorrhage from spurting varices.  However, endoscopic ligation may require more technical skill and the procedure may take longer than sclerotherapy. 
Prevent Further Bleeding
Once the acute bleeding episode has been controlled, it is important to prevent further haemorrhage as there is a 30 to 40% mortality rate associated with each bleeding episode. F ollow-up therapy is initiated to obliterate the varices and/or provide long-term control of portal pressure. 
Obliterate the Varices
Eradication of varices can be achieved either by repeated injection sclerotherapy or by band ligation. Although both methods are equally effective, band ligation requires fewer treatments, and is associated with fewer complications (oesophageal and gastric ulcers occur in about 70% of patients after sclerotherapy) and rebleeding episodes per treatment course than sclerotherapy.  However, these procedures do not reduce the portal pressure, and prophylactic drug therapy is usually initiated.
Lower Portal Pressure
Propranolol is the agent most frequently used to decrease portal pressure and prevent further bleeds in patients with oesophageal varices. [2,3] Propranolol acts by: 
- decreasing cardiac output and hence decreasing portal venous and hepatic arterial blood flow (a beta1-antagonist effect)
- blocking beta 2r eceptors in the splanchnic circulation leading to unopposed alpha-adrenergic vasoconstriction and a further reduction in portal blood flow.
The dosage of propranolol is adjusted until the heart rate is reduced by 25%, but not less than 55 beats/min.  Ideally, response to propranolol should be assessed by measuring the hepatic venous pressure gradient. In patients with a satisfactory reduction in hepatic venous pressure gradient with propranolol, there is a reduction in variceal bleeding episodes and a trend towards survival. 
In patients who cannot tolerate the adverse effects of propranolol, isosorbide dinitrate is a suitable alternative.  The dosage of isosorbide dinitrate should be titrated according to the change in blood pressure.
Shunts When Treatment Fails
If bleeding continues or recurs despite adequate endoscopic and drug therapy, then shunting must be considered. T ransjugular portosystemic stent-shunt (TIPS), a minimally invasive technique carried out transvenously under radiological control, is used by most centres. This technique is successful in >90% of patients. Procedure-related mortality is about 1%; mainly caused by abdominal haemorrhage resulting from hepatic capsular perforation.
Recurrent bleeding occurring after TIPS insertion is usually related to shunt insufficiency caused by thrombosis or stenosis of the stent. 
Because of the invasive nature of portacaval anastomosis and a high operative mortality rate of 5 to 15% in cirrhotic patients, this procedure is usually reserved for patients in whom other methods have failed to control the haemorrhage. 
Transplantation the Ideal?
Liver transplantation should be considered for all patients with end-stage chronic liver disease. Although variceal haemorrhage alone is not an indication for transplantation, this must be considered in those patients with poor liver function who have bled. 
The best survival rates in patients with oesophageal varices are achieved in those who receive a liver transplant (79% at 1 year and 71% at 5 years), with patients who are Child class C having the greatest survival advantage. 
Primary Prophylaxis for Those at Risk
Primary prophylaxis to try to prevent an index bleed should be considered in those patients at risk. T herefore, it has been suggested that a reasonable approach is to screen patients with cirrhosis for the presence of varices.
Beta-Blocker therapy should be started if moderate- or large-sized varices are identified and the patient has no contraindications to treatment.  A meta-analysis of propranolol therapy showed that the incidence of first time variceal bleeding was reduced by 47%, deaths due to bleeding by 45% and overall mortality by 22%. 
table 1. Different types of procedure used in the treatment of bleeding oesophageal varices [1-3]
ligation of varices
|Procedure||Thrombosis or necrosis of the varices is induced by injecting either a sclerosant (e.g. sodium tetradecyl-sulfate) or an adhesive [e.g. N-butyl-2-cyanoarylate (histoacryl)]||Thrombosis of the varices is caused by attachment of a tight rubber band around the varix and its surrounding mucosa||Placement of an expandable metal stent within the liver parenchyma to create a shunt between the portal and hepatic venous systems||Traditional surgical method of decompressing portal hypertension|
|Reduces portal hypertension?||No||No||Yes||Yes|
|Complications||Oesophageal strictures and ulcerations, pleural effusions and mediastinitis||Complications are those associated with the endoscopic procedure||Shunt insufficiency resulting in recurrent bleeds||Major surgical procedure with associated complications, including significant morbidity and mortality|
|There may be a risk of stroke following adhesive injection sclerotherapy ||Encephalopathy (10-20% of patients a )Transient worsening of liver function in [TM] 33% of patients
|Shunt occlusion and encephalopathy|
|a More frequent in those patients with severe underlying liver disease.|
Comparison of various features of drugs used to decrease portal pressure in patients with acute bleeding from oesophageal varices [2,3,5]
Feature Vasopressin † Terlipressin † Somatostatin Octreotide Mechanism of action Vasoconstrictor that reduces portal pressure by constricting the splanchnic bed and reducing inflow into the portal system Similar to vasopressin as terilpressin is a prodrug of vasopressin with some intrinsic activity. However, as terlipressin is converted by a saturable enzyme system, this agent results in a controlled release of active vasopressin Reduces splanchnic blood flow and has a modest effect on hepatic blood flow and wedged hepatic venous pressure with little effect on the systemic circulation Similar effects to somatostatin Route of administration Intravenous infusion Intravenous bolus Intravenous infusion Intravenous infusion Dosage 20 units over 15 minutes, then 0.4 units/minute until bleeding has stopped for 12 hours 1-2mg every 4-6 hours for 48 hours 250 mg/hour for 5 days 50 mg/hour for 2-5 days Adjunctive nitroglycerin (glyceryl trinitrate) a Yes Optional No No Adverse effects Myocardial, mesenteric and cutaneous ischaemia b As for vasopressin, but vasoconstrictor effects are less than vasopressin Occasionally causes altered liver function tests, hepatitis, pancreatitis and transient alopecia Efficacy shown in clinical trials (control of haemorrhage; % of patients) 50% (meta-analysis)  74% (vasopressin) vs 83% (vasopressin + nitrate) c 77.7%c ( vs 41.9% with placebo)  64% ( vs 41% with placebo) d 84% ( vs 90% with sclerotherapy)  and 85% ( vs 82% with sclerotherapy) 
† Vasopressin is not available in France and Spain; terlipressin is not available in Australia and Sweden. a A nitroglycerin patch 10mg is applied. Replace after 24 hours. b The incidence of adverse effects associated with vasopressin are reduced by the use of intravenous or transdermal nitrates. c A nitroglycerin patch was used in conjunction with the study drug. d An earlier study showed no benefit associated with somatostatin therapy; however, there was an unexpectedly high incidence of cessation of haemorrhage in the placebo group (83%). 
- Conn HO, Lebrec D, Terblanche J. The treatment of oesophageal varices: a debate and a discussion. J Intern Med 1997 Feb; 241: 103-8
- McCormack G, McCormick PA. A practical guide to the management of oesophageal varices. Drugs 1999 Mar; 57 (3): 327-35
- Trevillyan J, Carroll PJ. Management of portal hypertension and esophageal varices in alcoholic cirrhosis. Am Fam Physician 1997 Apr; 55: 1851-8
- Williams SGJ, Westaby D. Recent advances in the endoscopic management of variceal bleeding. Gut 1995; 36: 647-8
- British National Formulary. No. 376. London. The Pharmaceutical Press, 1999 Mar
- Burroughs AK, Panagou E. Pharmacological therapy for portal hypertension: rationale and results. Semin Gastrointest Dis 1995; 6 (3): 148-64
- Bosch J, Groszman RJ, Garcia-Pagan J, et al. Association of transdermal nitroglycerin to vasopressin infusion in the treatment of variceal haemorrhage: a placebo-controlled clinical trial. Hepatology 1989; 10 (6): 962-8
- Levacher S, Letoumelin P, Pateron D, et al. Early administration of terlipressin plus glyceryl trinitrate to control active upper gastrointestinal bleeding in cirrhotic patients. Lancet 1995; 346: 865-8
- Burroughs AK, McCormick PA, Hughes MD, et al. Randomized, double-blind, placebo-controlled trial of somatostatin for variceal rebleeding. Gastroenterology 1990; 99: 1388-95
- Sung JJY, Chung SCS, Lai CW, et al. Octreotide infusion or emergency sclerotherapy for variceal haemorrhage. Lancet 1993; 342: 637-41
- Jenkins SA, Shields M, Elias E, at al. A multicentre randomised trial comparing octreotide and injection sclerotherapy in the management and outcome of acute variceal haemorrhage. Gut 1997; 41: 526-33
- Valenzuela JE, Schubert T, Foger MR, et al. A multicenter, randomized, double-blind trial of somatostatin in the management of acute haemorrhage from esophageal varices. Hepatology 1989; 10 (6): 958-61
- Hope RA, Longmore JM. Oxford handbook of clinical medicine. Oxford University Press, 1985: 440
- D’Amico G, Pagliaro L, Bosch J. The treatment of portal hypertension: a meta-analytic review. Hepatology 1995; 22 (1): 332-54
- North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices. Prediction of the first variceal hemorrhage in patients with cirrhosis of the liver and esophageal varices. N Engl J Med 1988; 319: 983-9
- Gross M, Zoller WG. Medical prophylaxis of haemorrhage from oesophageal varices in patients with liver cirrhosis. Eur J Gastroenterol Hepatol 1997 Jun; 9: 603-12
- Lo GH, Lai KH, Cheng JS, et al. Emergency banding ligation versus sclerotherapy for the control of active bleeding from esophageal varices. Hepatology 1997; 25 (5): 1101-4
- Millikan WJ Jr, Henderson JM, Galloway JR, et al. Surgical rescue for failures of cirrhotic sclerotherapy. Am J Surg 1990; 160: 117-21
- Hayes P, Davis JM, Lewis JA, et al. Meta-analysis of value propranolol in prevention of variceal haemorrhage. Lancet 1990; 336: 153-6