Variceal bleeding and portal hypertensive gastropathy.

Cirrhosis can be the end stage of any chronic liver disease. At the time of diagnosis of cirrhosis varices are present in about 60% of decompensated and 30% of compensated patients. The risk factors for the first episode of variceal bleeding in cirrhotic patients are the severity of liver dysfunction, large size of varices and the presence of endoscopic red colour signs but only one-third of patients who have variceal haemorrhage have the above risk factors. Recent interest has been directed at identifying haemodynamic factors that may reflect the pathophysiological changes which lead to variceal bleeding, e.g. it has been confirmed that no bleeding occurs if HVPG falls below 12 mmHg and also a hypothesis has been put forward in which bacterial infection is considered a trigger for bleeding. Pharmacological treatment with beta-blockers is safe, effective and is the standard long-term treatment for the prevention of recurrence of variceal bleeding. Combination of beta-blockers with isosorbide-5-mononitrate needs further testing in randomized controlled trials. The use of haemodynamic targets for reduction in HVPG response needs further study, and surrogate markers of pressure response need evaluation. If endoscopic treatment is chosen, variceal ligation is the modality of choice. The combination of simultaneous variceal ligation and sclerotherapy does not offer any benefit. However, the use of additional sclerotherapy for the complete eradication of small varices after variceal ligation needs to be evaluated. The results of current prospective randomized controlled trials comparing variceal ligation with pharmacological treatment are awaited with great interest. Finally, the use of transjugular intrahepatic portosystemic shunt (TIPS) for the secondary prevention of variceal bleeding is not substantiated by current data, as survival is not improved and because of its worse cost-benefit profile compared to other treatments. In contrast, there still is a role for the selective surgical shunts in the modern management of portal hypertension. The ideal patients should be well compensated cirrhotics, who have had troublesome bleeding - either who have failed at least one other modality of therapy (drugs or ligation), have bled from gastric varices despite medical or endoscopic therapy, or live far from suitable medical services. Recently, ligation has been compared to beta-blockers for primary prophylaxis but so far there is no good evidence to recommend banding for primary prophylaxis, if beta-blockers can be given.