Variceal bleeding is a consequence of portal hypertension, which in turn is
the major complication of hepatic cirrhosis. Given the high rate of mortal
ity of the first bleeding episode, primary prophylaxis to prevent bleeding
from varices and portal hypertensive gastropathy is the current optimal the
rapeutic approach. The difficulty in identification of patients with varice
s who will bleed, before they do so, can justify a strategy of treating all
patients with varices prophylactically. We evaluated the various therapies
that have been assessed in randomized controlled trials for prevention of
first bleeding, using meta-analysis where applicable. The current first cho
ice treatment is non-selective beta-blockers; it is cheap, easy to administ
er, and is effective in preventing the first variceal haemorrhage and bleed
ing from gastric mucosa. Combination drug therapy of beta-blockers and nitr
ates looks promising, but needs further evaluation in randomized controlled
trials. The conflicting results of the randomized studies of endoscopic ba
nding ligation and the small number of patients and clinical events, as wel
l as the cost, do not warrant any change in current practice. However, endo
scopic banding ligation may be a reasonable alternative for patients who ca
nnot tolerate, or have contraindications to beta-blockers or no haemodynami
c response to the drug therapy, but this must be proved in randomized trial
s.