The aetiology of portal vein thrombosis (PVT) is heterogeneous. Important p
rimary risk factors for PVT are cirrhosis. hepatobiliary malignancies and p
ancreatitis. Newly discovered thrombotic risk factors, such as latent myelo
proliferative disorders acid prothrombotic genetic defects, have also been
identified as major risk factors for PVT. At least one-third of PVT patient
s demonstrate a combination of thrombotic risk factors. PVT, which does not
have a detrimental effect on liver function, usually becomes manifest as a
variceal haemorrhage in the oesophagus months to years after the developme
nt of thrombosis. Owing to intact coagulation variceal bleeding has a bette
r prognosis among patients with PVT than cirrhotics. Endoscopic sclerothera
py or band ligation is the primary therapeutic option for variceal bleeding
in patients with PVT. It is questionable whether anticoagulant therapy sho
uld be started, since it has not proven beneficial for most PVT patients. T
herapy with anticoagulants is only recommended for those with acute PVT (es
pecially in association with mesenteric vein thrombosis). those who recentl
y underwent a portosystemic shunt procedure, and those with other thromboti
c manifestations, particularly in case of proven hypercoagulability. Mortal
ity of patients with PVT may be associated with concomitant medical conditi
ons which lead to the PVT or with manifestations of portal hypertension, su
ch as variceal haemorrhage. Multivariate analysis of a large Dutch PVT popu
lation has shown that age, malignancy, ascites and the presence of mesenter
ic vein thrombosis are independently related to survival. Death due to a va
riceal haemorrhage is rare. Poor outcome of PVT thus appears to be associat
ed primarily with concomitant diseases which lead to PVT, and not the compl
ications of portal hypertension. It is therefore uncertain whether surgical
portosystemic shunting affects survival favourably.