Risk Factors for Portal Vein Thrombosis Identified
SAN FRANCISCO, CA—Risk factors for portal vein thrombosis, a complication of cirrhosis, include the use of beta-blockers and hepatic encephalopathy, investigators reported during The Liver Meeting® 2015.
“High albumin levels were a protective factor,” reported Marta López, MD, Hospital Universitario Puerta de Hierro, Madrid, Spain, and colleagues.
Portal vein thrombosis has an annual incidence of 4% in Spain, lower than the 5%–10% incidence described in other recent studies, the researchers noted.
Using two university hospital databases, the investigators retrospectively reviewed clinical and radiological data for 747 consecutive patients diagnosed with cirrhosis between September 2013 to September 2014, to evaluate the annual incidence of portal vein thrombosis and related risk factors. A total of 179 patients were excluded from the analysis; exclusion criteria included hepatocellular carcinoma, known portal vein thrombosis, transjugular intrahepatic portosystemic shunt (TIPS), and pregnancy.
“All patients with ultrasound diagnosis of portal vein thrombosis underwent MR or CT angiography,” Dr. Lopez Gomez noted.
During the study period, 23 of 568 patients (4%) presented with portal vein thrombosis. Significant baseline differences between patients with and without portal vein thrombosis were observed in levels of albumin (3.4 vs. 4.0; P<0.001), AST (41.5 vs. 62; P=0.04), hemoglobin (12.6 vs. 13.8; P=0.01) and prothrombin activity (0.64 vs. 0.77; P=0.01), respectively.
Other factors predictive of portal vein thrombosis in univariate analysis were hydropic decompensation (61% vs. 29.7%; P=0.02); hepatic encephalopathy (39.1% vs. 9.9%;P<0.01); spontaneous bacterial peritonitis (13% vs. 1.7%; P<0.01); variceal bleeding (56.5% vs. 20.8%; P<0.01); esophageal varices (77.3% vs. 39.1%; P<0.001); and use of beta-blockers (65.2% vs. 26.6%; P<0.01), the researchers reported.
"In the multivariate analysis, beta-blockers (OR 4.3; IC: 1.4–12.6; P=0.01) and hepatic encephalopathy (OR 3.2, IC:1.1, 8; P=0.03) were risk factors and high albumin levels (OR 0.3; IC: 0.2, 0.8; P=0.01)" were predictive, they reported.
"Although portal vein speed was not significantly lower in patients with portal vein thrombosis, patients with beta-blockers had significantly lower portal vein speed (15 vs. 16.4; P<0.003),” they noted.
A total of 20 patients (87%) received anticoagulation for a median of 9 months (range, 1–12) months; five achieved re-permeabilization and one presented with decompensation.