Nonselective beta-blockers and the risk of portal vein thrombosis in patients with cirrhosis: results of a prospective longitudinal study.

Background Nonmalignant portal vein thrombosis is a significant event in the course of cirrhosis that can contraindicate liver transplantation and even impact survival after the surgical procedure. Risk factors are not completely known or validated and are still debated. Aim To identify in patients with cirrhosis the risk factors for portal vein thrombosis that are assessable in clinical practice. Methods Between January 2014 and February 2017, 108 outpatients with cirrhosis and no portal vein thrombosis (78% Child A) were enrolled. Doppler ultrasound was performed every 3 or 6 months, for a median follow up of 19 months. Results Portal vein thrombosis developed in 11 patients. Nonselective beta-blockade (hazard ratio [HR] 10.56; 95% confidence interval [CI]: 1.35-82.73; P = 0.025), and medium or large-sized oesophageal varices (HR 5.67; 95% CI: 1.49-21.63; P = 0.011) at baseline were associated with portal vein thrombosis development. Although heart rate (P < 0.001) and portal blood flow velocity at baseline (P = 0.005) were significantly reduced by nonselective beta-blockers, they were not related to portal vein thrombosis development. Conclusions Our findings confirm an association between portal vein thrombosis development and oesophageal varices at baseline, but suggest that the association could be explained by exposure to nonselective beta-blockers, independently from effects on heart rate and portal blood flow velocity. The mechanisms that explain portal vein thrombosis development in patients on nonselective beta-blockers require elucidation in order to optimise targeting of nonselective beta-blockade in patients with cirrhosis.