Effects of anticoagulants in patients with cirrhosis and portal vein thrombosis: a systematic review and meta-analysis

BACKGROUND & AIMS: Liver cirrhosis is complicated by bleeding from portal hypertension but also by portal vein thrombosis (PVT). PVT occurs in approximately 20% to 50% of patients with cirrhosis, and is a warning sign for poor outcome. It is a challenge to treat patients with cirrhosis using anticoagulants, because of the perception that the coexistent coagulopathy could promote bleeding. We performed a systematic review and meta-analysis to determine the effects of anticoagulant therapy in patients with cirrhosis and PVT. METHODS: We searched the PubMed, ISI Web of Science, SCOPUS, and Cochrane databases through February 14, 2017, for studies that assessed the effect of anticoagulant therapy vs no treatment in patients with cirrhosis and PVT. We performed a meta-analysis to estimate the effect of anticoagulant treatment vs no therapy on recanalization and progression of PVT in patients with cirrhosis. We also assessed variceal and nonvariceal bleeding. RESULTS: We analyzed data from 8 studies, comprising 353 patients, that assessed the effects of anticoagulant therapy (low-weight heparin or warfarin vs no therapy) in patients with cirrhosis and PVT; these studies reported rates of complete and partial recanalization. A significantly higher proportion of patients treated with anticoagulants underwent PVT recanalization than patients who did not receive anticoagulants (71% vs 42%, respectively; P < .0001). From 6 studies (comprising 217 patients), 53% of patients treated with anticoagulants vs 33% of patients who did not receive anticoagulants had complete PVT recanalization (P ¼ .002). From 6 studies (comprising 225 patients), PVT progressed in 9% of patients treated with anticoagulants vs 33% of patients who did not receive these drugs (P < .0001). Six studies (257 patients) reported rates of any bleeding; there was no difference in the proportions of patients with major or minor bleeding between groups that did vs did not receive anticoagulants (11% for both groups). Four studies (comprising 158 patients) reported rates of spontaneous variceal bleeding, which occurred in a significantly lower proportion of patients who received anticoagulants vs those who did not (P ¼ .04). CONCLUSIONS: Based on a systematic review and meta-analysis, patients with cirrhosis and PVT who receive anticoagulant therapy have increased recanalization and reduced progression of thrombosis, compared with patients who do not receive anticoagulants, with no excess of major and minor bleedings and less incidence of variceal bleeding.