Editorial: Safety and Effects of Direct Oral Anticoagulants for Portal Vein Thrombosis

Portal vein thrombosis (PVT) is a serious complication in chronic liver disease, mainly occurring in patients with portal hypertension [1]. The pathophysiology of PVT includes a combination of venous stasis, hypercoagulability and endothelial dysfunction, leading to reduced portal flow velocity and stagnation of blood in the splanchnic circulation [2, 3]. PVT is associated with a range of complications in cirrhosis, including variceal bleeding and ascites and PVT may become a contraindication for liver transplantation [4]. Many cases are asymptomatic and detected incidentally during routine imaging. Symptomatic presentations include abdominal pain, worsening ascites or gastrointestinal bleeding. Treatment aims at achieving recanalisation with anticoagulation [5]. The safety of anticoagulation in cirrhosis requires consideration, especially in those with coagulopathy or varices. Acute-on-chronic PVT involves thrombus formation superimposed on a background of long-standing partial obstruction, often presenting more subtly but with potential for worsening liver function. PVT in people without cirrhosis is mainly acute and often due to an acquired or inherited prothrombotic condition [5, 6]. Although treatment principles are similar, non-cirrhotic cases generally have fewer risks related to coagulopathy, altering the safety and efficacy profile of anticoagulation therapy. In a study of 102 patients with acute PVT without cirrhosis, recanalisation of the portal vein was achieved in 39% following anticoagulation with vitamin K antagonists (VKA) and nine experienced gastrointestinal bleeding [6]. Treating PVT with direct oral anticoagulants (DOACs) offers several advantages over VKA including a more predictable effect eliminating the need for frequent monitoring. In a meta-analysis of 10 observational studies and one randomised trial including patients with cirrhosis and PVT, DOACs were associated with recanalisation in 87% [7]. Premkumar et al. report their experience with dabigatran for benign PVT [8]. The rate of recanalisation was assessed for 72 patients with cirrhosis and 47 without cirrhosis receiving dabigatran. Comparison groups consisted of 28 patients (18 with cirrhosis) who declined treatment and six patients with CP > 10 who received VKA. Recanalisation occurred in 56 (47.1%) patients treated with dabigatran and more frequently in acute compared to acute-on-chronic PVT. The large proportion of patients with acute on chronic PVT may partly explain the relatively low number of patients who achieved recanalisation. A large number of patients were identified through screening for portal hypertension rather than based on symptoms or incidental findings, which may impact results. None of the patients treated with VKA and 6 (21.4%) without treatment experienced spontaneous recanalisation. The overall adverse event rate of dabigatran was 6.7% including one patient with variceal bleeding and three with epistaxis [8]. This aligns with the result of a meta-analysis comparing anticoagulants versus no anticoagulation where 4% in the treatment group experienced variceal bleeding and 11% experienced major bleeding episodes [7]. Although the non-randomised design and the heterogeneity of patients should be considered when making conclusions, the overall result is very promising. The evidence suggests that DOACs are considered for the treatment of PVT, especially in patients with acute PVT. Additional evidence is needed to evaluate the benefit–risk ratio in patients with decompensated cirrhosis and in patients with acute on chronic PVT. Nina Kimer: writing – review and editing. Lise Hobolth: writing – review and editing. Lise Lotte Gluud: conceptualization, writing – original draft, writing – review and editing. N.K. and L.H. declare no conflicts of interest. L.L.G.: Novo Nordisk, Pfizer, Becton Dickinson, Gilead, Sobi, Alexion, Astra Zeneca, Boehringer Ingelheim, Immunovia and Norgine. This article is linked to Premkumar et al paper. To view this article, visit https://doi.org/10.1111/apt.18474. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.