Key learning points

1. Portal vein thrombosis can present in the acute or chronic setting in patients with or without cirrhosis
2. The cornerstone of therapy is anticoagulation, which is mandatory where there is intestinal ischaemia or an underlying pro-coagulant condition
3. Vitamin K antagonists and heparin are recommended, whereas direct oral anticoagulants should be used only in the context of research

Introduction

There is considerable heterogeneity in the clinical presentation and course of portal vein thrombosis (PVT). It is useful to differentiate between acute and chronic PVT (either cirrhotic or non-cirrhotic) based on clinical presentation (Table 1).^1–3

Acute PVT

Acute PVT is associated with local (e.g. pancreatitis) and/or systemic prothrombotic conditions (e.g. myeloproliferative neoplasm, antiphospholipid syndrome, paroxysmal nocturnal haemoglobinuria, and, occasionally, inheritable thrombophilia). Patients may present with symptoms of the precipitating condition (e.g. malignancy, pancreatitis) with PVT being found incidentally on ultrasound scans. Alternatively, presentation might be recurrent or severe abdominal pain, suggesting venous ischaemia of the gut and extension of the thrombus to the superior mesenteric vein (SMV). Due to compensatory increase in hepatic arterial blood flow, liver function is usually unaltered even in extensive acute PVT.

Chronic PVT

Chronic PVT generally occurs in the setting of cirrhosis. The annual incidence is estimated to be 10–15%. It commonly presents asymptomatically on routine surveillance ultrasound scans or as new-onset or worsening decompensation. The pathophysiological characteristics are reduced portal flow, haemostatic imbalance (elevation in factor VIII and decrease in natural anticoagulants, such as protein C, protein S or antithrombin III) and endothelial injury (elevated von Willebrand factor and circulating endothelial microparticles). All patients with chronic PVT must be assessed for hepatocellular carcinoma.^1,2

Unrecognised acute PVT can lead to chronic PVT. In the absence of cirrhosis, chronic PVT is an important cause of portal hypertension in children from developing parts of the world. Umbilical sepsis and intra-abdominal infection are precipitant in a small number of children. These children present with well-tolerated variceal bleeds and massive splenomegaly with hypersplenism. Varying severity of portal cavernoma and portal biliopathy can eventually develop in children with long-standing PVT due to biliary-tree obstruction secondary to peri-choledochal varices.¹

Classifying PVT as anatomical (Table 2)⁴ or an anatomic-functional (site of PVT, degree of occlusion, duration and presentation, extent of PV system involvement and type and presence of underlying liver disease) may help in establishing prognosis and in overall management decision making.⁵

Diagnosis

Doppler ultrasonography of the abdomen is usually the initial test to diagnose PVT (Table 1). Cross-sectional imaging, such as multiphase contrast-enhanced CT (Table 1, Figure 1) and MR cholangiography, helps in delineating the extent, acuity and often the local cause of PVT (hepatocellular carcinoma, pancreatitis, source of intra-abdominal sepsis, intestinal infarction, portal biliopathy). Most patients will benefit from an evaluation for systemic prothrombotic conditions, and the extent should be individualised and guided by clinical presentation.^1,2

Treatment

Anticoagulation and management of any predisposing factor(s) remain the mainstay of management of PVT (Figure 2). The management of complications of PVT (e.g. intestinal ischemia and portal biliopathy) and role of advanced therapies, such as placement of a trans-jugular intrahepatic portosystemic shunt (TIPSS) and surgery, are beyond the scope of this article. Anticoagulation is mandatory in patients with symptomatic PVT (intestinal ischemia) or a pro-coagulant condition.

Anticoagulants

Spontaneous resolution of acute symptomatic non-cirrhotic PVT is highly unlikely and immediate institution of anticoagulation is recommended.¹ The treatment duration is usually 6 months if a solely transient local cause is identified, but may have to be extended in absence of an identifiable local cause and/or identification of a persisting systemic cause.²

In patients with cirrhosis, anticoagulation can be considered in symptomatic patients with extension of the clot to the SMV or a definite pro-coagulant risk factor. Despite limited evidence, anticoagulation according to the European Association for the Study of the Liver (EASL) guidelines² and Baveno VI consensus³ may be used to support cirrhotic PVT patients while on the transplant waiting list. PVT progression can increase the complexity of surgery and adversely affect the post-transplant outcomes. The risks and benefits of anticoagulation in other settings remain unclear.

In patients with non-cirrhotic chronic PVT, current EASL and Baveno VI recommendations support indefinite anticoagulation.^2,3 This approach might not be applicable in patients of Asian origin, as very limited benefit from non-selectively anticoagulating these patients is reported.⁶

In all patients with chronic PVT, assessment and institution of adequate prophylaxis for gastroesophageal varices are recommended prior to anticoagulation.⁷

Are anti-coagulants effective?

For acute PVT, limited data suggest that prompt anticoagulation effectively prevents thrombus extension and can result in recanalisation in 38–45% of patients.⁸ Direct thrombolysis (via TIPSS) may be useful. A few data suggest a high rate of immediate recanalisation and good long-term patency.⁹ In a recent UK study of patients with non-cirrhotic acute PVT with intestinal ischaemia, recanalisation was reported in 86% of 22 patients with a stepwise approach of systemic thrombolysis and local clot dissolution therapy by TIPSS and thrombolysis.¹⁰

Retrospective studies have shown that anticoagulants are efficacious in achieving recanalisation and prevention of de-novo thrombotic events in patients with chronic non-cirrhotic PVT, without a substantial increase in bleeding complications.¹¹

In chronic cirrhotic PVT, patients might experience spontaneous complete recanalisation. Studies have demonstrated anticoagulation to be safe and effective in increasing the rate of recanalisation (72% versus 42% without anticoagulation), but overall survival is not altered.¹²

In a highly selected group of patients, use of anticoagulants is not associated with significantly increased bleeding risks. An adequate platelet count (>50,000 cells/mm³ or 50 × 10⁹/L) and prophylaxis for gastroesophageal varices is recommended at the time of initiation.^1–3,7

In a single randomised trial, anticoagulation with low-molecular-weight heparin (LMWH) in patients with advanced cirrhosis was efficacious in preventing PVT and reducing the risk of worsening decompensation.¹³

Which anticoagulants are safe and effective?

There is lack of clarity about the choice of anticoagulants in the literature. In clinical practice, most patients who warrant anticoagulation are initiated on LMWH or unfractionated heparin and overlapped to vitamin K antagonists (VKAs) to maintain a target international normalisation rate of 2–3.

Direct oral anticoagulants (DOACs) are an oral alternative to VKAs. These drugs have a quicker onset of action and lower risk of bleeding, obviating the need for monitoring of international normalisation rate (Table 3). DOACs have equivalent efficacy to VKAs in treating deep-vein thrombosis and pulmonary embolism and in stroke prevention in patients with atrial fibrillation. However, patients with cirrhosis have been excluded from most trials due to imbalanced haemostasis. Although all DOACs undergo hepatic elimination to a variable extent, studies have shown them not to be associated with a significant risk of liver injury.¹⁴ DOACs can be used for licensed indications in Child’s A disease, but are contraindicated in Child B and C cirrhosis.

In small retrospective studies, the use of DOACs was found to be safe and efficacious (equivalent to VKAs) in PVT, without an increase in overall bleeding events.^15,16 A randomised study in 80 patients with compensated cirrhotic PVT related to hepatitis C virus (secondary to splenectomy) showed that rivaroxaban was associated with significantly better results for recanalisation than warfarin (85% versus 45%).¹⁷ Survival was improved and the risk of bleeding complications was greatly reduced. The study was limited by small sample size and its inclusion only of post-splenectomy patients. Larger studies are warranted before strong recommendation can be made regarding the use of DOACs in the setting of cirrhosis.

It is the opinion of the authors that the use of DOACs in PVT should be restricted to research, as current evidence is lacking in quality and quantity.

Course and prognosis

Spontaneous resolution is common in patients with recent PVT, especially if it is non-occlusive. These patients are typically asymptomatic and resolution is usually demonstrated in the setting of ultrasound surveillance of cirrhosis.¹

The potential impact of PVT on worsening of portal hypertension remains a debatable point. A randomised study demonstrated that PVT recanalisation had a beneficial effect on hepatic decompensation.¹¹ By contrast, a large multicentre study found no association between recent PVT and hepatic decompensation.¹⁸

In cirrhosis, PVT reflects the severity of liver disease, and its impact on overall outcome is difficult to assess.

Patients with non-cirrhotic chronic PVT typically survive long-term with a clinical course characterised by variceal bleeding, portosystemic encephalopathy and portal cavernoma with biliopathy, which can present with recurrent cholangitis and eventually biliary cirrhosis.

Conclusions

PVT can present in the acute or chronic setting in patients with or without cirrhosis. The cornerstone of therapy is anticoagulation, which is mandatory if the patient has intestinal ischaemia or an underlying pro-coagulant condition. The role of anticoagulation in the setting of cirrhosis is debatable, but is generally recommended for patients who are on the transplant waiting list and have proven PVT. VKAs and heparin are recommended, whereas DOACs should only be used in the context of research.

Author Biographies

Dr Dhiraj Tripathi

Dr Ashish Goel

Dr Ashish Goel is a liver speciality doctor at University Hospitals Birmingham, UK. He is currently on leave from the department of Hepatology, Christian Medical College, Vellore, India.

Dr William Lester

Dr William Lester is a Consultant Haematologist at University Hospitals Birmingham. He is a member of the Royal College of Physicians, and a fellow of the Royal College of Pathologists. His main clinical interests include thrombosis and haemostasis, von Willebrand disease, and obstetric haematology.