In a study conducted between 1970 and 1982 on 23,796 autopsies revealed a portal vein thrombosis (PVT) prevalence of 1% wherein 28% of the subjects had cirrhosis, 23% of the subjects had primary hepatobiliary malignancy and 44% of the subjects had secondary hepatobiliary malignancy (1). The incidence of PVT in cirrhosis is unclear but, ranges from 4.4% to 15.8% (2). In a prospective study of 369 cirrhotic patients, the incidence of PVT was 1.6% at 1 year, 6% at 3 years and 8.3% at 5 years (3).

Portal Vein Thrombosis (PVT) is defined as a partial or complete obstruction of the portal vein due to thrombus formation (4). Although, there are several causes for PVT development whether it may local or systemic, they usually occur in concurrence to one another. A meta-analysis which included 2,436 cirrhotic patients suggests that PVT may increase mortality and ascites (5). In this review, we will discuss the etiopathogenesis, clinical features, diagnosis, management, and prevention of portal vein thrombosis in cirrhosis.

The etiopathogenesis of portal vein thrombosis (PVT) in cirrhosis is dynamic. Increased PVT risk is associated with decreased blood flow in relation to portal hypertension as per Virchow’s triad. The liver has a major role in the hemostatic system as it synthesizes most of the coagulation factors and proteins involved in fibrinolysis. It also synthesizes thrombopoietin which is responsible for platelet production. Consequently, acute and chronic liver disease have a profound impact on the hemostatic system. All three major components of the hemostatic process are disturbed in patients with cirrhosis. The derangements occur both on the procoagulant and the anticoagulation processes, leaving cirrhotic patients in a state of rebalanced hemostasis that can be altered in either direction in response to insults.

The extent of certain parameters correlating with PVT in cirrhosis has been shown in literature. Previously, PVT formation has been postulated as a consequence to a hypercoagulable and hyperinflammatory state due to the decrease of anticoagulants such as protein C and protein S with higher concentrations of certain factors such as Von Willebrand Factor (vWF) and factor VIII. Indeed, these risk factors are significant for systemic DVT formation. However, these same factors do not contribute to the formation of PVT. It was believed previously that the portal vein propagated PVT formation through the composition of an excessively inflammatory/hypercoagulable vascular bed. This belief was challenged in a prospective cohort study that showed that hypercoagulability and increased levels of inflammatory markers in the systemic circulation were not predictive of PVT development (6).

One risk factor for systemic DVT formation in the general public is non-O blood types. This is secondary to a hypercoagulable state due to increased von Willebrand Factor (VWF)/factor VIII levels in these individuals. This association is not present in non-O blood type patients with cirrhosis who develop PVT. A retrospective analysis of two large cohorts of cirrhosis patients with PVT found that non-O blood types was not a significant risk factor for PVT formation (7). In the first large cohort prospective study to address the risk factors for non-tumoral PVT development in cirrhotic patients, it was found that the severity of portal hypertension, history of variceal bleeding, low platelet count, and low portal blood flow velocity were the only significant factors contributing to PVT formation (8). It found that acquired or genetic hypercoagulable states were not significant risk factors. The dissociation of genetic hypercoagulability predisposing as a risk factor for PVT formation was further confirmed in a longitudinal prospective study (9).

The hemodynamic changes that occur in portal vein blood flow contribute in the formation of PVT in cirrhotic patients. As the severity of cirrhosis progresses, the intrahepatic vascular resistance increases proportionally and results in the clinical appearance of portal hypertension Consequentially, the increase in portal pressures leads to a compensatory splanchnic arteriolar vasodilation and the formation of porto-collateral vessels. The dilatory effect of the portal vein with shunting of blood away via the newly formed porto-collateral vessels leads to a substantial reduction of portal blood flow and portal flow velocity. The role of decreased portal flow velocity in the formation of PVT was demonstrated in a published prospective study of 100 cirrhotic patients in 2009 (10). This study found a flow velocity below 15 cm/s as a significant risk factor for the formation of PVT in cirrhotic patients. This parameter has been confirmed as a risk factor in other retrospective and prospective studies, therefore, establishing decreased portal flow velocity as a significant risk factor in the development of PVT in cirrhosis (3, 11, 12).

The role of endothelial intravascular vessel wall damage in hemostasis is well understood outside of the splanchnic territory. Although the territory is venous, the pathogenesis of thrombosis differs from that of systemic venous thrombosis. Most of information concerning the pathogenesis of venous thrombosis is systemic. This is largely because the splanchnic venous territory is vastly inaccessible (13, 14). During periods of stress, the endothelial cells can become damaged and express a pro-coagulant phenotype which propagates the initial formation of thrombus (15). The endothelial pro-coagulant phenotype has shown an upregulation of certain markers as the advancement of cirrhosis leads to a significant risk factor for PVT formation. Many studies have shown the upregulation of P-selectin (16, 17) and vWF (18, 19) in dysfunctional endothelium of the portal vein signifying a contributing factor for PVT. A study with 20 cirrhotic patients found elevated vWF in the portal venous circulation in comparison to the peripheral venous circulation, as well as increased levels of endothelial secreted Factor VIII (20). The significance of portal vein endothelial damage and upregulation of procoagulant factors is further supported by circulating factors such as sulphated glycosaminoglycans (GAGs), MP, annexin V+, CD62+, thrombomodulin and t-PAIC (21, 22). Although, the endothelial vessel wall damage plays a role in the formation of PVT, the exact mechanism is not completely elucidated and requires further research (10). The damage of the endothelial vessel wall from increased severity of portal hypertension also increases intimal hyperplasia. The association between formation of PVT secondary to upregulation of procoagulant factors, intimal hyperplasia, and increased severity of portal hypertension is not well understood. In sight of recent studies showing hypercoagulable states not significantly correlating with development of PVT in cirrhosis confirms that PVT and DVT/PE are two distinct disease processes with a distinct pathogenesis (3, 6, 7, 9). Thus, further elucidation of the pathophysiology of PVT formation is needed in literature.

The composition of a thrombosis in the portal vein is quite enigmatic in the sense that it can recanalize in the absence of treatment unlike systemic venous thrombi (10). A recent study analyzed 16 prospective and 64 retrospective portal vein segments from cirrhotic patients using histology and electron microscopy demonstrated that not only was the PVT contributing to the occlusion of the portal vein lumen, but rather the thickening of the tunica intima was also present (23). This appearance resembled intimal fibrosis and demonstrated that the thrombus formation found in the portal vein is more complex when compared to systemic venous thrombi. In addition, in 1/3^rd of the cases they found fibrinogen-rich blood clots which was distinct to those described in deep vein thrombi or arterial clots (23, 24). A study showed reduced incidence of PVT in patients given prophylactic ATIII and danaparoid sodium with liver cirrhosis and portal hypertension after splenectomy demonstrates that the etiopathogenesis of a thrombosis differs substantially in comparison to a systemic thrombosis and thus must be treated differently (25).

The formation of PVT in liver cirrhosis is associated with certain risk factors. A retrospective study with 98 cirrhotic patients with PVT and 101 cirrhosis without PVT demonstrated risk factors that contribute to the formation of PVT in cirrhotic patients. This study demonstrated an increase occurrence of PVT in advanced cirrhotic patients which confirmed previous studies. In this study it was demonstrated that patients with hyperglycaemia, hypoalbuminemia, anemia, hyperbilirubinemia, and elevated INR (international normalised ratio) levels, were statistically more likely to have thromboses compared with those of the controls (26, 27). Another significant risk factor leading to PVT that was seen in patient within this study was having HBV which was also seen in a similar study (28). These risk factors have been associated with PVT formation in patients with liver cirrhosis and portal hypertension which have been confirmed in prior studies (29–31). In sight of the many studies demonstrating risk factors for the etiopathogenesis of PVT development in cirrhosis, a meta-analysis of these studies would provide further insight on those that are truly significant.

The prevalence of PVT formation is generally low in the general population, but increased in individuals who have advanced liver cirrhosis with portal HTN. Studies have shown that the prevalence of PVT in cirrhosis with portal HTN increases proportionally with the severity of the disease (2, 32, 33). One study showed that 219 cirrhotic patients awaiting liver transplantation had an overall prevalence PVT that was 15.9% (2), supporting the reported prevalence range of 8%–25% of PVT in advanced liver cirrhosis found in similar studies (34, 35).

The clinical presentation of PVT is classified based on certain criteria. This includes whether the PVT is acute or subacute or chronic; occlusive vs. nonocclusive; benign vs. malignant, and intrahepatic vs. extrahepatic (2, 36–41). Depending on the criteria that is fulfilled on initial presentation determines the severity of symptoms. For example, a partially occluded acute portal vein thrombosis may be asymptomatic and present with nonspecific symptoms. In comparison, a completely occluded thrombosis in an acute phase can present with acute or progressive abdominal pain with signs of decompensation of chronic liver disease in the form of variceal bleeding, worsening ascites, bloody diarrhea, peritonitis, intestinal ischemia, or portal cholangiopathy. In a cirrhotic patient, sudden clinical deterioration such as the development of bacterial peritonitis can indicate the formation of an acute PVT through the unknown pathogenetic interaction of bacterial translocation, decreased portal blood flow velocity, and intimal hyperplasia of the portal vein wall. Intestinal infarction is a significant risk factor when the propagation of the thrombus extends to the superior mesenteric vein, mesenteric arches, and/or splenic vein (42).

The severity of portal hypertension in association with the extension of a portal vein thrombosis in cirrhotic patients correlates to an increased risk of complications. A cirrhotic patient with a PVT has greater than a threefold bleeding risk in comparison to a similar patient without a PVT irrespective of the use of endoscopic hemostasis or surgical shunting (43). In an acute complete occlusion of the portal vein, hepatic arterial vasodilatation typically preserves liver function (38, 42). Following a period of 3–5 weeks, the obstructed PVT is bypassed through the formation of venous collateral which is known as a portal cavernoma (39, 42).

Two separate studies showed that PVT in cirrhosis delayed the time to endoscopically fix esophageal varices, while also serving as an indicating factor of decompensation with poor diagnosis (44, 45). Contrary to these findings, two studies demonstrated no association between PVT development and prognosis (46, 47). A partial PVT that spontaneously resolved demonstrated a potential indication for improvement in liver function, but did not contribute to clinical outcome of a cirrhotic patient (48, 49). In comparison to benign, chronic PVT formation with a mortality rate of less than 10%, malignant PVT formation in the presence of liver cirrhosis secondary to HCC increases the mortality rate by 26% (50, 51). This demonstrates that PVT in liver cirrhosis increases the mortality risk of a patient irrespective of inciting factors.

There are many classifications of portal vein thrombosis. The Yerdel Classification is the most widely used because it can predict outcomes, correlates with surgical technique and with complication rates (33).

There are several different diagnostic modalities to evaluate for portal vein thrombosis as listed below:

When considering treatment, several factors need to be taken into consideration. Some studies suggest that PVT is associated with increased decompensation and mortality risk, while others indicate that it merely indicates cirrhosis progression. Bleeding remains a feared complication. In transplant candidates, the presence of PVT affects surgical technique and may affect survival.

Low molecular weight heparin (LMWH) and direct oral anticoagulants (DOACS) are considered safe in patients with cirrhosis and PVT. The drug selection needs to be individualized and adverse effects need to be discussed with the patient. Warfarin therapy has a narrow therapeutic window and the baseline elevation of the PT (prothrombin time) and INR in these patients creates difficulty for monitoring and dosing. Although, LMWH does not require monitoring, it does however include only injections which may be uncomfortable for the patient but, has potentially lesser side effects.

A meta-analysis on DOACs reported that the pooled rate of PVT recanalization among cirrhotic patients was 87.3% versus 44.1% in DOACs versus vitamin K antagonists (VKA) respectively. DOACs were associated with an overall lower pooled risk of major bleeding events when compared to VKAs but, have similar pooled risks of variceal bleeding and death (86). Rivaroxaban is contraindicated in patients with Child Pugh Group B and C cirrhosis with potential to be hepatotoxic (87–89). Apixaban has no warnings against its use in patients with Child Pugh Group A and B cirrhosis (87, 88). Edoxaban has both hepatic and renal clearance whereas dabigatran has predominantly renal clearance (90, 91). As per a review published in 2019, DOACs and LMWH may be safer and more efficacious than warfarin. To a large extent, it appears that DOACs may be safer and a more convenient option in PVT in cirrhotics all for the exception of rivaroxaban due to its hepatotoxic potential (89). Historically, warfarin and LMWH has been preferred due to familiarity and reversal agents. However, it is important to keep in mind that DOACs also have reversal agents in the event of major bleeding events. Despite this, further randomized clinical trials need to be conducted to assess safety and efficacy of DOACs in PVT in cirrhosis and to establish a new standard of care.

The AGA (American Gastroenterological Association Institute) guidelines recommend anticoagulation over no anticoagulation for the treatment of acute or subacute PVT in patients with cirrhosis (92). The EASL (European Association for the Study of the Liver) guidelines also have mentioned that anticoagulation have led to partial or complete recanalization of PVT in cirrhotic patients but, have not formally recommended this as larger randomized clinical trials would be needed to assess for morbidity and mortality (93). Although, EASL guidelines have not formally recommended DOACs for the treatment of PVT, they do recommend using DOACs in the treatment of deep vein thrombosis/pulmonary embolism (DVT/PE) in patients with Child-Pugh class A cirrhosis. They recommend using DOACs with caution in patients with Child-Pugh class B cirrhosis and in those with a creatinine clearance of less than 30 ml/min and advise against DOACs in Child-Pugh class C cirrhosis (93). This could be applied to the treatment for PVT as well although, data is limited. The AASLD (American Association for the Study of the Liver) guidelines have stated treatment for PVT is weak due to lack of clinical trials and the indication for anticoagulation should be dependent on the patient while considering the expected benefits for the patient and decreasing risk for clot extension (44). The AASLD guidelines also state that the non-portal hypertensive bleeding rates among cirrhotics compared to the general population on therapeutic anticoagulation appear to be similar with portal hypertensive bleeding rates among cirrhotics appear to be unchanged by anticoagulation (44).

Based on the guidelines mentioned above and the studies summarized in Table 1, it can be stated that anticoagulation for PVT in cirrhotics should not be feared and may even be recommended especially in the cases of acute and subacute PVT. Individualized bleeding risks should be analyzed prior to initiation of anticoagulation such as evaluating for portal hypertensive gastropathy, unbanded varices, etc. The type of anticoagulation used must be individualized based on their class of cirrhosis, renal function, and compliance with injectables and INR monitoring. Patients can be followed up closely to evaluate for medication related adverse effects, thrombus progression or bleeding complications.

Table 2 summarizes studies regarding the natural history of portal vein thrombosis in patients who did not undergo any form of intervention.

The suggested treatment algorithm of PVT is as described in Figure 1.

The prevention of PVT in cirrhosis is still an area that is currently undergoing research. A randomized controlled trial studying 70 cirrhotic patients showed that PVT was prevented at 96 weeks in the group that received 4,000 IU of enoxaparin versus 10 out of 36 controls developed PVT (130). Another study stratified risk of PVT based on antithrombin levels where high and highest risk cirrhotic patients received antithrombin III concentrates and danaparoid sodium which resulted in a low incidence of PVT (25). Another randomized controlled trial showed that warfarin is more effective than aspirin in preventing PVT after a laparoscopic splenectomy in cirrhotics while also providing hepato- and nephroprotective effects (131).

This review intends to evade the fear of anticoagulation and interventional strategies in patients with cirrhosis. It was written to enhance the knowledge of managing portal vein thrombosis in cirrhosis as it may help in reducing portal hypertensive complications. Despite the guidance on the management of PVT, an individualized assessment of risks vs. benefits is necessary when deciding between different treatment strategies.

All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.