EASL Clinical Practice Guidelines

Aetiological factors in splanchnic vein thrombosis in patients without underlying liver disease

In the last decades several aetiological factors for splanchnic vein thrombosis (SVT), including Budd-Chiari syndrome (BCS) and portal vein thrombosis (PVT), have been identified. These can be divided into local and systemic factors. Local risk factors for the development of BCS include solid malignancies or cysts that compress the venous tract [1]. PVT is most often seen as a complication of liver cirrhosis or hepatobiliary malignancies. Other local risk factors are intra-abdominal surgery and infections or inflammation in the abdomen. Systemic risk factors can be identified in most patients with SVT. In a large multicentre European En-Vie study on patients with BCS (n = 163) and PVT (n = 105), prothrombotic factors were present in up to 84% and 42%, respectively [[2], [3]] (Table 2). These data are consistent with earlier retrospective studies using similar diagnostic tools [[4], [5]]. In other parts of the world, especially in Asia other aetiological factors are observed, including Behçet disease, webs (also known as membranous obstruction) of the inferior vena cava (IVC) and hydatid cysts [[6], [7]]. Most studies have been performed in adults with SVT. In children with SVT prothrombotic factors seem to play an important aetiological role, however SVT may also be caused by age-specific factors, such as neonatal sepsis and umbilical catheterisation [8]. The aetiology of BCS and PVT is often multifactorial. In the En-Vie study a combination of two or more genetic or acquired prothrombotic factors occurred in 46% of BCS and 10% of PVT patients [[2], [3]]. In PVT a prothrombotic factor was found in 36% of patients with a local risk factor [3]. In BCS patients, 18% of the patients even had three risk factors. In over 60% of SVT patients diagnosed with inherited thrombophilia an additional risk factor was found.

Table 2
Aetiological factors in Budd-Chiari syndrome and portal vein thrombosis (references to articles of Murad and Plessier).

BCS, Budd-Chiari syndrome; PVT, portal vein thrombosis; PNH, paroxysmal nocturnal haemoglobinuria; n.d, no date.


Inherited and acquired thrombophilia

The term “thrombophilia” defines both inherited and acquired conditions that are associated with an increased risk of venous thrombosis, and is characterized by a hypercoagulable state [9]. Both inherited deficiencies of natural inhibitors of the coagulation system, increased levels of coagulation factors and genetic mutations of coagulant factors are associated with an increased risk of SVT. The prevalence of inherited deficiencies of antithrombin, protein C and protein S are difficult to assess in SVT patients, a result of decreased liver synthesis which is often encountered in these patients. Also treatment with vitamin K antagonists (VKA) hampers the diagnosis of protein C and protein S deficiency. The prevalence of antithrombin deficiency ranges between 0–5% in both BCS and PVT, of protein C deficiency between 4–20% in BCS and 0–7% in PVT, and of protein S deficiency between 0–7% in BCS and 0–30% in PVT [[2], [3], [4], [10], [11], [12]]. Because this is strikingly higher than in the general population, deficiencies of these coagulation inhibitors are considered an aetiological factor in the pathogenesis of BCS and PVT, and should be included in the diagnostic work-up.

In BCS patients the prevalence of Factor V Leiden mutation (FVL) ranges between 7% and 32%. Most of these BCS patients are heterozygous carriers, although homozygous patients have been described occasionally [13]. It is well known that homozygote carriers have a significantly higher risk of deep vein thrombosis compared to heterozygotes, however this has not been demonstrated for SVT. The prevalence of the FVL mutation in patients with PVT is lower, ranging between 3% and 9% [14]. FVL carriers have a 4- to 11-fold increased risk of BCS, and a 2-fold risk of PVT [15]. Prothrombin G20210A gene variant is more common in PVT than in BCS [14]. A meta-analysis reported a 4- to 5-fold increase in the risk of PVT in carriers of the prothrombin G20210A gene variant [15], whereas the risk of BCS is approximately 2-fold increased [10]. The mechanism for the difference in prevalence of FVL and the prothrombin G20210A gene variant in BCS and PVT remains unresolved. The prevalence of antiphospholipid antibodies (APA) in BCS and PVT has been estimated to be around 5–15% [[2], [3], [4]]. However, in most studies only one measurement of APA was carried out, whereas according to the current guidelines this measurement should be repeated after 12 weeks in order to confirm presence of APA [16].

In addition to the above mentioned risk factors for SVT, more recent studies have investigated whether increased levels of procoagulant factors or disorders of fibrinolysis are associated with an increased risk of SVT. Elevated factor VIII levels are found in patients with PVT [[17], [18]]. A significant increase of endogenous thrombin irrespective of the underlying prothrombotic or thrombophilic disorder was also observed in PVT [18]. Hypofibrinolysis, defined by an increase of clot lysis time, was also associated with an increased risk of BCS. This was mainly determined by increased plasminogen activator inhibitor-1 levels. So far the importance of these findings for prognosis and treatment of SVT has not been studied [19].


Myeloproliferative neoplasms

Myeloproliferative neoplasms (MPNs) are a common underlying cause of abdominal vein thrombosis. MPNs are chronic clonal haematopoietic stem cell disorders characterized by an overproduction of mature and functional granulocytes, red blood cells and/or platelets. One of the main complications of MPNs is the development of arterial and venous thrombotic complications caused by increased platelet aggregation and thrombin generation [[19], [20]]. It has previously been estimated that MPNs are observed in 30–40% of patients with BCS or PVT, whereas this is the cause in only a minority of other types of venous thromboembolism [[2], [3], [11], [21], [22]]. MPN is diagnosed based on several criteria including the characteristic peripheral blood cell changes (increased haemoglobin levels and thrombocytosis) and bone marrow findings. In SVT patients however the relevance of these commonly used criteria for the diagnosis of MPN is debated. Due to portal hypertension (PH) leading to hypersplenism and haemodilution the characteristic thrombocytosis and erythrocytosis may be masked [23]. Previously, diagnosis of MPNs in these patients relied on bone marrow (BM) biopsy findings and growth of erythroid colonies in the absence of exogenous erythropoietin, referred to as spontaneous endogenous erythroid colonies or EEC. This could also be used to identify patients at risk of aggravation of MPN [23]. Nowadays the JAK2V617F mutation, a common gain-of-function mutation leading to the development of MPN, is of major importance in the diagnostic strategy of MPN. This mutation is present in nearly all patients with polycythemia vera and in about 50% of patients with essential thrombocythemia and primary myelofibrosis. The JAK2V617F mutation has been detected in a large number of unselected BCS and PVT patients. In a recent meta-analysis the prevalence of MPNs and their subtypes as well as JAK2V617F and its diagnostic role in these uncommon disorders were reported [24]. In BCS, mean prevalence of MPNs and JAK2V617F was 40.9% and 41.1%, respectively. In PVT, mean prevalence of MPNs and JAK2V617F was 31.5% and 27.7%, respectively. MPN and JAK2V617F were more frequent in BCS compared to PVT. Polycythemia vera was more prevalent in BCS than in PVT. JAK2V617F screening in SVT patients without typical haematological MPN features identified MPN in 17.1% and 15.4% of screened BCS and PVT patients, respectively [24]. It can be concluded that in all patients with SVT BM histology and screening for JAK2V617F should be performed as part of the standard diagnostic work-up [25]. In some cases, MPN is difficult to diagnose and additional tests, such as peripheral blood smear, erythropoietin levels or endogenous erythroid colony formation in vitro may be added to the diagnostic algorithm, as suggested by the WHO [26]. Recently two research groups simultaneously reported the presence of somatic mutations in the gene encoding calreticulin (CALR), a protein present in the endoplasmic reticulum and involved in the regulation of STAT-signalling pathway [[27], [28]]. These mutations were detected using whole exome sequencing in the majority of patients with MPN with non-mutated JAK2. CALR mutations were absent in polycythemia vera patients, and occurred in up to 80% of patients with JAK2 negative essential thrombocythemia and primary myelofibrosis. In two recent studies [[29], [30]], CALR mutations were evaluated in patients with SVT being positive in 0.7 and 1.9% of patients respectively. The rate increased when only patients with MPN were considered (2.3 and 5.4% respectively). Indeed, CALR was found positive in respectively 9.1% (1 out of 11 patients) and 30% (4 out of 13 patients) of JAK2 negative MPN.

The exact pathogenetic mechanism of SVT in MPNs still remains to be resolved, but besides characteristic erythrocytosis and thrombocytosis, platelet and leukocyte functional abnormalities seem to have a pathogenetic role [31].


Other aetiological factors

Paroxysmal nocturnal haemoglobinuria (PNH) is a rare acquired haematological disorder of haematopoietic stem cells and is most strongly associated with BCS [32]. PNH has been reported in 9–19% of tested BCS patients [[11], [33]], whereas a prevalence of 0–2% has been reported in PVT [3]. The exact mechanism for the development of SVT is yet unknown [33]. Patients with a PNH cell population above 60% of the granulocytes appear to be at a greater risk for thrombosis [34]. Testing for PNH should routinely be performed in all BCS and considered in PVT patients [35]. Autoimmune-mediated diseases, inflammatory bowel disease, vasculitis, sarcoidosis and connective tissue disease may also be associated with SVT, although these disorders were hardly observed in the En-Vie study, Behçet’s disease is especially observed in the Mediterranean area [36]. Other rare causes of SVT include cytomegalovirus infections and celiac disease [[37], [38]].

Hormonal factors, including oral contraceptive use and pregnancy are considered risk factors for SVT. Oral contraceptives have been shown to be associated with at least a 2-fold risk for BCS [[10], [39]]. For PVT the risk may be slightly increased, but this has not yet been well-established [10]. It should be noted that in many patients other concomitant aetiological factors were identified.


Aetiological factors and their importance for treatment

Diagnosing the underlying aetiological factor for developing SVT is important, since it may have therapeutic or prognostic implications. For instance, the presence of a prothrombotic disorder may influence the duration of anticoagulant treatment in PVT patients. For patients with BCS, lifelong anticoagulant treatment is warranted considering the severity of the disorder. In individuals with acute PVT, anticoagulant therapy is given for 6 months. However, long-term treatment is sometimes given, depending upon the underlying disorder. In general, the duration of anticoagulant therapy is strongly dependent upon the risk of recurrent thrombosis. Although only a few retrospective studies have focused on the risk of recurrence in PVT, these studies revealed that an underlying prothrombotic state was an independent predictor of recurrent thrombosis [[40], [41], [42]]. On the other hand, the risk of bleeding in these patients, who frequently present with variceal bleeding, should be taken into account. Therefore recent guidelines have suggested long-term anticoagulant therapy only to those individuals with major underlying thrombophilic risk factors, such as homozygous FVL mutation and prothrombin gene variant [43]. However, other guidelines state that thrombophilic defects have an uncertain predictive value for recurrence and decisions regarding duration of anticoagulant treatment if the result of testing is not evidence-based [44]. Follow-up studies are needed to establish the duration of anticoagulant treatment especially those with no or mild thrombophilic disorders. Current guidelines do not support the testing of other family members in case a thrombophilia defect is identified [45].

In case of an underlying MPN, anticoagulant treatment with VKA should be given indefinitely for SVT. Nearly all MPN patients nowadays are treated with aspirin. However it is still unknown whether aspirin should be added to the treatment of SVT patients with MPN using VKA. Although a potential benefit of aspirin in patients with PVT and MPN was observed in a retrospective study, this should be confirmed in prospective studies [[44], [46]]. MPN patients should be treated with anti-proliferative therapy, such as alpha interferon or hydroxyurea, in order to normalise peripheral blood cell counts. In patients with polycythemia vera a haematocrit <45% should be aimed for [47]. The diagnosis of underlying PNH in patients with SVT may have important implications for treatment. Long-term treatment with eculizumab may be indicated in these individuals [35].