Advertisement

Current knowledge in pathophysiology and management of Budd-Chiari syndrome and non-cirrhotic non-tumoral splanchnic vein thrombosis

  • Virginia Hernández-Gea
    Affiliations
    Barcelona Hepatic Hemodynamic Laboratory, Liver Unit, Hospital Clínic de Barcelona, IDIBAPS, CIBERehd, European Reference Network for Rare Vascular Liver Diseases, Universitat de Barcelona, Spain
    Search for articles by this author
  • Andrea De Gottardi
    Affiliations
    Hepatology, University Clinic of Visceral Medicine and Surgery, Inselspital, and Department of Biomedical Research, University of Bern, Bern, Switzerland
    Search for articles by this author
  • Frank W.G. Leebeek
    Affiliations
    Department of Haematology, Erasmus University Medical Center, Rotterdam, the Netherlands
    Search for articles by this author
  • Pierre-Emmanuel Rautou
    Affiliations
    Service d'Hépatologie, Centre de Référence des Maladies Vasculaires du Foie, DHU Unity, Pôle des Maladies de l'Appareil Digestif, Hôpital Beaujon, AP-HP, Clichy, France

    Inserm, UMR-970, Paris Cardiovascular Research Center, PARCC, Paris, France
    Search for articles by this author
  • Riad Salem
    Affiliations
    Department of Radiology, Section of Interventional Radiology, Northwestern University, Chicago, IL, USA
    Search for articles by this author
  • Juan Carlos Garcia-Pagan
    Correspondence
    Corresponding author. Address: Barcelona Hepatic Hemodynamic Laboratory, Liver Unit, Hospital Clinic, Villarroel 170, Barcelona 08036, Spain. Fax: +34 932279856.
    Affiliations
    Barcelona Hepatic Hemodynamic Laboratory, Liver Unit, Hospital Clínic de Barcelona, IDIBAPS, CIBERehd, European Reference Network for Rare Vascular Liver Diseases, Universitat de Barcelona, Spain
    Search for articles by this author
Published:February 26, 2019DOI:https://doi.org/10.1016/j.jhep.2019.02.015

      Summary

      Budd-Chiari syndrome and non-cirrhotic non-tumoral portal vein thrombosis are 2 rare disorders, with several similarities that are categorized under the term splanchnic vein thrombosis. Both disorders are frequently associated with an underlying prothrombotic disorder. They can cause severe portal hypertension and usually affect young patients, negatively influencing life expectancy when the diagnosis and treatment are not performed at an early stage. Yet, they have specific features that require individual consideration. The current review will focus on the available knowledge on pathophysiology, diagnosis and management of both entities.

      Keywords

      Introduction

      Budd-Chiari syndrome (BCS) is defined as the obstruction of hepatic venous outflow regardless of its causative mechanism or level of obstruction. This obstruction can be traced to the small hepatic venules up to the entrance of the inferior vein cava (IVC) into the right atrium. Hepatic outflow obstruction related to cardiac disease, pericardial disease or sinusoidal obstruction syndrome have different pathophysiological and clinical implications and are excluded from this definition. BCS is classified as primary when the obstruction originates in the vein and thrombosis is the main cause, or secondary when the vein is externally compressed (abscess, tumour). The focus of this review is on primary BCS. Non-cirrhotic non-tumoral portal vein thrombosis (NCPVT) refers to the presence of a thrombus in the main portal vein trunk and/or the left or right intrahepatic portal vein branches that may extend to the splenic vein and/or the superior or inferior mesenteric veins. Isolated splenic or mesenteric vein thrombosis are beyond the scope of this review.

      BCS and NCPVT risk factors

      The estimated incidence of BCS and NCPVT in the absence of cirrhosis and cancer is 1 per million per year and 0.35–2.5 cases per 100,000 per year, respectively.
      • Ollivier-Hourmand I.
      • Allaire M.
      • Goutte N.
      • Morello R.
      • Chagneau-Derrode C.
      • Goria O.
      • et al.
      The epidemiology of Budd-Chiari syndrome in France.
      • Ageno W.
      • Dentali F.
      • Pomero F.
      • Fenoglio L.
      • Squizzato A.
      • Pagani G.
      • et al.
      Incidence rates and case fatality rates of portal vein thrombosis and Budd-Chiari syndrome.
      • Ki M.
      • Choi H.Y.
      • Kim K.-A.
      • Kim B.H.
      • Jang E.S.
      • Jeong S.-H.
      Incidence, prevalence and complications of Budd-Chiari syndrome in South Korea: a nationwide, population-based study.
      Most patients with BCS and NCPVT have identifiable thrombotic risk factors. However, both entities represent only ∼1% of all venous thromboembolic events, implying other local factors besides thrombophilic disorders for developing BCS or NCPVT.
      • Heit J.A.
      Epidemiology of venous thromboembolism.
      • Smalberg J.H.
      • Kruip M.J.
      • Janssen H.L.
      • Rijken D.C.
      • Leebeek F.W.
      • de Maat M.P.
      Hypercoagulability and hypofibrinolysis and risk of deep vein thrombosis and splanchnic vein thrombosis: similarities and differences.
      Although some risk factors are shared by both entities, others are specifically related to one or the other. Indeed, myeloproliferative neoplasm (MPN) and factor V Leiden mutation are prothrombotic conditions strongly associated with BCS but less frequently observed in NCPVT.
      • Deltenre P.
      • Denninger M.H.
      • Hillaire S.
      • Guillin M.C.
      • Casadevall N.
      • Briere J.
      • et al.
      Factor V Leiden related Budd-Chiari syndrome.
      Risk factors for BCS and NCPVT and their prevalence in 3 large European studies, including patients enrolled between 2003–2005 and 2013–2014, are presented in Table 1. Compared with older multicentre European studies, including patients between 2003 and 2005, we observed an overall stability in the prevalence of these risk factors over the last 15 years.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      • Murad S.D.
      • Plessier A.
      • Hernandez-Guerra M.
      • Fabris F.
      • Eapen D.E.
      • Bahr M.J.
      • et al.
      Etiology, management, and outcome of the Budd-Chiari syndrome.
      Work-up for these risk factors is presented in Table 2. Multiple prothrombotic conditions are found in 15–20% of patients with BCS or NCPVT suggesting that, when one causal factor is identified, additional factors should be investigated. Conversely, in some patients, no risk factor is found. However, in population-based databases, the reported percentage of patients with no risk factor is variable and has significantly decreased in recent studies, suggesting an improvement in their detection. Indeed, in the European study performed in 2009, which analysed 157 patients with BCS, 16% of patients did not show any prothrombotic risk factor. Interestingly, when these patients were re-analysed during follow-up, additional aetiological factors were diagnosed in 12 previously unclassified patients. Nonetheless, recent data from France and Italy describe the absence of a prothrombotic factor in 30%
      • Ollivier-Hourmand I.
      • Allaire M.
      • Goutte N.
      • Morello R.
      • Chagneau-Derrode C.
      • Goria O.
      • et al.
      The epidemiology of Budd-Chiari syndrome in France.
      and 61%
      • Ageno W.
      • Dentali F.
      • Pomero F.
      • Fenoglio L.
      • Squizzato A.
      • Pagani G.
      • et al.
      Incidence rates and case fatality rates of portal vein thrombosis and Budd-Chiari syndrome.
      of patients respectively, maybe due to the intrinsic limitations of population studies. Differences between Eastern and Western countries are found and the prevalence of prothrombotic disorders in China seems to be very low.
      • Qi X.
      • Wu F.
      • Ren W.
      • He C.
      • Yin Z.
      • Niu J.
      • et al.
      Thrombotic risk factors in Chinese Budd-Chiari syndrome patients. An observational study with a systematic review of the literature.
      However, over the years a higher detection of hypercoagulability conditions has been described,
      • Pati H.P.
      • Dayal S.
      • Srivastava A.
      • Pande G.K.
      • Acharya S.K.
      Spectrum of hemostatic derangements, in Budd-Chiari syndrome.
      reinforcing the need for more studies aimed at uncovering the causes of BCS in Asia.
      • Qi X.
      • Han G.
      • Guo X.
      • De Stefano V.
      • Xu K.
      • Lu Z.
      • et al.
      Review article: the aetiology of primary Budd-Chiari syndrome – differences between the West and China.
      Up to 30% of patients with NCPVT have no identifiable aetiological factor.
      Table 1Prevalence of acquired and inherited risk factors for BCS, NCPVT in 3 recent European cohort studies.
      • Turon F.
      • Cervantes F.
      • Colomer D.
      • Baiges A.
      • Hernandez-Gea V.
      • Garcia-Pagan J.C.
      • et al.
      Role of calreticulin mutations in the aetiological diagnosis of splanchnic vein thrombosis.
      • Poisson J.
      • Plessier A.
      • Kiladjian J.J.
      • Turon F.
      • Cassinat B.
      • Andreoli A.
      • et al.
      Selective testing for calreticulin gene mutations in patients with splanchnic vein thrombosis: a prospective cohort study.
      • Bureau C.
      • Laurent J.
      • Robic M.A.
      • Christol C.
      • Guillaume M.
      • Ruidavets J.B.
      • et al.
      Central obesity is associated with non-cirrhotic portal vein thrombosis.
      Underlying conditionNCPVT n = 432BCS n = 168
      n tested% positiven tested% positive
      Acquired conditions
       Myeloproliferative neoplasms43221%16841%
      JAK2V617F43216%16835%
       Antiphospholipid syndrome4296%16510%
       PNH3860.3%1527%
      Inherited conditions
       Factor V Leiden4293%1658%
       Factor II gene mutation4326%1683%
       Protein C deficiency4045%1505%
       Protein S deficiency4075%1474%
       Antithrombin deficiency4161%1531%
      External factors
       Recent pregnancy3532%1681%
       Recent oral contraceptive use35314%16822%
      Systemic disease
      Including connective tissue disease, coeliac disease, Behçet’s disease, mastocytosis, inflammatory bowel disease, human immunodeficiency virus infection, sarcoidosis, myeloma.
      4323%1686%
      Inflammatory intra-abdominal lesions
      Acute pancreatitis, biliary or intestinal infection or inflammation.
      43211%1682%
      Intra-abdominal surgery43210%1681%
      Abdominal trauma2924%1682%
      >1 risk factor43214%16819%
      No cause
      Oral contraception and pregnancy were not considered risk factors for NCPVT in all studies.
      21942%16824%
      BCS, Budd-Chiari syndrome; PNH, paroxysmal nocturnal haemoglobinuria; NCPVT, non-cirrhotic non-tumoral portal vein thrombosis.
      * Including connective tissue disease, coeliac disease, Behçet’s disease, mastocytosis, inflammatory bowel disease, human immunodeficiency virus infection, sarcoidosis, myeloma.
      ** Acute pancreatitis, biliary or intestinal infection or inflammation.
      *** Oral contraception and pregnancy were not considered risk factors for NCPVT in all studies.
      Table 2Investigations for thrombotic risk factors in patients with BCS or NCPVT.
      Underlying disordersSuggestive signs
      All causes should be tested in all patients. Do not restrict testing to patients displaying these signs.
      Work-up
      Systemic
      Acquired conditions
       MPNNormal platelet count



      Large spleen
      - In all patients, test first JAK2V617F in peripheral granulocyte DNA.

      - In patients without JAK2V617F, but with spleen height ≥16 cm and platelet count >200  × 109/L, test for CALR mutations. Propose a bone marrow biopsy in patients without CALR mutations.

      - In the remaining patients, i.e. those without JAK2V617F when spleen height is <16 cm and platelet count ≤200 × 109/L, MPNs are extremely uncommon. MPL and JAK2 exon 12 mutations are very rare. MPN diagnosis relies on bone marrow biopsy performed on a case-by-case basis.
       Antiphospholipid syndromeDiagnosis based on repeatedly detectable anticardiolipin antibodies at high level, or lupus anticoagulant, or antibeta2 glycoprotein 1 antibodies. Many patients with vascular liver disease have nonspecific fluctuating, low titer antiphospholipid antibodies in the absence of antiphospholipid syndrome.
       PNHSmall hepatic vein involvementCD55 and CD59 deficient clone at flow-cytometry of peripheral blood cells.
      Inherited conditions
       Factor V LeidenActivated protein C resistance. To be confirmed in patients with positive results, by molecular testing for factor V Leiden mutation
       Factor II gene mutationMolecular testing for G20210A mutation
       Protein C deficiencyResults can be interpreted only in patients with normal coagulation factor levels. Diagnosis based on decreased protein C activity levels. Inherited deficiency can be established only with a positive test in first degree relatives.
       Protein S deficiencyResults can be interpreted only in patients with normal coagulation factor levels. Diagnosis based on decreased free protein S levels. Inherited deficiency can be established only with a positive test in first degree relatives.
       Antithrombin deficiencyResults can be interpreted only in patients with normal coagulation factor levels. Diagnosis based on decreased antithrombin activity levels. Inherited deficiency can be established only with a positive test in first degree relatives.
       HyperhomocysteinemiaIncreased serum homocysteine level prior to disease. Uncertain value of C677T homozygous polymorphism. In many patients, a definite diagnosis for underlying hyperhomocysteinemia will not be possible. Blood folate and serum vitamin B12 levels may be useful.
      External factors
       Recent pregnancyMedical history
       Recent oral contraceptive useMedical history
      Systemic disease
       Behcet diseaseInvolvement of inferior vena cavaDiagnosis based on a set of conventional criteria. To be routinely considered in patients with inferior vena cava thrombosis, or originating from endemic areas, or having extrahepatic features suggestive of the disease.
       HIV infectionAnti-human immunodeficiency virus antibodies
       Celiac diseaseAnti-transglutaminase antibodies
       SarcoidosisBlack patients; respiratory symptomsSerum angiotensin-converting-enzyme level; Biopsy
       Cytomegalovirus infectionAnti-CMV IgM
      Local factor
       Inflammatory intra-abdominal lesionsMedical history. CT-scan. Increased circulating levels of C reactive protein and/or of fibrinogen and/or increased platelet count, although recent thrombosis can by itself induce systemic inflammation.
       Abdominal surgeryMedical history
       Abdominal traumaMedical history
      BCS, Budd-Chiari syndrome; NCPVT, non-cirrhotic non-tumoral portal vein thrombosis; PNH, paroxysmal nocturnal haemoglobinuria, MPN, myeloproliferative neoplasms.
      * All causes should be tested in all patients. Do not restrict testing to patients displaying these signs.
      In the following paragraphs we discuss the more relevant risk factors for developing BCS or NCPVT.

      Local risk factors

      In patients with BCS, abdominal infection or inflammation is more rarely identified than in patients with NCPVT,
      • Murad S.D.
      • Plessier A.
      • Hernandez-Guerra M.
      • Fabris F.
      • Eapen D.E.
      • Bahr M.J.
      • et al.
      Etiology, management, and outcome of the Budd-Chiari syndrome.
      possibly due to the impossibility of bypassing the liver. A complementary hypothesis could be that activated platelets and microvesicles generated at the site of inflammation/infection are cleared within the liver by liver endothelial cells and macrophages.
      • Rautou P.-E.
      • Mackman N.
      Deletion of microvesicles from the circulation.
      • Ma R.
      • Xie R.
      • Yu C.
      • Si Y.
      • Wu X.
      • Zhao L.
      • et al.
      Phosphatidylserine-mediated platelet clearance by endothelium decreases platelet aggregates and procoagulant activity in sepsis.
      However, it is still not fully understood why, even in the setting of a general prothrombotic condition, thrombosis arises at such an unusual site without any inflammatory or mechanical injury. In the European BCS cohort, the presence of local trauma, inflammatory diseases and abdominal infections was only reported in 11% of patients,
      • Murad S.D.
      • Plessier A.
      • Hernandez-Guerra M.
      • Fabris F.
      • Eapen D.E.
      • Bahr M.J.
      • et al.
      Etiology, management, and outcome of the Budd-Chiari syndrome.
      although in more recent data from the French survey it was found in 25% of patients.
      • Ollivier-Hourmand I.
      • Allaire M.
      • Goutte N.
      • Morello R.
      • Chagneau-Derrode C.
      • Goria O.
      • et al.
      The epidemiology of Budd-Chiari syndrome in France.
      In patients with NCPVT, a local cause should be exhaustively investigated, as it may be present in approximately 30% of cases. Initial imaging studies (CT or MRI) performed during the diagnostic work-up should be examined carefully as they may reveal signs of gastrointestinal (appendicitis, diverticulitis, intra-abdominal abscesses or infections) or biliopancreatic pathologies including pancreatic pseudocysts and gallbladder alterations.

      Abdominal infection and inflammation

      Although thrombosis by itself can induce systemic inflammatory responses, abdominal infection is a classical cause of PVT.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      Together with the inflammatory response, activation of coagulation is an important response in the host's defence against infection, as it prevents the dissemination of microorganisms. This implication of coagulation in infection is illustrated by the improved survival of patients with sepsis carrying heterozygous factor V Leiden compared with those without, a finding confirmed in animal models.
      • Kerlin B.A.
      • Yan S.B.
      • Isermann B.H.
      • Brandt J.T.
      • Sood R.
      • Basson B.R.
      • et al.
      Survival advantage associated with heterozygous factor V Leiden mutation in patients with severe sepsis and in mouse endotoxemia.
      Mechanisms by which infection triggers thrombosis have recently been reviewed in detail.
      • Iba T.
      • Levy J.H.
      Inflammation and thrombosis: roles of neutrophils, platelets and endothelial cells and their interactions in thrombus formation during sepsis.
      Briefly, monocytes and neutrophils play an important role: monocytes express tissue factor, the primary initiator of the coagulation cascade, and release tissue factor positive microvesicles; neutrophils release neutrophil extracellular traps (NETs), i.e. networks of extracellular fibres, primarily composed of DNA from neutrophils, which bind pathogens and activate coagulation and thrombosis by favouring FVIIa-mediated thrombin generation and activating the intrinsic pathway. Simultaneously, platelets are activated and contribute to clot formation. Endothelial cells lose their physiological antithrombotic phenotype following exposure to inflammatory mediators and to activated neutrophils, platelets, and other cells. In addition to the cellular components, alarmins such as histones, high mobility group box 1, microvesicles and secreted granule proteins, are all important for clot formation.
      • Iba T.
      • Levy J.H.
      Inflammation and thrombosis: roles of neutrophils, platelets and endothelial cells and their interactions in thrombus formation during sepsis.
      Inflammation without infection, e.g. acute pancreatitis or inflammatory bowel disease, shares many of the aforementioned features with infection and may also contribute to thrombosis.
      • De Caterina R.
      • D’Ugo E.
      • Libby P.
      Inflammation and thrombosis – testing the hypothesis with anti- inflammatory drug trials.
      PVT induced by inflammation or infection can be seen close to the site of infection/inflammation and may extend and or embolise to the portal trunk and branches (Fig. 1).
      Figure thumbnail gr1
      Fig. 158-year-old male patient with a perforated sigmoid diverticulitis. This patient had an abscess (lower green arrow), complicated with an inferior mesenteric vein thrombus (upper green arrow) with extension into intrahepatic portal branches (red arrow). (Courtesy of Dr Onorina Bruno, Hôpital Beaujon, Clichy, France).

      Abdominal malignancies

      Abdominal cancer is another common cause of NCPVT. The pathogenesis of the cancer-associated coagulopathy is complex and multifactorial.
      • Falanga A.
      • Russo L.
      • Milesi V.
      • Vignoli A.
      Mechanisms and risk factors of thrombosis in cancer.
      Most importantly, tumour cells gain the capacity to activate the host haemostatic system, a phenomenon driven by the same oncogenes responsible for the cellular neoplastic transformation. Indeed, cancer tissues express different procoagulant proteins including tissue factor, factor VII and cancer procoagulant (a molecule that, unlike tissue factor, directly activates factor X independently of coagulation factor VII), which contribute to the occurrence of the overt symptomatic coagulopathy in vivo. The shedding of procoagulant microvesicles is also regulated by oncogenic events and further adds to the pathogenesis of the cancer-associated hypercoagulable state. It is important to note that in one-third of patients with NCPVT exhibiting a recognized local factor, additional general prothrombotic risk factors are found.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      Notably, in one-third of patients with NCPVT and a recognized local factor, additional general prothrombic risk factors are present.

      Myeloproliferative neoplasms

      MPNs are chronic clonal haematopoietic stem cell disorders characterized by an overproduction of granulocytes, erythrocytes and/or platelets. Currently, 7 subcategories of MPN have been identified, of which polycythaemia vera (PV), essential thrombocythemia (ET) and primary myelofibrosis (PMF) have the highest prevalence. Patients with MPNs are at a high risk of arterial and venous thrombotic complications.
      • Marchetti M.
      • Castoldi E.
      • Spronk H.M.
      • van OR O.R.
      • Balducci D.
      • Barbui T.
      • et al.
      Thrombin generation and activated protein C resistance in patients with essential thrombocythemia and polycythemia vera.
      A recent meta-analysis revealed that MPNs are found in 40% of patients with BCS and 30% of those with NCPVT.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      • Murad S.D.
      • Plessier A.
      • Hernandez-Guerra M.
      • Fabris F.
      • Eapen D.E.
      • Bahr M.J.
      • et al.
      Etiology, management, and outcome of the Budd-Chiari syndrome.
      • Smalberg J.H.
      • Darwish M.S.
      • Braakman E.
      • Valk P.J.
      • Janssen H.L.
      • Leebeek F.W.
      Myeloproliferative disease in the pathogenesis and survival of Budd-Chiari syndrome.
      • Kiladjian J.-J.J.
      • Cervantes F.
      • Leebeek F.W.G.W.
      • Marzac C.
      • Cassinat B.
      • Chevret S.
      • et al.
      The impact of JAK2 and MPL mutations on diagnosis and prognosis of splanchnic vein thrombosis: a report on 241 cases.
      • Smalberg J.H.
      • Arends L.R.
      • Valla D.C.
      • Kiladjian J.J.
      • Janssen H.L.
      • Leebeek F.W.
      Myeloproliferative neoplasms in Budd-Chiari syndrome and portal vein thrombosis: a meta-analysis.
      More impressively, NCPVT and BCS are 2,000 and 10,000 times more common in patients with MPN than in the general population.
      • How J.
      • Zhou A.
      • Oh S.T.
      Splanchnic vein thrombosis in myeloproliferative neoplasms: pathophysiology and molecular mechanisms of disease.
      This high prevalence is not restricted to Western patients, since similar figures have been reported in India, Turkey and Egypt,
      • Shetty S.
      • Kulkarni B.
      • Pai N.
      • Mukundan P.
      • Kasatkar P.
      • Ghosh K.
      JAK2 mutations across a spectrum of venous thrombosis cases: table 1.
      • Deepak A.
      • Punamiya S.
      • Patel N.
      • Parekh S.
      • Mehta S.
      • Shah N.
      Prevalence of JAK29V617F) mutation in intra-abdominal venous thrombosis.
      • Shukla A.
      • Parikh H.
      • Modi T.
      • Abraham P.
      • Kamble S.
      • Majumder D.
      Hepatic vein obstruction is the most common type of hepatic venous outflow obstruction regardless of socioeconomic status.
      • Valla D.-C.
      Budd-Chiari syndrome/hepatic venous outflow tract obstruction.
      although less common in Chinese patients with BCS or NCPVT.
      • Qi X.
      • Wu F.
      • Ren W.
      • He C.
      • Yin Z.
      • Niu J.
      • et al.
      Thrombotic risk factors in Chinese Budd-Chiari syndrome patients. An observational study with a systematic review of the literature.
      • Wang H.
      • Sun G.
      • Zhang P.
      • Zhang J.
      • Gui E.
      • Zu M.
      • et al.
      JAK2 V617F mutation and 46/1 haplotype in Chinese Budd-Chiari syndrome patients.
      According to the current WHO 2016 guidelines, the diagnosis of PV is based on several criteria, the main criterion being an increased haemoglobin >16.5 g/dl in men or >16.0 g/dl in women, or a haematocrit >49% in men or >48% in women.
      • Arber D.A.
      • Orazi A.
      • Hasserjian R.
      • Borowitz M.J.
      • Le Beau M.M.
      • Bloomfield C.D.
      • et al.
      The 2016 revision to the World Health Organization classification of myeloid neoplasms and acute leukemia.
      For ET the main criterion is a platelet count >450 × 109/L. In addition, typical bone marrow findings including hypercellularity and an increased number of mature, enlarged, pleomorphic megakaryocytes with hyperlobulated nuclei, are present.
      In recent years, several underlying somatic mutations have been identified in MPN. In PV, the JAK2V617F or JAK2 exon 12 mutations are found in >95% of patients. In ET and PMF, JAK2V617F mutations are found in 50%. More recently, mutations in the calreticulin gene (CALR), encoding a protein present in the endoplasmic reticulum and involved in regulation of the STAT-signalling pathway, have been identified in 80% of patients with MPN that are JAK2V617F negative.
      • Klampfl T.
      • Gisslinger H.
      • Harutyunyan A.S.
      • Nivarthi H.
      • Rumi E.
      • Milosevic J.D.
      • et al.
      Somatic mutations of calreticulin in myeloproliferative neoplasms.
      • Nangalia J.
      • Massie C.E.
      • Baxter E.J.
      • Nice F.L.
      • Gundem G.
      • Wedge D.C.
      • et al.
      Somatic CALR mutations in myeloproliferative neoplasms with nonmutated JAK2.
      Based on these findings JAK2V617F, JAK2 exon 12 and CALR or thrombopoietin receptor (MPL) mutations have become major diagnostic criteria for MPNs.
      Compared to other patients with MPNs, those with BCS or PVT with underlying MPN are typically younger and more frequently female. In patients with splanchnic vein thrombosis who were diagnosed with MPN based on clinical, laboratory and/or morphological features of MPN in the bone marrow, as well as JAK2V617F mutation status, 80% of patients with BCS and 87% with PVT were JAK2V617F positive.
      • Smalberg J.H.
      • Arends L.R.
      • Valla D.C.
      • Kiladjian J.J.
      • Janssen H.L.
      • Leebeek F.W.
      Myeloproliferative neoplasms in Budd-Chiari syndrome and portal vein thrombosis: a meta-analysis.
      JAK2V617F mutation, irrespective of other MPN features, was present in 41% of patients with BCS and 28% with PVT. Subclassification of MPN in patients with BCS reveals 53% PV, 25% ET and 7% PMF. For patients with PVT and MPN this is 28% PV, 26% ET and 13% PMF, respectively. Other patients with MPN could not be classified. However, due to portal hypertension, leading to hypersplenism and haemodilution in patients with BCS or NCPVT, peripheral blood cell counts, haemoglobin or haematocrit can be normal or even reduced, even in the presence of other criteria for MPN. In patients without typical haematologic features, the JAK2V617F mutation can be found in 17.1% and 15.4% of patients, respectively.
      • Smalberg J.H.
      • Arends L.R.
      • Valla D.C.
      • Kiladjian J.J.
      • Janssen H.L.
      • Leebeek F.W.
      Myeloproliferative neoplasms in Budd-Chiari syndrome and portal vein thrombosis: a meta-analysis.
      CALR mutations are less frequent in patients with splanchnic vein thrombosis (<5%), however they may be of help to diagnose underlying MPN.
      • Turon F.
      • Cervantes F.
      • Colomer D.
      • Baiges A.
      • Hernandez-Gea V.
      • Garcia-Pagan J.C.
      • et al.
      Role of calreticulin mutations in the aetiological diagnosis of splanchnic vein thrombosis.
      • Plompen E.P.C.
      • Valk P.J.M.
      • Chu I.
      • Murad S.D.
      • Plessier A.
      • Turon F.
      • et al.
      Somatic calreticulin mutations in patients with Budd-Chiari syndrome and portal vein thrombosis.
      Due to the low incidence of CALR mutations, only screening for CALR mutations in patients who are JAK2V617F negative and have platelets >200 × 109/L with a spleen size of >16 cm has been suggested.
      • Poisson J.
      • Plessier A.
      • Kiladjian J.J.
      • Turon F.
      • Cassinat B.
      • Andreoli A.
      • et al.
      Selective testing for calreticulin gene mutations in patients with splanchnic vein thrombosis: a prospective cohort study.
      MPL mutations are rare, but they may be included in the diagnostic work-up, together with JAK2V617F, JAK2 exon 12 and CALR mutations.
      • Karalus N.C.
      • Dunn P.J.
      • Haslam A.J.
      • Burroughs A.
      Outcome of diabetic pregnancies in Waikato: five year experience.
      • Akpan I.J.
      • Stein B.L.
      Splanchnic vein thrombosis in the myeloproliferative neoplasms.
      • Iurlo A.
      • Cattaneo D.
      • Gianelli U.
      • Fermo E.
      • Augello C.
      • Cortelezzi A.
      Molecular analyses in the diagnosis of myeloproliferative neoplasm-related splanchnic vein thrombosis.
      • Langabeer S.E.
      • Haslam K.
      Incidence of CALR mutations in patients with splanchnic vein thrombosis.
      In a small subset of patients with BCS or PVT, peripheral blood counts are normal and molecular markers for MPN are negative, with MPN diagnosis made by bone marrow biopsy.
      • Smalberg J.H.
      • Arends L.R.
      • Valla D.C.
      • Kiladjian J.J.
      • Janssen H.L.
      • Leebeek F.W.
      Myeloproliferative neoplasms in Budd-Chiari syndrome and portal vein thrombosis: a meta-analysis.
      The role of bone marrow biopsy to diagnose MPN in patients with BCS and NCPVT, when all previous molecular markers are negative, remains challenging. Currently, decisions are made on a case-by-case basis. In the near future, it is possible that the introduction of extensive molecular diagnostic panels, that are able to simultaneously analyse multiple mutations, will change these recommendations.
      Over the last 10 years, several studies have shed light on the close relationship between MPN and BCS or NCPVT. A pivotal study reported on the existence of JAK2V617F in endothelial cells from hepatic veins in 2 patients with BCS.
      • Sozer S.
      • Fiel M.I.
      • Schiano T.
      • Xu M.
      • Mascarenhas J.
      • Hoffman R.
      The presence of JAK2V617F mutation in the liver endothelial cells of patients with Budd-Chiari syndrome.
      Subsequently, JAK2V617F has also been detected in splenic endothelial cells from patients with myelofibrosis.
      • Rosti V.
      • Villani L.
      • Riboni R.
      • Poletto V.
      • Bonetti E.
      • Tozzi L.
      • et al.
      Spleen endothelial cells from patients with myelofibrosis harbor the JAK2V617F mutation.
      JAK2V617F was found in circulating endothelial progenitor cells in 5 out of 17 patients with somatic JAK2V617F expression.
      • Teofili L.
      • Martini M.
      • Iachininoto M.G.
      • Capodimonti S.
      • Nuzzolo E.R.
      • Torti L.
      • et al.
      Endothelial progenitor cells are clonal and exhibit the JAK2V617F mutation in a subset of thrombotic patients with Ph-negative myeloproliferative neoplasms.
      Interestingly, patients harbouring JAK2V617F in circulating endothelial progenitor cells were those with a history of thrombosis (splanchnic vein thrombosis, deep vein thrombosis or stroke).
      • Teofili L.
      • Martini M.
      • Iachininoto M.G.
      • Capodimonti S.
      • Nuzzolo E.R.
      • Torti L.
      • et al.
      Endothelial progenitor cells are clonal and exhibit the JAK2V617F mutation in a subset of thrombotic patients with Ph-negative myeloproliferative neoplasms.
      Moreover, JAK2V617F mutated circulating endothelial progenitor cells showed significantly higher adhesion proficiency to mononuclear cells than normal circulating endothelial progenitor cells.
      • Teofili L.
      • Martini M.
      • Iachininoto M.G.
      • Capodimonti S.
      • Nuzzolo E.R.
      • Torti L.
      • et al.
      Endothelial progenitor cells are clonal and exhibit the JAK2V617F mutation in a subset of thrombotic patients with Ph-negative myeloproliferative neoplasms.
      The exact mechanism by which JAK2V617F in the endothelium leads to BCS or NCPVT remains unclear. Recent results from experiments using cultured endothelial cells transduced with a lentivirus expressing JAK2V617F and transgenic mice expressing JAK2V617F in their endothelial cells have filled this gap in knowledge.
      • Guy A.
      • Gourdou-Latyszenok V.
      • Le Lay N.
      • Peghaire C.
      • Kilani B.
      • Vieira Dias J.
      • et al.
      Vascular endothelial cell expression of JAK2V617F is sufficient to promote a pro-thrombotic state due to increased P-selectin expression.
      In this study, James and colleagues demonstrated that JAK2V617F induces the exposure of P-Selectin at the surface of endothelial cells, increasing endothelial adhesion of platelets, of neutrophils and of mononuclear cells and inducing in vivo thrombus formation. Interestingly, in mice, small concentrations of TNFα were required to uncover this increased adhesion, suggesting that a low level of inflammation may trigger thrombosis in MPN. Similar results were obtained by an independent group using pluripotent stem cells from patients with MPN redirected towards the endothelial lineage.
      • Guadall A.
      • Lesteven E.
      • Letort G.
      • Awan Toor S.
      • Delord M.
      • Pognant D.
      • et al.
      Endothelial cells harbouring the JAK2V617F mutation display pro-adherent and pro-thrombotic features.
      These results are a step forward in our understanding of the link between BCS or NCPVT and MPN. However, several questions remain unanswered: (a) Is JAK2V617F expressed in endothelial cells only in the digestive vascular bed or ubiquitously? If ubiquitous, additional factors are needed and might be inflammatory mediators derived from the gut. (b) Why is the somatic myeloid mutation JAK2V617F also found in endothelial cells? It is unlikely that endothelial JAK2V617F is due to the occurrence of the mutation in a common cell of origin for endothelial cells and myeloid cells, called haemangioblasts. Indeed, haemangioblasts exist in embryos, but not in adults, and JAK2V617F related BCS and NCPVT are rare in young children.
      • Poisson J.
      • Lemoinne S.
      • Boulanger C.
      • Durand F.
      • Moreau R.
      • Valla D.
      • et al.
      Liver sinusoidal endothelial cells: physiology and role in liver diseases.
      Patients with MPN and BSC or NCPVT are usually younger (≈30 years) than patients with MPN without thrombosis (≈60 years), suggesting the presence of haemangioblast mutations.
      In BCS, JAK2V617F is associated with worse prognostic features at presentation and the earlier requirement for hepatic decompression procedures.
      • Kiladjian J.-J.J.
      • Cervantes F.
      • Leebeek F.W.G.W.
      • Marzac C.
      • Cassinat B.
      • Chevret S.
      • et al.
      The impact of JAK2 and MPL mutations on diagnosis and prognosis of splanchnic vein thrombosis: a report on 241 cases.
      To determine whether endothelial JAK2V617F enhances liver injury and fibrosis induced by hepatic venous outflow obstruction, thus worsening BCS, a surgical model of BCS was applied to mice expressing endothelial JAK2V617F. It was observed that the expression of JAK2V617F in liver endothelial cells did not affect liver injury or liver fibrosis, meaning that endothelial JAK2V617F does not explain the more severe presentation of patients with BCS and JAK2V617F.
      • Poisson J.
      • Hilscher M.B.
      • Tanguy M.
      • Hammoutene A.
      • Boulanger C.M.
      • Villeval J.L.
      • et al.
      Endothelial JAK2V617Fdoes not enhance liver lesions in mice with Budd-Chiari syndrome.
      Therefore, the myeloid expression of JAK2V617F should be explored as the most likely explanation for this.
      NCPVT and BCS are 2,000 and 10,000 times more common in patients with MPN than in the general population, with JAK2V617F mutations found in >95% of patients with MPN.

      Other haematological conditions

      Paroxysmal nocturnal haemoglobinuria (PNH) is a rare acquired haematological disorder of haematopoietic stem cells which leads to complement-induced haemolysis and is strongly associated with an increased risk of venous thrombosis.
      • van Bijnen S.T.A.
      • van Rijn R.S.
      • Koljenovic S.
      • te Boekhorst P.
      • de Witte T.
      • Muus P.
      Possible high risk of thrombotic events in patients with paroxysmal nocturnal haemoglobinuria after discontinuation of eculizumab.
      BCS is one of the most common sites of thrombosis in patients with PNH, affecting 7–25%. More than one-fifth of the patients with PNH develop thrombosis in multiple sites.
      • Hill A.
      • Kelly R.J.
      • Hillmen P.
      Thrombosis in paroxysmal nocturnal hemoglobinuria.
      PNH has been reported in 9–19% of patients with BCS,
      • Hoekstra J.
      • Leebeek F.W.G.
      • Plessier A.
      • Raffa S.
      • Murad S.D.
      • Hadengue A.
      • et al.
      Paroxysmal nocturnal hemoglobinuria in Budd-Chiari syndrome: findings from a cohort study.
      whereas a prevalence of 0–2% has been reported in patients with NCPVT.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      Patients with a PNH cell population above 60% of the granulocytes are at a high risk of thrombosis.
      • Brodsky R.A.
      Narrative review: paroxysmal nocturnal hemoglobinuria: the physiology of complement-related hemolytic anemia.
      Testing for PNH should routinely be performed in all BCS and considered in NCPVT.
      • van Bijnen S.T.A.
      • van Rijn R.S.
      • Koljenovic S.
      • te Boekhorst P.
      • de Witte T.
      • Muus P.
      Possible high risk of thrombotic events in patients with paroxysmal nocturnal haemoglobinuria after discontinuation of eculizumab.

      Systemic thrombophilic disorders

      Factor V Leiden and factor II G20210A gene mutations are other frequently found prothrombotic factors in patients with BCS and NCPVT, respectively, with apparent site specificity. Indeed, the prevalence of factor V Leiden is twice as high in European patients with BCS (Table 1) than in the general population (∼4–5%), and is commonly associated with other risk factors for thrombosis.
      • Deltenre P.
      • Denninger M.H.
      • Hillaire S.
      • Guillin M.C.
      • Casadevall N.
      • Briere J.
      • et al.
      Factor V Leiden related Budd-Chiari syndrome.
      • Plompen E.P.C.
      • Valk P.J.M.
      • Chu I.
      • Murad S.D.
      • Plessier A.
      • Turon F.
      • et al.
      Somatic calreticulin mutations in patients with Budd-Chiari syndrome and portal vein thrombosis.
      Similar or even higher figures have been reported in India, Turkey and Egypt, but factor V Leiden is not found in Chinese patients with BCS.
      • Valla D.-C.
      Budd-Chiari syndrome/hepatic venous outflow tract obstruction.
      The G20210A mutation of the factor II gene is more common in patients with NCPVT (Table 1) than in the general Caucasian population (∼2%).
      • Plompen E.P.C.
      • Darwish Murad S.
      • Hansen B.E.
      • Loth D.W.
      • Schouten J.N.L.
      • Taimr P.
      • et al.
      Prothrombotic genetic risk factors are associated with an increased risk of liver fibrosis in the general population.
      By contrast, the role of factor V Leiden in patients with NCPVT and factor II gene mutation in patients with BCS appears to be negligible. The mechanism underlying this site specificity is unknown.
      Antiphospholipid syndrome is a third common risk factor for BCS or NCPVT. However, its diagnosis is difficult because of the poor specificity of antiphospholipid antibodies in chronic liver disease.
      • Qi X.
      • De Stefano V.
      • Su C.
      • Bai M.
      • Guo X.
      • Fan D.
      Associations of antiphospholipid antibodies with splanchnic vein thrombosis: a systematic review with meta-analysis.
      Primary antiphospholipid syndrome affects males and females, but a large percentage of patients are women with recurrent pregnancy loss, while secondary antiphospholipid syndrome occurs mainly in lupus, and about 90% of lupus patients are female
      Precise prevalence of inherited protein C, protein S or antithrombin deficiencies is difficult to estimate since the diagnosis of primary deficiencies is based on determination of plasma levels of these coagulation inhibitors that are synthesized by the liver. In patients with liver dysfunction, a non-specific decrease in plasma levels of these inhibitors makes interpretation of protein C, protein S or antithrombin levels quite challenging.
      • Janssen H.L.
      • Meinardi J.R.
      • Vleggaar F.P.
      • van Uum S.H.
      • Haagsma E.B.
      • Der Meer F.J.
      • et al.
      Factor V Leiden mutation, prothrombin gene mutation, and deficiencies in coagulation inhibitors associated with Budd-Chiari syndrome and portal vein thrombosis: results of a case-control study.

      Other systemic risk factors

      Behçet’s disease

      Behçet’s disease is a rare systemic disorder, clinically diagnosed by the presence of recurrent oral aphthous ulcers and genital ulcerations together with eye lesions, that may also be associated with the development of BCS.
      • Seyahi E.
      • Caglar E.
      • Ugurlu S.
      • Kantarci F.
      • Hamuryudan V.
      • Sonsuz A.
      • et al.
      An outcome survey of 43 patients with Budd-Chiari syndrome due to Behçet’s syndrome followed up at a single, dedicated center.

      Obesity

      In general, obesity is a risk factor for the first episode of venous thromboembolism (VTE) with an estimated overall odds ratio for VTE of 2.3.
      • Lentz S.R.
      Thrombosis in the setting of obesity or inflammatory bowel disease.
      Obesity is also a risk factor for recurrent VTE with an estimated hazard ratio of 1.6, a degree of risk similar to that of other risk factors for recurrent VTE.
      • Lentz S.R.
      Thrombosis in the setting of obesity or inflammatory bowel disease.
      In addition to clinical factors such as immobility, obstructive sleep apnoea, heart failure, and venous stasis, the major mechanisms proposed to be responsible for obesity-associated thrombosis are impaired fibrinolysis, and chronic inflammation.
      • Lentz S.R.
      Thrombosis in the setting of obesity or inflammatory bowel disease.
      Adipokines and proinflammatory cytokines secreted by M1 macrophages within adipose tissue contribute to the upregulation of procoagulant factors such as tissue factor and plasminogen activator inhibitor-1 (PAI-1), leading to increased thrombin generation, enhanced platelet activation, reduced fibrinolysis, and an increased risk of thrombosis.
      This effect of obesity on venous thrombosis might be even more marked for NCPVT.
      • Bureau C.
      • Laurent J.
      • Robic M.A.
      • Christol C.
      • Guillaume M.
      • Ruidavets J.B.
      • et al.
      Central obesity is associated with non-cirrhotic portal vein thrombosis.
      Indeed, concentrations of inflammatory molecules known to activate endothelial cells rendering them more prothrombotic, including interleukin 6, are higher in the portal vein than in the radial artery of obese patients.
      • Fontana L.
      • Eagon J.C.
      • Trujillo M.E.
      • Scherer P.E.
      • Klein S.
      Visceral fat adipokine secretion is associated with systemic inflammation in obese humans.
      • Magkos F.
      • Fabbrini E.
      • Patterson B.W.
      • Eagon J.C.
      • Klein S.
      Portal vein and systemic adiponectin concentrations are closely linked with hepatic glucose and lipoprotein kinetics in extremely obese subjects.

      Acute cytomegalovirus infection

      Acute cytomegalovirus (CMV) infection is a cause of NCPVT.
      • Kelkar A.H.
      • Jacob K.S.
      • Yousif E.B.
      • Farrell J.J.
      Venous thromboembolism related to cytomegalovirus infection: a case report and literature review.
      Several explanations have been put forward. One plausible mechanism of CMV-induced thrombosis involves formation of antiphospholipid syndrome antibodies observed in patients in response to CMV infection,
      • Uthman I.W.
      • Gharavi A.E.
      Viral infections and antiphospholipid antibodies.
      resulting in a transient hypercoagulable state. In mouse models, the immunological pathways have been studied in greater detail. Through this process, 1 CMV-derived peptide of particular interest, TIFI, was found to be an analogue of human beta-2-glycoprotein I (b2GPI). Mice injected with TIFI developed antiphospholipid antibodies and measurable lupus anticoagulant activity, resulting in more thrombotic events than controls. Translation to humans was postulated by formation of anti-b2GPI antibodies against TIFI which would bind endogenous human b2GPI on the surface of endothelial cells, leading to activation of the coagulation cascade.
      • Kelkar A.H.
      • Jacob K.S.
      • Yousif E.B.
      • Farrell J.J.
      Venous thromboembolism related to cytomegalovirus infection: a case report and literature review.
      • Gharavi A.E.
      • Pierangeli S.S.
      • Espinola R.G.
      • Liu X.
      • Colden-Stanfield M.
      • Harris E.N.
      Antiphospholipid antibodies induced in mice by immunization with a cytomegalovirus-derived peptide cause thrombosis and activation of endothelial cells in vivo.
      The composition of the CMV envelope might also contribute to thrombosis. Indeed, the CMV surface contains the necessary procoagulant phospholipid for assembly of coagulation cascade proteins, thus favouring coagulation activation.
      • Pryzdial E.L.
      • Wright J.F.
      Prothrombinase assembly on an enveloped virus: evidence that the cytomegalovirus surface contains procoagulant phospholipid.
      A third alternative or complementary mechanism relies on the ability of CMV to directly infect endothelial cells and induce endothelial tissue factor expression, as well as adhesion of monocytes and neutrophils to infected endothelial cells.
      • Evans P.C.
      • Coleman N.
      • Wreghitt T.G.
      • Wight D.G.
      • Alexander G.J.
      Cytomegalovirus infection of bile duct epithelial cells, hepatic artery and portal venous endothelium in relation to chronic rejection of liver grafts.
      • Squizzato A.
      • Gerdes V.E.A.
      • Büller H.R.
      Effects of human cytomegalovirus infection on the coagulation system.

      Portosinusoidal vascular disease/Idiopathic portal hypertension

      In the absence of an identifiable cause, especially when liver test abnormalities and/or hepatic dysmorphism on imaging are present, further aetiologic work-up may also include liver biopsy. In fact, idiopathic portal hypertension may be frequently associated with NCPVT. Among patients with idiopathic portal hypertension, the prevalence of NCPVT is 13–46% and the annual probability of developing NCPVT is 9%.
      • Siramolpiwat S.
      • Seijo S.
      • Miquel R.
      • Berzigotti A.
      • Garcia-Criado A.
      • Darnell A.
      • et al.
      Idiopathic portal hypertension: natural history and long-term outcome.
      Such a high incidence might be due to reduced blood flow velocity secondary to the increase in intrahepatic resistance, together with portal vein wall abnormalities. Systematic analysis of abdominal imaging of patients with portosinusoidal vascular disease revealed that portal vein abnormalities are 3 times more common than in patients with cirrhosis.
      • Glatard A.-S.
      • Hillaire S.
      • D’Assignies G.
      • Cazals-Hatem D.
      • Plessier A.
      • Valla D.C.
      • et al.
      Obliterative portal venopathy: findings at CT imaging.
      However, more detailed analyses are lacking.

      Natural history

      BCS

      Clinical manifestations of BCS are extremely heterogeneous and vary from severe forms of acute liver failure to asymptomatic forms incidentally diagnosed when studying mild alterations of liver enzymes. Generally, the diagnosis is made after portal hypertension related complications, mainly ascites. Ascites (83%), hepatomegaly (67%) and abdominal pain (61%) were the most frequent clinical manifestations in a European cohort of 163 patients with BCS. In this cohort, 58% of the patients had oesophageal varices
      • Murad S.D.
      • Plessier A.
      • Hernandez-Guerra M.
      • Fabris F.
      • Eapen D.E.
      • Bahr M.J.
      • et al.
      Etiology, management, and outcome of the Budd-Chiari syndrome.
      at diagnosis. Although less common, acute liver failure may be the initial presentation in around 5% of cases.
      • Dilawari J.B.
      • Bambery P.
      • Chawla Y.
      • Kaur U.
      • Bhusnurmath S.R.
      • Malhotra H.S.
      • et al.
      Hepatic outflow obstruction (Budd-Chiari syndrome). Experience with 177 patients and a review of the literature.
      • Langlet P.
      • Escolano S.
      • Valla D.
      • Coste-zeitoun D.
      • Denie C.
      • Mallet A.
      • et al.
      Clinicopathological forms and prognostic index in Budd-Chiari syndrome.
      This wide range of clinical presentations probably correlates with both time to establishment and extension of the hepatic vein thrombosis. A slight and gradually-formed thrombosis may be accompanied by the development of hepatic venous collaterals able to, at least partially, decompress the portal venous system
      • Hadengue A.
      • Poliquin M.
      • Vilgrain V.
      • Belghiti J.
      • Degott C.
      • Erlinger S.
      • et al.
      The changing scene of hepatic vein thrombosis: recognition of asymptomatic cases.
      and maintain the patient asymptomatic; a clinical form described in up to 15% of cases.
      • Murad S.D.
      • Plessier A.
      • Hernandez-Guerra M.
      • Fabris F.
      • Eapen D.E.
      • Bahr M.J.
      • et al.
      Etiology, management, and outcome of the Budd-Chiari syndrome.
      • Hadengue A.
      • Poliquin M.
      • Vilgrain V.
      • Belghiti J.
      • Degott C.
      • Erlinger S.
      • et al.
      The changing scene of hepatic vein thrombosis: recognition of asymptomatic cases.
      Conversely, an extensive and rapidly constituted thrombosis of the hepatic veins may produce a severe form of liver failure with renal impairment, coagulopathy and death if not adequately treated. However, in many instances, despite the initial form of presentation being acute, signs of chronic liver disease are frequently found (i.e. alterations of liver morphology with atrophy/hypertrophy of different liver segments at imaging studies). This results from hepatic veins (HV) being frequently thrombosed at different time-points in a progressive manner. Obstruction of one hepatic vein can promote the development of intrahepatic or extrahepatic collateral circulation aimed at bypassing the occluded vein; thus, the patient may remain asymptomatic. However, imaging studies may reveal chronic morphological changes in the hepatic lobe drained by the occluded vein. If the patient is misdiagnosed and not adequately treated, recurrent thromboses in additional patent veins may happen, promoting severe hepatic congestion and appearance of symptoms, a clinical scenario potentially misdiagnosed as acute disease.
      Typical laboratory findings are aminotransferase elevation, reflecting subjacent necrosis and a decrease in prothrombin time in severe cases. The biochemical characteristic of ascites are its low cellularity and high protein content.
      • Mitchell M.C.
      • Boitnott J.K.
      • Kaufman S.
      • Cameron J.L.
      • Maddrey W.C.
      Budd-Chiari syndrome: etiology, diagnosis and management.
      The site of occlusion and clinical presentation seem to be different in patients from Western countries and Asia. In the West, the available data suggest that the most frequent site of thrombosis is the hepatic veins,
      • Murad S.D.
      • Plessier A.
      • Hernandez-Guerra M.
      • Fabris F.
      • Eapen D.E.
      • Bahr M.J.
      • et al.
      Etiology, management, and outcome of the Budd-Chiari syndrome.
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      whereas IVC obstruction (mainly from membrane or web) or combined IVC-HV obstruction prevails in most of the reported cases from Asia.
      • Okuda K.
      Inferior vena cava thrombosis at its hepatic portion (obliterative hepatocavopathy).
      However, recent data from India indicates this is changing,
      • Shukla A.
      • Parikh H.
      • Modi T.
      • Abraham P.
      • Kamble S.
      • Majumder D.
      Hepatic vein obstruction is the most common type of hepatic venous outflow obstruction regardless of socioeconomic status.
      and more cases with HV obstruction have been identified in the last decade. Regarding clinical presentation, the most frequent clinical manifestation in the West is ascites and impaired liver function
      • Murad S.D.
      • Plessier A.
      • Hernandez-Guerra M.
      • Fabris F.
      • Eapen D.E.
      • Bahr M.J.
      • et al.
      Etiology, management, and outcome of the Budd-Chiari syndrome.
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      compatible with an acute course. In China, it is often diagnosed when complications of portal hypertension arise; the presence of abdominal varices and lower limb oedema or ulcers
      • Qi X.
      • Han G.
      • Guo X.
      • De Stefano V.
      • Xu K.
      • Lu Z.
      • et al.
      Review article: the aetiology of primary Budd-Chiari syndrome – differences between the West and China.
      are more frequently described in Eastern patients.
      Concomitant splanchnic vein thrombosis and BCS has also being described although the incidence varies from 3.8–21% in epidemiological studies.
      • Ollivier-Hourmand I.
      • Allaire M.
      • Goutte N.
      • Morello R.
      • Chagneau-Derrode C.
      • Goria O.
      • et al.
      The epidemiology of Budd-Chiari syndrome in France.
      Another clinical finding in patients with chronic BCS is the presence of benign hepatic regenerative nodules.
      • Sonomura T.
      • Sato M.
      • Kishi K.
      • Terada M.
      • Shioyama Y.
      • Kimura M.
      • et al.
      Balloon-occluded retrograde transvenous obliteration for gastric varices: a feasibility study.
      • Flor N.
      • Zuin M.
      • Brovelli F.
      • Maggioni M.
      • Tentori A.
      • Sardanelli F.
      • et al.
      Regenerative nodules in patients with chronic Budd-Chiari syndrome: a longitudinal study using multiphase contrast-enhanced multidetector CT.
      Although the pathogenesis remains unclear, the coexistence of focal defects of portal perfusion and hypervascularized areas of preserved venous outflow may be involved in their development. The reported prevalence of these nodules is highly variable; they have been described in around 60–70% of patients in pathology studies,
      • Tanaka M.
      • Wanless I.R.
      Pathology of the liver in Budd-Chiari syndrome: portal vein thrombosis and the histogenesis of veno-centric cirrhosis, veno-portal cirrhosis, and large regenerative nodules.
      • Wanless I.R.
      Benign liver tumors.
      but in only 36% in an imaging study.
      • Vilgrain V.
      • Lewin M.
      • Vons C.
      • Denys A.
      • Valla D.
      • Flejou J.F.
      • et al.
      Hepatic nodules in Budd-Chiari syndrome: imaging features.
      Typically, benign nodules are small (under 3–4 cm in diameter), multiple (more than 10 lesions), hypervascularized and disseminated throughout the liver. Benign nodules may not only increase in number during follow-up, but may also increase in size.
      • Flor N.
      • Zuin M.
      • Brovelli F.
      • Maggioni M.
      • Tentori A.
      • Sardanelli F.
      • et al.
      Regenerative nodules in patients with chronic Budd-Chiari syndrome: a longitudinal study using multiphase contrast-enhanced multidetector CT.
      Histologically, benign nodules have the macro and microscopic alterations of focal nodular hyperplasia (FNH) and they may display a map-like pattern of glutamine synthase expression.
      • Kim H.
      • Nahm J.H.
      • Park Y.N.
      Budd-Chiari syndrome with multiple large regenerative nodules.
      However, because the underlying liver is not healthy, these lesions are usually called FNH-like lesions.
      • Vilgrain V.
      • Paradis V.
      • Van Wettere M.
      • Valla D.
      • Ronot M.
      • Rautou P.E.
      Benign and malignant hepatocellular lesions in patients with vascular liver diseases.
      Similar to traditional FNH, these benign lesions are usually homogeneous and hypervascular at imaging and the presence of a central scar can be found in nodules larger than 1 cm in diameter.
      • Maetani Y.
      • Itoh K.
      • Egawa H.
      • Haga H.
      • Sakurai T.
      • Nishida N.
      • et al.
      Benign hepatic nodules in Budd-Chiari syndrome: radiologic-pathologic correlation with emphasis on the central scar.
      However, other imaging characteristics may be different from those of typical FNH such as hyperintensity on T1-weighted and variable signal T2-weighted MR images. In addition, benign nodules may have washout on contrast-enhanced CT or MR imaging during portal venous and/or delayed phase.
      • Maetani Y.
      • Itoh K.
      • Egawa H.
      • Haga H.
      • Sakurai T.
      • Nishida N.
      • et al.
      Benign hepatic nodules in Budd-Chiari syndrome: radiologic-pathologic correlation with emphasis on the central scar.
      Less frequently, patients with chronic BCS may also develop hepatocellular adenomas
      • Sempoux C.
      • Paradis V.V.
      • Komuta M.
      • Wee A.
      • Calderaro J.
      • Balabaud C.
      • et al.
      Hepatocellular nodules expressing markers of hepatocellular adenomas in Budd-Chiari syndrome and other rare hepatic vascular disorders.
      and hepatocellular carcinoma (HCC).
      • Moucari R.
      • Rautou P.E.
      • Cazals-Hatem D.
      • Geara A.
      • Bureau C.
      • Consigny Y.
      • et al.
      Hepatocellular carcinoma in Budd-Chiari syndrome: characteristics and risk factors.
      Moucari et al. reported a 5-year cumulative incidence of HCC of 7% in a large cohort of patients with BCS.
      • Moucari R.
      • Rautou P.E.
      • Cazals-Hatem D.
      • Geara A.
      • Bureau C.
      • Consigny Y.
      • et al.
      Hepatocellular carcinoma in Budd-Chiari syndrome: characteristics and risk factors.
      A recent systematic review of 16 studies that reported HCC prevalence in BCS highlights the huge difference in the reported rates ranging from 2.0–46.2%. This is probably due, at least in part, to the heterogeneity of the studies included: geographical differences, dissimilar follow-up time (ranging from 4.5 to 11.6 years), diverse enrolment periods, different diagnostic tools and treatments, and different survivals rates.
      • Ren W.
      • Qi X.
      • Yang Z.
      • Han G.
      • Fan D.
      Prevalence and risk factors of hepatocellular carcinoma in Budd-Chiari syndrome: a systematic review.
      Risk factors for developing HCC are not well defined, although a higher prevalence has been described in patients with long-term IVC obstruction when compared to patients with isolated HV involvement.
      • Moucari R.
      • Rautou P.E.
      • Cazals-Hatem D.
      • Geara A.
      • Bureau C.
      • Consigny Y.
      • et al.
      Hepatocellular carcinoma in Budd-Chiari syndrome: characteristics and risk factors.
      • Paul S.B.
      • Sreenivas
      • Shalimar V.
      • Gamanagatti S.R.
      • Sharma H.
      • Dhamija E.
      • et al.
      Incidence and risk factors of hepatocellular carcinoma in patients with hepatic venous outflow tract obstruction.
      Frequently, HCC appears as a hypervascular lesion, which is heterogeneous at arterial phase and hypoechoic on portal and delayed phase. However, the radiological pattern of HCC in patients with BCS is heterogeneous and differential diagnosis with benign regenerative nodules remains a challenge. As previously mentioned, benign nodules in BCS may present the typical radiological appearance and vascular enhancement pattern of HCC in cirrhosis
      • Moucari R.
      • Rautou P.E.
      • Cazals-Hatem D.
      • Geara A.
      • Bureau C.
      • Consigny Y.
      • et al.
      Hepatocellular carcinoma in Budd-Chiari syndrome: characteristics and risk factors.
      • Cazals-Hatem D.
      • Vilgrain V.
      • Genin P.
      • Denninger M.H.
      • Durand F.
      • Belghiti J.
      • et al.
      Arterial and portal circulation and parenchymal changes in Budd-Chiari syndrome: a study in 17 explanted livers.
      • Brancatelli G.
      • Federle M.P.
      • Grazioli L.
      • Golfieri R.
      • Lencioni R.
      Large regenerative nodules in Budd-Chiari syndrome and other vascular disorders of the liver: CT and MR imaging findings with clinicopathologic correlation.
      and may increase in number and size over time as part of the natural history of the disease.
      • Flor N.
      • Zuin M.
      • Brovelli F.
      • Maggioni M.
      • Tentori A.
      • Sardanelli F.
      • et al.
      Regenerative nodules in patients with chronic Budd-Chiari syndrome: a longitudinal study using multiphase contrast-enhanced multidetector CT.
      A recent study specifically evaluating the radiological pattern of nodules in BCS, showed that 29% of benign lesions presented washout and up to 18% of benign lesions greater than 1 cm showed both washout and arterial phase hyperenhancement.
      • van Wettere Morgane
      • Purcell Yvonne
      • Bruno Onorina
      • Payancé Audrey
      • Plessier Aurélie
      • Rautou Pierre-Emmanuel
      • et al.
      Low specificity of washout to diagnose hepatocellular carcinoma in nodules showing arterial hyperenhancement in patients with Budd-Chiari syndrome.
      Therefore, HCC diagnosis in BCS is always a challenge and should never rely only on imaging criteria but should require histological confirmation. Although alpha-fetoprotein above a cut-off value of 15 nm/ml has been suggested as a useful biomarker for HCC in the setting of BCS,
      • Moucari R.
      • Rautou P.E.
      • Cazals-Hatem D.
      • Geara A.
      • Bureau C.
      • Consigny Y.
      • et al.
      Hepatocellular carcinoma in Budd-Chiari syndrome: characteristics and risk factors.
      it needs to be validated in larger cohorts. Surveillance for HCC in patients with chronic BCS is recommended.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      Although specific data are lacking, the most widely endorsed strategy is to perform ultrasound every 6 months, as in cirrhosis.
      HCC surveillance is recommended in patients with chronic BCS; ultrasound every 6 months is the most widely endorsed strategy.

      Recent NCPVT

      Recent NCPVT relates to the new occurrence of a thrombus in the portal venous axis in a patient with a previous patent portal vein. However, it may also occur in patients who exhibit partial PVT, where progression of the thrombus is noted. (Fig. 2). Various imaging modalities including colour Doppler ultrasound, computed tomography and MRI can be used to identify the presence of a thrombus, collateral circulation, or a dilated portal vein with accuracy rates ranging from 88–98%, with sensitivity and specificity of 80–100%. Using grey-scale ultrasound, an acute portal vein thrombus typically appears as heterogeneous, mainly hyperechoic material in the vessel lumen. The advantage of CT scan in this case relates to the improved ability to detect possible aetiologies or complications. On CT, acute PVT appears as increased attenuation and lack of enhancement, with or without inhomogeneities in portal venous and parenchymal hepatic perfusion. On MRI, the portal vein may appear with edge enhancement because of blood flow surrounding the thrombus or because of an inflammatory response of the venous wall. T1-weighted images may reveal an isointense thrombus compared to muscle, while T2 images may show a hyperintense signal.
      • Sarin S.K.
      • Philips C.A.
      • Kamath P.S.
      • Choudhury A.
      • Maruyama H.
      • Nery F.G.
      • et al.
      Toward a comprehensive new classification of portal vein thrombosis in patients with cirrhosis.
      • Jha R.C.
      • Khera S.S.
      • Kalaria A.D.
      Portal vein thrombosis: Imaging the spectrum of disease with an emphasis on MRI features.
      The anatomical degree of occlusion of these vessels may be total or partial. While the occlusion by the thrombus may not be complete in imaging studies, the haemodynamic consequence of partial thrombosis may be relevant. Indeed, the portal vein can be compared to a relatively rigid tubular structure
      • White D.
      • Coombe D.
      • Rezania V.
      • Tuszynski J.
      Building a 3D virtual liver: methods for simulating blood flow and hepatic clearance on 3D structures.
      in which the Poiseuille’s law predicts a decreased blood flow rate proportional to the radius elevated to the fourth power. Thus, a thrombus maintaining 20% of the vessel radius free will result in a >98% decrease of the flow rate (http://hyperphysics.phy-astr.gsu.edu/hbase/ppois2.html). Consequently, partial PVT occupying more than 80% of the lumen corresponds to a nearly complete obstruction (Fig. 3).
      Figure thumbnail gr2
      Fig. 2Evolution of portal pressure in portal vein thrombosis. In recent portal vein thrombosis, the rapid rise in portal pressure (red line) may be associated with clinically relevant consequences including bowel infarction and ascites. However, when portal pressure is already increased due to cirrhotic or non-cirrhotic portal hypertension (violet line), the haemodynamic consequences of portal vein thrombosis are attenuated due to the presence of collateral vessels. When complete recanalization occurs, portal pressure returns to baseline levels (dashed lines).
      Figure thumbnail gr3
      Fig. 3Application of Poiseuille’s law to blood flow. According to Poiseuille’s law, and assuming that the portal vein is a rigid tube, following portal vein thrombosis, blood flow decreases proportionally to the fourth power of the vessel radius. For example, in the presence of a thrombus occupying 80% of the lumen, the blood flow is decreased by 98.4%.
      At presentation, NCPVT may involve an extensive obstruction of the portal vein and its right and left branches, superior mesenteric vein, and splenic vein in approximately one-third of patients, while obstruction of the portal vein or of its 2 branches was found in 87% of cases. Moreover, only a single obstructed portal vein branch (with or without splenic or superior mesenteric vein obstruction) was reported in 12% of patients and the splenic or superior mesenteric vein were obstructed in 43% and 58% of patients, respectively.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      Acute abdominal pain may be the initial manifestation of recent NCPVT. However, the intensity is variable among patients. Recent NCPVT may also be completely asymptomatic, delaying diagnosis for long periods until portal cavernoma develops. Liver function test abnormalities are usually mild and transient. A systemic inflammatory response syndrome is often present in cases of recent NCPVT due to inflammation related to thrombosis but recognized local or systemic infection is identified in only 20% of these cases. Transient ascites, often of low abundance (hence clinically non-relevant unless infection develops) and visible only using ultrasound, CT or MRI, is present in half of patients with NCPVT.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      Mesenteric infarction is the most severe and immediate complication of recent NCPVT. Its 60% mortality rate is high in the absence of anticoagulant treatment. Extended resection of the small bowel is sometimes necessary and is associated with a significant risk of short bowel syndrome. Early initiation of anticoagulant therapy is associated with a very low incidence of this complication.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      The diagnosis of venous mesenteric infarction is difficult because the clinical, biological and radiological manifestations are not specific. Severe and persistent abdominal pain, despite full anticoagulant treatment, signs of organ failure (shock, renal failure, metabolic acidosis with elevated arterial lactate), abundant ascites, and/or the presence of blood in stool must evoke the diagnosis of mesenteric infarction. Type 2 diabetes has been identified as a risk factor for mesenteric infarction,
      • Elkrief L.
      • Corcos O.
      • Bruno O.
      • Larroque B.
      • Rautou P.-E.E.
      • Zekrini K.
      • et al.
      Type 2 diabetes mellitus as a risk factor for intestinal resection in patients with superior mesenteric vein thrombosis.
      with a recent study identifying 3 readily available criteria that predict irreversible intestinal ischaemic injury requiring resection in the setting of acute mesenteric ischaemia, namely organ failure, serum lactate levels >2 mmol/L and bowel loop dilation on CT scan. The presence of only one of these criteria was associated with a 40% risk of in irreversible intestinal ischaemic injury requiring resection at 2 months.
      • Nuzzo A.
      • Maggiori L.
      • Ronot M.
      • Becq A.
      • Plessier A.
      • Gault N.
      • et al.
      Predictive factors of intestinal necrosis in acute mesenteric ischemia: prospective study from an intestinal stroke center.
      The most common manifestations of chronic NCPVT are related to the presence of portal hypertension. Oesophageal varices may develop in nearly half of patients,
      • Noronha Ferreira C.
      • Seijo S.
      • Plessier A.
      • Silva-Junior G.
      • Turon F.
      • Rautou P.-E.P.-E.
      • et al.
      Natural history and management of esophagogastric varices in chronic noncirrhotic, nontumoral portal vein thrombosis.
      whereas other complications, including rectal or ectopic varices, large portosystemic collaterals and splenomegaly, frequently occur during the course of the disease. In general, hepatic function is preserved, in apparent contrast with obvious manifestations of portal hypertension. The most common complications are variceal bleeding, recurrent thrombosis and biliary complications. The risk of variceal bleeding is, as in cirrhosis, higher in medium or large varices and in the presence of red signs. The recurrence or extension of thrombosis may often be asymptomatic and is consequently not only underdiagnosed, but also poorly evaluated. In chronic NCPVT, ascites, hepatic encephalopathy, and bacterial infections are rare and most often transient. They occur preferentially as complications after gastrointestinal bleeding.
      • Rangari M.
      • Gupta R.
      • Jain M.
      • Malhotra V.
      • Sarin S.K.
      • et al.
      Hepatic dysfunction in patients with extrahepatic portal venous obstruction.
      If specifically looked for, minimal hepatic encephalopathy is frequently present in patients with extrahepatic portal vein obstruction
      • Mínguez B.
      • García-Pagán J.C.
      • Bosch J.
      • Turnes J.
      • Alonso J.
      • Rovira A.
      • et al.
      Noncirrhotic portal vein thrombosis exhibits neuropsychological and MR changes consistent with minimal hepatic encephalopathy.
      • Sharma P.
      • Sharma B.C.
      • Puri V.
      • Sarin S.K.
      Minimal hepatic encephalopathy in patients with extrahepatic portal vein obstruction.
      but their clinical impact is unknown. Hepatic nodules of the “HNF-like” type were found in 21% of adult patients with cavernomatous transformation of the portal vein when studied by MRI.
      • Marin D.
      • Galluzzo A.
      • Plessier A.
      • Brancatelli G.
      • Valla D.
      • Vilgrain V.
      Focal nodular hyperplasia-like lesions in patients with cavernous transformation of the portal vein: prevalence, MR findings and natural history.
      However, the risk of HCC is low. Cardiovascular complications such as hepatopulmonary syndrome
      • Kaymakoglu S.
      • Kahraman T.
      • Kudat H.
      • Demir K.
      • Cakaloglu Y.
      • Adalet I.
      • et al.
      Hepatopulmonary syndrome in noncirrhotic portal hypertensive patients.
      and pulmonary arterial hypertension
      • Saunders J.B.
      • Constable T.J.
      • Heath D.
      • Smith P.
      • Paton A.
      Pulmonary hypertension complicating portal vein thrombosis.
      have been reported.
      The most common complications of NCPVT are variceal bleeding, recurrent thrombosis and biliary problems, while cardiovascular complications have also been reported.
      Portal cavernoma cholangiopathy is characterized by abnormalities of the intrahepatic and extrahepatic bile ducts. It is the consequence of the extrinsic compression of bile ducts by porto-portal collateral veins (cavernoma) and/or of an ischaemic injury of the bile ducts by thrombosis of the venous plexus of the biliary tree.
      • Dhiman R.K.
      • Saraswat V.A.
      • Valla D.C.
      • Chawla Y.
      • Behera A.
      • Varma V.
      • Portal
      • et al.
      cavernoma cholangiopathy: consensus statement of a working party of the Indian national association for study of the liver.
      Biliary tract abnormalities may be observed at cholangio-MRI in 77–100% of patients. Portal cavernoma cholangiopathy is most frequently asymptomatic and can be associated with initially mild abnormalities of liver enzymes, particularly cholestatic markers alkaline phosphatase and gamma glutamyltransferase. Severe biliary manifestations including biliary colics, cholecystitis, obstructive jaundice, cholangitis, pancreatitis, are rare, occurring in only 5–30% of cases.
      • Dhiman R.K.
      • Saraswat V.A.
      • Valla D.C.
      • Chawla Y.
      • Behera A.
      • Varma V.
      • Portal
      • et al.
      cavernoma cholangiopathy: consensus statement of a working party of the Indian national association for study of the liver.
      • Asselah T.
      • Toole D.O.
      • Rufat P.
      • Zappa M.
      • Condat B.
      • Moreau R.
      • et al.
      Portal cavernoma-associated cholangiopathy: a clinical and MR cholangiography coupled with MR portography imaging study.
      • Llop E.
      • de Juan C.
      • Seijo S.
      • García-Criado Á.
      • Abraldes J.G.
      • Bosch J.
      • et al.
      Portal cholangiopathy: radiological classification and natural history.

      Diagnostic strategy: staging and prognosis

      BCS

      Due to heterogeneous clinical presentation, BCS should be suspected and discarded in any patient with acute or chronic liver disease, especially when its aetiology is unknown and/or if there is an underlying prothrombotic condition. The key feature for the diagnosis of BCS is to demonstrate obstruction of the hepatic venous outflow. Non-invasive imaging techniques (Doppler ultrasound, CT or MRI) are the mainstay of an adequate diagnosis. Doppler ultrasound, performed by an experienced operator, has a sensitivity of >75% and should be the first choice option.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      MRI and CT evaluation have a role in diagnostic confirmation or in the absence of an experienced ultrasound operator. Both CT and MRI are also useful for mapping intrahepatic and extrahepatic collateral networks, identifying associated PVT, and planning future treatment. Numerous imaging features have been described in patients with BCS.
      • Bargalló X.
      • Gilabert R.
      • Nicolau C.
      • García-Pagán J.C.
      • Bosch J.
      • Brú C.
      • et al.
      Sonography of the caudate vein: value in diagnosing Budd-Chiari syndrome.
      • Miller W.J.
      • Federle M.P.
      • Straub W.H.
      • Davis P.L.
      Budd-Chiari syndrome: imaging with pathologic correlation.
      Direct signs included visualization of the occluded veins, presence of endoluminal thrombus in HV, non-visualization of the HV, stagnant or inverted venous flow and collateral networks; these signs are more frequently found in acute BCS. Chronic forms are usually associated with indirect signs such as hypertrophy of the caudate lobe, and a caudate vein >3 mm and concomitant atrophic lobes, dysmorphic liver, parenchymal heterogeneity, heterogeneous enhancement and benign regenerative nodules. CT and MRI can also identify a rapid clearance of contrast from the caudate lobe and patchy hepatic enhancement due to uneven portal perfusion. Acutely occluded veins demonstrate no enhancement following contrast administration on CT scan, whereas on MRI they display hyperintensity on spin echo and signal void on gradient echo sequences. A multidisciplinary approach is essential, as diagnostic efficiency is augmented when the radiologist is aware of the suspected clinical diagnosis.
      Liver damage in BCS is variable, and the histological changes (congestion, coagulative necrosis or simple loss of hepatocytes without inflammatory infiltrates and/or fibrosis) are not pathognomonic and do not reflect the severity of the disease. Therefore, liver biopsy is not usually necessary for the diagnosis. The sole scenario when biopsy is necessary is to confirm BCS due to small intrahepatic vein obstruction in the presence of preserved large veins on imaging techniques.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      Similarly, hepatic venography is only recommended if diagnosis remains uncertain despite the above investigations and classically reveals a spiderweb pattern, formed by a rich collateral circulation.

      BCS staging: prognostic scores

      The prognosis of BCS has dramatically changed in the last decade, with an improvement in survival as a consequence of a better management based on anticoagulation, transjugular intrahepatic portosystemic shunt (TIPS) and liver transplantation.
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      Better outcomes may also be related to a higher degree of suspicion leading to early stage diagnosis and consequently superior treatment response. Indeed, 3-year mortality in the 60s reached 90%,
      • Zeitoun G.
      • Escolano S.
      • Hadengue A.
      • Azar N.
      • El Younsi M.
      • Mallet A.
      • et al.
      Outcome of Budd-Chiari syndrome: a multivariate analysis of factors related to survival including surgical portosystemic shunting.
      whereas the current expected 5-year survival is above 80%.
      • Murad S.D.
      • Plessier A.
      • Hernandez-Guerra M.
      • Fabris F.
      • Eapen D.E.
      • Bahr M.J.
      • et al.
      Etiology, management, and outcome of the Budd-Chiari syndrome.
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      Information regarding the prognosis of BCS relies mostly on retrospective studies.
      • Hadengue A.
      • Poliquin M.
      • Vilgrain V.
      • Belghiti J.
      • Degott C.
      • Erlinger S.
      • et al.
      The changing scene of hepatic vein thrombosis: recognition of asymptomatic cases.
      • Zeitoun G.
      • Escolano S.
      • Hadengue A.
      • Azar N.
      • El Younsi M.
      • Mallet A.
      • et al.
      Outcome of Budd-Chiari syndrome: a multivariate analysis of factors related to survival including surgical portosystemic shunting.
      • Garcia-Pagan J.C.
      • Heydtmann M.
      • Raffa S.
      • Plessier A.
      • Murad S.
      • Fabris F.
      • et al.
      TIPS for Budd-Chiari syndrome: long-term results and prognostics factors in 124 patients.
      • Darwish Murad S.
      • Valla D.C.
      • de Groen P.C.
      • Zeitoun G.
      • Hopmans J.A.M.
      • Haagsma E.B.
      • et al.
      Determinants of survival and the effect of portosystemic shunting in patients with Budd-Chiari syndrome.
      • Plessier A.
      • Sibert A.
      • Consigny Y.
      • Hakime A.
      • Zappa M.
      • Denninger M.-H.H.
      • et al.
      Aiming at minimal invasiveness as a therapeutic strategy for Budd-Chiari syndrome.
      The largest prospective multicentre cohort of consecutive diagnosed patients with BCS comes from the European registry EN-Vie. 157 patients were prospectively diagnosed during a 2-year period and followed up for almost 5 years, reporting a 5-year survival of 85%.
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      As shown in Table 3, there are several parameters or combinations of them that are used to predict BCS prognosis. Liver function tests such as Child-Pugh
      • Tang T.J.
      • Batts K.P.
      • de Groen P.C.
      • van Hoek B.
      • Haagsma E.B.
      • Hop W.C.
      • et al.
      The prognostic value of histology in the assessment of patients with Budd-Chiari syndrome.
      and model for end-stage liver disease score
      • Darwish M.S.
      • Kim W.R.
      • de Groen P.C.
      • Kamath P.S.
      • Malinchoc M.
      • Valla D.C.
      • et al.
      Can the model for end-stage liver disease be used to predict the prognosis in patients with Budd-Chiari syndrome?.
      are able to predict outcomes in BCS. BCS-specific prognostic scores
      • Langlet P.
      • Escolano S.
      • Valla D.
      • Coste-zeitoun D.
      • Denie C.
      • Mallet A.
      • et al.
      Clinicopathological forms and prognostic index in Budd-Chiari syndrome.
      • Zeitoun G.
      • Escolano S.
      • Hadengue A.
      • Azar N.
      • El Younsi M.
      • Mallet A.
      • et al.
      Outcome of Budd-Chiari syndrome: a multivariate analysis of factors related to survival including surgical portosystemic shunting.
      • Darwish Murad S.
      • Valla D.C.
      • de Groen P.C.
      • Zeitoun G.
      • Hopmans J.A.M.
      • Haagsma E.B.
      • et al.
      Determinants of survival and the effect of portosystemic shunting in patients with Budd-Chiari syndrome.
      are useful for predicting transplant-free survival and invasive therapy-free survival and have been externally validated
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      • Rautou P.E.
      • Moucari R.
      • Escolano S.
      • Cazals-Hatem D.
      • Denie C.
      • Chagneau-Derrode C.
      • et al.
      Prognostic indices for Budd-Chiari syndrome: valid for clinical studies but insufficient for individual management.
      . The BCS-TIPS prognostic index score was developed to identify patients that would not tolerate TIPS. Indeed, this score identifies patients with poor outcomes despite TIPS, suggesting that liver transplant may be a better alternative. Nevertheless, despite showing a statistically significant association with survival and, permitting comparison among different cohorts, none of the BCS-specific prognostic indices has excellent discriminative capacity and none can be used to guide individualized management.
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      • Rautou P.E.
      • Moucari R.
      • Escolano S.
      • Cazals-Hatem D.
      • Denie C.
      • Chagneau-Derrode C.
      • et al.
      Prognostic indices for Budd-Chiari syndrome: valid for clinical studies but insufficient for individual management.
      Table 3BCS-specific prognostic index.
      ScoreFormulaCut-offPredicted survival rateReference
      Clichy PI(Ascites score
      Ascites score: 1, absent with free sodium intake and no diuretic agents; 2, easy to control with sodium restriction or diuretic agents; and 3, resistant to this treatment because of hyponatremia or functional renal failure.
       × 0.75) + (Child-Pugh score × 0.28) + (age × 0.037) + (creatinine × 0.0036)
      5.4 (range from 3.4 to 9.1)At 5 yr

      ≤5.4: 95%

      >5.4: 65%
      • Zeitoun G.
      • Escolano S.
      • Hadengue A.
      • Azar N.
      • El Younsi M.
      • Mallet A.
      • et al.
      Outcome of Budd-Chiari syndrome: a multivariate analysis of factors related to survival including surgical portosystemic shunting.
      New Clichy PI0.95 × ascites score + 0.35 × Child-Pugh score + 0.0 47 × age + 0.0045 × serum creatinine + 2.2 × type III
      Type III’ is a binary variable coded as 1 for patients with clinicopathological findings of acute injury superimposed on chronic lesions, and 0 for the other patients. BCS, Budd-Chiari syndrome.
      —2.6
      5.1 (range from 2.0 to 9.7)At 5 y

      <5.1: 100%

      ≥5.1: 65%
      • Langlet P.
      • Escolano S.
      • Valla D.
      • Coste-zeitoun D.
      • Denie C.
      • Mallet A.
      • et al.
      Clinicopathological forms and prognostic index in Budd-Chiari syndrome.
      Rotterdam BCS index1.27 × encephalopathy + 1.04 × ascites + 0.72 × prothrombin time + 0.004 × bilirubinClass I: 0–1.1

      Class II: 1.1–1.5

      Class III: ≥1.5 (range from 0.02 to 4.03)
      At 5 y

      Class I: 89%

      Class II: 74%

      Class III: 42%
      • Darwish Murad S.
      • Valla D.C.
      • de Groen P.C.
      • Zeitoun G.
      • Hopmans J.A.M.
      • Haagsma E.B.
      • et al.
      Determinants of survival and the effect of portosystemic shunting in patients with Budd-Chiari syndrome.
      TIPS-BCS PIAge (years) × 0.08 + bilirubin (mg/dl) × 0.16 + international normalized ratio (INR) × 0.6371-year OLT-free survival ≤7 95%

      >7 12%
      • Garcia-Pagan J.C.
      • Heydtmann M.
      • Raffa S.
      • Plessier A.
      • Murad S.
      • Fabris F.
      • et al.
      TIPS for Budd-Chiari syndrome: long-term results and prognostics factors in 124 patients.
      BCS-intervention-free survival prognostic scoreAscites [yes = 1, no = 0]*1.675 + ln creatinine [µmol/L]*0.613 + ln bilirubin [µmol/ L]*0.440)Interval I: ≤5

      Interval 2: 5–6

      Interval 3: ≥6
      Intervention-free survival

      Interval I: 78.3%

      Interval 2: 27.8%

      Interval 3: 6.8%
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      BCSurvival scoreAge/10*0.370 + ln creatinine [µmol/L]*0.809 + ln bilirubin [µmol/L]*0.496)Interval I: ≤7

      Interval 2: 7–8

      Interval 3 ≥8
      Probability survival

      Interval I: 87.5%

      Interval 2: 63.3%

      Interval 3: 42.9%
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      a Ascites score: 1, absent with free sodium intake and no diuretic agents; 2, easy to control with sodium restriction or diuretic agents; and 3, resistant to this treatment because of hyponatremia or functional renal failure.
      b Type III’ is a binary variable coded as 1 for patients with clinicopathological findings of acute injury superimposed on chronic lesions, and 0 for the other patients. BCS, Budd-Chiari syndrome.
      Histological features and aminotransferase levels have also been evaluated as prognostic factors. Pathological lesions are not useful for predicting outcomes in BCS due to heterogeneity of the lesions and the non-homogeneous pattern.
      • Langlet P.
      • Escolano S.
      • Valla D.
      • Coste-zeitoun D.
      • Denie C.
      • Mallet A.
      • et al.
      Clinicopathological forms and prognostic index in Budd-Chiari syndrome.
      • Tang T.J.
      • Batts K.P.
      • de Groen P.C.
      • van Hoek B.
      • Haagsma E.B.
      • Hop W.C.
      • et al.
      The prognostic value of histology in the assessment of patients with Budd-Chiari syndrome.
      The largest analysis of Western patients with acute liver failure due to BCS included 19 patients registered in the US Acute Liver Failure Study Group and showed poor outcomes. They evaluated patients from 1999–2015 and demonstrated that increased aminotransferases (aspartate aminotransferase [AST], alanine aminotransferase [ALT]) were predictors of poor outcomes, marking the severity and/or acuity of the damage. While mortality was high at 58%, cases reported after 2010 had a better outcome due to the improved diagnosis and management.
      • Parekh J.
      • Matei V.M.
      • Canas-Coto A.
      • Friedman D.
      • Lee W.M.
      Acute Liver Failure Study Group. Budd-chiari syndrome causing acute liver failure: a multicenter case series.
      Moreover, in a retrospective study including 96 patients with primary BCS, ALT ≥5 × the upper limit of normal was associated with more severe clinical presentation, a higher Child-Pugh score, Clichy score, Rotterdam BCS score, and higher mortality. However high levels of ALT that decrease rapidly can be considered a marker of imminent improvement in liver function, reflecting a good outcome.
      • Rautou P.
      • Moucari R.
      • Cazals-Hatem D.
      • Escolano S.
      • Denié C.
      • Douarin L.
      • et al.
      Levels and initial course of serum alanine aminotransferase can predict outcome of patients with Budd-Chiari syndrome.

      NCPVT staging and prognosis

      Following recent thrombosis of the portal vein, and in the absence of recanalization, a network of porto-portal collateral veins, defined as cavernoma, may develop within a few weeks. The presence of cavernoma is a direct consequence of NCPVT, and this term suggests that thrombosis of the portal vein was not recent. Moreover, the presence of portal vein cavernoma per se does not imply chronicity of this condition. Portosystemic collateral veins may also develop and contribute to the complications of portal hypertension. An accurate diagnosis and staging of NCPVT is important, because it represents the basis for its management. In addition, the purpose of correct classification is to guide therapeutic decisions aimed at obtaining recanalization and at preventing the extension and/or the relapse of PVT and of possible complications. Given the complexity of the clinical picture related to NCPVT, including a high number of possible aetiologic factors and different degrees of obstruction and extension of the thrombosis, the need for an individualized yet precise approach appears crucial. For these reasons, several classification models have been proposed. One of the most frequently used systems is, however, only based on localization and extension of thrombosis.
      • Yerdel M.A.
      • Gunson B.
      • Mirza D.
      • Karayalcin K.
      • Olliff S.
      • Buckels J.
      • et al.
      Portal vein thrombosis in adults undergoing liver transplantation: risk factors, screening, management, and outcome.
      More recently, Sarin et al. have proposed a new classification,
      • Sarin S.K.
      • Philips C.A.
      • Kamath P.S.
      • Choudhury A.
      • Maruyama H.
      • Nery F.G.
      • et al.
      Toward a comprehensive new classification of portal vein thrombosis in patients with cirrhosis.
      which takes into account not only the location and extension of the thrombus, but also the mode of onset and diagnosis of the thrombosis (recent or chronic), the type of presentation (signs and symptoms), and the possible underlying liver disease (cirrhosis or normal liver). The purpose of this classification (Box 1) is to serve as a starting point for the uniform reporting of NCPVT, to better define clinical endpoints, and to compare future studies. The prospective evaluation of this classification should allow a risk stratification to apply therapeutic or preventive strategies.
      Figure thumbnail gr5
      Fig. 4Minimal Invasiveness therapeutic strategy for BCS. BCS, Budd-Chiari syndrome.

      Treatment and management of complications

      Stepwise treatment for BCS

      Unfortunately, randomized clinical trials comparing different treatment options for BCS are still lacking. Recommendations are based on clinical experience, retrospective studies and expert consensus. Based on these, the current treatment strategy of BCS relies on progressively escalating invasiveness. The first step is based on medical management aimed at treating the complications of portal hypertension and the underlying disease. If no improvement or even further deterioration in BCS symptoms is observed, the next step is devoted to correcting hepatic venous outflow obstruction as described below.
      The current treatment strategy for BCS involves a progressive escalation in invasiveness.

      Portal hypertension complications

      Because no specific studies exist in patients with Budd-Chiari, the current recommendation is to treat and prevent complications related to portal hypertension, as recommended for patients with cirrhosis.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      • De Franchis R.
      • Abraldes J.G.
      • Bajaj J.
      • Berzigotti A.
      • Bosch J.
      • Burroughs A.K.
      • et al.
      Expanding consensus in portal hypertension: report of the Baveno VI Consensus Workshop: stratifying risk and individualizing care for portal hypertension.

      Treatment of the underlying disease

      The underlying disease should be promptly diagnosed and specifically treated. This usually requires a dedicated multidisciplinary team of hepatologist, haematologist and specialists in systemic disorders. As an example, early recognition and adequate treatment of underlying MPN, PNH or Behçet’s syndrome may markedly influence the outcome of the patients and/or prevent thrombosis progression.
      • Chagneau-Derrode C.
      • Roy L.
      • Guilhot J.
      • Gloria O.
      • Ollivier-Hourmand I.
      • Bureau C.
      • et al.
      Impact of cytoreductive therapy on the outcome of patients with myeloproliferative neoplasms and hepatosplanchnic vein thrombosis.
      • Desbois A.C.
      • Rautou P.E.
      • Biard L.
      • Belmatoug N.
      • Wechsler B.
      • Resche-Rigon M.
      • et al.
      Behcet’s disease in Budd-Chiari syndrome.

      Correcting hepatic venous outflow obstruction: Stepwise treatment

      Although there are small retrospective studies showing good outcomes after initial management with surgical shunts,
      • Orloff M.J.
      • Isenberg J.I.
      • Wheeler H.O.
      • Daily P.O.
      • Girard B.
      Budd-Chiari syndrome revisited: 38 years’ experience with surgical portal decompression.
      • Montano-Loza A.J.
      • Tandon P.
      • Kneteman N.
      • Bailey R.
      • Bain V.G.
      Rotterdam score predicts early mortality in Budd-Chiari syndrome, and surgical shunting prolongs transplant-free survival.
      • Bachet J.
      • Condat B.
      • Consigny Y.
      • Belghiti J.
      • Valla D.
      Long-term portosystemic shunt patency as a determinant of outcome in Budd-Chiari syndrome.
      most recommendations support the gradual escalation of management from least to most invasive (Fig. 4).
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      • Plessier A.
      • Sibert A.
      • Consigny Y.
      • Hakime A.
      • Zappa M.
      • Denninger M.-H.H.
      • et al.
      Aiming at minimal invasiveness as a therapeutic strategy for Budd-Chiari syndrome.
      • De Franchis R.
      • Abraldes J.G.
      • Bajaj J.
      • Berzigotti A.
      • Bosch J.
      • Burroughs A.K.
      • et al.
      Expanding consensus in portal hypertension: report of the Baveno VI Consensus Workshop: stratifying risk and individualizing care for portal hypertension.
      Figure thumbnail gr6
      Fig. 5Patient with NCPVT. (A) Before and (B) after angioplasty restoring physiological portal blood flow. NCPVT, non-cirrhotic non-tumoral portal vein thrombosis

      Anticoagulation

      The stepped approach starts with medical treatment using anticoagulation, in an attempt to achieve recanalization but mainly to prevent thrombosis progression, together with the treatment or prevention of the complications of portal hypertension. Anticoagulation should be administered to all patients with BCS, even those without an underlying prothrombotic disorder or in those that are initially asymptomatic. Long-term anticoagulation, started as soon as possible after the diagnosis, achieves a 5-year intervention free survival with disease-control in 25–30% of patients, particularly in mild/moderate cases. This is observed both in Western and in Asian patients.
      • Murad S.D.
      • Plessier A.
      • Hernandez-Guerra M.
      • Fabris F.
      • Eapen D.E.
      • Bahr M.J.
      • et al.
      Etiology, management, and outcome of the Budd-Chiari syndrome.
      • Mitchell M.C.
      • Boitnott J.K.
      • Kaufman S.
      • Cameron J.L.
      • Maddrey W.C.
      Budd-Chiari syndrome: etiology, diagnosis and management.
      • Dhiman R.K.
      • Saraswat V.A.
      • Valla D.C.
      • Chawla Y.
      • Behera A.
      • Varma V.
      • Portal
      • et al.
      cavernoma cholangiopathy: consensus statement of a working party of the Indian national association for study of the liver.
      • Rautou P.
      • Moucari R.
      • Cazals-Hatem D.
      • Escolano S.
      • Denié C.
      • Douarin L.
      • et al.
      Levels and initial course of serum alanine aminotransferase can predict outcome of patients with Budd-Chiari syndrome.
      Either low molecular weight heparin (LMWH) or vitamin K antagonists (VKAs) are the treatment of choice to initiate anticoagulation. Therapeutic doses of LMWH should be given together with VKAs until the therapeutic range (international normalized ratio between 2–3) is achieved. It should be noted that long-term LMWH cannot be used in patients with severe renal failure and that unfractionated heparin should be avoided due to the high incidence of heparin-induced thrombocytopenia in patients with BCS.
      • Plessier A.
      • Sibert A.
      • Consigny Y.
      • Hakime A.
      • Zappa M.
      • Denninger M.-H.H.
      • et al.
      Aiming at minimal invasiveness as a therapeutic strategy for Budd-Chiari syndrome.
      Direct oral anticoagulants (DOACs) are used to treat thrombotic diseases in other vascular areas and, although there are currently only a few reports,
      • Taguchi E.
      • Koyama J.
      • Kajiwara M.
      • Inoue M.
      • Horibata Y.
      • Nishigami K.
      • et al.
      Successful vascular intervention without embolic complications in Budd-Chiari syndrome.
      this option can also be considered in patients with BCS and normal liver function. However, DOACs are not registered for this indication, and therefore, if used, this must be done with caution, especially in patients with renal failure.

      Restoring hepatic venous outflow

      Thrombolysis

      The experience of thrombolysis in BCS is limited. Recombinant tissue plasminogen activator, streptokinase or urokinase have been used. These agents can be instilled through a peripheral vein or locally after catheterization of the thrombosed vein. There are no studies comparing the efficacy of local vs. systemic infusion. No systematic reviews have been published to evaluate the efficacy and risks of thrombolysis in BCS. A narrative review by Sharma et al. indicates that the best results are achieved in patients with a recent and incomplete thrombosis who are treated with local and early infusion combined with another interventional procedure (e.g. angioplasty, stenting) to restore venous outflow.
      • Sharma S.
      • Texeira A.
      • Texeira P.
      • Elias E.
      • Wilde J.
      • Olliff S.P.
      Pharmacological thrombolysis in Budd Chiari syndrome: a single centre experience and review of the literature.
      In the published series, the bleeding complications of thrombolysis were major, with 2 fatal outcomes.
      • Smalberg J.H.
      • Spaander M.V.
      • Jie K.S.
      • Pattynama P.M.
      • van Buuren H.R.
      • van den B.B.
      • et al.
      Risks and benefits of transcatheter thrombolytic therapy in patients with splanchnic venous thrombosis.
      Hence, this therapeutic option is contraindicated in patients with a potentially haemorrhagic condition, or patients who have had an invasive procedure, including paracentesis, in the previous 24 hours. In summary, thrombolysis should only be attempted in select cases with acute or sub-acute BCS, at experienced centres.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.

      Percutaneous angioplasty

      In some instances, BCS is due to partial or segmental stenosis in the cranial part of the HV or at the suprahepatic IVC.
      • Valla D.
      • Hadengue A.
      • el Younsi M.
      • Azar N.
      • Zeitoun G.
      • Boudet M.J.
      • et al.
      Hepatic venous outflow block caused by short-length hepatic vein stenoses.
      In these cases, percutaneous transluminal angioplasty is an effective and safe approach for restoring the physiological hepatic outflow, with or without stenting. This makes it mandatory to try and identify these patients once the diagnosis of BCS is established. In European patients, segmental stenosis is only found in a small percentage of patients and therefore, this technique only benefits a small proportion of patients with BCS (10% in the EN-Vie cohort).
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      In Asia, where IVC obstruction predominates, the reported applicability is much higher, and combining angioplasty and stenting can achieve patency in >80% of patients at 5 years.
      • Han G.
      • Qi X.
      • Zhang W.
      • He C.
      • Yin Z.
      • Wang J.
      • et al.
      Percutaneous recanalization for Budd-Chiari syndrome: an 11-year retrospective study on patency and survival in 177 Chinese patients from a single center.
      Angioplasty is the first step to restore hepatic outflow. However, post-angioplasty re-stenosis may occur, necessitating subsequent angioplasties. While stenting may reduce re-stenosis, stent misplacement may make future TIPS or liver transplantation more challenging. In the European cohort, 22 patients underwent angioplasty, 13 patients were treated with angioplasty, 7 with thrombolysis and 2 with both as the first invasive treatment. In 6 of these 22 patients, a vascular stent was placed at the time of angioplasty and in total 14 patients (64%) required further treatment (TIPS in 12 and orthotopic liver transplant [OLT] in 2 patients).
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      A recent retrospective study performed in China suggests higher efficacy and long-term patency of retrievable stents (retrieved after a median of 15 days). However, these results should be interpreted cautiously due to the short duration of stenting and because a significant number of patients with retrievable stents exhibited acute thrombosis (66.7% vs. 2.4%, p = 0.0004) and hence, were additionally treated with thrombolysis.
      • Bi Y.
      • Chen H.
      • Ding P.
      • Ren J.
      • Han X.
      Comparison of retrievable stents and permanent stents for Budd-Chiari syndrome due to obstructive inferior vena cava.

      Derivative techniques

      When the aforementioned treatments are not possible or fail to solve the obstruction of the hepatic blood flow, the portal system can be converted into an outflow tract by derivative techniques.
      Before the 90s, surgical shunts were the only derivative technique available. Mesocaval shunt was the most frequent shunt used, preferred to the porto-caval side-to-side shunt since it is easier to perform in the setting of caudate lobe hypertrophy.
      • Shaked A.
      • Goldstein R.M.
      • Klintmalm G.B.
      • Drazan K.
      • Husberg B.
      • Busuttil R.W.
      Portosystemic shunt versus orthotopic liver transplantation for the Budd-Chiari syndrome.
      In the setting of IVC compression, the presence of an infrahepatic caval pressure >20 mmHg or a gradient between it and the right atrium of 15 mmHg, are predictive of inadequate shunt function, unless the stenosis/compression of the IVC is simultaneously corrected.
      • Shaked A.
      • Goldstein R.M.
      • Klintmalm G.B.
      • Drazan K.
      • Husberg B.
      • Busuttil R.W.
      Portosystemic shunt versus orthotopic liver transplantation for the Budd-Chiari syndrome.
      In cases when the obstructed IVC cannot be bypassed, a meso-atrial shunt may be an alternative.
      • Klein A.S.
      • Molmenti E.P.
      Surgical treatment of Budd-Chiari syndrome.
      Decompressing the cava together with the portal venous system through a meso-cavo-atrial shunt has been proposed as a better alternative to a meso-atrial shunt.
      • Chen H.
      • Zhang F.
      • Ye Y.
      • Cheng Y.
      • Chen Y.
      Long-term follow-up study and comparison of meso-atrial shunts and meso-cavo-atrial shunts for treatment of combined Budd-Chiari syndrome.
      Overall, surgical shunts are associated with significant morbidity-mortality, and have not demonstrated a clear survival advantage.
      • Zeitoun G.
      • Escolano S.
      • Hadengue A.
      • Azar N.
      • El Younsi M.
      • Mallet A.
      • et al.
      Outcome of Budd-Chiari syndrome: a multivariate analysis of factors related to survival including surgical portosystemic shunting.
      • Langlet P.
      • Valla D.
      Is surgical portosystemic shunt the treatment of choice in Budd-Chiari syndrome?.
      However, in patients surviving surgery in whom the shunt remains patent, the outcome is excellent.
      • Bachet J.
      • Condat B.
      • Consigny Y.
      • Belghiti J.
      • Valla D.
      Long-term portosystemic shunt patency as a determinant of outcome in Budd-Chiari syndrome.
      • Panis Y.
      • Belghiti J.
      • Valla D.
      • Benhamou J.P.
      • Fekete F.
      Portosystemic shunt in Budd-Chiari syndrome: long-term survival and factors affecting shunt patency in 25 patients in Western countries.
      Since the 90s, surgical shunts have been replaced by the less-invasive TIPS in most places. TIPS has been demonstrated to be more effective in maintaining patency and is associated with lower morbidity and mortality than surgery in patients with failure on medical treatment or when recanalization has failed.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      • De Franchis R.
      • Abraldes J.G.
      • Bajaj J.
      • Berzigotti A.
      • Bosch J.
      • Burroughs A.K.
      • et al.
      Expanding consensus in portal hypertension: report of the Baveno VI Consensus Workshop: stratifying risk and individualizing care for portal hypertension.
      However, very good outcomes have been reported in selected patients with BCS treated with surgical portal decompression performed early after diagnosis.
      • Orloff M.J.S.
      • Daily P.O.
      • Orloff S.L.
      • Girard B.
      • Orloff M.S.
      A 27-year experience with surgical treatment of Budd-Chiari syndrome.
      Moreover, in cases with HV and concomitant IVC thrombosis or severe compression of the IVC by an enlarged liver, as mentioned, the traditional meso-caval surgical shunt may be ineffective and TIPS becomes a more feasible option.
      • Bachet J.
      • Condat B.
      • Consigny Y.
      • Belghiti J.
      • Valla D.
      Long-term portosystemic shunt patency as a determinant of outcome in Budd-Chiari syndrome.
      • Hemming A.W.
      • Langer B.
      • Greig P.
      • Taylor B.R.
      • Adams R.
      • Heathcote E.J.
      Treatment of Budd-Chiari syndrome with portosystemic shunt or liver transplantation.
      TIPS primary patency rate using PTFE-covered stents is 67% at 2-year follow-up.
      • Garcia-Pagan J.C.
      • Heydtmann M.
      • Raffa S.
      • Plessier A.
      • Murad S.
      • Fabris F.
      • et al.
      TIPS for Budd-Chiari syndrome: long-term results and prognostics factors in 124 patients.
      However, TIPS should be performed in experienced centres because of the increased difficulty and morbidity associated with the technique in patients with NCPVT compared to those with cirrhosis. Indeed, a transcaval approach for the portal vein puncture may be needed in up to 60% of patients given inaccessibility of the hepatic vein.
      • Garcia-Pagan J.C.
      • Heydtmann M.
      • Raffa S.
      • Plessier A.
      • Murad S.
      • Fabris F.
      • et al.
      TIPS for Budd-Chiari syndrome: long-term results and prognostics factors in 124 patients.
      • Hayek G.
      • Ronot M.
      • Plessier A.
      • Sibert A.
      • Abdel-Rehim M.
      • Zappa M.
      • et al.
      Long-term outcome and analysis of dysfunction of transjugular intrahepatic portosystemic shunt placement in chronic primary Budd-Chiari syndrome.
      In a European prospective cohort of 157 patients with BCS, after 5 years of follow-up, 40% of patients required TIPS because of failed medical treatment, most of them (73%) during the first 6 months after diagnosis. In this cohort, 5-year survival without liver transplantation was 72%.
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      The use of TIPS at an earlier time point in the recommended stepwise management has recently been suggested.
      • Mancuso A.
      In favour of early intervention for Budd-Chiari syndrome.
      However, there are no direct data supporting this recommendation. In addition, in the study by Seijo et al.
      • Seijo S.
      • Plessier A.
      • Hoekstra J.
      • Era A.D.
      • Mandair D.
      • Rifai K.
      • et al.
      Good long-term outcome of Budd-Chiari syndrome with a step-wise management.
      in the 62 patients receiving TIPS following the stepwise strategy, the median time from BCS diagnosis to TIPS was 1 month (range 0–38 months). Interestingly, no differences in survival were observed in patients receiving TIPS during or after the first months following diagnosis, with similar results observed when the cut-off time was 3 or 6 months after diagnosis. These results suggest that the stepwise strategy is effective and safe provided that patients are followed closely and TIPS is implemented soon after prior treatment ceases to confer an improvement or clinical deterioration hastens. no improvement, hastening further deterioration. In addition, instead of using TIPS for all patients with symptomatic BCS, following this strategy will prevent a significant number of patients from experiencing the potential side-effects of TIPS in cases of limited benefit. Evaluation of some criteria 2 weeks after treatment initiation has been proposed to identify the optimal point at which to move a given patient along the treatment algorithm (Table 4). However, this is always a challenge.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      This is why these patients are best managed in referral centres.
      Table 4Evaluation of the response to treatment (Adapted from Plessier A et al.
      • Plessier A.
      • Sibert A.
      • Consigny Y.
      • Hakime A.
      • Zappa M.
      • Denninger M.-H.H.
      • et al.
      Aiming at minimal invasiveness as a therapeutic strategy for Budd-Chiari syndrome.
      .
      Ongoing treatment response (2 weeks)Complete treatment response
      AscitesYesNo clinically detectable
      Without diuretic treatment or low dose (spironolactone 75 mg/d or furosemide 40 mg/d).
       CreatinineNormal or decreasingnormal
       SodiumNormal or increasingnormal
       Balance (water-Na)negative
       NaCl intakemoderatemoderate
      Factor V levelincreasingabove 40% of normal value
      Serum conjugated bilirrubinDecreasingbelow 15 µmol/L
      PHT related bleedingNo
      SB InfectionsNo
      BMI>20 kg/m2
      * Without diuretic treatment or low dose (spironolactone 75 mg/d or furosemide 40 mg/d).
      In Asian countries where IVC obstruction prevails, TIPS placement is not as frequent as in Europe. However, the use of TIPS in Asia is increasing, and available reports show similar positive results as in the West.
      • Shalimar
      • Gamanagatti S.R.
      • Patel A.H.
      • Kedia S.
      • Nayak B.
      • Gunjan D.
      • et al.
      Long-term outcomes of transjugular intrahepatic portosystemic shunt in Indian patients with Budd-Chiari syndrome.
      • Rathod K.
      • Deshmukh H.
      • Shukla A.
      • Popat B.
      • Pandey A.
      • Gupte A.
      • et al.
      Endovascular treatment of Budd-Chiari syndrome: single center experience.
      • Fan X.
      • Liu K.
      • Che Y.
      • Wang S.
      • Wu X.
      • Cao J.
      • et al.
      Good clinical outcomes in Budd-Chiari syndrome with hepatic vein occlusion.
      • He F.
      • Zhao H.
      • Dai S.
      • Wu Y.
      • Wang L.
      • Huang H.
      • et al.
      Transjugular intrahepatic portosystemic shunt for Budd-Chiari syndrome with diffuse occlusion of hepatic veins.
      • Qi X.
      • Guo W.
      • He C.
      • Zhang W.
      • Wu F.
      • Yin Z.
      • et al.
      Transjugular intrahepatic portosystemic shunt for Budd-Chiari syndrome: techniques, indications and results on 51 Chinese patients from a single centre.
      Current results suggest that the stepwise strategy to BCS treatment is effective and safe, as long as patients are followed closely, and TIPS is implemented as soon as needed.
      Recent data suggest that liver elastography measurements may be a good non-invasive test to evaluate the effectiveness of liver decompression. Currently this has been evaluated after invasive techniques such as balloon angioplasty (with or without stenting) in a small number of patients, showing a reduction of elastography measurement when the liver is adequately decompressed.
      • Wang H.-W.
      • Shi H.-N.
      • Cheng J.
      • Xie F.
      • Luo Y.-K.
      • Tang J.
      Real-time shear wave elastography (SWE) assessment of short- and long-term treatment outcome in Budd-Chiari syndrome: a pilot study.
      It is possible that this technology can also be used to monitor the response to medical treatment, although more studies are needed.

      Liver transplantation: indications and post-transplant approach

      In patients with BCS, LT represents the last therapeutic option following treatment failure on other less-invasive therapies. Although it may be a first step in patients who initially present with acute hepatic failure,
      • Doğrul A.B.
      • Yankol Y.
      • Mecit N.
      • Kanmaz T.
      • Acarlı K.
      • Kalayoğlu M.
      Orthotopic liver transplant for Budd-Chiari syndrome: an analysis of 14 cases.
      TIPS should be considered while waiting for LT as it may foster fast improvement and potentially avoid transplantation. It may be acknowledged that LT in patients with BCS represents a technical challenge mainly because of the presence of retroperitoneal fibrosis related to HV thrombosis, liver enlargement and adhesions. In addition, the classical “piggyback” technique for anastomosis becomes more challenging due to the increased size of the caudate lobe and occlusion of the HV ostia. However, TIPS did not worsen prognosis after LT in patients with BCS.
      • Garcia-Pagan J.C.
      • Heydtmann M.
      • Raffa S.
      • Plessier A.
      • Murad S.
      • Fabris F.
      • et al.
      TIPS for Budd-Chiari syndrome: long-term results and prognostics factors in 124 patients.
      • Segev D.L.
      • Nguyen G.C.
      • Locke J.E.
      • Simpkins C.E.
      • Montgomery R.A.
      • Maley W.R.
      • et al.
      Twenty years of liver transplantation for Budd-Chiari syndrome: a national registry analysis.
      Five-year survival rate after LT has improved over the years.
      • Ulrich F.
      • Pratschke J.
      • Neumann U.
      • Pascher A.
      • Puhl G.
      • Fellmer P.
      • et al.
      Eighteen years of liver transplantation experience in patients with advanced Budd-Chiari syndrome.
      • Mentha G.
      • Giostra E.
      • Majno P.E.
      • Bechstein W.O.
      • Neuhaus P.
      • Grady J.O.
      • et al.
      Liver transplantation for Budd – Chiari syndrome: a European study on 248 patients from 51 centres.
      • Srinivasan P.
      • Rela M.
      • Prachalias A.
      • Muiesan P.
      • Portmann B.
      • Mufti G.J.
      • et al.
      Liver transplantation for Budd-Chiari syndrome.
      • Raza S.M.
      • Zainab S.
      • Shamsaeefar A.R.
      • Nikeghbalian S.
      • Malek Hosseini S.A.
      Experience of liver transplant in patients diagnosed with Budd-Chiari syndrome.
      A large European study showed actuarial overall survival of 76%, 71% and 68% at 1-year, 5-years and 10-years, respectively.
      • Mentha G.
      • Giostra E.
      • Majno P.E.
      • Bechstein W.O.
      • Neuhaus P.
      • Grady J.O.
      • et al.
      Liver transplantation for Budd – Chiari syndrome: a European study on 248 patients from 51 centres.
      Notably, these outcomes are similar to those in patients treated with TIPS (88% and 78% OLT-free survival at 1 year and 5 years, respectively),
      • Garcia-Pagan J.C.
      • Heydtmann M.
      • Raffa S.
      • Plessier A.
      • Murad S.
      • Fabris F.
      • et al.
      TIPS for Budd-Chiari syndrome: long-term results and prognostics factors in 124 patients.
      reinforcing the benefit of the stepwise approach for selecting patients who undoubtedly require LT and saving organs for other indications.
      Once LT is considered, it will be the cure in cases of inborn errors of metabolism, such as antithrombin deficiency or homozygous factor V Leiden mutation. However, other prothrombotic disorders, such as MPN or PNH, can be a contraindication and/or impact post-LT outcome.
      MPN is not a contraindication for LT, since optimal treatment of these patients yields excellent long-term survival. Since patients with MPN have mature and well-differentiated granulocytes, the risk of infectious complications after LT is no higher than in LT patients without MPN. Several case series and retrospective studies
      • Doğrul A.B.
      • Yankol Y.
      • Mecit N.
      • Kanmaz T.
      • Acarlı K.
      • Kalayoğlu M.
      Orthotopic liver transplant for Budd-Chiari syndrome: an analysis of 14 cases.
      • Mentha G.
      • Giostra E.
      • Majno P.E.
      • Bechstein W.O.
      • Neuhaus P.
      • Grady J.O.
      • et al.
      Liver transplantation for Budd – Chiari syndrome: a European study on 248 patients from 51 centres.
      • Potthoff A.
      • Attia D.
      • Pischke S.
      • Mederacke I.
      • Beutel G.
      • Rifai K.
      • et al.
      Long-term outcome of liver transplant patients with Budd-Chiari syndrome secondary to myeloproliferative neoplasms.
      did not reveal evidence of accelerated MPN progression (leukemic transformation) after LT in a 10-year follow-up period, and the survival rates were excellent (71–>90% over 3 years, comparable to non-MPN patients with BCS).
      Patients with MPN who have undergone liver transplantation for BCS should be treated with anticoagulant drugs and/or aspirin, and with anti-proliferative treatment (hydroxyurea). Several groups have reported excellent outcomes in patients with MPN treated only with hydroxyurea and aspirin after LT, without recurrence of BCS.
      • Melear J.M.
      • Goldstein R.M.
      • Levy M.F.
      • Molmenti E.P.
      • Cooper B.
      • Netto G.J.
      • et al.
      Hematologic aspects of liver transplantation for Budd-Chiari syndrome with special reference to myeloproliferative disorders.
      • Chinnakotla S.
      • Klintmalm G.B.
      • Kim P.
      • Tomiyama K.
      • Klintmalm E.
      • Davis G.L.
      • et al.
      Long-term follow-up of liver transplantation for Budd-Chiari syndrome with antithrombotic therapy based on the etiology.
      However, in another study, recurrence of BCS and other thrombotic complications were reported despite anticoagulant therapy with warfarin.
      • Cruz E.
      • Ascher N.L.
      • Roberts J.P.
      • Bass N.M.
      • Yao F.Y.
      High incidence of recurrence and hematologic events following liver transplantation for Budd-Chiari syndrome.
      Based on the available data and the severity of thrombotic complications if they occur, it is advisable to treat patients with anticoagulants (warfarin), aspirin and hydroxyurea in order to prevent recurrent thrombotic complications after LT. In cases of high risk of bleeding or development of bleeding complications, it is suggested to give at least aspirin and anti-proliferative treatment, avoiding VKAs.
      When the underlying disorder is PNH, LT may be more challenging, as patients may present or develop aplastic anaemia, requiring an allogenic stem cell transplant, thereby increasing the risk of infection. Some cases of liver transplantation for BCS in patients with PNH have been reported, sometimes complicated by the recurrence of BCS.
      • Singer A.L.
      • Locke J.E.
      • Stewart Z.A.
      • Lonze B.E.
      • Hamilton J.P.
      • Scudiere J.R.
      • et al.
      Successful liver transplantation for Budd-Chiari syndrome in a patient with paroxysmal nocturnal hemoglobinuria treated with the anti-complement antibody eculizumab.
      • Bahr M.J.
      • Schubert J.
      • Bleck J.S.
      • Tietge U.J.
      • Boozari B.
      • Schmidt R.E.
      • et al.
      Recurrence of Budd-Chiari syndrome after liver transplantation in paroxysmal nocturnal hemoglobinuria.
      There is some evidence that living donor liver transplant is a viable choice for patients with BCS who have experienced treatment failure on other less-invasive therapies.
      Although the evidence comes from a small series of cases, living donor liver transplant is a viable choice with acceptable survival rates (>70% at 5 years).
      • Karaca C.
      • Yilmaz C.
      • Ferecov R.
      • Iakobadze Z.
      • Kilic K.
      • Caglayan L.
      • et al.
      Living-donor liver transplantation for Budd-Chiari syndrome: case series.
      • Ara C.
      • Akbulut S.
      • Ince V.
      • Karakas S.
      • Baskiran A.
      • Yilmaz S.
      Living donor liver transplantation for Budd-Chiari syndrome: overcoming a troublesome situation.

      Treatment of recent NCPVT

      The aim of therapy for recent NCPVT is to prevent the extension of the thrombus to mesenteric veins and intestinal infarction. In addition, therapy for recent NCPVT must try to achieve portal vein recanalization to prevent the development of portal hypertension.
      The general strategy for treating NCPVT is aimed at preventing the extension of the thrombus and achieving portal vein recanalization, to prevent portal hypertension and its complications.

      Anticoagulation

      Anticoagulation is the key treatment of recent NCPVT. Although randomized controlled trials are lacking, a landmark European prospective multicentre study reported no thrombus extension to the splenic or mesenteric vein in 95 patients with recent PVT. They were treated with early administration of LMWH, and rapidly replaced by oral anticoagulation with VKAs, targeting an international normalized ratio of 2–3.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      Only 2/95 patients developed intestinal infarction. There was no mortality related to NCPVT or its treatment. Full recanalization of the vein was only obtained in one-third of patients following 6 months of continued anticoagulation therapy. Interestingly, with a prolongation of anticoagulation, there was no further recanalization of the portal vein, but continued recanalization of the splenic and superior mesenteric vein. A year after the onset of NCPVT, 40% of the patients had a permanent obstruction of the portal vein and portal cavernoma.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      These findings independently validated previous retrospective single centre studies.
      • Amitrano L.
      • Guardascione M.A.
      • Scaglione M.
      • Pezzullo L.
      • Sangiuliano N.
      • Armellino M.F.
      • et al.
      Prognostic factors in noncirrhotic patients with splanchnic vein thromboses.
      • Condat B.
      • Pessione F.
      • Denninger M.H.
      • Hillaire S.
      • Valla D.
      • Condat B.
      • et al.
      Recent portal or mesenteric venous thrombosis: increased recognition and frequent recanalization on anticoagulant therapy.
      • Acosta S.
      • Alhadad A.
      • Svensson P.
      • Ekberg O.
      Epidemiology, risk and prognostic factors in mesenteric venous thrombosis.
      • Turnes J.
      • García-Pagán J.C.
      • González M.
      • Aracil C.
      • Calleja J.L.J.L.
      • Ripoll C.
      • et al.
      Portal hypertension-related complications after acute portal vein thrombosis: impact of early anticoagulation.
      Among baseline factors, splenic vein obstruction, ascites
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      and delays in initiating anticoagulation
      • Turnes J.
      • García-Pagán J.C.
      • González M.
      • Aracil C.
      • Calleja J.L.J.L.
      • Ripoll C.
      • et al.
      Portal hypertension-related complications after acute portal vein thrombosis: impact of early anticoagulation.
      have been associated with the absence of portal vein recanalization. Adverse events on anticoagulation therapy did not obviously differ from those expected from natural history.
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      The mortality rate was 2% and was not related to bleeding or NCPVT, with a median follow-up of 8 months after NCPVT diagnosis
      • Plessier A.
      • Darwish-Murad S.
      • Hernandez-Guerra M.
      • Consigny Y.
      • Fabris F.
      • Trebicka J.
      • et al.
      Acute portal vein thrombosis unrelated to cirrhosis: a prospective multicenter follow-up study.
      .
      Recently, a few cases of initial treatment of recent NCPVT with DOACs instead of LMWH have been reported.
      • Nery F.
      • Valadares D.
      • Morais S.
      • Gomes M.T.
      • De Gottardi A.
      Efficacy and safety of direct-acting oral anticoagulants use in acute portal vein thrombosis unrelated to cirrhosis.
      • Priyanka P.
      • Kupec J.T.
      • Krafft M.
      • Shah N.A.
      • Reynolds G.J.
      Newer oral anticoagulants in the treatment of acute portal vein thrombosis in patients with and without cirrhosis.
      • Janczak D.T.
      • Mimier M.K.
      • McBane R.D.
      • Kamath P.S.
      • Simmons B.S.
      • Bott-Kitslaar D.M.
      • et al.
      Rivaroxaban and apixaban for initial treatment of acute venous thromboembolism of atypical location.
      The largest study compared 26 patients treated with DOACs with 23 treated with enoxaparin. Although the data should be interpreted cautiously, since more than half of these thrombotic events occurred in a cancer setting, recurrence and major bleeding rates were not different between patients treated with DOACs and with enoxaparin.
      • Janczak D.T.
      • Mimier M.K.
      • McBane R.D.
      • Kamath P.S.
      • Simmons B.S.
      • Bott-Kitslaar D.M.
      • et al.
      Rivaroxaban and apixaban for initial treatment of acute venous thromboembolism of atypical location.

      Antibiotics

      Antibiotics are given in patients with NCPVT triggered by an abdominal infection. When septic pylephlebitis is diagnosed, prolonged treatment with antibiotics adapted to isolated bacteria or to anaerobic digestive flora is necessary.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      Recent data suggest that oral antibiotics should be given in patients with acute mesenteric ischaemia, since their use is associated with a lower risk of irreversible transmural intestinal necrosis.
      • Nuzzo A.
      • Maggiori L.
      • Ronot M.
      • Becq A.
      • Plessier A.
      • Gault N.
      • et al.
      Predictive factors of intestinal necrosis in acute mesenteric ischemia: prospective study from an intestinal stroke center.

      Treatment of underlying causes

      Retrospective studies suggested that rapid identification and treatment of risk factors for NCPVT might favourably influence NCPVT outcome. Indeed, in a retrospective multicentre study including 109 patients with MPN and NCPVT (n = 63) or BCS (n = 46), cytoreductive therapy was associated with less common severe liver-related events or vascular complications.
      • Chagneau-Derrode C.
      • Roy L.
      • Guilhot J.
      • Gloria O.
      • Ollivier-Hourmand I.
      • Bureau C.
      • et al.
      Impact of cytoreductive therapy on the outcome of patients with myeloproliferative neoplasms and hepatosplanchnic vein thrombosis.
      Although specific data on NCPVT are lacking, aetiological treatment in addition to anticoagulation in patients with other vascular liver diseases, such as corticosteroids and/or immunosuppressive therapy in Behçet’s disease,
      • Desbois A.C.
      • Rautou P.E.
      • Biard L.
      • Belmatoug N.
      • Wechsler B.
      • Resche-Rigon M.
      • et al.
      Behcet’s disease in Budd-Chiari syndrome.
      or eculizumab (a humanized monoclonal antibody directed against the terminal complement protein C5) in PNH,
      • Sicre de Fontbrune F.
      • Peffault de Latour R.
      Ten years of clinical experience with eculizumab in patients with paroxysmal nocturnal hemoglobinuria.
      improve patient outcomes.

      Thrombolysis and/or interventional radiology

      Pharmacological thrombolysis (local or systemic) has been proposed as an adjunct to anticoagulation. However, severe procedure-related morbidity and fatalities have been reported with recanalization rates similar to those achieved with anticoagulation alone.
      • Smalberg J.H.
      • Spaander M.V.
      • Jie K.S.
      • Pattynama P.M.
      • van Buuren H.R.
      • van den B.B.
      • et al.
      Risks and benefits of transcatheter thrombolytic therapy in patients with splanchnic venous thrombosis.
      • Hollingshead M.
      • Burke C.T.
      • Mauro M.A.
      • Weeks S.M.
      • Dixon R.G.
      • Jaques P.F.
      Transcatheter thrombolytic therapy for acute mesenteric and portal vein thrombosis.
      • Liu K.
      • Li W.-D.
      • Du X.-L.
      • Li C.-L.
      • Li X.-Q.
      Comparison of systemic thrombolysis versus indirect thrombolysis via the superior mesenteric artery in patients with acute portal vein thrombosis.
      Recent reports suggest better results with a combination of transjugular thrombectomy, local fibrinolysis and/or TIPS, as summarized (Table 5). Interestingly, patients treated with this approach rarely developed portal cavernoma and signs of portal hypertension. This invasive strategy does not fit all patients with acute NCPVT, but might be useful in patients with progressive thrombosis, clinical deterioration despite anticoagulation, or with a low likelihood of recanalization following therapeutic anticoagulation. This strategy might also be considered in patients with superior mesenteric vein thrombosis and features predictive of irreversible intestinal ischaemic injury, as detailed above.
      • Nuzzo A.
      • Maggiori L.
      • Ronot M.
      • Becq A.
      • Plessier A.
      • Gault N.
      • et al.
      Predictive factors of intestinal necrosis in acute mesenteric ischemia: prospective study from an intestinal stroke center.
      Table 5Interventional radiology treatment of acute extensive portal vein thrombosis without cirrhosis.
      Reference

      Number of patients
      ProcedureLong-term anticoagulationRecanalizationComplicationsRecurrence of thrombosis
      Hollingshead, 2005
      • Hollingshead M.
      • Burke C.T.
      • Mauro M.A.
      • Weeks S.M.
      • Dixon R.G.
      • Jaques P.F.
      Transcatheter thrombolytic therapy for acute mesenteric and portal vein thrombosis.
      n = 20
      Local thrombolysis aloneYes-Complete: 3/20

      -Partial: 12/20

      -No: 5/20
      -1 death 2 weeks after thrombolytic therapy.-11 major complications

      (bleeding)

      -2 liver transplantation
      Not mentioned
      Smalberg JH, 2008
      • Smalberg J.H.
      • Spaander M.V.
      • Jie K.S.
      • Pattynama P.M.
      • van Buuren H.R.
      • van den B.B.
      • et al.
      Risks and benefits of transcatheter thrombolytic therapy in patients with splanchnic venous thrombosis.
      n = 4
      Local thrombolysis, combined (in 1 out of the 4 patients) with TIPSYes-Complete: 1/4

      -Partial: 1/4

      -No: 2/4
      2 major bleedingNot mentioned
      Cao, 2013
      • Cao G.
      • Ko G.Y.
      • Sung K.B.
      • Yoon H.K.
      • Gwon D.
      • Kim J.H.
      Treatment of postoperative main portal vein and superior mesenteric vein thrombosis with balloon angioplasty and/or stent placement.
      n = 12
      Percutaneous transhepatic balloon angioplasty and/or stent placement without thrombolysis or thrombectomyNo11/121 death from acute respiratory distress syndrome 8 days after the procedure5/12
      Rosenqvist, 2016
      • Rosenqvist K.
      • Eriksson L.-G.
      • Rorsman F.
      • Sangfelt P.
      • Nyman R.
      Endovascular treatment of acute and chronic portal vein thrombosis in patients with cirrhotic and non-cirrhotic liver.
      n = 4
      Local thrombolysis, combined (in 3 out of the 4 patients) with TIPSNot availableLimited data.

      “3 recovered and have survived more than 6 years.”
      Klinger, 2017
      • Siramolpiwat S.
      • Seijo S.
      • Miquel R.
      • Berzigotti A.
      • Garcia-Criado A.
      • Darnell A.
      • et al.
      Idiopathic portal hypertension: natural history and long-term outcome.
      n = 17
      Combination of transjugular thrombectomy, local fibrinolysis and – depending on thrombus resolution – TIPSYes (≥12 months after recanalization)-Complete: 9/17

      -Partial: 7/17

      -No: 1/17
      -2 HIT, including 1 leading to segmental bowel resection)

      -1 additional bowel resection

      -1 hepatic artery pseudoaneurysm with spontaneous occlusion
      -2/17 recurrence of NCPVT at 28 months (1 recanalized)

      -3 TIPS obstructions, all recanalized

      -1 portal cavernoma
      Wolter, 2018
      • Glatard A.-S.
      • Hillaire S.
      • D’Assignies G.
      • Cazals-Hatem D.
      • Plessier A.
      • Valla D.C.
      • et al.
      Obliterative portal venopathy: findings at CT imaging.
      n = 9
      Thrombectomy, local fibrinolysis and/or TIPSNot available-Partial or complete: 7/9-No: 2/9

      (due to failure to access portal vein)
      None3 TIPS obstructions (1 recanalized)
      HIT, heparin-induced thrombocytopenia; TIPS, transjugular intrahepatic portosystemic shunt.

      Surgery

      Surgical thrombectomy is currently not an option for NCPVT given the invasiveness of the procedure, the low rate of recanalization achieved, and the favourable outcome yielded using only anticoagulation.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      Surgery has 2 main indications in patients with acute NCPVT: (a) treatment of a local factor responsible for NCPVT; (b) suspicion of mesenteric infarction.
      • Hmoud B.
      • Singal A.K.
      • Kamath P.S.
      Mesenteric venous thrombosis.

      Treatment of chronic NCPVT

      Management of complications of portal hypertension

      In general, the current recommendation is to treat and prevent complications of portal hypertension, as recommended for patients with cirrhosis.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      • De Franchis R.
      • Abraldes J.G.
      • Bajaj J.
      • Berzigotti A.
      • Bosch J.
      • Burroughs A.K.
      • et al.
      Expanding consensus in portal hypertension: report of the Baveno VI Consensus Workshop: stratifying risk and individualizing care for portal hypertension.

      Oesophageal varices

      In a recent study that included 178 non-cirrhotic patients with NCPVT, the natural history of oesophageal varices appeared similar to that observed in patients with cirrhosis.
      • Noronha Ferreira C.
      • Seijo S.
      • Plessier A.
      • Silva-Junior G.
      • Turon F.
      • Rautou P.-E.P.-E.
      • et al.
      Natural history and management of esophagogastric varices in chronic noncirrhotic, nontumoral portal vein thrombosis.
      In patients without varices at inclusion, the risk of developing them was 2% at 1 year and 22% at 3 years. Progression from small to medium or large varices was 13% at 1 year, and 54% at 5 years. In patients with medium or large varices, who received adequate prophylaxis, the risk of bleeding was 9% at 1 year and 32% at 5 years. Head to head comparison studies have never been conducted. Therefore, there are no strong data on the real impact of those treatments on the natural history of the disease.
      • Association E.
      EASL Clinical Practice Guidelines: vascular diseases of the liver.
      However, it has been suggested that the use of non-selective beta-blockers was associated with a decreased risk of bleeding
      • Condat B.
      • Pessione F.
      • Hillaire S.
      • Denninger M.-H.H.
      • Guillin M.C.
      • Poliquin M.
      • et al.
      Current outcome of portal vein thrombosis in adults: Risk and benefit of anticoagulant therapy.