JAK2 mutation: The best diagnostic tool for myeloproliferative disease in splanchnic vein thrombosis?


  • Harry L. A. Janssen,

    Corresponding author
    1. Department of Gastroenterology & Hepatology, Erasmus University Medical Center, Rotterdam, The Netherlands
    • Department of Gastroenterology and Hepatology, Erasmus University Medical Center, Gravendijkwal 230, Room Ca 326, 3015 CE Rotterdam, The Netherlands
    Search for more papers by this author
    • fax: (31) 10 436 5916

  • Frank W. G. Leebeek

    1. Department of Hematology, Erasmus University Medical Center, Rotterdam, The Netherlands
    Search for more papers by this author

  • See Article on Page 1528.

  • Potential conflict of interest: Nothing to report.

Budd-Chiari syndrome (BCS) is a severe liver disorder, characterized by hepatic venous outflow obstruction, as a result of the occlusion of hepatic veins or inferior vena cava.1–3 Occlusion of the hepatic venous system leads to an increase in the sinusoidal pressure, ischemia, and necrosis. The clinical course of BCS may differ markedly between patients. Some exhibit clinical signs of portal hypertension, such as refractory ascites with relatively intact hepatic function. Others present with liver failure, including hepatic encephalopathy, jaundice, and biochemical signs of severe hepatocellular dysfunction. Extrahepatic portal vein obstruction (EHPVO) is usually caused by thrombosis and may occur spontaneously or as a complication of pre-existent hepatobiliary disease.4 Complications of portal hypertension, such as hemorrhage from esophageal varices and splenomegaly appear to be the most important clinical manifestations of EHPVO. Furthermore, extension of thrombosis into the mesenteric system may induce intestinal infarction. Death in patients with EHPVO is currently primarily related to concurrent medical conditions which are often the cause of EHPVO, rather than to manifestations of portal hypertension.4, 5 Both BCS and EHPVO are rare and severe conditions, predominantly affecting young patients. Their respective presentations and course are highly heterogeneous. Lack of awareness of these conditions is responsible for their delayed recognition, and sometimes inappropriate use of available diagnostic and therapeutic procedures.


BCS, Budd-Chiari syndrome; EEC, endogenous erythroid colony; EHPVO, extrahepatic portal vein obstruction; MPD, myeloproliferative disease.

Regarding their etiology, BCS and EHPVO are considered primary when obstruction is the result of an endoluminal venous lesion.1 They are considered secondary when the obstruction of the venous tract results from material not originating from the venous system (malignant tumor or a parasitic mass invading the lumen) or from pressure of a tumor (abscesses, cysts, benign or malignant solid tumors). Primary disease is presumed to originate from primary thrombosis or involvement of the venous wall. Disorders in hemostasis, both inherited and acquired, can result in an increased tendency toward thrombosis. Hypercoagulability, resulting from either procoagulant conditions (e.g., factor V Leiden mutation, the G20210A prothrombin gene mutation, and the C677T MTHFR gene mutation), or a reduced inhibition of coagulation (e.g., deficiency of antithrombin, protein C, or protein S) predispose a patient toward thrombosis and also play a role in the pathogenesis of BCS and EHPVO.6, 7

The most important systemic etiological factors for BCS and EHPVO however are myeloproliferative diseases (MPDs). MPDs are chronic clonal hematopoietic stem cell disorders characterized by proliferation in the bone marrow of one or more of the myeloid (i.e., granulocytic, erythroid, and megakaryocytic) lineages. Proliferation is associated with relatively normal maturation, resulting in increased numbers of granulocytes, red blood cells, and/or platelets in the peripheral blood. Several earlier studies focused on the criteria for the diagnosis of underlying MPD in patients with BCS and EHPVO. In particular, these studies examined the potential added value of spontaneous endogenous erythroid colony (EEC) formation in detecting so-called occult MPD.8 This term has been designated for patients who do not fulfill the conventional criteria for MPD, as defined by the WHO or the Polycythemia Vera Study Group, but in whom bone marrow culture showed spontaneous EEC formation. In one study, the diagnostic yield of MPD in patients with BCS was reported to be 78% using spontaneous EEC formation as the sole criterion.9 The diagnostic criteria for MPD and the additive value of EEC formation, however, are still a matter of debate.10, 11 Recently, Chait et al. studied the value of EEC formation in diagnosing clinically relevant MPD in patients with splanchnic vein thrombosis, including 36 patients with BCS, and concluded that a combination of bone marrow biopsy findings and EEC formation allowed the diagnosis of MPD at risk of aggravation.10 The diagnosis of MPD in these patients may have therapeutic implications with regard to the anticoagulant treatment. If patients are diagnosed with polycythemia vera and essential thrombocythemia, low-dose aspirin should also be prescribed to reduce future thrombotic complications.12

Recently, a remarkable association between MPD and a somatic point mutation of the JAK2 tyrosine kinase (JAK2 V617F) has been described.13–15 JAK2 is a Janus kinase2 tyrosine kinase that is erythropoietin-receptor-bound. The recently described mutation is located on chromosome 9p and results in an amino acid substitution (valine to phenylalanine) in the auto-inhibitory domain of JAK2. The gain-of-function mutation leads to a constitutive kinase activity causing enhanced hematopoiesis, which is not properly controlled. The acquired mutation in JAK2 is found in more than 90% of patients with polycythemia vera and is seen in about half of the patients with essential thrombocythemia or idiopathic myelofibrosis. Several studies showed that MPD patients carrying the JAK2 mutation have a higher risk of thrombotic complications, possibly due to increased platelet and leukocyte activation.16, 17 Identification of the JAK2 mutation will lead to new approaches to the diagnosis, classification, and treatment of MPD and MPD-related diseases.

For patients with BCS or EHPVO, the discovery of this mutation may be of special importance. As mentioned above, the diagnosis of MPD in these patients has been difficult, and MPD as an etiological factor in BCS may have even been underestimated in the past. Several patients with BCS or EHPVO have marked splenomegaly attributed to portal hypertension, but this may also be due to MPD. In addition, whereas the production of platelets and/or leukocytes may be increased in the marrow of BCS patients with MPD, the peripheral blood cell counts may be normal due to hypersplenism. Therefore, conventional criteria of MPD, including spleen size and peripheral blood counts, are of limited value in patients with BCS or EHPVO. The importance of the JAK2 mutation has recently been demonstrated by Patel et al., who found the JAK2 mutation in 24 of 41 BCS patients.18 None of the 24 BCS patients fulfilled the classic MPD criteria.

Several issues regarding the JAK2 mutation and BCS or EHPVO require further investigation: (1) the prevalence of the JAK2 mutation should be assessed in these conditions; (2) the conventional criteria for diagnosis of MPD should be examined in BCS and EHPVO patients; and (3) the relation between the presence and absence of the JAK2 mutation, other diagnostic criteria (WHO and PVSG), and EEC formation should be evaluated. The latter may provide new insights on the issue of occult forms of MPD in patients with BCS and EHPVO.

The study by Primignani et al. in this issue of HEPATOLOGY explores the first 2 questions.19 They showed that in their patient population the JAK2 mutation was present in 36% of those with EHPVO and in 40% of those with BCS. They compared their findings with conventional clinical-hematological parameters, bone marrow histology according to WHO criteria, and clonality of hematopoiesis by granulocyte X-chromosome inactivation. Although the authors performed a thorough and well-described analysis of their patients, thereby providing a new view on the feasibility and usefulness of the JAK2 mutation in diagnosis of vascular liver disease, some of the results require a balanced interpretation.

Due to the small number of patients included in this study and the incomplete availability of data on bone marrow histology and clonality analysis, it is still unclear what the true value of the JAK2 mutation is in the diagnosis of MPD in BCS. Using only bone marrow findings, MPD was diagnosed in 55% of the patients, which is higher than assessed by JAK2 mutation analysis alone. Combination of the diagnostic tests revealed a diagnosis of MPD in 53% of the patients. Sensitivity of bone marrow biopsy for MPD (defined as a positive JAK2) was more than 90%, but was based only on a small number of patients. In addition, as mentioned by the authors, JAK2 is not the gold standard for MPD, as approximately 10% of patients with polycythemia vera and around 50% of patients with other MPDs are JAK2-negative. It is expected that other mutations in JAK2 or in other receptors of hematopoietic progenitor cells, such as the thrombopoietin receptor (MPL), will be found in MPD in the near future. This may reveal an even higher number of BCS patients with underlying MPD. Therefore, JAK2 mutation analysis presently does not replace other diagnostic tests for MPD, but should be considered an additional diagnostic tool.

A drawback of the study is that it regards a retrospective cohort of 93 patients as a whole and it is questionable whether patients with BCS and EHPVO can be merged into one group of patients with so-called splanchnic vein thrombosis. Although both BCS and EHPVO have a thrombotic origin, they are clearly different in etiology, clinical presentation, and outcome. In particular, the number of BCS (n = 20) patients in this study is limited and hardly sufficient for an appropriate risk analysis. The JAK2 prevalence found in BCS is lower than the almost 60% described recently by Patel et al.18 In fact, both studies harbor a relatively small population of BCS patients. Second, the retrospective nature of the study could well introduce a selection bias, especially because the period of patient identification spans approximately 13 years. All patients were selected after tertiary referral to either a specialized thrombosis and/or liver unit and thus were not identified hospital-wide or, even better, from a predefined general population. In particular, patients with portal vein thrombosis are not seldom treated at general hospitals and can also be referred to surgeons for operations after traumas or for drug refractory inflammatory bowel disease. Thirdly, cirrhosis which is an important risk factor for EHPVO, was excluded on clinical grounds and not based on liver biopsy. Liver tests are not provided and the surprisingly low rate of thrombotic risk factors in EHPVO (e.g., factor V Leiden mutation is even lower than in the general population) could imply that patients with compensated undiagnosed cirrhosis were present in part of the population and therefore not only patients with primary EHPVO may have been included.4

In conclusion, the study by Primignani provides interesting findings on the use of the recently discovered JAK2 mutation for the diagnosis of MPD in BCS and EHVPO. It clearly shows that MPD, both clinically overt and latent, is an important etiologic factor for BCS or EHPVO. Larger studies are required to establish the usefulness of the JAK2 mutation in BCS and EHVPO patients before it is routinely performed. In addition, it does not replace the conventional diagnostic assays, including examination of bone marrow histology, EEC formation, and clonal analysis.