Case 15-2006 — A 46-Year-Old Woman with Sudden Onset of Abdominal Distention
http://www.100md.com
《新英格兰医药杂志》
Presentation of Case
Dr. Patrick S. Yachimski (Gastrointestinal Unit): A 46-year-old woman was admitted to this hospital because of rapidly increasing abdominal girth. She had been well until five weeks earlier, when increasing abdominal distention developed over a period of several days and was accompanied by a rapid increase in weight (from a baseline of 56.7 kg to 70.3 kg), right upper abdominal discomfort, nausea, and vomiting. She saw her primary care physician, who referred her to a gastroenterologist. Abdominal paracentesis was performed, and 3.3 liters of ascitic fluid was removed. The serum–ascites albumin gradient was reported to be 1.2 g per deciliter. The patient was given prescriptions for furosemide and spironolactone.
Three days later, the ascites had reaccumulated, the patient's temperature rose to 38.1°C, and she was admitted to another hospital. Repeated paracentesis showed no evidence of spontaneous bacterial peritonitis. The results of laboratory tests are shown in Table 1. On the fifth hospital day, an abdominal Doppler ultrasonography was reported to show ascites, hepatofugal flow in the right portal vein, hepatopedal flow in the left portal and splenic veins, and a recanalized umbilical vein. The left and right hepatic veins were not visualized. A liver biopsy was performed; microscopical examination of the biopsy specimen showed evidence of venous outflow obstruction, with changes suggestive of partial or incomplete cirrhosis. An abdominal computed tomography (CT) performed on the eighth hospital day disclosed a narrow intrahepatic inferior vena cava.
Table 1. Results of Laboratory Tests.
Also on the eighth day, pain and swelling developed in the right leg. Ultrasonography disclosed deep venous thrombosis involving the right common femoral vein, superficial femoral vein, deep femoral vein, and popliteal vein. Heparin therapy was begun. The results of hypercoagulability testing, performed after the initiation of anticoagulation therapy, are shown in Table 1. The anticoagulation medication was gradually changed to warfarin, and the patient was discharged home on the 18th hospital day, with a referral for outpatient evaluation by a hepatologist. Her weight at discharge was 60.8 kg.
Three days after discharge, the patient came to the emergency department of this hospital because of increasing abdominal girth, despite continued diuretic therapy. She had also begun to have dyspnea, which was worse when she was lying flat but did not increase with exertion. She continued to have nausea, vomiting, anorexia, and early satiety.
During the six months preceding this illness, the patient had lost 9.1 kg in weight, which she attributed to a combination of diet and reduced appetite because of stress. She had no history of thrombosis and did not take oral contraceptives. She had had one pregnancy, with a normal vaginal delivery, and no spontaneous miscarriages. An intrauterine device, placed 12 years earlier, had been removed during her previous hospitalization. She had not had abnormal vaginal bleeding, melena, rectal bleeding, or easy bruising. A partial thyroidectomy had been performed at the age of 22 years because of what she described as "hyperactivity." An appendectomy and a tonsillectomy had been performed during childhood. She had no known allergies.
The patient was born in Egypt and had emigrated to the United States seven years earlier. She taught at a university near Boston, was divorced, and lived with her teenage son. She did not use alcohol or tobacco. Her maternal grandmother had had deep venous thromboses in her arm and leg when she was in her 60s, and intestinal cancer had developed when she was in her 80s. The patient's two sisters were well; one sister took aspirin, which she said was "to prevent clots." The patient's current medications were warfarin, spironolactone, zolpidem, and prochlorperazine as needed for nausea.
In the emergency department, the patient's temperature was 37°C, pulse 98 beats per minute, and blood pressure 120/82 mm Hg. The weight was 61.7 kg. The oxygen saturation was 98 percent while the patient was breathing room air. She appeared comfortable. There was no jaundice and no cutaneous signs of chronic liver disease. The jugular venous pressure was 8 cm. There was no lymphadenopathy. The lungs and heart were normal on auscultation. The abdomen was distended but not tense. The liver span was 10 cm, and the liver was neither tender nor pulsatile. The splenic tip was not palpable, and there were no abdominal masses. There was edema of the legs, which was more pronounced on the right side than on the left. On neurologic examination, she was alert and appropriate, without asterixis. She was admitted to the medical service.
Laboratory-test results are shown in Table 1. A urine pregnancy test was negative. Abdominal ultrasonography showed a coarse, echogenic liver and ascites; the spleen was enlarged, at 14.5 cm in length. Venography through a right transfemoral approach on the fourth hospital day, after the reversal of a supratherapeutic international normalized ratio, demonstrated narrowing of the intrahepatic portion of the inferior vena cava. Patent hepatic veins could not be visualized, despite the use of intravascular ultrasonography. Additional diagnostic studies were performed.
Differential Diagnosis
Dr. Raymond T. Chung: This 46-year-old woman presented to the gastroenterology service with the abrupt onset of right-upper-quadrant pain and distention, a new deep venous thrombosis, hepatomegaly, ascites and peripheral edema, an elevated serum–ascites albumin gradient of 1.2 g per deciliter, hepatic dysfunction, and heterozygosity for the factor V Leiden mutation.
The most useful way to approach this patient's initial presentation is around the abrupt onset of ascites. Ascites is currently classified according to the serum–ascites albumin gradient. This assessment has largely replaced the transudative and exudative concepts because it more effectively differentiates among pathophysiological causes (Table 2). In general, a serum–ascites albumin gradient of 1.1 g per deciliter or greater indicates the presence of portal hypertension, since an oncotic pressure gradient must counter the elevated hydrostatic pressure that drives blood through the sinusoids into the lymphatic vessels and the abdominal cavity as ascites. In contrast, a low serum–ascites albumin gradient (<1.1 g per deciliter) is associated with abnormalities of the peritoneum, including neoplasms, infections, and inflammation. This patient has a high serum–ascites albumin gradient, so the differential diagnosis is narrowed to causes associated with portal hypertension. Dr. Sahani, may we review the radiologic studies?
Table 2. Causes of Ascites According to the Serum–Ascites Albumin Gradient.
Dr. Dushyant V. Sahani: Ultrasonography of the abdomen revealed a coarse, echogenic liver (Figure 1A) with mild ascites. These findings, although nonspecific, are suggestive of chronic liver disease. The portal vein was patent but demonstrated mild reversal of Doppler color-flow mapping, and the hepatic veins were not well visualized. No focal liver lesions were identified. The spleen was mildly enlarged at 14.1 cm. On the fourth hospital day, a venogram showed narrowing of the intrahepatic inferior vena cava (Figure 1B). The hepatic veins could not be visualized even with intravascular ultrasonography.
Figure 1. Images from the Transverse Ultrasonography (Panel A) and Inferior Venacavography (Panel B).
A transverse ultrasonogram of the right upper quadrant shows a coarse echogenic liver with a patent portal vein (Panel A). Inferior venacavography (Panel B), performed through a right transfemoral route, demonstrates narrowing of the intrahepatic inferior vena cava (Panel B, arrow); the hepatic veins are not evident. HA denotes hepatic artery anterior patent to the patent portal vein.
Dr. Chung: The most common cause of hepatic outflow tract obstruction is the Budd–Chiari syndrome, defined as thrombosis of one or more of the large hepatic veins, the inferior vena cava, or both. However, other entities can produce outflow tract obstruction (Table 2). The absence of a history of stem-cell transplantation effectively rules out veno-occlusive disease, and the absence of cardiac findings such as regurgitant murmur, jugular venous distention, pericardial knock, or pulsatile liver on physical examination rules out cardiac causes. The clinical picture, together with the finding on radiography of absent hepatic veins and obstruction to inferior vena caval flow are consistent with the diagnosis of the Budd–Chiari syndrome.
The classic clinical triad of abdominal pain, hepatomegaly, and ascites was described by Budd in 1845,1 and the histopathological features were described by Chiari2 at the turn of the 20th century. The syndrome encompasses a variety of disorders that cause hepatic venous outflow occlusion. Occlusion of three of the major hepatic veins is seen in about two thirds of cases of the Budd–Chiari syndrome; isolated occlusion of the inferior vena cava causes about 10 percent of cases; occlusion of both the inferior vena cava and hepatic veins is observed in nearly a third of the cases. Occlusion of the hepatic veins or inferior vena cava leads to congestion of the centrilobular region, or zone 3, of the liver lobule, which in turn produces swelling of the liver and transmission of hydrostatic pressure across the sinusoids into the hepatic lymphatic vessels, with weeping of fluid into the abdominal cavity, or ascites.
Causes of the Budd–Chiari Syndrome
Myeloproliferative disorders, particularly polycythemia vera and essential thrombocythemia, are implicated in about half the cases of the Budd–Chiari syndrome (Table 3). Neither of these two disorders is overtly evident in this patient, although the platelet count was elevated on one occasion. Cancer may extend into the hepatic veins, causing occlusion, and in renal-cell or hepatocellular carcinomas, paraneoplastic overproduction of erythropoietin can produce secondary polycythemia. Hypercoagulable states can produce the Budd–Chiari syndrome; a factor V Leiden mutation was present in this patient and may have been a risk factor. The patient was neither pregnant nor taking oral contraceptives. Other conditions listed in Table 3 can be ruled out in this case. Finally, about 20 percent of cases cannot be readily explained. Although these cases are now categorized as idiopathic, emerging evidence suggests that many of these cases represent occult myeloproliferative disease.3,4,5
Table 3. Causes of the Budd–Chiari Syndrome.
Clinical Presentations of the Budd–Chiari Syndrome
The Budd–Chiari syndrome has three broad clinical presentations — acute, subacute, and chronic. The first form of presentation (20 percent) results from acute occlusion of all three hepatic veins, leading to the abrupt onset of right-quadrant pain, hepatomegaly, and ascites; in severe cases, fulminant hepatic failure develops. In the subacute form (40 percent), signs and symptoms have been present for less than six months. Ascites and necrosis may be minimal because of the formation of a collateral hepatic circulation. The chronic form (40 percent) tends to have been present for more than six months and is frequently accompanied by cirrhosis and its complications, particularly portal hypertension. In 50 percent of chronic cases, there is hypertrophy of the caudate lobe of the liver, which has independent venous drainage into the inferior vena cava and undergoes compensatory hypertrophy. This can compress the inferior vena cava and further aggravate venous outflow obstruction. Edema of the legs is often a symptom.
This patient had the relatively rapid onset of symptoms five weeks before admission, with ascites and abdominal discomfort; by the time she was seen at this hospital, she had occlusion of both the hepatic veins and the intrahepatic inferior vena cava, caudate-lobe hypertrophy, deep venous thrombosis, and pedal edema. This combination of findings suggests that the process antedated the onset of symptoms and puts her in the group with subacute-to-chronic Budd–Chiari syndrome.
Diagnosis of the Budd–Chiari Syndrome
The Budd–Chiari syndrome is typically associated with nonspecific elevations of liver enzymes, as seen in this patient; in the acute form, these can be dramatically elevated, in excess of five times the normal level. A high serum–ascites albumin gradient with low ascitic fluid total protein (<2.5 g per deciliter) is observed. Doppler ultrasonography, as was used in this case, is the most effective primary initial screening method to establish this diagnosis, with sensitivity and specificity at about 85 percent. CT and magnetic resonance angiography are both more sensitive than ultrasonography. The gold standard for diagnosis is hepatic venography, which should be performed when there is a high index of clinical suspicion and the results of noninvasive testing are either equivocal or negative. In addition to demonstrating venous occlusion, it may show a spider web–like pattern of futile collateral formation, as well as extrinsic inferior vena caval compression by an enlarged caudate lobe, as was seen in this patient.
To confirm the diagnosis and guide management, transjugular liver biopsy should be performed at the time of hepatic venography. In addition, since the underlying disorder that predisposed this patient to the Budd–Chiari syndrome could be a myeloproliferative disorder such as essential thrombocythemia, in conjunction with heterozygosity for factor V Leiden, testing for mutations in the Janus kinase 2 (JAK2) gene may be useful to uncover a latent myeloproliferative disorder. We asked Dr. Amrein to address this possibility.
Dr. Philip C. Amrein: In 75 percent of cases of the Budd–Chiari syndrome, an associated hypercoagulable state is found. For this reason, hematology consultation was requested for this patient, and tests were performed to try to determine whether she had an acquired or hereditary risk of thrombosis. The abnormal results of assays for levels of protein C, protein S, and antithrombin III can be discounted, since these tests were performed when this patient was already taking warfarin and heparin; these agents lower the levels of these proteins in plasma. Liver disease can also lower these values.
The finding that the patient was heterozygous for the factor V Leiden mutation is much more informative. Approximately 5 percent of the white population carries this mutation, and although it confers some increased risk of thrombosis, it is not as severe a defect as deficiencies of protein C, protein S, and antithrombin III. Because it is so prevalent, however, patients with venous disease have a high incidence of this mutation. In one study, the relative risk of the Budd–Chiari syndrome in carriers of the factor V Leiden mutation, as compared with noncarriers, was 11.3.6 However, this risk was much more pronounced when the factor V Leiden mutation was present in addition to another coagulation defect.
As many as 50 percent of patients with this syndrome have a myeloproliferative disorder, either preexisting or diagnosed at the time of the syndrome.3 Myeloproliferative disorders are clonal hematopoietic stem-cell disorders that result in overproduction of one or more types of hematopoietic cells. The diagnosis of myeloproliferative disorders is currently based on a combination of clinical and pathological criteria, the most important of which is the presence of abnormally high levels of peripheral-blood leukocytes, granulocytes, or platelets.7 However, some patients with the Budd–Chiari syndrome may have a latent myeloproliferative disorder, without elevated blood counts. Endogenous erythroid-colony formation (proliferation of erythroid progenitors in the absence of erythropoietin), which occurs in patients with a myeloproliferative disorder but not in persons without such a disorder,8 is common in patients with idiopathic Budd–Chiari syndrome,3,9 suggesting that they have a latent or subclinical myeloproliferative disorder. Recently, a mutation in the JAK2 gene, which is associated with several types of myeloproliferative disorder5,10,11,12 has been detected in a high proportion of patients with the Budd–Chiari syndrome, providing further evidence that these patients have a latent myeloproliferative disorder.13 Thus, testing for JAK2 mutations in all patients presenting with the syndrome and, indeed, all patients with a hypercoagulable state, is a reasonable approach.
The diagnostic procedures in this case were review of the slides of the biopsy specimen of the liver and testing of peripheral-blood leukocytes for evidence of the JAK2 mutation.
Dr. Raymond T. Chung's and Dr. Philip C. Amrein's Diagnosis
Budd–Chiari syndrome, associated with a mutation in JAK2 and the factor V Leiden mutation.
Pathological Discussion
Dr. Joseph Misdraji: We reviewed the slides of the liver-biopsy specimen from the other hospital, as well as those from another biopsy performed at this hospital one month later. Both showed centrilobular necrosis, congestion, and sinusoidal dilatation, classic findings in hepatic outflow obstruction (Figure 2A).14 One characteristic feature of hepatic outflow obstruction is extravasation of red cells into the space of Disse and into the liver-cell plate (Figure 2B). Focally, small hepatic veins contained an organized thrombus (Figure 2C). The two biopsy specimens had different levels of fibrosis. In the original biopsy specimen, an area of bridging fibrosis suggested the presence of incomplete cirrhosis, whereas in the more recent one, bridging fibrosis is absent. This may indicate variable fibrosis throughout the liver.
Figure 2. Liver-Biopsy Specimens (Hematoxylin and Eosin).
Panel A shows centrilobular necrosis (arrow), congestion, and sinusoidal dilatation consistent with the presence of hepatic venous outflow obstruction. A higher magnification (Panel B) of the same site shows red-cell extravasation into and replacing the hepatic plate, characteristic of hepatic venous outflow obstruction. Panel C shows an organizing thrombus (arrow) within a hepatic vein. The surrounding parenchyma shows centrilobular necrosis, congestion, and sinusoidal dilatation.
Dr. A. John Iafrate: Our understanding of the molecular pathogenesis of the chronic myeloproliferative disorders has advanced rapidly; specific causative genetic abnormalities involving tyrosine kinases have now been identified for several entities (Table 4).15,16 A single mutation (V617F) in the JAK2 tyrosine kinase has recently been found in most cases of polycythemia vera and in a subgroup of cases of essential thrombocythemia and chronic idiopathic myelofibrosis.5,10,11,12,17 JAK2 is a signal-transduction molecule, which acts downstream of several cytokine receptors, including the erythropoietin receptor. The V617F mutation gives rise to increased tyrosine kinase activity and allows erythroid precursors to grow in the absence of exogenous erythropoietin.
Table 4. Genetic Abnormalities in Chronic Myeloproliferative Diseases.
Since polycythemia vera and essential thrombocythemia often underlie the Budd–Chiari syndrome, DNA from this patient's peripheral blood was analyzed for JAK2 mutations. Allele-specific polymerase-chain-reaction (PCR) assay showed that the V617F mutation was present (Figures 3A and 3B), and direct sequence analysis revealed the mutant allele to be present in 20 to 30 percent of peripheral-blood nucleated cells (Figure 3C). Since 70 percent of peripheral-blood cells at the time of analysis were of myeloid origin, and since lymphoid cells have been shown not to harbor the JAK2 mutation, some normal myelopoiesis was still present in this patient.
Figure 3. Analysis of the JAK2 Gene.
Allele-specific polymerase chain reaction detected a G-to-T mutation at nucleotide residue 1849 in JAK2. PCR amplification was performed in two reactions (Panel A), one using a wild-type forward primer with a 3' G residue, the other using a mutant forward primer with a 3' T residue. The same reverse primer is used in both reactions, which both yield a 200-bp product. Agarose-gel analysis of PCR-amplification products (Panel B) with the use of peripheral-blood DNA reveals the mutant as well as wild-type JAK2 allele in this patient (shown in duplicate in lanes 4 and 5 and in lanes 6 and 7). Control reactions show no amplification products with water alone (lanes 2 and 3), wild-type JAK2 product only with non-tumor genomic DNA (lanes 8 and 9), and both mutant and wild-type products with DNA from a patient with polycythemia vera (PV) and JAK2 mutation (lanes 10 and 11). Sequence analysis confirmed the JAK2 mutation in peripheral-blood DNA from the patient (Panel C). An electropherogram reveals a new T peak at residue 1849 in the patient (arrow, upper panel), which is not present in the non-tumor genomic DNA control (arrow, lower panel). This mutation results in the substitution of phenylalanine for valine at amino acid position 617 (V617F). The peak height suggests that 20 to 30 percent of circulating white cells harbor the mutant JAK2 allele.
Since the clinical and hematologic findings in this case do not fulfill criteria for a diagnosis of either essential thrombocythemia or polycythemia vera, analysis to identify JAK2 mutations provides critical information supporting a myeloproliferative disorder as the cause for the Budd–Chiari syndrome. The classification of these disorders is currently in flux but will undoubtedly need to be revised to incorporate molecular findings. JAK2 analysis will be useful in differentiating myeloproliferative diseases from reactive thrombocythemia or erythrocytosis.
Management of the Budd–Chiari Syndrome
Dr. Chung: After establishing the diagnosis of Budd–Chiari syndrome, there are three goals of therapy for this patient: to prevent propagation of the underlying thrombotic condition that caused the syndrome, to decompress the liver to prevent further ischemic injury; and to relieve the ascites. The cornerstones of medical therapy include sodium restriction, administration of diuretics to manage the ascites and edema, and treatment of the underlying condition when known (e.g., the use of antiplatelet agents or cytoreductive therapies for essential thrombocythemia or polycythemia vera). When patients present with acute symptoms, thrombolysis may be of benefit.18 Long-term anticoagulation is indicated for known hypercoagulable states, although these are unlikely to lead to clinically meaningful vascular recanalization in patients with subacute or chronic cases. For patients with focal abnormalities such as membranous webs of the inferior vena cava, angioplasty and stenting may be useful and can be combined with thrombolytic therapy. The insertion of transjugular intrahepatic portosystemic shunts can be useful in patients in whom dilation cannot be performed and as a bridge to liver transplantation in patients with the acute Budd–Chiari syndrome.
Portosystemic-shunt surgery is reserved for patients who have necrosis or moderate fibrosis (in the absence of decompensated cirrhosis) and a pressure gradient from the portal vein to the vena cava in excess of 10 mm; it can lead to five-year survival rates of 94 percent.19 For patients with a patent inferior vena cava, a side-to-side portacaval shunt provides the best relief of symptoms. In patients such as this one with caudate hypertrophy compressing the intrahepatic inferior vena cava, a mesocaval shunt may be necessary, but it is associated with an increased rate of shunt thrombosis. For patients with a completely occluded inferior vena cava, a mesoatrial shunt may be inserted; however, this shunt is longer than other types and is thus associated with high occlusion rates.
In patients with hepatic decompensation, liver transplantation is appropriate and may also correct underlying thrombophilic states such as protein C, protein S, or antithrombin III deficiency. Five-year survival rates approach 80 percent.20 In most series, long-term anticoagulation has been applied empirically with excellent outcomes. The long-term prognosis of myeloproliferative disorders is sufficiently good to justify performance of liver transplantation in these patients.
In this patient, with a subacute-to-chronic form of the Budd–Chiari syndrome with necrosis and moderate fibrosis, the recommended approach would be a side-to-side portacaval shunt if technically feasible. If there is marked caudate-lobe hypertrophy, a mesocaval shunt may be necessary.
Dr. Martin Hertl (Department of Surgery): We planned to perform a side-to-side portacaval shunt, but the caudate-lobe hypertrophy necessitated the performance of a mesocaval shunt with a 12-mm Dacron (polyester) graft. The pressure gradient between the inferior vena cava and the portal vein was 11 mm Hg before the procedure, and it was 0 afterward. Anticoagulation therapy and diuretics were resumed postoperatively, and the patient's condition improved for about a week, but her ascites then returned. On postoperative day 14, partial shunt thrombosis was documented; thrombectomy, angioplasty, and stenting of the shunt were performed. On day 21 postoperatively, she had continued ascites and persistent narrowing of the inferior vena cava. Dilation and stenting of the intrahepatic inferior vena cava were performed by the interventional radiologists. The patient resumed taking diuretics and is now asymptomatic three months after the original procedure.
Dr. Nancy Lee Harris (Pathology): Dr. Gilliland, would you like to comment?
Dr. Dwight Gary Gilliland (Dana–Farber Cancer Institute): This case illustrates the point that the Budd–Chiari syndrome develops in patients who may have a mutated JAK2 allele, even in the absence of abnormal peripheral-blood counts. Thrombosis is one of the most devastating complications of myeloproliferative disorders, but we have very little understanding about the pathogenesis of this process or how the genetics of myeloproliferative disease relate to the genetics of coagulopathy. Now that the mutated JAK2 allele has been found in a portion of the cases, we can try to understand how it interfaces with alleles like factor V Leiden in potentiating hypercoagulability and thromboses. Not every person with a myeloproliferative disorder has the JAK2 mutation; thus, we need to continue our efforts to identify other mutations that contribute to the pathogenesis of these diseases.
Dr. Harris: What is the likelihood of recurrence of the Budd–Chiari syndrome or thromboses of other sites, particularly in patients with underlying myeloproliferative disorders? Is there any rationale for treating the myeloproliferative disorder at this point?
Dr. Chung: Recurrence of the syndrome is uncommon with long-term maintenance of anticoagulation therapy, as will be given in this case.
Anatomical Diagnosis
Budd–Chiari syndrome, with centrilobular hepatocellular necrosis, sinusoidal dilatation, congestion, and red-cell extravasation consistent with the presence of venous outflow obstruction.
Heterozygosity for factor V Leiden.
V617F mutation in JAK2.
Dr. Sahani reports having received grant support from Bracco Diagnostics. No other potential conflict of interest relevant to this article was reported.
Source Information
From the Gastrointestinal Unit (R.T.C.) and the Departments of Pathology (A.J.I., J.M.), Hematology–Oncology (P.C.A.), and Radiology (D.V.S.), Massachusetts General Hospital; and the Departments of Medicine (R.T.C.), Pathology (A.J.I., J.M.), Hematology–Oncology (P.C.A.), and Radiology (D.V.S.), Harvard Medical School — both in Boston.
References
Budd G. On diseases of the liver. London: John Churchill, 1845:146.
Chiari H. Ueber die selbstandige phlebitis obliterans der hauptstamme der venae hepaticae als todesursache. Beitr Z Pathol Anat 1899;26:1-18.
Valla D, Casadevall N, Lacombe C, et al. Primary myeloproliferative disorder and hepatic vein thrombosis: a prospective study of erythroid colony formation in vitro in 20 patients with Budd-Chiari syndrome. Ann Intern Med 1985;103:329-334.
Hirshberg B, Shouval D, Fibach E, Friedman G, Ben-Yehuda D. Flow cytometric analysis of autonomous growth of erythroid precursors in liquid culture detects occult polycythemia vera in the Budd-Chiari syndrome. J Hepatol 2000;32:574-578.
Kralovics R, Passamonti F, Buser AS, et al. A gain-of-function mutation of JAK2 in myeloproliferative disorders. N Engl J Med 2005;352:1779-1790.
Janssen HL, Meinardi JR, Vleggaar FP, 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. Blood 2000;96:2364-2368.
Vardiman JW, Brunning RD, Harris NL. Chronic myeloproliferative diseases. In: Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. Pathology and genetics of tumours of haematopoietic and lymphoid tissues. Vol. 3 of World Health Organization classification of tumours. Lyon, France: IARC Press, 2001:17-9.
Prchal JF, Axelrad AA. Bone-marrow responses in polycythemia vera. N Engl J Med 1974;290:1382-1382.
Pagliuca A, Mufti GJ, Janossa-Tahernia M, et al. In vitro colony culture and chromosomal studies in hepatic and portal vein thrombosis -- possible evidence of an occult myeloproliferative state. Q J Med 1990;76:981-989.
James C, Ugo V, Le Couedic JP, et al. A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera. Nature 2005;434:1144-1148.
Levine RL, Wadleigh M, Cools J, et al. Activating mutation in the tyrosine kinase JAK2 in polycythemia vera, essential thrombocythemia, and myeloid metaplasia with myelofibrosis. Cancer Cell 2005;7:387-397.
Baxter EJ, Scott LM, Campbell PJ, et al. Acquired mutation of the tyrosine kinase JAK2 in human myeloproliferative disorders. Lancet 2005;365:1054-1061.
Patel RK. Prevalence of the activating JAK2 tyrosine kinase mutation V617F in the Budd-Chiari syndrome. In: Proceedings of the 47th Annual Meeting of the American Society of Hematology, Atlanta, December 10–13, 2005:7282.
Wanless IR. Vascular disorders. In: MacSween RNM, Burt AD, Portmann BC, Ishak KG, Scheuer PJ, Anthony PP, eds. Pathology of the liver. 4th ed. London: Churchill Livingstone, 2002:539-74.
Nowell P, Hungerford D. A minute chromosome in human chronic granulocytic leukemia. Science 1960;132:1497-1497.
Cools J, DeAngelo DJ, Gotlib J, et al. A tyrosine kinase created by fusion of the PDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hypereosinophilic syndrome. N Engl J Med 2003;348:1201-1214.
Zhao R, Xing S, Li Z, et al. Identification of an acquired JAK2 mutation in polycythemia vera. J Biol Chem 2005;280:22788-22792.
Sharma S, Texeira A, Texeira P, Elias E, Wilde J, Olliff SP. Pharmacological thrombolysis in Budd-Chiari syndrome: a single centre experience and review of the literature. J Hepatol 2004;40:172-180.
Orloff MJ, Daily PO, Orloff SL, Girard B, Orloff MS. A 27-year experience with surgical treatment of Budd-Chiari syndrome. Ann Surg 2000;232:340-352.
Mentha G, Giostra E, Majno PE, et al. Liver transplantation for Budd-Chiari syndrome: a European study on 248 patients from 51 centres. J Hepatol 2006;44:520-528.(Raymond T. Chung, M.D., A)
Dr. Patrick S. Yachimski (Gastrointestinal Unit): A 46-year-old woman was admitted to this hospital because of rapidly increasing abdominal girth. She had been well until five weeks earlier, when increasing abdominal distention developed over a period of several days and was accompanied by a rapid increase in weight (from a baseline of 56.7 kg to 70.3 kg), right upper abdominal discomfort, nausea, and vomiting. She saw her primary care physician, who referred her to a gastroenterologist. Abdominal paracentesis was performed, and 3.3 liters of ascitic fluid was removed. The serum–ascites albumin gradient was reported to be 1.2 g per deciliter. The patient was given prescriptions for furosemide and spironolactone.
Three days later, the ascites had reaccumulated, the patient's temperature rose to 38.1°C, and she was admitted to another hospital. Repeated paracentesis showed no evidence of spontaneous bacterial peritonitis. The results of laboratory tests are shown in Table 1. On the fifth hospital day, an abdominal Doppler ultrasonography was reported to show ascites, hepatofugal flow in the right portal vein, hepatopedal flow in the left portal and splenic veins, and a recanalized umbilical vein. The left and right hepatic veins were not visualized. A liver biopsy was performed; microscopical examination of the biopsy specimen showed evidence of venous outflow obstruction, with changes suggestive of partial or incomplete cirrhosis. An abdominal computed tomography (CT) performed on the eighth hospital day disclosed a narrow intrahepatic inferior vena cava.
Table 1. Results of Laboratory Tests.
Also on the eighth day, pain and swelling developed in the right leg. Ultrasonography disclosed deep venous thrombosis involving the right common femoral vein, superficial femoral vein, deep femoral vein, and popliteal vein. Heparin therapy was begun. The results of hypercoagulability testing, performed after the initiation of anticoagulation therapy, are shown in Table 1. The anticoagulation medication was gradually changed to warfarin, and the patient was discharged home on the 18th hospital day, with a referral for outpatient evaluation by a hepatologist. Her weight at discharge was 60.8 kg.
Three days after discharge, the patient came to the emergency department of this hospital because of increasing abdominal girth, despite continued diuretic therapy. She had also begun to have dyspnea, which was worse when she was lying flat but did not increase with exertion. She continued to have nausea, vomiting, anorexia, and early satiety.
During the six months preceding this illness, the patient had lost 9.1 kg in weight, which she attributed to a combination of diet and reduced appetite because of stress. She had no history of thrombosis and did not take oral contraceptives. She had had one pregnancy, with a normal vaginal delivery, and no spontaneous miscarriages. An intrauterine device, placed 12 years earlier, had been removed during her previous hospitalization. She had not had abnormal vaginal bleeding, melena, rectal bleeding, or easy bruising. A partial thyroidectomy had been performed at the age of 22 years because of what she described as "hyperactivity." An appendectomy and a tonsillectomy had been performed during childhood. She had no known allergies.
The patient was born in Egypt and had emigrated to the United States seven years earlier. She taught at a university near Boston, was divorced, and lived with her teenage son. She did not use alcohol or tobacco. Her maternal grandmother had had deep venous thromboses in her arm and leg when she was in her 60s, and intestinal cancer had developed when she was in her 80s. The patient's two sisters were well; one sister took aspirin, which she said was "to prevent clots." The patient's current medications were warfarin, spironolactone, zolpidem, and prochlorperazine as needed for nausea.
In the emergency department, the patient's temperature was 37°C, pulse 98 beats per minute, and blood pressure 120/82 mm Hg. The weight was 61.7 kg. The oxygen saturation was 98 percent while the patient was breathing room air. She appeared comfortable. There was no jaundice and no cutaneous signs of chronic liver disease. The jugular venous pressure was 8 cm. There was no lymphadenopathy. The lungs and heart were normal on auscultation. The abdomen was distended but not tense. The liver span was 10 cm, and the liver was neither tender nor pulsatile. The splenic tip was not palpable, and there were no abdominal masses. There was edema of the legs, which was more pronounced on the right side than on the left. On neurologic examination, she was alert and appropriate, without asterixis. She was admitted to the medical service.
Laboratory-test results are shown in Table 1. A urine pregnancy test was negative. Abdominal ultrasonography showed a coarse, echogenic liver and ascites; the spleen was enlarged, at 14.5 cm in length. Venography through a right transfemoral approach on the fourth hospital day, after the reversal of a supratherapeutic international normalized ratio, demonstrated narrowing of the intrahepatic portion of the inferior vena cava. Patent hepatic veins could not be visualized, despite the use of intravascular ultrasonography. Additional diagnostic studies were performed.
Differential Diagnosis
Dr. Raymond T. Chung: This 46-year-old woman presented to the gastroenterology service with the abrupt onset of right-upper-quadrant pain and distention, a new deep venous thrombosis, hepatomegaly, ascites and peripheral edema, an elevated serum–ascites albumin gradient of 1.2 g per deciliter, hepatic dysfunction, and heterozygosity for the factor V Leiden mutation.
The most useful way to approach this patient's initial presentation is around the abrupt onset of ascites. Ascites is currently classified according to the serum–ascites albumin gradient. This assessment has largely replaced the transudative and exudative concepts because it more effectively differentiates among pathophysiological causes (Table 2). In general, a serum–ascites albumin gradient of 1.1 g per deciliter or greater indicates the presence of portal hypertension, since an oncotic pressure gradient must counter the elevated hydrostatic pressure that drives blood through the sinusoids into the lymphatic vessels and the abdominal cavity as ascites. In contrast, a low serum–ascites albumin gradient (<1.1 g per deciliter) is associated with abnormalities of the peritoneum, including neoplasms, infections, and inflammation. This patient has a high serum–ascites albumin gradient, so the differential diagnosis is narrowed to causes associated with portal hypertension. Dr. Sahani, may we review the radiologic studies?
Table 2. Causes of Ascites According to the Serum–Ascites Albumin Gradient.
Dr. Dushyant V. Sahani: Ultrasonography of the abdomen revealed a coarse, echogenic liver (Figure 1A) with mild ascites. These findings, although nonspecific, are suggestive of chronic liver disease. The portal vein was patent but demonstrated mild reversal of Doppler color-flow mapping, and the hepatic veins were not well visualized. No focal liver lesions were identified. The spleen was mildly enlarged at 14.1 cm. On the fourth hospital day, a venogram showed narrowing of the intrahepatic inferior vena cava (Figure 1B). The hepatic veins could not be visualized even with intravascular ultrasonography.
Figure 1. Images from the Transverse Ultrasonography (Panel A) and Inferior Venacavography (Panel B).
A transverse ultrasonogram of the right upper quadrant shows a coarse echogenic liver with a patent portal vein (Panel A). Inferior venacavography (Panel B), performed through a right transfemoral route, demonstrates narrowing of the intrahepatic inferior vena cava (Panel B, arrow); the hepatic veins are not evident. HA denotes hepatic artery anterior patent to the patent portal vein.
Dr. Chung: The most common cause of hepatic outflow tract obstruction is the Budd–Chiari syndrome, defined as thrombosis of one or more of the large hepatic veins, the inferior vena cava, or both. However, other entities can produce outflow tract obstruction (Table 2). The absence of a history of stem-cell transplantation effectively rules out veno-occlusive disease, and the absence of cardiac findings such as regurgitant murmur, jugular venous distention, pericardial knock, or pulsatile liver on physical examination rules out cardiac causes. The clinical picture, together with the finding on radiography of absent hepatic veins and obstruction to inferior vena caval flow are consistent with the diagnosis of the Budd–Chiari syndrome.
The classic clinical triad of abdominal pain, hepatomegaly, and ascites was described by Budd in 1845,1 and the histopathological features were described by Chiari2 at the turn of the 20th century. The syndrome encompasses a variety of disorders that cause hepatic venous outflow occlusion. Occlusion of three of the major hepatic veins is seen in about two thirds of cases of the Budd–Chiari syndrome; isolated occlusion of the inferior vena cava causes about 10 percent of cases; occlusion of both the inferior vena cava and hepatic veins is observed in nearly a third of the cases. Occlusion of the hepatic veins or inferior vena cava leads to congestion of the centrilobular region, or zone 3, of the liver lobule, which in turn produces swelling of the liver and transmission of hydrostatic pressure across the sinusoids into the hepatic lymphatic vessels, with weeping of fluid into the abdominal cavity, or ascites.
Causes of the Budd–Chiari Syndrome
Myeloproliferative disorders, particularly polycythemia vera and essential thrombocythemia, are implicated in about half the cases of the Budd–Chiari syndrome (Table 3). Neither of these two disorders is overtly evident in this patient, although the platelet count was elevated on one occasion. Cancer may extend into the hepatic veins, causing occlusion, and in renal-cell or hepatocellular carcinomas, paraneoplastic overproduction of erythropoietin can produce secondary polycythemia. Hypercoagulable states can produce the Budd–Chiari syndrome; a factor V Leiden mutation was present in this patient and may have been a risk factor. The patient was neither pregnant nor taking oral contraceptives. Other conditions listed in Table 3 can be ruled out in this case. Finally, about 20 percent of cases cannot be readily explained. Although these cases are now categorized as idiopathic, emerging evidence suggests that many of these cases represent occult myeloproliferative disease.3,4,5
Table 3. Causes of the Budd–Chiari Syndrome.
Clinical Presentations of the Budd–Chiari Syndrome
The Budd–Chiari syndrome has three broad clinical presentations — acute, subacute, and chronic. The first form of presentation (20 percent) results from acute occlusion of all three hepatic veins, leading to the abrupt onset of right-quadrant pain, hepatomegaly, and ascites; in severe cases, fulminant hepatic failure develops. In the subacute form (40 percent), signs and symptoms have been present for less than six months. Ascites and necrosis may be minimal because of the formation of a collateral hepatic circulation. The chronic form (40 percent) tends to have been present for more than six months and is frequently accompanied by cirrhosis and its complications, particularly portal hypertension. In 50 percent of chronic cases, there is hypertrophy of the caudate lobe of the liver, which has independent venous drainage into the inferior vena cava and undergoes compensatory hypertrophy. This can compress the inferior vena cava and further aggravate venous outflow obstruction. Edema of the legs is often a symptom.
This patient had the relatively rapid onset of symptoms five weeks before admission, with ascites and abdominal discomfort; by the time she was seen at this hospital, she had occlusion of both the hepatic veins and the intrahepatic inferior vena cava, caudate-lobe hypertrophy, deep venous thrombosis, and pedal edema. This combination of findings suggests that the process antedated the onset of symptoms and puts her in the group with subacute-to-chronic Budd–Chiari syndrome.
Diagnosis of the Budd–Chiari Syndrome
The Budd–Chiari syndrome is typically associated with nonspecific elevations of liver enzymes, as seen in this patient; in the acute form, these can be dramatically elevated, in excess of five times the normal level. A high serum–ascites albumin gradient with low ascitic fluid total protein (<2.5 g per deciliter) is observed. Doppler ultrasonography, as was used in this case, is the most effective primary initial screening method to establish this diagnosis, with sensitivity and specificity at about 85 percent. CT and magnetic resonance angiography are both more sensitive than ultrasonography. The gold standard for diagnosis is hepatic venography, which should be performed when there is a high index of clinical suspicion and the results of noninvasive testing are either equivocal or negative. In addition to demonstrating venous occlusion, it may show a spider web–like pattern of futile collateral formation, as well as extrinsic inferior vena caval compression by an enlarged caudate lobe, as was seen in this patient.
To confirm the diagnosis and guide management, transjugular liver biopsy should be performed at the time of hepatic venography. In addition, since the underlying disorder that predisposed this patient to the Budd–Chiari syndrome could be a myeloproliferative disorder such as essential thrombocythemia, in conjunction with heterozygosity for factor V Leiden, testing for mutations in the Janus kinase 2 (JAK2) gene may be useful to uncover a latent myeloproliferative disorder. We asked Dr. Amrein to address this possibility.
Dr. Philip C. Amrein: In 75 percent of cases of the Budd–Chiari syndrome, an associated hypercoagulable state is found. For this reason, hematology consultation was requested for this patient, and tests were performed to try to determine whether she had an acquired or hereditary risk of thrombosis. The abnormal results of assays for levels of protein C, protein S, and antithrombin III can be discounted, since these tests were performed when this patient was already taking warfarin and heparin; these agents lower the levels of these proteins in plasma. Liver disease can also lower these values.
The finding that the patient was heterozygous for the factor V Leiden mutation is much more informative. Approximately 5 percent of the white population carries this mutation, and although it confers some increased risk of thrombosis, it is not as severe a defect as deficiencies of protein C, protein S, and antithrombin III. Because it is so prevalent, however, patients with venous disease have a high incidence of this mutation. In one study, the relative risk of the Budd–Chiari syndrome in carriers of the factor V Leiden mutation, as compared with noncarriers, was 11.3.6 However, this risk was much more pronounced when the factor V Leiden mutation was present in addition to another coagulation defect.
As many as 50 percent of patients with this syndrome have a myeloproliferative disorder, either preexisting or diagnosed at the time of the syndrome.3 Myeloproliferative disorders are clonal hematopoietic stem-cell disorders that result in overproduction of one or more types of hematopoietic cells. The diagnosis of myeloproliferative disorders is currently based on a combination of clinical and pathological criteria, the most important of which is the presence of abnormally high levels of peripheral-blood leukocytes, granulocytes, or platelets.7 However, some patients with the Budd–Chiari syndrome may have a latent myeloproliferative disorder, without elevated blood counts. Endogenous erythroid-colony formation (proliferation of erythroid progenitors in the absence of erythropoietin), which occurs in patients with a myeloproliferative disorder but not in persons without such a disorder,8 is common in patients with idiopathic Budd–Chiari syndrome,3,9 suggesting that they have a latent or subclinical myeloproliferative disorder. Recently, a mutation in the JAK2 gene, which is associated with several types of myeloproliferative disorder5,10,11,12 has been detected in a high proportion of patients with the Budd–Chiari syndrome, providing further evidence that these patients have a latent myeloproliferative disorder.13 Thus, testing for JAK2 mutations in all patients presenting with the syndrome and, indeed, all patients with a hypercoagulable state, is a reasonable approach.
The diagnostic procedures in this case were review of the slides of the biopsy specimen of the liver and testing of peripheral-blood leukocytes for evidence of the JAK2 mutation.
Dr. Raymond T. Chung's and Dr. Philip C. Amrein's Diagnosis
Budd–Chiari syndrome, associated with a mutation in JAK2 and the factor V Leiden mutation.
Pathological Discussion
Dr. Joseph Misdraji: We reviewed the slides of the liver-biopsy specimen from the other hospital, as well as those from another biopsy performed at this hospital one month later. Both showed centrilobular necrosis, congestion, and sinusoidal dilatation, classic findings in hepatic outflow obstruction (Figure 2A).14 One characteristic feature of hepatic outflow obstruction is extravasation of red cells into the space of Disse and into the liver-cell plate (Figure 2B). Focally, small hepatic veins contained an organized thrombus (Figure 2C). The two biopsy specimens had different levels of fibrosis. In the original biopsy specimen, an area of bridging fibrosis suggested the presence of incomplete cirrhosis, whereas in the more recent one, bridging fibrosis is absent. This may indicate variable fibrosis throughout the liver.
Figure 2. Liver-Biopsy Specimens (Hematoxylin and Eosin).
Panel A shows centrilobular necrosis (arrow), congestion, and sinusoidal dilatation consistent with the presence of hepatic venous outflow obstruction. A higher magnification (Panel B) of the same site shows red-cell extravasation into and replacing the hepatic plate, characteristic of hepatic venous outflow obstruction. Panel C shows an organizing thrombus (arrow) within a hepatic vein. The surrounding parenchyma shows centrilobular necrosis, congestion, and sinusoidal dilatation.
Dr. A. John Iafrate: Our understanding of the molecular pathogenesis of the chronic myeloproliferative disorders has advanced rapidly; specific causative genetic abnormalities involving tyrosine kinases have now been identified for several entities (Table 4).15,16 A single mutation (V617F) in the JAK2 tyrosine kinase has recently been found in most cases of polycythemia vera and in a subgroup of cases of essential thrombocythemia and chronic idiopathic myelofibrosis.5,10,11,12,17 JAK2 is a signal-transduction molecule, which acts downstream of several cytokine receptors, including the erythropoietin receptor. The V617F mutation gives rise to increased tyrosine kinase activity and allows erythroid precursors to grow in the absence of exogenous erythropoietin.
Table 4. Genetic Abnormalities in Chronic Myeloproliferative Diseases.
Since polycythemia vera and essential thrombocythemia often underlie the Budd–Chiari syndrome, DNA from this patient's peripheral blood was analyzed for JAK2 mutations. Allele-specific polymerase-chain-reaction (PCR) assay showed that the V617F mutation was present (Figures 3A and 3B), and direct sequence analysis revealed the mutant allele to be present in 20 to 30 percent of peripheral-blood nucleated cells (Figure 3C). Since 70 percent of peripheral-blood cells at the time of analysis were of myeloid origin, and since lymphoid cells have been shown not to harbor the JAK2 mutation, some normal myelopoiesis was still present in this patient.
Figure 3. Analysis of the JAK2 Gene.
Allele-specific polymerase chain reaction detected a G-to-T mutation at nucleotide residue 1849 in JAK2. PCR amplification was performed in two reactions (Panel A), one using a wild-type forward primer with a 3' G residue, the other using a mutant forward primer with a 3' T residue. The same reverse primer is used in both reactions, which both yield a 200-bp product. Agarose-gel analysis of PCR-amplification products (Panel B) with the use of peripheral-blood DNA reveals the mutant as well as wild-type JAK2 allele in this patient (shown in duplicate in lanes 4 and 5 and in lanes 6 and 7). Control reactions show no amplification products with water alone (lanes 2 and 3), wild-type JAK2 product only with non-tumor genomic DNA (lanes 8 and 9), and both mutant and wild-type products with DNA from a patient with polycythemia vera (PV) and JAK2 mutation (lanes 10 and 11). Sequence analysis confirmed the JAK2 mutation in peripheral-blood DNA from the patient (Panel C). An electropherogram reveals a new T peak at residue 1849 in the patient (arrow, upper panel), which is not present in the non-tumor genomic DNA control (arrow, lower panel). This mutation results in the substitution of phenylalanine for valine at amino acid position 617 (V617F). The peak height suggests that 20 to 30 percent of circulating white cells harbor the mutant JAK2 allele.
Since the clinical and hematologic findings in this case do not fulfill criteria for a diagnosis of either essential thrombocythemia or polycythemia vera, analysis to identify JAK2 mutations provides critical information supporting a myeloproliferative disorder as the cause for the Budd–Chiari syndrome. The classification of these disorders is currently in flux but will undoubtedly need to be revised to incorporate molecular findings. JAK2 analysis will be useful in differentiating myeloproliferative diseases from reactive thrombocythemia or erythrocytosis.
Management of the Budd–Chiari Syndrome
Dr. Chung: After establishing the diagnosis of Budd–Chiari syndrome, there are three goals of therapy for this patient: to prevent propagation of the underlying thrombotic condition that caused the syndrome, to decompress the liver to prevent further ischemic injury; and to relieve the ascites. The cornerstones of medical therapy include sodium restriction, administration of diuretics to manage the ascites and edema, and treatment of the underlying condition when known (e.g., the use of antiplatelet agents or cytoreductive therapies for essential thrombocythemia or polycythemia vera). When patients present with acute symptoms, thrombolysis may be of benefit.18 Long-term anticoagulation is indicated for known hypercoagulable states, although these are unlikely to lead to clinically meaningful vascular recanalization in patients with subacute or chronic cases. For patients with focal abnormalities such as membranous webs of the inferior vena cava, angioplasty and stenting may be useful and can be combined with thrombolytic therapy. The insertion of transjugular intrahepatic portosystemic shunts can be useful in patients in whom dilation cannot be performed and as a bridge to liver transplantation in patients with the acute Budd–Chiari syndrome.
Portosystemic-shunt surgery is reserved for patients who have necrosis or moderate fibrosis (in the absence of decompensated cirrhosis) and a pressure gradient from the portal vein to the vena cava in excess of 10 mm; it can lead to five-year survival rates of 94 percent.19 For patients with a patent inferior vena cava, a side-to-side portacaval shunt provides the best relief of symptoms. In patients such as this one with caudate hypertrophy compressing the intrahepatic inferior vena cava, a mesocaval shunt may be necessary, but it is associated with an increased rate of shunt thrombosis. For patients with a completely occluded inferior vena cava, a mesoatrial shunt may be inserted; however, this shunt is longer than other types and is thus associated with high occlusion rates.
In patients with hepatic decompensation, liver transplantation is appropriate and may also correct underlying thrombophilic states such as protein C, protein S, or antithrombin III deficiency. Five-year survival rates approach 80 percent.20 In most series, long-term anticoagulation has been applied empirically with excellent outcomes. The long-term prognosis of myeloproliferative disorders is sufficiently good to justify performance of liver transplantation in these patients.
In this patient, with a subacute-to-chronic form of the Budd–Chiari syndrome with necrosis and moderate fibrosis, the recommended approach would be a side-to-side portacaval shunt if technically feasible. If there is marked caudate-lobe hypertrophy, a mesocaval shunt may be necessary.
Dr. Martin Hertl (Department of Surgery): We planned to perform a side-to-side portacaval shunt, but the caudate-lobe hypertrophy necessitated the performance of a mesocaval shunt with a 12-mm Dacron (polyester) graft. The pressure gradient between the inferior vena cava and the portal vein was 11 mm Hg before the procedure, and it was 0 afterward. Anticoagulation therapy and diuretics were resumed postoperatively, and the patient's condition improved for about a week, but her ascites then returned. On postoperative day 14, partial shunt thrombosis was documented; thrombectomy, angioplasty, and stenting of the shunt were performed. On day 21 postoperatively, she had continued ascites and persistent narrowing of the inferior vena cava. Dilation and stenting of the intrahepatic inferior vena cava were performed by the interventional radiologists. The patient resumed taking diuretics and is now asymptomatic three months after the original procedure.
Dr. Nancy Lee Harris (Pathology): Dr. Gilliland, would you like to comment?
Dr. Dwight Gary Gilliland (Dana–Farber Cancer Institute): This case illustrates the point that the Budd–Chiari syndrome develops in patients who may have a mutated JAK2 allele, even in the absence of abnormal peripheral-blood counts. Thrombosis is one of the most devastating complications of myeloproliferative disorders, but we have very little understanding about the pathogenesis of this process or how the genetics of myeloproliferative disease relate to the genetics of coagulopathy. Now that the mutated JAK2 allele has been found in a portion of the cases, we can try to understand how it interfaces with alleles like factor V Leiden in potentiating hypercoagulability and thromboses. Not every person with a myeloproliferative disorder has the JAK2 mutation; thus, we need to continue our efforts to identify other mutations that contribute to the pathogenesis of these diseases.
Dr. Harris: What is the likelihood of recurrence of the Budd–Chiari syndrome or thromboses of other sites, particularly in patients with underlying myeloproliferative disorders? Is there any rationale for treating the myeloproliferative disorder at this point?
Dr. Chung: Recurrence of the syndrome is uncommon with long-term maintenance of anticoagulation therapy, as will be given in this case.
Anatomical Diagnosis
Budd–Chiari syndrome, with centrilobular hepatocellular necrosis, sinusoidal dilatation, congestion, and red-cell extravasation consistent with the presence of venous outflow obstruction.
Heterozygosity for factor V Leiden.
V617F mutation in JAK2.
Dr. Sahani reports having received grant support from Bracco Diagnostics. No other potential conflict of interest relevant to this article was reported.
Source Information
From the Gastrointestinal Unit (R.T.C.) and the Departments of Pathology (A.J.I., J.M.), Hematology–Oncology (P.C.A.), and Radiology (D.V.S.), Massachusetts General Hospital; and the Departments of Medicine (R.T.C.), Pathology (A.J.I., J.M.), Hematology–Oncology (P.C.A.), and Radiology (D.V.S.), Harvard Medical School — both in Boston.
References
Budd G. On diseases of the liver. London: John Churchill, 1845:146.
Chiari H. Ueber die selbstandige phlebitis obliterans der hauptstamme der venae hepaticae als todesursache. Beitr Z Pathol Anat 1899;26:1-18.
Valla D, Casadevall N, Lacombe C, et al. Primary myeloproliferative disorder and hepatic vein thrombosis: a prospective study of erythroid colony formation in vitro in 20 patients with Budd-Chiari syndrome. Ann Intern Med 1985;103:329-334.
Hirshberg B, Shouval D, Fibach E, Friedman G, Ben-Yehuda D. Flow cytometric analysis of autonomous growth of erythroid precursors in liquid culture detects occult polycythemia vera in the Budd-Chiari syndrome. J Hepatol 2000;32:574-578.
Kralovics R, Passamonti F, Buser AS, et al. A gain-of-function mutation of JAK2 in myeloproliferative disorders. N Engl J Med 2005;352:1779-1790.
Janssen HL, Meinardi JR, Vleggaar FP, 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. Blood 2000;96:2364-2368.
Vardiman JW, Brunning RD, Harris NL. Chronic myeloproliferative diseases. In: Jaffe ES, Harris NL, Stein H, Vardiman JW, eds. Pathology and genetics of tumours of haematopoietic and lymphoid tissues. Vol. 3 of World Health Organization classification of tumours. Lyon, France: IARC Press, 2001:17-9.
Prchal JF, Axelrad AA. Bone-marrow responses in polycythemia vera. N Engl J Med 1974;290:1382-1382.
Pagliuca A, Mufti GJ, Janossa-Tahernia M, et al. In vitro colony culture and chromosomal studies in hepatic and portal vein thrombosis -- possible evidence of an occult myeloproliferative state. Q J Med 1990;76:981-989.
James C, Ugo V, Le Couedic JP, et al. A unique clonal JAK2 mutation leading to constitutive signalling causes polycythaemia vera. Nature 2005;434:1144-1148.
Levine RL, Wadleigh M, Cools J, et al. Activating mutation in the tyrosine kinase JAK2 in polycythemia vera, essential thrombocythemia, and myeloid metaplasia with myelofibrosis. Cancer Cell 2005;7:387-397.
Baxter EJ, Scott LM, Campbell PJ, et al. Acquired mutation of the tyrosine kinase JAK2 in human myeloproliferative disorders. Lancet 2005;365:1054-1061.
Patel RK. Prevalence of the activating JAK2 tyrosine kinase mutation V617F in the Budd-Chiari syndrome. In: Proceedings of the 47th Annual Meeting of the American Society of Hematology, Atlanta, December 10–13, 2005:7282.
Wanless IR. Vascular disorders. In: MacSween RNM, Burt AD, Portmann BC, Ishak KG, Scheuer PJ, Anthony PP, eds. Pathology of the liver. 4th ed. London: Churchill Livingstone, 2002:539-74.
Nowell P, Hungerford D. A minute chromosome in human chronic granulocytic leukemia. Science 1960;132:1497-1497.
Cools J, DeAngelo DJ, Gotlib J, et al. A tyrosine kinase created by fusion of the PDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hypereosinophilic syndrome. N Engl J Med 2003;348:1201-1214.
Zhao R, Xing S, Li Z, et al. Identification of an acquired JAK2 mutation in polycythemia vera. J Biol Chem 2005;280:22788-22792.
Sharma S, Texeira A, Texeira P, Elias E, Wilde J, Olliff SP. Pharmacological thrombolysis in Budd-Chiari syndrome: a single centre experience and review of the literature. J Hepatol 2004;40:172-180.
Orloff MJ, Daily PO, Orloff SL, Girard B, Orloff MS. A 27-year experience with surgical treatment of Budd-Chiari syndrome. Ann Surg 2000;232:340-352.
Mentha G, Giostra E, Majno PE, et al. Liver transplantation for Budd-Chiari syndrome: a European study on 248 patients from 51 centres. J Hepatol 2006;44:520-528.(Raymond T. Chung, M.D., A)