Transplantation for Budd-Chiari Syndrome

Etiology

BCS results from diverse causative factors ( Table 20-1 ). Their incidence varies significantly in different parts of the world. In India and other parts of Asia, many cases are idiopathic or caused by vena cava webs. Thus, in a study of 71 BCS patients in India seen between 1992 and 1997, 42 were idiopathic and 18 had vena caval membranes. Membranous webs may be congenital or caused by fibrosis from chronic thrombosis. Hepatic vein obstruction because of hepatocellular carcinoma is common in South Africa. Myeloproliferative disorders (MPDs) and other definable hypercoagulable states account for the majority of BCS cases in Europe and Western countries. The recent discovery of the V617F JAK2 mutation is helpful in identifying an MPD cause of BCS but does not distinguish between the different disease entities in this group. JAK2 mutation is present in 95% of patients with polycythemia vera and approximately 50% with essential thrombocythemia (ET) or primary myelofibrosis. The specific cause should be sought in every patient because the presurgical and postsurgical management and outcome are to a large extent dependent on prevention of further hepatic vein occlusion.

Pathology and Pathophysiology

The multiple underlying conditions listed in Table 20-1 do not define the morphological characteristics of the obstruction or the site involved. In Asian countries the common finding is a membranous web in the inferior vena cava, specifically at the suprahepatic region, resulting in progressive thrombosis of the inferior vena cava and the liver ostia. ^∗

∗ References .

In the United States and Europe, hematological disorders leading to thrombosis are the most frequent underlying mechanisms responsible for occlusion of the main hepatic veins. This thrombotic form of the syndrome accounts for up to 80% of the patients presenting for surgical treatment of BCS. Compression or invasion of the hepatic veins by tumors or granulomas occasionally results in acute or chronic occlusion or thrombosis of the hepatic vein, leading to clinical expression of BCS. The reason for the predominant localization of thrombosis at the level of the liver portion of the vena cava is not known, although for many years the underlying mechanism was believed to be an endophlebitis of the hepatic vein. The weight of the clinical evidence now indicates that the primary process is thrombotic rather than inflammatory. This conclusion is supported by autopsy studies demonstrating fibrous obstruction as the final product of thrombus organization.

Histological findings characteristic of rapidly progressive hepatic vein occlusion include intense congestion and cellular atrophy ( Fig. 20-1 ). Areas of necrosis and cell dropout are often superimposed. Centrilobular extravasation of red cells and necrosis can extend to the periphery of the lobules; however, the portal areas are preserved. In patients with chronic hepatic vein thromboses, cardiac cirrhosis develops. The typical finding in this group of patients is pericentral vein fibrosis.

Hepatic venoocclusive disease is a distinct form of BCS characterized by central venous dilatation, centrilobular necrosis, and intimal thickening throughout the smaller liver venules at the microscopic level. The principal cause of this disease is continuous exposure to hepatotoxic pyrrolizidine alkaloids. Hepatic venoocclusive disease has also been reported to develop after high-dose chemotherapy and in the setting of hematopoietic stem cell transplantation.

The wide range of topographical and pathological causes encompassed under the BCS label merits a better classification to facilitate patient evaluation, prognostic prediction, and rational treatment. Ludwig et al suggested a more accurate and simplified classification of BCS on the basis of morphological features (thrombotic versus nonthrombotic), site of the lesion (inferior vena cava, major or small hepatic veins), and cause. Such a classification would help clarify the pathophysiological abnormalities, histopathological characteristics, and spectrum of diseases that lead to the hepatic venous outflow obstruction and would aid in the selection of the appropriate medical management of patients with BCS.

Clinical Presentation and Diagnostic Evaluation

The clinical picture depends on the tempo and extent of hepatic vein occlusion and is directly related to liver congestion with subsequent hepatocyte necrosis and ultimately fibrosis.

Although the presentation of patients suffering from BCS is variable, the clinical history correlates with the acuity of disease onset. A minority of BCS patients present with the classical symptoms and findings of gradual onset of right upper quadrant abdominal pain, tender hepatomegaly, and ascites. The right upper quadrant pain is often preceded by weeks to months of vague abdominal distress. Splenomegaly and an enlarged caudate lobe (palpable epigastric mass) are frequent additional findings. Patients with occlusion of the inferior vena cava have lower extremity edema, distended abdominal flank and back veins, and albuminuria. Sudden-onset upper abdominal pain with vomiting and rapid accumulation of ascites with an enlarged, tender liver are less common and are associated with acute onset of disease. Only rarely will these patients experience fulminant liver failure, which is characterized by massive liver necrosis and consequent liver coma, severe coagulopathy, and hypoglycemia. In the chronic type of BCS the development of end-stage liver disease may not be associated with the typical stigmata of chronic liver disease seen in cirrhotic patients. Fatigue and poor nutritional status are common; however, spider angiomas and palmar erythema are unusual, and jaundice is usually mild. A common physical finding is lower extremity edema resulting from partial or complete occlusion of the retrohepatic vena cava by the hypertrophied caudate lobe. Caval obstruction may decrease kidney perfusion pressure, thereby contributing to the development of kidney failure. Despite severe portal hypertension in this patient population, variceal bleeding is uncommon.

Routine laboratory test results indicate variable degrees of liver dysfunction. Serum transaminase, bilirubin, and alkaline phosphatase levels are normal or mildly elevated. Serum albumin level may be decreased, and albumin levels correlate well with the severity of liver injury and the magnitude of protein loss into ascitic fluid. Prothrombin time may be mildly prolonged.

Liver biopsy typically demonstrates intense centrilobular congestion and pressure necrosis of the liver parenchyma. Cardiac cirrhosis is seen in the advanced stage of the disease. Liver biopsy may not be necessary in every patient; however, chronic BCS patients with biopsy findings of cirrhosis and marginal liver reserve should be considered for liver transplantation rather than a decompressive procedure.

Diagnostic Imaging

Noninvasive hepatic imaging has assumed a progressively important role in the diagnostic evaluation of BCS. Duplex ultrasonography and color Doppler ultrasound imaging accurately define the flow parameters of the hepatic veins, the portal vein, and the inferior vena cava in most patients ( Fig. 20-2 ). Often ultrasonography is the initial imaging modality in patients being evaluated for hepatic venoocclusive disease. However, the sensitivity of ultrasonography is affected by the patient’s body habitus, echo density of the liver, and the amount of upper abdominal bowel gas. Ultrasonography does provide valuable information in most patients, but it is less sensitive for the detection of hepatic masses than computed tomography (CT) or magnetic resonance imaging (MRI).

Multiphase infusion CT evaluation accurately images the hepatic parenchyma and gives valuable information regarding hepatic venous and portal venous flow ( Fig. 20-3 ). An enlarged caudate lobe is characteristically observed. CT arteriography and venography accurately assess the anatomical features of hepatic vasculature. CT also defines hepatic morphological features and helps distinguish BCS from other processes that may mimic venoocclusive disease.

MRI is also very sensitive in the detection of hepatic mass disease, as well as in evaluating flow characteristics in the hepatic vasculature ( Fig. 20-4 ). MRI can provide insight into diffuse hepatic disease. Gadolinium-based contrast agents are commercially available and are routinely employed in liver imaging. As with CT, the technique for contrast-enhanced MRI includes multiple phases of enhancement. In the preoperative evaluation, specific flow-sensitive sequences are often employed to evaluate splenoportal morphological features and hepatovenous morphological characteristics. Although not routinely necessary, magnetic resonance cholangiopancreatography may be added to preoperative protocols in those patients who are suspected of harboring biliary pathological conditions.

In patients with known or suspected BCS the noninvasive imaging evaluation is used to confirm the diagnosis and uncover unsuspected malignancies. This is particularly important because hepatic malignancy is more common in patients with BCS than in the general population.

Hepatic venography remains the gold standard for the imaging diagnosis of BCS. The hepatic veins may be cannulated from a femoral or jugular approach. A spiderweb appearance of the intrahepatic veins is confirmatory ( Fig. 20-5 ). Hepatic venography also allows measurement of the hepatic wedge pressure, and inferior venacavography can be performed during the same catheterization in the rare occasion shunt surgery is contemplated.

Thorough radiological evaluation of BCS is best accomplished with a multimodality approach. Cross-sectional imaging (CT or MRI) is used to examine the entire abdominal contents, as well as to confirm and characterize hepatic blood flow. The measurement of flow parameters is usually accomplished with ultrasonography. In patients who are not transplant candidates, venography confirms the diagnosis of BCS and provides preoperative assessment before the transjugular intrahepatic portosystemic shunt (TIPS) procedure. Surgical shunts are rarely used now.

Once hepatic vein obstruction is demonstrated, its cause should be established. Table 20-2 lists suggested diagnostic studies.

TABLE 20-2

Diagnostic Tests for Evaluation of Cause of Budd-Chiari Syndrome

Complete blood count and peripheral blood smear

Liver function blood tests

Liver imaging studies

Liver biopsy

V617F JAK2 mutation in peripheral blood or bone marrow

Bone marrow aspiration and biopsy

Factor V Leiden mutation

Prothrombin gene mutation (G20210A)

Antithrombin III

Protein C (total and functional)

Protein S (total and functional)

Lupuslike anticoagulant (anti-beta 2 glycoprotein 1 antibodies)

Anticardiolipin antibodies

Homocysteine level

Serum protein electrophoresis

Flow cytometry for CD55, CD59 expression (paroxysmal nocturnal hemoglobinuria)

Etiological Diagnosis: Myeloproliferative Disorders and Other Conditions

In Western countries the recognition of overt or occult MPDs—especially polycythemia vera, ET, and paroxysmal nocturnal hemoglobinuria—as causes of BCS has been documented for more than 20 years. Other MPDs (agnogenic myeloid metaplasia or primary myelofibrosis, chronic granulocytic leukemia, and erythroleukemia) are rarely associated with BCS. MPDs are the cause of 40% to 70% of BCS cases in the United States and Europe. Coagulation abnormalities, both hereditary and acquired, are also causes of BCS. Some of these disorders were recognized during the 1990s. If prevention of further thrombosis is to be achieved, complete hematological evaluation of each patient with BCS for evidence of MPDs and other hypercoagulable states becomes imperative. Consideration of one series of patients studied serially illustrates several pertinent points in this regard.

Between 1987 and 2007, 25 patients with BCS underwent orthotopic liver transplantation (OLT) at Baylor University Medical Center in Dallas, Texas ( Table 20-3 ). Data for this study were collected through a prospectively maintained longitudinal database and chart review. The diagnosis of BCS was confirmed by imaging studies, including MRI, CT, Doppler ultrasonography, and angiography. Pathological examination of the native liver after transplantation was accomplished in all cases.

Bone marrow examination was performed in all patients except the first. The diagnosis of MPD was based on bone marrow morphological abnormalities in conjunction with peripheral blood counts. Spontaneous erythroid colony-forming assays were not done. The JAK2 mutation was not identified until 2005, and so analysis for this abnormality was not carried out in these patients. Hypercoagulability evaluation was conducted in all patients. An expanded battery of studies was used as new causes for thrombophilia were established (see Table 20-2 ). The first 13 patients underwent transplantation before 1996, before widespread recognition of factor V Leiden, the prothrombin gene mutation, and hyperhomocysteinemia as causes of the hypercoagulable state. Laboratory studies performed on the earlier patients included functional and total protein C and protein S, antithrombin III, anticardiolipin immunoglobulin G and M antibodies, lupuslike anticoagulant, sucrose hemolysis, and serum protein electrophoresis. This protocol was not in place when the first patient with BCS was seen, but it was used subsequently. Causative diagnoses were established in 22 of the 24 remaining patients (92%).

The purpose of this study was to identify the origin of BCS and to determine whether antiplatelet treatment rather than anticoagulation would be effective in patients with underlying MPDs.

Table 20-3 lists the diagnosis, treatment, and outcome of the 25 patients in the Dallas study. Seventeen patients were women, and 8 were men. Age at time of OLT ranged from 9 to 61 (mean 33) years. Time from onset of BCS to OLT ranged from approximately 4 months to 4 years. Seventeen patients (68%) had evidence of an MPD as the cause of their BCS. Seven of these patients had a clinical picture consistent with polycythemia vera, 2 patients had ET, and 8 patients had unclassifiable MPDs. Cytogenetic analysis was performed in 5 patients (patients 2, 7, 8, 11, and 16), and all were normal. Two patients (patients 1 and 19) were classified as having idiopathic BCS. The 3 remaining patients had protein C deficiency (patient 13), sarcoidosis (patient 15), and the prothrombin gene (G20210A) mutation (patient 17). Three patients (patients 16, 22, and 24) were heterozygous for factor V Leiden in addition to having an MPD. One patient (25) had factor V Leiden plus antiphospholipid antibody syndrome.

It was possible to arrive at a causative diagnosis in nearly all patients with BCS. Sixty-eight percent (17 of 25) had evidence of MPD. Antiplatelet therapy with hydroxyurea and aspirin was employed as the antithrombotic regimen in the MPD patients (see Table 20-3 ).

Liver Transplantation