Telephone: 310-423-2641; FAX: 310-423-0565
Management of Budd-Chiari syndrome
Article first published online: 18 OCT 2006
Copyright © 2006 American Association for the Study of Liver Diseases
Supplement: Liver Transplantation in the 21st Century: Clinical, Pathologic and Radiologic Correlation in Liver Transplantation
Volume 12, Issue Supplement 2, pages S23–S28, November 2006
How to Cite
Klein, A. S. (2006), Management of Budd-Chiari syndrome. Liver Transpl, 12: S23–S28. doi: 10.1002/lt.20941
- Issue published online: 18 OCT 2006
- Article first published online: 18 OCT 2006
- 1Medical therapy alone is rarely sufficient for long-term management of patients with hepatic vein thrombosis.
- 2Enthusiasm for intravascular stents (transjugular intrahepatic portosystemic shunt [TIPS] or vena caval stents) for the management of Budd-Chiari syndrome must be tempered by the limited interval of expected utility, the likelihood of stent occlusion/revisions, and the potential complications that stent migration would impose upon a subsequent liver transplant.
- 3Both decompressive shunts and liver transplantation provide excellent long-term survival for patients with the Budd-Chiari syndrome. The determination of which surgical procedure is most appropriate is aided by assessment of the etiology of hepatic vein thrombosis, hepatic reserve, liver histology, and splanchnic venous anatomy.
- 4Progressive hepatic damage may develop in patients with Budd-Chiari syndrome who have patent surgical shunts or TIPS. Lifelong follow-up and tracking of hepatic function are indicated. Some patients with shunts will require salvage with liver transplantation.
- 5Long-term anticoagulation should be considered after transplantation, even in patients who do not have an identifiable coagulation disorder. Liver Transpl 12:S23–S28, 2006. © 2006 AASLD.
The primary goal of treatment for patients with Budd-Chiari syndrome (BCS) is reduction of hepatic congestion and associated sequelae such as ascites. A secondary but essential goal is the prevention of recurrence. It should be recognized that although membranous occlusion of the inferior vena cava (IVC) and/or hepatic veins is a common cause of hepatic venous outflow obstruction in Asian populations, it is an unusual entity in western countries. The intravascular webs responsible for the “eastern” form of BCS are often amenable to mechanical disruption and obliteration by surgical or interventional radiological techniques. Unfortunately, these same techniques are of little value in the management of the western form of BCS, which is characterized by hepatic venous thrombosis. The remainder of the discussion will focus on the role of surgery in the management of hepatic vein thrombosis.
The etiology of BCS must be considered when planning therapeutic options. Patients with paroxysmal nocturnal hemoglobinuria, for example, suffer from disseminated clotting, and prognosis is considered so poor with surgery that some would consider this disease a relative contraindication to transplantation or construction of a surgical shunt. Associated nonhematological pathologies at the time of presentation should also be taken into consideration when determining the best treatment for BCS. These include, but are not limited to, the presence of cirrhosis, previous attempts at decompressive surgery, hepatocellular carcinomas, overall physiologic status, and coexistent viral hepatic infection.
Initial management of BCS includes control of the underlying disease, reduction of ascites with medical therapy and paracentesis, and anticoagulation. Despite sporadic reports of successful treatment of BCS with anticoagulation and thrombolytic therapy alone, most of the patients described in these studies have relatively short periods of follow-up. Extrapolation of results from such studies to recommendations that medical therapy by itself is appropriate for long-term management is not warranted at this time. The natural history of unrelieved hepatic venous outflow obstruction is generally progressive liver failure with fibrotic/cirrhotic changes histologically.1, 2 Cameron et al.3 have demonstrated progressive hepatocyte atrophy and impaired cellular regeneration in the setting of persistent sinusoidal congestion. In a study reported by McCarthy et al.,2 12 of 14 patients with BCS who were managed nonsurgically died within 6 months of diagnosis. An exception to this poor prognosis may be realized by the subset of patients determined to have incomplete hepatic venous obstruction.
SURGICAL AND RADIOLOGICAL MANAGEMENT
Prior to the application of liver transplantation (LT) for patients with BCS, surgical therapy focused on decompression of the obstructed liver by the conversion of the portal vein from an inflow vessel to an outflow tract. This was accomplished by construction of a mesenteric—systemic or portal—systemic shunt. Some researchers have demonstrated excellent survival (75–94% at 5 years) with mesocaval or portocaval shunts.4–6 Hemodynamic and anatomic factors may limit the options for shunt selection. The caudate lobe, which drains directly into the IVC by means of several short veins, often hypertrophies in response to the dysfunctional state of the remaining part of the organ. This can result in external compression of the IVC and subsequent partial or complete obstruction. Marked obstruction, defined as a pressure gradient exceeding 20 mm Hg between the infrahepatic IVC and the right atrium, is thought by many to be a contraindication to standard mesocaval pr portocaval shunt because these forms of bypass will not effectively decompress the liver when the systemic venous reservoir is under high pressure. Shunt thrombosis is likely under these circumstances, and even if the shunt remains patent, sinusoidal hypertension will persist. A novel solution for patients with BCS with caval obstruction proposed by Cameron and Maddrey7 was to create a direct connection between the superior mesenteric vein and the right atrium (Fig. 1). The experience with mesoatrial shunts at Johns Hopkins was encouraging, with 5-year patient survival of 68%. However, the primary patency rate of the mesoatrial shunt was only 46%.4 With the advent of effective endovascular therapy, an alternative solution was proposed to eliminate the caval obstruction by placing a metallic stent in the IVC and then constructing a standard mesenteric-systemic surgical shunt.8 (Fig. 2). Although this was effective in the short term, in some cases, the indwelling caval stents have been shown to migrate above the diaphragm and even into the right atrium. This event can have dramatic consequences if LT is indicated in the future, as discussed below.
In the absence of vena caval obstruction, the selection of which shunt procedure is best has been a controversial topic. Orloff et al.6 reported 32 patients with BCS who underwent side-to-side portocaval shunt, 31 of whom were alive at 3.5 to 27 years after surgery. This success rate has not been replicated elsewhere and may speak to the particular expertise or patient population present in a single institution. Others have noted marked anatomic barriers to performing a portacaval shunt for patients with BCS; notably, the presence of caudate lobe hypertrophy often renders a side-to-side portocaval shunt technically difficult. In one report, 75% of patients required a caudate lobe resection before a side-to-side portocaval shunt could be constructed. Alternatively, the mesocaval shunt procedure can be performed rapidly, which leaves the porta hepatis unscathed, and in the event of a subsequent LT, can be easily ligated. The consequences of surgical shunting in the setting of LT are discussed below.
Regardless of which shunt procedure is preferred, in many instances, surgical options may be limited simply because local expertise is lacking. With increased availability of and enthusiasm for transjugular intrahepatic portosystemic shunt (TIPS), the requisite institutional expertise with surgically constructed portal systemic shunts is often absent. Even in training programs with access to a high volume of patients with portal hypertension, experience with portal venous decompression surgery is diminishing. Between 1995 and 1999, the surgical chief residents at Johns Hopkins finished their training with a combined experience of 4 portal-systemic decompression procedures.
Insertion of a TIPS has been shown to be an effective means of decompressing the congested liver in patients with BCS.9–11 Molmenti et al.12 have demonstrated that TIPS effectively decompresses the splanchnic circulation. In their series of 11 patients with BCS who were treated with TIPS, mean portal pressure was reduced by 43% and the portal vein–right atrium pressure gradient was reduced 73%. Attwell et al.9 reported 17 patients with BCS initially treated with TIPS, 14 (82%) of whom improved or stabilized initially and 3 of whom died. However, 5 (38%) of 14 survivors subsequently underwent LT. This series and others suggest that TIPS may be particularly useful as a temporizing measure, but by many accounts, it lacks the durability of LT as a long-term solution. TIPS has been used successfully in the emergent setting of managing patients with BCS with fulminant hepatic failure.13 Ryu et al.,13 from the University of Colorado, have shown that hepatic synthetic function improves and ascites is decreased after TIPS placement. However, TIPS occlusion was common in the Colorado series (50%), again emphasizing that in many cases, the interventional radiological approach is best viewed as a bridge to transplantation rather than a destination therapy.
From a practical standpoint, the radiological expertise required for successful TIPS insertion is far more likely to be present than is the local availability of a surgeon with the skills and experience necessary for construction of a mesenteric-systemic shunt. Kavanagh and associates14 have asserted that TIPS is more compatible with transplant surgery, should it be required, than are traditional surgical procedures. However, misplacement or migration of TIPS is not uncommon (Fig. 3). In a series of 12 patients from Duke University who received a TIPS before LT, 4 (33%) had misplacement of the stent inferiorly into the main portal vein.15 At the time of a subsequent LT, the blood transfusion requirements in the TIPS recipients were more than 2 times greater than that required for liver transplant patients who did not have a previous TIPS. As described by Slakey et al.,4 the penalty applied to patients with BCS treated initially by TIPS may not be limited to the need for multiple revisions and angioplasties. One patient who underwent 7 revisions after the placement of her first TIPS developed hepatic venous outflow obstruction from the multiple TIPS stents. The metallic stents, which extended from the patient's right atrium to the confluence of her superior mesenteric vein and splenic vein, were densely incorporated into her heart and her splanchnic vasculature. The technical demands of the LT itself were further complicated by the need after LT to place suprahepatic IVC stents to alleviate hepatic venous outflow obstruction in the new hepatic allograft caused by remnants of the TIPS (Fig. 4) Others have noted vena caval stenosis as a consequence of TIPS placement for BCS.16 In the largest clinical series to evaluating BCS patients who underwent LT, Mentha et al.17 determined that pre-LT TIPS was directly related to decreased patient survival.
Lifelong follow-up and tracking of hepatic function is indicated in patients with BCS treated with surgical shunts or TIPS. In some cases, progressive liver injury occurs despite portal venous decompression (which may be incomplete). It is the bias of a growing number of surgeons that portal-systemic shunts, which require dissection and/or anastomosis of the portal vein itself, are indicated in fewer situations. Because portacaval shunts require extensive dissection of the porta hepatis, they have the disadvantage of potentially complicating subsequent LT (e.g., in patients who develop cirrhosis, shunt occlusion, or hepatic neoplasm). Furthermore, nonselective shunts must be disconnected at the time of LT—a major undertaking for a portacaval or central splenorenal shunt involving considerable blood loss and increased operating time. These problems are avoided by mesocaval shunts, which are performed away from the porta hepatis, which is then not a reoperative field at the time of LT. At the conclusion of the LT, the mesocaval shunt is divided and oversewn or stapled. The mesocaval shunt may actually simplify the LT procedure by providing portal and systemic venous decompression during intraoperative veno-venous bypass through a single femoral vein cannula.
Brems et al.18 stratified LT recipients on the basis of the presence or absence of a prior portal-systemic shunt. Mean blood loss and hospital stay were highest in the subgroup of patients with a portacaval shunt. In a subsequent series from Pittsburgh, a preexisting mesocaval shunt appeared to confer a survival advantage.19 The 5-year posttransplantation survival for patients with a previously placed end-to-side portacaval shunt was 35%, 50% for patients with a side-to-side portacaval shunt, and 95% for patients with a mesocaval shunt. The survival for patients without shunts was 65%. As noted above, the effect of radiological therapies for BCS at subsequent LT cannot be neglected. It should be emphasized that intravascular stents (either a TIPS to directly decompress the hypertensive splanchnic circulation, or a transcaval stent placed to reduce the transhepatic vena caval pressure gradient caused by caudate lobe hypertrophy, thereby improving the efficiency of a surgically placed mesocaval shunt) become densely incorporated into their resident blood vessels. Should these stents migrate into the main portal vein, the suprahepatic IVC, and/or the right atrium, marked technical difficulties may be encountered during isolation of the hepatic vasculature and subsequent explantation of the liver.
The first LT for BCS was performed in 1974. Over the next 15 years, Halff and others reported 1- and 3-year posttransplantation patient survival rates for BCS (∼70 and 45%, respectively) that were inferior to what was observed for adult liver recipients in general.20 More recent series, however, have documented better survival rates, partly because of the decreased interval between the onset of symptoms and initiation of therapy and the early institution of anticoagulant and/or antithrombotic therapy after transplantation with lifelong use for patients with a definable hypercoagulable state (Table 1). LT series from Kings College, UCLA/Baylor, and Hanover have reported 3-year survival rates of 95, 76, and 69%, respectively.21–23 The largest clinical series of patients with BCS treated by LT is derived from data reported by Mentha et al.,17 who surveyed the European Liver Transplantation Registry between 1988 and 1999. A total of 295 patients transplanted for BCS were identified, and complete follow-up data were obtained for 248 patients. The actuarial survival was 75.6, 71.4, and 68% at 1, 5, and 10 years after transplantation, respectively. Late mortality was low in this study: only 9 patients died after 1 year. The concern that patients with myeloproliferative diseases (a common etiology of BCS) would die as a result of their underlying disease (e.g., due to leukemic transformation) is not supported by their data. However, 27 patients (11%) developed some form of venous thrombosis (6 patients had hepatic venous thrombosis) despite anticoagulation therapy.
|Center (Author)||Yr||# Pts||1-yr||3-yr||5-yr|
|King's College (Srinivasan)||2002||19||95||95||95|
|Univ. Colorado (Attwell)||2004||10||90||NA||80|
|European Registry (Mentha)||2006||248||76||NA||71|
In a recent study from UCSF, 3 (27%) of 11 patients developed recurrent BCS.24 Other series, however, have suggested that aggressive anticoagulation therapy has markedly reduced the incidence of recurrent BCS after LT (Table 2). Lifelong anticoagulation after LT for BCS is advocated by a growing number of transplant programs, even in the absence of an identifiable hypercoagulable state. This is not without penalty. The Cambridge series reported a 44% incidence of nonfatal hemorrhage when a policy of early posttransplantation anticoagulation was instituted, and the European multicenter clinical series identified 27 patients (11%) who had experienced clinically important hemorrhage that was related to anticoagulation.17, 25 Nonetheless, there is general agreement that complications that result from bleeding are generally more amenable to treatment than are complications that result from thrombosis.
|Author||# Patients||# (%) recurrences|
LT poses specific challenges for patients with BCS. The obstructed liver is generally enlarged, firm, and difficult to mobilize during the hepatectomy. A diffuse fibrotic reaction in the retroperitoneum, perhaps related to the hepatic vein thrombotic process, increases the difficulty of identifying, mobilizing, and controlling the IVC. Because the caudate lobe is enlarged and the hepatic veins orifices are occluded, the “piggyback” technique of LT may be particularly tricky. Control of the vena cava may actually require incision of the diaphragm and isolation of the vena cava within the pericardial sac (G.B. Klintmalm, personal communication). Yan et al.26 have described a case of adult-to-adult living donor LT for BCS that used the piggyback technique. However, it is not clear whether that procedure, performed in China, was for a patient with a vena caval web or a hepatic vein thrombosis.
Recognizing the spectrum of acuity and severity with which BCS presents, there is little controversy that LT offers the most effective therapy for the minority of individuals with either fulminant hepatic failure or the chronic cirrhotic form of BCS. Moreover, in patients who develop hepatic venous thrombosis as a result of metabolic defects localized to the liver (e.g., antithrombin III or protein C deficiency), LT offers the singular benefit of being curative. However, most BCS patients present with acute or subacute manifestations of hepatic venous outflow obstruction. Assessment of hepatic reserve and interpretation of the liver biopsy are the key elements in the algorithm used to select LT vs. alternative therapies. Although hepatic function may recover after shunting even when preoperative biopsies demonstrate the presence of fibrosis, there is general agreement that such patients will do best if transplanted. Unfortunately, clinical outcomes alone cannot be used to determine whether LT is advisable for patients with BCS. A more restricted use of LT is mandated by (1) the widening gap between the number of patients in need of LT and the static pool of donated organs, (2) the unpredictable availability of donor organs, (3) the need for and consequences of lifelong immunosuppression, and (4) the dramatically higher cost of transplant vs. nontransplant therapies.
- 14Acute Budd-Chiari syndrome with liver failure: the experience of a policy of initial interventional radiological treatment using transjugular intrahepatic portosystemic shunt. J Gastroenterol Hepatol 2004; 19: 1135–1139., , , , , .
- 26Adult-to-adult living donor liver transplantation using right lobe graft: report of 24 cases. Zhonghua Yi Xue Za Zhi 2006; 86: 411–415., , , , , , et al.