Management of Complex Patients with Budd–Chiari Syndrome
Article Outline
Therapy for patients with Budd–Chiari syndrome is well established. For those with commonly seen localized lesions, percutaneous transluminal angioplasty or stenting is the first-line treatment. Treatment methods for severely ill patients in whom intervention has failed, or those in a poor general condition, are worth exploring. From February 2002 to July 2008, 31 patients were referred to us. Eighteen patients had a failed intervention, 4 had undergone surgery, and 10 had conservative therapy. All had intractable ascites or/and hematemesis. The procedures carried out in this series included mesocavoatrial shunt in 10 patients, radical correction in 9, mesocavojugular shunt in 7 (including 2 mesojugular shunts), mesocaval shunt in 2, cavoatrial shunt in 2 (including a revision of cavoatrial shunt), and cavojugular shunt in 1. Surgical mortality and postoperative complications were both 3.2%. Twenty-eight patients had a mean follow-up of 40 months. Outcome of follow-up was measured as excellent, good, fair, poor, and death (28.6%, 53.6%, 10.7%, 3.6%, and 3.6%, respectively). The total mortality of the group is 6.5%. After appropriate preoperative evaluation and preparation, active and cautious treatment individualized to the underlying disease may help severely ill patients with Budd–Chiari syndrome.
Introduction
Budd–Chiari syndrome (BCS) is caused by occlusion of the hepatic veins (HVs) and/or the suprahepatic inferior vena cava (IVC). BCS is characterized by clinical manifestations of portal hypertension with or without IVC hypertension.1, 2 Medical therapy with diuretics, anticoagulation, or herbs has a very limited effect. Fibrinolytic therapy is instigated only if the disease is acute with a thrombotic etiology.
Surgical approaches include radiological treatment, collateral promoting, shunting, and radical correction procedures; liver transplantation would be indicated only at the end stage of the disease. Interventional treatment is the first choice if it is feasible.3 Other treatments should be investigated if intervention has failed and patients are extremely ill with intractable ascites or hematemesis.
The primary goal of treatment is radical clearance of the hepatocaval occlusive lesion or to decompress hypertension of the portal system and the IVC (particularly the former because it produces massive bleeding from esophagogastric varices and intractable ascites, which severely affect nutritional status and can lead to death). Based on our previous work,1, 2, 4 we report the treatment of 32 severely ill patients who were at high surgical risk or in whom interventional had previously failed.
Materials and Methods
From February 1, 2002, to July 31, 2008, 31 difficult-to-treat patients with BCS were referred to our surgical unit. Eighteen were male and 13 were female (age range, 3–63 years; mean age, 34.1 years). Illness duration was 2–72 months (mean, 48 months). Symptoms and signs are shown in Table I, and the preoperative laboratory data are given in Table II.
Table I. Clinical Features
| Symptoms and Signs | No. of Cases (N=31) | Percentage |
|---|---|---|
| Ascites | 31 | 100.0 |
| Pleural effusion | 11 | 35.5 |
| Varices in chest and abdominal wall | 22 | 71.0 |
| Hepatomegaly | 25 | 80.6 |
| Splenomegaly | 23 | 74.2 |
| Jaundice | 2 | 6.5 |
| Hepatic coma | 1 | 3.2 |
| Abdominal distention | 25 | 80.6 |
| Abdominal pain | 11 | 35.5 |
| Bleeding in upper gastrointestinal tract | 10 | 32.3 |
| Exertion palpitation | 13 | 41.9 |
| Cachexia malignant appearance | 7 | 22.6 |
| Edema of lower limb | 28 | 80.3 |
Table II. Preoperative Laboratory Data (N=31)
| Blood Test | Data (mean±SD) | Normal Value |
|---|---|---|
| ALT | 42.8±26.4 | 0–45 IU/L |
| AST | 41.8±17.5 | 0–40 IU/L |
| ALP | 155.7±100.6 | 45–150 IU/L |
| GGT | 66.6±30.4 | 7–50 IU/L |
| Total bilirubin | 33.3±12.2 | 1.7–26 μmol/L |
| Albumin | 33.6±7.6 | 35–55 g/L |
| Creatinine | 105.5±61.7 | 17.7–106 μmol/L |
| White blood cells | 4.8±1.6 | 4–10 g/L |
| Platelets | 107±34.5 | 100–300 g/L |
| Hemoglobin | 103.5±24.2 | 110–160 g/L |
Eight patients were in stage III and 23 patients in stage IV according to BCS severity.5 According to this classification, very early surgical mortality was 0%, 0%, 9%, and 21% for stage I, II, III, and IV, respectively.2 Figure 1 (right) shows a space-occupying lesion in the IVC extending to the right atrium; Figure 1 (left) shows a remarkably diminished size of the liver and intestine, and the abdomen is fully filled with ascites. Both patients were in stage IV, the end stage of the disease.
Fig. 1
Left: Computed tomography scan showing a remarkably diminished size of the live and intestine; the abdomen is full of ascites. Right: Cavography showing a space-occupying lesion in the inferior vena cava extending to the right atrium.
Eighteen patients had previously undergone percutaneous transluminal angioplasty (PTA) or stenting, or transjugular intrahepatic portosystemic stent-shunts (TIPPS). One subject had undergone two PTAs; another patient underwent three interventions. Figure 2 shows the patient who underwent two PTAs; computed tomography angiography (CTA) reveals stents that nearly (left) or completely (right) occlude the IVC (with very probable occlusion of the orifices of the HVs as well). Both patients had intractable ascites. Four patients recurred after previous splenectomy with devascularization, cavoatrial shunt, and mesocavoal shunt. The other nine patients were in very poor condition: intractable ascites with cachexia, acute bleeding in the upper gastrointestinal tract, or orthopnea because of thoracic effusions. They were all high-risk surgical patients.
Fig. 2
Computed tomography angiogram shows stent(s) completely (left) or almost completely (right) occludes the inferior vena cava, which likely occludes the hepatic veins.
Diagnosis was by clinical evaluation and confirmed by imaging: angiography in 24 patients, MRI in two patients, and CTA in five patients. Obstruction proved to be a relatively short segment of the IVC in 2 patients, long segment of the IVC in 26 patients, and an intrahepatic occlusion in 3 patients. All patients had HV involvement.
Preoperative Preparation
Autotransplantation of ascites was done for subjects with significant ascites. The frequency was from once a week to 3–4 times a week, using 1000–2000 mL per session. Albumin and plasma were used less often. Before autotransplantation, ascites was confirmed according to certain criteria.2 This was done by aspirating the ascites using a sterile technique into a blood transfusion bag and transfusing it back after filtering.
Surgical Technique
Six procedures were applied in this series: mesocavoatrial shunt in 10 patients (32.3% of study group); radical correction in 9 (29%); mesocavojugular shunt in 7 (22.6%, including 2 mesojugular shunts); mesocaval shunt in 2 (6.5%); cavoatrial shunt in 2 (6.5%, including 1 revision) and cavojugular shunt in 1 (3.2%).
Results
Immediate Results
Of the 31 patients, 1 had a graft infection after insertion of a mesocavojugular shunt. This resolved after a little more than 1 month through continuous irrigation and aspiration with 0.9% saline containing antibiotics. All patients were discharged from hospital except one, who died 24 hours after a radical correction. The average hospitalization days for the group is 46.4±9.8 days. The average operative bleeding is 1233±575 mL, and blood autotransfusion was given for every patient during operation.
Follow-up
Twenty-eight of 31 patients were followed up. Postoperative status was discerned mostly by telephone conversations or by letters. The main information we required was general health: ascites, hematemesis, and the ability to work. Quality of life was evaluated after surgery by means of validated questionnaires (36-Item Short Form Health Survey [SF-36]) and rated “excellent,” “good,” “fair,” “poor,” or “death” according to the score6 (Table III).
Table III. Operative Methods (N=31)
| Results | ||||||
|---|---|---|---|---|---|---|
| Surgical Procedure | Excellent, n (%) | Good, n (%) | Fair, n (%) | Poor, n (%) | Death, n (%) | Lost,∗ n (%) |
| Mesocavoatrial shunt (n=10) | 1 (10) | 7 (70) | 1 (10) | … | … | 1 (10) |
| Mesocavojugular shunt (n=7) | 2 (28.6) | 3 (42.9) | 1 (14.3) | 1 (33.3) | 1 (50) | 1 (14.3) |
| Cavoatrial (cavojugular) shunt (n=3) | 1 (33.3) | 1 (33.3) | 1 (11.1) | … | 1 (11.1) | … |
| Mesocaval shunt (n=2) | 1 (50) | 4 (44.4) | … | … | … | … |
| Radical correction (n=9) | 3 (33.3) | … | … | … | … | … |
∗Lost to follow-up. |
Duration of follow-up ranged from 4 months to 70 months (mean, 40 months). Eight patients (28.6%) were rated “excellent,” 15 (53.6%) were rated “good,” 3 (10.7%) were rated “fair,” 1 (3.6%) was rated “poor,” and 1 patient (3.6%) died. One death was recorded 3 years after insertion of a mesocavoal shunt. Total mortality of the group is 6.5%.
The Social Science version 11.5 for Windows (SPSS, Chicago, IL) was used for statistical analysis.
Discussion
Managing chronic BCS by medical means yields poor results. Outcome from a report of 48 patients who received conservative therapy stated that 39.6% of patients died, 43.8% did not improve, and only 16.7% improved.1 Surgery or endoluminal intervention is primarily designed to clear pathologic lesions at hepatocaval locations so as to reverse hepatofugal flow to forward hepatic venous flow, or to ease portal hypertension.
Many hospitals do not have expertise in BCS treatment because of a lack of experience in the treatment of such cases. These relatively simple cases were treated in a different hospital. All patients in this series had been treated by one or even several hospitals before referral to us. On admission, most were in poor physical condition, and management was difficult. The initial cause in this group was a membranous web obstructing the IVC but, before referral, such a simple lesion was not seen because they had already been treated (endoluminal therapy in 17 patients and surgery in 4 patients). Other causes included tumors in the IVC. The cause was unknown in some cases.
The clinical manifestations of the BCS are secondary to impaired liver function and the sequelae of portal hypertension.7 Intractable ascites, severe nutritional depletion, and hepatorenal failure are prominent features in almost all BCS patients at the late stage. In this series, all patients already had ascites. The other high-risk feature was bleeding in the upper gastrointestinal tract, which affected one-third of the study group. Preoperative management is very important and includes maintaining electrolyte balance and improving hepatic and nutritional status. Intravenous autotransfusion of ascites has been used with good results. These methods depend on pathophysiology, and all procedures were tailored to the underlying pathologic condition.
When the HVs and IVC were both occluded (particularly in cases of severe ascites), as long as the general condition permitted surgery and the patient actively asked for it, we applied the mesocavoatrial shunt. In 10 patients in this series, the result was good and suggested the grafts were patent, and no deaths occurred. The outcome depends on long-term observation, but we are confident because of our previous work on the mesocavoatrial shunt.2 This procedure requires strict attention to a sudden increase in cardiac preload because of a quick return of a large amount of congested blood in the lower body. This may incur acute cardiac failure, which we have experienced before.
When the IVC occlusion is longer than the part of the retrohepatic segment, and the patient has refractory ascites, even pleural effusion, which occurred in 35.5% in this series, becomes a very troublesome circumstance. A retrosternal mesocavojugular shunt (including mesojugular shunt) with externally ringed 14–16-mm ID grafts can be applied.8, 9 The vascular graft passed through a tunnel created with sufficient space behind the sternum. The abdominal end of the graft is anastomosed with the superior mesenteric vein and the cervical part with the internal jugular vein, both in an end-to-side fashion. The method remarkably reduces surgical trauma. We carried out this procedure on seven patients in this series and the result of follow-up was good (except for one patient who was lost to follow-up). The tunnel must be sufficiently wide to prevent graft compression. We suggest that a cardiac action gives a rhythmic compression to the graft, which pumps the blood flow in the graft upwards because of a pressure gradient. This procedure needs a sufficiently wide and fully dissected tunnel when the externally ringed graft is appropriately situated. We have confidence in this procedure, which is built on our previous success.
In one patient with repeated hematemesis, we decided on an emergency mesojugular shunt under compression by three-channel balloon while her hemoglobin was 40 g/L. During surgery, we could not find a patent portal vein and its tributary. We found that she had a remarkably large capitus medusa and that the pressure measured was as high as 59 cm H2O. A mesojugular shunt was carried out from this capitus medusa. Outcome was very good, and CTA confirmed graft patency (Fig. 3). After discharge, the patient subsequently took up employment, and has been working for more than 4 years.
Fig. 3
Computed tomography angiogram showing the patent graft of umbilical vein–jugular vein.
When the retrohepatic IVC occlusion leaves only the lowest short HV dilated and connected to the IVC, the cavoatrial shunt or cavojugular shunt can be carried out. There were two patients with a cavoatrial shunt and one patient with a cavojugular shunt in this series.
The mesocaval shunt is an abdominal procedure and is particularly appropriate for those with hepatic venous occlusion without hepatic parenchymal lesions or in whom the IVC was opened up by PTA or a stent. Encephalopathy is therefore unlikely to recur. Pure hepatic venous occlusion is rare in China (about 4%),1 whereas it is more common in Western countries. Long-term success depends on graft patency. Anticoagulation or antiplatelet therapy may help. The diameter of the graft also is important. In the case of cirrhosis, only an 8-mm-diameter graft was used to avoid encephalopathy. The liver function of patients with BCS is better than that for those with liver cirrhosis. The graft diameter we used was 13 mm or 14 mm. A better hemodynamic profile of the graft is important. It seems that the “H” or “C” shunt is not as good as the “(” shunt. From Figure 4, you may see that the shunt was switched from a “C” shunt to the “(” shunt. It gave better hemodynamics, as shown by the arrows showing the direction of flow.
Fig. 4
The mesocaval shunt at first was formed as a “C” (arrow). It was re-formed and became a “(” shape, which is much better hemodynamically. Left, photograph taken during surgery. Right, angiogram showing the flow direction (small arrows).
Radical correction was via a lateral thoracotomy. The IVC in the bare hepatic area and its lower part was longitudinally incised, and the lesions (e.g., membranous web or other occlusive material) were resected under direct vision. This procedure could be used to address occluded HVs and to reconstruct the upper part of occluded IVC.10, 11 It restores the normal anatomy in the hepatocaval area, so as to better reduce or normalize portal venous hypertension and IVC hypertension. The critical problem is the bleeding that is likely to occur during the procedure. We have prepared various balloons to temporarily stop bleeding. We recently used a balloon catheter inserted from the femoral vein and situated the balloon at the suprarenal venous level, which the distal part of the IVC is necessary for occlusion, and the balloon is opened up at this stage. Blood autotransfusion collects most of the lost blood and retransfuses it back through a right auricle intubation or other central vein. However, the risk of hemorrhage looms large.
We introduced this procedure in 1987.1, 9 In this series, we carried out such procedures in nine patients. The long-term patency for radical correction was 90% or less.4 In this series, the outcome of radical correction was good except for the one patient who died soon after surgery. It is vitally important to carefully select patients eligible for this procedure. The experience and skill of the surgeon also are very important. Radical correction targets the large HVs from their orifices but, if the orifices of the HVs cannot be found, the approach is to gradually resect liver tissue from inside the IVC until active venous bleeding is observed and then install a patch graft. Nothing can be done about the Chiari type of BCS by radical correction, but the portosystemic shunt works well. Five malignant tumors in the IVC can obstruct the IVC: renal cell carcinoma, hepatocellular carcinoma, adrenal carcinoma, hepatic metastasis, and primary leiomyosarcoma of the IVC.12 In cases of tumor obstruction, surgery could be permitted if the primary tumor could be resected and the patient's condition allows surgery.
The Denver shunt is an option to relieve ascites only.13 It is used only for very ill patients who cannot tolerate thoracotomy or laparotomy. Liver transplantation may be an appropriate treatment in the late stages of BCS when progressive hepatic venous decompensation or any intervention cannot be realized and liver disease is no longer reversible. Some studies showed that a previous portosystemic shunt (regardless of type) had no influence on the outcome of liver transplantation.14, 15 Transplantation does have drawbacks, such as the need for regular immunosuppressive medication, lifelong need for medical consultation, and susceptibility to infection. The choice is best determined by presentation, liver function tests, occlusion locations, and the available local expertise.16
For BCS patients who relapse after medical intervention or surgery and those with very severe conditions, further delay in decompression may reduce liver function and increase the chance of death.16 Timely treatment is essential.
Conclusion
During the treatment of complex patients with BCS, individualized means should be chosen depending on different pathologic lesions. The endovascular intervention should be the first choice if it is still feasible. Radical correction can be carried out for young patients after appropriate preoperative preparation. Less invasive surgical shunts are promoted for those with fairly bad general health condition. Various major shunts play important roles when the procedures are well performed. If the main trunk or tributaries of the portal system for shunting are unavailable, the captitus medusa can be used. It is of paramount importance to work to find ways to treat patients with complex lesions.
References
- Recognition and management of Budd-Chiari syndrome: report of 100 cases. J Vasc Surg. 1989;10:149–156
- . Budd–Chiari syndrome. Curr Probl Surg. 1996;33:83–211
- Primary Budd-Chiari syndrome: outcome of endovascular management for suprahepatic venous obstruction. J Vasc Surg. 2006;43:101–108
- Evolution on management of hepatocaval syndrome: a team view from 2654 patients. ANZJ Surg. 2005;75:55–63
- . Recognition and management of Budd-Chiari syndrome. Experience with 143 patients. Chin Med J (Engl). 1989;102:338–346
- . The MOS 36-Item Short Form Health Survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992;30:473–483
- Budd-Chiari syndrome: current management options. Ann Surg. 2001;233:522–527
- . Management of Budd-Chiari syndrome. Experience from 430 cases. Asian J Surg. 1996;19:23–30
- . Management of high risk Budd-Chiari syndrome with mesojugular shunt: report of 16 cases. Chin J Surg. 1994;32:611–614
- Radical correction of Budd-Chiari syndrome. Chin Med J. 2007;120:622–625
- . Evaluation of radical surgery as means of managing Budd-Chiari syndrome. Chin. J Thorac Cardiovas Surg (Chin). 1995;11:132–134
- . Budd-Chiari syndrome: in evolution. Eur J Gastroenterol Hepatol. 2005;17:33–35
- Peritoneovenous shunting for intractable chylous ascites complicated with lymphangioleiomyomatosis. Intern Med. 2008;47:281–285
- Comparison of transjugular and surgical portosystemic shunts on the outcome of liver transplantation. Arch Surg. 1994;129:1018–1024
- A 27-year experience with surgical treatment of Budd-Chiari syndrome. Ann Surg. 2000;232:340–352
- . To restrict indication for stenting of the inferior vena cava and liver transplantation in patients with Budd-Chiari syndrome. Front Med China. 2007;1:130–135
PII: S0890-5096(09)00180-0
doi:10.1016/j.avsg.2009.07.013
© 2010 Annals of Vascular Surgery Inc. Published by Elsevier Inc All rights reserved.
