Budd-Chiari syndrome, portal vein and mesenteric vein thrombosis in a patient homozygous for factor V Leiden mutation treated by TIPS and thrombolysis

Authors


Dr Leebeek Department of Haematology, University Hospital Rotterdam ‘Dijkzigt’, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands.

Abstract

We present the first case of Budd-Chiari syndrome in association with portal and mesenteric vein thrombosis in a patient homozygous for the factor V Leiden mutation. She was treated by transjugular intrahepatic portosystemic stent (TIPS) placement followed by local thrombolytic therapy. Venous outflow from the liver was established and the thrombi in the portal and mesenteric veins were lysed completely. This therapeutic approach may be used for such patients with this severe thrombotic event, who generally have a poor prognosis.

Acute Budd-Chiari syndrome (BCS) caused by hepatic vein thrombosis is a serious thrombotic event frequently associated with myeloproliferative disorders or hereditary coagulation disorders, such as antithrombin III, protein C and protein S deficiency ( Reynolds, 1993). Recently a few patients have been reported in whom BCS was associated with activated protein C (APC) resistance, most of whom were heterozygous for the factor V Leiden mutation ( Denninger et al, 1995 ; Mahmoud et al, 1995 , 1997a). Acute BCS leads to deterioration of liver function, portal hypertension and ascites. In some patients it results in acute liver failure for which liver transplantation is necessary. Patients in whom the hepatic vein thrombosis is associated with thrombosis of the portal vein or the superior mesenteric vein are considered a subgroup with a very poor prognosis and have a reported 1-year actuarial survival of 22% ( Mahmoud et al, 1997 b). We report a young patient presenting with hepatic vein, portal vein and superior mesenteric vein thrombosis, who was succesfully treated by transjugular porto-systemic stent (TIPS) placement followed by local thrombolytic treatment. She was found to be homozygous for the factor V Leiden mutation.

METHODS

Protein assays

Routine coagulation tests (APTT, PT, Fibrinogen, alpha2-antiplasmin, antithrombin III, lupus anticoagulant) were performed according to standard coagulometric and amidolytic assays. ELISA techniques were used to measure protein C, protein S and anticardiolipin antibodies. Latex agglutination tests were used to measure D-dimer and fibrin(ogen) degradation products (FDP). Activated protein C resistance was measured using a modified APC resistance test as described previously ( Engel et al, 1996 ).

DNA analysis: detection of factor V Leiden

A 220 bp fragment was amplified by PCR from genomic DNA. Both the PCR reaction mixture and cycling conditions used in this study have been described previously, as well as the nucleotide sequence of both the forward and reverse primers ( Bertina et al, 1994 ; Koeleman et al, 1994 ). The PCR products were digested with 6 U Mnl-I and electrophoresed on a 2% agarose gel. The following restriction digestion products were obtained: 67, 37 and 116 bp in the 1691 G allele (normal) and 67 and 153 bp in the 1691 A allele (mutated). Detection of G20210A prothrombin variant was done by PCR from genomic DNA followed by HindIII digestion as previously described ( Poort et al, 1996 ).

CASE REPORT

A 27-year-old female presented with 2 weeks of epigastric pain. She complained of nausea and vomiting without haematemesis. On physical examination there were no signs of ascites or hepatosplenomegaly. Laboratory investigations revealed a haemoglobin level of 14.5 g/dl (normal 11.8–15.0), haematocrit 0.44 l/l (normal 0.35–0.45), leucocyte count 10.1 × 109/l (normal 4.0–10.0) with a normal differential, and platelet count 374 × 109/l (normal 140–360). Renal function was normal. Liver function tests were as follows: aspartate aminotransferase 62 U/l (normal < 30), alanine aminotransferase 101 U/l (normal < 30), gamma-glutamyl transferase 32 U/l (normal < 30), alkaline phosphatase 108 U/l (normal 20–120), and total bilirubin 24 μmol/l (normal < 16). Ultrasound showed mild ascites and a partial thrombosis of the portal vein. Heparin treatment was started. Two days later ultrasound showed an increase of the size of the portal vein thrombus. She was then referred to our hospital for further treatment. Doppler-ultrasound examination revealed thrombosis of the three main hepatic veins. An accessory right hepatic vein showed an antegrade flow pattern. There was a partial thrombosis of the portal vein and complete thrombosis of the superior mesenteric vein. A transjugular intrahepatic portosystemic stent (TIPS) was placed to ascertain venous outflow from the liver. Through the stent a transjugular catheter was located in the superior mesenteric vein. Local thrombolytic treatment was given. She received a bolus of 10 mg tissue-plasminogen activator (t-PA, actilyse, Boehringer Ingelheim, Germany) locally, followed by 4 mg/h t-PA. Laboratory evaluation showed a marked decrease of fibrinogen (pretreatment value 3.6 g/l [normal 1.5–3.6]) to a nadir of 0.6 g/l after 18 h of treatment. Fibrin(ogen) degradation products (FDP) showed a marked increase from 10–40 mg/l (normal < 10) to 240–280 mg/l. D-dimer levels rose from 2.0–4.0 mg/l to > 8.0 mg/l. Because of the systemic effects (i.e. fibrinogenolysis) the dose of t-PA was reduced to 2 mg/h after 20 h of treatment. Venography using the local catheter showed a reduction of the thrombus in the portal vein after 16 h of treatment. After 24 h the thrombi in the portal vein and in the superior mesenteric vein were lysed completely. The hepatic veins remained occluded. t-PA was stopped after 24 h (total dose 96 mg). The patient suffered minor bleeding from puncture sites and some haematomas. During the next 24 h unfractionated heparin was given in a dose of 1250 U/h intravenously and was later continued at therapeutic levels. Phenprocoumon was commenced after 5 d. She left the hospital in good health 10 d later. At that time her liver function tests were normal. At follow-up at 2 months ultrasound examination showed a partial occlusion of the TIPS. A second stent was placed using a similar transjugular approach. After 9 months the patient is asymptomatic and the TIPS is patent. Because of the severity of the thrombotic event the patient will be treated with oral anticoagulants for the rest of her life.

Coagulation studies performed at admission to our hospital revealed resistance to activated protein C as shown by a decreased APC ratio (ratio 0.50 [normal 0.9–1.2]). Polymerase chain reaction (PCR) confirmed the presence of a homozygous factor V Leiden mutation. Other risk factors for thrombosis were as follows: antithrombin III activity 0.83 U/l (normal 0.8–1.20), protein C activity 0.42 U/ml (normal 0.7–1.4), protein C antigen level 0.74 U/ml (normal 0.7–1.4), protein S activity 0.59 U/ml (normal 0.7–1.4), free protein S 1.11 U/ml (normal 0.76–1.28), total protein S 1.04 U/ml (normal 0.7–1.4), plasminogen 0.91 U/ml (normal 0.8–1.2). Analysis of the prothrombin gene revealed a normal allele, without the prothrombin variant. There was no detectable lupus anticoagulant and the anticardiolipin antibodies were negative. Family studies showed a heterozygous factor V Leiden in both her parents and her sister. In these family members protein C, protein S and antithrombin III levels were normal.

No other underlying cause for BCS, such as a myeloproliferative disorder, was found.

DISCUSSION

The factor V Leiden mutation, recently established as the most frequently occurring risk factor for venous thrombosis, has only sporadically been described in association with BCS or portal vein thrombosis. In most patients an additional underlying disorder, such as a myeloproliferative disorder or other coagulation defects, next to the factor V Leiden mutation were found ( Denninger et al, 1995 ; Mahmoud et al, 1995 , 1997a).

Coagulation studies in our patient revealed a homozygous factor V Leiden mutation, confirmed by genotyping. The finding of slightly reduced levels of protein C and protein S activity may have been due to the presence of the ongoing thrombotic event. The use of oral contraconceptives may additionally contribute to the low protein S activity. Her parents had normal levels of protein C and protein S thereby excluding a hereditary deficiency. The use of an oral contraceptive may have contributed to the development of thrombosis.

Treatment of BCS depends on the underlying disorder and the severity of liver failure. Unfortunately in most cases anticoagulant treatment is insufficient to treat hepatic vein thrombosis. In many cases the obstruction persists, portal hypertension will develop, and liver failure occurs. Several surgical approaches have been used, including shunt operations or, in case of severe liver failure, even liver transplantation. TIPS has recently been introduced as treatment for acute BCS in a case of hepatic venous outflow obstruction ( Ochs et al, 1993 ); venous outflow of the liver is established by creating a side-to-side shunt connecting a main branch of the portal vein with the (intrahepatic) inferior caval vein. Several cases have been reported on the use of thrombolytic therapy in BCS or portal vein thrombosis ( Raju et al, 1996 ; Frank et al, 1994 ). Thrombolytic therapy has been given systemically or locally after transhepatic catheterization, sometimes complicated by severe bleeding ( Raju et al, 1996 ). The choice of drug used, as well as the dosage and the duration of treatment, vary considerably and no standard regimen of thrombolytic treatment for these diseases has emerged.

Our patient presented with BCS associated with portal vein thrombosis and superior mesenteric vein thrombosis. It is known that patients with such a massive thrombosis have a poor prognosis. These patients are especially difficult to treat and have a high mortality rate of about 70% ( Mahmoud et al, 1997 b). Our therapeutic approach was to place a TIPS, followed by local thrombolytic treatment, resulting in complete lysis of the thrombus in the portal vein and superior mesenteric vein within 24 h. Despite the low starting dose of t-PA, systemic effects (fibrinogenolysis) occurred after 20 h of treatment, for which the dosage of t-PA was reduced. The patient did not have a significant bleeding complication. Because of the low frequency in which BCS or portal vein thrombosis occurs it is not feasible to perform randomized controlled studies comparing standard anticoagulation treatment with thrombolytic therapy. We therefore suggest that our regimen of local t-PA treatment (starting dose 10 mg, followed by 4 mg/h, to a maximum of 100 mg) could be used also in patients with isolated portal vein thrombosis, after obtaining transhepatic access to the thrombosed vein. Considering the risk of bleeding the decision to use thrombolytic therapy should be made on an individual basis for each patient.

We have presented the first case of BCS in association with portal and mesenteric vein thrombosis in a patient homozygous for the factor V Leiden mutation. This challenging medical problem was treated by TIPS placement followed by local thrombolytic therapy. It resulted in restoration of the venous outflow from the liver and complete lysis of the thrombi in the portal and superior mesenteric veins. This approach seems suitable for patients with this disorder who have a poor prognosis.

Ancillary