Value of D-dimer and protein S for diagnosis of portal vein thrombosis in patients with liver cirrhosis

Abstract

Objective

To investigate the value of D-dimer and protein S plasma concentrations for diagnosis of portal vein thrombosis (PVT) in patients with liver cirrhosis.

Methods

D-dimer and protein S were quantified, PVT was diagnosed by dynamic enhanced computed tomography and liver function was classified using the Child–Pugh system. Receiver operating curve analysis was performed.

Results

D-dimer increased, and protein S decreased, with decreasing liver function in the total study population (n = 188). D-dimer concentrations were significantly higher and protein S concentrations were significantly lower in patients with (n = 51) than those without PVT (n = 137). D-dimer had high specificity and negative predictive value (NPV) in Child–Pugh class A or B patients (cut-off values>0.56 mg/l and >1.18 mg/l, respectively). In class C patients>0.77 mg/l D-dimer had high sensitivity and NPV. Protein S had high sensitivity but low specificity in class A or B patients (cut-off values < 17.4 mg/l and <19.2 mg/l, respectively).

Conclusion

Plasma D-dimer and protein S are potential biomarkers for PVT diagnosis in patients with cirrhosis. PVT can be excluded when D-dimer is low and protein S is elevated.

Keywords

Portal vein thrombosis D-dimer protein S liver cirrhosis diagnosis

Introduction

The liver plays a vital role in the coagulation process as it synthesizes and metabolizes the majority of fibrinolytic factors, as well as coagulation factors and inhibitors. Liver failure may disrupt the haemostatic system, leading to severe bleeding or thrombotic complications. Portal vein thrombosis (PVT) was originally thought to be a rare complication of decompensated cirrhosis, with an estimated prevalence of 0.57% of all patients with cirrhosis who had not undergone splenectomy.¹ Advances in imaging techniques have resulted in 5–27% of patients with liver cirrhosis being diagnosed with PVT.^2–5 PVT is more commonly seen in end-stage liver disease⁶ and may be associated with sclerotherapy, abdominal surgery or hepatocellular carcinoma.^7,8 Although patients with liver cirrhosis generally have impaired coagulation, it is possible for a hypercoagulable state to develop, due to increases in coagulation factor VIII and concurrent decreases in anticoagulant factors, such as protein C.^9,10 PVT formation may also be associated with inherited mutations of thrombophilic genes such as thrombin G20210A mutation, methylenetetrahydrofolate reductase 677CT mutation, or factor V Leiden mutation.^11–14 In addition, the presence of anticardiolipin antibodies and increased P-selectin concentrations have both been linked to PVT.^15,16 Life-threatening complications such as refractory ascites, upper gastrointestinal bleeding and intestinal ischaemia and necrosis can occur in patients with PVT.¹⁷ It is therefore necessary to develop efficient methods for detecting PVT, in order to allow earlier diagnosis and treatment and avoid severe complications.

D-dimer is formed by the factor XIII cross-linking and plasmin hydrolysis of fibrin monomer, and is an early diagnostic marker for thrombosis as well as a sensitive indicator of abnormal coagulation and fibrinolysis.¹⁸ D-dimer concentrations are routinely determined in the differential diagnosis of venous thromboembolism, including deep vein thrombosis and pulmonary embolism.^19,20 In addition, D-dimer concentrations increase with deteriorating liver function and may be associated with PVT.⁵

The plasma glycoprotein protein S is mainly synthesized in the liver and is present in platelet α-granules and vascular endothelial cells.^21,22 Protein S acts as an important cofactor for activated protein C (an anticoagulant) and enhances the protein C-mediated inactivation of factor Va and factor VIIIa.²³ Concentrations of protein S decline with deteriorating liver function,²⁴ but this is not well documented in patients with cirrhosis and PVT. Studies have indicated that protein S concentrations correlate with venous thromboembolism,²⁵ suggesting that it may also be involved in PVT formation in patients with cirrhosis. The aim of the present study was to explore the combined value of D-dimer and protein S as potential biomarkers for the diagnosis of PVT in patients with cirrhosis.

Patients and methods

Study population

The study recruited patients with liver cirrhosis caused by hepatitis B, hepatitis C or alcoholism, who were attending the outpatient department of Beijing Chao-yang Hospital Affiliate of Capital Medical University, Beijing, China, for routine medical examination between January 2008 and January 2011. Liver cirrhosis was diagnosed via unequivocal histochemical or histological examination of liver biopsy. Exclusion criteria were: presence of liver cancer; endoscopic therapy; splenectomy; autoimmune disease; other malignancies; liver cirrhosis due to autoimmune hepatitis or primary biliary cirrhosis. All patients were in a stable condition, with no changes in clinical or laboratory variables observed in the 3 months prior to enrolment. Patients underwent a physical examination, full blood cell count and blood biochemistry analysis. The severity of liver disease was assessed according to the Child–Pugh classification (Table 1), and patients were classified as class A, class B or class C.²⁶

Table 1.

Criteria used in the Child–Pugh Classification of liver function.²⁶

Parameter	1 point	2 points	3 points
Total bilirubin, µmol/l	<34	34–50	>50
Serum albumin, g/l	>35	28–35	<28
Prothrombin time (prolonged), s	<4	4–6	>6
Hepatic encephalopathy, grade	none	1–2	3–4
Ascites	none	mild	moderate-to-severe

Total score: class A, 5–6; class B, 7–9; class C, 10–15.

The study was carried out according to the principles of the Declaration of Helsinki and the guidelines of the Ethics Committee of Beijing Chao-yang Hospital Affiliate of Capital Medical University, Beijing, China. All patients provided written informed consent prior to enrolment in the study.

D-dimer and protein S quantification

Peripheral venous blood (15 ml) was collected from each patient into plastic tubes containing 1.5 ml of 0.109 mol/l sodium citrate. Blood was centrifuged at 2000 g for 10 min at 4℃ and the resulting plasma was stored at −80℃ until use. D-dimer and protein S were quantified using enzyme linked immunosorbent assay kits (Sun Biotech, Shanghai, China). The normal ranges of D-dimer and protein S are 0–0.5 mg/l and 19–26.8 mg/l, respectively.

PVT detection

Dynamic contrast enhanced computed tomography (CT) scans (GE LightSpeed® VCT, General Electric Company) were performed within 48 h following patient admission. PVT was localized in the trunk and the left and/or right branches, with possible extensions in the splenic and superior or inferior mesenteric veins. Thrombosis was considered whenever filling defects were recorded in the CT images.

Statistical analyses

Data were presented as mean ± SD or n (%) and analysed using χ²-test, Student’s t-test or Mann–Whitney U-test, as appropriate. Multiple groups were compared using single factor analysis of variance (least-significant difference test for group multiple comparisons) or Wilcoxon rank sum test. Receiver operating characteristic (ROC) curves, sensitivity, specificity, and positive and negative predictive values were determined to assess the diagnostic value of D-dimer and protein S. The area under the curve (AUC) was calculated, with a value <0.5 indicating the test was not valid. D-dimer and protein S cut-off values were chosen to yield tests with maximum AUC values. Statistical analyses were performed using SPSS® software, version 11.5 (SPSS Inc., Chicago, IL, USA) for Windows®. A P-value <0.05 was considered statistically significant.

Results

The study recruited 188 patients with liver cirrhosis (142 males/46 females; mean age 57.5 ± 11.9 years, age range 29–84 years), 51 (27.1%) of whom were diagnosed with PVT. Table 2 shows the demographic and clinical characteristics of the patients, stratified according to the presence or absence of PVT. There were no significant between-group differences in sex distribution, age, aetiology or Child–Pugh score. Patients with PVT had significantly higher D-dimer and lower protein S concentrations than those without (P < 0.01 for both comparisons, Table 2).

Table 2.

Demographic, clinical and laboratory characteristics of Chinese patients with liver cirrhosis, with or without portal vein thrombosis (PVT), included in a study to investigate the value of D-dimer and protein S plasma concentrations in PVT diagnosis.

Characteristic	PVT group n = 51	Control group n = 137
Males	39 (76.5)	103 (75.2)
Age, years	58.0 ± 10.5	57.3 ± 12.4
Aetiology
HBV	26 (51.0)	70 (51.1)
HCV	8 (15.7)	10 (7.3)
ALC	17 (33.3)	57 (41.6)
Child–Pugh classification²⁶
A	7 (13.7)	25 (18.2)
B	26 (51.0)	63 (46.0)
C	18 (35.3)	49 (35.8)
D-dimer, mg/l	1.13 ± 0.65**	0.66 ± 0.49
Protein S, mg/l	15.19 ± 4.92**	18.42 ± 5.41

Data presented as n (%) or mean ± SD.

P < 0.01 vs control group: Mann–Whitney U-test for D-dimer; Student's t-test for protein S.

HBV, hepatitis B virus; HCV, hepatitis C virus; ALC, alcoholic liver cirrhosis.

Data regarding D-dimer and protein S concentrations in the total study population, patients with PVT and control patients, stratified according to Child–Pugh classifications are shown in Tables 3 and 4, respectively. In the total study population, plasma D-dimer concentrations were significantly higher in Child–Pugh class B patients than in class A patients (P < 0.05), and in class C patients than class A patients (P < 0.01; Table 3). A similar pattern was observed in the control group, where D-dimer concentrations were significantly higher in group C patients than group A, and lower in group B than group C (P < 0.01 for both comparisons, Table 3). There were no significant differences in D-dimer concentrations between different Child–Pugh classifications in the PVT group, however. Concentrations of D-dimer were significantly higher in patients with PVT, compared with control patients with the same Child–Pugh classification (P < 0.05 for each comparison; Table 3).

Table 3.

Plasma D-dimer concentrations in Chinese patients with liver cirrhosis, with or without portal vein thrombosis (PVT), stratified according to Child–Pugh liver function classification.²⁶

Group	Child–Pugh classification
Group	A	B	C
Total study population	0.48 ± 0.35 n = 32	0.74 ± 0.57^a n = 89	1.00 ± 0.59^b n = 67
PVT group	0.81 ± 0.49 n = 7	1.10 ± 0.70 n = 26	1.28 ± 0.61 n = 18
Control group	0.38 ± 0.24^d n = 25	0.58 ± 0.43^c,d n = 63	0.90 ± 0.55^b,d n = 49

Data presented as mean ± SD.

P < 0.05 vs class A in same group, ^bP < 0.01 vs class A in same group, ^cP < 0.01 vs class C in same group, ^dP < 0.05 vs same Child–Pugh class in PVT group; Mann–Whitney U-test.

Table 4.

Plasma protein S concentrations in Chinese patients with liver cirrhosis, with or without portal vein thrombosis (PVT), stratified according to Child–Pugh liver function classification.²⁶

Group	Child–Pugh classification
Group	A	B	C
Total study population	19.46 ± 6.19 n = 32	17.23 ± 5.33^a n = 89	17.03 ± 5.15^a n = 67
PVT group	14.89 ± 5.18 n = 7	15.44 ± 5.17 n = 26	14.95 ± 4.73 n = 18
Control group	20.74 ± 5.91^b n = 25	17.97 ± 5.26^a,b n = 63	17.79 ± 5.13^a,b n = 49

Data presented as mean ± SD.

a,b

P < 0.05 vs class A in same group, P < 0.05 vs same Child–Pugh class in PVT group; single factor analysis of variance (least-significant difference test for group multiple comparisons).

Protein S concentrations decreased significantly with deteriorating liver function in the total study population and the control group (Table 4), such that both Child–Pugh class B and class C patients had significantly lower protein S concentrations than class A patients (P < 0.05 for all comparisons; Table 4). There were no significant between-class differences in protein S concentration in the PVT group, however. Protein S concentrations were significantly lower in patients with PVT, compared with control patients with the same Child–Pugh classification (P < 0.05 for each comparison; Table 4).

Results of receiver operating curve analysis of D-dimer and protein S concentrations for the total study population, and for patients stratified according to Child–Pugh classification, are shown in Table 5. In class A patients, a D-dimer cut-off value of >0.56 mg/l provided good specificity and negative predictive value (NPV), but low sensitivity and positive predictive value (PPV). The protein S cut-off value of <17.39 mg/l, resulted in good sensitivity and NPV, but low specificity and PPV.

Table 5.

Results of receiver operating curve analysis of D-dimer and protein S plasma concentrations for the diagnosis of portal vein thrombosis in Chinese patients with liver cirrhosis, stratified according to Child–Pugh liver function classification.²⁶

Parameter	Child–Pugh classification
Parameter	A n = 32	B n = 89	C n = 67	A + B + C n = 188	A + B + C n = 188
D-dimer
Cut-off value, mg/l	>0.56	>1.18	>0.77	>0.92	>0.24
Sensitivity, %	71.4	50.0	94.4	62.7	100.0
Specificity, %	84.0	92.1	46.9	75.9	30.7
PPV, %	55.6	72.2	39.5	42.9	16.1
NPV, %	91.3	81.7	95.8	84.6	100.0
AUC	0.809	0.737	0.691	0.782	0.000
Protein S
Cut-off value, mg/l	<17.39	<19.20	–	<16.36	<25.73
Sensitivity, %	85.7	76.9	–	60.8	100.0
Specificity, %	72.0	47.6	–	67.2	28.7
PPV, %	46.2	37.7	–	40.8	7.3
NPV, %	94.7	83.3	–	82.1	100.0
AUC	0.811	0.645	–	0.668	0.000

PPV, positive predictive value; NPV, negative predictive value.

In Child–Pugh class B patients, good specificity and NPV were achieved with a D-dimer cut-off value >1.18 mg/l. A protein S cut-off value of <19.2 mg/l provided good sensitivity and NPV, but low specificity and PPV. For class C patients, a D-dimer cut-off value of >0.77 mg/l resulted in high sensitivity and NPV, but low specificity and PPV.

In the total study population (classes A + B + C), a D-dimer cut-off value of >0.92 mg/l, resulted in good specificity and NPV, but low sensitivity and PPV. The protein S cut-off value of <16.36 mg/l provided good specificity and NPV, but a low sensitivity and PPV. D-dimer and protein S cut-off values of >0.24 mg/l and <25.73 mg/l respectively resulted in 100% sensitivity and NPV in the diagnosis of PVT, but very low specificity and PPV (Table 5).

Discussion

The overall prevalence of PVT formation in liver cirrhosis ranges between 5% and 27%,^2–5 or higher if hepatocellular carcinoma cases are included.^27,28 The prevalence of PVT in the current study was high (27%), possibly due to the severity of cirrhosis in the study cohort (83% of cases were Child–Pugh class B or C). This may not be the sole explanation, however, because a similar proportion of patients with and without PVT were classified as Child–Pugh class C.

The fibrin degradation product D-dimer is a sensitive marker of coagulation and fibrinolysis.²⁹ Patients with cirrhosis who have increased D-dimer concentrations may have hyper-fibrinolysis, including primary coagulation activation, delayed hepatic clearance of tissue plasminogen activator and reduced hepatic synthesis of fibrinolytic inhibitor.^30,31 Findings of the current study were in accordance with other data that revealed increased D-dimer concentrations with deteriorating liver function.³²

Presence of PVT was associated with a significant increase in the D-dimer concentration in the present study, but was not dependent on Child–Pugh classification. D-dimer evaluations have been shown to have high sensitivity and NPV for the diagnosis of venous thromboembolism, such that this complication can be excluded when the D-dimer concentration is <0.2 mg/l.²⁰ D-dimer concentrations were shown to have high specificity but low sensitivity for the diagnosis of PVT in Child–Pugh class A or B patients in the present study. In contrast, a D-dimer concentration >0.77 mg/l in class C patients had high sensitivity and NPV in the present study, but low specificity and PPV, possibly related to deterioration in liver function in severely affected patients.

Protein S is an important cofactor that enhances the protein C inhibition of factor Va and factor VIIIa.²³ Protein S concentrations were found to decrease with deteriorating liver function in the present study, in accordance with the findings of others.⁵ Protein S has been proposed as a sensitive indicator of liver cell dysfunction.³³

While coagulation function is altered by liver dysfunction, anticoagulant factors such as protein S decrease in a higher proportion in patients with cirrhosis.⁶ Thus, a relatively hypercoagulable state can develop with a large amount of protein S consumed after the formation of PVT. Both liver function deterioration and PVT formation can contribute to decreased protein S concentrations, which have been associated with venous thromboembolism and ischaemic stroke.^25,34

Protein S had a high sensitivity but a low specificity for the diagnosis of PVT in Child–Pugh class A or B patients in the present study. In Child–Pugh class C patients, there was no statistically significant difference between patients with, or without, PVT. It was possible to achieve 100% sensitivity and NPV in the diagnosis of PVT for the total study population with a D-dimer cut-off value of <0.24 mg/l and a protein S cut-off value of >25.73 mg/l, but the specificity and PPV were very low. The current study did reveal that significant decreases in D-dimer, or increases in protein S, could be used to exclude PVT in patients with liver cirrhosis.

The cause of PVT in cirrhosis remains unknown, and the prognosis of these patients is poor. Although anticoagulant therapy by low-molecular weight heparin and antithrombin III may be effective in theory, the risk of bleeding hampers the realization of large-scale clinical studies.^35,36 A transjugular intrahepatic portosystemic shunt may be the most effective therapeutic option in patients with cirrhosis who develop nonmalignant PVT.^37–39

In conclusion, plasma D-dimer and protein S concentrations are potential biochemical markers in the diagnostic strategy of PVT, in patients with cirrhosis. PVT can be excluded in cases where the D-dimer concentration is significantly low and the protein S concentration significantly enhanced. In the case of elevated D-dimer and low protein S concentrations, presence of PVT is suspected and specific imaging techniques should be performed to confirm the diagnosis and initiate early treatment.

Declaration of conflicting interest

The authors declare that there are no conflicts of interest.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Okuda

Ohnishi

Kimura

. Incidence of portal vein thrombosis in liver cirrhosis. An angiographic study in 708 patients. Gastroenterology 1985; 89: 279–286.

Amitrano

Brancaccio

Guardascione

. Inherited coagulation disorders in cirrhotic patients with portal vein thrombosis. Hepatology 2000; 31: 345–348.

Amitrano

Guardascione

Brancaccio

. Risk factors and clinical presentation of portal vein thrombosis in patients with liver cirrhosis. J Hepatol 2004; 40: 736–741.

Politoske

Ralls

Korula

. Portal vein thrombosis following endoscopic variceal sclerotherapy. Prospective controlled comparison in patients with cirrhosis. Dig Dis Sci 1996; 41: 185–190.

Zhang

Hao

Yang

. Protein C and D-dimer are related to portal vein thrombosis in patients with liver cirrhosis. J Gastroenterol Hepatol 2010; 25: 116–121.

Singhal

Karachristos

Bromberg

. Hypercoagulability in end-stage liver disease: prevalence and its correlation with severity of liver disease and portal vein thrombosis. Clin Appl Thromb Hemost 2012; 18: 594–598.

Mangia

Villani

Cappucci

. Causes of portal venous thrombosis in cirrhotic patients: the role of genetic and acquired factors. Eur J Gastroenterol Hepatol 2005; 17: 745–751.

Alkim

Ayaz

Sasmaz

. Hemostatic abnormalities in cirrhosis and tumor-related portal vein thrombosis. Clin Appl Thromb Hemost 2012; 18: 409–415.

Tripodi

Anstee

Sogaard

. Hypercoagulability in cirrhosis: causes and consequences. J Thromb Haemost 2011; 9: 1713–1723.

10.

Northup

. Hypercoagulation in liver disease. Clin Liver Dis 2009; 13: 109–116.

11.

Erkan

Bozdayi

Disibeyaz

. Thrombophilic gene mutations in cirrhotic patients with portal vein thrombosis. Eur J Gastroenterol Hepatol 2005; 17: 339–343.

12.

Amitrano

Guardascione

Ames

. Increased plasma prothrombin concentration in cirrhotic patients with portal vein thrombosis and prothrombin G20210A mutation. Thromb Haemost 2006; 95: 221–223.

13.

Gabr

Bessa

El-Zamarani

. Portal vein thrombosis in Egyptian patients with liver cirrhosis: Role of methylenetetrahydrofolate reductase C677T gene mutation. Hepatol Res 2010; 40: 486–493.

14.

Mahmoud

Elias

Beauchamp

. Prevalence of the factor V Leiden mutation in hepatic and portal vein thrombosis. Gut 1997; 40: 798–800.

15.

Oksüzoglu

Bayraktar

Arslan

. Portal vein thrombosis in cirrhotics: related with anticardiolipin antibodies? Hepatogastroenterology 2003; 50: 1527–1530.

16.

Wang

Liu

Chen

. Combined use of D-dimer and P-selectin for the diagnosis of splenic or portal vein thrombosis following splenectomy. Thromb Res 2010; 125: e206–e209.

17.

Amitrano

Guardascione

Brancaccio

. Portal and mesenteric venous thrombosis in cirrhotic patients. Gastroenterology 2002; 123: 1409–1410.

18.

Pabinger

. Biomarkers and venous thromboembolism. Arterioscler Thromb Vasc Biol 2009; 29: 332–336.

19.

Anderson

Wells

Stiell

. Management of patients with suspected deep vein thrombosis in the emergency department: combining use of a clinical diagnosis model with D-dimer testing. J Emerg Med 2000; 19: 225–230.

20.

Wells

Anderson

Rodger

. Evaluation of D-dimer in the diagnosis of suspected deep-vein thrombosis. N Engl J Med 2003; 349: 1227–1235.

21.

Di Scipio

Hermodson

Yates

. A comparison of human prothrombin, factor IX (Christmas factor), factor X (Stuart factor), and protein S. Biochemistry 1977; 16: 698–706.

22.

Fair

Marlar

Levin

. Human endothelial cells synthesize protein S. Blood 1986; 67: 1168–1171.

23.

High

. Antithrombin III, protein C, and protein S. Naturally occurring anticoagulant proteins. Arch Pathol Lab Med 1988; 112: 28–36.

24.

Zocco

Di Stasio

De Cristofaro

. Thrombotic risk factors in patients with liver cirrhosis: correlation with MELD scoring system and portal vein thrombosis development. J Hepatol 2009; 51: 682–689.

25.

Shen

Lin

Tsay

. Protein C and protein S deficiencies are the most important risk factors associated with thrombosis in Chinese venous thrombophilic patients in Taiwan. Thromb Res 2000; 99: 447–452.

26.

Pugh

Murray-Lyon

Dawson

. Transection of the esophagus for bleeding esophageal varices. Br J Surg 1973; 60: 646–649.

27.

Wang

Zhao

Liu

. Portal vein thrombosis. Hepatobiliary Pancreat Dis Int 2005; 4: 515–518.

28.

Ponziani

Zocco

Campanale

. Portal vein thrombosis: insight into physiopathology, diagnosis, and treatment. World J Gastroenterol 2010; 16: 143–155.

29.

Bick

. Disseminated intravascular coagulation and related syndroms: a clinical review. Semin Thromb Hemost 1988; 14: 299–338.

30.

Violi

Ferro

Basili

. Hyperfibrinolysis resulting from clotting activation in patients with different degrees of cirrhosis. The CALC Group. Coagulation Abnormalities in Liver Cirrhosis. Hepatology 1993; 17: 78–83.

31.

Leebeek

Kluft

Knot

. A shift in balance between profibrinolytic and anti fibrinolytic factors causes enhanced fibrinolysis in cirrhosis. Gastroenterology 1991; 101: 1382–1390.

32.

Gram

Duscha

Zurborn

. Increased levels of fibrinolysis reaction products (D-dimer) in patients with decompensated alcoholic liver cirrhosis. Scand J Gastroenterol 1991; 26: 1173–1178.

33.

Zurborn

Kirch

Bruhn

. Immunological and functional determination of the protease inhibitors, protein C and antithrombin III, in liver cirrhosis and in neoplasia. Thromb Res 1988; 52: 325–36.

34.

Chen

Lan

Chang

. The prevalence of protein C, protein S, and antithrombin III deficiency in non-APS/SLE Chinese adults with noncardiac cerebral ischemia. Clin Appl Thromb Hemost 2003; 9: 155–162.

35.

Romero-Gömez

Gutierrez-Tous

Delgado-Mije

. Anticoagulation therapy for recent portal vein thrombosis in a patient with liver cirrhosis suffering from variceal rebleeding. Gastroenterology 2001; 122: 2095–2095.

36.

Kawanaka

Akahoshi

Kinjo

. Impact of antithrombin III concentrates on portal vein thrombosis after splenectomy in patients with liver cirrhosis and hypersplenism. Ann Surg 2010; 251: 76–83.

37.

Senzolo

M Sartori

Rossetto

Burra

Cillo

Boccagni

. Prospective evaluation of anticoagulation and transjugular intrahepatic portosystemic shunt for the management of portal vein thrombosis in cirrhosis. Liver Int 2012; 32: 919–927.

38.

Han

. Transjugular intrahepatic portosystemic shunt for portal vein thrombosis with symptomatic portal hypertension in liver cirrhosis. J Hepatol 2011; 54: 78–88.

39.

Luca

Miraglia

Caruso

. Short- and long-term effects of the transjugular intrahepatic portosystemic shunt on portal vein thrombosis in patients with cirrhosis. Gut 2011; 60: 846–852.