Review

Current knowledge and management of portal vein thrombosis in

cirrhosis
Marco Senzolo1,*, Guadalupe Garcia-Tsao2,3, Juan Carlos García-Pagán4,5,6

Summary
Keywords: Cirrhosis; Portal vein thrombosis (PVT) is an increasingly recognised complication of cirrhosis whose incidence
Coagulopathy; Portal
increases in parallel with the severity of cirrhosis. Several risk factors have been associated with the
hypertension; Portal vein
thrombosis; TIPS; Transplant. occurrence and progression of PVT. Although the negative effect of complete PVT on the surgical outcome
of liver transplant recipients is clear, its impact on cirrhosis progression remains uncertain. Treatment
Received 30 September 2020;
options include anticoagulants and interventional thrombolytic therapies, which are chosen almost on a
received in revised form 20 April
2021; accepted 20 April 2021; case-by-case basis depending on the characteristics of the patient and the thrombus. In this manuscript,
available online 27 April 2021 we review current knowledge regarding the epidemiology, risk factors, diagnosis and classification,
natural history, clinical consequences and treatment of non-neoplastic PVT in cirrhosis.
© 2021 Published by Elsevier B.V. on behalf of European Association for the Study of the Liver.

1
Introduction
Multivisceral Transplant Unit-
Gastroenterology, Department of Cirrhosis, particularly at the decompensated stage, compensated cirrhosis,3 17% in patients with
Surgery, Oncology and is characterised by profound and complex haemo- mostly Child-Pugh B/C cirrhosis4 and up to 26% in
Gastroenterology, Padua University static alterations that are both procoagulant and liver transplant (LT) candidates.5 Earlier series of
Hospital, Padua, Italy
†
;
anticoagulant and could theoretically lead to both patients on the LT waiting list showed a lower
2
Section of Digestive Diseases, VA- thrombotic and bleeding complications.1 Therefore, prevalence of PVT (~7%) perhaps due to different
Connecticut Healthcare System, despite a low platelet count and elevated interna- policies regarding listing patients with PVT.6 In
West Haven, CT, USA;
3 tional normalised ratio (INR), patients with these studies, most PVTs are partial, but, in a study
Section of Digestive Diseases, Yale
University School of Medicine, New cirrhosis are not “auto-anticoagulated,” as previ- that pooled 589 patients listed for LT, occlusive PVT
Haven, CT, USA; ously thought, but actually have a greater pro- was observed in 44% of cases.7
4
Barcelona Hepatic Hemodynamic pensity than patients without cirrhosis to develop In prospective studies, the 1-year incidence of
Laboratory, Liver Unit, Hospital
Clínic, Barcelona, Spain;
thrombotic complications, the most important be- PVT ranges from 3.7% to 24.4% (Fig. 1),8–18 with a
5
Institut de Investigacions ing portal vein thrombosis (PVT), which is defined lower incidence in cohorts that included a majority
Biomèdiques August Pi i Sunyer as the presence of a thrombus in the lumen of the of patients with compensated cirrhosis.12,17
(IDIBAPS), University of Barcelona,
main portal vein that can extend into intrahepatic
Spain;
6
CIBEREHD (Centro de or extrahepatic venous branches. Risk factors
Investigación Biomédica en Red The pathogenesis of PVT in cirrhosis is probably
Enfermedades Hepáticas y Epidemiology multifactorial and results mainly from alterations
Digestivas), Spain
† PVT is a relatively rare disease in the general in the different components of Virchow’s triad:
†
Health Care Provider of the Eu-
population. In an autopsy study, representing 84% decreased portal vein flow, hypercoagulability and
ropean Reference Network on of all in-hospital deaths in Malmo, the lifetime damage of the vessel wall (Box 1).
Rare Liver Disorders (ERN-Liver). cumulative prevalence of PVT was 1.0%, but 40% of
* Corresponding author. Address:
these patients had cirrhosis and those with both Altered portal venous blood flow
Multivisceral Transplant Unit- liver disease and hepatocellular carcinoma A threshold portal vein flow <15 m/sec has been
Gastroenterology, Department of (HCC) had the highest risk of PVT (odds ratio [OR] described as the most predictive of PVT based on
Surgery, Oncology and Gastroen-
terology, Padua University Hospi- 17.1).2 both prospective and retrospective studies.8,19 An
tal, Via Giustninani 2, 35128, The prevalence of PVT in cirrhosis varies widely increase in portal blood inflow through portosys-
Padua, Italy. Tel.: +390498218726; among studies and depends on the type of diag- temic collaterals also seems to influence PVT
fax +390498218727.
nostic approach used, the stage of cirrhosis development. Indeed, a large flow volume
E-mail address: marco.senzolo@ (compensated or decompensated) and/or on the (>400 ml/min) and velocity (>10 cm/sec) in the
aopd.veneto.it (M. Senzolo).
inclusion or exclusion of patients with HCC. largest porto-collateral vessel visualised by
https://doi.org/10.1016/ Because HCC invasion of the portal vein (tumoral Doppler ultrasound have been shown to predict
j.jhep.2021.04.029 PVT) has entirely different implications than a PVT.9 Nevertheless, in another study,20 reversal of
blood clot in the portal vein, this review refers portal vein blood flow, the maximal expression of
solely to non-tumoral PVT. increased porto-collateral blood flow, did not
The prevalence of PVT increases in parallel with correlate with an increased risk of PVT. Further
the severity of cirrhosis: 10% in patients with prospective studies evaluating the role of reduced

Journal of Hepatology 2021 vol. 75 j 442–453

portal blood flow velocity and/or increased porto- thrombomodulin resistance in peripheral blood
collateral blood flow are needed. and PVT development.23 Results from larger pro-
Non-selective beta blockers (NSBBs), by spective studies, including a more comprehensive
reducing portal venous blood flow, could theoret- panel of haemostatic alterations in patients with
ically increase the risk of PVT, although data are cirrhosis should clarify this potential mechanism.
conflicting and a causal relationship has not been
firmly demonstrated. A recent meta-analysis of Inherited and acquired prothrombotic disorders
prospective and retrospective cohort studies sug- While disorders such as prothrombin or factor V
gested a positive association between NSBBs and gene mutations, and protein C, S or antithrombin
the risk of PVT (OR 4.62).21 However, most studies deficiencies play an important role in the devel-
were retrospective, abstracts were included and opment of PVT in patients without cirrhosis, their
there was significant heterogeneity. In a prospec- role in PVT in patients with cirrhosis is much less
tive single centre cohort of 108 patients, predictorsclear.24,25 Indeed, relevant studies have shown that
of PVT development were history of varices (grade most patients with cirrhosis and PVT do not have
18
>
−2) and use of NSBBs. However, the number of any of these inherited alterations. Similarly,
patients developing PVT was small (only 11) and although 1 study suggested a higher prevalence of
the analysis was not adjusted for confounders. JAK2 V617F mutations in patients with cirrhosis
Given strong data demonstrating the benefits of and PVT,26 no JAK2 mutations were found in a
NSBBs in cirrhosis, their use should not be limited cohort of 271 LT recipients irrespective of the
by their unconfirmed role in facilitating PVT. presence of PVT.27 Therefore, because it is not cost-
effective, screening for thrombophilia is not rec-
Altered coagulation ommended in patients with cirrhosis and PVT.28
Haemostatic alterations associated with cirrhosis Probably reflecting an increased prothrombotic
Patients with cirrhosis have a well-described state associated with neoplasia, a small prospective
derangement of the haemostatic balance due to a study showed a high (24.4%) 1-year incidence of
reduction of both anticoagulant and procoagulant non-neoplastic PVT in 41 patients with HCC.14
factors, together with increased levels of several Similarly, another study showed an incidence of
procoagulant factors such as factor VIII and von PVT of 25% over a median follow-up time of only
Willebrand factor (vWF).22 Overall, these alter- 4.4 months from the diagnosis of HCC.29 Further
ations indicate that the homeostatic balance leans exploration is warranted.
toward hypercoagulability, which would favour the
development of PVT. However, the association be- Local portal vein alterations
tween hypercoagulability and the development of In patients with cirrhosis and portal hypertension,
PVT has not been well studied. Indeed, only a single bacterial translocation, local inflammation and
retrospective study in 53 patients with cirrhosis endotoxemia could play a role in promoting a local
found a significant association between hypercoagulable state in the portal venous area. In

28 Total number
Partial PVT
Complete PVT
23

18
Incidence (%)

13

3
. (11)
. (10)

. 12)

. (18)
. (14)
. (8)

. (9)

. (16)
. (13)

. 14)

l. ( 1 5

. (17)
al (

al (

-2
al

al

al
al

al
al

al
al

ta

al
et

et

et
et

et

et
et

et
et

et

et
ie
o

ik
hn

ry

ry
tto

ira
lla

tto
cc

am

Cu

n
az
Ne

Ne

ro
Jo

ne
Vi

rre
Zo

ne
l-R
uy

Tu
Za

Za

Fe
ar

de
M

Ab

Patients (N) 100 243 150 1,243 95 36 35 41 109 108 241 320
Child-Pugh C (%) 5 NA 12 0 30.5 11 25.7 22 0 4.6 10.8 6
HCC (%) 0 0 0 0 0 0 0 100 0 0 0 0

Fig. 1. Incidence of PVT in prospective studies. HCC, hepatocellular carcinoma; PVT, portal vein thrombosis.

Journal of Hepatology 2021 vol. 75 j 442–453 443

Review

Box 1. Risk factors for PVT in cirrhosis. inside the vessel, more hyperechoic if chronic. The
addition of Doppler helps to evaluate the velocity
Cryptogenic and NASH cirrhosis of flow in the vessels, and to detect shunts and the
Child-Pugh class B and C
presence of small vessels in the extrahepatic main
Oesophageal varices grade > −2
Portal vein blood flow <15 cm/s PV that are characteristic of a cavernoma.40 How-
Large porto systemic collaterals with high blood flow ever, changes related to cavernous transformation
Non-selective beta blockers of the portal vein, which have been reported to
Local vascular damage occur as early as 6–20 days after acute PVT,41 are far
Previous splanchnic vein thrombosis
less common in patients with cirrhosis, since the
Sclerotherapy of oesophageal varices
Splenectomy pressure gradient before and after the clot is small,
Partial splenic embolisation leading to stasis of portal vein flow which hinders
the formation and dilatation of collateral
channels.42
NASH, non-alcoholic steatohepatitis; PVT, portal vein thrombosis. Ultrasound has a specificity of 92% and a
sensitivity of 89% for the detection of PVT,43 with
a lower sensitivity (14%–50%) for partial PVT.7
fact, a recent study in patients receiving a trans- The use of contrast-enhanced ultrasonography
jugular intrahepatic portosystemic shunt (TIPS) can be recommended in doubtful cases, since the
showed that there is a portal/systemic circulation sensitivity of this method is very high, exceeding
gradient for lipopolysaccharide (LPS), vWF and 95%.28 However, evaluation of the extension and
factor VIII. LPS may favour a hypercoagulable state occlusion of all portal vein branches (superior
by increasing endothelial secretion of factor VIII30 mesenteric and splenic veins) by Doppler ultra-
and decreasing thrombomodulin activity31 while sound can be limited by the presence of ascites,
also increasing platelet activation.32 Moreover, the obesity and bowel gas and is operator-
observation of a significantly higher concentration dependent; therefore, the use of cross-sectional
of microparticles of endothelial origin (CD62E- MP) imaging (angio-CT or angio-MRI) is recom-
in the portal vein compared to the systemic circu- mended to confirm the presence of PVT and for
lation suggested greater endothelial damage at this complete PVT staging.
level.33 Although all these arguments make it On CT, imaging features of acute PVT include
possible that local inflammatory conditions could increased attenuation in the portal vein on unen-
play a role in PVT development in patients with hanced scans and a lack of enhancement after
cirrhosis, this has not been demonstrated in pro- administration of IV contrast; associated changes
spective cause-effect studies. in hepatic perfusion may be seen with increased
hepatic enhancement in the arterial phase and
Other factors associated with PVT development decreased hepatic enhancement during the portal
Among other risk factors for PVT in cirrhosis, phase.44 MRI can be used as an alternative partic-
endoscopic therapy for oesophageal varices (with ularly since it is not associated with radiation
sclerotherapy or variceal band ligation)7 and a past exposure. However, it may have a lower specificity
history of variceal bleeding seem to play a role.20 in small calibre vessels and may lack the spatial
Endothelial damage or endotoxemia following resolution necessary to evaluate thrombus size.45
endoscopic sclerotherapy have been incriminated In the presence of HCC, exclusion of tumoral
in PVT development,34 however other studies invasion of the portal vein is essential. Recently,
failed to confirm this association.35 Previous non-invasive criteria with the acronym A-VENA
abdominal surgery, splenectomy36 and portosys- based on the presence of 3 or more of the
temic shunt surgery have also been reported as following: AFP >1,000 ng/dl; Venous expansion;
determinants of PVT due to venous injury and/or thrombus Enhancement; Neovascularity; and PVT
by altering portal venous flow. In addition, it has Adjacent to HCC have been shown to have 100%
been suggested that cryptogenic and NASH (non- sensitivity, 93.6% specificity, 80% positive predic-
Key point alcoholic steatohepatitis)-related cirrhosis are tive value, and 100% negative predictive value in
Both the incidence and associated with a higher risk of PVT than other diagnosing tumoral PVT and distinguishing it from
prevalence of portal vein aetiologies of cirrhosis37,38 because of increased non-tumoral PVT.46 However, these results need to
thrombosis (PVT) in pa- thrombin generation and hyperfibrinolysis.39 be validated.
tients with cirrhosis in-
crease in parallel with the Several different classifications, mainly based on
severity of liver disease.
Diagnosis and classification the anatomy, extension and degree of occlusion
The pathogenesis of PVT is Colour Doppler ultrasound is usually the first-line have been proposed47–55 (Table 1). A more recent
multifactorial but a imaging method used to diagnose PVT. In most classification also takes into account the presence
decrease in portal vein flow cases, the diagnosis is established in asymptomatic
seems to play the most
of symptoms related to portal hypertension and
patients during HCC screening. At ultrasound, the the time of onset of thrombosis.54 All these
important role.
thrombus appears as a hypo or iso-echoic material

444 Journal of Hepatology 2021 vol. 75 j 442–453

characteristics are not only important to address identify predictors of PVT regression or
the severity of thrombosis and its possible impact progression.
on portal hypertension and liver transplantation
(LT), but also in the formulation of a treatment al- Clinical consequences of PVT in cirrhosis
gorithm. The severity of the underlying liver Impact of PVT on complications/outcomes in
dysfunction and completeness of the occlusion, patients with cirrhosis
together with a delay in initiating treatment, can The extent to which PVT influences outcomes in
reduce the efficacy of anticoagulation. Current or patients with cirrhosis is still a matter of debate.
future candidacy for liver transplantation is Development of PVT, especially if complete, would
essential in the evaluation. In this context, the most exert a further increase in resistance to portal
widely used classification up to now was proposed blood flow, resulting in worsening portal hyper-
by Yerdel et al.56 Its importance lies in the surgical tension (proximal to the thrombus). However, in
approach employed, which is determined by the patients with extensive porto-collateral circulation
extent of thrombosis and the possibility to recon- and very low or absent hepatopetal portal blood
Key point
struct a physiological portal venous inflow. flow, the development of PVT may be of no
Recently, a new classification system has been consequence. Classification of PVT should
proposed, which could serve as a guide to multi- Few studies describe the impact of untreated be treatment-oriented and
include extension and
disciplinary decision-making before and during LT PVT on the natural history of cirrhosis and all
grade of occlusion of
in patients with diffuse PVT.55 This classification include very heterogeneous populations in terms thrombosis, severity of
incorporates an analysis of the presence of porto- of underlying liver disease and thrombus char- liver disease and evalua-
systemic collaterals, such as large gastric vein, acteristics. Although the prospective study by tion of feasibility of physi-
Nery et al. in mostly compensated patients ological reconstruction of
pericholedocal varices, splenorenal or other shunts,
portal vein inflow at liver
which could be used to guarantee portal vein showed that the presence of varices and a pro- transplantation.
inflow in complex PVT. longed INR predicted PVT, development of PVT
did not predict liver disease progression or
Natural history of PVT decompensation.12 In a large retrospective cohort
While spontaneous recanalisation of PVT is rare in of patients with cirrhosis admitted with a
a non-cirrhotic population, it can occur in patients decompensating event, PVT was not associated
with cirrhosis. In a recent meta-analysis comparing with 1-year mortality. In this study, causes of PVT
anticoagulated vs. non-anticoagulated patients were heterogeneous (1/3 being related to sple-
(excluding the study by Wang et al. that included nectomy or partial splenic embolisation) and
patients treated with TIPS in both arms57), 12% of liver disease was significantly less severe in the
patients had complete spontaneous PVT recanali- PVT group.62 In another prospective study there
sation while 48% had thrombosis progression.58 was no correlation between development of PVT
The largest study included in the meta-analysis and variceal bleeding or mortality but progres-
was performed by Nery et al. on 1,243 patients sive PVT did correlate with worsening ascites.9
with mostly compensated cirrhosis (70% Child- The fact that, at time of diagnosis, PVT is often
Pugh A) of whom 101 developed PVT, which asymptomatic63 has also been used as an argu-
showed that thrombi disappearance occurred at an ment against a major role of PVT in worsening
unexpectedly high rate of 70%. Nevertheless, outcomes in cirrhosis.
thrombi reappeared in 19 of these 70 patients.12 It Other studies provide arguments that support
is important to mention that most PVTs in this a deterioration of cirrhosis with PVT develop-
study were partial and disappearance of thrombi ment. In a study by Amitrano et al., among 79
was assessed by ultrasound and not confirmed by patients with cirrhosis and newly diagnosed PVT,
CT. In another study using CT scanning, PVT was 39% presented with an episode of portal
also shown to improve (but not resolve) in 45% of hypertension-related gastrointestinal bleeding.64
patients. Contrary to the study by Nery et al., these Similarly, in the previously mentioned large
patients were mostly decompensated, as 67% had retrospective cohort of 2,597 patients admitted
ascites.59 A much lower percentage of spontaneous with acute decompensation, patients with PVT
recanalisation (ranging from 5% to 19%) has been were more likely to have been admitted with
reported in patients with decompensated cirrhosis variceal haemorrhage compared to those without
and/or those listed for liver transplantation, with PVT.62 Moreover, in patients with variceal
higher rates of progression than regression.20,60,61 bleeding, PVT was found to be independently
Therefore, although spontaneous resolution of associated with a higher risk of failure to control
PVT may occur, especially if non-occlusive and in acute variceal bleeding, re-bleeding and short-
compensated cirrhosis, PVT is more likely to term mortality65–67
progress. Further prospective studies on well- Further arguments on the issue come from
characterised patient populations are required to studies that have correlated the response of PVT

Journal of Hepatology 2021 vol. 75 j 442–453 445

446

Review
Table 1. Classifications of non-neoplastic PVT in patients with cirrhosis.
Publication Grade I Grade II Grade III Grade IV
Stieber46 (Segmental) thrombosis of the PV trunk Thrombosis extension to SMV; patent SV Thrombosis extension to SV (partial or Diffuse, with involvement of SMV and SV,
(partial or complete) (partial or complete) complete) including the confluence (complete)
Nonami47 Thrombosis of intrahepatic PV branches Thrombosis of the right or left portal branch Partial obstruction of the PV trunk Complete obstruction of the PV trunk
of the PV or at the bifurcation
Gayowski48 Partial (mural) thrombosis of the main portal Complete thrombosis of the main portal Complete thrombosis of the main portal Complete thrombosis of the main portal
trunk extending to or below the confluence trunk not extending to the confluence of the trunk extending to the level of the trunk with extension below the
with residual flow SMV and SV confluence confluence
Yerdel49 Minimally or partially thrombosed PV >50% occlusion of the PV, including complete Complete thrombosis of both PVT and prox Complete thrombosis of the PV and proximal
(thrombosis <50% of the vessel lumen) with thrombosis, with or without minimal exten- imal SMV; patent distal SMV and distal SMV
or without minimal extension into the SMV sion into the SMV
Jamieson50 Thrombosis confined to the PV beyond the Thrombosis extending into the proximal SMV Diffuse thrombosis of the splanchnic venous Extensive thrombosis of the splanchnic
splenomesenteric confluence (partial or but with a patent vessel in the mesentery system but with large accessible collaterals venous system but with only fine collaterals
complete)
Charco51 Thrombosis confined to the PV (partial or Thrombosis that extends all the way to the Diffuse thrombosis of the splanchnic system Diffuse thrombosis with presence of fine
Journal of Hepatology 2021 vol. 75 j 442–453

complete) proximal portion of the SMV with perme- with presence of dilated collateral veins collateral veins
ability of the mesenteric confluence
Bauer* 52 Occlusion <25% (PVT, SV, or SMV) Occlusion 26%-50% (PVT, SV, or SMV) Occlusion 51%-75% (PVT, SV, or SMV) Occlusion >76% (PVT, SV, or SMV)
Sarin** 53 Thrombosis of one (type 2a) or both (2b) Thrombosis confined to the PV trunk (type 1) Thrombosis of PV trunk and branches Thrombosis of PV trunk and/or branches
intrahepatic PV branches (partial or (partial or complete) (type 3) (partial or complete) with involvement of SMV ± SV (partial or
complete) complete)
Bhangui54 Yerdel grade 1-3 Complex PVT (Yerdel grade 4, Jamieson grade Complex PVT (Yerdel grade 4, Jamieson grade 3-4 and Charco grade 3-4) in absence of
3-4 and Charco grade 3-4) in presence of spontaneous or surgical portosystemic shunts
spontaneous or surgical portosystemic shunt —
*Refers to percentage of occlusion of the vessel (a grade is assigned to each vessel, according to the degree of occlusion).
**Covers also functional aspects of PVT as duration (recent or chronic ± presence of a cavernoma), presentation (asymptomatic or symptomatic). PV, portal vein; PVT, portal vein thrombosis; SMV, superior mesenteric vein; SV,
splenic vein.
bosis, observed more often in patients with
risk of death.78 Early post-transplant re-throm-
5.65-fold increase in the 1-month post-surgical
meta-analysis in which complete PVT carried a
are decreased.7 This was also confirmed in a recent
30-day and 1-year post-transplant survival rates
(Grade III or IV according to Yerdel’s classification),
authors demonstrated that when PVT is complete

operative period.
essential, particularly during the initial post-
lishment of adequate portal flow to the graft is
post-transplant outcomes is major. The estab-
to be significant, its impact on transplant and
cirrhosis is not entirely clear and does not appear
While the impact of PVT on the course of
Impact of PVT on liver transplant outcomes

of both liver disease and PVT.

patients with a detailed evaluation of the severity
designed prospective studies on large cohorts of
natural history of cirrhosis will require well-

outcome.77
tality in 6/26 studies which reported this
coagulation was associated with decreased mor-
discarded. In a recent meta-analysis, anti-
history of cirrhosis irrespective of PVT cannot be
beneficial effect of anticoagulation on the natural
coagulated for atrial fibrillation.76 Thus, a direct
a large cohort of patients with cirrhosis anti-
cirrhosis74,75 and also in epidemiological studies of
fibrosis in different experimental models of
exert a beneficial effect on portal hypertension and
without PVT.13 Anticoagulation has been shown to
prophylactic heparin in patients with cirrhosis
observed in the study by Villa et al. who used
PVT.73 A similar beneficial effect on survival was
observed irrespective of resolution (or not) of
where a beneficial effect of anticoagulation was
was not the case in the study by Pettinari et al.,
coagulation in the study by La Mura et al.,70 this
patients in whom PVT resolved with anti-
complication-free survival was only observed in

improved (15.8%).59
gressive thrombus (39.1%) vs. those in whom PVT
mortality in patients who had a stable or pro-
vs. A).72 Moreover, Luca et al. reported increased
more advanced liver disease (Child-Pugh B and C
effect was especially marked in patients with
who had stable or progressive thrombosis. This
group of patients matched by severity of cirrhosis
(partial or complete) had better survival than a
treatment-associated portal vein recanalisation
patients
pensation.60,69–71 Moreover, in a recent study,
pressure68 and may reduce the risk of decom-
successful treatment of PVT decreases variceal
outcome. In this regard, it has been shown that
to treatment, mostly anticoagulation, and patient
In a systematic review by Rodriguez et al., the

To better understand the impact of PVT on the

It is important to point out that, while longer

who achieved spontaneous or

advanced pre-transplant Yerdel’s grades, is associ- However, not all studies are in agreement.70,71
ated with significant mortality, therefore anti- There is evidence that suggests that PVT is more
coagulation is recommended in the postoperative likely to resolve when the thrombus is partial than
Key point
period.7 when it is complete, although the difference (22%)
In fact, in the case of complete obstruction of was not significantly different, perhaps because of PVT increases the risk of
the portal vein with extension into the superior the small number of patients with complete PVT.70 acute variceal bleeding and
recurrence of bleeding,
mesenteric vein preventing the performance of a In 1 study, a model developed to predict suc-
although its impact on
physiological portal-to-portal vein anastomosis cessful anticoagulation in 65 patients correlated mortality has mostly been
during LT, several different technical approaches the lack of efficacy of anticoagulation with more shown in the post-trans-
have been proposed, such as portacaval hemi- severe liver disease, complete vs. partial PVT, age of plant setting.
transposition. However, most of these techniques the thrombus, and time from treatment start (>6
are unable to resolve the portal hypertensive state, months).81 Interestingly, in this study, no case of
which complicates the postoperative course (with recanalisation was observed in patients with Child-
variceal bleeding or persistent ascites) in approxi- Pugh C cirrhosis (only 5 patients).73
mately 50% of patients.79 Azoulay et al. proposed a Mean time to recanalisation ranges from 5.5 to 8
left renal vein to graft portal vein anastomosis in months and ACG and AISF guidelines suggest a
patients with spontaneous or surgical spleno renal treatment course of 6 months94,95; however, these
shunts (reno-portal anastomosis) as a successful recommendations are based on data that can be
strategy in patients for whom portal-to-portal biased given the very different schedule of follow-
anastomosis was not possible.55 All these strate- up imaging. It has been suggested that some pa-
gies highlight the need to discuss all potential tients may achieve recanalisation with extension of
means of achieving portal vein reconstruction in anticoagulation to up to 1 year,80 however, few
patients with PVT before deeming them non- patients will respond after 6 months.81 It has also
transplant candidates. been suggested that recanalisation is more likely to
occur in patients with partial PVT compared to
Treatment of PVT in cirrhosis those with complete PVT.70,81 Independent of
Anticoagulation length of anticoagulation therapy, it is important to
Efficacy mention that recurrence of thrombosis is frequent,
Several cohort studies have investigated the effi- ranging from 27% to 38% of cases.71,80,96 The time
cacy and safety of anticoagulation for the treat- elapsed between discontinuation of anti-
ment of PVT in patients with coagulation therapy and recurrence of thrombosis
cirrhosis15,16,20,60,70–73,80–92 (Table S1). However, ranges between 2 and 5 months.71,96 Therefore, in
most studies are retrospective with different in- order to prevent re-thrombosis, particularly in pa-
clusion criteria and coagulation regimens. Two tients listed for LT, long-term anticoagulation
separate meta-analyses summarise published should be considered.28
studies.58,93 The most recent includes data from 8
studies, comprising 353 patients. Excluding the Types of anticoagulation
study by Wang et al. that enrolled patients who LMWH is the initial treatment of choice. One study
underwent TIPS placement,57 a significantly higher investigated 2 separate doses of LMWH (1.5 mg/kg
proportion of patients treated with anticoagulants daily and 1 mg/kg twice daily) showing a similar
experienced some degree of PVT recanalisation rate of recanalisation with fewer bleeding compli-
compared to those who did not receive anticoag- cations in the twice daily dosage group.15 In 1
ulants (63% vs. 22%, respectively). In addition, in 5 study, reducing the LMWH dose by 30% for patients
studies (comprising 217 patients) that described with a platelet count lower than 50,000/ul did not
the rate of complete PVT recanalisation, it was significantly decrease the efficacy of
significantly higher in patients on anticoagulation anticoagulation.81
(42%) vs. those not on anticoagulation (12%). Meta- It should be noted that basing LMWH dosing on
regression analysis suggests that low molecular anti-Xa activity could inadvertently lead to higher
weight heparin (LMWH) could be more effective than expected anticoagulant levels because of the
than warfarin.58 innate decrease in antithrombin associated with
Among several published cohorts, pretreatment liver synthetic dysfunction.97 This effect is espe-
predictive factors of anticoagulant treatment effi- cially pronounced at higher Child-Pugh stages and
cacy were analysed in 6 studies, demonstrating a can lead to bleeding complications.98 Therefore,
relatively strong correlation between early anti- anti-Xa activity assays should not be used as a
coagulation and likelihood of recanalisation. Initi- LMWH dosing guide in patients with cirrhosis. In
ation of anticoagulation less than 6 months after the presence of renal dysfunction, LMWH requires
the diagnosis of PVT is the most important factor dose reduction. Fondaparinux (FPX) exerts its
predicting recanalisation.71,81 Other factors, such as function primarily through selective binding to
involvement of mesenteric veins and/or the antithrombin, causing rapid inhibition of activated
severity of baseline liver disease, have also been factor X99; it can be prescribed in fixed dosages
proposed as possible predictive factors.15,83 once a day with comparable or even greater

Journal of Hepatology 2021 vol. 75 j 442–453 447

Review

efficacy than LMWH. Furthermore, unlike unfrac- therefore data are lacking to make firm recom-
tionated heparin and LMWH, FPX does not bind to mendations in this respect.
Key point
platelet factor IV and the risk of heparin-induced According to a meta-analysis by Loffredo et al.
Anticoagulation should be thrombocytopenia is rare.100 Although reports on that included 4 studies describing the rate of acute
attempted for at least 6 its use for the treatment of PVT are scarce its effi- variceal bleeding,15,60,88 the risk of portal hyper-
months using low molecu-
cacy appears not to be inferior to LMWH.101 tensive bleeding was actually reduced in the anti-
lar weight heparin or ace-
nocumarol. Direct oral Given the parenteral administration of LMWH, coagulated group and this was confirmed in 2
anticoagulants can be used the compliance and quality of life of patients can be subsequent studies.70,72 On the contrary, patients
in special situations, but an issue, therefore vitamin K antagonists (VKAs) with cirrhosis receiving anticoagulation for in-
evidence is still lacking
can be used for long-term treatment. However, in dications other than PVT seem to have a higher risk
regarding their safety and
efficacy. Monitoring and patients with baseline alterations in INR, the of variceal bleeding than matched cohorts of non-
interpretation of standard therapeutic range of VKA can be difficult to anticoagulated patients with cirrhosis109,110
markers of anticoagulation maintain. Management of oesophageal varices (primary
is challenging given alter- In the last decade, direct oral anticoagulants and secondary prophylaxis) must differ from that
ations at baseline in
(DOACs) have emerged as a safe alternative to VKAs recommended for patients without PVT and should
cirrhosis.
when long-term anticoagulation treatment is follow the Baveno and AASLD guidelines.111,112
needed in patients with atrial fibrillation or venous Recent evidence suggests that band ligation can
thromboembolism. Underlying liver dysfunction be performed without the need for peri-procedure
severely affects the pharmacokinetics of DOACs withdrawal of anticoagulation as there are no dif-
because of plasmatic binding protein, cytochrome ferences in bleeding risk from post banding ul-
p450-mediated metabolism, biliary excretion and cers.113–115 However, the number of patients
renal clearance.102 For this reason, the FDA and included in these studies was small and therefore a
EMA do not recommend the use of these agents in formal recommendation cannot be made.
patients with Child-Pugh C cirrhosis, while they
recommend caution in patients with Child-Pugh B Thrombolysis
cirrhosis and no restrictions for those with Child- Systemic or local thrombolysis using streptokinase,
Pugh A cirrhosis. However, there are no specific urokinase or tissue plasminogen activator with
data on safety in these populations. concomitant administration of LMWH has been
Data on the use of DOACs in the clinical context attempted in a few cases68 to treat recent PVT but
of PVT in liver disease derives from small cohorts cannot be recommended because of a high rate of
and 1 RCT.103–107 From these studies, it seems that bleeding complications and a lack of evidence
DOACs could have at least the same efficacy as VKAs showing it to be superior to anticoagulation alone.
and similar side effects, but with much simpler
administration because no monitoring is required. Transjugular intrahepatic portosystemic shunt
More studies are needed to identify the specific role (TIPS)
of DOACs and the most appropriate one for the In previous decades, PVT had been considered a
treatment of PVT in patients with cirrhosis. contraindication to TIPS placement. However, since
the first reports in the early 1990s,116–118 the cu-
Complications of anticoagulation mulative evidence – derived from non-controlled
Anticoagulation does not seem to lead to a higher and retrospective cohort series – shows that TIPS
rate of bleeding complications in patients with can be performed successfully in 75%–100% of pa-
cirrhosis compared to anticoagulation in patients tients. In a recent meta-analysis, TIPS was able to
without cirrhosis. According to the meta-analysis reduce clot burden with partial recanalisation in
by Loffredo et al. no difference in the risk of 84% of patients with complete recanalisation in
119
bleeding was found between LMWH and VKA.58 In 73%. Overall, 95% of patients with complete
patients treated with VKA, a study showed that recanalisation after TIPS maintained patency of the
patients with a platelet count <50,000/ul had a portal vein. If portal vein flow is ensured after the
57
greater bleeding risk.70 However, this cut-off value procedure, anticoagulation is not mandatory.
has not been confirmed in other studies. In settings However, when discussing the potential role of
other than PVT in cirrhosis, it has been proposed to TIPS in PVT treatment, the characteristics of the
120
reduce the dose of anticoagulant (i.e. LMWH) in thrombus should be considered. TIPS is feasible in
121,122
patients with a platelet count below 50,000/ul and a patient with partial or even occlusive PVT,
to consider discontinuation in patients with provided that the portal vein can still be identified.
platelet counts below 30,000/ul.108 Reducing the The presence of portal cavernoma or an inability to
LMWH dose by 30% for patients with a platelet identify the intrahepatic portal trunk or intrahepatic
count below 50,000/ul did not result in a reduction portal vein branches markedly increases the tech-
in bleeding complications in a single study,81 nical difficulties of TIPS creation. These procedures

448 Journal of Hepatology 2021 vol. 75 j 442–453

 Patient with cirrhosis and PVT

Chronic PVT without PV >50% occlusion/progressive PVT of the trunk or both main
remnant or splenic vein branches in a potential transplant candidate or extension
access of thrombus into SMV*

Anticoagulation**
(LMWH, VKA, DOACs)
No anticoagulation If h/o variceal
nor TIPS*** haemorrhage or hard-
Evaluation at 3-6 months to-treat ascites
(CT scan or MRI)

Partial recanalisation or Progression of PVT or

resolution of occlusive PVT lack of improvement of
or stabilisation of partial PVT occlusive PVT

Continue anticoagulation Consider TIPS

for life or until transplant ± mechanical thrombolysis

Fig. 2. Approach to the patient with cirrhosis and PVT. *In patients who do not meet these criteria, a Doppler ultrasound
should be performed every 3 months and anticoagulation is not recommended; **Provided screening for varices/CSPH and
measures to prevent variceal haemorrhage are in place; if procedures are foreseen (e.g. EVL, LVP) in the near future, start with
LMWH and then transition to VKA/DOAC; ***In the presence of PV remnant or splenic vein access, experienced inter-
ventionalists may attempt portal vein recanalisation plus TIPS. DOAC, direct oral anticoagulant; EVL, endoscopic variceal
ligation; LMWH, low molecular weight heparin; LVP, large volume paracentesis; PVT, portal vein thrombosis; PV, portal vein;
SMV, superior mesenteric vein; TIPS, transjugular intrahepatic portosystemic shunt; VKA, vitamin K antagonist.

are technically challenging and radiation inten- increasing the risk of hepatic encephalopathy or
sive123; despite this, an increasing number of these adverse events. However, this benefit did not
patients have undergone successful TIPS placement translate into an improvement in survival.125,126
as familiarity with this procedure grows. In patients TIPS can be considered for the treatment of PVT
with cavernoma, the recanalisation of the portal vein in the case of an absolute contraindication to
remnant should be preferred to the placement of a anticoagulation or in the case of no response after a
stent in a collateral vessel as the latter poses an maximum of 6 months of anticoagulation treat-
increased risk of haemoperitoneum and is less ment.127 In LT candidates, the indications for TIPS
effective for portal system decompression. In the case placement are identical to those recommended for
of chronic complete/occlusive PVT or portal cav- non-listed patients. A specific aim of TIPS in can-
ernoma the absence of an appropriate landing zone at didates for LT is to guarantee patency of the portal
the distal end of the portal vein or at the spleno- vein, if the risk of progression to grade 3-4 Yerdel is
mesenteric confluence should be considered as a high or already present. To ensure physiological
Key point
relative contraindication to TIPS creation.122 The re- portal vein reconstruction in cases with occlusive
ported technical success rate of TIPS placement in PVT, portal vein recanalisation using a trans-splenic A transjugular intrahepatic
patients with PVT in highly specialised centres is TIPS procedure has been shown to be successful in portosystemic shunt (TIPS)
with or without thrombec-
75%–98%, with a major complication rate of 10%.119 98% of cases in a single cohort study, with portal
tomy is feasible in most
In cases of very extensive portal vein thrombus, vein patency obtained in 92% of patients.128 Only 1 cases and may reduce the
mechanical thrombolysis during the TIPS proced- RCT showed no advantage of anticoagulation in burden of thrombosis. The
ure may help to achieve portal vein maintaining portal vein patency after TIPS place- use of TIPS is generally
reserved for patients who
recanalisation.124 ment.57 However, this will probably depend on
do not respond to 6
In most of the reported studies, most patients in post-TIPS haemodynamics and no definitive rec- months of anticoagulation.
whom TIPS was placed still had an identifiable ommendations can be provided at this point. However, it could be
portal vein and the indication for TIPS was not the considered first line in the
PVT itself but the treatment of a portal Suggested treatment algorithm presence of significant
complications of portal hy-
hypertension-related complication (ascites, vari- Patients with thrombosis of the main portal vein or
pertension, old thrombus
ceal haemorrhage) in which PVT was present. In intrahepatic branches involving less than 50% of in a transplant candidate or
this regard, 2 RCTs in patients with cirrhosis, vari- the lumen of the vessels can be followed-up every contraindications to
ceal bleeding and PVT demonstrate that TIPS is 3 months to scan for progression of thrombosis anticoagulation.
more effective than ligation combined with pro- unless the patient is on the LT waiting list or there
pranolol and anticoagulation in preventing re- is involvement of the superior mesenteric vein
bleeding and achieving PVT resolution without which is, per se, an indication for anticoagulation.

Journal of Hepatology 2021 vol. 75 j 442–453 449

Review

Potential LT candidates or those patients with OR, odds ratio; PVT, portal vein thrombosis; SMV,
progressive or occlusive PVT with more than 50% superior mesenteric vein; TIPS, transjugular intra-
occlusion of the main portal vein branch or both heptic portosystemic shunt; VKA, vitamin K
intrahepatic branches should be treated. Anti- antagonist; vWf, von Willebrand factor.
coagulation is the first choice if the main portal
vein can be identified and there is no cavernoma. Financial support
The use of TIPS is generally reserved for patients The authors received no financial support to pro-
who do not respond to 6 months of anti- duce this manuscript.
coagulation. However, it could be considered first
line in the following circumstances: a) presence of Conflict of interest
significant complications of portal hypertension The authors declare no conflicts of interest that
(e.g. variceal haemorrhage, difficult to treat ascites) pertain to this work.
that would also require TIPS; b) identification of an Please refer to the accompanying ICMJE disclo-
old thrombus (>6 months), unlikely to respond to sure forms for further details.
anticoagulation, in a transplant candidate; and c)
contraindications to anticoagulation (Fig. 2). Authors’ contributions
MS, GGT and JCGP study concept and design, re-
Abbreviations view of the literature, writing of article.
DOACs, direct oral anticoagulants; FPX, fondapar-
inux; HCC, hepatocellular carcinoma; JAK, Janus Supplementary data
kinase; LMWH, low molecular weight heparin; MP, Supplementary data to this article can be found on-
microparticle; NSBBs, non-selective beta blockers; line at https://doi.org/10.1016/j.jhep.2021.04.029.

References [13] Villa E, Cammà C, Marietta M, Luongo M, Critelli R, Colopi S, et al.

Enoxaparin prevents portal vein thrombosis and liver decompensation in
Author names in bold designate shared co-first authorship
patients with advanced cirrhosis. Gastroenterology 2012;143. 1253-
1260.e1254.
[1] Intagliata NM, Argo CK, Stine JG, Lisman T, Caldwell SH, Violi F. Concepts [14] Zanetto A, Senzolo M, Vitale A, Cillo U, Radu C, Sartorello F, et al.
and controversies in haemostasis and thrombosis associated with liver Thromboelastometry hypercoagulable profiles and portal vein throm-
disease: proceedings of the 7th international coagulation in liver disease bosis in cirrhotic patients with hepatocellular carcinoma. Dig Liver Dis
conference. Thromb Haemost 2018;118:1491–1506. 2017;49:440–445.
[2] Rajani R, Björnsson E, Bergquist A, Danielsson A, Gustavsson A, Grip O, [15] Cui SB, Shu RH, Yan SP, Wu H, Chen Y, Wang L, et al. Efficacy and safety of
et al. The epidemiology and clinical features of portal vein thrombosis: a anticoagulation therapy with different doses of enoxaparin for portal
multicentre study. Aliment Pharmacol Ther 2010;32:1154–1162. vein thrombosis in cirrhotic patients with hepatitis B. Eur J Gastroenterol
[3] Tsochatzis EA, Senzolo M, Germani G, Gatt A, Burroughs AK. Systematic Hepatol 2015;27:914–919.
review: portal vein thrombosis in cirrhosis. Aliment Pharmacol Ther [16] Noronha Ferreira C, Marinho RT, Cortez-Pinto H, Ferreira P, Dias MS,
2010;31:366–374. Vasconcelos M, et al. Incidence, predictive factors and clinical signifi-
[4] Violi F, Corazza GR, Caldwell SH, Talerico G, Romiti GF, Napoleone L, et al. cance of development of portal vein thrombosis in cirrhosis: a pro-
Incidence and recurrence of portal vein thrombosis in cirrhotic patients. spective study. Liver Int 2019;39:1459–1467.
Thromb Haemost 2019;119:496–499. [17] Turon F, Driever E, Baiges A, Cerda E, García-Criado Á, Gilabert R, et al.
[5] Ponziani FR, Zocco MA, Senzolo M, Pompili M, Gasbarrini A, Avolio AW. Predicting portal thrombosis in cirrhosis: a prospective comprehensive
Portal vein thrombosis and liver transplantation: implications for wait- study of clinical, ultrasonografic and hemostatic factors. J Hep 2021. in
ing list period, surgical approach, early and late follow-up. Transpl Rev press.
(Orlando) 2014;28:92–101. [18] Nery F, Correia S, Macedo C, Gandara J, Lopes V, Valadares D, et al.
[6] Montenovo M, Rahnemai-Azar A, Reyes J, Perkins J. Clinical impact and Nonselective beta-blockers and the risk of portal vein thrombosis in
risk factors of portal vein thrombosis for patients on wait list for liver patients with cirrhosis: results of a prospective longitudinal study.
transplant. Exp Clin Transpl 2018;16:166–171. Aliment Pharmacol Ther 2019;49:582–588.
[7] Rodriguez-Castro KI, Porte RJ, Nadal E, Germani G, Burra P, Senzolo M. [19] Stine JG, Wang J, Shah PM, Argo CK, Intagliata N, Uflacker A, et al. Decreased
Management of nonneoplastic portal vein thrombosis in the setting of portal vein velocity is predictive of the development of portal vein throm-
liver transplantation: a systematic review. Transplantation bosis: a matched case-control study. Liver Int 2018;38:94–101.
2012;94:1145–1153. [20] Francoz C, Valla D, Durand F. Portal vein thrombosis, cirrhosis, and liver
[8] Zocco MA, Di Stasio E, De Cristofaro R, Novi M, Ainora ME, Ponziani F, transplantation. J Hepatol 2012;57:203–212.
et al. Thrombotic risk factors in patients with liver cirrhosis: correlation [21] Xu X, Guo X, De Stefano V, Silva-Junior G, Goyal H, Bai Z, et al. Nonse-
with MELD scoring system and portal vein thrombosis development. lective beta-blockers and development of portal vein thrombosis in liver
J Hepatol 2009;51:682–689. cirrhosis: a systematic review and meta-analysis. Hepatol Int
[9] Maruyama H, Okugawa H, Takahashi M, Yokosuka O. De novo portal vein 2019;13:468–481.
thrombosis in virus-related cirrhosis: predictive factors and long-term [22] Tripodi A, Mannucci PM. The coagulopathy of chronic liver disease.
outcomes. Am J Gastroenterol 2013;108:568–574. N Engl J Med 2011;365:147–156.
[10] John BV, Konjeti R, Aggarwal A, Lopez R, Atreja A, Miller C, et al. Impact [23] La Mura V, Tripodi A, Tosetti G, Cavallaro F, Chantarangkul V, Colombo M,
of untreated portal vein thrombosis on pre and post liver transplant et al. Resistance to thrombomodulin is associated with de novo portal
outcomes in cirrhosis. Ann Hepatol 2013;12:952–958. vein thrombosis and low survival in patients with cirrhosis. Liver Int
[11] Abdel-Razik A, Mousa N, Elhelaly R, Tawfik A. De-novo portal vein 2016;36:1322–1330.
thrombosis in liver cirrhosis: risk factors and correlation with the model [24] Ma SD, Wang J, Bezinover D, Kadry Z, Northup PG, Stine JG. Inherited
for End-stage Liver Disease scoring system. Eur J Gastroenterol Hepatol thrombophilia and portal vein thrombosis in cirrhosis: a systematic re-
2015;27:585–592. view and meta-analysis. Res Pract Thromb Haemost 2019;3:658–667.
[12] Nery F, Chevret S, Condat B, de Raucourt E, Boudaoud L, Rautou PE, et al. [25] Tremblay D, Naymagon L, Troy K, Cromwell C, Edwards C, Schiano T, et al.
Causes and consequences of portal vein thrombosis in 1,243 patients with The utility of thrombophilia testing in patients with newly diagnosed
cirrhosis: results of a longitudinal study. Hepatology 2015;61:660–667. portal vein thrombosis. Blood Coagul Fibrinol 2020;31:213–218.

450 Journal of Hepatology 2021 vol. 75 j 442–453

[26] Saugel B, Lee M, Feichtinger S, Hapfelmeier A, Schmid RM, Siveke JT. [48] Nonami T, Yokoyama I, Iwatsuki S, Starzl TE. The incidence of portal
Thrombophilic factor analysis in cirrhotic patients with portal vein vein thrombosis at liver transplantation. Hepatology 1992;16:1195–
thrombosis. J Thromb Thrombol 2015;40:54–60. 1198.
[27] Ayala R, Rapado I, Grande S, Moreno E, Martinez-Lopez J. JAK2 exon 12 [49] Gayowski TJ, Marino IR, Doyle HR, Echeverri L, Mieles L, Todo S, et al.
mutations were not found in liver transplant recipients with or without A high incidence of native portal vein thrombosis in veterans under-
pretransplant portal vein thrombosis. Eur J Gastroenterol Hepatol going liver transplantation. J Surg Res 1996;60:333–338.
2014:362–363. England. [50] Yerdel MA, Gunson B, Mirza D, Karayalçin K, Olliff S, Buckels J, et al.
[28] EASL Clinical Practice Guidelines. Vascular diseases of the liver. J Hepatol Portal vein thrombosis in adults undergoing liver transplantation: risk
2016;64:179–202. factors, screening, management, and outcome. Transplantation
[29] Connolly GC, Chen R, Hyrien O, Mantry P, Bozorgzadeh A, Abt P, et al. Inci- 2000;69:1873–1881.
dence, risk factors and consequences of portal vein and systemic throm- [51] Jamieson NV. Changing perspectives in portal vein thrombosis and liver
boses in hepatocellular carcinoma. Thromb Res 2008;122:299–306. transplantation. Transplantation 2000;69:1772–1774.
[30] Carnevale R, Raparelli V, Nocella C, Bartimoccia S, Novo M, Severino A, [52] Charco R, Fuster J, Fondevila C, Ferrer J, Mans E, García-Valdecasas JC.
et al. Gut-derived endotoxin stimulates factor VIII secretion from Portal vein thrombosis in liver transplantation. Transpl Proc
endothelial cells. Implications for hypercoagulability in cirrhosis. 2005;37:3904–3905.
J Hepatol 2017;67:950–956. [53] Bauer J, Johnson S, Durham J, Ludkowski M, Trotter J, Bak T, et al. The role
[31] Kapiotis S, Besemer J, Bevec D, Valent P, Bettelheim P, Lechner K, et al. of TIPS for portal vein patency in liver transplant patients with portal
Interleukin-4 counteracts pyrogen-induced downregulation of throm- vein thrombosis. Liver Transpl 2006;12:1544–1551.
bomodulin in cultured human vascular endothelial cells. Blood [54] Sarin SK, Philips CA, Kamath PS, Choudhury A, Maruyama H, Nery FG, et al.
1991;78:410–415. Toward a comprehensive new classification of portal vein thrombosis in
[32] Queck A, Carnevale R, Uschner FE, Schierwagen R, Klein S, Jansen C, et al. patients with cirrhosis. Gastroenterology 2016;151. 574-577.e573.
Role of portal venous platelet activation in patients with decompensated [55] Bhangui P, Lim C, Levesque E, Salloum C, Lahat E, Feray C, et al. Novel
cirrhosis and TIPS. Gut 2020:1535–1536. classification of non-malignant portal vein thrombosis: a guide to sur-
[33] Shalaby S, Simioni P, Campello E, Spiezia L, Gavasso S, Bizzaro D, et al. gical decision-making during liver transplantation. J Hepatol
Endothelial damage of the portal vein is associated with heparin-like 2019;71:1038–1050.
effect in advanced stages of cirrhosis. Thromb Haemost [56] Yerdel MA, Gunson B, Mirza D, Karayalcin K, Olliff S, Buckels J, et al.
2020;120:1173–1181. Portal vein thrombosis in adults undergoing liver transplantation: risk
[34] Amitrano L, Brancaccio V, Guardascione MA, Margaglione M, factors, screening, management, and outcome. Transplantation
Iannaccone L, D’Andrea G, et al. Inherited coagulation disorders in 2000;69:1873–1881.
cirrhotic patients with portal vein thrombosis. Hepatology [57] Wang Z, Jiang MS, Zhang HL, Weng NN, Luo XF, Li X, et al. Is post-TIPS
2000;31:345–348. anticoagulation therapy necessary in patients with cirrhosis and portal
[35] Politoske D, Ralls P, Korula J. Portal vein thrombosis following endo- vein thrombosis? A randomized controlled trial. Radiology
scopic variceal sclerotherapy. Prospective controlled comparison in pa- 2016;279:943–951.
tients with cirrhosis. Dig Dis Sci 1996;41:185–190. [58] Loffredo L, Pastori D, Farcomeni A, Violi F. Effects of anticoagulants in
[36] Gong C, Qin X, Yang J, Guo T. The best anticoagulation strategy for patients with cirrhosis and portal vein thrombosis: a systematic review
cirrhotic patients who underwent splenectomy: a Network meta-anal- and meta-analysis. Gastroenterology 2017;153. 480-487.e481.
ysis. Gastroenterol Res Pract 2017;2017:9216172. [59] Luca A, Caruso S, Milazzo M, Marrone G, Mamone G, Crino F, et al.
[37] Ghabril M, Agarwal S, Lacerda M, Chalasani N, Kwo P, Tector AJ. Portal Natural course of extrahepatic nonmalignant partial portal vein
vein thrombosis is a risk factor for poor early outcomes after liver thrombosis in patients with cirrhosis. Radiology 2012;265:124–132.
transplantation: analysis of risk factors and outcomes for portal vein [60] Senzolo M, Sartori TM, Rossetto V, Burra P, Cillo U, Boccagni P, et al.
thrombosis in waitlisted patients. Transplantation 2016;100:126–133. Prospective evaluation of anticoagulation and transjugular intrahepatic
[38] Stine JG, Shah NL, Argo CK, Pelletier SJ, Caldwell SH, Northup PG. portosystemic shunt for the management of portal vein thrombosis in
Increased risk of portal vein thrombosis in patients with cirrhosis due to cirrhosis. Liver Int 2012;32:919–927.
nonalcoholic steatohepatitis. Liver Transpl 2015;21:1016–1021. [61] Hidaka H, Kokubu S, SatoT, Katsushima S, Izumi N, Igura T, et al. Antithrombin
[39] Tripodi A, Fracanzani AL, Primignani M, Chantarangkul V, Clerici M, III for portal vein thrombosis in patients with liver disease: a randomized,
Mannucci PM, et al. Procoagulant imbalance in patients with non- double-blind, controlled trial. Hepatol Res 2018;48:E107–E116.
alcoholic fatty liver disease. J Hepatol 2014;61:148–154. [62] Zhang Y, Xu BY, Wang XB, Zheng X, Huang Y, Chen J, et al. Prevalence and
[40] Margini C, Berzigotti A. Portal vein thrombosis: the role of imaging in the clinical significance of portal vein thrombosis in patients with cirrhosis
clinical setting. Dig Liver Dis 2017;49:113–120. and acute decompensation. Clin Gastroenterol Hepatol 2020.
[41] De Gaetano AM, Lafortune M, Patriquin H, De Franco A, Aubin B, [63] Ageno W, Riva N, Schulman S, Beyer-Westendorf J, Bang SM, Senzolo M,
Paradis K. Cavernous transformation of the portal vein: patterns of et al. Long-term clinical outcomes of splanchnic vein thrombosis:
intrahepatic and splanchnic collateral circulation detected with Doppler results of an international registry. JAMA Intern Med 2015;175:1474–
sonography. AJR Am J Roentgenol 1995;165:1151–1155. 1480.
[42] Hernández-Gea V, De Gottardi A, Leebeek FWG, Rautou PE, Salem R, [64] Amitrano L, Guardascione MA, Brancaccio V, Iannaccone L, Ames PR,
Garcia-Pagan JC. Current knowledge in pathophysiology and manage- Balzano A. Portal and mesenteric venous thrombosis in cirrhotic pa-
ment of Budd-Chiari syndrome and non-cirrhotic non-tumoral tients. Gastroenterology 2002:1409–1410. United States.
splanchnic vein thrombosis. J Hepatol 2019;71:175–199. [65] Amitrano L, Guardascione MA, Martino R, Manguso F, Menchise A,
[43] Tessler FN, Gehring BJ, Gomes AS, Perrella RR, Ragavendra N, Balzano A. Hypoxic hepatitis occurring in cirrhosis after variceal
Busuttil RW, et al. Diagnosis of portal vein thrombosis: value of color bleeding: still a lethal disease. J Clin Gastroenterol 2012;46:608–612.
Doppler imaging. AJR Am J Roentgenol 1991;157:293–296. [66] Amitrano L, Guardascione MA, Manguso F, Bennato R, Bove A, DeNucci C,
[44] Mathieu D, Vasile N, Grenier P. Portal thrombosis: dynamic CT features et al. The effectiveness of current acute variceal bleed treatments in
and course. Radiology 1985;154:737–741. unselected cirrhotic patients: refining short-term prognosis and risk
[45] Jha RC, Khera SS, Kalaria AD. Portal vein thrombosis: imaging the factors. Am J Gastroenterol 2012;107:1872–1878.
spectrum of disease with an emphasis on MRI features. AJR Am J [67] D’Amico G, De Franchis R. Upper digestive bleeding in cirrhosis. Post-
Roentgenol 2018;211:14–24. therapeutic outcome and prognostic indicators. Hepatology
[46] Sherman CB, Behr S, Dodge JL, Roberts JP, Yao FY, Mehta N. Dis- 2003;38:599–612.
tinguishing tumor from bland portal vein thrombus in liver transplant [68] De Santis A, Moscatelli R, Catalano C, Iannetti A, Gigliotti F, Cristofari F,
candidates with hepatocellular carcinoma: the A-VENA criteria. Liver et al. Systemic thrombolysis of portal vein thrombosis in cirrhotic pa-
Transpl 2019;25:207–216. tients: a pilot study. Dig Liver Dis 2010;42:451–455.
[47] Stieber AC, Zetti G, Todo S, Tzakis AG, Fung JJ, Marino I, et al. The [69] Senzolo M, Riva N, Dentali F, Sartori MT, Ageno W. Portal vein throm-
spectrum of portal vein thrombosis in liver transplantation. Ann Surg bosis in decompensated cirrhosis: rationale for treatment. Clin Transl
1991;213:199–206. Gastroenterol 2019;10:e00026.

Journal of Hepatology 2021 vol. 75 j 442–453 451

Review

[70] La Mura V, Braham S, Tosetti G, Branchi F, Bitto N, Moia M, et al. Harmful anticoagulation is effective and safe. Clin Res Hepatol Gastroenterol
and beneficial effects of anticoagulants in patients with cirrhosis and 2019;43:395–402.
portal vein thrombosis. Clin Gastroenterol Hepatol 2018;16. 1146-1152. [92] Artaza T, Lopes M, Romero M, Gómez AZ, de la Cruz G, Sánchez JJ, et al.
e1144. Efficacy and safety of anticoagulation in non-malignant portal vein
[71] Delgado MG, Seijo S, Yepes I, Achécar L, Catalina MV, García-Criado A, thrombosis in patients with liver cirrhosis. Gastroenterol Hepatol
et al. Efficacy and safety of anticoagulation on patients with cirrhosis and 2018;41:611–617.
portal vein thrombosis. Clin Gastroenterol Hepatol 2012;10:776–783. [93] Qi X, Dai J, Jia J, Ren W, Yang M, Li H, et al. Association between portal
[72] Senzolo M, Riva N, Dentali F, Rodriguez-Castro K, Sartori MT, Bang SM, vein thrombosis and survival of liver transplant recipients: a systematic
et al. Long-term outcome of splanchnic vein thrombosis in cirrhosis. Clin review and meta-analysis of observational studies. J Gastrointestin Liver
Transl Gastroenterol 2018;9:176. Dis 2015;24. 51-59, 54 p following 59.
[73] Pettinari I, Vukotic R, Stefanescu H, Pecorelli A, Morelli M, Grigoras C, [94] Simonetto DA, Singal AK, Garcia-Tsao G, Caldwell SH, Ahn J, Kamath PS.
et al. Clinical impact and safety of anticoagulants for portal vein ACG clinical guideline: disorders of the hepatic and mesenteric circula-
thrombosis in cirrhosis. Am J Gastroenterol 2019;114:258–266. tion. Am J Gastroenterol 2020;115:18–40.
[74] Cerini F, Vilaseca M, Lafoz E, García-Irigoyen O, García-Calderó H, [95] Senzolo M, Riggio O, Primignani M. Vascular disorders of the liver:
Tripathi DM, et al. Enoxaparin reduces hepatic vascular resistance and recommendations from the Italian Association for the Study of the Liver
portal pressure in cirrhotic rats. J Hepatol 2016;64:834–842. (AISF) ad hoc committee. Dig Liver Dis 2011;43:503–514.
[75] Vilaseca M, García-Calderó H, Lafoz E, García-Irigoyen O, Avila MA, [96] Rodríguez-Castro KI, Senzolo M, Sartori MT. Optimal length of antico-
Reverter JC, et al. The anticoagulant rivaroxaban lowers portal hyper- agulant therapy in cirrhotic patients with portal vein thrombosis.
tension in cirrhotic rats mainly by deactivating hepatic stellate cells. Clin Gastroenterol Hepatol 2012:820–821. United States. author reply
Hepatology 2017;65:2031–2044. 821.
[76] Serper M, Weinberg EM, Cohen JB, Reese PP, Taddei TH, Kaplan DE. [97] Bechmann LP, Sichau M, Wichert M, Gerken G, Kröger K, Hilgard P. Low-
Mortality and hepatic decompensation in patients with cirrhosis and molecular-weight heparin in patients with advanced cirrhosis. Liver Int
atrial fibrillation treated with anticoagulation. Hepatology 2020. 2011;31:75–82.
[77] Valeriani E, Di Nisio M, Riva N, Cohen O, Porreca E, Senzolo M, et al. [98] Senzolo M, Rodriguez-Castro KI, Rossetto V, Radu C, Gavasso S, Carraro P,
Anticoagulant treatment for splanchnic vein thrombosis in liver et al. Increased anticoagulant response to low-molecular-weight heparin
cirrhosis: a systematic review and meta-analysis. Thromb Haemost 2021. in plasma from patients with advanced cirrhosis. J Thromb Haemost
[78] Zanetto A, Rodriguez-Kastro KI, Germani G, Ferrarese A, Cillo U, Burra P, 2012;10:1823–1829.
et al. Mortality in liver transplant recipients with portal vein thrombosis [99] Zhang Y, Zhang M, Tan L, Pan N, Zhang L. The clinical use of Fondapar-
- an updated meta-analysis. Transpl Int 2018;31:1318–1329. inux: a synthetic heparin pentasaccharide. Prog Mol Biol Transl Sci
[79] Ravaioli M, Zanello M, Grazi GL, Ercolani G, Cescon M, Del Gaudio M, 2019;163:41–53.
et al. Portal vein thrombosis and liver transplantation: evolution during [100] Lobo B, Finch C, Howard A, Minhas S. Fondaparinux for the treatment of
10 years of experience at the University of Bologna. Ann Surg patients with acute heparin-induced thrombocytopenia. Thromb Hae-
2011;253:378–384. most 2008;99:208–214.
[80] Amitrano L, Guardascione MA, Menchise A, Martino R, Scaglione M, [101] Senzolo M, Piano S, Shalaby S, Tonon M, Tonello S, Zanetto A, et al.
Giovine S, et al. Safety and efficacy of anticoagulation therapy with low Comparison of fondaparinux and low molecular weight heparin in the
molecular weight heparin for portal vein thrombosis in patients with treatment of portal vein thrombosis in cirrhosis. Am J Med 2021. In press.
liver cirrhosis. J Clin Gastroenterol 2010;44:448–451. [102] Turco L, de Raucourt E, Valla DC, Villa E. Anticoagulation in the cirrhotic
[81] Rodriguez-Castro KI, Vitale A, Fadin M, Shalaby S, Zerbinati P, Sartori MT, patient. JHEP Rep 2019;1:227–239.
et al. A prediction model for successful anticoagulation in cirrhotic [103] Hanafy AS, Abd-Elsalam S, Dawoud MM. Randomized controlled trial of
portal vein thrombosis. Eur J Gastroenterol Hepatol 2019;31:34–42. rivaroxaban versus warfarin in the management of acute non-neoplastic
[82] Werner KT, Sando S, Carey EJ, Vargas HE, Byrne TJ, Douglas DD, et al. portal vein thrombosis. Vasc Pharmacol 2019;113:86–91.
Portal vein thrombosis in patients with end stage liver disease awaiting [104] De Gottardi A, Trebicka J, Klinger C, Plessier A, Seijo S, Terziroli B, et al.
liver transplantation: outcome of anticoagulation. Dig Dis Sci Antithrombotic treatment with direct-acting oral anticoagulants in pa-
2013;58:1776–1780. tients with splanchnic vein thrombosis and cirrhosis. Liver Int
[83] Chen H, Liu L, Qi X, He C, Wu F, Fan D, et al. Efficacy and safety of 2017;37:694–699.
anticoagulation in more advanced portal vein thrombosis in patients [105] Yang H, Kim SR, Song MJ. Recurrent acute portal vein thrombosis in
with liver cirrhosis. Eur J Gastroenterol Hepatol 2016;28:82–89. liver cirrhosis treated by rivaroxaban. Clin Mol Hepatol 2016;22:499–
[84] Scheiner B, Stammet PR, Pokorny S, Bucsics T, Schwabl P, Brichta A, et al. 502.
Anticoagulation in non-malignant portal vein thrombosis is safe and [106] Intagliata NM, Henry ZH, Maitland H, Shah NL, Argo CK, Northup PG,
improves hepatic function. Wien Klin Wochenschr 2018;130:446–455. et al. Direct oral anticoagulants in cirrhosis patients pose similar risks of
[85] Kwon J, Koh Y, Yu SJ, Yoon JH. Low-molecular-weight heparin treatment bleeding when compared to traditional anticoagulation. Dig Dis Sci
for portal vein thrombosis in liver cirrhosis: efficacy and the risk of 2016;61:1721–1727.
hemorrhagic complications. Thromb Res 2018;163:71–76. [107] Nagaoki Y, Aikata H, Daijyo K, Teraoka Y, Shinohara F, Nakamura Y, et al.
[86] Zhang ZH, Zhang JW, He P, Zhou Y, Sun CY. Fondaparinux is effective for Efficacy and safety of edoxaban for treatment of portal vein thrombosis
acute portal vein thrombosis in decompensated cirrhotic patients. following danaparoid sodium in patients with liver cirrhosis. Hepatol Res
Medicine (Baltimore) 2017;96:e8256. 2018;48:51–58.
[87] Tripodi A, Primignani M, Braham S, Chantarangkul V, Clerici M, Moia M, [108] Leader A, Ten Cate V, Ten Cate-Hoek AJ, Beckers EAM, Spectre G,
et al. Coagulation parameters in patients with cirrhosis and portal vein Giaccherini C, et al. Anticoagulation in thrombocytopenic patients with
thrombosis treated sequentially with low molecular weight heparin and hematological malignancy: a multinational clinical vignette-based
vitamin K antagonists. Dig Liver Dis 2016;48:1208–1213. experiment. Eur J Intern Med 2020.
[88] Chung JW, Kim GH, Lee JH, Ok KS, Jang ES, Jeong SH, et al. Safety, efficacy, [109] Lee SR, Lee HJ, Choi EK, Han KD, Jung JH, Cha MJ, et al. Direct oral an-
and response predictors of anticoagulation for the treatment of ticoagulants in patients with atrial fibrillation and liver disease. J Am Coll
nonmalignant portal-vein thrombosis in patients with cirrhosis: a pro- Cardiol 2019;73:3295–3308.
pensity score matching analysis. Clin Mol Hepatol 2014;20:384–391. [110] Sasso R, Rockey DC. Anticoagulation therapy in patients with liver
[89] Hayashi T, Takatori H, Horii R, Nio K, Terashima T, Iida N, et al. Dana- cirrhosis is associated with an increased risk of variceal hemorrhage. Am
paroid sodium-based anticoagulation therapy for portal vein thrombosis J Med 2019;132:758–766.
in cirrhosis patients. BMC Gastroenterol 2019;19:217. [111] de Franchis R. Revising consensus in portal hypertension: report of the
[90] Naeshiro N, Aikata H, Hyogo H, Kan H, Fujino H, Kobayashi T, et al. Ef- Baveno V consensus workshop on methodology of diagnosis and therapy
ficacy and safety of the anticoagulant drug, danaparoid sodium, in the in portal hypertension. J Hepatol 2010;53:762–768.
treatment of portal vein thrombosis in patients with liver cirrhosis. [112] Garcia-Tsao G, Abraldes JG, Berzigotti A, Bosch J. Portal hypertensive
Hepatol Res 2015;45:656–662. bleeding in cirrhosis: risk stratification, diagnosis, and management:
[91] Bergère M, Erard-Poinsot D, Boillot O, Valette PJ, Guillaud O, Chambon- 2016 practice guidance by the American Association for the study of liver
Augoyard C, et al. Portal vein thrombosis and liver cirrhosis: long-term diseases. Hepatology 2017;65:310–335.

452 Journal of Hepatology 2021 vol. 75 j 442–453

[113] Bianchini M, Cavani G, Bonaccorso A, Turco L, Vizzutti F, Sartini A, et al. portosystemic shunt on portal vein thrombosis in patients with cirrhosis.
Low molecular weight heparin does not increase bleeding and mortality Gut 2011;60:846–852.
post-endoscopic variceal band ligation in cirrhotic patients. Liver Int [122] Han G, Qi X, He C, Yin Z, Wang J, Xia J, et al. Transjugular intrahepatic
2018;38:1253–1262. portosystemic shunt for portal vein thrombosis with symptomatic portal
[114] Ponthus S, Spahr L, Casini A, Berney T, Frossard JL, Majno P, et al. Safety of hypertension in liver cirrhosis. J Hepatol 2011;54:78–88.
variceal band ligation in patients with cirrhosis and portal vein throm- [123] Miraglia R, Maruzzelli L, Cannataci C, Gerasia R, Mamone G, Cortis K,
bosis treated with anticoagulant therapy: a retrospective study. Eur J et al. Radiation exposure during transjugular intrahepatic portosystemic
Gastroenterol Hepatol 2020;32:395–400. shunt creation in patients with complete portal vein thrombosis or
[115] Guillaume M, Christol C, Plessier A, Corbic M, Peron JM, Sommet A, et al. portal cavernoma. Radiol Med 2020;125:609–617.
Bleeding risk of variceal band ligation in extrahepatic portal vein [124] Chamarthy MR, Anderson ME, Pillai AK, Kalva SP. Thrombolysis and
obstruction is not increased by oral anticoagulation. Eur J Gastroenterol transjugular intrahepatic portosystemic shunt creation for acute
Hepatol 2018;30:563–568. and subacute portal vein thrombosis. Tech Vasc Interv Radiol
[116] Radosevich PM, Ring EJ, LaBerge JM, Peltzer MY, Haskal ZJ, Doherty MM, 2016;19:42–51.
et al. Transjugular intrahepatic portosystemic shunts in patients with [125] Lv Y, Qi X, He C, Wang Z, Yin Z, Niu J, et al. Covered TIPS versus endo-
portal vein occlusion. Radiology 1993;186:523–527. scopic band ligation plus propranolol for the prevention of variceal
[117] Walser EM, NcNees SW, DeLa Pena O, Crow WN, Morgan RA, Soloway R, rebleeding in cirrhotic patients with portal vein thrombosis: a rando-
et al. Portal venous thrombosis: percutaneous therapy and outcome. mised controlled trial. Gut 2018;67:2156–2168.
J Vasc Interv Radiol 1998;9:119–127. [126] Luo X, Wang Z, Tsauo J, Zhou B, Zhang H, Li X. Advanced
[118] Stein M, Link DP. Symptomatic spleno-mesenteric-portal venous cirrhosis combined with portal vein thrombosis: a randomized trial
thrombosis: recanalization and reconstruction with endovascular stents. of TIPS versus endoscopic band ligation plus propranolol for the pre-
J Vasc Interv Radiol 1999;10:363–371. vention of recurrent esophageal variceal bleeding. Radiology
[119] Rodrigues SG, Sixt S, Abraldes JG, De Gottardi A, Klinger C, Bosch J, et al. 2015;276:286–293.
Systematic review with meta-analysis: portal vein recanalisation and [127] Fagiuoli S, Bruno R, Debernardi Venon W, Schepis F, Vizzutti F, Toniutto P,
transjugular intrahepatic portosystemic shunt for portal vein throm- et al. Consensus conference on TIPS management: techniques, in-
bosis. Aliment Pharmacol Ther 2019;49:20–30. dications, contraindications. Dig Liver Dis 2017;49:121–137.
[120] Tripathi D, Stanley AJ, Hayes PC, Travis S, Armstrong MJ, Tsochatzis EA, [128] Thornburg B, Desai K, Hickey R, Hohlastos E, Kulik L, Ganger D, et al.
et al. Transjugular intrahepatic portosystemic stent-shunt in the man- Pretransplantation portal vein recanalization and transjugular intra-
agement of portal hypertension. Gut 2020;69:1173–1192. hepatic portosystemic shunt creation for chronic portal vein thrombosis:
[121] Luca A, Miraglia R, Caruso S, Milazzo M, Sapere C, Maruzzelli L, et al. final analysis of a 61-patient cohort. J Vasc Interv Radiol 2017;28. 1714-
Short- and long-term effects of the transjugular intrahepatic 1721.e1712.

Journal of Hepatology 2021 vol. 75 j 442–453 453