Introduction
Portal vein thrombosis (PVT) is defined as clot formation within the portal venous
system that leads to complete or incomplete portal venous obstruction. The underlying
mechanism contributing to the development of PVT is a hypercoagulable state, most
commonly found in patients with underlying liver cirrhosis, malignancy, acquired prothrombotic
disease, or an inflammatory condition. The estimated overall prevalence of PVT is
low (approximately 1 %), but research has shown that patients with cirrhosis or underlying
malignancy are at increased risk, with some studies demonstrating a prevalence of
PVT in this population ranging from 10 % to 26 % [1]. However, because PVT can remain asymptomatic in patients with chronic, gradual
thrombus development or limited thrombus extension, epidemiological studies tend to
underestimate the true incidence.
Treatment of PVT often begins with anticoagulation and escalates to endovascular interventions
if anticoagulation alone fails to reduce thrombus burden. Because PVT can be subdivided
into acute and chronic presentations, as well as cirrhotic and noncirrhotic etiologies,
evidence supporting the use of anticoagulation and endovascular approaches is not
standardized. This heterogeneity within the literature provides limited benefit to
interventionalists encountering this disease, and treatment approaches therefore often
have to be tailored to the individual. Because of this limitation, the present article
does not attempt to present strict guidelines, but rather uses a combination of institutional
experience and existing literature to highlight risks and benefits of various treatment
approaches.
Pathophysiology and Natural History
The pathophysiology of PVT is related to the disturbance of Virchow’s triad, in which
increased venous stasis, endothelial damage, and hypercoagulable states predispose
patients to thrombus formation. The development of PVT can be further divided into
acute and chronic PVT and into cirrhotic and noncirrhotic etiologies.
Cirrhotic and Noncirrhotic Cases
Cirrhotic patients are at an increased risk for PVT due to the presence of coagulopathy,
endothelial dysfunction, and venous stasis from portal hypertension. In liver disease,
the decreased production of factor VIII, Von Willebrand factor, and antithrombin initiates
platelet aggregation and thrombus formation. Hepatic endothelial dysfunction is the
sequela of increased oxidative stress and decreased nitric oxide bioavailability due
to increased free radical injury and disruption of vascular homeostasis respectively
[2]. The pathophysiology of underlying noncirrhotic causes of PVT is similar to that
of underlying cirrhotic causes, namely coagulopathy and disturbance of endothelial
integrity.
Inherited systemic prothrombotic disorders predispose patients to PVT in the absence
of underlying liver disease. For instance, JAK2 mutations are associated with myeloproliferative
disorders affecting the production of red blood cells or platelets and are implicated
in abnormal clot formation. Three recent European cohort studies testing approximately
432 cases of noncirrhotic PVT demonstrated that 16 % of cases had underlying JAK2
mutations and 20 % had some inherited prothrombotic mutations [3]. The inherited disorders included factor V Leiden mutations, protein C or S deficiency,
and antithrombin deficiencies, which contribute to PVT development through the derangement
of mechanisms related to clot breakdown.
Extrinsic factors increasing the risk of noncirrhotic PVT include intra-abdominal
inflammation or infection. The pro-inflammatory state generated by immune cell response
to pathogens underpins the prothrombotic state in infection. Neutrophils in particular
have recently been implicated in this process with the discovery of their role in
intrinsic pathway activation and tissue factor delivery by neutrophil extracellular
traps. Protease-activated receptors found on both leukocyte and platelet membranes
upregulate the expression of endothelial tissue factor, which is involved in clot
formation, further highlighting the ongoing communication between the endothelium
and inflammatory response [4].
Natural History
The most common sequela of acute PVT is spontaneous recanalization, which occurs in
45 % to 70 % of patients who do not receive treatment [5]. The absence of spontaneous recanalization is not clearly associated with poor clinical
outcomes or increased short-term mortality, likely because these outcomes are primarily
influenced by the underlying disease process itself and not directly by PVT [5]
[6]
[7]
[8]. The natural history of acute PVT also depends on the extent of initial thrombus
burden, with some severe cases presenting with splanchnic congestion and bowel necrosis.
The true incidence of mesenteric ischemia in cases of acute PVT is unclear, although
30-day mortality rates associated with bowel ischemia in mesenteric venous thrombosis
have been reported to approach 32 % [9]. Lastly, unrecognized or persistent acute PVT may progress to chronic PVT, which
is characterized by cavernous transformation of the portal vein and symptoms of worsening
portal hypertension such as ascites, variceal bleeding, or encephalopathy [8]. Treatment and outcome considerations for acute versus chronic PVT in cirrhotic
versus noncirrhotic patients will be elucidated in the following sections.
Diagnosis
The European Association for the Study of the Liver (EASL) guideline states that Doppler
imaging is the first-line diagnostic tool in the context of abdominal pain with a
suspicion for portal vein pathology. Subsequently, the diagnosis and extent of the
clot burden need to be confirmed with contrast-enhanced computed tomography (CT)/magnetic
resonance imaging (MRI) [10]. However, ultrasound imaging is operator-dependent. Therefore, the possibility of
false-negative reports remains a significant limitation, particularly as the thrombus
may not be well-appreciated on grayscale imaging alone. Conjunctive use of Doppler
imaging is therefore of utmost importance, as this modality may demonstrate greatly
reduced velocities with aphasic waveforms. Such findings are suggestive of portal
hypertension and PVT in the setting of intraluminal filling defects.
Cross-sectional imaging techniques such as CT and MRI are also helpful in evaluating
PVT, especially for cases in which extension of the thrombus into the superior mesenteric
vein or splenic vein is suspected. A new thrombus may appear with increased attenuation
on CT and will lack enhancement in the presence of intravenous contrast unless tumor
thrombus is present. In cases of chronic PVT, the presence of linear calcification
patterns, collaterals, and cavernous transformation can be seen on both CT and MRI
[11]. Thus, although liver Doppler ultrasound imaging is the preferred screening modality
for PVT, contrast-enhanced cross-sectional imaging allows for more extensive evaluation
of collateral vasculature, pathology contributing to the development of PVT, and procedural
planning [11].
Cases of acute noncirrhotic and acute-on-chronic cirrhotic PVT also require further
workup of local and systemic factors, such as cancer progression, extrinsic compression,
intra-abdominal infection, thrombophilia, and systemic hematologic disorders.
Anticoagulation
For initial treatment, the EASL guidelines recommend anticoagulation as the first-line
therapy for cirrhotic PVT and acute noncirrhotic, non-malignant PVT [10]. Similarly, the American Association for the Study of Liver Diseases recommends
anticoagulation therapy for recently occlusive or partially occlusive cirrhotic PVT
and acute noncirrhotic PVT without contraindications to anticoagulation, such as central
nervous tumor, gastrointestinal bleeding, and recent stroke [12]. However, the evidence cited within these guidelines to support the use of anticoagulant
therapy remains limited. One of the larger prospective studies evaluating the effect
of anticoagulation on PVT outcomes in 102 patients reported a 38 % recanalization
rate for patients receiving early systemic anticoagulation, whereas those who received
late or no anticoagulation had decreased rates of recanalization. Within this study,
however, only 2 patients did not receive anticoagulant therapy compared with 95 who
did [13].
Cirrhotic Patients
To date, three meta-analyses have compared anticoagulation with low-molecular-weight
heparin, vitamin K antagonists, and direct oral anticoagulation for PVT treatment
in acute or chronic cirrhotic PVT patients ([Table 1]) [15]
[18]
[19]. These studies consistently demonstrated higher rates of recanalization in treated
patients (42–77 %) than in nontreated patients (26–33 %) [15]
[18]
[19]. However, the analyses by Qi et al. [15] and Mohan et al. [19] demonstrated significant heterogeneity between the included studies. The reported
response rates also did not consistently distinguish between partial and complete
recanalization, which limits comparison. Because all three meta-analyses were performed
in the cirrhotic patient population, extrapolation of these results to noncirrhotic
PVT patients remains to be determined.
Table 1
Comparison of recanalization rates in cirrhotic patients with various anticoagulation
agents based on 3 meta-analyses [15]
[18]
[19].
|
treatment
|
study
|
recanalization rate
|
|
combined[*]
|
Qi et al. [15]
Loffredo et al. [18]
Mohan et al. [19]
|
41.5 %
53 %
66.7 %
|
|
low-molecular-weight heparin
|
Mohan et al. [19]
|
60.7 %
|
|
Vitamin K antagonists
|
Mohan et al. [19]
|
66 %
|
|
direct oral anticoagulation
|
Mohan et al. [19]
|
76.7 %
|
|
no treatment
|
Mohan et al. [19]
Loffredo et al. [18]
|
26 %
33 %
|
* Multiple anticoagulation agents used in treatment.
Noncirrhotic Patients
Similar data for acute noncirrhotic PVT patients are limited, with only one retrospective
study showing higher rates of recanalization and lower rates of bleeding events with
direct oral anticoagulation than with warfarin [20]. Anticoagulation use does appear to improve overall recanalization rates when compared
with nonuse of anticoagulation. However, the translation of such findings to patient
outcomes remains elusive. Further studies assessing the effects of these anticoagulants
on mortality or recurrence rates are warranted.
The risk of adverse bleeding events, particularly in patients with underlying coagulopathy,
is also a topic of contention, particularly when treating patients with liver disease
who are already prone to bleeding [7]
[17]. Limited evidence suggests that any increase in bleeding risk is often clinically
insignificant and nonvariceal in nature [21]
[22]
[23]
[24]. In cirrhotic patients with PVT, concomitant anticoagulation use does not appear
to increase the risk of complications. One purported hypothesis for this is that the
gradual resolution of the thrombus during anticoagulation treatment reduces portal
pressure and therefore variceal bleeding risk [21].
Anticoagulation is considered to have failed when patients experience persistent thrombosis
refractory to treatment after 6 months or mesenteric ischemia requiring escalation
of care. Furthermore, even with proper anticoagulation, patients with documented clot
resolution may still develop portal hypertension [25]. In such cases, more advanced interventions involving endovascular procedures can
be considered.
Endovascular Treatment
There is no established role for endovascular interventions in the management of PVT,
and the benefits of endovascular therapy are still under investigation. In our experience,
there are a few specific clinical scenarios in which endovascular therapy can play
a role.
Cirrhotic Patients
In cirrhotic patients with contraindication to or inadequate response to anticoagulation,
the benefits of endovascular therapy are two-fold: the therapy can salvage liver transplant
candidacy and treat symptomatic portal hypertension, which may include variceal bleeding,
refractory ascites, hepatic hydrothorax, among other symptoms. Per the EASL guidelines,
it is a level B2 recommendation that liver transplant patients should be referred
for TIPS placement if patients present with progressive PVT on anticoagulation ([Fig. 1], [2]) [10]. In the subset of acute cirrhotic PVT patients, endovascular therapy should be considered
to prevent intestinal ischemia and acute symptomatic portal hypertension, including
recurrent or impending variceal bleeding. The risks of impending bowel ischemia can
also be estimated with laboratory markers (inflammatory markers), radiologic assessment
(bowel wall thickening), and clinical assessment (worsening abdominal pain). It is
important to note that the incidence of acute PVT is much lower in the cirrhotic patients
versus noncirrhotic patients, due to the presence of collateral circulation [12].
Fig. 1 A 64-year-old man with nonalcoholic steatohepatic liver cirrhosis presented with
episodes of upper gastrointestinal bleeding. During workup for a liver transplant,
the patient was diagnosed with portal vein thrombosis with cavernous transformation
of the main portal vein. Thrombolysis was indicated to preserve liver transplant candidacy.
(A–C) Coronal CT images of the liver demonstrate chronic occlusion of the portal vein
with periportal collaterals as well as a large spleno-gastrorenal shunt (A: arterial phase; B and C: portal venous phase).
Fig. 2 Images from the same patient described in [Fig. 1]. A Splenic venogram demonstrates complete occlusion of the portal vein with large left
gastric vein leading to gastric varices. B, C Successful recanalization of the chronically occluded portal vein with placement
of a snare in the intrahepatic portal vein as the target during puncture from the
hepatic venous site. D Image obtained after TIPS placement and balloon venoplasty of the main portal vein.
E At the 6-month follow-up, portal venography reveals patency of the main portal vein
with resolved left gastric varices.
Noncirrhotic Patients
For noncirrhotic, nonmalignant patients with acute thrombus formation, the goal of
endovascular therapy is to prevent thrombus extension and its acute and chronic sequelae,
including the development of portal hypertension and intestinal ischemia. Therefore,
indications for endovascular therapy in noncirrhotic patients may include variceal
bleeding, refractory ascites, hepatic hydrothorax, among other symptoms ([Fig. 3], [4], [5], [6]).
Fig. 3 A 58-year-old man with alcoholic liver cirrhosis presented to the emergency department
with nonspecific right upper and lower quadrant abdominal pain. A, B Axial and coronal CT scans of the liver in the portal venous phase demonstrate a
nonocclusive thrombus in the main portal vein. C Coronal CT scan of the abdomen shows a partially occlusive thrombus in the superior
mesenteric vein with thickening of the terminal ileal loops, consistent with acute
mesenteric ischemia.
Fig. 4 Images from the same patient described in [Fig. 3]. A, B Splenoportal and portomesenteric venograms via the transjugular approach demonstrate
a partially occlusive thrombus in the main portal and superior mesenteric veins and
a completely occlusive thrombus in the right portal vein. C After mechanical thrombectomy and TIPS stent placement, complete resolution of the
main portal vein thrombus can be seen. D At the 6-month follow-up, color Doppler ultrasound demonstrates a patent main portal
vein without evidence of thrombus.
Fig. 5 A 65-year-old man with nonalcoholic steatohepatic liver cirrhosis and chronic portal
vein thrombosis with cavernous transformation of the portal vein presented to the
emergency department with acute upper gastrointestinal bleeding from gastroesophageal
varices. A, B Portal venous phase CT scans of the abdomen in the coronal and axial planes show
a chronically occluded portal vein with cavernous transformation.
Fig. 6 Images from the same patient described in Fig. 5. A Image demonstrates near-complete occlusion of the portal vein with a trace amount
of contrast passing through. A large left gastric vein leads to the gastroesophageal
variceal complex. B Image demonstrates successful recanalization of the main portal vein and placement
of a snare as the target in the right portal vein for a TIPS procedure. C Portal venogram after balloon venoplasty of the main portal vein demonstrates hepatopetal
flow. D Image demonstrates a patent main portal vein with brisk flow after TIPS placement.
It is important to note that endovascular therapy is associated with risks, and direct
comparison studies between endovascular and medical therapies are scarce. Recent studies
performed in small retrospective cohorts showed that despite successful recanalization,
outcomes may be affected by posttreatment complications such as re-thrombosis, symptom
recurrence, or bleeding [26]
[27]. Rossle et al. [28] found that complete response rates on imaging were higher for patients treated with
both endovascular therapy and anticoagulation than for those treated with anticoagulation
alone. Based on these data, there is a role for combination therapy involving both
anticoagulation and endovascular approaches.
Endovascular Access Approach
When considering endovascular intervention techniques, interventionalists must first
decide on the optimal approach to establish access to the portal venous system. This
can be performed directly using a transjugular (TIPS) approach or a percutaneous transhepatic
approach or indirectly using a transsplenic or transmesenteric route.
The transjugular intrahepatic approach is a popular method to establish portal vein
access and may be performed under fluoroscopy with intravascular ultrasound (IVUS)
or intracardiac echography (ICE) guidance. In our institution, IVUS is mainly used
to assess for thrombus burden in the portal system and ICE is used to guide access
into the portal branch. A recent study has shown that IVUS and ICE guidance could
decrease the rate of complications [29]. After the thrombus is reached via transjugular retrograde access with placement
of a TIPS stent, thrombolysis or thrombectomy can be performed. The advantage of this
approach is that it provides the opportunity to perform complementary procedures such
as variceal embolization and portal system decompression. The percutaneous approach
may be necessary, however, when the transjugular approach is too difficult due to
extensive thrombotic involvement of the portal venous system. Percutaneous access
has also been reported to be less technically challenging [30]
[31]
[32]. However, a disadvantage of the percutaneous approach involves the size of the sheath
that can be safely placed, which is typically approximately 8 French based on our
institutional experience. This sheath size limits possibilities for endovascular interventions,
as larger percutaneous sheaths traversing the liver capsule and parenchyma are associated
with an increased bleeding risk.
Other percutaneous approaches include the transsplenic and transmesenteric techniques.
These alternative routes may be of benefit when complete occlusion of the portal vein
precludes retrograde recanalization using the transhepatic or transjugular approach
alone ([Fig. 3], [4]). The transsplenic approach, which may be used to achieve antegrade access to the
portal vein, has demonstrated high technical success rates, with some authors arguing
it is safer to perform than a transjugular intrahepatic approach [33]
[34]. Additionally, one retrospective review comparing outcomes between transsplenic
and transhepatic access in 148 patients found no significant differences between groups
in procedural success rates or bleeding complications even though the spleen is highly
vascularized [35]. Transmesenteric access to the portal vein has also been described. With this technique,
a mini-laparotomy must first be performed to establish access to the mesenteric venous
vasculature. This procedure is infrequently performed because of the higher risk of
complications with this technique versus alternative approaches [32]
[36]
[37].
Transjugular Portosystemic Shunt Creation (TIPS Procedure)
TIPS placement remains an important adjunct therapy. It allows for management of symptomatic
portal hypertension or salvage of liver transplant candidacy. Additionally, once the
TIPS has been placed, repeated thrombectomy procedures can be performed through the
TIPS approach with relative ease of access into the portal venous system.
TIPS Placement in Cirrhotic Patients
In a study by Habib et al. [38], 11 patients with cirrhosis-induced chronic PVT underwent TIPS placement and portal
vein recanalization (PVR). All 11 patients had improved portal flow, ranging from
minor improvement to complete resolution (without further need of anticoagulation
therapy), and 3 patients went on to undergo successful liver transplant. Similarly,
in a study by Thornburg et al. [39], 24 of 61 patients with occlusive PVT went on to undergo successful transplant after
PVR plus TIPS placement. In a recent systematic meta-analysis of 13 studies, TIPS
placement was technically feasible in 95 % of patients, with a pooled recanalization
rate of 79 % and a TIPS patency rate of 80 % to 90 % at 12 months [40]. Similarly, Valentin et al. [41] reported a pooled success rate of 86 %. However, it is important to note that PVT
was not an indication for TIPS placement in the majority of the included cases. Therefore,
the success rate may be overestimated due to selection and reporting bias.
TIPS Placement in Noncirrhotic Patients
The use of TIPS placement in patients with PVT but without liver cirrhosis remains
controversial. It is mainly reserved for patients with impending bowel ischemia and
symptoms of portal hypertension or refractory thrombosis despite multiple interventions.
Some studies have reported favorable outcomes in a small number of acute or acute-on-chronic
noncirrhotic PVT patients requiring urgent intervention beyond systemic anticoagulation
[42‒45]. In a recent study by Sun et al. [44], technical success was achieved in 20 out of 22 chronic noncirrhotic PVT patients.
Notably, technical success was negatively affected by thrombus burden and presence
of cavernous transformation of the portal vein. Similarly, in a study by Klinger et
al. [45], 17 patients with chronic noncirrhotic, nonmalignant chronic PVT underwent PVR plus
TIPS placement, and 76.5 % of these patients had successful portal venous recanalization.
At 1 and 2 years, the portal venous and TIPS patency rates were both 69.5 %.
Limitations to the TIPS approach include the risk of worsening postprocedural hepatic
encephalopathy or right heart failure and technical failure when a completely occluding
thrombus is unable to be traversed and catheterized via the transjugular access [14]
[40]. Recently, a case series of 61 patients reported the novel use of a transsplenic
approach to TIPS placement, which was found to be less technically challenging [33]. Based on our institutional experience, we use the transsplenic approach for chronic
PVT recanalization but prefer the TIPS or percutaneous transhepatic approach for acute
PVT cases.
Thrombolysis
Endovascular thrombolysis using multihole infusion catheters can be performed when
systemic anticoagulation alone is insufficient to restore portal venous flow. Similar
to systemic anticoagulation, thrombolysis is contraindicated in patients with brain
tumors, recent hemorrhagic stroke, and gastrointestinal bleeding [46]. In particular, when patients exhibit signs of irreversible gut ischemia or infarction
and peritonitis, thrombolysis should not be pursued. Instead, surgical resection of
the infarcted bowel is indicated in these clinical scenarios.
This procedure can be performed in two ways and is frequently performed in combination
with other endovascular therapies, such as TIPS placement and mechanical thrombectomy.
The transvenous method involves localized application of fibrinolytics, such as tissue
plasminogen activator or heparin, directly into the portal venous thrombus. Thrombolysis
can also be performed using ultrasound-accelerated infusion catheters such as the
EKOS Infusion Catheter System (Boston Scientific). Acoustic microstreaming from ultrasound
allows for transportation of the lytic agent directly to the site of the clot, with
ultrasound pulses additionally disrupting the fibrin integrity of the thrombus [30]. Secondly, transarterial thrombolysis can be performed indirectly by catheterizing
the superior mesenteric artery and infusing thrombolytic agents through the visceral
arteries. In general, sole thrombolysis is not commonly performed for the treatment
of PVT due to the low rates of complete resolution of thrombus. Thrombolysis typically
achieves partial resolution only [47], and therefore it is beneficial to combine this technique with mechanical thrombectomy.
Data are lacking regarding the role of thrombolysis therapy in this setting. A recent
meta-analysis by Cheng and Tree [48] demonstrated that the recanalization rate was 84 % after thrombolysis with and without
other endovascular therapies. The major complication rate was 7 % and the overall
complication rate was 25 %. Given the small size of the patient sample, no direct
comparison between the transvenous and transarterial methods was performed. Similar
to other meta-analyses, the study results were negatively affected by heterogeneities
in patient selection and treatments.
Mechanical Thrombectomy
Mechanical thrombectomy involves physical disruption of the thrombus and is often
used in conjunction with systemic anticoagulation and endovascular thrombolysis when
systemic anticoagulation is insufficient. One of the more commonly used devices in
the portal venous system is the AngioJet system (Boston Scientific) [26]. This device uses a retrograde saline jet flow to generate a low-pressure system
at the catheter tip, thereby creating a vacuum effect to aspirate the clot (Venturi-Bernoulli
effect). Saline jets subsequently macerate the clot while delivering low-dose thrombolytics.
Similar rheolytic devices include the Hydrolyser device (Cordis) and the Oasis thrombectomy
catheter (Boston Scientific). Other devices use rotational components to mechanically
fragment the clot without aspiration capabilities or concomitant delivery of local
lytics. For instance, the Trellis Peripheral Infusion System (Covidien) uses spinning
wires, whereas the Amplatz Thrombectomy Device (Microvena), Arrow-Trerotola PTD (Arrow
International), and Cragg Brush (Microtherapeutics) use spinning impellers or brushes.
The efficacy of mechanical thrombectomy alone or in combination with other techniques
has not been well-established in the literature. Despite high technical success rates
noted in several case series using mechanical thrombectomy, the rates of re-thrombosis
and the need for additional interventions remain high [49‒51]. This risk of re-thrombosis
is likely due to underlying prothrombotic disease processes that are not addressed
during catheter-directed therapy. Another hypothesis suggests that mechanical thrombectomy
increases the risk of wall denudation and endothelial damage, further contributing
to future thrombus formation [31]. Additionally, a limitation of mechanical thrombectomy is the risk of partial recanalization
when the extent of the thrombus involves smaller branched vessels not amenable to
access using the mechanical thrombectomy device.
Comparing safety profiles among thrombectomy devices is challenging, but two studies
using animal models have attempted such an evaluation. One study compared the Akonya
Eliminator Device and the Arrow-Trerotola percutaneous thrombolytic device in porcine
models and found the latter device produced greater arterial injury on histology [52]. A second study in canine models found the Casteneda brush caused less arterial
wall damage than the Arrow-Trerotola device and Fogarty embolectomy catheter, although
all three devices demonstrated vascular wall lesions extending into the tunica media
[53]. Additional benchtop trials to evaluate the performance of these devices in vivo
may help optimize future treatment planning. Our institutional experience with mechanical
thrombectomy in the portal venous system has shown effective recanalization of the
portal veins using the AngioJet system, which can be advanced over a 0.035-inch guidewire.
Aspiration Thrombectomy
Aspiration thrombectomy is performed using a suction catheter that engages and extracts
the portal vein thrombus. To date, no formal studies have reported outcomes of aspiration
thrombectomy alone for the treatment of PVT specifically. The existing literature
discusses institutional experiences using this approach in treating superior mesenteric
vein thrombosis, so any extrapolation of reported success rates to PVT treatment is
theoretical at this point [54]
[55]. However, the use of aspiration thrombectomy in conjunction with systemic anticoagulation
or other endovascular interventions for the treatment of PVT has been reported [30]
[56]. In addition to concerns regarding low recanalization rates with aspiration thrombectomy
alone, this procedure is associated with blood volume loss because aspiration catheters
are unable to replace the aspirated fluid once activated [57].
How and When to Use Endovascular Therapy
As mentioned previously, there is a paucity of high-quality data to guide a standardized
clinical approach. Existing data are often derived from case series and small trials.
Systemic reviews and meta-analyses are often plagued by patient and treatment heterogeneity.
It is therefore of utmost importance to manage PVT in a multidisciplinary manner.
It is prudent to offer endovascular therapy as a second-line or adjunct therapy to
anticoagulation. In cirrhotic patients with symptoms of portal hypertension in whom
anticoagulation therapy has failed (defined as persistence or worsening symptoms after
six months), TIPS placement combined with other endovascular techniques may serve
as valuable adjunct therapies in managing chronic PVT, especially in pretransplant
patients. In particular, cirrhotic patients should be considered for TIPS based on
the severity of the portal hypertensive symptoms. The exact time of intervention is
less clear in the literature but should be guided by local clinical expertise and
clinical presentation, such as recurrent bleeding or developing bowel ischemia.
In noncirrhotic patients, endovascular therapy using thrombolysis and thrombectomy
with or without TIPS placement may be a viable option to prevent acute and long-term
sequelae, namely bowel ischemia and portal hypertensive symptoms. In the literature,
impending bowel ischemia, manifested as either radiographic changes or new/worsening
abdominal pain, has been listed as an indication for endovascular therapy along with
acute variceal bleeding or impending bleed. The exact time point at which endovascular
therapy is indicated is less clear but should be guided by local clinical expertise
and clinical judgement to prevent impending bowel ischemia or bleed. However, the
long-term efficacy of this strategy has not been validated and needs to be weighed
against the increased risk of periprocedural bleeding. More studies are warranted
to compare endovascular therapy to anticoagulation alone in managing PVT.
Thus far there has been no direct comparison between the different endovascular therapies,
likely because of the low number of patients treated and the heterogeneity of underlying
disease. Additionally, endovascular therapies are frequently used in various combinations
(TIPS placement plus thrombectomy/thrombolysis) in clinical practice, which further
complicates direct comparisons. Lastly, there are too few published case series and
retrospective studies to provide meaningful comparisons. As the endovascular therapy
approach gains in popularity, large patient cohorts should be available for analysis
in the near future.
