Interventional recanalization therapy in patients with non-cirrhotic, non-malignant portal vein thrombosis: comparison between transjugular versus transhepatic access

Purpose To compare the safety and outcome of transjugular versus percutaneous technique in recanalization of non-cirrhotic, non-malignant portal vein thrombosis. Methods We present a retrospective bicentric analysis of 21 patients with non-cirrhotic, non-malignant PVT, who were treated between 2016 and 2021 by interventional recanalization via different access routes (percutaneous [PT] vs. transjugular in transhepatic portosystemic shunt [TIPS] technique). Complication rates with a focus on periprocedural bleeding and patency as well as outcome were compared. Results Of the 21 patients treated (median age 48 years, range of 19–78), seven (33%) patients had an underlying prothrombotic condition. While 14 (57%) patients were treated for acute PVT, seven (43%) patients had progressive thrombosis with known chronic PVT. Nine patients underwent initial recanalization via PT access and twelve via TIPS technique. There was no significant difference in complete technical success rate according to initial access route (55.5% in PT group vs. 83.3% in TIPS group, p = 0.331). However, creation of an actual TIPS was associated with higher technical success in restoring portal venous flow (86.6% vs. 33.3%, p = 0.030). 13 (61.9%) patients received thrombolysis. Nine (42.8%) patients experienced hemorrhagic complications. In a multivariate analysis, thrombolysis (p = 0.049) and PT access as the first procedure (p = 0.045) were significant risk factors for bleeding. Conclusion Invasive recanalization of the portal vein in patients with PVT and absence of cirrhosis and malignancy offers a good therapeutic option with high recanalization and patency rates. Bleeding complications result predominantly from a percutaneous access and high amounts of thrombolytics used; therefore, recanalization via TIPS technique should be favored.


Introduction
Portal vein thrombosis (PVT) is a rare condition with a reported prevalence of 3.7 per 100.000 population, and half of the cases occur in a patient without liver cirrhosis or malignancy [1]. In contrast to cases with malignancy or cirrhosis, were the outcome is primarily dependent on the underlying disease, non-cirrhotic and non-malignant PVT have a different pathogenesis and the PVT itself is decisive for the prognosis of the patient [2]. Similar to thrombosis in other vessels, non-cirrhotic and non-malignant PVT is related to one or more features of the Virchow's triad [3][4][5].
Although PVT without has better survival rates than PVT in patients with cirrhosis or malignancy [1], it is associated with a significant risk of venous congestion of the gut in the 1 3 acute period and complications related to portal hypertension in the long term, especially life-threatening variceal bleeding, requiring lifelong specialized care [6,7]. The rarity of this condition precludes large-scale controlled trials, and the treatment algorithm of non-cirrhotic and nonmalignant PVT has not been standardized. The American Association For Study of Liver Diseases currently (AASLD) recommends the consideration of interventional portal vein recanalization (PVR) in patients with acute PVT and impending intestinal ischemia [2]. In patients with chronic PVT and recurrent bleeding and/or refractory ascites not manageable medically or endoscopically, interventional PVR followed by TIPS is also currently recommended [2]. Anticoagulation for six months is recommended in patients with reversible etiologies of PVT, life-long in patients with thrombophilia [8].
Several case series have shown that interventional recanalization with thrombolysis via superior mesenteric artery, percutaneous transhepatic portal vein thrombolysis [9] or thrombectomy [10], and recanalization via TIPS access with additional TIPS placement [11,12] have high technical and clinical success rates, especially in patients with a deteriorating clinical condition and persistent symptoms of portal hypertension despite anticoagulation (impending bowel ischemia or infarction). Recently, a multicenter study has compared the outcome of medical and interventional therapies in non-cirrhotic and non-malignant PVT and showed significantly higher recanalization rates after interventional treatment (37% vs. 71%, p < 0.001), despite the higher thrombus burden in the interventional arm at baseline [13]. These results suggest a wider use of interventional procedures in patients with non-cirrhotic and non-malignant PVT. However, further investigation of the optimal approach and technique for interventional recanalization still needs to be clarified, as severe complications including periprocedural bleeding have been reported. The aim of this retrospective, bicentric cohort study was to evaluate the outcome of interventional therapies in patients with non-cirrhotic and nonmalignant PVT, and thereby compare the safety and efficacy of different portal vein access routes.

Study design
This observational study includes a retrospective bicentric analysis of the clinical course in a total of 21 patients with non-cirrhotic and non-malignant PVT, who were treated at two tertiary care university hospitals with experience of > 50 TIPS procedures annually from the year 2016 to 2021. Preliminary results of the cohort treated at Center 1 have already been published in a first case series [14]. The study was approved by the institutional review board (Protocol number 2016-046-f-S). Informed consent was waived due to the retrospective character of the study.

Diagnosis and definitions
PVT was confirmed by contrast-enhanced CT in all patients and was graded according to the previously published grading system [13]: grade 1: incomplete occlusion of the vessel lumen; grade 2: complete occlusion or extended thrombosis; grade 3: the presence of cavernous transformation. PVT was considered acute when symptoms developed less than 60 days before presentation, and there is no radiological or endoscopic evidence of collateral circulation [2,[15][16][17]. Clinical and laboratory investigations were completed to identify the etiology of PVT, including screening for prothrombotic disorders. All patients were discussed in multidisciplinary rounds, including gastroenterology, interventional radiology, and liver/transplant surgery, and treatment decision was made by consensus in every case.
All patients with grade 2 thrombosis were heparinized with the aim of partial thromboplastin time (PTT) of 60-80 s after the establishment of the diagnosis. Portal vein recanalization procedures, either percutaneous (PT) or via TIPS access, were done under general anesthesia. PT access was preferred for recanalization in the initial procedure, except for the cases with ascites. Transjugular access as the first choice instead was performed in case of technical or clinical failure of PT access or in the presence of ascite due to increased risk of abdominal bleeding.
Technical success was described as complete recanalization of the entire portal venous system or complete bypass of the thrombus via a TIPS. In patients with residual thrombus causing less than a 25% decrease in the lumen, the outcome was described as a partial technical success and the rest of the cases as technical failure.

Interventional technique and concomitant treatments
Portal vein interventions were performed by interventional radiologists with > 5 years of experience in portal venous interventions. Puncture of the thrombosed portal vein was performed under ultrasound guidance either percutaneously (PT) or via the transjugular route in TIPS technique [11]. In case of inadequate visualization of the portal vein in ultrasound to guide TIPS puncture (n = 1), a guidewire was advanced as a fluoroscopic target into the portal vein via percutaneous puncture of the splenic vein. In case of percutaneous access, a 9F sheath was introduced into the portal vein, in case of TIPS access a 10F sheath was inserted. Aspiration thrombectomy was done in all patients with a largebore aspiration catheter (CAT-8 Indigo Catheter, Penumbra Inc., Alameda, California, USA). In both centers, aspiration thrombectomy was performed as the initial step and proceeded to local thrombolysis in case of incomplete recanalization. Pulse spray thrombolysis was done using dedicated 4F spray lysis catheters with 10-20 cm side holes at a dose of 1 mg/h rt-PA. If overnight thrombolysis was performed, patients were monitored on intensive care units, and control angiography was done on the following day. Balloon angioplasty and additional stenting of the portal vein, superior mesenteric and splenic vein were performed when needed to establish or maintain portal venous flow.
This step was followed by the creation of TIPS in patients with transjugular access, and in patients with insufficient recanalization after percutaneous thrombus aspiration, local thrombolysis and repeated aspiration were performed. This step was combined with rheolytic thrombolysis using the AngioJet rheolytic thrombectomy system in the power pulse mode (Boston Scientific, Marlborough, MA, USA) in some patients. Thrombolysis was also performed in patients with an inadequate flow after TIPS creation. In patients with insufficient inflow due to thrombosis in SMV or SV, visceral stenting was performed. In case of failed complete recanalization in patients with initial percutaneous access, secondary TIPS was created (Fig. 1).
All patients received therapeutic anticoagulation with unfractionated heparin during treatment (target PTT 60-80 s). Following complete or partial recanalization heparinization was switched to oral anticoagulation with phenprocoumon for at least six months followed by 100 mg/ day aspirin monotherapy for lifetime. For patients with an underlying prothrombotic mutation, lifelong oral anticoagulation was initiated. In these cases, novel oral anticoagulants were applied instead of phenprocoumon.

Statistical analysis
Preprocedural patient characteristics and technical details were grouped as categorical or nominal variables. The total dose of lysis in patients with PT and TIPS access were compared using the non-parametric Mann-Whitney U test. Univariate analysis of the relationship between periprocedural characteristics and technical success and bleeding was performed. Factors with a p-value < 0.1 were included in multivariate analysis using binary logistic regression. A p-value < 0.05 was considered significant. Statistical analysis was performed using R statistical software (R version 3.6.3).

Study population
21 patients presenting with non-cirrhotic, non-malignant PVT were treated according to the interventional recanalization scheme described in Fig. 1. Baseline characteristics of the patients are summarized in Tables 1 and 2. The 21 Fig. 1 Algorithm of interventional therapy in patients with non-cirrhotic, non-malignant portal vein thrombosis patients included were a median of 48 years, ranging from 19 to 78 years at the time of treatment. An underlying prothrombotic condition was identified in seven patients (33%), of these three had a Janus kinase 2 (JAK2) mutation (14%). In total, 14 patients were treated for acute PVT (57%) with median symptom onset to first intervention of 13 days (range of 3-49). Seven patients had chronic PVT with presence of cavernous transformation of the portal vein and were treated electively due to progressive thrombosis and escalating symptoms related to the increase of portal venous pressure, refractory to medical therapy (grade 3). Clinical presentations consisted of one or a combination of symptoms including abdominal pain (n = 12), esophageal varices (n = 7) leading to hematemesis (n = 3), ascites (n = 9), and paralytic ileus (n = 1). In the n = 7 patients with accompanying ascites treated at Center 1, the ascites was compensated in five patients, allowing for a percutaneous approach. Most of the patients treated in Center 1 underwent portal vein recanalization via percutaneous access (75%). In Center 2, the transjugular recanalization route in TIPS technique was performed in 77% of the cases. Two patients with acute PVT and venous congestion had to undergo open surgery before initiation of interventional treatment, one patient with resection of a gangrenous ileum segment, the other patient underwent emergent splenectomy due to rupture of the splenic capsule [18].

Outcomes
Median time from symptom onset to first intervention in patients with acute grade 2 PVT was 13 days (range of 3-49 days). Nine (42.8%) patients had initial recanalization via PT access, and twelve (57.2%) patients via TIPS access. Procedural details of each patient are given in Table 3. Three patients with initial PT access underwent sequential recanalization via transjugular access with secondary TIPS placement due to initially insufficient recanalization via the percutaneous access (defined as > 25% residual thrombosis of the portal vein or intrahepatic branches) ( Table 4). The complete technical success rate was 55.5% in patients with initial PT access and 83.3% in patients with initial TIPS access (p = 0.331). However, the creation of a TIPS was significantly associated with higher technical success (86.6% vs. 33.3%, p = 0.030). None of the other procedure-related parameters correlated with technical success (Table 5). One patient in the percutaneous group had partial recanalization only. Due to poorer flow restoration via the percutaneous access, higher amounts of thrombolytics [recombinant tissue-Plasminogen Activator (rt-PA)] were required to restore flow via percutaneous access as the first procedure (mean 77.3 mg vs. 25.6 mg, p = 0.037) (Fig. 2).
Nine (42.8%) patients experienced hemorrhagic complications. Of these, five patients had bleeding from the liver, one from spleen due to additional splenic access, one from the stump of the splenic artery, and the other one had disseminated abdominal bleeding without localized arterial extravasation. In univariate analysis, thrombolysis was significantly associated with bleeding (88.8% vs. 33.3%, p = 0.024). Also, the PT group had higher bleeding than the TIPS group (66.6% vs. 25%, p = 0.087). Multivariate analysis showed both thrombolysis (p = 0.049) and PT access (p = 0.045) were independent significant risk factors of bleeding (Table 6). Three patients in the PT group and one patient in TIPS group underwent further surgery with packing and evacuation of hematoma (p = 0.272). No significant difference was observed in technical success (p = 0.597) or bleeding complications (p = 0.338) between acute grade 2 and chronic grade 3 PVT with presence of cavernous transformation (Table 5).
One patient from Center 1 with failed recanalization had bowel perforation and underwent resection of 20 cm of ileum. Three patients (14.2%) died during hospitalization. One patient died due to multiorgan failure and sepsis despite complete recanalization of PVT. One patient was lost due to intracerebral hemorrhage and hemorrhagic shock after   Yes hepatic bleeding. The other patient died after failed recanalization attempts due to sepsis.

Patency
Primary patency (complete or partial) of the portal vein was reached in 16 patients (76.2%). The median follow-up was 365 days (range of 3-922 days). Two patients showed recurrence of PVT approximately after 30 weeks (one from each group). The patient from the TIPS group with recurrent PVT displayed minor thrombotic deposits after TIPS and stenting on invasive angiography and was treated with balloon angioplasty and local thrombolysis without any additional complications. All patients with a failed recanalization developed a chronic PVT with a subsequent cavernous transformation. None of the treated patients suffered from major long-term complications during the follow-up period; only one patient revealed mild gastroesophageal varices (grade I) on endoscopy. Follow-up laboratory values, clinical assessment, and ultrasound revealed no signs of cirrhosis formation.

Discussion
Non-malignant, non-cirrhotic portal vein thrombosis is a rare but potentially fatal condition requiring rapid and efficient treatment. Although systemic anticoagulation is recommended throughout different therapy guidelines as the method of choice in treating portal vein thrombosis [8,19], complete recanalization is only achieved in around 35% of the cases [20,21]. Recent case series have shown that interventional therapies, such as thrombectomy, thrombolysis or TIPS, are effective in patients with acute [22,23] or chronic PVT [24][25][26][27][28][29]. A systematic review of studies using TIPS in portal vein thrombosis showed a high rate of recanalization and long-term patency [30]. A recent multicentric study from Rössle et al. [13] comparing medical and interventional treatment in 65 patients showed low procedure-related mortality (2.9%) and a high success rate of the interventional treatment (17% vs. 54%). However, several techniques have been utilized in these studies, and the optimal interventional method needs to be defined. Our results have shown that the creation of TIPS resulted in significantly higher technical success, and percutaneous approach and thrombolysis are significantly associated with increased bleeding. Non-malignant, non-cirrhotic portal vein thrombosis constitute a special entity, because in patients with underlying cirrhosis or malignancy, portal vein thrombosis usually develops in the long term, which allows more time to develop collaterals. Additionally, the treatment of the underlying disease is the mainstay in management of these patients, and thrombosis-directed therapy is mostly restricted to systemic heparin. It is also important to view cirrhosis of the liver as an independent risk factor for thrombosis, mainly because loss of liver function impacts both procoagulant and anticoagulant factors; therefore, a differential approach of therapy in the presence of cirrhosis is required [8].
To avoid acute and long-term complications, patients who do not benefit adequately from initial anticoagulation need to be offered more aggressive therapy options. In this study we describe an escalating invasive therapy regime, starting with thrombaspiration, followed by local thrombolysis (bolus and overnight lysis), rheolysis, balloon angioplasty, stenting, and possibly TIPS implantation, if the latter methods did not lead to adequate flow restoration. In the treatment of noncirrhotic, non-malignant PVT, multiple studies with small cohorts have shown higher success rates for combined surgical/interventional and isolated interventional therapy than isolated anticoagulation therapy with recanalization rates ranging from 75% to 80%, respectively [11,14,31]. However, invasive recanalization is associated with significantly higher rates of major hemorrhagic complications, especially in patients who underwent invasive recanalization through a percutaneous transhepatic access [14]. In our case series, we compared two different access routes (PT vs. TIPS access) for PVT recanalization. In the past percutaneous access was predominantly used especially in case of one branch of the portal vein still being patent, facilitating the percutaneous puncture and allowing some degree of flow, whereas in extensive thrombosis of the entire portal venous system, TIPS is needed to achieve adequate flow restoration in order to maintain patency.
These recanalization routes offer different advantages and disadvantages. While recanalization of the portal vein through a percutaneous access allows the preservation of the endothelial wall and blood flow through the native portal vein and intrahepatic branches, recanalization via TIPS access requires the creation of an artificial connection. In patients with occluded intrahepatic branches, PT recanalization is more challenging, especially due to limited flow. In contrary, TIPS technique allows access to the portomesenteric system despite extension of thrombus in the intrahepatic branches, as seen in many PVT cases. Additionally, TIPS allows an easy access for further interventional procedures of the portal vein.
In this study, interventional recanalization of the portal vein through a transjugular approach showed significantly lower rates of major hemorrhagic complications. In this group, a significantly lower amount of thrombolysis was required to restore flow, while other procedural details such as use of additional stenting or rheolytic therapy were not associated with outcome. Percutaneous access requiring perforation of the liver capsule together with the increased amount of thrombolytics led to increased rates of major hemorrhagic complications. Previous studies of patients with impaired coagulation needing liver biopsy also showed 1 3 higher complication rates after percutaneous access compared to the transjugular route [32]. Furthermore, probably due to stable outflow after TIPS, patients with TIPS had significantly higher clinical success.
Limitations of this analysis include the retrospective character of the study and a certain selection and procedural bias between the two centers. The yet limited follow-up time may underrepresent long-term benefits of PVT recanalization. Heterogeneity of onset of PVT poses another limitation, since acute on chronic PVT offers different challenges in management, yet with the focus of this study on the route to restore blood flow, the inclusion of both groups seems feasible. Additionally, low enrollment rates mainly due to interventional therapy not being the first-line therapy method in patients with non-cirrhotic, non-malignant PVT contribute to smaller study populations. With current guidelines increasingly recommending escalating recanalization therapy in patients affected by acute or chronic PVT and impending serious complications that are not manageable conservatively [2], further research and a greater sample size may be desirable.

Conclusion
Invasive recanalization of the portal vein in carefully selected patients with PVT and absence of cirrhosis and malignancy offers a good therapeutic option with high recanalization and patency rates, compared to the moderate   Funding Open Access funding enabled and organized by Projekt DEAL. No funding was received for conducting this study.

Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.
Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent For this type of study formal consent is not required.
Consent for publication Consent for publication was obtained for every individual person's data included in the study.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.