Portal vein stent placement after hepatobiliary and pancreatic surgery

Purpose To evaluate the long-term outcomes of percutaneous transhepatic stent placement for portal vein (PV) stenosis after liver transplantation (LT) and hepato-pancreato-biliary (HPB) surgery. Methods Retrospective study of 455 patients who underwent LT and 522 patients who underwent resection of the pancreatic head between June 2011 and February 2016. Technical success, clinical success, patency, and complications were evaluated for both groups. Results A total of 23 patients were confirmed to have postoperative PV stenosis and were treated with percutaneous transhepatic PV stent placement. The technical success rate was 100%, the clinical success rate was 80%, and the long-term stent patency was 91.3% for the entire study population. Two procedure-related hemorrhages and two early stent thromboses occurred in the HPB group while no complications occurred in the LT group. A literature review of selected studies reporting PV stent placement for the treatment of PV stenosis after HPB surgery and LT showed a technical success rate of 78–100%, a clinical success rate of 72–100%, and a long-term patency of 57–100%, whereas the procedure-related complication rate varied from 0–33.3%. Conclusions Percutaneous transhepatic PV stent is a safe and effective treatment for postoperative PV stenosis/occlusion in patients undergoing LT regardless of symptoms. Due to increased risk of complications, the indication for percutaneous PV stent placement after HPB surgery should be limited to patients with clinical symptoms after an individual assessment.


Introduction
Portal vein (PV) stenosis is a well-known complication after liver transplantation (LT), and hepato-pancreato-biliary (HPB) surgery [1][2][3][4][5][6][7][8]. Potential symptoms related to PV stenosis are abdominal pain, liver failure, intractable ascites, and gastrointestinal bleeding [9][10][11] which render it as a potentially serious condition. Management of postoperative PV stenosis with percutaneous stent placement has previously been reported to be a viable and less invasive option than laparotomy and surgical revision [12]. Even though PV stenosis is a rare complication after LT [2,3], it has recently been shown to be as high as 19.6% after pancreatoduodenectomy [13]. Regarding LT, PV stenosis is more frequent after pediatric transplantation due to size mismatch [14].
With an increasing rate of surgical resection and vascular reconstruction in HPB surgery, an increasing incidence of complications related to the PV must be anticipated [15][16][17]. Percutaneous transhepatic stenting of the PV has gained acceptance as a safe minimally invasive procedure with promising results [12,[18][19][20][21][22][23][24]. The aim of this paper is to evaluate the results of stenting of PV stenosis after LT and HPB surgery within a high-volume tertiary referral center.

Materials and methods
This was a retrospective review of all patients undergoing PV stent placement due to postoperative PV stenosis at Rikshospitalet, Oslo University Hospital between June 2011 and February 2016. The hospital is a tertiary referral center for all HPB surgery in the southeast health region (3 million inhabitants) and the only national center for abdominal transplantation surgery in Norway (5.3 million inhabitants). Hospital records, including radiological reports, were reviewed. Type of surgery, time from surgery to PV stent placement, and complications to the procedure were assessed. Patency was evaluated according to latest available radiological modality, as follow-up was heterogeneous. Methods for pancreatic resections and LT were performed as described elsewhere [25,26]. The hospital review board approved the study (2016/8365) according to the general guidelines provided by the regional ethics committee. The manuscript was completed in accordance with the STROBE statement [27].

Indication for stent placement
For patients undergoing LT, ultrasonography (US) on postoperative day (POD) 1 was performed routinely. Subsequent radiology was obtained on clinical suspicion of adverse events. Patients undergoing pancreatoduodenectomy were not subjected for routine radiology in the immediate postoperative period, but patients with concomitant PV resection had US on POD 1 and the day before discharge. The criteria for definite diagnosis of significant PV stenosis were a 3-fold increase in portal blood velocity detected by US and a minimum of 50% PV diameter reduction on contrast-enhanced computed tomography (CT) or percutaneous transhepatic portography. PV occlusion was defined as the absence of contrast enhancement through the PV. Patients with radiological significant PV stenosis underwent PV stenting either because of clinical manifestation or as a preemptive treatment in order to avoid future complications. Hence, indication for PV stent placement was based on radiological findings, clinical manifestations, or suspicion of future adverse events related to untreated PV stenosis.

Stent placement procedure
Routinely, the procedures were performed with the patients under conscious sedation. In all patients, access to the PV was gained by percutaneous US-guided puncture of a subsegmental or segmental portal branch with introduction of a 6 Fr introducer (Merit Medical Systems, South Jordan, UT, USA). After diagnostic angiography with confirmation of significant stenosis, stenting was performed using selfexpanding nitinol stents 40-60 mm in length and oversized in diameter by 1-2 mm as compared with the adjacent PV. For diameters of 10 and 12 mm, we used Smart Control selfexpanding stents (Cordis, a cardinal health company, Dublin, OH, USA); in larger diameters, 14-20 mm Sinus-XL Flex stents (Optimed Medizinische Instrumente GmbH, Ettingen, Germany) were placed. In cases of residual stenosis, stents were dilated with balloons at the size of the prestenotic portal segment. During removal of the introducer sheaths, gelfoam pledgets were placed in the access channel in order to prevent hemorrhage. Technical success was defined as residual stenosis < 30% or < 10% difference in pressure gradient on immediate angiography and absence of significant stenosis on early follow-up at US or CT. After the procedure, the patients received prophylactic low molecular weight heparin based on their body weight (150 IU/kg/day) for a minimum of 3 months. Following PV stent placement, supplementary radiology in addition to routine follow-up was obtained on clinical indication.

Results
Between June 2011 and February 2016, 455 patients underwent LT and 522 patients underwent resection of the pancreatic head (pancreatoduodenectomy n = 494, total pancreatectomy n = 28). There were 14 patients (3.1%) in the LT group and eight patients (1.5%) in the HPB group confirmed to have postoperative PV stenosis. Six of the eight HPB patients underwent resection and reconstruction of the PV during the primary operation. In addition, one patient developed postoperative PV stenosis after a resection of the extrahepatic bile ducts. An overview of the patients is presented in Table 1. All of the 23 patients with PV stenosis/occlusion underwent a technical successful percutaneous transhepatic PV stent placement. The etiology of PV stenosis was anastomotic stenosis in 17 patients while the remaining six were caused by tumor recurrence and/or thrombosis. Only 10 of 23 patients had clinical manifestations while the remaining 13 patients had an asymptomatic PV stenosis discovered on routine radiological follow-up. Eight of 10 patients with clinical manifestations experienced disappearance of symptoms after percutaneous transhepatic PV stent placement resulting in a clinical success rate of 80%. There were two procedure-related hemorrhages and two early stent thromboses in the HPB group resulting in a complication rate of 17.4% for the entire study population. Long-term stent patency for the entire study population based on the last available radiological imaging was 91.3% with a median follow-up of 6 months in the HPB group and 49 months in the LT group (Table 2).

Discussion
In this study, we investigated the outcome of PV stent placement for postoperative PV stenosis after LT and HPB surgery. PV stent placement was performed with a procedure-related complication rate of 17.4% for the entire study population, confirming the results of previous publications [11]. The technical success rate was 100%, the clinical success rate was 80% among patients with clinical manifestations, and the long-term stent patency was 91.3% for the entire study population. This shows that PV stenting is effective in relieving PV stenosis after LT and HPB surgery. The results of this audit are comparable with other publications; [12,24,28] however, comparison must be done with caution as there are obvious inter-institutional differences in the postoperative follow-up and indication of PV stent placement. In our institution, patients with clinical and/or radiological signs of significant PV stenosis were subjected to PV stenting after evaluation by a multidisciplinary team. To the best of our knowledge, there are few studies investigating which patients benefit from postoperative PV stent placement.
Among the patients with PV stent placement after LT, there were no procedure-related complications and a long-term patency of 100%. This supports the decision to preemptive treat asymptomatic patients with radiological signs of significant PV stenosis after LT to reduce the risk of future complications. Importantly, PV complications after LT have been well documented because PV stenosis and thrombosis can potentially be devastating and lead to graft failure [14]. PV stenosis is more frequent after pediatric LT, and some centers have practiced an early approach to PV abnormalities during pediatric LT. A recent study showed that PV stent placement during the transplant or in the immediate postoperative setting through the inferior mesenteric vein offered both a high feasibility and satisfactory results in pediatric recipients [29].
In the HPB group, there were two procedure-related hemorrhages. One patient received PV stent 43 days after a pancreatoduodenectomy. During the PV stent placement, the patient suffered bleeding from a minor hepatic arterial branch and underwent embolization 8 days later. The patient eventually died in the intensive care unit 48 days after PV stent placement and 91 days after pancreatoduodenectomy due to pneumonia and respiratory failure due to complications after PV stent placement. The other hemorrhage was minor bleeding from the liver capsule that was successfully treated with percutaneous hemostatic applications. Another two patients experienced early stent thrombosis and permanent stent failure with a result comparable to the situation before stent placement. The possible reason for early stent occlusion in the first patient may be postoperative pancreatic fistula causing nearby inflammation and thus promoting thrombosis. The second patient had a recurrence of distal cholangiocarcinoma around the PV with resulting subtotal PV occlusion eight months after bile duct resection. Two days after PV stent placement, early stent thrombosis was confirmed with possible cause being thrombophilia in a palliative patient with advanced metastatic cancer surrounding the PV.
A summary of selected studies reporting PV stent placement for the treatment of PV stenosis after HPB surgery and LT are presented in Table 3. The studies show a technical success rate of 78-100%, a clinical success rate of 72-100%, and a long-term patency of 57-100%. The procedurerelated complication rate varied from 0-33.3%. In the current study, PV stent placement was performed on three HPB patients with asymptomatic PV stenosis. Of note, in other published studies, PV stent placement after HPB surgery was only performed in patients with symptoms of PV stenosis such as ascites, gastrointestinal bleeding, or liver dysfunction (Table 3). Thus, it may be questioned if patients with asymptomatic stenosis should undergo PV stenting after HPB surgery. Kang et al. showed that about 20% (162 of 826) of patients undergoing pancreatoduodenectomy developed PV stenosis/occlusion, with a significantly higher rate in patients who underwent PV resection [13]. Moreover, 13% (21 of 162) of the patients with PV stenosis/occlusion developed gastric or  hepaticojejunostomy varices. Accordingly, the authors recommended careful postoperative surveillance for PV stenosis/ occlusion after pancreatoduodenectomy. However, although 21% (5 of 21) of the patients with gastrointestinal bleeding experienced fatal recurrent bleedings, routine PV stenting is not generally recommended in patients with asymptomatic PV occlusion because the incidence of complications related to PV stenting is seemingly higher than the rate of gastrointestinal bleeding caused by portal hypertension. However, an aggressive approach including PV stenting or selective surgical therapy to lower PV hypertension is recommended in patients who develop recurrent gastrointestinal bleeding. For patients with symptomatic PV stenosis/occlusion caused by unresectable malignant tumors, attempts have been made to identify groups that benefit from PV stenting, but results remain inconclusive mainly due to small study samples [38].
Most of these patients have advanced disease and it is difficult to predict whether PV stent placement can provide a survival benefit in this subgroup of patients. Vascular interventional procedures may be contraindicated in the very early postoperative period (< 3 weeks) due to the risk of suture dehiscence and anastomotic bleeding. This did, however, not pose a problem in our study and the shortest time from primary operation to PV stent placement was 21 days. Concerns about the long-term stent patency have been reported [28,39,40]. The role of anticoagulation regarding PV stent placement is not well established. Our strategy was prophylactic low molecular weight heparin based on their body weight (150 IU/kg/day) for a minimum of three months. Aspirin was not routinely used; however, patients using aspirin before surgery continued their lifelong regimen throughout the hospital stay. In several studies [8,9,22,[40][41][42][43], no routine administration of anticoagulation was given as the risk of gastrointestinal bleeding was weighed against portal stent thrombosis. In our experience, the indication for anticoagulation after PV stent placement is very strong in patients with an intravascular thrombogenic foreign body in the PV sometimes combined with a malignant disease. The risk of gastrointestinal bleeding is also significantly reduced after a normal portal flow is reestablished. Our routine administration of anticoagulation may be a possible cause of acceptable long-term stent patency.
The limitations of this study lay in the retrospective design. The study sample was small and the study population was highly heterogeneous with respect to procedures being performed and causes of PV stenosis/occlusion. Nevertheless, we experienced an acceptable short-and long-term outcome. However, it is difficult to draw manifest conclusions and the findings of the study must be interpreted with caution. The literature review revealed that only two of 14 published papers have included more patients (29 and 59, respectively) than the 23 patients in the current study. Thus, it will be difficult to gather a larger collective of patients with this condition within a reasonable time span, even for high-volume referral centers. Management of anticoagulation after stenting and differences between collectives after liver transplant or HPB surgery are topics that should be further investigated. Moreover, a future meta-analysis of published studies on this topic could be useful. In conclusion, our results support the use of percutaneous transhepatic PV stent as a safe and effective treatment for postoperative PV stenosis or occlusion in patients undergoing LT regardless of symptoms. Due to increased risk of complications, the indication for percutaneous PV stent placement after HPB surgery should be limited to patients with clinical symptoms after an individual assessment.
Authors' contributions Study conception and design: KJL and DK; acquisition of data: all authors; analysis and interpretation of data: all authors; drafting the manuscript: AK, DK, and KJL; critical revision of the manuscript: all authors.
Funding Information Open Access funding provided by University of Oslo (incl Oslo University Hospital).

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent Informed consent was not obtained by all participants. The study was permitted by the hospital review board in accordance with the regional ethical committee.
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/.