The management of portosystemic shunts in liver transplant recipients relies on appropriate perioperative study. There are several strategies for shunt handling, ranging from preoperative interventional procedures to intraoperative surgical interruption or embolization. Appropriate management often results in a successful outcome, although wrong decisions could lead to serious consequences. Here, we report a liver transplant recipient with grade 2 portal vein thrombosis associated with 2 large portosystemic shunts (coronary and mesocaval), which were managed intraoperatively via thrombectomy without shunt ligation. Acute portal vein thrombosis developed early after transplant due to portal steal syndrome. The patient underwent a successful endovascular shunt embolization, with prompt restoration of hepatopetal portal flow and resolution of the portal steal. Use of interventional radiology in perioperative management of transplant patients has recently gained wider importance; our case reported here is particularly suggestive of the good outcomes of a multidisciplinary approach to a threatening complication such as postoperative acute portal vein thrombosis.
Key words : Complications, Embolization, Interventional radiology, Liver transplantation, Porto-systemic shunt
Liver transplant represents the best treatment for end-stage liver disease complicated by cirrhosis and portal hypertension. A frequent consequence of portal hypertension is the development of spontaneous portosystemic shunts, which are present in approximately 20% to 40% of patients with cirrhosis awaiting a liver transplant. The association between portosystemic shunts and portal vein thrombosis has been reported in 4.4% of transplant candidates.1
Replacement of a cirrhotic liver with a healthy graft with low vascular resistance usually leads to hepatopetal flow diversion and subsequent shunt detention and closure.2 However, the presence of large (> 1 cm) shunts may require further attention as they may not divert their flow after liver replacement, leading to low portal flow, inadequate graft perfusion, and increased risk of graft dysfunction and postoperative portal vein thrombosis.3 Here, we report the case of an acute posttransplant portal vein thrombosis that was successfully treated with an interventional approach.
Patient characteristics and preoperative imaging
A 60-year-old male patient affected by alcoholic cirrhosis presented with a history of hepatic encephalopathy, portal hypertension with previous variceal bleeding, moderate ascites, and Barcelona Clinic Liver Cancer stage A hepatocellular carcinoma (HCC). The patient had no other relevant comorbidities in past medical history. A pretransplant contrast-enhanced computed tomography (CT) revealed a cirrhotic liver associated with moderate ascites, splenomegaly, and the presence of 2 different portosystemic shunts through the coronary vein and through a 17.5-mm mesocaval shunt, associated with grade 2 portal vein thrombosis, and a 22-mm HCC in S8 (Figure 1).
After a thorough preoperative multidisciplinary evaluation, the patient was considered fit for transplant from a surgical and anesthesiologic point of view. At listing, the patient’s body mass index was 19.47 kg/m2 (weight: 53 kg/height: 1.65 m), his Child-Turcotte-Pugh score was 9 (class B), and his Model for End-Stage Liver Disease score was 17.
After 36 days on the wait list, the patient received an ABO-compatible 1683-g whole liver graft from a 70-year-old brain dead donor; no liver biopsies were retrieved during procurement due to a good macroscopic aspect of the liver graft.
The transplant procedure was performed using the piggyback technique; portal vein thrombectomy was carried out before portal anastomosis. Portal flow looked appropriate after thrombectomy, and the leading surgeon decided not to ligate either of the 2 shunts. Intraoperative Doppler ultrasonography showed sound anastomoses with adequate hepatopetal portal flow. Total operative time was 8 hours, graft cold ischemia time lasted 6 hours, and the patient required transfusion of 7 red blood cell units and 8 fresh plasma units.
At the end of surgery, blood lactate levels (which had increased to 7 mmol/L during the anhepatic phase) had decreased to 2 mmol/L and body temperature had increased to 36.2°C. The first posttransplant serum alanine aminotransferase (ALT) sample showed a level of 1151 IU/mL.
Early postoperative course and postoperative imaging
During the first day after transplant, liver function examinations showed that ALT levels had increased to 4570 IU/mL and lactate concentration had increased to 7 mmol/L. The patient underwent an urgent contrast-enhanced CT scan that revealed a complete portal vein thrombosis with periportal edema, intestinal edema, and mild ascites; the coronary vein and the mesocaval shunt were still patent. A coagulation workup, including protein S, protein C, and antiphospholipid antibody, revealed no clotting alterations.
Suspicion for a portal steal syndrome was raised on the basis of the CT scan, the patient’s clinical condition, and laboratory findings. The case was promptly discussed with the interventional radiology team to evaluate the best treatment option.
The interventional procedure considered 2 targets: portal vein stenting and shunt embolization. A foreign body in the porta was not desirable, considering the risk of a potential retransplant; therefore, the stenting option was excluded. The patient thus required angiographic suite planning for determination of portography plus shunt embolization for portal-flow diversion.
A 5F percutaneous transhepatic right portal access was obtained: the portography confirmed the presence of a high-flow mesocaval shunt with portal steal syndrome and subocclusion of the portal vein.
To avoid risk of nontarget embolization due to the high flow through the mesocaval shunt, a 9F jugular access was obtained to catheterize the origin of the mesocaval shunt on the caval side: a 14 × 40-mm balloon was inflated at shunt origin to stop the flow within the shunt, and an embolization with gelatin sponge and detachable coils and with N-butyl cyanoacrylate was performed under shunt exclusion (Figure 2).
After balloon deflation, complete embolization of the shunt with partial restoration of hepatopetal flow was observed, despite the finding of an increased flow through the coronary vein, which was subsequently occluded with embolization with detachable coils.
At the end of the procedure, a regular hepatopetal flow was obtained, with total occlusion of the mesocaval and coronary shunts (Figure 3). The interventional procedure lasted 3 hours, and no intraoperative complications were observed.
After the interventional procedure, there was a progressive decrease in liver enzymes levels with normalization on postoperative day 9. The postprocedural course was uneventful, and a Doppler ultrasonography on postoperative day 7 revealed hepatopetal flow within the graft. Shunt occlusion was confirmed by a contrast-enhanced CT on postoperative day 14; notably, there were no clinical or radiological signs of pulmonary embolism.
The patient was discharged 28 days after liver transplant in good general condition and with normal liver function tests. Six months after transplant, the patient had remained well with no evidence of ascites and with normal liver function tests. A follow-up contrast-enhanced CT scan showed normal liver graft, with regular portal flow associated with sustained mesocaval and coronary shunt closures.
The presence of spontaneous portosystemic shunts should be carefully studied during evaluation of transplant candidates so that the best perioperative strategy tailored on patient and shunt characteristics can be selected.
The pathophysiology of shunt formation relies on increased portal pressure due to the highly resistant cirrhotic liver, which facilitates shunt opening and flow diversion. These shunts are more frequent in patients with advanced cirrhosis, as proven by the direct correlation between cirrhosis severity and shunt formation.4 The impact of portosystemic shunts on transplant outcomes varies widely, depending on their influence on portal hemodynamics.
Most of these shunts close spontaneously due to pressure changes and flow diversion after the replacement of the cirrhotic liver with a healthy graft with adequate low-resistant vascular bed. On the other hand, in the presence of large and well-established shunts, spontaneous closure may be delayed or may not occur, carrying the risk of portal flow steal and portal vein thrombosis.5
The risk of such threatening complications is even higher in the setting of living-donor liver transplant because of the decreased vascular bed in a partial liver graft as well as the further increased intrahepatic vascular resistance caused by rapid regeneration of grafted liver, which delays spontaneous shunt closure.
The majority of literature on management of portosystemic shunts stresses the importance of flow/pressure monitoring to tailor intraoperative decisions for shunt closure/embolization or alternative anastomotic strategies in the setting of portal vein thrombosis (ie, renoportal anastomosis).1,6–9 In our patient, no intraoperative flow/pressure testing was carried out. There was consensus of adequate portal flow after a thrombectomy and flush test, and identification of hepatopetal flow on Doppler ultrasonography was considered sufficient for the decision process.
Such an approach was not appropriate in the setting of a thrombosed portal vein, when the portosystemic shunt usually exerts a significant hemodynamic effect and may not shrink rapidly enough to prevent flow reduction and development of portal vein thrombosis. It thus seems appropriate to suggest thorough intraoperative flow/pressure monitoring to tailor the intraoperative approach toward shunt management.
It must be noted that shunt isolation may be challenging and may increase the risk of intraoperative complications (eg, bleeding) in patients with severe cirrhosis. Intraoperative portography and shunt embolization have been reported as an appealing choice in this setting.8 Another interesting option for transplant candidates with portal vein thrombosis and portosystemic shunts who are awaiting liver transplant is pretransplant transjugular intrahepatic portosystemic shunt. This procedure has been recently shown to be a safe and effective approach for portal recanalization with the potential for pretransplant shunt embolization and closure.10
The role of interventional radiology in perioperative management of transplant patients has recently gained wider importance.11 Our case is particularly suggestive of the good outcomes of a multidisciplinary approach to a threatening complication such as postoperative acute portal vein thrombosis. Here, our multidisciplinary analyses of pre- and postoperative imaging together with intraoperative findings allowed a prompt diagnosis and a successful endovascular treatment.
An interesting technical aspect of the procedure was the utilization of an inflated balloon on the caval side of the mesocaval shunt during the embolization maneuver. This technique allowed the achievement of complete shunt embolization, avoiding any risk of pulmonary embolism, which would have been hazardous in the setting of a high-flow mesocaval shunt, as documented during initial portography.
This complex but extremely tailored interventional procedure granted proper embolization of the shunts and restoration of the hepatopetal portal flow with reduced risk of pulmonary embolism, allowing a proper functional recovery of the injured graft within several days.
We recommend careful perioperative evaluations of the hemodynamic effects of portosystemic shunts, based on preoperative imaging and intraoperative flow/pressure monitoring of the portal system. In addition, we suggest application of intraoperative ligation or embolization in the presence of high-flow portosystemic shunts.1,3
Today, interventional radiologists have gained wider importance in the perioperative management of liver transplant patients, providing a minimally invasive, safe, and effective treatment to overcome complex and threatening complications such as those presented in this case.11
DOI : 10.6002/ect.2019.0273
From the 1Department of General Surgery and Transplantation and the
of Interventional Radiology, Niguarda Ca’ Granda Hospital, Milan, Italy; and the
3School of Medicine, University of Milan-Bicocca, Milan, Italy
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare.
Corresponding author: Leonardo Centonze, Department of General Surgery and Transplantation, Niguarda Ca’ Granda Hospital, Piazza dell’Ospedale Maggiore, 3, 20162 Milan, Italy
Phone: +39 3334127815
Figure 1. Pretransplant Contrast-Enhanced Computed Tomography Scan
Figure 2. Catheterization and Embolization of Mesocaval Shunt With Balloon at Shunt Origin
Figure 3. Regular Hepatopetal Flow Shown After Embolization