Hepatic encephalopathy in a liver transplant recipient with stable liver function


  • Juan Pablo Arab,

    1. Departamento de Gastroenterología, Facultad de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
    Search for more papers by this author
  • Luis Meneses,

    1. Departamento de Radiología, Facultad de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
    Search for more papers by this author
  • Rosa M. Pérez,

    1. Departamento de Gastroenterología, Facultad de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
    Search for more papers by this author
  • Marco Arrese,

    1. Departamento de Gastroenterología, Facultad de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
    Search for more papers by this author
  • Carlos Benítez

    Corresponding author
    1. Departamento de Gastroenterología, Facultad de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile
    • Departamento de Gastroenterología, Facultad de Medicina, Pontificia Universidad Católica de Chile, Marcoleta 367, 833-0024 Santiago, Chile===

    Search for more papers by this author
    • fax: 56-2-6397780

  • Potential conflict of interest: Nothing to report.

  • Partially funded by grants from the Fondo Nacional De Ciencia y Tecnologia de Chile (Fondecyt #11100113 to C.B. and #1110455 to M.A.).


Postshunt hepatic encephalopathy after liver transplantation (LT) is an infrequent condition and is commonly associated with portal occlusion or stenosis and the presence of a patent portosystemic shunt. Portal vein stenosis (PVS) or thrombosis (PVT) are uncommon complications after LT. The overall frequency of both complications is reported to be less than 3%.1, 2 When PVS or PVT develop early after LT, the occlusion of the portal vein can have catastrophic consequences to the graft including acute liver failure and graft loss. Late PVT/PVS are asymptomatic in ∼50% of the cases2 and mainly diagnosed by a routine ultrasound. Symptomatic postshunt hepatic encephalopathy (HE) is a very infrequent condition after LT that has been scarcely reported in the literature.3, 4 We present here the case of a liver recipient with normal graft function who presented with hepatic encephalopathy 3 months after LT with stable liver function but a severe portal stenosis and the presence of a spontaneous portosystemic shunt whose successful endovascular treatment was followed by the complete resolution of the HE.


HE, hepatic encephalopathy; LT, liver transplantation; MRI, magnetic resonance imaging; PVS, portal vein stenosis; PVT, portal vein thrombosis.

Case Report

A 58-year-old man with cryptogenic cirrhosis (Child-Pugh score A, MELD 9) was listed for LT in June 2010 due to an early stage 2.5 cm hepatocellular carcinoma. He also had experienced a few episodes of grade 2 episodic HE which were properly treated with lactulose. He underwent an LT with a deceased-donor liver graft on January 2011. The patient had an uneventful postsurgical evolution and was discharged on a cyclosporine-based immunosuppression protocol. Four weeks after LT a severe stenosis of the biliary anastomosis was diagnosed and successfully treated employing an endoscopic approach.

Two months later, the patient was admitted because of evident bradypsychia, confusion, and flapping tremor. Serum ammonia levels were raised (80 μmol/L, normal range 16-60). Liver function tests were in the normal range with preserved synthetic and excretory function and renal function was normal. Also, common precipitating factors of HE (infection, constipation, gastrointestinal bleeding, not-prescribed medications, etc.) were excluded. At admission his 2-hour postdose cyclosporine blood level (C2) was 601 ng/mL, which ruled out calcineurin inhibitor-related neurotoxicty. A brain magnetic resonance imaging (MRI) only showed high signal intensity in the globus pallidum on T1-weighted images with no other abnormalities. An abdominal MRI showed a normal liver graft but a severe stenosis of the portal anastomosis and a patent splenorenal shunt (Fig. 1), which was confirmed by angiography (Fig. 2A). Persistent isolated small type 1 gastric varices (diagnosed before LT) were visualized with an upper digestive endoscopy. Employing a balloon, an endovascular dilation of the stenosis was successfully performed. Simultaneously, the splenorenal shunt was successfully occluded with a plug Amplatzer II and metallic coils (Fig. 2B). Twenty-four hours after the procedure, neurological examination was completely normal with no flapping tremor.

Figure 1.

Abdominal MRI showing a patent splenorenal shunt after orthotopic LT (white arrow).

Figure 2.

Portal venogram in a patient with HE due to a patent splenorenal shunt after LT. (A) Severe stenosis of the portal anastomosis (black arrow) and a patent splenorenal shunt (white arrow) are shown. (B) Endovascular dilation of the stenosis (black arrow) and occlusion with a braided-nitinol vascular plug (Amplatzer II) and coils of splenorenal shunt (white arrow) were performed. (C) Anastomotic portal stenosis (black arrow) and new high-flow collateral veins (white arrow). (D) Collateral veins occluded with coils (black arrows) and a self-expansible stent deployed in the portal vein (white arrow).

Fifteen months later the patient presented new self-limited mild confusion episodes. An MRI showed a recurrent stenosis of the portal vein anastomosis and new high-flow collateral veins. These findings were confirmed by angiography (Fig. 2C) and successfully treated with the placement of a self-expansible stent in the portal vein and the use of coils in collateral veins (Fig. 2D). The patient become asymptomatic and was discharged with anticoagulation therapy that was discontinued after 3 months. After 13 months, the patient remains asymptomatic with normal neurologic function.


Portal vein stenosis after deceased donor LT is an uncommon complication, having a reported incidence of about 1%.2 Clinically, PVS presents mainly as portal hypertension-related complications including ascites and esophageal varices.5 HE associated with PVS after LT has only been rarely described. In fact, we were able to found only one similar case in the literature.6 All other reported cases were associated with PVT. Moreover, the recurrent nature of the HE in our patient and its resolution after definitive treatment of PVS with stenting confirmed that the vascular complication was the only cause of HE.

Postshunt HE after LT is an uncommon disorder produced by the presence of two concomitant factors: a patent portosystemic shunt and a high enough portal pressure able to derive a significant flow to the shunt. Notably, this condition can be generated in the presence of normal graft function. This suggests the pathophysiological importance of the high flow through the portosystemic shunt; in fact, the condition can be completely reversed by reducing or eliminating the flow through the portosystemic shunt. Interestingly, our patient already had a high flow splenorenal shunt before LT, as shown in pretransplant abdominal MRI (Fig. 3). Thus, we believe that, in the context of well-preserved liver function and portal hypertension, the relative contribution of the splenorenal shunt to the development of pretransplant HE was significant. Of note, after LT HE developed only in the context of PVS and recurred after restenosis and formation of new collaterals. Thus, this case illustrates a prototypical type B HE7 with portal-systemic bypass and no intrinsic hepatocellular disease being the main mechanism of central nervous system derangement.

Figure 3.

Pretransplant abdominal MRI showing: (A) a cirrhotic liver with hepatocellular carcinoma (black arrow), and (B) a patent splenorenal shunt before orthotopic LT (white arrow).

HE with normal liver function has been reported in patients with surgically created or spontaneous large splenorenal shunts after LT.8 Indeed, HE is more frequent in this context when PVT is present. In fact, in a short series published by Braun et al.,3 all the recipients had a portal occlusion caused by thrombosis. Interestingly, it has been suggested that large splenorenal shunts may be a risk factor for PVT since its presence would decrease portal flow after LT and promote portal occlusion.9, 10 For that reason, if a large splenorenal shunt is present in an LT candidate, as in our case, its ligation could be beneficial. Shunt ligation is not associated with any detrimental effects on either hepatic or renal function.10 However, whether these shunts should be routinely closed during LT needs to be further studied.

Finally, therapy of PVT and PVS in the transplant setting requires a team approach to make a timely diagnosis and treatment. While PVTs are usually detected in the immediate postoperative period and could be managed with thrombolytic therapy or surgical repair, PVS is a more late complication with a state-of-the art endovascular approach being the first option considering safety and efficacy.6, 11 Surgery is then reserved for cases where a less invasive approach fails.12 If detected, occlusion of the splenorenal shunts and collaterals is indicated using vascular plugs and/or metallic coils.

In conclusion, the presence of HE after LT should first prompt verifying proper graft function. If graft function is normal and the usual triggering factors of HE are ruled out, the presence of portosystemic shunts and factors perpetuating its patency (PVT and PVS) should be evaluated using imaging modalities. It must be kept in mind that the absence of PVT in a first-line diagnostic modality such as color Doppler imaging does not exclude an increased resistance to the portal flow and the presence of PVS must be evaluated with complementary imaging techniques. An interventional radiology approach is a safe and effective approach to treat both PVS and persistent portosystemic shunts after LT.