Portosystemic shunts (PSS) reportedly occur in 18% to19% of patients evaluated for a liver transplant.1, 2 In almost all of the patients with large spontaneous PSS, the flow in the portal vein becomes hepatofugal.1 Although smaller PSS do not cause serious problems and will undergo involution over time following transplantation, larger PSS are associated with increased risk of portal vein thrombosis and primary nonfunction from decreased portal flow into the graft.1, 2 To prevent complications, ligation of larger PSS is recommended following graft perfusion. To the authors' best of knowledge, PSS causing clinical problems several years following liver transplantation have not been reported previously.
A 40-year-old male presented with progressively worsening hepatic encephalopathy nine years following a liver transplant for postnecrotic hepatitis C cirrhosis. Initially, he was thought to have graft failure from recurrent hepatitis C. A liver biopsy showed mild, chronic, active hepatitis C with no rejection (Fig. 1A). Trichrome stain showed no evidence of fibrosis (Figure 1B). Laboratory studies showed total bilirubin, 0.9 mg%: albumin, 3.6 gm%: alanine and aspartate aminotransferases, 26 U/L and 33 U/L, respectively: prothrombin international normalized ratio, 0.8: and venous ammonia, 181 μmol/L. Doppler ultrasound showed patent hepatic vasculature, a large splenorenal shunt, and hepatopedal flow in the portal vein. Magnetic resonance imaging of the abdomen after gadolinium administration showed a large splenorenal shunt arising near the hilum of the spleen and entering the left renal vein, which was markedly enlarged (Fig. 2). The portal vein pressure measured via a percutaneous transhepatic approach was 8 mm Hg. Contrast injected at the junction of the splenic and portal veins showed hepatopetal flow as well as flow into the shunt (Fig. 3A). Contrast injected into the splenic vein demonstrated that all of the flow was into the shunt, the left renal vein, and the inferior vena cava with no flow toward the liver (Fig. 3B). At a different session, the renal end of the shunt was approached via a percutaneous transfemoral vein route, and the shunt was occluded successfully with a combination of Guglielmi Detachable Coil (Boston Scientific, Natick, MA) and n-butyl cyanoacrylate liquid embolic system (Trufill, Cordis, Miami, FL) (Fig. 4A and 4B).
Following the shunt occlusion, the patient's encephalopathy improved dramatically, and venous blood ammonia decreased to 50 μmol/L. Presently, nine months following the procedure, the patient's encephalopathy is resolved, and the liver chemistries have remained stable. A Doppler ultrasound showed no evidence of the splenorenal shunt.
The unique features in this case are the occurrence of a large PSS nine years following a liver transplant in the absence of current cirrhosis and portal hypertension. Furthermore, the shunt was responsible for the occurrence of hepatic encephalopathy severe enough to initially consider retransplantation. The final interesting feature of our case is successful treatment of the problem with nonsurgical interventional radiological technique with complete resolution of the hepatic encephalopathy.3