A 39-year-old man presented in April 2010 with an 11 mo history of slowly progressive spastic paraparesis of his lower extremities. The patient in question had a history of hepatitis B disease. In March 2008, he underwent a surgical splenorenal shunt and splenectomy due to recurrent esophagogastric variceal bleeding and thrombocytopenia. In May 2009, fourteen months after his surgical splenorenal shunt, gait impairment was first noticed. An episode of HE was not seen during this period. The patient received conservative medical management, including protein restriction, non-absorbable antibiotics, lactulose and physical therapy, after the onset of gait impairment. However, these measures did not prevent the progressive decline in his mobility and this necessitated the use of a cane because of marked leg weakness, gait imbalance, and instability.

Physical examination revealed normal mental status and cranial nerve function, and the absence of asterixis. Kayser-Fleischer rings were absent, upper extremity strength and tendon reflexes were normal, and Hoffman’s sign was absent. In the lower extremities, the legs were spastic, the tone was more noticeably increased with brisk tendon reflexes and there was clonus at the ankles. Plantar reflexes were extensor and the patient was unable to move his legs. There was no atrophy or any fasciculations, nor any evidence of ascites or peripheral edema.

Laboratory investigations showed normal serum levels of electrolytes, glucose, vitamin B12, and creatinine. He had mild anemia (hemoglobin level of 112 g/L; normal range, 120-160 g/L) and thrombocytopenia (85 × 10³/mm³; normal range, 100 × 10³-300 × 10³/mm³). Prothrombin time was 16 s (normal range, 11-14.5 s). Other values were as follows: 28 μmol/L total bilirubin (normal range, 3-20 μmol/L); 15 μmol/L conjugated bilirubin (normal range, 0-7 μmol/L); 156 μmol/L ammonia (normal range, 18-45 μmol/L); 56 U/L alanine aminotransferase (normal range, 5-40 U/L); 65 U/L aspartate aminotransferase (normal range, 15-45 U/L); 126 U/L alkaline phosphatase (normal range, 30-115 U/L); 36 g/L albumin (normal range, 35-50 g/L); 45 g/L globulin (normal range, 21-40 g/L); international normalized ratio 1.3 (normal range, 0.9-1.2). His liver was graded as Child-Pugh class A.

Serum examinations for hepatitis B surface antigen and hepatitis B e-antigen were positive, but a test for hepatitis B core antigen was negative. Antibodies to hepatitis C virus and human immunodeficiency virus were negative, syphilis serology was negative, and an analysis of the cerebrospinal fluid was normal.

Magnetic resonance imaging of the brain and spine was normal, and an abdominal computed tomography (CT) revealed a cirrhotic liver. An abdominal ultrasound examination showed features of cirrhosis of liver and a Doppler study showed a widely-open splenorenal shunt. Electro-encephalography revealed no definite abnormality and an endoscopic examination revealed no esophagogastric varices.

Given the aforementioned extensive neurological evaluation, which did not identify any alternative explanation for the patient’s spastic paraparesis of the lower extremities, the diagnosis of HM secondary to the surgical splenorenal shunting was made. After admission, with protein restriction and medication with lactulose, neomycin, B-vitamins, physical therapy, and antispastic agents for 3 wk, the patient showed no improvement in his spastic paraparesis, and his hyperammonemia remained within the range of 140-180 μmol/L.

After discussion with hepatologists, gastroenterologists, and vascular surgeons at the authors’ hospital, we decided to occlude the surgical splenorenal shunt to prevent further neurologic deterioration. Informed consent was obtained from the patient.

Under local anesthesia (2% lidocaine) and fluoroscopic guidance, venous access was obtained through the right internal jugular vein, and a 9 Fr with an angled-tip 80-cm long introducer sheath (Cook, Bloomington, Indiana) was advanced into the inferior vena cava. The splenorenal shunt was then catheterized with a 5 Fr Cobra catheter (Terumo Corporation, Tokyo, Japan) and an angled-tip 0.035-inch hydrophilic guidewire (Terumo). Initial portography and pressure measurements were performed with a 5 Fr multiple-side-hole catheter (Cook) positioned in the superior mesenteric vein. At the portal venogram, a widened patent shunt was seen (Figure 1), with a portacaval pressure gradient of 8.0 mmHg. The portal vein pressure was 12.5 mmHg.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Digital subtraction portogram obtained via a catheter inserted through the distal splenorenal shunt into the superior mesenteric vein (arrows) using the right jugular vein approach. The wide patent splenorenal shunt (arrow heads) is shown. Note the inferior vena cava (curved arrow).

We initially placed an occlusion balloon catheter into the splenorenal shunt tract to test if the patient could be tolerable to the full embolization procedure. Following the angiographic procedure, over a 0.035-inch Amplatz wire (Cook), the 9 Fr introducer sheath advanced into the splenorenal shunt. A 7 Fr latex balloon catheter, with a maximum outer diameter of 16 mm (Boston Scientific, Watertown, Mass) was then inserted into the splenorenal shunt through the sheath. Under the inflated balloon, injection of contrast material through the sheath showed complete occlusion of the surgical splenorenal shunt. After the balloon occlusion, heparin infusion was given via a peripheral vein, at a dosage of 120 mg/24 h, to avoid catheter-related thrombosis.

Following the balloon occlusion procedure, the patient’s plasma ammonia level decreased to 110 μmol/L, 70 μmol/L, 60 μmol/L and 50 μmol/L, on post-procedural days 2, 3, 5 and 6, respectively. He reported a mild improvement in his leg strength and balance, while the other HE-related symptoms remained unchanged. Follow-up CT on post-procedural day 5 showed the balloon positioned appropriately (Figure 2A). Repeat liver function tests were normal, and repeat endoscopic examination confirmed no induction of esophagogastric varices. There was no evidence of ascites. Six days after the balloon occlusion procedure, we decide to occlusion the splenorenal shunt permanently with endovascular embolization.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Coronal maximum intensity projection reconstruction of a contrast enhanced 64-slice helical computed tomography scan of the abdomen. A: Obtained 5 d after placement of an occlusion through the splenorenal shunt and shows the occlusion balloon placed within the shunt tract (arrows), note the inferior vena cava (curved arrow) and the left renal vein (arrow heads); B: Obtained 3 mo after the Amplatzer vascular plugs (AVP) occlusion procedure and shows the correct AVP position (arrow heads), opacification of the portal system (arrows), and no opacification of the surgical shunt, as well as the varices.

Given the relatively larger diameter of the splenorenal shunt, we decided to use an Amplatzer vascular plug (AVP) instead of conventional embolization materials (i.e., coils, glue, and particles).

The diameter of the surgical splenorenal shunt was approximately 13 mm in diameter, calculated on the angiography workstation. Given the diameter of the shunt, we used an 18-mm AVP (AGA Medical Corp., Golden Valley, MN, United States) with the 30%-50% oversizing recommended by the manufacturer, to prevent plug migration. For deploying the AVP, the 9 Fr introducer sheath advanced over a 0.035-inch Amplatz wire (Cook) into the most distal extreme of the splenorenal shunt. The AVP device was then advanced through the 9 Fr sheath and initially deployed by retracting the sheath, but without detaching the device. A control injection was performed, to ensure satisfactory positioning of the device and to enable repositioning. When in a satisfactory position, the AVP was detached by anticlockwise rotation of the AVP guide wire to unscrew it from the AVP device proper. Within minutes of deployment, the shunt had completely stopped. Injection of contrast material through the sheath showed complete occlusion of the surgical shunt.

Following AVP embolization, an indirect digital subtraction portography, carried out during the same procedure and obtained with a 4 Fr catheter (Terumo) inserted into the superior mesenteric artery (SMA) via the right femoral artery approach, was used to evaluate portal vein system. Digital subtraction SMA angiogram at the portal venous phase demonstrated a marked improvement in the opacification of the portal vein in comparison with that of the portography pre-deployment of the AVP (Figure 3), indicating an increase in antegrade intrahepatic portal vein perfusion. No contrast material passed through the surgical splenorenal shunt.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 After deploying the Amplatzer vascular plugs, indirect digital subtraction portography, obtained through the catheter inserted into the superior mesenteric artery using the right femoral artery approach, demonstrates opacification of the portal vein (arrows). Note the correct Amplatzer vascular plugs position (arrow heads) and no visualization of varices.

No complication was noted during and after the procedure. Five days after the AVP embolization, the patient’s serum ammonia level had normalized to 40 μmol/L.

The patient was discharged 6 d post-AVP embolization. Doppler US confirmed complete occlusion of the splenorenal shunt. In June 2010, one month after the embolization, the patient reported a gradual improvement in strength; he was still moderately weak but was able to walk short distances (50-100 m) with crutches.

By 3 mo after the procedure, he was able to walk about 300-500 m with crutches. He still complained of stiffness and spasms in his legs, but improved compared to 3 mo ago. A follow-up contrast enhanced CT scan at the venous phase confirmed that the AVP position was correct, that there was opacification of the intrahepatic portal veins, that the splenorenal shunt tract was thrombosed, and that there was no opacification of the surgical shunt or varices (Figures 2B and 4).

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Axial image of contrast enhanced 64-slice helical computed tomography scan of the abdomen, showing the correct Amplatzer vascular plug position (curved arrow), opacification of the portal system (arrow heads), and a thrombosed shunt tract (arrows).

The patient continued to show improvement in his HM symptoms over the next few months. In December 2010, 7 mo after the splenorenal shunt embolization, his strength improved significantly, he was able to walk 1 to 2 km aided by crutches, and had only mild stiffness in both legs. Esophagogastroscopy showed no evidence of esophagogastric varices.