This case report highlights two important clinical issues: portal vein arterialization prior to embolization of the aneurysm may be a feasible therapeutic strategy for a pseudoaneurysm that develops after hepatectomy for hepatobiliary malignancy to guarantee arterial inflow to the remnant liver, and early embolization of arterioportal shunting after confirmation of arterial inflow to the liver should be performed to prevent morbidity induced by portal hypertension.
The remnant liver oxygenation must be cautiously considered when an intervention is performed for a pseudoaneurysm after hepatobiliary resection. Previous reports have shown that ligation of the hepatic artery to treat a pseudoaneurysm after hepatectomy results in fatal liver failure [2, 4]. Selective embolization of an aneurysm is an established procedure, but there is a risk of migration of the embolic agent to the hepatic artery, potentially resulting in hepatic parenchyma infarction [5,6,7]. Although the use of a stent graft for a hepatic artery aneurysm allows maintenance of the hepatic arterial flow, it cannot be attempted for a narrow artery or at a bifurcation of vessels [5, 8,9,10]. In the present case, the pseudoaneurysm was located close to the bifurcation of the LHA to A2 plus A3 and A4, and the LHA was extremely narrow; therefore, stent graft placement could not be attempted. Embolization of the pseudoaneurysm was associated with a risk of occluding the LHA by migration of the embolus material. Therefore, we performed arterioportal shunting before embolization of the pseudoaneurysm to maintain the oxygen level of the remnant liver, even if the LHA was accidentally occluded. Fortunately, we were able to selectively embolize the aneurysm. We selected liquid thrombin as the embolization material because coil embolization developed halation of CT that made follow-up of the pseudoaneurysm difficult, and n-butyl cyanoacrylate with lipiodol embolization had a possibility to become indistinct of size of the pseudoaneurysm.
The arteriovenous shunt uses the mesenteric vessels and resolves the liver ischemia caused by thrombosis or ligation of the hepatic artery after hepato-biliary-pancreatic surgery or liver transplantation [2, 11,12,13]. Although previous reports have shown that arterioportal shunting can salvage the totally dearterialized liver and prevent fatal liver failure, Iseki et al.  described the risk of reperfusion liver injury caused by arterioportal shunting after ligation of the hepatic artery . Therefore, we performed arterioportal shunting between the ileocolic artery and vein before the interventional radiology. This is the first report of performing arterioportal shunting prior to embolization of a pseudoaneurysm. This procedure would be a feasible therapeutic strategy for a pseudoaneurysm after massive hepatectomy.
Arterioportal shunting is associated with postoperative complications of shunt occlusion and portal hypertension with significant ascites, hyperbilirubinemia, and gastrointestinal bleeding [2, 13]. Bhangui et al.  reported that the mortality after arterioportal shunting for the totally dearterialized liver was high, but most causes of death were not related directly to arterioportal shunting because patients were already very severely ill before shunting. They reported that only 1 of 52 patients died from a complication after arterioportal shunting (shunt occlusion and reperfusion injury). Significant ascites, variceal bleeding, hyperbilirubinemia, and liver fibrosis can occur after arterioportal shunting, but the long-term outcomes of arterioportal shunting have been reported to be acceptable when post-embolization management was adequate [2, 13,14,15]. Our patient also developed refractory ascites associated with portal hypertension, and embolization of the shunt rapidly resolved it without any problem. Thus, early embolization of arterioportal shunting after confirmation of arterial inflow to the liver should be performed to prevent morbidity induced by portal hypertension, and arterioportal shunting should be a temporary and the final option in an emergency situation.
In our patient, the cause of the pseudoaneurysm was not known. Notably, our patient did not develop bile leakage, a pancreatic fistula, or an intra-abdominal abscess; thus, development of the pseudoaneurysm was not influenced by pancreatic juice, bile, or infection. The inflammation around the LHA was severe due to preoperative cholangitis, and dissection of the LHA was difficult. Therefore, we postulate that a retraction injury or thermal injury by the electrocautery device intraoperatively may have occurred.
The patient in our case was asymptomatic, and dynamic CT revealed the pseudoaneurysm. It is vitally important to discover an asymptomatic aneurysm and perform a swift intervention to prevent mortality. Therefore, screening CT or ultrasonography is recommended after major hepatectomy with extrahepatic duct resection to check for the presence of an aneurysm.