Biliary leaks are known complications of cholecystectomy, hepatobiliary interventions, and liver transplantation. Treatment via surgical intervention, percutaneous drainage, or endoscopic therapy is widely accepted. In the setting of a failed surgical intervention and continuous biliary leakage (specifically in cases involving an isolated ductal system contributing to the bile leak), surgical re-exploration can be difficult because of adhesions, edema, or infections. Therefore, percutaneous therapies such as biliary stenting and biloma drainage have become acceptable options in such scenarios. When the percutaneous diversion of biliary drainage is not an option, isolated biliary ductal system ablation is a valid treatment alternative.
We report a case in which a continuous biliary leak from an isolated segmental biliary system in a transplant liver after bile duct revision was successfully treated with N-butyl cyanoacrylate (NBCA) ablation. Our patient had a history of nonalcoholic steatohepatitis, end-stage liver disease, and hepatocellular carcinoma and underwent orthotopic liver transplantation without immediate complications. The bile duct anastomosis was performed in a duct-to-duct fashion with a donor duct with a low confluence of the right and left hepatic ducts. The initial posttransplant Doppler sonogram of the liver was unremarkable with hepatic resistive indices of 0.73 to 0.87 that continued to be within the normal range. However, 2 weeks after transplantation, he developed a nonanastomotic stenosis in the segmental right hepatic lobe duct and a high-grade nonanastomotic stenosis at the confluence of the right and left hepatic ducts for which bilateral percutaneous biliary drains were placed (Fig. 1A,B). The biliary stricture was thought to be due to the fact that the bile duct anastomosis was performed with a donor duct with a low confluence. Because the stenoses were refractory to repeat biliary plasty, the patient underwent Roux-en-Y hepaticojejunostomy for 4 bile ducts, with a single anastomosis sewn with running 5-0 polydioxanone sutures to the hepatic plate. After the hepaticojejunostomy, the patient developed a subhepatic biloma (Fig. 1C). Initially, the biloma was percutaneously drained for a week. A bilateral antegrade sheath cholangiogram via 2 percutaneous access points demonstrated a patent hepaticojejunostomy and no evidence of a biliary leak (Fig. 2A). A 1-week follow-up study showed an improved biloma cavity with direct communication of the biloma with an isolated segmental (segment VIII) hepatic duct system (Fig. 2B). The isolated biliary ducts were retrogradely catheterized. A Kumpe catheter was advanced through the biloma cavity over an angled glide wire to the excluded hepatic ducts. Several subsegmental bile duct branches were catheterized with a microcatheter (3-Fr mass transit catheter, Cordis, Miami Lakes, FL). The catheters and microcatheters were flushed with a nonionic dextrose solution. A 1:3 mixture of NBCA (Trufill, Cordis) and Ethiodol (Savage Laboratories, Melville, NY) was injected under fluoroscopic guidance in a pullback fashion. The rate of microcatheter retraction depended on how fast the biliary duct was filled with the glue solution (the 1:3 mix of NBCA and Ethiodol). The ablation was performed in 2 staged injections. The final fluoroscopy image demonstrated an NBCA/Ethiodol cast within the excluded biliary ducts and complete resolution of the biloma cavity (Fig. 2C). Therefore, the drainage catheter was removed. The night after the bile duct embolization, the patient had a transient low fever and slightly abnormal liver function test results. Initially, there was an increase in the alkaline phosphate level, which normalized. No change in the bilirubin level was observed. At the 1-year clinical follow-up for postdrainage catheter removal, the patient was doing well without evidence of cholangitis, and the results of liver function tests continued to be within normal limits.
Biliary leaks due to a transected duct disconnected from the main duct may require surgical intervention, fibrin glue occlusion, or acetic acid ablation.[1, 2] Fibrin, acetic acid, ethanol, and NBCA have been used for biliary tree ablation.[3-7] Several agents, including a combination of coils, vascular plugs, and liquid agents (Onyx and NBCA), have been used to embolize biliary fistulas and biliary leaks. However, the goal of ablation is to eliminate the biliary duct epithelium and biliary secretion with a subsequent decrease in the blood perfusion of the affected segment and atrophy. The use of ethanol as an ablating agent causes necrosis of the bile duct epithelium and requires multiple staged injections in order to accomplish complete biliary tree occlusion. Ethanol is complicated by the induction of fever, leukocytosis, and systematic hypotension at the time of the injection. Fibrin ablation does not permanently occlude the biliary tree, and other approaches may be required.
NBCA has been advocated as a safe and successful method of biliary ablation.[5-9] NBCA is a tissue adhesive and a permanent liquid embolic agent. NBCA or glue is diluted with Ethiodol. Ethiodol is used to opacify the embolic agent and to slow the polymerization time. The time of polymerization depends on the glue concentration in the NBCA/Ethiodol mixture. Embolization with NBCA occurs as a result of the polymerization of the cyanoacrylate agent in an ionic environment such as normal saline, blood, or bile. NBCA is flushed with a 5% dextrose nonionic solution to prevent the premature contact of the glue mixture with the ionic fluid and thus premature polymerization. NBCA induces an inflammatory exudate, and this is followed by granulation tissue that grows around the fragments of the polymer and results in luminal occlusion, scarring, and tissue atrophy.
Our patient presented with a bile leak that originated from an isolated segmental hepatic duct system. We believe that this isolated duct was missed at the time of the hepaticojejunostomy partly because the hepaticojejunostomy was performed high up in the hilum with multiple ducts. A ductal variant is another possible etiology, but this was not thoroughly investigated. The purpose of the NBCA injection was to ablate the biliary epithelium (without causing luminal narrowing) and the isolated systems. This is the reason that we selected the isolated ducts with a microcatheter very peripherally and injected the glue in a pull-through fashion; this promoted ablation instead of luminal narrowing. This, in turn, decreased the risk of cholestasis, cholangitis, and abscess formation. It was important to achieve a continuous column of glue in order to prevent recanalization at a later stage and provide a definitive treatment.
Possible complications that may arise from the NBCA ablation of a biliary tree are the need to re-ablate in case of failure, cholestasis, cholangitis, and abscess formation. Our patient did not exhibit any of these complications and will be followed routinely at the clinic with regular liver function assays.
The use of NBCA as an ablating agent of isolated biliary ducts in liver transplantation appears to be a safe and cost-effective approach. Although in this case NBCA was used in an isolated ductal system, it can be used to ablate nonisolated bile ducts as long as this is done under strict fluoroscopic control with the additional use of a coil or vascular plug to prevent the reflux of glue into nontargeted areas. Overall, this approach is less invasive than hepatic resection and can serve as a permanent alternative to chronic external biliary drainage.
Mehrzad Zarghouni, M.D.1
Marco Cura, M.D.2
Peter T. W. Kim, M.D.3
Giuliano Testa, M.D.3
Goran B Klintmalm, M.D.3
Departments of 1Diagnostic Radiology, 2Interventional Radiology, and 3Transplant Surgery
Baylor University Medical Center