Congenital hepatic arterioportal fistula.

Abstract

To the Editor: The report by Marchand et al. (1) of a 3-year-old boy with a congenital hepatic arterioportal fistula was accompanied by a detailed literature review of this rare and challenging problem. However, the authors did not include one previously reported case of an infant with the potentially lethal combination of congenital arterioportal fistula associated with type 3 extrahepatic biliary atresia (2). This infant underwent a successful Kasai portoenterostomy with eventual clearance of jaundice, but a large, congenital intrahepatic arterioportal venous fistula centred in segment 5 of the right lobe of the liver resulted in intractable portal hypertension. The fistula could not be controlled adequately by repeated coil embolisation. Subsequently, deteriorating liver function was associated with the development of ascites and gastrointestinal bleeding from portal hypertension. He underwent a successful liver transplantation at Birmingham Children's Hospital at the age of 2 years and remains well 2 years later. We have also treated another patient with biliary atresia associated with a congenital arterioportal venous fistula. This boy had a successful Kasai portoenterostomy at 8 weeks of age in 1989, but gastrointestinal bleeding developed within 2 weeks of surgery, thought to be related to an occluded portal vein. An aortoportogram did not demonstrate flow in the superior mesenteric, splenic, and portal veins. In 1994, he was referred to King's College Hospital, London, for consideration of liver transplantation. Further evaluation of his venous anatomy by selective angiography revealed retrograde filling of the portal vein through a hepatic arterioportal fistula. This was confirmed by a percutaneous transhepatic portogram. The fistula was embolised with coils, which reduced the retrograde flow in the portal vein. However, his portal venous system had become arterialised by this stage and, during the next year, the portal enteropathy did not improve. A liver transplantation was performed, but he has continued to have problems from unresolved portal hypertension. Congenital hepatic arterioportal fistula occurring in conjunction with extrahepatic biliary atresia is particularly hazardous. Any attempt to formally ligate the hepatic artery, which may be an effective treatment for lesions with bilateral arterial supply, is likely to be fatal in infants with biliary atresia who are abnormally dependent on arterial rather than portal venous hepatic inflow (3). Liver transplantation is the only solution in this situation, and it should be considered before the portomesenteric venous system becomes arterialised. Otherwise, this will jeopardise the outcome. M. D. Stringer P. McClean N. D. Heaton J. Karani G. Mieli-Vergani Departments of Paediatric Surgery and Hepatology; Leeds Teaching Hospitals Trust; Leeds, United Kingdom; Departments of Paediatric Hepatology and Radiology, and the Liver Transplantation Unit; King's College Hospital; London, United Kingdom