Uncoupling the systemic circulation and the Ex Vivo circuit during experimental isolated liver perfusion.

Performing an ex vivo liver perfusion as a transient liver support requires perfusing the liver with a flow of 1 ml/min per kg of liver, which could reach 25% of the cardiac output when a human liver is used. This high flow could be detrimental in patients with acute liver failure. Therefore, in an ischemic-induced liver failure pig model, we developed a circuit allowing low flows going out of and into the systemic circulation, whereas the flow going through the ex vivo liver is maintained at a high value. This was obtained by uncoupling the ex vivo circuit from the systemic circulation. Ex vivo liver perfusion was performed in a closed circuit with a flow averaging 1 ml/min per kg of ex vivo liver (700 to 800 ml/min, according to the weight of the livers we used). It was linked to the systemic circulation with input and output flows equal to that used during hemodialysis (200 ml/min). Compared with previously reported direct circuits, this perfusion system was well tolerated from a circulatory point of view. After the induction of ischemic liver failure, the ex vivo liver perfusion led to an increase in urea, branched amino acids to aromatic amino acid ratio, and fractional clearance of indocyanine green and galactose and to a decrease in ammonia and lactic acid. This system allowed the ex vivo liver to keep its clearing properties despite a low extracorporeal flow. It represents an extracorporeal circuit that could be used in place of the direct extracorporeal high-flow liver perfusion.

Liver function during extracorporeal whole liver perfusion in a pig model of acute ischemic liver failure.

The shortage of livers for transplant has renewed interest in the potential of temporary liver support such as extra corporeal whole liver perfusion. In an ischemic induced liver failure model we perfused an extra corporeal liver through only a portal vein and assessed the function of this ex vivo liver by using hepatic tests to estimate elimination as well as synthesis capacities. Acute liver failure was performed in five control pigs by a hepatic devascularization associated to an end to side portocaval shunt. In a treated group, 5 to 6 h after this hepatic devascularization, animals were connected to an extra corporeal liver perfused via the portal vein with blood withdrawn from the ischemic liver animal from its portal vein. Devascularization of the liver induced an increase in liver enzymes and ammonia, a drop in the ratio of branched chain amino acids to aromatic amino acids, and a decrease in blood urea and indocyanine green and galactose clearances. In treated animals, urea, amino acid ratio, and clearances increased after the ex vivo liver perfusion. In this group, mean bile production and mean liver oxygen consumption were 13.7 +/- 3.6 ml/h and 16.1 +/- 7.7 ml/min, respectively. In an acute ischemic liver failure pig model, an extra corporeal whole liver perfusion demonstrated detoxification properties as well as synthesis capacities.