Liver failure is a fatal disease. Liver transplantation is the only established treatment for liver failure; however, donor shortages remain problematic. In the United States and Europe, artificial livers as a bridge to liver transplantation are being considered. In Japan, we have taken a different approach to the treatment of end-stage liver diseases because of the characteristics of the health-care insurance system, regulated by the government. Furthermore, cadaveric liver transplantations are unsuited to the social mores of Japanese culture. Practically speaking, we believe that plasma exchange (PE) and continuous hemodiafiltration (CHDF) are the most effective therapies for the treatment of liver failure, although randomized controlled studies are needed to determine their effects. Overall, we believe that the first line of treatment for liver failure should be PE and CHDF, and the second line should be bioartificial liver support. In the near future, we hope that both gene therapy and regenerative medicine will contribute to the development of a functional artificial liver.

Liver transplantation and blood purification therapy, including plasmapheresis, hemodiafiltration, and bioartificial liver support, are available to treat patients with severe liver failure. The two mainstream systems developed for bioartificial liver support are extracorporeal whole-liver perfusion (ECLP) and the bioreactor system (BIS). We developed a method of plasma cross-perfusion, in which plasma is exchanged between the blood circuit of the patient and that of a hepatic function unit, that is, whole liver or a bioreactor through which immunologically free whole human blood is perfused. From the aspects of efficacy and safety, the best system of bioartificial liver support for clinical use is considered to be ECLP in cross plasma perfusion. However, a social objection about zoonosis has consistently been raised, with controversy surrounding the use of xenogeneic organs for human treatment, and this might be a final obstacle to the development of system efficacy. The combination therapy of hemodiafiltration with the administration of human serum albumin and anticoagulant factors can minimize the economic and medical resource costs through the development of transgenic livestock that secrete human pharmaceuticals systemically. It is possible that this therapy will become the most practical treatment for patients with severe hepatic failure.

To treat patients with severe liver failure, liver transplantation and blood purification therapy, including plasmapheresis, hemodiafiltration, and bioartificial liver support, are available. The two mainstream systems developed for bioartificial liver support are extracorporeal whole liver perfusion (ECLP) and the bioreactor system (BIS). We developed a method of cross-plasma perfusion, in which plasma is exchanged between the blood circuit of the patient and that of a hepatic functioning unit, through which immunologically free whole human blood is perfused. From the aspects of efficacy and epidemic safety, the best system of bioartificial liver support for clinical use is considered to be ECLP in cross-plasma perfusion. In opposition, a social antagonist for zoonosis has consistently been raised, with controversy surrounding the use of xenogeneic organs for human treatment, which might be final obstacle. It is possible that the combination therapy of hemodiafiltration and the administration of human serum albumin and anticoagulant factors, which minimizes the economic and medical resource costs through the development of transgenic livestock that secrete human pharmaceuticals systemically, will become a more desirable and practical treatment for patients with severe liver failure.

We developed a new method of xenogeneic direct hemoperfusion of a bioartificial liver support system consisting of a leukocyte-adsorbent column, an immunoglobulin-adsorbent column, and the substitute unit for hepatic function. By this method, we performed xenogeneic direct hemoperfusion experiment using resected whole swine liver for treatment of a canine liver failure model, and compared the contribution of each adsorbent column both by blood analysis and from the histological point of view.

Canine liver failure model was produced by portocaval shunting and ligating the entire hepatoduodenal ligament. The xenogeneic direct hemoperfusion system was constructed using a roller pump, a leukocyte-adsorbent column, an immunoglobulin-adsorbent column, a combined device of oxygenator and warmer, the resected whole swine liver accommodated in a chamber, and a dissolved oxygen meter through which canine whole blood leaving the external jugular vein circulated in this order.

Xenogeneic direct hemoperfusion was successfully performed for 3 h without hyperacute rejection occurring. Adequate ammonia detoxification and bile juice secretion were exhibited, and no findings of hepatocyte destruction by immunological cells and proteins were detected. Blood data showed that the immunoglobulin adsorbent were more effective than the leukocyte adsorbent in avoiding hyperacute rejection. This result indicates that hyperacute rejection has a closer relation to humoral immune responses, especially regarding removal of complements than to cellular immune responses.

We successfully performed xenogeneic direct hemoperfusion of the whole swine liver without hyperacute rejection using our method.

Despite more than 30 yr of research and development, an artificial liver has still not yet become clinical reality. Although previous attempts using a multiplicity of techniques including hemodialysis, hemoperfusion, plasma exchange, extracorporeal perfusion, and crosshemodialysis have shown minor improvement in patients with acute hepatic failure, limited clinical trials have failed to demonstrate any survival benefit. Encouraged by the progress on techniques that maintain long-term cultures of hepatocytes, more recent efforts have been directed at the use of hepatocytes as the basis of liver support. This review takes a critical look at past and present concepts in the development of artificial liver supports and both qualitatively and quantitatively evaluates the advantages and disadvantages of the available methodology.

The effects of a prostacyclin analog OP-41483 on energy metabolism were studied in an isolated porcine liver perfused with human blood for 8 h. OP-41483 was administered intravenously at a rate of 0.3 microgram/kg/min during the procurement and into the perfusate at a rate of 1.0 microgram/min during perfusion. Acetoacetate, beta-hydroxybutyrate, lactate, and pyruvate were measured before perfusion and at 1, 2, 3, 5, and 8 h after perfusion, from which values the ketone body ration (acetoacetate/beta-hydroxybutyrate, KBR), reflecting the redox state of liver mitochondria, was calculated. In the OP-41483 group, KBR increased rapidly from 0.34 to 0.95, 1.61, 1.51, 2.35, and 2.04, and lactate decreased rapidly from 9.81 to 6.30, 4.51, 3.22, 2.39, and 1.33 mmol/L at the respective hours after perfusion. There were significant differences after 3 h of perfusion as compared with the control group (p less than 0.05). These results suggest that administration of OP-41483 causes an increase in mitochondrial NAD+/NADH ratio (oxidized and reduced forms of free nicotinamide-adenine dinucleotides), leading to an enhancement of the metabolic capacity of the perfused liver.

Using a modified alginate-polylysine membrane, we have successfully encapsulated rat hepatocytes with little loss of viability. Urea and albumin release from encapsulated liver cells was comparable to that from non-encapsulated cells during the first 4 days in culture. Histological studies also showed that more than 50% of the encapsulated hepatocytes remained viable 35 days after implantation in the peritoneal cavity of both normal rats and rats with galactosamine induced fulminant hepatic failure. Transplantation of microencapsulated hepatocytes provides a potential clinical treatment for liver failure.

The present study emphasizes the principle of using liver support to restore the blood ketone body ratio (acetoacetate/beta-hydroxybutyrate), which reflects the redox potential of liver mitochondria and correlates with hepatic energy charge (ATP + 0.5ADP/ATP + ADP + AMP). Eleven surgical patients with grade IV hepatic coma were treated by an ex vivo pig or baboon liver cross-hemodialysis with an interposed Cuprophan membrane when their blood ketone body ratios had decreased to below 0.4 compared with the normal of above 0.7. Three patients were treated by cross-hemodialysis using a standard Cuprophan membrane dialyzer without increase of blood ketone body ratio and without marked beneficial effect. However, five of eight patients who had blood ketone body ratios of above 0.25 became fully alert after treatment by cross-hemodialysis using the larger pore size and greater surface area Cuprophan membrane, concurrent with a rise in the decreased blood ketone body ratio, and three of them were later discharged. By contrast, in the three patients with blood ketone body ratios below 0.25, there was no restoration of consciousness and no improvement in their blood ketone body ratios by this liver support. It is suggested that, as long as the blood ketone body ratio remained over 0.25, this metabolic liver support is effective in restoring grade IV hepatic coma.