Bridging the gap: Advances in artificial liver support


  • Potential conflict of interest: Nothing to report.


Key Points

1. The goals of liver support therapy include the following:

• To provide detoxification and synthetic function during liver failure.

• To remove or reduce the production of proinflammatory cytokines to correct the systemic inflammatory response of liver failure.

• To stimulate the regeneration of the injured liver and increase the likelihood of spontaneous recovery.

2. There is a large unmet need for a liver support device because of the shortage of organs for liver transplantation and the risks of major surgery.

3. Liver support devices can be divided into 2 groups: purely mechanical artificial devices and cell-based bioartificial devices. Both provide detoxification, but bioartificial liver devices provide the option of synthetic function and biotransformation activities that are not possible with a purely mechanical device.

4. An abundant high-quality supply of human hepatocytes is not currently available for liver cell therapy. However, such a supply is essential for successful bioartificial liver therapy. Novel options are under development for the unlimited production of high-quality human hepatocytes. Liver Transpl, 2012. © 2012 AASLD.

This syllabus is divided into 4 general categories: the general goals of therapy, a description of specific artificial liver systems, a description of specific bioartificial liver systems, and areas for future improvement. Artificial and bioartificial liver support systems are used for the extracorporeal treatment of acute liver failure (ALF), which is a devastating disorder characterized by jaundice, encephalopathy, and coagulopathy in the setting of a previously normal liver. These devices have also been used to support patients with liver failure in the setting of established cirrhosis (so-called acute-on-chronic liver failure). Artificial liver support refers to purely mechanical devices, including albumin dialysis, whereas bioartificial liver support refers to devices with cellular components such as primary hepatocytes or hepatic cell lines. Artificial systems remove toxins by filtration and absorption, whereas bioartificial liver systems perform these functions along with biotransformation and synthetic functions of biochemically active hepatocytes.


ALF, acute liver failure; ECAD, extracorporeal albumin dialysis; ELAD, extracorporeal liver assist device; FAH, fumarylacetoacetate hydrolase; FDA, Food and Drug Administration; FPSA, fractionated plasma separation and absorption; MARS, molecular adsorbents recirculating system.


The incidence of ALF is approximately 2500 cases per year in the United States and is much higher worldwide.1, 2 The majority of patients with acute hepatitis progressing to ALF will die without liver transplantation. A common cause of mortality in this patient population is brain death secondary to cerebral edema and intracranial hypertension. The outcome of ALF depends on several variables, including the disease etiology, patient's age, extrahepatic manifestations, severity of encephalopathy, and availability of liver transplantation. Three subgroups of ALF can be defined according to the time between the onsets of jaundice and encephalopathy: hyperacute (≤7 days), acute (8-28 days), and subacute (29 days to 12 weeks). ALF subgrouping is useful for predicting morbidity, mortality, and the likelihood of spontaneous regeneration and thus can be used to determine the appropriateness of liver support therapy or the need for replacement therapy such as liver transplantation.

Currently, no satisfactory treatment for ALF exists other than liver transplantation. The only Food and Drug Administration (FDA)–approved extracorporeal therapy is the use of albumin dialysis [molecular adsorbents recirculating system (MARS), Gambro] for the treatment of drug overdose and toxicity. Otherwise, no extracorporeal devices have yet been approved for the support of patients suffering from liver failure in the United States. In addition, the outcomes of liver transplantation for ALF are significantly poorer than the outcomes of transplantation for chronic liver disease or metabolic liver disease because of the scarcity and urgency for donor organs. It is critical to consider other means of treatment for liver failure, including artificial and bioartificial liver support.


The goals of liver support therapy are to prevent the manifestations of liver failure and to bridge patients with liver failure to liver transplantation or allow time for the recovery of native liver function and thus avoid liver transplantation. To accomplish these goals, a liver support device should be able to remove toxic substances from the blood and provide temporary liver function until either transplantation or functional recovery occurs. Currently, artificial and bioartificial modalities exist and are being tested in the clinical arena. Artificial liver devices detoxify blood without the use of cellular material. As yet, no artificial support device has been shown to reliably reduce the mortality rate of patients with liver failure.3, 4 Therefore, many believe that the treatment of liver failure requires the provision of whole liver function, which would entail the use of cellular material.

Ideally, a bioartificial liver would provide detoxification while also assuming synthetic roles for the failing liver. The concept of biological replacement of hepatic function was first put forth more than 50 years ago, and researchers showed that fresh liver homogenate or intact parenchyma was capable of ureagenesis and other metabolic functions under extracorporeal conditions.5, 6 Since that time, numerous investigators have used hepatocytes from various sources in attempts to create a bioartificial liver. However, various factors, including the high cost, complexity, and difficulty associated with both obtaining and maintaining hepatocytes in a differentiated state, have prevented the successful application of bioartificial liver devices in clinical practice. With respect to cell-based therapy, several of these devices have been tested. Currently, the Vital Therapies extracorporeal liver assist device (ELAD) is in clinical trials in the United States. A variety of other bioartificial liver systems are in various stages of development.

The 3 primary goals of liver support therapy include (1) decreased waste accumulation, (2) the quenching of the systemic inflammatory response to ALF, and (3) the enhanced regeneration of the acutely injured liver. A successful liver support device will address each of these 3 impairments of liver failure. Kjaergard et al.7 systematically reviewed artificial and bioartificial support systems for both acute and acute-on-chronic liver failure. Data from 12 randomized trials and 483 patients were reviewed. Ten trials assessed artificial liver systems for the treatment of ALF and acute-on-chronic liver failure, whereas 2 assessed bioartificial liver systems for the treatment of ALF. This meta-analysis suggested a survival benefit of liver support devices in the setting of acute-on-chronic liver failure; however, no clear survival benefit was observed in the setting of ALF. Liver support devices also served as an effective treatment for hepatic encephalopathy with improved mental status in the majority of cases. There also was a trend toward bridging to successful liver transplantation, although this finding was not reproducible in all studies.


As for mechanical devices, the current options include plasma exchange and albumin dialysis. The 2 most frequently used forms of albumin dialysis are extracorporeal albumin dialysis (ECAD; eg, MARS) and fractionated plasma separation and absorption (FPSA; eg, Prometheus). The first clinical use of the ECAD system was in 1999, and the first use of FPSA was in 2003, so both are relatively new treatment systems. ECAD uses a membrane with a 70-kDa molecular weight cutoff and a dialysate flow rate of 200 mL/minute, whereas the FPSA system uses a 250-kDa membrane with a dialysis flow rate of 300 mL/minute. Because of the membrane's larger pore size, the treatment efficacy of FPSA appears greater than that of ECAD,8 and the treatment stability appears greater with FPSA versus ECAD.9 The hemodynamic benefit of therapy appears greater with ECAD,10 and complications from coagulopathy also appear to be fewer with the ECAD system versus the FPSA system.11

The most recent study of survival after FPSA therapy was reported in 2012 by Kribben et al.4 in Gastroenterology. This study included 145 patients with acute-on-chronic liver failure who were randomly assigned to combination therapy with FPSA and standard medical therapy (n = 77) or to standard medical therapy alone (n = 68). Patients were provided 8 to 10 rounds of therapy for a minimum of 4 hours each for 3 weeks. The primary endpoint of this study was survival at 28 and 90 days. In an intention-to-treat analysis, the probability of survival on day 28 was 66% for the FPSA group and 63% for the standard medical therapy group (P = 0.70); at 90 days, the survival rates were 47% and 38%, respectively (P = 0.35). The subgroup analysis suggests that positive benefits may be observed in the sickest patients (Model for End-Stage Liver Disease score > 30) and patients with hepatorenal syndrome. Subsequent studies are required to validate this preliminary observation. Notably, MARS was tested in a similar cohort of patients with acute-on-chronic liver failure in the Recompensation of Exacerbated Liver Insufficiency with Encephalopathy and/or Renal Failure (RELIEF) trial, but no benefit of therapy was observed.3 Prior studies of MARS have suggested a reduction in hepatic encephalopathy.12, 13 No significant survival benefit has been reported for a randomized, prospective cohort with either form of albumin dialysis. As stated earlier, MARS albumin dialysis (an ECAD system) has 510(k) approval for treating drug overdose and poisoning. A decision from the FDA regarding indications for this device in patients with cirrhosis and hepatic encephalopathy is being awaited. The MARS albumin device has also been used as a treatment for resistant pruritus from cholestasis. Using a visual analogue scale before and after treatment and 30 days after therapy, several case reports have indicated improvements in the severity of pruritus with therapy.14 Albumin dialysis is not yet FDA-approved as a treatment for severe pruritus, although an FDA-sponsored trial has been approved in the United States (S.L.N., unpublished data, 2012). The mechanism behind the benefit of reduced pruritus with albumin dialysis is unclear.


It is unlikely that the complex mechanism by which the liver ensures hemostasis can be replaced by means of a nonbiological detoxification system alone. Many investigators have concluded that biological tissue will be necessary for the treatment of ALF. A bioartificial liver that incorporates hepatocytes from various sources has the theoretical advantage of not only providing blood purification through dialysis but also providing the hepatocyte-specific functions that are lost with ALF. These include protein synthesis, ureogenesis, ureagenesis, and glucogenesis detoxification through P450 activity. Many cell types have been studied as means of providing bioartificial liver support. Upon an initial inspection, primary human hepatocytes appear to be the most appropriate cells for use in a bioartificial liver. They should, in theory, provide all the metabolic functions necessary to support an ALF patient with a limited immunological response. However, several factors have prevented the widespread use of primary human hepatocytes. Paramount is the fact that human hepatocytes are not as readily available as hepatocytes from other sources. In addition, although human hepatocytes readily regenerate in vivo, their regenerative capacity in vitro is limited. Therefore, human hepatocytes in culture have a decreased level of function because of the loss of gap junctions when they are cultured. This problem has been approached by the attachment of hepatocytes to various matrices to provide a parenchymal echotexture and stimulate cell interactions and clarity. Because of these obstacles to the use of primary hepatocytes and cultures, various alternatives to primary hepatocytes have been used. Immortalized cell lines of the C3A human hepatoblastoma line have been studied in vitro and used in a bioartificial liver device. Although C3A cells have the theoretical benefit of unlimited expansion in vitro and the capacity for high albumin production and growth in glucose-deficient media, there are many problems that will likely preclude their use in a therapeutically successful bioartificial liver. These immortalized cell lines do not express normal metabolic profiles (eg, ureagenesis) in comparison with primary hepatocytes. A recent study has shown that although C3A cells produce urea, they do not express significant levels of all urea cycle genes and, therefore, do not detoxify ammonia. Urea production appears to be due to the single step of arginase rather than the complete urea cycle.15 Although the transmission of oncogenic cells during therapy has not been observed, these factors have precluded the widespread acceptance of immortalized cells in clinical therapy. Porcine hepatocytes are an alternative cell option for bioartificial livers. Unlike immortalized cells, porcine hepatocytes are primary cells and maintain a higher level of differentiated metabolic function (including ammonia detoxification). Porcine hepatocytes are currently more readily available than human hepatocytes. However, ongoing concerns about xenozoonosis have limited the clinical acceptance of porcine hepatocytes in human therapy in some countries. No cases of xenozoonosis from a bioartificial liver or any pig cell therapy have been reported.


More than 30 different cell-based support devices have been reported since 1987, and more than 14 systems have been reported in clinical trials. However, none of these bioartificial liver systems have yet obtained FDA approval for the treatment of liver failure. In all, more than 400 patients have been treated with bioartificial liver systems.

HepatAssist by Circe was the first biologically based liver assist device to be tested on a large clinical scale. More than 200 patients were treated with this device, which employed a hollow-fiber extracorporeal bioreactor loaded with approximately 7 billion cryopreserved porcine hepatocytes. These cells were separated from the patient's blood by a macroporous hollow-fiber membrane with a pore size of 0.15 μm. Initial animal studies and a phase 1 trial showed promising results, including improvements in both clinical and biochemical parameters. A large, multicenter, phase 3, randomized controlled clinical trial was then conducted with patients with fulminant/subfulminant liver failure and patients with primary graft nonfunction. In all, 171 patients were enrolled in this study at 20 sites; 85 underwent treatment with the HepatAssist device.16 This study demonstrated favorable safety; however, it failed to demonstrate improved 30-day survival in the overall study population. When adjustments were made for patients undergoing liver transplantation, patients treated with the device had improvements in survival in comparison with controls receiving standard medical therapy alone (P < 0.05). Furthermore, the subgroup of bioartificial liver–treated patients with ALF of known causes had a significant survival advantage in comparison with controls with ALF of known causes who were not treated with a bioartificial liver (P = 0.009). Despite this significant survival benefit (which was identified in a post hoc subgroup analysis), the HepatAssist device was not approved by the FDA.


The bioartificial liver device with the largest clinical experience is the ELAD device by Vital Therapies, which uses multiple hollow-fiber cartridges loaded with the C3A human hepatoblastoma cell line.17 The most current version of this device combines conventional hemodialysis with filtration through 4 ELAD cartridges. Each ELAD cartridge contains approximately 200 g of cells. The Vital Therapies device has been tested in numerous clinical settings.18-20 One of the largest clinical trials of ELAD to date, which was conducted at a United Nations hospital in Beijing, used a 4-cartridge device with more than 400 g of CA3 cells.20 Transplant-free survival at 28 days was 84.1% for treated patients and 50.2% for the control group. At 84 days, transplant-free survival was 67.9% for treated patients and 43.9% for the control group. Both comparisons were statistically significant. This study has been reported in abstract form; the complete study is not yet available. More recently, the ELAD device was tested in the pivotal US Stabilization in Liver Failure trial. The results of this phase 2B study of the safety and efficacy of this ELAD human cell–based biological device in patients with acute-on-chronic liver failure due to either acute alcoholic hepatitis or acute decompensation of cirrhosis were not yet available at the time of this writing. It is hoped that the results of this trial will be made available before the publication of this article.

A number of cell-based systems are currently under development; these include the spheroid reservoir bioartificial liver, which is based on the suspension of multicellular aggregates of primary hepatocytes.21 The spheroid reservoir bioartificial liver device has minimal complexity in comparison with other devices and can achieve substantial cell doses in the range of 200 to 400 g per treatment. The primary advantage of this device is the fact that it uses primary cells with normal urea cycle activity, and it also has an option for continuous treatment. Spheroids are formed by a novel rocking technique that allows cell-to-cell collision after hepatocyte isolation. Isolated hepatocytes adhere to one another through a process involving the cadherin E surface adhesion molecule. These spheroid aggregates form multicellular, 3-dimensional structures with typical bile canaliculi and polarized hepatocytes. Hepatocytes have been maintained in spheroid cultures for up to 28 days with relatively stable gene expression and stable functionality.20

In the future, an abundant high-quality supply of primary hepatocytes will be needed for bioartificial liver support devices. A number of techniques, including the use of liver progenitor stem cells, have been identified for the production of hepatocyte-like cells. However, the complete differentiation of these progenitor cells in vitro has not yet been reported. It is possible that in vivo conditions are necessary for the full differentiation of liver progenitor cells. One option that has been reported for the production of primary hepatocytes is the metabolically defective and immunodeficient Fah−/−/Rag2−/−/Il2rg−/− (FRG) triple knockout.22 The robust expansion of primary human hepatocytes can be observed in the livers of these mice after transplantation and the slow withdrawal of the protective drug 2-(2-nitro-4-trifluoromethylbenzoyl)−1,3-cyclohexanedione (NTBC). More recently, even greater liver chimerism with up to 95% human hepatocyte repopulation in FRG mice has been reported after the transplantation of a higher dose of human hepatocytes (5 × 106 per mouse).

On the basis of the encouraging results with FRG mice, investigators at Oregon Health Science University and the Mayo Clinic have collaborated to produce a fumarylacetoacetate hydrolase (FAH)–deficient pig.23 This larger animal model of FAH deficiency may address the significant demand that exists for an abundant and routinely available high-quality source of human hepatocytes for therapeutic and diagnostic applications. The pig offers a 1000-fold scale-up for the production of primary human hepatocytes in comparison with mice. A strategy for the cloning and herd development of FAH-deficient pigs has been reported.24 The engraftment and expansion of human hepatocytes in pigs are low in part because of the xenogeneic immune response of the recipient. One option for allowing expansion without a cytotoxic immune response is the creation of a hyporesponsive state through the transplantation of the human cells in utero before maturation of the recipient's immune system.25, 26

In summary, despite notable achievements with artificial and bioartificial liver therapy in treating liver diseases and bridging patients to transplantation, shortcomings exist for both forms of therapy. The limited supply of primary hepatocytes remains a major barrier to the successful use of the bioartificial liver. As for liver progenitor cells as a source of hepatocytes, it is likely that the normal and complete differentiation of liver progenitor cells into mature hepatocytes requires an in vivo milieu. A critical analysis of the long-term safety, tolerability, and effectiveness of the engraftment of primary hepatocytes and/or stem cells in diseased livers is needed. The efficacy of novel cell-based treatments must be established in clinical practice. Clinical trials are needed to evaluate unresolved issues of extracorporeal liver cell therapies. These hurdles are likely to be overcome, and this will be associated with the wider acceptance and application of cell-based extracorporeal therapies for the treatment of liver failure. Meanwhile, parallel strategies for increasing organ donation are required in conjunction with cell therapies to combat liver disease and address the current shortage of support therapies for liver failure.