Transplantation

The liver performs numerous functions that are essential to maintain metabolic homeostasis, such as oxidative detoxification, biotransformation, excretion, protein and macromolecule synthesis, immune and hormonal modulation (¹).

Therefore, acute liver failure (ALF) is a devastating clinical syndrome occurring in approximately 2000 cases per year in the United States and is associated with a mortality reaching 80% (²). A particularly severe form of hepatic failure—fulminant liver failure (FLF)—was defined as the appearance of encephalopathy less than 8 weeks after onset of symptoms and occurs in patients without previous history of liver disease (³). The majority of cases are viral in origin or secondary to drugs or toxins (⁴). The most frequent causes of death in ALF or FLF are intracranial hypertension secondary to brain edema, infection, or multiple organ failure.

Acute-on-chronic liver failure (AoCLF) represents an acute onset or acute deterioration of liver function in patients with chronic liver disease. It is caused by a diet rich in protein, gastrointestinal hemorrhage, obstipation, use of benzodiazepines, or narcotics. In case of AoCLF, causes of death are mostly secondary to sepsis and multiple organ failure. In these forms of liver failure, regeneration of native liver is impaired by accumulation of toxins, such as ammonia or nitric acid (⁵).

Management of ALF remains supportive and is an interdisciplinary challenge. Emergency whole organ liver transplantation is currently the only effective therapy for those patients who are unlikely to recover spontaneously with survival rates of 70% to 85% (⁶). However, shortage of deceased human organ donors limits the number of possible liver transplantation and is responsible for approximately 15% mortality in patients on waiting list. Despite resourceful use of donor livers through split liver transplantation and living related donors, patient’s needs are not being met.

Treatment of chronic end-stage liver failure and various metabolic liver diseases is also based on whole liver transplantation, but scarcity of organ donors and invasiveness of procedure limit this approach.

Alternative therapies for treatment of liver failure are currently being developed and explored: (i) extracorporeal artificial liver devices based on detoxification using membranes and adsorbents; (ii) bioartificial liver (BAL) devices using hepatocytes, and designed to provide synthetic functions in addition to detoxification; and (iii) hepatocyte transplantation based on implantation of hepatocytes into the patient (^7–9).

The objective of artificial, BAL support devices, and hepatocyte transplantation is to bridge patients with liver failure until a suitable liver allograft is obtained for transplantation or the patient’s own liver regenerates sufficiently to resume normal function.

In this review, we discuss these three strategies to treat of liver failure and summarize the current status of clinical experience.

Extracorporeal Artificial Liver Support

Most early attempts to provide temporary liver support for ALF involved artificial systems based on plasma detoxification for water-soluble and protein-bound toxins that accumulate during liver failure. These products can lead to multiple organ failure and inhibit regeneration of native liver. Various techniques for artificial detoxification have been evaluated clinically.

Albumin Dialysis

Albumin binds toxins that have been postulated to contribute to the pathogenesis of liver damage, hemodynamic instability, and multiple organ failure. These toxins include fatty acids, bile acids, tryptophan, bilirubin, aromatic amino acids, and nitric oxide (¹⁰). The ability to bind such molecules forms the basis for the use of albumin as a dialysate in an attempt to treat ALF or AoCLF. Albumin dialysis is the basis of two systems: (i) molecular absorbent and recirculating system (MARS); (ii) Prometheus system. Clinical trials using albumin dialysis are summarized in Table 1.

Molecular Absorbent and Recirculating System

This system combines a 20% albumin dialysis solution containing a membrane with a cutoff of 50 kDa, sorbent columns (charcoal and anion exchangers), and a conventional dialysis unit. MARS allows eliminating protein-bound toxins and water-soluble products. The patient’s blood passes through the MARS membrane and is cleaned sequentially by conventional hemodialysis removing water-soluble substances and adsorber columns containing activated charcoal and anion exchange resin removing most of the albumin bound substances (^11–14).

Since 1993, several thousand patients have been treated with the MARS device for liver failure of various etiologies, for example, advanced malignancy (¹⁵), acute alcoholic hepatitis (¹⁶), drug-induced liver failure (¹⁷), primary nonfunction after liver transplantation (^18–22), cytotoxic mushroom poisoning (²³), septic multiple organ dysfunction (²⁴), ALF (^{17–19, 25–29}), and AoCLF (^{17–18, 29–38}). Results have shown consistent improvement of biochemical parameters, decrease of intracranial pressure (^{39, 40}), and serum levels of inflammatory cytokines, such as tumor necrosis factor-α and interleukin-6 (⁴⁰).

In an initial phase 1 trial using MARS for patients with ALF, Awad et al. (⁴¹) showed a significant decrease of ammonium and creatinine, intracranial pressure, and severity of encephalopathy. Increase of serum factor VII and albumin was observed. However, no modification of hemodynamic status was observed. Of the nine patients, three were bridged to transplantation and one regained full liver function.

A controlled study included 18 patients suffering of AoCLF, who were randomized to receive standard medical therapy (SMT) alone or combined with MARS over 7 days. This study was designed to determine the effect of MARS on the cytokine profile, oxidative stress, nitric oxide, and ammonia. Results showed a significant improvement of encephalopathy in patients treated with MARS compared with the control group, but no changes on mean arterial pressure, renal function, cytokine level, and ammonia were observed (³⁰).

In a randomized trial, Mitzner et al. (⁴²) treated 13 patients presenting AoCLF and hepatorenal syndrome by SMT or by MARS. All patients had multiorgan failure and a Child-Pugh Score more than or equal to 12. Mortality at 7 days was 100% in the control group (n=5) compared with 62% in MARS group (P<0.01). Authors observed improvement of serum level of sodium, creatinine, bilirubin, prothrombin time, but no differences in albumin, urine output, or mean arterial pressure. Mean survival time was 25.2 days in the MARS group and 4.6 days in the control group (P<0.05).

Heeman et al. (⁴³) performed a randomized trial, including 24 patients presenting AoCLF and marked hyperbilirubinemia (bilirubin >340 μmol/L), who were randomized to be treated by SMT alone (n=12) or combined to MARS (n=12). Serum levels of bilirubin, bile acids, creatinine decreased and mean arterial pressure and encephalopathy improved in the MARS group compared with the control group. Use of MARS was associated with a significant improvement of survival at 30 days (11 of 12 vs. 6 of 12, respectively).

Report from the International MARS Registry, describing treatment of 176 patients who underwent treatment with MARS demonstrated that MARS did not led to severe adverse effect and was beneficial in the treatment of encephalopathy (⁴⁴).

A meta-analysis reviewed out of four randomized and out of two nonrandomized trials and showed that one randomized trial and two nonrandomized trials demonstrated a significant survival benefit (⁴⁵).

Prometheus System

The Prometheus system combines fractional plasma separation and an adsorption method using various adsorbents (neutral resin and an anion exchange resin) with high-flux hemodialysis using an albumin semipermeable membrane with a cutoff of 250 kDa. Prometheus aims improved purification of albumin-bound toxins and water-soluble low-molecular weight substances. Recently, Rifai et al. (⁴⁶) reported the first clinical trial using the Prometheus system in 11 patients with AoCLF and hepatorenal syndrome and showed a significant decrease in serum levels of bilirubin, bile acids, urea, creatinine, ammonia, cholinesterase, and blood pH but no improvement of neurologic status. A drop in blood pressure in two patients and uncontrolled bleeding in one patient were the observed adverse events.

In 2006, Laleman et al. (⁴⁷) compared MARS, Prometheus system, and SMT in patients presenting AoCLF. Eighteen patients were randomized to receive SMT alone (n=6), or combined with MARS therapy (n=6) or with Prometheus (n=6) on three consecutive days in a randomized controlled study. Groups were comparable for baseline hemodynamic status and levels of vasoactive substances. Both MARS and Prometheus decreased serum levels of bilirubin (P<0.005 vs. SMT), Prometheus device being more effective than MARS (P<0.002). Only MARS showed significant improvement in the mean arterial pressure (P<0.05) and in systemic vascular resistance index (P<0.05), whereas the cardiac index and central filling remained constant. This circulatory improvement in the MARS group was paralleled by a decrease in plasma renin activity, and serum levels of aldosterone, norepinephrine, vasopressin, and nitrate/nitrite. Stadlbauer et al. (⁴⁸) have demonstrated that both MARS and Prometheus cleared cytokines from plasma, but there were no changes in serum cytokine levels. This discrepancy was probably due to a high rate of cytokine production in patients with AoCLF.

Extracorporeal Bioartificial Liver Support

As artificial methods have shown limited benefit on patient survival because of their inability to replace synthetic and metabolic functions (⁴⁹), hybrid BAL devices using hepatocytes have been developed. Clinical trials using extracorporeal liver support are summarized in Table 2.

Extracorporeal Bioartificial Liver Support

This device is composed by a bioreactor loaded with hepatocytes which is perfused with patient’s plasma or blood. This system consists of a hollow-fiber cartridge (bioreactor), similar to a hemodialysis device, containing numerous hollow fibers of semipermeable membranes. The fibers divide two compartments, an intramembrane space, which is perfused with plasma or blood, and an extramembrane space loaded with hepatocytes. This membrane allows passage of blood substances of low molecular weight. This device can be placed extracorporeally or internally. It has been demonstrated that key metabolic functions, which are impaired or lost during severe liver failure, might be provided by isolated hepatocytes, such as protein synthesis, ureagenesis, conjugation, and detoxification (cytochrome P450 activity) (⁴⁹). Some studies have demonstrated that hepatocytes remain viable and functional in these systems for long periods of time and that bio-livers are able to provide metabolic and physiologic support in animal models of liver failure (^50–54). These bioartificial devices are based on various cell sources, culture techniques, and types of bioreactors.

Extracorporeal Liver Assist Device

Introduced by Sussman et al. (⁵⁵), extracorporeal liver assist device (ELAD) is based on a conventional hollow-fiber bioreactor loaded with the cell line C3A derived from a human hepatoblastoma (HepG2) chosen for its high albumin production and its capacity to grow in a glucose-deficient medium.

Ellis et al. (⁵⁶) performed the first randomized, controlled clinical trial, in United Kingdom, including 24 patients with ALF, 12 treated with ELAD, and 12 controls, comprising 17 patients not fulfilling criteria for transplantation (predicted survival 50%) and seven patients fulfilling criteria (predicted survival <10%). Arterial ammonia decreased marginally in the ELAD-treated group and rise in serum bilirubin was more pronounced in the control group. Encephalopathy was decreased in ELAD-treated patients but there was no benefit on patient survival.

Millis et al. (⁵⁷) performed a randomized controlled trial in patients with FLF. Patients were separated in two groups. Eleven of 12 (92%) of ELAD-treated patients received a liver graft, whereas only three of seven (43%) of controls were transplanted (P<0.05). Ten of 12 (83%) of ELAD patients achieved the primary endpoint of 30-day survival compared with 3 of 7 (43%) of control group (P=0.12). In conclusion, there was a significant advantage for patients receiving liver transplantation in the ELAD group.

The HepatAssist

Demetriou et al. (⁵⁸) developed a device combining a perfusion circuit containing cellulose-coated activated charcoal columns and hollow-fiber modules with porcine hepatocytes on microcarriers in the extracapillary space.

The system was evaluated in a large randomized trial including 171 patients presenting FLF or primary nonfunction of liver transplant. Patients were randomized in a 1:1 ratio receiving standard therapy or HepatAssist liver support system in addition to SMT. The primary endpoint of the study was patient survival, with or without liver transplantation. The 30-day patient survival rates were 62% in the control group vs. 71% in the group treated with HepatAssist, a difference that did not reach statistical significance (P=0.26). However, when adjusting for the impact of liver transplantation, BAL-treated patients had a statistically significant improved survival compared with controls receiving standard medical care. Furthermore, HepatAssist therapy reduced the risk of pretransplant death by 67% in patients with drug toxicity and by 47% in patients with rapid onset of FLF (P<0.0428) (⁵⁹).

The Amsterdam Medical Center-Bioartificial Liver

Developed by Chamuleau and coworkers, The Amsterdam Medical Center-bioartificial liver consists of a hollow-fiber bioreactor (polysulfon housing) and a plasmapheresis system using 10 to 14×10⁹ porcine hepatocytes attached in a three-dimensional configuration. This device was used in a pilot study of 12 patients performed in Italy. Kerkhove et al. (^{60, 61}) successfully bridged 11 patients to liver transplantation and one patient survived after two treatments without a transplant. Four patients died within a month after transplantation. Treatment of all patients was tolerated and associated with improvement in neurologic and hemodynamic status and lowering of plasma bilirubin and ammonia levels. The only adverse effect observed was transient hypotension in two patients.

Modular Extracorporeal Liver Support

The Modular Extracorporeal Liver Support system was introduced by the group of Gerlach in 1996. This system utilizes either porcine hepatocytes or human hepatocytes (⁶²). For porcine hepatocytes, Sauer et al. (⁶³) recently reported a phase I clinical trial including eight patients with ALF. The procedures were well tolerated by the patients, thrombocytopenia being the only adverse event encountered. All patients were successfully bridged to transplantation. Three patients were alive with a follow-up to 3 years. No patients showed any evidence of porcine endogenous retrovirus infection. For human hepatocytes, Sauer et al. (⁶⁴) treated eight patients using Modular Extracorporeal Liver Support loaded with cells isolated from organs unsuitable for transplantation. For all patients, improvement of neurologic status and blood coagulation parameters was observed.

Bioartificial Liver Support System

The BAL Support System developed at the University of Pittsburgh Medical Center uses primary porcine hepatocytes in a 100 kDa cutoff hollow-fiber bioreactor with whole blood perfusion. In a report of a phase I clinical trial including four patients with ALF or AoCLF, Mazariegos et al. (⁶⁵) showed that BAL Support System was well tolerated. The main adverse event during the procedures was hypotension responsive to fluid administration. Improvement of plasma bilirubin and ammonia levels was observed.

Radial Flow Bioreactor

The Radial Flow Bioreactor system uses a polyester screen with a cutoff 1 μm with a barrier filter of 0.4 μm and containing freshly isolated primary porcine hepatocytes (⁶⁶). In a phase I clinical trial, Morsiani et al. (⁶⁷) treated seven patients (four with ALF and three with primary nonfunction) during 24 hours. The results showed an improvement of neurologic status, plasma ammonia, and bilirubin levels. However, no improvement of survival was observed.

Hepatocyte Transplantation

Clinical hepatocyte transplantation has been performed for three indications: (i) ALF, (ii) chronic liver failure, and (iii) metabolic disorders. Clinical trials using hepatocyte transplantation liver support are summarized in Table 3.

Most cases of hepatocyte transplantation resulted in measurable benefits for patients suffering of acute or chronic liver failure or metabolic liver disease. However, randomized clinical trials are necessary to demonstrate a survival benefit.

Acute Liver Failure

Bilir et al. (⁶⁸) reported five patients presenting ALF with contraindication for liver transplantation who underwent intrasplenic allogeneic hepatocyte transplantation under cyclosporine immunosuppression. Two recipients received an additional intraportal infusion. Patients improved encephalopathy from grades IV to II or 0, prothrombin time, ammonium serum level, and cerebral edema. A team at the Medical College of Virginia performed allogeneic hepatocyte transplantation in nine patients with ALF. Seven patients were brought up successfully to liver transplantation or achieved regeneration of native liver. Two of nine patients died due sepsis due to herpes virus or to cardiopulmonary complications (⁶⁹).

Habibullah et al. (⁷⁰) transplanted intraperitoneally human fetal hepatocytes in patients suffering of FLF. They reported a decrease of plasma levels of ammonia and bilirubin and improved survival slightly.

Recently, Khan et al. (⁷¹) described a case of a patient diagnosed with acute fatty liver during pregnancy and encephalopathy grade IV. The patient received an intraperitoneal transplantation of human fetal hepatocytes and subsequently improved level of consciousness within 24 hours followed by full recovery within 7 days. Schneider et al. (⁷²) reported an intraperitoneal transplantation of cryopreserved human hepatocytes in a 64-year-old woman who accidentally ingested liver toxin–producing amanita phalloides mushrooms. Immunosuppression was started by intravenous application of steroids and cyclosporine A. After hepatocyte transplantation, the patient improved consciousness and levels of plasmatic ammonium and bilirubin. At 8 weeks, the patient fully recovered and an abdominal ultrasound showed a normal liver architecture and a normal portal blood flow.

Chronic Liver Failure

During the last decade, 20 patients with chronic liver diseases were transplanted with human hepatocytes (⁷³). Mito et al. (⁷⁴) performed the first clinical trial of hepatocyte autotransplantation including 10 patients. Hepatocytes were obtained from the recipients’s native cirrhotic left lateral liver segment and were infused into splenic artery by an open surgical procedure. They observed an improvement of encephalopathy in only two patients.

Metabolic Liver Disease

Twenty-one patients suffering of inherited liver enzyme deficiencies have been transplanted with hepatocyte infusions to correct their metabolic liver disease (⁷³). Fox et al. (⁷⁵) treated a 10-year-old girl with Crigler-Najjar syndrome type 1 by infusion through the portal vein of a volume of 7.5×10⁹ allogeneic hepatocytes representing approximately 5% of the normal hepatic mass to treat. Tacrolimus and corticosteroids were given as immunosuppressive treatment. Bilirubin levels decreased at day 35 and stabilized at 14.0 mg/dL even after 11 months after hepatocyte transplantation.

More recently, Horslen et al. (⁷⁶) transplanted 4×10⁹ allogeneic hepatocytes intraportally to treat a patient with ornithine transcarbamylase deficiency. Immunosuppression consisted of tacrolimus and corticosteroids. They observed a transient improvement of hyperammonemia and protein intolerance. Sokal et al. (⁷⁷) treated a patient with Refsum’s disease by transplantation of 2×10⁹ hepatocytes into the portal vein. Tacrolimus and basiliximab were given as immunosuppressive treatment. They observed a decrease of abnormal total bile acids and dihydroxycoprostanoic acid for 1 year.

Dhawan et al. (⁷⁸) reported hepatocyte transplantation in two brothers with severe inherited coagulation factor VII deficiency. The first patient received a total of 1×10⁹ cryopreserved hepatocytes, and the second patient received 2×10⁹ fresh and cryopreserved hepatocytes in the inferior mesenteric vein. These patients received tacrolimus and prednisolone as immunosuppression. Authors observed an improvement of the coagulation defect and a decrease of the requirement for exogenous recombinant factor VII (rFVIIa) to approximately 20% of pretransplantation need. However, within 6 months after transplantation, higher rFVIIa doses were again required, suggesting loss of transplanted hepatocytes and orthotopic liver transplantation (OLT) was performed in both cases.

Stéphenne et al. (⁷⁹) reported hepatocyte transplantation in a child with urea cycle disorder poorly equilibrated by conventional therapy as a bridge to transplantation. A 14-month-old boy with ornithine transcarbamylase deficiency received 0.24×10⁹ cryopreserved cells per kilogram over 16 weeks. Tacrolimus and steroids were given as immunosuppressive treatment. Mean blood ammonia level decreased significantly after the seven first infusions, whereas urea levels became detectable. After seven infusions, an ammonium peak up to 263 μg/dL, clinically well tolerated, was observed. Blood urea levels increased continuously to reach 25 mg/dL, after the last infusions. Finally, he benefited from elective OLT and the postsurgical course was uneventful. The authors concluded that the use of cryopreserved cells allowed short- to medium-term metabolic control and urea synthesis while waiting for OLT.

DISCUSSION

OLT is the only effective treatment for patients with ALF, AoCLF, and end-stage chronic liver diseases. However, because of scarcity of donor organs, alternative therapies have been developed. Numerous studies reported efficacy of artificial methods (MARS and Prometheus) for removing albumin-bound toxins and improving patient’s hemodynamics. Unfortunately, these improvements had no benefit on patient survival, probably due to their inability to replace synthetic and metabolic functions. The clinical utility of artificial liver devices is uncertain and over the last 5 years, few randomized controlled trials have been reported, and have not shown significant novelty in the field (^80–82).

Progress has been achieved in the development of bioartificial extracorporeal systems (⁸³). Randomized controlled clinical trials have shown an improvement of neurologic status and biochemical parameters in patients with end-stage liver diseases, but failed to demonstrate significant survival benefit. Over the last years, one prospective randomized controlled international multicenter trial was reported in 2004 for patients with ALF and was not associated with improved survival (⁵⁹). Assessment of quantity of hepatocytes required to sustain life is complicated by the fact that severity and duration of ALF vary according to origin, but approximately 20% to 40% of normal hepatocyte mass is required. The number of cells used by various investigators in three BAL systems varied from 0.5×10⁸ to 1×10¹¹, representing approximately 1% to 30% of normal hepatocyte mass. The difference between the need of hepatocyte mass during liver failure and the number used in bioartificial devices could explain partially the modest results obtained clinically with these systems.

In the field of hepatocyte transplantation, small clinical trials had limited success. For treatment of acute and chronic liver failure, these studies have shown improvement of neurologic status and biochemical parameters, but failed to show significant survival benefit, except one trial (^68–72,74). For treatment of liver-based metabolic diseases, hepatocyte transplantation was more successful. However, correction of metabolic defect was transient, suggesting a progressive loss of transplanted cells (^75–79).

In conclusion, clinical trials using artificial, bioartificial, and hepatocyte transplantation for patients with end-stage liver diseases showed improvement of neurologic status and biochemical parameters, but failed to improve survival. Several technical challenges remain for development of successful internal or external liver supports: for example, to develop bioreactors with stable long-term liver cell culture systems to enhance and stabilize high levels of liver-specific functions; to explore new sources of hepatocytes, such as fetal, embryonic, bone marrow cells, or reversibly immortalized cells; to find new implantation sites for hepatocyte transplantation; and to develop new biomaterials for encapsulation to allow long-term function.

Methods to prolong hepatocyte function have been developed in preclinical settings and include (i) addition of growth factors and hormones to culture medium, such as corticosteroids, cytokines, vitamins, or amino acids (⁸⁴); (ii) culturing hepatocytes in presence of extracellular matrix component such as Matrigel, a tumor-derived basement membrane-like gel (^{85, 86}); (iii) culturing hepatocytes between layers of type I collagen mimicking environment seen by hepatocytes in vivo (^{87, 88}); (iv) coculture of hepatocyte with nonparenchymal cells, such as fibroblasts, sinusoidal endothelial cells, stellate cells, biliary epithelial cells, which have been shown to improve hepatocyte viability and function (^{89, 90}); and (v) hepatocyte aggregates or spheroids, a three-dimensional cell–cell adhesion, such as microcarrier or microencapsulation within collagen or alginate-polylysine (⁹¹). The ideal device should maximize mass transfer, that is, allow nutrients and toxins from patient’s blood or plasma to reach hepatocytes and future bioreactors need to recreate this microenvironment to eventually assume functions of a normal liver.

REFERENCES