Numerous laboratory studies have shown that hepatocyte transplantation may serve as an alternative to organ transplantation for patients with life-threatening liver disease. Because of the successes of experimental hepatocyte transplantation, institutions have attempted to use this therapy in the clinic for the treatment of a variety of hepatic diseases. Unfortunately, unequivocal evidence of transplanted human hepatocyte function has been obtained in only one patient with Crigler-Najjar syndrome type I, and, even then, the amount of bilirubin-UGT enzyme activity derived from the transplanted cells was not sufficient to eliminate the patient's eventual need for organ transplantation. A roadmap for improving patient outcome following hepatocyte transplantation can be obtained by a re-examination of previous animal research. A better understanding of the factors that allow hepatocyte integration and survival in the liver and spleen is needed to help reduce the need for repeated cell infusions and multiple donors. Although clinical evidence of hepatocyte function can be used to indicate function of transplanted hepatocytes, definitive histologic evidence is difficult to obtain. In order to assess whether rejection is taking place in a timely fashion, a reliable way of detecting donor hepatocytes will be needed. The most important issue affecting transplantation, however, relates to donor availability. Alternatives to the transplantation of allogeneic human hepatocytes include transplantation of hepatocytes derived from fetal, adult or embryonic stem cells, engineered immortalized cells, or hepatocytes derived from other animal species.
Orthotopic liver transplantation has proven to be effective in the treatment of a variety of life-threatening liver diseases; however, significant morbidity and mortality remains. In addition, the growing disparity between the number of donated organs and the disproportionately large number of patients awaiting transplantation has provided an impetus for developing alternative therapies for the treatment of liver failure (1). Novel strategies designed to increase the number of organs transplanted, such as the use of adult living donors, are not without significant risk to both the donor and recipient (2).
Extensive laboratory work, performed over the last three decades, and recent clinical studies suggest that hepatocyte transplantation may be useful in the treatment of metabolic liver diseases and as a bridge for selected patients with liver failure to transplantation. Hepatocyte transplantation has several practical and theoretical advantages over whole liver transplantation. While intact livers can only be transplanted within a short time after procurement, isolated liver cells may be cryopreserved for later use in emergencies (3). A single donor could potentially provide hepatocytes for several patients, and hepatocyte transplantation should not interfere with subsequent orthotopic liver transplantation, should it be needed, or liver directed gene therapy, should it become available. Like other minimally invasive medical procedures, hepatocyte transplantation may be performed in the future as an outpatient procedure and repeated with low morbidity, if necessary. Thus, morbidity, mortality, and initial costs are significantly less than those for whole organ transplantation.
Despite unequivocal evidence of function in some patients, the general efficacy of hepatocyte transplantation has been difficult to prove. Although many of the promises of hepatocyte transplantation are yet to be realized, the issues hindering the development of this therapy have been identified and are being addressed.
Individual hepatocytes are isolated by collagenase digestion of the liver and are usually transplanted fresh. Techniques for storage by cryopreservation are becoming increasingly successful, and following transplantation in rodents, cryopreserved hepatocytes form proliferating hepatocyte colonies within the liver substance (4). It is still not clear, however, whether cryopreserved hepatocytes either engraft or function as well as fresh hepatocytes.
The liver and spleen are the most reliable sites for hepatocyte engraftment and function. The peritoneal cavity may also be a site for transplantation of encapsulated or matrix-attached hepatocytes (5), but other ectopic sites appear to be far less favorable for hepatocyte engraftment. Hepatocytes can be seeded into the liver by portal vein infusion or by injection into the splenic pulp from which cells translocate to the liver through the splenic vein. Following injection through the portal circulation, transplanted hepatocytes integrate into the liver cords, leaving the hepatic architecture intact (6). Hepatocytes engrafted in the liver receive the benefits of exposure to portal nutrients, contact with other hepatocytes and nonparenchymal cells, proximity to paracrine factors and have the ability to secrete bile into the native biliary system. In rodent recipients, engrafted hepatocytes can survive and function throughout life. Between one and four percent of the hepatocyte mass can be transplanted into a structurally normal liver at any one time while producing only mild or transient increases in portal pressure (7).
The expanded extracellular matrix associated with liver cirrhosis increases the endothelial barrier to engraftment in the liver. However, transplanted hepatocytes can migrate into cirrhotic nodules and integrate into liver plates following intraportal infusion in rodents. Furthermore, transplanted hepatocytes express enzymes associated with normal liver function, such as glucose-6-phosphatase and glycogen, and are capable of significant expansion following transplantation, as long as there is no ongoing injury to the liver (8). The relative percentage of cells that enter the cirrhotic liver is most likely significantly less than the percentage entering the normal liver, but the data indicates that that transplanted hepatocytes that are resistant to the underlying disease could potentially repopulate a severely diseased cirrhotic liver. It is possible that such an approach could be applied to the treatment of Wilson's disease as an example. Several issues, however, may limit hepatocyte transplantation into the cirrhotic liver. Portal-systemic shunts will result in translocation of hepatocytes to the pulmonary circulation. While hepatocytes do not engraft in this location and are rapidly cleared, translocation of a large number of transplanted cells may produce pulmonary emboli with resultant cardiopulmonary compromise. More importantly, the presence of portal hypertension increases the risk of portal vein thrombosis, potentially further compromising host liver function. Finally, it unclear whether the transplanted cells can function within cirrhotic nodules when there is ongoing injury or whether enough cells can engraft in the decompensated cirrhotic liver to significantly affect overall liver function.
While the majority of hepatocytes transplanted into the spleens of normal animals migrate to the liver, a small fraction of the cells engraft within the spleen, giving rise to structures resembling hepatic cords. A significant hepatocyte mass can develop and is associated with the appearance of bile cannuliculae, sinusoidal structures, and endothelial and stellate cells, producing morphologic similarity to the liver (9,10). The spleen is considered the most appropriate site for transplantation of hepatocytes in cirrhotic recipients. Excellent engraftment of the cells in the red pulp of the spleen has been achieved by direct injection through the capsule of the spleen, and has resulted in marked improvement in liver function in cirrhotic animals (11).
Some of the earliest studies on the efficacy of hepatocyte transplantation were performed by transplantation into the peritoneal cavity. The attraction of this site for transplantation is based on its large capacity and easy access. Normally, isolated hepatocytes do not anchor or survive following direct injection in the peritoneal cavity. However, prolonged survival has been demonstrated following encapsulation of hepatocytes with microcarriers or hydrogel-based hollow fibers (12). Transplantation of hepatocytes with nonparenchymal cells also results in moderate long-term engraftment in the peritoneal cavity.
Transplantation in animal models of liver failure
Conceptually, hepatocyte transplantation should be especially suitable for treating acute liver failure because the liver remains architecturally normal and it has considerable potential for recovery. As a result, hepatocyte transplantation has been studied in animal models of liver failure since the early 1970s. Transplantation of hepatocytes has been shown to significantly improve the survival of animals with both chemically and surgically induced acute liver failure (13) and to prevent the development of intracranial hypertension in pigs with acute, ischemic liver failure (14). However, the outcome of hepatocyte transplantation in patients with acute liver failure has been disappointing, except when used as a ‘bridge’ to liver transplantation.
The explanation for this discrepancy may be explained by a careful examination of the laboratory studies performed. Interestingly, survival in rodent models of acute liver failure has also been improved by injection of bone marrow cells and hepatocyte lysates (15,16). Thus, it is possible that hepatocyte-derived substances may be responsible for the beneficial effect. This possibility is further reinforced by the fact that most of these animal models do not produce significant inhibition of host-liver regeneration, a critical prognostic factor in patient recovery from acute liver failure. Thus, in the animal models tested, it is likely that release of substances, such as glucose, from the injected hepatocytes may simply provide transient life support until there is adequate hepatic regeneration for survival. These animal studies therefore do not conclusively predict that hepatocyte transplantation will improve the survival of patients with acute liver failure. Recently, a transgenic mouse, whose native hepatocytes lack the capacity to regenerate the liver, has been developed and more accurately reflects the natural history of fulminant liver failure in man (17). Thus far, studies using these animals indicate that a single infusion of donor hepatocytes is unlikely to alter survival. Further studies will be required to determine the exact role hepatocyte transplantation may play in this disease process.
Other studies show that hepatocyte transplantation can improve hepatic physiology in animal models of chronic liver disease. Following formation of a surgical portacaval shunt, rats develop subtle encephalopathy evidenced by changes in spontaneous activity and neurologic reflexes. Injection of hepatocytes into the spleens of these animals improves their behavior and partially corrects their amino acid imbalances (18). In addition, hepatocyte transplantation protects portacaval shunted rats from developing hepatic coma when given exogenous ammonia (19). Finally, intrasplenic hepatocyte transplantation has also been shown to be effective in rats with decompensated liver cirrhosis induced with carbon tetrachloride and phenobarbital (11). In those studies, hepatocyte transplantation improved and stabilized liver function and prolonged survival.
Transplantation in animal models of liver-based metabolic disease
Much of our knowledge about the function of transplanted hepatocytes comes from studies in animals with genetic defects in liver function. The Gunn rat, which lacks hepatic bilirubin UDP-glucuronosyl transferase activity (UGT1A1), and is a model for human Crigler–Najjar syndrome type I, and the Nagase analbuminemic rat are among the best-studied models (20,21). Transplantation of normal donor hepatocytes results in the novel production of bilirubin glucuronides in the bile of Gunn rats and reduces their serum bilirubin levels. Hepatocyte transplantation also increases the serum albumin in Nagase analbuminemic rats, and corrects the metabolic abnormalities in animal models of hereditary tyrosinemia, familial hypercholesterolemia and a number of other liver-based metabolic disorders (22–24).
As many biologically active liver proteins are present in excess, transplantation of a relatively small number of liver cells would be expected to correct liver-based metabolic deficiencies. This, however, has not been the case. Correction of the genetic abnormalities in experimental animals has required a significant degree of native liver cell replacement by engrafted cells. Multiple hepatocyte infusions can improve the response to transplantation (25), but more dramatic results have been obtained by preferential expansion of the engrafted hepatocytes over the host cells. This has been accomplished by inhibiting proliferation of the native liver cells while providing a strong proliferative signal to the transplanted cells.
This process is best exemplified in fumarylacetoacetate hydrolase (FAH) mutant mice. Hepatocytes from FAH knockout mice contain a genetic defect similar to the one that produces hereditary tyrosinemia type I in man, and are destroyed as a result of this metabolic abnormality (22). When hepatocytes from normal mice are transplanted into the livers of FAH mutant mice, almost all of the native liver cells become replaced with donor cells. Unfortunately, FAH knockout mice, like patients with hereditary tyrosinemia, are susceptible to the development of hepatocellular carcinomas and susceptibility to cancer persists despite extensive repopulation of their livers with normal hepatocytes.
To make liver repopulation with donor hepatocytes more feasible clinically, strategies are being developed in the laboratory to exogenously inhibit regeneration of host hepatocytes while stimulating donor hepatocyte expansion. In animals, regeneration of host hepatocytes can be inhibited by administration of the chemical, retrorsine, which blocks the hepatocyte cell cycle (26), or by preparative irradiation of the liver (27,28), and donor cell proliferation can be stimulated by partial hepatectomy or its equivalent (29), or by the use of pharmacological doses of thyroid hormone (30).
Following on the heels of encouraging laboratory results, several centers have instituted clinical hepatocyte transplantation trials. In the earliest test of hepatocyte transplantation efficacy in acute liver failure, intraperitoneal transplantation of human fetal liver cells produced a mild but significant improvement in survival compared to matched controls (31). All patients treated with allogeneic fetal hepatocytes that had Grade 3 hepatic encephalopathy survived, whereas only 50% of matched controls did.
In the United States, hepatocyte transplantation has generally been used to ‘bridge’ patients with acute liver failure to liver transplantation (32,33). Since isolated hepatocytes have not been shown to survive longterm following direct injection into the peritoneum, investigators in the U.S. have used other sites for engraftment and, because of ethical concerns, have used adult hepatocytes rather than fetal cells for transplantation. Patients have been treated with between 107 and 1010 allogeneic hepatocytes, or from less than 1% to 4% of the native hepatocyte mass, where infusion has been performed through either the splenic artery or the portal vein. Transplanted cells have been found in the liver and in the spleen, and transplantation has been associated with anecdotal improvements in ammonia, prothrombin time, encephalopathy, cerebral perfusion pressure and cardiovascular stability. Complications have been few, but include transient hemodynamic instability during intraportal hepatocyte infusion, sepsis, and embolization of hepatocytes into the pulmonary circulation (34). Portal hypertension as a result of transplantation through the portal vein has been generally transient.
Although transplanted hepatocytes may have provided clinical benefit, convincing evidence of engraftment and function of the transplanted cells has been difficult to prove, especially since up to 20% of patients with acute liver failure thought to require transplantation survive without it. The relatively modest results are not surprising since the clinical trials were initiated based on the results of experimental studies in animal models of acute liver failure where the clinical course does not correspond with that seen in man. An additional issue could also be the relatively small numbers of hepatocytes transplanted in many of the patients. Far better results may be seen in the future if multiple hepatocyte infusions are performed.
Treatment of chronic liver failure by hepatocyte transplantation has also been studied at a few centers. The first clinical experience with hepatocyte transplantation for cirrhosis was reported from Japan. Ten patients were treated with hepatocytes recovered from their own left lateral segments (35,36). Patients received hepatocytes by direct splenic puncture, or by splenic artery or portal vein infusion. In addition, according to the investigator, four patients also had their hepatic arteries ligated to control ascites. Unfortunately, the use of such a large number of variables has made the results of this experience difficult to interpret. Hepatocyte engraftment was detected by radioisotope uptake in the spleen of one patient 11 months after transplantation. In the United States, an additional eight patients with decompensated chronic liver disease have received hepatocyte transplants, all via the splenic artery (34). Two of these patients were also shown to have engrafted hepatocytes in their spleens as documented by radioisotope uptake. All patients appear to have tolerated the infusions well, and improvements in encephalopathy, synthetic and renal function were observed.
The reason for poor outcomes following hepatocyte transplantation for end-stage chronic liver disease secondary to cirrhosis is not clear. One possible explanation for the discrepancy between the laboratory and clinical outcomes may relate to the route of hepatocyte delivery. Following infusion using direct splenic puncture, dramatic corrections in liver function have accompanied hepatocyte transplantation in laboratory animals. In patients with cirrhosis, however, allogeneic hepatocytes have been delivered to the spleen exclusively through the splenic artery (34). Animal studies do not support this engraftment strategy, since injection of hepatocytes infused into arterial beds are lost rapidly because of a lack of vessel wall anchorage and shear injury (10). Similarly, in patients, very few hepatocytes have been found in the spleen after infusion through the splenic artery.
Liver-Based Metabolic Disease
The first attempt at correcting a liver-based metabolic disorder with transplanted hepatocytes was performed using genetically modified, autologous hepatocytes. The study was performed in five patients with familial hypercholesterolemia as part of an ex vivo gene therapy trial (37). In situ hybridization of liver tissue 4 months after transplantation in at least one patient revealed evidence of engrafted transgene expressing cells. Clinical benefit, however, was controversial. Additional patients have undergone allogeneic hepatocyte transplantation to correct metabolic liver disorders. Treatment of patients with ornithine transcarbamylase (OTC) deficiency, alpha-1- antitrypsin deficiency, glycogen storage disease type Ia and Crigler–Najjar syndrome type I by transplanting allogeneic hepatocytes has been described. Children with OTC deficiency have shown transient evidence of enzyme activity (38,39), and an adult with glycogen storage disease type Ia has experienced stable improvement in glucose control after liver cell transplantation (40). Unequivocal evidence of function of transplanted human hepatocytes, however, has been obtained in only one patient with Crigler–Najjar syndrome type I (UGT1A1 deficiency). After transplantation, the child's serum bilirubin level fell more than 50%, hepatic bilirubin glucuronidating activity increased from essentially unmeasureable levels to approximately 5% of normal and, after transplantation, more than 30% of the child's bile pigments consisted of bilirubin glucuronides (Figure 1). Although long-term engraftment and function of transplanted allogeneic hepatocytes was accomplished with standard Tacrolimus-based immunosuppression, the level of hepatic UGT1A1 activity derived from the transplanted cells was not sufficient to eliminate the patient's need for phototherapy, and the patient ultimately underwent successful auxiliary liver transplantation (41) (Figure 2).
The most important factor affecting clinical hepatocyte transplantation is a lack of donor availability. The number of livers available for hepatocyte isolation and liver transplantation is limited. Residual segments from reduced liver transplants and organs not suitable for transplantation are the usual source of hepatocytes. Unfortunately, fatty livers do not consistently yield cells of good quality or provide cells in sufficient number to transplant. The ability to preserve and bank hepatocytes would allow pooling of cells from multiple donors to increase cell numbers for transplantation. Unfortunately, cryopreserved liver cells have not yet been shown to engraft as well as fresh hepatocytes (42) and human hepatocyte viability following cryopreservation by current technology is quite variable.
Methods to maintain long-term hepatocyte growth in culture have improved (43). However, at this time, only hepatocytes that have been immortalized by gene transfer are capable of long-term growth and correcting metabolic abnormalities and liver failure after transplantation. Transplanted rat hepatocytes, conditionally immortalized with a temperature-sensitive mutant SV-40 large T antigen have been shown to improve the survival of rats with surgically induced acute liver failure (16); to correct the bilirubin conjugation defect in Gunn rats (44); to protect portacaval shunted rats from hyperammonemia-induced hepatic encephalopathy (19); and to improve liver function and prolong the survival of cirrhotic rats with decompensated liver failure (45).
An alternative solution for the scarcity of human donor cells is the transplantation of hepatocytes from other animal species (46). Animal donors could provide a nearly unlimited supply of hepatocytes of predictable quality when needed. There are, however, major barriers to the use of animal liver cells for transplantation in man. Xenogeneic hepatocytes are susceptible to immunologic processes that are not active following allotransplantation (47,48). In addition, the extent to which xenogeneic hepatocytes can restore normal liver function in patients may be limited (49). Several studies now indicate that the immunologic barrier to xenogeneic hepatocyte transplantation can be overcome with conventional immune suppression (50,51). Preliminary studies in primates appear to confirm this contention (unpublished observations) and engraftment on the order of months following transplantation of pig hepatocytes into the spleens of cirrhotic rodents has been reported to require no immune suppression. Finally, while liver disease can recur following transplantation, it is possible that transplanted nonhuman hepatocytes could be resistant to disease recurrence.
The maximum number of hepatocytes that can be transplanted at any time needs to be better defined. Currently, only 30% or less of transplanted hepatocytes engraft. A better understanding of the factors that allow hepatocyte integration and engraftment may lead to techniques for transplanting a larger number of hepatocytes. This would help reduce the need for repeated cell infusions and multiple donors. While there are some diseases such as hereditary tyrosinemia, where host cells die spontaneously and transplanted normal cells will eventually repopulate the entire liver, in most cases, transplanted hepatocytes do not have a survival advantage over the host cells.
Transplanted syngeneic hepatocytes significantly prolong the survival of animals with cirrhosis but appear to lose function over time. It is not clear how liver disease affects intrasplenic engraftment, survival, and function of transplanted hepatocytes. Finally, although clinical evidence of hepatocyte function can be used to indicate function of transplanted hepatocytes, definitive histologic evidence is difficult to obtain. Definitive evidence of engraftment is essential for the identification and treatment of rejection. At this time, it appears that conventional immunosuppression is effective at controlling rejection of transplanted allogeneic hepatocytes. For liver-based metabolic disease, biopsy measurement of enzyme activity in the liver can be useful. However, for early detection of rejection in the treatment of either liver failure or metabolic disease, less cumbersome methods of detecting the donor hepatocytes will be needed.
Laboratory studies indicate unequivocally that hepatocyte transplantation should be an effective alternative to whole liver transplantation for the treatment of a variety of liver disorders. Significant progress is being made; however, until an adequate supply of donor hepatocytes is identified it will be difficult to prove its efficacy in patients.
This work was supported in part by grants DK48794 and AI131641 from the National Institutes of Health and ATP-70NANBOH3008 from the Department of Commerce.