Xenotransplantation : A potential solution to the critical organ donor shortage

Can J Gastroenterol Vol 13 No 4 May 1999 311 KH Sim, A Marinov, GA Levy. Xenotransplantation: A potential solution to the critical organ donor shortage. Can J Gastroenterol 1999;13(4):311-318. The success of allotransplantation as a treatment for end-stage organ failure has resulted in the need for an increasing number of organ donors. Attempts to meet this need include the use of organs from living related and unrelated donors, financial or other incentives for the donor family, and even the reuse of transplanted organs. Despite these initiatives, the supply of organs for transplantation still falls far short of the demand, as evidenced by longer waiting times for transplantation and decreasing transplantation rates. Even if Canada were able to increase its organ donor rate to that of Spain (40 to 50/million), where organ donation is governed by ‘presumed consent’ legislation, this would not alleviate the problem of donor shortage. Interest in xenotransplantation stems from the need to overcome this increasingly severe shortage of human organs. Indeed, some argue that xenotransplantation is the only potential way of addressing this shortage. As immunological barriers to xenotransplantation are better understood, those hurdles are being addressed through genetic engineering of donor animals and the development of new drug therapies. However, before xenotransplantation can be fully implemented, both the scientific/medical communities and the general public must seriously consider and attempt to resolve the many complex ethical, social and economic issues that it presents.

O rthotopic liver transplantation is the treatment of choice for many patients with end-stage liver disease.Advances in immunosuppression and perioperative care have resulted in a dramatic improvement in survival of patients undergoing liver transplantation (1).However, the need for donor organs far outweighs the available supply (Table 1) (2)(3)(4)(5).This need has prompted the search for alternatives to conventional allotransplantation (6,7).Xenotransplantation, the transplantation of cells, tissues or organs between members of different species, has emerged as a potential solution to the shortage of human organs.The concept of transplanting animal organs into humans is not new.The first such attempt was made by a Russian physician in 1682, who reportedly repaired the skull of a wounded nobleman by using a bone from a dog (8).In 1964, kidneys from chimpanzees were transplanted into 13 patients (9).Most died within days, although one survived for nine months on an immunosuppressive regimen of azathioprine, actinomycin C and steroids.More recently, two patients received liver transplants from baboons.One patient survived for more than two months, but ultimately both patients died due to infection (10).
The preferred source of the xenograft is a concordant species that is closely related to the recipient (eg, nonhuman primate to human).Discordant xenotransplantation (eg, pig to human) involves widely divergent species.While it is much easier to achieve xenograft acceptance across concordant combinations, the widespread use of primates as donors (such as baboons and chimpanzees) poses a number of problems.These problems include ethical issues (eg, 'humanization of primates'), limited availability, small size and risk of transmitting infectious agents (11).For these reasons, it is likely that a species discordant with humans will offer a source of organs for xenotransplantation.The use of pigs as a source of donor organs is attractive for reasons including their similar size and physiology, their availability in sufficient quantities and the fact that they can be bred under controlled conditions where their make-up can be genetically modified.
In addition to overcoming the immunological barriers raised by cross-species transplantation, a number of complex ethical, social, legal and economic principles must be addressed.A thorough evaluation and understanding of the safety, efficacy and risks of xenotransplantation are necessary prerequisites to the widespread clinical application of this potential future therapy.

HYPERACUTE REJECTION
The first immunological barrier to xenotransplantation is hyperacute rejection (HAR), characterized by complement activation leading to widespread thrombosis and interstitial hemorrhage of the donor organ (12).This phenomenon is initiated by the binding of antibodies of the recipient to the vascular endothelium of the donor organ (13).These naturally occurring 'xenoreactive' antibodies (XRAs) recognize cells from discordant species, and are present in humans and other Old World primates (14).
XRAs react predominantly to the carbohydrate epitope Gal-alpha-1,3-Gal, which is expressed as a terminal modification on glycoproteins and glycolipids of New World monkeys and all lower mammals (15).Synthesis of Galalpha-1,3-Gal is catalyzed by alpha-1,3galactosyltransferase, an enzyme that is absent in humans, apes and baboons (16).Evidence that Gal-alpha-1,3-Gal is a critical target in HAR of pig to primate xenografts came from the demonstration by Collins et al (17) that transplantation of a New World monkey heart into a baboon results in HAR, and that virtually all the antibodies that bind to the rejected heart are specific for Galalpha-1,3-Gal.Sachs and Sablinski (18) subsequently demonstrated that specific depletion of anti-Galalpha-1,3-Gal antibodies from primates prevents HAR of porcine xenografts.
Activation of complement is a crucial step in the development of HAR.Administration of complement inhibitors such as cobra venom factor (19,20), soluble complement receptor type 1 (21) or gamma globulin (22) markedly prolongs the survival of a discordant xenograft.HAR does not occur in recipients with inherited deficiencies of complement (23).In porcine to primate xenografts, complement activation appears to occur predominantly via the classical pathway (24).Because complement activation plays a pivotal role in the pathogenesis of HAR, the ability to regulate human complement may help overcome a major hurdle in the clinical application of xenotransplantation.One way to regulate complement activation may be through membrane-bound regulators of complement activity that inhibit the complement cascade at various critical steps (25).Examples of such regulators are decay accelerating factor (DAF [CD55]) (26), which promotes dissociation of convertase complexes; membrane cofactor protein (MCP [CD46]), which inhibits complement activation at the C3 convertase step; and membrane inhibitor of reactive lysis (CD59), which prevents assembly of lytic membrane attack complexes (27).These glycoproteins help protect autologous (ie, host) cells against inadvertent complement-mediated injury targeted against xenogeneic cells.Complement regulatory proteins appear to function poorly against heterologous complement and thus may make xenografts particularly susceptible to complementmediated injury (28).For example, human DAF inhibits human C3 convertases effectively but rabbit C3 convertases poorly (29).The importance of complement regulation is suggested by the observation by McCurry et al (30) that organs from transgenic pigs expressing human CD55 and CD59 generally resist HAR after transplantation into baboons.
The precise mechanism by which complement activation leads to destruction of the donor organ is not well understood but appears to involve loss of function of endothelial cells.Complement-mediated lysis of endothelial cells alone is not the sole mechanism because lysis is not a major finding early in the course of HAR (12).One noncytotoxic change in the endothelium that may lead to loss of endothelial barrier function is the formation of intercellular 'gaps' induced by C5b67 complexes -gaps that may allow platelet adhesion and activation (31).This change may also lead to loss from endothelium of heparan sulphate, a protein polysaccharide conjugate that contributes to the endothelial barrier to blood cells and proteins, anticoagulation, and protection against complement and oxidants (32,33).Other complement-mediated changes that may contribute to the manifestations of HAR are the secretion of von Willebrand factor (34), which may trigger platelet aggregation, and the expression of P-selectin (35), which can serve as a ligand for neutrophils and platelets.
Therefore, efforts to prevent HAR can focus on modifying the immune system of the host (either targeting natural antibody reactivity or inhibiting complement activation) or modifying the donor organ to diminish their immunogenicity.Depletion of XRA with plasmapheresis, plasma exchange, anti-idiotypic antibody and extracorporeal immunoadsorption in pig to primate xenografting has, at best, prolonged graft survival from several hours to two weeks, making these approaches to block antibodyinduced HAR clinically unattractive (36).Neutralizing anti-Gal-alpha-1,3-Gal antibody by continuous intravenous infusion of Gal-alpha-1,3-Gal oligosaccharides leaves few antibodies to bind to the xenograft and has met with some success in prolonging graft survival (37).Attention has recently turned towards suppressing the activity of B cells involved in anti-Gal-alpha-1,3-Gal antibody production by administering a Gal-alpha-1,3-Gal oligosaccharide to which a toxin has been attached (38).Decomplementing the xenograft recipient with cobra venom factor or similar agents also only provides several days of graft survival and is, therefore, a less than ideal therapy (19,20).
Another strategy for disrupting antibody-antigen reactions focuses on reducing the expression of antigen in xenograft donors.An alpha-1,3-galactosyltransferase knockout mouse has been developed that does not express Gal-alpha-1,3-Gal, although this process has yet to be carried out in pigs (39).A similar approach is to exploit natural variations in antigen expression by breeding donors with low antigen expression of Gal-alpha-1,3-Gal (40).The alternative approach of competitive glycosylation replaces the alpha-Gal molecule with an oligosaccharide against which humans do not possess antibody (such as the O blood group oligosaccharide alpha-Fuc), and has been used in mice (41) and pigs (42).Finally, much attention has recently focused on the creation of pigs transgenic for one of the human regulators of complement activation.Organs from transgenic pigs expressing low levels of human CD55 and CD59 resisted complement-mediated injury and HAR after transplantation into baboons (30).

DELAYED XENOGRAFT REJECTION
If HAR is averted, xenografts are still rejected in days instead of minutes or hours by a process referred to as 'delayed xenograft rejection' (DXR), also termed 'acute vascular xenograft rejection' (19).This process is characterized pathologically by infiltration of leukocytes (particularly monocytes and natural killer [NK] cells), focal ischemia and diffuse microvascular coagulation (43).There is increasing evidence that DXR may be caused by the ongoing interaction of XRA with the graft.Development of DXR coincides with a marked increase in the synthesis of XRA after exposure to xenogeneic cells (44).Furthermore, removal and/or inhibition of XRA from xenograft recipients delays or prevents DXR (45).In contrast to HAR, the role of the complement system in the pathophysiology of DXR is unclear.DXR typically occurs in xenograft recipients depleted of complement (43).Yet the pathologic lesions characteristic of DXR appear to reflect complement-mediated changes in endothelial structure and function.Hancock (46) hypothesized that even the most potent inhibitors of complement, such as cobra venom factor, are incompletely effective.While such treatments may reduce complement to levels that do not result in HAR, there may still be sufficient complement activity in serum to contribute to the development of DXR.Other possible effector mechanisms for DXR include the action of host platelets and leukocytes on the xenograft.Platelets activated by exposure to complement fragments express P-selectin, resulting in the release of potent chemotactic signals, such as RANTES and MCP-1, that attract leukocytes, particularly monocytes (47).The presence of both monocytes and NK cells has been documented in organs with DXR (19,43).Treatment of xenograft recipients with antibodies that inhibit the function of inflammatory cells has been shown to prolong graft survival (48).Some studies suggest that NK cells act by disrupting the integrity of the endothelial monolayer (49) and by activating endothelial cells (50).Endothelial activation in animal models of DXR has been associated with a shift to a procoagulant state with induction of tissue factor, production of chemokines such as MCP-1, and induction of leukocyte adhesion molecules such as intercellular adhesion molecule 1 (43).
The Multiorgan Transplant Research Unit at the Toronto Hospital has an interest in the role of immunemediated thrombosis in the pathogenesis of DXR (Figure 1).Using a rat liver model of xenotransplantation, we have shown that prophylactic infusion of monoclonal antibodies to a cellular procoagulant, fgl-2 prothrombinase, a glycoprotein that cleaves prothrombin to thrombin, prevents thrombosis, platelet adherence and fibrin deposition (51).Studies are underway to determine whether organs from animals in which the prothrombinase gene has been disrupted undergo DXR when transplanted into recipient animals.Another potential therapeutic strategy is to vaccinate animals with soluble prothrombinase so that they may produce neutralizing antibodies to this procoagulant before xenotransplantation.The likely role of endothelial activation in the pathogenesis of DXR suggests additional approaches for modulating endothelial responses, for example, by regulated targeting of NF-kappa-B-dependent endothelial cell pathways (52).In addition, targeting of platelet aggregation and recruitment and activation of host monocytes -early events in the development of DXR -are potential therapeutic approaches.

CELLULAR/HUMORAL REJECTION
Xenografted organs that are not destroyed by HAR or DXR are subject to elicited cellular and/or humoral rejection (18).In pig to primate transplants, Alexandre et al (53)  Conversely, the indirect T cell response to a xenotransplant might be expected to be stronger than that towards an allotransplant, given the wide array of foreign peptides it would give rise to (56).Potent immunosuppressants such as methotrexate, cyclosporine and rapamycin can and have been systemically administered to blunt host T cell responses to allografts, albeit with a significant risk of infection, malignancy and organ toxicity.A 'tolerant' state, in which the recipient's immune system regards donor antigens as 'self' would obviate the need for chronic immunosuppressive therapy.One approach to induce tolerance involves introducing donor-derived hematopoietic cells before grafting the recipient with a solid organ from the same animal (57).The interaction of two coexisting donor and recipient cell populations is believed to result in mutual down-regulation of both recipient and donor immune systems, a phenomenon known as 'microchimerism' (10).This approach appears to have worked across rodent species but has not been successfully applied across broader species lines (58).
Little is known about the role of the elicited humoral response to a xenograft and whether it differs from what occurs with allografts (53).In a limited number of xenografts in which recipient XRAs are temporarily removed, HAR does not occur even after antibodies return to the circulation (27).This apparent ability of the humoral system and the graft to coexist, known as 'accommodation', may involve a change in the properties or specificity of the antibodies, a change in the antigen, or an acquired resistance of endothelium to humoral injury and complement activation (59).Further understanding of the biological basis of accommodation may prove critical to the clinical application of xenotransplantation.

INFECTIOUS CONCERNS
Infections have proven to be a substantial cause of morbidity and mortality after allotransplantation.This is owing in large part to the immunosuppressive treatment needed to prevent graft rejection.Interspecies transplantation may require greater levels of immune suppression than are currently used for allotransplantation, thereby further increasing the risk of opportunistic infection.An additional risk of xenotransplantation is that of 'xenosis', the transfer of animal-derived infectious pathogens with xenotransplants (Figure 2) (60).This is in contrast to the natural spread of infection between species, referred to as 'zoonosis'.(61) While xenograft recipients are susceptible to infection by bacteria, fungi and parasites, much attention has focused on infection by viral pathogens (Table 2).
Several mechanisms of xenogeneic infection have been identified.A pathogen may be infectious for both the donor species and the human recipient (eg, Toxoplasma gondii).Some animal viruses that are similar to their human counterparts, such as primate cytomegalovirus (CMV), have been documented to produce clinical disease in hu-man xenograft recipients (62).A major concern is retroviruses such as simian immunodeficiency virus, which can be transmitted across species and produce a more virulent reaction in the new host (63).Some pig retroviruses have been shown to reactivate after radiation exposure and might similarly reactivate when exposed to immunosuppressive medication (64).Indeed, many of the conditions associated with retroviral activation (eg, immune suppres-Can J Gastroenterol Vol 13 No 4 May 1999 315 Xenotransplantation and transplantation Retroviruses sion, graft rejection, cytotoxic therapy) are present in the transplant recipient (65,66).Latent animal viruses present in the xenograft, such as porcine CMV, may be unable to infect human tissue but may later reactivate in the animal organ, resulting in graft failure (60).Finally, concern has been raised regarding the possibility of crossover of an animal virus with a human virus, leading to a more virulent recombinant organism.Indeed, dual infections can lead to recombination, as has been observed with in vitro mixing of CMV isolates from transplant recipients (67).Concurrent inoculation of two avirulent herpes simplex viruses into mice has been demonstrated to produce lethal recombinations (68).Recognition of xenogeneic infections may be complicated by the presence of novel pathogens for which laboratory testing is not available, new clinical syndromes and altered behaviour of these pathogens in the immunocompromised recipient.

TABLE 2 Zoonosis in xenotransplantation
Pathogens considered most likely to cause human disease should be screened for and excluded from source animals.Careful thought must be given to determining the organisms to be screened.Algorithms have been developed for baboon bone marrow transplantation into a human to help classify potential microorganisms in source baboons (69).Pathogens were designated as absolute contraindications if they were known to causes zoonoses in primates and known to be hazardous to humans.In particular, organisms that could be secondarily transmitted from an infected recipient to close contacts, such as simian immunodeficiency virus, T gondii and Mycobacterium tuberculosis, were placed in this category.Viruses at high risk for recombination, such as parvovirus and rotavirus, were also designated absolute contraindications.Relative contraindications included organisms suspected to be transmissible but of unproven or unclear clinical significance, such as baboon herpesviruses.Treatable infections were those that could be identified and eradicated successfully before harvest, such as Babesia species.Lastly, a category called 'unavoidable' included baboon endogenous virus and all organisms yet to be recognized.Similar categories have been designed for microbial agents in pigs (70).Molecular techniques to screen source animals for retroviruses, herpesviruses and other organisms are being developed, but much remains to be learned in this field (71).Source animals should be raised in a gnotobiotic environment to eliminate microbial agents.Maintaining pathogen-free environments is an arduous task because colonies can easily be contaminated by unwanted organisms introduced from outside animals or human caretakers (72).
Prevention of disease from human donors has relied on protocols for donor screening, recipient prophylaxis and intensive post-transplantation surveillance.Similar protocols for screening of source animals and surveillance of xenograft recipients can help prevent and monitor infections.Research must, therefore, focus on the behaviour of known pathogens in the transplant recipient.Newer po-tential pathogens must be isolated and characterized, and diagnostic tests must be developed.

ETHICAL, SOCIAL AND ECONOMIC
CONSIDERATIONS Xenotransplantation is potentially a reliable, long term solution to the current donor shortage for allotransplantation.At the same time, it presents a number of complex ethical, social and economic issues that must be resolved before it can be fully implemented.Unlike other areas of clinical medicine, xenotransplantation raises a variety of ethical issues that have not been formally discussed or agreed upon in the public arena.The general public must play a role in defining ethical policies regarding xenotransplantation, and it is the duty of the medical and scientific communities to provide as much information as possible to them, so that public policies can be formulated in an informed manner.
The clinical application of xenotransplantation is predicated on the assumption that the use of animals in medical research and therapy is morally acceptable.Even if this debatable principle is accepted, the use of animals in xenotransplantation, particularly nonhuman primates, raises additional issues.That these animals may need to be genetically altered to overcome certain immunological obstacles further complicates their use as experimental subjects.As a minimum requirement, the use of source animals for xenotransplantation have to comply with the Canadian Council on Animal Care's Guide to the Care and Use of Experimental Animals (73,74).
The issue of informed consent in xenotransplantation, as in any other area of medicine, requires that the subject be made aware of its potential benefits and risks.In the case of xenotransplantation, there are additional considerations that may violate some of the principles of informed consent.Early recipients of xenografts typically have poor rates of survival, so treatment of these subjects in experimental protocols may raise serious ethical concerns.While the potential benefits and risks of rejection may be estimated, the risk of xenozoonoses cannot reliably be predicted, particularly when DNA from different species is mixed, and may result in new and unpredictable infectious agents.For example, it was only recently determined that porcine endogenous retroviruses can infect human cells (75).The transplant recipient has to agree to the requirement for strict postoperative monitoring, which may result in quarantine or other physical constraints should the recipient develop an infection potentially communicable to the community.This requirement may also extend to household members and other close contacts, and would violate another proviso of informed consent -the right to withdraw from the experiment at any time (76).Hence, in the clinical application of xenotransplantation, 'informed consent' may need to be replaced with a more binding 'contract' between doctor and patient.
For xenotransplantation to become a clinical reality, substantial financial resources have to be invested to establish facilities for rearing and monitoring source animals, developing laboratory tests, training staff, monitoring recipients and close contacts, and establishing infection control measures.Use of these resources in the development of xenotransplantation technology might result in fewer dollars being available for other equally important medical interventions.If, in the long term, the cost of xenotransplantation drops sufficiently to justify its widespread introduction, there would be additional cost implications for an already cash-strapped Canadian health care system.