Xenogeneic transplantation is transplantation between members of different species. A discussion of xenogeneic transplantation, or ‘xenografting’, enters a textbook of surgery partly as history and partly as speculation, since clinical xenografting is not currently actually taking place. However, the notion that animals might donate organs to human beings has been a source of fascination since transplantation was conceived and the prospects for clinical xenografting being performed successfully in the future are good.

Xenogeneic transplantation has been attempted approximately 30 times in human beings. Although none of these transplants has survived for a full year, a kidney donated by a chimpanzee maintained good function for 9 months, until the patient died.

Reemtsma was the first to report a xenograft of a chimpanzee kidney to a human, in 1963; he went on to perform the procedure 12 times. Starzl reported a series of baboon kidney transplants to humans which started shortly afterward. Starzl also performed three chimpanzee to human liver transplants without success. A xenogeneic heart transplant was first reported in 1964 by Hardy. Several additional heart transplants were attempted including the case of ‘Baby Fae’ in 1985. All but one of these clinical efforts involved genetically closely-related primate donors and most were performed during the 1960s, when the immunosuppression available was less powerful than that used now.

It was an historical accident that the first clinical attempts at xenogeneic transplantation were performed just as the phenomenon of hyperacute rejection was characterized. Hyperacute rejection is caused by pre-existing antibodies in a recipient's serum that are specific for donor antigens. Human beings may have antibodies specific for human blood group antigens; these antibodies can be induced by pregnancy, blood transfusion, or transplantation of other tissues. Human beings and other species also have ‘natural’ antibodies in their serum: these are specific for antigens of many other species and exist even without prior immunization. These natural antibodies may result in the hyperacute rejection of xenografts. Recognition of the importance of antibody-mediated rejection contributed to the decline in clinical xenografting, and the presence of natural antibodies is still the major obstacle to its widespread application.

Natural antibodies and their target antigens are not well characterized. Most of the antibodies are of the IgM subclass and probably react with endothelial glycoproteins. They may arise from cross-reactions with common environmental pathogens. In some respects natural antibodies are similar to the antibodies that react with ABO blood group antigens, although their target antigens are not the same. Very closely related species do not have natural antibodies reactive with each other: xenotransplantation between these so-called ‘concordant’ combinations is not followed by hyperacute rejection. The presence of natural antibodies causing hyperacute rejection defines ‘discordant’ combinations.

Although the existence of ‘natural’ antibodies hinders clinical xenografting, their presence makes xenotransplantation an excellent model to study the mechanisms and prevention of hyperacute rejection. Dozens of xenografting studies have been performed in which one or another substance has been used to delay rejection. These studies have provided information on the role of complement and vasoactive substances in hyperacute rejection, but their results have nonetheless been discouraging for clinical application. Whenever natural antibody has been present, hyperacute rejection has always occurred with only minimal delay after treatment. Even extensive absorption of antibody or its near-complete elimination by plasmapheresis, which have occasionally been successful in achieving allogeneic transplantation across blood group barriers, have not been successful in experimental xenotransplantation. The development of strategies to remove natural antibodies effectively is one of the major goals of research in xenografting.

While hyperacute rejection is a bigger factor in xenotransplantation than in allotransplantation it is still uncertain whether induced antibody responses to xenografts, which occur after transplantation, are also stronger than those induced by allografts. Some evidence suggests that they are. Since chronic graft rejection is thought to be frequently mediated by induced antibody, chronic rejection may prove an even greater problem for xenografts than for allografts, even if hyperacute rejection can be avoided by elimination of natural antibody. The induced antibody response to xenogeneic antigens probably does not reflect a secondary response of the natural antibodies, but rather a new response to xenogeneic antigens, including those of the major histocompatibility complex. Antibodies induced by concordant xenografting recognize the same determinants as those defined by allografting. Because of the likelihood that induced xenoantibody responses are stronger than those seen in allogenic combinations, strategies to achieve long-term clinical xenotransplantation will probably require forms of immunosuppression aimed at limiting this response.

The special importance of antibody-mediated rejection in xenografting is in keeping with the simple notion that the more disparate a set of antigens is from those of the responding host the more powerful the immune response will be. When considering cell-mediated immune responses, however, the immunological discoveries of the past decade suggest that this simple dogma may not hold. The T-cell response to allogeneic MHC antigens is extraordinarily strong relative to the response to normal environmental pathogens. Late in the 1970s the basis of this strength was identified when immunologists realized that T cells recognize environmental pathogens as peptides of these antigens presented in association with ‘self’ antigens of the major histocompatibility complex and that developing T cells are selected for their ability to recognize such modified forms of self antigens. Allogeneic major histocompatibility complex antigens express determinants which are similar to those formed by the association of environmental pathogens and self antigens. The strength of the immune response to alloantigens therefore depends, in part, on their similarity to the responding host's own antigens. In the light of this concept, the possibility emerges that xenogeneic major histocompatibility complex antigens might be sufficiently dissimilar from those antigens of a responding species that they may not be well recognized by T cells. As a result, the cell-mediated response to xenografts may be weaker or of a different character compared with the response to allografts.

Three approaches have been used to test the strength of cell-mediated immunity to xenogeneic antigens in the absence of simultaneous humoral immunity. First, in-vitro assays of T-cell function, including mixed lymphocyte proliferation, interleukin-2 production, and cell-mediated cytotoxicity, have measured T-cell responses to xenogeneic stimulation. It is not certain, however, that these assays accurately reflect the processes of graft rejection. Secondly, in-vivo grafts exchanged between concordant species can test cellular immunity in the absence of preformed antibody, although not without induced antibody responses. Thirdly, certain kinds of in-vivo grafts may be resistant to antibody-mediated rejection, leaving them susceptible only to T-cell or other cell-mediated destruction. For example, skin grafts are very resistant to humoral attack, the liver is unusually resistant to antibody-mediated destruction, and cultured pancreatic islets may also be resistant to humoral immunity. Each of these approaches has been used to test the strength of cell-mediated responses to xenogeneic antigens.

In-vitro assays of T-cell responses to xenoantigens have provided mixed results. Some have suggested strong primary cellular responses to xenoantigens equivalent to those for allogeneic MHC antigens. In many cases, however, well-designed studies of T-cell immunity have shown substantially diminished responses to xenogeneic antigens compared with allodeterminants. For example, quantitative assays of cytotoxic T cells directed against xenoantigens have shown diminished precursor frequencies compared to those for alloantigens. In addition, assays of interleukin-2 production have shown absent or minimal responses to xenoantigens but strong responses to allodeterminants. The differences in these results may stem in part from the different species combinations and assay conditions used. Overall, however, evidence is emerging to support the notion that xenogeneic cellular immunity for some species combinations, as measured by in-vitro assays, is not as powerful as is that to allografts.

Additional analysis of the in-vitro cell-mediated response to xenoantigens has demonstrated that powerful secondary responses to xenoantigens can be generated after in-vivo priming, but that these secondary responses require the presence of antigen-presenting cells from the responding species. This result suggests that secondary recognition of xenoantigens involves recognition of peptides of these antigens presented in association with responder-type major histocompatibility antigens, on responder-type antigen-presenting cells, a mechanism similar to that used for recognition of environmental pathogens. These in-vitro studies, therefore, provide support for the notion that cell-mediated immunity directed against xenografts may not only be weaker, but also of different character to that directed against allografts.

Despite these in-vitro results, in-vivo experiments using concordant species to test the cell-mediated response to xenografts show that rejection is rapid. In addition, all standard forms of immunosuppression aimed at preventing cell-mediated rejection have been found to be less effective for xenografts than for allografts. Combinations of reagents have also been tested, in an attempt to produce optimal programmes for xenograft immunosuppression. Splenectomy, 15-deoxyspergualin, and other types of immunosuppression aimed at preventing B-cell responses, in addition to anti-T cell reagents, are often useful in achieving more prolonged xenograft survival. These results probably reflect a contribution of induced antibody in concordant xenograft rejection. It is difficult to tell whether in-vivo cellular immunity to xenoantigens is stronger or weaker than to alloantigens in these studies involving concordant species.

In-vivo studies of cell-mediated rejection of xenografts performed between discordant species have also been reported using skin grafts to avoid humoral rejection. Graft rejection without immunosuppression has again been rapid and poorly controlled by standard immunosuppression. However, antibody treatment of recipients to remove CD4⫀ T cells has allowed prolonged survival of discordant xenogeneic skin and resulted in better survival of xenografts than of allografts placed on the same immunosuppressed recipients. These results demonstrate that in-vivo xenograft rejection is especially dependent on CD4⫀ T cells and that the remaining cell-mediated response to xenografts is weaker than that to allografts, in the absence of CD4⫀ cells.

The combined results of the in-vitro and in-vivo studies of cell-mediated xenograft rejection are difficult to interpret. On the one hand there appears to be a measurably weaker in-vitro T-cell response to xenoantigens in many cases and a particular requirement for CD4⫀ T cells in vivo. On the other, cell-mediated xenograft rejection is rapid in vivo and is more difficult to control by standard immunosuppression than is allograft rejection. This apparent conflict may reflect imperfect measurement of T-cell mechanisms of xenograft rejection by in-vitro assays or the existence of other cellular processes of graft rejection which are dependent on CD4⫀ T cell function, but otherwise independent of T cells. In either case, the results suggest that if B-cell responses to xenografts can be controlled or avoided, cell-mediated xenograft rejection may be prevented by more selective immunosuppression than is required for allografts.

The ability to induce tolerance to xenoantigens has been studied by induction of neonatal tolerance and induction of tolerance in adult life by creation of bone marrow chimeras after whole body irradiation. The results of these studies have suggested that the ability to achieve tolerance to xenografts is similar to that for allografts in principle, but far more difficult to achieve in practice. This situation seems to result from the difficulty in achieving lasting survival of xenogeneic cells in the recipient even when the recipient is immunologically incompetent as judged by allogeneic cell survival. Perhaps the effector mechanisms of xenograft rejection arise earlier in development or are more resistant to whole body irradiation.

Both of these standard approaches to tolerance induction have practical limitations when considered for application to adult human patients. A modification of the technique for creating bone marrow chimeras has been developed in rodents to produce mixed chimeras, reconstituted (after whole body radiation) partially with syngeneic marrow cells and partially with donor (in this case xenogeneic) stem cells. In addition, the requirement for toxic whole-body irradiation has been avoided by using monoclonal antibodies, combined with more selective radiation, to prepare the recipient for marrow reconstitution. Further investigation of such practical strategies to achieve tolerance to xenoantigens may help make clinical xenografting possible.

The difficulty in achieving long-term survival of xenogeneic cells in recipients of another species raises the possibility that there may be non-immunological incompatibilities which prevent some forms of xenografting. Animal organs may not function adequately in humans because of enzyme restrictions, receptor incompatibilities, differences in metabolic products, or for many other reasons. Little information is available regarding such possible incompatibilities. The kidneys of chimpanzees and baboons were capable of supporting human life for several months and pig livers were found to produce bile, to secrete porcine albumin, and perhaps to improve human neurological function when they were perfused with blood from patients with hepatic failure. Beyond these limited data, identification of the limits of xenograft survival and function will require control of immunological causes of graft failure. It is likely that such limitations do exist.

Closely related to the question of xenograft function in a human host is the question of which species might prove the best donor for clinical xenogeneic transplantation. Presumably the more closely related species will show fewer incompatibilities of organ function. In addition, more closely related species are concordant to humans, thus avoiding the problem of natural antibodies. On the other hand, the subhuman primates, especially chimpanzees, are not available in large numbers, breed slowly in captivity, may have organs too small for human beings, and their use would present ethical problems today. Many investigators have therefore considered pigs as potential donors well suited for clinical xenotransplantation. These animals breed rapidly and are of appropriate size. However, man possesses natural antibody against pig antigens. Another consideration in selecting animal donors is whether endemic pathogens or non-pathogenic retroviruses of another species might prove dangerous to human beings.

The idea of clinical xenotransplantation provokes much controversy regarding the ethical issues, as was demonstrated in the ‘Baby Fae’ case in 1985. The issues are of several kinds. From the point of view of the recipient, xenotransplantation would be entirely experimental and a successful outcome, as this chapter should indicate, cannot be confidently predicted. Thus appropriate informed consent for an uncertain undertaking would be necessary; obtaining informed consent would be particularly uncertain if the potential recipient were a child. With respect to the donor animals, ethical issues also arise. Some critics have objected especially to the use of subhuman primates because these animals have easily recognizable human characteristics. How can one know, however, where a line should be drawn that makes an animal too close for comfortable use by human beings. Others have objected to the breeding and use of any animals for the purpose of supporting human life by organ donation. In this case it is not clear how this would differ from the breeding of livestock for human food and clothing, an activity generally accepted by our society.

In many respects this chapter presents the uncertain prospects for clinical xenografting in the near future. Nonetheless, several approaches might result in successful xenogeneic transplantation being accomplished in this century. First, the many months of primate graft survival accomplished during the early 1960s makes it seem quite likely that some successful xenografts could be undertaken even today using closely related donors and modern techniques of immunosuppression. As still newer immunosuppressive agents become available, such as 15-deoxyspergualin and FK506, these prospects may become even better. Animal kidneys would probably not be used for humans as a first step, because dialysis can maintain most recipients satisfactorily until human organs become available. Patients with heart disease, however, might be saved in some circumstances by use of a primate donor, even if that were only a temporary bridge until allotransplantation could be performed. On the other hand, the limited number of animals available from concordant species will prevent widespread application of this approach, making it questionable whether it should be tried at all.

A second approach to xenotransplantation would be through achievement of better techniques to eliminate natural antibody and prevent humoral rejection, thereby allowing use of more disparate donors. Since successful allotransplantation across blood group barriers has been accomplished in a few cases, this approach may be feasible. Although plasmapheresis has never been sufficiently effective to achieve long-term xenotransplantation, the precise identification of the target antigens of natural antibody might allow more complete immunoabsorption. On the other hand, the effort to eliminate natural antibody has not been successful in over 20 years of work, making this approach to xenotransplantation very uncertain.

A third potential approach to achieve successful xenogeneic transplantation would be to achieve neonatal tolerance to xenogeneic antigens. For example, diagnosis of life-threatening cardiac birth defects in utero by ultrasound might provide the opportunity for prenatal introduction of xenogeneic cells, potentially rendering the newborn infant tolerant to antigens of this species. Sufficient experimental data to demonstrate the feasibility of this approach are not yet available.

A fourth potential approach would be to use xenogeneic tissues resistant to antibody-mediated attack. The liver might be one such organ, especially for children, who are at considerable risk of dying while awaiting a human donor. More likely, animal donors could provide pancreatic islets to treat diabetes mellitus: cultured pancreatic islets may be resistant to humoral rejection and they have the additional advantage of being depleted of antigen-presenting cells. There has been considerable difficulty in isolating large quantities of human islets, making the gap between the potential supply of human islets and the potential demand of thousands of diabetic patients very large. Since the cell-mediated response to discordant pancreatic islets may be more easily controlled than for allogeneic islets, this approach is promising.

While successful xenotransplantation has not yet been achieved, it is still not too early to imagine the use of xenogeneic tissues, not for their organ function, but rather as vectors to introduce the products of genes not normally carried by the animal donor. For example, techniques of molecular biology make it possible to imagine the creation of transgenic pigs expressing products of human genes which may be absent or defective in a human recipient. Transplantation of cells from these animals would avoid the malignant potential of long-term cultured human cell lines. It is also possible to imagine the selective breeding of herds of animal donors or the induction of genetic mutation so as to avoid expression of particular deleterious transplantation antigens. In these ways xenogeneic transplantation may be the key to a new era in the field of transplantation.

Auchincloss HJ. Xenogeneic transplantation. A review. Transplantation 1988; 46: 1–20.

Auchincloss HJ. Xenografting: a review. Transplant Rev 1990; 4: 1–7.

Calne RY. Observations on renal homotransplantation. Br J Surg 1961; 48: 384.

Hardy MA. Xenograft 25. Netherlands: Elsevier Science Publishers BV (Biomedical Division) 1989.

Ildstad ST, Wren SM, Sharrow SO, Stephany D, Sachs DH. In vivo and in vitro characterization of specific hyporeactivity to skin xenografts in mixed xenogeneically reconstituted mice (B10 + F344 rat to B10). J Exp Med 1984; 160: 1820.

Moses RD, Pierson RNI, Winn HJ, Auchincloss HJ. Xenogeneic proliferation and lymphokine production are dependent on CD4⫀ helper T cells and self antigen-presenting cells in the mouse. J Exp Med 1990; 172: 567–75.

Perper RJ, Najarian JS. Experimental renal heterotransplantation: I. In widely divergent species. Transplantation 1966; 4: 377.

Pierson RN, Winn HJ, Russell PS, Auchincloss HJ. Xenogeneic skin graft rejection is especially dependent on CD4⫀ T cells. J Exp Med 1989; 170: 991–6.

Platt JL, Lindman BJ, Chen H, Spitalnik SL, Bach FH. Endothelial cell antigens recognized by xenoreactive human natural antibodies. Transplantation 1990; 50: 817–22.

Reemtsma K. Renal heterotransplantation from nonhuman primates to man. Ann NY Acad Sci 1969: 162: 412.