On Cardiac Xenotransplantation and the Role of Xenogeneic Tolerance

An increasing number of patients die without a life-saving transplant as the gap between individuals awaiting transplantation and available human donor hearts widens. Cross-species transplantation (xenotransplantation) of pig hearts into human recipients offers the prospect of an unlimited supply of organs and would permit elective transplantation using quality-controlled organs. Although initial results in the 1960s and 1980s using non-human primate donors for cardiac xenotransplantation were poor, several advances in genetic engineering and immunosuppression have helped mitigate previously insurmountable obstacles. The initial success of the first pig-to-human heart transplant this year¹ ushers in newfound excitement for the promise of xenotransplantation, but also highlights several uncertainties which require further investigation.

Pediatric vs. Adult Recipients

Probably the best known cardiac xenotransplant was that by Leonard Bailey, who transplanted a baboon heart into an infant girl, known as Baby Fae, in 1983. Infant organs at the time were nearly impossible to obtain, so xenotransplantation offered a potential solution. The surgical procedure was technically successful, but the patient died from acute rejection 20 days later.² Since then, advances in mechanical circulatory support markedly improved survival to transplant as well as survival among non-transplant candidates; two-year survival free from disabling stroke or reoperation exceeds 75% in adults.³ Although satisfactory outcomes are achievable among larger children and young adults with mechanical circulatory support devices, neonates and infants represent a particularly challenging group. Even in experienced centers, mortality after ventricular assist device implantation in single ventricle patients approaches 50–75% in neonates, and only 30–50% survive to discharge.⁴ But with wait times of 4 to 6 months, and wait-list mortality rates exceeding 30%, we believe this population to be that in most need of a viable alternative such as xenotransplantation.

Neonates and infants pose unique challenges, but also offer important advantages. Heart size and growth rate become very important considerations for translatability; pigs reach adult human size within 6 months. Miniature swine or growth hormone receptor knockout pigs likely offer advantages in regard to transplantation to humans, including young children. On the other hand, neonates and infants do not yet produce antibodies to T-cell-independent antigens, which may mitigate organ rejection. As a result, graft survival after heart transplant in neonates and infants is significantly longer than in older individuals, even after ABO-incompatible heart transplantation.⁵ The potential to achieve tolerance – long-term graft survival without immunosuppression – will be fundamental to the ultimate success of xenotransplantation. The increased susceptibility of neonatal and infant immune systems to tolerance induction may facilitate this goal.

Immunologic Barriers to Xenotransplantation

The earliest human xenotransplants – including cardiac – used non-human primate (NHP) organs because of their phylogenetic proximity. However, the very limited survival, coupled with ethical and virologic concerns, rendered this approach impractical. Pigs were subsequently chosen as organ sources due to their comparable size, anatomy, and physiology. But hyperacute rejection ensued in NHP recipients, whereby preexisting natural antibodies (NAbs) to epitopes on porcine endothelial cells activated complement and coagulation cascades, leading to organ ischemia and death. NAbs exist irrespective of a previous exposure to pig antigens; instead, they exist due to cross-reactivity with shared antigens on common microbes. Most human and NHP anti-pig NAbs recognize a single carbohydrate, galactose-α1,3-galactose (Gal),⁶ and adsorption of these NAbs helped prolong pig organ survival from hours to days or weeks in NHP recipients in the 1990s. A decade later, nuclear transfer-based cloning methods permitted creation of pigs deficient in the enzyme which produces Gal, further prolonging pig organ survival in NHP recipients to weeks or even months (Figure 1).

Figure 1:

Chronology of Xenotransplantation: General timescales of obstacles and advances in pig to primate xenotransplantation are shown. HAR, hyperacute rejection; DXR, delayed xenograft rejection; AMR, antibody-mediated rejection; CR, chronic rejection; Ab, antibody; CRP, complement regulatory protein; GalT, α1,3 galactosyl transferase; KO, knockout; costim, costimulatory

Over the last twenty years, CRISPR/Cas9-mediated gene editing⁷ and improvements in immunosuppression further extended pig organ survival in NHPs. Additional carbohydrate moieties targeted by anti-pig NAbs – such as NeuGc and SDa – have been identified and the genes which encode the enzymes that produce them have been deleted (reviewed in⁸). Further genetic modification by introduction of human transgenes for coagulation and complement regulatory proteins, namely thrombomodulin and CD46, have facilitated prolonged cardiac xenograft survival in NHPs up to 945 days in the heterotopic position.⁹ Additional genetic modifications include: incorporation of the human macrophage-inhibitory protein (CD47) and other genes encoding anti-inflammatory molecules, as well as deletion of the growth hormone receptor gene to mitigate cardiac hypertrophy. It is important to note that the need for each individual genetic modification remains unproven, and it is possible that certain modifications (or combinations thereof) may be detrimental. Successful immunosuppression regimens typically include depletion of recipient T and B lymphocytes, and post-transplant maintenance with mycophenolate mofetil and anti-inflammatory agents. Monoclonal antibody blockade of the CD40-CD154 “co-stimulation” pathway is also important. Using these approaches, Langin and colleagues performed life-sustaining orthotopic heart transplants from Gal-knockout pigs that express human thrombomodulin and CD46 into baboons which survived up to 195 days.¹⁰ It is important to note that Langin and colleagues also used continuous ex-vivo perfusion to ensure donor organ preservation, and found that continuous mTOR inhibition was necessary to prevent pathologic myocardial hypertrophy and diastolic heart failure.

Human Investigations

The first pig-to-human solid organ xenotransplants were performed last year (Figure 1). Two research groups, one from New York University (NYU) and one from University of Alabama Birmingham (UAB), used brain-dead recipients to support pig kidney xenografts.^{11, 12} The NYU group used a Gal knockout donor and the UAB group used a donor with multiple genetic modifications. Both were short-term experiments lasting 3 days or less, owing to the experimental design. Early antibody-mediated rejection, cytokine storm, and coagulopathy did not occur, nor did transmission of zoonotic infections. The kidneys in the experiments from NYU functioned. Although the information gathered from these investigations is limited by the short duration of the studies, they do help substantiate the case for clinical trials.

In January of 2022, physicians and scientists at the University of Maryland performed the first pig-to-human orthotopic heart transplant.¹ The recipient was reliant on venoarterial ECMO and was deemed ineligible for human heart transplantation or durable ventricular assist device. The heart originated from a pig with 10 genetic modifications (10-GE), akin to that from the UAB group; the heart supported the recipient for 49 days, and the patient died on post-transplant day 60. The team used a T-cell and B-cell-depleting induction strategy with maintenance monoclonal antibody blockade of the CD40 pathway and mycophenolate mofetil (which ultimately transitioned to tacrolimus due to neutropenia). No acute cellular or antibody-mediated rejection was detected on myocardial biopsy by ISHLT criteria, but the patient was treated empirically for possible atypical antibody-mediated rejection. Levels of donor-specific antibodies remained low, but rose after administration of intravenous immune globulin (which in laboratory studies contains anti-pig NAbs). It is still unclear what precipitated the recipient’s demise, but reactivation of latent porcine CMV, which is known to accelerate xenograft rejection in NHPs,^{13, 14} is presently the leading theory. If so, improved screening and husbandry practices could alleviate such risks. Despite the numerous genetic modifications, requisite immunosuppression remains significant and highlights the need to better understand mechanisms to induce tolerance.

It remains unclear how many of the 10 genetic modifications are beneficial, and if any are actually harmful. Preclinical studies use NHPs as recipients, which have been instrumental in advancing human organ transplantation. However, biological differences between NHPs and humans limit the extent to which NHP xenotransplant models can predict outcomes in humans. The array of human anti-pig NAbs does not completely overlap with NHP anti-pig NAbs. For example, anti-pig NAbs to the carbohydrate NeuGc are present in humans but not NHPs because NHPs and pigs both express NeuGc. Knocking out NeuGc in pigs increases NAb binding in NHPs and increases xenograft rejection.^{15, 16} Variable amounts of data support each modification in preclinical models (some only in mice), but equivalent survival in NHP models of cardiac xenotransplantation have been achieved with fewer modifications.¹⁰ Therefore, human investigations, such as the early trials in brain-dead recipients, are essential to identifying the ideal donor organ. Additionally, more detailed NHP studies may reveal unexpected and possibly organ-specific effects of certain genetic modifications. For example, a human CD47 transgene, which was present in the pig heart transplanted to a patient, was associated with systemic inflammation, possibly related to increased levels of thrombospondin-1, an alternate CD47 ligand, in NHPs receiving pig kidney transplants with widespread tubular hCD47 expression.¹⁷

Tolerance

Despite advances in genetic engineering of donor organs to mitigate the innate immune response, many xenografts in NHPs continue to demonstrate rejection despite chronic immunosuppression. Herein lies a major opportunity: inducing tolerance, or re-educating the host immune system to recognize the donor organ as “self”, offers substantial theoretical benefits towards successful and durable xenotransplantation. Methods to induce xenograft tolerance currently under investigation include: 1) thymic transplantation, wherein recipient T lymphocytes are educated within the pig thymus in recipients treated with T cell-depleting antibodies; and 2) mixed chimerism, wherein bone marrow-derived cells from both pig and recipient co-exist.

Thymic transplantation supports recipient thymopoiesis in the presence of both donor and recipient antigen-presenting cells, thereby resulting in T-cell tolerance of xenograft and self (reviewed in⁷⁸,¹⁸). Early investigations demonstrated that pig thymus transplantation into thymectomized, T cell-depleted mice generated mature, murine T-cells that were tolerant of pig skin grafts¹⁹. This tolerance reflected a combination of intrathymic deletion of donor-reactive T-cells and production of regulatory cells recognizing pig antigens.¹⁸ In large animal models, non-vascularized thymic xenografts did not survive long-term but did achieve initial donor-specific hyporesponsiveness(reviewed in ⁷⁸). Vascularized thymic grafts, such as vascularized thymic lobe grafts or thymokidney composite grafts, have been more successful in large animals. Concomitant heterotopic transplant of fully MHC-mismatched cardiac allografts with vascularized thymic lobes facilitated long-term graft acceptance after four weeks of tacrolimus.²⁰ En-bloc heterotopic heart-thymus allografts also demonstrated survival without rejection until the study endpoint (200 days).²¹ In xenotransplant models, NHPs receiving pig vascularized thymus plus kidney xenotransplants have survived beyond 6 months and were found to harbor new thymic emigrants with donor-specific T-cell unresponsiveness (reviewed in⁸). Despite these very encouraging results, complete withdrawal of immunosuppression has not yet been attempted in this model.

Cultured thymus tissue implantation is an alternative to vascularized thymic transplantation, wherein slices of thymic tissue are cultured and then implanted into the quadriceps muscle as non-vascularized grafts following heart transplantation.²² In August of 2021, a team at Duke University performed the first orthotopic human heart transplant followed by cultured thymus tissue implantation in a recipient with heart failure and T-cell deficiency. Whether this recipient will tolerate withdrawal of immunosuppression is unknown, but perhaps similar methods for tolerance should be investigated for xenotransplantation.

Although thymic transplantation can induce T-cell tolerance across xenogeneic barriers, it does not directly induce B-cell or NK-cell tolerance in pig-to-NHP models. An alternative approach is mixed hematopoietic chimerism, wherein a recipient produces both self and donor hematopoietic cells through hematopoietic stem cell transplantation following non-myeloablative conditioning. Transient mixed chimerism, when combined with donor kidney transplantation has achieved renal allograft tolerance across HLA barriers in humans.²³ In xenotransplantation models, even low levels of more durable mixed chimerism have been shown to not only tolerize recipient T cells, but also to induce B-cell and NK-cell tolerance in a rat to mouse model (reviewed in¹⁸) and, more recently, among human lymphocytes generated in immunodeficient mice with human immune systems.^24–26 However, sustained mixed chimerism has not been achieved between pigs and NHPs. Different genetic modifications, such as the addition of human CD47, may prolong chimerism,²⁷ but the necessary extent and duration of chimerism to achieve tolerance for cardiac xenotransplantation is unknown and this may vary with the age of human recipients. For example, B cell tolerance may not be needed prior to cardiac xenotransplantation in infants who have not yet established adult NAb repertoires and who may spontaneously develop B cell tolerance, as observed in ABO-mismatched heart transplantation. Indeed, vascularized thymic transplantation, with its ability to tolerize T lymphocytes, might prove to be sufficient to achieve cardiac xenograft tolerance in this subgroup of patients.

Partial Heart Xenotransplantation

Many donor hearts are declined due to poor ventricular function and fewer are declined due to valve disease. Transplanting a component of a heart, such as the semilunar or atrioventricular valves, may offer the potential of a living, growing valve substitute for patients with valve disease. Of course, the immunologic considerations which underpin solid organ transplantation also exist with partial heart transplantation, but the extent to which immunosuppression is required to maintain a functional valve as opposed to a functional ventricle is unknown. Valve disease is far more prevalent than end-stage heart failure, so the potential for shortages in donor valve availability is high should this therapy become widely adopted. Herein lies a major opportunity for xenotransplantation; transplantation of genetically-engineered pig valves (or other cardiovascular components such as pulmonary arteries, pulmonary veins, or systemic arteries) to avoid the drawbacks of conventional valve prostheses, with possibly fewer immunosuppression requirements than conventional total heart replacement.

Conclusion

Numerous immunologic and developmental biology questions remain to be answered before xenotransplantation can be truly optimized. Efforts spanning from mechanistic to preclinical, translational studies must continue in order to improve this potentially groundbreaking solution to end-stage heart failure. Yet some questions may only be answerable in humans, and the rapid progress and durable success of orthotopic heart xenotransplantation in pig-to-NHP pre-clinical models justifies continuation of human clinical testing in both children²⁸ and adults.

Central Message:

Although several questions remain unanswered, the success of xenotransplantation in preclinical models supports the continuation of clinical testing in children and adults.

Perspective statement:

Advances in non-human primates have brought xenotransplantation closer to solving the severe shortage of organs for transplantation. Xenotransplantation could overcome this shortage and allow elective transplantation under ideal conditions. This year the first pig-to-human heart transplant raised hope that clinical trials are imminent. Tolerance may ultimately optimize xenotransplantation.

Abbreviations:

Ab

antibody

AMR

antibody-mediated rejection

CD

cluster of differentiation

CMV

cytomegalovirus

CR

chronic rejection

CRISPR

clustered regularly interspaced short palindromic repeats

CRP

complement regulatory protein

DXR

delayed xenograft rejection

ECMO

extracorporeal membrane oxygenation

Gal

galactose

GalT

α1,3 galactosyl transferase

HAR

hyeracute rejection

HLA

human leukocyte antigen

ISHLT

International Society for Heart and Lung Transplantation

KO

knockout

MHC

major histocompatibility complex

mTOR

mammalian target of rapamycin

NAb

natural antibodies

NeuGc

N-Glycolylneuraminic acid

NK

natural killer

NHP

non-human primate

SDa

sid antigen