Surveillance and prevention of infection in clinical xenotransplantation

SUMMARY

Xenotransplantation, the transplantation of living organs, tissues, or cells between species, carries the potential to address the global shortage of human organs for patients with end-stage organ failure. Recent advances in genetic engineering have improved prospects for clinical xenotransplantation by reducing immune and inflammatory responses to grafts, controlling coagulation on endothelial surfaces, and modifying viral risks, including the porcine endogenous retrovirus (PERV). Management of infectious risks posed by clinical xenotransplantation requires meticulous attention to the biosecure breeding and microbiological surveillance of source animals and recipients and consideration of novel infection control requirements. Infectious risks in xenotransplantation stem from both known human pathogens in immunosuppressed transplant recipients and from porcine organisms for which the clinical manifestations, microbial assays, and therapies are generally limited. Both known and unknown zoonoses may be transmitted from pigs to humans. Some pig-specific pathogens do not infect human cells but have systemic manifestations when active within the xenograft, including porcine cytomegalovirus/porcine roseolovirus (PCMV/PRV), which contributes to graft rejection and consumptive coagulopathy. The role of porcine endogenous retrovirus (PERV) in humans remains uncertain despite the absence of documented transmissions and the availability of swine with inactivated genomic PERV. New technologies, such as metagenomic sequencing and multi-omics approaches, will be essential for detection of novel infections and for understanding interactions between the xenograft, the host’s immune system, and potential pathogens. These approaches will allow development of infection control protocols, pathogen surveillance requirements, and tailored antimicrobial therapies to enhance the safety and success of clinical xenotransplantation.

INTRODUCTION

Clinical xenotransplantation has the potential to address the global shortage of organs for individuals with end-stage organ failure (1, 2). Other replacement technologies may reach clinical application in the future (3–5). Xenotransplantation is the implantation of viable cells, tissues, or organs between different species (6). The first documented attempt at xenotransplantation was in Lyon, France, in 1906 by Dr. Mathieu Jaboulay who transplanted a pig kidney onto the elbow of a 48-year-old woman who died 3 days later (7). Since then, advances in immunosuppression, life-supporting technologies, understanding of the immunology of xenotransplant rejection, and genetic modification of pigs have made clinical deployment of xenotransplantation feasible (8, 9). In particular, use of clustered regularly interspaced short palindromic repeat (CRISPR) technology has been used to delete antigenic targets of the human immune response, inactivate genomic copies of the porcine endogenous retrovirus (PERV), exclude growth hormone receptor, and express human genes for complement regulation, coagulation, and innate immunity and inflammation (10). Pig-to-non-human primate and human decedent model data have advanced the understanding of immune tolerance, immunosuppression, and xenograft function (11, 12). This has recently culminated in a small number of clinical xenotransplantation procedures (13). Two human recipients of pig hearts (University of Maryland, USA) with 10 gene edits exhibited life-supporting graft function for over a month before the xenografts failed (14, 15). The recipient of a kidney from a genetically engineered pig carrying 69 gene edits (Massachusetts General Hospital, USA) had stable life-supporting function until dying from cardiac complications at 52 days (in press). A GalTKO kidney transplanted 10 days after implantation of a ventricular assist device (NYU Langone Medical Center, USA) was removed after approximately 50 days (in press). A second renal xenograft was recently performed at the same center. A genetically modified pig liver was implanted as an auxiliary graft (in press) after removal of the right lobe of a patient’s liver for malignancy with survival of over 1 month.

Prevention of infection before, during, and after organ transplantation is an essential component of transplant clinical care (16, 17). Overall infectious risk in transplant patients relates to the net state of immunosuppression and epidemiological exposures to infectious agents (Fig. 1) (18–20). Unique to xenotransplant recipients are potential exposures to both human and porcine pathogens (21). There are a small number of known zoonotic pathogens (e.g., Hepatitis E virus, swine influenza virus, Streptococcus suis) transmitted to humans from pigs (22). In addition, there are porcine pathogens that have not been documented to cause disease in humans but can bind certain human cells in vitro (e.g., PERV) or may infect only pig cells but cause systemic adverse effects in xenograft recipients (e.g., porcine cytomegalovirus/porcine roseolovirus, PCMV/PRV) (23). Advancement of clinical safety in xenotransplantation will require improved understanding of the components of the “swine infectome” as well as validated microbiological tools for the surveillance of breeding herds and xenograft recipients (24, 25). As additional clinical and decedent xenotransplants are untaken, essential data will become available regarding the importance of various potential pathogens for the recipient, as well as any risks to their contacts and the community. Further data are needed regarding the efficacy of antimicrobial agents for potential human pathogens from swine. These data will provide a basis for safety assessments, regulatory frameworks, public education, and informed consent for future clinical trials.

Fig 1.

Timeline of potential infections after pig-to-human organ xenotransplantation. Standard immunosuppression for organ transplantation is associated with a predictable pattern of human infections over time based on the intensity of immune deficits and epidemiological exposures. A similar pattern of infectious risk is expected following xenotransplantation. Introduction of newer immunosuppressive agents for induction or maintenance therapy (e.g., costimulatory blockade, cytokine inhibition, complement inhibitors) is expected to alter the risk of various infections over time.

IMMUNE BARRIERS TO XENOTRANSPLANTATION

There are at least four major immunological barriers to successful organ xenotransplantation between pigs and humans and non-human primates. The first is hyperacute rejection (HAR), which is caused by xenoreactive natural (preformed in all humans) antibodies and complement in the recipient reacting against the vascular endothelial cells of the pig organ (26–29). Second, acute vascular rejection (AVR) is caused by the combined effect of elicited xenoreactive antibodies and activated innate immune host cells, e.g., natural killer cells and monocytes. In combination, these stimuli (anti-graft antibodies and activated host cells) result in endothelial activation within the pig organ, resulting in inflammation, thrombosis (platelet aggregation and activation of the coagulation cascade), and vascular injury (30–32). Third, delayed xenograft rejection (DXR) occurs, the more variable counterpart of classical T-cell-mediated rejection seen with allografts. Finally, xenografts may also be subject to chronic rejection (CXR) in a manner analogous to that which occurs in allografts, which is largely antibody mediated but includes T lymphocytes and chronic inflammation. These will be considered briefly.

Antibody-mediated rejection and complement activation

Antibodies directed against xenoantigens cause hyperacute xenograft rejection (HAR) within minutes of reperfusion. Pre-formed antibodies target glycan xenoantigens, galactose-a 1,3-galactose (Gal), N-glycolylneuraminic acid (Neu5Gc), and Sda, which are present in human and non-human primate sera (33–35). IgM isotypes are the main contributor to acute antibody-mediated rejection (36). Anti-Gal antibodies represent approximately 1% of circulating immunoglobulins in humans. The complement cascade is activated by either binding of antibody to xenoantigens (classical pathway) or in the absence of antibody as part of the innate immune response (alternative and mannose binding lectin pathways) (37, 38). Porcine complement regulatory proteins in the xenograft are less efficient in cleaving human complement products (39); in preclinical studies, improved xenograft survival has been achieved by mitigating complement activation via monoclonal antibodies targeting C1-esterase or C5 (40, 41). Anti-pig antibodies directed against swine leukocyte antigen (SLA) classes I and II may also be important in later phases of xenograft immunity and is not targeted by the genetic removal of alpha-gal sugars (42–44). A serosurvey study of individuals awaiting kidney transplantation demonstrated that up to 27% will have anti-SLA I antibodies (45).

Cellular immune responses

Adaptive and innate cell-mediated immune responses are elicited by xenografts and contribute to rejection (46). NK cells and macrophages have also been found to infiltrate solid organ xenografts (47, 48). T-cell-dependent and T-cell independent macrophage activation occurs in preclinical models (49, 50). Human monocytes may upregulate inflammatory cytokines and cause activation of porcine endothelial cells (51). Pig cells do not express a human-compatible CD47 and are not able to recognize macrophage inhibitory receptors (52). NK cells fail to recognize major histocompatibility complex (MHC) molecules on xenogeneic cells (“not self”), which leads to NK cell-mediated destruction by granzyme and perforin (53, 54). Human T cells can directly recognize swine leukocyte antigen (SLA) molecules or xenoantigens presented on antigen-presenting cells (55–57). CD4+ cells elicit a more vigorous immune response to xenoantigen compared with alloantigens (58). T cells also promote NK cell responses and antibody production to xenoantigens.

Coagulation dysfunction

Coagulation dysfunction is a major barrier to xenotransplantation (59, 60). Immune injury to vascular endothelium with exposure of subendothelial tissue and tissue factor initiates an influx of inflammatory cells with upregulation and secretion of proinflammatory and procoagulant factors (61). This process results in thrombotic microangiopathy (TMA) of the xenograft (62). Regulatory proteins involved in the porcine coagulation cascade are highly conserved but interact less effectively with human regulatory proteins. Porcine tissue factor pathway inhibitor and thrombomodulin fail to inhibit factor Xa and thrombin, respectively, in non-human primates and in vitro with human inhibitors (63–65). Such accelerated procoagulant states may precipitate systemic consumptive coagulopathies.

PREVENTION OF XENOGRAFT REJECTION: GENETIC EDITING AND IMMUNOSUPPRESSION

Two major approaches have been used to prevent rejection, including (i) genetic editing of the organ-source pig and (ii) the administration of novel immunosuppressive regimens that block the CD40/CD154 T cell co-stimulation pathway and prevent T-cell activation to prevent both cellular and elicited antibody-mediated responses (26, 66). Gene editing generally includes combinations of (i) blocked expression of the three known glycan xenoantigens (triple knockout or TKO) against which humans have natural anti-pig antibodies and are the targets of HAR (67); (ii) the introduction into the pig of human genes that provide protection against human immune responses and coagulopathy, e.g., complement-regulatory or coagulation-regulatory genes (27); (iii) inactivation of genomic copies of porcine endogenous retrovirus (PERV); and (iv) other modifications, such as inactivation of the genes encoding growth hormone receptors. To date, the respective roles of the various genetic edits and newer immunosuppressive therapies remain unclear (68).

Gene editing

Engineering of the pig genome in vitro and generation of live pigs through cloning (somatic cell nuclear transfer) was accelerated by “gene targeting” using homologous recombination with the discovery of programmable nucleases, including the CRISPR/Cas9 system. As a result, three genetic engineering strategies have been used to reduce antibody binding and improve regulation of the complement, coagulation, and inflammatory pathways (Table 1).

TABLE 1.

Genetic modifications used in swine breeding for xenotransplantation

Genetic manipulations	Target	Potential gene targets for xenotransplantation
Pig breed		Various pig breeds
Endogenous retrovirus	Virus	Porcine endogenous retrovirus (PERV A, B, C, AC)
Knockout	Carbohydrate antigens	GGAT1 (α−1,3-glycosyltransferase)
	Carbohydrate antigens	B4GalNT2 (glycosyltransferase)
	Carbohydrate antigens	CMAH (cytidine monophosphate-N-acetylneuraminic acid hydroxylase)
	Organ growth	Growth hormone receptor
Added human transgenes	Complement regulation	CD46 (hMCP, human membrane cofactor protein)
	Complement regulation	CD55 (hDAF human decay-accelerating factor)
	Coagulation	THBD (human thrombomodulin gene)
	Coagulation	EPCR (endothelial cell protein C receptor)
	Innate Immunity	CD47 (block SIRPα tyrosine phosphorylation)
	Inflammation, apoptosis	HO1 (heme oxygenase-1)
	Inflammation, apoptosis	HA20 (human A20)

In the 1990s, advances in embryo microinjection and in vitro fertilization allowed the surface expression of one of three human complement pathway regulatory proteins (hCPRPs), hCD46, hCD55, or hCD59 (66, 69–72). This resulted in reduction of complement-mediated cell injury and prolonged survival of pig cells and organs (73–76). Surprisingly, even with a tolerogenic “mixed chimerism” (bone marrow plus xenograft) strategy, DXR was observed even without elicited anti-pig antibodies (69, 74–76) and with donor-specific cellular immune unresponsiveness. The exact mechanism of such delayed injury remains uncertain. To reduce anti-pig antibody binding to the xenograft, the three principle carbohydrate antigen targets were removed from pigs by triple “gene knock-outs” (TKO), including the Gal-1-3α,galactosyl transferase gene (GalTKO), the cytidine monophospho-N-acetylneuraminic acid hydroxylase (CMAHKO), and α,4 galactosyl transferase (α4GalKO) (77–79). Organs with GalTKO, the TKO, or with complement regulatory gene modifications alone generally avoided HAR and survived for days or weeks in immunosuppressed primates before exhibiting DXR.

Molecular incompatibilities between pig thromboregulatory molecules and human blood proteins result in dysregulation of the coagulation cascade via inefficient deactivation of human thrombin by porcine thrombomodulin (TBM), ineffective binding of activated human protein C (aPC) to porcine endothelial protein C receptor (EPCR), and dysregulated adenosine metabolism consequent to low surface CD39 expression. Thrombodysregulation is manifested both within the organ xenograft (thrombotic microangiopathy, TM) and systemically (consumptive coagulopathy) and is enhanced by anti-pig antibody and associated complement cascade activation. Infection due to porcine cytomegalovirus (PCMV) may also contribute to endothelial activation with thrombosis and consumptive coagulopathy (31, 32). Consequently, multiple groups produced pigs that express one or more human coagulation pathway regulatory genes. TKO plus complement regulatory protein-expressing heart and kidney xenografts survive over 1 year in primates, which were often preselected for low levels of anti-pig antibodies and using experimental costimulation-based immunosuppression (80–84).

A third genetic breakthrough was reported in 2015 when CRISPR-based gene editing was successfully used to inactivate 62 copies of pig genes encoding porcine endogenous retrovirus (PERV), first in cell lines and then to generate viable, fertile pigs (85, 86). Subsequent to the cloning of PERV and the demonstration of the presence of cell surface receptors for PERV-A and B on normal cells as well as productive infection of transformed, permissive human cell lines, there was concern that pig-to-human xenografts might cause infection in the xenograft recipient and risk pandemic viral infection (87–90). PERV infection of normal human cells and of humans exposed to porcine tissues has not been demonstrated (discussed below).

Subsequent cloned pig lines include many genetic modifications (see Table 1), including TKO, multiple human complement regulatory (hCD46, hCD55, and hCD59), thromboregulatory molecules (hTBM, hEPCR, hCD39, and hTFPI), molecules to inhibit phagocytosis (hCD47) or NK cell activation (HLA-E for “self-recognition”) and inhibit adaptive immunity (PD-L1), and with PERV polymerase gene inactivation. Some lines carry KO of growth hormone receptor (not required in miniature swine) and swine histocompatibility antigens. Testing of cells and organs from pigs with various combinations of these genetic modifications is currently underway by multiple groups.

The role of costimulatory blockade

Intensive immunosuppression has been deployed in pig to non-human primate studies, including induction with antithymocyte globulin, anti-CD20, and a C1 esterase inhibitor, and maintenance with CD40/CD154 costimulatory blockade, rapamycin, methylprednisolone, and, often, anti-IL6 (91–93). The requirement for each component of these regimens is uncertain (94). The allospecific T-cell population responding to a transplanted organ consists of both naïve and memory lymphocytes. Combined antigen-specific T-cell receptor stimulation and costimulation through various pathways (CD28 and CD40) are needed for complete T-cell activation (95). Costimulatory blockade has been an attractive target for organ transplantation, including xenotransplantation, given the potential to reduce or eliminate (anergy) stimulation of naïve T lymphocytes in the setting of new antigenic stimulation (96, 97). While the families of costimulatory molecules are complex, there are two families of costimulatory pathways most important for adaptive immune responses: the Ig-related superfamily (e.g., CD28-CD80/86, inducible costimulator-inducible costimulator ligand) and the TNF-related superfamily (e.g., CD40-CD40L/CD154). There are coinhibitory pathways related to each family. Inhibition of naïve T-cell responses by costimulatory blockade is consistent with the high risk of EBV replication and EBV-induced PTLD. Early preclinical success of anti-CD40L therapy could not be immediately translated into clinical practice due to thromboembolic complications, likely due to direct activation of platelets through CD40L on the surface and formation of immune complexes. The success of newer anti-CD40L/CD154 molecules in preclinical xenotransplant models has stimulated adoption of these molecules for clinical trials (82, 93, 97, 98).

Clinical experience with available costimulatory blockade molecules suggests a pattern of infections different from that observed with calcineurin-inhibitor (CNI)-based therapies (99, 100). In renal allotransplantation, tacrolimus in combination with belatacept reduced the incidence of graft rejection but with an increased risk of infections, including cytomegalovirus, Pneumocystis pneumonia, progressive multifocal leukoencephalopathy (JC polyomavirus), and central nervous system Epstein–Barr virus-induced post-transplant lymphoproliferative disease (PTLD) (101). Addition of induction therapy with either antithymocyte globulin or alemtuzumab appears to reduce these risks (99, 100, 102, 103). The viral infections encountered under costimulatory blockade also appear to be more severe and more resistant to therapy (103). The use of costimulatory blockade in induction or maintenance immunosuppression for xenotransplantation raises the potential for more frequent or more severe viral infections. This will necessitate prolonged monitoring of xenograft recipients.

RISK OF INFECTION IN XENOTRANSPLANTATION VS ALLOTRANSPLANTATION

The risk for infections in solid organ transplant recipients is determined by the epidemiologic exposures of the organ donor and recipient and the recipient’s overall immune status or “the net state of immunosuppression.” The risk is largely determined by the intensity of immunosuppression but also by other factors (e.g., lines, drains, viral coinfections, technical issues, nutritional status), which may contribute to poor infectious clearance. As a result of standardized immunosuppressive regimens, a pattern of common opportunistic infections has been recognized, which has provided a basis for prophylaxis (i.e., vaccination, antimicrobial use) and differential diagnosis (18, 20, 104). Changes in the immunosuppressive regimen, intensification for the treatment of graft rejection, or in epidemiological exposures will alter the timeline of infection (105, 106). Unfortunately, other than therapeutic drug monitoring of immunosuppressive agents to prevent toxicity and minimize infectious risk, immune function assays have not yet emerged to guide the management of immune suppression prospectively (107–109). The impact of factors, such as genetic polymorphisms in immunoregulatory genes or nutritional status, are not easily measured (110, 111). Genetic modifications of donor swine have included insertion of functional human complement regulatory transgenes for CD46 (hMCP and human membrane cofactor protein) and CD55 (hDAF human decay-accelerating factor) to reduce complement activation and deposition on porcine endothelial cells. These glycoproteins also serve as receptors for many common human viruses, including human CMV (one of many), adenovirus, human herpesvirus 6 (HHV6), and rubeola (measles). The clinical impact of such modifications is unknown. Measles is included in standard vaccinations (MMR). However, HHV6 and human CMV viremias are common among immunosuppressed transplant recipients, which might be increased by intensified immunosuppression (see section on costimulatory blockade above), and could, potentially, utilize these receptors. The impact of other genetic modifications in donor swine remain to be clarified.

SCREENING OF DONOR SWINE

The risk for infection in porcine xenotransplantation is a function of latent or unknown pathogens activated in immunosuppressed human transplant recipients, porcine organisms carried by the xenograft and the nature and intensity of the immunosuppressive regimen (Fig. 1) (16). Swine may carry organisms with the capacity for which microbiological assays and therapies may be limited. For example, the common colonizer of swine, Streptococcus suis, may be pathogenic in humans but may not be routinely speciated by clinical microbiology laboratories. Testing of swine requires studies for both latent (i.e., serologies) and active infections (nucleic acid, NAT, or antigen tests). The optimal tissues (e.g., blood, biopsies), frequency, and animal age for testing remain to be established. Given the lack of validated diagnostic assays for pig organisms in humans, both pathogen-specific and broad-range, unbiased (metagenomic sequencing) surveillance may be recommended for clinical studies (112).

Screening for most porcine bacteria, fungi, and parasites can be performed in veterinary or clinical microbiologic laboratories and using standard laboratory techniques. Repeated testing for species-specific organisms (e.g., porcine cytomegalovirus/porcine roseolovirus, PCMV/PRV, porcine circoviruses, PCV, porcine pestivirus, Salmonella spp..) may be needed to avoid asymptomatic transmission events (113, 114). Fewer assays are available for pig viruses, and data on the biology of these are largely limited to those of some commercial impact. Studies of the infectivity of various porcine organisms for immunosuppressed humans are needed (16, 115).

Consensus is needed regarding optimal methods for testing. The potential importance of specific swine-derived pathogens in immunosuppressed humans, and the urgency for the development of clinical-grade microbiological assays, can be considered based on extensive experience with similar pathogens in human allotransplantation and in xenotransplantation in non-human primates (16). Possible groups include:

1. Known, common, zoonoses for immunologically normal hosts (hepatitis E virus, swine influenza virus, Mycobacteria spp., rabies virus) and geographically restricted zoonotic organisms (e.g., Nipah virus, Schistosoma spp., African swine fever virus, Menangle virus, porcine circovirus type 4 [PCV4]).

2. Zoonotic pathogens of immunosuppressed hosts (e.g., Toxoplasma gondii, Strongyloides spp., Aspergillus spp., Cryptococcus spp., Cryptosporidium spp.).

3. Porcine organisms similar to common pathogens of immunosuppressed human transplant recipients (e.g., porcine adenovirus, porcine parvovirus 1, porcine respiratory coronavirus, parainfluenza virus 3, porcine pestivirus).

4. Swine-specific pathogens that replicate only in pig cells and, as a result, in the xenograft (porcine cytomegalovirus/porcine rosiolovirus [PCMV/PRV], porcine circovirus [PCV 1–4], porcine lymphotropic herpesvirus [PLHV 1,2], porcine endogenous retrovirus [PERV A, B, C, AC]).

5. Organisms routinely tested for the health status of swine (e.g., porcine enterovirus spp., Lawsonia intracellularis, porcine epidemic diarrhea virus, transmissible gastroenteritis virus, porcine delta coronavirus, Brucella suis, porcine reproductive and respiratory syndrome virus, porcine epidemic diarrhea virus, pseudorabies virus).

Lists of potential pig-derived pathogens to be excluded from source animal herds for clinical xenotransplantation have been published as a basis for pig breeding for xenotransplantation, termed “designated pathogen-free” swine (116–118). These lists may vary with the geographic origin of the swine facility (115). Such lists have been adopted by regulatory authorities to govern pig breeding (16, 115, 119–121). Management of infectious risk includes serial monitoring of donor herds during breeding to exclude designated potential human pathogens and to assure animal health. Retesting of source animals just before organ procurement is also desirable. Multiple assays have been developed for screening of swine. It is worth noting that each assay may need to be validated separately for pig screening and diagnosis and for monitoring of human xenograft recipients. NAT assays, while sensitive, are limited to detection of actively replicating species. Stool samples may be used for screening for pathogenic bacteria (e.g., Salmonella spp.) and carriage of some viruses. Serologic testing is of special utility given detection of latent or prior infection of otherwise healthy swine (16, 122). For example, Fischer, Halecker and others describe the need for a combination of PCR-based (young pigs) and antibody detection methods (adults) for PCMV/PRV in pigs for xenotransplantation (113, 114, 123). Of interest, cross-reaction of antibodies between PCMV and HHV6 has been described (124). Plotzki et al. describe an ELISA and a Western blot assay using recombinant glycoprotein B of PLHV-1 and identified antibodies in tested swine but not humans (125). Characterization of genomic and circulating PERV has been described by multiple authors (87, 88, 126, 127). Screening for porcine circoviruses has been performed by viral metagenomic testing of stool samples and NAT testing of blood (128–130). Optimal screening requirements have not yet been defined but will generally include organisms important for swine health, known human pathogens, swine organisms known to impact graft function, such as PCMV, and PERV. Such animals are raised in biosecure facilities to prevent introduction of potential pathogens from other animals (including birds, amphibians, rodents, swine) and caretakers (25, 131–133). Biosecure facilities may receive filtered air, sterile water, and irradiated food, with specialized infection control entry requirements for personnel (134, 135). Swine introduced into the breeding facility may be subject to early weaning protocols and cesarean derivation to reduce pathogen transfer (32, 136, 137).

Regulations governing clinical xenotransplantation vary among countries and regulatory agencies (138). A xenotransplant product is defined as live cells, tissues, or organs from a non-human animal source. In most cases, genetically modified pig organs are considered a medicinal product (139). The aspects of regulation are typically broken down into (i) source herd, (ii) source animal, (iii) product manufacturing, and (iv) patient-based aspects (140, 141). Regulatory documents have been produced by the U.S. Food and Drug Administration (FDA), European Medicines Agency (EMA), World Health Organization (WHO), International Xenotransplantation Association (IXA), and International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) (138, 142, 143). Some controversy surrounds the optimal subjects for clinical trials; generally, children are prohibited from receiving xenografts, and non-human primates are prohibited as a source of xenograft organs (144). Requirements for preclinical experience in NHP is common, but decedent experience may suffice.

PORCINE ORGANISMS OF IMPORTANCE TO CLINICAL XENOTRANSPLANTATION

PCMV/PRV does not infect human cells although the virus may infect xenografts and cause systemic complications. In preclinical studies in immunosuppressed non-human primate (NHP) xenograft recipients, PCMV was activated in the porcine xenograft with shedding of pig cells (“chimerism”), virus, and viral DNA into the circulation with endothelial activation and resultant systemic consumptive coagulopathy and accelerated renal xenograft rejection (30–32, 145–148). These effects may have been observed in the first cardiac xenotransplant recipient emphasizing the role of optimal microbiological assays. The source animal was PCMV negative by sensitive nucleic acid test (NAT), but PCMV infection was detected in the recipient; serologic screening assays were not available (14, 15). Graft rejection and injury were similarly observed with renal xenografts in decedent recipients from swine NAT-negative for PCMV/PRV prior to procurement, but with reactivation of latent infection following transplantation and immunosuppression (149).

There are no reports of porcine circovirus (PCV 1–4) disease in humans. However, rising PCV3 viral loads in baboons receiving cardiac xenografts are consistent with porcine graft infection (128, 150, 151). Porcine lymphotropic herpesviruses (PLHV) can cause post-transplant lymphoproliferative disease in swine, although do not replicate in human cells (152). In a pig-to-non-human primate (NHP) xenotransplantation study, three kidney recipients developed rapid graft loss in the setting of newly acquired adenovirus infection with post-mortem kidney specimens staining positive for adenovirus (153). Despite data suggesting a role for swine in the evolution of SARS coronavirus (CoV)-2, there is no evidence that swine carry CoVs pathogenic for humans. However, pigs carry the common cellular receptor for the viral spike or S-glycoprotein of SARS-CoV-2, angiotensin-converting enzyme 2 (ACE2), suggesting that infection is a theoretical possibility. Six CoVs cause significant infection in pigs; CoVs in swine are largely enteric pathogens, but porcine respiratory coronavirus causes mild respiratory infections in swine (154, 155). In vitro, porcine coronaviruses can replicate on some transformed and primary human cell lines (156, 157). Recombinant swine enteric viruses have been reported. Some potential pathogens (e.g., porcine pestivirus) may be excluded by herd vaccination. Recent reports of novel human strains of circovirus causing hepatitis in immunocompromised hosts demonstrates the need to continually monitor source animals for changing epidemiology of zoonotic pathogens (158–161).

Porcine endogenous retrovirus (PERV)

Special consideration must be given to the porcine endogenous retrovirus (PERV), which exists in three homologs (PERV-A, -B, and -C) (162, 163). PERV is a ubiquitous genomic provirus of pig cells; viral replication and reinsertion into the pig genome are ongoing, asymptomatic processes. PERV mRNAs are expressed in all pig tissues from most swine; the size and quantity of PERV mRNA transcripts vary between strains and tissues (87, 164–167). Between 7 and 14 full-length proviral copies of PERV-A and B exist in domestic pig genomes; not all are infectious (87, 168). The three homologous C-type porcine endogenous retroviruses (PERV A, -B, and -C) have different host ranges (87, 89, 169–173). In vitro, PERV-A and -B can infect pig cells and adenovirus-5 transformed, permissive human target cells (HEK293); PERV-C infects only pig cells (87–89, 169, 172–177). Increased efficiency of viral replication on HEK293 cells was observed when viral recombination occurred between genes encoding the receptor-binding domain of PERV-A and transmembrane region of the envelope of PERV-C (PERV AC). Spontaneous development of PERV AC occurs in vivo in swine with reintegration into the host genome (127). Most primary human cells have not been infectable by PERV (174–182). In a clinical study, despite persistent circulation of pig cells in 23 patients for up to 8.5 years, no PERV infection of human cells occurred (183). In xenotransplantation studies of human xenograft recipients and in those exposed to pig cells or organs via extracorporeal perfusion, no evidence of viral replication, genomic integration, or development of anti-PERV antibodies was observed (121, 184). Swine lacking functional PERV-C have been used in clinical studies of porcine islet xenotransplantation without evidence of transmission to recipients (121, 162, 184–187). PERV expression in pancreatic islets does not correlate with expression in circulating leukocytes; the impact of islet encapsulation on transmission risk is uncertain (188, 189). No evidence of PERV infection of human hosts has been detected in recent clinical xenotransplants (personal communication), although with relatively short exposures to xenografts. PERV replication is defective in non-human primates, making primate studies less informative regarding PERV infectious risk (190, 191).

Quantitative NAT or whole genome sequencing (WGS) can assess PERV infections in vivo (192). Should transmission occur, PERV is susceptible in vitro to nucleoside and non-nucleoside reverse transcriptase inhibitors in common clinical use (176, 193–198). Using CRISPR–Cas9 to target the polymerase gene of PERV elements, inactivation of PERV in the immortalized porcine kidney epithelial cell line PK15 was demonstrated (85). Cells carrying inactivated PERV polymerase genes lacked reverse transcriptase activity; such cells demonstrate resistance to retroviral superinfection (199). PERV C-deleted pig fibroblasts were used in somatic cell nuclear transfer to generate PERV-inactivated embryos; the offspring are PERV negative and do not appear to carry off-target genomic CRISPR modifications (86).

MONITORING THE XENOGRAFT RECIPIENT FOR INFECTIOUS DISEASES

Strategies for infectious disease surveillance and antimicrobial prophylaxis will enhance safety in clinical trials (118, 200). Routine pretransplant screening of human xenograft recipients for common pathogens of immunocompromised transplant recipients is required (Table 2) (16–18, 201). Prophylaxis is based on serologic risk stratification (e.g., for human CMV) and standard practices (e.g., for Pneumocystis and Toxoplasma spp.).

TABLE 2.

Routine microbiological testing of xenograft recipients^a

Virus	Testing method
Porcine endogenous retrovirus (PERV) A, B, C, ACb	Qualitative and quantitative (QNAT) nucleic acid testing (NAT); antibody-based tests (serology, ELISA, Western blot)b
Porcine lymphotropic herpesvirus type 2 (PLHV-1–2)	QNATb
Porcine circovirus (PCV 1–4)	QNAT
Porcine cytomegalovirus/porcine roseolovirus (PCMV/PRV)	NATb; serology
Porcine adenovirus	NAT
Porcine pestivirus	NAT
Human cytomegalovirus (HCMV) – per risk status	QNAT, serology
Human Epstein–Barr virus (EBV) – per risk status	QNAT, serology
BK polyomavirus (kidney recipients) – per protocol	QNAT
Pig cell chimerism in circulation (PBMC)	QNATb (e.g., P-MHC class I gene; p-mtCOII gene) in recipient PBMC DNA
Unknown pathogens	Metagenomics or next generation sequencing (202–205)

^a Additional testing may be needed for individuals with infectious syndromes. QNAT: quantitative nucleic acid test; normalized or corrected for circulating pig cells. P-MHC: porcine major histocompatibility complex; p-mtCOII: pig mitochondrial cytochrome c oxidase subunit II gene.

^b Quantitative NAT for PERV, PCMV, PCV, and other viruses must be normalized against chimerism studies to correct for the number of circulating infected pig cells in blood samples. Modified from reference 17.

Most porcine pathogens are eliminated by strict breeding practices in biosecure pig farms, although a conservative approach to pathogen screening is still advised (133, 206). Lists of organisms used for pig screening may be applied to monitoring of recipients, although the optimal screening paradigm for clinical xenotransplantation is unresolved (118, 121, 206–209). Commercial development of NAT for pig organisms is limited by lack of agreement on target pathogens, sequences, and clinical specimens for testing (e.g., blood vs tissues), and requires optimization of nucleic acid extraction and reaction kits, regulatory approval and laboratory certification for new assays. In addition, the financial incentive for xenotransplantation NAT testing has not yet been demonstrated.

The intensity and duration of immunosuppression required to prevent rejection in xenotransplantation patients may necessitate longer periods of microbial surveillance (68). Published data from human recipients of various genetically modified porcine xenografts and decedent recipients have demonstrated a limited spectrum of infections, including human viruses, bacteria (associated with wound infection), and fungal (Candida) species, using both metagenomics and routine culture systems (data in press). Assay selection is critical. For example, despite negative NAT testing of a cardiac xenograft donor animal, in the absence of serologic testing, PCMV viremia was detected in the recipient (14). Although the source animal was positive for PERV A and B, serial plasma microbial cell-free DNA testing (mcfDNA testing; Karius) did not detect PERV or PCV infections (14). Metagenomic testing of a human recipient of a pig kidney (with negative serology and PCR for PCMV and PCV and with inactivated PERV) revealed no porcine infections in the recipient but reactivation of known, prior human viral infections, including BK polyomavirus and Epstein–Barr virus (Fishman, data in preparation). Microbiologic surveillance was negative for PERV A, B, and C in two kidney xenotransplants from PERV A- and B-positive pigs in a decedent recipient using blood and plasma viral amplification (210). Data from other xenokidney decedent recipients have not yet been published (12, 211).

Quantitative NAT for PERV, PCMV, PCV, and other pig viruses must be normalized against chimerism studies (e.g., porcine MHC class I gene; p-mitochondrial COII gene) to correct for the number of circulating, sloughed, infected pig cells in blood samples (212). Knowledge of antiviral treatment options and the impact on xenograft rejection for porcine viruses will also dictate the clinical response (201). Surveillance can be supplemented by unbiased metagenomic testing (e.g., cell-free DNA and RNA sequencing compared with sequence databases) for unknown or unsuspected pathogens (134, 213, 214). Some of the limitations of metagenomic sequencing are worth noting. These include challenges in sample procurement, storage, nucleic acid extraction and library preparation, limitations in microbial sequence databanks and bioinformatics, background human and contaminant nucleic acids, sequencing (costs, turnaround times, quality), and data interpretation by nonexpert clinicians.

For the symptomatic xenograft recipient, common sources of human infections associated with technical issues or reactivation of latent or colonizing human infections are most likely. For these, bacterial and fungal cultures and NAT for human viral infections are deployed. Assessment for swine-specific viral pathogens requires both serologic testing and NAT methods, which are currently available only in the veterinary and research environments. Seroconversion for both pig and human pathogens may be delayed under intensive immunosuppression in the immunosuppressed recipient.

Advance archiving of biological specimens is required to be used in subsequent epidemiological investigations of the recipient as well as social contacts and clinical staff exposed to donor swine and recipient body fluids. Such samples are governed by regulatory authorities but should be maintained initially for at least 10 years or the survival of the recipients.

MULTI-OMICS APPROACHES FOR EARLY TRACKING OF INFECTION, GRAFT FUNCTION AND REJECTION

Transplant medicine is beginning to take a pre-emptive approach to infectious diseases screening and monitoring of at-risk patients developing graft rejection or infection (215, 216). One of the tools for post-transplant monitoring is multi-omics, which integrates multiple simultaneous techniques, such as targeted and unbiased DNA and RNA sequencing, transcriptomic, epigenetic, metabolomic, and proteomic analyses (217, 218). Serial studies provide insights into the evolution of biological processes in individuals, including infection, graft function, or rejection, with the potential of tailoring antimicrobial prophylaxis and immunosuppression strategies (216, 219). For example, in a small study of lung transplant recipients, multi-omics analyses of bronchoalveolar lavage specimens were used to predict changes in FEV1 (220). Multi-omics analyses have examined interactions between infection and acute rejection in kidney transplant recipients (218, 221). Single-cell multi-omics analyses of peripheral blood mononuclear cells in liver transplant recipients identified unique patterns of acute cellular rejection (149). Serial multi-omics studies of peripheral blood cells and tissues in human decedent recipients of pig xenografts demonstrated evolving changes associated with ischemia reperfusion injury, antibody-mediated rejection and graft dysfunction, (222). Coupled with targeted and unbiased metagenomic sequencing, multi-omic approaches may help to define immunologic and infectious risks in the individual.

EMERGING INFECTIOUS DISEASES, GENOMIC EVENTS, AND XENOTRANSPLANTATION

Infectious diseases of swine pose both public health and economic challenges (161). Over the past three decades, many new organisms have been identified in pigs ((e.g., porcine delta coronavirus, PDCoV; swine acute diarrhea syndrome coronavirus, SADS-CoV; and influenza viruses ([influenza D virus, IDV)]), although most have not (yet) been associated with human disease (115, 223). Contacts between humans, swine, rodents, birds, and amphibians in markets and farms provide new exposures to humans as well as opportunities for recombinations producing novel pathogens for which microbiologic assays may not be available (224, 225). Zoonotic organisms associated with outbreaks carry high rates of capacity for mutation (e.g., Nipah virus, porcine respiratory coronavirus, PRCV, transmissible gastroenteritis virus, TGEV). The global impact of such new pathogens may be amplified by international travel and inadequate regulation of animal sales and markets. Microbiologic surveillance will require pathogen agnostic techniques, such as unbiased metagenomic sequencing. The impact of immunosuppression on the manifestations of such novel pathogens is unknown. Lists of pathogens to be excluded from breeding colonies will need constant revision.

THERAPEUTICS FOR PORCINE VIRUSES

Therapeutic options available for porcine viral infections are limited. Some in vitro and in vivo evaluations of antiviral activity have been performed (198, 226–228). PCMV replication is poorly inhibited in vitro by ganciclovir and acyclovir and to a limited degree by cidofovir and foscarnet with their associated toxicities (227, 229, 230). Ganciclovir can be used at therapeutic doses for prophylaxis in the pig-to-non-human primate (NHP) model and in vitro but is ineffective for active infection in the immunosuppressed NHP recipient (231, 232). No antivirals active against porcine lymphotropic herpesviruses (PLHV) or porcine circoviruses (PCV) currently exist, although these agents are not known to be pathogenic in human hosts. For PERV, antiretroviral drugs, which include reverse transcriptase inhibitors (e.g., tenofovir) and integrase inhibitors (e.g., dolutegravir), have in vitro activity (194, 233). Some authors recommend combination antiretrovirals for postexposure prophylaxis (e.g., needlestick injury) or empiric therapy for PERV in the absence of PERV inactivation (17, 201).

INFECTION CONTROL CONSIDERATIONS

Human exposure to pigs, pig organs, and recipients of porcine xenografts, requires a protocol-based program of infection control. Zoonotic infections acquired from pigs have been well described for common viruses (e.g., influenza A, hepatitis E), bacteria (e.g., Erysipelothrix rhusiopathiae, Leptospira spp.), fungi (e.g., dermatophytes), and protozoa (e.g., Cryptosporidium spp.) (22, 223). The actual degree of risk for infection due to less common organisms is unclear; guidance is based on experience with exposure to common pathogens from blood or bodily fluids (234). Thus far, there are no reports of transmission of microorganisms unique to swine (e.g., PCMV, PERV, PCV) to humans frequently exposed (e.g., veterinary workers or pig farmers) (235). Similarly, there have been no transmission events to care providers or clinical staffs associated with the breeding of swine as source animals or with clinical and decedent xenotransplants (12, 14, 15, 210). With implementation of standard infection control practices, acquisition of swine pathogens under routine conditions seems unlikely as has generally been the case for transmissible human pathogens. Guidance documents require baseline specimens from surgical teams and others with potential body fluid exposures (e.g., sexual contacts) should be archived, including leukocytes and plasma samples to support future investigations. Active surveillance of these groups should not be required. Standard post-exposure procedures (i.e., occupational health protocols) must be developed for significant unprotected contacts with the xenograft or the blood or bodily fluids of a xenotransplant recipient (25, 236, 237). Given the potential for PERV to bind human cells, post-exposure prophylaxis for (e.g.) needlestick exposures associated with PERV-positive animals may be considered. In the event of an infectious syndrome developing in the xenograft recipient, appropriate infection control precautions should be implemented based on the presenting clinical syndrome (17, 201, 238).

Precautions are needed for the transportation of animals and procurement of organs to mitigate infectious exposures (239). Organ procurement for clinical xenotransplants and decedent studies have generally been performed at biosecure facilities separate from the hospital where tissues are handled as a potential biohazard (240, 241). Standard facility cleaning, disinfection, and sterilization practices used for human material have then been applied. Some programs have used disposable equipment to undergo incineration, while others have not. No transmission events have been reported. Swine are considered to be resistant to prion diseases; thus, standard universal precautions with appropriate personal protective equipment should be observed (14, 15). Some have advocated contact tracing for early experiences.

EDUCATION AND INFORMED CONSENT FOR THE RECIPIENT, CLINICAL CARE TEAMS AND THE PUBLIC

Risk assessment for xenotransplantation is incomplete. The impacts of pig-derived pathogens, immunosuppression, and genetic engineering of swine over remain to be defined as clinical success increases. Standardized educational materials for patients and medical staffs are under development. Patients must understand any requirements for monitoring for infection or neoplasia in addition to those used in allotransplantation (Table 3). While the issues of informed consent are well described elsewhere, experience suggests that some assessment be made of recipients and family health literacy with independent counselling prior to trial enrollment (242, 243). Topics may include potential infectious risks, graft rejection, unprotected sexual contacts, blood donation, potential social stigmatization, the unknown aspects of genetic engineering, and the waiving of rights to withdraw from routine infectious monitoring in the face of unknown public health considerations. The risks and benefits of xenotransplantation and any alternative therapies for the individual as well as for the knowledge gained must be explained.

TABLE 3.

Considerations for informed consent in xenotransplantation

Possible considerations	Examples
Genetic modifications	Possible off-target effects of CRISPR/Cas9 Genetic mosaicism incomplete penetration in xenografts
Immunosuppression	Novel immunosuppression may precipitate new patterns of opportunistic infections Side-effects related to immunosuppressive agents
Infection risk to recipient	Activation of porcine viruses Requirement for validated assays
Infection risk to contacts (close, social, sexual, healthcare)	Few known risks to community Need for post-exposure prophylaxis? Monitoring, testing and any restrictions for recipients and contacts (with or without infection) Lifelong risk of xenozoonses Limitations on sexual contacts and blood donation
Participant selection	Understanding of risks and unknowns of xenotransplantation Retention of eligibility for allotransplantation Freedom from coercion or financial incentives

UNKNOWNS, UNKNOWN UNKNOWNS, AND FUTURE DIRECTIONS

Clinical trials will be essential to assess advances in xenotransplantation. Patient selection will also impact the success of these trials. Each inbred source animal has unique native and genetically modified traits associated with a specific immunosuppressive regimen. Each potential subject carries a set of unique comorbidities common to transplant recipients and epidemiological exposures that may impact outcomes. While the initial experience has been positive, further studies are needed to optimize immunosuppression, microbial assays, metagenomic testing, antimicrobial therapies, and other newer technologies (e.g., multi-omics) to maximize safety and the data from each study. Deployment of immune tolerance strategies may improve transplant outcomes.

The safe development of this technology will address the shortage of organs for transplantation worldwide. Novel techniques for organ perfusion and preservation may allow these advances to address issues of access to clinical care (equity). This opportunity, like all of biomedicine, poses unknown risks that can only be addressed in clinical trials. Xenotransplantation, like all of transplantation, sits at the interface between basic science and clinical application. The opportunity may no longer be remote.

Biographies

Dr. Adam G. Stewart is an Australian-trained Infectious Diseases Physician and Clinical Microbiologist who is the clinical Fellow in the Transplant Infectious Disease and Compromised Host Program at Massachusetts General Hospital. He has completed a Master of Public Health and is awaiting conferral of the Doctor of Philosophy degree for a thesis incorporating work on antimicrobial resistance, clinical trial methodology and antibiotic susceptibility testing. His career is devoted to clinical care of compromised hosts and academic interests including invasive fungal diseases, infectious risk in immunocompromised hosts, and immunotherapy.

Jay A. Fishman, M.D., is Professor of Medicine at Harvard Medical School, Director of the Transplant Infectious Diseases and Compromised Host Program at the Massachusetts General Hospital (MGH), and Associate Director of the MGH Transplant Center. Dr. Fishman received his MD from the Johns Hopkins University School of Medicine, internal medicine training and Infectious Disease Fellowship at MGH, and Fellowships in Molecular Biology and Genetics at MGH and Harvard Medical School. Dr. Fishman established the Transplant and Immunocompromised Host Program of the MGH, the first such training program worldwide. His studies defined the use of ganciclovir for cytomegalovirus infection in transplant recipients, the molecular biology of human Pneumocystis and porcine endogenous retrovirus and the biology of porcine cytomegalovirus (PCMV) in primate models of xenotransplantation. Dr. Fishman is a Fellow of the American College of Physicians, the American Society of Transplantation, and the Infectious Disease Society of America. He is Past-President of the American Society of Transplantation, President-elect of the International Xenotransplantation Association, and Councillor of the Transplantation Society. He has received career achievement awards from AST and TTS.