Current status of porcine islet xenotransplantation

Abstract

Purpose of review

Human islet transplantation has proven to be a highly effective treatment for patients with labile type 1 diabetes mellitus, which can free patients from daily glucose monitoring and insulin injections. However, the shortage of islet donors limits its’ broad application. Porcine islet xenotransplantation presents a solution to the donor shortage and recent advances in genetic modification and immunosuppressive regimens provide renewed enthusiasm for the potential of this treatment.

Recent findings

Advances in genetic editing technology are leading to multigene modified porcine islet donors with alterations in expression of known xenoantigens, modifications of their complement and coagulation systems, and modifications to gain improved immunological compatibility. Recent NHP-based trials of costimulation blockade using CD154 blockade show promising improvements in islet survival, whereas results targeting CD40 are less consistent. Furthermore, trials using IL-6 receptor antagonism have yet to demonstrate improvement in glucose control and suffer from poor graft revascularization.

Summary

This review will detail the current status of islet xenotransplantation as a potential treatment for type I diabetes mellitus, focusing on recent advances in porcine xenogeneic islet production, assessment in nonhuman primate preclinical models, the outcome of human clinical trials and review barriers to translation of xenoislets to the clinic.

INTRODUCTION

Human pancreatic islet transplantation is an effective treatment for patients with type 1 diabetes mellitus (T1DM) with hypoglycemia unawareness [1]. In 2000, a leap forward in the quest for an effective cell-based β-cell replacement occurred with the development of the Edmonton islet transplantation protocol. Since then, further efforts by international groups have optimized the procedural techniques and immunosuppressive regimens leading to long-term insulin independence for many patients [2-4]. Despite these encouraging results, human islet transplantation is limited by the scarcity of suitable pancreata from deceased donors [5]. The current organ donor supply will never meet the clinical demand, thus alternative islet sources of unlimited supply are being explored, such as bioartificial devices, gene therapy, stem cell-derived islet transplantation and xenotransplantation.

Many aspects of islet xenotransplantation have been recently reviewed as the field reaches toward broad clinical relevance [6-10]. Although most islet transplants occur via infusion into the portal system, Smood et al. [11^■] evaluated the optimal site for transplantation, questioning if the well-accepted intraportal route is truly superior to other sites, such as the renal subcapsular space. Kemter et al. [12^■] recently reviewed work related to optimal porcine islet isolation and culture and alternate strategies to improve maturation and engraftment. Kemter [13^■] also summarized recent advances in the genetic engineering of porcine islet donors and the important distinctions between source of islets, mode of delivery and site of transplantation. In this review, we will outline the rejection processes important in islet xenotransplantation, review optimal targets for porcine genetic modification, recent preclinical nonhuman primate models, human clinical trials, and the persistent challenges of clinical utilization of porcine islet xenotransplantation in the treatment of T1DM.

MAJOR BARRIERS

Instant blood mediated inflammatory reaction

Both human and porcine islet transplantation must overcome the immediate inflammatory response called instant blood-mediated inflammatory reaction (IBMIR) upon intraportal transplantation [14]. Although the cause of IBMIR is not perfectly understood, one theory suggests that in the normal state, only bone marrow-derived blood cells and endothelium are exposed to blood components. Exposure to other cells and to the extracellular matrix is non-physiologic, triggering inflammation, coagulation, and complement. Tissue factor, which is constitutively expressed on many cell types, including vessel adventitia, renal glomeruli and islets, and inducible by the inflammatory response on vessel endothelium, platelets, and neutrophils, has been demonstrated to be integral in the initiation of IBMIR [15,16]. This reaction is not unique to xenotransplantation and can be recapitulated ex-vivo in studies that have been informative about the underlying mechanisms that include activation of immune mediators, the complement and coagulation systems and platelet aggregation [17-19]. IBMIR leads to immediate loss of a significant portion of transplanted islets [20]. This is partially overcome by increasing the number of islets transplanted to sustain an adequate islet mass. IBMIR is particularly pronounced in islet xenotransplantation, in which immediate platelet consumption, activation of complement and a rapid dysregulated release of insulin representing islet injury is seen [21,22]. Efforts to overcome IBMIR include encapsulation strategies, genetic modifications of porcine donors and improved immunosuppressive regimens.

Chronic rejection

Sustaining long-term function of islets in both preclinical pig to non-human primate (NHP) trials and in human alloislet transplantation has proven challenging because of immune responses as well as detrimental effects of necessary immunosuppressive medications. In numerous studies, histologic evaluation of xenoislets has shown T-cell infiltration representing residual rejection not controlled by current immunosuppression regimens [23-26]. In addition to T-cell-mediated rejection, there are additional, poorly understood, nonimmune mechanisms resulting in the late loss of insulin independence seen even in autologous islet transplant recipients, in which alloimmunity and immunosuppression confounders are avoided [27].

PORCINE GENETIC MODIFICATIONS

Genetic modifications to target rejection

The difference in the expression of cell surface carbohydrate epitopes between pigs and humans and the role of human preformed antibodies in hyperacute rejection (HAR) of xenografts allowed for an actionable target for porcine genetic modifications [28,29]. The major carbohydrate antigen targets of human preformed antiporcine antibodies include galactosyl-α1,3-galactose, synthesized by α-1,3-galactosyltransferase, N-acetylneuraminic acid, synthesized by cytidine monophosphate-N-acetylneuraminic acid hydroxylase, and an Sd(a)-like glycan made by β-1,4-N-acetyl-galactosaminyl transferase 2 [13^■]. The development of pigs deficient in these carbohydrates occurred initially by single genetic modification, cloning, and natural breeding, but with advancements in gene-editing techniques, such as clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9, modifications can now occur through rapid, multiple molecular techniques [10,30]. These initial modifications allowed researchers to largely overcome the first hurdle evident within xenotransplantation of HAR; however, further barriers persisted, despite reducing antibody binding to porcine endothelial cells by more than 90%. In addition to residual antibodies to as yet undefined targets, activation of complement occurs during both HAR and IBMIR and there are species incompatibilities in complement components and natural complement pathway regulatory proteins (CPRPs). Thus, further efforts in porcine genetic modification focused on complement regulators [17,31]. Xenoislets from pigs with the human complement regulators, CD46, CD55, and CD59, have been evaluated in preclinical models and manifest less early islet loss because of IBMIR but inconsistent improvement in long-term islet survival [32-35]. This finding is compatible with the results of islet transplantation using multitransgenic pigs [36].

Genetic modifications to target infection transmission

The theoretical risk of transmission of porcine infectious diseases to humans was a central concern that tempered initial interest in xenotransplantation in the 1990s and remains a concern today. Dedicated efforts to eliminate clinical and subclinical infection of porcine donors with various porcine viruses include delivery via cesarean section and maintenance in designated pathogen-free facilities [37^■■]. These steps can effectively reduce transmission of a multitude of viruses; however, porcine endogenous retroviruses (PERVs) are not eliminated as these viruses are integrated into the porcine genome in varying copy number in all strains tested to date. Whether PERVs have a physiologic role in the pig host remains uncertain [38]. Recently, PERV inactivation was achieved through CRISPR/Cas9 mediated editing of multiple genomic sites [39]. Furthermore, it is evident that different porcine tissues and organs harbor variable levels of infectious agents. Crossan et al. [37^■■] evaluated neonatal and adult porcine islets in culture and found that even when PBMC tested positive for specific viruses, the islet cells remained free of the pathogen, implying that islet cells may pose a lower infectious risk than other organs.

Genetic modifications of porcine insulin

Both bovine and porcine insulin was historically utilized in the treatment of diabetes prior to the transition to synthetic human insulin [40]. Porcine insulin has a similar structure to human insulin, with only a single amino acid difference (alanine in pigs and threonine in humans at position B30) [41]. Given that porcine insulin is more immunogenic and less effective, efforts are underway to develop genetically modified pigs that express human insulin [42^■■,43]. Yang et al. [43] successfully generated pigs that expressed human insulin by utilizing somatic cell nuclear transfer (SCNT) technology with transcription activator-like effector nucleases or CRISPR/Cas9, combined with single-stranded oligonucleotides to modify the pig insulin gene. These pigs expressed human insulin, but porcine C-peptide, which have different amino acid sequences and lengths [44]. More recently, Cho et al. [42^■■] used CRISPR/Cas9 to generate insulin-deficient pigs and subsequent SCNT to develop piglets that expressed both human C-peptide and insulin. Of note, the insulin expression levels were low and the piglets died after three days of life, thus further optimization of this strategy is necessary.

PRECLINICAL NON-HUMAN PRIMATE STUDIES

Prior work

Over the past twenty years, many groups have investigated pig islet cell xenotransplantation in NHP, recently reviewed by Kemter and Liu [8,13^■]. Table 1 summarizes free, intraportal islet xenotransplantation trials since 2006, with recent studies and novel approaches reviewed below.

Table 1.

Porcine-free, intraportal islet xenotransplantation in nonhuman primates since 2006

Donor pig	Recipient type	Average number of islets (IEQ/kg)	Immunosuppression	Maximum graft survival (days)	References
SNU miniature pigs Adult	STZ-induced diabetic rhesus monkeys (n = 9)	96 090	ATG, sirolimus, tacrolimus, anti-CD40 mAb, belatacept, CVF, adalimumab	>320	[45■■]
SNU miniature pigs Adult	STZ-induced diabetic rhesus monkeys (n = 5)	93 575	ATG, sirolimus, tacrolimus, anti-CD40 mAb, tocilizumab, CVF, adalimumab	176	[50■■]
SNU miniature pigs Adult	STZ-induced diabetic rhesus monkeys (n = 5)	100 000	ATG, sirolimus, tacrolimus, anti-CD154 mAb, CVF, adalimumab	603	[23]
GTKO/hCD46/hCD39 or GTKO/CD46/TFPI/CTLA4-Ig or GTKO/CD46/TFPI/CTL4-Ig/CD39 Adult	STZ induced diabetic cynomolgus monkeys (n = 5)	85 000	ATG, MMF, anti-CD154 mAb	365	[36]
Large white × Landrace GTKO/hCD55/hCD59/HT Neonatal	Nondiabetic baboons WT n = 4 Genetically modified n = 5	7644 (WT) 17889 (genetically modified)	WT: none Genetically modified: ATG, MMF, tacrolimus	30	[34]
hCD46 genetically modified adult	STZ-induced diabetic cynomolgus monkeys (n = 9)	85–100000	ATG, MMF, anti-CD154 mAb	396	[35]
Outbred wild type or GTKO adult	STZ induced diabetic cynomolgus monkeys (n = 10)	40 000	ATG, tacrolimus, rapamycin, anti-CD154 mAb, MMF	>58	[71]
Duroc or Large White Crossbreeds Neonatal	Diabetic rhesus macaques s/p pancreatectomy (n = 9)	50 000	Anti-CD 154 mAb, basiliximab, sirolimus, belatacept	>260	[25]
Outbred swine and inbred miniature swine Adult	STZ-induced diabetic cynomolgus monkeys (n = 12)	25 000	Basilixumab, FTY720 or tacrolimus, everolimus, anti-CD154 mAb, leflunomide	>187	[46]

ATG, Antithymocyte globulin; CVF, cobra venom factor; IEQ, insulin equivalents; mAb, monoclonal antibody; MMF, mycophenolate mofetil; SNU, Seoul National University; STZ, streptozotocin; WT, wild type.

Costimulation blockade

Costimulation blockade through anti-CD40 and anti-CD154 monoclonal antibodies (mAb) is thought to have great potential as an immunosuppressive strategy by interfering with a key signaling pathway required for optimal activation of T cells, B cells, macrophages, and dendritic cells. This immunosuppressive approach has led to recent progress in porcine islet xenotransplantation, as well as xenotransplantation in general. Shin et al. [45^■■] completed nine xeno islet transplantations from Seoul National University (SNU) miniature pigs to Streptozotocin (STZ)-induced diabetic rhesus monkeys via the portal system. The immunosuppressive regimen was composed of anti-CD40 mAb in conjunction with belatacept or tacrolimus. This regimen achieved one long-term survivor (>6 months); however, the results were inconsistent, apparently because of a lack of control of both IBMIR and cell-mediated rejection with this immunosuppressive regimen. This contrasts with their experience with anti-CD154 mAb-based immunosuppression in which five STZ-induced diabetic rhesus monkeys underwent islet cell transplantation from SNU miniature pigs [23]. In this study, graft survival with adequate glycemic control was achieved for greater than six months in four of the five monkeys. This is consistent with prior studies using anti-CD154 mAb-based immunosuppressive regimens; however, clinical advancement of CD154 targeting antibodies was halted because of thromboembolic complications [25,46-48]. More recently, a number of companies have developed modified antibodies against CD154 that are designed to eliminate thromboembolic triggering. The Pittsburgh group reviewed their experience with 14 STZ-induced diabetic cynomolgus monkeys, all of which were maintained on an anti-CD154 mAb-based immunosuppressive based regimen and experienced no thromboembolic complications, using only Aspirin 81 mg per day and a low-dose continuous heparin infusion to maintain intravenous catheter patency [49].

IL-6 receptor antagonist

In addition to optimizing immunosuppression, Min et al. [50^■■] evaluated the effects of addition of tocilizumab, an IL-6 receptor antagonist. Theoretically, IL-6 has both positive and negative effects with regards to islet xenotransplantation. It is an important mediator in angiogenesis leading to revascularization of the islets but it also plays a role in the deleterious inflammatory response and coagulation dysregulation. Five STZ-induced diabetic rhesus monkeys underwent portal vein islet transplantation with islets derived from SNU miniature pigs. The immunosuppressive regimen included anti-CD40 mAb, sirolimus, tacrolimus, thymoglobulin, cobra venom factor, and adalimumab (TNF-α neutralizing mAb). The monkeys treated with tocilizumab were noted to have decreased levels of CRP production, representing less initial inflammation, but poorer revascularization of the islets. Graft survival was variable, with the longest surviving 176 days with 134 days independent of insulin.

TRANSLATION TO HUMANS

Human islet xenotransplantation

Trials of porcine islet xenotransplantation in humans have varied in their approach and success (Table 2). The first human islet xenotransplantation was conducted by Groth et al. [51] in 1994. Ten patients who had received a kidney allotransplant, on standard immunosuppression, received fetal pig islet cell clusters either under the kidney capsule (n = 2) or via the portal system (n = 8). Each transplant required pancreata from 39 to 100 fetuses, with the average yield per fetus of 10000 islet cell clusters. The patients tolerated the procedures well, but the porcine islets did not improve their glycemic control.

Table 2.

Human clinical trials utilizing porcine islets

Islet source	Recipient	Islet type	Average number of islets	Site	Immunosuppression	References
Auckland Island	T1DM (n = 8)	Neonatal Encapsulated	5000 and 10 000 IEQ/kg	Peritoneum	None	[58]
Auckland Island	T1DM (n = 14)	Neonatal Encapsulated	5000 or 10 000 or 15 000 or 20 000 IEQ/kg	Peritoneum	None	[57]
Xeno-1	T1DM (n = 22)	Neonatal Free	55 000 IEQ/kg	Intraportal	Cyclosporine, prednisolone, MMF, tacrolimus, sirolimus, OKT-3	[56]
New Zealand bred	T1DM adolescents (n = 23)	Neonatal with Sertoli cells	13 000–20 000 IEQ/kg (first transplant)	Subcutaneous	None	[54,55]
Cross-White Breed	T1DM (n = 1) with kidney transplant (n = 1)	Neonatal Encapsulated	15 000 IEQ/kg	Peritoneum	None (n = 1) and cyclosporine, prednisone, azathioprine	[52,53]
Swedish Landrace	T1DM with kidney transplant (n = 10)	Fetal Free	200 000–1 million islets	Kidney capsule or intraportal	ATG, 15-deoxyspergualin, cyclosporine, prednisolone, azathioprine	[51]

ATG, Antithymocyte globulin; IEQ, insulin equivalents; MMF, mycophenolate mofetil; T1DM, type 1 diabetes mellitus.

Subsequently, Elliott et al. [52] completed two intraperitoneal transplants of genetically unmodified neonatal alginate-encapsulated islets in one patient without immunosuppression and one who was on standard immunosuppression because of a prior kidney transplant. Importantly, neither patient showed evidence of PERV transmission and both had short-term improvement in their insulin requirements and glycosylated hemoglobin (HbA1c). The nonimmunosuppressed patient was followed for over nine years and underwent laparoscopic islet harvest, revealing intact nodules with encapsulated islets that were similar in size to initial implantation, but opaque and rigid. When assessed for functionality, the islets were capable of low-level insulin production [53]. Whether the modest improvement in A1c was attributable to the graft or to the improved diabetes care and attention associated with the transplant is unclear.

Valdes-Gonzalez et al. [54] instituted a novel approach, creating a vascularized collagen environment prior to islet xenotransplantation by implanting a subcutaneous collagen-generating steel rod. Twelve adolescent diabetic patients underwent transplant with an average of 17 798 insulin equivalents (IEQ)/kg body weight combined with Sertoli cells, derived from New Zealand 7–10-day-old piglets. Eleven of the patients underwent retransplant 6–9 months later and four underwent a third transplant. Immunohistochemistry on three explanted devices showed the presence of insulin-producing cells and glucagon positive cells, with CD3 T-cell infiltrate. At four years, half of the patients achieved reduction in their insulin needs, and at 5–7 years of follow-up of a similar cohort, the team reported sustained improvement in HbA1c with less chronic complications of diabetes, including neuropathy, retinopathy, and nephropathy [55].

Wang et al. [56] reported 21 patients who underwent islet xenotransplantation utilizing neonatal islets from the xeno-1 pig bred in the Animal House of Central South University of Changsha, China. The islets, 55 000 IEQ/kg body weight, were implanted in the liver via the hepatic artery and patients were immunosuppressed. Patients saw a reduction in their insulin requirements and improvement in HbA1c in the first three months. Although none achieved long-term control, one patient was able to discontinue insulin therapy for seven days. No significant complications developed and there was no evidence of PERV transmission, with up to six years posttransplant follow-up.

In 2011, the group in New Zealand completed a Phase 1/2a xenotransplantation trial of encapsulated neonatal porcine islets transplanted to the peritoneum of 14 nonimmunosuppressed diabetic patients [57]. Patients received 5000, 10 000, 15 000, or 20 000 IEQ/kg; however, there were no apparent outcome differences between the groups. Patients experienced a reduction in unaware hypoglycemic events; however, minimal changes to HbA1c or daily insulin doses were evident at one year. This work led to a trial conducted in Argentina in which eight patients underwent intraperitoneal transplant of either 5000 or 10 000 IEQ/kg of encapsulated neonatal porcine islets [58]. Again, there was no evidence of PERV transmission [59]. Patients experienced fewer episodes of unaware hypoglycemia and improved HbA1c; however, marginal change to their daily insulin requirement.

Persistent challenges

Prior to moving forward with further human clinical trials of unencapsulated porcine islets, it will be necessary to obtain sustained (>6 months), consistent improvement in diabetes management, with either insulin independence or a greatly reduced insulin requirement, in a porcine to NHP model using clinically tolerable immunosuppression [60,61]. Current success in NHP preclinical models has been achieved using immunosuppressive regimens that are not clinically applicable, predominately based upon costimulation and complement blockade. As the ability to gain more extensive genetically modified donor pigs increases, there is hope that the addition of human CPRPs and coagulation regulatory elements will help overcome this barrier. Additionally, the ability to cure clinical diabetes and achieve insulin independence in NHPs under current xenotransplantation techniques has been inconsistent. It is clear that a method to mitigate IBMIR and immediate xenograft damage in a clinically relevant manner is necessary prior to clinical translation, as is effective therapy to combat cellular immunity.

In humans, the number of transplanted islets has been demonstrated to be the greatest predictor of insulin independence after islet transplantation [62]. In the recent Clinical Islet Transplantation Consortium trial, an average transplanted islet mass of more than 16 000 IEQ/kg had been delivered in those achieving insulin independence [63]. There are clear species differences between pigs, NHPs, and humans regarding metabolic demands, islet insulin release, and glycemic control [64]. NHPs have been shown to have a higher daily insulin requirement/kg than humans, related to differences in dietary energy requirements and eating habits, and in most preclinical islet xenotransplantation studies, a high islet mass is utilized, typically 50 000–100 000 IEQ/kg [64]. For allotransplantation in NHP, a mass of 25 000 IEQ/kg is generally curative. Translating this number to a human dose/kg of 50 000 IEQ per 70 kg person would require 10 adult pigs and greater than ninety juvenile pigs per recipient [65]. Although it is anticipated that a smaller islet mass will be required for human islet xenotransplantation, it is still a concern as to the practicality and ability to obtain sufficient numbers of islets for transplant.

These concerns are compounded by other emerging technologies to treat diabetes. In particular, recent advances allow stem cell differentiation into islet-like clusters (ILCs) that are glucose responsive and approach near-normal patterns of insulin secretion [66]. The composition of these ILCs differs from that of normal islets, lacking antigen-presenting cells and endothelium, and in some cases, expansion of other endocrine cell types (such as enterochromaffin cells) [67^■,68^■]. The impact of these differences remains to be determined but in small and large animal studies, glycemic control can be excellent. Despite these issues, a clear advantage of ILCs is that they can be generated in massive numbers and could be further genetically modified to reduce immunogenicity [69^■■,70^■].

CONCLUSION

Recent advances in porcine genetic modification and immunosuppressive regimens in preclinical pig-to-NHP islet xenotransplantation studies have led to new insights into the pathogenesis of islet cell rejection and progress toward achieving successful islet xenotransplantation. There are persistent barriers to clinical human islet xenotransplantation and further efforts should remain focused on establishment of an optimal genetically modified porcine islet source and establishment of long-term functioning islets with clinically relevant immunosuppression.

KEY POINTS.

• Instant blood-mediated inflammatory reaction and chronic T-cell-mediated rejection persist in both human islet transplantation and porcine islet xenotransplantation.

• Genetic modification of porcine islet donors target coagulation and complement regulatory elements, reduction of infectivity, and optimization of insulin type and release.

• Recent preclinical pig-to-NHP trials demonstrate improved outcomes with anti-CD154 mAb-based immunosuppression when compared with anti-CD40 mAb-based treatment.

• Inconsistent outcomes in NHP preclinical trials, the large number of xenoislets needed for insulin independence and progress made in alternative stem cell-derived β-cell replacement may impede progress toward clinical relevance of porcine xenotransplantation