Rabbit antithymocyte globulin induction and sirolimus monotherapy supports prolonged islet allograft function in a nonhuman primate islet transplantation model

BACKGROUND: We reported that rabbit anti-thymocyte globulin (RATG) induction followed by maintenance immunosuppression with sirolimus supports human kidney allograft survival and asked if this combination would promote islet allograft survival in our primate model. METHODS: Using intra-arterial streptozotocin infusion, we rendered four cynomolgus primates diabetic with undetectable C-peptide levels. Animals were maintained on insulin therapy for at least 1 month, and then islets from mixed lymphocyte reaction mismatched primates were infused into the portal vein. Immediately before the islet allotransplant and for 6 additional days, primates were infused with RATG (20 mg/kg) and given a sirolimus dose to achieve a 24-hr trough level of 8 to 14 ng/mL. RESULTS: The regimen resulted in profound peripheral and lymph node lymphocyte depletion for up to 1 month. Repopulation was gradual thereafter. One primate remained insulin-independent for 169 days and rejected after a sirolimus-dose reduction. Two primates died on day 23 while insulin independent because of wound dehiscence, and a third died on day 30 with high sirolimus levels. Liver sections revealed well-vascularized islets with no signs of inflammation. CONCLUSION: Using a nonhuman primate islet transplant model, RATG plus sirolimus supports islet survival as long as proper sirolimus levels are maintained, but the therapy is limited by sirolimus toxicity. Our findings suggest that RATG is not toxic for islets and thus may be considered in future clinical trails while recognizing that sirolimus monotherapy, with its difficult-to-achieve therapeutic dosing, may not be sufficient to maintain long-term islet allograft function in an autoimmune environment.     

The standardization of pancreatic donors for islet isolations

BACKGROUND: Islet transplantation has proven to be a successful treatment for type 1 diabetes mellitus. The aim of this study was to establish an algorithm by which the combination of the donor quality and pancreas quality was given a numerical score from 0 to 100 for use in determining the quality of a pancreas for islet isolation. METHODS: We retrospectively analyzed 326 pancreases and the outcomes of their respective isolations. Specific donor variables and physical characteristics were identified and weighted according to their influence on the success of the isolation. For each variable, ranges and point weightings were established based on our laboratory experience and literature review. RESULTS: Analysis of the data showed a strong association of the donor point with isolation outcome. Pancreases with lower donor point scores had lower transplant success rates, whereas higher donor point scores in turn produced higher transplant rates. CONCLUSIONS: This scoring system has proven to be useful in guiding the decision process as to whether to accept a particular pancreas for a favorable isolation outcome.     

Interference with tissue factor prolongs intrahepatic islet allograft survival in a nonhuman primate marginal mass model

BACKGROUND: Tissue factor (TF) expression on islets can result in an instant blood-mediated inflammatory reaction (IBMIR) that contributes to early islet loss. We tested whether peritransplant protection of islets from IBMIR with a monoclonal anti-TF antibody (CNTO859) would enhance engraftment in our nonhuman primate marginal mass model. METHODS: Each of six pairs of cynomolgus monkeys (CM) with streptozotocin-induced diabetes was closely matched for metabolic control and was transplanted with 5,000 IEQ/kg allogeneic, ABO-compatible islets from the same donor under the cover of steroid-free immunosuppression. For each pair, experimental animals received islets cultured with 20 microg/mL anti-TF and were dosed with 6 mg/kg anti-TF intravenously, 10-25 min before islet infusion; control monkeys received an equal number of islets from the same preparation cultured without anti-TF and no in vivo treatment. RESULTS: Early fasting C-peptide (CP) values were different between (P<0.01), but not within, pairs and correlated with in vitro functional capacity of islets as assessed by perifusion (r=0.60; P=0.022). Compared to their matched controls, experimental animals had decreased posttransplant markers of coagulation, higher fasting CP levels (1 month posttransplant and end of study) and prolonged graft function. CONCLUSIONS: These data suggest that pretreatment of islets and the recipient with anti-TF may limit the effects of IBMIR, thereby enhancing islet engraftment and survival.     

Total hypophysectomy in the nonhuman primate

A new microsurgical technique for total hypophysectomy in nonhuman primates is described. This procedure is relatively simple and reliable. It can be used as a model in the study of the hypothalamic-pituitary-gonadal axis, particularly in the area of infertility research.     

Extended survival versus accelerated rejection of nonhuman primate islet allografts: Effect of mesenchymal stem cell source and timing

Mesenchymal stem cells (MSC) have been shown to be immunomodulatory, tissue regenerative, and graft promoting; however, several questions remain with regard to ideal MSC source and timing of administration. In this study, we utilized a rigorous preclinical model of allogeneic islet cell transplantation, incorporating reduced immune suppression and near to complete mismatch of major histocompatibility antigens between the diabetic cynomolgus monkey recipient and the islet donor, to evaluate both the graft promoting impact of MSC source, that is, derived from the islet recipient, the islet donor or an unrelated third party as well as the impact of timing. Co-transplant of MSC and islets on post-operative day 0, followed by additional IV MSC infusions in the first posttransplant month, resulted in prolongation of rejection free and overall islet survival and superior metabolic control for animals treated with recipient as compared to donor or third-party MSC. Immunological analyses demonstrated that infusion of MSC from either source did not prevent alloantibody formation to the islet or MSC donor; however, treatment with recipient MSC resulted in significant downregulation of memory T cells, decreased anti-donor T cell proliferation, and a trend toward increased Tregulatory:Tconventional ratios.     

Tolerance in a concordant nonhuman primate model

BACKGROUND: We have previously demonstrated that induction of mixed lymphohematopoietic chimerism resulted in donor specific renal allograft tolerance without the need for chronic immunosuppression in nonhuman primates. Here we have tested whether tolerance can be similarly induced for baboon to cynomolgus renal xenografts. METHODS: After preconditioning with anti-thymocyte globulin (ATG), nonlethal total body irradiation, and thymic irradiation, cynomolgus monkeys underwent splenectomy, native nephrectomies, and baboon marrow and renal transplants. Postoperative cyclosporine was given for 28 days. RESULTS: In Group 1 (n=2, survival= 13, 14 days), both animals developed anti-donor immunoglobulin G, had biopsy findings consistent with humoral rejection, and showed rapidly progressive xenograft failure. In Group 2 (n=5, survival=1, 16, 33, 112, 190 days), 15-deoxyspergualine was added to the regimen (Day 0-13). In one long-term survivor, donor specific hyporesponsiveness was first observed (mixed lymphocyte culture [(MLR]) on Day 48. MLR reactivity returned on Day 64 together with the development of anti-donor antibody and subsequent xenograft failure on Day 112. Donor specific T-cell hyporesponsiveness was detected in the other long-term survivor for the first 133 days, after which a donor-specific skin xenograft was placed, (survival 24 days). Following the skin graft rejection, a rise in the MLR, development of anti-donor antibody and progressive rejection of the renal xenograft were observed. CONCLUSIONS: Antibody-mediated rejection seems to constitute the major difference between concordant xenografts and allografts. Addition of 15-deoxyspergualine for 2 weeks posttransplant extended concordant primate xenograft survival to 6 months without chronic immunosuppression. In contrast to the allogeneic model, renal transplant acceptance in this xenogeneic system was interrupted by placement of a donor-specific skin graft.     

The immunobiology of pig‐to‐nonhuman primate islet xenotransplantation: insights,innovation, and impact

Previous studies of pig‐to‐non‐human primate (NHP) islet xenotransplantation have provided important insights into the immune recognition and effector pathways operative in this relevant preclinical model. The specifics of the xenograft product, microenvironment at the implantation site, and the immunosuppressive regimen significantly influence the mechanisms underlying the rejection of xenogeneic islets. Our current understanding of the immunological barriers to survival of pig islets in NHPs is largely based on studies on intraportal islet xenografts and on comparisons with islet allografts. The demonstration of cell‐mediated rejection of intraportal porcine islet xenografts at about 1 month posttransplant in monkeys immunosuppressed with the same protocols that prevent monkey islet allograft rejection indicates that islet xenograft rejection involves cellular mechanisms that are not present in acute islet allograft rejection. While these mechanisms remain poorly defined the demonstration of long‐term diabetes reversal after intraportal islet xenotransplantation in non‐human primates immunosuppressed with anti‐CD40L but not with anti‐CD40 antibody‐based protocols suggests that the therapeutic efficacy of anti‐CD40L in this transplantation setting likely involves the depletion of donor‐reactive, activated T cells besides CD40:CD40L costimulation blockade. Rejection of intraportal islet xenografts in NHPs immunosuppressed with CTLA4‐Ig and rapamycin was mediated largely by IL‐15‐primed, CXCR3+CD8+ memory T cells recruited by IP‐10 (CXCL10) positive pig islets and macrophages that showed staining for IL‐12 and iNOS. Adding basiliximab induction and tacrolimus maintenance therapy to this protocol prevented rejection in 24 of 26 recipients followed for up to 275 days. Comparison of both groups suggests, though by no means conclusive, that prolongation of graft survival in this large cohort was associated with reduced direct T cell responses to xenoantigens, reduced proportion of intrahepatic (intragraft) B cells and IFN‐γ+ and IL‐17+ CD4 and CD8 T cells, and increased local production of immunoregulatory molecules linked with Tregs, including TGF‐β, Foxp3, HO‐1, and IL‐10. Anti‐pig non‐Gal IgG antibody elicitation was suppressed in both groups. We are currently exploring the concept of negative vaccination to markedly minimize the need for immunosuppression in islet xenotransplantation. Peritransplant administration of donor apoptotic cells extended pig‐to‐mouse islet xenograft survival to >250 days when combined with peritransplant B cell‐depletion and rapamycin. This costimulation blockade‐sparing, antigen‐specific immunotherapy is expected to cause rapid pretransplant clonal deletion of indirect and anergy of direct xenospecific T cells while inducing regulatory T cells. As anti‐CD40L antibodies, B cell depleting antibodies are expected to interfere with indirect antigen presentation, costimulation, and cytokine production required for optimal T cell proliferation, memory formation, and intragraft CD8+ effector function. It is conceivable that additional strategies must be employed in NHPs and eventually in diabetic patients to achieve – as previously with anti‐CD40L antibodies – more complete, yet selective depletion of donor‐reactive, activated T‐cells for the purpose of stable xenograft acceptance.     

Efficacy of the oxygen-charged static two-layer method for short-term pancreas preservation and islet isolation from nonhuman primate and human pancreata

Previous reports indicate that the two-layer method (TLM) of human pancreas preservation is superior to University of Wisconsin solution (UW) when pancreata are preserved for extended periods (i.e., >24 h) prior to islet isolation. In this study, the efficacy of using the TLM for preserving pancreata for short periods (i.e., <13 h) was evaluated using both nonhuman primate and human pancreata preserved with a TLM kit precharged with oxygen. An oxygen precharged TLM (static TLM) was established and compared with the original TLM with continuous oxygen supply. For the static TLM, the perfluorochemical was fully oxygenated and the oxygen supply removed prior to pancreas preservation. In the primate model, pancreata were preserved by the static TLM, the original TLM, and UW for 5 h prior to islet isolation. In the human model, pancreata were preserved with the static TLM or the original TLM or UW for 4-13 h. Both primate and human pancreata were processed by intraductal collagenase injection and digestion followed by continuous density gradient purification to isolate islets. Islets were assessed for islet yield, purity, viability, and in vitro functionality. In the primate model, islet yield, viability, and in vitro functionality were significantly improved by both the static TLM and the original TLM with similar results. Postculture islet yields were 23,877 +/- 3619 IE/g in the static TLM, 21,895 +/- 3742 IE/g in the original TLM, and 6773 +/- 735 IE/g in UW. In the human model, both the static TLM and the original TLM significantly increased islet yield compared with UW with postculture islet yields of 2659 +/- 549 IE/g in the static TLM, 2244 +/- 557 IE/g in the original TLM, and 1293 +/- 451 IE/g in UW. Nonhuman primate and human pancreata stored in the static TLM, immediately upon procurement, yield isolated islets of a substantially higher quantity than when pancreata are stored in UW. Thus, the use of the static TLM should replace the use of UW for storage of pancreata during transport prior to islet isolation.     

Multi-lot analysis of custom collagenase enzyme blend in human islet isolations

INTRODUCTION: Variability currently in Liberase HI from lot to lot limits the ability to effectively isolate islets with consistency. Roche Diagnostics Inc (Indianapolis, Ind, USA) has developed a Custom Collagenase enzyme blend in hopes that producing collagenase II and I and thermolysin separately will eliminate variability. In this study we examined the variability in Custom Collagenase lots in respect to isolation results and isolation success rates and compared those to Liberase HI. METHODS: We retrospectively analyzed records from 68 islet isolations where either Liberase HI (lot A: n = 23, Lot B: n = 20) or Custom Collagenase blend (Lot C: n = 10, Lot D: n = 15) was employed. Human islets were isolated from cadaveric pancreata using standardized methods performed in a controlled islet isolation facility. RESULTS: Analysis of Liberase HI and Custom Collagenase using Student t test showed no difference between the two groups. Comparison of the two Custom Collagenase lots using the t test showed a statistical difference between undigested pancreas weight and pancreas digestion times. Using chi-square test, no statistical significance was found in isolation success rates from lot to lot. CONCLUSION: Although the Custom Collagenase blend is comparable to Liberase HI in its ability to isolate human islets, variability still exists from lot to lot when used conventionally as Liberase HI is. The ability to predetermine doses is beneficial, and as techniques to manipulate the activity levels prior to isolations improve so to will the enzymes' ability to isolate islets on a consistent basis.     

Pancreatic islet transplantation using the nonhuman primate (rhesus) model predicts that the portal vein is superior to the celiac artery as the islet infusion site

We've established a nonhuman primate islet allotransplant model to address questions such as whether transplanting islets into the gut's arterial system would more safely and as effectively support long-term islet allograft survival compared with the traditional portal vein approach. We reasoned that islets make up <2% of pancreatic cell mass but consume an estimated 20% of arterial blood flow, suggesting an advantage for the arterial site. Access to the arterial system is also easier and safer than the portal system. Pancreatectomized rhesus macaques were transplanted with allogeneic islets infused into either the portal vein (n = 6) or the celiac artery (n = 4). To prevent rejection, primates were given daclizumab, tacrolimus, and rapamycin. In five of six portal vein experiments, animals achieved normoglycemia without exogenous insulin. In contrast, none of the animals given intra-arterial islets showed even transient insulin independence (P = 0.048). Two of the latter animals received a second islet transplant, this time to the portal system, and both achieved insulin independence. Thus, intraportal islet transplantation under conventional immunosuppression is feasible in primates and can result in long-term insulin independence when adequate immunosuppression is maintained. Arterial islet injection, however, does not appear to be a viable islet transplantation technique.     

Pig islet xenotransplantation into non-human primate model

Allogeneic islet transplantation faces difficulties because (1) organ shortage is recurrent; (2) several pancreas donors are often needed to treat one diabetic recipient; and (3) the intrahepatic site of islet implantation may not be the most appropriate site. Another source of insulin-producing cells, therefore, would be of major interest, and pigs represent a possible and serious source for obtaining such cells. Pig islet grafts may seem difficult because of the species barrier, but recent reports demonstrate that pig islets may function in primates for at least 6 months. Pig islet xenotransplantation, however, must still overcome several hurdles before becoming clinically applicable. The actual consensus is to produce more preclinical data in the pig-to-primate model as a necessary requirement to envisage any pig-to-human transplantation of islets; therefore, a summary of the actual acquired knowledge of pig islet transplantation in primates seemed useful and is summarized in this overview.     

Experimental immune complex-mediated glomerulonephritis in the nonhuman primate.

This study was undertaken to develop a model of immune complex (IC)-mediated glomerulonephritis (GN) in the nonhuman primate that could be used in subsequent studies to examine critically the role of the erythrocyte complement receptor (E-CR) in the pathogenesis of IC-mediated disease. Cynomolgus monkeys were chosen for study because they constitutively express E-CR levels that are either less than, equal to, or greater than that seen in normal man. After immunization with bovine gamma globulin (BGG), the GN induction protocol was begun in 10 cynomolgus by initiating daily i.v. administration of BGG in amounts sufficient to achieve or exceed antigen/antibody equivalence (assessed by the quantitative precipitin assay) for precipitating antibody present in the plasma volume. We found that within eight weeks of daily BGG administration of all the cynomolgus developed IC-mediated GN, irrespective of the initial E-CR level of the animals. However, the high E-CR cynomolgus tended to receive the higher BGG doses because of higher initial antibody levels to BGG. When the total number of glomerular deposits (determined by morphometric studies) per total BGG dose for each animal was plotted against the initial CR/E of that animal, there was a tendency for the animals with higher CR/E levels to have a lower number of glomerular deposits/BGG dose (r = 0.62, P = 0.06). Also, the total number of glomerular deposits correlated with the severity of the GN. During the early weeks of the GN induction protocol, the IC that formed in vivo (assessed by infusion of 125I-BGG) bound in large amounts to the circulating erythrocytes of the cynomolgus with medium or high E-CR levels. However, when tested after the onset of heavy proteinuria, which occurred between weeks 5 and 8 of daily BGG administration, the IC that formed in the circulation bound only poorly to circulating erythrocytes. By this time the E-CR levels had declined to 43 +/- 9% of initial values (P less than 0.01). This study demonstrates that: 1) A workable model of IC-mediated GN has been developed in the nonhuman primate. 2) During the induction of GN, CR/E and the ability of the erythrocyte to bind IC in vivo are decreased significantly. This suggests that an intact E-CR system could play a role in the protection against IC-mediated disease. However, further study will be needed to test that hypothesis critically. The present model should be useful in such studies.     

Parameters favouring successful adult pig islet isolations for xenotransplantation in pig-to-primate models

Dufrane D, D'hoore W, Goebbels R‐M, Saliez A, Guiot Y, Gianello P. Parameters favouring successful adult pig islet isolations for xenotransplantation in pig‐to‐primate models.
Xenotransplantation 2006; 13: 204–214. © Blackwell Munksgaard, 2006 Abstract: Background: In the near future, adult porcine islets of Langerhans appear as an unlimited source of insulin‐producing cells which could play a major role for treating diabetes mellitus. There is, however, an obvious lack of pre‐clinical results and data in the pig‐to‐primate model. One of the main hurdles of this model is certainly related to the difficulty of reproducing regularly successful porcine islet isolation. This experimental work was designed to provide guidelines applicable in pig pancreas procurement and islet isolation for successful islet xenotransplantation into primates. Methods: Pancreases were harvested from adult Belgium Landrace pigs (n = 79) in a single centre. The impact on islet yield of (1) pancreas procurement (blood exsanguination and warm ischaemia time (WIT)), (2) cold storage solutions (classic UW and modified UW (without hydroxyethyl starch and inverse K⁺/Na⁺ concentration)), (3) a dynamic or static method of pancreas digestion, and (4) the endotoxin content and enzymatic activity from five different batches of Liberase PI was studied. In addition, pancreatic biopsies (n = 18), performed before isolation, were retrospectively analyzed to study the impact of histomorphometry on porcine islet yield. Finally, two diabetic cynomolgus monkeys were transplanted without immunosuppression with 15 000 pig islet equivalents/kg body weight of recipient to assess in vivo the function of freshly isolated islets. Univariate and multivariate analyses were performed. Results: By multiple linear regression, the most significant variables that significantly improved islet yield were, firstly, the presence of <30 EU (endotoxin units) of endotoxin in Liberase batches, followed by a WIT under 10 min and the use of blood exsanguination before pancreas harvesting (P<0.005). In contrast, isolation method (dynamic vs. static) and the solution used for storage (short‐term) (UW vs. modified UW) did not significantly influence islet yield. The correlation of retrospective histomorphometry analysis of native pancreas and extemporaneous biopsy before isolation clearly determined a positive relationship between isolated islet number and the number of islets/cm² (r = 0.708, P<0.01) or with the percentage of large islets (r = 0.680, P<0.01) found in pancreas biopsies. Pig pancreases containing more than 82 islets/cm² and more than 42% of large islets (>100 μm) thus enabled more than 120 000 islet equivalents to be obtained in 90% of the cases, which is an ideal amount of islets to transplant into a primate of 4 to 5 kg. In vivo, a reduction of blood glucose (<200 mg/dl), associated with porcine C‐peptide production, was observed in two primates after transplantation with adult pig islets. At day 7 post‐transplantation, however, loss of islet function was associated with graft destruction and immune reaction. Conclusions: Morphological screening of the pig pancreas before isolation, optimal blood exsanguination, WIT <10 min, and an endotoxin content <30 EU/mg in Liberase PI batches determine successful pig islet isolation for xenotransplantation in primates.     

An evaluation of the activation of endogenous pancreatic enzymes during human islet isolations

Despite advances in human islet isolations, there remain inconsistencies in human islet yield and viability after collagenase digestion. It has been suggested that trypsin may contribute to the proteolysis of collagenase and the destruction of islet cells, or possibly exert indirect effects on the pancreas by activating other endogenous serine proenzymes. This study evaluated the effects of serine proteases on collagenase activity and profiled the kinetics of serine protease activity throughout human islet isolations with and without addition of Pefabloc, a serine protease inhibitor. Cadaveric pancreases were perfused in the presence (n = 12) and absence of Pefabloc (0.4 mmole; n = 8). Samples were collected before and throughout the digestion process and were assayed for trypsin, chymotrypsin, and elastase activity. A study of the enzyme kinetics of serine proteases throughout human islet isolations showed an increase in activity levels throughout the digestion period. There was a significant difference in the chymotrypsin (1342 +/- 503 and 384 +/- 71 units) and elastase (7.94 +/- 1.1 and 2.761 +/- 0.69 units) levels between the control and Pefabloc-supplemented isolations, respectively. There was no significance difference noted among the trypsin (88 +/- 27 and 54 +/- 18 units) levels between the control and Pefabloc-supplemented isolations, respectively. This demonstrates that serine proteases are effectively inhibited by Pefabloc during the islet isolation process. These data show that the presence of serine proteases may likely damage the islets upon prolonged digestion of the pancreatic tissue.     

Demonstration of multilineage chimerism in a nonhuman primate concordant xenograft model

Abstract: Prior studies from our laboratory have demonstrated that a nonmyeloablative conditioning regimen can induce transient mixed chimerism and renal allograft tolerance between MHC disparate cynomolgus monkeys. We have also shown that this preparative regimen can be extended to a concordant baboon to cynomolgus xenograft model by adding, to the post transplant protocol, therapy designed to prevent antibody production. Here we examine the use of brequinar (BQR) for this purpose and the efficacy of two new reagents developed to demonstrate the establishment of chimerism in the xenograft recipients. The cynomolgus recipients were conditioned with WBI (300 cGy), TI (700 cGy), ATG, cyclosporine, and brequinar sodium. To detect engraftment of the donor marrow, we prepared a polyclonal cynomolgus anti-baboon antibody (CABA) and a monoclonal antibody (215.1), which distinguish baboon and cynomolgus lymphocytes and granulocytes. We employed flow cytometry analysis to detect multilineage chimerism in the xenograft recipients. Five of the six recipients monitored using our new reagents (CABA and 215.1) developed detectable chimerism and only one of these animals lost its kidney to rejection. However, other complications have not permitted assessment of long-term outcome. The features of the multilineage chimerism included the detection of donor granulocytes (1.8–77.4%) and lymphocytes (2.4–22.2%) for 9 to 37 days. Our new reagents permit the detection of multilineage mixed chimerism, which may be a predictor of xenograft tolerance. We also conclude that brequinar may be effective in preventing antibody formation, but because of its toxicity, it is probably not the drug of choice for extension of the mixed chimerism protocol to concordant xenografts.