Influence of Cold Ischemia Time, Pretransplant Anti-Porcine Antibodies, and Donor/Recipient Size Matching on Hyperacute Graft Rejection After Discordant Porcine to Cynomolgus Kidney Transplantation

Organs transplanted between phylogenetically disparate species, such as from the pig into the primate, are subject to hyperacute rejection (HAR). This form of xenograft rejection is mediated by preformed natural antibodies and is believed to occur invariably in discordant xenografts thus leading to rapid destruction and complete thrombosis of the graft. Recent data, however, have shown that in the porcine to cynomolgus monkey setting, HAR is not inevitably seen after porcine kidney transplantation. The influence of preoperative antiporcine antibody levels in the recipient, cold ischemia time, and donor organ weight on the onset of HAR was investigated by using unmodified large white pigs (aged 3–12 weeks) as organ donors and adult cynomolgus monkeys (aged 1.5–3.5 years) as recipients. Porcine kidney xenotransplantation was performed in either a non-life-supporting model (n=7) or in a life-supporting model (n=8). In both models, no correlation was found between cold ischemia time and HAR. When preoperative anti-porcine antibody levels were investigated, a significant increase in incidence of HAR was observed in animals with elevated anti-porcine IgM (P<0.05) but not IgG levels (P=NS). Interestingly, although 5 of 12 grafts with an organ weight of less than 50 g underwent HAR, none of three grafts with a donor organ weight of more than 70 g showed signs of HAR. In addition, all three larger grafts showed intraoperative and postoperative urine production, although only in 1 (48 g) of the 12 grafts weighing less than 50 g primary graft function was observed. In one animal, a second porcine kidney (23 g) was successfully transplanted (without HAR) immediately after HAR and subsequent removal of a first porcine kidney (20 g). These results indicate that in the porcine to cynomolgus monkey setting anti-porcine IgM rather than IgG anti-porcine antibody levels seem to be of predominant importance for the induction of HAR. By increasing the donor organ size and weight the frequency of the onset of HAR can be at least reduced. This is most likely due to immunoabsorption of the recipients preformed antibodies in the porcine kidney without lethal damage for the graft. After transplantation of Galα1–3Gal positive solid organs into Galα1–3Gal negative recipients, a rapid destruction of the donor organ mediated by preformed recipient antibodies with subsequent complement activation may occur within minutes to hours after reperfusion (1, 2). This reaction has been classified as hyperacute xenograft rejection (HAR) (3, 4). Although there was general consensus that HAR almost invariably occurs in the Galα1–3Gal positive to Galα1–3Gal negative donor/recipient combination (e.g., pig-to-primate or pig-to-human transplantation), recent evidence suggests that this is not always the case (5, 6). We therefore analyzed, by using a pig-to-cynomolgus monkey heterotopic xeno-kidney transplantation model, the influence of various factors such as pretransplant anti-porcine antibody titer, cold ischemia time, and donor organ weight on the onset of HAR. MATERIAL AND METHODS Animals. Fifteen unmodified large white land race pigs (Schweinezuchtverband Weser-Ems, Oldenburg, Germany) were used as donor animals. Pigs were 3–12 weeks old with a weight of 4.1–25.0 kg. As recipients 15 adult cynomolgus monkeys (Macaca fascicularis) weighing 3.8 to 8.0 kg and aged 1.5 to 3.5 years were used. Animals were purchased from the German Primate Center, Göttingen, Germany. All surgical procedures and postoperative care were carried out in accordance with the German National Institute for animal care and were approved by the local authorities for animal welfare. Surgical technique. Donor pigs were anesthetized with ketamine and intubated. Anesthesia was maintained with isoflurane and N2O/O2. Through a midline laparotomy and after anticoagulation with heparine 300 I.E./kg body weight, the abdominal aorta was canulated and the kidneys were perfused in situ with cold preservation solution (HTK, Köhler-Chemie GmbH, Alsbach-Höhnlein, Germany). After perfusion, the kidneys were dissected, removed, and stored in ice cold perfusion solution until they were used. The cold ischemia time was between 75 and 310 min. Recipient cynomolgus monkeys were anesthetized using ketamine and propofol and intubated. Anesthesia was maintained with isoflurane and N2O/O2 and supplemented with boli of fentanyl. Through a midline abdominal incision the inferior aorta and inferior vena cava were exposed. Transplantation was carried out by end-to-side anastomosis of the donor vessels to the aorta and inferior vena cava. The ureter was implanted into the bladder using a modified Gregoir technique. Transplantation was performed creating either a non-life-supporting (n=7) or a life-supporting model (ligature of native ureters; n=8) (7). HAR was defined as thrombosis of the graft 2 hr after reperfusion with standard- and immuno-histological signs of HAR (5, 6). Primary kidney function was defined as production of urine at the end of operation. Immunosuppressive therapy. Immunosuppression was commenced on the day before the operation and continued until day 28 of the experiment. Immunosuppression consisted of cyclosporine (CyA), cyclophosphamide (CyP), and prednisolone (8, 9). CyA was started orally on day −1 at a dose of 100 mg/kg. After transplantation CyA was administered orally at a dose adjusted to achieve a blood trough level of 400–600 ng/liter. CyP was administered i.v. on days −1, 0, 2 at 40, 10, and 20 mg/kg respectively. Further CyP dosing was adjusted to white blood cell count (WBC) in a way to maintain levels of 2–4×109 leukocytes/liter (9, 10). Steroids were administered as follows: 1 mg/kg methylprednisolone i.v. on day 0, followed by 1 mg/kg prednisolone p.o. in the first week after transplantation, then reduced by 0.05 mg/kg on each subsequent day to reach a maintenance therapy of 0.2 mg/kg. Postoperative blood monitoring. Blood samples were taken every day for assessment of full blood count, urea, creatinine, and electrolytes. CyA blood levels were determined using a monoclonal antibody based method (EMIT 2000, Behringwerke, Liederbach, Germany). Flow cytometry. Preoperative sera from all monkeys were screened for antiporcine antibody titer using a flow cytometric assay for analysis (11). After blood drawing, serum was prepared by centrifugation at 3400×g for 10 min at 20°C. For detection of anti-porcine xenoreactive antibodies, frozen aliquots of porcine peripheral blood leukocytes obtained from an individual large white pig were thawed and 0.5×105 cells were stained by 20 μl of the respective cynomolgus or human serum in different solutions. After 20 min incubation at 4°C, cells were washed two times with phosphate-buffered saline containing 1% bovine serum albumin and 0.1% sodiumazide. Bound cynomolgus or human antibody was detected with goat anti-human fluorescein isothiocyanate secondary antibodies detecting IgG (Dianova, Hamburg, Germany) and IgM (Dianova), respectively. These antibodies are known to cross-react with cynomolgus immunoglobulins. The antibodies had been preabsorbed using porcine serum. After incubation for 20 min at 4°C cells were washed again two times and then analyzed on a FACScan (Becton Dickinson, Mountain View, CA) Cytometer. 5×103 cells were analyzed using list mode data. Statistics. For statistical analysis SPSS Version 8.0 (SPSS Inc., Chicago, IL) was used. Mann-Whitney U test was used to estimate significance. P <0.05 was considered statistically significant. RESULTS HAR as defined by complete thrombosis of the graft after 2 hr after reperfusion with vessel anastomoses patent occurred in only 5 of 15 experiments (Table 1). Table 1: Influence of kidney weight, cold ischemia time (CIT), and preoperative anti-porcine antibody levels on HAR and primary kidney functionAnalysis of standard hematoxylin and eosin pathology in these five grafts showed classical signs for HAR such as thrombosis, intragraft hemorrhage and structural disintegration (data not shown). Additional staining for immunoglobulins as well as complement components by immunohistochemistry demonstrated strong deposition of IgM and C5b-9 (Fig. 1) as well as IgG and C3 (data not shown) in these kidneys. Analysis of cold ischemia times showed no clear-cut correlation between organ storage time and susceptibility toward HAR (Table 1). Figure 1: Immunohistochemical deposition of IgM and C5b-9 in a porcine kidney during HAR (left) and in a porcine kidney with primary graft function (right). a, IgM deposition in a porcine kidney during HAR. b, IgM deposition in a porcine kidney with primary function. c, C5b-9 deposition in a porcine kidney during HAR. d, C5b-9 deposition in a porcine kidney with primary function.Analysis of antiporcine antibody levels in the monkeys by FACS failed to show any correlation between preoperative IgG antibody activity and HAR. There was, however, a significant influence of elevated anti-porcine IgM levels on the onset of HAR (Table 1). Four of five animals with HAR had preoperative antiporcine IgM levels of >200 mean fluorescence intensity (MFI). The anti-porcine IgM levels in animals with HAR were significant (P <0.05) higher than those in animals without HAR. Among the 5 animals with IgM levels of more than 200 mean fluorescence intensity four experienced HAR. The only animal experiencing HAR despite a preoperative IgM level of<200 mean fluorescence intensity had an exceptional high IgG level (1286 mean fluorescence intensity) indicating presensitization. The anti-porcine antibody levels as detected in the 15 monkeys were compared with those measured in 5 healthy human volunteers (Table 1). Anti-porcine IgM levels were found to be substantially lower in humans if compared to the monkeys (38.1±35.2 vs. 163.9±79.0; mean±SD). In contrast IgG levels were in general higher in the humans tested (971.9±765.9 vs. 172.2±232.8; mean±SD). When the influence of kidney organ size on the development of HAR was investigated, all 5 episodes of HAR were observed in the 12 animals receiving relative small kidneys (20–48 g). In addition, 11 of 12 of these kidneys never produced any urine, the remaining kidney in this group which showed primary function weighted 48 g. In contrast, all three kidneys obtained from larger donors, weighting 75, 85, and 90 g, showed primary urine production; no evidence of HAR was observed in these kidneys. Analysis by immunohistochemistry demonstrated that in all grafts (also in the large kidneys not undergoing HAR) a profound deposition of cynomolgus immunoglobulin IgM (Fig. 1) and IgG (data not shown) could be detected. Moreover, FACS analysis of recipient sera demonstrated a sharp decrease of antiporcine antibody levels after transplantation during the first 2 postoperative days in all animals (data not shown). Interestingly, when in the larger grafts C5b-9 deposition was analyzed by immunohistochemistry, only scare amounts of C5b-9 were demonstrated indicating incomplete activation of the complement cascade (Fig. 1). In one animal (no. 411) a right porcine kidney weighting 20 g was hyperacutely rejected. In this animal, after removal of the hyperacutely rejected graft, a second kidney (left kidney from the same donor) was transplanted into the same recipient (23 g). This second organ did not undergo hyperacute rejection and showed primary urine production. DISCUSSION After interspecies transplantation of Galα1–3Gal positive vascularized organs into Galα1–3Gal negative recipients a rapid destruction of the graft with intragraft hemorrhage and thrombosis may occur. This immunological phenomenon is thought to be mediated primarily by binding of preformed recipient antibodies and subsequent complement activation and is termed HAR (1–3). Up to recently the available evidence suggested that HAR invariably occurrs in any xenotransplantation model with Galα1–3Gal positive to Galα1–3Gal negative species combinations. In 1998 Zaidi et al. (5) for the first time described that in adult cynomolgus monkeys receiving unmodified porcine kidneys for control purposes, five of six organs in this series obviously did not undergo HAR and functioned up to 30 days. The pathophysiological factors determining the onset of HAR on the one hand or protecting a given graft from HAR on the other hand remained unknown. Using newborn baboons with low serum levels of cytotoxic anti-pig antibodies as organ recipients Michler et al. (13) demonstrated that HAR in a model of pig to newborn baboon heart transplantation could be avoided. This was shown to be due to the relative absence of anti-porcine IgM antibodies (predominantly anti-Galα1–3Gal) in the newborn primates (12, 13). In our experiments mature cynomolgus monkeys age 1.5–3.5 years were used as organ recipients. The anti-pig PBL antibody levels as measured by FACS (11) in these monkeys were compared with the xenoreactivity in five human control sera. These data demonstrated that the anti-porcine IgM but not IgG levels were considerably higher in the cynomolgus recipients than in the healthy human controls investigated. IgM rather than IgG anti-porcine antibodies are known to primarily induce HAR in the pig to primate xenotransplantation setting (14–16). Therefore, the absence of HAR in most of the experiments reported seems not to be due to an immunological immaturity of the recipients as in the model of Michler et al. (12, 13). When in our experiments different variables such as cold ischemia time, pretransplant anti-porcine antibody levels, or donor organ weight were analyzed as to their influence on HAR the duration of cold ischemia time (75 to 310 min) did not had any influence on the frequency of HAR. The question whether prolonged cold ischemia time might impair the quality of the primary organ function, however, has to be investigated further by generation of even more prolonged storage times such as 12 or 24 hr. This would be of particular interest for eventual future shipments of transgenic porcine organs; our data (up to 5 hr storage) point to the possibility of the shipment of xenografts that had been procured in a central institution rather than transferring transgenic pigs to the transplant center. As shown by others (14, 16), we have also found that preoperative antiporcine antibodies of the IgM subtype might be more effective in mediating HAR than of the IgG subtype. However, we have also observed one HAR in an animal with low IgM levels but high IgG levels. It has been postulated that in addition to anti-porcine antibody levels in the recipient other factors such as alternative complement activation might also play a role in HAR (17). In fact, the development of HAR might be multifactorial (with antiporcine antibody levels as measured here being only one contributing factor) rendering the analysis of individual parameters as to their influence of HAR more complicated. In our series of 15 experiments, the most pronounced effect as to abrogation of HAR was exerted by donor organ weight. In fact, no HAR was observed in donor organs exceeding 70 g of weight corresponding to a donor age of 12 weeks or more; all these organs showed primary graft function. In one experiment two small kidneys were transplanted sequentially into the same recipient. Although the first graft was hyperacutely rejected, the second graft did not undergo HAR and showed initial function. In accordance with the results obtained by the use of the larger grafts, this would be in line with an in vivo absorption of preformed antibodies from the recipients circulation by the first graft leading to HAR of the first organ and allowing for primary graft function of the subsequent kidney. Absorption of preformed antiporcine antibodies by ex vivo organ perfusion using livers or kidneys for abrogation of HAR has been described several times by various groups (18, 19). Taking into account the demonstrated deposition of anti-porcine antibodies in all the grafts reported and the parallel decrease in systemic anti-porcine antibody titers after graft reperfusion as shown by FACS, in all likelihood the protective effect of increasing organ size toward onset of HAR could therefore reflect an in vivo immunoabsorption of preformed recipient antibodies without lethal damage to the graft. This theory would be in line with the fact that the larger grafts are not protected against later rejection mechanisms such as acute xenograft rejection mediated by induced antibodies. Currently, most preclinical models of discordant kidney xenotransplantation are described by using the pig to cynomolgus monkey combination. The observed protective effect of oversized donor kidneys on HAR described in our experiments might be helpful for further preclinical research in this area.