Anti-CD3-immunotoxin (α-CD3-IT*) has shown outstanding potential to facilitate tolerance induction in nonhuman primates, a property attributed to its unusual efficiency in depleting the sessile lymphoid T cells as well as circulating T cells (1-3). A conjugate of monoclonal IgG1 anti-rhesus CD3ε (FN18) and the mutant diphtheria toxin CRM9, α-CD3-IT has <0.1% of the systemic toxicity of wild-type diphtheria toxin (4, 5). The conjugate is directed predominantly to T cells via the antibody's high affinity for the CD3 receptor. Toxicity is achieved by CD3ε-mediated internalization of bound CRM9 toxin by way of the endosomal pathway and intracellular translocation (6). Other immunotoxins used to treat cancer or graft versus host disease have exhibited side effects of vascular leak syndrome, hepatotoxicity, and nephrotoxicity, dose-related complications that constrain effectiveness and clinical application (7-9). Before clinical testing of a cognate anti-human CD3ε-IT construct (10), it seems important to have available a comprehensive documentation of preclinical tolerability and side effects of the anti-nonhuman primate CD3ε-IT construct (10). Thus, we summarize our preclinical experience of tolerability and side effects of α-CD3-IT treatment in nonhuman primate kidney and islet transplant recipients.
Normal, 3-4 kg male rhesus monkeys (Maccaca mulatatta) were obtained from breeding colonies al LABS (Yemassee, SC), University of Wisconsin (Madison, WI), and Hazelton (Alice, TX). After quarantine conditioning for 60 days, 37 transplant recipients were selected to be free of measurable antidiphtheria toxin antibodies at a 1:100 serum dilution (5). Surgical and nonsurgical procedures were carried out in accordance the National Institutes of Health guide for The Care and Use of Primates under supervision of the Institutional Animal Care and Use Committee. The animals were tranquilized with 10 mg/kg ketamine for handling and received Butorphanol analgesia postoperatively for 2 days. Heterotopic kidney allografts were transplanted as described with intrinsic bilateral nephrectomy performed to provide a working allograft model (1). Intravenous 0.45% saline was given on days 0-3 and as needed for an additional 7 days to prevent dehydration. On the day of surgery, the volume infused was adjusted to replace the volume of urine output. On days 1-3 posttransplant, the volume of saline was 30 ml/kg. Cepahzolin was given as prophylactic antibiotic therapy on days 0-5 at 12.5 mg/kg twice a day. Nutritional support was administered 0-2 days postsurgery (Ensure, 30-50 ml twice daily). In addition to kidney transplants recipients, five spontaneously diabetic primates (three Macacca fasicularis, one M. mulatta, and one Ceropithecus aethiops) also received α-CD3-IT, three of which had an pancreas islet transplant via the portal vein as described (11).
α-CD3-IT was administered i.v. at a total dose of 133-300 μg/kg given over 3 days beginning on the day of transplant. Some recipients received additional treatment with 15-deoxyspergualin (DSG, 2.5 mg/kg/i.v. for 4-14 days) and methylprednisolone (MP for 3 days at a tapering dose from 7 to mg/kg/i.v. to 0). Another group received cyclosporine (20 mg/kg/i.v. on day 0 and Neoral 120 mg/kg/oral, dose adjusted to achieve 400-600 ng/ml for 4 days), MP (15 mg/kg/i.v. for 4 days), and Simulec (20 mg/kg/i.v. on day 0). Also, nine animals in the DSG plus MP group received donor peripheral blood CD34+ cells (at 1×105-4×107) on days 1-3. Five kidney transplant recipients given the parent α-CD3-monoclonal antibody (FN18 without the toxin) instead of α-CD3-IT served as controls. The University of Alabama at Birmingham Hospital Outreach Laboratory performed liver, kidney, and hematology test on days 1, 2, 3, 7, 14, 21, and 28 after treatment. Open liver and kidney biopsies for histopathology were performed in three recipients at 24 hr after α-CD3-IT administration.
Vascular leak syndrome was the main side effect of α-CD3-IT when initiated on the day of transplant without MP or other immunosuppressive agents. This syndrome was characterized by hypoalbuminemia without proteinuria, weight gain, and edema, which was aggravated by posttransplant infusion of 0.45% saline. Vascular leak syndrome was accompanied by significant prominent proinflammatory cytokine release in transplant recipients and was entirely prevented by coadministration of a short course of low dose MP and DSG as previously reported (3) or more recently with high dose MP and cyclosporine A (Thomas JM, unpublished observations).
Hepatotoxicity is the most commonly documented side effect of other immunotoxins (7, 8). To examine the potential hepatotoxic effect of α-CD3-IT, we compared liver function test from animals treated with α-CD3-IT (Fig. 1A) or the parent α-CD3-mAb (Fig. 1B). No significant difference was observed in aspartate amino transferase, alanine amino transferase, alkaline phosphatase, or γ-glutamyl transpeptidase between the α-CD3-IT and α-CD3 mAb treated animals. Similarly, no significant liver function test differences were observed in recipients treated with α-CD3-IT versus α-CD3-IT in combination with a short course of immunosuppressive drugs. Furthermore, the findings were similar in kidney and islet transplant recipients and in those treated with low (total 150 μg/kg) versus high (total 300 μg/kg) dose of α-CD3-IT (data not shown). The absence of hepatotoxicity was corroborated by histopathological analysis (Fig. 2, A and B) 24 and 72 hr after α-CD3-IT treatment. Thus, there was no evidence for hepatotoxicity associated with peritransplant administration of this α-CD3-IT construct.
Renal dysfunction is another potential complication of immunotoxin administration (9). In these monkeys, no difference in serum creatinine (Fig. 1C) or blood urea nitrogen (Fig. 1D) was observed in kidney transplant recipients treated either with α-CD3-IT or α-CD3 monoclonal antibodies. Normal kidney transplant histopathology was observed 3 days after the immunotoxin treatment (Fig. 2, B and C), indicating a lack of nephrotoxicity.
Despite a 95-99% decrease in lymph node T cells and in peripheral blood at 5-7 days after the treatment (1), no evidence of alterations in the total white blood cell, hematocrit, or platelet counts were observed in any animals treated with α-CD3-IT versus α-CD3-mAb (Fig. 3, A, B, and C, respectively). Furthermore, increased incidence of infection or neoplasia was not seen after α-CD3-IT treatment. Thus, the toxic effect of α-CD3-IT on T cells was exquisitely specific without sustained risk of apparent immune deficiency, presumably by virtue of sparing alternative components of innate and acquired immunity.
Erythema and skin induration are lesser side effects reported after immunotoxin treatment in humans (7, 8). Similar to earlier observations in monkeys (1, 3), 20% of our series exhibited a rash at 24 hr after α-CD3-IT treatment, characterized by facial erythema that disappeared at 48 hr without specific treatment. α-CD3-induced rash, therefore, appears to be self-limiting.
Compared to control recipients given the α-CD3 mAb (Fig. 4C), transient but significant weight loss (14±5%) occurred in all recipients after α-CD3-IT administration. The data in Figure 4A were from transplant recipients given α-CD3-IT without infusion of donor CD34+ cells. Of those followed for >3-48 months, all but one had >100% weight recovery. The single exception died of an unrelated accident during his weight recovery phase. In contrast to weight recovery in these animals, the recipients given donor CD34+ cells, exhibited a different pattern. Of eight recipients surviving without rejection after infusion of donor CD34+ cells (Fig. 4B), 50% (4/8) showed failure to survive or recovery toward pretransplant weight and had to be euthanized for humane purposes. Importantly, only the recipients given donor CD34+ cells developed wasting. The cause of weight loss in α-CD3-IT-treated animals is unknown and is currently under study in our laboratory. Multiple factors seems to be involved. Rhesus donors, for example, presented weight loss (8.1±1.1%) after the nephrectomy procedure (Fig. 4D). Similar weight loss (8.0±1%) was observed in animals treated with α-CD3-IT without kidney transplantation. In our most recent series of kidney transplantation in recipients treated with α-CD3-IT, a reduction of weight loss (10±3%, Fig. 4D) was observed with enteral nutritional support early after transplantation (Ensure 50-100 mL/BID/for 2 weeks). Thus, significant (≥15%) weight loss can be prevented by transient enteral nutritional support.
The wasting syndrome that developed in recipients given α-CD3-IT plus donor CD34+ cells was accompanied by immune deficiency and focal, lymphoid, nonsystemic chimerism (Thomas JM, unpublished observations). These conditions likely represent chronic graft versus host disease induced by a small percentage (2-10%) of mature T cell contaminants present in the allogeneic CD34+ cell suspension (Thomas JM, manuscript in preparation). Small numbers of contaminating donor T cells in CD34+ populations or in mini-infusions of donor leukocytes are known to precipitate chronic graft versus host disease in humans (12, 13). Consistent with this notion are observations of failure to survive in rhesus organ transplant recipients given rabbit antithymocyte globulin with posttransplant lymphoid irradiation and splenectomy in combination with low doses of donor bone marrow that was not purged of T cells (Thomas JM, manuscript in preparation). Thus, in outbred primate species, even tiny numbers of mature donor T cells introduced into recipients that lack lymphoid T cells may have graft versus host disease potential. It is also possible for biologically significant numbers of mature T cells to be transferred within an organ allograft. If so, a clinical caveat for α-CD3-IT application would be its administration on the day of transplant to eliminate passenger T cells.
The application of brief, peritransplant α-CD3-IT therapy as a strategy for organ allograft tolerance induction in this outbred primate species is designed to ostensibly "reset" the immune system by highly efficient T cell depletion of populations in lymphoid tissue in the absence of maintenance therapy. Thus, conceptually, new T cell clones that repopulate and mature posttransplant should recognize the graft as self. Previously, we have shown that peritransplant coadministration of MP and DSG for 14 days effectively prevents proinflammatory cytokine release, leading to IL-4 production and stable, rejection-free graft survival in 80% of α-CD3-IT-treated recipients without chronic immunosuppressive drugs (3), the mechanism of which is discussed elsewhere (Thomas JM, manuscript in preparation). In our study, we simply report no measurable hepatotoxic or nephrotoxic side effects related to this brief α-CD3-IT induction treatment in rhesus monkey transplant recipients.
In summary, for α-CD3-IT immunotoxin-based strategies to become effective drugs for human transplantation, prevention of immunotoxin-induced side effects is mandatory. We have demonstrated in a preclinical primate model, that the side effects of α-CD3-IT are manageable and should not prevent therapeutic clinical application. Unique characteristics of α-CD3-IT, the stability of antibody-mutant diphtheria conjugate, the high specificity for CD3+ cells, and the efficient intracellular translocation of immunotoxin via the cycling CD3ε receptor (4-6), collectively appear to obviate the troublesome side effects that have hindered immunotoxin therapy to date.
Acknowledgments. The authors thank Nat Borden and Sharon Mcknight for skilled primate care assistance.
1. Thomas JM, Neville DM, Contreras JL, et al. Preclinical studies of allograft tolerance in rhesus monkeys: a novel anti-CD3-immunotoxin given peritransplant with donor bone marrow induces operational tolerance to kidney allografts. Transplantation 1997; 64: 124.
2. Knechtle SJ, Vargo D, Fechner J, et al. FN18-CRM9 immunotoxin promotes tolerance in primate renal allografts. Transplantation 1997; 63: 1.
3. Contreras JL, Wang PX, Eckhoff DE, et al. Peritransplant tolerance induction with anti-CD3-immunotoxin: a matter of proinflammatory cytokine control. Transplantation 1998; 65: 1159.
4. Neville DMJ, Scharff J, Srinivasachar K. In vivo T-cell ablation by a holo-immunotoxin directed at human CD3. Proc Natl Acad Sci USA 1992; 89: 2585.
5. Neville DMJ, Scharff J, Hu HZ, et al. A new reagent for the induction of T-cell depletion, anti-CD3-CRM9. J Immunother Emphasis Tumor Immunol 1996; 19: 85.
6. Neville D Jr, Scharff J, Srinivasachar K. Anti-T cell immunotoxins: a look at post-endocytic receptor-mediated routing. J Controlled Release 1993; 24: 133.
7. Thrush GR, Lark LR, Clinchy BC, Vitetta ES. Immunotoxins: an update. Annu Rev Immunol 1996; 14: 49.
8. Pastan IH, Pai LH, Brinkmann U, Fitzgerald DJ. Recombinant toxins: new therapeutic agents for cancer. Ann NY Acad Sci 1995; 758: 345.
9. Vallera DA, Panoskaltsis-Mortari A, Blazar BR. Renal dysfunction accounts for the dose limiting toxicity of DT390anti-CD3sFv, a potential new recombinant anti-GVHD immunotoxin. Protein Eng 1997; 10: 1071.
10. Thompson J, Hu H, Scharff J, Neville DMJ. An anti-CD3 single-chain immunotoxin with a truncated diphtheria toxin avoids inhibition by pre-existing antibodies in human blood. J Biol Chem 1995; 270: 28037.
11. Thomas FT, Ricordi C, Contreras JL, et al. Reversal of naturally occuring diabetes in primates by unmodified islet xenografts without chronic immunosuppression. Transplantation 1999 (in press). Transplantation 1999; 67: 846.
12. Storek J, Gooley T, Siadak M, Besinger WI, Maloney DG, et al. Allogeneic peripheral blood stem cell transplantation may be associated with a high risk of chronic graft-versus-host disease. Blood 1997; 90: 4705.
13. Smit WM, Rijnbeek M, van Bergen CA, Willemze R, Falkenburg JH. Generation of leukemia-reactive cytotoxic lymphocytes from HLA-identical donors of patients with chronic myeloid leukemia using modifications of a limiting dilution assay. Bone Marrow Transplant 1998; 21: 553.
* Abbreviations: α-CD3-IT, anti-CD3-immunotoxin; DSG, 15-deoxyspergualin; MP, methylprednisolone.