International Journal of Cell Cloning 4: 122-126 Suppl 1 (1986) Manipulation of T Cells in the Thnsplant Inoculum J. Kersey Bone Marrow Transplantation Program, University of Minnesota, Minneapolis, MN, USA Key Words. T cell depletion GVHD - Immunotoxin Ricin Bone marrow transplantation I was asked to speak on the topic of depletion of marrow T lymphocytes as a method of preventing graft-versus-host disease (GVHD). I will discuss with you both murine and human results. I would like mostly though to confine my remarks to the Minnesota experience. The Minnesota experience is in large part due to the fine efforts of Doctors Daniel Vallera, Tucker LeBien, Norma Ramsay, Philip McGlave and Alexandra Filipovich. Now let me begin with the murine model, which we have found over the years an extremely useful model for studying human marrow transplantation. I begin with some data that have been generated in that model in which we studied lethal GVHD. The model utilizes total body irradiation and therefore these animals are not immunocompetent. This model is one that utilizes H2 incompatible animals, that is, animals that differ in both class I and class 11 antigens. Animals which are C57b/6 recipients are irradiated with 900 rads of total body irradiation. They then receive BALB/c marrow which has been treated and we look at GVHD mortality. Since this model utilizes a mismatched combination, the GVHD mortality is very high. This is an excellent model for early acute GVHD mortality. If we look at these animals that receive untreated allogeneic bone marrowspleen cells, the mortality is extremely high. We pretreat the donor marrow to prevent lethal GVHD, using certain antibodies for T cell depletion. If we use an antibody against the LY 1 determinant, which is found on many if not all T cells in the mouse, we can abrogate 90% of the acute GVHD mortality. If we use other Correspondence: Dr. John Kersey, University of Minnesota, Box 86, Mayo Memorial Building, 420 Delaware St. SE, Minneapolis, MN 55455 (USA). 0737-1454/86/$2.00/0 @AlphaMed Press, Inc.
T Cells in the Transplant Inoculum 123 antibodies, for example, LY 2, the mortality rate is reduced somewhat but not nearly to the extent that it is with the pan-t antibody. Depending on the degree of mismatching, there are somewhat variable results with varying antibodies in the presence of complement. It is still not clear which subsets are important for acute mortality due to GVHD. However, based on many studies, it appears that pan-t antibodies are effective in preventing GVHD. The negative side of T cell depletion is that marrow engraftment is often inhibited when the recipient is immunocompetent and the donor marrow is T celldepleted. For example, if the mouse receives somewhat less immunoconditioning than in total body irradiation (e.g., with total lymphoid irradiation and cyclophosphamide) and the marrow is not T celldepleted, essentially au of these animals will engraft. If, on the other hand, the marrow is T cell-depleted, many of these animals do not develop long-term grafts. In contrast, when we gave more immunosuppression in the form of total body irradiation, greater than 90% of animals engrafted. We have attempted to overcome this failure of engraftment by giving more cells, and still only 14% of the animals were engrafted. This failure of engraftment in T-depleted animals appears to be due to cells in the marrow inoculum which are rejection stoppers. They are cells that stop rejection from occurring. Most experiments to date have used antibody plus complement. Antibody plus complement is a relatively effective way to eliminate cells from the marrow inoculum. However, complement is quite variable; there are problems with potency, consistency and possible stem cell activity. In our view, in the use of antibodies to remove cells from donor marrow, an effector other than complement would be useful. We have been working with the toxic molecule ricin, which we and others have covalently linked to antibody. Ricin is an extremely interesting molecule. Derived from the castor bean, it is probably the most potent poison that exists in nature. One molecule is sufficient to kill a single cell. It is composed of two 30 kda subunits, an A chain which catalytically inhibits protein synthesis and a B chain which binds galactose residues on the cell surfaces and also facilitates entry of an A chain into the cytosol. Anti-T cell antibodies are covalently linked by thioether linkage to the ricin molecule. The antibody is directed to T cells in the marrow inoculum. Through binding of the antibody, ricin is directed toward this cell, but not to stem cells which lack the determinant detected by the monoclonal antibody. The immunotoxin binds first to the antigen on the cell surface and the antigenricin conjugate is then internalized. The ricin is subsequently released and binds the 60s ribosomes where it inhibits protein synthesis. Cells are killed by protein
Kersey 124 starvation. One can use either intact ricin or the A chain of ricin. We find that the anti-t cell antibodies conjugated to ricin are more potent than the A chain as an antibody-conjugate. In our studies of the murine bone marrow system, we asked whether we could prevent fatal GVHD using antibody-ricin conjugate, just as we had used antibody and complement. We found that there was excellent protection from fatal GVHD with antibody-ricin. Overall, several of these methods were satisfactory in terms of removing T cells from the bone marrow. AU of them resulted essentially in preservation of bone marrow precursors, but antibody ricin had the advantage of being simpler and the most reproducible in our hands. We have used three antibodies as ricin conjugates: T101, a CD5 antibody which is very similar to the LY 1 of the mouse; UCHT 1, a CD3 antibody; and TA 1, which is directed against a determinant found on granulocytes, monocytes and T cells. TA 1 is of interest in that this determinant is found on natural killer cells in addition to T cells. The three antibodies have very broad anti-t reactivity and also anti-nk activity. These antibody-rich conjugates are quite effective in inhibiting the generation of cytotoxic T lymphocytes, in that as little as 300 ng per ml of the antibody-ricin is effective in completely abrogating the generation of cytotoxic T lymphocytes. At lower doses of each, the degree of inhibition is significantly reduced. With this preclinical data, we began our clinical pilot studies. The immunotoxin treatment of the bone marrow is quite simple. The marrow is heparinized and depleted of red cells by heta starch, and nucleated cells are prepared on a Ficoll-Isopaque layer. The immunotoxin treatment lasts two hours, cells are washed to eliminate residual immunotoxin and then administered to the transplant recipient. In the first two patients studied, the PHA response, the MLR response, natural killer cells, alloantigen and sensitized cytotoxicity showed no residual response following this immunotoxin treatment of donor marrow. The patients studied to date in these pilot studies have been patients with highrisk leukemia. Patients have been conditioned for transplantation using cyclophosphamide followed by total body irradiation of 165 cgy per dose, twice a day. They receive eight doses for a total of 1,320 cgy. Two patients received alternative conditioning that utilized cytosine arabinoside rather than cyclophosphamide. Our pilot study included 17 high-risk leukemia patients, 12 with acute leukemia and 5 with chronic myelogenous leukemia (CML). All of them had T cell immunotoxin-treated marrow and HLA-matched donors. Three of the patients with acute leukemia were in relapse. Patients ranged in age from 7 to 53, with a total of 7 patients over the age of 30, representing a relatively large number of older patients. There was no toxicity from antibody-ricin immunotoxin in any patient.
T Cells in the Transplant Inoculum 125 We also carried out a pilot study with nine patients who were mismatchedi.e., the donors were haploidentical. All but two donors were parents. One patient had Wiskott-Aldrich syndrome, one had Wolman s disease, another had SCID, and there were three patients with CML and three with acute leukemia. These patients ranged in age from 1 to 29. In this group, only one developed GVHD of any kind and that patient developed grade III. We do not know why that patient developed grade III because the T cell depletion seemed adequate; the only possible explanation is that the patient had a severe concurrent parainfluenza inktion which may have contributed to the GVHD picture. Graft rejection was a major problem in the patients; only one patient did not have evidence of a problem with his graft and that was the patient with SCID. Mixed chimerism was found frequently in this group. One patient is alive at 16 months, and two patients developed lymphoma, one at 24 months and the other at an earlier point. Another approach to the use of antibody-ricin immunotoxins is autologous transplantation. We have had six patients with T cell leukemia who have had their marrow treated with antibodies conjugated to ricin. Of the six patients, five of them were TALL and the sixth was a T lymphoma (they were children and young adults). All of them had prompt evidence of engraftment. Of these six patients, two patients are alive and relpase-free. One is at 16 months and the other is at 8 months. In summary, our experience suggests that antibody-ricin immunotoxins can be potent and specific reagents and that ex vivo depletion of T cells is relatively safe and simple. Immunotoxins can be produced in large quantities both for autologous and allogeneic transplantation. I believe that the T cell depletion method does reduce the incidence of fatal GVHD. On the other hand, the graft rejections in the patients receiving T cell-depleted marrow strongly suggest that T cells do contribute to engraftment in allogeneic transplantation. The challenge is to try to distinguish the lymphoid cells which produce GVHD from the graft facilitating cells. Discussion Keating: Were the T cell ALLs purged, in relapse or in remission? Kersey: They were in remission. They were all patients who had failed primary therapy and then were purged in remission. Shustick: Were the haplotype mismatches that were T cell-depleted and developed the two lymphomas EBV-positive?
Kersey 126 Kersey: Both of them contained EBV and both of them were B cell neoplasms. Whether or not EBV produced the neoplasm, I don t know. Ho: Perhaps I should just mention our current approach to this problem of a failure of engraftment or graft rejection that we ve seen with the T cell manipulated marrows. In randomized studies with patients with leukemia who received HLA identical grafts, half of them with T cell depletion using antibodies and complement, we showed that there was a definite effect on graft-versus-host disease. I think this is widely accepted now. In the groups that are doing T cell depletion, there is this definite effect on the reduction of graft-versus-host disease. But the overall survival of the patients in the two groups was not changed. Patients were dying for various reasons. In patients who received the intact marrow, the main cause of death was graft-versus-host disease, and in those patients who got T celldepleted marrow, the main cause of death was leukemia relapse. Those patients who had their counts decreasing had either overt leukemic cells present in their marrows or they had cytogenic genetic markers that showed an abnormality similar to the overt leukemia that the patients had previously. Because of this, we felt it might be worthwhile to continue with the T cell depletion in order to have the effect we wanted on graft-versus-host disease, and to try to get rid of residual leukemia by further immunosuppression of the recipient prior to the transplant. We are in the middle of a pilot study in which the patients are receiving increased radiation. We are giving high-dose ara-c together with six doses of total body irradiation and an additional dose of total lymphoid radiation to the patients. We have done 11 patients in this fashion. The first 9 or 10 actually engrafted very well. We recently had one patient who showed the devastating effect of lack of engraftment. The patient actually did have a transient rise in counts about two weeks after the marrow was given, and then became cytopenic with an empty marrow once again. We were forced to give some more untreated marrow to that patient, together with more chemotherapy.