| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Cancer Therapy: Preclinical |
T cells, hu14.18 Antibody, and the Immunocytokine Fc-IL7 in Disseminated Neuroblastoma
Authors' Affiliations: Departments of 1 Hematology-Oncology and 2 Animal Resources Center, St. Jude Children's Research Hospital, Memphis, Tennessee, and 3 EMD Lexigen Research Center, Billerica, Massachusetts
Requests for reprints: Mario Otto, Department of Hematology-Oncology, St. Jude Children's Research Hospital, Mailstop 321, 332 North Lauderdale Street, Memphis, TN 38105-2794. Phone: 901-495-3695; Fax: 901-495-4023; E-mail: mario.otto{at}stjude.org.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
T cells with the humanized anti-GD2 antibody hu14.18 and a novel fusion cytokine, Fc-IL7.
Experimental Design: T cells were large-scale enriched from leukapheresis product obtained from granulocyte colony-stimulating factor–mobilized donors. T cell cytotoxicity was tested in a europium-TDA release assay. The effect of Fc-IL7 on T-cell survival in vitro was assessed by flow cytometry. NOD.CB17-Prkdcscid/J mice received 1 x 106 NB-1691 neuroblastoma cells via the tail vein 5 to 6 days before therapy began. Treatment, for five consecutive weeks, consisted of injections of 1 x 106 T cells weekly, 1 x 106 T cells weekly, and 20 µg hu14.18 antibody four times per week, or 1 x 106 T cells weekly with 20 µg hu14.18 antibody four times per week, and 20 µg Fc-IL7 once weekly.
Results: The natural cytotoxicity of T cells to NB-1691 cells in vitro was dramatically enhanced by hu14.18 antibody. Fc-IL7 effectively kept cultured T cells viable. Combination therapy with T cells and hu14.18 antibody significantly enhanced survival (P = 0.001), as did treatment with T cells, hu14.18 antibody, and Fc-IL7 (P = 0.005). Inclusion of Fc-IL7 offered an additional survival benefit (P = 0.04).
Conclusions: We have shown a new and promising immunotherapy regimen for neuroblastoma that requires clinical evaluation. Our approach might also serve as a therapeutic model for other malignancies.
T cells), a tumor-targeting antibody, and, perhaps, a cytokine (which might enhance survival or cytotoxicity of the effector cells) may be required, especially when the patient's immune system is not yet reconstituted.
We propose the transfusion of potent, nonalloreactive, allogeneic T cells, a small subset of T cells that exert MHC-unrestricted natural cytotoxicity against a variety of tumors (e.g., neuroblastoma, colon carcinoma, rhabdomyosarcoma, and some subsets of leukemias and lymphomas) in vitro (14, 15). In vivo, T cells also respond to certain bacterial, viral, and parasitic infections, and their numbers expand significantly in organ allografts infected with cytomegalovirus or herpes simplex virus. T cells are a unique cell type combining features of both the innate and the acquired immune systems. They constitute about 2% to 10% of all peripheral T cells in humans but are much more widespread in certain tissue types, such as the skin, the intestine, and the reproductive tract, and their population can expand to constitute >50% of the circulating T-cell population within 1 week of appropriate antigenic stimulation (16). Like natural killer cells, T cells exert cytotoxicity by producing IFN-, perforin, RANTES, macrophage inflammatory protein-1
, and macrophage inflammatory protein-1ß. Their ability to exert ADCC and apparent nonalloreactivity make them an attractive candidate as a cell-based immunotherapy agent (17). Furthermore, because the effect of cellular immunotherapy based on T cells may be prolonged by the use of interleukin-7 (IL-7), a cytokine that supports the survival and activity of human T cells in vitro and in vivo (18, 19), we propose a potential therapeutic approach combining immunotherapy using a humanized anti-GD2 antibody with the use of human T cells and IL-7 (in the form of Fc-IL7) in a mouse model of disseminated neuroblastoma. Fc-IL7 is a novel fusion cytokine that has the mature sequence of human IL-7 attached to the COOH terminus of the Fc portion of human IgG1 and has the advantage of a longer half-life in serum than does recombinant IL-7 (20). We report here the results of our novel combined therapeutic approach.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
T cells. To obtain enough human T cells for long-term treatment, we developed a large-scale method for the enrichment of T cells from leukapheresis product on the CliniMACS device (Miltenyi Biotec, Auburn, CA; ref. 21). Leukapheresis products were obtained from healthy adult volunteers who had given written informed consent on a protocol that was approved by the institutional review board. Donors received s.c. injections of 480 µg granulocyte colony-stimulating factor daily for four consecutive days, and on day 5, a single leukapheresis was done on a Cobe Spectra apheresis system (Cobe, Lakewood, CA). Cells were washed twice with PBS (Bio Whittaker, Walkersville, MD) supplemented with 0.5% human serum albumin (Bayer, Elkhart, IN). T cells were labeled with a hapten-conjugated antibody to the T-cell receptor (TCR), and bound antibody was labeled with microbead-conjugated anti-hapten antibody (Miltenyi Biotec) according to the manufacturer's protocol but using 
35% smaller volumes of all recommended agents and buffers. T cells were then positively enriched by magnetic separation using the automated CliniMACS device according to the manufacturer's protocol. Viability was assessed either by conventional trypan blue staining or by flow cytometry using propidium iodide. After the enrichment procedure, cells were promptly injected into the animals (via the tail vein) or divided into aliquots at a concentration of 7.5 x 106 cells/mL in freezing medium containing DMSO (Invitrogen, Grand Island, NY) and frozen at –80°C. Before injection, frozen T cells were quickly thawed, washed twice with RPMI 1640 containing 10% fetal bovine serum (Invitrogen), and tested for viability. The appropriate number of viable cells was then injected into the mouse's tail vein.
Cell lines and culture. For the in vitro assays and to establish the disseminated neuroblastoma xenograft, we used the previously described cell line NB-1691, which was originally obtained from a patient with stage D, MYCN-amplified disease that originated in the adrenal gland (22). The cell line was kindly provided by the Children's Oncology Group (Arcadia, CA). We cultured the erythroleukemic cell line K562 (American Type Culture Collection, Manassas, VA), the neuroblastoma cells, and the T cells in RPMI 1640 (Invitrogen) supplemented with 10% fetal bovine serum, 100 IU/mL penicillin, 50 µg/mL streptomycin, and 2 mmol/L L-glutamine (all additives from Bio Whittaker), in a humidified atmosphere of 5% CO2, 95% air.
Mouse model of disseminated neuroblastoma. We obtained 6- to 8-week-old, female NOD.CB17-Prkdcscid/J mice from The Jackson Laboratory (Bar Harbor, ME). These mice are homozygous for the spontaneous mutation resulting in severe combined immune deficiency (Prkdcscid, commonly referred to as scid) and are characterized by an absence of functional T cells and B cells, lymphopenia, hypogammaglobulinemia, and a normal hematopoietic microenvironment. The NOD background results in natural killer cell deficiency. Therefore, the thymus, lymph nodes, and splenic follicles in these mice are virtually devoid of lymphocytes. The acceptance of allogeneic and xenogeneic grafts by these mice makes them ideal for cell transfer experiments (23). Mice were caged under standard pathogen-free and climate-regulated conditions in the Animal Resources Center at St. Jude Children's Research Hospital. They received pelleted food and tap water and were cared for strictly according to criteria outlined in the NIH Guide for Care and Use of Laboratory Animals and other federal regulations. NB-1691 neuroblastoma cells were grown as described above. When about 90% confluent, cells were harvested after treatment with Cellgro trypsin EDTA reagent (0.25% trypsin/2.21 mmol/L EDTA; Mediatech, Herndon, VA) and filtered to produce a single-cell suspension. Mice were injected via the lateral tail vein with 1 x 106 tumor cells suspended in 200 µL PBS, monitored daily for signs of illness (e.g., visible tumor mass, extended abdomen, arched back, and impaired movement), and killed when showing any of these signs.
Humanized chimeric antibody hu14.18 and Fc-IL7. The humanized chimeric antibody hu14.18 is a genetically engineered antibody that, like its predecessor ch14.18, recognizes the disialoganglioside GD2, which is expressed on NB-1691 cells and many other neuroblastoma cell lines (24, 25). It has been successfully used alone or fused to cytokines, such as IL-2, in animal studies and has entered clinical trials. Fc-IL7 is a novel fusion protein that has the mature sequence of human IL-7 fused to the COOH terminus of the Fc portion of human IgG1 (20). Because of the distinct orientation of the Fc part, the molecule is not removed by the reticuloendothelial system. Furthermore, the specific Fc isotype and the lack of glycosylation as a result of genetic mutation mean that this molecule cannot fix complement or exert ADCC. These characteristics result in a significantly longer serum half-life when compared with human recombinant IL-74. Both reagents were kindly provided by the manufacturer (EMD Lexigen, Billerica, MA).
Treatment schedule. All mice were injected with 1 x 106 NB-1691 neuroblastoma cells via the tail vein at the start of the experiment. Treatment began 5 to 6 days after the injection of tumor cells. In a first experiment, we compared control group 1a (n = 6) that received tumor cells only with group 2 (n = 7), which also received 1 x 106 T cells once per week. Next, we compared control group 1b (tumor cells only, n = 9) with group 3 (n = 10), which also received 1 x 106 T cells once per week and 20 µg hu14.18 antibody four times per week, and group 4 (n = 7), which also received 1 x 106 T cells and 20 µg hu14.18 antibody four times per week and 20 µg Fc-IL7 once per week. To investigate the effect of antibody alone, we compared control mice (group 1c, n = 6) that received tumor cells only with mice (group 5, n = 7) that received tumor cells and 20 µg hu14.18 antibody four times per week. Control groups were run concurrently with their corresponding treatment groups. Although overall the mice received T cells from several donors (because the number of cells available was limited by the yield from the enrichment procedure), each experimental group received T cells from the same donors. Therapeutic injections were given for five consecutive weeks, and mice were killed on becoming moribund.
Postmortem pathologic examination. Postmortem examination by gross necropsy was done on all animals to ensure xenograft development. Tissue specimens from sites of tumor growth in two representative animals were taken for further histologic examination. Tissues were fixed in 10% buffered formalin, embedded in paraffin wax, sectioned, and routinely stained with H&E.
Cytotoxicity assays. The ability of T cells to exert natural and antibody-dependent cytotoxicity was assessed in a conventional 2-hour europium-TDA release assay (Perkin-Elmer Wallac, Turku, Finland) as described previously (26). In brief, target cells were labeled with a fluorescence-enhancing ligand (BATDA). This hydrophobic ligand quickly penetrates the cell membrane. Within the cell, hydrolysis of ester bonds results in the ligand becoming hydrophilic and therefore unable to pass through the cell membrane. Cytolysis, however, results in release of ligand and reaction of the ligand with the europium to form a stable, fluorescing chelate, which is evaluated fluorometrically. The following formulas were used to calculate spontaneous and specific cytotoxicity:
 |
 |
T cells in different effector-to-target ratios in triplicate, with or without 5 µg/well of hu14.18 antibody. After isolation and before the cytotoxicity assay, the isolated T cells were stimulated with 200 IU/mL human recombinant IL-2 (R&D Systems, Minneapolis, MN) and cultured overnight. We scheduled our cytotoxicity assays for the day after the isolation procedure and coincubated the cells overnight with IL-2 to allow the cells recovery time, because during the remainder of the 36-hour processing time (donor leukapheresis, storage overnight, and CliniMACS enrichment procedure), the cells had to be constantly refrigerated.
Viability assays. Using flow cytometry, we evaluated the effect of Fc-IL7 on the survival of T cells in vitro and compared this effect with that of human recombinant IL-2 and IL-7 (R&D Systems). In a 12-well plate, 2 x 106 freshly isolated T cells per well (at a concentration of 1 x 106/mL) were cultured with human recombinant IL-2 (200 IU/mL), human recombinant IL-7 (50 ng/mL), or Fc-IL7 (50 ng/mL), or without cytokine. Cytokine was added every 48 hours. Viability was measured at four time points (after 24, 96, 168, and 240 hours) by staining the cells with allophycocyanin-conjugated Annexin V (Annexin V-APC; BD Biosciences, Mountain View, CA), counterstaining them with propidium iodide (Roche Diagnostics, Penzberg, Germany), and analyzing them on a fluorescence-activated cell sorting Aria flow cytometer (BD Biosciences) according to the manufacturer's protocol. Viable cells are negative for Annexin V-APC and propidium iodide (27).
Statistics. The time interval (in days) from initiation of therapy to death from any cause (disease or killing) was defined as the event-free survival period, which was estimated by using the method of Kaplan and Meier. The probability of survival was calculated by using the log-rank test, using the commercially available Stata software (Stata Corp., College Station, TX). Two-sided P values are presented.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
T cells. We processed an average number of 11.3 x 109 (range, 6-17 x 109) peripheral mononuclear cells through the CliniMACS device. Large-scale enrichment yielded an average of 255 x 106 T cells (range, 68-512 x 106) with a mean purity of 91% (range, 77-98%); the main contaminants were CD14+ cells. The viability of the T cells was consistently >80%. Although the cells were obtained by leukapheresis from peripheral blood, in some donors, an unusually high proportion of the cells were of the TCR V9V1 subtype, with V9V2 cells constituting the remaining fraction. In addition, there was an unusually high level of coexpression of CD8 and CD56.
Cytotoxicity assays. The T cells (purity >94%, from three different individuals) showed natural cytotoxicity against K562 and NB-1691 cells (Fig. 1). Although they showed substantial natural cytotoxicity against the erythroleukemic cell line K562, they showed low cytotoxicity against the NB-1691 cells. However, their cytotoxicity to NB-1691 cells was substantially enhanced by the addition of 5 µg/well of the humanized anti-GD2 antibody hu14.18 to induce ADCC. The antibody alone had no cytotoxic effect on the NB-1691 cells (data not shown).
|
T cells in vitro. The results of the survival assays are shown in Fig. 2. Without cytokine, the viability of the T cells rapidly decreased. Treatment with IL-2, IL-7, or Fc-IL7 kept, on average, 60% to 80% of cells alive over a cultivation period of 10 days. IL-2 treatment seemed to have a slight advantage over IL-7 or Fc-IL7. Due to the small number of samples, we did not perform statistical analysis, but the results clearly show the protective effect of all tested cytokines and support the use of Fc-IL7 in our animal studies.
|
Animal survival. We found no significant difference (P = 0.33) in survival between mice in control group 1a (median survival period, 61.5 days) and mice in group 2, which were treated only with weekly injections of T cells (median survival period, 62 days; Fig. 3A). Treatment with weekly injections of T cells and hu14.18 antibody for 5 weeks (group 3), however, significantly prolonged the survival period (median survival period, 75 days) compared with that of the control mice (group 1b; median survival period, 62 days; P = 0.001). Further enhanced survival (median period, 89 days) resulted from treatment with T cells, hu14.18 antibody, and Fc-IL7 (group 4; P = 0.005). The results are illustrated in Fig. 3B. Comparison of the overall survival probabilities of groups 3 and 4 showed that the addition of Fc-IL7 provided an additional advantage (P = 0.04). To rule out an effect of the antibody alone, we compared control group 1c with mice that received only antibody for 5 weeks (group 5; Fig. 3C). There was no statistically significant difference between these groups (median survival periods, 67.5 and 65 days, respectively; P = 0.26).
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
T cells combined with a humanized anti-GD2 antibody and the novel fusion cytokine Fc-IL7.
T cells exert MHC-unrestricted natural cytotoxicity against a variety of tumors in vitro. However, their number in the peripheral blood of potential donors is low, and enrichment techniques have to be applied to yield enough cells suitable for immunotherapy. Although ex vivo expansion is possible, this approach is time consuming and harbors the risk of microbial contamination. We therefore developed a method to enrich the population of T cells in leukapheresis products by using a clinical-grade closed system (21). The cytotoxic capabilities against neuroblastoma were illustrated by a standard europium-release assay using K562 (erythroleukemia) and NB-1691 (neuroblastoma) cells as targets (Fig. 1). The T cells' natural cytotoxicity to K562 cells was inherently high, whereas their spontaneous cytotoxicity to the neuroblastoma cells was very low. In other experiments with T cells from different donors, we and others have published before that the natural cytotoxicity of these cells to NB-1691 cells and other neuroblastoma cell lines is, on average, somewhat higher, and this difference probably reflects normal interindividual variability (21, 28). Despite the relatively low inherent cytotoxicity of these cells to the neuroblastoma cells, their cytolytic capacity was enormously enhanced by the addition of the humanized anti-GD2 antibody hu14.18 to induce ADCC. That is, at an effector ( T cells) to target (NB-1691 cells) ratio of 20:1, the cytolytic capacity of the T cells increased from 6% to 78%. This finding illustrates the remarkable capacity of T cells to up-regulate the production of cytotoxic factors when properly stimulated. Their natural cytotoxicity to neuroblastoma cells could also be somewhat increased (up to 15-30% at an effector to target ratio of 20:1) by prestimulating them with IL-2 for several days (data not shown). However, we did not pursue this strategy because our goal was to maintain a clinically relevant strategy involving the transfusion of freshly isolated cells with no further external manipulation.
Once effector cells are transfused into a recipient in a cytotherapy regimen, it will be important to support their vitality, thus preserving their cytotoxic capacities for as long as possible. Even more crucial might be providing the host with immunomodulatory cytokines in the early days and weeks after stem cell transplantation, when certain cytokine levels might still be low due to the lack of cytokine producers. A key factor in promoting T-cell survival is the cytokine IL-7, which also plays a crucial role in T-cell homeostasis (29). The signaling pathways induced by the binding of IL-7 to its receptor (30, 31) regulate and control the development and the differentiation of T cells. Moreover, by activating several pathways, IL-7 binding increases the expression of genes that promote lymphocyte survival and proliferation while down-regulating genes responsible for apoptosis and cell cycle arrest (32). These facts seem to hold true also for the T-cell subset (32, 33).
We have shown that the novel fusion protein Fc-IL7 has properties similar to those of IL-2 in vitro (Fig. 2). After 240 hours in culture, an average of 64% of T cells treated with Fc-IL7 remained alive, in contrast to only 8% of nontreated cells. We tested our therapeutic concept in a mouse model of disseminated neuroblastoma. Although the metastatic pathway in these animals is probably influenced by the route of administration of the tumor cells (in our case, tail vein injection), the metastatic distribution we observed in these mice resembles, with some exceptions, that seen in childhood neuroblastoma (34). Treatment of the animals solely with T cells did not prolong their survival relative to that of the control group. Perhaps this is not surprising, given that our in vitro experiments had shown that the natural cytotoxicity of our isolated T cells to NB-1691 cells was much lower than that shown to the erythroleukemic cell line K562. However, the cytotoxicity of the T cells was dramatically increased by initiating ADCC by using the humanized anti-GD2 antibody hu14.18. The efficacy of this antibody and its predecessor, the chimeric anti-GD2 antibody ch14.18, has been investigated in numerous animal and, in the case of ch14.18, clinical phase I/II studies, either alone or as a fusion protein attached to a cytokine such as IL-2. This strategy proved efficacious in the treatment of GD2-positive tumors like neuroblastoma and subsets of melanoma (11, 12, 35–37). Our study also showed therapeutic benefit to the addition of hu14.18 antibody and Fc-IL7 to the treatment with T cells. Significantly increased survival periods resulted from the addition of hu14.18 antibody alone (P = 0.001) and from the addition of hu14.18 antibody and Fc-IL7 (P = 0.005) to the treatment with T cells. Moreover, use of Fc-IL7 offered an additional advantage over the use of antibody with the T cells, albeit at marginal statistical significance (P = 0.04). Note, however, that the number of animals in this experimental group was relatively small.
We ruled out a direct cytotoxic effect of hu14.18 antibody or any unanticipated activation of the mouse's remaining immune system (natural killer cells, granulocytes, complement, etc.) by treating a set of animals with the antibody alone. Compared with a control group that received no treatment, treatment with antibody alone had no influence on survival.
Our T-cell preparation contained, on average, 7% CD14+ contaminants (data not shown). Upon proper ex vivo stimulation, these cells can further differentiate into macrophages and macrophage-like cells, which also can promote ADCC (38). However, it is very unlikely that our results were influenced by this population of cells, because the number of contaminants was small and similar in each group and we did not subject the cell preparation to a pretreatment that would have stimulated the differentiation of the CD14+ cells.
Granulocyte colony-stimulating factor stimulation leads to an increased harvest of leukocytes and, consequently, of T cells from a single donor. Granulocyte colony-stimulating factor stimulation seems not to diminish the cytotoxic capabilities of the harvested cells. However, because we did not have access to T cells from each donor before granulocyte colony-stimulating factor stimulation, we were not able to address this systematically.
Compared with other studies that used adoptively transferred cytotoxic effector cells in mouse tumor models, we injected a relatively small amount of cells but gave treatment injections weekly for 5 weeks. However, we also injected only a small number (1 x 106) of human neuroblastoma cells into the animals, via the tail vein, and we allowed 5 to 6 days for the tumor cells to disseminate before starting treatment. This use of a small amount of tumor cells, which, nevertheless, caused systemic and bulky disease in the control mice, probably more closely reflects the MRD condition. By using our large-scale method of enrichment of the T cell population, it would be easy to repeatedly obtain and transfuse supraphysiologic numbers of T cells into patients. It is noteworthy that the administration of T cells, hu14.18 antibody, and Fc-IL7 was well tolerated by all animals. Moreover, no animal became moribund during treatment. These observations strongly argue for an extended period of maintenance treatment. Our procedure can easily be transferred to the clinic, because all reagents are available at good manufacturing practices quality, and the isolation procedure is already well established for the clinical grade enrichment of CD34+ and CD133+ stem cell populations. Our T-cell product contained a maximum of 1.5% remaining ß T cells (data not shown). This number is small enough to allow the dose of T cells to be easily adjusted to avoid a graft-versus-host–like reaction. Alternatively, the leukapheresis product could be depleted of ß T cells before being enriched with T cells.
Recently, several studies have shown efficient in vivo stimulation of T cells in cancer patients by systemic administration of aminobisphosphonates, thereby improving the antitumor effect and proliferative capacity of the T cells (39, 40). Including such an agent in our therapeutic strategy might further augment the life-prolonging effect we observed and, in our opinion, is worth considering in the future.
Finally, we conclude that a therapy regimen using clinical-scale enrichment of the human T-cell population from leukapheresis product and adoptive transfer of the T cells with a humanized anti-GD2 antibody and the novel fusion cytokine Fc-IL7 can significantly enhance survival in a mouse model of human disseminated neuroblastoma. Our regimen should be clinically evaluated as a new element in the treatment of neuroblastoma, especially in the context of low tumor burden or MRD, as is encountered directly after stem cell transplantation.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
4 Unpublished data provided by the manufacturer. 
Received 6/ 2/05; revised 8/26/05; accepted 9/ 3/05.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
