Autologous attenuated T-cell vaccine (Tovaxin) dose escalation in multiple sclerosis relapsing-remitting and secondary progressive patients nonresponsive to approved immunomodulatory therapies.

An open-label dose escalation study of T-cell vaccination in multiple sclerosis patients was conducted using attenuated myelin reactive T-cells (MRTC) selected with six myelin peptides, two each from MBP, PLP and MOG. The dose range of subcutaneous injections given at weeks 0, 4, 12 and 20 was 6-9E6, 30-45E6 and 60-90E6 irradiated MRTC. Assessments were over 52 weeks for MRTC levels, EDSS, MSIS-29, brain MRI and relapses. The 30-45E6 dose was the most effective with reductions in MRTC ranging from 92.4% at week 5 to 64.8% at week 52. The reduction in relapses compared to baseline for the M-ITT and evaluable per-protocol analyses were 63.5%, and 85.0% at week 52. The MRI lesions were stable while there was an improvement trend in the EDSS and MSIS-29 physical subscore following the second injection. Adverse events were mild to moderate in intensity with mild injection site reactions occurring with increasing dosage. The mid-dose was selected for further clinical development studies because of the rapid depletion of peripheral blood MRTC and a trend for improvements in clinical outcomes following immunization.

Adenovector-induced expression of human-CD40-ligand (hCD40L) by multiple myeloma cells. A model for immunotherapy.

OBJECTIVE: CD40L restores the antigen-presenting cell (APC) function of some B-cell tumors and induces professional APC maturation. We therefore evaluated the effects of transgenic CD40L expression on the behavior and immunogenicity of human multiple myeloma (MM) cells. MATERIALS AND METHODS: CD40L expression was induced in a CD40(+) (RPMI 8226) and a CD40(-) (U266B1) human myeloma cell line (HMCL) by adenoviral vector gene transfer. The viability and proliferative activity of control HMCL and HMCL/CD40L were determined by daily trypan blue staining and methyl-3H-thymidine incorporation. Mixed lymphocyte reaction (MLR) with allogeneic mononuclear cells (MNCs) and coculture of allogeneic dendritic cells (DCs) with HMCL expressing transgenic CD40L were used to evaluate the APC function of modified HMCL as well as the role of bystander DCs in inducing an anti-tumor immune response. RESULTS: CD40L expression significantly inhibited the growth of the CD40(+) HMCL and induced apoptosis. These effects were less evident for the CD40(-) HMCL. There was no upregulation of costimulatory molecules on either HMCL following CD40L expression. Both HMCL expressing transgenic CD40L induced maturation of bystander DCs and enhanced their ability to stimulate the proliferation of MNCs. DCs cultured with the poorly immunogenic RPMI 8226 expressing CD40L upregulated T-lymphocyte release of IFN-gamma and other Th1 cytokines (interleukin-2, tumor necrosis factor-alpha). CONCLUSIONS: Our data suggest that transgenic expression of CD40L exerts a dual effect favoring generation of an immune response to human MM. Where the tumor cells are CD40(+), the engagement of CD40 antigen by CD40L on tumor cells induces their apoptosis, allowing uptake of tumor-associated antigen by professional APC. Independently of tumor-cell expression of CD40, transgenic expression of CD40L on tumor cells allows them to stimulate CD40(+) APC, to increase their maturation and their capacity to stimulate cytotoxic T lymphocytes (CTL) that recognize the tumor-derived antigens the APC may have engulfed.

Transgenic expression of CD40L and interleukin-2 induces an autologous antitumor immune response in patients with non-Hodgkin's lymphoma.

The malignant B cells of non-Hodgkin's lymphoma (B-NHL cells) express peptides derived from tumor-specific antigens such as immunoglobulin idiotypes, and also express major histocompatibility complex antigens. However, they do not express co-stimulatory molecules, which likely contributes to their protection from host antitumor immunity. To stimulate NHL-specific immune responses, we attempted to transfer the human CD40 ligand (hCD40L) gene to B-NHL cells and enhance their co-stimulatory potential. We found that an adenoviral vector encoding human CD40L (AdhCD40L) was ineffective at transducing B-NHL cells because these cells lack the coxsackievirus B-adenovirus receptor and alpha(v) integrins. However, preculture of the B-NHL cells with the human embryonic lung fibroblast line, MRC-5, significantly up-regulated expression of integrin alpha(v)beta 3 and markedly increased their susceptibility to adenoviral vector transduction. After prestimulation, transduction with AdhCD40L increased CD40L expression on B-NHL cells from 1.3+/-0.2% to 40.8+/-11.9%. Transduction of control adenoviral vector had no effect. Expression of transgenic human CD40L on these CD40-positive cells was in turn associated with up-regulation of other co-stimulatory molecules including B7-1/-2. Transduced B-NHL cells were now able to stimulate DNA synthesis of autologous T cells. However, the stimulated T cells were unable to recognize unmodified lymphoma cells, a requirement for an effective tumor vaccine. Based on previous results in an animal model, we determined the effects of combined use of B-NHL cells transduced with AdhCD40L and AdhIL2 vectors. The combination enhanced initial T-cell activation and generated autologous T cells capable of specifically recognizing and killing parental (unmodified) B-NHL cells via major histocompatibility complex--restricted cytotoxic T lymphocytes. These findings suggest that the combination of CD40L and IL2 gene-modified B-NHL cells will induce a cytotoxic immune response in vivo directed against unmodified tumor cells.

Autologous antileukemic immune response induced by chronic lymphocytic leukemia B cells expressing the CD40 ligand and interleukin 2 transgenes.

Although the B cells of chronic lymphocytic leukemia (B-CLL cells) express both tumor-specific peptides and major histocompatibility complex (MHC) class I antigens, they lack the capacity for costimulatory signaling, contributing to their protection against host antitumor immunity. To stimulate CLL-specific immune responses, we sought to transfer the human CD40 ligand (hCD40L) gene to B-CLL cells, using an adenoviral vector, in order to upregulate costimulating factors on these cells. Because efficient gene transduction with adenoviral vectors requires the expression of virus receptors on target cells, including the coxsackievirus B-adenovirus receptors (CAR) and alpha(v) integrins, we cocultured B-CLL cells with human embryonic lung fibroblasts (MRC-5 line). This exposure led to increased expression of integrin alpha(v)beta3 on B-CLL cells, which correlated with higher transduction rates. Using this novel prestimulation system, we transduced B-CLL cells with the hCD40L gene. The Ad-hCD40L-infected cells had higher expression of B7 molecules and induced activation of autologous T cells in vitro, but these T cells could not recognize parental leukemic cells. By contrast, an admixture of Ad-hCD40L-positive cells and leukemic cells transduced with the human interleukin 2 (IL-2) gene produced greater T cell activation than did either immunostimulator population alone. Importantly, this combination generated autologous T cells capable of specifically recognizing parental B-CLL cells. These findings suggest that the combined use of genetically modified CD40L-expressing B-CLL cells in combination with IL-2-expressing B-CLL cells may induce therapeutically significant leukemia-specific immune responses.

Adenovirus-mediated BMP2 expression in human bone marrow stromal cells.

Recombinant adenoviral vectors have been shown to be potential new tools for a variety of musculoskeletal defects. Much emphasis in the field of orthopedic research has been placed on developing systems for the production of bone. This study aims to determine the necessary conditions for sustained production of high levels of active bone morphogenetic protein 2 (BMP2) using a recombinant adenovirus type 5 (Ad5BMP2) capable of eliciting BMP2 synthesis upon infection and to evaluate the consequences for osteoprogenitor cells. The results indicate that high levels (144 ng/ml) of BMP2 can be produced in non-osteoprogenitor cells (A549 cell line) by this method and the resultant protein appears to be three times more biologically active than the recombinant protein. Surprisingly, similar levels of BMP2 expression could not be achieved after transduction with Ad5BMP2 of either human bone marrow stromal cells or the mouse bone marrow stromal cell line W20-17. However, human bone marrow stromal cells cultured with 1 microM dexamethasone for four days, or further stimulated to become osteoblast-like cells with 50 microg/ml ascorbic acid, produced high levels of BMP2 upon Ad5BMP2 infection as compared to the undifferentiated cells. The increased production of BMP2 in adenovirus transduced cells following exposure to 1 microM dexamethasone was reduced if the cells were not given 50 microg/ml ascorbic acid. When bone marrow stromal cells were allowed to become confluent in culture prior to differentiation, BMP2 production in response to Ad5BMP2 infection was lost entirely. Furthermore, the increase in BMP2 synthesis seen during differentiation was greatly decreased when Ad5BMP2 was administered prior to dexamethasone treatment. In short, the efficiency of adenovirus mediated expression of BMP2 in bone marrow stromal cells appears to be dependent on the differentiation state of these cells.

IL-2 adenovector-transduced autologous tumor cells induce antitumor immune responses in patients with neuroblastoma.

In many different murine models, the immunogenicity of tumor cells can be increased by transduction with a range of immunostimulatory genes, inducing an immune response that causes regression of pre-existing unmodified tumor cells. To investigate the relevance of these animal models to pediatric malignancy, we used autologous unirradiated tumor cells transduced with an adenovirus-IL-2 to immunize 10 children with advanced neuroblastoma. In a dose-escalation study, we found that this tumor immunogen induced a moderate local inflammatory response consisting predominantly of CD4(+) T lymphocytes, and a systemic response, with a rise in circulating CD25(+) and DR+ CD3(+) T cells. Patients also made a specific antitumor response, manifest by an IgG antitumor antibody and increased cytotoxic T-cell killing of autologous tumor cells. Clinically, five patients had tumor responses after the tumor immunogen alone (one complete tumor response, one partial response, and three with stable disease). Four of these five patients were shown to have coexisting antitumor cytotoxic activity, as opposed to only one of the patients with nonresponsive disease. These results show a promising correlation between preclinical observations and clinical outcome in this disease, and support further exploration of the approach for malignant diseases of children.

Interleukin-2 gene-modified allogeneic tumor cells for treatment of relapsed neuroblastoma.

Tumor cells that have been genetically modified to express immunostimulatory genes will induce effective antitumor responses in a range of syngeneic animal models. For human applications, transduced autologous tumor cell lines are often difficult or impossible to prepare, so that there are strong incentives for substituting a standardized allogeneic tumor cell line. However, such lines may be inferior immunogens if they differ from host tumors in the antigens they express. We have evaluated the safety, immunostimulatory, and antitumor activity of an interleukin-2-secreting allogeneic neuroblastoma cell line in 12 children with relapsed stage IV neuroblastoma. They received two to four subcutaneous injections of cells in a dose-escalating schedule, up to a maximum of 10(8) cells per injection. There was induration and pruritus at the injection site, and skin biopsies revealed mild panniculitis with CD3+ cells surrounding scanty residual tumor cells. There was a limited but significant peripheral monocytosis. No patient showed any increase in direct cytotoxic effector function against the immunizing cell line, but 3 patients had a rise in the frequency of neuroblastoma-reactive cytotoxic T lymphocyte precursor cells. One child had > 90% tumor response (PR), 7 had stable disease, and 4 had progressive disease in response to vaccine alone. Although these results offer some encouragement for the continued pursuit of allogeneic vaccine strategies in human cancer, the antitumor immune responses we observed are inferior to those obtained in an earlier immunization study using autologous neuroblastoma cells. Hence, we suggest that this earlier approach remains preferable, its difficulties notwithstanding.

A novel herpes vector for the high-efficiency transduction of normal and malignant human hematopoietic cells.

Herpes simplex viruses (HSVs) would offer numerous advantages as vectors for gene transfer, but as yet they have not proved capable of transducing hematopoietic cells. Using a genetically inactivated form of HSV that is restricted to a single cycle of replication (disabled single-cycle virus, [DISC-HSV]), we have transduced normal human hematopoietic progenitor cells and primary leukemia blasts with efficiencies ranging from 80% to 100%, in the absence of growth factors or stromal support. Toxicity was low, with 70% to 100% of cells surviving the transduction process. Peak expression of transferred genes occurred at 24 to 48 hours after transduction with the DISC-HSV vector, declining to near background levels by 14 days. Despite this limitation, sufficient protein is produced by the inserted gene to permit consideration of the vector for applications in which transient expression is adequate. One example is the transfer of immunostimulatory genes, to generate leukemia immunogens. Thus, murine A20 leukemia cells transduced with a DISC-HSV vector encoding granulocyte-macrophage colony-stimulating factor were able to stimulate a potent antitumor response in mice, even against pre-existing leukemia. The exceptional transducing ability of the DISC-HSV vector should therefore facilitate genetic manipulation of normal and malignant human hematopoietic cells for biological and clinical investigation.

Gene marking and gene therapy for transplantation medicine.

The classic application for gene therapy is in the correction of single gene defects, although this has been complicated by the low efficiency of gene transfer into hematopoietic cells. Gene therapy, however, has potential for the modulation of tumor cell growth, drug sensitivity, and antitumor immune responses. In addition, gene marking can be used, in spite of this limited transfer efficiency, to provide information on hematopoiesis, sources of cancer relapse after stem cell transplant, and the relative efficacy of graft manipulation techniques. This article reviews the applications of gene therapy and gene marking in transplantation medicine.

Transfer of marker genes into hemopoietic progenitor cells.

Ex vivo gene marking of normal and malignant hemopoietic cells allows the cells to be subsequently tracked in vivo. Marking has shown that even when marrow is in remission, it may contain malignant cells that contribute to relapse. These studies have also shown that it is possible to obtain long-term gene expression in human long-lived hemopoietic progenitor cells and T lymphocytes in vivo. Current marker studies use two distinguishable vectors to track two distinctively treated cell populations in a single individual. This modification greatly increases the power of the technique. It is now possible to study the effects of purging on residual malignant cells in marrow, to determine the action of growth-promoting agents (such as cytokines and stroma) on short- and long-term repopulation by transduced marrow, and to discover which phenotypic subsets of hemopoietic progenitor cells have long-term repopulating potential. The information gained will be invaluable for improving therapeutic gene transfer protocols in which marrow-derived cells are the targets.

Use of gene marking in bone marrow transplantation.

We have used gene marking to investigate the mechanism of relapse and biology of reconstitution following bone marrow transplantation (BMT). The rationale for our initial protocols was to learn if residual malignant cells in autologous marrow contribute to subsequent relapse. Marked malignant cells were found at the time of relapse in 6/8 patients relapsing after autologous BMT for AML or neuroblastoma showing the infused marrow contributed to disease recurrence. Modifications of this marker approach with two distinguishable vectors are now being used to compare the efficacy of purging techniques. We were also able to evaluate gene transfer to normal progenitors and demonstrated that the marker gene was expressed for up to 36 months. Gene marking is also being used to trace the fate of EBV-specific CTLs that we are administering to recipients of allogeneic BMT and has provided evidence of persistence of adoptively transferred CTL for up to 10 months.

Immunomodulatory effects of human neuroblastoma cells transduced with a retroviral vector encoding interleukin-2.

We have investigated whether retroviral mediated transfer of the IL-2 gene renders human neuroblastoma cells immunogenic, justifying their use in a clinical tumor immunization study. Fourteen neuroblastoma cell lines were established from patients with disseminated neuroblastoma and transduced with the vector G1Ncvl2, which contains the neomycin phosphotransferase gene and the cDNA of the human interleukin-2 gene. Clones secreting > 150 pg/10(6) cells/24 h of IL-2 were selected for further study. Secretion of IL-2 was maintained for at least 3 weeks in nonselective media, implying that production of the cytokine would continue under in vivo conditions. Co-culture of IL-2 transduced cell lines with patient lymphocytes induced potent cytotoxic activity against both transduced and parental neuroblastoma cell lines. This activity was HLA unrestricted, and predominantly mediated by CD16+ or CD56+ and CD8- lymphocytes. These data form the preclinical justification for our current immunization protocol for patients with relapsed or resistant neuroblastoma.

Direct demonstration that autologous bone marrow transplantation for solid tumors can return a multiplicity of tumorigenic cells.

Patients with solid tumors are increasingly being treated by autologous bone marrow transplantation (BMT). Although response rates appear to be increased, disease recurrence is the commonest cause of treatment failure. Whether relapse is entirely due to residual disease in the patient or arises also from infiltrating malignant cells contained in the autologous marrow transplant has not been resolved. If the latter explanation is correct, then purging would be required as part of the transplantation procedure. We used retrovirally mediated transfer of the neomycin-resistance gene to mark BM harvested from eight patients with neuroblastoma in clinical remission. The marked marrow cells were subsequently reinfused as part of an autologous BMT. At relapse, we sought the marker gene in malignant cell populations. Three patients have relapsed, and in each the marker gene was detected by phenotypic and genetic analyses of resurgent malignant cells at medullary and extramedullary sites. Analysis of neuroblast DNA for discrete marker gene integration sites suggested that at least 200 malignant cells, each capable of tumor formation, were introduced with the autologous marrow transplant and contributed to relapse. Thus, autologous BMTs administered to patients with this solid tumor may contain a multiplicity of malignant cells that subsequently contribute to relapse. The marker-gene technique we describe should permit evaluation of the mechanisms of relapse and the efficacy of purging in patients receiving autologous marrow transplantation for other solid tumors that infiltrate the marrow.

Gene marking and autologous bone marrow transplantation.

If residual cancer cells in harvested bone marrow could be marked and subsequently detected in patients at relapse, valuable information would be obtained about the source of recurrent disease after autologous marrow transplantation. If normal progenitor cells were also marked, the study would provide useful data on the susceptibility of these human cells to gene transfer and their capacity to express newly introduced genes. We transferred the neomycin-resistance gene (NeoR) into bone marrow cells harvested from 20 children with acute myeloid leukemia (n = 12) or neuroblastoma (n = 8) in clinical and cytological remission using a retrovirus vector. The cells were then returned to the patients as part of an autologous bone marrow transplantation protocol. Two AML and three neuroblastoma patients have relapsed. In all, the resurgent cells contained the NeoR marker by analysis with PCR. These results prove that so-called remission marrow can contribute to relapse in patients who receive autologous transplants. The gene marking technique is now being used to evaluate techniques of pretransplant purging.

Gene marking to improve the outcome of autologous bone marrow transplantation.

The neomycin resistance gene in a retroviral vector has been used to mark the marrow of patients receiving autologous bone marrow transplantation for neuroblastoma and acute myeloid leukemia. We have shown that the marker gene can be detected in the resurgent malignant cells at the time of relapse, directly demonstrating that tumorigenic cells contaminate "remission" marrow. We have also shown effective gene transfer to normal primitive progenitor cells and demonstrated a long lasting contribution of the infused marrow to hematopoiesis. The gene marking technique is now being used to evaluate the efficacy of purging and the impact of growth factor treatment on long and short term hematopoietic reconstitution.

Gene marking to determine whether autologous marrow infusion restores long-term haemopoiesis in cancer patients.

The contribution of infused bone marrow cells to long-term haemopoietic recovery in patients undergoing autologous bone marrow transplantation is unknown. Such information would help to clarify the role of this procedure in cancer therapy and would aid in the development of strategies to reduce the risk of subsequent aplasia. By transferring a neomycin resistance marker gene into the marrow cells of 20 patients before transplantation, we were able to trace the pattern of haemopoietic reconstitution postinfusion. The marker gene was present and expressed in all haemopoietic lineages in vivo in 15 of 18 evaluable patients at 1 month post-transplantation, in 8 of 9 patients at 6 months, and in 5 of 5 at 1 year. The marker has remained detectable for up to 18 months--the duration of our study. Our findings indicate that harvested bone marrow consistently contributes to long-term multilineage recovery of haemopoiesis after autologous marrow transplantation in cancer patients. These results provide a rationale for the continued exploration of more ablative preparative regimens with single or sequential autologous marrow transplants.