Expression of co-stimulatory molecules by tumor cells decreases tumorigenicity but may also reduce systemic antitumor immunity

Chimeric PD-1:28 Receptor Upgrades Low-Avidity T cells and Restores Effector Function of Tumor-Infiltrating Lymphocytes for Adoptive Cell Therapy

Inherent intermediate- to low-affinity T-cell receptors (TCR) that develop during the natural course of immune responses may not allow sufficient activation for tumor elimination, making the majority of T cells suboptimal for adoptive T-cell therapy (ATT). TCR affinity enhancement has been implemented to provide stronger T-cell activity but carries the risk of creating undesired cross-reactivity leading to potential serious adverse effects in clinical application. We demonstrate here that engineering of low-avidity T cells recognizing a naturally processed and presented tumor-associated antigen with a chimeric PD-1:28 receptor increases effector function to levels seen with high-avidity T cells of identical specificity. Upgrading the function of low-avidity T cells without changing the TCR affinity will allow a large arsenal of low-avidity T cells previously thought to be therapeutically inefficient to be considered for ATT. PD-1:28 engineering reinstated Th1 function in tumor-infiltrating lymphocytes that had been functionally disabled in the human renal cell carcinoma environment without unleashing undesired Th2 cytokines or IL10. Involved mechanisms may be correlated to restoration of ERK and AKT signaling pathways. In mouse tumor models of ATT, PD-1:28 engineering enabled low-avidity T cells to proliferate stronger and prevented PD-L1 upregulation and Th2 polarization in the tumor milieu. Engineered T cells combined with checkpoint blockade secreted significantly more IFNγ compared with T cells without PD-1:28, suggesting a beneficial combination with checkpoint blockade therapy or other therapeutic strategies. Altogether, the supportive effects of PD-1:28 engineering on T-cell function make it an attractive tool for ATT. Cancer Res; 77(13); 3577-90. ©2017 AACR.

Combination viroimmunotherapy with checkpoint inhibition to treat glioma, based on location-specific tumor profiling

BACKGROUND

Systemic delivery of a complementary cDNA library expressed from the vesicular stomatitis virus (VSV) treats tumors by vaccinating against a wide range of tumor associated antigens (TAAs). For subcutaneous B16 melanomas, therapy was achieved using a specific combination of self-TAAs (neuroblastoma-Ras, cytochrome c, and tyrosinase-related protein 1) expressed from VSV. However, for intracranial B16 tumors, a different combination was therapeutic (consisting of VSV-expressed hypoxia-inducible factor [HIF]-2α, Sox-10, c-Myc, and tyrosinase-related protein 1). Therefore, we tested the hypothesis that tumors of different histological types growing in the brain share a common immunogenic signature which can be exploited for immunotherapy.

METHODS

Syngeneic tumors, including GL261 gliomas, in the brains of immune competent mice were analyzed for their antigenic profiles or were treated with systemic viroimmunotherapy.

RESULTS

Several different histological types of tumors growing intracranially, as well as freshly resected human brain tumor explants, expressed a HIF-2α(Hi) phenotype imposed by brain-derived CD11b+ cells. This location-specific antigen expression was exploited therapeutically against intracranial GL261 gliomas using systemically delivered VSV expressing HIF-2α, Sox-10, and c-Myc. Viroimmunotherapy was enhanced by immune checkpoint inhibitors, associated with the de-repression of antitumor T-helper cell type 1 (Th1) interferon-γ and Th17 T cell responses.

CONCLUSIONS

Since different tumor types growing in the same location in the brain share a location-specific phenotype, we suggest that antigen-specific immunotherapies should be based upon expression of both histological type-specific tumor antigens and location-specific antigens. Our findings support clinical application of VSV-TAA therapy with checkpoint inhibition for aggressive brain tumors and highlight the importance of the intracranial microenvironment in sculpting a location-specific profile of tumor antigen expression.

Phase I study of the intratumoral administration of recombinant canarypox viruses expressing B7.1 and interleukin 12 in patients with metastatic melanoma

The objective of this study was to evaluate the safety and activity of the intratumoral administration of the immune costimulatory molecule, B7.1, encoded by a vector derived from the canarypox virus, ALVAC (ALVAC-B7.1), alone and with the intratumoral injection of ALVAC encoding the immune-stimulatory cytokine, interleukin 12 (ALVAC-IL-12). Fourteen patients with metastatic melanoma who had s.c. nodules received intratumoral injections on days 1, 4, 8, and 11. Nine patients were given escalating doses of up to 25 x 10(8) plaque-forming units of ALVAC-B7.1. Five patients were given 25 x 10(8) plaque-forming units of ALVAC-B7.1 combined with ALVAC-IL-12 50% tissue culture infective dose of 2 x 10(6). Toxicity was mild to moderate and consisted of inflammatory reactions at the injection site and fever, chills, myalgia, and fatigue. Higher levels of B7.1 mRNA were observed in ALVAC-B7.1-injected tumors compared with saline-injected control tumors. Higher levels of intratumoral vascular endothelial growth factor and IL-10, cytokines with immune suppressive activities, were also observed in ALVAC-B7.1- and ALVAC-IL-12-injected tumors compared with saline-injected controls. Serum levels of vascular endothelial growth factor increased at day 18 and returned to baseline at day 43. All patients developed antibody to ALVAC. Intratumoral IL-12 and IFN-gamma mRNA decreased. Changes in serum IL-12 and IFN-gamma levels were not observed. Tumor regressions were not observed. The intratumoral injections of ALVAC-B7.1 and ALVAC-IL-12 were well tolerated at these dose levels and at this schedule and resulted in measurable biological response. This response included the production of factors that may suppress the antitumor immunologic activity of these vectors.

The role of B7-2 (CD86) in tumour immunity

Tumour cells engineered to express co-stimulatory molecules on their surface provide researchers with powerful new tools to manipulate antitumour responses. It has been demonstrated that B7-1+ and B7-2+ tumour cells can elicit effective responses against their wild-type counterparts. This response is primarily mediated by CD8+ cytolytic T-lymphocytes. The co-stimulatory ability of B7-2+ tumour cells is comparable to that of B7-1+ tumour cells, though with some exceptions. However, on host antigen-presenting cells (APC), B7-2 plays a dominant role in inducing T-cell-mediated immune responses. Up-regulation of B7-2 on host APC may, therefore, present an effective means of generating potent antitumour immunity.

Vaccine Therapy for Murine Glioma Using Tumor Cells Genetically Modified to Express B7.1

OBJECTIVE

In a syngeneic mouse brain tumor model, we tested the hypothesis that vaccination with tumor cells genetically modified to express B7.1 molecules induces tumor-specific T cells and immunological antitumor effects.

METHODS

Malignant glioma cells (RSV-MG) derived from a C3H/He mouse induced by Schmidt-Ruppin Rous sarcoma virus (RSV) were infected with an adenovirus encoding the B7.1 gene (AdB7). To investigate the effects of B7.1 expression on the tumorigenicity of RSV-MG cells, infected cells were implanted subcutaneously into C3H/He mice. The C3H/He mice were vaccinated with AdB7 transfectants injected subcutaneously and 2 weeks later were challenged intracerebrally with wild-type RSV-MG cells to determine whether or not the expression of B7.1 would enhance the immunogenicity of RSV-MG cells.

RESULTS

Immunocytochemistry confirmed the expression of B7.1 and major histocompatibility complex Class I antigen on the infected cells. The growth of subcutaneous tumors was markedly retarded in the AdB7 group, whereas tumors had formed and progressively increased in size in the other control groups. In the vaccine experiments, the mice immunized with AdB7 transfectants survived longer than did the mice of the other groups, and a significant difference in survival times was noted. Immunocytochemistry revealed that brain tumors in mice previously vaccinated with AdB7 infectants had been infiltrated by a larger number of CD3(+) lymphocytes and that these CD3(+) lymphocytes contained not only CD4(+) and CD8(+) T cells but also CD25(+)-activated T cells. In addition, a cytotoxicity assay confirmed that vaccination with the AdB7 transfectants induced tumor-specific cytotoxicity.

CONCLUSION

These results demonstrate the therapeutic potential of vaccination with tumor cells expressing B7.1 for the treatment of malignant glioma.

Adenovirus-mediated gene transfer of B7.1 induces immunological anti-tumor effects in a murine brain tumor

The purpose of the present study was to determine if adenovirus-mediated transfection of a syngeneic mouse brain tumor with the gene encoding B7.1 enhances immunogenicity against tumor. Malignant astrocytoma cells were transfected with adenoviral vectors carrying the B7.1 gene (AdB7). Immunocytochemical analysis confirmed the expression of B7.1 in vitro and in vivo. To investigate the effects of B7.1 expression on tumorigenicity of the malignant astrocytoma, mice were implanted intracerebrally with B7.1-transfected glioma cells. There was no significant difference in proliferation between B7.1-transfected cells and controls in vitro. Nevertheless, mice implanted with B7. 1-transfected cells survived significantly longer than those in the control groups. Immunocytochemical analysis of the tumors showed that there was infiltration of a number of CD8+ T-cells and CD25+ activated T-cells in the brain implanted with B7.1-transfected glioma cells. The results showed the possibility that adenovirus-mediated B7.1 gene transfection to a brain tumor induced activation of CD8+ cytotoxic T-lymphocytes.

Immunotherapy for murine K1735 melanoma: combinatorial use of recombinant adenovirus expressing CD40L and other immunomodulators

We have constructed and tested five recombinant adenoviruses (Ads) that express a variety of immunomodulators, including CD40 ligand (CD40L), a potent costimulator of several components of the immune system. We demonstrate that CD40L expressed from Ad in K1735 mouse melanoma cells leads to a strong reduction in tumorigenicity and to efficient protective immunity in a vaccination setting. Subsequently, using a therapeutic approach, we found that local, intratumoral coinjection of CD40L- and IL-2-expressing Ads was superior to any other agents tested and resulted in an at least 1.9-fold increase in mean survival time, in contrast to systemic application of recombinant CD40L or GM-CSF proteins, which had no significant effects. When using vaccination as a therapeutic approach, the combinations of CD40L plus IL-2 or GM-CSF plus IL-2 from Ad gave rise to an extended (2.8-fold) increase in mean survival time. A detailed analysis of immune cells present within regressing tumors indicated that mainly CD4(+) and CD8(+) T cells, and to a lesser extent dendritic cells, infiltrated the tumor mass, but not NK cells, macrophages, or granulocytes. These results propose that a combination of CD40L plus IL-2 has an improved efficacy over the use of single agents when applied for direct in situ therapy or vaccination therapy.

Antigenic and immunologic characterization of an allogeneic colon carcinoma vaccine

We report the immunological characterization of three colon carcinoma cell lines, COLO 205, SW620 and SW403, which we selected to combine with cytokine-secreting fibroblasts for the development of an allogeneic tumour cell vaccine. The cell lines expressed HLA-A2 as well as shared tumour-associated antigens (TAAs) representative of colon carcinomas: CEA, Ep-CAM, MUC1, HER2/neu and MAGE antigens. They did not secrete high levels of the immunosuppressive factors TGF-beta, IL-10 or prostaglandins. The lines presented TAAs in a manner recognized by immune effector cells, which was demonstrated by the lysis of SW620 by HLA-A2-restricted anti-p53 cytotoxic T lymphocytes (CTL). COLO 205 and SW620 were genetically modified to express the co-stimulatory molecule CD80 (B7.1), which increased the ability of the cells to stimulate CTL in vitro. CTL clones derived from HLA-A2+ peripheral blood mononuclear cells stimulated with the CD80-expressing lines lysed the stimulator cell and an HLA-A2+ colon cancer cell line, but did not lyse an isogeneic fibroblast line or an HLA-A2- colon cancer cell line. CTL clones derived from colon carcinoma patients immunized with an allogeneic vaccine containing these lines demonstrated killing of autologous tumour cells, the vaccine cell lines and other HLA-A2+ colon cancer cell lines, but not fibroblasts isogeneic to certain of the target cell lines. Our studies demonstrate that these colon carcinoma cell lines express shared TAAs that can induce CTLs which recognize and lyse other colon carcinoma cells, and support the continued clinical evaluation of the CD80 gene modified allogeneic colon cell/cytokine-secreting fibroblast carcinoma vaccine.

Local costimulation reinvigorates tumor-specific cytolytic T lymphocytes for experimental therapy in mice with large tumor burdens

Cytotoxic T cells recognize tumor Ags and destroy cancer cells in vitro. Adoptive transfer studies with transgenic T cells specific for tumor Ags have demonstrated that CTL are effective only in mice with small tumor burdens and thus appear to have limited potential in cancer immunotherapy. Here we used transgenic mice that express the TCR specific for an unmutated tumor Ag P1A and multiple lineages of P1A-expressing tumors to address this critical issue. We found that local costimulation, either by expression of B7-1 on the tumor cells or by local administration of anti-CD28 mAb 37N, reinvigorated the function of CTL specific for the tumor Ag, as it substantially increased the efficacy of CTL therapy for mice with large tumor burdens. Our study suggests that CTL-based immunotherapy can be manipulated to deal with large tumors.

Direct retroviral delivery of human cytochrome P450 2B6 for gene-directed enzyme prodrug therapy of cancer

Human cytochrome P450 2B6 (CYP2B6) metabolizes the prodrug cyclophosphamide (CPA) to produce phosphoramide mustard that cross-links DNA leading to cell death. We have constructed a novel retroviral vector encoding CYP2B6 (designated "MetXia-P450") and used it to transduce the human tumor cell lines HT29 and T47D. MetXia-P450 transduction sensitised these cells to the cytotoxic effects of the prodrug CPA. Results from in vitro experiments demonstrated adverse effects on the clonogenic survival of cyclophosphamide-treated cells transduced with MetXia-P450. Cytotoxic activity accompanied by bystander effect was particularly evident in 3-D multicellular spheroid models suggesting that this in vitro system may be a more appropriate model for assessing the efficacy of gene directed-enzyme prodrug therapy (GDEPT). We have applied this approach in a clinically relevant gene therapy protocol on established subcutaneous tumor xenografts. These studies show for the first time the efficacy of a P450-based GDEPT strategy mediated by a direct retroviral gene transfer in vivo.

B7.1 expression by the weakly immunogenic F98 rat glioma does not enhance immunogenicity

Enhanced immunogenicity has been reported following transfection of a variety of immunogenic tumors with the B7.1 co-stimulatory molecule. The purpose of the present study was to determine if transfection of a weakly immunogenic rat brain tumor, the F98 glioma, with the gene encoding B7.1 could enhance its immunogenicity. F98 cells were transfected with a plasmid containing the B7.1 gene, and stable transfectants (F98/B7.1) were obtained. Flow cytometric analysis confirmed the expression of B7.1 and MHC class I antigens on the cell surface. To investigate the effects of B7.1 expression on the tumorigenicity of the F98 glioma, Fischer rats were implanted intracerebrally with either F98 (wild-type) or F98/B7.1 transfected cells. No significant differences in survival times were noted. Mean survival times of 21.8 and 24.0 days were observed for the respective groups at a challenge dose of 103 cells. These differences in survival time were not significant. To determine if expression of B7.1 enhanced the immunogenicity of the F98 glioma, rats were vaccinated weekly for 3 weeks with 107 mitomycin C-treated F98 or F98/B7.1 cells injected subcutaneously and then challenged intracerebrally with F98 cells 1 week later. Unvaccinated animals or those that received wild-type F98 cells as a vaccine had a survival time (mean +/- s.d.) of 22.3 +/- 1.5 days following tumor challenge versus 20.0 +/- 1.7 days for rats that had been vaccinated with F98/B7.1. Although we recognize that it might be possible to design more effective vaccination regimes, nevertheless, our data indicate that transfection of the B7.1 gene into the F98 rat glioma did not enhance its immunogenicity, and that other approaches will be required.

Gene transfer to liver cancer cells of B7-1 plus interleukin 12 changes immunoeffector mechanisms and suppresses helper T cell type 1 cytokine production induced by interleukin 12 alone

To investigate the cooperative effect of B7-1 and IL-12 in the induction of antitumor activity, we have developed retroviral vectors encoding human B7-1, murine IL-12, or both B7-1 and IL-12 coordinately. Murine transformed liver cells (BNL) were engineered to stably express B7-1, IL-12, or both by infection with corresponding retroviruses. No tumor was observed in 20, 75, and 95% of mice receiving, respectively, B7-1-, IL-12-, and B7-1/IL-12-modified tumor cells after 250 days of inoculation. In contrast, injection of parental BNL or BNL/Neo cells resulted in lethal tumor progression in all mice. Protection against rechallenge with parental tumor cells was observed only in mice who had rejected BNL/IL-12, but not in animals that rejected BNL/B7-1 or BNL/B7-1-IL-12. Growth of parental tumor cells was significantly delayed by simultaneous injection in a distant site of irradiated tumor cells engineered to express IL-12 or both B7-1 and IL-12 but not B7-1 alone. BNL/B7-1 and BNL/B7-1-IL-12 showed similar efficacy in these experiments. Antitumor immunity induced by B7, with or without IL-12, was found to depend mainly on CD4+ T cells with a minor contribution of a non-T cell mechanism; whereas the effect of IL-12 was dependent on CD8+ T cells and on non-T cell effectors. Immunization of mice with IL-12-modified BNL cells induced secretion of a Thl pattern of cytokines while immunization with cells expressing both IL-12 and B7-1 resulted in inhibition of IFN-gamma production. Immunization with BNL/B7-1-IL-12 cells in the presence of anti-human B7-1 MAb resulted in restoration of IFN-gamma production to the levels found in animals injected with BNL/IL-12 cells. To summarize, in our model coexpression of B7-1 and IL-12 in tumor cells does not result in improved antitumoral activity as compared with expression of IL-12 alone. This may be related to the fact that B7-1 changes the mechanisms of antitumor immunity and inhibits IFN-gamma production induced by IL-12 in vivo.

Progress in cancer gene therapy

Many technical difficulties have to be overcome before effective gene therapy can be achieved. Strategies for gene therapy include 'suicide' gene therapy, transfer of a tumor suppressor gene, inhibition of activated oncogenes by antisense mechanisms, and cytokine gene transfer and tumor cell vaccination. Gene therapy will have a major impact on the healthcare of our population only when vectors are developed that can safely and efficiently be injected directly into patients as drugs. One of the most promising areas of vector development is that of non-viral vectors, which consist of liposomes, molecular conjugates, and naked DNA delivered by mechanical methods. Future research should be focused on modifying viral vectors to reduce toxicity and immunogenicity, increasing the transduction efficiency of non-viral vectors, enhancing vector targeting and specificity, regulating gene expression, and identifying synergies between gene-based agents and other cancer therapeutics. The evaluation of gene therapy combinations is another important area for future research. The identification of tumor rejection antigens from a variety of cancers and the immune response that is defective in cancer patients are important topics for future studies. A universal gene delivery system has yet to be identified, but the further optimization of each of these vectors should result in each having a unique application. Gene therapy has still a long way to go and requires the efforts of investigators in the basic and clinical sciences. Despite substantial progress, a number of key technical issues need to be resolved before gene therapy can be effectively applied in the clinic.

Combination Gene Therapy with CD86 and the MHC Class II Transactivator in the Control of Lung Tumor Growth

Early reports suggest that the costimulatory molecule CD86 (B7-2) has sporadic efficacy in tumor immunity, whereas changes in cancer immunity mediated by the MHC class II transactivator (CIITA) have not been extensively investigated. CIITA activates MHC class II expression in most cells; however, in the Line 1 lung carcinoma model system, CIITA activates MHC class I and well as class II. Here we show that CD86 is very effective in inducing a primary immune response against Line 1. Tumor cells expressing CD86 grew in only 50% of the mice injected with live cells, and those mice that developed tumors did so with significantly delayed kinetics. Furthermore, irradiated CD86-expressing Line 1 cells served as an effective tumor vaccine, demonstrating that CD86 is effective in inducing tumor immunity in the Line 1 system. These data suggest that if CIITA and CD86 cooperate, enhanced tumor immunity could be achieved. CIITA alone was mildly beneficial in slowing primary tumor growth but only when expressed at low levels. Clones expressing high levels of class II MHC grew as fast as or faster than parental tumor, and CIITA expression in a tumor vaccine assay lacked efficacy. When CIITA and CD86 were coexpressed, there was no cooperative immune protection from tumor growth. Cells that coexpress both genes also failed as a cancer vaccine, suggesting a negative role for CIITA in this lung carcinoma. These data suggest that human cancer vaccine trials utilizing CIITA gene therapy alone or in combination with CD86 should be approached with caution.

Enhanced interleukin-2 gene transfer immunotherapy of breast cancer by coexpression of B7-1 and B7-2

The capability of B7-1 to augment the antitumor activity of some cytokines has been shown primarily for such cytokines as interleukin-12 (IL-12), IL-7, and to a lesser extent IL-2. In this study, we investigate the ability of B7-1 and B7-2 to augment the antitumor activity of IL-2. Considering the affinity of both molecules for CD28 (T cell receptor for B7-1 and B7-2), we postulated that their potential to augment IL-2 antitumor activity would be similar. Two murine transgenic adenocarcinoma models were chosen to investigate the activity of adenoviral vectors constructed to express either B7-1 and IL-2 or B7-2 and IL-2. Before administering the vector intratumorally to tumor-bearing mice, we determined the expression of B7-1, B7-2, MHC I, and MHC II on these tumor cells and demonstrated positive expression of only MHC I. Intratumoral injection of the vector expressing B7-1 and IL-2 resulted in complete regression of all tumors treated. In contrast, the vector expressing B7-2 and IL-2 was significantly less effective at regressing PyMT tumors, whereas both double recombinant vectors demonstrated similar levels of complete regression in the Neu (NDL 8142) model. Regressed mice were all protected for rechallenge in both models and demonstrated antigen-specific cytotoxic T lymphocytes (CTL) in the PyMT model. These findings indicate that the combination of IL-2 with B7-1 or B7-2 significantly enhances the antitumor activity of IL-2.

Expression of B7 co-stimulatory molecules by B16 melanoma results in a natural killer cell-dependent local anti-tumour response, but induces T-cell-dependent systemic immunity only against B7-expressing tumours

In an attempt to enhance the anti-tumour immune response, the co-stimulatory molecules B7-1 or B7-2 were expressed on the surface of B16 melanoma cells. B7-expressing tumours grew more slowly in both syngeneic immunocompetent mice and athymic T cell-immunodeficient nude mice. The delay in growth of B7-expressing tumours was dependent on natural killer (NK) cells, as reductions in tumour growth rates were minimized in mice depleted of NK cells. Systemic immunity to B16 melanoma was examined by vaccination with irradiated tumour cells. Inoculation with irradiated B16 B7-1 cells failed to protect against a subsequent challenge with live parental B16 cells, but conferred partial protection against challenge with live B16 B7-1 cells. In contrast to the local anti-tumour reaction, this protective response was dependent on T cells. The results presented here reveal some of the mechanisms involved in the in vivo response to a poorly immunogenic tumour modified to express co-stimulatory molecules.

Gene therapy for melanoma in humans

As melanoma evolves, it interacts with the immune system. Based on this immunobiology, there are now a number of rationally designed attempts to develop genetically modified melanoma vaccines. This article outlines immunologic and other strategies in gene therapy for melanoma and provides an overview of current clinical trials.

Lack of correlation between rejection of tumor cells co-expressing interleukin-2 and B7.1 and vaccine efficiency

Genetically modifying tumor cells to express a variety of cytokines such as interleukin-2 (IL-2) or the co-stimulatory molecule B7.1 leads to increased immunogenicity and reduced tumorigenicity of tumors in several models with T cells involved in the process. We have previously reported decreased tumorigenicity of the murine plasmacytoma J558L [major histocompatibility complex (MHC) class I+ and class II-] expressing IL-2 or B7.1. When systemic immunity was analyzed, immunization with either J558-IL2 or J558-B7.1 cells generated moderate protection against unmodified J558L tumor cells, comparable to immunization with a tumor cells/adjuvant Corynebacterium parvum mixture. In this study, we asked whether the co-expression of IL-2 and B7.1 in tumor cells would augment vaccine potency, cytotoxic T lymphocyte (CTL) activity and protective immunity. Rejection of single IL-2 or B7.1 or co-transfected IL-2/B7.1 cells occurred in most syngeneic animals but not in T cell-deficient nude mice, thus confirming that T cells were required for tumor rejection. We knew from previous experiments that CD8+ T cells were responsible for rejection. Surprisingly, immunization with J558-IL2/B7.1 cells followed by challenge with parental J558L caused a reduction in systemic protection as compared to J558-B7.1 or J558-IL2 alone. We examined the mechanism underlying this unexpected result: 6 days after injection of J558-IL2/B7.1 cells, tumor were nearly completely destroyed and were almost devoid of CD8+ cells, while CD8+ cells were increased in both IL-2- and B7.1-transfected tumors. In addition, immunization with J558-IL2/B7.1 tumors had an adverse effect on the generation of CTL. Mice immunized with J558-B7.1 and to a lesser extent J558-IL2 cells mounted a CTL response against J558L cells while, in contrast, no CTL activity could be detected in mice immunized with J558-IL2/B7.1, thus showing a correlation between the absence of CTL activity and the lack of in vivo protection. We demonstrate that "hyperstimulation" of the immune response by genetically modified cancer vaccines can have adverse effects on tumor immunity, even though the mechanism is not yet completely understood.

Granulocyte-macrophage colony-stimulating factor and B7-2 combination immunogene therapy in an allogeneic Hu-PBL-SCID/beige mouse-human glioblastoma multiforme model

Glioblastoma multiforme is the most common primary central nervous system neoplasm. Its dismal prognosis has led to investigation of new treatment strategies such as immunogene therapy. We transduced the human glioblastoma cell line D54MG in vitro with genes encoding the proinflammatory cytokine granulocyte-macrophage colony-stimulating factor (GM-CSF), the T cell co-stimulatory molecule B7-2, or both (in a bicistronic vector) via retroviral vectors. Therapeutic gene expression by D54MG was high after transduction and selection (30 ng/10(6) cells/day for GM-CSF and > 2 orders of magnitude fluorescence shift on flow cytometry for B7-2). The effect of GM-CSF and/or B7-2 transduction on D54MG tumor growth in vivo was monitored in a novel allogeneic human peripheral blood lymphocyte-severe combined immunodeficiency mouse (Hu-PBL-SCID) model. GM-CSF- or B7-2-transduced tumors showed growth suppression in hu-PBL-reconstituted mice compared to untransduced and/or unreconstituted controls. Growth suppression was greatest for B7-2. Furthermore, vaccination with irradiated GM-CSF/B7-2-transduced tumor cells markedly inhibited growth of wild-type tumors at distant sites. Thus, this study illustrates a potential gene therapy strategy for glioblastoma multiforme patients using GM-CSF and/or B7-2 transduced tumor vaccines. Although extension of these allogeneic studies to an autologous system is critical, this is the first demonstration of in vivo efficacy of combination GM-CSF and B7-2 immunogene therapy for human glioblastoma multiforme.