Oral Abstracts

Synergistic anti-tumor responses against metastatic cancer using antibodies to CD40 and IL2: coordination of dendritic cell and T cell arms of the immune response

W.J. Murphy1, L.A. Welniak2, J. Hixon1*, T. Back1*, J. Sublewski1*, J. Wigginton2, S. Wilson3, B.R. Blazar4, N. Seki2*, T. Sayers1*, R. Wiltrout2*
1 SAIC-Frederick, and 2 NCI-Frederick, Frederick. MD, 3 Chiron Corporation, Emeryville, CA, 4University of Minnesota Cancer Center, Minneapolis, MN

Background: Due to the weak immunogenicity of human cancers, the coordinate engagement of professional antigen-presenting cells (i.e. dendritic cells) and cell-mediated effector cells (i.e. CD8+ T cells) may be vital for inducing effective anti-tumor responses. CD40 is critical for, but not limited to, dendritic cell differentiation and function whereas IL2 is pivotal for, but also not limited to, T cell expansion.

Objective: To determine if the combination of CD40 stimulation with IL2 may result in synergistic anti-tumor effects in vivo.

Methods: Mice were given metastatic renal cell carcinomas and, after significant tumor growth, were subsequently treated with agonistic antibodies to CD40 (intraperitoneal administration; 10-100 µg/mouse/day for 14 days) plus human recombinant IL2 (intraperitoneal administration; 300,000 IU/mouse BID, twice a week for a total of 8 injections).

Results: In this study, protective effects were observed when either anti-CD40 or IL2 were administered as single agents to these advanced tumor-bearing mice. However, a marked and highly synergistic effect on tumor regression was observed following combined administration of CD40 agonist monoclonal antibody and IL-2 such that 70% of the advanced tumor-bearing mice survived the tumor and were immune to subsequent re-challenge with the original tumor. No overt toxicity was noted after treatment. Marked increases in both dendritic cell and CD8+ T cell numbers were observed after combination treatment compared to either treatment alone (p<0.005 and p<0.0001, respectively) and the recovered CD8+ T cells from this treatment group alone produced significantly greater amounts of interferon-gamma specifically in response to the tumor (p<0.05). The anti-tumor effects were independent of NK cells but were dependent on both interferon-gamma and IL12 since therapeutic synergy was not observed in interferon-gamma or IL12 knockout mice.

Discussion: Combination therapy using co-stimulatory molecules and cytokines such as CD40 stimulation in combination with IL2 administration can result in synergistic and efficacious immune responses against advanced metastatic cancer in mice.

Recurrent Murine Neuroblastomas Increase or Decrease MHC Class I Expression to Respectively Escape NK- or T Cell-Dependent Immunotherapy

Zane C. Neal1, Michael Imboden1, Alexander L. Rakhmilevich1, Kyung-Mann Kim1, Jean Surfus1, Holger N. Lode2, Ralph A. Reisfeld2, Stephen D. Gillies3, Paul M. Sondel11University of Wisconsin Comprehensive Cancer Center, University of Wisconsin-Madison, WI; 2Department of Immunology, The Scripps Research Institute, La Jolla, CA; 3Lexigen Pharmaceuticals Corp., Lexington, MA

We have evaluated tumors that recur following distinct immunotherapies that preferentially induce NK- or T cell-mediated in vivo control of the NXS2 neuroblastoma in tumor bearing mice.  Recurrent tumors that had escaped an NK-dependent antitumor response were generated by intentionally using a sub-optimal regimen of hu14.18-IL2, a humanized IL2 immunocytokine targeted to the GD2 ganglioside, which initially induced a clear antitumor response followed, however, by delayed tumor recurrence in some mice.  Importantly, we demonstrated that such recurrent NSX2 tumors revealed markedly enhanced (5-7 fold) MHC class I antigen expression.  The H-2high phenotype of these tumors is likely dependent on immune selective pressures in vivo since it is not maintained following 1 week in tissue culture.  Additionally, this enhanced H-2 antigen expression on NSX2 cells was associated with reduced susceptibility to both NK cell-mediated tumor cell lysis and antibody-dependent cellular cytotoxicity in vitro.  In contrast to the increased MHC class I antigen levels of NSX2 cells escaping after sub-optimal hu14.18-IL2 therapy, most NSX2 tumors that have escaped F1t3-Ligand treatment, which was able to induce a T cell-dependent memory response, revealed a decreased expression of MHC class I antigens.  While NSX2 tumors are highly susceptible to destruction following either hu14.18-IL2 or Flt3-Ligand mediated therapies, these results indicate that some tumor cells may respond to these in vivo immunotherapies by modulating their level of H-2 expression up or down, allowing them to respectively escape either NK cell-or T cell-mediated antitumor responses.

Combined Administration of Antibodies to Human Interleukin-8 and Epidermal Growth Factor Receptor Results in Increased Anti-metastatic Effects on Human Breast Carcinoma Xenografts

Salcedo, Rosalba1; Martins-Green, Manuela2; Gertz, Barry1; Oppenheim, Joost J1; Murphy, William J31Laboratory of Molecular Immunoregulation, NCI Frederick, Frederick, MD; 2Biology Department, University of California Riversite, CA; 3Intramural Research Program, SAIC, NCI at Frederick, Frederick, MD

Therapeutic approaches directed towards blocking the epidermal growth factor receptor (EGFR) based on its anti-proliferative activities are currently being evaluated clinically in cancer. Chemokines have been shown to affect the progression of various tumors, particularly metastatic spread and angiogenesis. The aim of this study was to assess the effects of a combined blockade with antibodies to EGFR and interleukin 8 (IL8). IL8 has been reported to augment the progression of some human tumors, and we therefore evaluated the anti-tumor effects of a human IL8 antibody, ABXIL8 in combination with anti-EGFR, ABXEGFR in a SCID mouse model bearing a metastatic human breast carcinoma. Whereas anti-IL8 alone had no appreciable anti-tumor effect, the combination of anti-IL8 significantly enhanced the anti-tumor effects of anti-EGFR, resulting in greater survival of SCID tumor bearing mice. This effect on survival was correlated with decreased metastatic spread in mice receiving both antibodies. Moreover, the combination of anti-IL8 and anti-EGFR markedly inhibited matrix metalloproteinase activity associated with MDA-231 cells to a greater degree than either antibody singly. Combined administration of these two human antibodies may thus provide a more effective approach for treatment of metastatic human breast carcinoma.

Repopulating stem cells carrying chemotherapy resistance-inducing genes

Stanton L Gerson, MD, Division of Hematology-Oncology and Comprehensive Cancer Center, Case Western Reserve University and University Hospitals of Cleveland, Cleveland, OH

Stem cell drug resistance gene therapy has yet to produce definitive results in patients. Drug resistance gene transfer into stem cells is designed to protect stem cells from cytotoxic damage and to allow enrichment of stem cells during myelosuppression. In contrast, most cancer gene therapy is designed to directly attack tumjor cells. While the first goal is to improve tolerance to chemotherapy, successful stem cell selection may lead to the use of these genes to enrich for cells carrying a second therapeutic correction gene be it to correct a genetic disorder, or target a malignant cell by introduction of a cytokine, suicide enzyme, antisense to an oncogene, or a wild type tumor suppressor gene. The most common approaches to gene therapy include murine oncoretroviruses although a number of investigators have introduced adenovirus, adeno-associated virus or plasmids by electroporation. For stem cells, retroviruses are the most promising, but these other methods have yielded recent success. More recently, lentivirus have been used because of the potential to introduce these vectors without the requirement of cell division and the use of cytokines.

In our own studies, we have used a series of mutant forms of MGMT identified as resistant to the alkyltransferase [AGT] inhibitor, O6-benzylguanine [BG]. These mutants target the region of the AGT protein that is responsible for the transfer of the alkyl group to cysteine, the key residue in the active site. The most potent mutants consist of one or more [up to 5] amino acid substitutions in the active site region of the protein that collapses the pocket making it inaccessible to BG, either by becoming more polar or by reducing the size of the pocket.. The most potent of these contains 5 amino acid substitutions and was identified using Darwinian molecular evolution and selection from a pool of over ten million mutants [3]. Others, containing G156A and P140K substitutions are also potent BG-resistant mutants [4]. Gene transfer of one modified murine retroviral backbone, MFG, into which has been placed the of one mutant MGMT, either G156A or P140 K results in a 120 to >2000 fold resistance to BG inhibition and an increase in the IC50 of about 6 fold for BCNU and up to 20-fold for methylating agents such as temozolomide [5]. We have made a series of observations in this system that proof its value in vitro and in animal models. First, we showed that transduced murine and human hematopoietic progenitor cells are resistant to BG & BCNU. We next showed that mice transplanted with transduced cells are protected from the lethal effects of BCNU and BG and that these cells are selected for during drug treatment. In fact we found that the mice were protected from myelosuppression and gained in cell counts with each dose of treatment. We then moved to a setting of cell infusion without prior myeloablation, and asked whether it was possible to enrich for transduced cells in mice that received only a small inoculum of cells. In a study performed in mice that had not received myelosuppression prior to cell infusion, animals infused with as few as 50,000 transduced cells could be completely repopulated with MGMT+ cells after 2-3 doses of BG& BCNU [6]. When we asked how much enrichment at the level of the long term repopulating cell had taken place, we calculated that the enrichment was 1000 fold in favor of the transduced cells. These mice can serve as the donors for secondary transplant recipients, and these secondary recipients continue to express the MGMT gene and to contain a high proportion of the transduced cells, indicating that repopulating stem cells have been transduced and that these cells express the MGMT gene. Next , we turned to a pre-clinical model of stem cell drug resistance gene therapy. In nude mice carrying the SW480 human colon cancer xenograft, improved tolerance to drugs and better tumor control over more cycles of chemotherapy was observed [7]. We have now initiated a clinical trial in which patients with advanced malignancy will have collection of CD34 cells that are then transduced with the MFG G156A mutant MGMT and reinfused into the patient. Patients will then receive standard phase II doses of BG [120 mg/m2] and BCNU [33 to 40 mg/m2]. The BG dose was selected based on our phase I trial that indicated that tumor alkyltransferase activity was completely inhibited without toxicity [8]. During sequential treatments with BG and BCNU, monitoring of marrow CD34 cells will determine whether transduced cells can be identified and whether there is enrichment of these cells over time. Our results suggest that MGMT gene transfer remains a promising approach to stem cell drug resistance gene mediated selection that may make possible significant degrees of stem cell protection and selection in vivo with application in both chemotherapy treatment for cancer patients as well as therapeutic introduction of other genes.

    1. Abonour, R., et al., ]. Nat Med, 2000. 6(6): p. 652-8.
    1. Allay, J.A., et al., . Nat Med, 1998. 4(10): p. 1136-43.
    1. Davis, B.M., et al., Proc Natl Acad Sci U S A, 2001. 98(9): p. 4950-4.
    1. Davis, B.M., et al., . Hum Gene Ther, 1999. 10(17): p. 2769-78.
    1. Reese, J.S., et al., . Clin Cancer Res, 1999. 5(1): p. 163-9.
    1. Davis, B.M., et al., . Blood, 2000. 95(10): p. 3078-84.
    1. Koc, ON, et al.,. Human GeneTher, 10:1021-1029, 1999
  1. Spiro, T.P., et al., . Cancer Res, 1999. 59(10): p. 2402-10.

Clinical Results of a PSA-Selective, Replication-Competent Adenovirus for the Treatment of Prostate Cancer.

T. L. DeWeese, Drew, R., Li, S., Goemann, M. A., Mikhak, B., Rodriquez, R., Hamper, U., DeJong, M.R., Detorie, N., Simons, J.W.
The Johns Hopkins University School of Medicine, Baltimore, MD 21287

Purpose: To determine the maximal tolerated dose (MTD) of CV706, a replication-competent cytolytic adenovirus, when delivered into the prostate by a transrectal ultrasound-guided transperineal technique in patients with prostate cancer (PCa) that is locally-recurrent following radiation therapy. Secondary objectives included evaluation of antitumor activity as determined by serum PSA and evaluation of post-treatment prostate biopsies, time to progression, PSA velocity, and evaluation of biodistribution and immune response to therapy.

Materials and Methods: CV706 was constructed by deleting E3 from an Ad5 adenovirus and adding a minimal promoter-enhancer construct of the human PSA gene 5’ of E1A resulting in regulation of E1A expression and, thus, regulated replication. All patients had biopsy-proven locally-recurrent PCa with biochemical failure as defined by the ASTRO criteria or a PSA > 10 ng/ml on 2 successive occasions. Other entry criteria included: normal serum testosterone, no evidence of metastatic disease on bone and CT scans. Dose escalation method: single administration of 1X1011 to 1X1013 viral particles divided in 20-80, 0.1 cc deposits injected with 10-40 needles. Pre-treatment planning was performed using a modified 3-D prostate brachytherapy treatment planning system. Toxicity was graded using the NCI CTC. Serial post-treatment biopsies were obtained to assess for treatment effect, PSA staining and evidence of viral replication. Serial urine and serum samples were also obtained to assist in PK/PD measurements.

Results: Twenty patients were treated over 5 dose levels. Treatment was well tolerated. Adverse events consistently noted included: Day 1, Grade 2 fever (treated with antipyretics), perineal bruising/edema/pain, urinary irritation (largely due to the indwelling Foley catheter). Five patients had mild, transient Grade 1 transaminase elevation. No clinically-significant alteration in hematology or coagulation parameters noted. There were 5 reversible Grade 3 toxicities (1 patient with hypertension, 1 patient with short episode psychosis, 3 patients with blood clots in urine). There was no dose-limiting toxicity. Four patients exhibited a PR (> 50% reduction in PSA for at least 4 weeks). Note that all responders occurred in final 2 dose levels, i.e. 4/11 patients. Viral replication confirmed by EM of post-treatment biopsy samples. Quantitative PCR documented CN706 in blood at two different time points, also suggestive of viral replication.

Conclusions: CV706, administered by a modified prostate brachytherapy technique, exhibits an acceptable safety profile with both biochemical and histologic evidence of antitumor activity in patients with prostate cancer that is locally-recurrent following radiation. In vivo replication was established in these patients. As we have recently demonstrated substantial in vitro and in vivo synergy when CV706 treatment is followed by radiation, these results support continued clinical development in a Phase I/II study combining treatment with CV706 plus radiation thearpy for initial management of patients with clinically-localized prostate cancer.

Supported by NIH S.P.O.R.E. in Prostate Cancer, NIH GCRC Johns Hopkins and a clinical research agreement only with Calydon, Inc.

Vaccines for Cancer Treatment: From Mice to Men

Elizabeth Jaffee, Department of Oncology, Johns Hopkins University, Baltimore, Maryland 21231, USA

The goal of cancer treatment is to develop modalities that specifically target tumor cells, thereby avoiding unnecessary side effects to normal tissue. Vaccine strategies that result in the activation of the immune system specifically against antigens expressed by a cancer have the potential to be effective treatment for this purpose. An early vaccine approach that was developed by our group involves the use of autologous tumor cells that are genetically modified to express the cytokine granulocyte-macrophage colony stimulating factor (GM-CSF), as the immunogen to induce systemic antitumor immunity. Studies in our animal model demonstrated that the paracrine secretion of GM-CSF at the site of tumor cells induced a CD4 and CD8 T cell dependent systemic immune response potent enough to cure mice of most histologic tumor types. We first tested this vaccine in patients with renal cell carcinoma (RCC) and found that this vaccine approach can activate RCC specific immunity in some patients. This induction of antitumor immunity was associated with clinical responses. Unfortunately, we also found that it is not technically feasible to develop an autologous vaccine approach for most types of human cancer including pancreatic adenocarcinoma. More recent data from our group and others has demonstrated that the immunizing tumor cell does not need to be HLA compatible with the host to induce effective antitumor immunity. These data therefore provide the immunologic rationale for using allogeneic cells as the antigen delivery system.

We have completed testing of an allogeneic pancreatic tumor vaccine approach in patients with stage 2 and 3 pancreatic cancer. Our findings demonstrated that GM-CSF secreting allogeneic vaccine cells can induce specific antitumor immunity comparable to an autologous vaccine. The findings that will be presented are similar to what we observed in our animal models and in patients with RCC thereby providing evidence for a similar mechanism of immune activation. These findings lay the ground work for further testing of this vaccine approach to determine if it will result in clinical responses. Furthermore, lymphocytes isolated from vaccinated patients may assist in the identification of the immune relevant antigens that are expressed by the vaccine and patient tumors. Finally, we are utilizing a murine model of tumor tolerance to develop methods of improving on the potency of our vaccine approach. Plans for translating new findings from this model are underway in pancreas and breast cancer.