Oral Abstracts

In vivo tumor imaging using near infrared labeled Endostatin

K. Camphausen1, C. Menard1, M. O’Reilly2, J. Folkman2
1Radiation Oncology Branch, National Cancer Institute; 2Surgical Research, Children’s Hospital, Harvard University

Introduction: Endostatin is a 20 kDa C-terminal fragment of collagen XVIII and is a potent inhibitor of angiogenesis. The exact mechanism of action for endostatin is unknown. Imaging technologies that use near-infrared (NIR) fluorescent probes are well suited to the laboratory setting. The goal of this experiment was to determine if endostatin labeled with a NIR probe could be detected in an animal after a sub-cutaneous injection.
Methods: Endostatin was conjugated to Cy5.5 monofunctional dye and purified from free dye by gel filtration. LLC a murine tumor, and BxPC-3 a human pancreatic tumor, were implanted in C57BL/6 and SCID mice respectively. Tumors were allowed to grow to 350mm2 when mice were injected with endostatin-Cy5.5 and imaged at various time points. Imaging was performed using a lightproof box affixed to a fluorescent microscope mounted with a filter in the NIR bandwidth (absorbance max 675nm and emission max 694nm). Images were captured by a CCD and desktop computer and stored as 16 bit Tiff files.
Results: The endostatin-Cy5.5 was quickly absorbed and rendered a NIR fluorescent image of the entire animal within 2 hours. Ex vivo imaging of multiple organs at 24 hours demonstrated no NIR image. In contrast, both the LLC and the BxPC-3 tumors emitted persisting NIR signal at 24 hours. Unlike previous analogous studies with GSAO-Cy5.5, whose tumor image faded with time, the endostatin-Cy5.5 emitted NIR signal from the tumor up to 7 days after injection, at the last time point examined. Cy5.5 dye alone had no tumor signal enhancement over background at any time point.
Conclusion: This study demonstrates that endostatin covalently bound to Cy5.5 will migrate from a distant sub cutaneous injection site to a tumor. This may indicate that the mechanism of action of endostatin is within the tumor or tumor vasculature and is not a systemic effect. This imaging technique is ideal for in-laboratory animal imaging and will be useful as the Radiation Oncology Branch begins the molecular credentialing project.

Peptide-Targeting IL-12 to avb3 Integrin Mediates Potent Antiangiogenic Activity

Dickerson EB1, Akhtar N1, Wang Z-Y2 , Steinberg H3, Padilla M1, Auerbach R4,5, Helfand SC1,5
Departments of 1Medical Sciences, 2Surgical Sciences, 3Pathobiological Sciences, School of Veterinary Medicine, 4Laboratory of Developmental Biology, University of Wisconsin-Madison, Madison, WI, and 5University of Wisconsin Comprehensive Cancer Center, Madison, WI

Background: Interleukin-12 (IL-12) is a cytokine with promising antiangiogenic activity mediated by the induction of interferon-g (IFN-g) and antiangiogenic chemokines. Although IL-12 does not have a direct effect on endothelial cell (EC) proliferation or migration, it triggers ECs to release IFN-g that can activate immune cells. Stimulated immune cells release factors that can inhibit EC proliferation suppressing angiogenesis. We developed a bifunctional fusion protein, murine recombinant IL-12 vascular homing peptide (mrIL-12vp) that simultaneously targets vascular and immune cell compartments. The fusion protein contains a small peptide sequence, arginine-glycine-aspartic acid (RGD), a ligand for the integrin avb3 that specifically directs mrIL-12vp to avb3 expressed on angiogenic endothelial cells.

Objective: We tested the hypothesis that directly targeting IL-12 to avb3 integrin on proliferating endothelial cells would be more effective at inhibiting angiogenesis and less toxic than systemic administration of non-targeted IL-12. Methods: mrIL-12vp was over expressed from CHO cells and an IL-12 ELISA verified production. ELISA was also used to validate activity of mrIL-12vp using production of IFN-? as a readout. Immunoflourescence labeling was used to verify specific targeting of mrIL-12vp to avb3 integrin expressed on malignant ECs and other tumor cells. Toxicity experiments were performed in DBA mice to evaluate tolerability of several doses of mrIL-12vp vs. mrIL-12. Corneal pocket neovascularization assays compared the antiangiogenic effects of mrIL-12vp with those of mrIL-12 in response to angiogenic stimulation. Results: mrIL-12vp elevated levels of IFN-g in vitro and in vivo indicating that it has biological activity indistinguishable from that of mrIL-12. Immunofluorescent assays using fluorescence-labeled anti-IL-12 antibody showed mrIL-12vp only bound avb3+ cells but not avb3- cells, indicating that mrIL-12vp specifically targets avb3. Systemic administration of mrIL-12vp (0.5 µg/day IL-12 equivalent) or mrIL-12 (0.5 µg/day) by continuous subcutaneous infusion in DBA mice resulted in hepatic necrosis in the mice treated with mrIL-12 but not mrIL-12vp. In two mouse models, corneal angiogenesis induced by bFGF showed at least a four-fold greater inhibition in angiogenesis in the mice treated at equivalent dosages of mrIL-12vp compared to those treated with mrIL-12. mrIL-12vp is a potent antiangiogenic agent with a favorable toxicity profile in mice.

Signal and transcriptional activation of proinflammatory and proangiogenic factor expression by cancer cells as a target for biological therapy

Van Waes C, Chen, Z, Dong G, Bancroft C, Sunwoo J
Tumor Biology Section, Head and Neck Surgery Branch, NIDCD, NIH, Bethesda, MD

Background: Angiogenesis, inflammation and immune dysregulation are important features of the pathogenesis of cancer, and these responses have been associated with activities of a diverse array of angiogenesis and immunoregulatory factors. The limited effectiveness of therapy targeted against individual angiogenesis and immunomodulatory factors or receptors has been found to arise from expression of this diversity of factors. The underlying basis for expression of such a diverse repertoire of factors in cancer is not well understood, but the identification of common mechanisms could provide a more limited number of targets for biologic therapy. Alterations in signal pathway and transcription factor activation have been implicated in oncogenesis, and such alterations within common pathways could regulate multiple genes, including angiogenesis and inflammatory factors.

Methods: mRNA differential display, microarray and ELISA were used to identify a repertoire of cytokines and angiogenesis factors expressed by murine and human SCC. Comparison of sequences of the promoter regions revealed common transcription factor binding sites, and the role of these transcription factors and upstream signal pathways in expression of cytokines was determined using genetic mutational analysis and inhibitors.

Results: IL-6, IL-8, GRO, GM-CSF and VEGF were detected in tumor, cell lines or serum of patients with head and neck squamous cell carcinomas, and their homologues were expressed with metastatic progression of murine SCC. These factors were found to be co-expressed as a result of signal activation of transcription factors Nuclear Factor kappa B and Activator Protein-1 by IL-1, Epidermal Growth Factor and MET receptors. Inhibition of these receptors and common signal pathways inhibited cell survival, proliferation, inflammation, angiogenesis and tumorigenesis.

Discussion: The molecular mechanisms responsible for oncogenesis of SCC promote co-expression of a diverse array of proinflammatory and proangiogenic cytokines. Inhibition of common signal receptors, kinases and transcription factors which regulate cytokine factors and other genes inhibit tumorigenesis and may provide a target for biological therapy.

Preclinical Strategies to Improve Dendritic Cell-Based Tumor Vaccines

J. J. Mulé, Department of Surgery and the Tumor Immunology and Immunotherapy Program, Comprehensive Cancer Center, University of Michigan Medical Center, Ann Arbor, Michigan 48109-0666

Tumor lysate-pulsed dendritic cells (TP-DC) serving as stimulators can educate naive primary lymphocytes in vitro, which can result in the generation of tumor-specific proliferative, cytokine producing, and cytolytic T cells. Moreover, immunization of syngeneic mice with TP-DC has resulted in potent specific priming and antitumor effects on micrometastatic pulmonary nodules in several histologically-distinct tumors, which are mediated by CD8+ and, to a lesser extent, CD4+ host-derived T cells. We have also found tumor lysates to be equivalent to apoptotic tumor cells as a source of TAA for pulsing of DC. The systemic administration of relatively low doses of recombinant IL-2 in combination with tumor TP-DC has resulted in markedly enhanced therapeutic effects against well-established tumors at either subcutaneous or pulmonary sites. We have also evaluated whether KLH can augment the antitumor efficacy of TP-DC immunization in vivo. In addition to being used as a “surrogate antigen” in vaccine approaches to measure immunologic response in cancer patients, KLH has also been shown to be a strongly immunogenic carrier protein to elicit T cell help. Indeed, the addition of KLH to TP-DC immunization can both profoundly augment IFN-gamma production by tumor-specific T cells and result in enhanced antitumor therapeutic efficacy in vivo. We have also shown for the first time that secondary lymphoid tissue chemokine (SLC) can induce a strong antitumor response that results in significant infiltration of immune effector cells into treated tumors and that genetic modification of DC to express SLC can enhance their capacity to elicit tumor rejection in vivo. Of importance, we have shown for the first time that SLC-secreting DC can effectively prime tumor-reactive T cells at the tumor site in the complete absence of functional lymph nodes. The potential of combining DC-based vaccines with bone marrow transplantation (BMT) for the treatment of metastatic disease is currently being considered for clinical evaluation, based on our recent successful results from preclinical studies. Indeed, in a lymphopenic environment, naive T cells can undergo homeostasis-driven proliferation and can acquire increased sensitivity to antigen stimulation. Our new findings demonstrate that it is possible to promote effective antitumor immunity in a defined lymphopenic environment following BMT through DC-based immunization. We have also focused effort on designing alternative strategies to overcome the potential limitation of sufficient tumor from an individual to produce the DC-based vaccine. In preclinical studies, we have found that approaches or agents that selectively elicit apoptosis of tumors in vivo may profoundly augment both the therapeutic efficacy and immune stimulatory capacity of injected DC alone.

Dendritic cells as vectors and targets for therapy

A. Karolina Palucka, Joseph W. Fay, Virginia Pascual and Jacques Banchereau
Baylor Institute for Immunology Research, Dallas, TX

DCs can be utilized either as vectors or as targets for therapy. Patients with metastatic melanoma received CD34-DC vaccine, that contains Langerhans cells and Interstitial DCs. DCs were pulsed with MART-1, tyrosinase, MAGE-3, gp100 and Flu-MP peptides, and KLH. DCs induced an immune response to control antigens in 16/18 patients. An enhanced immune response to 1 or more melanoma antigens (MelAg) was seen in these 16 patients. The 2 patients failing to respond experienced rapid tumor progression. 6/7 patients with immunity to 2 or less MelAg had progressive disease 10 weeks after study entry, in contrast to tumor progression in only 1/10 patients with immunity to > 2 MelAg. The tumor immunity score correlated with clinical outcome. Since tumor immunotherapy targets autologous antigens we can learn from systemic autoimmunity such as SLE. As opposed to normal monocytes, SLE monocytes induce proliferation of allogeneic CD4 T cells. SLE sera induce monocyte differentiation towards DCs in IFN-a dependent mechanism. Spiking autologous serum with IFN-a reproduces DC differentiation. 50% of SLE patients have high serum levels of IFN- a, which could explain T/B lymphopenia.Yet, plasmacytoid DCs, a major IFN-a source, are 80% decreased. pDCs and IFN- a may play a role in SLE pathogenesis and therapy.

Multi-epitope Peptide Vaccines with GMCSF and IL2: CTL Responses in the Peripheral Blood and in the Sentinel Immunized Node

Craig L. Slingluff, Jr., University of Virginia

Dermal administration of a melanoma vaccine should result in activation and trafficking of epidermal Langerhans cells to draining lymph nodes. There, lymphocytes may be activated and induced to proliferate. Despite this central role of the draining node in the response to vaccination, little is known about the T-cell response in these nodes, after tumor vaccines. Thus, we have performed two clinical trials in which a lymph node draining a vaccine site was harvested using sentinel node technology (sentinel immunized node, SIN), and was evaluated by ELIspot and by tetramers for reactivity to defined melanoma antigens, in conjunction with evaluation of the peripheral blood.

All vaccines used a mixture of 4 melanoma peptides, from tyrosinase and gp100, restricted by HLA-A1, A2, or A3, plus a modified tetanus helper peptide that induces Th1-type responses.

In the first of these trials (Mel31), two vaccine delivery systems were evaluated: (1) dendritic cells pulsed with peptides, and (2) injection of an emulsion of peptides plus GMCSF in incomplete Freund’s adjuvant. Patients had measurable stage IV disease or unresectable stage III disease. The emulsion of peptides and GMCSF was substantially more immunogenic that the dendritic cell vaccine, with CTL detected in 80-90% of patients receiving the former vaccine. CTL were detectable twice as often in the SIN as in the PBL.

In the second trial (Mel36), we evaluated the same peptides in GMCSF-plus-adjuvant, in high-risk patients (Stage IIB, III, or IV) after resection, but without clinical evidence of disease. Interestingly, CTL responses detected in the peripheral blood were much more consistent with CTL responses detected in the SIN. This finding suggests that, in the absence of measurable tumor, CTL are not depleted from the peripheral circulation in the same way that appears to occur in patients with advanced melanoma. In this study, we also evaluated the impact of low-dose interleukin-2 on the CTL response detected in the SIN and in the peripheral blood. These results will be presented.

Our newest trial will include a panel of 12 melanoma peptides, as a prelude to more complex multi-epitope vaccines incorporating epitopes for both CD4+ and CD8+ T-cells.