Annual Forum 2008 – Immunotherapy and Breaking Tolerance

James Allison, Ph.D., Memorial Sloan Kettering Cancer Center, New York, NY
Stanley Riddell, MD, Fred Hutchinson Cancer Research Center, Seattle, WA

Prof. Thomas Blankenstein, Charite, Berlin, Germany
Philip D. Greenberg, MD, University of Washington, Seattle, WA
Michael Jensen, MD, City of Hope Medical Center, Duarte, CA
Carl June, MD, University of Pennsylvania, Philadelphia, PA
Cornelis J. M. Melief, MD, Ph.D., Leiden University Medical Center, Leiden, Netherlands
Jeffrey Molldrem, MD, University of Texas MD Anderson, Houston, TX
David Munn, MD, Medical College of Georgia, Augusta, GA
Drew Pardoll, MD, Ph.D., Johns Hopkins, Baltimore, MD
Nick Restifo, MD, National Cancer Institute, Bethesda, MD
Steven A. Rosenberg, MD, Ph.D., National Cancer Institute, Bethesda, MD
Michael Sadelain, MD, Ph.D., Memorial Sloan-Kettering Cancer Center, New York, NY

Layout and Goals

  1. Local and Systemic Host Response to Cancer
  2. Immunologic Checkpoints and Cancer Vaccines
  3. Adoptive Cellular Therapy
  4. Genetic Modifications of Immune Cells for Cancer Therapy

The immune system in our bodies keeps us safe from our environment that is full of bacteria and viruses. When we catch a cold it is our immune system that fights the infection. When we cut ourselves our immune system ensures that foreign invaders do not take over our bodies.

For the last thirty years scientists and clinicians have tried to exploit the body’s immune system to treat patients with specific forms of cancer. Antibody therapy for diseases such as lymphoma and even breast cancer has been shown to be effective in many instances and these reagents are sold today for main line therapy. In academia efforts continue to exploit the cellular arm of the immune system and we are closer than ever to making an impact on specific cancers. Responses have been seen in patients but they remain unpredictable. Nevertheless the potential has been identified and the 2008 forum is intended to bring together a group of academics and clinicians to discuss manipulation of the immune system to assist in the treatment of cancer. This is the second time the Foundation has focused on this area of medicine indicating its importance in the development of new treatments for cancer and how quickly the field is moving.

Outcome Report
The 2008 Forbeck Symposium focused on new developments in cancer immunotherapy. Recent findings in tumor immunology have provided a clearer understanding of the complex interaction between the host immune system and cancer. Significant progress has been made in elucidating the function and regulation of immune cells, in identifying molecules that are expressed by tumor cells and can be targeted for immune-based therapy, and in understanding the mechanisms by which tolerance to self-antigens is maintained. These insights have resulted in a significant transformation in the field of tumor immunology over the past several years, and led to the development of novel therapies that are now yielding encouraging results in a subset of human cancers. The goals of this meeting were to review the progress made in understanding the host immune response to a developing cancer and the mechanisms that growing tumors employ to impede the host response, and discuss strategies that can be used to overcome regulatory checkpoints and promote effective immunity to cancer, either by adoptive T cell transfer or by vaccination.

The first session of the meeting was chaired by Jim Allison and dealt with the local and systemic host response to cancer. Drew Pardoll led off the session with a presentation provocatively entitled “Immunotherapy: saving our field to save cancer patients.” He reviewed data demonstrating how oncogenic signaling pathways involving the transcription factor Stat 3 can shift the immune response from an adversary of tumor progression to one that supports tumor growth. He illustrated how studies in animal tumor models have revealed the importance of combining therapeutics that disrupt pathways that inhibit effective antitumor immunity and with those that enhance antitumor immunity. He argued that combination therapies to simultaneously disrupt inhibition of immunity and to augment host responses to the tumor are likely be most effective and expressed concern that it will be difficult for academic scientists to assemble the reagents to perform such trials due to the proprietary interests of pharmaceutical companies. He expressed concern that even though the tools available for clinical translation are unprecedented, well conceived approaches such as vaccines that have proven effective in preclinical models, may fail in the clinic if the trial design and patient populations are not appropriately selected. David Munn focused his presentation on the role of indoleamine 2,3-diosygenase (IDO) in tumor induced tolerance. IDO has been shown to play a key role in tolerance to the fetus in pregnancy and in mucosal tolerance to foreign antigens. David presented data showing that IDO is expressed by some tumors, and is upregulated in dendritic cells in tumor bearing animals. These IDO expressing dendritic cells suppress effector T cell responses to the tumor and drive the expansion of regulatory T cells in the local tumor environment that inhibit the immune response to the tumor. Based on these observations, IDO is a potential target for pharmacologic inhibition and David described a small molecule IDO inhibitor drug that is now in Phase I clinical trials. This drug (1-methyl-Dtryptophan) has synergism in animal tumor models when combined with chemotherapy or tumor vaccines. Russell Jones, one of the Forbeck Scholars, presented his work on the metabolic adaptations to cellular stress. The ability of a tumor cell to grow depends on the cell utilizing available energy resources, and the increased proliferation of cancer cells requires changes in energy metabolism to meet metabolic demands. The mammalian AMP activated protein kinase (AMPK) is an evolutionarily conserved energy rheostat that alters the transcription of many genes key to mediating the response to cellular energy stress. Cancer cells utilize this pathway to enable their rapid proliferation. He explained how AMPK may function as a transcriptional activator by the modification of chromatin and this pathway could provide a target for novel therapies. Stefanie Sarantopolos was the second Forbeck Scholar to present at the meeting. Her work has focused on alterations in B lymphocyte homeostasis after allogeneic hematopoietic stem cell transplantation, and how B cell responses may contribute to the pathogenesis of chronic graft versus host disease and a graft versus leukemia effect. The premise of this work is that the severe depletion of B cells resulting from intensive chemoradiotherapy results in elevated levels of B cell activating factor (BAFF) in an effort to restore B cell homeostasis. High BAFF levels may contribute to the defective censoring of autoreactive B cells. Dr. Sarantopoulos explored a potential relationship between BAFF levels and B cell numbers in transplant recipients with and without chronic GVHD and found that an elevated BAFF/B cell ratio was associated with chronic GVHD. This suggests that high BAFF levels in the setting of allogeneic stem cell transplant may contribute to the development of alloantibodies and autoantibodies that are characteristic of chronic GVHD and identifies BAFF as a potential target for therapy. Thomas Blankenstein returned to the topic of T cell recognition of tumors and described a very elegant model of cancer in transgenic mice that develop sporadic tumors as a consequence of rare spontaneous activation of a dormant oncogene. These tumors express a strong antigen to which the mice are not initially tolerant. If immunized prior to tumor development, an immune response is elicited and the mice are protected from cancer development. In non-immunized mice, potentially tumor-reactive T cells are rendered tolerant by the growing tumor. These results reveal a failure of immunosurveillance of an immunogenic tumor, and illustrate some of the obstacles to developing immune based therapies for sporadic tumors in humans. This generated considerable discussion about the relevance of murine tumor models for recapitulating the issues relevant to human cancer. There was general agreement that the models that are currently available, while superior to previous models for deriving basic insights into mechanisms that may be responsible for the failure of the immune system to recognize tumors, still have limitations.

The afternoon session on day one was chaired by Drew Pardoll and reviewed what has been learned about immunologic checkpoints and how therapeutics are being developed based on inhibiting regulation of the immune system and the progress in cancer vaccines. Jim Allison led off this session and described the role of CTLA-4, which is a transmembrane protein expressed on T cells that serves as a key negative regulator of adaptive immune responses. CTLA-4 acts as a brake on T cell responses and plays a central role in maintenance of peripheral tolerance and shaping a developing T cell response. He presented data on treatment of patients with metastatic melanoma with an anti-CTLA-4 antibody to allow activation of anti-tumor immune responses. With this agent alone, there is a 16% response rate in advanced melanoma with one third of the responses consisting of complete and durable tumor regression. A high proportion of responding patients develop autoimmune toxicity that can be treated with systemic corticosteroids. Jim made the point that anti-CTLA-4 is an agent that may be substantially more effective if used in combination with vaccines or cell therapy. This prompted a vigorous debate about the difficulty developing combination therapies prior to FDA approval of the individual components of an experimental regimen. Kies Melief presented data on the use of a synthetic long peptide vaccine for human papilloma virus (HPV)-associated premalignant lesions. HPV causes cervical cancer and can also chronically infect the vulva where it induces a lesion termed vulval intraepithelial neoplasia (VIN) that eventually progresses to cancer. Many vaccines have been tried unsuccessfully, and Melief proposed that one reason for failure is due to inadequate stimulation of the CD4 helper T cell arm of the immune system. Studies in mice showed that immunization with a vaccine comprised of long overlapping peptides from the HPV E6 and E7 antigens that can elicit CD4+ and CD8+ T cells mediated the eradication of established HPV positive tumors. This approach was then tested for efficacy in women with high-grade VIN. Each patient received three immunizations of E6/E7 peptides formulated in an adjuvant. Vaccinated patients did not have side effects, E6 and E7-specific T cell responses increased after the vaccine, and objective clinical responses were observed in a high proportion of patients, providing encouraging evidence that premalignant lesions might be eradicated by immunotherapy. Joshua Brody was the third Forbeck Scholar to present and discussed the development of vaccination for B cell lymphoma in preclinical models and in patients. Dr. Brody employed direct intratumoral injection of CpG oligonucleotides that activates Toll like receptor 9 (TLR9) and enhances the ability of antigen presenting cells to induce an immune response to tumor-associated antigens. In a murine model of lymphoma this approach elicited tumor-specific CD8+ cytotoxic T cells that mediated tumor regression but also activated CD4+ regulatory T cells. The effect of regulatory T cells could be mitigated by transferring T cells from vaccinated animals into lymphodepleted mice, which resulted in expansion of the transferred effector T cells and improved the antitumor effect. The CpG intratumoral vaccine was administered to patients without toxicity. Several objective clinical responses were observed. Responses occurred at non-injected tumor sites and correlated with induction of a tumor-specific CD8+ T cell response. The investigators are planning a future trial incorporating storage of post vaccine lymphocytes for cell transfer after lymphodepletion. Jeff Moldrem presented his studies targeting PR1, an antigenic epitope derived from the proteinase 3 gene that is overexpressed in acute and chronic myeloid leukemias. Dr. Moldrem discovered the PR1 epitope several years ago and has performed clinical trials of PR1 peptide vaccination in patients with myeloid leukemia that demonstrated the induction of immune responses in 53% of patients and objective clinical responses in 18% of patients. Patients with antitumor responses had PR-1 specific T cells with a higher functional avidity than non-responders. Dr. Molldrem generated a monoclonal antibody to the HLA A2/PR-1 peptide complex, one of very few monoclonal antibodies to defined MHC/peptide determinant. He presented data showing the HLA A2/PR1 monoclonal antibody specifically stained acute myeloid leukemia cells that express proteinase 3 but not promyelocytes suggesting that leukemia cells process the PR3 protein differently than their normal bone marrow counterparts. This monoclonal antibody also mediated complement dependent cytotoxicity of leukemic cells both in vitro and in mice engrafted with human leukemia. These results support efforts to target PR1 by vaccination or adoptive T cell therapy and suggest a potential therapeutic role for the HLA A2/PR1 monoclonal antibody. The second day of the symposium focused on cell based immunotherapy for cancer.

Steven Rosenberg opened the session on adoptive T cell therapy by discussing work at the Surgery Branch of the National Cancer Institute on of T cell therapy for metastatic melanoma. This approach relies on the isolating T cells from surgically resected melanoma specimens, expanding these T cells in interleukin-2 and administering large numbers of cells selected for tumor reactivity back into patients. The partial and complete response rate after adoptive immunotherapy was dramatically improved when lymphodepleting chemotherapy was administered prior to adoptive T cell transfer. Lymphodepletion results in an increase in serum levels of IL-15 and IL-7 that promote T cell survival and proliferation and depletes regulatory T cells and suppressor cells. Several lymphodepletion regimens have been examined including the use of myeloablative total body irradiation requiring hematopoietic stem cell support. The toxicity with this approach is tolerable and response rates of 50-72% in patients with advanced and otherwise incurable melanoma are now being achieved. A key determinant of the antitumor response is the persistence of transferred T cells and the reason T cells fail to persist in some patients is unclear. There are obstacles to making this therapy available to more patients including the inability to obtain tumor-infiltrating lymphocytes (TIL) from many patients. A solution to this problem is to use gene transfer approaches to express the T cell receptor genes from highly avid tumor reactive T cells in peripheral blood T cells from the patient, Data demonstrating that this can be effective in humans was presented, although response rates in patients that received gene-modified T cells were significantly lower than those from whom TIL could be obtained. The remarkable success of T cell therapy for advanced melanoma validates the potential to employ the immune system to treat even advanced solid tumors. Stanley Riddell continued on the issue of T cell persistence and presented experiments in a nonhuman primate model that addressed the intrinsic qualities of antigen-specific T cells that are required for their persistence in vivo after adoptive transfer. T cells can be divided into antigen inexperienced naïve T cells (TN) and into broad classes of memory cells termed central memory (TCM) and effector memory (TEM). These cells have a distinct phenotype, homing properties, and transcription profile, and serve distinct functions in the immune response. T cells that are generated for cancer immunotherapy differentiate into effector cells while in culture and their origin (from TN, TCM, or TEM) has not been known for certain in any clinical trial. Cell sorting was used to purify TCM and TEM, and antigen specific effector T cell clones were derived from each subset and used for adoptive transfer. T cell clones from both memory subsets had equivalent cytolytic and proliferative capacity. However, CD8+ TE clones derived from TEM survived in the blood for only a short duration after adoptive transfer, and failed to persist in lymph nodes, bone marrow, or peripheral tissues. By contrast, TE clones derived from TCM persisted in the blood long-term after adoptive transfer, migrated to memory T cell niches, and responded to antigen challenge These results have implications for the types of T cells that should be selected for adoptive transfer, and for strategies to derive tumor-reactive T cells for immunotherapy of cancer by gene insertion. Nicholas Restifo continued on the theme of qualitative properties of T cells that enable their survival in vivo and antitumor activity. He presented data using T cell receptor transgenic mice as a source of naïve tumor-specific T cells and a murine model of melanoma. In this model, transfer of naïve T cells is effective in eradicating the tumor providing the tumor bearing mice are vaccinated to activate the T cells in vivo and interleukin 2 is administered. This model was used to determine the importance of lymphodepletion for the efficacy of transferred T cells and suggested that higher doses of total body irradiation resulted in more effective T cell mediated antitumor activity. The model also enables the analysis of culture conditions for activating the T cells ex vivo and analyzing their antitumor activity in vivo. T cells cultured with antigen in IL-2 exhibited greater cytotoxicity and cytokine production in vitro than T cells cultured in IL-15 but were less effective in mediating tumor regression on a per cell basis after adoptive transfer. Remarkably, the most effective T cells were those derived by culture in antigen and IL-21, which promoted proliferation of the cells but not differentiation to effector cells. These results suggest that the typical measures used to assess tumor reactivity of T cell products may not provide information on the quality of the cells or their potential for antitumor activity in vivo. Carl June presented the first talk on the use of genetically modified T cells for adoptive therapy of human malignancies and viral diseases. He discussed the construction of chimeric antigen receptors based on the fusion of a single chain antibody that targets a tumor cell surface molecule to the signaling components of the T cell receptor. He presented data showing that these artificial receptors can be readily expressed in human T cells using lentiviral vectors and confer recognition and lysis of tumor cells in vitro. Dr. June showed that the surface membrane glycoprotein mesothelin is overexpressed on mesothelioma, ovarian, and pancreatic tumors and described the engineering T cells with a lentiviral vector encoding a chimeric antigen receptor that targeted mesothelin. This receptor also incorporates the signaling domains of CD28 and CD137, which provide T cell proliferative and survival signals upon recognition of mesothelin positive tumor cells. These engineered T cells were transferred into immunodeficient mice that were engrafted with established mesothelin bearing tumors and could mediate complete tumor eradication at low effector to target ratios. The incorporation of the CD137 domain was important for T cell persistence in this model. This work illustrates the potential to target solid tumors by engineering immune effector cells for tumor recognition. The final session of the symposium continued the theme of genetic modification of immune cells for cancer therapy. Philip Greenberg reviewed the impediments to effectively treating patients with established malignancies, and discussed how genetic modification of T cells could be used to circumvent these impediments. He initially focused on the problem of isolating T cells with high functional avidity for the tumor from tumor bearing patients, and described how this might be overcome by assembling a library of T cell receptor genes that could be introduced into patient T cells. He described the use of yeast expression display as a method for rapidly screening for mutations in low avidity T cell receptors that enhance avidity. Expression of a tumor-specific T cell receptor in patient T cells can result in cross pairing with endogenous receptors. This both reduces the levels of the tumor-specific receptor and could result in a deleterious specificity. He described ways in which pairing, expression, and function of the introduced receptor could be optimized by modifications to the introduced receptor chains. Finally, he described an elegant mouse model in which T cells expressing dual receptors for a tumor antigen and a viral antigen respectively, were used to define how tolerance to tumor antigens develops, is maintained, and can be broken to promote tumor eradication. Michael Jensen described a variety of genetic engineering strategies for T cells including a novel approach to treating human glioblastoma, a malignant brain tumor that is notoriously resistant to radiation and chemotherapy. Glioblastoma expresses a form of the IL13 receptor and Jensen developed a tumor targeting receptor that consisted of a mutated IL-13 molecule that binds to the tumor form of IL-13R, and is linked to the T cell receptor zeta chain. T cells that express this “zetakine” after gene transfer recognize and lyse glioblastoma cells in vitro and in mice. He is now investigating this approach in an ongoing clinical trial in which patients with recurrent glioblastoma receive direct intratumoral injection of engineered T cells. This treatment has had remarkably little toxicity in the initial patients, and he has observed encouraging evidence of antitumor activity. He then described efficient methods for expressing multiple genes in primary T cells to enhance their antitumor activities and described a novel approach for editing regulatory genes using zinc finger nucleases to permanently disrupt the coding sequence at a targeted site. Lili Yang was the last Forbeck Scholars to present at the meeting. Her work focused on the use of lentiviral vectors to target tumor antigen expression in dendritic cells. For this purpose, the lentivirus envelope consisted of a viral glycoprotein from the Sindbis virus that was engineered to be specific for a dendritic cell surface protein, termed DCSIGN. This resulted in efficient introduction of tumor antigen and maturation of dendritic cells. Injection of this lentiviral vector into mice with established tumors promoted tumor regression and induced a CD8+ T cell response to tumor antigen. Thus, lentiviral targeting of tumor antigen to dendritic cells could provide a novel alternative to current vaccine regimens for inducing immune responses to tumors. The formal meeting concluded with a presentation from Michele Sadelain on redirecting T cells to be specific for leukemia by expressing a chimeric antigen receptor that targets the CD19 molecule on leukemia cells. He described novel constructs of the receptor that incorporate co-stimulatory ligands for T cells and provide for auto- and trans- co-stimulation of tumor-reactive T cells. These receptors provide improved antitumor activity in murine models, and may provide for more efficient activation of T cells in human tumor therapy. Michele is conducting a clinical trial of engineered T cells specific for CD19 to treat patients with chronic lymphocytic leukemia and presented the results in the initial patients suggesting that the engineered T cells can infiltrate very large tumors become activated into effector cells. Many B cell malignancies express CD19 and this appears to be a very promising approach for such patients.

The format of the Forbeck meeting provided for vigorous discussion and debate amongst the participants, and there was unanimous enthusiasm amongst the participants for this type of small focused meeting on a topic in cancer biology. The level of scientific discussion was outstanding and identified many opportunities for future collaborations that will hopefully result in the development= and rapid translation of improved approaches to treat cancer through immune manipulation.