URPP Fellow: Anna Rita Liuzzi

1. CTLA-4 checkpoint inhibition plus local IL-12 delivery for glioblastoma (Becher/Chahwan/Weller)
Patients with glioblastoma, the most common malignant primary brain tumor, experience a median survival in the range of 12 months despite multimodal therapy including surgery, radiotherapy and alkylating agent chemotherapy. This tumor has become a paradigmatic tumor for cancer-associated immunosuppression for decades. The first pivotal observations were the decrease in immune reactivity of peripheral blood cells harvested from glioblastoma patients. This prompted the search for soluble factors produced by glioblastoma cells that were supposed to be strong enough to confer systemic immunosuppression. Among these, transforming growth factor (TGF)-β assumed a central role in the concept of immunosuppression in this disease, but other factors like interleukin 10 or prostaglandin E2 received attention, too. Later on it became apparent that there are also cell surface expressed molecules, which may inhibit immune cell responsiveness, such as CD95 ligand, regeneration and tolerance factor (RTF) or, more recently, programmed death ligands (PD). Finally, in recent years it has been increasingly recognized that glioblastoma cells are capable of reshaping the phenotype of tumor-infiltrating host cells which compose a significant proportion of cells within the microenvironment of glioblastoma to support their growth and maintain an immunosuppressive milieu. Accordingly, and because of disappointing results with traditional cancer therapy in this disease, various approaches of immunotherapy have recently gained a lot of interest, and various phase II and phase III clinical trials of immunotherapy are on the way or have already been completed. The most advanced program, vaccination targeting the epidermal growth factor receptor vIII variant, has resulted in the completion of a phase III registration trial with results awaited in late 2015, led by the Brain Tumor Center in Zurich. Furthermore, there is now great interest in integrating antibodies controlling the immune checkpoint receptors CTLA-4 or PD-1 into the treatment of glioblastoma, and again Zurich is participating in clinical trials exploring the use of anti-PD-1 antibodies nivolumab or pembrolizumab in patients with recurrent or newly diagnosed glioblastoma.
In that regard, we have developed a novel approach to immunotherapy for glioblastoma that combines peripheral CTLA-4 immune checkpoint inhibition using ipilimumab with the intratumoral delivery of IL-12, a potent immune stimulatory cytokine that, however, was not tolerated by human cancer patients upon systemic delivery. Local delivery of toxic payloads including immunotoxins or antisense oligonucleotides has been explored in various clinical trials for glioblastoma, but investigators have so far failed to establish this approach in the clinic. Reasons for failure include mainly insufficient delivery, i.e., coverage of the target region, as well as the choice of inefficient or poorly tolerated therapeutic agents. Yet, the technical equipment to perform controlled, convection-enhanced local delivery of therapeutic molecules to brain tumor patients has been developed, is currently being explored in various clinical trials, and would thus be available in principle.
Thus, in the second funding period of the URPP, we focused our attention on how to translate our glioma pre-clinical experiments into patient clinical trials. In order to use IL-12 as a converter of a cold into a hot tumor microenvironment, we combined a single injection of IL-12: FC in the tumor site with the peripherical injection of anti-checkpoint blockade such as anti-LAG-3, TIM-3, PD-1. Preliminary results show the improved survival of the combinatorial approach IL-12:FC+anti-LAG-3 compared to the single therapies. To easily translate IL-12: FC pre-clinical research in clinical trial, we also generate a murine IL-12:Fc mRNA to test in mouse model of glioma and compare different formulations of mRNA to achieve the best transfection in vivo. We found that IL-12: FC mRNA is better transfected in vivo when present in a formulation of Ringer Lactate. Lastly, in cooperation with the University College London Cancer Research Center we found that CART cells directed against glioma cells had a mild capacity to reach the tumors, but the intratumoral application of IL-12: FC had a dramatic synergistic effect and led to durable anti-tumor responses.
Despite the encouraging outcome of IL-12 in preclinical trials, the administration of recombinant IL12 into patients with cancer is associated with severe side effects. To increase the efficiency of this cytokine in the tumor microenvironment and avoid its leakage in the periphery we will fuse the IL-12 sequence to the antibody specific for PD-L1. In glioma PD-L1 is one of the main immune checkpoint inhibitors which is upregulated in the tumor microenvironment on both myeloid and tumor cells. The fusion protein will then be tested in our GL261 glioma model and compared to the IL-12: FC for efficiency.
Previous data in our lab showed that the myeloid compartment in the model of glioblastoma contribute to the tumor rejection induced by IL-12. We will use single cells RNA-sequencing technique on individually sorted myeloid populations to study more deeply how IL-12 shape the tumor microenvironment and to identify new target responsible for the poor and good prognosis in glioma.

2. PD-L1 checkpoint inhibition for mesothelioma and non-small cell lung cancer (Curioni)
Malignant pleural mesothelioma (MPM) is an incurable cancer with a worldwide increasing incidence due to the use of asbestos. This is true also despite the use of different treatment modalities. Neither surgery nor radiotherapy alone has resulted in increased survival with a median of less than 12 months after diagnosis. Due to the unacceptable bad prognosis of patients with MPM new therapeutic approaches have to be developed. Immunotherapy with immune checkpoint inhibitors, such as anti-PD1, have shown some activity in patients with advanced mesothelioma and treatments combining chemotherapy with immune checkpoint inhibitors are currently under investigation for this disease. Lung cancer is the leading cause of cancer-related death worldwide. A recent breakthrough in the treatment of lung cancer has been achieved with immunotherapy based on immune checkpoint inhibitors. In metastatic NSCLC, immune checkpoint inhibition by targeting programmed death 1 (PD-1) (Nivolumab) doubled the responses of patients compared to standard chemotherapy in the second line setting and has been widely adopted. Despite initial response, primary and secondary resistance to immunotherapy is common and can have different causes. In patients with NSCLC under immunotherapy with immune checkpoint inhibitors, the use of daily dexamethasone was associated with worse outcome. Using a model of mesothelioma in immunocompetent C57BL/6, we showed that anti-CTLA-4 with anti-PD-1 synergize with the chemotherapeutic gemcitabine, the latter being a second line treatment for mesothelioma. We found that the synergistic combination led to an increased number of tumor infiltrating CD8+ T-cells and that dexamethasone abolished T cell infiltration as well as therapeutic efficacy. Based on these results, we administered a combination of an anti-PD-1 antibody plus gemcitabine in two patients who did not respond to gemcitabine or anti-PD-1 alone, and observed that the combination led to tumor shrinkage. Based on these results, a clinical trial has been designed and approved by the Swiss Group for Clinical Cancer Research with the use of gemcitabine in combination with checkpoint-inhibitors for solid tumors (as depicted below).  

3. Phase I/II clinical trial using IL-2 complexes in patients with metastatic melanoma (Onur Boyman, Richard Chahwan (AP), Reinhard Dummer, Steve Pascolo)
Rationale: Within D2 (project 1) we have generated anti-human IL-2 mAbs suitable for clinical development. These anti-human IL-2 mAbs are able, when combined with recombinant human IL-2, to form human IL-2 complexes that show the same features as murine IL-2 complexes in mice, to exert potent anti-tumor immune responses, while toxic adverse effects are minimal in comparison to standard IL-2 immunotherapy. We have recently developed these anti-human IL-2 mAbs in collaboration with a Swiss industrial partner in order to obtain fully humanized, good manufacturing procedure (GMP)-conform, quality checked material for fully human IL-2 complexes. Moreover, we are currently generating a single-molecule version of these fully human IL-2 complexes by engineering a single-chain molecule made of recombinant human IL-2 and the fully humanized anti-human IL-2 mAb interposed by a short flexible linker, as previously established. Notably, we and others have previously shown using recombinant murine IL-2 plus S4B6 anti-IL-2 mAb that a single-molecule version of murine IL-2 complexes shows the same in vivo properties as murine IL-2 complexes. The advantage of a single-molecule version of IL-2 complexes over standard IL-2 complexes is that IL-2 cannot fully dissociate from the anti-IL-2 mAb although the receptor binding sites of IL-2 remain unmodified, which addresses certain concerns of pharmaceutical companies.
Experimental strategy: Firstly, we will aim for the production of clinical-grade, fully humanized IL-2 complexes and a single-molecule version thereof and test these in mouse melanoma models to confirm their efficacy and safety as previously reported for their murine analogs. Then, we will investigate these complexes in purified human lymphocyte subsets to confirm their selectivity for anti-tumor effector T cells. Subsequently, we will plan and conduct a phase I/II clinical trial. This will encompass enrollment of suitable patients with metastatic melanoma, followed by IL-2 complex immunotherapy and measurement of primary endpoints, including tolerability and safety, as well as assessment of secondary endpoints following IL-2 complex immunotherapy in these patients, such as immune response and efficacy. As alternative to GMP produced single chain proteins, the systemic injection of liposome-formulated GMP mRNA coding for the desired protein will be evaluated.
Expected output: This first-in-human trial of IL-2 complexes in patients with metastatic melanoma will constitute a crucial step in the decision making of whether and how to further develop IL-2 complexes for cancer immunotherapy.