Abstracts from the 4th ImmunoTherapy of Cancer Conference

Abstract

Preclinical modelsA1 Intratumoral immunotherapy. B16-F10 murine melanoma modelJ. Ženka1, V. Caisová1, O. Uher1, P. Nedbalová1, K. Kvardová1, K. Masáková1, G. Krejčová1, L. Paďouková1, I. Jochmanová2, K. I. Wolf3, J. Chmelař1, J. Kopecký11Department of Medical Biology, Faculty of Science, University of South Bohemia, Česke Budějovice, Czech Republic; 21st Department of Internal Medicine, Medical Faculty of P. J. Šafárik University in Košice, Košice, Slovakia; 3University of Michigan Medical Center, Ann Arbor, MI, United StatesCorrespondence: J. ŽenkaBackground: Cancer immunotherapy based on direct intratumoral injection of immunomodulators has been established at the end of 19th century using Coley’s toxin. Our novel therapeutic strategy is based on the intratumoral injection of optimized mixture of TLR agonists, causing strong inflammatory infiltration. Infiltrating cells (mainly phagocytes) are directed to artificially opsonized tumor cells covered by phagocytosis stimulating ligands.Materials and methods: Immunotherapy was tested using B16-F10 murine melanoma model. Inflammatory infiltration was achieved using the mixture of resiquimod, poly(I:C), and lipoteichoic acid. Artificial opsonisation of tumor cells was elicited by mannan anchored to cell membranes using a hydrophobic anchor. The course of tumor infiltration was studied using flow cytometry. Cytotoxic effect of infiltrating immune cells on opsonized tumor cells was studied in vitro. Participation of acquired immunity was elucidated on the basis of intracellular IFN-gamma production by lymphocytes.Results: Optimized cancer immunotherapy based on the synergy of TLR agonists with artificially induced tumor opsonisation resulted in complete cure of 83% of mice with advanced melanomas. Moreover, cured mice acquired resistance to retransplantation of tumor cells. Applied therapy has a strong antimetastatic effect. In the first phase of immunotherapy, granulocyte predominance was observed, followed by involvement of acquired immunity.Conclusions: Intratumoral immunotherapy initiated by innate immunity based attack followed by joining of mechanisms of acquired immunity is a promising approach for future application in clinical practice as compounds used are safe for humans.A2 Assessing humanized mouse models for cancer immunotherapyL. Loumagne, J. Mestadier, S. D’agostino, A. Rohaut, Y. Ruffin, V. Croize, O. Lemaître, S. S. SidhuSanofi, Vitry, FranceCorrespondence: L. LoumagneBackground: Preclinical models that can recapitulate a functional human immune system are essential tools for the continued investigation of novel immunotherapy approaches. In this regard, immunodeficient mice offer a unique opportunity to reconstitute, at least partially, a human immune system together with human tumors to characterize immunomodulator compounds in a more physiologically relevant setting. The aim of this project was to perform a head to head comparison of several highly immunodeficient mouse strains engrafted with human CD34+ Hematopoietic Stem Cells (HSCs) for their ability to develop and sustain a multi-lineage engraftment of human immune cells. The most promising mouse model(s) will then be assessed for their capacity to support human tumor xenografts (CDX and PDX). Finally, we will determine if we can elicit an immune response to selected immunotherapeutic agents to drive anti-tumor activity.Materials and methods: The mice strains analyzed are NSG, NOG, NSG-SGM3, and NOG-EXL; the latter two being transgenic for the expression of human cytokines important in myeloid cell development. Humanization was performed in-house by injecting human umbilical cord blood CD34+ HSCs from different donors into the tail-vein of 4–5 week old female mice post-irradiation. To monitor human immune system reconstitution, flow cytometry analysis on blood was performed at approximately 4 week intervals beginning at week 4 post-injection of human CD34+ HSCs. At week 18, a terminal analysis was performed, consisting of a thorough immune cell profiling of blood, spleen and bone marrow by flow cytometry. Tissues were collected for the immunohistochemical staining of human immune cells and ex vivo functional assays will be performed.Results: Comparison of the immune system from different humanized mouse models show that all strains display high levels of engraftment of human CD45+ cells ranging from 50 to 70% at 18 weeks post-humanization. Population such as T CD4+ and CD8+ as well as B cells are well reconstituted and minor populations such as monocytes/macrophages, NK, and subsets of DC are also present at a lower percentage. Complete results obtained from different mouse strains will be available on the congress day.Conclusions: Overall, this project will provide a head-to-head comparison of several humanized mouse models that will be of value in the immunotherapy field according to the potent advantages of these models and the paucity of studies directly comparing different immunodeficient mouse strains.A3 A predictive PD-L1*CD8 cell density score for NSCLC patients treated with durvalumabS. Althammer1, K. Steele2, M. Rebelatto2, T. Tan1, T. Wiestler1, A. Spitzmueller1, R. Korn1, G. Schmidt1, B. Higgs2, X. Li2, L. Shi2, X. Jin2, K. Ranade21DEFINIENS, Muenchen, Germany; 2Medimmune, Gaithersburg, MD, United StatesCorrespondence: S. AlthammerBackground: Current PD-1/PD-L1 checkpoint inhibitor based immuno-therapies for advanced non-small cell lung cancer (NSCLC) show impressive response rates and long survival. We used automated image1 and data analysis of immunohistochemically (IHC) stained tissue sections within the Tissue Phenomics methodology to determine whether CD8+ and PD-L1+ cells densities could better identify patients most likely to respond to durvalumab than using a manual PD-L1 scoring of positive tumor cells (25% cutoff). Durvalumab is a human IgG1 monoclonal antibody that inhibits programmed death ligand-1 (PD-L1) binding to programmed death-1 and B7.1/CD80, restoring antitumor immunity2,3.Materials and methods: CP1108/NCT01693562 was a nonrandomized phase 1/2 trial evaluating durvalumab in advanced NSCLC and other solid tumors4. Digital slides from 163 baseline tumor biopsies, IHC stained for PD-L1 (Ventana SP263) and CD8 (Ventana SP239), were fully automatically scored with Definiens’ Developer XD 2.1.4 software using the product of PD-L1+ and CD8+ cell densities. The image analysis results were quality controlled using ground truth annotations provided by a panel of three pathologists. A discovery set (n = 84) was used to identify the scoring algorithm and the cutoff value, which was then tested on a validation set (n = 79). The robustness of the scoring algorithm was pre-validated using Monte-Carlo (MC) cross-validation.Results: The most accurate and robust scoring method in terms of MC performance metrics (positive predictive value, prevalence, pOS, pPFS) was identified as the product of average CD8+ and PD-L1+ cells densities in the pathologist annotated tumor center. Patients with high CD8*PD-L1 scores show significantly improved outcome compared to the CD8*PD-L1 low subgroup (ORR: 0.39 vs 0.07, OS log rank p-value: 0.0099 (training set) and 0.00053 (confirmation set)). CD8*PD-L1 further appears to predict patients’ outcome better than the PD-L1 status visually determined by pathologists (ORR: 39% vs 27%, median OS: 24.3 vs 17.8).Conclusions: Although further validation studies are required, we observe that a scoring algorithm based on automated image analysis of CD8+ and PDL1+ cell densities in baseline tumor biopsies improves the stratification of NSCLC patients treated with durvalumab when compared to visual histopathology scoring. References 1. Brieu, et al. Proc. SPIE 9784 Medical Imaging 2016: doi:10.1117/12.2208620.2. MedImmune/AstraZeneca. Data on file.3. Ibrahim R, et al. Semin Oncol 2015;42(3):474–83.4. Rizvi NA, et al. J Clin Oncol 2015;33(Suppl.):[Abstract 8032].A4 The influence of specific cytokines on cancer microtissue infiltrating cytotoxic T lymphocyte subpopulationsS. Koeck1, A. Amann1, G. Gamerith1, M. Zwierzina2, E. Lorenz3, H. Zwierzina1, J. Kern31Medical University of Innsbruck, Department of Internal Medicine V, Innsbruck, Austria; 2Medical University of Innsbruck, Department of Plastic, Reconstructive and Aesthetic Surgery, Innsbruck, Austria; 3Tyrolean Cancer Research Institute, Innsbruck, AustriaCorrespondence: S. KoeckBackground: Cancer associated fibroblasts are known to highly interact with infiltrating immune cells within the tumor microenvironment. In this work, we investigate the influence of specific fibroblast derived cytokines on the infiltration capacity of CD3 + CD8+ cytotoxic T lymphocyte subpopulations using a multicellular 3D co-culture System.Materials and methods: 3D tumor microtissues were cultivated using a hanging drops system. Human A549 and Calu-6 cancer cell lines were incubated alone or together with the human fibroblast cell line SV80 for 10 days to form microtissues. On day 10, peripheral blood mononuclear cells (PBMC) were added for 24 h. Endogenous cytokine expression of interleukin-2 (IL-2), IL-4, IL-5, IL-6, IL-12p70, interferon gamma (IFN-γ) and tumor necrosis factor alpha (TNF-α) in microtissue mono-, co- and tri-cultures was analyzed with a multi-cytokine immunoassay. Subsequently, measured cytokines were added to cancer cell/PBMC co-cultures to investigate the effect of these cytokines on infiltrating immune cell subpopulations. Infiltrating subpopulations were investigated by flow cytometry.Results: In both A549/SV80 and Calu-6/SV80 co-cultures and in SV80 monocultures, the cytokines TNF-α, IL-2, IL-5, IL-6 and IL-12p70 were measured. Minimal or no concentrations of the investigated cytokines were detected in cancer cell and PBMC monocultures. IL-4 and IFN-γ were not measured in all approaches. Additional application of the detected cytokines showed that IL-2 inhibited infiltration of naïve T lymphocytes into c