GLIOBLASTOM - Anni Hofmann Stiftung

Final report of the first funding period 2013 - 2015

Prof. Dr. Herold-Mende (Heidelberg): "Immunological changes and adaptation of the tumor microenvironment during glioblastoma progression."

Our research project aimed at characterizing the immune-relevant microenvironment during glioblastoma progression and the identification of immunogenic structures, which could be utilized for future immunotherapies. Therefore, tumor samples of patients with a quite opposite survival and 42 paired samples of primary and recurrent glioblastoma (pGBM, rGBM) were included in our studies. Thus, the initially intended study sample of 40 pGBM/rGBM pairs could be achieved.

Because microglia cells represent an especially dominant immune cell population in glioblastoma and are assumed to influence T cell-mediated immune responses, we first analyzed proportion and polarization of microglia cells and if this is related to patient survival. Therefore, cryosections of all tumor samples were prepared to perform multi-color-immunofluorescent stainings and subsequent TissueFAXS-based evaluation. These analyses showed a significant accumulation of immunosuppressive M2-polarized microglia cells in tumors of glioblastoma patients with a shorter survival (< 10 months) as compared to patients who survived substantially longer (> 3 years, Geisenberger et al., Acta Neuropathol. 2015). While comparison of pGBM and rGBM showed no change regarding the amount of infiltrating microglia cells, there were again marked differences regarding their polarization. Also, the simultaneous analysis of infiltrating T cells revealed striking changes in rGBM. To better understand the underlying mechanisms, microarray analyses of the same tumor samples have been performed. Thereby, massive changes of the transcriptome from pGBM to rGBM could be determined suggesting tumor cell selection processes, which result in an adverse immunological milieu. To obtain more information about the changes of immune-relevant chemokines and cytokines during glioblastoma progression, tissue lysates of 38 corresponding pairs of pGBM and rGBM have been analyzed by Luminex. A pronounced decrease of proteins, which promote tumor rejection, could be detected in rGBM. These data suggest a negative influence on anti-tumor immune responses in recurrent tumors. In the following funding period we will study this phenomenon in more detail.  

To identify immunogenic future vaccine antigens as well as changes in the immunogenic repertoire between pGBM and rGBM, a combination of protein fractionation (PF2D) and T cell activation assays (IFN‐γ ELISpot Assay) was used (n = 5). As already observed in the previous transcriptome analyses, all patients showed severe changes during tumor progression ranging from a decreased number of immunogenic tumor protein fractions in recurrent tumors to an increase. In total, more than 1500 proteins were identified in immunogenic fractions by mass spectrometry. An intensive filter process resulted in the selection of 19 proteins, which could be detected in immunogenic fractions of primary as well as recurrent tumors. To validate the therapeutic potential of these proteins as vaccine antigens, analyses in the patient of origin, in healthy donors as well as further glioblastoma patients have been performed. Thereby, four proteins showed a high immune response in more than 30% of all tested glioblastoma patients but no or only a low immune response in healthy donors. Overexpression of three of these proteins and an inverse relation to survival of two of them strengthen their suitability as vaccine antigens in future immunotherapies.

In order to finally obtain a more comprehensive impression about immune responses in glioblastoma patients and to test the efficiency of immunotherapies against defined antigens, we developed a diagnostic tool to examine immune responses in the peripheral blood (Mock et al., Oncotarget 2015). Six GBM-associated antigens have been printed on glass arrays and a pre-existing antibody-mediated immune response in the serum was tested. Thereby, antibody responses against the Tenascin C peptide VCED were identified as prognostic biomarker and will now be further validated.