Mapping of Tumor Stem Cells in the Resection Marigin During Extirpation of Highly Malignant Gliomas Using GlioStem

Background Glioblastoma (GBM) is the most malignant primary brain tumor with an annual incidence of around 3 in 100 000. The tumors are often found in a late stage of the disease and untreated patients have a median survival of only 3 months. Current state-of-the-art treatment includes maximal surgical resection, followed by concomitant radiochemotherapy. However, despite state-of-the-art treatment the prognosis remains dismal with median survival of less than two years. At recurrence, there is no standard of care, and further treatment including renewed surgery, chemotherapy re-challenge or bevacizumab offers only very limited prolongation of survival.

Emerging research suggests that failure to target glioma stem cells (GSCs) rather than the inability to remove tumors through surgery, radiation, or chemotherapy, explains the poor survival of GBM patients. It is convincingly demonstrated that GSCs possess tumor initiating abilities. GSCs also seem to escape conventional therapy due to their slow metabolism, and they can be in quiescent states for long times. GSCs can extend into the healthy tissue from the actual tumor, which is challenging for the surgeon as the healthy cells and the GSCs cannot be distinguished in real-time by 5-aminolevulinic acid (5-ALA) fluorescence guided resection or other proven methods or by other proven methods. Furthermore, as it is known that GSCs are resistant to chemo- and radiotherapy - and targeting of GSCs significantly reduces the risk for lethal relapse - being able to detect and eliminate GSCs during tumor resection would mean a crucial step towards increased patient survival. Hence, there is a need for better methods for precise removal of GSCs with minimum damage to healthy tissue to improve GBM prognosis and quality of life of operated patients.

Other studies in vitro and in animal models have pointed to the importance of GSCs for the recurrence and therapy resistance of gliomas. Targeting these cells specifically or tumor regions specifically dense in such tumor initiating cells - either macroscopically during a surgical procedure or with drugs specifically aimed at these subpopulation of cells - may therefore be an important aim in efficiently preventing tumor regrowth.

Celluminova's core technology is centered around the molecular luminescent biomarker GlioStem, which, unlike 5-ALA, is indicative of GSCs. GlioStem is an oligothiophene derivative sprung from research at KI and Linköping University (LiU), which can penetrate physiological cell membranes and selectively binds to structures inside the GSCs. GlioStem is conveniently administered onto the investigated tissue or cell culture by a pipette. After only a few minutes, the GSCs will emit green light from GlioStem molecules under blue illumination. The high luminescence specificity for the GSCs has been verified in large in vitro studies, including a great variety of human and animal cells, and in animal models. Moreover, we have shown that the high GlioStem specificity permits efficient separation of GSCs from astrocytes and other GBM cells by fluorescence-activated cell sorting (FACS). As GlioStem detects the quiescent slow-dividing GSCs that are not seen with Gliolan or by other methods, it can give valuable complementary information to Gliolan during fluorescent-guided tumor resection. As previously mentioned, it is known that GSCs are resistant to chemo- and radiotherapy. Therefore, being able to detect and eliminate GSCs during tumor resection would mean a crucial step towards increased patient survival.

In previous studies, we have shown that the cancer stem cell marker GlioStem can efficiently and with high specificity identify tumor cells with stem cell genotype in tissue samples from GBM (unpublished data). We now intend to study the extent to which this can be used to guide the procedure in a tumor resection to identify areas in the marginal zone where cancer stem cells can be identified and where extended resection would be particularly beneficial.

Project plan and methods

We will collect tissue samples from surgical resection of newly diagnosed or recurrent cerebral tumors, at Department of Neurosurgery, Karolinska University Hospital. We will collect samples from diffuse astrocytic and oligodendroglial human brain tumors WHO grade II-IV in adults. Samples will be pseudonymized. Tissue samples will be collected from the proliferating peripheral part of the tumor which is rich in viable tumor cells, as well as small tissue samples from the border zone just outside the dissection plane.

Clinical parameters that will be collected includes:

• Age
• Gender
• Diagnosis
• Treatment protocol
• Progression free survival
• Overall survival
• Status of histopathologically important markers such as IDH1, MGMT, EGFR, p53, BRAF and other markers
• Site of tumor

We will perform single-cell RNA sequencing using the 10x Genomics Chromium assay, collecting transcriptome data for about 10,000 single cells per sample. For a number of specimens, we will perform fluorescence-assisted cell sorting (FACS, RT-PCR and microarray analysis).

This dataset will be analyzed to (1) characterize the phenotypes of sorted cell populations by gene expression microarray to compare with the genotyping of the tumor and, (2) assess tumor heterogeneity and stem cell composition on a cell-by-cell basis and (3) Assess the correlation of the differences in gene expression (phenotype) between the cell populations within a tumor sample to the genotype. Our goal is to gain a detailed picture of the gene expressions of individual cells is provided, giving complementary information to the averaged genotype data for the whole cell populations. This data will be analyzed for spatial regions or subclones of cell with gene expression patterns indicating a "tumor initiating" or stem cell-like phenotype, to understand how these cell clones that are thought to propagate tumor recurrence are spatially distributed and how therapies specifically targeting these cells could be designed.

We will perform immunofluorescence microscopy of tissue samples using the stem cell marker GlioStem (developed by Celluminova AB). This marker can be applied to tissue biopsies, and specifically stains stem cell-like tumor cells in gliomas. We will investigate if this can reliably identify regions within tumor tissue samples that are dense with such GSCs. If so, this marker could potentially in the future be developed for patient administration during surgery, allowing the surgeon to already intraoperatively be able to identify regions enriched with cells with a propensity to drive tumor recurrence. This will also be informative for diagnostic and prognostic purposes

Aims

• The overall aim is to detect and visualize stem cell-like cells using biomarkers or based on tumors' genetic material or proteins, thereby developing improved treatment strategies for patients with brain tumors, and in the long run improving the survival and quality of life of these patients.
• Characterize the spatial presence of stem cell-like cells in tumors and their periphery.
• Map whether there are any molecular or genetic differences between tumor stem cells in central tumor compared to stem cells in the periphery. This could in the long run be used to identify areas in the tumor that are particularly important to remove during surgery.
• Investigating if cancer stem cells or other tumor initiating cells could be reliably visualized using immunofluorescence techniques that could potentially be developed for intraoperative use.
• Map the composition of brain tumors at the cellular level, with the goal of finding proteins (genes) that are active in tumor cells but not in normal cells in the brain. It is also important to map signaling mechanisms between tumor cells.
• The development of specific markers (e.g. fluorescent molecules) that can identify tumor tissue and, in optimal cases, specifically the tumor cells that are essential for tumor growth.
• Mapping of spatial distribution of gene expression to understand regional differences within tumors with regards subtypes of cells and propensity to drive tumor proliferation.