Glioblastoma Tumors Research Articles

Abstract B046: Pyk2 and MEK/ERK signaling regulate the extracellular vesicle release and tumor-associated myeloid Cells in glioblastoma

Abstract Glioblastoma (GBM) is often fatal within a year despite treatment, with immune checkpoint blockade therapies benefiting only about 10% of patients. Non-responsive GBMs exhibit an immunosuppressive profile and higher levels of tumor-associated myeloid cells (TAMs). Addressing this immunosuppressive microenvironment is a major challenge in GBM immunotherapy. Our recent studies have identified a positive correlation between Proline-Rich Tyrosine Kinase 2 (Pyk2) activation in GBM tumor cells and the cytokine expression profile of TAMs. This identified Pyk2 as a potential prognostic marker for the immune state in GBM tumor microenvironment. We hypothesize that Pyk2 and MEK/Erk signaling in GBM cells regulate the release of chemokines and cytokines through extracellular vesicles (EVs) by modulating the actin cytoskeleton, thereby impacting TAM activation. The study aimed to investigate the role of Pyk2/MEK/Erk signaling in EV release in GBM cells, potentially revealing new therapeutic targets. Two humans primary GBM cell lines, with and without Pyk2 CRISPR/Cas9 knockout (Pyk2KO), were used. Flow cytometric analysis of EVs, enriched from cell-conditioned medium, identified a shift towards larger in diameter EVs populations in Pyk2 KO cells, compared to wild-type (WT) cells. Analysis using Integrin as a plasma membrane marker showed that 84.70% of Integrin+ and 15.30% of Integrin- EVs were derived from WT cells, whereas Pyk2KO cells produced 89.07% Integrin+ and 10.93% Integrin- EVs. Treatment of WT cells with the Erk/MEK inhibitor Avutometinib (1µM) altered the EV population’s ratio to 79.11% Integrin+ and 20.89% Integrin-. Similarly, Avutometinib treatment of Pyk2KO cells resulted in 78.83% Integrin+ and 21.17% Integrin- EVs, similar to the ratio observed in Avutometinib-treated WT cells. Western blot and PCR analysis of EVs content revealed a significant reduction in CCL2, CCL5, tumor necrosis factor (TNF), and vascular endothelial growth factors (VEGF) in EVs from Pyk2KO GBM cells, compared to WT. In vivo studies using GL261/C57Bl/6 mouse glioma implantation model and flow cytometric analysis of TAMs from tumors generated with WT and Pyk2KO GL261 cells demonstrated increased infiltration of TNF+/IFNg+ CD8+ lymphocytes and Ly6C+/CD206- myeloid cells, alongside a reduction in CD45+/CD11b+/CD33+/MHCII- myeloid-derived suppressor cells. PCR analysis of TAM isolated from Pyk2KO tumors revealed lower gene expression of TNF, CCL5, CCL2, VEGFa, and epidermal growth factor (EGF) compared to WT tumors. The study highlights that Pyk2/MEK/Erk signaling regulates the immune microenvironment in GBM by influencing EV release, which impacts TAM cell activation. Pyk2 primarily regulates the release of Integrin- EVs, while MEK/Erk predominantly regulates EVs shed from the plasma membrane. The findings underscore the interplay between Pyk2 and MEK/Erk signaling in regulating EV biogenesis and composition. Citation Format: Neisha M. Ramirez. Pyk2 and MEK/ERK signaling regulate the extracellular vesicle release and tumor-associated myeloid Cells in glioblastoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Tumor-body Interactions: The Roles of Micro- and Macroenvironment in Cancer; 2024 Nov 17-20; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2024;84(22_Suppl):Abstract nr B046.

Differential Replication and Oncolytic Effects of Zika Virus in Aggressive CNS Tumor Cells: Insights from Organoid and Tumoroid Models

Central nervous system (CNS) cancers are responsible for high rates of morbidity and mortality worldwide. Malignant CNS tumors such as adult Glioblastoma (GBM) and pediatric embryonal CNS tumors such as medulloblastoma (MED) and atypical teratoid rhabdoid tumors (ATRT) present relevant therapeutic challenges due to the lack of response to classic treatment regimens with radio and chemotherapy. Recent findings on the Zika virus’ (ZIKV) ability to infect and kill CNS neoplastic cells draw attention to the virus’ oncolytic potential. Studies demonstrating the safety of using ZIKV for treating malignant CNS tumors, enabling the translation of this approach to clinical trials, are scarce in the literature. Here we developed a co-culture model of mature human cerebral organoids assembled with GBM, MED or ATRT tumor cells and used these assembloids to test ZIKV oncolytic effect, replication potential and preferential targeting between normal and cancer cells. Our hybrid co-culture models allowed the tracking of tumor cell growth and invasion in cerebral organoids. ZIKV replication and ensuing accumulation in the culture medium was higher in organoids co-cultured with tumor cells than in isolated control organoids without tumor cells. ZIKV infection led to a significant reduction in tumor cell proportion in organoids with GBM and MED cells, but not with ATRT. Tumoroids (3D cultures of tumor cells alone) were efficiently infected by ZIKV. Interestingly, ZIKV rapidly replicated in GBM, MED, and ATRT tumoroids reaching significantly higher viral RNA accumulation levels than co-cultures. Moreover, ZIKV infection reduced viable cells number in MED and ATRT tumoroids but not in GBM tumoroids. Altogether, our findings indicate that ZIKV has greater replication rates in aggressive CNS tumor cells than in normal human cells comprising cerebral organoids. However, such higher ZIKV replication in tumor cells does not necessarily parallels oncolytic effects, suggesting cellular intrinsic and extrinsic factors mediating tumor cell death by ZIKV.

Beyond the blood: expanding CAR T cell therapy to solid tumors.

Chimeric antigen receptor (CAR) T cell therapy stands as a transformative advancement in immunotherapy, triumphing against hematological malignancies and, increasingly, autoimmune disorders. After a decade of relatively modest results for solid tumors, recent clinical trials and patient reports have also started to yield promising outcomes in glioblastoma and other challenging solid tumor entities. This Perspective seeks to explore the reasons behind these latest achievements and discusses how they can be sustained and expanded through different strategies involving CAR engineering and synthetic biology. Furthermore, we critically analyze how these breakthroughs can be leveraged to maintain momentum and broaden the therapeutic impact of CAR T cells across a variety of solid tumor landscapes.

TMIC-82. CHARACTERIZATION OF GENETIC MARKERS AND BIOLOGICAL PROCESSES UNDERLYING THE TRANSITION OF GLIOBLASTOMA CELLS FROM A DIFFERENTIATED TO A STEM-CELL-LIKE STATE

Abstract BACKGROUND The main drivers of Glioblastoma (GBM) tumor growth are elevated angiogenesis and heightened characteristics of stemness. These driving factors make GBM an extremely invasive and aggressive tumor. While the GBM tumor microenvironment can be classified into several molecular subtypes, one common characteristic is the atypical induction of various signal transduction pathways including the receptor tyrosine kinases (RTKs), pro-angiogenic, and stem-cell-determining transcription factors, which all play important roles in GBM tumor formation, growth, and progression. OBJECTIVE To investigate the regulations of RTKs, pro-angiogenic, and stem-cell–determining transcription factors in two different tumor environments - monolayer cultures that induce a more differentiated state and neurosphere cultures that select for stem-like cells. METHODS Neurosphere- and monolayer-forming cell samples were cultured from four primary GBM cell lines including HK-336, HK-374, U87, and GS-154. RNA sequencing was performed at the Technology Center for Genomics and Bioinformatics at the University of California, Los Angeles. The transcriptomic data findings were validated by qPCR and western blot assays. RESULTS Our transcriptomic data revealed the significant upregulation of EGFR, PDGFR-α, VEGFA, and SOX2 in the neurosphere-forming HK-336, HK-374, U87, and GS-154 primary GBM cells. Consistent with the transcriptomic data, western blot, and qPCR assays demonstrated significantly higher expressions of these genes in the neurosphere-forming compared to the monolayer-forming GBM cells. Neurosphere-forming GBM cells possess stem-cell-like characteristics with the ability to self-renew and differentiate via the induction of various genes and disease progression pathways. CONCLUSIONS Neurosphere-forming GBM cells demonstrate upregulation of SOX2, EGFR, PDGFR-α, and VEGFA in RNA sequencing, western blot, and qPCR assays. Multipotent drug therapies designed to target multiple gene regulation pathways involving RTKs, pro-angiogenic, and stem-cell-determining transcription factors may be necessary to effectively treat GBM. Understanding the underlying mechanisms driving these aberrant cellular pathways could provide valuable insights for developing targeted therapies against this aggressive brain cancer.

TMIC-03. CARD9 BLOCKADE OFFERS A NEW OPPORTUNITY TO REPROGRAM TUMOR-ASSOCIATED MACROPHAGES IN GLIOBLASTOMA

Abstract BACKGROUND Glioblastoma (GBM), one of the deadliest primary tumors of the central nervous system, is characterized by a highly immunosuppressive tumor microenvironment (TME). Tumor-associated macrophages and microglia (TAMs) are a dominant population of immune cells in the GBM TME and substantially contribute to immunosuppression, tumor progression, and treatment resistance. Elucidating the molecular mechanisms underlying TAM behavior is crucial to the development of effective immuno-therapeutic strategies for treating GBM. The adaptor protein, Caspase Recruitment Domain-containing protein 9 (CARD9) is primarily expressed in myeloid cells and its activation triggers assembly of the CARD9-BCL10-MALT1 (CBM) complex, which then promotes activation of NF-κB and/or MAPK pathways. The aim of this study was to investigate the role of the CBM complex in the GBM TME and assess its potential as a therapeutic target. METHODS/RESULTS In silico analyses revealed that GBM tumors demonstrate increased CARD9 and MALT1 expression in comparison to non-tumor brain tissue, with CARD9 expressed only within the myeloid compartment. Using co-culture, scratch wound, and transwell assays, we found that GBM tumor cells induce macrophage M2 polarization and migration via a mechanism that is dependent on CARD9/MALT1 signaling within macrophages. scRNA-Seq and flow-cytometric analysis of orthotopic syngeneic immunocompetent GBM tumor models revealed that the GBM TME is dominated by macrophages. Disruption of MALT1 protease activity within the TME leads to reduced tumor infiltration by immunosuppressive TAMs, remodeling of TAMs toward an immunoreactive anti-tumor phenotype, and improved survival of GBM-bearing mice. Additionally, we found that tumor growth was abrogated when GBM-bearing mice were treated with either a MALT1-protease inhibitor or a CARD9 inhibitor. CONCLUSIONS Our study suggests that blockade of the CARD9-MALT1 signaling axis impairs GBM-induced macrophage recruitment and M2 polarization and inhibits tumor progression. These findings nominate the CARD9-MALT1 signaling axis as a new target for TAM-directed immunotherapy for the treatment of GBM.

TMIC-20. UNLOCKING THE POTENTIAL OF HYPOXIA REDUCTION IN RESHAPING THE GLIOBLASTOMA TUMOR MICROENVIRONMENT FOR ENHANCED THERAPEUTIC SYNERGY

Abstract Glioblastoma (GBM) tumors respond poorly to nearly all cancer therapies due to their highly immunosuppressive tumor microenvironment (TME). Hypoxia exacerbates this immunosuppression by upregulating the hypoxia-inducible factor (HIF) pathway in glioma-associated myeloid cells (GAMs) which suppressively polarizes them to maintain this immune “cold” state. This hypoxia-driven programming inhibits CD8+ T cell infiltration and effector function by inducing metabolic dysfunction, thereby increasing resistance to chemotherapy and T cell immune checkpoint blockade (ICB). Targeting hypoxia thus emerges as a promising strategy to overcome such resistance. Our study suggests that GBM hypoxia ablation can convert the immunosuppressive tumor microenvironment into a more pro-inflammatory one. Tumor hypoxia ablation using the hypoxia-activated prodrug Evofosfamide significantly enhances survival in both GL261 (p<0.0001) and the checkpoint blockade-resistant CT2A murine GBM models (p<0.01). Evofosfamide treatment boosts tumor infiltrating Ki67+ proliferative CD8+ T cells (p<0.001) and heightens cytotoxicity, indicated by increased Granzyme B expression (p<0.01). It also reduces the co-expression of exhaustion markers LAG-3 and TIM-3 (p<0.001), signaling decreased T cell exhaustion. Antigen-specific CD8+ T cells from Evofosfamide-treated primary GBM tumors exhibit enhanced ex vivo killing capacity when co-cultured with GL261-OVA cells (p<0.001). Furthermore, Evofosfamide enhances microglia proliferation and repolarizes them towards a pro-inflammatory state, marked by upregulation of M1 markers such as CD80 (p<0.001) and downregulation of M2 markers like Arginase (p<0.05). Hypoxia reduction also increases chemotherapy sensitivity, synergistically improving survival when combined with the blood-brain barrier-penetrating, immunogenic chemotherapy Berubicin (CT2A; vehicle MS=26d, Evo MS=36d, Berubicin MS=55d, Combo MS=70.5d). Future research will investigate the mechanisms underlying this shift towards a pro-inflammatory state, focusing on metabolic changes in CD8+ T cells and the impact of altered HIF pathway signaling on GAM polarization following hypoxia ablation. Understanding the therapeutic potential of reversing hypoxia-mediated immunosuppression could lead to novel combination therapies capable of overcoming resistance to both chemotherapy and immunotherapy in GBM.

TMIC-25. MICROGLIA IN THE GLIOBLASTOMA MICROENVIRONMENT SECRETE ACETYLATED AMINO ACIDS WHICH ASSIST IN THE REPAIR OF RADIATION INDUCED DNA DAMAGE

Abstract Rapid repair of DNA damage mediates genotoxic therapy resistance and poor prognosis in glioblastoma (GBM). The GBM tumor microenvironment (TME) is heterogenous, and we hypothesized that non-malignant cells support the repair of DNA damage in glioblastoma cells. Consistent with this hypothesis, we found that GBM tumors grown intracranially in mice had rapid DNA repair and were resistant to radiation compared to tumors of identical genotype grown extracranially in mouse flanks. Further interrogation of irradiated intracranial tumors using gamma-H2AX immunofluorescence revealed spatial heterogeneity in DNA damage repair, with spatially separate regions of rapid and slow gamma-H2AX resolution. Paired spatial transcriptomic analysis and gamma-H2AX immunofluorescence showed that tumor regions with rapid DNA repair were rich in microglia. Having nominated microglia from these results, we cocultured GBM cells and microglia in direct contact as well as in transwell coculture. We found that GBM cells cocultured with microglia repair DNA damage faster in a contact-independent manner. We then assessed the supernatant metabolome of microglia and GBM monocultures and cocultures and found that acetylated amino acids are secreted by microglia and consumed by GBM cells. Supplementation with acetylated amino acids promotes radiation resistance by speeding the repair of radiation-induced DNA damage. Mechanistically, acetylated amino acid supplementation allows glioma cells to re-route glucose utilization away from the TCA cycle and towards nucleotide synthesis. Hence, we show that microglia in the GBM microenvironment assist GBM cells to repair radiation induced DNA damage by secreting acetylated amino acids which are consumed by GBM cells and used to fuel the TCA cycle allowing utilization of glucose for nucleotide synthesis, thereby making the GBM cells resistant to radiation therapy.

EXTH-12. POTENTIATING THE EFFICACY OF IMMUNE CHECK-POINT INHIBITORS IN GLIOBLASTOMA BY INHIBITION OF CXCL12

Abstract BACKGROUND The tumor immune microenvironment (TIME) in glioblastoma (GBM) is rich in CXCL12, a chemokine known for stimulating angiogenesis. CXCL12 also controls immune cell trafficking and promotes polarization to an immunosuppressive phenotype. We postulated that inhibiting CXCL12 will modulate the immunosuppressive TIME in GBM, thereby augmenting the efficacy of immunotherapy agents like immune checkpoint inhibitors (ICIs). We examined the immunomodulatory and therapeutic efficacy of CXCL12 inhibition, using a small molecule NOX-A12, with and without ICIs in pre-clinical murine GBM models. METHODS B6 immunocompetent mice bearing intracranial (i.c.) or subcutaneous (s.c.) SB28 tumors received either vehicle, NOX-A12, anti-PD1 and anti-CTLA4 (dual ICI) or NOX-A12 + dual ICI (combination). Tumor growth was measured by IVIS imaging and immune cells in brain were analyzed using high dimensional flow cytometry. Survival was analyzed using Kaplan-Meier curves and log rank test. RESULTS Combination treatment resulted in tumor regression and long-term survival in 40% of s.c. tumor bearing mice compared to 10% treated with dual ICI only. Three of 4 mice rechallenged with contralateral s.c. tumor remained tumor free while all naive mice reached endpoint. TIME from treated s.c. tumors revealed increase in CD4 and effector memory CD8 T cells by combination treatment compared to dual ICI. In i.c. GBM tumors, while no survival benefit or growth inhibition was seen, combination treatment induced an early increase in effector memory CD8 T cells and PD-L1+ B cells, and a later increase in MHC-II+ microglia compared to dual ICI. CONCLUSION Inhibition of CXCL12 enabled the increase of effector cytotoxic T cells by ICI, improving survival in the s.c. GBM model, but not in i.c. GBM model. This might be due to differences in CXCL12 effect between extra and intra-CNS tumor or due to robust immune response causing excessive brain edema and death. These will be further explored in ongoing and future studies.

EXTH-58. BLOCKING IMMUNE CHECKPOINT— LAIR1: DISRUPTING THE LAIR1—FXIII-A—COLLAGEN IMMUNOSUPPRESSIVE CIRCUIT FOR ENHANCED ANTITUMOR THERAPEUTICS

Abstract BACKGROUND Recent evidence indicates that tumor-associated myeloid cells (TAMCs), including tumor-associated macrophages (TAMs) and myeloid-derived suppressor cells (MDSCs), play a crucial role in glioblastoma (GBM) immunosuppression and progression. These cells can constitute up to 50% of the total mass in GBM tumors and are pivotal in dampening the immune response, which negatively impacts patient survival. Targeting TAMCs represents a promising strategy to overcome the limitations of current cancer treatments. However, drug development in this area remains limited. Our study focuses on Leukocyte-Associated Immunoglobulin-like Receptor 1 (LAIR1), a novel immune checkpoint overexpressed on TAMCs within GBM. We comprehensively analyze LAIR1’s inhibitory role in cancer progression, highlighting its potential as a therapeutic target. METHODS Both in vitro and in vivo experiments were conducted. LAIR1 wild-type (LAIR1+/+) and knock-out (LAIR1-/-) tumor models were tested for antitumor response and intratumoral immune landscape. We also utilized 2D and 3D cultures to evaluate the effects of antibody blockade (aLAIR1) alone and in combination with CAR-T therapy (8R-70CAR) in both human and mouse GBM models. RESULTS We have uncovered a previously unrecognized immunosuppressive LAIR1→ Factor XIII A (FXIII-A)→ Collagen IV circuit across cancer types. Inhibiting LAIR1, either through genetic LAIR1-/- or aLAIR1, effectively disrupts this immunosuppressive pathway and results in enhanced peripheral and tumor-infiltrating memory CD8 T cell populations, a phenotypic shift of TAMs towards an M1 profile, and a decrease in tumor collagen deposition. These collective effects contribute to the normalization of the TME and facilitate improved interactions between T cells and tumor cells, leading to a more effective antitumor response. Notably, using aLAIR1 as a standalone intervention or combined with CAR T cell therapy demonstrates enhanced antitumor efficacy in preclinical models resistant to chemo-radiation and anti-PD-1 treatments. CONCLUSIONS We discovered a novel mechanism by which LAIR1 contributes to GBM progression. LAIR1 blockade holds promise in reversing suppressive TME and serving as a therapeutic strategy.

RTID-02. PERSONALIZED PEPTIDE VACCINATION AMONG 173 PATIENTS WITH GLIOBLASTOMA

Abstract Current treatment outcome of patients with glioblastoma (GBM) remains poor. Following standard therapy, recurrence is universal with limited survival. GBM tumors from 173 patients were analyzed for somatic mutations to generate a personalized peptide vaccine targeting tumor-specific neoepitopes. In agreement with their treating physician, patients added a personalized peptide vaccine to their treatment as an individual healing attempt. Patients were monitored from October 2015 until August 2023. We retrospectively evaluated their clinical courses and the results of their immune monitoring data. Among all patients, including 70 treated prior to progression (primary) and 103 treated after progression (recurrent), the median overall survival from first diagnosis was 31.9 months (95% CI: 25.0-36.5). Adverse events were infrequent and were predominantly grade 1 or 2. An immune response to at least one of the vaccinated peptides was detected in blood samples of 87 of 97 (90%) monitored patients. T-cell responses to vaccinated neoepitope peptides were durable in most patients. Significantly prolonged survival was observed for patients with multiple vaccine-induced immune responses (53 months) compared to those with no/low induced responses (27 months; P=0.03). Altogether, our results highlight that the application of personalized neoantigen-targeting peptide vaccine is feasible and represents a promising potential treatment option for GBM patients. This is the largest real-world observation involving GBM patients safely treated with a personalized peptide vaccine to date. This real-world observation will be translated into a clinical trial to evaluate the specific contribution of the neoantigen vaccine.

TMOD-22. MULTIREGION BIOPSIES AND CORRESPONDING NEUROSPHERE CULTURES REVEAL SPATIAL DIVERGENCE IN GLIOBLASTOMA EVOLUTION

Abstract Spatial organization of normal tissues is key for their function, yet the role of distribution of distinct clones of cancer cells in tumor is not usually studied as a potential vulnerability in cancer. This is of particular interest in glioblastoma (GBM), in which extensive intratumor heterogeneity hampers effectiveness of treatment. To better understand tumor evolution and the complexity of cancer ecosystem in GBM, we performed profiling of biopsies collected from MRI-defined regions of the tumors, spanning the surface, tumor core and deep tumor margins, which drive local recurrence after surgery. To test how spatially distinct biopsies differ in their cancer cell composition and to assess the tumor microenvironment heterogeneity, we took advantage of 5-ALA metabolite fluorescence, which accumulates in GBM cells and allows the surgeon to visualize cancer under blue light. We used it to separate cancer cells (predominantly 5-ALAhigh) from tumor microenvironment (5-ALAlow) and subjected the fractions to bulk and single cell transcriptomics, respectively. We found a gradient of cellular states associated with location in the tumor, with the deep margin samples having more 5-ALAlow cancer cells and generally less defined phenotype. We have also developed neurosphere cultures from our collection of spatially distant biopsies. Despite the selection pressure of the in vitro culture conditions, cells from different regions of the tumor adopted different morphologies, suggesting retention of some of their differential features. The differences between cultures were reflected in metabolism, proliferation, and differential sensitivity to a drug panel. Whole exome sequencing and transcriptomic analysis were also performed to identify the selection bottleneck of culture conditions, establish evolutionary trajectory of the distinct regions and neurosphere cultures, and test how strongly our cultures represent both the 5-ALAhigh and 5-ALAlow subpopulations of GBM cells. Our study identified properties associated with different spatial locations within GBM tumors, which will help uncover the cellular and microenvironmental dependencies that could lead to novel therapeutic targets for this highly heterogeneous tumor.

DNAR-09. GB13 IS A POTENT NEOADJUVANT TREATMENT OPTION FOR IL13RA2-EXPRESSING GBM AND IMPROVES STANDARD-OF-CARE THERAPY

Abstract Glioblastoma (GBM) is the most common primary adult brain cancer and is uniformly fatal. Despite maximum safe surgical resection followed by radiation and temozolomide chemotherapy, current median survival is 14-18 months. GB13 is a novel therapy for treatment of GBM and received Orphan Drug Designation by the Food and Drug Administration (FDA) for treatment of malignant gliomas. GB13 specifically targets the tumor-restricted IL13Ra2 receptor that is highly expressed in GBM, among other cancers, but not on non-cancerous cells. IL13Ra2 represents a clinically viable target for directed therapies and is being explored by numerous approaches. Previous data demonstrate that GB13 potently kills GBM and diffuse midline glioma (DMG) cells and tumors, IL13Ra2 correlates to GB13 sensitivity and, GB13 can significantly improve the cytotoxic effects of ionizing radiation (IR) when tested in models of pediatric diffuse midline glioma. In the current studies, the ability of GB13 to enhance the effects of IR were explored using IL13Ra2-positive and negative GBM models. Major findings demonstrate that neoadjuvant GB13 treatment increases cytotoxic effects of IR, enhances cell apoptosis and decreases cell viability. These results were not observed in IL13Ra2-negative settings, demonstrating the necessity for this clinical target. Mechanistically, increased cell death was caused by sustained apoptotic signaling through caspase-mediated pathways. Preclinical models support the use of GB13 prior to radiation for decreased tumor proliferation. Using multiple models, our results identify a temporal benefit for treating GBM tumors prior to IR and this should be tested in future clinical studies.

TMET-24. MULTI-OMIC CHARACTERIZATION OF THE TEMPORAL METABOLIC ADAPTATIONS UNDERPINNING THERAPY RESISTANCE AND TUMOR RECURRENCE IN GLIOBLASTOMA

Abstract Therapy resistance and tumor recurrence are barriers to achieving long-term survival in cancer patients, particularly evident in glioblastoma (GBM), known for its dismal 5-year survival rate below 10%. Despite aggressive treatment regimens, including maximal resection, alkylating chemotherapy, and radiation, the majority of GBM patients experience relapse within 7-9 months post-initial intervention. Recurrent GBM poses surgical challenges and resists subsequent treatments, leading to an almost universally fatal outcome and an average survival of 12-18 months. Metabolic networks can serve as ancient stress-protection mechanisms that operate independently of genomic changes, offering an initial line of defense against exogenous pressures. Although the role of metabolism in cancer has gained prominence, our understanding of the metabolic adaptations following chemoradiotherapy and their contributions to therapy resistance and tumor recurrence in highly aggressive GBM tumors remains incomplete. We postulate that metabolic adaptations play a pivotal role in facilitating therapy resistance and GBM recurrence. Exploring the metabolic dependencies of these persistent cells may unveil vulnerabilities for therapeutic exploitation to counteract chemoradiotherapy resistance. We established clinically relevant patient-derived in vitro and in vivo models mirroring the standard GBM treatment protocol. Using state-of-the-art single-cell and bulk multi-omics technologies, we profiled the temporal transcriptomic and metabolomic adaptations from therapy-naïve primary GBM to chemoradiotherapy-resistant recurrent tumors. Our findings revealed unique metabolic transitions that occurred during chemoradiotherapy treatment in GBM cells that were conserved in vitro and in vivo across ten genetically diverse GBM patients. In a pre-clinical animal trial, a novel metabolism-targeting combinatorial treatment regimen prevented the emergence of recurrent tumors and significantly prolonged survival compared to chemoradiotherapy alone. Altogether, our results uncover distinct metabolic adaptations mediating therapy resistance and tumor recurrence in GBM. These findings emphasize the need to consider metabolic flexibility and dependencies for effective therapeutic strategies, offering potential avenues to overcome chemoresistance in recurrent GBM tumors.

TMIC-61. DEVELOPMENT OF A NOVEL NOTCH-TARGETING ONCOLYTIC HSV FOR GBM TUMORS

Abstract Glioblastoma (GBM) is the most aggressive brain malignancy, which despite continuing worldwide efforts to develop new therapies, remains a deadly disease with limited treatment options. The NOTCH signaling pathway is often hyperactive in GBM tumors, and its activity contributes to tumor progression and resistance to treatment. The role of NOTCH signaling in tumor growth and resistance is well studied, but the significance of this pathway for immunotherapy of GBM is not very well understood. We have recently shown that NOTCH activation following virotherapy contributes to tumor regrowth and induces recruitment of monocytic MDSC cells into the tumor, which limit optimal immune activation to kill tumor cells. For this study, we have discovered a NOTCH interfering intracellular ligand (NIIL) that blocks NOTCH activation from inside glioma cells. RNA sequencing shows the expression of NIIL downregulates NOTCH signaling in vitro. To evaluate the effect of NIIL on virotherapy, we generated an oncolytic herpes simplex virus (oHSV) that also encodes for NIIL. This virus showed a slower spread in vitro but improved therapeutic efficacy in vivo in multiple models of intracranial tumors. In immune competent mice there was a significant improvement in the survival of mice treated with the NIIL-encoding virus, with a fraction of mice showing a complete response after a single treatment. Single-cell RNA sequencing analysis of immune infiltrates is ongoing to understand how NIIL impacts tumor microenvironment and anti-tumor immune therapy.

IMMU-62. INVIGORATING EFFECTOR IMMUNE CELLS WITH HIGHLY SELECTIVE IL-2R AGONISTS AND POTENTIAL SYNERGY WITH TUMOR TARGETING THERAPEUTICS FOR TREATMENT OF GLIOBLASTOMAS

Abstract Glioblastoma (GBM) is an aggressive brain tumor with a median overall survival (mOS) of 14 months. Treatment options are limited with no new approved therapies over the past 3 decades and no standard of care for a majority of patients (> 90%) who experience recurrence. GBM has a highly immune suppressive tumor microenvironment (TME) comprising of myeloid derived suppressor cells (MDSCs) and tumor associated macrophages (TAMs) that are capable of suppressing the activity of anti-cancer CD8+ T and NK cells. We have engineered highly selective IL-2 superkines (IL-2SK) that preferentially expand and activate CD8+ T and NK cells with limited effects on immune suppressive regulatory T cells (i.e., Tregs). MDNA11 is an IL-2SK -albumin fusion protein designed to increase half-life and promote tumor accumulation. MDNA223 is a bi-functional anti-PD1-IL-2SK designed to stimulate effector immune cells while preventing immune exhaustion. MDNA11 and MDNA223 extended survival of mice harboring orthotopic GBM tumors with accompanying increase in CD8+ T and NK cells within the TME. When patient-derived GBM tumor explants were treated with MDNA11 and MDNA223 ex vivo, there was clear evidence of activation among resident CD8+ T cells characterized by increased levels of intra-cellular Granzyme B responsible for tumor cell killing. There was also increased release of soluble Fas ligand and granulysin, consistent with an activated anti-tumor immune response within the TME. Ongoing studies include in depth immune profiling to further understand the mechanism of MDNA11 and MDNA223 in GBM as well as to test potential synergy with tumor targeting therapeutics and other treatment modalities capable of eliciting immunogenic cell death.

Epidemiology of glioblastoma in Pakistan: a secondary analysis of the Pakistan Brain Tumor Epidemiology Study (PBTES).

The incidence and outcomes of glioblastoma (GBM) patients in Pakistan remain unassessed owing to a lack of cancer registries and the absence of population-based studies. For any specific population-based oncological intervention, epidemiology must be studied. Therefore, this study aims to examine the epidemiological characteristics of glioblastoma patients in Pakistan, as part of a secondary analysis of a nationwide epidemiological study. Data comprising of sociodemographic, tumor and treatment characteristics of 2750 patients from the Pakistan Brain Tumor Epidemiology Study were extracted and analyzed for cases between January 1, 2019, and December 31, 2019. Chi-square tests identified outcome and treatment differences. Data analysis was performed using SPSS version 26. A total of 260 GBM cases were analyzed, with a mean diagnosis age of 45 years. Males accounted for 68.8%. Most patients were from a middle- (39.6%) or lower-income (42.7%) socioeconomic background and received care from a public institution (63.8%). GBM tumors were mainly located in the frontal lobe with similar proportions of right and left laterality. A median distance of 119km was traveled for oncological care, and the mean time to surgery from the initial radiological diagnosis was 72 days. Gross total resection was achieved in 47.3% of first surgeries, with 23 reoperations for recurrence. At the end of the study period, 33% of the GBM cohort was recorded as alive with 47% being lost to follow-up. Our analysis is the first population-based analysis of GBM in Pakistan. This epidemiologic study can serve as a basis for future research in etiology, treatment, and outcomes for glioblastoma in the Pakistani population.

TMIC-55. CHARACTERIZATION OF THERAPEUTICS EFFECTS IN GLIOBLASTOMA THROUGH INTEGRATED SPATIAL SINGLE-CELL MULTIOME PROFILING OF LONGITUDINAL SAMPLES: UNVEILING CLINICALLY RELEVANT SUBTYPES OF PATHOLOGICAL CHANGES

Abstract Glioblastoma (GBM) remains a highly malignant, intrinsically resistant and inevitably recurring brain tumor with dismal prognosis. The aggressiveness and lack of effective GBM treatments can be attributed to the highly heterogeneous and plastic nature of GBM tumor cells, which easily confer resistance to standard-of-care (SOC) therapy. While tumor progression has also been attributed to interactions with the tumor microenvironment, quantitative data describing these interactions are still largely missing. Here, we used an multiome approach combining high-dimensional, multiplexed immunohistochemistry and spatial transcriptomics to map evolutions in the spatial, single-cell tissue architecture of 357 paired adult GBM tumor samples derived from 52 patients at diagnosis (ND) and upon recurrence (REC) following SOC treatment. We mapped the spatial distribution of a multitude of GBM tumoral subtypes across this multicentric cohort, through which we identified a high level of heterogeneity defined by specific tumoral niches within and across patients and which evolved when subjected to SOC therapy. In addition, we describe the relationship of the various tumoral niches with their local immune-infiltrates, highlighting an even more immunosuppressive environment following SOC resistance. Finally, by aligning these findings to the observed genomic aberrations and the clinical data of the patients, we are now able to more precisely describe the heterogeneous landscape of glioblastoma and how it evolves under SOC treatment at spatial, single-cell resolution.

IMMU-70. GLIOBLASTOMA DERIVED IL-8 INSTRUCTS IMMUNE CELL IDENTITY IN LATERAL VENTRICLE CONTACTING TUMORS

Abstract Adult IDH-wildtype glioblastoma (GBM) is an aggressive brain tumor with no established immunotherapy. Glioblastoma tumors that contact the ventricular-subventricular zone (V-SVZ) stem cell niche have especially poor clinical outcomes (PMC5771712, PMC5262526) and abnormal immune microenvironments filled with CD32+ HLA-DRhi macrophages that have displaced resident microglia (PMC10371245). Identifying the origin and role of these CD32+ macrophages is likely critical to developing successful GBM immunotherapies. Here, we present a mechanistic origin for these CD32+ cells as M_IL-8 macrophages. In ex vivo experiments with conditioned medium from primary human tumor cells, IL-8 was sufficient and necessary for tumor cells to instruct healthy macrophages, and inhibitory antibodies to IL-8 blocked the generation of CD32+ CD163+ M_IL-8 cells. Additionally, IL-8 protein was present in GBM tumor cells in vivo and especially common in tumors contacting the V-SVZ (p<0.0001, N=192 cores from 73 patients). Surface proteins CXCR1 and CXCR2, the primary receptors for IL-8, were detected on cells that were spatially separated from IL-8+ glioblastoma cells in tumors contacting the V-SVZ. These results suggest the hypothesis that glioblastoma-cell IL-8 instructs incoming CXCR1+ hematopoietic macrophages to adopt a suppressive M_IL-8 identity in V-SVZ-contacting GBM. Abundant HLA-DR (MHC II) observed on the CD32+ M_IL-8 cells closely matched the signature phenotype of cells in human tumors (PMC10371245) and contrasted with lower surface MHC II expression typically observed on human myeloid derived suppressor cells (PMC6608074). M_IL-8 cells might especially target CD4+ helper T cells required for effective cancer immunotherapy (PMC5312823). IL-8 and CD32+ macrophages should now be explored as targets in combination with GBM immunotherapies, especially for patients whose tumors present with radiographic contact with the V-SVZ stem cell niche. Some results here have been shared in a bioRxiv preprint by the these authors (PMC10996638).

TMET-02. CELL-INTRINSIC METABOLIC PHENOTYPES IDENTIFIED IN GLIOBLASTOMA PATIENTS USING MASS SPECTROMETRY IMAGING OF 13C-LABELED GLUCOSE METABOLISM

Abstract Glioblastoma (GB) is an inherently heterogeneous and invasive primary brain cancer. Genomic and transcriptomic studies have attempted to classify GB into subtypes that predict survival and that have different therapeutic vulnerabilities. An outstanding question and technical challenge is to visualize the metabolism of a tumour within its native microenvironment and understand to what extent its metabolism is driven by the microenvironment. Patients with GB tumours were infused with [U-13C]glucose and intra-operative multi-regional tumour samples were rapidly snap frozen for analysis using mass spectrometry imaging (MSI). Tissue microenvironment was assessed using multiplexed antibodies in immunocytochemistry (IMC) of contiguous sections. Isotope tracking in human-derived neurospheres and patient-derived xenografts were used to confirm the presence of metabolic phenotypes detected in the human samples and to test the susceptibility of each phenotype to a panel of drugs targeting cellular metabolism. Three metabolic signatures were identified in patient tumours: glycolytic, oxidative and a mixed glycolytic/oxidative phenotype. The phenotypes did not correlate with microenvironmental features, including proliferation rate, immune cell infiltration, and vascularisation. The phenotypes were retained when patient-derived cells were grown in vitro, or as orthotopically implanted xenografts, and were robust to changes in oxygen concentration, demonstrating their cell intrinsic nature. The spatial extent of the regions occupied by cells displaying these distinct metabolic phenotypes are large enough to be detected using clinically applicable metabolic imaging techniques. The phenotypes also demonstrate differential drug sensitivities, suggesting imaging of these phenotypes may be used to stratify patients for personalised therapy trials. This is the first report to demonstrate high resolution imaging of metabolic fluxes in a human tumour in vivo and the presence of different metabolic phenotypes within GB that are tumour cell intrinsic and largely independent of the tumour microenvironment.

IMMU-23. NEOADJUVANT ANTI-PD-1 AND ANTI-CTLA-4 COMBINATION THERAPY INDUCED INTRATUMORAL IMMUNOLOGIC ALTERATIONS IN RECURRENT GBM PATIENTS

Abstract Glioblastoma (GBM) is a highly aggressive brain tumor that responds poorly to checkpoint blockade inhibitors (ICIs), such as anti-PD-1. We previously reported that neoadjuvant anti-PD-1 therapy activated and recruited T cells into recurrent GBM (rGBM) tumors, but also increased immune checkpoint interactions between T cells and myeloid cells. To overcome this resistance mechanism, we are currently conducting a phase 1B clinical trial of neoadjuvant anti-PD-1 (Nivolumab, BMS) +/- anti-CTLA-4 (Ipilimumab, BMS) combination therapy for rGBM patients (NCT04606316). To evaluate the local tumor microenvironment effects of these checkpoint inhibitors, we used multiplex immunofluorescence (mIF), bulk RNA sequencing, and T cell receptor sequencing to analyze the intratumoral immune profile of tumors. Using mIF, we are able to determine the density of key immune cell phenotypes within the section of tumor sample as well as the proximity between phenotypes to better understand their interactions. Patients who received the combination therapy (n=15) trended to have a higher median density of cytotoxic T cells (CD45+CD8+; median density of 27.13 cells/mm2 for combination therapy and 13.78 cells/mm2 for anti-PD-1), helper T cells (CD45+CD4+; median density of 33.73 cells/mm2 for combination therapy and 9.82 cells/mm2 for anti-PD-1), monocytes (CD14+), and HLA-DR+ macrophages (CD45+CD14+HLA-DR+) while patients treated with anti-PD-1 alone (n=16) trended to have densities similar to placebo treated patients (n=6). Patients treated with combination therapy are also trending to have a higher density of macrophages in close proximity (≤20μm) to T cells, particularly HLA-DR+ macrophages, suggesting a higher frequency of interactions within their tumor microenvironments. This data set is currently incomplete, but the final data set will be completed by the meeting. Nonetheless, these preliminary results, as well as the bulk RNA sequencing and T cell receptor sequencing, suggest distinct immunologic effects of monotherapy vs. combination neoadjuvant anti-PD-1 and anti-CTLA-4 therapy in the tumor microenvironment of rGBM patients.