Search

Background A recent phase III trial (NCT01372774) comparing use of stereotactic radiosurgery [SRS] versus whole-brain radiation therapy [WBRT] after surgical resection of a single brain metastasis revealed that declines in cognitive function were more common with WBRT than with SRS. A secondary endpoint in that trial, and the primary objective in this secondary analysis, was to identify baseline biomarkers associated with cognitive impairment after either form of radiotherapy for brain metastasis. Here we report our findings on APOE genotype and serum levels of associated proteins and their association with radiation-induced neurocognitive decline. Methods In this retrospective analysis of prospectively collected samples from a completed randomized clinical trial, patients provided blood samples every 3 months that were tested by genotyping and enzyme-linked immunosorbent assay, and results were analyzed in association with cognitive impairment. Results The APOE genotype was not associated with neurocognitive impairment at 3 months. However, low serum levels of ApoJ, ApoE, or ApoA protein (all P < .01) and higher amyloid beta (Aβ 1–42) levels (P = .048) at baseline indicated a greater likelihood of neurocognitive decline at 3 months after SRS, whereas lower ApoJ levels were associated with decline after WBRT (P = .014). Conclusions Patients with these pretreatment serum markers should be counseled about radiation-related neurocognitive decline.

Background: The development of successful immunotherapy targeting antigens in glioblastoma multiforme (GBM) remains a challenge owing to antigen heterogeneity and the low mutation burden in GBM. Cancer immunogenomics represents a complementary approach to the application of genomics in developing novel immunotherapies for GBM. Our goal is to develop a personalized RNA-based therapeutic vaccine that simultaneously targets a selected pool of tumor rejection antigens including all neoantigens and tumor-associated antigens (TAAs) as identified using cancer immunogenomics. We evaluated the treatment efficacy of these RNA therapeutics in combination with adoptive cellular therapy and checkpoint inhibitors. Methods: We identified immunogenic neoantigens and TAAs that are aberrantly overexpressed in KR158-luc and GL261 murine GBM tumors using our cancer immunogenomics pipeline called the Open Reading Frame Antigen Network (O.R.A.N.). A tumor antigen-specific RNA library was created for each tumor using a novel gene enrichment strategy and validated by RNAseq. Tumor-bearing animals were treated with adoptively transferred ex vivo expanded lymphocytes and dendritic cells loaded with the tumor antigen-specific RNA as vaccines. Additionally, we treated animals with the tumor antigen-specific RNA vaccines in combination with anti-PD-1 checkpoint-blockade therapy. Tumor progression and survival outcomes were analyzed to evaluate the treatment efficacies. We determined the peripheral and tumor-infiltrating lymphocyte responses using flow cytometry, TCR sequencing, and single cell RNAseq. Results: The tumor antigen-specific RNA vaccines were significantly effective in slowing the progression of murine GBM tumors in combination with both the adoptive cellular therapy and the checkpoint blockade therapy and improved the outcome of these therapies as compared to monotherapy alone. Additionally, we identified antigen-specific T cells targeting several of our predicted antigens and an increase in tumor-infiltrating lymphocytes and memory T cells in the treated animals. Conclusion: Here, we developed an RNA-based immunotherapy that simultaneously targets numerous tumor-specific antigens and demonstrated its therapeutic efficacy in preclinical models of GBM. Citation Format: Vrunda Trivedi, Changlin Yang, Oleg Yegorov, Kelena Klippel, Graeme Fenton, Christina von Roemeling, Lan Hoang-Minh, Elizabeth Ogando-Rivas, Paul Castillo, Kyle Dyson, Ginger Moore, Duane Mitchell. Development of personalized RNA-based immunotherapeutics for glioblastom. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5114.

INTRODUCTION Glioblastoma (GBM) contains cell populations with distinct metabolic requirements, with fast-cycling cells harnessing aerobic glycolysis, and treatment-resistant slow-cycling cells (SCCs) preferentially engaging lipid metabolism. The interaction between immune and tumor cells, and how their metabolic heterogeneity shapes the immune landscape in GBM has yet to be understood. Objectives: The primary objective of this project is to spatially and molecularly decode the GBM microenvironment heterogeneity with a specific focus on unraveling the metabolic links that underlie the interaction of SCCs with the immune compartment. METHODS Multiple murine glioma cell lines coupled with geospatial profiling were used to establish metabolic heterogeneity and communications, while various genetic and pharmacological approaches were applied to assess the effect of disrupting the metabolic interplay between SCCs and the immune system. RESULTS We determined that SCCs exhibit distinct metabolic dependencies, involving preferential lipid metabolism supported by enhanced fatty acid uptake. We also found that SCCs exhibit specific geospatial distribution and that tumor progression is regulated by their interactions with immune suppressive cells, which in turn work against tumor immune rejection by inhibiting T cell anti-tumor activity. The immune microenvironment shaped by SCCs is marked by specific metabolic features enhancing lipid exchange capacities that are exploited by SCCs to support their survival and functions. Importantly, disrupting lipid metabolic exchange sensitized tumors to chemotherapy. CONCLUSION Our results reveal that metabolic interactions between SCCs and tumor-associated immune suppressive cells within the GBM microenvironment play a critical role in the development of drug and immune resistant tumors. This study delineates these metabolic communications and assesses the potential therapeutic effect of disrupting these interactions to treat GBM. The insights generated from this project uncover fundamental principles of the emerging connections between the tumor microenvironment, cell metabolism, anti-tumor immunity, and associated therapeutic vulnerabilities.

BACKGROUND Recurrent PCNSBLs represent a therapeutic challenge. Up to 60% of PCNSBL patients relapse to later face survival rates as low as 22%. Unfortunately, tumor heterogeneity and off-target effects have limited the success of immunotherapy against PCNSBL. METHODS We propose a novel immunotherapy to overcome PCNSBL heterogeneity and off-target effects in an exquisitely tumor specific manner using nanoparticle vaccination, capable of delivering personalized tumor derived mRNAs, that induces systemic orchestration of innate and adaptive immunity. We target tumor antigens derived from the B cell receptor (i.e., heavy chain immunoglobulin - IgH) clonotypes. IgH clonotypes are hypervariable gene rearrangements clonally generated by B cells. Tumor IgH clonotypes are unique for each malignant B cell clone and hence attractive immune targets, not shared by normal B cell clones avoiding undesirable off-target effects. RESULTS RNA-NPs can reprogram tumor microenvironment while activating the innate immunity via IFN type I (i.e., IFNα) and priming of hypervariable region clonotype specific T cell responses in naïve mice. We determined the rearranged IgH sequences (predominant clone 99% and nine additional clones with frequencies < 1%) of clinically relevant inbred murine PCNSBL models (BAL17 and A20) by PCR. The number of identified clonotypes confirmed the IgH variability observed in human B cell hematological malignancies. In preliminary experiments targeting lymphoma derived single clonotypes with RNA-NPs, we showed the feasibility of priming in-vivo T cells specific against hypervariable regions after 3 weekly i.v. RNA-NPs (median IFNγ: 58 pg/ml; range: 50-70 pg/ml vs controls < 30 pg/ml; p=0.008). Targeting of clonotype RNA-NPs was associated with decreased tumor growth (p=0.04). Interestingly, we have observed tumor reactive lymphangiogenesis that communicates with regional skull bone marrow observed in 3D microscopy that might direct future routes of RNA-NP administration. CONCLUSION Our RNA-NP systemic vaccination platform can induce PCNSBL clonotype specific T cell responses, sparing normal tissues.

SIGNIFICANCE Promising clinical trials for glioblastoma are evaluating the efficacy of neoadjuvant immunotherapy in the context of recurrent tumor surgery. OBJECTIVE We investigated the effects of neoadjuvant PD1 blockade on the glioma tumor microenvironment after surgery in a clinically relevant murine model of recurrent tumor. RESULTS Using a mouse resection model of KR158 glioblastoma that we established, we show that neoadjuvant PD1 inhibition and surgery enhance mouse survival and enhance the recruitment of CD8+ and CD4+ T cells at recurrent tumor sites following bulk tumor resection when compared with surgery followed by adjuvant immunotherapy. Transcriptome and spatial genomic analyses reveal alterations in immune exhaustion and activation pathway signaling genes after neoadjuvant anti-PD1 treatment when compared with adjuvant anti-PD1-treatment or surgery alone. CONCLUSIONS These studies provide insights into the effects of neoadjuvant PD1 blockade on the tumor microenvironment after surgery and uncover additional treatment targets that could be used in combination with this neoadjuvant therapy.

BACKGROUND Although mRNA vaccines have been deployed with great success against COVID-19, unlocking this technology against glioblastoma will necessitate new lipid-nanoparticle formulations that overcome cancer tolerance and immunosuppression. OBJECTIVE/METHODS We sought to develop a novel mRNA vaccine system to make tolerogenic tumor antigens appear more dangerous through use of unmodified nucleosides (pathogen associated molecular patterns, PAMPs) and highly cationic lipid shells that elicit a systemic damage response against cancer antigens. RESULTS We developed a novel vaccine formulation that increases payload packaging of tumor amplified mRNA into multilamellar (onion-shaped) particles for systemic (intravenous) administration. We demonstrate significant immunogenicity and efficacy of multilamellar RNA-NPs in syngeneic murine models for high-grade glioma (KR158b-pp65), and diffuse midline glioma (H3K27M DMG). Remarkably, RNA-NPs significantly improve median survival outcomes of DMG bearing mice beginning therapy at endpoint (Day 35 after midline intracranial implantation). Unlike prototypical mRNA vaccines that activate endosomal toll-like receptors (i.e. TLR7), multilamellar RNA-NPs elicit immunologic response predominantly through intracellular pathogen recognition receptors (RIG-I); long-term survival benefits from RNA-NPs were completely abrogated in RIG-I knockout mice. In canines (pet dogs) with spontaneous gliomas, RNA-NPs elicit massive recruitment/activation of peripheral blood mononuclear cells (PBMCs) which correlate with their trafficking into lymphoreticular organs (in follow-up murine studies). In canines receiving neoadjuvant RNA-NPs, prior to glioma biopsy, we see significant reprogramming of the glioma microenvironment with increased gene signatures for antigen processing/presentation, interferon signaling and cytotoxicity. Upon translation into human clinical trials for glioblastoma patients (NCT04573140), RNA-NPs elicit rapid (within hours) release of cytokines (e.g. IL-1, IL-6, IL-12 TNF-α, interferons) and chemokines (e.g. MIP1α, MCP-1, IP-10), which correlate with mobilization of PBMCs and activation of dendritic cells/CD8 lymphocytes. CONCLUSION First-in-human application of systemic multilamellar RNA-NP vaccines results in significant biologic effects and rapid immunologic reprogramming.

BACKGROUND Although immunotherapy holds promise for producing durable outcomes in GBM, a major obstacle is the limited ability of cytotoxic T lymphocytes (CTLs) to home to the tumor. CTLs are recruited via long-range signaling mediated by the diffusion of chemokines present in inflammatory environments. During glioma formation, tumors manufacture immune-suppressive chemokines and cytokines that co-opt resident cells, resulting in the preferential recruitment of immune suppressor cells from the periphery. We explored a novel strategy to selectively reprogram the tumor microenvironment (TME) using focal delivery of rAAV encoding a powerful call-and-receive chemokine for CTLs: CXCL9. HYPOTHESIS: rAAV transduction of the GBM TME to express CXCL9 (rAAV6-CXCL9) will enhance CTL tumor infiltration and improve responsiveness to immunotherapy. METHODS Proteomic arrays were used to identify CTL chemokines absent in human glioma. 3D immunohistochemistry and flow cytometry was used to evaluate geospatial rAAV transgene expression and lymphocyte trafficking in vitro and in vivo. Survival studies of rAAV6-CXCL9 alone and in combination with anti-PD-1 checkpoint blockade was performed in preclinical models of GBM. Tumor immunogenicity following treatment was evaluated by scRNAseq and proteomic analysis. RESULTS We identify rAAV6 as a reliable capsid for transducing murine tumor-reactive astrocytes. Transgene expression following a single intra-tumor injection is focal, stable, and results in high levels of CXCL9 production. CXCL9 is effective at promoting tumor T-cell chemotaxis in vitro and in vivo. rAAV6-CXCL9 treatment sensitizes GBM to PD-1 blockade, improving survival in two distinct syngeneic preclinical models- an effect that is largely dependent on CD8 T-cells. CONCLUSIONS We combine the excellent safety profile of rAAV with focal delivery, enhancing exposure within the tumor compartment and mitigating systemic toxicities seen with conventional therapies. rAAV therapy was successful at stimulating the intended biological response to our encoded transgene (CXCL9): increased lymphocyte recruitment. This strategy has far-reaching potential across other immunotherapy platforms.

BACKGROUND Glioblastoma multiforme (GBM) remains a disease with debilitating survival outcomes. Cancer immunogenomics represents a complementary approach to the application of genomics in developing novel immunotherapies for brain malignancies. Using a cancer immunogenomics approach that we developed, called the Open Reading Frame Antigen Network (O.R.A.N.), we identified the immunogenic neoantigens and tumor-associated antigens (TAAs) including cancer-testis and developmental antigens, that are aberrantly overexpressed in murine models of GBM. The aim of this study is to evaluate the immunologic effects and safety of immunotherapy vaccines targeting neoantigens and TAAs in preclinical models of GBM. METHODS RNAseq and WES were performed for murine glioblastoma tumors- KR158-Luc and GL261, and the results were fed to the O.R.A.N pipeline for antigen prediction. A tumor antigen-specific RNA library was created for each tumor using a gene enrichment strategy and validated by RNAseq. Tumor-bearing animals were treated with adoptively transferred ex vivo expanded lymphocytes and dendritic cell vaccines loaded with the tumor antigen-specific RNA. Tumor volume, and thus progress, was determined using in vivo luciferase imaging technique and the survival outcomes were noted. We also evaluated the efficacy of the tumor vaccines in combination with checkpoint blockade therapy by treating tumor-bearing animals with dendritic cell vaccines loaded with the tumor antigen-specific RNA combined with an anti-PD-1 antibody. RESULTS The dendritic cell vaccines loaded with tumor antigen-specific RNA were significantly effective in slowing the progression of murine GBM tumors in combination with both the adoptive cellular therapy as well as checkpoint blockade therapy. Additionally, we identified antigen-specific T cells targeting several of our predicted antigens and an increase in tumor-infiltrating lymphocytes and memory T cells in the treated animals. CONCLUSION We developed a dendritic cell-based vaccination approach targeting neoantigens and TAAs identified as being tumor-specific and evaluated the efficacy of immunotherapy vaccines in preclinical models of GBM.

BACKGROUND In recent years, emerging evidence indicates that tumor-associated myeloid cells (TAMCs) affect cancer progression. But the molecular mechanisms and signaling pathways involved in the TAMCs’ interaction with extracellular matrix (ECM) and tumor stroma are incompletely understood. Checkpoint inhibition therapy targets immune inhibitory receptors, such as CTLA-4 and PD-1, which have become a major weapon in fighting cancer. These antibodies demonstrate apparent advantages such as being easy to use, broad applicability, and durable clinical response. Leukocyte-Associated Immunoglobulin-like Receptor 1 (LAIR1), also called CD305, a collagen-binding immunoreceptor tyrosine-based inhibition motifs (ITIM)-bearing inhibitory receptor, was identified to present on almost all immune cell populations and overexpressed in cancers of human patients. However, LAIR1’s role and regulation in solid cancers are poorly defined. OBJECTIVE To identify LAIR1 as a new class of immune checkpoints in cancers that impacts TAMCs-associated tumor immunity. Further functional investigation warrants understanding the crosstalk between LAIR1-related immune checkpoint blocking agent(s), immune micro-environment, and its underlying mechanisms for targeted therapy development. METHODS Murine GBM cell line KR158-Luciferase cell line was used to set up the mouse model. Tumor-bearing mice were administered the Anti-LAIR1 blockade and IgG, followed by IVIS imaging for tumor growth and survival were recorded. The presentation and phenotype of immune cell populations in tumors and spleens were measured through scRNA-Seq. RESULTS GBM is one of these tumors that overexpress LAIR1 based on the TCGA analysis. LAIR1 molecules express highly on macrophages, monocytes, DCs, and activated T cells, not on the naive T cells. We found that LAIR1 blockade enhanced survival in preclinical GBM models. LAIR1 blockade reduced the presentation and function of TAMCs in tumors. What's more, LAIR1 blockade provided a synergetic antitumor effect with CD70 CAR T cells. CONCLUSION This study suggests that we identified a novel immune checkpoint molecule, LAIR1, which can limit tumor progression.

Melanoma brain metastases (MBM) remain a primary driver of melanoma morbidity and mortality. It is estimated that as many as 60% of all metastatic melanoma patients will develop MBM during the course of their disease. Once found, dual checkpoint inhibitor therapy or BRAF/MEK inhibition in BRAF mutated cancers, are the best options currently available. Even with these treatments, the response to therapy is less than what is seen with disease outside the. In BRAF mutated cancers, which constitute 50% of all patients, initial response to targeted therapy is rapid but resistance to therapy develops 3 months sooner than resistance outside of the brain. Novel strategies to prevent onset of resistance or to work in concert with BRAF/MEK inhibition are needed for this treatment population. The exact mechanism of resistance to targeted therapy in MBM is not known, but mutations in EphA2, PI3K, pAKT, and re-activation of MAPK signaling through secondary mechanisms have all been identified as drivers of resistance. One pathway of MAPK reactivation is through activation of phospho-ERK (eIF2alpha kinase) which also activates the JNK pathway. Our group has previously shown that MBM express high levels of the inflammatory transcription factor IRAK-4, the terminal chaperone of the inflammatory myddosome downstream of TLR and IL1R signally. This increase in IRAK-4 results in constitutive activation of NF-KB, JNK, and MAPK. We have additionally shown the novel oral IRAK-4 inhibitor CA-4948 can reach therapeutically relevant concentrations past the blood brain barrier in both tumor-bearing and naïve mice. We additionally show CA-4948 inhibits MAPK expression in melanoma and suppresses the escape pathways pERK and JNK signaling which can contribute to the prevention of resistance to targeted therapy. This study supports further investigation into the combination potential of CA-4948 with BRAF/MEK inhibitors as triplet therapy to delay intracranial resistance.

Our lab previously published that adoptive hematopoietic stem cell (HSC) therapy administered in conjunction with aPD1 increases median survival in multiple models of CNS malignancies. Tumor-infiltrating lymphocytes (TIL) abundance is independently associated with better outcomes in many cancers. Impressively, approximately 25% of TILs are HSC-derived in early disease. Importantly, we show HSC therapy generates de novo TILs in a thymus-dependent fashion and offer evidence of altered central tolerance following aPD1 therapy, which may permit immune activation against a larger pool of tumor-associated antigens. Therapeutic HSCs were prepared from constitutive GFP-expressing (CD45.2+) and PD1KO (CD45.2+) mice and transferred into CD45.1+ hosts. Sublethal 5Gy total body irradiation was administered day 5 post-intracranial tumor implantation. The next day, 1E6 HSCs from a 1:1 – GFP:PD1KO cocktail were injected intravenously. Descriptive statistics are reported as mean ± SD. Transferred HSCs engraft, expand, differentiate, and migrate to form approximately 25% of the TIL compartment in only 15 days. We found that, 56.2% ± 20.3 of CD3+ TILs display a thymocyte-like CD4+CD8+ (DP) phenotype in early disease, and 43.8% ± 17.4 of these DP-TILs are HSC-derived. Interestingly, DP-TILs were virtually absent in advanced disease 29 days after treatment. At this stage, HSC-derived TILs most commonly fell within the single-positive (SP) TIL subsets, comprising 9.8% ± 5.6 of SP-CD8+ TILs. Notably, HSCPD1KO-derived cells were 4x more abundant amongst SP-CD8 TILs when compared to HSCGFP-derived cells (6.5% vs 1.6%, p < 0.05). SP-CD8 thymocytes were similarly enriched for HSCPD1KO progeny (41.1% vs 24.7%, p < 0.05) which may suggest PD-1 blockade directly promotes thymopoiesis. The privileged maturation of PD1-deficient HSCs is intriguingly reversed when GFP:PD1KO HSCs are co-transferred into non-tumor-bearing mice. Thus, we show the thymus is an active site where PD-1-blocking agents function, and that HSC + aPD1 therapy is uniquely positioned to leverage thymus-dependent anti-tumor immunity.

BACKGROUND PCNSL is an aggressive brain tumor accounting for 3% of all CNS malignancies and is associated with poor prognosis. Standard of care treatment includes induction with high-dose methotrexate based chemoimmunotherapy followed by consolidation with autologous stem cell transplant or whole brain radiation. However, more than half of PCNSL patients cannot tolerate this intensive therapeutic intervention. Novel treatments are urgently needed. Toll-like receptor signaling pathway via MyD88/IRAK-4 signalosome is constitutively active in PCNSL secondary to MYD88 L265P mutation and represents an excellent therapeutic target. Emavusertib (CA-4948) is an oral first-in-class small molecule inhibitor of IRAK-4 that demonstrates clinical activity in patients with systemic Non-Hodgkin’s Lymphoma, however it has not been evaluated in the CNS space. HYPOTHESIS: Emavusertib has anti-tumor activity against preclinical PCNSL. METHODS Using multiparameter immunohistochemistry (IHC) and proteomics analysis, IRAK-1, IRAK-4, and NF-κB pathway activation were examined in patient PCNSL tissues as compared to normal controls. Plasma, cerebrospinal fluid, and brain tissue were assessed via UPLC-MS/MS for emavusertib drug concentration following single oral high dose in a murine model. Survival responses in preclinical PCNSL models were measured in response emavusertib treatment, along with in vivo tumor proliferation and downstream biomarker expression by multi-parameter IHC. RESULTS Our data confirm elevated IRAK-1, IRAK-4, and NF-κB signaling in patient PCNSL. We show that emavusertib can achieve therapeutically relevant concentrations in the brain and CSF. Emavusertib impairs tumor cell proliferation in vitro and in vivo, and shows dose-dependent single-agent activity in preclinical PCNSL models. We confirm decreased ERK1/2, MAPK and NF-κB activation, indicative of downstream IRAK-4 inhibition. CONCLUSIONS Our data supports that emavusertib demonstrates anti-tumor activity in the CNS space, warranting further preclinical and clinical evaluation of this agent for the treatment of PCNSL.

2011 Background: Melanoma is a growing clinical health concern where the incidence has doubled over the last 30 years. A major driver of melanoma associated deaths is the development of brain metastases (MBM). It is estimated that 40-60% of patients with metastatic melanoma will develop MBM. Current treatment strategies have a great local control rate, but do not prevent recurrent disease, with a distant intracranial failure rate of 50% at 1 year. Novel treatments and an improved understanding of the MBM tumor microenvironment are needed. Toll-like receptor signaling is activated in cutaneous melanoma, triggering innate inflammatory activation downstream through the adapter proteins MyD88 and interleukin (IL)-1 receptor-associated kinase (IRAK-1 and -4). Increased expression of IRAK-4 stimulates transcription of multiple cellular kinases and transcription factors, and production of inflammatory chemokines and cytokines. This places IRAK-4 as a central driver of both intrinsic tumorigenicity and immune evasion in melanoma. In this study we examine IRAK-4 activation in MBM, as well as preclinical responsiveness to CA-4948, an oral first-in-class small molecule inhibitor of IRAK-4 that has demonstrated clinical activity and has an acceptable safety in cancer patients. Methods: Using multiparameter IHC and proteomics analysis IRAK-1, IRAK-4, and NF-κB pathway activation were examined in patient MBM tissues and compared to normal controls. Plasma, cerebrospinal fluid, and brain tissue were assessed via UPLC-MS/MS for CA-4948 drug concentration following single oral high dose in a murine model. Using an aggressive preclinical model of MBM, B16F10, downstream biomarker response to treatment including phospho-NF-κB, phospho-ERK1/2, and phospho-P38 was measured in control and CA-4948 treated animals using multi-parameter IHC. Survival response was also tested using the B16F10 model. Results: Our data confirm elevated IRAK-1, IRAK-4, and NF-κB signaling in patient MBM. We also demonstrate that CA-4948 is capable of achieving therapeutically relevant concentrations in the brain parenchyma, and shows single agent activity in an aggressive preclinical model of MBM. We further confirm decreased ERK1/2, MAPK and NF-κB activation in response to CA-4948 treatment. Conclusions: We present here for the first time that IRAK-4 is a strong candidate for targeted therapy in MBM. We further validate CNS penetrance of the oral IRAK-4 inhibitor CA-4948, where single agent markedly reduces the proliferative capacity of aggressive preclinical MBM, resulting in improved survival. These data warrant further investigation of CA-4948 for the treatment of melanoma brain metastases, with far reaching potential as an adjunct to standard of care.

BACKGROUND Innate immune cells comprise the majority of the immune microenvironment in glioblastoma (GBM) tumors where they are chiefly thought to foster a hospitable environment for cancer cells by regulating immune suppression and driving resistance to immunotherapy. Of these, myeloid-derived suppressor cells (MDSCs) are regarded as one of the most potent contributors to immune suppression in GBM and thus have become a focus of targeted therapy. Signal transducer and activator of transcription 3 (STAT3) is a key phenotypic regulator of MDSCs. Therefore, we sought to examine if targeted STAT3 inhibition may augment the immunogenicity of these tumors. HYPOTHESIS: Targeted STAT3 inhibition reduces GBM tumor infiltration by MDSCs. METHODS Using syngeneic murine models of GBM, we performed pharmacological inhibition analyses using a specific small molecule inhibitor of STAT3, LLL12B. Circulating numbers of immune cells were assessed in tumor bearing animals with or without concomitant focal radiation. Treated tumors were examined for immune infiltrates, and additional phenotyping analyses were performed. Therapeutic responses to LLL12B alone and in combination with immune checkpoint blockade were evaluated. RESULTS STAT3 is activated in the bone marrow of tumor-bearing animals, preferentially by Gr-1 positive granulocytic myeloid cells. Increased circulating numbers of these cells were also detected. These observations were markedly enhanced in tumor-bearing animals following cranial irradiation. Therapeutic inhibition with LLL12B could mitigate these effects, indicating a dependency on STAT3. Within the tumor compartment, granulocytic myeloid cells that successfully infiltrated following treatment demonstrated a pro-inflammatory phenotype denoted by interferon-gamma expression. Improved survival was also observed following combination treatment with LLL12B and radiation or immune checkpoint blockade. CONCLUSIONS These findings advocate a critically important role for STAT3 in regulating granulocytic myeloid cell mobilization and trafficking to GBM tumors. It further illustrates the plasticity of these cells within these tumors, which may be useful in designing successful immunotherapeutic strategies.

BACKGROUND Glioblastoma tumors (GBM) are comprised upwards of 30 percent by glioma-associated myeloid cells (GAMs). A large fraction of GAMs are myeloid-derived suppressor cells (MDSC) thought to traffic in from peripheral circulation. MDSCs foster a hospitable environment for cancer cells by regulating immune suppression and driving resistance to immunotherapy. While signal transducer and activator of transcription 3 (STAT3) is considered to be a specific marker for MDSCs and is a known player in their phenotypic polarization in cancer, we sought to closely examine its specific role in MDSC trafficking and tumor recruitment. HYPOTHESIS Targeted inhibition of STAT3 blocks MDSC from the bone marrow, mitigating their accumulation within GBM tumors. METHODS Human and murine GBM samples were analyzed for immune-specific STAT3 phosphorylation (P-STAT3). Immune cell characterization and quantification in peripheral and GBM tissue from syngeneic murine models treated with a novel STAT3 inhibitor (LLL12B) was done. Therapeutic responses to LLL12B alone and in combination with immune checkpoint blockade were evaluated. RESULTS We found P-STAT3 to be predominantly expressed by tumor-associated immune cells in both human and murine GBM, particularly MDSCs. Circulating numbers of MDSCs dramatically increase in response to tumor burden. P-STAT3 is elevated in these cells, indicating that its activation occurs prior to tumor infiltration. LLL12B blockade of STAT3 significantly reduced the number of circulating and tumor-infiltrating MDSCs. As a single agent LLL12B reduced tumor volume and extended the survival of murine GBM models. When combined with anti-PD-1 checkpoint blockade, we observed durable survival in a significant fraction of treated mice. CONCLUSIONS These findings advocate a critically important role for STAT3 in regulating MDSC mobilization and trafficking to GBM tumors. It additionally reveals a new therapeutic mechanism of action for LLL12B that may be relevant for other direct inhibitors of STAT3, influencing their clinical applicability to treat GBM.

SIGNIFICANCE Adoptive T cell therapy (ACT) has emerged as the most effective treatment against advanced malignant melanoma, eliciting remarkable objective clinical responses in up to 75% of patients with refractory metastatic disease, including within the central nervous system. Immunologic surrogate endpoints correlating with treatment outcome have been identified in these patients, with clinical responses being dependent on the migration of transferred T cells to sites of tumor growth. OBJECTIVE We investigated the biodistribution of intravenously or intraventricularly administered T cells in a murine model of glioblastoma at whole body, organ, and cellular levels. METHODS gp100-specific T cells were isolated from the spleens of pmel DsRed transgenic C57BL/6 mice and injected intravenously or intraventricularly, after in vitro expansion and activation, in murine KR158B-Luc-gp100 glioma-bearing mice. To determine transferred T cell spatial distribution, the brain, lymph nodes, heart, lungs, spleen, liver, and kidneys of mice were processed for 3D imaging using light-sheet and multiphoton imaging. ACT T cell quantification in various organs was performed ex vivo using flow cytometry, 2D optical imaging (IVIS), and magnetic particle imaging (MPI) after ferucarbotran nanoparticle transfection of T cells. T cell biodistribution was also assessed in vivo using MPI. RESULTS Following T cell intravenous injection, the spleen, liver, and lungs accounted for more than 90% of transferred T cells; the proportion of DsRed T cells in the brains was found to be very low, hovering below 1%. In contrast, most ACT T cells persisted in the tumor-bearing brains following intraventricular injections. ACT T cells mostly concentrated at the periphery of tumor masses and in proximity to blood vessels. CONCLUSIONS The success of ACT immunotherapy for brain tumors requires optimization of delivery route, dosing regimen, and enhancement of tumor-specific lymphocyte trafficking and effector functions to achieve maximal penetration and persistence at sites of invasive tumor growth.

SIGNIFICANCE New promising clinical trials for glioblastoma are evaluating the efficacy of neoadjuvant immunotherapy in the context of recurrent tumor surgery. OBJECTIVE We investigated the effects of neoadjuvant PD1 blockade on the glioma tumor microenvironment in a clinically relevant murine model of recurrent tumor. RESULTS Using an orthotopic mouse KR158 resection model of glioblastoma that we have established, we show that neoadjuvant anti-PD1 and surgery enhance animal survival and increase the recruitment of CD8+ and CD4+ T cells at recurrent tumor sites following bulk tumor resection compared to surgery followed by adjuvant immunotherapy. Transcriptome and spatial genomic analyses reveal alterations in immune exhaustion and activation pathway signaling after neoadjuvant anti-PD1treatment when compared with adjuvant anti-PD1-treatment or surgery alone. CONCLUSIONS These results provide insights into the effects of neoadjuvant PD1 blockade on the tumor microenvironment and uncover additional treatment targets.