Neuro-immune Mechanisms in CNS Tumors — A Living Review

Overview

CNS tumors — particularly glioblastoma (GBM) and brain metastases — exploit neural circuit integration to drive proliferation, invasion, and immune evasion through a bidirectional neuro-immune axis. Glioma cells form authentic electrochemical synapses with neurons, receive activity-regulated mitogens such as neuroligin-3 (NLGN3), and propagate oncogenic calcium signals through tumor microtube (TM) networks, collectively sustaining tumor growth while remodeling the local excitatory-inhibitory balance¹²³. Concurrently, glutamatergic and GABAergic neurotransmitter signaling shapes myeloid polarization, promotes T cell exclusion, and suppresses antitumor immunity within the tumor microenvironment (TME), creating an “immune-excluded” phenotype that limits immunotherapy efficacy⁴⁵. Disrupting these neural-immune circuits — through AMPA receptor antagonists, TM inhibitors, or combined neuromodulation and checkpoint blockade — represents an emerging therapeutic paradigm that bridges neuroscience and immuno-oncology⁶⁷. Neuroimmuno-oncology has emerged as a distinct interdisciplinary field recognizing that neural circuit integration, neurotransmitter signaling, and immune regulation are inseparable axes of CNS tumor biology that must be co-targeted for durable therapeutic benefit⁴⁸.

Glioma Neural Circuit Integration

Glioma cells integrate into host neural circuits through direct electrochemical coupling, establishing bona fide synaptic connections with neurons. Ultrastructural studies of surgically resected human GBM tissue using the Patch2MAP technique — combining patch-clamp electrophysiology with super-resolution protein imaging — confirm presynaptic vesicles in neurons and postsynaptic densities in glioma cell membranes, while simultaneously demonstrating that electrophysiological AMPA-to-NMDA receptor ratios correspond tightly to protein expression levels at neuron-to-glioma synapses [human]⁹. Postsynaptic currents are electrophysiologically detectable in tumor cells, validating functional integration into neural circuitry [human]⁹.

• Activity-dependent optogenetic stimulation of cortical neurons in patient-derived pediatric GBM xenograft models significantly promotes tumor proliferation and growth in vivo, establishing a direct causal link between neuronal firing and glioma expansion [mouse]¹.
• High functional connectivity within glioma networks, assessed by functional imaging and local field potential recordings in awake patients, correlates with shorter patient survival, positioning circuit integration as a prognostic biomarker [human]⁸.
• Glioma cells hijack activity-dependent oligodendroglial precursor cell (OPC) proliferation signals — sharing a putative cellular origin with many gliomas — through paracrine and synaptic mechanisms, with neuronal activity regulating OPC proliferation and myelination and dysregulated versions of these mechanisms promoting glioma growth [mouse]¹⁰.
• Functional imaging and awake intraoperative subcortical mapping reveal that glioma-induced circuit remodeling is frequent in language and motor networks, depending on tumor molecular parameters, and directly influences functional patient outcomes [human]⁸.
• Neuro-cancer interactions drive intratumoral heterogeneity by conferring neural-like transcriptional states and diverse cellular phenotypes — spanning neurogliomal synapses, glioma networks, and neuronal-like motility — that synergistically fuel invasion and therapy resistance [human]¹¹.
• Glioma cells resemble oligodendrocyte precursor cells in exploiting neural network-derived signals for growth, migration, and stem-like properties, while simultaneously disrupting normal neural conduction and creating a permissive hyperexcitable microenvironment for progression [human]¹².
• The tumor microenvironment both maintains malignant cells and is maintained by them through cancer stem cell (CSC) signaling mechanisms; in GBM, glioma stem cells (GSCs) integrate synaptic network signals and represent critical nodes linking neural circuit activity to immune evasion and tumor maintenance [human]¹³.
• Glioma cells acquire neural-like features through paracrine and synaptic communication, producing diverse cellular phenotypes and transcriptional states that exhibit remarkable resemblance to neural cells, providing a conceptual framework linking neuro-cancer interactions to intratumoral heterogeneity and therapeutic resistance [human]¹¹.

Neuron-Glioma Synaptic Signaling and Ion Channels

The molecular machinery of neuron-glioma synapses centers on glutamate release by neurons activating AMPA receptors (AMPARs) on tumor cells as the primary driver of oncogenic signaling, with additional contributions from NMDA receptors, GABAergic inputs, synaptogenic factors, and epigenetic mechanisms.

• Neuronal glutamate activates AMPARs on glioma cells, triggering PI3K-mTOR pathway activity and feedforward NLGN3 expression in tumor cells; NLGN3 is sufficient and necessary for activity-regulated HGG proliferation, and its expression negatively correlates with patient overall survival [mouse]¹.
• NLGN3 acts as a secreted, activity-regulated mitogen: conditioned medium from optogenetically stimulated cortical slices promotes patient-derived HGG proliferation, and this effect is abolished by NLGN3 knockdown [mouse]¹.
• AMPAR activation on glioma cells activates downstream PI3K-AKT and MAPK pathways, enhancing proliferation, invasion, and resistance to temozolomide and radiotherapy; AMPAR-driven calcium influx is a key second-messenger event [in-vitro]¹⁴.
• In diffuse midline gliomas (DMGs), altered chloride homeostasis — mediated by upregulated NKCC1 and reduced KCC2 expression — renders GABAergic signaling depolarizing and pro-proliferative rather than inhibitory, a subtype-specific oncogenic mechanism [in-vitro]¹⁵.
• Network connectivity in GBM is regulated through chloride homeostasis and GABAergic signaling alterations, as demonstrated in preclinical glioblastoma models [in-vitro]¹⁶.
• Epigenetic decoding of neuronal cues in glioma involves enhancer reprogramming, chromatin remodeling, and rewiring of 3D genome organization; transcription factors SMAD3 and PITX1 orchestrate transcriptional programs sustaining neuron-to-glioma communication downstream of both glutamatergic and GABAergic inputs [in-vitro]¹⁷.
• Multi-omics integration highlights gene regulatory networks linked to GABAergic signaling as contributors to GBM pathogenesis; in GBM specifically, GABA-related metabolic and paracrine mechanisms rather than classical synaptic input may be the dominant drivers [in-vitro]¹⁷.
• BDNF, acting alongside NLGN3 as a synaptogenic factor, further enhances glioma proliferation and synapse formation, contributing to cognitive impairment and epileptogenesis in glioma patients [human]¹⁵.
• Glioma-secreted C1QL1 binds BAI3 receptors on neighboring neurons and GBM cells, activating Rac1-mediated cytoskeleton rearrangement to prune normal synapses and expand tumor microtubes, promoting malignant synapse formation and driving infiltrative recurrence; targeted Rac1 inhibition rescues this synaptic pruning and impedes glioma recurrence [mouse]¹⁸.
• Epigenetic dysregulation in high-grade gliomas contributes not only to molecular classification but to malignant functional biology by modifying the neuron-glioma network; DNA methylation, histone modifications, chromatin remodeling, and non-coding RNAs collectively shape intratumoral heterogeneity through their influence on synaptic integration programs [human]¹⁹.
• Excitatory glutamatergic and inhibitory GABAergic signaling pathways are both hijacked by glioma and brain metastatic cells to enhance their proliferation and survival; the harmonious orchestration of these pathways is disrupted by tumor cells that exploit ionotropic (NMDAR, AMPAR, kainate) and metabotropic receptor subtypes [in-vitro]²⁰.
• Glioma-associated epilepsy (GAE) arises in part through IDH-mutation-associated alterations in tumor metabolism, increased glutamatergic activity, and neuroinflammation; IDH and mTOR signaling pathways are implicated in both epileptogenesis and tumor biology, creating a shared therapeutic opportunity [human]²¹.

Tumor Microtubes and Intra-Tumor Connectivity

Tumor microtubes (TMs) are thin, dynamic cytoplasmic protrusions that form the structural backbone of the glioma cellular network, enabling multicellular communication, treatment resistance, and integration into the neural microenvironment.

• TMs are enriched in AMPA receptors and mediate both glioma-to-glioma and glioma-to-neuron connectivity; ultrastructural and electrophysiological studies confirm their role in propagating calcium signals and action potential-like activity throughout the tumor network [in-vitro]¹⁴.
• TM-connected tumor cells exhibit markedly increased resistance to radiotherapy and surgical resection, while TM-unconnected cells show increased sensitivity to cytotoxic therapy but greater invasive capacity, establishing TM connectivity as a dual driver of resistance and invasion [mouse]²².
• PKC modulators robustly inhibit TM formation and pacemaker tumor cell-driven, TM-mediated GBM cell network communication in a combined in vitro/in vivo anti-TM drug screening approach; PKC activator TPPB combined with radiotherapy demonstrated anti-TM and antitumor effects in intravital two-photon microscopy [mouse]²².
• TPPB treatment decreases expression of tweety family member 1 (TTYH1), identified as a key driver of invasive TMs, and spatially resolved multiomics confirms network-level antitumor effects [mouse]²².
• Glioblastoma-secreted C1QL1 drives coordinated TM expansion and malignant synapse formation via BAI3-Rac1 signaling, with first-in-class Rac1 inhibition rescuing synaptic pruning and constraining TM-dependent glioma recurrence [mouse]¹⁸.
• Tumor Treating Fields (TTFields) at 200 kHz reduce GBM cell density by 85–88%, disrupt TM network interconnectivity, and reduce global calcium activity by 51–83% in primary GBM cultures; similar reductions are confirmed in patient-derived tumoroids [in-vitro]²³.
• PTPRZ1 in the neural tumor microenvironment modulates GBM cell fate: PTPRZ1 knockdown in the nTME increases the mesenchymal cell fraction, enriches EMT gene programs, and elevates TM length in co-cultured patient tumors, implicating nTME receptor tyrosine phosphatase signaling in TM plasticity [in-vitro]²⁴.
• Connexin46 levels increase in temozolomide-resistant GBM cells, and prolonged hypoxia increases both connexin46 and connexin43 levels through reduced proteasomal degradation, linking gap junction protein dynamics to microenvironmental stress and chemotherapy resistance [in-vitro]²⁵.
• Connexin43 displays paradoxical roles in GBM — acting as a tumor suppressor by reducing proliferation while promoting invasion — whereas Pannexin1 predominantly supports tumor progression and Pannexin2 exerts tumor-suppressive effects [in-vitro]²⁶.
• Mathematical modeling of TM-driven glioma regrowth after surgical resection identifies growth-inducing wound-healing mechanisms and TM-mediated proliferative advantage as the dominant re-growth drivers, with TMs providing orientational guidance from untreated tissue into the resection cavity [mouse]²⁷.
• Microglia depletion with CSF1R inhibitor PLX5622 reduces GBM cell migration and constrains TM plasticity at the far infiltration zone, establishing a functional link between microglial activity and TM-driven invasion [mouse]²⁸.
• Calcium signaling in GBM networks is sustained by rare “periodic cells” with KCa3.1 pumps and more TM connections than average; mathematical modeling demonstrates that scale-free and small-world GBM networks are more vulnerable to periodic cell elimination than to random TM damage, and that KCa3.1 inhibition can significantly impair network communication [in-vitro]²⁹.
• Pacemaker tumor cells drive TM-mediated calcium wave propagation throughout the GBM network; network topology — scale-free and small-world properties observed in vivo — determines vulnerability to targeted versus stochastic disruption strategies [in-vitro]²⁹.
• The human organoid tumor transplantation (HOTT) co-culture system recapitulates core features of patient tumor cell types and key aspects of neural cell-enriched TME gene programs, enabling identification of receptor-ligand interactions — including PTPRZ1 — that modulate TM length and mesenchymal state transitions [in-vitro]²⁴.

Neural Activity and the Immune Microenvironment

Synaptic activity within the glioma TME shapes immune infiltration, myeloid polarization, and T cell function through multiple interconnected mechanisms, ultimately fostering an immunosuppressive, circuit-driven immune exclusion phenotype.

• AMPAR signaling in glioma cells facilitates immune escape by promoting an “immune-excluded” phenotype that impairs T cell infiltration and function; AMPAR antagonism reverses this exclusion and sensitizes tumors to immunotherapy in preclinical models [mouse]⁴.
• Glutamate released by glioma cells causes excitotoxicity and alters the local excitatory-inhibitory balance, creating a hyperexcitable pro-tumorigenic microenvironment that indirectly suppresses immune surveillance by sustaining peri-tumoral neuroinflammation [in-vitro]¹⁵.
• NLGN3, beyond its mitogenic role, modulates immune cell function within the TME; neuronal-derived NLGN3 and BDNF promote an immunosuppressive TME in part by influencing myeloid polarization states and limiting effector T cell activity [human]⁴.
• Neurotransmitters GABA and glutamate, along with neuron-derived BDNF and NLGN3, modulate immune cell function and promote formation of an immunosuppressive TME, with neuronal electrical activity through AMPAR signaling being a key driver of immune evasion in glioma [mouse]⁴.
• Circuit-driven immune exclusion is linked to PD-L1 upregulation in glioma; COX-2/PGE2 signaling — enhanced by EP1 receptor activation on GBM cells through neural excitation — mediates CAR-T and checkpoint inhibitor resistance by impairing NK and T cell cytotoxic function [in-vitro]³⁰.
• PGE2, produced via the COX-2 pathway in the excitatory glioma TME, elevates intracellular cAMP through EP2/EP4 receptor Gαs signaling, downregulating NK cell activating receptors (NKG2D, NKp30), inducing CD8⁺ T cell exhaustion, and promoting Treg expansion [in-vitro]³⁰.
• Antiepileptic drugs with activity against the neural-tumor axis — including valproic acid and levetiracetam — exert direct antitumor effects through metabolic reprogramming, epigenetic regulation, and modulation of the tumor immune microenvironment, beyond their antiseizure properties [human]³¹.
• Neuronal activity and synaptic communication between neurons and brain tumor cells reshape neuronal signaling to favor tumor growth through excitatory glutamatergic and inhibitory GABAergic pathways; their dysregulation drives a pro-tumorigenic hyperexcitable milieu that simultaneously impairs immune surveillance [in-vitro]²⁰.
• The neuro-immune axis links neural signaling to immune regulation through neurotransmitter-driven modulation of T cell, macrophage, and NK cell phenotype and function; neuronal-derived molecules including BDNF and NLGN3 influence immune cell function to promote an immunosuppressive TME [mouse]³².
• Glioma peri-tumoral cortex undergoes severe alterations during disease progression — including genetic mutations, anomalous synaptic remodeling, inflammatory changes, and neurotransmitter imbalance — that together define the nidus of cortical hyperexcitability; these pathological mechanisms simultaneously provide pro-tumorigenic immune microenvironmental signals [human]³³.

Microglia, Macrophages, and Brain-Resident Myeloid Cells

Glioma-associated microglia/macrophages (GAMs) constitute the dominant immune population in the GBM TME, exhibiting profound spatiotemporal and metabolic heterogeneity that sustains immunosuppression and limits immunotherapy efficacy through direct interactions with the neural microenvironment.

• Single-cell and spatial multi-omics reveal that TAMs are not uniformly M2-polarized but encompass functionally distinct subpopulations — including lipid-associated, interferon-responsive, pro-angiogenic, and HMOX1⁺ metabolically reprogrammed subsets — with immunosuppressive TAMs enriched in the tumor core and BAMs localizing to perivascular zones [human]³⁴.
• During GBM progression, the TAM ecosystem shifts from microglia-dominated at initial stages to BMDM-dominated at recurrence, driven by dynamic niche-instructive signals including hypoxia, perivascular cues, and tumor-derived metabolites [mouse]³⁴.
• Microglia exhibit a biphasic response to GBM infiltration: increased surveillance in sparsely infiltrated areas, but reduced surveillance with higher GBM cell density; CX3CR1 deficiency enhances microglial reactivity while paradoxically limiting GBM cell migration [mouse]²⁸.
• CSF1R inhibitor PLX5622-mediated microglia depletion reduces GBM cell migration and constrains TM plasticity at the far infiltration zone, demonstrating that microglia actively modulate GBM dissemination [mouse]²⁸.
• Glioma cells manipulate TAMs to suppress anti-tumor functions through glioblastoma-derived extracellular vesicles (EVs) that reprogram microglia; microglia-mediated modulation of neuron-glioma cell interactions in turn promotes tumor progression, establishing bidirectional neuro-immune crosstalk [in-vitro]³⁵.
• Neuronal activity and neurotransmitter signaling directly influence myeloid phenotype: glutamate-driven excitotoxicity and PGE2 signaling promote M2-like TAM polarization, while AMPAR antagonism has been shown to partially reverse immunosuppressive myeloid states [mouse]⁴.
• Lactate signaling from metabolically active glioma cells — fueled partly by neural activity-driven metabolic demands — drives histone lactylation in TAMs, reinforcing their immunosuppressive phenotype and constituting a metabolic-epigenetic resistance mechanism [in-vitro]³⁶.
• In DMG/DIPG, TAMs are the predominant immune population and adopt a profoundly immunosuppressive pro-tumor state; reprogramming strategies including CSF1R inhibition, HDAC/BET inhibitors, and CD47/SIRPα axis blockade are under investigation to render the TME immunologically accessible for CAR T cell therapy [mouse]³⁷.
• Microglia-T cell crosstalk in brain cancers is bidirectionally suppressive: microglia impair T cell infiltration and function through checkpoint ligand expression and cytokine suppression, while tumor-educated T cells further limit microglial antigen-presentation capacity [human]³⁸.
• Astrocytes undergo pro-tumor state transitions driven by IL-6/STAT3, NF-κB, and TGF-β signaling, contributing to immune exclusion and invasion; IL-6 family signaling couples to metabolic rewiring and chromatin reinforcement to stabilize astrocyte pro-tumor phenotypes [human]³⁹.
• An international consensus statement on GAM function challenges the M1/M2 paradigm, emphasizing instead the cellular, spatial, and temporal heterogeneity of TAMs and their functional plasticity; glioma stem cells interact with TAMs to suppress anti-tumor functions, while microglia-mediated modulation of neuron-glioma interactions promotes tumor progression [human]³⁵.
• Myeloid-derived suppressor cells (MDSCs) in the GBM TME suppress T cell and NK cell anti-tumor functions, providing a rapid pathway for adaptive resistance to immunotherapy; interactions between MDSCs and the neural microenvironment through neurotransmitter-driven metabolic signaling represent an underexplored immunosuppressive axis [human]⁴⁰.
• Tumor-associated microglia in GBM engage in directional migration toward GBM cells restricted to defined spatial ranges, revealing heterogeneous microglial reactivity patterns that are anatomically organized relative to TM density and tumor cell invasion speed [mouse]²⁸.

CAR T and ICI Resistance Mechanisms in CNS Tumors

GBM and other CNS tumors resist CAR T cell therapy and immune checkpoint blockade through interconnected mechanisms encompassing T cell exhaustion, physical exclusion, antigen heterogeneity, and neuro-immune suppression operating within the brain’s unique anatomical and immunological environment.

• Phase III trials of anti-PD-1 monotherapy (nivolumab) have not demonstrated survival benefit in GBM, reflecting profound myeloid-driven immunosuppression, low neoantigen burden, intratumoral heterogeneity, and adaptive resistance [clinical-trial]⁴¹.
• Antigen heterogeneity and escape remain a primary barrier to CAR T efficacy in GBM: single-target CAR T approaches (e.g., EGFRvIII) show initial responses followed by antigen loss in clinical studies; multi-antigen and logic-gated CAR designs are in active development to address this [clinical-trial]⁴².
• Analysis of 23 active interventional CAR T trials for GBM commenced after January 2020 reveals a paradigm shift toward multi-antigen OR-gated approaches, conditional AND-gated synNotch logic, and armored CAR-T cells with cytokine payloads or resistance to suppressive mediators such as TGF-β [clinical-trial]⁴³.
• The myeloid-enriched GBM TME — composed predominantly of immunosuppressive GAMs — rapidly limits CAR T cell infiltration, persistence, and function; CAR T-derived inflammatory mediators (IFN-γ, TNF, Type I IFNs) can license microglia/TAMs for antigen presentation and chemokine secretion, suggesting a systems-level engagement strategy [human]⁴³.
• PGE2/COX-2 pathway activation in GBM mediates resistance to CAR T therapy by suppressing cytotoxic immunity through EP2/EP4-cAMP signaling; COX-2 inhibition combined with CAR T approaches is a proposed strategy to overcome this resistance [in-vitro]³⁰.
• Oncolytic virus OVV-03 synergizes with B7H3- or HER2-targeted CAR-T cells in orthotopic GBM models, inducing superior tumor regression and prolonged survival; OVV-03 downregulates PD-L1 by inhibiting the JNK-c-Fos/c-Jun axis and remodels the TME toward cytotoxic T cell activity [mouse]⁴⁴.
• T cell exhaustion in GBM is reinforced by the AMPAR-driven immune-excluded phenotype; neuronal activity through AMPAR signaling promotes an immunosuppressive glioma state that limits both endogenous and adoptively transferred T cell function, supporting the rationale for combining AMPAR antagonists with immunotherapy [mouse]⁴.
• Tumor-intrinsic chromatin programs enforce immune evasion in GBM by transcriptionally silencing immune recognition machinery; epigenetic modulation represents a mechanistic basis for the “cold” GBM immune phenotype that undermines checkpoint blockade [human]⁴⁵.
• STING pathway activation in GBM drives type I interferon production and immune cell infiltration, demonstrating preclinical antitumor efficacy across multiple GBM models; STING agonists are being evaluated in combination with checkpoint inhibitors and CAR T approaches to “heat up” the cold GBM TME [mouse]⁴⁶.
• CAR-γδ T cells targeting B7-H3 offer an MHC-independent cytotoxic strategy that partially circumvents antigen-presentation deficits in GBM, leveraging innate-like recognition mechanisms [in-vitro]⁴⁷.
• GBM-induced systemic immunosuppression — manifested by severe lymphopenia and depletion of effector immune cells — represents an underappreciated barrier to immunotherapy that operates independently of local TME mechanisms, further limiting CAR T cell manufacturing quality and in vivo persistence [human]⁴⁸.
• Combining PD-1 inhibitors with chemotherapy, radiotherapy, oncolytic viruses, vaccines, other immune checkpoint inhibitors, and CAR-T cell therapy shows better efficacy and safety profiles in clinical studies than monotherapy, though no combination has yet demonstrated phase III survival benefit [clinical-trial]⁴⁹.
• CAR-NK cells offer distinct advantages over CAR-T cells in GBM — including reduced risk of GvHD and cytokine release syndrome — but are limited by poor infiltration and persistence in the immunosuppressive CNS environment; emerging engineering strategies to resist TME-induced dysfunction are in development [in-vitro]⁵⁰.
• Locoregional intracranial delivery, viral vector platforms, and nanotechnology-enabled systems are emerging delivery paradigms designed to enhance BBB penetration and intratumoral CAR T retention, addressing the anatomical constraints that limit systemic cell therapy efficacy in GBM [clinical-trial]⁴².

Brain Metastasis Immune Microenvironment

Brain metastases (BrM) establish a distinct immune microenvironment that diverges from both primary CNS tumors and extracranial metastases, shaped by brain-resident myeloid populations, blood-brain barrier constraints, and unique neuro-immune interactions that collectively impair antitumor immunity.

• Brain metastases display an “immune desert” phenotype: reduced T cell infiltration relative to extracranial sites, microglia-mediated immunosuppression, and BBB-enforced limitations on immune surveillance and therapeutic delivery [human]⁵¹.
• Single-cell and spatial transcriptomic studies of BrM reveal a spectrum of TAM functional subpopulations beyond M1/M2 — including lipid-associated, interferon-responsive, and pro-angiogenic subtypes — with M2-like states prevailing to mediate immunosuppression in the brain niche [human]⁵².
• TAMs in brain metastases disrupt the BBB and facilitate immune evasion through molecules including ANGPTL4 and MMP9; breast and lung cancer brain metastases illustrate how TAMs actively remodel the brain niche to support metastatic colonization [mouse]⁵².
• Melanoma brain metastases achieve limited molecular, metabolic, and electrical similarity to neural circuitry through vascular co-optation and synaptic mimicry, reactivating neural crest-derived migration and communication pathways; this neuro-mimicry creates a metastable tumor-host system defined by stabilizing and destabilizing forces shaping therapeutic vulnerability [human]⁵³.
• B cells play multifaceted and underexplored roles in the BrM TME — contributing to antigen presentation, cytokine secretion, and tertiary lymphoid structure (TLS) formation — with their immunostimulatory versus immunoregulatory functions dependent on differentiation state and local CNS cues [human]⁵⁴.
• Lung adenocarcinoma brain metastases specifically display exhausted CD8⁺ T cells, expanded Tregs, and high PD-L1 expression overlaid on the baseline “immune desert” of the CNS niche, with microglia-mediated immunosuppression as the dominant feature distinguishing brain from other metastatic sites [human]⁵¹.
• Cycloastragenol (CAG) ameliorates radiation-induced brain injury in tumor-bearing mice with lung cancer brain metastases by attenuating pro-inflammatory microglia/macrophage polarization through JAK/STAT and IKK/NF-κB pathway suppression, demonstrating that microglial state modulation can simultaneously enhance radiotherapy efficacy and reduce neurotoxicity [mouse]⁵⁵.
• Advances in personalized immunotherapy for BrM — guided by biomarker-driven approaches and spatial immune profiling — are required to overcome the CNS immune privilege enforced by the BBB, specialized myeloid populations, and reduced lymphatic drainage [human]⁵⁶.
• Systems immunology approaches leveraging spatial and single-cell technologies reveal dynamic interplay between metastatic cancer cells and brain immunity, identifying potential immunotherapeutic targets and strategies to boost antitumor immunity within the CNS niche [human]⁵⁷.
• Breast cancer brain metastasis (BCBM) is particularly prevalent in HER2-positive and triple-negative subtypes; tumor-stroma crosstalk with astrocytes and microglia in the brain niche promotes immune evasion, therapeutic resistance, and metastatic outgrowth beyond what is seen at extracranial sites [human]⁵⁸.
• Tissue-specific macrophage heterogeneity in brain metastases from breast cancer involves distinct polarization states, metabolic characteristics, and bidirectional effects on metastatic organotropism; TAM-targeting therapies including CSF1R inhibitors are being evaluated in combination with immune checkpoint blockade in clinical trials [clinical-trial]⁵⁹.
• The intricate connections established between GBM cells and the brain parenchyma, paired with the ability of peripheral metastatic cells to form functional synapses with neurons, challenge the concept of brain tumors as disconnected from the CNS; tumor cell integration to the brain parenchyma alters brain functionality and accelerates cancer progression [human]⁶⁰.

Therapeutic Implications

AMPA Receptor Antagonists

• Perampanel, an FDA-approved AMPAR antagonist, is under clinical investigation for GBM based on preclinical evidence that AMPAR blockade reduces glioma proliferation, invasion, PI3K-AKT/MAPK pathway activation, and immune exclusion; its dual antitumor and antiseizure profile makes it a compelling candidate for repurposing [clinical-trial]⁶.
• Riluzole (glutamate release modulator) and sulfasalazine (system xCT inhibitor reducing glutamate release) are additional pharmacological strategies to reduce excitatory input into glioma cells and limit peri-tumoral hyperexcitability [in-vitro]¹⁵.
• AMPAR modulation must account for the paradox that AMPAR activation can produce excitotoxicity in tumor cells under some conditions; the therapeutic window and context-dependence of AMPAR antagonism versus agonism warrants careful examination in both primary GBM and peripheral tumors expressing AMPAR [in-vitro]⁶¹.
• Chloride cotransporter inhibitors targeting NKCC1 (bumetanide and related compounds) are proposed to reverse the depolarizing GABAergic signaling that drives proliferation in DMG/DIPG, representing a subtype-specific adjunct to standard therapy [in-vitro]¹⁵.

Anti-NLGN3 Approaches

• NLGN3 shedding from neurons is mediated by ADAM10 protease; ADAM10 inhibitors block activity-regulated NLGN3 release and abrogate its mitogenic effect on HGG in preclinical xenograft models, representing a first-in-class neuro-oncological target [mouse]¹.
• Feedforward NLGN3 expression in glioma cells, induced by soluble neuronal NLGN3, constitutes a self-amplifying oncogenic loop that could be targeted at the receptor or downstream PI3K-mTOR level [mouse]¹.

CAR T Strategies for GBM

• Multi-antigen and logic-gated CAR T designs (OR-gated, AND-gated synNotch), armored CAR T cells secreting cytokines or resistant to TGF-β, and locoregional intracranial delivery are the dominant engineering strategies in 23 active post-2020 GBM CAR T trials registered on ClinicalTrials.gov [clinical-trial]⁴³.
• CAR-γδ T cells targeting B7-H3 offer MHC-independent cytotoxicity that may bypass antigen-presentation barriers in GBM [in-vitro]⁴⁷.
• Combining CAR T therapy with oncolytic virotherapy (e.g., OVV-03) enhances TME remodeling, upregulates tumor antigen expression, and synergistically prolongs survival in orthotopic GBM models [mouse]⁴⁴.
• CAR-macrophage (CAR-M) and CAR-NK platforms are gaining prominence as alternatives or complements to CAR-T in GBM, with CAR-NK offering reduced GvHD and CRS risk while CAR-M leverages natural BBB-crossing and tumor-homing capacity [in-vitro]⁵⁰.
• In DMG/DIPG, TAM reprogramming combined with CAR T is a priority combinatorial strategy; molecular approaches including CSF1R inhibition, HDAC/BET inhibitors, and CD47/SIRPα axis blockade are under preclinical and early clinical evaluation to render the immunologically cold DMG/DIPG TME CAR T-accessible [mouse]³⁷.

ICI Combinations with Circuit Modulation

• Combining AMPAR antagonists (perampanel) with immune checkpoint inhibitors is proposed to simultaneously reverse circuit-driven immune exclusion and reinstate T cell infiltration; this neuromodulation-immunotherapy combination is supported by preclinical evidence of AMPAR-mediated immune-excluded phenotype reversal [mouse]⁴.
• Repetitive transcranial magnetic stimulation (rTMS) and deep brain stimulation (DBS) are proposed as neuromodulatory strategies to reprogram the neuro-immune-tumor axis and remodel the immune landscape in glioma, potentially synergizing with ICI [human]⁴.
• COX-2/PGE2 axis inhibition combined with NK/T cell immunotherapy is a rational strategy to overcome PGE2-mediated suppression of cytotoxic immunity in GBM [in-vitro]³⁰.
• STING agonist combinations with checkpoint blockade and CAR T are in preclinical development for GBM, leveraging type I interferon production to convert the cold TME [mouse]⁴⁶.
• PD-1 inhibitor combinations with chemotherapy, radiotherapy, oncolytic viruses, vaccines, and CAR-T show improved efficacy and safety in clinical studies; the optimal partner for neural circuit modulation remains to be established in prospective trials [clinical-trial]⁴⁹.

Microglial Reprogramming

• CSF1R inhibitor (PLX5622) approaches targeting TAM survival and recruitment are being evaluated in GBM; however, microglia depletion also paradoxically constrains TM plasticity and GBM cell migration, revealing dual effects of microglial modulation [mouse]²⁸.
• In DMG/DIPG, TAM reprogramming strategies combining CSF1R inhibition with HDAC/BET inhibitors, CD47/SIRPα axis blockade, and nanoparticle delivery are under investigation to render the immunologically “cold” TME accessible to CAR T cell therapy [mouse]³⁷.
• GBM-derived EV reprogramming of microglia represents a novel therapeutic target; bioengineered EVs are being explored as a complementary approach to deliver reprogramming signals [in-vitro]³⁵.
• Astrocyte state normalization through pharmacologic modulation and RNA therapeutics (siRNA lipid nanoparticles targeting IL-6/STAT3, NF-κB pathways) is a proposed strategy to restore protective barrier functions while limiting immune exclusion and invasion [human]³⁹.

Gap Junction Inhibitors and TM-Targeting

• Gap junction inhibitors targeting connexin43-mediated intercellular communication are in preclinical development; the paradoxical dual role of Cx43 (tumor suppressor for proliferation, promoter of invasion) necessitates careful context-dependent targeting [in-vitro]²⁶.
• PKC modulators (TPPB and related compounds) identified through the novel anti-TM screening pipeline simultaneously inhibit TM formation, network communication, and TTYH1 expression, representing a new class of neuroscience-instructed cancer therapeutics poised for clinical development [mouse]²².
• TTFields at 200 kHz demonstrate network-level disruption by reducing TM interconnectivity and calcium signaling in GBM; TTFields are already FDA-approved and their anti-TM mechanism represents an incompletely exploited therapeutic axis [in-vitro]²³.
• KCa3.1 pump inhibition is proposed as a strategy to disrupt calcium signaling in GBM networks by targeting periodic pacemaker cells; combined elimination of periodic cells and KCa3.1 blockade is predicted to maximally degrade network communication [in-vitro]²⁹.
• Rac1 inhibition using non-GEF-targeting first-in-class inhibitors rescues C1QL1-mediated synaptic pruning, simultaneously inhibiting TMs and malignant synapses to impede GBM recurrence [mouse]¹⁸.

Open Questions

1. Does AMPAR-driven immune exclusion operate through a single defined molecular pathway, and can simultaneous AMPAR antagonism plus ICI produce durable remissions in IDH-wildtype GBM clinical trials? The mechanistic link between AMPAR activation, PD-L1 upregulation, and T cell exclusion is supported by preclinical data, but no randomized trial has tested perampanel plus checkpoint inhibitor as a combination, leaving the translational magnitude entirely unknown.

2. What is the causal relationship between tumor microtube network connectivity and the spatiotemporal organization of TAM subsets? Microglia depletion constrains TM plasticity, and TMs mediate calcium-driven crosstalk, but whether specific TM-connected versus TM-unconnected tumor cell populations differentially recruit or polarize myeloid subsets — and whether disrupting this relationship improves immunotherapy response — has not been resolved.

3. How do epigenetic mechanisms (SMAD3/PITX1 enhancer reprogramming, chromatin remodeling) downstream of neuron-glioma synaptic activity specifically regulate immune evasion gene programs, and can these be pharmacologically reversed without disrupting normal neural plasticity? The epigenetic decoding of neuronal cues has been mapped in glioma cells, but the downstream effects on immune checkpoint ligand expression and myeloid licensing remain uncharacterized.

4. Do brain metastases from different primary tumor types (breast, lung, melanoma) use distinct neuro-synaptic integration mechanisms, and does the degree of neural circuit integration in BrM predict response to ICI or CAR T therapy? Melanoma BrM exhibits neural crest-derived synaptic mimicry, but whether other BrM subtypes form functional electrochemical synapses with neurons and whether this correlates with immune exclusion depth or immunotherapy resistance is unknown.

5. Can real-time monitoring of glioma network activity (calcium imaging, local field potentials, functional connectivity MRI) serve as a pharmacodynamic biomarker for neural-immune circuit-modulating therapies, and what are the minimal network disruption thresholds required to achieve measurable immune reinstatement? The correlation between functional connectivity and patient survival is established, but no prospective trial has used circuit activity metrics as a pharmacodynamic endpoint for any neural-targeting or immunomodulatory intervention.

References