Search What Studies Actually Found

Concise summary

Recent, high quality work shows multiple complementary routes advancing GBM immunotherapy: (1) TIME subtyping that enables patient/tumor stratification and subtype-specific combination therapies, (2) cell fate reprogramming of GBM cells into dendritic cell like APCs that reheat the tumor and synergize with PD1 blockade, (3) improved CAR and bispecific strategies (multiantigen, synNotch, armored constructs and locoregional delivery), (4) new myeloid targets (TIGIT, PI3K gamma) and tumor secreted factors (GDF15) amenable to genetic/nanoparticle targeting, (5) renewed vaccine progress including personalized neoantigen/peptide/DC platforms, and (6) ferroptosis and nanodelivery used to sensitize tumors and remodel the TME — each supported by recent preclinical and early clinical evidence cited below.

Key primary sources: TIME subtypes (2025) TIME subtypes in GBM (preprint) and iDC conversion (2024) iDC reprogramming paper.

Latest advancements in immunotherapy for glioblastoma (GBM) — evidence based, November 20, 2025

Below is a concise synthesis of the most actionable, high-quality developments from 2024–2025 papers and reviews. Each claim is explicitly cited; clinical vs preclinical status and major limitations are highlighted.

1) Tumor immune microenvironment (TIME) subtyping guides therapy selection

Large single-cell and spatial proteomic analyses identified three reproducible TIME subtypes in GBM with distinct vascular and immune features: TIME-low (immune desert, leaky vasculature), TIME-med (angiogenic but immune active) and TIME-high (heavy, immunosuppressive myeloid infiltrates with dysfunctional T cells). These subtypes show different responses to immunomodulatory combinations: TIME-med is the most immunotherapy responsive; TIME-high resists anti-angiogenic plus anti-PD-L1 but responds when myeloid PI3Kγ/δ is inhibited; CD40 agonists can paradoxically worsen TIME-high because they increase angiogenesis and myeloid suppression. The paper provides murine models that match human TIME subtypes and shares scRNA accessions for reproducibility.
Implication: stratifying patients by TIME could avoid harmful interventions and direct myeloid-targeting combinations to TIME-high patients.

Key citation: TIME subtypes in GBM (preprint)

2) In situ reprogramming: converting GBM cells into dendritic cell like APCs (iDC-APC)

One of the most novel preclinical advances is ML-directed cell fate engineering that reprograms GBM cells into DC-like antigen presenting cells (iDC-APCs). In murine models a PU.1/IRF8/ID2/BATF3 combination produced cells that cross-present tumor neoantigens, prime CD8 T cells, reshape the TME, and synergize with soluble PD1 decoys and DC vaccines to extend survival — effects absent in immunodeficient hosts, indicating immune-mediated tumor control. A human-optimized combination (PU.1/IKZF1) generates CD45+MHCII+ iDC-APCs from human GBM lines and GSCs; raw data and code (NETZEN-classic) are available (GEO). This approach converts the tumor itself to a local APC factory, potentially solving antigen presentation scarcity and intratumoral antigen heterogeneity.
Limitations: delivery, long-term safety, off-target reprogramming and potential neuroinflammation need careful translational testing.

Key citation: iDC conversion to APCs

3) CAR T and engineered cell therapies: multi antigen, armored, locoregional, and synthetic circuits

Progress in CAR T approaches for GBM focuses on three practical advances: (A) multiantigen and synNotch logic to address antigen heterogeneity and minimize off-tumor toxicity; (B) armored CARs and cytokine/chemokine secretion to resist the suppressive TME; and (C) locoregional delivery and improved manufacturing (point of care, allogeneic, in vivo engineering) to increase CNS exposure and speed. Reviews and 2025 analyses report promising intracranial responses in early trials but persistent antigen escape and T-cell exhaustion remain central problems. New synthetic ideas (elapsed-time circuits, synNotch cascades) are being proposed to dynamically reprogram CAR specificity in vivo but remain conceptual or early preclinical.
Caveat: most clinical CAR-T GBM work has small cohorts and transient responses; safety and durable persistence are unresolved.

Representative review citation: CAR T review 2025

4) Myeloid reprogramming and novel myeloid targets (TIGIT, PI3Kgamma, S100A4 etc.)

GBM TME is myeloid-dominant; targeting myeloid suppression is emerging as a priority. A 2025 study shows TIGIT expression on myeloid cells (induced by tumor EVs) drives immunosuppressive polarization via NLRP3-related pathways; TIGIT knockdown reverts suppression and improves T cell function in vitro. The TIME subtype work also shows PI3Kγ/δ inhibition can convert TIME-high tumors to be more responsive. These myeloid-directed strategies are attractive because T cell–focussed therapies often fail due to persistent myeloid suppression.

Key preclinical citation: TIGIT in GBM myeloid cells

5) Genetic and nanoparticle targeting of immunosuppressive tumor factors (GDF15, CRISPR delivery)

Targeting tumor-derived secreted factors is practical: 2025 work using tumor microenvironment responsive CRISPR-Cas9 nanoparticles to edit GDF15 in murine GBM models reduced tumor growth and boosted T cell activation and survival. This shows nanoparticles can be engineered for tumor selective gene editing to remove suppressive signals — promising but still preclinical and requiring off-target and immunogenicity safety assessment.

Key citation: GDF15 CRISPR nanoparticle study

6) Vaccines: personalized neoantigen, multipeptide, and DC platforms

Vaccine strategies remain a major area: dendritic cell vaccines, multi-peptide and neoantigen approaches (including microbiome-mimic peptides) have shown safety and some signals of improved PFS/OS in subgroups. Personalized peptide vaccines using tumor sequencing show promising PFS/OS in small phase I trials. EO2401 (microbiome homology peptides) and other combination vaccine trials with checkpoint blockade and anti-angiogenic agents have interim signals; large randomized evidence remains limited and variable across platforms.

Comprehensive review citation: GBM vaccine review 2025

7) Oncolytic viruses and virotherapy combinations

Oncolytic viruses continue to mature (e.g. G47Δ licensed in Japan for malignant glioma) and are being tested in combination with immune checkpoint inhibitors and myeloid-targeted agents; they provide intratumoral lysis plus antigen release — essentially an in situ vaccine. Clinical signals are variable but promising when combined with agents that neutralize myeloid suppression or checkpoint blockade.

Representative review citation: Oncolytic virotherapy review

8) Ferroptosis and immunotherapy cross-talk — new adjuvant strategies

Reviews from 2025 position ferroptosis induction (GPX4, SLC7A11, FSP1 modulation) as a strategy to kill resistant GBM cells and to reshape the TME to improve immunotherapy efficacy. Nanoparticle delivery to overcome the BBB and combination scheduling with TMZ/radiation/ICB are actively explored. Key challenges include tumor heterogeneity, toxicity, and finding predictive biomarkers.

Key review citation: Ferroptosis in GBM review

9) Biomarkers and AI for patient selection (radiogenomics, multiomics)

Strong recent work uses multiomics and radiogenomic AI to predict immune states non-invasively and to prioritize patients for immunotherapy trials (radiomic signatures that predict T cell, TAM, DC scores). Multiomics plus AI is also used to select neoantigens for vaccines in other cancers, a paradigm applicable to GBM personalized vaccines.

Examples: radiogenomic predictive models and multiomics neoantigen selection reviews support prospect of preoperative stratification of TIME to guide neoadjuvant immunotherapy selection in GBM trials (see multiple Neuro-Oncology abstracts and MR/AI work in 2025). Representative recent work summarized in bibliometric and method papers (2025) underscores the field movement toward biomarker-driven selection.

Visual summary graphs

Interactive dataset and links

• TIME subtyping study and dataset: TIME subtyping study
• iDC conversion GEO and code: iDC GEO and code
• Preclinical targeting of TIGIT on myeloid cells: TIGIT myeloid GBM study

Limitations, biases, and missing information

• Most positive translational signals remain preclinical or early-phase clinical; randomized, phase III level proof of durable survival benefit for immunotherapy in GBM is still limited Vaccine review cautions.
• Publication and sponsor biases, small cohort sizes, and heterogeneous endpoints are frequent; sample sizes for many mechanistic human studies are small and require external validation.
• BBB penetration and reliable, safe delivery (viral vectors, nanoparticles, CED, FUS) remain key translational hurdles; see pharmacokinetic review for a PK framing CNS PK and delivery review.

Where would new evidence change the picture?

Large randomized trials showing OS benefit for any of these modalities (eg vaccine + ICB, CAR T + myeloid targeting, iDC conversion + PD1) would move therapies to standard care; conversely, negative randomized trials stratified by TIME would falsify some proposed subtype-guided strategies. Safety signals (neuroinflammation, off-target editing, severe CRS/ICANS) would materially constrain translational paths.

Practical recommendations for researchers and trial designers

1. Stratify trials for TIME or use validated radiogenomic classifiers to preselect likely responders to ICB/anti-angiogenic combos (TIME subtyping).
2. Combine myeloid-targeting agents (eg PI3Kgamma inhibitors, TIGIT-blockade strategies) with T cell–centric therapies rather than T cell therapies alone for myeloid-rich GBMs TIGIT myeloid study.
3. Prioritize delivery research (CED, focused ultrasound, nanoparticles) parallel to immunotherapy discovery — BBB remains the dominant translational barrier CNS PK review.
4. Design early phase combination trials to include robust immune and CSF pharmacodynamic monitoring and standardized steroid-sparing management to avoid confounding immune readouts CAR T clinical insight review.

Tools, datasets and resources referenced

• TIME paper accessions and murine GEO (GSE301073) and public human scRNA accessions listed in the TIME preprint methods TIME datasets
• iDC conversion GEO: GSE270855, GSE270857 and GitHub NETZEN-classic for CFD code and pipelines iDC GEO and code

Selected reference citations (evidence snippets)

1. TIME subtyping in GBM
2. iDC conversion
3. CAR T review
4. TIGIT myeloid GBM
5. GDF15 CRISPR nanoparticles
6. GBM vaccine review
7. Ferroptosis GBM review

If you want a full bioinformatic re-analysis of any public dataset above or a custom figure reproduced exactly from a paper, click Run AI Biology Analysis below.