Late 2025 Clinical & Translational Update — EGFRvIII-Targeted CAR-T, IDH Inhibitors, Tumor Treating Fields, Oncolytic Virotherapy, Blood-Brain Barrier Drug Delivery, Neoantigen Vaccines & Liquid Biopsy Strategies
Glioblastoma (GBM, WHO Grade 4 astrocytoma) remains the most lethal primary brain tumor in adults. Despite four decades of incremental progress, median overall survival has improved only modestly — from approximately 12 months with radiotherapy alone to roughly 15 months with the addition of temozolomide (STUPP protocol, 2005). The defining biological challenges are profound intra-tumoral heterogeneity, a uniquely immunosuppressive tumor microenvironment, near-impenetrable blood-brain barrier (BBB) pharmacokinetics, and near-universal resistance mechanisms driven by EGFR amplification, PTEN loss, and adaptive stem-cell reprogramming. However, Q4 2025 marks a genuine inflection: EGFRvIII-directed CAR-T cells have demonstrated in-tumor activity for the first time, IDH-mutant glioma treatment is being redefined by ivosidenib and olutasidenib, and multi-antigen vaccine combinations are entering pivotal trials building on mRNA platform momentum. The field is no longer asking whether GBM can be targeted — it is engineering the tools required to do so across the BBB.
Temozolomide · Radiotherapy · Bevacizumab · Tumor Treating Fields
Ivosidenib (IDH1) · Olutasidenib (IDH1) · Enasidenib (IDH2)
GCT02 CAR-T · CARv3-TEAM-E · AMG 596 (EGFR×CD3) · EGFRvIII ADC
DNX-2401 · G207 · CAN-3110 · Oncolytic HSV · PVSRIPO
mRNA-4157 (autogene cevumeran) · SurVaxM · GBM6-AD · Poly-ICLC primed vaccines
CED · Focused Ultrasound · Nanoparticles · Antibody-Drug Conjugates
Updated INDIGO data presented at SNO 2025 confirm sustained PFS benefit of ivosidenib in IDH1-mutant non-enhancing Grade 2 glioma post-radiotherapy. FDA approved May 2024 — the first molecularly targeted approval for low-grade glioma. Median PFS 33.1 vs 11.1 months (HR 0.35). Importantly, ivosidenib delays radiographic progression without the toxicity burden of early chemotherapy — offering a tolerable oral treatment option for a predominantly young, high-functioning patient population. OS data remain immature but trend favorably.
Landmark case series from City of Hope (NEJM 2024; updated SNO 2025): intra-tumoral + intraventricular infusion of tandem CAR-T cells targeting EGFRvIII and IL13Ra2, co-expressing a secreted anti-EGFR bispecific T-cell engager. First patient achieved near-complete radiographic response sustained beyond 6 months — unprecedented in recurrent GBM. Multi-antigen approach directly addresses the EGFRvIII antigen escape problem. Expanded Phase 1 cohort data presented Q3 2025: 3/7 patients with objective responses; all responders showed multi-antigen simultaneous targeting.
SurVaxM targets survivin — an anti-apoptotic protein overexpressed in GBM. Phase 2 in newly diagnosed MGMT-methylated GBM in combination with standard TMZ and radiotherapy. Median OS 25.9 months — approximately 10 months above historical controls on STUPP protocol alone in MGMT-methylated GBM. Two-year OS rate 41%. Vaccine-specific T-cell responses correlated with prolonged survival. Phase 3 (SURVIVE) now enrolling as of Q4 2025. If confirmed, SurVaxM would represent the first positive Phase 3 immunotherapy trial in newly diagnosed GBM.
PVSRIPO intra-tumoral infusion in recurrent GBM (Duke University): 21% OS at 36 months vs ~4% historical controls. Long-term survivor plateau confirmed — approximately 20% of patients appear to achieve durable disease control beyond 24 months. Updated Q3 2025 data show the survival plateau is maintained at 5-year analysis, identifying a subset of potential long-term survivors. PVSRIPO + ipilimumab combination Phase 2 now active — rationale based on PVSRIPO-mediated innate immune activation priming for CTLA-4 blockade.
Implantable focused ultrasound device (SonoCloud-9) for repeat, reversible BBB opening enables systemic carboplatin to achieve therapeutic intratumoral concentrations otherwise unachievable. Phase 2 data presented at AANS 2025: 5–7-fold increase in CNS carboplatin exposure versus systemic delivery alone; median OS 10.4 months in heavily pretreated recurrent GBM — a meaningful signal in a setting with historical median OS of ~7 months. Phase 2b expansion ongoing; combination with temozolomide and bevacizumab evaluating additive BBB opening strategies.
EF-14 five-year follow-up (ASCO 2025): TTFields + TMZ achieves 13% 5-year OS vs 5% with TMZ alone — identifying a clinically meaningful long-term survivor subgroup. Compliance (>75% daily use) strongly correlated with survival benefit; adherence support programs critical. EF-32 (Phase 3 recurrent GBM): TTFields + pembrolizumab vs pembrolizumab alone — interim data Q3 2025 show PFS HR 0.72; OS data immature. If positive, EF-32 would establish TTFields + checkpoint inhibitor as a recurrent GBM standard.
The BBB excludes >98% of systemically administered small molecules and virtually all large-molecule biologics from achieving therapeutic CNS concentrations. Nearly every failed GBM trial can be partly attributed to inadequate intratumoral drug delivery. Focused ultrasound BBB opening, convection-enhanced delivery, intranasal delivery, and intrathecal/intraventricular administration are the leading strategies to circumvent this barrier — each with distinct feasibility, safety, and coverage profiles requiring tumor location-specific selection.
GBM is one of the most heterogeneous tumors in oncology — with spatially distinct clonal subpopulations expressing different molecular profiles across a single tumor. Single-antigen targeting (EGFRvIII, IL13Ra2) fails due to antigen loss under selective pressure in ~50% of treated patients. Multi-antigen CAR-T constructs (CARv3-TEAM-E), neoantigen vaccines spanning multiple tumor subclones, and pan-EGFR targeting strategies directly address this fundamental challenge.
The GBM TME is dominated by tumor-associated microglia/macrophages (TAMs, ~30–50% of tumor mass), regulatory T-cells, myeloid-derived suppressor cells, and high TGF-beta/IL-10 cytokine secretion. Checkpoint blockade monotherapy (nivolumab, pembrolizumab) has consistently failed in GBM — CheckMate 143, CheckMate 498, and KEYNOTE-589 all negative. TME reprogramming via CD47 "don't-eat-me" signal blockade, CSF1R inhibition (microglial depletion), and IDO inhibition are under investigation as sensitization strategies.
Temozolomide + radiotherapy induces radiographic pseudoprogression in ~20–30% of patients — indistinguishable from true progression on conventional MRI. Misclassification as true progression leads to premature treatment discontinuation in patients who may be deriving benefit. Advanced imaging (MR spectroscopy, perfusion MRI, amino acid PET with FET or FDOPA) and CSF ctDNA monitoring are improving pseudoprogression vs true progression discrimination — but no gold-standard non-invasive tool has achieved broad clinical implementation.
MGMT-unmethylated GBM (~60% of newly diagnosed cases) has a median OS of ~13–15 months — the lowest of any major cancer subtype. TMZ provides minimal benefit; no alternative systemic therapy has demonstrated superiority to TMZ in this population. TTFields provides a modest but real benefit regardless of methylation status. MGMT-unmethylated patients represent the most urgent target population for novel trials: KRAS/RTK pathway inhibitors, oncolytic viruses, and TME-targeting approaches are being specifically designed for this population.
Median GBM diagnosis age is 64 — yet most clinical trials enrich for ECOG 0–1 patients under 70. Elderly and/or frail patients (ECOG 2–3, age >70) represent ~40% of real-world GBM diagnoses but are underrepresented in trials. Short-course hypofractionated radiotherapy (40 Gy/15 fractions or 25 Gy/5 fractions) ± temozolomide (MGMT-methylated: TMZ benefit confirmed; unmethylated: RT alone) are validated for this population. TTFields feasibility in elderly patients is being studied in dedicated frailty-stratified sub-studies.
Fit patients (ECOG 0–1, age <70), newly diagnosed GBM: Maximum safe surgical resection → concurrent radiotherapy (60 Gy/30 fractions) + temozolomide → adjuvant temozolomide × 6 cycles (STUPP). Add TTFields (Optune) throughout adjuvant phase regardless of MGMT status (EF-14 benefit confirmed). Test all tumors for MGMT methylation, IDH1/2 mutation, EGFR amplification/EGFRvIII, TERT promoter, 1p/19q, and BRAF V600E at initial surgical resection. Enroll in clinical trials incorporating neoantigen vaccines (autogene cevumeran, SurVaxM SURVIVE) or TME-targeting combinations where available.
MGMT-methylated GBM (newly diagnosed or recurrent): Temozolomide benefit confirmed — full STUPP protocol standard. At recurrence post-TMZ: rechallenge with TMZ if extended treatment-free interval (>6 months); bevacizumab for steroid-sparing/symptom management; consider lomustine. SurVaxM (SURVIVE Phase 3 — enroll). IDH-mutant Grade 2–3 glioma (MGMT-methylated subset): ivosidenib maintenance post-radiotherapy (INDIGO — FDA approved). Olaparib + TMZ combination (MGMT-methylated GBM) — Phase 2 data ongoing; rationale based on PARP trapping + TMZ DNA damage synergy.
First recurrence post-STUPP + TTFields: Bevacizumab ± lomustine (modest PFS benefit; no OS); re-irradiation (selected cases, <6 months from prior RT); TTFields continuation or re-initiation. EGFRvIII+ recurrent GBM: Enroll on CARv3-TEAM-E or GCT02 CAR-T trials where available. Oncolytic virus (PVSRIPO, DNX-2401) + checkpoint inhibitor combinations: clinical trial priority. Focused ultrasound + carboplatin (SONOCLOUD platform) in centers with FUS capability. Bevacizumab alone if functional status decline and palliation-oriented goals.
IDH1-mutant Grade 2 glioma (previously treated): Ivosidenib maintenance (FDA approved 2024, INDIGO) — oral, well-tolerated. IDH1-mutant recurrent glioma (further lines): Olutasidenib; IDH inhibitor + temozolomide combinations under study. BRAF V600E-mutant GBM / pediatric HGG: Dabrafenib + trametinib (FDA approved pediatric HGG; adult use per ROAR trial data). MSI-H (<1% GBM): Pembrolizumab. NTRK fusion (rare): Larotrectinib/entrectinib. CSF ctDNA-positive post-surgery: consider adjuvant intensification protocol on trial.
Age >70 or ECOG 2, newly diagnosed GBM: Short-course hypofractionated RT (40 Gy/15 or 25 Gy/5 fractions). MGMT-methylated: add temozolomide concurrent + adjuvant (NORDIC/NOA-08 data). MGMT-unmethylated: RT alone (TMZ adds toxicity without survival benefit in elderly unmethylated). TTFields feasibility in elderly being evaluated; use when functional status and compliance support it. Enroll in frailty-stratified clinical trials — historically excluded population with significant unmet need.
Post-STUPP + post-bevacizumab failure: Clinical trial enrollment is the unconditional priority. Investigational options available at NCI-designated centers: PVSRIPO + ipilimumab; autogene cevumeran mRNA neoantigen vaccine Phase 2; CARv3-TEAM-E or GCT02 CAR-T (EGFRvIII+); AMG 596 bispecific; SONOCLOUD-9 FUS-enabled BBB opening; LNP-delivered intratumoral mRNA platforms; CSF1R inhibitor (pexidartinib) for microglial reprogramming. Best supportive care and palliative radiotherapy for ECOG 3–4. Hospice referral discussions should be initiated proactively.
Glioblastoma treatment is undergoing its most consequential scientific evolution in two decades — driven not by incremental chemotherapy optimization but by fundamentally new therapeutic modalities targeting GBM's unique biology. The CARv3-TEAM-E tandem CAR-T data represent the most compelling single-patient proof of concept in GBM history: a near-complete radiographic response in a disease where single-digit percentage objective response rates have been the norm. The multi-antigen design — simultaneously targeting EGFRvIII and IL13Ra2 while secreting an anti-EGFR bispecific — directly addresses antigen escape and is the template for the next generation of CNS cellular therapy.
Simultaneously, ivosidenib's approval in IDH-mutant Grade 2 glioma has established that molecularly targeted oral therapy can replace or delay the toxicity of chemotherapy in a predominantly young patient population — a paradigm shift with profound quality-of-life implications. The INDIGO HR of 0.35 is among the strongest PFS signals ever reported in a glioma trial.
The unifying scientific bottleneck remains the blood-brain barrier — and focused ultrasound BBB opening is emerging as the most clinically scalable solution. SONOCLOUD-9's 5–7-fold CNS drug exposure amplification demonstrates that pharmacokinetic barriers previously considered insurmountable are now tractable engineering problems. If BBB opening can be safely combined with CAR-T or neoantigen vaccines, the combinatorial therapeutic space expands dramatically.
Key trials to watch: SURVIVE (SurVaxM Phase 3), CARv3-TEAM-E Phase 1 expansion, autogene cevumeran Phase 2 GBM cohort, EF-32 OS data, SONOCLOUD-9 Phase 2b, PVSRIPO + ipilimumab Phase 2, olutasidenib recurrent IDH-mutant glioma, and GCT02 pan-EGFR CAR-T Phase 1 at SNO 2025 and ASCO 2026.