The Blood-Brain Barrier as a Therapeutic Impasse
Glioblastoma, IDH-wildtype, CNS WHO grade 4, and other malignant gliomas remain among oncology's most formidable challenges, with median overall survival for newly diagnosed glioblastoma generally limited despite maximal surgical resection, radiotherapy, and temozolomide chemotherapy 3. Recurrence exceeds 90%, driven in large part by a biological constraint that renders systemic pharmacotherapy largely ineffective: the blood-brain barrier (BBB).
The BBB is not a simple membrane but a dynamic neurovascular unit (NVU) comprising specialized endothelial cells interconnected by tight junctions—formed by claudins (especially claudin-5), occludins, and junctional adhesion molecules anchored by zonula occludens scaffolding proteins—reinforced by surrounding pericytes and astrocytes 1. This architecture maintains highly restrictive paracellular permeability, allowing mainly small lipophilic molecules to diffuse passively while limiting the entry of most hydrophilic compounds, macromolecules, and charged agents unless they use specific transport pathways. Compounding this structural exclusion, active efflux transporters—particularly ABCB1 (P-glycoprotein) and ABCG2 (breast cancer resistance protein)—actively expel a broad range of xenobiotics, including standard cytotoxics, from the CNS compartment 1. As a result, most monoclonal antibodies and many targeted agents have limited CNS and intratumoral exposure after systemic administration, although delivery can vary by agent, tumor-region barrier disruption, and use of delivery-enhancing strategies.
Within glioblastoma, the blood-tumor barrier (BTB) introduces an additional layer of complexity. Tumor-associated vasculature exhibits marked regional heterogeneity: while some vessels retain claudin-5 and occludin expression, others display substantial reductions or complete loss of tight junction proteins 2. VEGF produced by tumor and stromal cells simultaneously promotes aberrant angiogenesis and disrupts tight junction integrity via claudin-5 degradation and occludin internalization, while hypoxia-inducible factor-1 (HIF-1) drives further BTB remodeling. The practical consequence is a spatially heterogeneous barrier—leaky near the tumor core but often intact at the infiltrative margin—that creates therapeutic sanctuary sites where systemically delivered drugs cannot accumulate at tumoricidal concentrations, even as intracranial pressure and edema complicate drug distribution. Tumor molecular heterogeneity, including MGMT promoter methylation, EGFR amplification, TERT promoter alterations, and variable transcriptional states such as proneural, classical, and mesenchymal programs, further undermines the assumption that any single systemically delivered agent can uniformly address glioblastoma; IDH-mutant grade 4 astrocytoma is discussed as a distinct entity under the current WHO classification 3.
Nanocarrier-Based Delivery: Platforms, Mechanisms, and Translational Realities
Nanocarrier engineering has emerged as the predominant systemic strategy for overcoming BBB constraints. Multiple platforms are under active investigation, each with distinct mechanisms of BBB penetration, tumor targeting, and controlled release.
| Platform | Key Mechanism of BBB Penetration | Representative Recent Example |
|---|---|---|
| Polymeric nanoparticles (PLGA) | Encapsulation protection and passive accumulation in regions with disrupted BTB/EPR-like permeability; BBB penetration remains limited without active targeting or barrier-modulating strategies | PLGA-nintedanib NPs inhibiting autophagy via VPS18 5 |
| Lipid nanoparticles (LNPs) | Receptor-mediated transcytosis (LRP1 via Angiopep-2) | Angiopep-2-LNPs delivering PLK1-siRNA; 2.18-fold survival extension 7 |
| Liposomes (layer-by-layer) | CED-assisted tumor penetration and retention | LbL-MMAF liposomal NPs via CED; superior survival vs. free drug 6 |
| Peptide-functionalized LNPs | Peptide ligands targeting brain endothelial and/or neuronal receptors, such as RVG29–nicotinic acetylcholine receptor, T7–transferrin receptor, AP2/Angiopep-2–LRP1, and mApoE–LDLR | Enhanced brain/neuronal mRNA transfection and reduced hepatic delivery in non-tumor brain-delivery models 14 |
| Lactoferrin NPs | LRP-1-mediated endocytosis at BBB | LF-V4 NPs inducing autophagy and ferroptosis in U87-MG cells 20 |
| Exosome-mimetic biomimetic NPs | Immune evasion + Angiopep-2 receptor-mediated transcytosis | Ru/Pt-TiOx NPs wrapped in macrophage exosome membranes; photothermal DOX release 22 |
| Trimethyl chitosan (TRIOZAN) NPs | Absorptive-mediated transcytosis via ionic interactions | IGF-Trap delivery; ~90% brain signal retention at 24h vs. ≤50% for free protein 29 |
| Gold nanorod/lipid hybrid NPs | Transferrin receptor-mediated BBB penetration | Transferrin-functionalized HNPs; 78% tumor volume inhibition in orthotopic GBM 24 |
Receptor-mediated transcytosis represents the most mechanistically rational approach for systemic BBB penetration. Angiopep-2-conjugated LNPs targeting LRP1 achieved approximately 2.23% injected dose accumulation in the brain and extended median survival 2.18-fold in a mouse GBM model through PLK1-siRNA silencing 7. Similarly, lactoferrin-vanadium nanoparticles exploit LRP-1 binding to enter brain endothelial cells, subsequently inducing mitochondrial ROS generation, GSH depletion, and GPX4 downregulation to trigger ferroptosis in glioblastoma cells 20. Biomimetic exosome-membrane engineering addresses the parallel challenge of opsonization and reticuloendothelial system clearance, providing immune evasion alongside active targeting 22.
Intranasal delivery represents a non-invasive alternative route, exploiting olfactory and trigeminal nerve pathways to bypass systemic circulation. Platforms including polymeric nanoparticles, liposomes, solid lipid nanoparticles, and self-assembled micelles have been evaluated for nose-to-brain TMZ delivery, demonstrating improved brain retention and reduced systemic toxicity in preclinical models 23. However, limitations in targeting specificity, restricted drug loading capacity, mucociliary clearance, and heterogeneous mucosal absorption have collectively slowed clinical translation 12.
Critical translational barriers apply across all nanocarrier platforms: manufacturing scalability and batch-to-batch consistency, protein corona formation reducing targeting fidelity, immunogenicity, and the persistent challenge that preclinical mouse models inadequately recapitulate human BTB heterogeneity. These remain the dominant obstacles separating promising preclinical signals from meaningful clinical outcomes 15.
Convection-Enhanced Delivery: Principles, Evidence, and Limitations
Convection-enhanced delivery circumvents BBB constraints entirely by infusing therapeutics directly into brain parenchyma via stereotactically placed catheters, using positive pressure to drive bulk flow through interstitial spaces well beyond the limits of passive diffusion 1. The volume of distribution to volume of infusion ratio (Vd/Vi), catheter design, infusion rate, and proximity to low-resistance structures such as white matter tracts all critically determine drug distribution.
Recent clinical trial evidence underscores both the promise and the procedural complexity of CED. A multicenter Phase 1 trial of rhenium-186 nanoliposomes (186RNL, mean diameter ≤130 nm) delivered via CED in 21 patients with recurrent glioma demonstrated no dose-limiting toxicities across six escalating cohorts (1–22.3 mCi). Patients achieving ≥100 Gy absorbed tumor dose demonstrated median overall survival of 17 months versus 6 months for those receiving lower doses, with median progression-free survival of 6 versus 2 months respectively, suggesting an exploratory dose-response association that requires confirmation in larger controlled studies 10. In pediatric diffuse intrinsic pontine glioma (DIPG)—a tumor both inaccessible to surgery and refractory to systemic chemotherapy—a multicenter Phase 2 trial of CED-administered nimustine hydrochloride (ACNU) achieved a 1-year survival rate of 55.0% (95% CI 31.3–73.5%), significantly exceeding the prespecified historical threshold of 30%, with median overall survival of 15 months from radiotherapy initiation 11. A separate Phase 1 trial of 124I-Omburtamab, a radiolabeled anti-B7-H3 antibody delivered via CED to 50 pediatric DIPG patients, achieved a median overall survival of 15.29 months with a mean lesion-to-whole-body absorbed dose ratio of 816, reflecting exceptional tumor selectivity 30.
Procedural optimization remains an active frontier. A Phase 1 trial of MRI-guided CED of liposomal irinotecan in 18 patients with recurrent high-grade glioma demonstrated that personalized volume dosing—adapting infusion volume to tumor size during the procedure—improved mean tumor coverage to 41–95% in later cohorts, compared to 12–58% under fixed-volume protocols 28. Real-time MRI guidance proved critical for detecting catheter backflow and enabling intra-procedural trajectory correction, underscoring that image guidance is not ancillary but integral to CED efficacy 3132. Procedural risks remain real: serious adverse events including intracranial hemorrhage from catheter manipulation, wound infection, and stroke have been reported in CED trials, though incidence has been low and most events were manageable 11.
Regulatory Milestones and Emerging Hybrid Approaches
The regulatory landscape reflects growing confidence in these strategies. In April 2026, the US FDA granted Orphan Drug Designation to REYOBIQ (rhenium Re-186 obisbemeda nanoliposomes) for pediatric malignant gliomas, following an IND clearance in June 2025 for the Phase 1/2a ReSPECT-PBC trial supported by a $3.0 million Department of Defense grant 3536. The EMA granted Orphan Drug Designation to G2B-002, a peptide-based BBB-penetrating platform from Gate2Brain, for DIPG and pediatric GBM 34. At AACR 2025, NanoValent Pharmaceuticals presented preclinical data on NV103, a CD99-targeting nanoparticle antibody-drug conjugate with irinotecan payload, demonstrating at least 5-fold enhanced tumor versus brain delivery, detectable plasma persistence at 24 hours versus rapid clearance of free drug, and at least 3-fold survival extension in orthotopic PDX mouse models—with no dose-limiting toxicity at five times the efficacious dose 37.
Future Outlook
Both nanocarrier and CED strategies address the BBB problem through fundamentally different mechanisms, each with distinct clinical implications. Nanocarriers offer non-invasive systemic administration and potential for broad tumor coverage but remain critically dependent on achieving sufficient BBB penetration and face formidable manufacturing and translational hurdles. CED provides direct intratumoral access with demonstrated dose-dependent clinical efficacy signals, but is procedurally demanding, carries hemorrhagic risk, and cannot easily reach infiltrating tumor margins.
A promising near-term trajectory may lie in hybrid approaches: CED-administered nanocarrier formulations that combine the direct delivery advantage of CED with the potential tumor-retention properties of engineered nanoparticles. Layer-by-layer polymer-coated MMAF liposomal nanoparticles administered via CED exemplify this strategy, achieving superior intratumoral retention and GBM-selective uptake compared to free drug or conventional conjugates 6. Integration of real-time MRI guidance and AI-assisted planning for catheter optimization, alongside advanced nanocarrier engineering informed by receptor biology and tumor microenvironment modulation, represents the most likely pathway toward meaningful clinical progress. Immediate research priorities include rigorous human pharmacokinetic and pharmacodynamic characterization, standardization of CED infusion protocols, manufacturing scale-up for complex nanocarrier systems, and careful patient selection based on tumor-specific BTB status and molecular profiling.