Introduction

The treatment of glioblastoma has limited clinical progress over the past decade, partly due to the lack of effective drug delivery strategies across the blood-brain-tumor barrier (BBTB). Moreover, discrepancies between preclinical and clinical outcomes demand a reliable translational platform that can precisely recapitulate the characteristics of human glioblastoma. The focus of this study is to evaluate the BBTB heterogeneity with two genetically engineered models that recapitulate two important glioma phenotypes, including the diffusely infiltrative tumor margin and angiogenic core. We further ask whether methods to increase the barrier permeability such as our recently developed optical modulation of the blood-brain barrier leads to improved therapeutic outcome.

Methods

Mice were injected with 73C (solid tumor) or PS5A1 (infiltrative tumor) in the cortex, followed by GBM treatment. To modulate the BBTB, 37 µg/g of the AuNP-BV11 was administrated by intravenous (i.v.) injection into the tumor-bearing mouse. One hour later, the mice received either vehicle or taxol via i.v. injection. Then the mice received picosecond-laser excitation (532 nm, 28 ps, 40 mJ/cm², 1 pulse) in the tumor area. The treatment was repeated every 4 days during 4-12 dpi. The tumor size was measured by MRI. Similar treatment groups were used with seven mice in each group to obtain the survival rate.

Results

To characterize optical modulation of the blood-brain-tumor barrier (optoBBTB), the accumulation of tight junction-targeting AuNP-BV11 was analyzed by ICP-MS and a fluorescent dye (Evans blue) was injected via i.v. to visualize the increase in BBTB permeability. We found AuNP-BV11 in the tumor (0.19±0.02%ID/g) was almost 2-fold higher than that in the normal brain (0.10±0.01 %ID/g). The BBTB permeability can be successfully increased after single pulse laser excitation in both the 73C tumor and PS5A1 tumor after BBTB modulation. Comparing two tumor models, the PS5A1 model has an infiltrative growth pattern with vessel co-option development, and the 73C model recapitulates GBM features such as an angiogenic tumor core and intratumoral heterogeneity in terms of BBTB function and tight junction composition. After GBM treatments, the tumor volume was reduced by 6 and 2.4-fold, and the survival was prolonged by 50% and 33% in the 73C tumor and PS5A1 tumor, respectively.

Conclusion

In this study, we present an effective therapeutic approach for GBM treatment utilizing optoBBTB and taxol delivery in both clinically relevant infiltrative and angiogenic tumor models. Since paclitaxel does not penetrate the blood-brain-tumor barrier and is abandoned for glioblastoma treatment following its failure in early-phase clinical trials, our results raise the possibility of reevaluating a number of potent anticancer drugs by combining them with strategies to increase blood-brain-tumor barrier permeability. Our study reveals that optoBBTB significantly improves therapeutic delivery and has the potential to facilitate future drug evaluation for cancers in the central nervous system.

The work presented in this abstract is accepted at Nature Communications.