Vascular Normalization Equilibrates the Distribution of Model Small Molecule Drug Delivery to Brain Tumors with Focused Ultrasound

Introduction: Despite aggressive combination therapy, patients with glioblastoma (GBM), typically survive only 16 months post-diagnosis. One major contributor to this poor prognosis is drug delivery limitations created by the blood-tumor barrier (BTB), comprised of regions of leaky microvessels that yield high tumor interstitial pressure and poor convective transport gradients, as well as regions with an intact blood-brain barrier (BBB) that limits both diffusive and convective delivery. Moreover, the GBM tumor microenvironment is highly heterogeneous, limiting the efficacy of therapies that do make it through these transport barriers. MRI-guided focused ultrasound (FUS) in conjunction with microbubbles has emerged as a valuable non-invasive technique to transiently open the BBB/BTB and increase delivery of anti-cancer agents. However, despite the promise of FUS approaches for treating GBM, the highly dysregulated and heterogeneous GBM tumor vasculature remains a challenge. Herein, we used neoadjuvant, low dose, anti-angiogenic therapy to restore the structure and function of tumor vessels, a process called vascular normalization (VN), in combination with FUS to enhance low-molecular weight model drug delivery and/or dispersion within tumor tissue.

Materials and

Methods: GL261-Luc2 syngeneic mouse GBM cells were implanted into the right striatum of the brains of C57BL/6 mice. DC101, a VEGFR2 inhibitor that elicits VN effects, or IgG control treatment were administered intraperitoneally on d8, 11, and 14 post-tumor implantation. Quantitative 9.4T T1-mapping MRI was used to track the spatiotemporal concentration of MultihanceTM (MH, ~1kDa), a low molecular weight gadolinium-based model drug on three cohorts of mice to assess paired baseline and FUS-mediated delivery and penetration of a model small molecule drug at Time Point (TP) 1 (d11-12), TP2 (d14-15), and TP3 (d16-17).

FUS-mediated blood-brain tumor barrier opening was achieved using a 1.1 MHz single-element transducer with a 10 ms burst length over a 2000 ms period. A total of 60 sonications were administered during a 2-min sonication duration. The feedback control system parameters were set as follows: a starting pressure of 0.2 MPa, pressure increment of 0.05 MPa, maximum pressure of 0.4 MPa, 20 sonication baselines without microbubbles, area under the curve (AUC) bandwidth of 500 Hz, AUC threshold of 10 standard deviations, pressure drop of 0.95, and frequency selection of the subharmonic, first ultraharmonic, and second ultraharmonic. Microbubbles were intravenously injected as a bolus dose of 10^5 microbubbles per gram of body weight. Four non-overlapping sonication targets were placed over the GL261-Luc2 tumor. Injection of MH agent was given immediately prior to MRI acquisition for FUS- control studies and immediately following the initiation of FUS for FUS+ studies.

Results, Conclusions, and Discussions:

Results: Quantitative T1 mapping MRI tracked intratumoral MH concentration during DC101 treatment (Fig. 1A-C). Baseline tumor vessel leakiness decreased by TP2 with DC101 treatment (Fig. 1B top); an effect lost by ~3 days post-treatment termination (Fig. 1C top). Despite the decrease in baseline tumor leakiness at TP2, FUS increased MH concentration in both IgG and DC101 treated groups to similar levels across all treatment time points (Fig. 1A-C bottom). Further, the foldchange in post-FUS over baseline MH concentration for DC101 treated tumors at TP2 was higher than IgG control (Fig. 1D). MRI concentration maps at TP2 show differences in intratumoral model drug distribution with FUS (Fig. 1E; GL261-Luc2 tumors denoted with white arrows). Assessment of radiomic features shows no changes in coefficient of variation (COV), but a decrease in both distribution skewness and kurtosis in DC101 treated tumors (Fig. 1F-H).

Discussion: We observed a temporal difference in small molecule model drug concentration at baseline and post-FUS over the DC101 treatment time course, elucidating a potential therapeutic window. VN by DC101 decreased the baseline leakiness of the tumor to a low molecular weight model drug one day after DC101 treatment termination, identifying treatment TP2 as the peak effect of DC101. Importantly, even with the decrease in baseline leakiness, the changes in tumor vasculature with DC101 did not limit FUS BBB opening in comparison to IgG control. Assessment of MRI concentration maps showed different patterns of intratumoral model drug accumulation and subsequent quantification of such features suggest DC101 treatment better equilibrated the spatial distribution of model drug within the tumor microenvironment after FUS-mediated delivery when compared to FUS negative controls. This suggests that this combination approach can improve the penetration of a small molecule drug into previously inaccessible pockets of tumor.

Conclusions: DC101 combined with FUS BBB/BTB opening did not appreciably augment the overall concentration of a low molecular weight model drug over FUS control; however, the combination of DC101+FUS improved the uniformity of intratumoral model drug distribution, a key effect that can enhance the efficacy of low molecular weight cancer therapies, such as chemotherapies.

Acknowledgements (Optional): Supported by NIH R01EB030409, R01EB030744, and R21NS118278.