Purpose

To develop safe and reproducible technique to disrupt local BBB in mice through guidance of interventional MRI and to improve efficacy of IA chemotherapy of brain tumors including glioblastoma.

Methods

Naïve or human glioblastoma inoculated scid mice (20-25g, Jackson Laboratory) were anesthetized with 2% isoflurane gas. Briefly, the occipital artery was cauterized, the ECA and pterygopalatine arteries (PPA) were ligated temporarily with 4-0 silk sutures. The common carotid artery (CCA) was catheterized with a microcatheter. The mice with intra-arterial (IA) catheter secured were positioned in a Bruker 11.7T MRI scanner. Baseline T2 (TR/TE 2500/30) and T1 (TR/TE 100/1.88) weighted and dynamic GE-EPI (TE=3.0ms, TR=60ms, FOV=18, matrix=128, and acquisition time=92s and 16 repetitions) images of the brain were acquired. IA Feridex (dissolved in saline at 1:100; 0.1mg Fe/ml) was infused under dynamic GE-EPI to predict perfusion territory at specific speeds. 25% mannitol was delivered via IA route over one minute at the speed determined by previous Feridex injection. MRI (T1, T2-weighted) and histology were used to assess status of the blood brain barrier (BBB) and any consequences compromising safety of the procedure. For the glioblastoma-bearing mice, the BBB opening was followed by subsequent IA melphalan infusion (0.05mg/mouse) as a one-stop-shop.

Results

We have previously reported a strong association between IA infusion speed and perfusion territory in large animals1. Here, we observed similar dependence for mice and real-time MRI was essential for adjusting the injection speed for safe and predictable brain targeting. Initially our transcatheter SPIO contrast delivery, at the infusion rate between 0.0008-0.0016ml/s resulted in inconsistent cerebral perfusion. However, when the speed was increased to the level between 0.0015 and 0.0025 ml/s, desired brain perfusion was obtained as visualized by characteristic reduction in signal intensity for the duration of injection bolus (Fig.1a-b). Gadolinium-enhanced T1 weighted scan showed hyperintensity in the region previously highlighted by the contrast infusion at the same rate (Fig.1c). This indicates successful BBB opening by IA mannitol as predicted by perfusion pre-scan. In histopathological validation using Evans Blue, which is a gold standard for BBB assessment and rhodamine, which was used a surrogate marker of therapeutic agent demonstrated pattern of extravasation that was consistent with the observed by MRI (Fig.1d). There was strong correlation of the SPIO perfusion pre-scan and T1+Gd enhancement (Fig.1e). In assessment of safety of the BBBO procedure T2-weighted MRI 3 days post BBBO showed no obvious abnormalities and the T1+Gd enhancement was not observed any more(Fig.2a). Immunohistochemistry for GFAP and IBA-1 7 days post BBBO showed no activation of astrocytes or microglia (Fig.2b-c). Similarly, there was no evidence of neuronal damage based on NeuN staining (Fig.2d). Altogether these data indicate excellent safety profile of BBB opening procedure. Subsequent experiments were performed in mice implanted with luciferase-expressing glioblastoma (Fig.3a). Tumor bearing mice were subjected to BBB disruption under interventional MRI (Fig.3b) followed by 0.05mg IA melphalan treatment. Longitudinal bioluminescent imaging revealed marked reduction of signal intensity six days after IA chemotherapy compared to no chemotherapy controls indicating treatment response. However, after this initial reduction the signal in treated mice continued to grow.

Discussion

The goal of BBBO is to maximize CNS targeting of the therapeutic agent with minimizing systemic toxicity. However, osmotic BBBO proved highly variable and inconsistent2. This is also reflected by inconsistent methodology of preclinical studies. The infusion velocity of mannitol into carotid artery varied particularly, and in many animal studies extremely exceeded physiological rate of brain perfusion and had to be damaging3-4. Our MRI-guided IA injection experiments in several species show the value and importance of imaging allowing adjustment of infusion rate and open BBBO at physiologically relevant non-damaging infusion rates5-6. Our work showed that IA melphalan following BBBO can reach and affect the tumor growth. Transient nature of this effect is likely due to chemoresistance. In the future studies, we will potentiate the anti-tumor effect by adjusting parameters, for example, higher drug concentration, longer duration of mannitol infusion and will administer different chemotherapeutic agents with the same targeted approach.

Conclusions

MRI guidance enables precise fine-tuning of the infusion rate to achieve safe and effective local BBBO at a high reproducibility. Bypassing the BBB obstacle resulted in treatment response, but the effect was transient.

Acknowledgements

Supported by NIH/NINDS R01NS091110

References

1. Janowski, M.; Walczak, P.; Pearl, M. S., Predicting and optimizing the territory of blood–brain barrier opening by superselective intra-arterial cerebral infusion under dynamic susceptibility contrast MRI guidance. Journal of Cerebral Blood Flow & Metabolism 2016, 36 (3), 569-575. 2. Joshi, S.; Ergin, A.; Wang, M.; Reif, R.; Zhang, J.; Bruce, J. N.; Bigio, I. J., Inconsistent blood brain barrier disruption by intraarterial mannitol in rabbits: implications for chemotherapy. J Neurooncol 2011, 104 (1), 11-9. 3. Foley, C. P.; Rubin, D. G.; Santillan, A.; Sondhi, D.; Dyke, J. P.; Crystal, R. G.; Gobin, Y. P.; Ballon, D. J., Intra-arterial delivery of AAV vectors to the mouse brain after mannitol mediated blood brain barrier disruption. J Control Release 2014, 196, 71-78. 4. Yang, W. L.; Barth, R. F.; Leveille, R.; Adams, D. M.; Ciesielski, M.; Fenstermaker, R. A.; Capala, J., Evaluation of systemically administered radiolabeled epidermal growth factor as a brain tumor targeting agent. Journal of Neuro-Oncology 2001, 55 (1), 19-28. 5. Walczak, P.; Wojtkiewicz, J.; Nowakowski, A.; Habich, A.; Holak, P.; Xu, J.; Adamiak, Z.; Chehade, M.; Pearl, M. S.; Gailloud, P.; Lukomska, B.; Maksymowicz, W.; Bulte, J. W.; Janowski, M., Real-time MRI for precise and predictable intra-arterial stem cell delivery to the central nervous system. J Cereb Blood Flow Metab 2016. 6. Janowski, M.; Lyczek, A.; Engels, C.; Xu, J.; Lukomska, B.; Bulte, J. W.; Walczak, P., Cell size and velocity of injection are major determinants of the safety of intracarotid stem cell transplantation. J Cereb Blood Flow Metab 2013, 33 (6), 921-7.