Hindered diffusion of Gadolinium-based Contrast Agents in rat brain extracellular micro-environment after ultrasound-induced delivery
Allegra Conti1,2, Rémi Magnin1,3, Matthieu Gerstenmayer1, François Lux4, Olivier Tillement4, Sébastien Mériaux 1, Stefania Della Penna2, Gian Luca Romani2, Erik Dumont3, Denis Le Bihan1, and Benoît Larrat1

1CEA/DSV/I2BM/NeuroSpin, Gif Sur Yvette, France, 2Department of Neuroscience, Imaging and Clinical Sciences, G. D'Annunzio, University of Chieti and Pescara, Chieti, Italy, 3Image Guided Therapy, Pessac, France, 4Université Lyon 1, Lyon, France

Synopsis

We present here a new method to study the diffusion process of Gadolinium-based Contrast Agents within the brain extracellular space after the artificial Blood-Brain Barrier opening induced by ultrasound. Four compounds were tested (MultiHance, Gadovist, Dotarem and AGuIX). By estimating the Free Diffusion Coefficients from in vitro studies, and the Apparent Diffusion Coefficients from in vivo experiments, an evaluation of the tortuosity (λ) in the right striatum of 11 Sprague-Dawley rats has been performed. The values of λ are in agreement with literature and demonstrate that the chosen permeabilization protocol maintains the integrity of brain tissue.

Purpose

The in vivo characterization of Gadolinium (Gd) based MRI Contrast Agents (MR-CA) diffusion within brain extracellular space after a transient and local Blood Brain Barrier permeabilization induced by ultrasound is of great interest for the understanding of drug transport in brain parenchyma in the framework of new pharmaceutical developments for Central Nervous System diseases.

Methods

Four Gd-chelates with different hydrodynamic diameters (dH) were tested: three commercially available MR-CA (MultiHance, Gadovist and Dotarem) and a new class of Gd-based nanoparticles (AGuIX, University of Lyon). These latest compounds are composed of a core of polysiloxane, grafted with two or three Gadolinium chelates. They are sufficiently small (dH <5nm) to escape hepatic clearance and can be easily functionalized1. Diffusion Light Scattering (DLS) measurements were performed to estimate the hydrodynamic diameter of all compounds (Table 1). The MRI acquisitions were performed with a 7T/90 mm Pharmascan scanner (Bruker). To evaluate the CAs’ longitudinal relaxivities (r1) at 7T and 37°C, bundles of tubes containing different CA-concentrations in 0.3 % w/w agar gel were prepared. The T1 values of these tubes were measured using an IR-FGE sequence2(TE/TR1= 2.5/5 ms, 6 segments, 90 TI from 75ms to 8975ms, FA =5°, resolution = 0.250 x 0.250 x 1.25 mm3, delay between two segments TR2 = 15000ms, NA = 6). Measured values of 1/T1 for different CA-concentrations [CA] were fitted using the following equation3: 1/T1=1/T10+r1x[CA], where T10 is the T1 of the media without Gd (see Fig.1). Relaxivity values are summarized in Table 1 for all compounds. The Free Diffusion Coefficients (DFree) of these compounds were estimated by injecting 10µL of a 5mM solution in a tube filled with 0.3 % w/w agar gel. The diffusion was followed for 1 hour by acquiring five T1-maps (IR-FGE, TE/TR1 = 2.5/5 ms, 6 segments, 60 TI from 88ms to 5100ms, FA = 5°, resolution = 0.225x0.225x1 mm3, TR2 = 9000ms, NA = 1, total duration = 12.5min). The tubes were kept at 37°C during the acquisition. A T10-map acquired before the injection was used as a reference. MR-CA concentration maps were then calculated from T1-maps using the previous equation. On each CA map, a 2D Gaussian function was fitted4. The square of the widths of the Gaussian fit was plotted along time and fitted with the relationship σ2x,y= 2t x Dx,y,vitro for both σx and σy. For each compound DFree was estimated as the average DFree= (Dx,vitro+ Dy,vitro)/2. Hydrodynamic diameter (dH) of the CA was deduced from DFree using the Stokes Einstein formula DFree=kT/(3πηdH) where k = 1.38x1023 Pa.m3.K-1 is the Boltzmann constant, T is the temperature in Kelvin degrees and η is the viscosity of the agar gel(6.92·10-4 Pa.s). Focal ultrasound induced BBB permeabilization was also performed5 in the right striatum of 11 Sprague-Dawley rats (120g, Janvier, France), followed by intravenous MR-CA injections. The Gd chelates diffusion starting from the BBB disruption site was followed by repeatedly acquiring T1-maps for about 1 hour, as for the in vitro measurements. The same Gaussian fitting procedure was applied and the Apparent Diffusion Coefficients (ADC) of all the compounds in the striatum were estimated as the average ADC = (Dx,vivo+ Dy,vivo)/2.

Results

Figure 2a shows examples of in vitro CA-maps obtained for MultiHance. The fitted 2D-Gaussian functions are presented in Fig.2b whereas in Fig.2c the linear fit on their spreads is plotted. Fig.3 shows an example of in vivo dataset acquired by injecting Dotarem in one rat. In Fig.3a the original CA maps acquired within 66 minutes after the CA injection are pictured, and their respective Gaussian fits are shown in Fig.3b. The linear fit over these Gaussian spreads is given in Fig.3c. As can be noticed in Table 1, both the DFree and the ADC are decreasing with increasing hydrodynamic diameters ( Dotarem > Gadovist > MultiHance > AGuIX). Furthermore, quantitative values of hydrodynamic diameters deduced from DFree measurements are really consistent with DLS measurements (see Table 1).

Discussion and Conclusions

The ADC values have been used to estimate tissue tortuosity $$$\sqrt{\frac{D_{Free}}{ADC}}$$$ (Table 1), showing a very good agreement with the tortuosities evaluated with more standard techniques6. This agreement confirms the validity of this method to estimate the ADC values in the tortuous regime, but also that the diffusion properties of brain tissue are not altered by the chosen blood-brain barrier permeabilization protocol unlike by direct intracerebral injection4.

Acknowledgements

No acknowledgement found.

References

1. Lux F. et al. Ultrasmall Rigid Particles as Multimodal Probes for Medical Applications. Angewandte Chemie International Edition. 2011;50(51):12299–12303. 2. Deichmann R. et al. Fast T1 mapping on a whole-body scanner. Magn. Reson. Med. 1999;42(1):206–209. 3. Swift T. J. and Connick R. E.. NMR-Relaxation Mechanisms of O17 in Aqueous Solutions of Paramagnetic Cations and the Lifetime of Water Molecules in the First Coordination Sphere. J. Chem. Phys. 1962;37(2) :307. 4. Marty B. et al., Hindered diffusion of MRI contrast agents in rat brain extracellular micro-environment assessed by acquisition of dynamic T1 and T2 maps. Contrast Media Mol. Imaging. 2013;8(1):12–9. 5. Marty B. et al. Dynamic study of blood-brain barrier closure after its disruption using ultrasound: a quantitative analysis. JCBFM. 2012;32(10):1948-58. 6. Nicholson C. and Tao L..Hindered diffusion of high molecular weight compounds in brain extracellular microenvironment measured with integrative optical imaging. Biophys. J. 1993;65(6):2277–2290.

Figures

Table 1: Extracted parameters for each compound (Dotarem, Gadovist, MultiHance and AGuIX). Notably, the apparent dH estimated from the Stokes-Einstein equation is in agreement with DLS measurements, and both DFree and ADC values decrease when molecular size increases. The tortuosities are consistent for all compounds and in agreement with literature6.

Figure 1: The r1 values were estimated by fitting the IR-FGE signals as a function of TI (S(TI)=|A-Bexp(-TI/ T1*)|, T1=T1*x[B/A–1]). Fig.a shows an IR-FGE MR image for one TI, and Fig.b shows the signal fits in each tube. From the T1-maps (Fig.c) r1 was extracted by the fit: R1=1/T1=1/ T10+r1x[CA].

Figure 2: In vitro diffusion of MultiHance: a) concentration maps and their respective 2D Gaussian fits (b) acquired during 1 hour after CA injection in 0.3 % w/w agar gel. Fig.c shows the squared Gaussian spreads as a function of time and their linear fits ( σ2x,y = 2t x Dx,y,vitro).

Figure 3: In vivo diffusion of Dotarem in the frontal hemisphere of a rat brain after ultrasound-induced BBB permeabilization (a) and the corresponding 2D Gaussian fits of CA-spots (b). Fig.c shows the linear fits of the squared Gaussian spreads as a function of time after CA injection.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
3586