The animal experiments were approved by our institutional animal committee. Male New Zealand white rabbits (weight: approximately 4 kg) were anesthetized using intramuscular injections of a mixture of 12 mg of sodium xylazine (Xyla-ject; Phoenix Pharmaceuticals, St Joseph, MO, USA) and 48 mg of ketamine hydrochloride (Abbott Laboratories, North Chicago, IL, USA) given per kg of body weight per hour. A craniotomy (approximately 2×2 cm) performed at least two weeks before the experiments provided an acoustic path into the brain. The craniotomy was performed to allow more accurate estimation of the peak focal pressure amplitude in the brain and to simplify the experiments (i.e., we did not have to correct for skull aberrations). The fur in the ultrasound beam path was removed before the experiments with clippers and depilatory cream. Ultrasound exposures (sonications) were then applied under MRI guidance at two non-overlapping locations in each hemisphere. These targets were one cm deep in the thalamus and approximately 2–3 mm from the midline. Forty-four locations were sonicated over the course of the experiments in 11 rabbits. Four locations were excluded from the analysis due to operator errors during the experiments.
The ultrasound field was generated with an air-backed spherically curved transducer (diameter/radius of curvature: 10/8 cm) with a center frequency of 0.69 MHz which was operated at its 3rd
harmonic (2.04 MHz). It was driven by a function generator (Model 395, Wavetek, San Diego, CA, USA) and RF amplifier (model 240L, ENI Inc, Rochester, NY, USA). Electrical power was measured in each experiment with a power meter (model 438A, Hewlett Packard, Palo Alto, CA, USA) and dual directional coupler (model C173, Werlatone, Brewster, NY, USA). The transducer’s electrical impedance was matched to the output impedance of the amplifier (50Ω) by an external matching network. The ultrasound field was calibrated using methods described previously (Hynynen et al. 1997
). The peak negative pressure amplitude was measured in a tank of degassed water using a calibrated membrane hydrophone (spot diameter 0.5 mm, GEC-Marconi Research Center, Chelmsford, England) for the entire pressure amplitude range used. Reported values for the rabbit experiments are estimates in the brain based on ultrasound attenuation through 10 mm of brain, assuming an attenuation coefficient of 5 Np/m/MHz (0.43 dB/cm/MHz) (Goss et al. 1978
). The values for the rat brain are based on attenuation through 5 mm of brain using the same attenuation coefficient and take into account the insertion loss of the rat skull (Treat et al. 2007
). Ultrasound attenuation through 1 mm of skin (measured in a rabbit using digital calipers) was also taken into account, assuming an attenuation coefficient of 40 Np/m/MHz (4.6 dB/cm/MHz), the median value reported in a review by Hynynen (Hynynen 1990
Sonication parameters were: 20 s duration; burst length: 10 ms; pulse repetition frequency: 1 Hz. Ten seconds before each sonication, a bolus of ultrasound contrast agent (Optison, GE Healthcare, Milwaukee, WI) was injected intravenously through the ear vein at a dosage of 50 μl per kg of body weight. This dosage was selected as it in the range recommended for human use (0.5–5.0 ml; i.e., 7.1–71 μl/kg for a 70 kg adult). Optison injection was followed by a bolus injection of approximately two ml of saline to flush extension tubing that extended out of the MRI. Time between each sonication was approximately five minutes. This delay allowed for the agent to be mostly removed from the circulation.
The transducer was mounted in a three dimensional manually operated positioning system and was submerged in a tank of degassed, deionized water. The rabbit was placed supine on a plastic tray mounted above this tank. The ultrasound beam propagated vertically out of the water tank through a hole in this tray. Acoustic coupling between the water tank and the animal was provided by a plastic bag filled with degassed water. The experiments were performed in a clinical 1.5T MRI scanner (GE Healthcare, Milwaukee, WI, USA). A receive-only surface coil (7.6 cm diameter, GE Healthcare, Milwaukee, WI, USA) was placed below the head to provide imaging with a high signal-to-noise ratio. The ultrasound beam propagated through a hole in this coil. The experimental setup is diagrammed in .
Diagram of the experimental setup.
BBB disruption was detected in contrast-enhanced T1-weighted fast spin echo images (parameters: TR/TE: 500/15–23 ms; ETL: 4; BW: 16 kHz; matrix size: 256×256; NEX: 4; FOV: 10 cm; slice thickness: 1.5 mm; interslice spacing: 1.5 mm). Immediately after the last sonication, a baseline set of axial T1-weighted images was acquired. Following the injection of contrast agent (: Magnevist®, Berlex Laboratories, Inc., Wayne, NJ, USA administered IV at a dose of 0.125 mmol per kg of body weight as a bolus injection), a time series of these images was acquired to evaluate the BBB disruption. Imaging was acquired for approximately 30 min after contrast injection.