In their latest work, the researchers have developed a closed-loop controller that enables more precise control of bubble activities during BBB disruption. They demonstrated in both healthy and tumour-bearing rats that their system enables targeted delivery of predefined drug concentrations within a therapeutically effective range, while maintaining a safe exposure level (PNAS 114 E10281)

Bubble control

Microbubbles oscillate when exposed to focused ultrasound, a process known as acoustic cavitation. Stable cavitation serves to disrupt the BBB, but if the bubbles destabilize and collapse (inertial cavitation), they can damage the critical vasculature in the brain. To monitor and control this cavitation behaviour during treatment, the team employed passive cavitation detection (PCD).

PCD works by measuring the characteristic Fourier spectra of microbubble emission signals. By monitoring the harmonic emission (HE), a hallmark of stable cavitation, and the broadband emission (BE), seen in inertial cavitation, the researchers can characterize the strength of the two types of bubble behaviour. They can then adjust the ultrasound input instantly to sustain stable cavitation while suppressing inertial cavitation.

"We want to be able to monitor our ability to open the blood-brain barrier in real-time by listening to echoes – this could give us immediate information on the stability of the microbubbles oscillations and give us fast, real-time control and analysis," explained lead author Tao Sun, a PhD candidate advised by McDannold and Eric Miller from Tufts.

Testing the controller

Using a focused ultrasound system operating at a clinically relevant frequency, Sun and colleagues investigated the feasibility of controlling cavitation behaviour via PCD. First, they optimized the controller performance during in vivo BBB disruption by measuring the stability of the HE signals under feedback control for two pulse repetition frequencies and two microbubble infusion protocols.

Using the optimized settings, the researchers then investigated the association between HE strength and delivery of a model drug (a fluorescent tracer). They showed that the integrated HE signal correlated with the amount of a tracer delivered to the brain and that, by sonicating until the integrated HE reached a preset goal, the controller could deliver a predetermined amount of tracer to the brain.

To further validate their controller, the team tested it in a rat glioma model with a larger chemotherapy agent, liposomal doxorubicin (DOX), in addition to the tracer. They sonicated a tumour and a non-tumour location in the same hemisphere in the hippocampus, with non-sonicated locations in the contralateral hemisphere serving as controls. The controller minimized inertial cavitation by decreasing the pressure amplitude once any BE was detected above the preset threshold.

The animals were killed 24 h after sonication, to allow release of the naturally fluorescent DOX from the liposomes. The fluorescence intensities of DOX and the tracer were higher in the sonicated regions than control regions, in both the tumour and non-tumour targets. In the tumours, the researchers observed a clear threshold of enhanced delivery for both agents and a strong correlation between fluorescence and total HE. The estimated DOX concentrations reached about 10 µg/ml, exceeding levels previously shown to induce tumour regression and improve survival in rat glioma.

The authors note that in previous work without exposure level control, they observed occasional vascular/tissue damage. "With control, we were able to deliver a higher DOX dose while maintaining a safe exposure level. This result confirmed the ability of the proposed closed-loop system to tailor the drug delivery dosage within a therapeutically effective range, while improving safety control," they wrote.

Further research will be needed to adapt the technique for humans, but the approach could offer improved safety and efficacy control for human clinical trials, which are now underway in Canada. "We are currently working on implementing the controller into a clinical focused ultrasound setup, the INSIGHTEC Exablate Neuro," Sun told medicalphysicsweb.

"The human skull will attenuate the emission signals, and the detectors will be located farther away from the focal region. Using multiple detectors or a passive array may be necessary to compensate for the resulting lower signal-to-noise ratio," Sun explained. "Our current clinical system has a hemisphere phased array, so there is limited space for another passive recording array. Additionally, the location of the focal region and the detectors may vary if we use the phased array to steer the beam. These factors will be taken into account."