A new safety index based on intrapulse monitoring of ultra-harmonic cavitation during ultrasound-induced blood-brain barrier opening procedures

Abstract

Ultrasound-induced blood-brain barrier (BBB) opening using microbubbles is a promising technique for local delivery of therapeutic molecules into the brain. The real-time control of the ultrasound dose delivered through the skull is necessary as the range of pressure for efficient and safe BBB opening is very narrow. Passive cavitation detection (PCD) is a method proposed to monitor the microbubble activity during ultrasound exposure. However, there is still no consensus on a reliable safety indicator able to predict potential damage in the brain. Current approaches for the control of the beam intensity based on PCD employ a full-pulse analysis and may suffer from a lack of sensitivity and poor reaction time. To overcome these limitations, we propose an intra-pulse analysis to monitor the evolution of the frequency content during ultrasound bursts. We hypothesized that the destabilization of microbubbles exposed to a critical level of ultrasound would result in the instantaneous generation of subharmonic and ultra-harmonic components. This specific signature was exploited to define a new sensitive indicator of the safety of the ultrasound protocol. The approach was validated in vivo in rats and non-human primates using a retrospective analysis. Our results demonstrate that intra-pulse monitoring was able to exhibit a sudden appearance of ultra-harmonics during the ultrasound excitation pulse. The repeated detection of such a signature within the excitation pulse was highly correlated with the occurrence of side effects such as hemorrhage and edema. Keeping the acoustic pressure at levels where no such sign of microbubble destabilization occurred resulted in safe BBB openings, as shown by MR images and gross pathology. This new indicator should be more sensitive than conventional full-pulse analysis and can be used to distinguish between potentially harmful and safe ultrasound conditions in the brain with very short reaction time.

Figure 1

(A) Representation of a typical ultrasound sequence for BBB opening. In general, multiple bursts of 3–10 ms are repeated at a pulse repetition frequency of 5–10 Hz, for a total exposure time ranging from 0.5 to 5 minutes. (B) The ultrasound backscattered signal originated from the cavitation of microbubbles is decomposed into n temporal windows for comparative frequency analysis.

Figure 2

Representative frequency responses of a microbubble cavitation signal measured at different periods (before and after 3.3 ms) following ultrasound bursts. Transcranial ultrasound was emitted at 0.5 MHz into the NHP brain, and the cavitation signal was measured using a PCD transducer centered at 1.5 MHz. The occurrence of ultra-harmonic frequency components (1.5 f₀, 2.5 f₀, 3.5 f₀) is a very sudden event (<256 µs), which can occur over the burst period.

Figure 3

T2-weighted images in the focal region for safe (A) and hemorrhagic (C) 2 weeks after BBB opening in NHP. Evolution of IUD measured over ultrasound bursts (0.5 MHz, 10 ms) for safe (B) and hemorrhagic (D) BBB opening. Examples of IUD (E) and IHD measurements (F) for safe (blue) and hemorrhagic (red) BBB opening. Cavitation doses were normalized by a reference value corresponding to the ultra-harmonic cavitation dose from the 1^st window of 128 µs.

Figure 4

T2-weighted images one month after BBB opening shows a suspicious contrast in the focal region (dashed circle). No residual BBB leakage was detected.

Figure 5

Representative safe and unsafe ultrasound-induced BBB opening in rats. A real-time feedback control based on the detection of inertial cavitation events set the sonications to safe levels. MRI sequences (T2-weighted, 7 T) identified the presence of post-opening edema. Rat brain image (A) without edema and (B) with edema (red circle) following sonication. Similarly, post-mortem analysis confirms the safety of ultrasound-induced BBB opening without apparent damages using Evans Blue dye (C) and with visible hemorrhages following sonication (D). These results show the need for more robust and more reliable controls.