The development of focused ultrasound (FUS) was initially tailored for intracranial ablation and has recently been expanded for the treatment of various brain pathologies
[1]. Ultrasounds are longitudinal (alternating compressions and rarefactions) mechanical waves that transmit their energy onto target tissues. A coupling agent, such as gel, between the transducer and biological tissue, ensures the ideal propagation of the ultrasound wave
[2]. Early treatment of intracranial lesions using FUS was performed with a craniotomy since the skull caused ultrasound wave distortion, diminishing the energy transfer and efficacy
[1][3][1,3]. However, the recent development of hemispheric distribution of phased arrays and concave focusing has resolved the issue of wave distortion, allowing for noninvasive procedures
[3][4][3,4]. FUS acts on biological tissues through mechanical (cavitation) and thermal mechanisms of action
[5]. Cavitation is the rapid formation and collapse of vapor bubbles. This leads to cellular fragmentation in a process defined as histotripsy
[6]. As ultrasound travels through tissue, the wave attenuates. The diminished wave is converted to thermal energy, is absorbed by the target tissue, and ultimately increases the temperature
[5]. The amount of heat absorbed can be calculated and a thermal dose can be determined. This process of sonication causes coagulation necrosis of the target tissue, subsequently ablating the targeted area
[1].
2. Indications for Use of FUS
One of the greatest obstacles to the treatment of brain tumors is the impermeability of the brain to therapeutic agents secondary to the BBB. There is a growing interest in the use of FUS to provide a non-invasive, spatially targeted method to locally increase the vascular permeability of the BBB, thus allowing for drug entry
[12][19][20][12,19,20]. This technology is frequently combined with ultrasound contrast agents known as microbubbles in clinical studies
[20][21][20,21]. When microbubbles interact with FUS sonication, microbubbles will expand and contract rapidly. This contraction results in a force generation onto the capillary walls. Gradually, the endothelial cells of the BBB will become more permeable, thereby briefly opening the BBB
[19].
Rodent research
[20] indicates that this technology is useful for the successful delivery of immunotherapy agents. For example, when treated with FUS, the delivery of small chemotherapy drugs in the brain tumor microenvironment was 3.9-fold higher than the control. Schoen et al.
[20] also illustrated a 2.7-fold increase in antibody delivery to brain tumor models with FUS. Wei et al.
[22] successfully showed an increased delivery and efficacy of the glioblastoma treatment, etoposide, alongside the opening of the BBB mediated by FUS application in rodent models. A study demonstrated the successful conversion of the immunosuppressive tumor environment to an immuno-stimulatory tumor environment in glioblastoma patients with FUS-dependent drug delivery in the presence of microbubbles
[21].
Another use for FUS is to enhance liquid biopsy by enriching circulating brain-derived biomarkers. Since the BBB limits the amount of material present within the circulation, FUS-BBB opening increases cancer-related biomarkers within the circulation
[23]. Meng et al.
[23] found that FUS treatment enhanced extracellular vesicles of neurons and brain-specific proteins without any adverse effects. Besides the application to tumor treatment and drug delivery, FUS technology shows potential use in vessel occlusion, movement disorders, and psychiatric disorders
[17].
3. FUS—From Past to Present—Preclinical Investigations
Ultrasound was first discovered in 1935 by Johannes Gruetzmacher by attaching a concave lens to an ultrasonic generator. In 1942, John Lynn et al. used a container with a crystal at the bottom, in which the ultrasonic wave was transmitted via transformer oil to a cellophane wrap against which the animal’s head was placed. They had to use maximal intensities to produce any cerebral effects, as noted by transient behavioral changes. However, with their treatments of 835 kHz for 5–15 min, all animals experienced significant scalp damage. They noted that a minimum time threshold at this intensity was necessary for these changes but proposed that a lower frequency would result in less superficial damage.
In the early 1950s, William and Francis Fry developed a FUS device that consisted of four independent transducers with planoconcave lenses focused by a polystyrene lens to a converging focal point. The beam was transmitted to the tissue via degassed saline to avoid distortion by steaming gas bubbles. These transducers could be moved vertically to change the focal point while the animal was held in a stereotactic frame. However, the entire apparatus took up two rooms, with the controls in an upper room and the transducer array coming through the ceiling of the lower room. In their animal experiments, they made a skull flap to obviate the effects of delivering the beam through the skull. They targeted the thalamus and internal capsule in cats and showed that white matter tracts were more susceptible to the effects of FUS than grey matter
[24][25].
4. FUS: Past to Present
In the 1950s, Peter Lindstrom worked with a stereotactic frame designed by Lars Leksell to study FUS as an alternative to lobotomy in patients
[25][26]. The Fry brothers later worked with Russell Meyers from the University of Iowa to perform the first ablations in humans, targeting the substantia nigra and ansa lenticularis in Parkinson’s patients
[26][27][27,28]. Even with relative successes, Lindstrom and Meyers saw that the need for a craniotomy was a limiting factor.
Lynn et al.
[28][24] noted that the heating of the scalp limited the energy that could be applied during FUS. Clement et al.
[29] solved this problem by placing transducers along a hemispherical surface to maximize the scalp surface available for absorption and reduce heating. The next development in FUS technology solved the problem of having to transmit ultrasonic waves through different tissues, which had been shown to result in phase distortions. After a 16-element needle hydrophone proved useful in monitoring acoustic feedback during ultrasonic applications
[30], Hynenen and Jolesz
[31] used this to calculate the distortion caused by an intact human skull at different frequencies. They found that the skull drastically weakened focal points when transducers at >1 MHz were used. However, using two transducer arrays with 64 elements, they showed that higher frequencies (up to 1.58 MHz) could still result in a sharp focal point when the phase shifts of the skull compensated for each element. This was the first time a multidimensional array was used with phase correction circuitry. Clement and Hynenen later showed that measurements from CT scans of the head could be used to calculate the phase shift of each transducer in a 320-element array, causing beams to converge within 2 mm of the intended focal point
[32].
The desire to measure intracranial temperatures was paramount due to the desire to avoid unnecessary damage to surrounding brain tissue. A trial by Guthkelch et al. in 1991
[33] using flexible thermocouples during FUS therapy in patients with brain tumors showed that desired intratumoral temperatures could be achieved when a craniotomy was performed. In 1997, Hynenen et al.
[34] showed that the temperature increase induced by sonications could be detected by MR thermometry. With the knowledge that a temperature of 55–60 °C was needed to cause cellular death by thermal coagulation, this lab then went on to design a model that allowed for the calculation of the lesion size induced by sequential sonications
[35]. Hynenen et al.
[36] brought these advances together by showing that the hemispherical transducer array could be placed in an MRI scanner to accurately measure the temperature change immediately after the sonications.
The first neurosurgical clinical trial using MRgFUS was a Phase 1 study done by Ram et al.
[37] to treat patients with recurrent glioma. However, the investigators performed a craniotomy in their approach. The first transcranial application of MRgFUS occurred in 2010 when McDannold et al.
[38] used the ExAblate 3000 and techniques developed by the Hynenen lab to assess the clinical feasibility of using FUS in patients with glioblastoma (GBM)
[38]. They could not achieve the temperatures required for thermal ablation due to the maximum power setting of the machine at the time. Thermal ablation would have required a 1200 W sonication delivered for 20 s—longer durations had been shown to be proportional to the extent of surrounding tissue damage. In 2014, Coluccia et al.
[39] reported they were able to use the ExAblate to induce thermal ablation in a patient with recurrent GBM (rGBM) by using 25 sonications of 150–950 W for 10–25 s. It has been shown that therapeutic temperatures may not be achieved in every patient treated with the same acoustic energy dose
[40].
Since then, InsighTec, the first company to design a commercial MRgFUS device, has increased the number of transducer elements, increased the maximum power available, and reduced the ultrasound frequency in the ExAblate. The current version, ExAblate 4000, was FDA approved in 2021, and is currently reimbursed by insurance companies for treatment of essential tremor and Parkinson’s disease tremor. It has 1024 elements in a hemispheric transducer array encircling the head that can be electronically controlled in four degrees of movement. A detachable accessory machine pumps degassed water through this “helmet” to achieve both scalp cooling and high-fidelity transmission of the waves. The machine operates at 620–720 kHz, is compatible with multiple General Electric and Siemens MRI scanners, has MR thermometry software, and allows pre-treatment CT scans to be fused with the intraoperative MRI scan. It can achieve a maximum temperature of 58.5 °C ± 2.5 °C, with a precision of <2 mm for lesions approximately 4 mm in diameterb
[2].
Another mechanism by which FUS can be used to treat brain tumors lies in opening the BBB to allow increased delivery of chemotherapeutic drugs to the tumor. The cavitation, in which microbubbles within the bloodstream are excited by low-power, pulsed sonications at the area of interest, causes shear forces in the adjacent endothelial cells that allow for the transient opening of the BBB. Hynynen et al. showed that the BBB opening induced by an ultrasound contrast agent with albumin-coated microbubbles could be identified by contrast enhancement on MRI scans
[41][42][41,42]. Multiple studies have shown that various large molecules can be delivered across the BBB in this manner
[43][44][45][43,44,45]. There are currently multiple clinical trials assessing the efficacy of using the MRgFUS-mediated BBB opening to deliver various chemotherapeutic agents
[15]. An alternative to microbubbles is nanodroplets which are generated by microbubble condensation
[46]. Chen et al.
[47] showed that nanodroplets had a higher threshold to undergo linear oscillation (the shearing forces that open the BBB) while not showing the fragmentation associated with microbubbles at higher acoustic pressures. Histologic imaging of the brain tissue of mice in their study showed that this fragmentation caused minor cellular damage. There are currently multiple commercial microbubble manufacturers in the US.
Work by Canney et al.
[48] had shown that a catheter with multiple transducers could potentially be used to treat tumors greater than 3 cm. Acoustic MedSystem, Inc designs interstitial FUS applicators which contain multiple MRI-compatible transducers spaced evenly along their length. These transducers can be independently activated and allow for circumferential US application. The placement of the transducers allows the surgeon to configure the shape of the lesion, especially helpful with large tumors. The applicator is encased in a Celcon catheter, which allows degassed water to circulate over the transducers. McDannold et al.
[49] paired this device with an automated robotic system similar to a stereotactic frame and demonstrated that the delivery of the catheters could be made more efficient.
The Sonocloud 9, designed by Carthera (Paris, France), is a miniature device with nine ultrasound transducers that is implanted in a skull window below the skin. This MRI-compatible device is attached to an external control module via a transdermal needle and allows for multiple treatments. This is currently being used in a Phase 2 trial in coordination with Northwestern University
[50] to investigate the enhanced drug delivery of paclitaxel in patients with rGBM.
NaviFUS (Chang Gung, Taiwan) developed a FUS system with incorporated neuronavigation that requires a pre-treatment CT but no intra-operative MRI. Once the patient’s cranial anatomy is registered to provide a 3D anatomical image in a manner similar to Brainlab, a headpiece is placed on the patient’s head which can deliver FUS to multiple deep locations. The pre-treatment imaging allows the device to calculate the energy needed for transcranial penetration at different points. This device, designed for outpatient use, is used for BBB opening and neuromodulation, with efficacy studies underway at Chang Gung Memorial Hospital in Taiwan to enhance the delivery of bevacizumab
[51] and assess the effect of combining FUS with irradiation in patients with rGBM
[52]. The device allows the focal point to be adjusted up to 20 mm from a central point and incorporates a custom-designed, multi-channel, hemispherical phased array to generate beams capable of covering a large tumor volume.