Targeted manipulation of pain neural networks:The potential of focused ultrasound for treatment of chronic pain

Abstract

Focused ultrasound (FUS) is a promising technology for facilitating treatment of brain diseases including chronic pain. Focused ultrasound is a unique modality for delivering therapeutic levels of energy into the body, including the central nervous system (CNS). It is non-invasive and can target spatially localized effects through the intact skull to cortical or subcortical regions of the brain. FUS can achieve three different mechanisms of action in the brain that are relevant for chronic pain treatment: (1) localized thermal ablation of neural tissue; (2) localized and transient disruption of the blood-brain barrier for targeted drug delivery to CNS structures; and (3) inhibition or stimulation of neuronal activity in targeted regions. This review provides an in-depth look at the technology of FUS with emphasis placed on applications to CNS-based treatments of chronic pain. While still in the early stages of clinical translation and with some technical challenges remaining, we suggest that FUS has great potential as a novel approach for manipulating CNS networks involved in pain treatment.

Introduction

The experience of pain results from a remarkably complex set of actions and interactions. Under normal conditions a nociceptive stimulus detected at the periphery of the nervous system initiates a signal that is transmitted and modulated through systems in the spinal cord and ultimately integrated into a conscious experience in the brain (Dubin and Patapoutian, 2010; Garland, 2012; Woolf and Ma, 2007). However, in chronic pain, abnormalities in brain function, structure and chemistry may result in persistent pathological pain (Borsook et al., 2018; May, 2008; Vania Apkarian et al., 2013). Many current treatments for chronic pain use drugs targeted to the central nervous system (CNS) in an attempt to either produce symptomatic relief, or as disease modifiers to normalize the pain connectome (i.e., back to the pre-pain state) (Ashburn and Staats, 1999). As many drugs have serious side effects or limited efficacy for certain chronic pain conditions, the use of medical devices is often integrated into patient care. Devices such as Transcranial Magnetic Stimulation (TMS; (Klein et al., 2015; Young et al., 2014)), transcutaneous electrical nerve stimulation (TNS; (DeSantana et al., 2008)), transcranial direct current stimulation (tDCS; (Pinto et al., 2018)); or deep brain stimulation (DBS; (Farrell et al., 2018)) may produce their effects through alteration in brain connectivity, neurochemistry or neuronal activity (Dossantos et al., 2018; Hanlon et al., 2018; Lozano et al., 2019).

Focused ultrasound (FUS) is an emerging technology that has great potential to be added to the set of medical devices used in the treatment of chronic pain. The underlying principle of the technique is contained in the name – a convergence of ultrasound energy to create an amplified effect at one location while leaving the surrounding region unaffected. FUS devices can perform non-invasive and targeted delivery of energy into the brain, even through the intact skull. Ultrasound parameters can be set to achieve one of three primary mechanisms of action in the brain: (1) thermal ablation of tissue; (2) transient disruption of the blood-brain barrier; or (3) excitation/inhibition of neuronal activity. Focused ultrasound treatments in the brain have now entered the clinic (Krishna et al., 2018). The technology is being used successfully for non-invasive neurosurgical ablative treatments (Elias et al., 2013); clinical trials for targeted drug delivery through the disrupted blood-brain barrier have begun (Lipsman et al., 2018; Mainprize et al., 2019); and the first human studies on neuromodulation of brain activity have been published (Lee et al., 2015; Legon et al., 2018). However, applications to chronic pain treatment remain largely unexplored and represent fertile ground for researchers and clinicians to investigate.

A thematic summary of this review is presented in Figure 1. The chronic pain state is depicted as aberrant functioning within, and connections between, particular regions of the brain involved in processing the experience of pain. Focused ultrasound can target these regions for different treatments that attempt to restore their proper functioning and connectivity. In this review, we start a broad overview of chronic pain and provide The Rational for focused ultrasound as a therapeutic strategy to treat chronic pain. We then summarize the Physical Principles of Focused Ultrasound and give a brief perspective on the Historical Development of Focused Ultrasound. Development of the technology culminated in the recent clinical successes of using focused ultrasound for neurosurgical applications, described in the section, Into the Clinic: Transcranial MR-Guided Focused Ultrasound Ablation. Building on the above sections we then move to more recent applications of focused ultrasound: Targeted Delivery of Pharmaceuticals: Focused Ultrasound Disruption of the Blood-Brain Barrier andAltering Neural Circuitry: Focused Ultrasound Neuromodulation. Finally, in the section onClinical Translation and Future Directions for Pain Applications we summarize the current hurdles the technology faces in maximizing its clinical utility and discuss how the technology could be applied to pain applications for symptomatic treatment or disease modification today and in the future. Throughout the review we emphasize the few studies that have specifically looked at applications to pain conditions and other studies that have not been pain specific but have targeted their therapy to regions of the brain known to be a part of the pain network.

Figure 1:

Thematic overview of focused ultrasound treatments for chronic pain. Chronic pain can result in aberrant activity within, and connections between, brain regions that collectively give rise to the experience of pain. Focused ultrasound is uniquely suited to selectively target any of these pain network nodes to deliver a therapy with the goal of normalizing the network back to the pre-pain state.

1. The Rational for focused ultrasound as a therapeutic strategy to treat chronic pain

In 2015, a task force from the International Association for the Study of Pain converged on a definition for chronic pain as a condition with “persistent or recurrent pain lasting longer than 3 months” (Treede et al., 2015). This simple definition reflects the fact that pain enduring past the normal time of healing is no longer performing the necessary function of alerting the body to potential acute harm, but is rather a pathological state that serves no protective purpose. Chronic pain can be classified according to the underlying cause (e.g., cancer), the physiological mechanism (e.g., neuropathic pain), or the perceived location in the body (e.g., headache). The International Association for the Study of Pain task force identified seven classifications of chronic pain: Chronic primary pain, chronic cancer pain, chronic postsurgical and posttraumatic pain, chronic neuropathic pain, chronic headache and orofacial pain, chronic visceral pain, and chronic musculoskeletal pain (Treede et al., 2015). It is estimated that 20% of the global adult population suffers from some form of chronic pain (Goldberg and McGee, 2011), including 100 million Americans (Simon, 2012).

While an initial painful injury can occur anywhere in the body, the transition to chronic pain involves changes to the CNS (Harte et al., 2018; Woolf, 2011). This evolution from acute to chronic pain is considered to be a process where an initial sensory barrage on normal nociceptive circuits evolves into driving changes in other brain circuits, including reward, salience and cognitive networks (Simons et al., 2014). Changes to these various brain networks, collectively referred to as the pain connectome, can manifest as patient behaviors that may include depression, enhanced anxiety or fear of pain (Borsook et al., 2007). It is therefore not always sufficient to treat only the initial injury site in order to resolve chronic pain; the altered connections and chemistry of the affected brain networks must also be brought back to the pre-pain state. The brain regions and pathways identified as being a part of the perception and processing of the pain experience are detailed in Figure 2.

Figure 2:

Brain regions and pathways involved in the perception and experience of pain. Nociceptive signals from the periphery travel from the spinal cord into the brain along pathways that include the spinothalamic, spinoparabrachio–amygdaloid and spinoreticulo–thalamic tracts. Information entering the thalamus is relayed to cortical areas that include the primary somatosensory cortex, insula and cingulate cortex. Further connections within the brain to regions such as prefrontal cortex and the amygdala allow for attentional and emotional factors to be incorporated in the experience of pain. Integrated signals from various higher order areas of the brain are transmitted via the periaqueductal gray to the spinal cord to act as descending modulatory influences on the incoming nociceptive signals. Figure adapted from (Simons et al., 2014).

Treating chronic pain has proven to be very difficult (Borsook et al., 2014). To date for most chronic pain patients, treatments are psychological, pharmacological or interventional. Several recent meta-analyses from the Cochrane Database of Systematic Reviews and other sources illustrate that these treatments have varying, and often limited, efficacy. See, for example, reviews on the evidence for gabapentin for neuropathic pain (Wiffen et al., 2017), topical analgesics for the treatment of pain (Derry et al., 2017), opioids for noncancer pain (Busse et al., 2018), various neurostimulation techniques for the management of chronic pain (Hofmeister et al., 2020), or nonpharmacologic therapies for back pain (Chou et al., 2017). Perhaps the issue of needing new and more effective treatments is best considered in the context of the efficacy of opioids and the current opioid crisis that in part has emanated from the lack of alternative treatments available for pain.

This lack of evidence for effective treatment outcomes extends to device-based interventions. Another paper from the Cochrane Database of Systematic Reviews considered the evidence for “non-invasive brain stimulation techniques aim to induce an electrical stimulation of the brain in an attempt to reduce chronic pain by directly altering brain activity” (O’Connell et al., 2014). The technologies considered were repetitive transcranial magnetic stimulation (rTMS), cranial electrotherapy stimulation (CES), transcranial direct current stimulation (tDCS), transcranial random noise stimulation (tRNS) and reduced impedance non-invasive cortical electrostimulation (RINCE). Overall, the review concluded that there was a lack of high-quality evidence to support, or refute, the use of these modalities for effectively treating chronic pain (O’Connell et al., 2014).

Focused ultrasound falls into the domain of manipulation of the CNS from a device-based therapy. While other device-based treatments aimed at normalizing pain-related networks in the brain have yet to demonstrate long lasting success in the treatment of chronic pain, focused ultrasound has several unique features that make it a promising candidate for continued research. As this review will detail, focused ultrasound can deliver energy into the brain that is both localized and deeply penetrating, allowing targeting of pain nodes in cortical and subcortical structures. It further has the flexibility to harness this energy delivery for different mechanisms of action that are not available to existing modalities. Very little research has been done with focused ultrasound on the specific application of pain so it remains to be seen whether these attributes can be translated into successful treatments. There is enough initial promise to make further research into pain applications a worthwhile endeavor.

2. Physical Principles of Focused Ultrasound

A brief overview of the physical principles of ultrasound is given to explain several concepts that appear throughout this review: how FUS energy propagates through tissue to a small volume focused deep in the body and how the FUS energy interacts with biological tissues to produce several different bioeffects. These basic principles are important in the context of utilizing this approach safely in patients. While each of the three methods have been safely used in humans, they all carry potential concerns that need to be carefully deployed as the technology advances for pain treatment.

Ultrasound is a mechanical longitudinal wave in soft tissue, displacing particles in the direction of wave propagation. The particles in the medium oscillate back and forth creating microscopic and rapidly switching regions of compression (high pressure) and rarefaction (low pressure). By definition, an ultrasound wave can have any frequency above 20 kHz, however all of the applications considered here use a frequency between 200 kHz and 3 MHz. The user has control over the center frequency, physical geometry of the ultrasound transducer, input power, pulse duration, pulse repetition frequency, and number of pulses applied. These parameters combine with the physical properties of the tissue to determine the spatiotemporal distribution of energy deposition in the body and the bioeffects on the tissue.

2.1. Absorption and Reflection

As the ultrasound wave travels through tissue it is affected by the phenomena of absorption, reflection, refraction and scattering. Absorption is the process of vibrational energy being converted into heat, which raises the temperature of the tissue. The combination of absorption and scattering attenuates the ultrasound energy as it propagates through the tissue and is the primary factor in determining how deeply the energy can be delivered into the body. Attenuation is highly dependent on frequency and tissue properties. In brain tissue, the intensity of an ultrasound wave at 1 MHz will be attenuated by 50% for each 5 cm of distance traveled (Goss et al., 1979). For comparison, 50% attenuation of visible light in the brain occurs over less than 1 mm (Jacques, 2013). The attenuation of ultrasound energy in skull bone is much more severe, to the extent that only approximately 10% of the original energy remains after traversing the roughly 8 mm thick human skull. Reflections occur when the wave encounters a boundary between two tissue types with differing acoustic impedances (the tissue’s speed of sound multiplied by its density). At soft tissue layers, the ultrasound tissue properties are similar enough that the amount of reflection is only a few percent. At tissue-bone interfaces, the reflected energy is closer to 30% when the wave is normal to the surface and even higher for oblique incidences. The highest reflectance occurs at tissue-gas interfaces, where essentially all of the energy is reflected. Such an interface could be at the skin/air surface, or inside the body if the lungs or other gas-filled regions are encountered.

2.2. Focusing

The focusing of the ultrasound beam is most often achieved by using a transducer, or array of transducers, with a curved geometry. For a spherically curved transducer, the geometric focus of the beam will lie at a distance equal to the radius of curvature. Modern transducers for human transcranial applications consist of many (up to ~1,000) small individual piezoelectric sources arrayed onto a hemispherically shaped surface. The size of the focus is determined by the transducer geometry and operating frequency, with smaller focusing achievable at higher frequencies. A focal diameter of ~2 mm can be achieved for the frequency range used in transcranial FUS applications (roughly 200 kHz to 1.0 MHz). The length of the focus depends on the ratio of the transducer diameter to radius of curvature and can be twice to upwards of 20 times larger than the focal diameter. As an example, the ExAblate 4000 system from InSightec, a transcranial focused ultrasound system with FDA approval, is a hemispherical design with 1,024 elements, 30 cm diameter and 15 cm radius of curvature. It operates at 670 MHz and quotes a focal size of 1.5 × 1.5 × 3.0 mm.

Figure 3 provides a conceptual approach to demonstrating the benefits of focusing. Ultrasound intensity at a particular location is the most important property for eliciting bioeffects. The important point is that intensity is a per area value, specifically quantifying power per unit area (W/cm²) at a given location. The schematic shows a hemispherical transcranial FUS system with a 30 cm diameter and a 15 cm radius of curvature. An assumed power on concentric hemispheres (W) is plotted as a function of radial distance from the transducer, showing attenuation of the energy as it propagates through the water bath, skull and brain tissue. The next plot shows the corresponding hemispherical surface area (cm²) as a function of radial distance from the transducer face, with a minimum surface area at the focus based on the ExAblate 4000 focal diameter of 1.5 mm. The intensity (W/cm²) plot is then obtained as the ratio of these two plots. At the skull surface, the total power in the ultrasound beam is 468 W, but this is spread out over a surface area of 443 cm² and therefore the intensity impinging on the skull at any location is only 1.06 W/cm². For reference, the American Institute of Ultrasound in Medicine (AIUM) has proposed 1.0 W/cm² as a limit below which heating from ultrasonic exposure is not a concern (Fowlkes, 2011). When the ultrasound beam reaches the focus, it’s total power has been attenuated down to only 17.3 W, however this is concentrated over a surface area of only 0.04 cm² and therefore the intensity at the focus is the much stronger value of 432 W/cm². It is the ratio of surface areas (approximately 10,000 to 1 for skull to focus in this example) that allows the transmitted ultrasound energy to pass through intervening tissue without having an effect until it reaches the focus where a large effect can be generated.

Figure 3:

A conceptual demonstration of ultrasound focusing in the brain. The ultrasound energy should be considered along concentric hemispheric surfaces starting at the transducer face and propagating with decreasing radius towards the focus. The first plot shows the total power (W) on these surfaces as a function of radial distance from the transducer where the power gets attenuated sharply by the skull and mildly by the brain tissue. The second plot shows the corresponding surface areas (cm²) of the hemispheres. The bottom plot of ultrasound intensity is obtained by taking the ratio of the power over the area (W/cm²).

2.3. Bioeffects

Ultrasound affects tissue primarily in two ways: the absorbed energy being converted into heat and the pressure wave interacting with the microenvironment to induce local mechanical forces. Heating is the much slower process of the two, occurring on the order of seconds to tens of seconds. The rate of heating depends on the ultrasound intensity, the absorption properties of the tissue (which is frequency dependent), and the thermal properties of the tissue (which are independent of any ultrasound parameters). It is highly influenced by the dissipative effects of the surrounding vasculature. Heating rates in brain tissue can be as high as 1.8 °C/s, achieving a temperature rise from 37°C to 55°C in 10 seconds. The damage done to cells by heating can be quantified by thermal dose (Sapareto and Dewey, 1984), an integrative metric that accounts for both the temperature rise and the duration of elevated temperature. Thermal coagulative necrosis of cells occurs if the temperature reaches ~55–60 °C. Irreversible damage can also occur at lower temperatures held for a longer time.

The mechanical forces most relevant for this review are those that are produced when the ultrasound field interacts with gas-filled microbubbles in the tissue. Microbubbles can be purposefully injected into the blood stream, as is done in imaging applications to improve signal contrast or can be created spontaneously in the tissue when a region of large amplitude negative pressure causes an existing gas pocket to grow into a microbubble. The oscillating pressure field of the ultrasound wave causes the microbubble to expand and contract in size, a process referred to as stable cavitation. The changing size of the microbubble is what creates the forces on the adjacent tissue. If the ultrasound pressure amplitude is too large, the compression cycle of the pressure wave can cause the microbubble to collapse, a much more violent phenomenon known as inertial cavitation that causes shock waves strong enough to damage local tissue. The chain of nucleation, inertial cavitation and cell damage happens at much higher intensities than stable cavitation and is frequency dependent. For example, empirical results in dog muscle show a cavitation threshold of 1,000 W/cm² at 1 MHz and 2,800 W/cm² at 3 MHz (Hynynen, 1991).

Mechanical forces can also be exerted on tissue without the presence of microbubbles. Focused ultrasound is known to produce a measurable radiation force at the focus, a push on tissue acting in the direction of propagation that is maintained throughout the sonication duration. Another type of force, acoustic streaming, is the phenomenon whereby ultrasound forces cause a liquid medium to circulate. These forces also act in the direction of propagation and can create movement of free floating small molecules and organelles (Johns, 2002).

3. Historical Development of Focused Ultrasound

Adoption of new technologies across disease entities can proceed in fits and starts. The feasibility of delivering energy to the brain was realized early in the development of FUS, but only very recently has it begun to be applied transcranially for functional brain diseases. It can target superficial or deep regions of the brain to alter behavior (Kubanek, 2018), but still little work has been done to date in chronic pain. The modern study of ultrasonics dates back to World War I when acoustic methods were being developed for underwater detection and communication purposes. Early investigations into possible medical applications soon followed as ultrasound had potential advantages over electromagnetic radiation for the non-invasive delivery of energy into the body. At the wavelengths needed for suitable focusing of the energy (~1 mm), electromagnetic radiation has a very shallow penetration depth and cannot reach deep lying tissue. Much higher electromagnetic frequencies can penetrate deeply, but these fall into the realm of ionizing radiation that would potentially affect all tissues in the beam path, not just at the focus. Ultrasound was seen as having the advantageous combination of being a nonionizing energy source with the capability to achieve deep penetration at the 1 mm wavelength scale necessary for tight focusing of the energy. The early development work was driven by thermal ablation applications as other mechanisms of action were not discovered until much more recently.

Applications of focused ultrasound in the brain faced two major challenges: focusing the energy through the skull and receiving some form of feedback that could aid in targeting and monitoring the treatment. Initially both problems were addressed by performing a craniotomy to create a window for unimpeded propagation of the focused ultrasound and through which to perform ultrasound imaging. The first study to demonstrate that focused ultrasound could destroy brain tissue was done in animals in the early 1940s by Lynn and Putnam (Lynn et al., 1942). The technology was pushed forward during the 1950s through the 1970s by the brothers William and Francis Fry. They demonstrated focused ablation of brain tissue in a primate model and brought the technology into the clinic in an attempt to treat Parkinson’s disease and brain cancer patients (Fry and Fry, 1960; Fry and Meyers, 1962). By the 1970s, the treatments were still done with craniotomy, but the systems now incorporated multiple ultrasound transducers for better focusing, a stereotactic device to help with targeting and ultrasound b-mode imaging to help guide the treatment (Jagannathan et al., 2009). In these early years a handful of other groups also used focused ultrasound with craniotomy for neurosurgical applications, attempting to treat disorders such as brain tumors, depression, epilepsy, and cancer-related pain (Ballantine et al., 1960; Lele, 1967; Lindstrom, 1954; Meyers et al., 1959). The ability to selectively ablate brain tissue with focused ultrasound was established, but the treatments still required craniotomies and thus were far from the goal of complete non-invasiveness.

Two technical advances in the 1990s enabled fully non-invasive treatments: focused ultrasound systems were integrated with magnetic resonance imaging (MRI) scanners for much improved treatment guidance, and phased array transducers were developed to focus the energy through the intact skull. MRI is a much more challenging environment to work in than using ultrasound imaging, but the benefits proved to outweigh the difficulties as images of the brain could now be obtained without having to perform a craniotomy. MRI provides good spatial resolution and superior soft tissue contrast for identification of anatomical landmarks that can guide targeting, and, crucially, MRI is sensitive to temperature changes in tissue so it can provide real time temperature mapping of the targeted region to monitor thermal treatments for safety and efficacy (Poorter et al., 1995). The first demonstration of MR-guided focused ultrasound thermal ablation was performed under the leadership of Kullervo Hynynen and Ferenc Jolesz in the early 1990s (Cline et al., 1992).

The problem of focusing through the skull was solved by the combination of two technological advances. First, transducers were designed to be big enough to spread out the energy impinging on the skull over a large surface. As detailed in Figure 3, this would create relatively low ultrasound intensities on the skull surface even for a large amount of initial total energy. The second challenge, which is not captured in the simple schematic of Figure 3, is that the varying shape and thickness of the skull aberrates the ultrasound beam path such that coherent wave fronts leaving the transducer are no longer coincident and in phase at the focus. This leads to a less sharp and less intense pressure field at the intended focus. This problem was solved by designing transducers made up of hundreds of individual piezoelectric elements that could be independently driven and phased. By applying appropriate corrective phase offsets to the ultrasound field emanating from each element, a sharp focus could be recovered after propagation through the skull (Hynynen and Jolesz, 1998).

Figure 4 shows all of the components of a modern MR-guided transcranial focused ultrasound system. This includes the large phased array transducer, a skull CT image that is used for skull aberration correction, MRI temperature mapping for real-time monitoring and feedback, and high resolution anatomical MRI images for post-treatment evaluation. With the technical pieces in place, a commercial system for MR-guided transcranial focused ultrasound treatments was designed and built by InSightec in the early-2000s and clinical trials for truly non-invasive treatments were ready to commence.

Figure 4.

Elements of a modern MR-guided focused ultrasound brain tissue ablation procedure. A) The FDA-approved ExAblate Neuro MR-guided transcranial focused ultrasound system from InSightec. B) Treatment planning step using patient-specific CT for skull aberration correction. C) Real-time temperature mapping during the treatment. D) T2-weighted MRI evaluation of lesion at one day post-treatment.

For a more complete overview of the history of focused ultrasound, see the excellent reviews by Jagannathan et al. (Jagannathan et al., 2009), Harary et al. (Harary et al., 2018) and O’Brien et al. (O’Brien and Dunn, 2015).

4. Into the Clinic: Transcranial MR-Guided Focused Ultrasound Ablation

Of the three primary applications for focused ultrasound in the brain, ablation is the most well understand and furthest advanced for clinical use. The ultrasound is applied continuously at high power and can heat tissue at the focus to above 60 °C in under 30 seconds. It is a completely non-invasive approach to lesioning that could theoretically be used in place of invasive neurosurgical ablation methods.

4.1. Clinical trials

The first patients to undergo a transcranial MR-guided focused ultrasound ablation procedure were treated at Brigham and Women’s Hospital in 2005 (McDannold et al., 2010). The treatments were performed on three patients with inoperable recurrent glioblastoma tumors. The study established the feasibility of focusing energy through the human skull to a targeted location inside the brain while using MRI to measure the temperature rise at the focus and on the brain surface. The experience also revealed the challenges in sonicating a large volume of tissue that was not near the center of the brain (see Section 3.3, Limitations).

The experience in tumors led investigators to turn their attention to functional neurosurgical applications in which only a small volume of white matter near the center of the brain would be targeted for ablation. The first application was to neuropathic pain (see below), followed by treatment of essential tremor. For essential tremor patients, a single lesion was targeted unilaterally to the ventral intermediate nucleus of the thalamus. Results from the first 15-patient clinical trial were published in 2013 demonstrating the safety and efficacy of the procedure. Since the success of these first clinical trials, the scope of treatments has been expanded to include treatments for Parkinson’s disease, epilepsy, neuropsychiatric disorders such as obsessive compulsive disorder and depression, and a return to brain tumors. The treatment for essential tremor received FDA approval in 2016. See Table 2 for a full list of completed and ongoing clinical trials as of December 2019.

Table 2:

FUS Neurosurgery ablation clinical trials. Trials were obtained from clinicaltrials.gov using search for Other terms “ExAblate OR transcranial focused ultrasound OR transcranial MRgFUS”. Up to date through December 2019. VIM = ventral intermediate nucleus, ACC = anterior cingulate cortex.

Disease	Trial Number	Dates	Patients	Target	Primary Outcome Measure
Essential Tremor	NCT01932463	2011 – 2012	6	Thalamotomy	Accuracy of Thalamotomy lesion
	NCT01304758	2011–2012	15	VIM Thalamus	Adverse events
	NCT02037217	2011–2014	10	VIM Thalamus	Adverse events
	NCT01698450	2012 – 2013	10	Thalamus	Lesion size / Safety
	NCT01827904	2013 – 2017	72	Thalamotomy	Upper limb CRST Subscore
	NCT03253991	2014 – 2023	100	Thalamotomy	Device Related Complications / Effectiveness
	NCT02289560	2015 – 2017	50	Thalamotomy	Adverse Events
	NCT02252380	2015 – 2017	10	VIM Thalamus	Device and Procedure related complications
	NCT04074031	2019 – 2023	15	VIM Thalamus	Contralateral upper limb tremor at 3-months
	NCT03935581	2019 – 2020	20	VIM Thalamus	Adverse Events
	NCT03465761	2019 – 2021	30	Bilateral Thalamotomy	Clinical Rating Scale for Tremor
	NCT02692183	2018 – 2019	100	Thalamotomy	EEG/ ERP recordings
	NCT03560622	2018 – 2019	30	Thalamotomy	Connectivity correlates of tremor
Parkinson’s Disease	NCT01772693	2012 – 2017	27	VIM Thalamus	Device and Procedure related complications
	NCT02347254	2014 – 2017	6	Pallidotomy	Adverse events
	NCT02246374	2014 – 2018	10	Subthalamic Nucleus	Device and Procedure related complications
	NCT02263885	2015 – 2017	20	Globus Pallidum	Device and Procedure related complications
	NCT02912871	2016 – 2017	10	Subthalamic Nucleus	Adverse events
	NCT03454425	2018 – 2019	40	Subthalomotomy	Motor MDS-UPDRS score; Adverse events
	NCT04002596	2017 – 2020	50	Thalamotomy	Clinical Rating Scale for Tremor
	NCT03964272	2019 – 2020	10	Bilateral subthalamotomy	Adverse events; Motor MDS-UPDRS score
	NCT04112381	2020 – 2021	10	Bilateral thalamotomy	Adverse events
Dyskinesia	NCT02003248	2013 – 2017	5	Pallidotomy	Unified Dyskinesia Rating Scale
	NCT03319485	2018 – 2020	116	Pallidotomy	Unified Dyskinesia Rating Scale
Pain	NCT01699477	2008 – 2013	30	Thalamus	Precision of lesioning / Adverse events / Lesion size
	NCT03309813	2017 – 2019	10	Thalamotomy	Device and Procedure related complications
	NCT03111277	2017 – 2021	5	Thalamus	Adverse events
Brain Tumors	NCT01698437	2011 – 2016	3	Brain tumors	Adverse events / Lesion size
	NCT01473485	2011 – 2017	10	Brain gliomas	Adverse events
	NCT00147056	2012 – 2017	10	Brain tumors	Adverse events
	NCT03028246	2017 – 2019	10	Benign tumors	Adverse events / Tumor volume
Obsessive Compulsive Disorder	NCT01986296	2013 – 2019	15	ACC, internal Capsule, Ventral Striatum or Subgenual Cingulate Cortex	Effectiveness of treatment
Depression	NCT02348411	2014 – 2017	5	Bilateral Anterior Capsulotomy	Hamilton Depression Rating Scale-17
	NCT03421574	2018 – 2020	12	Capsulotomy	Adverse events
Epilepsy	NCT02804230	2016 – 2019	10	Subcortical focal epileptic target	Device and procedure related complications
	NCT03417297	2018 – 2020	10	Anterior Nucleus	Feasibility; Safety
Graves’ disease	NCT03013257	2017 – 2020	266	Thyroid parenchyma	Rate of remission

4.2. Ablation Pain studies

Chronic pain was among the first conditions that researchers aimed to treat with ablative transcranial MR-guided focused ultrasound. The first trial was carried in nine patients with chronic therapy-resistant neuropathic pain (Martin et al., 2009). The target was the central lateral thalamus, which was successfully heated to between 51 °C and 60 °C by continuous ultrasound sonications lasting 10 to 20 seconds. The sonications were performed under the guidance of real-time MR temperature monitoring and the patients remained awake throughout the procedure. Lesions at the targeted location of 3–5 mm in diameter could be seen on follow up MR imaging immediately after the treatment and at 48 hours later. No adverse events were reported at the time of procedure.

A second paper from the same group expanded their results to 12 patients and included follow up out to one year post treatment (Jeanmonod et al., 2012). They report successful lesioning in 9 out of the 12 patients; the first patient in the study was purposefully heated to only 42 °C and two other patients had lesions that were too small to fully cover the posterior part of the thalamic central lateral nucleus. Those 9 patients experienced clinical benefits from the treatments as measured by their visual analog scale score (49% improvement at 3 months, 57% improvement at 1 year), moderate reduction in drug intake, and some reports of significant somatosensory improvements around the pain area.

Two separate clinical trials using transcranial MR-guided focused ultrasound to treat neuropathic pain patients have begun at the University of Maryland and the University of Virginia in the United States (NCT03111277, NCT03309813). Both trials will enroll therapy-resistant neuropathic pain patients with the goal of lesioning the Central Lateral Thalamic nucleus, each trial aims to treat a total of 10 patients.

4.3. Limitations

One of the major limitations of focused ultrasound ablative treatments is that they are currently constrained to targets only in central regions of the brain. This small treatment envelope is due to the inherent tradeoffs in heating at the focus versus heating at the skull surface. As the target moves closer to the skull, the ratio of skull-to-focus surface areas decreases. The same amount of energy is getting to the focus, but it is passing through a smaller area of skull surface which leads to greater intensity and more skull heating. Expanding the treatment envelope is an area of on-going research (McDannold et al., 2013; Odéen et al., 2014).

Another challenge is treating large volumes of tissue. The dimensions of the focal region were given above as 1.5 × 1.5 × 3.0 mm, a volume of 6.75 mm³, which is ideal for precise lesioning and required in order to achieve the skull-to-focus surface area ratio necessary for reaching high temperatures. However, this volume is about 150 times smaller than the fairly modest volume of 1 cm³. Ablating a volume with that number of repeated sonications would be prohibitively long, primarily due to the time that is allocated to allow the skull to cool. For example, a case report of using focused ultrasound for brain tumor ablation reported using 25 high-power sonications with “cooling periods of several minutes” between sonications to achieve an ablation volume of 0.7 CC in 4 hours of treatment time (Coluccia et al., 2014). Current research is investigating beam steering and focus shaping strategies that can increase the volume treated in a clinically reasonable amount of time.

4. Targeted Delivery of Pharmaceuticals: Focused Ultrasound Disruption of the Blood-Brain Barrier

By modifying the ultrasound parameters towards lower power and shorter bursts repeated at a low duty cycle over a longer duration, it is possible to produce biophysical effects on tissue other than heating. In this section we describe how focused ultrasound can be combined with systemically circulating microbubbles to be used for transient disruption of the blood-brain barrier (BBB).

4.1. Mechanism of action

A seminal paper from 2001 demonstrated that pulsed focused ultrasound applied in the presence of systemically circulating gas-filled microbubbles could open the BBB to agents that normally do not cross from the vasculature into the brain (Hynynen et al., 2001). The hypothesis at the time was that the ultrasound pressure wave would induce the microbubbles into a state of stable cavitation and that this rapid physical oscillation in size would create forces on the vessel walls strong enough to disrupt the tight junctions between endothelial cells that help to form the BBB. Experimental evidence validated that focused ultrasound BBB opening did indeed make the BBB permeable to normally non-penetrant agents. Later studies investigating the underlying mechanisms implicated several changes at the cellular level including disruption of tight junctions, a reduction in drug efflux mechanisms, and an increase in the number of transcytotic vesicles (Aryal et al., 2014; Burgess et al., 2015; Hynynen et al., 2005; Sheikov et al., 2008, 2006, 2004). Figure 5 shows a schematic of a pre-clinical system for focused ultrasound BBB opening, a conceptual diagram of microbubbles oscillating in the blood stream, and MRI contrast imaging for confirmation of cortical BBB disruption in a rat.

Figure 5.

Focused ultrasound BBB opening. A) Pre-clinical MRI-compatible focused ultrasound system. B) Schematic of microbubbles undergoing stable cavitation within blood vessel to induce disruption of the blood-brain barrier. C) Gadolinium-based contrast MRI imaging for characterization of location and extent of BBB opening.

4.2. Extent of BBB Opening, Safety, and Clinical Trials

Focused ultrasound opening of the BBB has been used in pre-clinical studies to deliver a wide range of agents to the brain, from the chemotherapeutic agent Temozolomide with a molecular weight of 194 Da (Wei et al., 2013) up to antibodies as large as ~150 kDa (Aryal et al., 2014; Jordão et al., 2010). For comparison, molecules that are able to naturally cross the intact BBB have a molecular weight under 400 Da and high lipid solubility (Banks, 2009). The BBB remains permeable over several hours, with closing following an exponential decay curve (Conti et al., 2019; Marty et al., 2012; Park et al., 2012). Depending on the ultrasound parameters, MRI contrast agent molecular weight and animal model, BBB opening has been seen to persist for over 24 hours in a few cases (Marquet et al., 2014). The location and extent of BBB opening can be determined in vivo using contrast MRI. Following systemic injection of a gadolinium-based contrast agent, regions of hyperintense signal on a T1-weighted image correspond to locations where the agent has leaked out of the blood stream and into the brain parenchyma. The size of BBB opening matches the dimensions of the ultrasound focal zone. It therefore has the same trade-offs as ablation – very precise opening in a small targeted region can be achieved but opening larger volumes will be time consuming if done in a point-by-point fashion.

Numerous studies in pre-clinical animals have demonstrated the primary safety outcome that focused ultrasound BBB opening can make the BBB permeable without causing damage to vascular or neuronal tissues (Baseri et al., 2010; Downs et al., 2015; McDannold et al., 2012). This includes studies of focused ultrasound BBB opening at titrated ultrasound power levels (Sun et al., 2015), and studies of repeated focused ultrasound BBB opening in rodents (Tsai et al., 2018) and non-human primates (Downs et al., 2015; McDannold et al., 2012). Damage due to thermal heating is not a risk. The ultrasound is applied at lower power and a 1–2% duty cycle so that any heat generated would dissipate before building up to increase the temperature of the tissue. Because heating is not a concern, BBB opening can be targeted to both cortical and sub-cortical regions. The primary risk is inertial cavitation, the rapid collapse of microbubbles that can produce shock waves violent enough to damage tissue. The probability of cavitation occurring increases with higher ultrasound power and lower ultrasound frequency (Shin et al., 2018). It has also been shown recently that focused ultrasound BBB opening induces a local inflammatory response that is resolved after 24 hours (Kovacs et al., 2017). The extent, duration, variation across species, and dependence on ultrasound parameters of this response is still under debate (Kovacs et al., 2018; McMahon and Hynynen, 2018, 2017; Sinharay et al., 2019).

Clinical tests of focused ultrasound BBB opening using a transcranial device from InSightec that is similar to the one used for ablation have begun. The first clinical trials were in Alzheimer’s disease patients, brain tumor patients, and amyotrophic lateral sclerosis patients. In the case of Alzheimer’s disease, the BBB was opened but no agent was delivered. This approach was based on pre-clinical results that showed focused ultrasound disruption of the BBB alone helped to reduce plaque burden in an Alzheimer’s disease mouse model, most likely due to the initiation of an immune response at the site of opening (Burgess et al., 2014; Leinenga and Götz, 2015). The results published on the human studies for Alzheimer’s disease and ALS showed that the BBB could be opened without any adverse events (Abrahao et al., 2019; Lipsman et al., 2018). The study in tumor patients delivered the chemotherapy agent doxorubicin or temozolimide and similarly showed safe opening of the BBB was possible (Mainprize et al., 2019). Additional clinical studies on ultrasound-mediated BBB disruption have used a surgically implanted transducer (Carpentier et al., 2016). A list of completed and ongoing clinical trials using focused ultrasound BBB opening as of December 2019 is shown in Table 3.

Table 3: FUS Blood-Brain Barrier Opening clinical trials.

Trials were obtained from clinicaltrials.gov using search Other terms “(ExAblate AND Blood brain barrier) OR (transcranial focused ultrasound AND Blood brain barrier) OR (transcranial MRgFUS AND Blood brain barrier)”. Up to date through December 2019.

Disease	Trial Number	Dates	Patients	Target	Primary Outcome Measure
Brain Tumor	NCT02343991	2014 – 2017	10	Brain tumors	Adverse events
	NCT03322813	2018 – 2020	15	Glioblastoma	Adverse events
	NCT03551249	2019 – 2021	20	Glioma	Adverse events; Feasibility of repeated BBB disruption
	NCT03616860	2018 – 2021	20	Glioblastoma	Adverse events
	NCT03714243	2018 – 2020	10	Breast cancer metastases in the brain	Adverse events
	NCT03712293	2018 – 2021	10	Glioblastoma	Adverse events
	NCT03626896	2018 – 2019	9	Glioblastoma	Adverse events
	NCT04063514	2019 – 2020	15	Low grade gliomas	Adverse events
Alzheimer Disease	NCT02986932	2016 – 2017	6	Right frontal lobe	Change in contrast enhancement (intensity) following BBB disruption
	NCT03739905	2018 – 2019	30	Not specified	Adverse events
	NCT03119961	2017 – 2019	10	Not specified	Florbetapir SUVr and Fluorodeoxyglucose MUV changes
	NCT03671889	2018 – 2020	10	Not specified	Adverse events
Amyotrophic Lateral Sclerosis	NCT03321487	2018 – 2019	8	Primary motor cortex	Adverse events
Parkinson’s disease	NCT03608553	2019 – 2020	10	Parieto-occipito-temporal cortex	Adverse Events

4.3. Pain Studies

To our knowledge, no focused ultrasound BBB opening studies have been published that were carried out in animal models of chronic pain or with the goal of alleviating acute pain. Our group has investigated the possibility of delivering gamma aminobutyric acid (GABA) through the disrupted BBB to the rat brain somatosensory cortex as a way of suppressing neuronal activity (Todd et al., 2019). This was done is healthy rats undergoing non-noxious electrical hind paw stimulation. Functional MRI measurements demonstrated that the BBB opening alone had a suppressive effect on the functional response to stimulation and GABA delivery further attenuated the response. We are currently pursuing similar experiments in the spared nerve injury model of chronic pain in rats (Decosterd and Woolf, 2000) and plan to investigate the effects of GABA delivered to both cortical and thalamic targets. We hypothesize that GABA induced suppression of activity in a critical node of the pain network will work as an analgesic. Unlike focused ultrasound for thermal ablation, the application of BBB opening is not restricted to central areas of the brain. BBB opening can be achieved in deep sub-cortical structures as well as cortical regions.

There have been several studies applying focused ultrasound to open the blood-spinal cord barrier (Payne et al., 2017; Weber-Adrian et al., 2015). This means that agents could be targeted for delivery to essentially any point in the pain pathway in the CNS. This approach could allow for targeted delivery of an analgesic to a particular hub in the pain network. Another possibility would be the ability to give a lower systemic dose of a drug that has adverse side effects in the periphery while still achieving a therapeutic effect due to the increased delivery at the targeted region in the brain. Given the lack of use of FUS in pain studies, targeted delivery of analgesics and other chronic pain therapeutics by focused ultrasound BBB opening is an area of fertile research.

5. Altering Neural Circuitry: Focused Ultrasound Neuromodulation

It has been known for some time that ultrasound can be used to manipulate the activity of peripheral nerves. Ultrasound has been used for both peripheral nerve stimulation and peripheral nerve conduction block. Peripheral applications will not be covered here; see the excellent reviews by Bobola et al. and Blackmore et al. (Blackmore et al., 2019; Bobola et al., 2018). Accumulating evidence now suggests that focused ultrasound can also be used to directly affect the activity of neurons in the brain. Modulation of neuronal function is achieved by applying the focused ultrasound in repeated bursts at lower power, similar to the case of BBB opening but without the use of microbubbles. The phenomenon is caused by the direct action of ultrasound-induced mechanical forces on cells. This application is the newest and least well understood of the three, but a picture is starting to emerge of how focused ultrasound can be applied directly to alter brain function.

5.1. Mechanism of action

Reversible effects that ultrasound may have on neuronal function has fascinated neuroscientists for over 50 years (Ballantine et al., 1960; Fry et al., 1958), and this application has received renewed interest in the last decade. There is in vitro and in vivo empirical evidence that direct application of bursts of focused ultrasound can both excite and suppress neuronal activity, depending on the specific parameters used. The evidence includes in vitro work on brain slices and individual neurons (Choi et al., 2013; Khraiche et al., 2008; Kim et al., 2017; Tyler et al., 2008), in vivo work stimulating the motor cortex of rodents (King et al., 2013; Ye et al., 2016; Yoo et al., 2011), and in vivo work in non-human primates (Deffieux et al., 2013; Folloni et al., 2019; Kubanek et al., 2019; Verhagen et al., 2019; Yang et al., 2018) and humans (Lee et al., 2016b, 2016a, 2015; Legon et al., 2014). These studies were done using a wide range of ultrasound parameters and readout modalities, and although they each showed an effect, there is not yet a theoretical framework for understanding how the effect is generated. Most studies have been performed at a low ultrasound power level and duty cycle combination that does not heat the tissue appreciably and therefore excludes heating as a cause (Constans et al., 2018), however others have reported thermally-induced suppression (Darrow et al., 2019). Other recent studies have all but ruled out the possibility that the motor responses seen in rodents were due exclusively to a startle response through the auditory pathway (Mohammadjavadi et al., 2019).

It is known that ultrasound can impart a mechanical force on the lipid bilayer of cell membranes. Several papers have modelled the biophysics of this interaction to provide a theoretical framework for understanding how ultrasound may affect the electrical activity of neuronal cells (Krasovitski et al., 2011; Naor et al., 2016; Plaksin et al., 2016). These models predict that absorption of ultrasound mechanical energy leads to oscillations of the intramembrane space (Krasovitski et al., 2011). These deformations could lead to several effects that would explain neuronal excitation or inhibition, including changes to the membrane’s capacitance that could drive a current, the opening of pores in the membrane lipid bilayer through which ions could flow, or the activation of mechanosensitive ion channels (Blackmore et al., 2019). There is currently the most empirical evidence for the mechanism of activation of mechanosensitive ion channels. Studies have demonstrated the effect on voltage-gated sodium and calcium channels in an in vitro preparation (Tyler et al., 2008), sodium and potassium channels in xenopus oocytes (Kubanek et al., 2016), the MEC-4 ion channel in Caenorhabditis elegans (Kubanek et al., 2018), opening of TRPA1 channels in astrocytes that leads to release of glutamate and activation of nearby neurons (Oh et al., 2019) and TRP-4 channels expressed in ASH neurons of Caenorhabditis elegans (Ibsen et al., 2015). More work remains to be done to determine which of these is the primary mechanism of action and what are the optimal ultrasound parameters for initiating the effect.

5.2. Non-Human Primate and Human Studies

Based on the promising results from in vitro and rodent work, several groups have investigated using focused ultrasound neuromodulation in non-human primate and human studies. In one of the first studies to demonstrate a behavioral effect, Deffieux et al. showed that focused ultrasound could modulate the latency time of an antisaccade task in awake rhesus macaque monkeys (Deffieux et al., 2013). Non-human primate studies using functional MRI as their readout have demonstrated that focused ultrasound can induce a response in somatosensory cortex similar to what is produced by traditional tactile stimulation (Yang et al., 2018), and can modify brain functional connectivity by targeting areas such as the supplementary motor area, frontal polar cortex (Verhagen et al., 2019), anterior cingulate cortex, or amygdala (Folloni et al., 2019). Importantly, these last two papers reported that the effects of functional connectivity modulation lasted for tens of minutes after the ultrasound had been applied. Further work from this group demonstrated that stimulation of the anterior cingulate cortex could modulate the outcome of a counterfactual choice task (Fouragnan et al., 2019).

Thus far, studies in humans have mostly shown effects related to sensory stimulation and perception, with some reports of modulation of task performance. Legon et al. showed that focused ultrasound applied to the somatosensory cortex could modulate the evoked potential response to median nerve stimulation (Legon et al., 2014). This same study reported that the ultrasound sonications could alter performance on touch and frequency discrimination tasks. Separate studies reported that focused ultrasound applied to the primary somatosensory cortex could induce perceptions in the hand and finger such as numbness or tingling (Lee et al., 2015), while application to the visual cortex produced the phenomenon of visual phosphenes, the experience of seeing white light without light actually entering the eye (Lee et al., 2016b). Subcortical regions have also been targeted, where ultrasound applied to thalamic regions showed an effect on EEG and somatosensory evoked potential measurements of median nerve stimulation (Legon et al., 2018).

A summary of focused ultrasound neuromodulation applications currently in clinical trials is shown in Table 4. Epilepsy is the first neurological disorder that researchers are aiming to treat. Results have not yet been published on these studies. Other clinical treatment applications include Alzheimer’s disease, where patients with mild cognitive impairments will receive four successive low intensity pulsed ultrasound treatments while undergoing fMRI scans (NCT03347084); ultrasound targeted to the thalamus as a “neurorestorative stimulation” method for patients in conditions of coma, Vegetative State, or Minimally Conscious State (NCT02522429); and amygdala targeting for treatment of anxiety (NCT03782194).

Table 4: FUS neuromodulation clinical trials.

Trials were obtained from clinicaltrials.gov using search Other terms “Focused ultrasound AND (modulation OR neuromodulation OR stimulation)”. Up to date through December 2019.

Disease	Trial Number	Dates	Patients	Target	Primary Outcome Measure
Epilepsy	NCT03868293	2019 – 2020	10	Temporal lobe	Patient tolerability; Reducing seizure frequency
	NCT03657056	2019 – 2021	3	Temporal lobe	Change in BOLD fMRI signal during sonication
	NCT02151175	2014 – 2020	12	Temporal lobe	Absence of histological changes
	NCT03860298	2019 – 2019	6	Site of surgery	Safety and tolerability
	NCT04192149	2019 – 2021	300	Various	Electrophysiological Changes as a result of Ultrasound
Neuroscience	NCT03192436	2018 – 2021	80	Motor/Somatosensory	Changes in EEG signal
	NCT03634631	2017 – 2018	5	Cortical/Subcortical	BOLD signal
	NCT03882931	2019 – 2020	90	DL prefrontal cortex	Performance Error on Visuomotor Task
	NCT03717922	2019 – 2022	50	Entorhinal cortex	Rey Auditory Verbal Learning Task
	NCT04168086	2019 – 2020	45	Cerebellum	Performance Error on Visuomotor Task
	NCT04168762	2019 – 2020	80	Motor cortex	Change in Motor Evoked Potential Amplitude
	NCT04190940	2019 – 2021	39	Thalamus (ablation)	Performance Error on Visuomotor Task
Alzheimer’s disease	NCT03347084	2018 – 2020	8	Hippocampus	Change from baseline in fMRI brain scan
Acute Brain Injury	NCT02522429	2015 – 2019	15	Thalamus	fMRI; Coma Recovery Scale Revised
Psychiatric Disorders	NCT03782194	2019 – 2022	50	Amygdala	Emotion Reactivity
	NCT02685488	2015 – 2016	26	Right fronto-temporal	Depressive Symptoms Assessed With the Beck Depression Inventory-II

5.3. Studies targeting pain-related CNS regions

Almost no work has been done using focused ultrasound to stimulate or suppress brain activity for the applications of alleviating acute or chronic pain. The closest study was one by Hameroff et al. in which a standard 8 MHz imaging ultrasound probe was used in B mode to try to target the frontal cerebral cortex through the temporal bone in chronic pain patients (Hameroff et al., 2013). The authors measured a subjective measure of pain, a rating for mood, and several physiological readouts after ultrasound and placebo trials, and report a few differences that just meet the significant threshold of p < 0.05 (uncorrected for multiple comparisons), including the rating for mood.

Despite not addressing the issue of pain directly, several of the studies reviewed above applied ultrasound stimulation to regions of the brain that are known to be involved in the experience of pain. This includes primary somatosensory cortex, frontal cortical regions, thalamic targets, and the amygdala. The ability of focused ultrasound to reach subcortical regions is one of its main advantages over other stimulation modalities. Stimulation or suppression of regions such as the habenula, nucleus accumbens or amygdala would have high relevance for pain studies. Also of note are the recent studies showing that modulation of functional connectivity between brain regions lasts for tens of minutes after the sonications were applied (Folloni et al., 2019; Verhagen et al., 2019). Most studies measure the effects of focused ultrasound stimulation as it is being applied, but a longer lasting effect indicates that we need further research into how the ultrasound sonications are altering the local environment and how these changes can be harnessed for treatments.

5.4. Indirect approaches for ultrasound neuromodulation

Beyond direct sonication, several groups are looking at ways to use focused ultrasound to indirectly modulate brain activity. One approach is to target neuromodulatory drugs to a specific brain region. This can be achieved by giving a systemic dose of a BBB-permeable drug packaged in a carrier and then using focused ultrasound to “uncage” it from the carrier at the desired location (Airan et al., 2017), or by a systemic dose of a free non-BBB permeable drug and using focused ultrasound to open the BBB at the targeted location (Todd et al., 2019). Another application is to use focused ultrasound BBB opening to deliver existing methods for neuromodulation such as optogenetics (Wang et al., 2017) or chemogenetics (Szablowski et al., 2018). Finally, based on the idea that ultrasound can activate mechanosensitive ion channels, several groups are investigating the possibility of “sonogenetics” where genes for optimally responsive mechanosensitive ion channels would be delivered to the brain and expressed in neurons (Huang et al., 2020; Maresca et al., 2018; Ye et al., 2018). This would allow their activation by ultrasound in the same way that optogenetics relies on light activation, with all of the pros and cons that come with ultrasound energy delivery versus light energy delivery.

6. Clinical Translation and Future Directions for Pain Applications

Each of the three ways discussed here that focused ultrasound is used in the CNS, ablation, BBB disruption and neuromodulation, has made it to human use. They can each be applied, under a specific set of parameters, in a safe manner that has not yet led to any adverse events. In this sense they have achieved clinical translation. The challenge now is to expand the range of ultrasound protocols that can be safely used and thereby unlock new applications for the technology. This includes more regions of the brain that can be targeted for ablation, transducers optimally designed for BBB opening and neuromodulation applications where improved ease of use is required for multiple repeated treatments, methods for targeting and monitoring of the ultrasound energy deposition that do not require MRI, and optimization of ultrasound parameters that can reliably induce neuronal stimulation and suppression.

While some of these improvements may be a few years away, there are still many ways in which the technology of focused ultrasound in its current state can be applied for the treatment of chronic pain. As noted above, there have only been a few ablation studies targeting pain, while BBB-opening and neuromodulation remain almost entirely unexplored. Of all the approaches to manipulating brain physiology, focused ultrasound provides the unique combination of being non-invasive, highly localized and able to reach deep lying structures. These attributes, combined with the flexible ways that it can be deployed to achieve different effects on brain tissue, make it a very attractive candidate for developing novel treatments of chronic pain.

Table 1:

FUS mechanisms of action in the brain.

	Acoustic Power	Pulse Duration / Repetition Frequency / Number Repeats	Mechanism of action	Duration of effect
Ablation	1000 W	15 s / Continuous / 1	Energy absorption converted into thermal heating	Permanent
BBB Opening	10 – 50 W	10 ms / 1 Hz / 100	Mechanical forces due to expansion and contraction of microbubbles	4 – 8 hours
Neuromodulation	10 – 50 W	1 ms / 500 Hz / 150	Likely mechanical forces on cell membranes opening mechanical-sensitive ion channels.	10s ms to minutes

Highlights.

• Focused ultrasound treatments in the brain have recently entered the clinic.

• Applications to the treatment of chronic pain are currently understudied.

• The technology can target aberrantly functioning nodes of the pain network.

• A review of the pre-clinical and clinical literature suggests an opportunity.