Minimally Invasive Brain Intervention: What all the FUS is About
Brain surgery has been used for the treatment of movement disorders for roughly half a century. Beginning in the 1960s, thalamotomies were performed with either cryo-probes that killed brain tissue by freezing or radiofrequency thermal probes that coagulated and were used for the treatment of essential tremor (ET) and Parkinson’s disease (PD). During the same period, pallidotomy was used for PD. Yet, while these procedures were effective, they also had the potential to worsen speech, swallowing, and balance, particularly when done bilaterally.
Fast forward to the 1990s: Deep Brain Stimulation (DBS) of the ventral intermediate (VIM) nucleus of the thalamus was found effective for the treatment ET and Parkinsonian tremor, while DBS in the subthalamic nucleus (STN) and of the globus pallidus interna (GPi) was commonly used to treat Parkinson’s disease and dystonia. The advantage of this form of surgical therapy is that it is both effective and adjustable to allow for changing symptoms as the condition worsens. When worsening of symptoms, such as speech, occur with bilateral DBS, it can frequently be reduced by changing the stimulation parameters. While there is no intentional brain lesion created with DBS, unintended surgical complications of this open surgical procedure, such as bleeding and infection, do occur.
The development of modern neurosurgery for movement disorders was made possible through advances in stereotactic apparatuses, MRI, and intra-operative neurophysiology. A precisely machined stereotactic frame mounted on the head along with MRI allows the surgeon to accurately place the DBS lead in a deep and unseen brain center. Physicians then confirm the target by physiologic recording and stimulation in an awake patient. After the electrodes are implanted in the brain, a second procedure is required under general anesthesia to implant the pulse generator and link it to the electrodes. This second procedure is very similar to that of placing a cardiac pacemaker and is typically performed as an outpatient about two to seven days after the implantation of electrodes.
For the past 20 years, DBS has been the standard of care for patients with movement disorders, who in spite of best medical management seek a surgical therapy. A group of patients remain that may be candidates for surgical intervention but are unwilling or unable to tolerate this open surgical procedure with the implantation of a complex device. Thus, technological innovations have been sought to develop minimally invasive procedures that deliver similar results to DBS. The first of these less invasive strategies was radiosurgery using the stereotactic focused array known as the Gamma Knife that is now widely used for treatment of brain tumors. Widespread use of radiosurgery for movement disorders was limited by relatively inconsistent effects and complications likely related to that lack of effect in real time to monitor the lesion. Recently though, a more promising and potentially effective minimally invasive procedure has emerged: MRI guided high intensity focused ultrasound (MRgFUS). This procedure has recently been granted its first FDA approval for functional neurosurgery—unilateral thalamotomy for the treatment of essential tremor.
A New Option for Movement Disorders: Microfocused Ultrasound
Like any form of wave-energy, sound can be focused. Sonic energy can penetrate, reflect (i.e., scatter), or it can be absorbed and converted into thermal energy/heat. A moderate rise in brain temperature (40 to 50 degrees) results in a temporary, reversible change) in function, similar to that of electrical stimulation, whereas a higher rise in temperature (60 degrees) for even several seconds can result in irreversible tissue coagulation, with an effect similar to the placement of a radiofrequency thermal probe. Since the absorption of sonic energy is related to tissue density, the greatest challenge to this approach is the need to minimize the absorption of ultrasound energy by the skull to allow the far less dense brain tissue target to be heated to coagulation levels.
Although the concept of FUS for movement disorders is relatively new, therapeutic ultrasound has been used on the brain for many years and in a number of capacities. Experiments in the 1940s and 1950s in animals explored its potential for neuromodulation and coagulation of deep lesions, and in the 1960s FUS was employed as an early treatment for brain tumors. These early experiments could not be termed minimally invasive, since part of the skull needed to be removed.
More recently, therapeutic ultrasound has been applied to the Central Nervous System (CNS) for the possible treatment of movement disorders, stroke, epilepsy, pain, and other disorders. This has become possible because of MRI-guided devices that employ an array of aligned sonic emitters, where the point of intersection receives 1,000 times the sonic energy—the same principle as gamma knife. These devices can target deep brain regions with millimeter accuracy in spite of the distortional effects of the skull on sonic energy. MRI can measure local brain temperature after each sonic treatment in real time, leading to the controlled creation of a temporary and then permanent lesion, while monitoring of both brain temperature and patient response for both symptom relief and off-target side effects. MRI Thermography generally correlates well with a thermal lesion on post-procedure MRI, although the persistence of an MRI lesion does not correlate well with persistence of a clinical effect.
The application of this technology to neurologic disease is moving very quickly. The first reports—including a 15-patient study of FUS based thalamotomy showing improvement in patients with severe, disabling, medically refractory essential tremor of the dominant hand—appeared as recently as 2013. This result was validated with a double-blind sham controlled, multi-center study leading to the first FDA approval for a brain treatment in July 2016.1
The use of MRgFUS for the creation of stereotactic ablation is only one of the many potential applications of this technology to neurologic disease. The goals of FUS treatment can be grouped in relation to the level of sonic energy employed:
• High Intensity (HIFU): Lesioning for ET, PD, chronic pain, OCD, epilepsy, tumors, clotlysis
• Medium Intensity: Opening the blood–brain barrier (BBB) to enhanced chemotherapy for brain tumors, cell and molecular therapies for neurodegenerative diseases
• Low intensity: Neuro-modulation resulting in neuronal inhibition or excitation, modification of behavior
The major limitation of MRgFUS for stereotactic ablation remains the absorption and distortion of sonic energy by the skull. Part of the rationale for using this technology to target brain regions involved in movement disorders like ET and PD is that these regions are distant from the skull and its effects. Specifically, the skull absorbs energy and heats up, effectively reducing the maximum energy at the target. Even with the current technology, at least 10 percent of ET patients have skull density properties on CT scan that prevent treatment. Along with a stereotactic frame, patients have a silicone bag of chilled water on their shaved head to reduce skull heating during the procedure. This is much less an issue for other brain indications that require lower intensity sonication such as blood–brain barrier opening and neuromodulation.
Stacking Up to DBS
MRgFUS for movement disorders can be viewed as a new technology to perform an established surgical goal: the creation of a stereotactic brain lesion. However, since the introduction of DBS, stereotactic lesioning for movement disorders is rarely performed. Any comparison of these two technologies suffers for the great difference in their stage of development. DBS is the current recognized standard of care as a surgical approach to movement disorders, with over 150,000 patients treated worldwide over the last 20 years, while the total number of patients treated with MRgFUS to the brain is less than 1,000. Both require the use of a head frame and patient cooperation is needed for both in spite of pain or discomfort. MRgFUS is a less invasive procedure done as a single outpatient visit, while DBS requires not only an open brain surgical procedure with the risk of bleeding and infection, followed by a second surgical procedure to implant the pacemaker like pulse generator in the chest. The goal of MRgFUS is to make a permanent brain lesion, with the possibility that the benefit will be not sustained, and the risk that an off target neurological symptom may be long-lasting. Side effects related to the thalamic brain lesion have been common in MRgFUS treatment of movement disorders, but they have usually been transient and rarely rated as serious or severe. DBS does not create an intentional brain lesion, relying on highly adjustable high frequency stimulation for its effects. Stimulation parameters can be changed in an outpatient visit to either reduce off-target side effects or improve symptom control. The placement of a DBS requires passing a rigid wire through centimeters of brain with the clear opportunity for creation of an unintended off-target lesion. The placement of the programmable system also carries with it the possibility for electrode breakage and device failure, which require further surgical correction. MRgFUS currently is fulfilling its potential in avoiding the major risks of any open neurosurgical procedure that of intracranial infection or bleeding (about one percent for each with DBS). Only the passage of time and increasing numbers of treated patients will determine the true safety profile of MRgFUS.
It has been suggested that the amount of tremor reduction in the pivotal study of MRgFUS for ET (about 40 percent after one year) although substantial, is smaller than usually reported for DBS. It is notable that for most of the centers involved in that study (including our own), the published patients represented their first clinical experience ever with MRgFUS. Further studies will be needed to see if center experience will result in a “learning curve” with improved patient outcomes using this novel and complex technology.
The optimal target for treatment of PD may be different between MRgFUS and DBS. The most common target for DBS is the subthalamic nucleus (STN) for patients with a severely fluctuating response to medications. However, the STN has rarely been targeted with stereotactic lesioning, because of the risk of producing severe involuntary movements such as hemiballismus. A feasibility trial of MRgFUS is currently underway targeting the globus pallidus interna (GPi) a brain target with extensive previous experience using either stereotactic lesioning or DBS. The initial patients treated also have levodopa-induced involuntary movements (dyskinesias) that interfere with their lives, an aspect of PD that has been well treated with interventions targeting the GPi.
At this point, virtually all MRgFUS studies of movement disorders have only treated one brain side, while DBS is commonly performed on both brain sides to give symptom relief on both body sides. The reluctance to use MRgFUS on both brain hemispheres is a reflection of the experience with bilateral thalamic lesions that extends back to the 1960s where high frequency of worsening of speech and balance occurred. More recent studies suggest that staged bilateral thalamotomy in carefully selected patients with bilateral ET may be worth further investigation with MRgFUS.
FUS and the Blood-Brain Barrier
One aspect of FUS with the potential for high impact on neurologic disease is its capacity to open the BBB in a controlled, safe, and targeted manner. Most of the “high tech” therapies, i.e. molecular and cellular for brain disorders, are too large to cross the BBB. Current routes of brain delivery include directly into the CSF (intra-thecal or intra-cerebro-ventricular), which is invasive and with good distribution but poor penetration, and intra-cerebrally, which is even more invasive, with good penetration but poor distribution. Sadly, although there have been advances in the devices and parameters used for brain delivery, they can still be considered variations on the hypodermic needle that was invented in 1835. Molecular and cellular therapies such as gene therapy have been very difficult to translate from promising studies in small animal models to a successful clinical trial for any neurological disease. There has been speculation that poor distribution from even highly invasive brain injection may play a role in the lack of efficacy of these therapies. Advancements in methods for opening the blood brain barrier have also languished. Edward Neuwelt introduced the use of injecting a solution of hypertonic mannitol into a brain artery to open the BBB in 1989, and recently commented that he is still waiting for this method to be replaced.
The normal BBB consists of brain endothelial cells connected together with continuous tight junctions. Moderate levels of ultrasound energy can open the BBB when used in combination with commercial solutions of microbubbles (i.e., gas filled lipid vesicles), which are widely used as a contrast agent in diagnostic ultrasound. When activated by FUS, the microbubbles expand and vibrate transiently, separating endothelial tight junctions. By controlling the levels of sonic energy, the BBB can be safely opened for a short time and allow large molecules and cells to enter with no brain pathology.
Many pre-clinical studies in recent years have demonstrated the potential of FUS-mediated opening of the BBB in combination with an intravenously injected therapy for a variety of condition. For instance, anti-cancer antibodies in combination with FUS can shrink brain metastasis in a rodent model.2 The effect of FUS in combination with anti-amyloid antibodies in reducing amyloid plaque burden and improving cognition in transgenic mice, has generated interest in a pilot clinical trial to attempt to reduce amyloid in patients with Alzheimer’s disease.3 Additionally, one recent study found that FUS mediated BBBD allows IV Viral Vector Gene therapy for GDNF in a mouse model of PD.4
FUS-BBB disruption (D) has even been used to deliver experimental brain therapies as large as stem cells. FUS-BBBD-mediated stem cell delivery is far less invasive and has the potential for better distribution than intracerebral needle injection, since stem cells injected into the adult brain frequently show little migration from the needle track. FUS-mediated delivery uses the brain’s natural delivery system: the vasculature. Not unexpectedly, it also highly inefficient, with less than 0.05 percent of cells injected into the blood found in brain after FUS. The current technology is limited in its ability to improve efficiency of BBD opening, since greater intensity and longer duration of sonication raise safety concerns, particularly regarding the entry of cytotoxic serum proteins into the brain. Our group has been working with another novel delivery technology called magnetic targeting, which could improve the efficiency of movement of stem cells from blood into brain during FUS-mediated BBB opening. Human stem cells grown in culture will engulf non-toxic super paramagnetic iron oxide nanoparticles (SPION) and can then be attracted to an external magnet. A small “refrigerator” magnet mounted to the head of rats would attract SPION-loaded stem cells from the blood to a region of brain injury where the BBB is also damaged. Stem cells loaded with SPION are much more likely than non-magnetic cells to cross from blood to brain of FUS opening of the BBB. These experiments open the way to see if this combination strategy can deliver enough stem cells to reverse animal models of diseases like PD and AD. Sonication is also known to stimulate production of vessel/tissue adhesion molecules (integrins), growth factors, and cytokines that promote homing of stem cells to tissues such as muscle and kidney. If FUS-stimulated homing also occurs in the brain it would have importance for stem cell therapy for brain diseases.
The potential benefits of FUS begin with a new option for patients with movement disorders, but extend to a wide range of conditions that have been treated with stereotactic neurosurgery. Its capacity to open the blood-brain barrier may also help bring the many potential molecular and cellular therapies for neurologic disease to the clinic. The application of FUS to brain disease will need a growing awareness and knowledge by neurologists, but will be very rewarding for all who become involved in this powerful and multidisciplinary technology.
Paul S. Fishman, MD, PhD is a Professor of Neurology at the University of Maryland School of Medicine. He is also Chief of Neurology at the VA Maryland Health Care System.
1. Elias, et al. A Randomized Trial of Focused Ultrasound Thalamotomy for ET. NEJM, 2016 Aug 25; 375(8): 730-9.
2. Park, McDannold, et al. The kinetics of blood brain barrier permeability and targeted doxorubicin delivery into brain induced by focused ultrasound. 2012 Aug 20;162(1):134-42.
3. Leinenga, Goetz, et al. 2016. Nat Rev Neurol. 2016 Mar; 12(3): 161-74.
4. Lin, et al. Non-invasive, neuron-specific gene therapy by focused ultrasound-induced blood-brain barrier opening in Parkinson’s disease mouse model. J Control Release. 2016 Aug 10; 235: 72-81.