Blood Brain Barrier Opening with Focused Ultrasound in Experimental Models of Parkinson’s Disease

Abstract

Parkinson’s disease has many symptomatic treatments but there is no neuroprotective therapy currently available. The evolution of this disease is inexorably progressive and halting or stopping the neurodegenerative process is a major unmet need. Parkinson’s disease motor features at onset are typically limited to one body segment, i.e. focal signs, and the nigrostriatal degeneration is highly asymmetrical and mainly present in the caudal putamen. Thus, clinically and neurobiologically the process is fairly limited early in its evolution. Tentatively, this would allow the possibility of intervening to halt neurodegeneration at the most vulnerable site. The recent use of new technologies such as focused ultrasound provides interesting prospects. In particular, the possibility of transiently opening the blood brain barrier to facilitate penetrance of putative neuroprotective agents is a highly attractive approach that could be readily applied to Parkinson’s disease. However, because there are currently effective treatments available (i.e. dopaminergic pharmacological therapy), more experimental evidence is needed to construct a feasible and practical therapeutic approach to be tested early in the evolution of Parkinson’s disease patients. In this review, we provide the current evidence for the application of blood brain barrier opening in experimental models of Parkinson’s disease and discuss its potential clinical applicability.

Introduction

Early treatment for Parkinson’s disease (PD) remains a challenge. PD has several symptomatic treatments but there is no neuroprotective therapy currently available. We still do not know its etiology, and we cannot stop the progression of the disease¹. At onset, there are very few regions which have obvious neuronal loss, and motor symptoms are typically highly asymmetrical and limited to one limb². In fact, it is clear that PD motor features are closely associated with the degeneration of the dopaminergic neurons in the substantia nigra pars compacta (SNc), especially those in the ventrolateral part, and with the loss of dopaminergic terminals predominantly in the posterior putamen². The process thus seems to be fairly limited clinically and anatomically in the evolution of the disease.

The other pathological hallmark of PD is the presence of proteinaceous inclusions that are rich in fibrillary forms of alpha-synuclein (α-syn), commonly named Lewy pathology (LP)³. LP is observed more widely in the brain but to variable degrees. However, it is commonly accepted that LP generally occurs in selected nuclei including the olfactory bulb, dorsal motor nucleus of the vagus nerve, locus coeruleus, SNc, raphe nuclei, amygdala or cortex, leaving many other brain regions unaffected². Therefore, developing therapeutic strategies aiming to delay the clinical onset of PD in the most vulnerable regions of the brain looks reasonable.

Unfortunately, a number of difficulties, particularly the inaccessibility of the degenerating brain, have made this impossible so far. In this regard, new technologies such as focused ultrasound (FUS) provide an interesting opportunity. In particular, the possibility of transiently opening the blood brain barrier (BBB) to facilitate the penetrance of putative neuroprotective agents is a highly attractive approach, which could be applied readily to PD and other neurodegenerative diseases⁴. Gene delivery is a clear example of a promising therapeutic modality that would benefit^5,6. However, because there are currently effective treatments for PD patients available (i.e. dopaminergic pharmacological therapy) more experimental evidence is needed in order to construct a feasible and practical therapeutic approach to be tested in the early symptomatic phase of PD patients. In this Review, we provide current evidence of the application of BBB opening for the delivery of therapeutic agents to the brain in experimental models of PD and discuss its potential clinical applicability in the future.

The blood-brain barrier

The BBB is a term used to describe the unique properties of the microvasculature of the central nervous system (CNS). Blood vessels are made up of two main cell types: endothelial cells that form the walls of the blood vessels, and mural cells that sit on the ablumenal surface of the endothelial cell layer. Interactions with mural cells, immune cells, astrocytes, pericites and neural cells in the capillary basement membrane maintain intact the properties of the BBB⁷.

Importantly, the BBB prevents the brain from the entry of neurotoxic plasma components, blood cells, and pathogens in normal conditions. The BBB also serves to regulate the transport of different molecules though the CNS, maintaining control of the environment and homeostasis of the brain required for correct neuronal functioning. However, at the same time the BBB represents an obstacle for drug delivery to the CNS, and thus major efforts have been made to generate methods to bypass the BBB for delivery of therapeutic compounds and drugs⁸.

The blood-brain barrier opening with focused ultrasound

The treatment of CNS diseases involves the synergistic action of the BBB to transport therapeutic agents. The BBB hinders the transcellular diffusion path, which is confined only to lipid soluble compounds smaller than 400 Da with fewer than nine hydrogen bonds crossing via lipid-mediated transport. To overcome this obstacle, current treatment strategies involve transcranial injection or infusion, and employment of medicinal chemistry to chemically alter the nature of the compound so it can cross the BBB through carrier-mediated, receptor-mediated or active efflux transport⁹. However, all of these methods are either invasive, non-targeted and/or involve alteration of the drug composition. Direct injection, convection enhanced delivery and osmotic BBB disruption are some examples of targeted but invasive techniques while biological and chemical approaches and intranasal drug delivery are non-invasive but non-targeted methods^10,11.

FUS technology has emerged as a promising alternative in delivering pharmacological agents into the brain by overcoming the impermeable BBB and concurrently the adverse events associated with excessive drug administration. FUS coupled with the administration of microbubbles has been proposed as the only noninvasive technique to transiently, locally and reversibly disrupt the BBB, allowing a temporal and spatial window for molecules to cross to the brain parenchyma^12,13 (Figure 1).

Figure 1:

Mechanism of BBB opening disruption. (a) The microbubbles (white) and the therapeutic agent (purple) follow the circulation after intravenous injection. (b) The microbubble oscillate when reaching the focus of the beam that exerts mechanical forces to the endothelial cells and loosens the tight junctions. (c) The therapeutic agent diffuses through the disrupted barrier into the brain parenchyma.

The microbubbles described herein are perfluorocarbon-filled, lipid-coated microspheres on the order of a few microns in diameter, characterized by slow solubility and dissolution kinetics, attributed to their shell composition¹⁴. Their stability in the blood vessels has been improved by increasing the hydrocarbon chain-length of the coat-constituent lipids, improving their physicochemical properties and their overall efficiency^14–17.

Ultrasonic energy focused at the geometrical center can be tightly deposited deeply within the brain tissue while minimizing skull energy absorption¹⁸. During focused ultrasound, application of the transmitted acoustic pulse generates a radiation force that drives the expansion and contraction (or collapse) of the microbubbles¹⁴, characterized as acoustic cavitation. Controlled oscillation of the microbubbles results in increased vascular permeability, while rapture of the bubbles¹⁹ has been associated with increased risk for damaging the surrounding microenvironment^20,21 (Figure 1).

Hence, the size-dependent resonance behavior of the microbubbles and their response to the acoustic field varies according primarily to the center frequency, the applied pressure and the pulse length¹⁴. The relatively low frequencies combined with various pressures have been shown to successfully induce BBB openings of different sizes depending on the weight of the deliverable agent²². The effect of the ultrasound parameters on the vascular permeability is linearly dependent on the microbubble concentration, whereas the functional outcomes are more predictable for narrower size distributions of microbubbles²³.

The closing timeline and the reversibility of the BBB opening have been extensively investigated to assess the safety profile of the intervention. The time necessary for the barrier to be fully restored has been found proportionally related to the opening volume assuming the induction of a single opening²⁴. The decoupling of this dependence has been achieved by substituting large openings with small multifoci openings decreasing the necessary time for the barrier to be restored^25,26. Longitudinal studies on rodents and primates have shown that repeated ultrasound-induced BBB opening in the absence of vascular damage is a transient and reversible application^27,28. Restoration of the barrier was not only macroscopically evaluated by MRI but also neurologically by visual, cognitive, motivational and motor function behavioral testing^27,28.

Although the interaction of systemically-administered microbubbles with the capillary walls has been proposed to drive the disruption of the BBB with FUS, the mechanism is not entirely clear as not fully understood are the downstream bioeffects. Disassembling of the tight junctional (TJs) molecular structure has been placed at the beginning of the induced biological cascade, explaining the paracellular passage of molecules that has been reported^29,30. Moreover, transcriptomic analysis in the acute stages following sonication revealed a transient upregulation in proinflammatory cytokines and chemokines including CCL2, CCL3, and Tnf that have been found to promote migration, proliferation, differentiation and survival of neural progenitor cells favoring neurogenesis³¹. The prominent presence of BrdU-positive cells in animals that survived for a week after the last of six sonications has been linked to enhanced neurogenesis attributed to the expression of trophic factors such as brain-derived neurotrophic factor in the targeted brain³². Concurrent with the overexpression of inflammatory markers that mostly resolved within 24 hours, was the increase in angiogenetic-related genes and astrocytic activation³¹.

Despite the positive impact of the technique, reports on elevated microtubule-associated protein tau, phosphorylated at the Thr231 epitope³², and activated Nf-kB pathway³³ following repeated sonications, alarmed the ultrasound field and surfaced important unmet needs. Although the uncontrolled phosphorylation of tau protein has been linked to Alzheimer’s disease, phosphorylation has to occur at specific epitopes and proven to lead to pathological outcomes³⁴. Tau phosphorylation at the Thr231 epitope is associated with both physiological and pathological processes³⁴, and further experimentation is required to assess association with Alzheimer’s disease as the upregulation alone does not suffice. Regarding the initiation of the Nf-kB pathway, several reports argue on whether it is a byproduct of the sonication regime³³ or it is completely dissociated from the intervention³¹. Contradictory findings and ambiguous interpretations signify the need to fully characterize the biological changes specific to the selected sonication protocol.

So far, FUS has been studied extensively in a multitude of experiments involving the safe disruption of the BBB of various animal species (including rabbits^35,36 mice³⁷, rats³⁸ and primates^39,40) and in different PD animal models (Table 1). The integrity of the BBB is restored within hours and it remains intact²⁴ depending on the ultrasound parameters, regardless of the pathological state of the brain at least for early stages⁴¹. FUS-mediated BBB opening has been proven indifferent in terms of energy requirements to achieve permeability and closing timeline between transgenic and wild type mice⁴¹.

Table1:

Summary of studies undertaking BBB Opening in animal models of Parkinsońs disease.

Animal Model	Deliverable	Delivery vehicle	Delivery method	FUS applications	Staining	Behavioral	References
MPTP mice	GDNF	AAV	IV	1	Yes	Yes	Karakatsani et al. 2019
MPTP mice	GDNF	gene-liposome-microbubbles	IV	1	Yes	Yes	Lin et al. 2016
6-OHDA rats	GDNF	brain-penetrating nanoparticles	IV	1	Yes	Yes	Mead et al. 2017
6-OHDA rats	GDNF	Cationic microbubbles	IV	1	Yes	Yes	Fan et al. 2017
Wild type mice	NTN	Direct	IV	1	Yes	No	Samiotaki et al. 2015
MPTP mice	NTN	Direct	IV	3	Yes	No	Karakatsani et al. 2019
Wild type mice	BDNF	Direct	IN	1	Yes	No	Chen et al. 2016
MPTP mice	BDNF	Direct	IN	3	Yes	Yes	Ji et al. 2018
A53T α-syn mice	anti α-syn antibody	Direct	IV	3	Yes	No	Zhang et al. 2018
WT α-syn mice	α-syn shRNA	AAV	IV	1	Yes	No	Xhima et al. 2018

This intervention has shown efficacy in delivering various compounds of different molecular weights into the brain parenchyma including contrast agents⁴², sugars⁴³, antibodies⁴⁴, chemotherapeutics⁴⁵ and neurotrophic factors^{6,25,43,46–49}. Aside from direct delivery of the pharmacological agent, FUS-facilitated viral and non-viral vector-based gene delivery has been proven feasible to promote long term expression of endogenous proteins^6,46,49.

Focused ultrasound-induced BBB opening is an innovative and non-invasive approach to achieve drug delivery within the CNS by providing significant advantages compared to other approached in terms of targeting, non-invasiveness, and reversibility. The recent advances in technical optimization and preclinical validation support the immense potential of the intervention as the drug-delivery technique of choice in neurodegeneration models¹⁸.

Focused ultrasound-induced BBB opening in experimental models of Parkinson’s disease

When considering opening the BBB in experimental models of PD, we have to consider i. the animal model of choice and ii. what we are going to deliver. One obvious option is to target the dopaminergic system, preferentially at the level of the SNc or striatum. So far, the toxin-based models (MPTP and 6-OHDA) are the best choice for those studies designed to test neuroprotection, since there is clear dopaminergic neuronal loss in the SNc and dopamine loss in the striatum⁵⁰. Looking ahead, the other therapeutic option to consider in PD patients is to modulate α-syn aggregation in the brain. Unfortunately, these models do not recapitulate the α-syn aggregation observed clinically in PD patients⁵⁰. In this case we have the following different options: i. intra-parenchymal inoculations of exogenous α-syn (e.g., synthetic α-syn fibrils), ii. transgenic mice, and iii. animals in which α-syn overexpression is induced by viral vector injections⁵¹.

Targeting the dopaminergic system: trophic factors

Despite highly positive evidence from pre-clinical studies⁵², clinical trials of PD testing the delivery of different neurotrophic factors have been largely ineffective for several reasons including dosage, poor distribution in the brain, poor retrograde transport and late time points of delivery⁵³. FUS can partially address some of these issues primarily by improving the distribution of the deliverable molecule in the targeted location^43,54 (Figure 2) and thus adjusting the dosage to balance sufficient deposition and saturation, making trophic factors a possible alternative for PD treatment⁵⁵.

Figure 2:

Figure adjusted from Samiotaki et al. (# 39). Horizontal section at the striatum, outlined by the red dotted lines, immunostained against Neurturin (brown color) and counterstained with hematoxylin (purple color): (a) The sonicated striatum developed higher intensity of the anti-Neurturin antibody suggesting higher concentration of the trophic factor compared to the contralateral side. (b–c) Higher magnification at sonicated and contralateral side, respectively, (d–e) Extraction of the brown color only corresponding to Neurturin for b and c, respectively.

Different studies have tried to restore the integrity of the dopaminergic system by combination of FUS with trophic factors in combination with different particles or vectors. Normally, this approach is thought to be either neuroprotective, delaying the death of the dopaminergic neurons in the SNc, or restorative, by restoring or improving the capacity of the brain to produce dopamine again. Glial-derived-neurotrophic-factor (GDNF) is usually the first choice, but neurturin (NTN) and brain-derived-neurotrophic-factor (BDNF) have also been tested.

Following the initial feasibility study, Karakatsani and Wang et al.⁴⁷ delivered adeno-associated virus (AAV) – GDNF (AAV1-CAG-eGFP-GDNF) in the left striatum and midbrain of subacute MPTP treated mice. Dopaminergic neuronal cell bodies demonstrated a 58.4% upregulation with more than a twofold increase in their projections’ density after the administration of AAV-GDNF and its diffusion through the FUS-induced BBB opening, as evidenced by tyrosine hydroxylase (TH) immunostaining. Similar upregulation was observed in the striatum where the evidenced upregulation of the terminal density reached 30%. Dopamine-related behavioral changes demonstrated by amphetamine-elicited unilateral rotations, confirmed the physiological advances of the FUS-facilitated viral delivery (Figure 3).

Figure 3:

Figure adjusted from Karakatsani et al. (# 61). The nigrostriatal pathway includes the Substantia Nigra (SN), where the dopaminergic neuronal cells (SNc) and dendrites (SNr) lie and the Striatum, where the neuronal terminals can be found. The nigrostriatal pathway is downregulated upon MPTP induction in a “dying back” regime, from the terminals to the cell bodies. Application of focused ultrasound coupled with the administration of microbubbles results in increased BBB permeability allowing the diffusion of neurotrophic factors. Single administration of Neurturin resulted in a 19% increase in the TH expression at the striatal site while 4% and 20% in the SPc and SNr sites respectively. Triple administration of NTN with the equivalent number of FUS applications resulted in an increased effect compared to the single treatment on the order of 50%, 17% and 22% for the striatum, SNc and SNr respectively. Diffusion of AAV-GDNF resulted in a similar to the triple treatment effect on the order of 32%, 14% and 76% for the corresponding structures.

Similarly, Lin et al.⁴⁸ delivered gene-carrying liposomes in combination with FUS to improve the GDNF gene-delivery efficiency in the MPTP-mice model⁴⁸. The FUS-induced BBB opening was verified by contrast-enhanced MRI, and gene expression was verified by in vivo imaging. The focal delivery of gene-liposome complexes successfully served as gene carrier and BBB-opening catalyst. Immunoblotting and histological staining confirmed the expression of reporter genes in neuronal cells leading to reduced expression and progression of motor abnormalities. Post-mortem analysis confirmed preserved dopaminergic metabolism associated with the improvement of motor abnormalities.

In another study, Mead and colleagues⁴⁹ used FUS in combination with brain-penetrating nanoparticles to induce widespread and focal GDNF transgene expression in the brain following systemic administration in 6-OHDA-treated rats. After only a single treatment, this strategy led to therapeutically relevant levels of GDNF protein content in the striatum that lasted at least for 10 weeks. This strategy restored both dopamine levels at the striatum and dopaminergic TH neuron density in the SNc and reversed behavioral abnormalities with no evidence of local or systemic toxicity.

Fan et al.⁵ loaded cationic microbubbles with GDNF and FUS was used to allow transient gene permeation and induce local GDNF expression. In this study, FUS made is possible to achieve higher titer GDNF genes than with intracerebral injections. The combination of GDNF-loaded microbubbles and FUS resulted in restored behavioral motor deficits and ameliorated neuronal death in the SNc and dopamine loss in the striatum of 6-OHDA-trated rats. In another similar approach the delivery of GDNF alone or in combination with Nurr1 with PEGylated liposomes-coupled microbubbles using FUS alleviated the behavioral deficits and neuron loss in the 6-OHDA rats^56,57. The multistep process involved in gene transfection requires successful delivery, translation, and release followed by receptor ligation. It is therefore subjected to limited efficacy, shifting the scientific interest towards direct protein delivery.

Samiotaki et al.²⁵ demonstrated enhanced non-invasive local FUS delivery of NTN in wild-type mice at the level of striatum and midbrain, confirmed by immunostaining²⁵ (Figure 2). Briefly, the area of NTN bioavailability was 5.07 ± 0.64 mm² in the striatum and 2.25 ± 1.14 mm² in the midbrain, with NTN being present across the entire ultrasound-treated brain region in contrast to the relatively smaller region reached by direct injection (Figure 2). Furthermore, NTN bioactivity was evaluated by tracing the local activation of the downstream signaling pathway through the detection of increased phosphorylation of the Ret receptor, cytoplasmic kinase Erk 1 and 2 and CREB transcription factor in structures associated with their abundance. This finding was particularly significant since the nigrostriatal pathway, which connects the ventral midbrain region with the striatum, is the most severely affected dopaminergic pathway in PD⁴³.

To determine the potential value of using FUS to improve brain penetrance of bioactive molecules of the compromised neurons in the neurodegenerated brain, Karakatsani et al.⁴⁷ compared TH-based parameters between hemispheres of MPTP-injected mice that received single or triple systemic NTN injections coupled with the equivalent number of FUS applications (Figure 3). Upregulation of TH expression in the midbrain was initiated by both single and triple administrations, hence the 20–22% denser dendritic network, with only multiple exposures achieving restoration of neurotransmission, evidenced by a 50% increase in the immunoreactivity of the innervating dopaminergic neurons at the striatal level. Significantly increased dopamine levels in the treated ventral midbrain were confirmed with HPLC analysis, strengthening the relevance of the two techniques in achieving functional outcomes⁴⁷.

Another promising, non-invasive drug delivery approach that has been developed and evaluated in animal models and clinical trials is intranasal delivery (IN), which circumvents the impermeable BBB by employing the olfactory epithelium to reach the central nervous system¹¹. Low delivery specificity, the major drawback of this method, was resolved by coupling it with FUS. Chen et al.⁴³ administered BDNF to wild type mice through the nostrils, before exposing their left striatum to ultrasound⁴³. Immunohistochemical findings revealed the increased bioavailability of BDNF in the exposed striatum compared to the contralateral side, showing its immense potential as a surrogate for intravenous delivery especially when systemic exposure to pharmacological agents needs to remain contained.

To explore the translational potential of intranasal BDNF delivery, Ji et al.⁵⁸ has been investigating the functional outcomes of this delivery approach in subacute MPTP mice employing multiple administrations and exposures, following the rationale of extended dopaminergic upregulation achieved by multiple treatments shown by Karakatsani et al.⁴⁷. Expectedly, the fiber density in the ventral midbrain was increased by 13% as evidenced by TH-immunopositivity, whereas the projections of the dopaminergic neurons in the striatum were upregulated by 20% after the experimental procedure. Animals tested in the circular open-field task after amphetamine injection showed a significant preference towards ipsilateral-rotations, suggesting an amelioration of the pathology in the treated hemisphere.

Similar approaches with molecules other than trophic factors have been adopted with similar results. For example, delivery of nuclear factor E2-related factor 2 (Nrf2, NFE2L2) with FUS and microbubbles in the SNc of 6-OHDA rats results in reduced reactive oxygen species levels, thereby protecting dopaminergic neurons⁵⁹.

The therapeutic alternatives explored so far have shown a gradual increase in beneficial outcome with direct protein delivery proven efficient in upregulating neuronal function. However, translation to clinical practice entails multiple applications that would significantly improve neuronal integrity along the entire nigrostriatal pathway. Gene therapy comes as an alternative to multiple applications allowing for constant release of the neurotrophic factor with only one ultrasound session.

Targeting α-synuclein aggregation

Several studies have shown the effectiveness of the use of FUS in animal models of AD targeting both β-amyloid and tau^4,44,60–62. Indeed, in some of these cases there was a clear beneficial influence of BBB opening alone (in the absence of any therapeutic agent) on the clearance of β-amyloid plaques^4,44,60 or tau⁶¹. On the other hand, FUS enhanced antibody delivery, increasing clearance of proteins⁶¹. Subsequently to these experiments, BBB opening with FUS was started in five AD patients with early to moderate AD in a phase I safety trial⁶³. In all patients, the BBB within the target volume was safely, reversibly, and repeatedly opened. However, despite the encouraging preliminary results in AD, to date FUS-facilitated therapy against the abnormal α-syn accumulation remains uncharted. Only a few studies have targeted this so far, all of them with positive results.

In a pioneering study in cell culture, Karmacharya et al.⁶⁴ show that FUS decreases α-syn aggregation by attenuation of mitochondrial reactive oxygen species in MPP(+)-treated PC12 cells. Regarding in vivo studies, Zhang et al.⁶⁵ designed a study on A53T transgenic mice that overexpress human α-syn with prominent inclusions by the pre-symptomatic age of 9 months. The animals that received three exposures to FUS coupled with the administration of anti α-syn antibody experienced a 1.5 fold decrease in the α-syn load in the treated hemisphere compared to the untreated brains one month after the delivery completion. These preliminary findings suggest the feasibility of such an approach in clearing accumulated proteins from the degenerated brain.

In a more recent study, transgenic mice expressing WT human α-syn were subjected to MRI-guided FUS focally in different brain regions susceptible to α-syn aggregation (hippocampus, SNc, olfactory bulb, and dorsal motor nucleus) in tandem with intravenous microbubbles and an AAV9 bearing a shRNA targeting α-syn⁶⁶. One month after treatment, α-syn immunoreactivity was decreased, whereas other neuronal markers such as synaptophysin or TH were unchanged, and cell death and glial activation remained at baseline levels. These results demonstrate that FUS can effectively deliver viral vectors targeting α-syn to multiple brain areas. This approach might be useful to alter the progression of LP in PD patients particularly in those diagnosed early, thus improving the evolution of the disease.

Limitations and Prospect

The inconsistency in the pathological outcomes stemming from different sonication protocols surfaces the necessity to establish standardized methods to evaluate the sonication regime. Aside the ultrasonic parametric space, microbubble dosing and distribution have to be fully characterized and brain-structure susceptibility to ultrasound is crucial in understanding the biological effects that occur during drug delivery and fairly compare findings⁶⁷. Furthermore, a larger number of molecules should be tested given that so far, only GDNF has been widely tested in several independent laboratories. Further studies are needed to assess the effects of antibody administration and other drug types such as anti-inflammatory compounds or drugs regulating genes/enzymes related to PD (i.e. glucocerebrosidase), alone or in combination with FUS, to the brain. In this regard it is important to remember that none of the available PD animal models recapitulates all the pathologic abnormalities observed in PD clinical cases (i.e. absence of LP in toxin-based models or absence of extensive neurodegeneration in α-syn models)⁶⁸. For example, up to now, experiments aimed to reduce α-syn expression with FUS have not addressed if there is neuroprotection. There are a lot of promising candidates, some of which are “old friends” such as GDNF, but others that are new, like GBA modulators, iron-chelators, antibodies against α-syn or anti-inflammatory drugs⁶⁹, which can benefit a lot from the FUS-facilitated delivery. More gene therapy experiments should be attempted. For example, recent clinical trials involved invasive putaminal injections^53,70. AAV-AADC delivery proved to improve motor function⁷⁰ whereas AAV-GDNF did not meet its primary endpoint and did not provide clinical benefits⁵³. These approaches could be more effective (and safer) in PD patients if the viral vectors are delivered focally with FUS.

Finally, moving towards a possible clinical application in PD, several important questions remain unresolved. These include defining regions of the brain that should be targeted, how many times this procedure can be performed safely in patients and the duration of action of the delivered agent(s) to have a significant effect against the neurodegenerative process. Another fundamental underlying issue is the interaction between the described increased permeability of the BBB in the SNc and striatum in PD patients^71–74 and the impact of BBB opening via ultrasound. Indeed, one could question the need for opening the BBB if it is already deranged. However, we would argue that such abnormalities would not guarantee local and sufficient delivery of putative neuroprotective/restorative agents. In any case, this remains a topic in need for further investigation.

Conclusion

Currently the use of FUS for ablative purposes shows some promise for the treatment of PD. This includes targeting either the ventral intermediate thalamus, the subthalamic nucleus or the internal pallidum^75–78). Phase I trials exploring both of these indications are currently underway (NCT03608553). However, the use of FUS to disrupt the BBB in a focal and temporary way and facilitate the entry of different compounds such as GDNF, anti-inflammatory drugs or antibodies seems to be an extremely promising therapeutic option for the treatment of PD not only as a symptomatic treatment but also to impact on the mechanisms underlying neurodegeneration. This would also hold for other neurodegenerative disease including AD or ALS^63,79,80. Here we have reviewed several studies that display encouraging results regarding the possible application of this technology for PD. However, more experimental evidence is needed before clinical applications of these approaches can be developed.

In sum, FUS-facilitated drug delivery is a highly promising therapeutic approach, but probably still in need of further experimental work before reaching clinical trials. Thus, different protocols for FUS exposure must be explored in valid animal models in the near future.