Focused ultrasound for opening blood-brain barrier and drug delivery monitored with positron emission tomography

Focused ultrasound (FUS) is a minimally-invasive technology used for treatment of many diseases, including diseases related to the colon, uterus, prostate, and brain. Although it has been mainly used for ablative procedures, the ability of FUS to open the blood-brain barrier (BBB) presents a promising new application. However, the mechanism of BBB opening by FUS remains unclear. This review focuses on the use of FUS to open the BBB for enhancing drug delivery and investigating how Positron Emission Tomography (PET) provides insight into the underlying mechanism.


Introduction
The blood-brain barrier (BBB) is a physical barrier composed of endothelial cells connected by tight junctions, which regulates brain homeostasis and protects the brain from harmful agents [1]. The BBB regulates drug entry into the brain via transporters, either by active or passive mechanisms [1].
Although these transporters protect the brain from neurotoxic effects, they reduce the efficacy of targeted cerebral drugs. To overcome this obstacle, several approaches have been introduced to enhance BBB permeability. One approach is to increase drug lipophilicity to improve its penetration through the BBB [2]. However, this method cannot be applied to molecular therapies targeting local areas of the brain [3]. Another obstacle is the molecular weight of the agent. The molecular weight of a drug may increase due to drug modifications, making it difficult for the drug to cross the BBB if it exceeds the threshold of 400 Da [2]. To circumvent the BBB, a technique termed convection-enhanced delivery was developed [4], which is an invasive method that involves inserting a cannula through untargeted tissues to reach a subcortical structure and then injecting the drug directly [4]. Although this method can target specific areas in the brain, it can cause complications, such as chemical meningitis, infection, or brain tissue damage [4]. As such, safety concerns surround this method, as it is difficult to apply.
In contrast, high-intensity focused ultrasound (FUS) is a therapeutic extra-corporeal thermoablative technique, which has been used as an alternative to radiotherapy and surgery for the treatment of several diseases. In fact, FUS has been applied to treat uterine fibroids with a lower risk of haemorrhage and many types of cancers, including brain, kidney, liver, prostate, and bone metastases [4] [6]. In addition to the thermo-ablative approach, low-intensity FUS has also been recently proposed as a safe and reversible approach for focally opening the BBB. Thus, ultrasound arises as a potential novel technique for improving drug delivery to selected targets in the brain [7]. FUS in combination with microbubbles can lead to a transient and focal opening of the BBB, thus enabling the passage of therapeutic agents across the BBB without relying on the enhanced permeability and retention effect [8][9] [10]. There is a strong debate about the exact physical mechanisms underlying the BBB opening, with the proposed explanations ranging from prolonged stable cavitation of the microbubbles to more lower intensity of the latter together with the pulsed wave cycles result in only 4-5°C heating within the focused area, rendering the impact on brain tissue harmless and the BBB opening temporary [6], (see Table 2). In addition, animal studies demonstrated that lower intensities elicited neuromodulative effects (either inhibition or stimulation), such as activated motor responses [33] and decreased cortical excitability, to suppress epileptogenic discharge [34]. Legon et al. noticed that low intensity FUS modulated cortical activity and enhanced sensory discrimination ability in healthy human volunteers [35]. Furthermore, Monti et al. [36] investigated the feasibility of using this method to awaken patients suffering from traumatic disorders. An ongoing clinical trial at the University of California in Los Angeles (ClinicalTrials.gov Identifier: NCT02151175) is investigating the use of low-intensity FUS as a therapeutic modality to treat patients with temporal lobe epilepsy.

FUS Effects
FUS induces three different types of effects: thermal, cavitation, and mechanical and streaming effects. At high intensities, this technique generates a discrete thermal lesion at the focal point of the FUS. Conversely, at medium intensities, due to a limited increase in tissue temperature, FUS is able to disrupt the BBB in the sonicated area for hours [5]. The principle underlying this BBB disruption involves the mechanical effects of FUS or cavitation [37] [38]. Combining focused ultrasound with contrast agents, such as stabilized microbubbles, facilitates this procedure and reduces the energy required to disrupt the BBB. Contrast microbubbles are optimally designed for stable cavitation, which is associated with safe BBB opening . Several possible mechanisms, including vessel wall displacement due to expansion and contraction, have been proposed [39]. After the procedure, an MRI scan using intravenous gadolinium contrast injection allows delineating the areas enhance d due to BBB opening. Finally, at lower intensities, FUS can also induce neuromodulation by activating neuronal circuits. This may be generated by several mechanisms, such as microcavitation of the internal membranes and plasma, which modifies voltage-gated ion channels or neurotransmitter receptors [39]. It is important to note that the potential side effects of FUS are based on several factors: exposure duration, tissue type, and FUS frequency and intensity [38]. Thermal effects may J o u r n a l P r e -p r o o f cause skin burns, whereas mechanical or cavitation effects can rupture vessel walls and lead to haemorrhage [40].

Microbubbles
Using ultrasound scans to open the BBB requires a large amount of energy to overcome the diffraction and attenuation of the skull, which increases the risk of permanent tissue damage. Therefore, some studies have recommended using FUS in combination with ultrasound contrast agents [5], which can be in gaseous form (microbubbles) or liquid form (nanodroplets). Further, these agents can be used in conjunction with FUS to increase its efficiency in disrupting the BBB. The following are two common types of ultrasound contrast agents (UCA) that are approved by the FDA: lipid-coated UCA Definity® [41] and protein-coated UCA Optison™ [42]. It is important to note that Definity® is more responsive to ultrasounds because of its more flexible lipid shell [15].

The mechanism of FUS in opening the BBB using microbubbles
The exact mechanism by which FUS enhances BBB permeability is not fully understood but some insight has been provided by previous studies, in which the BBB remains disrupted for a duration of approximately four hours [43]. The main hypothesis regarding the mechanism is that microbubbles vibrate due to the FUS waves and cause mechanical action exerting force on the capillary walls, which consequently widens the tight junctions in the BBB. Another explanation involves the presence of vacuoles, which are spaces within the cytoplasm of a cell that are enclosed by a membrane. FUS can temporarily open this membrane to allow the drug to be transported to the cells in the interstitial space [37]. Other studies hypothesized that certain biochemical substances may be released by endothelial and glial cells after sonication as a reaction to protect the brain, thus enhancing BBB opening. For example, Cucullo et al. [44] found a transitory increased release of α2-macroglobulin after BBB breakdown.

1-First-in-human Study :
From the ten eligible articles that were reviewed, only one study was applied on humans [55]. [ 18 F] Florbetaben was used to measure Aβ deposition in five patients with early to moderate Alzheimer's disease. This phase I clinical trial showed that it was feasible and safe to temporarily open the BBB within the targeted area, which was the superior frontal gyrus white matter of the dorsolateral prefrontal cortex. However, in the exploratory analysis, no differences were observed in Aβ levels before and after sonication, as shown in Figure 2, in contrast to the preclinical studies that were previously mentioned. The difference between the clinical and the preclinical findings regarding Aβ clearance with FUS can be attributed to several reasons. First, the study by Lipsman et al. [55] is considered the first of its kind, since it used human subjects; thus, the primary focus was on the feasibility and safety of opening the BBB, rather than the kinetics and timing of Aβ clearance. In addition, the sample size and age of patients may also impact results. A small sample size may sometimes lead to insignificant differences in results. Moreover, with age, the function of BBB transporters may be negatively affected. Further, preclinical studies typically sonicate several large areas, compared to studies that involve humans. For example, Lipsman et al.
applied FUS on three spots that were each 3mm apart, unlike other animal studies which sonicated 4 spots that were 1.5mm apart [47] [49] [55]. [55]. Furthermore, Aβ deposition in animal models was found to clear out more easily compared to patients with Alzheimer's [56]. Chen et al. [3] mentioned several obstacles in their review, which limited the translation of the preclinical FUS studies to clinical trials in humans. One of these obstacles is variations in the anatomical structure, biochemical characteristics, and responses between species and individuals, which led to the use of different physical parameters, such as the amount of ultrasound dose. Other challenges include the different types of medical devices, microbubbles, and drugs that are used for delivery into the brain, and the lack of real-time monitoring during BBB disruption [3]. Thus, further clinical studies are necessitated to calibrate physical parameters as maximally as possible and establish a standard protocol for every specific situation .  64 Cu-labeled gold nanoclusters (AuNCs) to evaluate BBB permeability after applying FUS in mice (see Table1). All four studies succeeded in opening BBB by FUS and effectively delivered the nanoclusters into the brain, as shown in Figure 3 [57] [58] [59] [60]. Gold nanoclusters are metal nanoclusters with a size range of 1 to 100 nm [61]. Although it has not yet been tested on humans, preclinical studies show that 64 Cu-AuNCs can be an accurate guide to therapy [59] [60]. Sultan et al. [57] investigated the effects of surface charges of 64 Cu-AuNCs on its efficacy to penetrate the BBB. The results indicate that the nanostructure with neutral charge is optimal for use in theranostic application [57]. However, the application of 64 Cu-labeled gold nanoclusters in clinical studies is expensive and has two major drawbacks: poor therapeutic efficacy and difficulty in degradation. Thus, the toxicity level increases, making it difficult for repeated use as a therapeutic modality. Moreover, a specific cyclotron is needed to produce 64 Cu. An analysis should, therefore, be performed to ensure that its use is valid, reliable, and safe for humans by testing for toxicity, bio-distribution, and stability. 3-Potential Radiotracers to evaluate BBB integrity : Goutal et al. [62] investigated the effects of FUS on BBB integrity and function using [ 11 C]erlotinib. The uptake of [ 11 C]erlotinib was not found to increase after sonication. However, after applying an inhibitor (elacridar), the uptake of the radiotracers increased, and the drug (erlotinib) was delivered to the brain (with and without FUS), as shown in Figure 4 [62]. These results indicate that FUS can affect BBB integrity, but not BBB function. Okada et al. [28] concluded in their animal study that 2-amino-[3-11 C] isobutyric acid ([3-11 C]AIB) has tremendous potential for evaluating BBB disruption, given that a 1-MHz single sine wave is applied with the aid of microbubbles. The PET tracer [3-11 C]AIB is a neutral amino acid that does not cross BBB rapidly. However, it is absorbed by brain cells after opening of BBB [63]. Meanwhile, the efflux of [3-11 C]AIB from the brain to the blood is negligible, and thus, the amount of this unidirectional J o u r n a l P r e -p r o o f amino acid in the BBB can be quantified [63]. Moreover, a large increased uptake on the sonicated side was observed, compared to the collateral side over time ( Figure 5). In addition, [3-11 C]AIB was shown to be stable in arterial plasma [28]. [3][4][5][6][7][8][9][10][11] C]AIB, can be suitable for assessing brain mechanisms after sonication, since 11 C has a sufficient half-life of 20 minutes and the radiotracer can be easily produced and is metabolically stable. The radiotracer is transported unidirectionally from the blood to the brain and has preferable kinetic properties for assessing BBB opening [28]. Moreover, [3-11 C]AIB was shown to be more sensitive than [ 18 F]FDG in differentiating between tumors and inflammation, especially in brain lesions, which is useful in monitoring treatment responses [64].
[ 18 F]-FBPA-Fr, as a radiotracer, has the ability to show specific brain tumor uptake in F98 gliomabearing rats [66]. The results show that the uptake in the sonicated tumor area was significantly higher than the uptake in the non-sonicated tumor area. Moreover, [ 18 F]-FBPA-Fr can typically pass through the BBB, but with FUS, the concentration of [ 18 F]-FBPA-Fr in the tumor area was significantly higher than that without FUS in the same targeted area [65] (Figure 6). Thus, [ 18 F]-FBPA-Fr seems to be a promising radiotracer for evaluating brain mechanisms following sonication due to the favorable halflife of 18 F at 109.8 min [65]. In addition, the combination of phenylalanine (BPA) and fructose was found to increase BPA solubility, which aids in increasing the efficacy of Boron Neutron Capture Therapy (BNCT) in the tumor [66]. Moreover, [ 18 F]-FBPA-Fr, in the preclinical studies, demonstrates high tumor-to-normal tissue uptake [66]. However, only a few studies have been published on FUS in combination with [ 18 F]-FBPA-Fr and [3-11 C]AIB, and the limitations of these radiotracers are still unrevealed. In addition, these two radiotracers were tested only in preclinical studies. Under normal physiological conditions, antibodies are unable to pass the BBB. Bevacizumab is a monoclonal antibody that affects the vascular endothelial growth factor and aids in reducing tumor J o u r n a l P r e -p r o o f size [67]. Although it has been approved as a treatment in recurrent glioblastoma, its use offered no significant benefit due to the difficulty in crossing the BBB. However, Liu et al. [68] conducted a study with animals to investigate whether the use of FUS enhanced the accumulation of [ 68 Ga] bevacizumab in brain tumors. The results show a significant accumulation in the sonicated area compared to the non-sonicated area, and the tumor progression with bevacizumab and FUS was significantly reduced compared to with bevacizumab alone (Figure 7). Thus, FUS is noted to enhance drug delivery in animal studies, especially when passing the BBB is difficult, and thus improves treatment.

4-Potential Radiotracers to evaluate BBB transporters:
PET allows understanding the mechanism of FUS and its effects on BBB transporters, such as P-gp function and GLUT-1. However, the optimal radiotracers to monitor these effects remain to be determined. For example, fluorine-18 fluorodeoxyglucose ([ 18 F]FDG) as a GLUT-1 tracer may not be suitable to assess BBB opening in the brain after sonication, since glucose uptake immediately after FUS is low [29]. Further, [ 18 F]FDG can cause non-specific uptake and false positive results [66][69], whereas [ 68 Ga]ethylenediaminetetraacetate (EDTA) was successfully used to assess BBB leakage after mannitol solution was used in Rhesus monkeys [70]. It is known that [ 68 Ga]EDTA cannot cross BBB in normal conditions, which makes it a suitable radiotracer to assess FUS effects on the BBB. [ 18 F]FLT can be potentially used to assess the permeability of the BBB, since it does not easily cross the BBB [71].
[ 11 C]-N-desmethyl-loperamide is a radiotracer with high potential for use to assess ABC transporters, especially P-gp. It is known as a potent P-gp substrate that is often used in clinical studies on PET [72]. This radiotracer was successfully used by Goutal et al. [62] to assess the function of P-gp. The PET tracers [ 11 C]Metoclopramide and [ 18 F]MC225 are defined as weak P-gp substrates that result in a higher brain uptake value in baseline conditions, and thus potentially are more sensitive to detect changes in P-gp function [73] [31]. Due to the higher initial brain uptake of the tracer,  [29]. This study measured glucose metabolism using a [ 18 F]FDG micro PET scan after applying FUS and microbubbles, as shown in Table 5. The results demonstrate a reversible reduction of glucose uptake after sonication compared to control brains, followed by a drop in GLUT-1 protein expression in the brain (Figure 8) [29]. It is known that [ 18 F]FDG can cross the BBB. It has been proven that, following sonication, the brain starts to re-establish the barrier function beginning at 8 hours from the first sonication [29]. We can conclude that [ 18 F]FDG is sufficiently sensitive to detect the metabolic changes in the brain following FUS. The cause of the decreased glucose and GLUT-1 protein levels after sonication remains unclear and requires further investigation [29]. However, alteration of glucose uptake in the brain was evident in patients with neurological diseases; thus, glucose metabolism can be used as a biomarker to detect brain deterioration or BBB disruption [74] [75].

Figure 8.
Given the aforementioned studies, out of ten studies, only seven featured radiotracers to monitor the effects of FUS eon BBB. Further, a portion of these radiotracers showed promising results in evaluating BBB integrity after applying FUS and may be translated to clinical studies in the future. However, with only several studies, it remains insufficient to determine which radiotracer is best to understand the physiology of the BBB after applying FUS and observing its effects on BBB transporters. Thus, further preclinical and clinical studies are needed to address the role of FUS in relation to PET and to assess BBB transport and its role in drug delivery.

Is FUS safe and ready for clinical application?
As the preclinical studies showed promising and safe results in opening BBB by FUS, the first study on humans was performed [55]. The main objective of this human study was to assess the safety of FUS in opening BBB in 5 subjects. In this clinical study, no major adverse events were detected during the procedure or during the follow-up. Moreover, no neurological disorder, hemorrhages, swelling, or deaths were observed. However, discrete round hypointensities were observed on gradient echo in two patients immediately after sonication, but they were no longer evident in the 24-hour J o u r n a l P r e -p r o o f follow-up MRI [55]. In another clinical study, for the first time, BBB was successfully opened temporarily in the primary motor cortex with no serious adverse events in 4 amyotrophic lateral sclerosis patients [76]. Furthermore, transient BBB opening was also safe and feasible in 5 patients with primary brain tumor and increased the efficacy of chemotherapy [77].
Translating FUS into the clinical field, especially in neurology and drug delivery, may benefit a large range of patients, especially those who are unable to undergo surgery [78]. In addition, FUS could enhance drug efficacy in the brain and improve treatment responses in patients and patients with conditions such as cancer or psychiatric and neurodegenerative diseases [3] [6]. However, before establishing FUS as a routine clinical procedure, its use should be monitored by neuroimaging modalities such as PET and MRI, for safety purposes [78].

Conclusion and future perspectives
Given the studies that were reviewed in this paper, we can conclude that FUS in combination with microbubbles is a feasible and safe method to reversibly enhance BBB permeability. The studies showed significant improvement in the manipulation of BBB permeability in cerebral drug delivery and therapy. However, to use these advancements in the context of neurodegenerative disease treatment, existing preclinical work needs to be translated into optimal protocols and clinical trials.

Rat
Delivered to Her2expressing tumor cells in the brain Successfully transplanted to the targeted brain region J o u r n a l P r e -p r o o f