Ultrasound-emitting catheters have been applied to a wide variety of pathologies (Doomernik et al. 2011). The catheter is inserted into the thrombus allowing combined exposure to ultrasound and local drug infusion. A relatively uniform ultrasound field covering large clots can be obtained using a series of miniaturized unfocused transducers aligned in the lumen of the catheter (Soltani et al. 2007). The acoustic pressure decreases rapidly with distance from the catheter because the transducers are small and unfocused. Nevertheless, effective ultrasound-catheter sonothrombolysis was achieved in large thrombi caused by intracerebral hemorrhage (Newell et al. 2011).

Extracorporeal insonation of the thrombus, combined with an intravenous injection of a thrombolytic drug and/or microbubbles, is an appealing minimally invasive approach for the treatment of ischemic stroke. The cranial bone attenuates, reflects and distorts the ultrasound waves (Fry et al. 1970). Hence, bone constitutes a major obstacle for transcranial sonothrombolysis. The temporal bone acoustic window is routinely used for transcranial Doppler (TCD) examination of cerebral blood flow (Aaslid et al. 1982). TCD, or transcranial color-coded sonography (TCCS), exposure through the temporal bone accelerates rt-PA thrombolysis (Cintas et al. 2002; Alexandrov et al. 2004; Eggers et al. 2008). TCD and TCCS sonothrombolysis use similar frequencies (around 2 MHz) and similar pulsed-ultrasound waveforms, and comparable results were obtained in clinical trials (Saqqur et al. 2013).

TCD is operator dependent, and the lack of trained operators is an obstacle for widespread adoption (Barlinn et al. 2012). To overcome this limitation, a sonothrombolysis device has been designed in order to be simpler to put into place and is less operator-dependent than TCD (Barlinn et al. 2013). A frame is fixed to the patient’s head using simple landmarks. Eighteen transducers are mounted on the head frame in order to insonify the most frequent locations of thrombi during ischemic stroke through the temporal and suboccipital acoustic windows. The acoustic parameters used by this device are based on those used in FDA-approved TCD devices (2 MHz, 100 kPa derated peak rarefactional pressure). The safety and relevance of this device has been demonstrated in phase I and II clinical trials (Barreto et al. 2013).

Ultrasound waves at 2 MHz are transmitted poorly through the bone (Ammi et al. 2008), which prevents TCD examination in 18% of patients (Wijnhoud et al. 2008). Barlinn et al. (2012) estimated the in-situ acoustic pressure for 20 subjects treated with TCD in the TUCSON trial using a simple attenuation model based on CT data. The calculated in-situ acoustic pressures were higher for patients functionally independent at 3 months. Hence, some patients may receive an insufficient acoustic pressure for thrombolysis enhancement with TCD insonification.

Efficient acoustic transmission through the skull can be achieved in the sub-megahertz frequency range. Ammi et al. (2008) measured a pressure reduction resulting from bone between 77.1 and 96.6% at 2 MHz in 5 human skull specimens. A pressure reduction between 22.5 and 45.5% was observed at 120 kHz. Therefore, sub-megahertz sonothrombolysis may allow a more consistent insonation than 2-MHz TCD. Sub-megahertz (300 kHz) insonation was used in the TRUMBI clinical trial (Daffertshofer et al. 2005). However, a significantly higher rate of symptomatic hemorrhage occurred in the ultrasound group. Baron et al. (2009) numerically simulated TRUMBI trial insonation parameters and concluded that standing waves, due to acoustic reflections in the cranial cavity, could have caused unexpected local pressure maxima. Reflection from contralateral bone and associated constructive interference are less likely for 2-MHz TCD because the wave is significantly attenuated while propagating in the brain tissue. The TRUMBI trial and subsequent studies revealed the necessity of producing a well-controlled transcranial ultrasound field for safe and efficient sonothrombolysis.

Taking into account these findings, revised sub-megahertz transcranial sonothrombolysis devices were designed. A dual-frequency array capable of performing 2.5-MHz TCCS and emitting a 500-kHz sonothrombolysis beam was developed for transcranial insonation (Azuma et al. 2010). The transcranial ultrasound field produced by this device was evaluated in-vitro and the safety of this approach was demonstrated in a healthy primate model (Shimizu et al. 2012). Bouchoux et al. (2014) simulated 120 and 500-kHz transcranial ultrasound fields from the head CT scans of 20 ischemic stroke patients using a validated acoustic propagation numerical model (Bouchoux et al. 2012). Consistent and homogeneous insonation of the M1 segment of the MCA was achieved at both 120 and 500 kHz. A positioning strategy based on external head landmarks, which did not require the knowledge of the position of the thrombus, was proposed by Bouchoux et al. and was found to perform similarly to an optimized transducer positioning technique based on CT data analysis. Local acoustic pressure maxima, due to reflection from the contralateral bone, were well controlled and similar to the amplitude in the targeted M1 segment of the MCA.

Transcranial exposure with sub-megahertz ultrasound may allow simple and consistent insonation of the thrombus in ischemic patients, without excluding subjects with an insufficient temporal bone window for TCD or TCCS. Sub-megahertz sonothrombolysis devices should be developed and evaluated carefully. The transcranial ultrasound field can be evaluated by simulation or in-vitro measurements in several skulls. Moreover, sub-megahertz sonothrombolysis devices should be evaluated in a large animal model accounting for constructive interference due to standing waves. Random frequency modulation of the wave can also be employed to suppress standing waves (Tang and Clement 2010; Furuhata and Saito 2013).

Thrombolysis with focused ultrasound is also a promising approach. A highly focused ultrasound beam allows accurate spatial targeting of the clot. High-pressure amplitudes can be applied locally with a low risk of side effects outside the focal zone. Therefore, thrombolysis based on erosion of the clot by inertial cavitation may be possible with focused ultrasound without a thrombolytic drug. Maxwell et al. (2011) treated acute thrombi in juvenile pig femoral arteries using a 1-MHz transducer with a 1.9 × 13.5 mm focus (−6 dB). Similarly, a 1.5-MHz transducer (0.9 × 7.1 mm focal zone) was used to lyse clots in rabbit femoral arteries (Wright et al. 2012a). These studies demonstrated the feasibility of high intensity focused sonothrombolysis using accurate clot targeting. However, the focal zone is usually larger than the targeted vessels. Therefore careful studies on the side effects on tissue surrounding the clot are needed. The treatment of ischemic stroke with a MR-guided transcranial focused ultrasound device (ExAblate™ 4000, InSightec Inc., Tirat Carmel, Israel) was also proposed (Hölscher et al. 2013). This device is a hemispheric 1000-element 220-kHz phased array that can produce an electronically-steered 4 × 6 mm transcranial focal spot. Nevertheless, the work reported by Hölscher et al. (2013) is in a very early stage and more studies are needed to assess the efficacy and safety of this approach. Safe and efficient intracerebral hemorrhage (ICH) clot liquefaction with a similar MR-guided transcranial ultrasound device was demonstrated in an in-vivo porcine model and in a human cadaver model (Monteith et al. 2013). Though large aperture, focused ultrasound for transcranial insonation appears promising, numerical studies by Pajek and Hynynen (2012) indicate that these devices currently appear to be limited to treatment near the center of the brain (basilar artery and the proximal end of the M1 segment of the middle cerebral artery).

Thresholds for cavitation activity depend on the purity of the medium (Roy et al. 1985). However, cavitation thresholds in even the purest sources of water (Herbert et al. 2006; Bader et al. 2012; Maxwell et al. 2013) are much less than theoretically predicted (Church 2002). This discrepancy is generally attributed to nanometer-sized nuclei (Bader et al. 2012; Maxwell et al. 2013) that are difficult to extract from the medium (Flynn 1964). Such nuclei are however not likely to be found in-vivo due to the body’s natural filtration system. Instead, Yount (1979) proposed nuclei composed of an elastic organic skin of amphiphilic molecules which would stabilize bubbles against dissolution. The cavitation thresholds of endogenous nuclei in tissue are summarized by Church (2015). Maxwell et al. (2013) measured the thresholds for a wide range of ex-vivo tissues with short 1-MHz pulses, including clots, using a regime of therapeutic ultrasound termed histotripsy (Xu et al. 2004). The cavitation threshold of clots was similar to that of pure water, suggesting similar nanometer-sized nuclei to that observed in water.

Cavitation from endogenous nuclei has been observed during sonothrombolysis. Maxwell et al. (2011) observed microbubble clouds as regions of hyperechogenicity using 10-MHz diagnostic ultrasound imaging while insonating thrombi in a porcine femoral vein artery at 1 MHz. Additional studies have observed acoustic emissions from inertial (Everbach and Francis 2000; Wright et al. 2012a) or stable cavitation (Datta et al. 2006) during ultrasound-enhanced thrombolysis.

Introducing exogenous nuclei allows control of the type and location of cavitation activity. Hydrophobic particulates (Soltani 2013), perfluorocarbon droplets (Pajek et al. 2014) and focused laser pulses (Cui et al. 2013) have been used to nucleate cavitation effectively during sonothrombolysis. However, ultrasound contrast agents (UCAs), or stabilized microbubbles, are also used to nucleate cavitation. Excellent reviews of the principles of UCAs are provided by Stride and Saffari (2003) and Cosgrove (2006). The specific use of UCAs for sonothrombolysis is reviewed by de Saint et al. (2014). Briefly, most commercially available UCAs are composed of a high-molecular-weight inert gas core surrounded by a lipid shell (Faez et al. 2013). The use of a gas with low solubility in the blood stream and pegylation of the lipid shell increases the circulation lifetime (Sarkar et al. 2009). The large compressibility of the gas compared to the surrounding tissue results in a high backscatter of incident acoustic pulse. UCAs are excellent blood pooling agents for left ventricular opacification (Radhakrishnan et al. 2013), detection of focal lesions in the liver (Claudon et al. 2013) and evaluation of brain perfusion (Claudon et al. 2008). UCAs range in size from 1 to 10 μm in diameter (Bouakaz and de Jong 2007), which prevents filtration of UCAs by the lungs (Hogg 1987). In addition, UCAs of this size are resonant at diagnostic imaging frequencies (1–10 MHz). Current FDA-approved use of UCAs is restricted to left ventricular opacification, although there is a growing interest in using UCAs for therapeutic applications as well (Cosgrove and Harvey 2009; Stride et al. 2009).

To gauge the potential for generating therapeutic cavitation activity from UCAs, Bader and Holland (2012) investigated the nucleation and detection of stable cavitation numerically. The cavitation index (I _CAV ) was devised to gauge the likelihood of stable cavitation activity from UCAs in terms of the peak rarefactional pressure in MPa (P _r ) and center frequency in MHz (f):
Bader and Holland found there was an increased likelihood of nucleating cavitation activity by rupturing the UCA shell when 0.09 $$ ” src=”https://radiologykey.com/wp-content/uploads/2017/06/A317357_1_En_19_Chapter_IEq2.gif”>. Subharmonic emissions are more likely from inertial cavitation when 19.5).

In-vitro studies minimize the complexity associated with biological systems to assess the interaction of ultrasound with clots in a controlled setting. Clot models have been developed through strict protocols in order to produce consistent, clinically-relevant clots (Holland et al. 2008). Coagulation is initiated by incubating fresh or recalcified blood at physiologic temperature (37 °C) (Roessler et al. 2011). Coagulation can also be initiated by the addition of thrombin to plasma (Lauer et al. 1992; Suchkova et al. 1998). Fibrin clots, which are optically translucent, are also used for optical studies (Acconcia et al. 2013). The degree of clot retraction can be modified by altering the properties of the surface in contact with incubated blood (Sutton et al. 2013b), or by storing the clot at low temperatures (<4 °C) for several days post-incubation (Shaw et al. 2006). Thrombolytic metrics have included mass loss (Datta et al. 2006), dimensional reduction on an image (Cheng et al. 2005; Kim et al. 2012; Petit et al. 2012a), or the presence of fibrin degradation products (Francis et al. 1992; Kimura et al. 1994; Pfaffenberger et al. 2003; Alonso et al. 2009). Interested readers are invited to review an exhaustive list of recent in-vitro studies compiled by Petit et al. (2012b).

Ex-vivo studies incorporate excised living tissue as a first step towards understanding the interaction of vascular tissue exposed to thrombolytics and ultrasound. Excised arteries have been employed to determine the effect on the endothelium during sonothrombolysis (Rosenschein et al. 2000; Hitchcock et al. 2011). Clot models are similar to those used in-vitro, and mass loss is used as the thrombolytic metric.

In-vivo animal models allow the full biological response of a living organism to ultrasound and thrombolytics to be assessed, including the efficacy of the treatment, the ability to track and monitor treatment progress, physiologic alterations, and the potential for collateral damage. These studies are required prior to U.S. Food and Drug Administration approval for new clinical devices (Harris 2009). A variety of in-vivo thrombosis models have been developed and are summarized by Hossmann (1998), Verbeuren (2006), and Mousa (2010). Small animal models, such as rodent (Daffertshofer et al. 2004) and rabbit (Hölscher et al. 2012), are used in sonothrombolysis due to their low cost, ease of handling, and accumulation of data from previous studies. Large animal models, such as swine (Culp 2004) and primates (Shimizu et al. 2012), are needed to model the propagation of ultrasound through the skull and brain architecture. Thrombus formation can be initiated by incubation of autologous blood within a stenotic artery (Culp 2004; Damianou et al. 2014). High-intensity laser pulses can also be used to induce clot formation (Yamashita et al. 2009; Chen et al. 2013). Flow velocity (Maxwell et al. 2011), degree of thrombus size burden (Stone et al. 2007), or the presence of fibrin degradation products (Xie et al. 2005; Yamashita et al. 2009) can be used to assess thrombolytic efficacy. Potential side effects are evaluated using imaging techniques, post-mortem histology and immunohistochemistry techniques.

Insonation of thrombi to initiate lysis have been employed with and without thrombolytic drugs. UCAs have also been used to nucleate cavitation activity in both cases. The degree of thrombolysis is dependent on the ultrasound exposure conditions (Holland et al. 2008; Petit et al. 2012a).

Ultrasound is well established to enhance thrombolysis in-vitro (Datta et al. 2006; Prokop et al. 2007), ex-vivo (Hitchcock et al. 2011), and in-vivo (Culp 2004; Xie et al. 2013). Mechanical effects, particularly acoustic cavitation (Petit et al. 2012a; Wu et al. 2014; Bader et al. 2015), are clearly responsible for enhanced thrombolysis. These bench top experiments provide proof of concept for a particular insonation scheme, but the results are restricted to the particular clot model employed (Xie et al. 2011). Given the variability in thrombi composition in-vivo (Liebeskind et al. 2011), future studies should focus on both the insonation regime and subtype of clot in order to employ the optimal treatment regime. For example, thrombolytic enhancement was not possible for low-amplitude (<0.25 MPa peak negative pressure), sub-megahertz insonation of retracted clots treated with Definity^® and rt-PA (Sutton et al. 2013b). Other insonation schemes need to be developed for these stiff, retracted thrombi. Studies are also needed to determine the thrombus type, for example using elastography techniques (Viola et al. 2004) to determine the optimal ultrasound approach. Finally, methods to monitor and quantify the dose of cavitation energy, possibly employing PCI (Haworth et al. 2012), need to be further developed in order to monitor treatment progress.

Significant progress has been made over the past decade to translate ultrasound from the bench to the bedside. Catheter-directed ultrasound-accelerated thrombolysis has been successfully employed to treat peripheral arterial occlusions (Greenberg et al. 1999), stroke (The IMS II Trial Investigators 2007), deep venous thrombosis (Raabe 2006), and pulmonary embolism (PE) (Engelberger and Kucher 2014). Catheter-based ultrasound initiated complete or partial recanalization in 87.9% of cases in a review of 340 clinical cases (Doomernik et al. 2011). Overall rates of non-response (8.2%), complications (7.1%), bleeding (4.1%) and re-occlusion (2.1%) were low compared to conventional thrombolytic treatment. No distal embolization was reported during treatment. Engelberger and Kucher (2014) compiled the results from clinical studies of PE treatment with ultrasound catheter-directed thrombolysis and found similar clinical outcomes compared to administration of the thrombolytic tenecteplase, a genetically modified version of rt-PA (Baruah et al. 2006). In the first randomized trial employing an ultrasound catheter device, 59 patients with acute PE were treated with either the EKOS Endovascular System and rt-PA, or anticoagulation therapy alone (Kucher et al. 2013). The authors concluded ultrasound-assisted thrombolysis was the superior treatment versus anticoagulation at 24 h, without an increased bleeding risk.