INITIAL ASSESSMENT OF BOILING HISTOTRIPSY FOR MECHANICAL ABLATION OF EX VIVO HUMAN PROSTATE TISSUE

Abstract

Boiling histotripsy (BH) is a focused ultrasound technology that uses millisecond-long pulses with shock fronts to induce mechanical tissue ablation. The pulsing scheme and mechanisms of BH differ from cavitation cloud histotripsy, which was previously developed for benign prostatic hyperplasia (BPH). The goal of this work is to evaluate the feasibility of using BH to ablate fresh ex vivo human prostate tissue as a proof of principle for developing BH for prostate applications. Fresh human prostate samples (N=24) were obtained via rapid autopsy (<24 hours after death, IRB exempt). Samples were analyzed using shear wave elastography (SWE) to ensure that mechanical properties of autopsy tissue were clinically representative. Samples were exposed to BH using 10 ms or 1 ms pulses with 1 % duty cycle under real-time B-mode and Doppler imaging. Volumetric lesions were created by sonicating 1 – 4 rectangular planes spaced 1 mm apart, containing a grid of foci spaced 1 – 2 mm apart. Tissue then was evaluated grossly and histologically, and the lesion content was analyzed using transmission electron microscopy (TEM) and scanning electron microscopy (SEM). Observed SWE characterization of ex vivo prostate tissue (37.9 ± 22.2 kPa) was within the typical range observed clinically. During BH, hyperechoic regions were visualized at the focus on B-mode and BH-induced bubbles were also detected using Power Doppler. As treatment progressed, hypoechoic regions of tissue appeared suggesting successful tissue fractionation. BH treatment was twofold faster using shorter pulses (1 ms vs 10 ms). Histological analysis showed lesions containing completely homogenized cell debris, consistent with histotripsy-induced mechanical ablation. It was shown therefore that BH is feasible in fresh ex vivo human prostate tissue producing desired mechanical ablation. The study supports further work aimed at translating BH technology as a clinical option for prostate ablation.

INTRODUCTION

Histotripsy is a developmental high intensity focused ultrasound (HIFU) technology that induces non-thermal mechanical tissue ablation (Parsons et al. 2006, Khokhlova et al. 2011, Lin et al. 2014, Khokhlova et al. 2015, Eranki et al. 2018, Xu et al. 2021). Compared to existing clinical thermal HIFU regimes, histotripsy delivers sequences of shorter pulses (from microseconds to a few milliseconds) of higher acoustic energy (10–100-fold) at low duty cycle (< 1%). Central to the mechanism of histotripsy is formation of vapor/gas bubbles at the focus that leads to mechanical effects through ultrasound-bubble interactions (Maxwell et al. 2011, Simon et al. 2012, Pahk et al. 2017). The presence of these bubbles produces hyper-echogenicity during each pulse and the ensuing mechanical destruction of tissue eliminates tissue scatterers producing a hypoechoic cavity as treatment completes (Wang et al. 2009, Khokhlova TD et al. 2019). Collectively, this allows for real-time targeting, treatment monitoring, and evaluation of treatment outcomes on B-mode ultrasound, which provides minimal real-time feedback with purely thermal HIFU. Additionally, due to its non-thermal mechanism and rapidity of bioeffects, histotripsy minimizes heat-sinking and thermal spread that can limit the consistency and precision of HIFU thermal ablation. In general, the precision of thermal ablative techniques is highly dependent on the target tissue, the thermal dose, and the area over which the thermal dose is applied (Elhelf et al 2018; Gschwend et al 2021). However, it is generally accepted that a margin of at least 5 mm beyond a tumor region is needed to provide less frequent recurrence given the limitations of contemporary tumor localization techniques (re, MRI). As a result, histotripsy may offer a strategy to improve HIFU ablation.

Accordingly, histotripsy has been under development for several clinical applications including ablation of the prostate for both benign prostatic hyperplasia (BPH) and prostate cancer (PCa) (Schade et al. 2012a, 2012b, Hempel et al. 2010, Khokhlova VA et al. 2019, Dubinsky et al. 2019). Initial prostate studies used “cavitation cloud” histotripsy, which relies on microsecond-duration pulses and very high peak rarefactive (negative) pressures to induce microbubbles at the focus through shock scattering (Maxwell et al. 2011, Hall et al. 2009). Early pre-clinical work showed promise and the technology proceeded to phase 1 clinical trials for treatment of BPH using the Vortx Rx device (HistoSonics, Ann Arbor, MI) through a transperineal approach. Though men demonstrated a significant subjective improvement in their lower urinary tract symptoms for up to 6 months after treatment, there were no objective improvements in prostate volume, flow rate, or post-void residual (Schuster et al. 2018). The relatively small boney acoustic window and significant depth of the prostate for the transperineal approach, combined with devices pulse parameters may have limited the ability to produce significant ablation of BPH.

Our group has been developing an alternative histotripsy regime, termed “boiling histotripsy” (BH). BH uses milliseconds-long (<20 ms) pulses and nonlinear propagation effects to generate high-amplitude shock fronts in acoustic pressure waveform at the focus (Khokhlova et al. 2011). Super-efficient shock-wave focusing and shock-induced heating results in formation of a mm-sized vapor bubble at the focus within each pulse (Canney et al. 2010). Interaction of subsequent shock fronts with this bubble produces mechanical tissue ablation with negligible thermal effects through prefocal cavitation, acoustic atomization, and micro-fountaining (Khokhlova TD et al. 2011; Maxwell et al. 2011, Simon et al. 2012; Pahk et al. 2017). Ultrasonic atomization, i.e., emission of fine droplets, and micro-fountain formation are well-known phenomena that occur when a focused ultrasound wave propagating in liquid encounters an interface with air (Simon et al. 2012). Focal waveforms in BH fields are asymmetric and have very high peak positive pressures following the shock fronts. When reflected from the pressure-release boundary of the vapor-filled boiling bubble, the wave changes polarity, which leads to creation of very high negative pressures in tissue in front of the bubble. This results in prefocal cavitation and weakening or partial disintegration of the tissue. Acoustic radiation force caused by the BH beam pushes this tissue inside the vapor cavity forming a miniature acoustic fountain of tissue fragments, and consequent atomization occurs inside the vapor bubble. The process is repeated with each incident pulse, resulting in complete tissue liquefaction.

The BH method therefore relies on the shock amplitude instead of peak rarefactional pressure and has lower power requirements, which may make it more conducive to transducer miniaturization and a transrectal application for prostate indications. Herein, building on prior experiments (Khokhlova VA et al. 2019), we evaluated the feasibility of using BH to ablate fresh ex vivo human prostate tissue as a proof of principle for developing BH for prostate applications.

MATERIALS AND METHODS

Fresh human prostate tissue samples (N = 24) were obtained via rapid autopsy (<24 hours after death, IRB exempt). The sizes of samples varied from 4 cm³ to 15 cm³. Tissue samples were placed in phosphate buffered saline solution and degassed in a desiccant chamber for at least one hour with residual pressure < 0.1 bar. For all BH experiments, the samples then were embedded in 1% agarose gel (N = 16). For shear wave elastography measurements, the samples (N = 8) were kept in degassed water to perform imaging.

Shear Wave Elastography Imaging Tissue Characterization

Mechanical properties of the autopsy prostate tissue were analyzed in N = 8 samples using the shear wave elastography (SWE) imaging to ensure that no significant change in tissue stiffness occurred within 24 hours after death. To reduce artifacts at the edges of tissue samples, which were much stronger when they were embedded in agarose gel, the measurements were performed in a container filled with degassed water. Each tissue sample was placed in water on an absorbing silicone rubber layer, positioned so that the imaging was performed from the same direction as the BH exposure, and the areas of interest for quantitative measurement were chosen in the middle of the samples away from the artifact areas. An Aixplorer ultrasound system with grey-scale ultrasound imaging and SWE (SuperSonic Imagine, Aix-en-Provence, France) and a linear probe SL15-4 with an effective bandwidth of 4 to 15 MHz were used in the measurements.

Settings were optimized for depth of penetration using a thyroid gland preset with elasticity scale of 100 kPa. SWE images were obtained and recorded in the imaging plane through the center of each sample. For each measurement, the transducer was maintained in a steady position for 4 seconds until the images stabilized. In each image, the elastic (Young’s) modulus was measured within three 4 mm diameter circular regions of interest centered at 6 mm, 12 mm, and 18 mm depth in the sample. Each Young’s modulus measurement was performed three times and the average value was recorded. For analysis, the values obtained in all 24 measurements were used.

Boiling Histotripsy Sonication

Agarose embedded prostate tissue samples (N = 16) were placed in a custom holder submerged in degassed water and attached to a 3D positioning system (Precision Acoustics, Dorchester, UK) as shown in Figure 1A. BH pulses were delivered using a 1.5-MHz custom-made transducer of 80 mm diameter, 60 mm focal length, and 24 mm diameter central opening (Fig. 1B).

FIGURE 1.

(A) Experimental arrangement for BH treatment of ex vivo tissue under real time US imaging; (B) Photo of the BH transducer with P7-4 imaging probe placed in its central opening.

To ensure successful tissue liquefaction in all samples, the transducer was operated at the highest driving voltage of 240 V provided by a custom-made electronic driving system similar to the one described and characterized in our previous study (Maxwell AD et al. 2017). Acoustic power of the transducer and in situ focal pressures were estimated using measurement-based nonlinear modeling with an equivalent single-element source as a boundary condition (Rosnitskiy et al, IEEE UFFC 2027). Pressure distributions were measured and reconstructed from acoustic holography measurements (Sapozhnikov et al, JASA 2015) performed in a plane between the transducer and the focus using a calibrated hydrophone (HNA-0400, 1 mV/kPa at 1.5 MHz; Onda Corporation, Sunnyvale, CA) at low driving voltage (3 V). Geometrical parameters of an equivalent single-element source with a central opening and uniform distribution of the vibrational velocity on its surface were defined so that its axial pressure amplitude distribution matched the focal lobe (above – 6 dB level) of the corresponding experimentally obtained axial distributions as shown in Fig. 2a. Then, assuming linear dependence of the source pressure on the driving voltage (Maxwell et al. 2017), nonlinear simulations were performed for the equivalent source at the operational output (240 V) using a HIFU-beam software (Yuldashev et al. IEEE 2021). Simulations were performed in a layered medium “water-prostate” with the focus located 1 cm deep inside the prostate (Fig. 2B). Focal waveform was also derated from simulations in water, performed at lower voltage to compensate for attenuation in tissue (Khokhlova et al JASA 2011).

FIGURE 2.

(A) Axial pressure amplitude distribution reconstructed from acoustic holography measurements (dashed-dotted curve), directly measured by a hydrophone (solid curve), and simulated for an equivalent single-element source (dashed curve) at low transducer driving voltage of 3 V. (B) Geometry of nonlinear simulations for the equivalent source transducer in a “water-prostate” layered medium with the focus located 1 cm deep in tissue at the BH driving voltage of 240 V. (C) Two cycles of the focal waveform obtained from direct nonlinear simulations (B) (solid curve) and derated from simulations in free field in water (dashed curve).

The following parameters were used in acoustic simulations in water and tissue, correspondingly: sound speed 1490.6 m/s and 1559.5 m/s, density 997 kg/m3 and 1045 kg/m3, nonlinear parameter 3.5 and 4.8, thermoviscous absorption (sound diffusivity) 4.33 × 10–6 m2/s; and additional absorption in tissue of 1.2 dB/cm at 1.5 MHz with a power law of 1.1 (Duck, 1990). The aperture of the equivalent source was 65.8 mm, focal length 56 mm, central opening aperture 20 mm, characteristic source pressure amplitude that corresponded to 3 V driving voltage was10 kPa. Acoustic output of 240 V used in BH experiments corresponded to 734 W acoustic power of the equivalent source, 0.8 MPa characteristic pressure amplitude or 21 W/cm2 intensity on its surface. Peak pressures in the focal waveform, directly simulated and derated from simulations in water, were P+/P− = 122/−22 MPa and the shock amplitude was 135 MPa (Fig. 2C).

BH exposures were delivered to a rectangular grid containing 2 – 11 foci in the transverse directions along the scanning plane (Fig. 3A). Depending on the size and geometry of the samples, foci were spaced 1 –2 mm apart. Volumetric lesions were obtained by sonicating 1 – 4 scanning planes with 1 mm between them. Two BH exposure protocols with the same peak acoustic power were evaluated (N = 8 samples each). For both protocols, bubbles were detected at the focus by B-mode and Doppler imaging in all treatments, 100% of the treatment time. The first protocol (Fig. 3B) used pulses of 10 ms duration delivered at 1% duty cycle, 20 – 40 pulses per focus. This is the “standard” BH protocol that has been used for performing treatments of different tissues (Khokhlova TD et al. 2019, Wang et al. 2018), including an initial pilot study in human prostate (Khokhlova VA et al. 2019). The second protocol used shorter pulses of 1 ms duration, 1% duty cycle, and 75 – 150 pulses per sonication point, with the aim to accelerate the treatment (Fig. 3C).

FIGURE 3.

(A) Tissue samples embedded in transparent gel and scheme of the scanning grid; Diagrams of time sequences for (B) 10 ms long pulses and (C) 1 ms long pulses with 1% duty cycle.

In the first N = 4 samples tested for each BH pulse duration regime (N = 8 total), different spacing and number of pulses per focus were assessed to optimize BH treatments. Focus spacing and the minimum number of pulses needed to produce uniformly liquefied lesions were selected. Subsequently, N = 4 samples (N = 8 total) were treated for each BH pulse duration with the following BH sonication parameters: 1 mm spacing between the treatment foci and between the sonication planes; 30 pulses per focus for 10 ms pulses and 150 pulses per focus for 1 ms pulses.

Real-time-imaging feedback of the BH sonications and monitoring of the BH treatments were performed using a Verasonics V1 Ultrasound Engine (Kirkland, WA, USA). A P7–4 ATL probe operating in B-mode/power Doppler was placed within the central opening of the BH transducer (Fig. 1A, B). Doppler sequences were triggered by the BH driving electronics to generate images right after the end of each BH pulse (Li et al. 2014). B-mode images were also collected after the exposure for evaluating the outcome of the treatment.

Specimen Processing

After the initial N = 4 “optimization” BH exposures for each pulse duration regime, the samples were bisected for gross evaluation of lesion formation. Subsequently, for the optimized pulse parameters, N = 2 samples per pulse duration regime were formalin-fixed after BH exposures and processed for histologic assessment with Masson’s trichrome staining. Additionally, N = 2 samples for each regime were bisected for gross evaluation and then the lesion content was collected for ultrastructural analysis using transmission electron microscopy (TEM) and scanning electron microscopy (SEM).

RESULTS

Observed SWE imaging characterization of prostate tissue specimens was within the typical range observed clinically (Barr et al. 2017). Specifically, the measured Young’s moduli ranged from 11.9 kPa to 91.7 kPa, with a mean ± SD of 37.9 ± 22.2 kPa (Table 1). A representative photograph of an ex vivo prostate tissue sample with evidence of BPH and an example of B-mode and SWE images illustrating the range of tissue stiffness are shown in Figure 4.

TABLE 1.

SWE measured Young’s moduli (kPa) in three locations of eight ex vivo human prostate tissue samples.

Loc./Samp.	1	2	3	4	5	6	7	8	Total
Proximal	24.5	27.7	20.8	11.9	45.5	18.4	80.6	34.8
Central	29.8	23.6	12.6	19.8	47.8	16.6	82.7	55.2
Distal	40.7	38.4	24.3	26.7	59.6	21.9	54	91.7
Mean (SD)	31.7 (6.7)	29.9 (6.3)	19.2 (4.9)	19.4 (6.0)	51.0 (6.2)	19.0 (2.2)	72.4 (13.1)	60.6 (23.5)	37.9 (22.2)

FIGURE 4.

(A) Photograph of an ex vivo human prostate tissue sample with evident presence of BPH used in BH experiments; Representative B-mode (B) and SWE (C) images of a prostate sample showing the tissue stiffness (Young’s modulus) ranging from 30 to 75 kPa.

The time of BH treatments varied from 5 to 33 minutes for 1 ms pulses and from 10 to 60 minutes for 10 ms pulses, depending on the number of sonication points in one layer (2 – 11), and the number of layers (1 – 4). The geometry of the sonication grid was chosen following the size and shape of the samples. In all treatments, hyperechoic regions were visualized during BH sonications at the focus on B-mode (Fig. 5A) and BH-induced bubbles were also detected using Power Doppler mode (Fig. 5B). These areas persisted for at least 10–15 minutes after the BH exposure. As treatment progressed, hypoechoic regions of tissue appeared suggesting successful tissue fractionation and liquefaction of the lesion content (Fig. 5C). On gross inspection of BH treated tissue (Fig. 5D), lesions contained liquefied regions of a homogeneous suspension consistent with mechanical fractionation of tissue. As expected, treating with a higher “dose” of 100 pulses indicated by yellow dots in Fig. 5D, produced larger lesions compared to 30 pulses per focus, depicted by blue dots.

FIGURE 5.

(A) Appearance of hyperechoic regions on B-mode and (B) Power Doppler images during BH exposures; (C) Hypoechoic appearance of a volumetric lesion of merged BH lesions on B-mode images 15 min after BH exposure; (D) Bisected BH lesion consisting of one line of 6 discrete foci with 2 mm spacing. Two foci on the left were irradiated with 30 pulses/point (blue dots), four foci on the right – with 100 pulses per point (yellow dots) resulting in larger lesions. The red arrow shows the direction of the BH administration.

On gross inspection, lesions produced with longer pulses (10 ms vs 1 ms) showed appearance of slightly whitened tissue around the lesion cavity (Fig. 6B) which may indicate some thermal effect on the surrounding tissues (Fig. 6A,B). Such color change is a known effect resulting from thermal treatment of tissue. Above a specific thermal dose, tissue proteins will denature and coagulate. The changes in protein conformation and coagulation results in a color change in addition to an increase in opacity (Wang et al. 2018, Park et al. 2018, Zhou et al. 2021).

FIGURE 6.

Gross view (left column) and histologic appearance of Masson’s trichrome stained ex vivo human prostate tissue sections treated with (A) 1 ms and (B) 10 ms BH pulses obtained with 1 mm spacing between foci, and 1% duty cycle. i) Low power magnification of BH lesions with the lesion border indicated by the yellow dotted line. ii) High power magnification of the lesion contents (solid black box in i) demonstrating uniform ablation for both parameters in the lesion centers with debris <50 μm (arrowheads) in size. iii) High power magnification of the lesion borders (dashed black box in i) demonstrating an ~200 μm region containing frayed collagen bundles and spared fibromuscular components (arrowheads) between the completely ablated lesion and intact untreated tissue containing intact fibrillar collagen (F). The two treatments resulted in varying amounts of homogenization; the 1 ms treatment produced smaller tissue fragments (white arrowheads) than the 10 ms treatment where larger tissue fragments (black arrowheads) are present. The border for the 1 ms treatment was more defined compared to the 10 ms treatment. The 10 ms treatment resulted in a wide border of frayed collagen fibers (black arrows). In contrast, there was a narrower region of frayed collagen fibers at the border (white arrows) for the 1 ms. Fibrillar collagen (F) can be observed outside of the lesion.

The resulting lesion volumes closely approximated the rectangular geometry of the planned sonication grid for all treatments. With 10 ms pulses, the dimensions were ~1.5 mm larger than the grid size in all dimensions and with 1 ms the dimensions were ~1 mm larger in all dimensions. This expansion was expected and is consistent with the known focal lesion size (for each pulse duration) extending out from centers of the grid points at the lesion margins. Lesion evaluation performed grossly showed that tissue was fully liquefied within the sonicated volumes.

On histological analysis, lesions containing homogenized cell debris were observed for both 1 ms and 10 ms pulses, consistent with histotripsy induced mechanical ablation of glandular elements (Fig. 6). BH treatment with both pulse lengths resulted in uniform tissue homogenization containing <50 μm tissue fragments centrally. However, the use of shorter 1 ms pulses (with the same 1% duty cycle) required a larger number of pulses per focus vs. 10 ms pulses (150 vs 30, respectively) to achieve complete ablation but resulted in 2-fold acceleration of the treatment. The treatment speed evaluated grossly was 9 ± 1.7 mm³/min for 1 ms treatments and 4.5 ± 0.7 mm³/min for 10 ms treatments. The lesion of completely ablated tissue was surrounded by a margin, measuring <200 μm, of incompletely ablated tissue. The margin contained regions of intact smooth muscle and collagen fibrils, which is consistent with sparing of fibromuscular elements. Beyond that margin, glandular and fibromuscular elements appeared normal consistent with viable untreated tissue.

In TEM images of the liquefied lesion content, both treatment protocols displayed discrete regions with loss of cellular structure and indistinguishable cellular components with a gradient of remaining electron dense matter. However, with 10 ms pulses (Fig. 7B), there were regions where the cellular debris became more condensed and electron-dense vs. 1 ms pulses (Fig. 7A), which may indicate some thermal effect in the liquefied lesion content (Wang et al. 2018). In the SEM images, 10 ms pulses produced layers of fibrillar fragments of collagen covered by a thick layer of cell and protein debris, whereas 1 ms pulses resulted in thinner and cleaner fibrillar fragments with discrete globules of cellular and protein remnants.

FIGURE 7.

Transmission electron microscopy (TEM) and scanning electron microscopy (SEM) images of samples taken from BH lesions produced with 1ms and 10 ms pulses. Both treated lesions display loss of cellular structure. With the 10 ms pulses, there are regions where the cellular debris have become more condensed and electron-dense (*) which may indicate some thermal effect. In SEM images corresponding to the 10 ms pulses multi-layered fragments of fibrillar collagen covered by thick layer of cell and protein debris were observed whereas the 1 ms pulse treatment resulted in thinner and cleaner fibrillar fragments with discrete globules of cell and protein remnants.

DISCUSSION

Herein we report successful results of applying BH for mechanical ablation of fresh ex vivo human prostatic tissue. BH produced reproducible mechanical homogenization in prostate tissue samples using two pulse parameter sets, which was confirmed with histologic and ultrastructural analysis. BH treatment targeting, progression, and outcomes were readily monitored in real time on B-mode and Doppler ultrasound. The presence of BH-generated bubbles resulted in the appearance of a hyperechoic region at the treatment site during each pulse and mechanical destruction of tissue eliminated tissue scatters producing a hypoechoic cavity as treatment completed.

With the longer 10 ms pulses there may have been subtle thermal effects on the liquefied tissue debris seen grossly, and on TEM and SEM. As the same high acoustic power output (0.7 kW) used in this study was sufficient to generate BH for both 1 ms and 10 ms pulses, longer tissue heating with 10 ms and initiation of boiling early within each pulse may have resulted in whitening of the lesion border and specific changes of the liquefied lesion content observed in SEM and TEM images and related to the thermal effect. Further optimization experiments will aim on minimizing power output required for generating BH lesions with longer pulses and minimizing “dose” requirements for various pulse lengths.

This study builds on prior pre-clinical work developing cavitation cloud histotripsy as a treatment for BPH in which that technique was successfully used for transabdominal prostate ablation in in vivo canine studies (Schade et al. 2012a, 2013, Hempel et al. 2011, Hall et al. 2009) and ultimately led to the creation of a clinical device (Vortx ®, Histosonics) (Schuster et al. 2018). However, this device failed to produce objective evidence of successful prostate ablation (e.g. change in prostate volume, urine flow rate, etc.) (Schuster et al. 2018) and to date, no studies have detailed the effects of the shock scattering approach on human prostate tissue. As a result, this study represents a significant advance in the prostate histotripsy literature by providing definitive evidence of the ability of the BH approach to mechanically ablate human prostate tissue. It also offers the first ultrastructural characterization of the effects of BH on human prostate tissue on both TEM and SEM confirming that BH can disrupt prostate tissue into subcellular debris. Additionally, similar to what has been observed in canine prostate with cavitation cloud histotripsy (Hempel et al. 2011, Hall et al. 2009), BH ablation of ex vivo human prostate is precise with <200 μm border.

The results presented here were obtained using fresh human ex vivo tissue, therefore the possibility exists that the same BH treatments would not have the same effects in patients. However, several studies by our group have shown good correlation between ex vivo and in vivo tissue sensitivities in pre-clinical models evaluating BH ablation in porcine liver and kidney (Khokhlova TD et al. 2019, Khokhlova et al. 2014) and rat kidney (Schade et al. 2019). Additionally, the rapid autopsy tissue collected for these experiments had similar mechanical properties on SWE imaging to what one would expect clinically in BPH and prostate cancer patients (Barr et al. 2017, Rouvière et al. 2017). Specifically, in young patients without prostatic disorders, the stiffness of the peripheral and central zones ranges from 15 to 25 kPa, whereas the transitional zone exhibits stiffness below 30 kPa. With the development of benign prostate hyperplasia (BPH), the peripheral zone remains soft, whereas the transition zone becomes heterogeneous and stiff, with elasticity values ranging from 30 to 180 kPa (Barr et al 2017). A stiffness value of the peripheral zone greater than 35 kPa is suggestive of a malignancy (Barr 2017; Correas et al. 2015). Wei C et al. 2018 reported even higher stiffness of 82.6 kPa in malignant tumors compared with benign areas. All measurements presented in these papers, were performed by SE12-3 multi-frequency intracavitary probe with an effective 3–12 MHz bandwidth. Our studies were performed using an SL15-4 linear array transducer with an effective bandwidth of 4–15 MHz. Thus, the frequency of the transducer we used was similar to the frequency of the transducer used in clinics and our range of Young’s moduli (37.9 ± 22.2 kPa) measured in the samples was similar to clinical values for benign tissues and certain malignancies. As a result, we anticipate that the observations and developments from this study (with respect to BH pulse parameters) will translate into patients though the optimal BH parameters for clinical prostate ablation remain to be determined.

Prostate cancer tumors and BPH therefore are typically stiffer mechanically (greater than 35 kPa) (Barr et al 2017, Correas et al. 2015, Wei C et al. 2018), compared to normal prostate tissue (from 15 to 30 kPa) (Barr et al 2017). In previous studies, it was shown that the sensitivity of tissues to histotripsy correlates with their mechanical stiffness and collagenous tissue is more resistant (Vlaisavljevich et al 2014; Wang et al 2018, Khokhlova et al 2019). Correspondingly, BPH and prostate cancer tissue may have higher resistance to histotripsy, i.e. would require larger number of pulses per focus. Most of samples used in the experiments contained BPH identified grossly and by palpation and the BH treatment parameters were sufficient for liquefying them.

Non-thermal ablation with BH offers several potential advantages over existing minimally invasive treatment options for prostate diseases. First, BH’s mechanism of action using bubbles and mechanical bioeffects enable real-time ultrasound control not possible with other technologies. Second, the precision of BH ablation and lack of heat-sink and thermal diffusion effects (owing to the rapidity of its mechanism) are a major advantage over existing thermal based techniques (such as thermal HIFU, cryotherapy, etc.). We anticipate it will enable very tightly controlled ablation near critical structures such as the urinary sphincter and neurovascular bundles to minimize side effects such as urinary incontinence and erectile dysfunction when thinking about PCa focal therapy. Third, as this study demonstrated, refinement of BH pulse parameters enables tailoring of pulse parameters and bioeffects, accelerating treatments, and facilitating clinically relevant rates of non-thermal tissue ablation. Collectively, these advantages suggest that BH, unlike any other available minimally invasive therapy for prostate pathology, can be translated into a clinical treatment of prostate diseases.

The positive results of this study, combined with the potential advantages of the BH approach, support further investigation and development of BH for both PCa and BPH. Compared to cavitation cloud histotripsy, BH pulse parameters may be more amenable to miniaturization to facilitate the transrectal application that has been successfully used for thermal HIFU in prostate (Huber et al. 2020, Guillaumier et al. 2018). Indeed, simulations by our group suggested that BH of the prostate is feasible using a probe with similar geometry to existing thermal transrectal HIFU devices (Khokhlova et al. 2018). Following these studies, a preclinical system has been developed and pilot experiments in ex vivo and in vivo tissue are underway (Schade et al. 2020).

CONCLUSIONS

These data represent successful application of the BH method with real-time B-mode and Doppler-type imaging in fresh ex vivo human prostate tissue and suggest that BH can be accelerated without the loss of efficiency using shorter pulses of sufficient shock amplitude (1 ms vs 10 ms). Based on these encouraging results, further evaluation of BH for prostate applications is underway.