Histotripsy induces apoptosis and reduces hypoxia in a neuroblastoma xenograft model

Abstract

Background

Neuroblastoma (NB) is the most common extracranial solid tumor of childhood, and high-risk disease is resistant to intensive treatment. Histotripsy is a focused ultrasound therapy under development for tissue ablation via bubble activity. The goal of this study was to assess outcomes of histotripsy ablation in a xenograft model of high-risk NB.

Methods

Female NCr nude mice received NGP-luciferase cells intrarenally. Under ultrasound image guidance, histotripsy pulses were applied over a distance of 4–6 mm within the tumors. Bioluminescence indicative of tumor viability was quantified before, immediately after, and 24 h after histotripsy exposure. Tumors were immunostained to assess apoptosis (TUNEL), endothelium (endomucin), pericytes (αSMA), hypoxia (pimonidazole), vascular endothelial growth factor A (VEGFA), and platelet-derived growth factor-B (PDGF-B). The apoptotic cytokine TNFα and its downstream effector cleaved caspase-3 (c-casp-3) were assessed with SDS-PAGE.

Results

Histotripsy induced a 50% reduction in bioluminescence compared to untreated controls, with an absence of nuclei in the treatment core surrounded by a dense rim of TUNEL-positive cells. Tumor regions not targeted by histotripsy also showed an increase in TUNEL staining density. Increased apoptosis in histotripsy samples was consistent with increases in TNFα and c-casp-3 relative to controls. Treated tumors exhibited a decrease in hypoxia, VEGF, PDGF-B, and pericyte coverage of vasculature compared to control samples. Further, increases in vasodilation were found in histotripsy-treated specimens.

Conclusions

In addition to ablative effects, histotripsy was found to drive tumor apoptosis through intrinsic pathways, altering blood vessel architecture, and reducing hypoxia.

GRAPHICAL ABSTRACT

Keywords:

• Histotripsy
• neuroblastoma
• hypoxia
• VEGF
• PDGFB
• apoptosis

Introduction

Neuroblastoma (NB) is the most common extracranial solid tumor in children, arising from the sympathetic nervous system and most commonly found in the adrenal medulla. Stratification depends on a myriad of factors, including grade of tumor differentiation, stage, patient age, and MYCN-amplification [1–5]. High-risk NB patients have a 5-year event-free survival rate of 50% [6] despite being subjected to an aggressive combination therapy regime, including chemotherapy, surgical resection, radiation therapy, autologous stem cell transplant, and immunotherapy [7,8]. Patients that do respond are susceptible to acute and long-term morbidity associated with treatment [9], motivating the development of alternative or adjuvant approaches.

A factor contributing to treatment resistance in high-risk NB is tumor hypoxia [10]. Low oxygen microenvironments in NB are associated with dedifferentiation of neural crest cells and stemness [11], deregulation of cellular processes leading to higher rates of proliferation, and apoptosis bypass via HIF1α, RhoA, VEGF, and ERK1/2 pathways [10]. These factors are associated with tumors resistant to chemotherapy and radiotherapy [12], increased angiogenesis [13], and increased metastasis [10,14,15]. The reduction of hypoxia has been used as a two-pronged strategy to both reverse these oncogenic mechanisms and improve efficient drug delivery [16,17]. Therefore, methods that promote favorable changes in tumor hypoxia could support existing NB interventions.

Ablation may supplement current NB therapies by targeting hypoxic cores and immediate surrounding tissue in solid tumors. Therapeutic ultrasound has gained traction for ablation of localized abdominal tumors in the adult population [18], in part because of its success in other conditions (e.g. kidney stones [19–21]). Histotripsy is a focused ultrasound therapy that disintegrates tissue into acellular debris through bubble cloud activity generated with short duration (∼ 1–20 µs) and highly tensile pressure pulses (peak negative pressures ∼ 25 MPa or greater) [19,22]. Several aspects of histotripsy may be advantageous for pediatric patients, including its noninvasive and nonionizing nature, and ability to spare critical structures such as large vessels and nerves [23]. Further, innate and adaptive immune cells are upregulated after histotripsy relative to other ablative therapies [24,25].

We hypothesized that histotripsy will cause ablation and apoptosis in NB xenograft tumors followed by an innate immune response. Based on our previous observations that histotripsy induces vascular dilation [26], we also anticipate changes in tumor oxygenation.

Materials and methods

Cell culture

NGP-Luciferase cells (DSMZ) were cultured in RPMI-1640 (Gibco) with 10% FBS, 1% pen/strep (Gibco), and 1% gentamycin, in 5% CO₂ maintained at 37 °C.

Xenograft model

We used the well characterized xenograft model consisting of 1x10⁶ NGP-Luciferase cells injected intrarenally in 6–8 weeks old female NCr nude mice [17]. Mice were randomly allocated into histotripsy or untreated groups. Tumors were treated (histotripsy or control) once they reached 1 to 2 grams (5–6 weeks after implant).

Luciferin

1.5 mg/10g of D-Luciferin (GoldBio) in DPBS without Ca²⁺ and Mg²⁺ was injected intraperitoneally into mice 18–25 min prior to imaging.

Histotripsy methodology

Figure 1 shows the experimental setup. Animals were anesthetized with ketamine/xylazine and hair was removed. Histotripsy pulses were generated with a 1-MHz focused source (75-mm focal length, 100-mm diameter) driven by a custom class D amplifier [27]. The animals were placed prone on a stage over a tank of degassed (∼20% dissolved oxygen), filtered (∼ 0.2 µm pore size), reverse osmosis water maintained at 37.3 ± 0.5 °C with a custom-built temperature controller circuit (ITC-308, Inkbird, Pengji Industrial Zone, Luohu District, Shenzhen, China). Prior to histotripsy exposure, the distal and proximal ends of the tumor boundary were identified using an imaging probe (L11-5v, Verasonics, Kirkland, WA, USA) coaxial to the histotripsy source driven by a research imaging scanner (Vantage 128, Verasonics, Kirkland, WA, USA). A customized MATLAB script (Mathworks, Natick, MA, USA) was used to identify a path along the tumor length (approximately 4 to 6 mm). The histotripsy source scanned along this path at a 0.33 mm/s rate using motorized stages (BiSlide, Velmex Inc., Bloomfield, NY, USA). During the translation, bubble clouds were generated within the tumor using ultrasound pulses of 35 MPa peak negative pressure and 2-µs duration applied at a 50 Hz rate. All mice were euthanized 24 h after histotripsy treatment.

Figure 1. Histotripsy reduces tumor viability and is a safe procedure. A) Experimental setup for histotripsy bubble cloud generation. Histotripsy pulses of 2-cycle duration and 35 MPa peak negative pressure were applied to a NB orthotopic murine model with a 1-MHz focused source. A coaxial imaging probe ensured bubble clouds were contained within the tumor as shown in inset, where histotripsy pulses propagated from top to bottom. The focused source was translated along the tumor with motorized stages. The streaks are constructive interference between the imaging and therapy pulses. Bar represents 1 cm. B) Representative images of luciferase tumor radiance measured at baseline, immediately after, and 24 h after histotripsy on IVIS or at equivalent time points for untreated controls. Red arrowheads point at the luciferase decrease after histotripsy. C) Radiance of Untreated controls (blue bars) had no significant change in radiance over 24 h (p = 0.8, n = 7) while radiance decreased by over 50% compared to baseline after histotripsy (green bars, p = 0.03, n = 7). Luciferase had not returned to baseline 24 h after treatment. Graph represents mean ± standard deviation. D) Treated mice lost less than 10% body weight compared to their pre-histotripsy weight over 24 h (p = n/s, n = 7). *=p < 0.05.

In Vivo imaging systems

An IVIS system 9 (Xenogen, Roche) captured bioluminescence 18–25 min post-injection at medium binning, automatic exposure, FOV D, and Fstop1. Images were acquired immediately before histotripsy, immediately after histotripsy, and 24 h after insonation.

Volumetric 3D images were captured using the automated settings from the IVIS system 9, and reconstruction using the default settings. These volumetric datasets were used to identify liver metastases [28], which were excluded from the study.

Pimonidazole

Animals received a 60 mg/kg intraperitoneal injection 30 min prior to sacrifice.

Histology

Paraffin-embedded tumors were sectioned to 5 µm thickness for immunohistochemistry. Apoptosis was interrogated using the Terminal deoxynucleotidyl transferase mediated dUTPNick End Labeling (TUNEL) kit ApopTag (Millipore).

Immunostaining antibodies

Endomucin (Santa Cruz SC-65495, clone V.7C7, 1:500), followed by rabbit anti-rat AlexaFluor488 (Invitro® clone A21210, 1:1000), αSMA (Alpha smooth muscle actin labeled with Cy3. Sigma C6198, clone 1A4, 1:1000), VEGFA (Epredia RB-9031-P1-A, 1:1000), PDGF-B (Abcam ab23914, 1:200), pimonidazole (Hypoxyprobe®, 1:100), c-casp-3 (Cell Signaling 9661 1:200).

Image acquisition and processing

Endothelium to pericyte interactions and c-casp-3 were imaged on Marianas spinning disk confocal microscope (Zeiss), using 40X oil objective, capturing z-stacks (10 images separated by 0.34 µm). Eight to ten stacks of each section/tumor were captured by focusing on the endomucin channel. Projections of maximum intensity of the z-stacks were visualized using FIJI (NIH). Digital image scans were captured with a VS200 Research Slide Scanner (Olympus) and ORca-Fusion camera (Hamamatsu Photonics). Individual images were created with the OlyVIA Viewer software (Olympus). TUNEL signals directly adjacent to the ablation zone were quantified using OlyVIA VS200 (Olympus). Regions of the tumor distal to the ablation zone were analyzed using FIJI to gate and quantify apoptotic cell signal counts. Endothelium to pericyte coverage was quantified on the z-stacks using FIJI, with distance measurement of αSMA immunostains adjacent to endomucin-positive signals as a percentage of the total endothelium length. Measurements of blood vessel lumens were performed with OlyVIA VS200 on endomucin-stained slides, measuring the widest diameter inside of endomucin stains. All vessels within a 20x random field of view were counted, with 100 measurements per tumor (n = 6 per group). VEGF area was normalized to the endomucin area of each blood vessel, and untreated controls were used as 100%. PDGF-B images captured with Axioskop2 (Zeiss) at 40X air objective, and intensity quantified with FIJI.

SDS-PAGE

Tumors were homogenized in RIPA buffer (Sigma) with PMSF and protease inhibitor (Roche) using a polytron (Kinematica). BCA (Pierce) was used to assess protein concentration: 50 µg of protein per tumor was loaded onto a 15% acrylamide gels and transferred to nitrocellulose membrane (Lycor). Primary antibodies: TNFα (Cell signaling 11948 1:1000), c-casp-3 (Cell Signaling 9661 1:1000), β − actin (Cell Signaling 3700 1:1000). Secondary antibodies: goat anti-rabbit 800CW and goat anti-mouse 680CW (IRdye, LiCOR) captured using Azure (Azure Biosystems). Band density was analyzed on FIJI.

Figures

Generated with Biorender.com and Adobe Illustrator.

Statistics

Radiance quantification in identical ROIs was analyzed by Repeated Measures ANOVA on Prism (Graphpad). Vascular lumen measurements were not normally distributed, and thus were log transformed. Linear regression with a random intercept was calculated using R statistical software 4.1.3, adjusting for the fact that measurements within the same tumor are more similar than those from different subjects. Student t-tests for TUNEL staining quantification and Fisher’s exact test were performed on Graphpad. PDGF-B data were not normally distributed; median and interquartile ranges were analyzed with the Mann-Witney test in Graphpad. Significance was ascribed to p values <0.05.

Results

Histotripsy is a safe procedure

Baseline tumor viability was assessed using an In-Vivo Imaging System (IVIS), after which mice received either histotripsy or no treatment (control group). Bioluminescence was used to determine time of treatment, selecting tumors between 4x10⁷ and 2x10⁸ p/sec/cm²/sr for histotripsy treatment, which corresponds to 1 to 2 grams. Histotripsy exposure was well tolerated by the animals, taking care to avoid the gas-filled intestine distal to the tumor. Bubble cloud generation was evident throughout the insonation as indicated by the presence of hyperechoic signals within the tumor under ultrasound imaging (Figure 1(A), Bubble Cloud). Gross examination of the animals at sacrifice indicated ablation was confined within the tumor, indicating that abdominal tumors were safely and effectively targeted [27]. Mice maintained over 90% of their baseline body weight 24 h post histotripsy (Figure 1(D), n = 7). Furthermore, mice had no change in body scoring over a 24-h period, and no animal required early removal from the study.

Histotripsy decreases tumor viability

Bioluminescence was captured before, immediately after and 24 h after histotripsy. Untreated controls were imaged at equivalent time points. Figure 1(B) depicts representative IVIS images of controls and histotripsy-treated animals: bioluminescence was reduced immediately after histotripsy (Figure 1(B), lower middle panel, red arrowheads). More than 50% reduction in the baseline luciferase radiance was observed immediately after histotripsy exposure (Figure 1(C) green bars, 43 ± 24% of baseline p = 0.03, n = 7). In contrast, no change in luciferase signal was observed in untreated mice over the same time (Figure 1(C) blue bars, 107 ± 45% of baseline, p = n/s, n = 7).

Histotripsy-treated tumors did not return to their baseline radiance after 24 h (Figure 1(C), green bars, 86 ± 70% of baseline, p = 0.03, n = 7), while untreated controls had no significant change relative to baseline (Figure 1(C) blue bars 134 ± 66% of baseline, p = 0.8, n = 7). Further volumetric analysis indicated histotripsy successfully ablated tumor areas with the highest radiance (Supplementary Videos 2 and 3, loss of red and orange signal). Bioluminescence remained low in the targeted regions during the 24 h imaging period despite being adjacent to viable, non-targeted portions of the tumor.

Histotripsy exposure induces apoptosis in NGP xenografts after 24 h

Histotripsy contributions to fractionation and apoptosis within tumors were assessed via the nuclear stain DAPI (blue, Figure 2(A) and the apoptosis marker TUNEL (red, Figure 2(A)), respectively. Tumors exposed to histotripsy displayed regions devoid of nuclei (fractionation), indicative of ablation and complete elimination of cells (Figure 2(A) bottom right panel, indicated by the * symbol). Regions of apoptosis as indicated by TUNEL staining were noted directly adjacent to the ablation zone (Figure 2(A), lower panels, labeled ‘Direct’). The width of the apoptotic region measured 1192 ± 470 µm (Figure 2(A), lower right panel, yellow line). To quantify the apoptosis in the areas directly targeted by histotripsy, we estimated the percentage of tumor positive for TUNEL stain within 1662 µm of the fractioned tissue (mean plus one standard deviation of apoptosis width across all tumors). Insonation resulted in 6-fold higher apoptosis rates compared to controls (Figure 2(B), 1 ± 0.1% of controls vs 27.8 ± 5.6% of histotripsy, p < 0.01, n = 3). Apoptosis was found throughout the tumor for histotripsy exposure (Figure 2(A), arrowheads, ‘Indirect’). The number of apoptotic cells outside of the direct fractionation zone and apoptotic margins were increased 2.5-fold for histotripsy exposure relative to the baseline apoptosis for controls (Figure 2(B), 985 ± 113, n = 3, vs 2490 ± 473 apoptotic cells/field, n = 6, p = 0.02).

Figure 2. Histotripsy induces fractionation and apoptosis directly as well as via TNFα 24 h after treatment. A) Representative images of the apoptosis marker TUNEL in controls (left panels) and histrotripsy-treated tumors (right panels), shown at low magnification (top row, bar = 1mm) or high magnification of white box (bottom row, bar = 200µm). Histotripsy resulted in a fractionated area devoid of nuclei denoted by *, bottom right panel), and surrounded by a rim of TUNEL-positive cells ‘Direct effect’. Tumor regions distant from the ablation site also had more TUNEL-positive cells (labeled ‘Indirect effect’) than controls. B) Direct apoptosis (areas adjacent to fractionation) was 28% higher in histotripsy (p < 0.0001). Apoptosis also increased 2.5-fold in Indirect areas (p = 0.02). C) SDS-PAGE of untreated controls vs histrotripsy-treated tumors under denaturing conditions reveals that histotripsy induced a four-fold increase in TNFα relative to controls, both full-length (25 kDa) as well as soluble forms (17 kDa) (lanes 1-6 are control tumors; lanes 7-12 are histrotripsy-treated tumors); graph quantifies both forms normalized to β-actin housekeeping gene as a % of controls (p < 0.05, n = 6 per group). D) Representative sections of c-casp-3 (green) in control or histotripsy-treated tumors: histotripsy increased c-casp-3 in tumor cells within Indirect areas (arrowhead). E) SDS-PAGE of tumor protein extracts revealed that histotripsy induced a four-fold increase in c-casp-3 (19 kDa) compared to controls (p < 0.05, n = 6 per group) (Lanes 1-6 are control tumors; lanes 7-12 are histrotripsy-treated tumors) Graph quantifies c-casp-3 normalized to beta actin housekeeping gene as a % of controls. *=p < 0.05. All graphs represent means and standard deviation.

Histotripsy triggers tumor necrosis factor alpha (TNFα) and cleavage of caspase-3

A potential mechanism for the observed histotripsy-induced apoptosis distal to the ablation zone could be macrophage polarization and cytokine release (nude mice lack mature T cells). Thus, the cytokine tumor necrosis factor alpha (TNFα) was compared for histotripsy-treated tumors to untreated controls. The cell-associated, full-length form of TNFα (26 kDa) and its soluble, cleaved (17 kDa) form were assessed. Figure 2(C) shows SDS-PAGE of tumor protein extracts, demonstrating full-length TNFα was increased almost five-fold in histotripsy-treated tumors compared to controls (100% ± 152 control vs 472 ± 323% histotripsy TNFα/β- actin, arbitrary units, p = 0.04, n = 6). Similarly, the cleaved form was increased four-fold in histotripsy-treated tumors (Figure 2(C), 100% ± 144 control vs 442% ± 299 histotripsy TNFα/beta actin arbitrary units, p = 0.04, n = 6). Depending on the biological context, TNFα either increases cell survival or apoptosis [29]. The cleavage of caspase-3 (c-casp-3) promotes TNFα-induced apoptosis, and is shown in Figure 2(D). An overall increase in c-casp-3 was observed in histotripsy-treated tumors relative to controls (Figure 2(D), green), and was elevated in tumor cells (Figure 2(D), arrowhead). SDS-PAGE revealed a four-fold increase in c-casp-3 after histotripsy exposure relative to controls (Figure 2(E), 100 ± 139% controls vs 429 ± 330% histotripsy c-casp-3/β-actin, arbitrary units, p = 0.048, n = 6). Together, these data suggest that histotripsy increases intratumoral levels of TNFα cytokine and induces cleavage of caspase-3, thereby promoting apoptosis in areas not directly targeted by histotripsy.

Histotripsy decreases hypoxia and VEGF in neuroblastoma xenografts

Control tissues displayed areas of positive pimonidazole staining consistent with tumor hypoxia (Figure 3(A), brown, black arrowheads). In contrast, no positive pimonidazole were observed in treated tumors (4/5 controls vs 1/7 histotripsy, p = 0.03). The absence of pimonidazole extended throughout the entire tumor (up to 5 mm diameter), including regions that were not exposed to bubble activity.

Figure 3. Histotripsy reduces hypoxia and VEGF in NGP xenografts 24 h post-treatment. A) Control tumors display regions that stain for the hypoxia marker pimonidazole (dark brown, arrowheads, bar = 1mm), while histotripsy resulted in its absence. B) VEGF expression is upregulated by hypoxia; representative images of control tumors with high VEGF (red, arrowheads) adjacent to endomucin-positive endothelial cells (green), while VEGF was reduced in histrotripsy-treated tumors (bar = 10µm). C) VEGF levels were signifcantly lower in histrotripsy-treated tumors compared to controls (p < 0.001, n = 4), confirming decreased hypoxia. Graphs represent means and standard deviation. **=p < 0.01.

Vascular endothelial growth factor A (VEGF) is upregulated by hypoxia as a HIF1α target [30], and regulates endothelial cell survival, pericyte recruitment, and tumor angiogenesis [31,32]. Treated tumors had reduced VEGF levels relative to controls (Figure 3(B), red color and white arrowheads). Quantitative analysis indicated VEGF was reduced to 12 ± 7% for histotripsy arms relative to control tumors (Figure 3(C), p < 0.001, n = 4).

Histotripsy disrupts endothelial cells and endothelial cell-pericyte interactions

PDGF-B is a VEGF target and regulates pericyte recruitment, survival, and proliferation in a paracrine manner [33]. Figure 4(A) shows histotripsy resulted in less PDGF-B (red color, arrowheads) adjacent to α-SMA positive pericytes (green). Quantification of PDGF-B confirmed a decrease after histotripsy (Figure 4(B), controls = 4 [3.2–5.9] vs histotripsy = 0.8 [0.32–2.3], median [IQR percentage of area], p = 0.016, n = 4). Given the role of PDGF-B in pericyte recruitment to endothelium, we investigated the relationship between endothelial cells and pericytes using the pericyte marker α-SMA and the endothelial marker endomucin. Confocal microscopy revealed histotripsy treatment resulted in a reduced coverage of endomucin-positive endothelial cells (Figure 4(C), green) by α-SMA pericytes (Figure 4(C), white arrows). Quantification revealed the percentage of endothelium with adjacent pericytes decreased by 46% after histotripsy, from 69 ± 6% to 32 ± 9% (Figure 4(D), p < 0.005, n = 4). The distance between pericytes and endomucin-positive endothelium increased two-fold after histotripsy (Figure 4(E), 1.1 ± 0.26 Untreated controls vs 3.3 ± 1.3 µm in histotripsy, p < 0.01, n = 3). No TUNEL staining was found to coincide with pericytes in indirectly targeted areas, confirming apoptosis or pericyte loss does not explain the loss of coverage (data not shown). To assess whether decreased pericyte coverage of endothelium resulted in vasodilation, the vascular lumens (delineated by the endothelial cell marker endomucin) were measured in both control and histotripsy-treated tumors (Figure 4(C), ‘Lumen’). Linear regression of log-transformed values resulted in an estimated increase of 53% after histotripsy (CI: 0.11–2.52) p < 0.0001, n = 6. Together, these results suggest histotripsy reduction in PDGF-B leads to decreased pericyte recruitment and vasodilation.

Figure 4. Histotripsy reduces pericyte coverage and induces vasodilation in NGP xenografts 24 h after treatment. A) Representative images of the VEGF target and pericyte recruitment ligand PDGF-B (red, arrowheads), lower in histrotripsy-treated tumors than controls. B) Quantification of PDGF-B ligand stains depicts lower PDGF-B levels as a result of histotripsy (p < 0.05, n = 4). C) Z-stacks captured with confocal microscopy shown as a 2D projection of the endothelial marker endomucin (green), the pericyte marker α-SMA (red) and nuclear DAPI stain (blue). Endothelial cells of untreated control tumors have a continual coverage of α-SMA pericytes lining (left panel). In contrast, histotripsy resulted in endothelial cells (green) lacking pericyte coverage (red), as well as higher separation between remaining pericytes and endothelial cells (yellow arrowhead). Scale bar: 10 µm. D) Quantification shows decreased endothelial cells (green) surrounded by neighboring α-SMA positive pericytes (red), p = 0.001). E) Pericytes distance from endothelium increases 3-fold after histotripsy (p < 0.01). F) Median vascular lumen of histrotripsy-treated tumors was two-fold wider than untreated controls, p < 0.01). *=p < 0.05. Graphs D and E represent means and standard deviation, B and F represent medians.

Discussion

Here, we show that a single histotripsy treatment decreases tumor viability. A margin of dense apoptotic cells directly adjacent to fractionated areas was observed, representing over 25% of the total tumor area (Figure 2(B)). These direct effects may be attributed to bubbles generated just outside the focal zone that cause sublethal damage to adjacent cells. Indirect effects of histotripsy were also observed, including an increase in TNFα levels associated with apoptosis via cleaved-caspase-3. Finally, histotripsy was found to decrease hypoxia, VEGF, and PDGF-B levels. These changes resulted in a decreased coverage and attachment of pericytes to the endothelium, which was associated with vasodilation. Together, our data indicate that tissue disintegration via histotripsy is localized to the focal region, though alterations to vasculature and apoptosis extend well beyond the immediate ablation zone.

It is important to highlight that histotripsy induced apoptosis both directly and indirectly. Apoptosis does not trigger multiple pro-inflammatory cytokines that lead to adverse side effects, and thereby is a desirable process to eradicate tumor cells [34]. Given that this study utilized nude mice lacking mature T cells, TNFα upregulation suggests a shift in macrophage polarization to anti-tumor activity after histotripsy [35,36]. Natural killer cells are known to secrete TNFα, and have pro-apoptotic effects, which may have contributed to the indirect effect observed here [37]. Abscopal effects following histotripsy ablation have been noted in pre-clinical studies [38] and a recent case report [39]. The precise mechanism for immune activation is still under investigation and will likely vary based on tumor type. Our results indicate that the innate immune system may be an important contributor to histotripsy-induced abscopal effects.

Although pericytes are known to express TNF receptors after stimulation [40], no evidence of pericyte apoptosis was observed in non-targeted areas in this study. Given the loss of PDGF-B after histotripsy, the most likely explanation for the changes in pericytes observed here is a lack of recruitment signaling. In turn, vasodilation is also promoted due to the histotripsy-induced reduction in PDGF-B and pericytes. The vasodilatory effects observed here could also be due to histotripsy-induced changes in the tumor structure and pressure [41–43]. Histotripsy liquefaction will reduce the tumor stiffness, which may reduce interstitial pressure and relax the vasculature resulting in a decrease in hypoxia. Histotripsy increased tumor cell apoptosis, which would decrease tumor oxygen requirements, contributing to the loss of hypoxia as demonstrated previously [44]. Hypoxic tumors are correlated with poor outcomes and resistance to chemotherapy, immunotherapy, and radiation, all of which are standard of care in high-risk NB treatment. Thus, reduction in hypoxia suggests histotripsy is well poised to be tested as an adjuvant therapy for NB. However, based on the experiments conducted in our study, it is not possible to determine what causes the reduction in hypoxia. Long-term studies will determine the duration of the observed decrease in hypoxia and its consequences. Clarifying the optimal window for this phenomenon may contribute to enhanced drug delivery and better therapeutic outcomes. Finally, the near loss of VEGF suggests tumor angiogenesis will be hampered as a result of histotripsy, providing additional tumor growth control.

Taken together, these data suggest that there is a robust global response of NB xenografts after histotripsy treatment, suggesting susceptibility to combination therapy treatment. These data are promising for future studies examining the combined effect of histotripsy with chemotherapy and immunotherapy.

Classification

major: Cell Biology, minor: applied Physical Sciences.

Author contributions

Conception and design: HSL, BKB.

Development of methodology: BKB, HSL, IJI, WLL, FFG.

Acquisition of data: HSL, BKB, IJI, WLL, FFG, SR, VGP, CSC.

Analysis and interpretation of data: HSL, BKB, IJI, WLL, FFG, NR, JJK.

Writing, review, and/or revision of the manuscript: HSL, BKB, AM, CC, WL, IJI, JJK.

Administrative, technical, or material support: HSL, BKB, IJI, WLL, FFG.

Acknowledgments

We thank The University of Chicago Integrated Light Microscopy Core, especially Shirley Bond and Christine Labno, for their assistance with slide scanning and image analysis, which receives financial support from the Cancer Center Support Grant (P30CA014599). RRID: SCR_019197.

Disclosure statement

None of the authors have competing interests.

Data availability statement

All data and protocols are available: 10.6084/m9.figshare.22557934.

Additional information

Funding

This work was supported by Focused Ultrasound Foundation; Feis Family Fellowship; Sorkin Fellowship. Funding for this project was provided by the Pediatric Cancer Foundation, The University of Chicago Human Tissue Resource Center, the Feis Family, the American Cancer Society (Grant # RSG-21-171-01-ET) and National Cancer Institute (Grant # R01 CA235756).