Ultrasound-responsive micro/nanobubbles based intelligent theranostic systems for precision tumor therapy Authors: Deng Liu, Yi Ling, Wanqian Liu, Jun Deng, Yanli Guo Published: 2025 ABSTRACT Ultrasound-targeted micro/nanobubble cavitation (UTMC/UTNC) has emerged as a highly promising ultrasound-based strategy for precision tumor therapy. This technique harnesses microbubbles and nanobubbles as cavitation nuclei that respond to ultrasound, inducing cavitation effects, generating mechanical forces that transiently permeabilize biological barriers, enhance vascular and cellular permeability, and induce localized tumor cell disruption. These cavitation-mediated effects enable spatiotemporally controlled drug release, markedly improving intratumoral drug accumulation and maximizing therapeutic efficacy while reducing systemic toxicity. This review provides a comprehensive overview of the mechanistic foundations and therapeutic applications of UTMC/UTNC. The physical mechanisms by which bubbles act as cavitation nuclei, undergoing oscillation, expansion, collapse, or rupture under ultrasound stimulation, and their resulting biological effects are discussed in detail. Furthermore, emphasizing the roles of UTMC/UTNC in enhancing chemotherapy, sonodynamic therapy, immunotherapy, gene delivery, radiotherapy, and ferroptosis is reviewed. These therapeutic enhancements are primarily attributed to improved drug delivery and cavitation-induced tumor cell disruption. Finally, key challenges and limitations associated with UTMC/UTNC-mediated tumor therapy are discussed, along with prospects for clinical translation.Graphical Abstract Schematic illustration of MBs/NBs-based theranostics: bioimaging, ultrasound-triggered physical mechanisms with consequent biological effects, and enhanced cancer therapies via UTMC/UTNC. (Created with bioRender.com) Introduction Ultrasound-targeted micro/nanobubble cavitation (UTMC/UTNC) is a non-invasive ultrasound-mediated therapeutic strategy that employs low-frequency ultrasound to induce bubble oscillation and collapse. The resulting cavitation generates mechanical effects, such as microjets and shock waves, that increase cell membrane and vascular permeability, disrupt microvessels, promote drug penetration and perfusion, and directly induce tumor cell death [1–4]. Compared to conventional high-intensity focused ultrasound (HIFU), UTMC/UTNC induces less thermal damage because it operates at a lower acoustic intensity, and its bioeffects are predominantly mediated by mechanical cavitation. This technique has been extensively investigated for treating a wide range of diseases, including cancer [5, 6], central nervous system disorders [7–9], cardiovascular diseases [10–12], and metabolic disorders [13–15], demonstrating promising therapeutic potential. UTMC/UTNC enables spatiotemporally controlled, reversible, and non-invasive modulation of biological barriers and local drug release, effectively minimizing off-target toxicity and improving therapeutic outcomes. Moreover, cavitation-induced mechanical injury and physical disruption elicit stronger immunogenic responses than apoptosis, potentially enhancing tumor immunotherapy via robust adaptive immune activation [16–18]. Despite progress in this field, a systematic understanding of the mechanisms by which UTMC/UTNC augments tumor therapy remains limited, which hinders broader clinical translation. Cavitation nuclei and ultrasound are central to the UTMC/UTNC process. The introduction of exogenous cavitation nuclei, primarily conventional gas-filled bubbles, reduces the cavitation threshold (acoustic pressure) and increases cavitation probability. Other nanomaterials with cavitation potential include phase-change nanodroplets [19, 20], gas vesicles (GVs) [21–24], and hollow solid nanoparticles [25, 26]. These agents can also be used for contrast-enhanced ultrasound (CEUS) imaging. Among them, bubbles have become prominent ultrasound contrast agents in biomedical imaging and therapy and can be classified by size into microbubbles (MBs) and nanobubbles (NBs). Clinically approved MB agents such as SonoVue and Sonazoid are pure blood-pool agents; following intravenous administration, they enable dynamic real-time assessment of perfusion in normal and pathological tissues. These agents are extensively employed for tumor diagnosis, boundary delineation, and therapy monitoring, demonstrating excellent biosafety and a low incidence of adverse reactions [27–29]. In contrast, the smaller size of NBs allows them to extravasate through tumor vasculature into the interstitium, facilitating tumor-specific molecular ultrasound imaging and treatment monitoring [30, 31]. Bubbles can also be engineered to carry bioactive cargos, including genes, drugs, proteins, and gene-silencing agents, and further modified for targeting or bionic modification to integrate diagnosis and therapy. The outcomes of UTMC/UTNC are influenced by multiple parameters, including bubble dose, ultrasound intensity and frequency, and bubble composition. In clinical ultrasound applications, distinguishing between acoustic environments for imaging and therapy is essential. Low mechanical index (MI) ultrasound is typically used for contrast-enhanced imaging, where bubbles undergo stable oscillation without rupture, thereby enhancing the acoustic contrast between blood and tissue. In comparison, high-MI ultrasound intentionally induces inertial cavitation, resulting in bubble collapse, microjetting, and localized mechanical disruption, the core mechanisms underlying UTMC/UTNC-mediated therapy. This review primarily focuses on the latter, summarizing how controlled bubble destruction enhances intratumoral drug delivery and therapeutic efficacy, while briefly introducing diagnostic imaging as a complementary tool for real-time monitoring and guidance. Thus, understanding bubble physicochemical properties, their ultrasound-responsive behaviors, and interactions with surrounding biological environments is essential for optimizing targeted drug delivery and designing next-generation bubble platforms. Given the pivotal role of ultrasound and bubbles in oncology, this review aims to elucidate their unique contributions to tumor diagnosis and therapy. We first compare the physicochemical and functional properties of MBs and NBs. We then examine the physical mechanisms underlying UTMC/UTNC and their consequent biological effects, including cavitation, sonoporation, and drug delivery. We also summarize recent advances in chemotherapy, sonodynamic therapy (SDT), immunotherapy, gene delivery, radiotherapy, and ferroptosis under UTMC/UTNC conditions. Finally, we highlight current challenges and future directions for clinical translation, offering insights into the design of synergistic ultrasound-bubble strategies for precision cancer theranostics (Fig. 1). Fig. 1Schematic illustration of MBs/NBs-based theranostics: bioimaging, ultrasound-triggered physical mechanisms with consequent biological effects, and enhanced cancer therapies via UTMC/UTNC. (Created with bioRender.com) Schematic illustration of MBs/NBs-based theranostics: bioimaging, ultrasound-triggered physical mechanisms with consequent biological effects, and enhanced cancer therapies via UTMC/UTNC. (Created with bioRender.com) Bioimaging of MBs and NBs in tumors Ultrasound contrast agents generate both strong fundamental and harmonic signals via nonlinear oscillation of their gas-filled cores under an acoustic field. The harmonic signals are predominantly utilized in nonlinear imaging modes to markedly enhance imaging contrast. These agents have been widely applied in clinical diagnostics for various diseases. Based on core composition and structure, they can be classified as gas-filled bubbles (MBs, NBs), phase-changed nanodroplets, GVs, and hollow solid nanoparticles, all of which encapsulate gas that serves as both the acoustic contrast medium and cavitation nucleus. Gas-filled bubbles typically consist of a shell surrounding a gaseous core; their hydrodynamic diameters define them as NBs (< 1 μm) or MBs (1–10 μm). Bubbles exceeding 10 μm cannot traverse capillary networks, posing risks of microcirculatory obstruction. In addition, they are rapidly sequestered by pulmonary capillaries, exhibit short circulation half-lives, and vibrate nonlinearly if driven at a low (sub-MHz) frequency, rendering them unsuitable for CEUS. Similarities and differences between MBs and NBs A summary of different chemical formulations and synthesis strategies for MBs and NBs is provided in Table 1, and a comparison of MBs and NBs differences is summarized in Table 2. Both share core-shell composition, targeting ligand conjugation, and drug-loading strategies. However, their stability is governed by internal Laplace pressure(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:P$$\end{document}). The calculation formula is as follows:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:P=\frac{2\sigma\:}{r}$$\end{document} where σ is surface tension and r is bubble radius. Under identical shell materials and ambient fluid pressure P0, larger MBs exhibit lower internal pressure than smaller NBs (PMB=P0+\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:P$$\end{document}MB< PNB=P0+\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:P$$\end{document}NB), rendering MBs intrinsically more stable (Fig. 2a) [32]. The stability of MBs strongly depends on the properties of their shell. PEGylated lipids prolong circulation time by introducing steric hindrance, which prevents bubble coalescence [33–35]. Furthermore, a fluorinated inner layer (PEG2k–PFC) with low interfacial tension and low Laplace pressure can further enhance MB stability when in contact with the internal gas phase [36, 37]. Multilayer shells further retard gas diffusion and dissolution [38]. Bubble resonance in liquid is described by the Minnaert equation (ignoring shell elasticity and damping):\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{f}_{0}=\frac{1}{2\pi\:r}\sqrt{\frac{3\gamma\:{P}_{0}}{\rho\:}}$$\end{document} where γ is the specific heat ratio, P0 is ambient pressure, ρ is liquid density, and r is the bubble radius [39, 40]. Larger bubbles resonate at lower frequencies; maximum oscillation amplitude and thus strongest scattering occurs when the ultrasound frequency approaches f0. Because clinical frequencies used in clinical ultrasound probes (2–15 MHz for diagnostic applications) align closely with the resonance range of MBs (1–5 μm diameter, f0 ≈ 1–6 MHz), MBs provide superior acoustic responsiveness. In contrast, NBs (0.1–0.5 μm diameter, f0 ≈ 12–60 MHz) exhibit resonance frequencies above conventional diagnostic ranges [40, 41]. The acoustic scattering cross-section of a single spherical gas bubble scales with its radius, and a 200 nm bubble exhibits a scattering cross-section approximately four orders of magnitude lower than that of a 1 μm bubble. Consequently, MBs yield markedly stronger CEUS signals than NBs under standard diagnostic frequencies. The reported apparent echogenicity of NBs in both linear and nonlinear modes may partially originate from residual microbubbles. However, gas absorption and coalescence of NBs under high acoustic pressures may promote the formation of larger cavitation bubbles, enhancing CEUS contrast. Despite this limitation, NBs retain distinct advantages for drug and gene delivery, though higher bubble concentrations, ultrasound frequencies, or acoustic intensities may be required to achieve comparable imaging or therapeutic outcomes [42]. Moreover, MBs remain confined to the intravascular space, whereas NBs exploit the enhanced permeability and retention effect to extravasate into the tumor interstitium; when functionalized with targeting ligands, NBs can selectively bind tumor cell markers for precision molecular imaging [30, 31, 43]. Tumor-leaked NBs also demonstrate prolonged retention relative to MBs, and targeted NBs exhibit further delay in decorrelation time due to receptor-ligand interactions (Fig. 2b) [44]. Fig. 2The difference between MBs and NBs. (a) Under identical solution conditions (pressure = P0), the Laplace pressure of a gas bubble is inversely related to its radius; thus, given rMB > rNB, the Laplace pressure PMB < PNB. (b) Schematic illustration of vascular wall penetration by intravenously administered MB, NB, or targeted NB. (Created with bioRender.com) The difference between MBs and NBs. (a) Under identical solution conditions (pressure = P0), the Laplace pressure of a gas bubble is inversely related to its radius; thus, given rMB > rNB, the Laplace pressure PMB < PNB. (b) Schematic illustration of vascular wall penetration by intravenously administered MB, NB, or targeted NB. (Created with bioRender.com) Table 1Summary of the physicochemical and functional characteristics of MBs and NBs formulations with different chemical compositionsEncapsulation typeRepresentative shell materialsCommon ligand-conjugation chemistriesBiocompatibility and toxicityClinical statusRecommendationReferencesLipid-based shellsPhospholipids (e.g., DSPC, DPPC), Cholesterol, PEGylated lipids (e.g., DSPE-PEG2000)Lipid-anchored maleimide-thiol coupling, EDC/NHS amide coupling, and biotin–streptavidin interactionExcellent biocompatibility and prolonged circulation timeWidely used clinically (e.g., lipid-based SonoVue and Sonazoid)Flexible lipid shells facilitate nonlinear oscillations and enhance CEUS signals; preparation methods are mature, though drug-loading capacity and long-term stability remain limited.[30, 31, 45, 46]Polymer-based shellsPLGA, Polyvinyl alcoholEDC/NHS amide coupling, thiol–maleimide linkage, carbodiimide chemistries and click chemistryGood mechanical stability and biodegradability, albeit with risks of acidic degradation byproducts and subsequent inflammatory responsesPrimarily in preclinical stage, favored for mechanical robustnessPolymer shells offer high mechanical strength and stability, enabling controlled drug release, though their low compressibility may reduce cavitation efficiency.[12, 46]Protein-based shellsAlbuminEDC/NHS amide coupling, thiol–based couplingExcellent biocompatibility and prolonged circulation time, although protein antigens may induce immune responsesSome clinically approved platforms historically used albumin shells (e.g., Albunex, Optison)Protein-based shells provide uniform rigidity and biocompatibility but exhibit variability in mechanical uniformity.[38, 47]Hybrid lipid–polymer shellsLipid, PLGA, etc.Lipid insertion of ligand-PEG or EDC/NHS-mediated covalent coupling to the polymer shellBalanced biocompatibility and mechanical robustness; increased manufacturing complexityIncreasingly reported in preclinical studiesHybrid lipid–polymer shells combine the acoustic advantages of lipids with the high stability and loading capacity of polymers, albeit with more complex fabrication processes.[48]DSPC, 1,2-Distearoyl-sn-glycero-3-phosphorylcholine; DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; DSPE-PEG2000, 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000]; PLGA, Poly (lactic-co-glycolic acid); EDC/NHS, 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide / N-Hydroxysuccinimide Table 1Summary of the physicochemical and functional characteristics of MBs and NBs formulations with different chemical compositionsEncapsulation typeRepresentative shell materialsCommon ligand-conjugation chemistriesBiocompatibility and toxicityClinical statusRecommendationReferencesLipid-based shellsPhospholipids (e.g., DSPC, DPPC), Cholesterol, PEGylated lipids (e.g., DSPE-PEG2000)Lipid-anchored maleimide-thiol coupling, EDC/NHS amide coupling, and biotin–streptavidin interactionExcellent biocompatibility and prolonged circulation timeWidely used clinically (e.g., lipid-based SonoVue and Sonazoid)Flexible lipid shells facilitate nonlinear oscillations and enhance CEUS signals; preparation methods are mature, though drug-loading capacity and long-term stability remain limited.[30, 31, 45, 46]Polymer-based shellsPLGA, Polyvinyl alcoholEDC/NHS amide coupling, thiol–maleimide linkage, carbodiimide chemistries and click chemistryGood mechanical stability and biodegradability, albeit with risks of acidic degradation byproducts and subsequent inflammatory responsesPrimarily in preclinical stage, favored for mechanical robustnessPolymer shells offer high mechanical strength and stability, enabling controlled drug release, though their low compressibility may reduce cavitation efficiency.[12, 46]Protein-based shellsAlbuminEDC/NHS amide coupling, thiol–based couplingExcellent biocompatibility and prolonged circulation time, although protein antigens may induce immune responsesSome clinically approved platforms historically used albumin shells (e.g., Albunex, Optison)Protein-based shells provide uniform rigidity and biocompatibility but exhibit variability in mechanical uniformity.[38, 47]Hybrid lipid–polymer shellsLipid, PLGA, etc.Lipid insertion of ligand-PEG or EDC/NHS-mediated covalent coupling to the polymer shellBalanced biocompatibility and mechanical robustness; increased manufacturing complexityIncreasingly reported in preclinical studiesHybrid lipid–polymer shells combine the acoustic advantages of lipids with the high stability and loading capacity of polymers, albeit with more complex fabrication processes.[48]DSPC, 1,2-Distearoyl-sn-glycero-3-phosphorylcholine; DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; DSPE-PEG2000, 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000]; PLGA, Poly (lactic-co-glycolic acid); EDC/NHS, 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide / N-Hydroxysuccinimide Summary of the physicochemical and functional characteristics of MBs and NBs formulations with different chemical compositions DSPC, 1,2-Distearoyl-sn-glycero-3-phosphorylcholine; DPPC, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine; DSPE-PEG2000, 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000]; PLGA, Poly (lactic-co-glycolic acid); EDC/NHS, 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide / N-Hydroxysuccinimide Table 2Summary of differences between MBs and NBsFeatureMBsNBsReferencesSize1–10 μm (most clinically MBs are 1–5 μm)< 1 μm [49]Laplace pressureLower pressure (PMB) due to a larger radiusHigher pressure (PNB) due to a smaller radius [32]StabilityClinically used MBs often exhibit rapid signal decay; Modification of the shell composition with a fluorinated inner layer enhances bubble stability.Under identical shell conditions, NBs exhibit poorer stability; with optimized lipid formulations, NBs can exhibit slower decay and more persistent ultrasound signals. [36, 37]Gas volumeLarger per-bubble gas volumeSmaller per-bubble gas volume [50]Acoustic scattering efficiencyA scattering cross-section up to four orders of magnitude greater than that of NBsSignificantly weaker scattering [42]CEUSExcellent for CEUS and perfusion assessmentLimited CEUS ability; better suited for targeted imaging [42]Circulation timeShorter circulation time; rapid pulmonary filtration (e.g., SonoVue)Longer circulation time; slower clearance by the reticuloendothelial system [5, 51, 52]Cavitation thresholdLower acoustic pressure threshold and conventional clinical ultrasound sufficeHigher acoustic pressure is required to induce cavitation [42]Vascular permeabilityRemain intravascular due to large diameter; Lesser leakage in capillariesCapable of extravasation through leaky tumor vasculature; Greater leakage in capillaries [44, 53]Decorrelation time0.09 ± 0.03 s in tumor0.82 ± 0.59 s in skin6.04 ± 3.57 s in tumor1.06 ± 0.89 s in healthy tissue [44] Table 2Summary of differences between MBs and NBsFeatureMBsNBsReferencesSize1–10 μm (most clinically MBs are 1–5 μm)< 1 μm [49]Laplace pressureLower pressure (PMB) due to a larger radiusHigher pressure (PNB) due to a smaller radius [32]StabilityClinically used MBs often exhibit rapid signal decay; Modification of the shell composition with a fluorinated inner layer enhances bubble stability.Under identical shell conditions, NBs exhibit poorer stability; with optimized lipid formulations, NBs can exhibit slower decay and more persistent ultrasound signals. [36, 37]Gas volumeLarger per-bubble gas volumeSmaller per-bubble gas volume [50]Acoustic scattering efficiencyA scattering cross-section up to four orders of magnitude greater than that of NBsSignificantly weaker scattering [42]CEUSExcellent for CEUS and perfusion assessmentLimited CEUS ability; better suited for targeted imaging [42]Circulation timeShorter circulation time; rapid pulmonary filtration (e.g., SonoVue)Longer circulation time; slower clearance by the reticuloendothelial system [5, 51, 52]Cavitation thresholdLower acoustic pressure threshold and conventional clinical ultrasound sufficeHigher acoustic pressure is required to induce cavitation [42]Vascular permeabilityRemain intravascular due to large diameter; Lesser leakage in capillariesCapable of extravasation through leaky tumor vasculature; Greater leakage in capillaries [44, 53]Decorrelation time0.09 ± 0.03 s in tumor0.82 ± 0.59 s in skin6.04 ± 3.57 s in tumor1.06 ± 0.89 s in healthy tissue [44] Summary of differences between MBs and NBs 0.09 ± 0.03 s in tumor 0.82 ± 0.59 s in skin 6.04 ± 3.57 s in tumor 1.06 ± 0.89 s in healthy tissue Clinical value of ultrasound contrast agents in tumor diagnosis Conventional ultrasound is widely used for tumor diagnosis and treatment monitoring because it is cost-effective, noninvasive, nonionizing radiation, and provides real-time visualization. However, B-mode ultrasound provides only two-dimensional anatomical images, and color Doppler is limited to large-vessel (> 100–200 μm) perfusion assessment, failing to resolve microvascular flow. CEUS enables the imaging of blood flow and its local changes by introducing ultrasound contrast agents into the bloodstream, which act as powerful ultrasound scatterers. CEUS enables real-time visualization of capillary-level perfusion, dramatically improving diagnostic sensitivity and specificity for malignant tumors [54]. MBs represent the most commonly used clinical ultrasound contrast agents; commercially available formulations such as SonoVue and Sonazoid are routinely employed across a range of indications [46]. SonoVue is comprised of a sulfur hexafluoride gas core stabilized by a phospholipid shell, whereas Sonazoid consists of a perflubutane/perfluorobutane gas core coated and encapsulated within a hydrogenated egg phosphatidylserine sodium shell. As a purely intravascular tracer, SonoVue remains confined to the vascular space, permitting dynamic microcirculation assessment. A retrospective analysis of 463,434 CEUS examinations reported an overall adverse-reaction rate of 0.034% for SonoVue, with serious events being exceedingly rare, underscoring its excellent safety and tolerability [29]. CEUS characteristically reveals malignant tumors as heterogeneous, centripetal enhancement with ill-defined margins, perfusion defects, and distorted vasculature. Arteriovenous shunts within tumors produce a “rapid in, rapid out” enhancement pattern, whereas increased peritumoral microvessel density leads to characteristic peripheral enhancement. Quantitative CEUS parameters shortened time to peak, elevated peak intensity, and steeper wash-in slope, provide objective metrics for distinguishing benign from malignant lesions. Beyond lesion characterization, CEUS offers unique advantages in predicting molecular tumor subtypes, evaluating lymphatic channels and sentinel lymph nodes, and forecasting responses to neoadjuvant chemotherapy, making it an indispensable component of precision oncology. Targeted MBs and NBs for tumor diagnosis Conventional ultrasound contrast agents lack specificity, remaining confined to the vasculature and providing a short-term enhancement. This limitation hampers the comprehensive evaluation of microvascular perfusion and precludes molecular-level disease recognition, thereby restricting the utility of CEUS in diagnostic and differential applications. With the rapid development of ultrasound imaging, molecular biology and nano-medicine, a new field of ultrasound molecular imaging came into being. This technique can perform ultrasonic imaging at the living molecular level, and qualitatively and quantitatively analyze the changes of structure and molecular expression level in tissues. Ultrasound molecular imaging uses a targeted ultrasound contrast agent as a molecular probe, and the specific ligand carried on its surface specifically binds to the molecules in the target tissue, so that the contrast agent can be enriched in the target area for a long time [43, 48, 55]. Specific binding of these ligands to target molecules in diseased tissue enables prolonged retention of contrast agents at the site of interest, markedly amplifying ultrasonic signal and allowing accurate detection of molecular or cellular changes, and further enhances the accuracy and application value of ultrasound technology in early diagnosis of diseases [31, 45, 56]. Ligand-decorated MBs and NBs, engineered to bind molecular markers overexpressed in tumors or vascular diseases, have emerged as promising tools for assessing angiogenic markers and detecting early-stage cancers. Targeted bubbles are typically synthesized via biotin-avidin coupling, amide conjugation, and maleimide-thiol chemistry, etc. Targeted MBs Because MBs are restricted to the intravascular space, all molecular targets must be located inside the blood vessel. Current research in tumor ultrasound molecular imaging focuses on endothelial and angiogenesis-related markers (e.g., αvβ3 integrin, anti-vascular endothelial growth factor receptor 2 [VEGFR2], Neuropilin-1, Endoglin), tumor endothelial antigens (e.g., secreted frizzled related protein 2, B7-H3, Nucleolin, Thy1), and adhesion molecules (e.g., vascular cell adhesion molecule-1, P-selectin, E-selectin) [57]. To image tumor angiogenesis, soft-shell MBs conjugated with iRGD peptides (L-MBα) have been developed to target αvβ3 integrin, while hard-shell PLGA-lipid hybrid MBs bearing VEGFR2 antibodies (LP-MBv) selectively bind VEGFR2. Following intravenous co-injection of L-MBα and LP-MBv, low-intensity ultrasound (0.65 MPa, 5.6 MHz) was used to disrupt L-MBα, and high-intensity ultrasound (1.52 MPa, 5.6 MHz) was used to disrupt LP-MBv. Differential enhancement before and after each destruction event allowed quantitative comparison of targeting efficiency and simultaneous detection of both biomarkers, thereby improving diagnostic precision (Fig. 3a) [48]. In hepatocellular carcinoma models, MBs targeting C-X-C chemokine receptor type 4 (CXCR4) via LFC131 peptide achieved high ligand coupling efficiency (99.77 ± 0.15%) and exhibited superior enhancement and peak intensity compared to non-targeted MBs. Moreover, LFC131 binding blocked CXCL12-CXCR4 interactions, inhibiting tumor cell survival, proliferation, migration, and invasion (Fig. 3b) [55]. Beyond diagnosis, targeted MBs can predict the tumor immune microenvironment. Tumor neovascularization is morphologically abnormal and functionally anergic, limiting immune cells’ trafficking across the endothelial barrier and suppressing antitumor immune responses. CD93-targeted MBs, fabricated using Multimerin-2 containing hybrid cell membrane/lipid shells, demonstrated altered enhancement kinetics in R848-treated (immune adjuvant) versus PBS-treated controls, earlier time-to-peak, higher peak intensity, and reduced post-destruction signal reflecting immune activation within the tumor (Fig. 3c) [58]. Fig. 3Vascular-targeted MB. (a) Schematic diagram of the serial collapse strategy for ultrasound molecular imaging of αvβ3-targeting soft-shell lipid MBs (L-MBα) and VEGFR2-targeting hard-shell lipid-PLGA MBs (LP-MBv) [48]. Copyright 2022, Wiley-VCH. (b) The design of CXCR4-targeted MB and CEUS examinations in orthotopic liver tumor-bearing nude mice [55]. Copyright 2025, Elsevier. (c) The proposed CD93-targeted MBs could target CD93 with high efficiency. Ultrasound imaging of CD93 using the MBs provides a noninvasive and accurate strategy to predict the tumor immune microenvironment [58]. Copyright 2024, Wiley-VCH Vascular-targeted MB. (a) Schematic diagram of the serial collapse strategy for ultrasound molecular imaging of αvβ3-targeting soft-shell lipid MBs (L-MBα) and VEGFR2-targeting hard-shell lipid-PLGA MBs (LP-MBv) [48]. Copyright 2022, Wiley-VCH. (b) The design of CXCR4-targeted MB and CEUS examinations in orthotopic liver tumor-bearing nude mice [55]. Copyright 2025, Elsevier. (c) The proposed CD93-targeted MBs could target CD93 with high efficiency. Ultrasound imaging of CD93 using the MBs provides a noninvasive and accurate strategy to predict the tumor immune microenvironment [58]. Copyright 2024, Wiley-VCH Targeted NBs Due to their submicron size, NBs can extravasate into tumor interstitium via the enhanced permeability and retention effect [59]. Active targeting is achieved by conjugating antibodies, peptides, aptamers, or small-molecule ligands against tumor-associated antigens or neoantigens, enabling molecular ultrasound imaging and, when combined with ultrasound irradiation, simultaneous therapy (Fig. 4a-b) [56, 60–63]. Recently, biomimetic NBs generated by fusing tumor cell membranes with lipid shells (NBCCM) have attracted attention. These constructs display native adhesion molecules, tumor antigens, and homing receptors, conferring homologous targeting to tumors. Moreover, cell-membrane CD47 mediates immune evasion, prolonging circulation time. NBCCM exhibited the highest CEUS signal in homologous tumor models, underscoring their potential for precise molecular imaging (Fig. 4c-e) [64]. Fig. 4Tumor tissue targeted NB. (a) Illustration of the targeted NBs (ACP/ICG-NBs) that bind specifically to RCC cells expressing CAIX antigen and enhance the ultrasound, photoacoustic and fluorescence imaging of RCC xenograft [63]. Copyright 2023, Dove Medical Press. (b) Schematic diagram showing the experimental approach for targeted NB therapy with PSMA-NB injection followed by non-focal therapeutic ultrasound irradiation [62]. Copyright 2024, Elsevier. (c) Cancer cell membrane-based NB for ultrasound molecular imaging in triple-negative breast cancer. d-e) In vivo effect of NBCCM and NBctrl on tumor enhancement by contrast-mode imaging. Representative ultrasonographic images and mean time intensity curves of both SK-BR3 and MDA-MB-231 tumors after NB injection [64]. Copyright 2022, Springer Nature Tumor tissue targeted NB. (a) Illustration of the targeted NBs (ACP/ICG-NBs) that bind specifically to RCC cells expressing CAIX antigen and enhance the ultrasound, photoacoustic and fluorescence imaging of RCC xenograft [63]. Copyright 2023, Dove Medical Press. (b) Schematic diagram showing the experimental approach for targeted NB therapy with PSMA-NB injection followed by non-focal therapeutic ultrasound irradiation [62]. Copyright 2024, Elsevier. (c) Cancer cell membrane-based NB for ultrasound molecular imaging in triple-negative breast cancer. d-e) In vivo effect of NBCCM and NBctrl on tumor enhancement by contrast-mode imaging. Representative ultrasonographic images and mean time intensity curves of both SK-BR3 and MDA-MB-231 tumors after NB injection [64]. Copyright 2022, Springer Nature The physical and biological mechanisms of UTMC/UTNC Physical mechanism: cavitation effect In an acoustic field, gas-filled bubbles suspended in liquid serve as cavitation nuclei. When exposed to ultrasound, these nuclei undergo oscillation, growth, contraction, and collapse, collectively referred to as the cavitation effect. The high acoustic impedance (Acoustic impedance [Z], which quantifies a medium’s resistance to sound propagation, is the product of its density [ρ] and the speed of sound [c]) mismatch between the encapsulated gas and surrounding liquid medium, along with the viscoelastic nature of the bubble shell, endows bubbles with high compressibility (Compressibility measures a substance’s tendency to reduce in volume under pressure and is the inverse of its bulk modulus [K]) and highly susceptible to ultrasound-driven dynamic volume oscillations [59]. When the ultrasound frequency matches their intrinsic resonance frequency, the oscillatory response is markedly amplified. These oscillations can initiate dynamic cycles of expansion, contraction, and, under sufficient acoustic pressure, eventual collapse. During collapse, mechanical energy is released as radiation pressure, microjets, bubble tunneling, and shock waves [65, 66]. Both bubble properties and ultrasonic parameters critically influence cavitation. Higher acoustic pressure and MI promote cavitation, whereas increasing ultrasound frequency suppresses it [59]. The MI, a dimensionless parameter, reflects the likelihood of mechanical bioeffects caused by acoustically driven bubbles or gas bodies [67, 68]. It is defined as:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathrm{M}\mathrm{I}=\frac{P}{\sqrt{f}}$$\end{document} where P is the peak rarefaction pressure (in MPa) and f is the center frequency (in MHz) of the ultrasound wave [67, 68]. The U.S. Food and Drug Administration limits MI values to < 1.9 for most human applications (and < 0.23 for ophthalmic imaging) to prevent tissue damage [68]. During cavitation, rapid gas compression transiently increases the bubble’s internal temperature and pressure (sustained time < 1 µs), which can result in light emission (sonoluminescence) and production of high reactive chemical species (e.g., reactive oxygen species [ROS]) [66]. Although frictional heat transfer between the bubble and liquid dissipates energy, the thermal time constant greatly exceeds the oscillation period; consequently, thermal effects are negligible in biomedical contexts [66]. Two cavitation regimes are distinguished: stable cavitation, which occurs at low MI (MI ≤ 0.3, bubbles oscillate linearly when MI < 0.1, whereas oscillations become nonlinear, producing harmonic backscattering, within the MI range of 0.1–0.3) and is characterized by periodic oscillation with induced microstreaming in the surrounding fluid; and inertial cavitation, which arises as acoustic intensity and pressure increase with high MI (MI > 0.5, violent vibrations occur, potentially resulting in bubble destruction), leading to violent collapse, shock-wave formation, microjet emission, and localized thermal effects (Fig. 5a) [69–71]. High-speed imaging and acoustic detection techniques have been employed to visualize these phenomena. Ultra-high-speed cameras capture MB morphology at successive time points, revealing rapid bubble growth followed by collapse (Fig. 5b) [72]. As acoustic intensity rises, the maximum area of oscillating vapor-bubble clouds correspondingly increases (Fig. 5c) [73]. Piezoelectric pressure acoustic transducers have been used to record bubble oscillation amplitude and collapse timing, demonstrating a transition from uniform oscillation to explosive collapse (Fig. 5d) [74]. Acoustic monitoring techniques, particularly passive cavitation detection, are employed to monitor multi-bubble dynamics and cavitation phenomena in vivo [75, 76]. Passive cavitation detection operates by analyzing the transient nonlinear acoustic emissions generated from ultrasound-activated bubbles. This method distinguishes between stable and inertial cavitation in real time by quantifying the distinct Fourier spectral features of the emission signals: the strength of harmonics, subharmonics, and ultraharmonics for stable cavitation, and the level of the broadband response for inertial cavitation [77]. When bubbles adhere to cell surfaces, unsymmetry is disrupted under ultrasound irradiation: the adherent site in contact with the cell (W1) expands less than the free site (W2), though overall expansion ratios remain comparable between bulk and targeted bubbles (Fig. 5e) [2]. Indirectly responsive bubbles, such as GVs, collapse under low-frequency ultrasound to release bubble nuclei, which then undergo inertial cavitation, producing strong mechanical forces (Fig. 5f). High-speed imaging has captured GVs immediately before collapse (1.4 µs), at collapse (1.8 µs), and after bubble formation (4.2 µs), confirming their role in enhancing cavitation nucleation (Fig. 5g) [78]. Phase-change nanodroplets and hollow solid nanoparticles can also act as exogenous cavitation nuclei. Nanodroplets vaporize under ultrasound to form bubbles, while hydrophobic cavities in solid particles trap gas to provide nucleation sites. Both require higher acoustic pressures to activate and cannot be imaged before activation. Fig. 5The physical mechanism of MB/NB under ultrasound stimulation and the monitoring of cavitation effect. (a) MB dynamics under ultrasound stimulation, including linear oscillation, nonlinear oscillation and collapse [69]. Copyright 2021, Royal Society of Chemistry. (b) Optical images showing the collapse process of a MB (radius 10 μm) [72]. Copyright 2012, Elsevier. (c) Bubble clouds (peak intensity in a cycle captured by the high-speed camera) were increased with ultrasound intensity. Scale bar, 100 μm [73]. Copyright 2022, AAAS. (d) Acoustic spectrum of the bursting of a floating bubble at an air/water interface with a contact size of 0.30 ± 0.05 cm depicted by piezoelectric pressure acoustic transducers in mV vs. time and shown in high-speed photos [74]. Copyright 2019, Elsevier. (e) Ultrahigh-speed imaging of oscillating targeted microbubbles (TMBs) adherent to cells. R0 indicates the MB resting radius (here 0.75 μm), and W1 and W2 are the maximal expansion of the MB’s adherent and free sides, respectively [2]. Copyright 2020, National Academy of Sciences. (f) Proposed mechanism of GV-seeded cavitation. (g) High frame rate camera images immediately before GV collapse (1.4 µs), immediately after GV collapse (1.8 µs) and after the formation of bubbles (4.2 µs) [78]. Copyright 2021, Springer Nature The physical mechanism of MB/NB under ultrasound stimulation and the monitoring of cavitation effect. (a) MB dynamics under ultrasound stimulation, including linear oscillation, nonlinear oscillation and collapse [69]. Copyright 2021, Royal Society of Chemistry. (b) Optical images showing the collapse process of a MB (radius 10 μm) [72]. Copyright 2012, Elsevier. (c) Bubble clouds (peak intensity in a cycle captured by the high-speed camera) were increased with ultrasound intensity. Scale bar, 100 μm [73]. Copyright 2022, AAAS. (d) Acoustic spectrum of the bursting of a floating bubble at an air/water interface with a contact size of 0.30 ± 0.05 cm depicted by piezoelectric pressure acoustic transducers in mV vs. time and shown in high-speed photos [74]. Copyright 2019, Elsevier. (e) Ultrahigh-speed imaging of oscillating targeted microbubbles (TMBs) adherent to cells. R0 indicates the MB resting radius (here 0.75 μm), and W1 and W2 are the maximal expansion of the MB’s adherent and free sides, respectively [2]. Copyright 2020, National Academy of Sciences. (f) Proposed mechanism of GV-seeded cavitation. (g) High frame rate camera images immediately before GV collapse (1.4 µs), immediately after GV collapse (1.8 µs) and after the formation of bubbles (4.2 µs) [78]. Copyright 2021, Springer Nature Sonoporation and mechanical damage in cells Under ultrasound stimulation, MB/NB oscillation, microjetting, and collapse induce transient rearrangements of the lipid bilayer, forming nanoscale pores and tunnels in cell membranes. This sonoporation effect facilitates intracellular delivery of drugs, genes, and nanoparticles. Cells are subjected to various stresses from cavitation events, including jet impact pressure, oscillation impact pressure, oscillation shear stress, and acoustic radiation pressure, each oscillating periodically with the bubble dynamics (Fig. 6a-b). Microjets and these mechanical stresses can inflict localized damage on adjacent cells. Scanning electron microscopy reveals that the control group and NB group only treatments leave cell membranes smooth and intact, the NB + US treatment group yields abundant nanoscale poration sites (Fig. 6c) [79]. Confocal microscopy has captured the full cycle of sonopore formation and resealing in real time (Fig. 6d). Upon bubble collapse, membrane perforations appear; pores up to 30 μm² readily reseal within 1 min (Fig. 6e), trapping delivered cargo intracellularly. However, excessively large pores may fail to close pore diameter can increase from ~ 12.5 μm at 10 s post-collapse to ~ 32.6 μm at 3.5 min before gradual reduction (Fig. 6f). The efficacy of sonoporation is also influenced by the bubble-cell distance (D) relative to the bubble diameter (d), and reducing the interfacial distance between bubbles and cells can significantly enhance sonoporation efficiency [80]. When D/d ≤ 0.75, cavitation reliably induces membrane permeabilization [72]. Under comparable interfacial conditions, larger bubbles tend to generate stronger cellular permeability due to increased mechanical force upon collapse [81]. Cell cycle stage further modulates responsiveness: S-phase cells, being flatter and more compliant (lower elastic modulus), exhibit maximal transient permeability and rapid cytoskeletal disassembly, whereas G1-phase cells, with greater height and stiffness, display reduced sonoporation [82]. Excessive mechanical forces during UTMC/UTNC can cause irreversible cellular injury and tissue disruption. At the cellular level, treated cells exhibit loss of membrane integrity, mitochondrial swelling, nuclear fragmentation, and cytoplasmic leakage (Fig. 6g) [83]. Cell viability inversely correlates with ultrasound intensity and bubble concentration: at 108 NBs/mL, 1 MHz, 1 W/cm2 conditions, survival falls to ~ 50%; at 109 NBs/mL, viability drops below 10%, indicating severe damage [5, 76]. In vivo, UTMC/UTNC generates zones of apoptosis and necrosis within tumors [62]. When MBs are loaded with annexin-V-indocyanine green nanoparticles (AV-ICG-NPs@MBs), UTMC-triggered release of AV-ICG-NPs binds to exposed phosphatidylserine on apoptotic membranes, amplifying fluorescence. Concurrent delivery of the antioxidant gastrodin mitigates cavitation-induced ROS, reversing UTMC-mediated apoptosis [84]. Fig. 6The effect of UTMC/UTNC on the cell. (a) Different stress mechanisms elicited by an ultrasound-driven MB on a cell. (b) Stress evolution over time in the logarithmic scale of each stress mechanism, inferred from the radial dynamics of the MB. The solid lines represent stresses directed towards the substrate or outwards from the bubble, whereas the dashed lines indicate the opposite [3]. Copyright 2025, Springer Nature. (c) Effects of UTNC on the cell surface visualized under a scanning electron microscope. White arrows indicate pores on the cell surface [79]. Copyright 2016, Taylor & Francis. (d) Time-series confocal fluorescence images revealing localized membrane perforation and recovery induced by ultrasound-triggered collapse of a single MB. (e) Time sweep plots of membrane fluorescence for the specified scan line. (f) Membrane perforation induced by ultrasound-triggered MB collapse can be unrecoverable for large pore sizes [85]. Copyright 2013, Elsevier. (g) Transmission electron microscopy images of RM1 cells obtained immediately after treatment with US, NB or USNB. The blue arrow indicates the fragmented Golgi apparatus; the yellow arrow indicates swollen mitochondria; and the red arrow indicates fragmented nuclei. The scale bars in the global and magnified images represent 5 and 1 μm, respectively [83]. Copyright 2022, BMJ Publishing Group Ltd The effect of UTMC/UTNC on the cell. (a) Different stress mechanisms elicited by an ultrasound-driven MB on a cell. (b) Stress evolution over time in the logarithmic scale of each stress mechanism, inferred from the radial dynamics of the MB. The solid lines represent stresses directed towards the substrate or outwards from the bubble, whereas the dashed lines indicate the opposite [3]. Copyright 2025, Springer Nature. (c) Effects of UTNC on the cell surface visualized under a scanning electron microscope. White arrows indicate pores on the cell surface [79]. Copyright 2016, Taylor & Francis. (d) Time-series confocal fluorescence images revealing localized membrane perforation and recovery induced by ultrasound-triggered collapse of a single MB. (e) Time sweep plots of membrane fluorescence for the specified scan line. (f) Membrane perforation induced by ultrasound-triggered MB collapse can be unrecoverable for large pore sizes [85]. Copyright 2013, Elsevier. (g) Transmission electron microscopy images of RM1 cells obtained immediately after treatment with US, NB or USNB. The blue arrow indicates the fragmented Golgi apparatus; the yellow arrow indicates swollen mitochondria; and the red arrow indicates fragmented nuclei. The scale bars in the global and magnified images represent 5 and 1 μm, respectively [83]. Copyright 2022, BMJ Publishing Group Ltd Enhanced drug delivery in vivo via UTMC/UTNC Efficient delivery of therapeutics, particularly larger nanomaterials, requires traversal of endothelial barriers and dense tumor extracellular matrices, with the plasma membrane presenting an additional obstacle to intracellular uptake. Ultrasound, with its spatiotemporal precision, noninvasiveness, and deep-tissue penetration, has emerged as a potent external stimulus to overcome these barriers [86]. UTMC/UTNC transiently increases vascular and membrane permeability, facilitating localized accumulation of drugs and macromolecules; systemic dosing can thus be reduced while maintaining therapeutic efficacy and minimizing toxicity. Under ultrasound exposure, MBs generate both observable and submicroscopic pores and tunnels in cellular barriers. When 3 μm MBs contact endothelial cells under buoyancy in the presence of propidium iodide (PI), low-pressure ultrasound (60 kPa, 1 MHz) induces stable cavitation alternating expansion and compression without detectable PI uptake (Fig. 7a). At higher pressure (100 kPa, 1 MHz), the same bubbles undergo inertial cavitation, producing recirculating microjets and larger radial displacements; each collapse event perforates the membrane, permitting PI entry and strong intracellular fluorescence (Fig. 7b) [3]. Microjetting is initiated once bubble expansion exceeds ~ 1 μm, and inertial cavitation thresholds are reached when radial expansion doubles the equilibrium radius (Fig. 7c-d) [3]. Receptor-mediated binding and buoyancy enhance bubble internalization under ultrasound. αvβ3-under MB bound beneath endothelial cells exhibit greater uptake depth and PI delivery than non-targeted or IgG-coated controls, and outperform αvβ3-above MB. Bubble internalization correlates positively with cell thickness and bubble size, and targeted MBs yield enhanced transfection efficiency at reduced cell viability, attributable to asymmetric oscillation and collapse forces at the adhesion site (Fig. 7e-f) [2]. Beyond sonoporation, cavitation can disrupt microvascular integrity. Ultrasound-mediated MB collapse has been shown to create transcellular pores through both apical and basal membranes, as well as to transiently open intercellular junctions, promoting paracellular drug transport into the extravascular space [87, 88]. In vivo multiphoton microscopy tracking of fluorescent NBs revealed that, following intravenous injection and 2 min of ultrasound, tumor vessels exhibited increased permeability and structural disruption. NBs extravasated passively yet intact into tumor interstitium, with significantly greater accumulation than in healthy tissues (Fig. 7g) [89]. Furthermore, UTMC/UTNC has been applied to transiently open the blood-brain barrier (BBB). Deep learning segmentation of tumor vessels indicated that in capillaries 2–6 μm in diameter, NBs (200 nm) achieved higher extravasation than MBs (1.5 μm), whereas both performed similarly in larger vessels [53]. However, excessive inertial cavitation may cause adverse biological effects, including erythrocyte extravasation, sterile neuroinflammation, and tissue injury [75, 90]. Therefore, it is essential to balance acoustic pressure to achieve therapeutic efficacy while minimizing collateral damage to surrounding tissues. Fig. 7UTMC/UTNC UTMC/UTNC promotes drug delivery. a) The applied ultrasound pressure induces spherical oscillations in the bubble, followed by asymmetric deformation (pa = 60 kPa, f = 1 MHz). However, this does not result in cell membrane poration and drug uptake. b) A higher ultrasound pressure causes the bubble to develop cyclic piercing microjets directed towards the cell (pa = 160 kPa, f = 1 MHz), resulting in cell membrane poration and drug uptake. c) Experimental and theoretical radial motion of the MB. d) Experimental and theoretical vertical positions of the MB centroid [3]. Copyright 2025, Springer Nature. e) Orthogonal views of typical examples of confocal microscopy z-stacks showing the 3D morphology of the endothelial cell membrane and different MBs. f) Internalized depth for the distinct MB types and experimental set-up orientations; statistically significant differences indicated by horizontal lines (all at p < 0.001) [87]. Copyright 2022, Elsevier. g) Rapid (within seconds) focal vascular disruption during ultrasound exposure (which begins at t=10 s). Slower (within minutes), widespread vascular permeabilization occurring following ultrasound exposure [89]. Copyright 2020, IVYSPRING. UTMC/UTNC UTMC/UTNC promotes drug delivery. a) The applied ultrasound pressure induces spherical oscillations in the bubble, followed by asymmetric deformation (pa = 60 kPa, f = 1 MHz). However, this does not result in cell membrane poration and drug uptake. b) A higher ultrasound pressure causes the bubble to develop cyclic piercing microjets directed towards the cell (pa = 160 kPa, f = 1 MHz), resulting in cell membrane poration and drug uptake. c) Experimental and theoretical radial motion of the MB. d) Experimental and theoretical vertical positions of the MB centroid [3]. Copyright 2025, Springer Nature. e) Orthogonal views of typical examples of confocal microscopy z-stacks showing the 3D morphology of the endothelial cell membrane and different MBs. f) Internalized depth for the distinct MB types and experimental set-up orientations; statistically significant differences indicated by horizontal lines (all at p < 0.001) [87]. Copyright 2022, Elsevier. g) Rapid (within seconds) focal vascular disruption during ultrasound exposure (which begins at t=10 s). Slower (within minutes), widespread vascular permeabilization occurring following ultrasound exposure [89]. Copyright 2020, IVYSPRING. UTMC/UTNC-Mediated enhancement of tumor therapy UTMC/UTNC-Enhanced chemotherapy Systemic chemotherapy remains a cornerstone of clinical oncology but is limited by nonspecific distribution, off-target cytotoxicity (e.g., myelosuppression, gastrointestinal toxicity, alopecia), uneven drug accumulation, and acquired resistance. To address these shortcomings, precision delivery platforms and stimulus-responsive carriers have been developed to maximize antitumor efficacy while minimizing harm to healthy tissue (Fig. 8). Fig. 8Multimodal strategies for augmenting chemotherapy via ultrasound-responsive MBs/NBs. (Created with bioRender.com) Multimodal strategies for augmenting chemotherapy via ultrasound-responsive MBs/NBs. (Created with bioRender.com) Enhancement of free chemotherapy drug delivery UTMC/UTNC enables high local drug concentrations within tumors while avoiding nonspecific damage to the surrounding normal tissues. In one study, 0.1 mg of free doxorubicin (DOX) was co-administered intravenously with 5 × 107 NBs, followed by ultrasound irradiation at the tumor site in a PC-3 xenograft model. High performance liquid chromatography analysis revealed that DOX accumulation in tumor tissue was significantly higher in the DOX + NB + US group compared to DOX alone, whereas DOX concentrations in the heart and liver were substantially reduced, thereby enhancing tumor suppression while minimizing cardiotoxicity [79]. Furthermore, the timing of UTMC/UTNC relative to drug administration affected intratumoral drug levels. Maximum gemcitabine accumulation in tumors was achieved when UTMC was applied immediately post-infusion rather than 10 min later or separated entirely, underscoring the importance of synchronized ultrasound application for optimal drug deposition [91]. Using Evans blue (MW ≈ 960 Da) as a small-molecule surrogate, UTMC markedly increased dye penetration depth and fluorescence intensity in tumor cores versus non-irradiated controls, visually confirming enhanced extravasation and accumulation (Fig. 9a-d) [92]. The transient opening of the BBB by UTMC offers a viable approach to enhance drug delivery for treating intracranial tumors, Alzheimer’s disease, and Parkinson’s disease, which is restricted by the BBB [93–95]. Mechanical forces from ultrasound-mediated cavitation increased DOX concentrations in the human and mouse brains by 2-fold and 3.9-fold, respectively. Similarly, anti-PD-1 antibody levels in the brain and peritumoral areas were elevated, accompanied by upregulation of interferon-gamma (IFN-γ) and major histocompatibility complexes class I in microglia and macrophages, thereby enhancing glioma immunotherapy (Fig. 9e–g) [96]. Systemic chemotherapy remains the mainstay for patients with advanced cancer. Clinical studies have demonstrated that UTMC enhances the efficacy of chemotherapeutics. The use of ultrasound contrast agents (e.g., SonoVue, Definity) combined with ultrasound irradiation to deliver drugs such as gemcitabine, paclitaxel, and carboplatin has shown improved survival and safety across various tumor types [97]. In patients with stage III/IV pancreatic cancer, UTMC combined with chemotherapy yielded longer overall survival, progression-free survival, and disease control rates compared to chemotherapy alone, without increasing adverse events [98, 99]. UTMC-mediated transient disruption of the BBB enables the safe and repeated penetration of cytotoxic agents into the brain parenchyma, increasing the local concentration of chemotherapeutics at glioma sites and effectively suppressing tumor progression [100–102]. These results suggest that UTMC, when applied using clinically available ultrasound equipment and FDA-approved MBs and chemotherapeutics, holds substantial potential for clinical translation. Fig. 9UTMC/UTNC promotes chemical drug delivery. (a) The detailed schematic illustration of the sequence of intravenous drug administration and MB-assisted ultrasonication technology treatment. (b) The digital photos of tumors after different treatments as depicted. The blue color in tumors indicates Evans blue accumulation. c-d) Microscopic observation of tumor slices collected from CT-26 tumor-bearing mice at 2 h post various treatments at different sequences (blue: nucleus, red: Evans blue) [92]. Copyright 2023, Elsevier. e) Clinical course of recurrent GBM patients analyzed. f) Representative MRI scans showing the biopsy sites in the sonicated (top) and non-sonicated (bottom left) peri-tumoral brain determined by whether these sites were immediately deep to an ultrasound emitter, as well as acoustic simulation of the ultrasound beams for this case (bottom right). g) Bar plot representing the fold change in DOX concentration in non-sonicated (n = 4 brain samples) and sonicated peritumoral brain tissues (n = 8 brain samples) from 2 GBM patients [96]. Copyright 2024, Springer Nature UTMC/UTNC promotes chemical drug delivery. (a) The detailed schematic illustration of the sequence of intravenous drug administration and MB-assisted ultrasonication technology treatment. (b) The digital photos of tumors after different treatments as depicted. The blue color in tumors indicates Evans blue accumulation. c-d) Microscopic observation of tumor slices collected from CT-26 tumor-bearing mice at 2 h post various treatments at different sequences (blue: nucleus, red: Evans blue) [92]. Copyright 2023, Elsevier. e) Clinical course of recurrent GBM patients analyzed. f) Representative MRI scans showing the biopsy sites in the sonicated (top) and non-sonicated (bottom left) peri-tumoral brain determined by whether these sites were immediately deep to an ultrasound emitter, as well as acoustic simulation of the ultrasound beams for this case (bottom right). g) Bar plot representing the fold change in DOX concentration in non-sonicated (n = 4 brain samples) and sonicated peritumoral brain tissues (n = 8 brain samples) from 2 GBM patients [96]. Copyright 2024, Springer Nature Therapy with Drug-Loaded MBs/NBs Encapsulation of chemotherapeutics within lipid shells enhances pharmacokinetics until payload release [103]. Common chemotherapeutic agents such as DOX, docetaxel, paclitaxel, and camptothecin can be incorporated into MBs/NBs via thin-film hydration or mechanical agitation methods. Depending on drug hydrophobicity or hydrophilicity, they are distributed within the amphiphilic lipid layer (Fig. 10a). Under UTMC/UTNC, drug-loaded MBs/NBs undergo cavitation; the resulting mechanical and thermal effects trigger controlled, site-specific drug release, and this spatiotemporal control over payload liberation overcomes systemic exposure limitations [104]. These drug-loaded bubbles and droplets also enhance ultrasound imaging capabilities, facilitating image-guided drug delivery and therapy monitoring, and potentiate therapeutic outcomes through ultrasound irradiation-mediated augmentation of local drug delivery to tumor sites (Fig. 10b) [105–107]. Comparative studies in liver tumors using DOX-loaded NBs (hDOX-NB, ≈ 280 nm) and MBs (hDOX-MB, ≈ 1100 nm) under unfocused ultrasound revealed similar hepatic drug accumulation, but the hDOX-NB + US group exhibited larger zones of tumor apoptosis and necrosis with reduced off-target toxicity in healthy liver parenchyma [108]. In 2–6 μm capillaries, NBs extravasated more effectively than MBs under ultrasound, suggesting superior delivery in microvasculature, with parity observed in larger vessels [53]. To further enhance tumor localization and controlled release, integrated strategies such as active targeting ligands, tumor microenvironment responsive release mechanisms, prodrug activation cascades, and multifunctional nanomaterials have been developed. These designs improve intratumoral drug enrichment and reduce systemic toxicity. Tumor-targeting ligands or biomimetic membrane coatings promote tumor accumulation while evading phagocytic clearance, enabling ultrasound-triggered molecular imaging and therapy. Upon ultrasound exposure, cavitation enhances transmembrane drug transport and tumor cell disruption, offering an effective theranostic approach [45, 64, 109]. For example, macrophage membrane coated MBs coloaded with DOX and perfluoropentane (TFHPD) were used in a two-stage ultrasound protocol: initial UTMC to permeabilize tumor vasculature and enhance TFHPD accumulation, followed by a second ultrasound to induce perfluoropentane phase transition and vigorous cavitation, collapsing the extracellular matrix and releasing drug deep within the tumor core (Fig. 10c-d) [110]. Environmental- and thermo-responsive nanocarriers have also been formulated to exploit tumor hypoxia, acidic pH, and elevated glutathione (GSH) for selective drug release, further potentiating chemotherapeutic efficacy while mitigating off-target effects [111–114]. UTMC-mediated oxygen delivery has been shown to alleviate hypoxia, elevate ROS, induce apoptosis, deplete GSH, downregulate HIF-1α/P-gp, and inhibit epithelial-mesenchymal transition, thus overcoming multidrug resistance and suppressing metastasis [115, 116]. Finally, ultrasound-activated prodrug systems, such as Ce6-sensitized platforms that generate ROS to trigger in situ chemotherapeutic activation, have achieved cascade-release kinetics, amplifying DNA damage and cytotoxicity with minimal systemic toxicity [117, 118]. In hypoxic conditions exacerbated by ROS consumption during SDT, an azo-linked camptothecin prodrug releases its active form selectively in tumors, outperforming standalone SDT [119, 120]. Furthermore, co-loading NBs with ICG enables ultrasound/photoacoustic/fluorescence tri-modal imaging for diagnosis and real-time monitoring of therapeutic response, offering a multifunctional platform for integrated cancer management [121]. Fig. 10UTMC/UTNC facilitates the drug release and delivery of drug-loaded bubbles. (a) Structural diagrams of camptothecin-loaded MBs (CPT-MBs) and doxorubicin-loaded MBs (DOX-MBs) [104]. Copyright 2025, Elsevier. (b) Fluorescence micrographs showing LIVE/DEAD staining of the 3D cultured MDA-MB-231 cells in the spheroid tumour models treated with free curcumin and curcumin-loaded MBs with and without ultrasound. Scale bar = 1 mm [105]. Copyright 2025, Elsevier. (c) Schematic representation of sequential US-activated MB-NP hybrid (MMB@TFHPD) for robust tumor suppression in PDAC. (d) Characterization of tumor spheroid penetration using a confocal laser scanning microscope. Images were captured using a Z-stack at 25 μm intervals [110]. Copyright 2025, Wiley-VCH UTMC/UTNC facilitates the drug release and delivery of drug-loaded bubbles. (a) Structural diagrams of camptothecin-loaded MBs (CPT-MBs) and doxorubicin-loaded MBs (DOX-MBs) [104]. Copyright 2025, Elsevier. (b) Fluorescence micrographs showing LIVE/DEAD staining of the 3D cultured MDA-MB-231 cells in the spheroid tumour models treated with free curcumin and curcumin-loaded MBs with and without ultrasound. Scale bar = 1 mm [105]. Copyright 2025, Elsevier. (c) Schematic representation of sequential US-activated MB-NP hybrid (MMB@TFHPD) for robust tumor suppression in PDAC. (d) Characterization of tumor spheroid penetration using a confocal laser scanning microscope. Images were captured using a Z-stack at 25 μm intervals [110]. Copyright 2025, Wiley-VCH UTMC/UTNC-Enhanced sonodynamic therapy Enhancement of sonosensitizer and O2 delivery SDT exploits the deep-tissue penetration of ultrasound to activate sonosensitizers, generating ROS via cavitation, sonoluminescence, and pyrolysis. Sonoluminescence, brief light emission during bubble collapse, is a principal ROS-forming mechanism in SDT [47, 122]. These ROS directly induce DNA damage, lipid peroxidation, and mitochondrial dysfunction, culminating in cell death. However, SDT efficacy is often constrained by the intrinsic properties of sonosensitizers, high intratumoral GSH levels, and tumor hypoxia. Consequently, various nanoscale strategies have been developed to amplify ROS production and ROS-mediated cytotoxicity [122–124]. Here, we focus on how exogenous cavitation nuclei introduced via UTMC/UTNC potentiate SDT. Bubbles and droplets serve as carriers for both sonosensitizers and oxygen. Upon ultrasound exposure, these constructs undergo cavitation, passively targeting tumors and promoting local penetration and cellular uptake of the sonosensitizer. Concurrently, dissolved oxygen in encapsulated perfluorocarbons mitigates hypoxia and enhances ROS generation (Fig. 11a) [125, 126]. Both MBs and NBs can serve as oxygen carriers and enhance the delivery of sonosensitizers under UTMC/UTNC exposure. The smaller size of NBs enables passive extravasation through the leaky tumor vasculature, promoting deeper penetration and accumulation of both oxygen and therapeutic agents within the tumor parenchyma. Active targeting ligands attached to the bubble surface can further increase tumor specificity [127–129], while strategies such as extracellular matrix modulation [130] and GSH consumption [131] amplify intracellular delivery and oxidative stress, thereby improving SDT outcomes. In vivo, a two-stage ultrasound regimen is employed: the first sonication induces droplet vaporization and cavitation to facilitate sonosensitizer extravasation and accumulation; the second activates the sonosensitizer to generate ROS. When combined with immune checkpoint blockade, this approach has significantly suppressed pancreatic tumor growth (Fig. 11b-d) [132]. Fig. 11UTMC/UTNC potentiates SDT efficacy by augmenting the delivery of sonosensitizers and O2. (a) Schematic of the IR780@O2 NBs-based, UTMC-assisted SDT strategy [126]. Copyright 2021, Wiley-VCH. (b) Schematic illustration of the synthesis route of PMPS NDs and the in vivo cavitation-assisted endoplasmic reticulum targeted SDT and aPD-L1 immunotherapy at orthotopic and distant tumor models. c-d) The quantitative data of fluorescence intensity and tumor weight [132]. Copyright 2022, Springer Nature UTMC/UTNC potentiates SDT efficacy by augmenting the delivery of sonosensitizers and O2. (a) Schematic of the IR780@O2 NBs-based, UTMC-assisted SDT strategy [126]. Copyright 2021, Wiley-VCH. (b) Schematic illustration of the synthesis route of PMPS NDs and the in vivo cavitation-assisted endoplasmic reticulum targeted SDT and aPD-L1 immunotherapy at orthotopic and distant tumor models. c-d) The quantitative data of fluorescence intensity and tumor weight [132]. Copyright 2022, Springer Nature Enhancement of radical quantum yields While bubbles mainly function as sonosensitizer vehicles, they are limited by the inherently low quantum yields of radical generation. Hollow or porous solid counterparts can further boost radical quantum yields by enhancing electron transfer and cavitation efficacy. Inertial cavitation intensity is positively correlated with ROS production, and even minimal ROS generated during cavitation can significantly amplify SDT efficacy [19, 133]. For example, mesoporous organosilica nanoparticles featuring hollow, high-surface-area frameworks provide a robust loading platform for gaseous and molecular components. Mesoporous organosilica nanoparticles have been loaded with protoporphyrin IX (PpIX), L-arginine, and CO2 (MON-PpIX-LA-CO₂). The sustained release of CO2 provides continuous cavitation nuclei, enabling repeated inertial cavitation under ultrasound and substantially increasing singlet oxygen (1O2) production versus MON-PpIX alone (Fig. 12a-c) [133]. Similarly, hollow covalent organic framework nanocages (COFNs), composed of tetra-(4-aminophenyl) porphyrin and 2,5-dihydroxyterephthalic acid shells, combine large π-conjugated systems with a porous, hollow architecture (Fig. 12d-e) [135]. These features facilitate efficient charge separation and ROS generation, provide a confined reaction environment for porphyrin redox reactions, and promote gas adsorption, thereby augmenting both cavitation and SDT efficacy [135, 136]. Furthermore, mixing the sonosensitizer hematoporphyrin monomethyl ether with MBs has been shown to enhance ROS generation and induce significant mechanical damage to bacterial membranes during UTMC/UTNC, outperforming SDT alone [137]. Carbon-doped ZnO nanoparticles (C-ZnO) with hollow morphologies exhibit enhanced cavitation, prevent electron-hole recombination, prolong the lifetime of charge carriers (Fig. 12f-g), and thereby enhance the production of toxic ROS, 1O2 generation of C-ZnO was increased by 134.9% and hydroxyl radicals (•OH) were increased by 60.3% compared with ZnO under US irradiation (Fig. 12h-j). The increased surface area and porosity of C-ZnO also intensify mechanical cavitation effects, producing pronounced membrane perforation and high-efficiency cytotoxicity against tumor cells (Fig. 12g) [138]. Collectively, these advances demonstrate that the introduction of exogenous cavitation nuclei, whether via lipid, hybrid, or solid-state nanomaterials, can markedly enhance SDT by increasing ROS yield and combining mechanical disruption with chemical cytotoxicity. Fig. 12Enhanced SDT mediated by cavitation effects of hollow solid nanoparticle. (a) Schematic image of ROS production ability of MON-PpIX-LA-CO2 under different US irradiations. (b) Electron spin resonance spectra of different groups. (c) Ultraviolet–visible spectra of 1,3-diphenylisobenzofuran after different treatments [133]. Copyright 2021, Elsevier. (d) Schematic diagram of the SDT based on porphyrin-based COFNs. (e) Schematic illustration showing the synthetic process of COFNs [135]. Copyright 2023, American Chemistry Society. (f) The synthesis process of ZnO and C-ZnO. (g) SEM images of mechanical damage to cancer cells induced by PBS, ZnO, or C-ZnO under US irradiation. Scale bar = 5 μm. h-i) ESR spectra demonstrating (h) 1O2 and (i) •OH generation of ZnO or C-ZnO with or without US. (j) Rate constants for 9,10-diphenylanthracene and MB decomposition in the presence of ZnO or C-ZnO under US irradiation [138]. Copyright 2024, Wiley-VCH Enhanced SDT mediated by cavitation effects of hollow solid nanoparticle. (a) Schematic image of ROS production ability of MON-PpIX-LA-CO2 under different US irradiations. (b) Electron spin resonance spectra of different groups. (c) Ultraviolet–visible spectra of 1,3-diphenylisobenzofuran after different treatments [133]. Copyright 2021, Elsevier. (d) Schematic diagram of the SDT based on porphyrin-based COFNs. (e) Schematic illustration showing the synthetic process of COFNs [135]. Copyright 2023, American Chemistry Society. (f) The synthesis process of ZnO and C-ZnO. (g) SEM images of mechanical damage to cancer cells induced by PBS, ZnO, or C-ZnO under US irradiation. Scale bar = 5 μm. h-i) ESR spectra demonstrating (h) 1O2 and (i) •OH generation of ZnO or C-ZnO with or without US. (j) Rate constants for 9,10-diphenylanthracene and MB decomposition in the presence of ZnO or C-ZnO under US irradiation [138]. Copyright 2024, Wiley-VCH UTMC/UTNC-Enhanced immunotherapy UTMC/UTNC induced Immunogenic cell death Tumor cell death that provokes adaptive immunity, immunogenic cell death, requires both antigenicity and adjuvanticity [139]. Compared with thermal ablation by HIFU, which induces apoptosis and protein denaturation, mechanical HIFU generates necrotic cell death and stronger adaptive immune responses marked by increased dendritic cell, T cell, and NK cell infiltration, as well as M1 macrophage polarization [140, 141]. Introduction of exogenous cavitation nuclei dramatically lowers the ultrasound pressure threshold for cavitation, allowing vigorous bubble collapse to lyse tumor cells, release damage-associated molecular patterns (DAMPs) and tumor antigens, and mature antigen-presenting cells. Application of blank NBs with low-frequency, low-intensity ultrasound disrupts both cell and nuclear membranes, liberates DAMPs, and, via spatiotemporally controlled mechanical injury, inhibits growth in multiple solid tumor models (Fig. 13a-c) [76, 83]. Phase-change nanodroplets, owing to their submicron size, extravasate into tumors after intravenous injection; upon ultrasound activation, they vaporize and undergo cavitation, generating ROS and inducing apoptosis and necrosis in cancer cells (Fig. 13d-e) [19]. Compared with chemotherapy, ROS therapy, or hyperthermia alone, UTMC/UTNC provokes more robust adaptive immunity. Necrotic apoptosis and release of mitochondrial and nuclear DNA activate the cyclic GMP-AMP synthase and stimulator of interferon genes (cGAS-STING) pathway, leading to type I IFN production and dendritic cell maturation (Fig. 13f-i) [142]. Cavitation of NBs under ultrasound induces necrotic apoptosis, direct membrane rupture, and DAMPs release, preserving immunogenic epitopes from oxidative or proteolytic degradation (Fig. 13j-k) [16–18]. Nuclear membrane disruption liberates nuclear DAMPs (e.g., IL-33, High mobility group protein B1, gasdermin E), which recruit eosinophils for direct tumor cell killing and mobilize NK and CD8+ T cells to enhance antitumor immunity (Fig. 13l) [143]. When combined with programmed death-1/programmed death ligand 1 (PD-1/PD-L1) blockade, UTMC/UTNC overcomes T cell suppression, stimulates CD8+ T cell maturation, and increases tumor necrosis factor-α (TNF-α), IFN-γ, perforin, and granzyme release, improving the tumor immune microenvironment and inhibiting both primary and distant tumor growth [19, 83]. Fig. 13UTMC/UTNC induces tumor cell death through cavitation-mediated mechanical disruption while promoting immunogenic cell death and sustaining high immunogenicity. (a) Schematic illustration showing the effects of ultrasound-stimulated NB on a mouse tumor model, which leads to systemic antitumor immunity and enhances the therapeutic effect of anti-PD1 treatment. (b) Western blot of tumor antigen OVA levels and immunogenic cell death marker levels in the supernatants, respectively. c) ATP concentration in the supernatants of RM1 cells in the different treatment groups [83]. Copyright 2022, BMJ Publishing Group Ltd. d) Schematic illustration of the two-step ND-mediated histotripsy method combined with ati-PD1 administration [19]. Copyright 2025, Wiley-VCH. e) Distribution of FITC labelled PPF-loaded nanodroplets in tumor vessels and stroma at different time points after tail vein injection. Scale bar = 100 μm [76]. Copyright 2025, Elsevier. f) The scheme of design and underlying therapeutic mechanism of nano-oncolytic system. g) Confocal laser scanning microscope images showing the staining of CT26 cells with Annexin V and PI to assess cell death modalities. h) Flow cytometry analysis of CRT exposure of CT26 cells after different treatments. i) Representative flow cytometry of in vitro dendritic cell (DC) maturation induced by CT26 cells after different treatments [142]. Copyright 2025, Wiley-VCH. j) Schematic illustration of necroptosis-inducible NBs for antitumor immune response. k) NB-induced necroptosis and DAMP release of CT26 cells [16]. Copyright 2020, Wiley-VCH. l) Schematics for the mechanism of NanoBurst-induced cancer cell death [143]. Copyright 2024, Wiley-VCH UTMC/UTNC induces tumor cell death through cavitation-mediated mechanical disruption while promoting immunogenic cell death and sustaining high immunogenicity. (a) Schematic illustration showing the effects of ultrasound-stimulated NB on a mouse tumor model, which leads to systemic antitumor immunity and enhances the therapeutic effect of anti-PD1 treatment. (b) Western blot of tumor antigen OVA levels and immunogenic cell death marker levels in the supernatants, respectively. c) ATP concentration in the supernatants of RM1 cells in the different treatment groups [83]. Copyright 2022, BMJ Publishing Group Ltd. d) Schematic illustration of the two-step ND-mediated histotripsy method combined with ati-PD1 administration [19]. Copyright 2025, Wiley-VCH. e) Distribution of FITC labelled PPF-loaded nanodroplets in tumor vessels and stroma at different time points after tail vein injection. Scale bar = 100 μm [76]. Copyright 2025, Elsevier. f) The scheme of design and underlying therapeutic mechanism of nano-oncolytic system. g) Confocal laser scanning microscope images showing the staining of CT26 cells with Annexin V and PI to assess cell death modalities. h) Flow cytometry analysis of CRT exposure of CT26 cells after different treatments. i) Representative flow cytometry of in vitro dendritic cell (DC) maturation induced by CT26 cells after different treatments [142]. Copyright 2025, Wiley-VCH. j) Schematic illustration of necroptosis-inducible NBs for antitumor immune response. k) NB-induced necroptosis and DAMP release of CT26 cells [16]. Copyright 2020, Wiley-VCH. l) Schematics for the mechanism of NanoBurst-induced cancer cell death [143]. Copyright 2024, Wiley-VCH Enhanced delivery of nanomaterials and antibodies Aberrant tumor vasculature, dense stroma, and elevated interstitial fluid pressure (IFP) hinder drug and immune cell infiltration [144]. Using a Millar Mikro-Tip pressure catheter (SPR-1000, 0.33 mm), tumor IFP was shown to drop immediately after UTMC, reaching a nadir (~ 40% below baseline) at 20 min, then gradually returning toward baseline by 80 min (Fig. 14a) [145]. This IFP reduction enhances intratumoral delivery and accumulation of PEOz-Lip@R837-IR820 and downregulates immunosuppressive checkpoint molecules (Foxp3, PD-L1, CTLA-4), alleviating the tumor’s immunosuppressive microenvironment. Modulating ultrasound parameters to induce MB cavitation can also dilate tumor microvessels, normalize perfusion, and increase CD8+ T cell infiltration and anti-PD-L1 antibody penetration, yielding superior tumor control and survival versus anti-PD-L1 alone (Fig. 14b) [146]. Perfluoropentane nanodroplets undergo ultrasound-induced vaporization and MB collapse, stripping away plasma-derived protein coronas, restoring ligand-receptor targeting, and enhancing deep tumor delivery [147]. Simultaneously, coaccumulated sonosensitizers generate ROS under ultrasound to induce immunogenic death and dendritic cell activation, synergizing with anti-PD-L1 to suppress pancreatic tumors [132]. Fig. 14UTMC/UTNC enhanced tumor immunotherapy. (a) Tumor interstitial fluid pressure in mice bearing Panc02 tumors following ultrasound-mediated cavitation treatment [145]. Copyright 2022, Elsevier. (b) Schematic illustration of the combination therapy of ultrasound-stimulated MB cavitation with anti-PD-L1. The cavitation effect of ultrasound-stimulated MB cavitation promotes the tumor blood perfusion, which enhances CD8+ T cell infiltration and anti-PD-L1 delivery, consequently boosting the number and the activity of CD8+ T cells for effective cancer immunotherapy [146]. Copyright 2021, Elsevier. (c) Schematic illustration showing the synthesis of DTX-R837@NBs and the mechanism of the antitumor response induced by DTX-R837@NBs-based ultrasound in combination with immune checkpoint blockade. (d) DC maturation in tumor-draining lymph nodes induced by DTX-R837@NBs + US in mice bearing RM-1 tumors, as assessed by flow cytometry after staining for CD11c, CD80, CD86 and CD16/32. d-e) Primary and distant tumor growth curves of different groups of mice after various treatments [5]. Copyright 2025, Elsevier UTMC/UTNC enhanced tumor immunotherapy. (a) Tumor interstitial fluid pressure in mice bearing Panc02 tumors following ultrasound-mediated cavitation treatment [145]. Copyright 2022, Elsevier. (b) Schematic illustration of the combination therapy of ultrasound-stimulated MB cavitation with anti-PD-L1. The cavitation effect of ultrasound-stimulated MB cavitation promotes the tumor blood perfusion, which enhances CD8+ T cell infiltration and anti-PD-L1 delivery, consequently boosting the number and the activity of CD8+ T cells for effective cancer immunotherapy [146]. Copyright 2021, Elsevier. (c) Schematic illustration showing the synthesis of DTX-R837@NBs and the mechanism of the antitumor response induced by DTX-R837@NBs-based ultrasound in combination with immune checkpoint blockade. (d) DC maturation in tumor-draining lymph nodes induced by DTX-R837@NBs + US in mice bearing RM-1 tumors, as assessed by flow cytometry after staining for CD11c, CD80, CD86 and CD16/32. d-e) Primary and distant tumor growth curves of different groups of mice after various treatments [5]. Copyright 2025, Elsevier Synergy with other therapies Combining UTMC/UTNC with other therapies (Chemotherapy, SDT, photothermal therapy, photodynamic therapy, etc.), immune adjuvants, and checkpoint blockade amplifies antitumor immunity [148]. DOX-loaded MBs under ultrasound enhance tumor uptake and apoptosis and release DAMPs to activate adaptive immunity [149]. UTMC-mediated delivery of pH-sensitive adjuvant R837 induces dendritic cell maturation and T cell activation in situ [145]. NBs coloaded with docetaxel and R837, upon ultrasound irradiation, release both agents and create an in situ tumor vaccine via DAMPs, R837, and tumor antigens, driving adaptive immunity. Checkpoint blockade added to this regimen yields robust tumor suppression and durable memory that prevents recurrence (Fig. 14c-f) [5, 150]. Furthermore, photothermal, photodynamic, and sonodynamic therapies can be integrated to induce immunogenic cell death, improve tumor immunogenicity, and inhibit metastasis, demonstrating significant translational potential [151, 152]. Vaccine delivery and dendritic cell activation Lipid MB vaccines targeting dendritic cells can also prime adaptive immunity. OVA-conjugated MBs (OVA-MBs) are internalized by dendritic cells for antigen processing and presentation, enabling specific recognition of OVA-expressing tumors [153]. Robust immunity requires codelivery of multiple antigens and adjuvants via mannose, aptamer, or antibody targeting to mature dendritic cells and initiate adaptive responses [154, 155, 156, 157]. DCs matured in vitro with tumor cell membranes were used to generate DC membrane-coated MBs (DC-iMBs). These DC-iMBs migrated to secondary lymphoid organs in vivo, inducing robust antitumor immunity and reducing tumor growth through T cell-mediated apoptosis [158]. Nanocomplex-conjugated MBs (ncMBs) bearing anti-CD11b and the STING agonist cGAMP target macrophages, activate cGAS-STING, drive CD8+ T cell proliferation, and IFN-γ release, when combined with anti-PD-1, markedly suppress breast cancer and melanoma growth and metastasis [159, 160]. Direct T cell activation Cavitation-induced shear stress can potentiate immune function. Streptavidin MBs targeting HEK293T cells under 2 MHz ultrasound triggered Piezo1 channel opening, Ca2+ influx, and activation of CD19 chimeric antigen receptor T cells, as evidenced by CD69 upregulation following co-culture with CD19+ targets [161]. Ultrasound-induced hyperthermia can also activate chimeric antigen receptor T cells engineered with heat shock promoters, inhibiting tumors without systemic immune activation [162]. Shear stress-activated Piezo1 signaling increases key T cell activation proteins (ZAP70, NFAT, NF-κB, AP-1) and cytokines (TNF-α, IL-2, IFN-γ), further enhancing T cell responses [163]. Thus, UTMC/UTNC may serve not only to deliver therapies but also to directly modulate T cell activation and augment antitumor immunity. Enhancement of gene delivery Intracellular gene delivery UTMC/UTNC enhances the precision and efficiency of gene therapy through a synergistic combination of cavitation and sonoporation. Transient shock waves generated by cavitation detach genetic payloads from bubble surfaces and accelerate their diffusion through the interstitium, while shear forces from bubble oscillation create membrane pores that facilitate cytosolic entry of nucleic acids. This noninvasive, precisely controlled approach represents a highly efficient gene-therapy platform. Optimizing gene delivery efficiency is critical for maximizing therapeutic outcomes. Ligand-decorated targeted MBs adhere selectively to tumor cells and undergo asymmetric oscillation under ultrasound, producing higher amplitude mechanical forces than untargeted bubbles. Intratumoral injection of targeted bubbles loaded with an interferon-beta (IFN-β) plasmid achieved superior transfection and local IFN-β expression, reducing tumor cell viability and recruiting CD8+ T cells and macrophages to inhibit tumor growth (Fig. 15a) [2]. Cationic DOTAP lipids on NBs bind negatively charged siRNAs (e.g., SIRPα or STAT3 siRNA), and folate functionalization directs these nanocarriers to macrophages. Upon ultrasound activation and phase transition, siRNA release and uptake suppress “don’t eat me” signaling (SIRPα) and M2 polarization (STAT3), thereby enhancing phagocytosis, M1 polarization, T cell activation, and antitumor immunity in non-small cell lung cancer [164, 165]. Although PEGylation stabilizes bubbles and prolongs circulation, excess PEG impedes cellular uptake and endosomal escape. Fluorinated PEG lipids on nanodroplets can be selectively shed during cavitation, boosting cellular internalization, membrane fusion, and mRNA release. This strategy increased splenic transfection of full-length ovalbumin mRNA approximately fourfold, elicited potent CD8+ T cell responses, and produced strong antitumor efficacy in OVA-expressing models, extending the clinical potential of mRNA therapeutics (Fig. 15b-c) [36]. Intranuclear gene delivery Conventional ultrasound-mediated delivery strategies facilitate gene transport from the extracellular space into the cytoplasm but typically fail to deliver genetic material directly into the nucleus, limiting transfection efficiency. A novel approach utilizing intracellular cavitation from biological GVs has been proposed to overcome this barrier. GVs were surface-modified to load plasmid E-cadherin DNA (pEcad). Following co-incubation with cells, the complexes entered the cytoplasm and, upon ultrasound exposure, generated cavitation capable of transiently perforating the nuclear membrane. This strategy enhanced nuclear gene delivery and improved transfection efficiency by 47%. (Fig. 15d) [166]. Interestingly, the efficacy of this method varied with the cell cycle. Although transfection efficiency remained comparable across G1, S, and G2/M phases (Fig. 15e), ultrasound-mediated nuclear delivery of E-cadherin during the G2/M phase significantly downregulated Fam50a expression, resulting in improved inhibition of tumor invasion and metastasis (Fig. 15f-g) [166]. To minimize irreversible damage from rapid cavitation, surface acoustic waves have been employed to induce mild sonothermal effects that transiently permeabilize both plasma and nuclear membranes, enabling safe, efficient plasmid transfection in the absence of MBs [167]. By leveraging ultrasound’s spatiotemporal precision, these strategies overcome low efficiency and poor targeting in gene delivery, paving the way for clinical translation. Fig. 15UTMC/UTNC enhances intracellular and intranuclear gene delivery. (a) Schematic illustration of tumor transfection steps of operation [2]. Copyright 2020, National Academy of Sciences. (b) Schematic illustration of the spleen-targeted LNP-mRNA delivery. (c) Spleen-targeted F19-LNPs induced gene editing in the spleen and other organs. tdTomato fluorescence in the main organs was detected 2 days after intravenous injection of Cre mRNA-loaded LNPs [36]. Copyright 2025, Wiley-VCH. (d) Schematic illustration of nuclear gene delivery via intracellular cavitation of pDNA-GV triggered by acoustic irradiation. (e) Schematic illustration of nuclear gene delivery in a running cell cycle model, in which nocodazole-synchronized pEcad-GVs@C6 cells are released, followed by nocodazole-free medium release and acoustic irradiation at the SynG2/M, SynG1, or SynS phases. Flow cytometry assays of mCherry gene expression efficiency in the asynchronized or synchronized pEcad-GVs@C6 tumor cells received with or without acoustic irradiation at SynG2/M (0 h after cell cycle release), SynG1 (8 h after cell cycle release), or SynS phases (14 h after cell cycle release) in the operating cell cycle. (f) The quantitative results of Transwell assay, revealing a stronger inhibitory effect of tumor cell migration when nuclear E-cadherin delivery was applied in the G2/M phase. (g) Number of lung metastatic nodules from each group [166]. Copyright 2023, Springer Nature UTMC/UTNC enhances intracellular and intranuclear gene delivery. (a) Schematic illustration of tumor transfection steps of operation [2]. Copyright 2020, National Academy of Sciences. (b) Schematic illustration of the spleen-targeted LNP-mRNA delivery. (c) Spleen-targeted F19-LNPs induced gene editing in the spleen and other organs. tdTomato fluorescence in the main organs was detected 2 days after intravenous injection of Cre mRNA-loaded LNPs [36]. Copyright 2025, Wiley-VCH. (d) Schematic illustration of nuclear gene delivery via intracellular cavitation of pDNA-GV triggered by acoustic irradiation. (e) Schematic illustration of nuclear gene delivery in a running cell cycle model, in which nocodazole-synchronized pEcad-GVs@C6 cells are released, followed by nocodazole-free medium release and acoustic irradiation at the SynG2/M, SynG1, or SynS phases. Flow cytometry assays of mCherry gene expression efficiency in the asynchronized or synchronized pEcad-GVs@C6 tumor cells received with or without acoustic irradiation at SynG2/M (0 h after cell cycle release), SynG1 (8 h after cell cycle release), or SynS phases (14 h after cell cycle release) in the operating cell cycle. (f) The quantitative results of Transwell assay, revealing a stronger inhibitory effect of tumor cell migration when nuclear E-cadherin delivery was applied in the G2/M phase. (g) Number of lung metastatic nodules from each group [166]. Copyright 2023, Springer Nature UTMC/UTNC-Enhanced radiotherapy Radiotherapy exerts antitumor effects primarily via DNA damage-induced cell death. However, DNA repair, signal transduction, hypoxia, and angiogenesis often confer radioresistance. When combined with UTMC/UTNC, radiotherapy’s cytotoxicity exceeds that of radiation alone by an order of magnitude [168]. MBs coloaded with oxygen and sulfur hexafluoride (OS MBs) retain excellent ultrasound contrast and stability (Fig. 16a-c). Upon ultrasound-triggered cavitation (OS MBs + US), tumor hypoxia is alleviated: photoacoustic imaging demonstrates significant increases in oxygenated hemoglobin signals throughout the tumor, including deep regions (Fig. 16d). When OS MBs are employed alongside 125I brachytherapy (BT + OS MBs), maximal tumor suppression is achieved (Fig. 16e) [169]. Further studies showed that UTMC/UTNC combined with radiotherapy reduced vessel oxygenation, CD31-stained microvessel density, and vessel perimeter significantly within 24 h post-treatment. Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining revealed increased apoptotic areas, with UTNC exhibiting superior radiosensitization compared to UTMC (Fig. 16f-h) [170]. These effects are attributed to enhanced tumor cell death stimulating adaptive immunity, cavitation- and heat-mediated improvements in perfusion overcoming hypoxia, and sonoporation-facilitated delivery of radiosensitizers across endothelial barriers collectively reversing radioresistance [171, 172]. Early clinical investigations corroborate these findings. In a Phase I trial, MRI-guided focused ultrasound with Definity MBs induced vascular perturbation, increased perfusion, and pro-apoptotic signaling, leading to endothelial apoptosis, reduced microvessel density, and increased tumor hypoxia-induced cell death. Among 18 breast cancer patients, nine achieved complete remission, six partial remission, two stable disease, and one progression, demonstrating radiotherapy enhancement [173]. UTMC combined with yttrium 90 transarterial radioembolization has demonstrated improved response rates and prolonged survival in patients with hepatocellular carcinoma, intrahepatic cholangiocarcinoma and/or liver metastases, without compromising safety [174 , 175, 176]. Nevertheless, radiation-related adverse events should be carefully monitored and managed according to established clinical guidelines [177]. Fig. 16UTMC/UTNC potentiates radiotherapy efficacy. (a) Schematic representation of the emulsion freeze-drying strategy employed to fabricate the MBs, and the illustration of the radiosensitizing effect of the O2/SF6 MB (OS MB) for cancer treatment with 125I brachytherapy. (b) Ultrasound images of SF MBs, OS MBs, and OX MBs in the gel phantom at specific time intervals. (c) Relative average gray values of ultrasound images of SF MBs, OS MBs, and OX MBs to initial average gray values (t = 0 s) obtained from the time-series ultrasound images in panel. (d) Ultrasound images of tumor region and photoacoustic images of tumor region in different groups. (e) Tumor growth curves of different groups with various treatments for twelve days [169]. Copyright 2021, Wiley-VCH. f-h) PA estimations in tumor sO2, CD31 positive area and TUNEL positive area relative to the untreated control as a function of treatment and imaging timepoint [170]. Copyright 2020, Elsevier UTMC/UTNC potentiates radiotherapy efficacy. (a) Schematic representation of the emulsion freeze-drying strategy employed to fabricate the MBs, and the illustration of the radiosensitizing effect of the O2/SF6 MB (OS MB) for cancer treatment with 125I brachytherapy. (b) Ultrasound images of SF MBs, OS MBs, and OX MBs in the gel phantom at specific time intervals. (c) Relative average gray values of ultrasound images of SF MBs, OS MBs, and OX MBs to initial average gray values (t = 0 s) obtained from the time-series ultrasound images in panel. (d) Ultrasound images of tumor region and photoacoustic images of tumor region in different groups. (e) Tumor growth curves of different groups with various treatments for twelve days [169]. Copyright 2021, Wiley-VCH. f-h) PA estimations in tumor sO2, CD31 positive area and TUNEL positive area relative to the untreated control as a function of treatment and imaging timepoint [170]. Copyright 2020, Elsevier UTMC/UTNC-Induced ferroptosis Ferroptosis is an iron-dependent form of regulated cell death characterized by lipid peroxidation and ROS accumulation. Glutathione peroxidase 4 (GPX4), a glutathione-dependent enzyme, prevents ferroptosis by detoxifying lipid hydroperoxides. UTMC promotes ferroptosis via two mechanisms: cavitation-generated ROS (1O2, •OH) disrupts the redox balance and depletes GSH, thus inhibiting GPX4 activity [134], and cavitation-induced mechanical stress activates Piezo1 channels, triggering iron influx independent of transferrin receptors, elevating intracellular iron and mitochondrial ROS to induce ferroptosis [178]. Although UTMC alone generates limited ROS and modest ferroptosis, codelivery of simvastatin in phase-change nanodroplets enhances GSH depletion while inhibiting mevalonate pathway-derived intermediates, downregulating GPX4 and promoting tumor cell ferroptosis [134]. Further amplification is achieved by combining UTMC with exogenous iron. Hollow mesoporous silica nanoparticles coloaded with air and FeOOH and cloaked in cancer cell membranes (MSF@CCM) serve as cavitation nuclei and Fenton catalysts. Under ultrasound, cavitation enhances deep tumor penetration, while FeOOH-mediated Fenton reactions produce 140% more 1O2 and 186% more •OH, intensifying oxidative stress, lipid peroxidation, mitochondrial disruption, and downregulation of GPX4 and ACSL4, robustly triggering ferroptosis [179]. Thus, UTMC facilitates ferroptosis both by generating ROS to consume GSH and by enhancing the intratumoral accumulation of FeOOH and O2, thereby increasing the frequency of Fenton reactions, a key mechanism for achieving robust ferroptotic tumor cell death. Conclusions and perspectives Ultrasound offers precise, noninvasive, and deep-tissue penetration advantages, and when combined with exogenous cavitation nuclei (bubbles, GVs, nanodroplets and hollow solid nanoparticles), it enables substantial advances in tumor imaging and therapy. The approach is operationally straightforward, cost-effective, and compatible with existing clinical workflows, providing a readily translatable platform for localized, precision treatment of solid tumors. Although bubbles were originally developed as ultrasound contrast agents for real-time diagnostic imaging, their controlled cavitation under ultrasound has established a novel strategy for targeted drug delivery and localized tumor ablation. Consequently, this review focuses on UTMC/UTNC-mediated therapeutic strategies, summarizing their integration with chemotherapy, SDT, immunotherapy, gene delivery, radiotherapy, and ferroptosis induction. Several critical challenges must be addressed to enable successful clinical translation and further development: (1) Parameter Standardization and Mechanism Clarification: Considerable variation exists across studies regarding UTMC/UTNC application parameters, including MBs/NBs concentration, ultrasound frequency, intensity, duration, and duty cycle. A systematic exploration of the relationships between these parameters and the induction of stable versus inertial cavitation is urgently needed. Understanding these dynamics will guide the optimization of treatment conditions for specific therapeutic goals, such as efficient drug delivery, safe and reversible biological barrier opening, and mechanical tumor cell ablation, thereby providing a theoretical framework for precision ultrasound therapy. (2) Bubble Stability and Biocompatibility Enhancement: Improving the in vivo stability and biosafety of MBs/NBs, while avoiding rapid clearance by mononuclear phagocytes, is a key challenge in achieving effective diagnosis and therapy. Although the shell materials used for blank bubble fabrication are generally safe, drug-loaded formulations are prone to premature release in circulation. In addition to relying on passive tumor targeting via UTMC/UTNC, the design of intelligent nanocarriers with precise targeting and controlled drug release capabilities is crucial for enhancing therapeutic efficacy and biocompatibility. Notably, lipid-based particles, especially those sized 100 nm to 1 μm with negatively charged surfaces, tend to be rapidly sequestered by the reticuloendothelial system, primarily in the liver and spleen [103]. To prolong systemic circulation and improve tumor-specific accumulation, several strategies can be employed: surface modification (e.g., PEGylation, CD47 biomimetic coating), tuning physicochemical properties (e.g., size, charge, mechanical stiffness), and bioinspired membrane camouflage (e.g., tumor cell membranes, red blood cell membranes, or exosomes). Additionally, dosing regimens can be optimized, for example, by pre-injecting blank lipid materials to reduce macrophage-mediated clearance. Current clinically approved ultrasound contrast agents are limited by short imaging duration and poor in vivo stability. Therefore, future efforts should focus on optimizing core-shell composition, applying multilayered surface modifications, and developing next-generation formulations to improve stability and enable advanced biomedical applications. Further optimization of bubble fabrication techniques is essential for generating MBs/NBs with more uniform particle sizes and improved physicochemical stability, which is critical for CEUS imaging in vivo performance [42]. (3) Biomimetic Materials as a Future Direction: Biomimetic materials represent a promising frontier for advancing ultrasound-mediated tumor theranostics. Beyond traditional cell membrane-coated bubbles, biologically derived GVs have demonstrated considerable potential in both imaging and therapeutic applications. GVs, naturally produced by microorganisms, exhibit excellent biocompatibility, low cavitation thresholds due to their rigid protein shells, genetic encodability, and inherent immunogenicity, along with the ability to penetrate biological barriers. However, their clinical translation is currently limited by insufficient imaging signal, weak drug-loading capacity, and structural instability. Genetic engineering of GVs offers opportunities to enhance their acoustic responsiveness, expand their applicability in ultrasound imaging, and develop novel nanomedicine strategies. Moreover, given the heterogeneity of tumor tissues, biomimetic modification of nanocarriers holds great promise for enabling personalized and precision oncology. (4) Stimuli-Responsive and Multifunctional Smart Nanomaterials: Smart nanomaterials capable of responding to tumor microenvironment cues, such as hypoxia, acidic pH, elevated GSH levels, and enzymatic activity, offer promising strategies for enhancing therapeutic specificity and efficacy. By enabling site-specific drug release, these materials can increase local drug concentrations while reducing systemic toxicity, thus improving treatment outcomes and biosafety. Furthermore, the integration of multiple imaging modalities, such as ultrasound, photoacoustic, and fluorescence imaging into bubble-based platforms, facilitates real-time monitoring, image-guided delivery, and adaptive treatment planning. Future developments may focus on coupling endogenous TME responsiveness with external triggers (e.g., ultrasound, light, or magnetic fields) to achieve highly controllable, individualized, and efficient cancer theranostics. In summary, MBs/NBs, when coupled with UTMC/UTNC, have demonstrated notable advances in cancer theranostics. These systems significantly enhance tumor-specific drug delivery, facilitate BBB opening, and amplify the efficacy of chemotherapy, sonodynamic therapy, and immunotherapy via cavitation-mediated mechanisms. The multifunctional properties of MBs/NBs, such as their ability to serve as ultrasound contrast agents, cargo carriers, and immune modulators, offer precise spatiotemporal control over therapeutic payloads while minimizing off-target toxicity. Despite these achievements, the mechanical damage imposed on vascular and adjacent normal tissues induced during UTMC/UTNC remains a persistent safety challenge. Continued optimization of ultrasound parameters, bubble formulations, and real-time cavitation monitoring is required to ensure therapeutic efficacy while minimizing adverse bioeffects. Clinical translation is still constrained by inconsistencies in treatment parameters, suboptimal bubble stability, and limited drug-loading capacity. Moreover, tumor heterogeneity and immune evasion remain major obstacles. Future success will rely on interdisciplinary collaboration to optimize bubble design, acoustic protocols, and delivery strategies. Emphasis should be placed on developing bioinspired or gene-encoded materials, scalable synthesis methods, and intelligent nanoplatforms tailored to the tumor microenvironment. Together, these efforts may unlock the full clinical potential of UTMC/UTNC in next-generation, non-invasive oncology therapeutics.