Biosensors and Nanotheranostics

1. Introduction

Primary liver cancer sits stubbornly among the leading causes of cancer death worldwide, and hepatocellular carcinoma (HCC)—roughly 85% of cases—accounts for most of that burden (Du et al., 2025; Zhao & Zhao, 2026). The problem is not a shortage of treatment options but the ceiling each one hits. Surgical resection can be curative, yet fewer than a third of patients are operable at diagnosis, typically because the disease is detected late, has already invaded the vasculature, or sits on a background of cirrhosis that will not tolerate further loss of functioning liver (Du et al., 2025; Wei et al., 2024). Systemic chemotherapy is undercut by intrinsic multidrug resistance and by off-target toxicity—hepatotoxicity and nephrotoxicity chief among them—that limits the dose a patient can actually receive (Zhao & Zhao, 2026). Even immune checkpoint blockade, which reshaped the outlook in several other malignancies, delivers only modest response rates in HCC, held back by dense extracellular matrix, persistent intracellular signaling, and an intensely immunosuppressive tumor microenvironment (TME) that keeps cytotoxic lymphocytes at bay (Du et al., 2025; Wei et al., 2024).

The common denominator across these failures is access: the liver is deep, mechanically shielded, and immunologically defended. What the field has lacked is a modality that is at once non-invasive, spatially precise, and able to act at depth without scorching the healthy parenchyma it must pass through (Peng et al., 2025; Uinarni et al., 2026). That requirement is what motivates the turn toward acoustically driven therapy.

Sonodynamic therapy (SDT) answers the depth problem by replacing light with ultrasound. Photodynamic therapy, for all its elegance, is confined to the first centimeter or so of tissue because photons scatter and are absorbed long before they reach a visceral organ (Y. Huang et al., 2024; Uinarni et al., 2026). Low-intensity focused ultrasound, by contrast, propagates through tens of centimeters of soft tissue with little attenuation, so a sonosensitizer sequestered in a liver tumor can be excited from outside the body (Jeong et al., 2024; Zhang et al., 2026). The therapeutic payload is the same family of cytotoxic reactive oxygen species (ROS)—singlet oxygen (1O2), hydroxyl radicals (•OH), and superoxide anions (•O2–)—that drive lipid peroxidation, DNA strand breaks, mitochondrial collapse, and programmed cell death (Bai et al., 2026; Uinarni et al., 2026; Zhao & Zhao, 2026).

Mechanistically, the bridge between an acoustic wave and a chemical radical is acoustic cavitation—the nucleation, growth, and collapse of microbubbles in the tissue fluid (N. Yang et al., 2024). Inertial cavitation, the violent form, generates transient micro-hotspots exceeding 5000 K and 80 MPa, emits short-wavelength light (sonoluminescence) that excites adjacent sensitizer molecules, and pyrolyzes water into free radicals (Jeong et al., 2024; N. Yang et al., 2024). Stable cavitation, the gentler oscillatory form, produces microstreaming and shear that transiently permeabilize cell membranes—sonoporation—and so improves the intracellular delivery of the therapeutic agents themselves (Jeong et al., 2024; Liu et al., 2026; Uinarni et al., 2026).

The promise of SDT was, for years, held hostage by the sonosensitizers available to deliver it. Classical small molecules such as the porphyrin derivatives are poorly water-soluble, clear quickly from blood, lack organelle specificity, and convert acoustic energy into ROS with disappointing efficiency (Peng et al., 2025; Wei et al., 2024). Nanotechnology changed the calculus. Liposomes, polymeric micelles, metal–organic frameworks (MOFs), inorganic semiconductor particles, and micro/nanobubbles improve solubility and stability while concentrating the sensitizer in the tumor through the enhanced permeability and retention (EPR) effect and through active ligands such as cRGD, SP94, and hyaluronic acid (Du et al., 2025; Meng et al., 2025; Wei et al., 2024). More unusual strategies push accumulation further still—repurposing the clinical vasodilator isosorbide mononitrate (ISMN) to dilate hepatic microvessels raises nanosonosensitizer uptake by roughly 1.8-fold (Peng et al., 2025; Zhang et al., 2026).

Delivery, however, is only half the battle. The hepatic TME is hypoxic and chemically reducing: elevated glutathione (GSH) mops up ROS almost as fast as SDT can make them (Meng et al., 2025; Zhao & Zhao, 2026). The answer has been to build nanoplatforms that remodel the microenvironment as they treat it—oxygen-releasing nanobubbles, catalase-bearing bacterial hybrids, and transition-metal nanozymes (MnO2, FeMoO4, TiZrRu-MON) that convert intratumoral H2O2 into O2 while draining GSH (Jiang et al., 2025; Meng et al., 2025; T. Wang et al., 2026). By restoring redox balance, these systems amplify ROS and engage several cell-death programs at once—apoptosis, ferroptosis, pyroptosis, and cuproptosis (Jiang et al., 2025; Liang et al., 2025; Meng et al., 2025). Many of the same particles carry an imaging function, so that a single construct supports ultrasound, photoacoustic, and magnetic resonance imaging for real-time localization and response monitoring—true theranostics (Jiang et al., 2025; Uinarni et al., 2026; Zhang et al., 2026).

Perhaps the most consequential development is that SDT does not merely kill cells quietly. Ultrasound-driven ROS and cavitation provoke immunogenic cell death (ICD), exposing calreticulin and releasing high-mobility group box 1 (HMGB1) and ATP (Du et al., 2025; N. Yang et al., 2024; Zhao & Zhao, 2026). These damage-associated molecular patterns recruit and mature dendritic cells, prime CD8+ cytotoxic T lymphocytes, and convert a cold tumor into a hot one—an effect that pairs naturally with anti-PD-L1 blockade to suppress both primary and distant lesions (Du et al., 2025; Wei et al., 2024). Combinations with chemodynamic therapy and gaseous transmitters (CO, NO, SO2) extend the cascade further (Liang et al., 2025; Zhao & Zhao, 2026).

Against this backdrop, the present review synthesizes recent preclinical progress in ultrasound-activated sonodynamic nanoparticles for deep-tissue theranostics of HCC. Our aims are fivefold: to clarify the biophysical, chemical, and immunological mechanisms of SDT at depth; to organize the expanding catalog of acoustic nanoplatforms by design logic; to evaluate the imaging modalities that guide them; to map the synergistic combinations that overcome the immunosuppressive liver TME; and to confront the translational bottlenecks—dosimetry, biosafety, and model fidelity—that will decide whether any of this reaches patients (Uinarni et al., 2026; Zhao & Zhao, 2026).

2. Ultrasound-Activated Nanotheranostics Reprogram the Hypoxic Liver Tumor Microenvironment and Prime Antitumor Immunity

This section examines the state of the evidence in four movements: the evolution of sonosensitizer chemistry, the biophysics and microenvironmental regulation of deep-liver SDT, the synergistic combination strategies now dominating the literature, and the theranostic imaging that ties diagnosis to treatment. The underlying mechanistic logic is summarized in Figure 1, and the material classes, acoustic parameters, combination regimens, and imaging platforms discussed here are tabulated in Tables 1–4, respectively.

2.1. Evolution of Sonosensitizers: From Small Molecules to Engineered Nanostructures

The efficacy of SDT depends, before anything else, on the sonosensitizer—its acoustic responsiveness, its stability in blood, and its willingness to accumulate where it is needed (Zhao & Zhao, 2026). First-generation agents were borrowed wholesale from photodynamic therapy: hematoporphyrin, protoporphyrin IX (PpIX), and chlorin e6 (Ce6) (Jeong et al., 2024). Their planar 18-π aromatic macrocycles do generate singlet oxygen under acoustic activation, but their translation to a deep target like HCC is frustrated by poor aqueous solubility, rapid clearance, cutaneous photosensitivity, and low tumor uptake (Jeong et al., 2024; Zhao & Zhao, 2026). Worse, hydrophobic molecules stack through π–π interactions in physiological media, and the resulting aggregation-caused quenching (ACQ) depresses ROS yields precisely when concentration is highest (Zhao & Zhao, 2026).

Second- and third-generation designs addressed these faults through chemistry and then through architecture (Du et al., 2025). Modified porphyrins—hematoporphyrin monomethyl ether (HMME), sinoporphyrin sodium (DVDMS), and zinc phthalocyanines—improved acoustic behavior (Y. Huang et al., 2024; Jeong et al., 2024). More recent organic sonosensitizers exploit narrow-bandgap acceptor–donor configurations (e.g., BTz) and aggregation-induced emission (AIE) luminogens, which invert the ACQ problem: restricting intramolecular rotation on self-assembly suppresses non-radiative decay and raises ROS output under ultrasound (Peng et al., 2025; Zhang et al., 2026; Zhao & Zhao, 2026). In parallel, inorganic nanosonosensitizers—TiO2, MnO2, noble-metal nanocages, and two-dimensional dichalcogenides—introduced a semiconductor mechanism, in which cavitation assists electron–hole separation and the excited electrons reduce oxygen or water to •O2– and •OH (Y. Huang et al., 2024; Zhao & Zhao, 2026). These materials are thermally robust and resist photobleaching, though their slow hepatobiliary clearance has pushed the field toward organic–inorganic hybrids such as TiZrRu-MON and TCPP-based MOFs (Jiang et al., 2025; Zhang et al., 2026). The comparative strengths and liabilities of these classes are laid out in Table 1.

Carrier engineering completes the picture. Liposomes, polymeric micelles, biomimetic membrane coatings, and mesoporous silica shield the payload from enzymatic attack and extend circulation, while active ligands—cRGD against αvβ3 integrins, hyaluronic acid against CD44, SP94 against HCC membranes—sharpen targeting beyond passive EPR leakage (Du et al., 2025; Meng et al., 2025; Wei et al., 2024; Zhang et al., 2026). Where perfusion itself is the barrier, pharmacology helps: ISMN transiently dilates tumor microvessels, lifting nanosonosensitizer accumulation about 1.8-fold and raising tumor growth inhibition from 39% to 78% in murine HCC (Peng et al., 2025; Zhang et al., 2026).

2.2. Biophysical Mechanics and Microenvironment Regulation in Deep Liver Tissues

The physical foundation of SDT is the ability of low-intensity focused ultrasound (LIFU) to traverse intact anatomy and excite a localized sensitizer without harming the tissue it crosses (Uinarni et al., 2026; L. Wang et al., 2026). That excitation runs through acoustic cavitation, which comes in two regimes (Jeong et al., 2024; L. Wang et al., 2026). Stable cavitation—sustained, low-amplitude bubble oscillation—drives microstreaming and shear that

Figure 1. Biophysical and sonochemical mechanism of ultrasound-activated sonodynamic therapy at depth. Low-intensity focused ultrasound drives acoustic cavitation at the nanosonosensitizer; inertial cavitation produces micro-hotspots, sonoluminescence, and pyrolytic radicals, while stable cavitation generates microstreaming and sonoporation that enhance delivery. The two pathways converge on a reactive-oxygen-species burst (1O2, •OH, •O2–) that inflicts lipid peroxidation, DNA damage, and mitochondrial and endoplasmic-reticulum injury, culminating in apoptosis, ferroptosis, pyroptosis, cuproptosis, and immunogenic cell death of hepatocellular carcinoma cells (Jeong et al., 2024; N. Yang et al., 2024; Zhao & Zhao, 2026).

Figure 2. Hepatic tumor-microenvironment barriers to sonodynamic therapy and the nanoplatform strategies devised to overcome them. The left column lists the three principal obstacles—intratumoral hypoxia, elevated glutathione, and poor perfusion and delivery—while the right column pairs each with its engineered countermeasure: catalytic or carrier-based oxygen supply, redox-active metal-mediated glutathione depletion, and ligand- or vasodilator-assisted delivery. The combined effect restores oxygenation and collapses the antioxidant buffer, sustaining the reactive-oxygen-species cascade and sensitizing tumor cells to ferroptosis and cuproptosis (Jiang et al., 2025; Li et al., 2026; Meng et al., 2025; Peng et al., 2025; Zhao & Zhao, 2026).

loosen tight junctions and membranes, the sonoporation effect that raises vascular permeability and nanoparticle uptake in dense hepatic tissue (Liu et al., 2026; L. Wang et al., 2026). Inertial cavitation—the asymmetric implosion of bubbles—creates the extreme transient conditions (>5000 K, >80 MPa), microjets, and sonoluminescence that excite sensitizers indirectly and pyrolyze water into hydroxyl radicals (Jeong et al., 2024; N. Yang et al., 2024). The parameter windows that separate these regimes, and their biological consequences, are compiled in Table 2; the full mechanistic chain is depicted in Figure 1.

The limiting factor in the liver is rarely the physics and usually the microenvironment (Meng et al., 2025; Zhao & Zhao, 2026). Rapid proliferation, disordered vasculature, and high oxygen consumption leave solid liver tumors hypoxic (Y. Huang et al., 2024; N. Yang et al., 2024). Because Type II sonodynamic reactions require ground-state molecular oxygen to produce 1O2, hypoxia directly throttles ROS output (Zhang et al., 2026; Zhao & Zhao, 2026). Compounding this, malignant hepatocytes carry GSH at up to ten times normal levels, and this antioxidant buffer scavenges ROS before they can do damage (N. Yang et al., 2024; Zhao & Zhao, 2026).

Advanced nanoplatforms therefore remodel the TME as a condition of working at all, and the countermeasures fall into two broad families illustrated in Figure 2 (Li et al., 2026; Meng et al., 2025). Hypoxia is relieved either by exogenous oxygen carriers—perfluorocarbon nanodroplets and oxygen-loaded nanobubbles that release their cargo on cavitation—or by in situ catalysis, in which MnO2, Pt, FeMoO4, or CaO2 decompose endogenous H2O2 into O2 during irradiation (Jiang et al., 2025; Li et al., 2026; Meng et al., 2025; N. Yang et al., 2024). GSH is drained in parallel by redox-active metal ions (Mn4+, Fe3+, Cu2+) that oxidize it to GSSG while being reduced to lower valence states; the loss of GSH inactivates glutathione peroxidase 4 (GPX4), weakening antioxidant defense and amplifying lipid peroxidation (Jiang et al., 2025; Li et al., 2026; Meng et al., 2025; Zhao & Zhao, 2026). These microenvironmental maneuvers are catalogued alongside their acoustic parameters in Table 2.

2.3. Synergistic Multimodal Combination Strategies

Given the molecular heterogeneity of HCC, SDT alone rarely eradicates a tumor, and most contemporary designs are explicitly combinatorial (N. Yang et al., 2024; Zhang et al., 2026; Zhao & Zhao, 2026). The representative regimens, platforms, and reported efficacies are summarized in Table 3.

The most common partnership is with chemodynamic therapy (CDT) (Jiang et al., 2025; Zhao & Zhao, 2026). CDT uses transition-metal ions (Fe2+, Cu+, Mn2+) to run Fenton or Fenton-like reactions, converting H2O2 into •OH without any external light or oxygen (Jiang et al., 2025; Zhang et al., 2026). Housed in one nanostructure—Fe-doped nanosheets such as TiZrRu-MON, or bimetallic FeMoO4—the two modalities reinforce each other: acoustic microstreaming lowers the activation barrier for Fenton cycling, SDT-generated ROS feed CDT substrates, and CDT-driven GSH depletion protects SDT-generated ROS from scavenging (Jiang et al., 2025; Li et al., 2026; Meng et al., 2025). Gas therapy adds a third axis; activatable SO2 nanosonosensitizers (Aza-DNBS NPs), for instance, release SO2 only on dual GSH-and-ultrasound activation, damaging mitochondria, downregulating immunosuppressive genes (Timd4, Foxp1), and upregulating MHC-I/II in a trimodal sono-gaseous-checkpoint scheme (Liang et al., 2025). Across these systems, SDT increasingly drives non-apoptotic death—ferroptosis, pyroptosis, and cuproptosis—thereby bypassing the apoptotic resistance typical of aggressive HCC (Chen et al., 2025; Li et al., 2026; Meng et al., 2025).

The combination with greatest translational weight, however, is immunological. As a trigger of ICD, ultrasound-activated SDT induces severe endoplasmic reticulum stress and membrane rupture, prompting dying cells to expose or release calreticulin, HMGB1, ATP, and heat-shock proteins (Du et al., 2025; Wei et al., 2024; N. Yang et al., 2024). These DAMPs, together with liberated tumor-associated antigens, engage dendritic cells, which mature and present antigen in the draining lymph nodes, priming CD3+ and CD8+ cytotoxic T lymphocytes (Du et al., 2025; Wei et al., 2024). The net effect—summarized in Figure 3—is conversion of a cold TME, rich in regulatory T cells and myeloid-derived suppressor cells, into a hot, infiltrated one (Du et al., 2025; Wei et al., 2024; Zhao & Zhao, 2026). Pairing this with immune checkpoint blockade, typically anti-PD-L1, blocks the PD-L1/PD-1 axis and preserves CTL function; preclinical orthotopic and bilateral HCC models (e.g., IR780@FOM-cRGD, or RCM-Lip co-loading Ce6 with the STING agonist MSA-2) report elevated IL-6, TNF-α, and IFN-γ, fewer FoxP3+ Tregs, primary tumor suppression, and abscopal control of

Table 1. Classification, architectural design, and physicochemical properties of sonosensitizers used in sonodynamic cancer theranostics. The table groups agents into four generations—first/second-generation organic porphyrins, cyanine dyes and aggregation-induced-emission luminogens, inorganic semiconductors and metal-oxide particles, and metal–organic-framework hybrids—and contrasts their nanocarrier formats, acoustochemical mechanisms of reactive-oxygen-species generation, and the trade-offs between biocompatibility and clearance that govern their suitability for deep-tissue hepatocellular carcinoma therapy (Du et al., 2025; Jeong et al., 2024; Jiang et al., 2025; Zhao & Zhao, 2026).

Table 2. Acoustic exposure parameters, cavitation bioeffects, and tumor-microenvironment modulation mechanisms in deep-tissue sonodynamic therapy. For each mechanism—inertial cavitation with sonoluminescence, stable cavitation with sonoporation, catalytic hypoxia relief, glutathione depletion, and vasodilator-assisted delivery—the table lists representative ultrasound frequency, intensity, and duty-cycle windows alongside the resulting physical and chemical effects, the molecular pathways engaged, and the therapeutic consequence, highlighting how acoustic dosimetry is tuned to a specific biological target (Jeong et al., 2024; Li et al., 2026; Peng et al., 2025; Shao et al., 2025; N. Yang et al., 2024).

untreated metastases (Li et al., 2026; Wei et al., 2024). These immunological outcomes are tabulated in Table 3.

2.4. Multimodal Theranostic Guidance and Clinical Translation Horizons

A defining advantage of acoustic nanomedicine is that one platform can both see and treat, replacing sequential workflows with image-guided procedures that localize the tumor, confirm nanoparticle arrival, control acoustic activation, and assess response in real time (Uinarni et al., 2026; L. Wang et al., 2026). The complementary modalities and their preclinical outcomes are summarized in Table 4 and the guidance loop is depicted in Figure 4. Ultrasound and contrast-enhanced ultrasound (CEUS) exploit microbubbles and phase-changing fluorocarbon droplets as echo contrast, while passive cavitation detection quantifies sub- and superharmonic emissions to keep energy on target (Liu et al., 2026; Uinarni et al., 2026; L. Wang et al., 2026). Photoacoustic imaging (PAI) pairs optical absorption contrast—from melanin, polydopamine, or AIE polymers—with acoustic resolution to map microvasculature, oxygen saturation, and drug accumulation before activation (Cheng et al., 2026; J. Huang et al., 2018). Magnetic resonance imaging, enabled by paramagnetic Fe3+, Mn2+, or Gd3+, resolves deep hepatic anatomy, and sonoafterglow converts acoustically generated singlet oxygen into delayed NIR-II signals for background-free margin delineation (Chen et al., 2025; Cheng et al., 2026; Jiang et al., 2025).

A concrete illustration is the phthalocyanine-conjugated mesoporous silicate nanoparticle (PAMSN), whose allylhydrazine moieties react with ROS under ultrasound to generate nitrogen microbubbles; these produce CEUS signal and passive cavitation emissions that mark the exact zone of ROS generation while confirming renal clearance (M. Wang et al., 2025; Zhang et al., 2026). Yet for all this sophistication, the path to the clinic is guarded by familiar obstacles (Uinarni et al., 2026; Zhao & Zhao, 2026). Acoustic dosimetry remains unstandardized—center frequency, intensity, duty cycle, and exposure time vary widely between studies, producing inconsistent ROS yields (Li et al., 2026; Uinarni et al., 2026; N. Yang et al., 2024). Biosafety, metabolic fate, and GMP-compliant manufacturing of multi-component constructs are incompletely characterized (Uinarni et al., 2026; L. Wang et al., 2026; Zhao & Zhao, 2026). And most damningly, the preclinical evidence rests heavily on subcutaneous models in immunocompromised mice, which cannot reproduce cirrhosis, fibrotic stroma, altered hepatic hemodynamics, or an intact immune system; immunocompetent orthotopic models, patient-derived xenografts, and immuno-organoids are needed to predict clinical behavior (Bai et al., 2026; Uinarni et al., 2026; Zhao & Zhao, 2026).

3. Methods

This work was conducted and is reported as a structured narrative review of preclinical literature on ultrasound-activated sonodynamic nanoparticles for hepatocellular carcinoma. To make the process transparent and reproducible, the search strategy, eligibility criteria, screening workflow, and data-extraction scheme are described below in a manner consistent with the reporting conventions of indexed biomedical reviews (Bai et al., 2026; Uinarni et al., 2026).

3.1. Information Sources and Search Strategy

Four electronic databases were interrogated: PubMed/MEDLINE, Scopus, Web of Science Core Collection, and Embase. The search covered records published between January 2018 and June 2026, a window chosen to capture the emergence of nanostructured sonosensitizers while retaining foundational theranostic work (J. Huang et al., 2018). The search combined controlled vocabulary and free-text terms across three concept blocks using Boolean operators: (i) disease—“hepatocellular carcinoma” OR “liver cancer” OR “HCC”; (ii) modality—“sonodynamic therapy” OR “sonosensitizer” OR “ultrasound-activated” OR “acoustic cavitation” OR “sono-immunotherapy”; and (iii) platform—“nanoparticle” OR “nanoplatform” OR “nanomedicine” OR “metal–organic framework” OR “nanobubble.” The three blocks were intersected with AND. Reference lists of retrieved reviews were hand-searched to recover records missed by the database queries (Uinarni et al., 2026; Zhao & Zhao, 2026).

3.2. Eligibility Criteria

Records were eligible for inclusion if they satisfied all of the following:

• reported original preclinical (in vitro, in vivo, or ex vivo) or mechanistic data on sonodynamic therapy using a nanostructured sonosensitizer, or were authoritative reviews defining mechanism or translational status (Jeong et al., 2024; N. Yang et al., 2024);

• addressed hepatocellular carcinoma directly, or reported mechanisms (cavitation physics, ROS chemistry, immunogenic cell death, theranostic imaging) generalizable to deep-tissue liver application (Du et al., 2025; Uinarni et al., 2026);

• specified the acoustic exposure conditions and/or the nanoplatform composition in sufficient detail to permit mechanistic interpretation (Shao et al., 2025);

• were peer-reviewed and published in English.

Records were excluded if they concerned diagnostic ultrasound without a therapeutic sonosensitizer, addressed photodynamic therapy without an acoustic component, were conference abstracts without full data, or were non-retrievable in full text.

3.3. Study Selection and Data Extraction

Titles and abstracts were screened against the eligibility criteria, and candidate records were then assessed in full text. From each retained study, the following fields were extracted into a standardized matrix: sonosensitizer class and chemical identity; nanocarrier architecture and targeting ligand; acoustic parameters (center frequency in MHz, intensity in W/cm², duty cycle, and exposure duration); tumor-microenvironment modulation strategy (oxygenation, GSH depletion); combination modality (CDT, gas therapy, checkpoint blockade); imaging modality; experimental model (cell line, subcutaneous versus orthotopic, immune status); and quantitative antitumor and biosafety outcomes (Peng et al., 2025; Wei et al., 2024; Zhang et al., 2026). Extracted acoustic and microenvironmental parameters were organized by mechanism to allow cross-study comparison despite heterogeneous reporting (Li et al., 2026; Uinarni et al., 2026).

3.4. Synthesis Approach

Because the included studies differed widely in nanomaterial, model system, and acoustic dosimetry, a quantitative meta-analysis was neither feasible nor appropriate; evidence was therefore synthesized narratively and thematically (Uinarni et al., 2026). Findings were grouped into four analytic domains—sonodynamic cytotoxicity and ROS generation, tumor-microenvironment remodeling, synergistic multimodal combinations, and theranostic imaging with biosafety—mirroring the organization of the literature and the summary tables. Where multiple studies reported a comparable endpoint (for example, fold-change in tumor accumulation or percentage tumor growth inhibition), values are reported as stated in the primary source without recalculation, and apparent inconsistencies in acoustic dosimetry are flagged rather than reconciled (Shao et al., 2025; Zhao & Zhao, 2026).

4. Performance of Acoustic Nanoplatforms Across the Sonodynamic Cascade

4.1. Sonodynamic Cytotoxicity and ROS Generation In Vitro and In Vivo

Read across the preclinical datasets, one pattern recurs almost without exception: putting a sonosensitizer inside an engineered nanoparticle raises ultrasound-driven ROS yields and tumor cytotoxicity above what the free molecule can manage (Du et al., 2025; Peng et al., 2025; N. Yang et al., 2024). The reason is not mysterious. Free PpIX, hematoporphyrin, and Ce6 are hampered by poor solubility, fast clearance, and the aggregation-caused quenching that suppresses both 1O2 and •OH generation (Du et al., 2025; Zhao & Zhao, 2026). Nanoscale confinement either separates the chromophores spatially or restricts their rotation, and in doing so it rescues the energy-transfer pathways that cavitation is meant to drive (Du et al., 2025; Shao et al., 2025). The sensitizer classes behind these gains, and their characteristic acoustochemical features, are summarized in (Table 1), while the physical chain from cavitation to ROS is depicted in (Figure 1).

The numbers bear this out, if unevenly. In HepG2 and Hepa1-6 cells, IR820 nanobubbles under low-intensity ultrasound produced intracellular ROS fluorescence nearly double that of unencapsulated IR820, an effect attributed to microbubble cavitation and sonoluminescence-driven energy transfer (H. Yang et al., 2019). The same construct was essentially inert in the dark—viability stayed above 90%—yet ultrasound drove HepG2 survival down to about 25% (H. Yang et al., 2019). Pairing it with the dual PI3K/mTOR inhibitor PI-103 lowered the combination index to 0.18, which is to say the synergy was real rather than merely additive, with mitochondrial depolarization and impaired migration alongside (H. Yang et al., 2019). A structurally very different agent, the narrow-bandgap A–D–A′–D–A organic nanosonosensitizer BTz, likewise degraded DPBF efficiently and lit up DCFH-DA on acoustic exposure (Peng et al., 2025). What makes the BTz work notable is the delivery trick layered on top: oral ISMN at 5 mg/kg transiently widened tumor microvessels and raised nanoparticle accumulation about 1.8-fold, and the combined regimen reached 78% tumor growth inhibition—close to twice what SDT managed alone—with extensive coagulative necrosis on histology (Peng et al., 2025).

4.2. Remodeling the Tumor Microenvironment: Hypoxia Relief and Glutathione Depletion

If there is a single recurring obstacle in these studies, it is the microenvironment rather than the chemistry (Meng et al., 2025; Zhao & Zhao, 2026). Type II energy transfer needs ground-state oxygen to make 1O2, so hypoxia caps the achievable ROS, and the elevated GSH of malignant hepatocytes scavenges much of what is made before it can act (Jiang et al., 2025; Shao et al., 2025; Zhang et al., 2026). The platforms that perform best are, almost by definition, the ones that attack both problems at once—a logic set out schematically in (Figure 2) and parameterized in (Table 2).

Catalytic self-oxygenation is the more mature of the two strategies. MnO2-functionalized hyaluronic acid nanoparticles, directed at CD44-overexpressing HepG2 cells, decomposed endogenous H2O2 into O2 while oxidizing GSH to GSSG; oxygen-saturation measurements confirmed that in situ oxygen generation reversed hypoxia-driven resistance, raised intracellular ROS, and triggered calreticulin exposure (Du et al., 2025). Bimetallic and MOF nanosheets—TiZrRu-MON and GMTF Mn-TCPP among them—lean on multivalent metal cycling (Mn4+/Mn2+, Fe3+/Fe2+) to drain GSH, and the consequent inactivation of GPX4 sensitizes cells to lipid peroxidation, ferroptosis, and cuproptosis (Chen et al., 2025; Jiang et al., 2025; Li et al., 2026; Zhao & Zhao, 2026). It is worth noting that the mechanical side of SDT contributes here too: stable-cavitation microstreaming opens transient 50–200 nm membrane pores that improve extravasation and uptake, so the delivery and redox effects are not cleanly separable (Liu et al., 2026; L. Wang et al., 2026).

4.3. Synergistic Outcomes: Sono-Immunotherapy, Chemodynamic, and Gas Cascades

Across the in vivo work, the strongest responses come not from SDT in isolation but from SDT wired into a secondary cascade (Du et al., 2025; Wei et al., 2024). The initiating event is consistent—ultrasound-driven ROS and ER stress provoke immunogenic cell death, with surface calreticulin, HMGB1 secretion, and ATP efflux (Du et al., 2025; Wei et al., 2024; Zhao & Zhao, 2026). What happens next is captured in (Figure 3): released DAMPs recruit and mature dendritic cells, which present tumor antigens to naive T cells and seed a systemic cytotoxic response (Du et al., 2025; Wei et al., 2024).

The checkpoint combinations are where the effect sizes become striking. In mice bearing subcutaneous tumors treated with targeted micellar IR780@FOM-cRGD under ultrasound, adding anti-PD-L1 raised serum IL-6, TNF-α, and IFN-γ while suppressing IL-10 and TGF-β (Wei et al., 2024). Histology told a coherent story—elevated γ-H2AX, high TUNEL positivity, reduced Ki-67, dense CD3+ and CD8+ infiltration, and a marked fall in FoxP3+ Tregs—and, importantly, the regimen produced abscopal control of un-irradiated distant lesions rather than merely ablating the treated one (Du et al., 2025; Wei et al., 2024; Zhao & Zhao, 2026). The chemodynamic and gas combinations add further mechanisms, as compiled in (Table 3): NO-releasing PIH-NO platforms generate peroxynitrite, repolarize M2 macrophages toward M1, and reach up to 92% tumor inhibition (Meng et al., 2025), while activatable SO2 nanosonosensitizers (Aza-DNBS NPs) release gas only on dual GSH-and-ultrasound activation, downregulate Timd4 and Foxp1, upregulate MHC-I/II, and do so with a hemolysis rate below 5% (Liang et al., 2025; Zhao & Zhao, 2026). The consistency of the immunological readouts across chemically unrelated platforms is, arguably, the most reassuring signal in the dataset.

4.4. Theranostic Imaging, Cavitation Detection, and Biosafety

Precise targeting is only useful if it can be verified, and the better platforms build verification in (Uinarni et al., 2026; Zhang et al., 2026). CEUS and passive cavitation detection, using gas-filled nanobubbles or phase-changing fluorocarbon droplets, outline tumor microvasculature and track where acoustic energy is actually deposited (Uinarni et al., 2026; Zhang et al., 2026). The PAMSN system makes the point concretely: reacting with ROS under ultrasound, it generates N2 microbubbles that raise CEUS signal up to 2.5-fold and produce 5–10 dB higher harmonic emissions confined to the orthotopic liver tumor zone (Uinarni et al., 2026; M. Wang et al., 2025; Zhang et al., 2026). Photoacoustic imaging, demonstrated with folate-functionalized HMME melanin nanoparticles (FHMP NPs), maps deep microvascular sO2 and tracks accumulation in real time

Figure 3. The sono-immunotherapy cascade converting a cold hepatocellular-carcinoma microenvironment into a hot, checkpoint-responsive one. Ultrasound-activated sonodynamic therapy, alone or combined with chemodynamic or gas therapy, induces immunogenic cell death with release of calreticulin, HMGB1, ATP, and tumor-associated antigens; these mature dendritic cells and prime CD8+ cytotoxic T lymphocytes. The response remodels the microenvironment by depleting regulatory T cells and repolarizing macrophages, and synergizes with immune checkpoint blockade to achieve primary tumor regression, abscopal control of distant lesions, and immune memory (Du et al., 2025; Liang et al., 2025; Wei et al., 2024; Zhao & Zhao, 2026).

Figure 4. Multimodal theranostic guidance workflow for image-guided sonodynamic therapy of deep-seated hepatocellular carcinoma. A single multifunctional acoustic nanoplatform supplies contrast for ultrasound/contrast-enhanced ultrasound with passive cavitation detection, photoacoustic imaging, and magnetic resonance imaging with sonoafterglow. These feed an image-guided treatment loop that localizes the tumor before treatment, provides real-time cavitation feedback during activation, and evaluates response afterward, enabling dose adjustment or re-treatment and ultimately precise, monitorable tumor ablation (Cheng et al., 2026; Uinarni et al., 2026; M. Wang et al., 2025; Zhang et al., 2026).

(J. Huang et al., 2018; Li et al., 2026), and paramagnetic Fe3+/Mn2+ ions supply the T1/T2 MRI contrast that confirms pH/GSH-triggered disassembly at depth (Jiang et al., 2025; Zhao & Zhao, 2026). These modalities and their reported efficacies are consolidated in (Table 4) and the overall guidance loop in (Figure 4).

On safety, the preclinical record is encouraging though, one should concede, early (Peng et al., 2025; Wei et al., 2024; Zhao & Zhao, 2026). Treated cohorts across several studies maintained steady body weight and normal core temperature through the monitoring window (Wei et al., 2024; Zhao & Zhao, 2026). Serum chemistry—ALT, AST, BUN, and creatinine—stayed within baseline ranges, and H&E staining of heart, liver, spleen, lung, and kidney showed no acute inflammation, hemorrhage, or necrosis (Peng et al., 2025; Wei et al., 2024; Zhao & Zhao, 2026). These are the right endpoints, but they are short-horizon ones in small, mostly immunocompromised animals, and that caveat carries directly into the discussion that follows.

5. Interpreting the Promise and the Gaps of Sonodynamic Nanomedicine for HCC

5.1. What the Evidence Establishes

Taken as a whole, the literature makes a reasonably firm case for three claims. First, nanostructuring genuinely rescues the sonosensitizer problem that stalled SDT for years—the solubility, clearance, and quenching liabilities of free porphyrins are real, and encapsulation or self-assembly measurably reverses them (Du et al., 2025; Zhao & Zhao, 2026). Second, the depth argument holds: ultrasound reaches the liver in a way light cannot, and the cavitation machinery that converts an acoustic pulse into ROS and sonoporation is well enough understood to be engineered deliberately rather than stumbled upon (Jeong et al., 2024; Uinarni et al., 2026; N. Yang et al., 2024) (Figure 1). Third, and most consequentially, SDT is a dependable trigger of immunogenic cell death, and the immunological readouts—DAMP release, DC maturation, CTL infiltration, Treg suppression—recur across chemically unrelated platforms, which is exactly the kind of convergence one hopes to see before trusting a mechanism (Du et al., 2025; Wei et al., 2024) (Table 3; Figure 3).

5.2. The Tumor Microenvironment as the Central Design Problem

If the field has a unifying insight, it is that the hepatic TME must be treated as an adversary to be engineered against, not a passive backdrop (Meng et al., 2025; Zhao & Zhao, 2026). The hypoxia-and-GSH double bind explains why first-generation SDT underperformed in solid liver tumors, and the platforms that now lead the field are precisely those that oxygenate and deplete GSH in the same construct (Jiang et al., 2025; Li et al., 2026) (Table 2; Figure 2). That said, one should be careful not to over-read the mouse numbers. Fold-changes in tumor accumulation and the headline tumor-inhibition percentages—the 1.8-fold uptake and 78% inhibition with ISMN, the 92% with NO platforms—are impressive, but they come from heterogeneous models with incompatible dosimetry, and it is not obvious how they would translate to a cirrhotic human liver with altered perfusion (Peng et al., 2025; Meng et al., 2025; Shao et al., 2025). The consistency of direction is more convincing than the magnitude of any single figure.

5.3. Translational Bottlenecks

Three obstacles stand out, and none is trivial. The first is dosimetry. Acoustic parameters—frequency, intensity, duty cycle, exposure time—vary so widely between studies that cross-comparison is often guesswork, and without a calibrated standard it is hard to separate a better nanoparticle from a more aggressive ultrasound protocol (Li et al., 2026; Uinarni et al., 2026; N. Yang et al., 2024). The second is biosafety over a realistic horizon. The acute panels look clean, but long-term hepatobiliary clearance, reticuloendothelial sequestration, degradation kinetics, and immunogenicity of multi-component constructs remain thinly characterized, and GMP-compliant, reproducible manufacturing of such constructs is its own unsolved problem (Uinarni et al., 2026; L. Wang et al., 2026; Zhao & Zhao, 2026). The third, and perhaps the most limiting, is model fidelity. Much of the evidence rests on subcutaneous tumors in immunocompromised mice—a model that, almost by construction, cannot capture cirrhosis, fibrotic stroma, altered hepatic hemodynamics, or an intact immune system, which is awkward for a modality whose chief selling point is immune activation (Uinarni et al., 2026; Zhao & Zhao, 2026).

5.4. Future Directions

The corrective steps follow fairly directly from the gaps. Immunocompetent orthotopic liver models, patient-derived xenografts, and patient-derived immuno-organoids would test the immunological claims in settings that resemble the

Table 3. Synergistic multimodal combination strategies built around sonodynamic therapy. The table pairs each combination—SDT with chemodynamic therapy, with gaseous transmitters (NO, CO, SO2), with immune checkpoint blockade, and with regulated cell-death inducers (ferroptosis, cuproptosis, disulfidptosis)—with a representative engineered nanoplatform, its cascade mechanism, the resulting immune and microenvironmental reprogramming, and the quantitative preclinical efficacy reported, illustrating how combinatorial design converts local sonodynamic injury into systemic antitumor responses (Chen et al., 2025; Du et al., 2025; Jiang et al., 2025; Liang et al., 2025; Meng et al., 2025; Wei et al., 2024; Zhao & Zhao, 2026).

Table 4. Multimodal theranostic imaging guidance and preclinical outcomes in deep-tissue cancer models. The table summarizes four complementary modalities—ultrasound and contrast-enhanced ultrasound with passive cavitation detection, photoacoustic imaging, magnetic resonance imaging, and sonoafterglow luminescence—specifying for each a representative theranostic nanoplatform, the contrast-generating mechanism, the real-time diagnostic guidance it provides, and the tumor model and antitumor efficacy reported, thereby mapping how imaging is embedded within the sonodynamic treatment workflow (Cheng et al., 2026; J. Huang et al., 2018; Jiang et al., 2025; Uinarni et al., 2026; M. Wang et al., 2025; Zhang et al., 2026; Zhao & Zhao, 2026).

disease, and organoid platforms in particular offer a tractable bridge between mouse and patient (Bai et al., 2026; Zhao & Zhao, 2026). Standardized acoustic dosimetry—reported as fully as drug dose now is—would let the field compare platforms rather than protocols (Shao et al., 2025; Uinarni et al., 2026). And the theranostic designs that already report cavitation zones and ROS generation in real time (PAMSN being the clearest example) point toward a plausible clinical workflow in which treatment is titrated against live imaging feedback rather than delivered blind (M. Wang et al., 2025; Zhang et al., 2026) (Table 4; Figure 4). Whether sonodynamic nanomedicine becomes a clinical reality for HCC will depend less on new chemistry—there is already plenty—and more on the unglamorous work of standardization, long-term safety, and honest models.

6. Conclusion

Ultrasound-activated sonodynamic nanoparticles offer a genuinely distinctive answer to the central difficulty of hepatocellular carcinoma—reaching a deep, defended tumor non-invasively and precisely. The evidence reviewed here shows that nanostructuring overcomes the chemical liabilities of early sonosensitizers, that cavitation-driven ROS generation can be engineered rather than merely observed, and that microenvironment-remodeling platforms which relieve hypoxia and deplete glutathione substantially amplify therapeutic effect (Jiang et al., 2025; Zhao & Zhao, 2026). Most importantly, SDT reliably provokes immunogenic cell death, converting cold liver tumors into checkpoint-responsive ones and, in several models, controlling distant metastases (Du et al., 2025; Wei et al., 2024). What remains is not primarily a problem of invention but of consolidation: standardized acoustic dosimetry, long-term biosafety in large animals, GMP-scalable manufacturing, and immunocompetent, clinically faithful models (Uinarni et al., 2026). Addressing these will determine whether a compelling preclinical story becomes a practical, image-guided option for liver cancer patients.

References