Theranostic potential of ultrasound-guided sonodynamic therapy for cancer: a systematic review of preclinical studies

Abstract

Ultrasound-guided sonodynamic therapy (US-guided SDT) is an emerging theranostic approach that combines noninvasive tumor treatment with real-time imaging. By activating sonosensitizers through focused US to produce reactive oxygen species (ROS), SDT enables targeted tumor ablation with minimal systemic toxicity. This systematic review summarizes preclinical advancements in US-guided SDT from 2015 to 2025. Comprehensive searches across PubMed, Scopus, Web of Science, ScienceDirect, and Google Scholar identified 24 eligible studies. Extracted data included nanoplatform characteristics, parameters, imaging modalities, therapeutic outcomes, and translational potential. Across studies, US served both as an imaging and activation modality, enabling simultaneous diagnosis and therapy. Nanoparticle-based sonosensitizers enhanced ROS yield, tumor selectivity, and imaging contrast. Most studies reported significant tumor inhibition (> 70%), precise targeting, and favorable biosafety. Multifunctional nanoplatforms integrating oxygen generation, immunomodulation, or chemotherapeutic co-delivery further improved efficacy. US-guided SDT exhibits robust preclinical evidence as a dual-function imaging and treatment platform. However, lack of standardized US parameters, limited long-term safety data, and reliance on subcutaneous tumor models hinder clinical translation. Future studies should emphasize standardized dosimetry, advanced tumor models, and comprehensive biosafety evaluation to advance this promising theranostic modality toward clinical application.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s12672-026-04638-5.

Introduction

Although surgery, chemotherapy, radiotherapy and immunotherapy have achieved a series of breakthroughs, the problems of low treatment efficiency in patients with advanced or drug-resistant tumors, high recurrence rate and serious systemic toxicity of cancer remain a major challenge in clinical practice [1, 2]. The need for developing new therapeutic approaches that are highly effective, minimally invasive, with tumor selectivity and real-time monitoring is becoming an urgent task. Sonodynamic therapy (SDT), as an emerging and promising non-invasive therapeutic strategy for cancer, has received more and more attention [3, 4].

SDT employs low-intensity ultrasound (US) to trigger sonosensitizers that generate reactive oxygen species (ROS) which destroy tumor cells [5]. SDT represents a distinct approach among cancer treatments as it selectively targets tumor cells without harming adjacent healthy tissues [6]. Sonosensitizers can be activated locally as US waves can be spatially focused within a region of interest, e.g. tumor or infected tissues [7, 8]. Furthermore, the energy from US is able to penetrate deeply into biological tissue, in contrast to light in photodynamic therapy (PDT), which is only able to penetrate to superficial tissue [9, 10]. This enables SDT to non-invasively activate sonosensitizers at a deep focus point, and thus is a promising platform for treatment of tumors in areas that are hard to reach with other methods.

US serves as an established medical imaging technique used in clinical settings because of its real-time imaging capabilities and noninvasive nature alongside its affordability and wide accessibility. SDT transforms US into both a therapeutic activator and an imaging method which enables a single system that delivers diagnosis and treatment monitoring [11]. This theranostic ability is also particularly attractive, as it enables clinicians to image tumor location, guide delivery of therapy, and then actively monitor for treatment response, ultimately improving precision and reducing off-target delivery [12]. This unification of therapeutic activation and real-time imaging thus has the potential to make US-guided SDT a paradigm shifting technique.

Preclinical research has played a pivotal role in advancing the field of US-guided SDT [13]. Over the past decade, significant progress has been made in the design of novel nanoparticle (NP)-based sonosensitizers with improved tumor selectivity, stability, and ROS yield [14]. Many of these nanosystems are engineered to respond to US stimuli while simultaneously carrying imaging contrast agents, thereby reinforcing the theranostic concept [15]. Examples include liposomes, polymeric micelles, inorganic nanostructures, and hybrid platforms designed to enhance circulation time, tumor accumulation, and controlled activation. In addition, targeted delivery strategies such as ligand modification and biomimetic coatings have further improved the specificity and safety of SDT in preclinical models [16].

Mechanistically, the anti-cancer effects of SDT are not limited to direct ROS-mediated cytotoxicity. For example, recent work has shown that SDT can elicit immunogenic cell death (ICD), trigger anti-tumor immune responses, and even remodel the tumor microenvironment [17, 18]. These properties suggest that the scope of SDT is broad and can be considered for combination regimens with immunotherapy, chemotherapy, or radiotherapy. Synergistic with this is the capacity of US to affect vascular permeability and enhance NP accumulation.

Even with these advances, clinical translation of US-guided SDT is still in its infancy. One of the primary challenges is the lack of standardization in protocol. Currently, there is no consensus in sonosensitizer design, US parameters, or treatment regimens. Variation in acoustic frequency, intensity, duty cycle, and exposure time between studies has led to difficulty in reproducing results and in developing universal treatment guidelines. The variety of preclinical models and endpoints for evaluation has also presented a challenge for translating preclinical findings to human cancers. Additionally, the long-term safety and off-target effects of some of these nanoconstructs still need to be evaluated before they can be clinically translated. Given these opportunities and challenges, it has become timely to perform a comprehensive assessment of recent preclinical studies that will summarize current information and demonstrate research progress while identifying remaining gaps. While prior reviews have highlighted the general concepts of SDT or evaluated the therapeutic efficacy of individual probes, the theranostic role of US guidance in preclinical models was not fully discussed. By centering on US-guided SDT, this review aims to offer a broad perspective of US employed as both an imaging and therapeutic modality in preclinical cancer research.

Unlike currently existing reviews on SDT, or US-based theranostics, this review put special emphasis on US-guided SDT as an integrated imaging-therapy paradigm. This study encompasses a more detailed system of interpreting preclinical efficacy of SDT and translational relevance by collaboratively scrutinizing imaging guidance, therapeutic results, and methodological quality. To the best of our knowledge and based on existing research in English publications no comprehensive systematic review paper explored US-guided SDT in cancer treatment. Therefore, the following sections of this article will present the methodology, findings, and implications of this systematic review, with a focus on the emerging theranostic paradigm of US-guided SDT.

Methods

Literature search

A comprehensive literature search was conducted in PubMed, Scopus, Web of Science, ScienceDirect, and Google Scholar to identify relevant preclinical studies published between January 2015 and September 2025. Only articles published in English were considered eligible for inclusion. The search strategy combined Medical Subject Headings (MeSH) and free-text keywords, including: “sonodynamic therapy,” “ultrasound-triggered therapy,” “ultrasound activation,” “sonosensitizer,” “nanoparticles,” “reactive oxygen species,” “cancer treatment,” “image-guided therapy,” and “theranostics.” Boolean operators (AND, OR) were applied to refine the query.

The search was designed according to the PICO framework: preclinical cancer models (P), interventions using US-guided sonodynamic therapy (I), compared with control groups (e.g., untreated, US only, or sensitizer only) ©, and outcomes related to therapeutic efficacy, imaging guidance, and biosafety (O). The full PICO framework is presented in Table 1. Reference lists of included studies and relevant reviews were manually screened to capture additional eligible publications.

Table 1.

PICO framework for search strategy

Component	Description
P (Population)	Preclinical cancer models (murine xenografts, orthotopic models, syngeneic tumors, etc.)
I (Intervention)	Ultrasound-guided sonodynamic therapy (SDT) using chemical or nanoparticle-based sonosensitizers
C (Comparison)	Control groups (untreated, ultrasound only, sensitizer only, or other conventional interventions)
O (Outcomes)	ROS generation, tumor suppression, therapeutic efficacy, survival outcomes, imaging guidance, systemic toxicity, and translational relevanc

The following search string was applied across all databases:

(“sonodynamic therapy” OR sonodynamic OR SDT OR “ultrasound-activated therapy” OR “ultrasound-induced therapy” OR “Sonotherapy”)

AND

(ultrasound OR ultrasonography OR “ultrasound imaging” OR “image-guided” OR “ultrasound-guided” OR “US-guided”)

AND

(cancer OR tumor OR tumour OR neoplasm OR malignancy OR carcinoma OR sarcoma)

Date of last search: September10, 2025.

Inclusion and exclusion criteria

Eligible studies met the following criteria:

• Study type: Original in vivo preclinical studies investigating US-guided SDT in cancer models.

• Intervention: Use of US to activate sonosensitizers, with or without NP-based delivery platforms.

• Outcomes: Must report at least one of the following—tumor growth inhibition, ROS measurement, imaging guidance, systemic toxicity, survival benefit, or mechanistic insights.

• Publication criteria: Articles published in English between January 2015 and September 2025.

• We excluded reviews, case reports, editorials, conference abstracts, in vitro-only studies, purely theoretical works, and studies without US-guided intervention.

Study selection

All identified records were exported to EndNote (Version X9, Clarivate Analytics) for reference management and removal of duplicates. Two reviewers independently screened the title and abstract of the retrieved records against the eligibility criteria. Potentially eligible records underwent full-text assessment for inclusion. Any disagreement was resolved by consensus between the reviewers. If consensus could not be reached, a third reviewer was consulted.

Data extraction

A standard data extraction form was created and piloted. Data extraction was performed by two reviewers independently, with disagreements resolved by consensus or adjudication by a third reviewer. The extracted data were summarized in two tables:

• Study Characteristics Table: This table included the following variables for each study: author, year, country, cell line / animal model, sample size (treated / control), cancer type / model, sonosensitizer / nanoplatform, dosage / concentration, US parameters.

• Comparative Imaging and Therapeutic Efficacy Table: This table summarized theranostic performance focusing on sonosensitizer / nanoplatform, imaging findings, therapeutic outcomes, advantages, and limitations.

Risk of bias assessment

The methodological quality and risk of bias (ROB) of the included studies were independently assessed by two reviewers using Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE)’s ROB tool for this study [19]. Several domains relevant to preclinical research were examined for all included. For each domain, judgments were categorized as low risk, high risk, or unclear risk. A designation of unclear risk was applied when relevant methodological details were not reported or insufficiently described, rather than indicating confirmed bias. The assessments were done independently by two reviewers. Any disagreements were resolved by discussion.

Results

Study selection

Searching yielded a total of 1310 records in international databases (PubMed, Scopus, Web of Science, ScienceDirect, and Google Scholar). Following the removal of duplicates, 849 articles remained for title screening. Titles and abstracts were screened by two independent investigators for inclusion and exclusion criteria, and 31 studies were chosen for full-text review. After a full-text evaluation, 7 articles were excluded based on irrelevant interventions (n = 2), irrelevant outcomes (n = 3), or unavailable full text (n = 2). In total, 24 preclinical studies met all the eligibility criteria and were included in this systematic review. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 flowchart of the selection process is shown in Fig. 1. Detailed reasons for full-text exclusions are provided in Supplementary Table S1, and adherence to PRISMA 2020 reporting guidelines is documented in Supplementary Table S2.

Fig. 1.

Flowchart of eligible studies included in this systematic review

Study characteristics

In Table 2, a summary of the study characteristics is provided [20–43].

Table 2.

Summary of study characteristics

Author, Year	Country	Cell line / animal model	In vivo sample Size	Cancer type/model	Sonosensitizer/nanoplatform	Dosage/concentration	Ultrasound parameters
M. Chen et al. 2018 [20]	China	HT-29 (human colorectal cancer) xenograft in Balb/c nude mice	NA	Colorectal cancer	Power cavitation factor- microbubbles (PCF-MBs) (porphyrin/camptothecin-floxuridine triad microbubbles)	5 µM (in vitro); 1 mg mL− 1 (in vivo)	Imaging: 7.0 MHz (B-mode/harmonic); Therapeutic (SDT): 1.0 MHz, 1 W cm− 2, 50% duty cycle, 3 min
S. Chen et al. 2018 [21]	China	SKOV3 / Balb-c nude mice	8 per group (5 groups)	Ovarian cancer	PTX/ICG/O2-loaded PLGA NP (PIO_NPs)	5 mg kg− 1 PTX; ICG 1.73 µg/mL (in vitro); PTX 7.28 µg/mL (in vitro)	Therapeutic (SDT): 1.0 MHz, 1 W cm− 2, 1 min; Imaging (USI/PA): B-mode + PA
Feng et al., 2018 [22]	China	MCF-7 / BALB-c nude mice	5 per group (8 groups)	Breast cancer	TPZ/HMTNPs-SNO (tirapazamine-loaded hollow mesoporous TiO2 NPs modified with S-nitrosothiol)	TPZ 4 mg kg− 1; HMTNPs-SNO 20 mg kg− 1	Therapeutic (SDT): 1 W cm− 2, 1 min; Imaging (USI): B-mode contrast, NO-induced echogenicity, 1 MHz; 3 h post-injection
Feng et al. 2018 [23]	China	MCF-7 / BALB-c nude mice	5 per group (8 groups)	Breast cancer	HMME/MCC-HA (hematoporphyrin monomethyl ether–loaded mesoporous CaCO3 NPs with hyaluronic acid)	HMME 5 mg kg− 1; MCC-HA 10 mg kg− 1	Therapeutic (SDT+cavitation): 1 MHz, 1 W cm− 2, 1 min; Imaging (US): B-mode contrast via CO2 bubbling, 3 h post-injection
Liu et al. 2018 [24]	China	MDA-MB231/BALB-c nude mice	3 per group (2 groups)	Breast cancer	FA-PEG-PLGA-Ptx@ICG-Pfh (folate-targeted PLGA NPs co-loaded with paclitaxel, indocyanine green (ICG), and perfluorohexane)	50 mg kg− 1 i.v.; NIR laser 2 W cm− 2, 5 min (808 nm)	Therapeutic: 808 nm laser for photothermal therapy (PTT) and triggered Ptx release; Imaging (US + PA): CEUS enhancement after NIR-triggered Pfh vaporization
Zhang et al. 2018 [25]	China	4T1/4T1 tumor-bearing BALB/c female mice	6 per group (7 groups)	Breast cancer	HMME/RBC-HPBs/HMME/PFH (hematoporphyrin monomethyl ether-loaded hollow Prussian blue NPs coated with red blood cell membrane and perfluorohexane)	HMME 5 mg kg− 1 i.v.	Therapeutic (SDT): 3 MHz, 5.0 W cm− 2, 30 s; Imaging (USI): Vevo 2100 small animal US imaging system
He et al. 2019 [26]	China	HepG2, 4T1/4T1 tumor-bearing BALB/c female mice	5 per group (7 groups)	Breast cancer	mTiO2@PPY-HNK (honokiol-loaded polypyrrole-coated mesoporous TiO2 NPs)	HNK 10 mg kg− 1 i.v.	Therapeutic (SDT): 1 MHz, 1.5 W cm− 2, 60 s; Imaging (USI/PA): US imaging with Mylab 90 system, PA imaging at 715 nm
Lin et al. 2019 [27]	China	MCF-7 (in vitro); MCF-7 tumor xenograft in nude mice (in vivo, subcutaneous)	5 per group (4 groups)	Breast cancer	AIPH-loaded liposome (PEGylated phospholipid/cholesterol liposome encapsulating AIPH; FITC used as tracer in some experiments)	In vivo: 200 µL of 500 µg mL− 1 Lip-AIPH IV (i.e., 100 µg per injection); In vitro: 0, 30, 60, 250, 500 µg·mL− 1	Therapeutic in vitro: 1 min or 20 s (lysosome permeability assay); in vivo imaging used 1.0 MHz, 2.5 W cm− 2; in vivo therapy: 1.0 MHz, 2.5 W cm− 2, 50% duty cycle, 10 min; Imaging device/settings for B-mode: transducer 40 MHz, power intensity 100%, gain 10 dB
Xie et al. 2019 [28]	China	ID8 (mouse ovarian cancer cell line); female C57BL/6 mice	NA	Ovarian cancer	OI_NPs, PLGA phase-transition NPs loaded with perfluoropentane (PFP) + ICG + OXP	ICG: 8.5 µg mL− 1; OXP: 6.0 µg mL− 1 (used in in-vitro assays); OI_NPs OXP loading = 1.42 ± 0.04% w/w	Therapeutic: 1.0 W·cm− 2, 1 min (low-intensity) — applied sequentially after laser for PSDT; Imaging: MyLab 90 (US)
Zhang et al. 2019 [29]	China	4T1 murine breast-cancer cells; BALB/c nude mice bearing 4T1 xenografts	5 per group (4 groups)	Breast cancer	IR780-NDs: lipid/PFP nanodroplets (DPPC/DSPEmPEG2000/cholesterol) loaded with IR780	In vitro: IR780 0–5 µg mL− 1 (CCK-8, ROS). In vivo: 200 µL of 3 mg mL− 1 IV (24 h before US)	Therapeutic: 650 kHz, 2.4 W cm− 2, 50% duty cycle, 3 min on / 3 min off × 4 cycles; in vitro tests 30–90 s at 2.4 W cm− 2; Imaging: B-mode and CEUS performed after injection
Ho et al. 2019 [30]	China	TRAMP (murine prostate cancer) cells; C57BL/6JNarl mice	3–5 per group (4 groups)	Prostate cancer	FMSNs-Dox (Superhydrophobic mesoporous silica NPs loaded with Doxorubicin and capped with β-cyclodextrin)	In vitro: Dox 10 µg mL− 1 (LD50); In vivo: 100 µL of 1 mg mL− 1 (≈ 150 µg Dox/mouse) via retro-orbital injection	Therapeutic: 2 MHz high intensity focused US (HIFU), 5–6 megapascal (MPa), 5000-cycle pulses, PRF = 18 Hz, 20–30 min; Imaging: B-mode US (7 MHz) used for intratumoral monitoring
Hou et al. 2020 [31]	China	HT-29 human colorectal adenocarcinoma cells; BALB/c nude mice bearing HT-29	4 per group (6 groups)	Colorectal cancer	RB-MBs: Rose Bengal microbubbles made from Span-60/PEG-40 S/ARB (79:19:2) encapsulating C₃F₈ gas; convertible to RB-NPs (~ 130 nm) upon US exposure	In vitro: 1 × 108 MBs mL− 1 (ARB ≈ 140 µg mL− 1); In vivo: 50 µL of 1 × 109 MBs mL− 1 IV; drug loading 6.8 ± 0.9%. MB	Therapeutic: 1 MHz, 1.5 W·cm− 2, 50% duty cycle, 900 s; Imaging: B-mode & CEUS at 5 MHz
Liu et al. 2020 [32]	China	SKOV3 human ovarian cancer (FR⁺) and A549 lung cancer (FR⁻) cells; BALB/c nude mice bearing SKOV3	Imaging: 3 mice/group; therapy: 5 mice/group)	Ovarian cancer	FA-OINPs: Folate-targeted, oxygen/ICG-loaded lipid NPs (DPPC/DPPG/DSPE-PEG(2000)-FA/CHOL, 5:2:2:1) with PFH core	In vitro: ICG 4 µg mL− 1 (for ROS and cytotoxicity); In vivo: 200 µL, 64 µg mL− 1 ICG IV (tail vein)	Therapeutic: Laser 808 nm, 1 W cm− 2, 2 min; US 300 kHz, 1 W cm− 2, 30 s (photo-sonodynamic mode); Imaging: B-mode and CEUS (5–12 MHz)
Zhao et al. 2020 [33]	China	4T1 murine mammary carcinoma cells; BALB/c mice bearing 4T1 tumors	6 per group (7 groups)	Breast cancer	LIP3 (C- Arginylglycylaspartic acid or RGD-modified lipid NPs co-loaded with HMME, PFH and AQ4N)	In vitro 1 mg mL− 1; in vivo 200 µL LIP3 suspension (5 mg kg− 1 HMME equiv)	Therapeutic: 1 MHz, 6 W cm− 2, 50% duty cycle, 3 min; imaging: CEUS; 1 MHz, 2–6 W cm− 2, 50% duty cycle, 1–4 min after LIFU exposure
Zhang et al. 2021 [34]	China	4T1 murine breast cancer cells; BALB/c mice bearing 4T1 tumors	3 per group (4 groups)	Breast cancer	CPDP NPs: PLGA core–shell NPs encapsulating chlorin e6 (Ce6, SDT), perfluoropentane (PFP, imaging), and docetaxel (DTX, chemotherapy)	In vitro: 1 mg mL− 1; in vivo: 1 mg mL− 1 IV	Therapy / Imaging: LIFU 1 MHz, 1–2 W cm− 2, 50% duty cycle, 1 s pulse duration, 120 s exposure; Imaging: 2D and CEUS performed after LIFU irradiation
Zheng et al. 2021 [35]	China	ID8 syngeneic ovarian cancer cells; female C57BL/6 mice (6–8 weeks)	8 per group (4 groups)	Ovarian cancer	OIX_NPs: Oxygen + ICG + Oxaliplatin (OXP) in PLGA/PFP	300 µL (64 µg mL− 1 ICG; 2 mg kg− 1 OXP), IV every 5 days	Theraputic: 1 MHz, 1 W cm− 2 × 1 min; exposure 4 h post-injection; Imaging: B-mode and PA imaging performed pre- and post-laser
Kang et al. 2022 [36]	China	Murine 4T1 breast cancer cells; female BALB/c mice (6–8 weeks)	n = 5 per group (treatment); n = 3 (imaging)	Breast cancer	A-DPPs: AS1411–DOX/PFH–PEG@PLGA NPs (PFH phase-change + DOX chemotherapy + Focused ultrasound (FUS)	4.0 mg mL− 1; DOX 1.5 mg kg− 1, IV via tail vein	Theraputic FUS: 1.0 MHz; 80–150 W cm− 2; duration 5 s; exposure performed 24 h post-injection; CEUS and B-mode imaging before/after FUS
Qin et al. 2023 [37]	China	Murine 4T1 breast cancer cells; 4T1 tumor–bearing BALB/c mice (primary and contralateral tumors)	5 per group (7 groups)	Breast cancer	SPIOs + RPPs (superparamagnetic iron oxide NPs combined with resiquimod-loaded, perfluoropentane-encapsulated PLGA phase-transition nanodroplets)	10 mg ml− 1	Magnetic field–responsive US imaging via magnetic droplet vaporization (MDV); alternating magnetic field (AMF) used for mild hyperthermia (~ 43–44 °C, 10 min); US imaging (B-mode & CEUS) during MDV; IR thermal imaging for temperature monitoring
Samani et al. 2023 [38]	Iran	4T1 breast cancer cells and L929 normal cells	NA	Breast cancer	AuNCs/PFH NDs (BSA-functionalized gold nanoclusters-loaded perfluorohexane nanodroplets)	0.5 mg mL− 1	Theraputic: 1 MHz, 0.5–1.5 W cm− 2 × 4 min; exposure; Imaging: B-mode imaging performed
Zhang et al. 2023 [39]	China	4T1 breast cancer cells / 4T1 tumor-bearing BALB/c nude mice	5 per group (6 groups)	Breast cancer	SMISO NPs (Rattle-type SiO2 loaded with Mn-doped In2S3/InOOH)	15 mg kg− 1 (i.v. in mice); 0–200 µg mL− 1 (in vitro)	Theraputic: US applied 12 h post-injection at tumor site; 3 cycles every 5 days; Imaging: B-mode; 1.0 MHz; 1.0 W cm− 2 × 3 min (in vitro); 1.5 W cm− 2 × 2 min (in vivo)
Yang et al. 2024 [40]	China	4T1, HeLa, HepG2 cells / 4T1 tumor-bearing BALB/c mice	6 per group (5 groups)	Breast cancer	a-CoBiMn-LDH-PEG NPs (acid-etched amorphous CoBiMn layered double hydroxide nanoparticles with PEG modification)	200 µg mL− 1 (in vitro); 10 mg kg− 1 (i.v. in vivo)	Theraputic: US applied 30 min post-injection; 1 cycle per day for 16 days; 40 kHz; 3 W cm− 2; 6 min
Guo et al. 2024 [41]	China	Hepa 1–6 (HCC) and AML-12 cells; BALB/c mice with Hepa 1–6 xenografts	5 per group (5 groups)	Hepatocellular carcinoma	C/DCNB (Dual-drug (curcumin + doxorubicin-CDM-PEG) loaded acid-sensitive PEG-modified nanobubbles)	200 µg mL− 1 (in vitro); 200 µL of 200 µg mL− 1 solution (i.v. in mice)	Theraputic: 1.0 MHz; 2 W cm− 2; 50% duty cycle; 1 min exposure immediately and 8 h after injection; imaging: 5–12 MHz probe; CEUS mode
Xie et al. 2025 [42[	China	SKOV-3 ovarian cancer cells; NIH-3T3, 4T1, RAW264.7 cells / SKOV-3 tumor-bearing BALB/c mice	5 per group (7 groups)	Ovarian cancer	AMBC: AIPH–MSTN@BSA–MnO2@CCM (US and GSH dual-responsive nanoplatform)	20–500 µg mL− 1 (in vitro); 10 mg kg− 1 (i.v. in mice)	Theraputic: US applied 5 h post-injection; 1 MHz; 1.5 W cm− 2; 60 s per exposure; imaging: B-mode
Wang et al. 2025 [43]	China	HepG2 (human hepatocellular carcinoma) cells; orthotopic HepG2-Luc liver tumor in BALB/c nude mice	3–6 per group (4 groups)	Liver cancer	PAMSN (Phthalocyanine-conjugated mesoporous silicate NP loaded with allylhydrazine)	10 µM Pc / 100 µL i.v. injection	Theraputic FUS: 500 kHz; 400 kPa; 25% duty cycle; 2 min; imaging: B-mode & CEUS with passive cavitation detection ; time–intensity imaging over 2 min

A total of 24 preclinical studies published from 2015 to 2025 were identified as eligible, most of which were conducted in China, with the remaining studies originating from Iran. The majority of these studies used murine xenograft or syngeneic tumor models, such as 4T1 breast, SKOV3 ovarian, HepG2 liver, HT-29 colorectal, with a typical group size of 3–8 animals. The nanoplatforms employed in all studies were US-responsive with co-integrated therapeutic and imaging modalities. Frequently used architectures encompassed polymeric and lipid NPs, and inorganic nanostructures (such as TiO₂, SiO₂, LDH). Multiple systems co-encapsulated such as chemotherapeutics, photothermal, or oxygen-generating agents to improve therapeutic efficacy.

The US parameters reported in these studies were different. The parameters were generally in the range of 0.5–6 W/cm2, 40 kHz-3 MHz and 30 s–10 min of exposure. Low intensity focused US (LIFU) was the most common activation methods for therapy. The imaging modalities included B-mode, contrast-enhanced US (CEUS), and photoacoustic (PA) imaging. The imaging is performed generally 1–6 h after injection, when the accumulation of the nanoprobes in the tumor is at its peak. Overall, these studies show a clear trend toward multifunctional nanosystems that can be used for real-time tumor visualization and local activation with US in preclinical cancer models.

Theranostic outcomes of included studies

In Table 3, the comparative imaging and therapeutic efficacy of USI-guided sonodynamic cancer therapy are provided.

Table 3.

Comparative imaging and therapeutic efficacy of USI-guided in sonodynamic cancer therapy

Contrast agent/ sonosensitizer	Imaging findings	Therapeutic outcomes	Advantages	Limitations	Ref
PCF-MBs	Strong US/fluorescence enhancement; 5-fold tumor signal increase after US	90% tumor inhibition (US + Light); safe, no organ damage	Synergistic chemo-PDT; ABCG2 suppression	PDT depth limit	[20]
PIO_NPs	Strong US enhancement at tumor (peak 6 h post-injection)	Tumor reduction to ~ 250 mm3−, 61% apoptosis; median survival 50 d (+ 78% vs. control)	Dual-mode imaging; high ROS generation; O2 delivery and reduce hypoxia; synergistic chemo-PSDT	Limited to subcutaneous model; no long-term toxicity data reported	[21]
TPZ/HMTNPs-SNO	US contrast enhanced via NO bubbles; real-time tumor visualization	Tumor reduction to 79 mm3 (vs. 743 m m3 control); strong apoptosis, low PCNA; minimal systemic toxicity	Sequential tritherapy (SDT + NO + hypoxia-activated TPZ); enhanced tumor accumulation; on-demand NO release	Imaging weaker than clinical agents; limited long-term biosafety data	[22]
HMME/MCC-HA	Bright US contrast at tumor (acidic pH + US); enhanced CO2-mediated signal	Tumor volume ratio V/V₀ ≈ 0.87; extensive apoptosis+necrosis	Dual pH/US-triggered release; targeted (CD44) delivery; combined SDT+cavitation+vascular occlusion	Mechanism of cavitation-induced necrosis needs long-term study; imaging weaker than microbubbles	[23]
FA-PEG-PLGA-Ptx@ICG-Pfh	Strong PA signal at 800 nm; CEUS enhancement post-irradiation; high tumor accumulation	Tumor reduction of 63% vs. control; complete regression by day 14; minimal systemic toxicity	Dual PA/US imaging; targeted chemo-PTT synergy; laser-triggered drug release; high biocompatibility	Thermal therapy, not pure SDT; limited deep-tissue penetration	[24]
RBC-HPBs/HMME/PFH	US contrast enhanced via PFH phase transition to bubbles; real-time tumor visualization; strongest signal at 8 h post-injection	Tumor volume reduced to 20.5 mm3 (vs. control); high apoptosis via ROS-mediated mitochondrial pathway; relieved tumor hypoxia; reduced microvessel density; minimal systemic toxicity	O2 self-supplied SDT; enhanced ROS generation; long circulation and enhanced tumor accumulation; multi-mechanism treatment (apoptosis, tumor starvation, metastasis inhibition); US-triggered on-demand release	Limited to subcutaneous model; no long-term toxicity data reported	[25]
mTiO2@PPY-HNK	Strong US signal increasing with concentration; clear PA signal at 715 nm; enhanced tumor visualization in vivo	Tumor growth significantly suppressed, gradual ablation by day 16; no lung metastatic nodules with combined therapy; minimal systemic toxicity	Triple therapy (SDT + PTT+chemotherapy); dual US/PA imaging; high ROS generation; good photothermal conversion (~ 35.3% efficiency); inhibits tumor metastasis	Not pure SDT (combined with PTT and chemotherapy); limited to subcutaneous model; no long-term toxicity data reported	[26]
AIPH-loaded liposome	In vitro: generation of bright gas bubbles after US, strong echo and increased gray value versus PBS/blank liposome; In vivo: IV accumulation ~ 24 h post-injection; US contrast (microbubble-enhanced) increased with US exposure, average gray value increased from 33.96 ± 3.45 to 61.37 ± 1.03 a.u. after 30 min US irradiation	In vitro: Under US Lip-AIPH produced ROS and alkyl radicals; MCF-7 viability decreased with concentration and maximum inhibition 74.8% at 500 µg·mL− 1; In vivo: Lip-AIPH + US group showed marked tumor growth suppression; Histology/TUNEL showed increased apoptosis and necrosis in Lip-AIPH + US tumors	Generates gas bubbles (echogenic) and alkyl radicals on US activation → simultaneous US imaging guidance and oxygen-independent SDT (works under hypoxia)	Long-term biosafety beyond 40-day monitoring not fully addressed in the paper; (These limitations are stated or explicitly flagged as future directions)	[27]
OI_NPs (PFP + ICG+OXP in PLGA)	Dual-mode imaging: Strong phase-change CEUS enhancement and increased Echo Intensity; PA signal significantly higher for OI_NPs vs. PBS, free ICG or Blank NPs	In vitro: OI_NPs+PSDT produced the highest apoptosis (67.01 ± 4.60%) and > 50% cell death at the tested dose (ICG 8.5 µg/mL / OXP 6.0 µg/mL)	Combines chemotherapy + photo therapy + sono(chemical) therapy in one NP; enables PA + US dual imaging (ICG stability improved by PLGA); reported ICD markers (enhanced immunogenicity) and shows synergy between OXP and PSDT	Main therapeutic data are in vitro plus vaccination/rechallenge; explicit therapeutic tumor regression data / group sizes for in vivo treatment are not detailed	[28]
IR780-ND	US imaging: Strong CEUS enhancement via acoustic droplet vaporization (ADV), producing microbubbles (~ 2 μm) and marked echo-intensity gain; PA and US contrast maximized 24 h post injection	In vitro: ROS-dependent cytotoxicity increased with IR780 (0–5 µg mL⁻¹) and US duration; In vivo: Tumor growth significantly suppressed after 16 days in US + IR780-ND group (p < 0.05); H&E and TUNEL showed extensive necrosis and apoptosis	Enables deep tumor penetration via ADV and mitochondrial-targeted SDT with strong ROS generation; provides multimodal (US/PA/FL) imaging guidance	Lack of active targeting ligands (beyond IR780 affinity); long-term efficacy and repeat-dose data not yet established	[29]
FMSNs-Dox	US imaging confirmed persistent INB cavitation; Cavitation repeatable after single injection due to slow biodegradation	In vitro: Viability reduced to ~ 41% at 5 MPa (INB cavitation + Dox); In vivo: Vessel disruption (≤ 40 μm), increased intratumoral ROS, and 2.98× higher drug penetration; Tumor inhibition: Single injection + multiple US significantly slowed tumor growth	Combines anti-vascular, chemo-, and sonodynamic effects; allows repeatable imaging and therapy with one injection; long-lived INBs enable day-scale image-guided SDT	Partial lung embolism observed; hydrophobicity slows biodegradation and requires optimization for long-term safety	[30]
RB-MBs	US imaging: strong tumor contrast lasting > 4 min after injection; local US (1 MHz, 1.5 W·cm− 2) triggered RB-MB collapse and conversion to RB-NPs with reduced signal followed by re-perfusion	In vitro: ~84% cell death with RB-MBs + US; ROS generation confirmed via ADPA assay; In vivo: Tumor growth inhibition ~ 76.5% for RB-MBs + US vs. 23.8% (MBs + US) and 49.2% (RB-NPs + US)	Combines theranostic US contrast and SDT in one system; in situ micro-to-nano conversion via US enhances tumor delivery (~ 7.5× ↑); high ROS generation efficiency	Limited tumor penetration depth tested (subcutaneous model only); no long-term toxicity or repeat-dose data reported	[31]
FA-OINPs	Dual-mode US/PA imaging: After 808 nm irradiation, EI (B-mode) increased from 33.05 ± 2.78 to 83.09 ± 12.36 a.u. (~ 2.5×), CEUS increased by 22.5×, PA intensity increased vy > 2×; In vivo, FA-OINPs peaked at 6 h post-injection (faster and stronger accumulation than OINPs)	In vitro: Cell viability = 31.0 ± 4.8% and apoptosis/necrosis = 75.5 ± 0.9% (FA-OINPs + laser + US). In vivo: Tumor inhibition strongest among all groups; apoptosis index = 69.5 ± 4.5%	Combines targeted delivery, oxygenation, and photo-sonodynamic/ therapy; enables real-time US/PA imaging; enhances tumor accumulation and ROS yield under dual stimulation	Tested only in subcutaneous xenograft; no deep-tissue or orthotopic model data; laser exposure limits penetration depth	[32]
LIP3	Bright CEUS contrast after LIFU; visualized phase transition and localized drug release in vitro and in vivo	Cytotoxicity to 4T1 cells > 5 × LIP1 (AQ4N only) and > 2 × LIP2 (HMME only); cell viability ≈ 42.9%. In vivo tumor growth significantly inhibited under cascade SDT	Dual-mode US/PA imaging and LIFU-triggered drug release; cascade amplification therapy (AQ4N + SDT)	Subcutaneous model only; no deep-tissue or long-term safety data reported	[33]
CPDP NPs	LIFU-triggered phase transition of PFP enhanced both 2D and CEUS contrast in vitro and in vivo; grayscale intensity significantly increased after LIFU	In vitro ROS generation and cell apoptosis maximal in CPDP + LIFU group (~ 3× higher); In vivo, tumor growth inhibition significant after 18 days; lung metastases markedly reduced	Combines US imaging, SDT and chemotherapy in one platform; LIFU-triggered drug release and ROS generation enable synergistic tumor killing and anti-metastasis effect; good biocompatibility	Evaluated only in subcutaneous 4T1 model; no long-term toxicity data reported	[34]
OIX_NPs	B-mode and PA signals increased markedly after laser; peak tumor accumulation and maximum PA contrast observed 4 h post-injection; enhanced dual-modal US/PA imaging due to PFP phase transition and oxygen core	In vivo tumor growth inhibition rate increased to > 80%; apoptotic index 81.9%; cell infiltration in primary and distant tumors increased → abscopal effect	Enables oxygen-supplied chemo-photo-sonodynamic therapy with real-time PA/US guidance; dual anti-tumor and anti-metastatic efficacy with low toxicity	Distant tumors not fully eliminated; no data on long-term immunity or peritoneal metastases; mechanism of antigen-presenting cells not yet evaluated	[35]
A-DPPs	Strong CEUS signal enhancement and bright B-mode contrast at tumor site 24 h after injection; grayscale intensity significantly increased post-FUS compared with saline or A-DPs	Extensive coagulative necrosis in FUS + A-DPP group; tumor inhibition rate highest among all treated groups	Dual diagnostic–therapeutic capability; low EEF indicates efficient ablation energy use	Evaluated only in subcutaneous model; no long-term immune or metastasis data; limited tissue-depth validation	[36]
SPIOs + RPPs	Strong contrast enhancement (~ 67% signal increase) after AMF-induced magnetic droplet vaporization (MDV) in SPIO + RPPs group, both in vitro and in vivo; real-time guidance of mild MHT achievable	In vitro: apoptosis > 90%, strong HMGB1/ATP release, DC maturation increased to 69.7%; In vivo: tumor inhibition and metastasis suppression; Reported 100% survival at 90 days in the treated group	Combines magnetic hyperthermia, immunotherapy, and US imaging; enables mild-temperature therapy (~ 44 °C); safe and biocompatible; real-time monitoring through magnetic field–responsive US	Conducted only in 4T1 murine model; intratumoral injection route; translation to systemic delivery not verified	[37]
AuNCs/PFH NDs	Marked echogenic enhancement under US; AuNCs increased brightness; stable signal observed at 1 MHz confirming efficient droplet activation	Enhanced 4T1 cell apoptosis and colony inhibition under US (cell viability ≈ 56.8% at 1.5 W cm− 2; ≈ 2.5-fold increase in apoptotic cells vs. control; US exposure improved NDs uptake ≈ 2.6-fold	Reported good biosafety (> 80% cell viability at ≤ 120 µM); AuNCs surface facilitated cavitation; PFH core enabled ADV and enhanced imaging–therapy coupling	Only in-vitro evaluation; no in-vivo imaging validation; single US frequency tested	[38]
SMISO NPs	Enhancement of US reflection and scattering due to r-SiO2 multi-interface structure and MISO’s high acoustic impedance; in vivo gray value peaked at 12 h, confirming enhanced permeability and retention (EPR)-driven accumulation	Significant tumor inhibition (88.2%); marked apoptosis and ROS production; strong DHE red fluorescence in tumor sections; Ki67 expression reduced	Dual-function theranostic platform integrating high-contrast imaging and SDT; Mn doping enhanced ROS yield and good biocompatibility	Lacks long-term survival and immune-response analysis	[39]
a-CoBiMn-LDH-PEG NPs	US signal enhancement due to Mn4+-catalyzed O2 generation; contrast at tumor site increased within 30 min post-injection; O2 bubble formation visible especially at pH 6.5	In vitro cell viability ≈ 14.8% after US (3 W cm− 2, 6 min) and ≈ 9.8% with H2O2; in vivo complete tumor regression; H&E and TUNEL showed extensive apoptosis	Defect-rich amorphous structure promotes e−/h+ separation and high ROS yield (~ 8.2 × TiO2); Mn4+ mediated O2 generation relieves hypoxia and enhances imaging; good biocompatibility and metabolism	Only 4T1 model tested; lack of biodistribution in other tumors and long-term follow-up; single US frequency (40 kHz)	[40]
C/DCNB	Significant CEUS signal enhancement in vitro and in vivo; signal appeared within 30 s and maintained > 3 min	Cell viability < 20% in Hepa 1–6 cells after C/DCNB + US; marked TUNEL-positive apoptosis; in vivo tumor growth suppression > 70% vs. control	Dual-drug synergy (CUR + DOX) enhances chemosensitivity and ROS generation; CEUS guidance enables real-time monitoring and localized release	Limited biodistribution duration (24 h); no long-term recurrence or immune response assessment	[41]
AMBC	Dual MR/US responsiveness; T1-weighted magnetic resonance imaging (MRI) signal significantly enhanced after GSH exposure due to MnO₂→Mn²⁺ reduction; US gray-scale intensity markedly increased within 60 s post-irradiation, confirming N2 microbubble generation from AIPH	In vitro SKOV-3 cell viability < 30% after US (1.5 W cm− 2, 60 s); in vivo tumor inhibition > 95%; tumor volume reduced to 41.1 ± 26.3 mm3 vs. 1257.4 ± 248.4 mm3 (control)	Dual GSH/US responsiveness; MnO₂ catalyzes H2O2 to O2 improving hypoxia; AIPH releases alkyl radicals + N2 for cavitation enhancement; high biosafety and stability	Only SKOV-3 model evaluated; short observation (16 days); limited biodistribution analysis	[42]
PAMSN	CEUS signal increased up to 2.5× baseline; passive cavitation detection revealed 5–10 dB higher harmonics; fluorescence localized to liver tumor within 2–4 h post-injection	In vitro: enhanced cavitation and ROS generation; in vivo: marked tumor inhibition	Enables real-time ROS and cavitation monitoring via CEUS; orthotopic model validation; high stability and biocompatibility	Requires sophisticated multimodal FUS–USI platform; study limited to short-term efficacy	[43]

In all studies, US was validated to be theranostic in nature and was used as a noninvasive imaging modality as well as a trigger for sonodynamic or combination therapies. The concomitant diagnostic and therapeutic function facilitated accurate tumor targeting, activation control of the sonosensitizer, and monitoring of therapeutic efficacy. From an imaging perspective, enhanced US contrast was consistently reported following nanoplatform administration, often due to phase-transition mechanisms (e.g., PFH or PFP vaporization) or gas generation (e.g., CO₂, NO, or O₂ release). Many nanoplatforms also provided multimodal imaging capabilities, combining B-mode US, CEUS, PA, fluorescence, or magnetic resonance (MR) imaging to improve sensitivity and spatial resolution. Several systems—such as FA-OINPs, OIX_NPs, and AMBC—exhibited time-dependent imaging enhancement, typically peaking within 4–8 h post-injection, correlating with maximum tumor accumulation.

Therapeutically, nearly all studies showed significant tumor inhibition, high apoptosis, and ROS generation. The inhibition rates were higher than 70–90% in most studies. Hybrid approaches that combined chemotherapy, PTT, PDT, or magnetic hyperthermia had additive or synergistic anti-tumor efficacy. SPIOs + RPPs caused complete 90-day survival via immunomodulatory mild hyperthermia. a-CoBiMn-LDH-PEG NPs and AMBC elicited complete or near-complete tumor regression upon US activation.

Major advantages included targeted accumulation, hypoxia relief, biocompatibility, and spatiotemporal control of activation. However, limitations persist—namely, reliance on subcutaneous tumor models, nonstandardized US parameters, and limited long-term or deep-tissue evaluations. Overall, these findings affirm US-guided SDT as an emerging precision theranostic modality, integrating imaging and therapy within a single noninvasive framework.

Figure 2 gives a combined graphic overview of the most commonly reported study characteristics and outcomes as taken out of Tables 2 and 3. The left panel outlines the most important features of the experiment, such as the most common types of nanosensitizers (polymeric/lipid, inorganic and hybrid systems), the US regimes (LIFU, HIFU, and US in the kHz range), most common models of cancer (breast, ovarian, liver, and colorectal), as well as the most common types of imaging (B-mode, CEUS, PA imaging, and multimodal). The central panel reflects the conceptual flow of US-guided sonodynamic theranostics, which are an US-induced phase change/gas generation, improved imaging feedback, ROS generation, and the subsequent tumor cell apoptosis. The right card shows the range of most frequently described therapeutic outcomes (moderate to near complete tumor inhibition), most frequently used combination approaches (chemotherapy, PTT, and immunotherapy), recurrent benefits (strong antitumor activity, multimodal imaging and hypoxia restoration) and significant shortcomings (heterogeneous ultrasound parameters, use in subcutaneous tumors and limited follow-up times). Significantly, the elements presented in this figure reflect common trends as opposed to a comprehensive list of all findings in all included studies and have been made in order to offer a high-level, intuitively summary of the current preclinical situation.

Fig. 2.

Overview of study chracteristics and therapeutic outcomes of included studies

Risk of bias assessment

The quality of the methodology of the 24 preclinical studies included was determined using the SYRCLE ROB tool. A domain was rated as ‘unclear risk’ when the relevant methodological information (e.g., blinding, allocation concealment) was not explicitly reported in the main text or supplementary materials and could not be reasonably inferred from the study description. Table 4; Fig. 3 provide a summary of the study results carried out with the aim of obtaining a consistent overview of the ratings of various SYRCLE domains.

Table 4.

Risk of bias assessment

Author, Years	Sequence generation	Allocation concealment	Random housing	Baseline characteristics similar	Blinding (Investigators)	Random outcome assessment	Blinding of outcome assessor	Incomplete outcome data	Selective reporting	Other bias
M. Chen et al. 2018 [20]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
S. Chen et al. 2018 [21]	Unclear	Unclear	Low risk	Low risk	Unclear	Low risk	Unclear	Low risk	Low risk	None
Feng et al., 2018 [22]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Low risk	Low risk	Low risk	None
Feng et al. 2018 [23]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Liu et al. 2018 [24]	Low risk	Unclear	Low risk	Low risk	Unclear	Unclear	Low risk	Low risk	Low risk	None
Zhang et al. 2018 [25]	Unclear	Unclear	Low risk	Low risk	Unclear	Low risk	Low risk	Low risk	Low risk	None
He et al. 2019 [26]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Lin et al. 2019 [27]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Xie et al. 2019 [28]	Unclear	Unclear	Unclear	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Zhang et al. 2019 [29]	Low risk	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Ho et al. 2019 [30]	Unclear	Unclear	Unclear	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Hou et al. 2020 [31]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Liu et al. 2020 [32]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Low risk	Low risk	Low risk	None
Zhao et al. 2020 [33]	Low risk	Unclear	Unclear	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Zhang et al. 2021 [34]	Unclear	Unclear	Low risk	Low risk	Unclear	Low risk	Unclear	Low risk	Low risk	None
Zheng et al. 2021 [35]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Low risk	Low risk	Low risk	None
Kang et al. 2022 [36]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Qin et al. 2023 [37]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Samani et al. 2023 [38]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	High risk	Low risk	None
Zhang et al. 2023 [39]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Low risk	Low risk	Low risk	None
Yang et al. 2024 [40]	Unclear	Unclear	Low risk	Low risk	Unclear	Low risk	Unclear	Low risk	Low risk	None
Guo et al. 2024 [41]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Xie et al. 2025 [42]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None
Wang et al. 2025 [43]	Unclear	Unclear	Low risk	Low risk	Unclear	Unclear	Unclear	Low risk	Low risk	None

Fig. 3.

Distribution of ROB domains across all included studies

The general comparison of baseline characteristics was similar among the experimental groups, and selectivity of outcome reporting was not observed in the retrieved literature. Moreover, outcome information was mostly complete, and only one study [38] had been considered at high risk of attrition bias because of the lack of quantitative in vivo efficacy data and extended follow-up. Conversely, areas on selection and performance bias such as sequence generation, allocation concealment, random housing and blinding of caregivers or investigators were commonly rated as unclear risk. This was mainly due to lack of reporting and not necessarily due to lack of methodological flaw. In a similar vein, domains of detection bias, especially random outcome assessment and outcome assessor blinding were mostly rated as unclear with none of the studies reporting blinding in outcome interpretation on the imaging or histopathological assessment. Notably, all studies provided detailed descriptions of US imaging–guided SDT protocols, including imaging modality, treatment timing, and acoustic parameters, enabling reproducibility of the experimental workflow. Also, nanoplatforms, including size distribution, morphology, surface charge, and stability, were characterized in most reports, which indicated a relatively high standard of material and protocol standardization.

Altogether, the SYRCLE evaluation shows that although the reports of the outcomes and the characterization of the experiments were strong, some important methodological information connected with sequence generationran, domization, and blinding was frequently not reported explicitly. The results indicate that the current preclinical evidence behind theranostic performance of US-SDT is sound as per its methodology on the outcome level yet could use a better transparency in reporting the measures aimed to reduce the effects of selection, performance, and detection bias.

Disscussion

Interpretation of main findings

This section first discusses the fundamental therapeutic mechanisms underlying US-guided SDT, providing a biological and physical framework for interpreting subsequent advances in nanoplatform design and imaging guidance.

This systematic review of 24 preclinical studies shows that US-guided SDT is a reliable and reproducible theranostic strategy in rodent cancer models. A consistent message across a variety of nanoplatform designs, laboratories, and cancer types is: [1] US can be used reliably both as a noninvasive imaging modality and as a local trigger for sonosensitizer activation and [2] NP-based sonosensitizers incorporating imaging-responsive elements (phase-change liquids, gas-generating moieties, or acoustic scatterers) can significantly enhance tumour visualization and spatially restricted therapeutic action. In the majority of studies reviewed, these design features led to strong antitumour effects—typical tumor inhibition rates were often > 70% and several studies reported near-complete regression or survival benefits when SDT was combined with complementary modalities (chemotherapy, PTT, oxygenation strategies, or mild hyperthermia).

Mechanistically, the reviewed evidence supports three interlinked advantages of current US-guided SDT platforms. First, phase-transition or gas-generating components (e.g., PFH/PFP, NO, CO₂) provide immediate and measurable US contrast that both confirms tumour accumulation and enables on-demand activation. Second, oxygen-supplying or hypoxia-modulating strategies (e.g., Mn-based catalysts, oxygen cores) increase ROS yield and overcome a key limitation of ROS-based therapies in hypoxic tumors. Third, multifunctional combinations (SDT + chemo/PTT/immune-modulation) show additive or synergistic efficacy, suggesting that SDT is most powerful when embedded in a multimodal therapeutic design rather than as a monotherapy.

Beside these encouraging signs, some caveats merit consideration. There was significant heterogeneity in experimental conditions (particularly US frequency, intensity, duty cycle, exposure duration and timing of imaging after injection), which hampers quantitative synthesis and comparison of effect sizes. Studies typically relied on subcutaneous xenograft models using small groups of 3–8 animals and short observation periods; thus, it remains uncertain whether these findings can apply to orthotopic or metastatic models or immunocompetent conditions and deeper tumor depths. Long-term biodistribution and repeat-dose safety data were often absent, as was a more comprehensive assessment of immune-mediated outcomes; only a minority of studies reported long-term survival or 90-day follow-up. On the methodological side, the identified ROB patterns suggest that reported SDT efficacy may be influenced by methodological limitations, including insufficient blinding, selective outcome reporting, and variability in US exposure parameters. These factors may contribute to overestimation of therapeutic effects and limit cross-study comparability, underscoring the need for standardized experimental protocols and transparent reporting in future investigations.

Preclinical and review evidence on Ultrasound-Guided SDT

Building on these mechanistic principles, this subsection focuses on sonosensitizer and nanoplatform engineering strategies, including phase-change materials, microbubble-assisted systems, and ROS generating constructs that modulate therapeutic efficacy across existing literature.

The recent literature further recapitulates NP-enabled SDT, where therapeutic ROS generation is paired with concurrent imaging. Recent reviews place an emphasis on the expanding role of multifunctional sonosensitizers in synergistic diagnosis and therapy [44–47]. For example, Zhang et al. categorize acoustic‐responsive nanomaterials (carbon‐, iron‐, metal-organic framework‐based, etc.) that enable SDT with real‐time US guidance [44], and and Liang et al. review sonosensitizers engineered for US, CT (computed tomography), and MRI imaging as well as therapy [45]. Both those surveys and the present one concur that ROS generation is the fundamental cytotoxic mechanism: under US, sensitizers produce singlet oxygen and radicals by acoustic cavitation and sonoluminescence [48, 49]. Cavitation in particular is widely postulated – e.g. micro/nanobubbles loaded with sensitizers (or perfluorocarbon cores) act as cavitation nuclei to amplify SDT effects. In multiple studies, augmenting cavitation or oxygen supply increased ROS yield and efficacy. Lin et al. engineered Au–Pt nanobowls which produces O₂ bubbles that accumulate in the bowl rim, dramatically enhancing inertial cavitation and ROS under US [50]. Likewise, Feng et al. loaded quaternary ammonium salt NBs with Ce6: the membrane‐disrupting QAS shells synergistically augment SDT by pre‐sensitizing cells while US triggers Ce6‐mediated ROS [51]. These novel strategies build on the mechanisms identified in our review but extend them: in addition to standard ROS and cavitation, these studies exploit catalytic gas generation and chemical membrane disruption. Datta et al. also emphasize type I/II photochemical pathways for SDT (analogous to PDT) as central to cell killing [48].

The reviewed studies highlighted that the majority of US-guided SDT platforms resulted in CEUS imaging after phase-change (i.e. PFH/PFP vaporization) or gas release. The recent studies similarly demonstrate robust US imaging: Zhu et al. leveraged 18 MHz B‐mode US to localize magnetic black phosphorus microbubbles prior to SDT activation [52]. PA imaging is also favored: Li et al.’s DVDMS+IR780@PFP NPs provided both US (via PFP) and PA (via IR780/DVDMS) contrast [53], and Feng et al.’s Ce6@QAS NBs showed strong PA signal in vitro and in vivo [51]. Thus, recent work frequently combines the US/PA imaging of earlier SDT systems with new modalities.

Promising therapeutic effects were reported in these new studies as well, consistent with previous results in terms of high efficacy, but now also indicating immunological effects. Our review identified that nearly all USI-SDT platforms reported very high tumor inhibition (70–90% or greater) and extensive apoptosis/necrosis. The new reports generally agree: e.g. Feng et al. observed ~ 76% tumor growth inhibition in a melanoma model with Ce6@QAS NBs under US exposure, which was significantly higher than controls [51]. Crucially, a number of recent studies directly assess immunological responses. Cheng et al. found that SDT (Mn–porphyrin liposomes) induced ICD, M1 repolarization of macrophages, and activation of dendritic/T/NK cells [49]. These findings extend the review’s note that SDT can induce ICD and may synergize with immunotherapy.

Limitations

With the therapeutic and materials foundations established, this subsection discusses the current limitations and provide a future direction for ongoing studies in the field.

One of the biggest methodological limitations this review has found is the absence of the standardization of US dosimetry in the US guided SDT preclinical studies. There is also a lot of variability in the reported US parameters, such as frequency (kHz -MHz range), acoustic output (reported as intensity either in Wcm^− 2 or as pressure in MPa) and duty cycle, exposure duration, and transducer configuration. These were not compared across studies in a systematically way because the experimental setups varied significantly, in the areas of transducer design, focal geometry, calibration conditions, and coupling conditions. Notably, the conversion of acoustic intensity (Wcm^− 2) to acoustic pressure (MPa) is not a universal law and was not performed in this review because conversion cannot be done without exact knowledge of transducer-specific data and medium-dependent acoustic properties which were not always reported. This heterogeneity highlights the importance of standardizing the reporting of US parameters in future research to enhance their reproducibility and allow cross-study comparisons of US-SDT.

Moreover, the high diversity in sonosensitizer type and tumor models used in studies produces inconsistent outcomes which makes it difficult to perform comparative evaluations of different methods. In addition, the majority of these experiments are conducted in small animal xenograft models rather than more clinically relevant orthotopic or immunocompetent models. Long-term safety, biodistribution, and pharmacokinetics are often underreported across studies makes it difficult to establish standardized treatment protocols.

Although several studies reported marked tumor growth suppression, often described as exceeding 70% relative to controls, these findings were not suitable for formal quantitative synthesis. Across the included literature, tumor volume outcomes were reported using heterogeneous models, measurement intervals, calculation formulas, and graphical-only data presentations, frequently without complete numerical datasets or variance measures. As a result, direct numerical comparison or standardized categorization of inhibition rates across studies was not methodologically appropriate. Accordingly, this review presents a qualitative synthesis of tumor inhibition trends as reported by individual studies rather than pooled quantitative effect estimates.

Conclusion

The presented systematic review identifies US-guided SDT as a fast developing form of theranostics with robust preclinical results with various nanoplatform designs and tumor models. US-guided SDT provides a competitive solution to a number of limitations associated with conventional cancer treatments, especially to deep seated and treatment resistant tumors since it combines real time US imaging with spatially regulated therapeutic activation. Despite the fact that the field has a high potential, more standardization of US parameters, biological modeling, and long-term safety analysis will be needed to translate the field to clinical applications. Overall, the results of this review encourage the further development of US-guided SDT as a flexible and noninvasive image-guided cancer therapy with a great potential in translation.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ADV

Acoustic droplet vaporization

CEUS

Contrast-enhanced ultrasound

CI

Confidence interval

DOX

Doxorubicin

EPR

Enhanced permeability and retention

FMSNs

Superhydrophobic mesoporous silica NPs

FUS

Focused ultrasound

HIFU

High-intensity focused ultrasound

IC50

Half maximal inhibitory concentration

ICG

Indocyanine green

LIFU

Low-intensity focused ultrasound

MBs

Microbubbles

MRI

Magnetic resonance imaging

MPa

Megapascal

NO

Nitric oxide

NPs

Nanoparticles

PA

Photoacoustic

PEG

Polyethylene glycol

PDT

Photodynamic therapy

PFP

Perfluoropentane

PCF

Power cavitation factor

PRF

Pulse repetition frequency

PLGA

Poly(lactic-co-glycolic acid)

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

PTT

Photothermal therapy

PS

Photosensitizer

RGD

Arginylglycylaspartic acid

ROS

Reactive oxygen species

ROI

Region of interest

SDT

Sonodynamic therapy

SPIO

Superparamagnetic iron oxide particle

SYRCLE

Systematic Review Centre for Laboratory Animal Experimentation

US

Ultrasound

US-SDT

Ultrasound-guided sonodynamic therapy

W cm⁻²

Watts per square centimeter

Author contributions

Conceptualization of the study: H.U. and M.I.; Regulatory affairs: T.F.H.; Preparation of paper-based questionnaire: M.S.J. and W.R.K.; Manuscript draft and generation of table/figures: H.U., M.S.J. and W.R.K.; Critical revisions to the manuscript: H.U., M.I., and T.F.H. All authors read and approved the final manuscript.

Funding

None.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.