Sonocavitation therapy for pancreatic cancer: Research progress, clinical application and future perspectives

Summary

Pancreatic cancer (PC) remains one of the most lethal malignancies, characterized by a dense desmoplastic stroma, an immunosuppressive microenvironment, and poor response to conventional therapies. Sonocavitation therapy, based on ultrasound microbubble (USMB) technology, has emerged as a potential strategy to improve intratumoral drug delivery, remodel the tumor microenvironment (TME), and enhance chemotherapy, immunotherapy, gene therapy, and radiotherapy. Proposed mechanisms include sonoporation, cascade cavitation, vascular normalization, and mechano-acoustic remodeling of the stroma. Early clinical studies suggest that USMB-assisted chemotherapy is feasible and may improve disease control, although the evidence remains heterogeneous and is based largely on small, single-center or early-phase studies. Major barriers to translation include nonstandardized acoustic parameters, pharmacokinetic mismatch between microbubbles and co-administered drugs, incomplete representation of human disease in preclinical models, and limited long-term safety data. Future work should prioritize standardized acoustic dosing, biomarker-guided patient selection, clinically practical formulations, rigorous pharmacokinetic and biodistribution studies, and adequately powered randomized trials. This review summarizes the mechanistic rationale, preclinical evidence, and clinical progress of sonocavitation therapy in pancreatic cancer and outlines priorities for its clinical translation.

Graphical abstract

Health sciences

Introduction

Overview of pancreatic cancer therapy

Pancreatic cancer, particularly pancreatic ductal adenocarcinoma (PDAC), ranks among the most aggressive gastrointestinal malignancies, with a continuously rising global disease burden. According to statistics, there were approximately 64,050 new cases and 50,550 deaths in the United States in 2023, making it the third leading cause of cancer-related mortality.¹ In China, the pancreatic cancer incidence also presents a significant upward trend, whereas the overall 5-year survival rate remains approximately 8.5%.^2,3

The primary reason for the high mortality rate of pancreatic cancer is the extreme difficulty of early diagnosis. Approximately 80% of patients are diagnosed at a locally advanced or metastatic stage, losing the opportunity for surgical resection,⁴ which remains the only potentially curative treatment. But only 15%–20% of early-stage patients have the opportunity for surgical treatment, not to mention that even after successful resection, about 60%–80% of patients will experience recurrence, with 70% of them progressing to metastatic disease within one year.⁵ Therefore, metastatic disease, characterized by multi-organ dissemination is the terminal stage of most patients with pancreatic cancer, with the liver being the most common site of metastasis.⁶ Current clinical guidelines recommend systemic chemotherapy regimens based on gemcitabine or fluorouracil drugs, combined with other chemotherapy drugs. However, due to mechanical barriers posed by the tumor microenvironment (TME), achieving therapeutic drug concentrations locally within the tumor remains challenging.^7,8,9 Beyond chemotherapy, targeted therapy for pancreatic cancer is also effective but only in specific subgroups. The PARP inhibitor Olaparib for patients with BRCA1/2 mutations and the KRAS G12C inhibitor Sotorasib achieved an objective response rate of 21% and a disease control rate of 80% in second-line therapy.^10,11,12 For KRAS wild-type patients, nimotuzumab combined with gemcitabine has been shown to significantly prolong survival.¹³ Immunotherapy demonstrates low efficacy as a single agent in most pancreatic cancer patients (response rate < 5%), primarily attributable to the highly immunosuppressive TME.^14,15,16,17 In recent years, some studies have focused on tumor vaccines and CAR-T cell therapy for immunotherapy in pancreatic cancer. Various tumor vaccines targeting tumor-associated or tumor-specific antigens (such as personalized mRNA vaccines, DC vaccines, and KRAS-targeted vaccines) can specifically activate long-lasting anti-tumor T cell responses. When combined with radiotherapy/chemotherapy or immune checkpoint inhibitors, they can significantly extend recurrence-free survival in patients with pancreatic cancer. Multi-antigen targeted CAR-T cells can overcome the limitations of single-antigen targeting, improve tumor infiltration, and eliminate micrometastases. Both approaches can reshape the immunosuppressive TME. However, there are still many challenges, such as the dense fibrotic stroma in the pancreas blocking immune cell infiltration, tumor antigen heterogeneity causing immune escape, multiple suppressive cells such as Tregs/MDSCs continuously exhausting effector cells, off-target toxicity and cytokine storms with CAR-T therapy, high cost and long preparation time for personalized vaccines, and half of the patients not being able to mount an effective immune response.^18,19

The therapeutic challenges in pancreatic cancer stem from its unique tumor biology. First, a dense stromal barrier with extensive fibrotic connective tissue proliferation impedes drug penetration.²⁰ Second, strong genetic heterogeneity, with KRAS mutations present in 83.2% of patients (predominantly p.G12D (43.6%) and p.G12V (30.1%)) often accompanied by co-mutations in TP53, CDKN2A, and so forth,²¹ complicating targeted drug development. Third, a profoundly immunosuppressive TME, characterized by high expression of inhibitory factors such as CD47 and VEGF, mediates immune evasion.¹⁴ These features underscore the need for strategies that improve tumor-selective delivery and integrate local treatment effects with systemic antitumor immunity.

Overview of sonocavitation therapy for tumor

Based on the cavitation effect, ultrasound microbubbles (USMB) technology utilizes gas-core microbubbles (0.5–10 μm in diameter with lipid, polymer, or albumin shells) and their dynamic response to low-frequency ultrasound fields to achieve therapeutic functions,²² which provides a theoretically mature and clinically practical therapy—sonocavitation therapy. Specifically, microbubbles oscillate under low-frequency ultrasound (typically 1–3 MHz), expanding during the low-pressure phase and compressing during the high-pressure phase, ultimately undergoing inertial cavitation and collapse.²³ This process releases energy, which generates local high pressure (>1000 atm), high temperature (>5000 K), and microjets, capable of transiently disrupting vascular endothelial or cell membrane structures, and reversibly opening physiological barriers (e.g., blood-brain barrier, blood-tumor barrier).^24,25 For cells, cavitation energy induces the formation of transient, reversible pores in cell membranes (lasting minutes to hours), which is called Sonoporation,^26,27 significantly enhancing membrane permeability and promoting intracellular transport of drugs/genes.²⁸ Furthermore, during inertial cavitation, the released high temperatures and pressures can lead to the production of free radicals in the surrounding medium, generating reactive oxygen species (ROS) in aqueous environments.^29,30 If not sufficiently scavenged, oxidative stress induces severe metabolic dysfunction in cells, thereby modulating TME. Therefore, sonocavitation therapy not only has the potential to break through the dense matrix barrier of pancreatic cancer and promote drug delivery to the tumor site, but also has the potential to reshape the TME and enhance the therapeutic effect of other therapies such as immunotherapy.^31,32

Compared to other tumor treatments such as chemotherapy, targeted therapy, and immunotherapy, sonocavitation therapy offers several significant advantages. As illustrated in Figure 1, first, it enables precise targeted drug delivery.³³ Active targeting can be achieved by modifying microbubbles with antibodies (e.g., against PD-L1,³⁴ LHRH receptor³⁵), or passive accumulation can be leveraged via the enhanced permeability and retention (EPR) effect in tumor tissue, thereby increasing drug concentration at the target site. Second, it allows for minimally invasive and controllable drug release. The treatment process involves no ionizing radiation, and ultrasound parameters can be adjusted in real-time. Localized ultrasound application enables controlled drug release at the target site, maximizing local drug concentration while minimizing systemic toxicity. Finally, it facilitates synergistic enhancement of anti-tumor effects. Microbubbles themselves could serve as ultrasound contrast agents, which can enable visual drug delivery.³⁶ The capabilities of promoting drug delivery and modulating TME are conducive to enhancing combination therapies and achieving increased efficiency and decreased toxicity.

Figure 1.

The role of sonocavitation therapy in creating synergistic effects for pancreatic cancer treatment

Overall, the development of sonocavitation therapy has progressed from diagnostic contrast enhancement and ultrasound-targeted microbubble destruction to drug-loaded carriers and early clinical sonochemotherapy in recent years.³⁷ Although there are still some inevitable issues in this field, such as short microbubble circulation time, incomplete tumor penetration, uncertain dose-exposure relationships, device-dependent acoustic output, and the lack of standardized reporting and regulatory pathways for combination products, it still shows huge potential in cancer treatment. As a highly malignant digestive system tumor, the pancreas is accessible to ultrasound with sufficient penetration depth (∼10 cm), offering a unique advantage for its treatment.³⁸ Research on using USMB technology to improve the diagnosis and treatment efficiency of pancreatic cancer is extensive, with numerous clinical translation studies underway. This review focuses on the therapeutic applications of USMB technology for pancreatic cancer, discussing current research and its potential to improve the therapeutic paradigm for pancreatic cancer.

Mechanisms of sonocavitation therapy in synergizing pancreatic cancer treatment

Overcoming the dense stroma of pancreatic cancer to promote drug delivery

Sonoporation and targeted delivery

Sonocavitation therapy provides an effective strategy to overcome the dense stromal barrier of pancreatic cancer and enhance drug delivery. The core mechanism lies in utilizing the cavitation effect generated by microbubbles under ultrasound irradiation. Studies show that applying USMB treatment within a specific time window after intravenous chemotherapy drug infusion can significantly promote drug penetration into tumor tissue and improve therapeutic efficacy,^{39,40,41,42,43,44} and the relevant studies are summarized in Table 1. Diane Bressand and colleagues explored the efficiency of USMB technology to deliver Nab-paclitaxel to pancreatic cancer by intravenously injecting Nab-paclitaxel first and immediately injecting BG8214 or BG8610 microbubbles.⁴⁰ The results showed that the growth of tumor cells in the group treated with Nab-paclitaxel combined with USMB was significantly inhibited both in vitro and in vivo, without any side effects even at high-dose Nab-paclitaxel (20 mg/kg). Additionally, drug-loading strategies of conjugating chemotherapeutic drugs to the microbubble surface or encapsulating them within the core, combined with simultaneous ultrasound-triggered destruction, also effectively enhance drug delivery efficiency.^{52,53,54,55,56} Table 2 summarizes the relevant studies. For instance, John F. Callan and colleagues developed a microbubble-liposome conjugate (IRMB-OxLipo) for co-delivery of irinotecan (IR) and oxaliplatin (Ox).⁵³ Utilizing ultrasound-targeted microbubble destruction (UTMD) technology, IRMB-OxLipo enhanced drug delivery efficiency and efficacy in pancreatic cancer while reducing toxicity. Both above strategies rely on local ultrasound irradiation of the tumor area, causing circulating microbubbles to rupture at the target site. As shown in Figure 2, the mechanical forces generated by cavitation (e.g., microjets, shock waves) transiently increase vascular endothelial gaps and cell membrane permeability, driving drug extravasation into tumor tissue. Notably, this cavitation effect loosens the pancreatic cancer stroma and increases vascular permeability, promoting not only the enrichment and penetration of small chemotherapeutic molecules but also rather large drug-loaded nanoparticles such as Nab-paclitaxel (130 nm in diameter).^40,44 This suggests sonocavitation therapy is not limited by drug molecular weight or size and can be applied to various clinical drugs.

Table 1.

Preclinical researches of co-administration drug and microbubbles for pancreatic cancer treatment

Treatment	Drug & microbubbles	Animal models	US device	US parameters	Results	Reference
Chemotherapy	Gemcitabine Clinical SonoVue®	orthotopic MIA PaCa-2-luc tumor model in nude mice	Laboratory customized single-element focused ultrasound transducers	1.0 MHz; MI 0.26; DC 40%; Acoustic intensity 8.80 mW/cm2; PNP 0.20 MPa; 4 min	significantly reduced primary tumor volume; increased median survival by 10%; delayed metastases to 7–8 weeks and reduced the number of metastases by 40%	Kotopoulis et al.39
Chemotherapy	Nab-paclitaxel Preclinical BG8214 or BG8610 MBs	subcutaneous BxPC-3 tumor model in nude mice	Laboratory 1 MHz single-element focused ultrasound transducer	1.0 MHz; DC 40%; PNP: 400 kPa; PRF 10 kHz; 3 min	significantly reduced tumor volume (p < 0.05)	Bressand et al.40
Chemotherapy	Gemcitabine Preclinical VISISTAR® +SONAZOID®	subcutaneous Pan-02 tumor model in C57BL/6 mice	Clinical Olympus GF-UCT 260 endoscopic ultrasound (EUS) system with 4 MHz convex array probe	Sonoporation: 4.0 MHz; MI 0.16; PNP 0.65 MPa; PRF 10 Hz; 3–6 min MB destruction: 1.0 MHz; MI 0.2	significantly extended tumor diameter doubling time (TDT)	Shimamoto et al.41
Chemotherapy	FOLFIRINOX Clinical SonoVue®	orthotopic KPC001S gLuc/green fluorescent protein tumor model in B6 albino mice	Laboratory 1 MHz single-element focused ultrasound transducer (Imasonic)	1.0 MHz; MI 0.3/0.6; PRF 0.25 Hz; 7 min	increased Pt uptake by 1.2–2.5 times; no significant difference in tumor volume	Haram et al.42
Chemotherapy-gene therapy	PEI nanogels loaded with Gem and periostin siRNA Clinical SonoVue®	subcutaneous SW1990+ PSCs tumor model in BALB/c nude mice	Laboratory Customized 1 MHz focused ultrasound transducers	1.0 MHz; DC 20%; Power density: 0.4 W/cm2; PRF 1.0 kHz	significantly decreased tumor volume and weight (p < 0.001)	Liu et al.43
Chemotherapy	CA19-9 targeted PTX-loaded nanoparticles Clinical SonoVue®	subcutaneous Capan-1 tumor model in BALB/c nude mice	Laboratory Topteam161 Physioson-Basic Therapeutic Ultrasound Machine (Physioson Elektromedizin AG)	1.0 MHz; DC 20%; Acoustic density 2.0 W/cm2; 2 min	a tumor inhibition rate of 90.51%; the longest median survival time (75 ± 7.12 days); no significant toxicity	Xing et al.45
Sonocavitation therapy	Preclinical DFB MBs	subcutaneous 4662 tumor model in C57BL/6 mice	Laboratory therapy and imaging probe system (TIPS, Philips research)	1.0 MHz; DC 10%; PNP 0.5 MPa; PRF 100 Hz; 7–10 min	increased activated macrophages (MHCII+), CD4+ T cells, and CD8+ T cells; reduced PD-1+ CD8+ T cells in draining lymph nodes; significant tumor growth suppression	Joiner et al.46
Sonocavitation therapy	Clinical SonoVue®	orthotopic SW1990 tumor model in BALB/c nude mice	Laboratory PHYSIOSON-Basic Ultrasound Therapy Instrument (Physioson Elektromedizin AG, Germany)	1.0 MHz (±5%); DC 20%; Incident intensity 1.2 W/cm2 (±20%); PRF 1.0 kHz (±10%); 2 min per session, 5 times per week	inhibited tumor growth and reduced lung metastasis by 60%	Lin et al.47
Chemotherapy, Immunotherapy	Dox, Gem, or αPDL1 Preclinical NH002 MBs	orthotopic UN-KC-6141 tumor model in C57BL/6J mice	Clinical Philips CX50 Portable Clinical Ultrasound System	Diagnostic/localization (CEUS/B-mode): MI 0.8 (B-mode), 0.08 (CEUS) Treatment (Power Doppler mode): MI 0.4; 4 min	extended median survival to 33.5 days (vs. 27 days with chemotherapy alone); prolonged survival (46 days)	Wu et al.48
Chemotherapy-gene therapy	Gem and miR-21 inhibitor via dendrimer-entrapped gold nanoparticles Clinical SonoVue®	subcutaneous SW1990 tumor model in BALB/c nude mice	Laboratory PHYSIOSON-Basic Ultrasound Therapy Instrument (Physioson Elektromedizin AG, Germany)	1.0 MHz; Power density 0.4 W/cm2; 2 min per session	significant tumor volume reduction (p < 0.001); median survival extended to 120 days (vs. Control 75 days)	Lin et al.49
Chemotherapy	Nab-paclitaxel+Gem or Liposomal Irinotecan Preclinical PS101	subcutaneous patient-derived xenograft (PDX) tumor model in nude mice	Laboratory custom dual-frequency Ultrasound Transducer (Relab, Norway)	Step 1 (Activation): 2.7 MHz; PRF 1 kHz; MI 0.3 Step 2 (Enhancement): 0.5 MHz; PRF 1 kHz; MI 0.20	reduced the median tumor volume at day 50; increased complete response rate at day 120	Ng et al.50
Radiotherapy	150 kVp X-ray, TiOx NPs Clinical Sonazoid™	In vitro MIAPaCa-2, PANC-1 pancreatic cancer cells	Clinical GE LOGIQ E Portable Ultrasound System	4.0 MHz; MI 1.0; 3 min	the enhanced effects of radiotherapy occur in relation to specific cell species, microbubble concentration, and radiation dose	Nakayama et al.51

Gem, gemcitabine; 5-FU, 5-fluorouracil; RB, rose bengal; Dox, doxorubicin; PTX, paclitaxel; FOLFIRINOX, fluorouracil + irinotecan + oxaliplatin + calcium folinate; IR, irinotecan; Ox, oxaliplatin; PEI, polyethyleneimine; MI, mechanical index; DC, duty cycle; PRF, pulse repetition frequency; PNP, peak negative pressure; CEUS, contrast-enhanced ultrasound.

Table 2.

Preclinical researches of Drug-loading microbubbles for pancreatic cancer treatment

Treatment	Drug	Animal models	US device	US parameters	Results	Reference
Chemotherapy-gas therapy	Gem and H2S-loading MBs	orthotopic PaTu 8902 tumor model in BALB/c nude mice	Laboratory Sonoplus 190 Ultrasound Therapy Device	3.0 MHz; 2.0 W/cm2; DC 50%; 1 min	decreased tumor volume; no significant weight change or biological toxicity	Huang et al.57
Chemotherapy-sonodynamic therapy	O2, 5-FU and RB-loading MBs	orthotopic BxPC-3 tumor model in nude mice	Laboratory Sonidel SP100 Sonoporator	1 MHz; 3.5 W/cm2; 3.5 min; 30% duty; PRF 100 Hz; PNP 0.48 MPa	decreased tumor volume by 48.3% (p < 0.05); no significant weight change or toxicity	Sheng et al.58
Chemotherapy	DOX and iron oxide NPs-loading liposome-MBs conjugate	subcutaneous Panc02 tumor model in nude mice	Laboratory Customized 1 MHz focused ultrasound transducers	1 MHz; DC 30%; 3.5 W/cm2; 3.5 min	resulting in the smallest tumor volume; deeply penetrated iron oxide NPs; lower cardiac toxicity	Dwivedi et al.59
Chemotherapy	PTX-loading MBs	subcutaneous PANC-1 tumor model in BALB/c nude mice	Clinical Chattanooga 2776 Intelect Mobile Ultrasound	1 MHz; DC 10%; 3 min; 2.5 W/cm2	resulting in the smallest tumor volume; a tumor inhibition rate of 89.8%; no obvious damage to major organs	Zhang et al.60
Chemotherapy-sonodynamic therapy	O2 and Gem-loading MBs, O2 and RB-loading MBs	subcutaneous MIA PaCa-2 tumor model in nude mice	Laboratory Sonidel SP100 Sonoporator	1 MHz; 3.5 W/cm2; 3.5 min; DC 30%; PRF 100 Hz; PNP 0.48 MPa; MI 0.48	reduced cell viability; Delayed Tumor growth by >80% (p < 0.001) without liver or kidney damage	Nesbitt et al.52
Chemotherapy	IR and Ox-loading MBs-liposome conjugate	subcutaneous BxPC-3 tumor model in nude mice	Laboratory Sonidel SP100 Sonoporator	1 MHz; 3.5 W/cm2; 3.5 min; DC 30%; PRF 100 Hz; PNP 0.48 MPa; MI 0.48	reduced tumor volume by 136% with no significant weight loss	Gao et al.53
Chemotherapy	Gem-loading MBs	subcutaneous MIA PaCa-2 tumor model in nude mice	Clinical S3000 Helx Evolution	Imaging:4 MHz; MI 0.06; UTMD:MI 1.35; 4 s	no significant difference in tumor growth	Delaney et al.54
Chemotherapy	Gem and PTX-loading MBs	subcutaneous BxPC-3 tumor model in nude mice	Laboratory Sonidel SP100 Sonoporator	1 MHz; 3.5 W/cm2; 3.5 min; DC 30%; PRF 100 Hz; PNP 0.48 MPa	reduced tumors volume by 40 ± 30% with stable body weight	Logan et al.55
Sonodynamic-gas-gene therapy	MBs loaded with PARP1 siRNA, Pyropheophorbide and DSPE-PEG2000-SNO	subcutaneous SW1990 tumor model in nude mice	Clinical Chattanooga 2776 Intelect Mobile Ultrasound	1.0 MHz; 1.0 W/cm2; DC 20%; 3 min	68.3% tumor inhibition rate	Tang et al.56

Gem, gemcitabine; 5-FU, 5-fluorouracil; RB, rose bengal; Dox, doxorubicin; PTX, paclitaxel; NPs, nanoparticles; PRF, pulse repetition frequency; PNP, peak negative pressure; MI, mechanical index; DC, duty cycle; UTMD, ultrasound-targeted microbubble destruction.

Figure 2.

Two strategies of sonocavitation therapy for enhancement of drug delivery

To further optimize delivery efficiency, microbubbles can be modified for active targeting.^45,58,60 By conjugating antibodies or ligands specific to pancreatic cancer-associated molecular targets (e.g., vascular endothelial growth factor receptor type 2 (VEGFR2),⁶¹ integrin αvβ3⁶⁰) onto the microbubble surface, active targeted delivery can be achieved. Contrast-enhanced ultrasound confirmed that such targeted modifications significantly enhance specific accumulation of microbubbles in tumor tissue,⁶² which helps to create sufficient cavitation within the tumor to promote more drug leakage into the tumor parenchyma, and ultimately significantly enhance drug efficacy. Zhang et al. constructed paclitaxel (PTX)-loaded microbubbles which were modified with both cCLT1 peptides targeting fibronectin in the tumor matrix and cRGD peptides targeting the tumor cell membrane, forming dual-targeted drug-loaded microbubbles (PTX-RCMBs).⁶⁰ USMB technology first enhanced tumor vascular permeability through the cavitation effect and promoted the conversion of microbubbles into nanoparticles (PTX-RCNPs). Subsequently, the cCLT1 peptide mediates the penetration of PTX-RCNPs into the matrix and binds to tumor cell surface integrin (α₅β₁); The final cRGD peptide further targets integrin (αvβ₃) to promote cellular uptake. This strategy increased cellular uptake by 3 times compared with only microbubbles group in vitro and reduced tumor volume to 1/4 of that of the non-targeted microbubble group in vivo, achieving an 89.8% tumor growth inhibition rate in a mouse model of pancreatic cancer, providing an efficient strategy for targeted therapy of pancreatic cancer.

Cascade cavitation

While active targeting increases microbubble accumulation in tumor tissue, enhancing cavitation efficiency and drug delivery, the complex vascular endothelial and stromal barrier system of pancreatic cancer still limits the efficiency of a single cavitation event. Addressing this, Huang et al. proposed a “cascade ultrasonic cavitation” strategy, providing an innovative solution to overcome multiple biological barriers in pancreatic cancer.⁶³ As displayed in Figure 3, the team developed a sequentially ultrasound-responsive composite system of macrophage membrane-modified microbubbles-mesoporous silica nanoparticles (MMB@TFHPD). The macrophage membrane coating confers tumor tropism via ligand-receptor recognition, promoting preferential accumulation in the tumor region. Then, spatiotemporally separated sequential cavitation effects achieve dual breakthrough. In detail, the first ultrasound irradiation (1.0 W/cm², 3 MHz) induces microbubble cavitation and rupture, selectively opening the tumor vascular endothelial barrier and promoting nanocarrier accumulation in the target area. Subsequently, triggered by the weakly acidic tumor microenvironment, the carrier structure transforms, enabling a second ultrasound irradiation (2.0 W/cm², 3 MHz) to drive the liquid-gas phase transition of perfluoropentane (PFP), generating a novel cavitation effect that degrades the dense extracellular matrix. This cascade mechanism increased drug penetration depth in PDAC by 4.1 times, significantly enhancing the killing efficacy against deep-seated tumors.

Figure 3.

A cascade ultrasonic cavitation strategy of MMB@TFHPD to overcome multiple biological barriers in pancreatic cancer

(A) The procedure for preparing and fabricating MMB@TFHPD and the drug release mechanism enabled by the sequential ultrasound-triggered cascade ultrasonic cavitation effect.

(B) A diagram demonstrates the sequential ultrasound-induced cascade ultrasonic cavitation, which successively breaches the blood-tumor barrier and the dense extracellular matrix barrier to facilitate the deep delivery of chemotherapeutic agents. Reproduced with permission.⁶³ Copyright 2025, John Wiley and Sons.

Mechano-acoustic synergy

The cascade cavitation strategy tackles the vascular endothelial and stromal barriers sequentially, but the excessive deposition of collagen and hyaluronic acid within pancreatic cancer leads to tumor stiffening. The resulting mechanical stress compresses blood vessels, causing up to 95% vessel compression and 80% complete collapse, forming hypoperfused regions that also hinder microbubble perfusion. To address this, the team of Constantina Neophytou proposed an innovative strategy of “mechanotherapy combined with sonopermeation”.⁶⁴ As shown in Figure 4, this strategy achieves dynamic regulation through sequential intervention. First, a TGF-β inhibitor (e.g., ketotifen) was used to modulate cancer-associated fibroblast function, reducing the tumor elastic modulus by 42%. Then ultrasound irradiation excited microbubble cavitation, further reducing the elastic modulus by 28%. This synergy creates a positive feedback loop—mechanotherapy-opened vascular networks promote microbubble delivery, while mechanical waves from cavitation continuously decompose collagen fibers, significantly improving perfusion by 3.2 times. Notably, the combined intervention not only increased nanodrug tumor enrichment by 2.7 times but also simultaneously remodeled the immune microenvironment. CD8⁺ T cell infiltration density increased by 4.1 times, and the Treg proportion decreased to 31% of the control group, with these effects being significant in both primary and metastatic sites. This strategy of transforming the tumor from a “dense state” to a “permeable state” has already enabled the conversion of 60% of unresectable pancreatic cancers to resectable status via a losartan combination regimen, marking the successful clinical translation of the mechanotherapeutic index concept.

Figure 4.

A “mechanotherapy combined with sonopermeation” strategy to enhance drug delivery in pancreatic cancer

Reproduced with permission.⁶⁴ Copyright 2025, Springer Nature.

The above ultrasound-assisted strategies primarily focus on breaking the physical stromal barriers of pancreatic cancer to improve intratumoral drug accumulation and penetration. Beyond overcoming tumor delivery barriers, achieving precise, on-demand drug release at tumor sites is another critical prerequisite for maximizing therapeutic efficacy and minimizing systemic toxicity.

Ultrasound-controlled drug release

Traditional chemotherapy for pancreatic cancer involves systemic administration of chemotherapeutic drugs, leading to widespread distribution and inevitable significant toxic side effects. Effective delivery of chemotherapeutic agents with minimal toxicity and maximum effect is crucial in clinical practice but technically challenging. Many researchers have developed a series of ultrasound microbubble-responsive drug delivery systems that enable controlled release upon exposure to ultrasound pulses.^57,59,60 Ultrasound-triggered release is governed by a coupled set of variables: frequency and mechanical index influence whether stable or inertial cavitation dominates; acoustic intensity, pulse length, duty cycle, and exposure duration determine cumulative mechanical and thermal dose; and microbubble size, shell composition, gas core, concentration, drug-loading method, and stability determine oscillation, rupture, circulation, and release. Comparisons across studies are therefore difficult when intensity or duty cycle is not reported. Stable cavitation is generally preferred for controlled sonoporation and vascular permeability, whereas inertial cavitation may be useful for stronger disruption, matrix remodeling, or ROS generation but carries greater risk of hemorrhage and tissue injury. Parameter selection should consequently be linked to the intended biological effect, with the hope of increasing drug concentration within the tumor while reducing systemic distribution, allowing effective tumor cell killing even at lower doses and improving therapeutic outcomes.

Ronald X. Xu’s team constructed a DOX-ML-MBs complex system,⁵⁹ where doxorubicin (DOX)-loaded magnetic liposomes (DOX-MLs) were covalently conjugated via a thiol-maleimide bond to perfluorocarbon gas microbubbles, forming composite microbubbles approximately 4 μm in size. When exposed to ultrasound irradiation, DOX-ML-MBs undergo stable or inertial cavitation. This cavitation causes microbubble rupture, breaking the covalent bond and releasing the DOX-loaded DOX-MLs. After intravenous injection, an external magnetic field enriches DOX-ML-MBs in the tumor region. Microfluidic vascular model experiments showed significantly higher retention of DOX-ML-MBs under a magnetic field compared to the no-field group (p < 0.01). Afterward, the tumor was irradiated with 1–3 MHz ultrasound, which acts through dual mechanisms of cavitation and thermal effects. On one hand, it disrupted the microbubble structure to promote drug release. On the other hand, it enhanced tumor vascular permeability via sonoporation, allowing magnetic liposomes to penetrate the dense pancreatic cancer stroma. The results showed that DOX-ML-MBs reduced tumor volume by approximately 60% compared to the control group (p < 0.001) while significantly reducing cardiotoxicity through a combined magnetic-ultrasound treatment strategy in vivo.

Zhao et al. used microfluidic electrospray technology to prepare ultrasound-responsive microbubbles (Gem-H₂S-UDMs) based on an alginate/poly(N-isopropylacrylamide) (ALG/PNIPAM) hydrogel shell, encapsulating the chemotherapeutic drug gemcitabine and H₂S therapeutic gas within the hydrogel shell.⁵⁷ Utilizing the thermosensitive contraction property of PNIPAM under ultrasound-induced mild hyperthermia enabled dual controlled release of gemcitabine and H₂S gas. Results showed that Gem-H₂S-UDMs combined with ultrasound treatment reduced PaTu-8902 cell viability to 45.37%, significantly lower than single-agent groups. The group receiving intratumoral injection of Gem-H₂S-UDMs combined with ultrasound showed a Ki67 positive rate of only 21.05% and an apoptosis rate of 10.26%, achieving low-toxicity, high-efficiency precision therapy.

Ultrasound-controlled drug release technology offers a promising solution to address the challenges of traditional pancreatic cancer chemotherapy. By leveraging ultrasound-responsive drug delivery systems, it successfully enhances drug concentration in tumors while minimizing systemic distribution, effectively reducing toxic side effects. Through innovative designs such as the DOX-ML-MBs complex system and Gem-H₂S-UDMs, researchers have demonstrated significant tumor volume reduction, decreased cell viability, and improved therapeutic outcomes. Nevertheless, the intrinsically hypoxic and immunosuppressive TME remains a key factor driving tumor progression, drug resistance, and immune escape, which severely restricts the ultimate therapeutic outcome. To further enhance anti-tumor efficacy, emerging ultrasound-microbubble technologies have been developed to remodel the tumor microenvironment and reverse its malignant characteristics.

Remodeling the tumor microenvironment

Alleviating hypoxia

The TME of pancreatic cancer is characterized by hypoxia due to vascular abnormalities, interstitial fibrosis, and high interstitial fluid pressure. This reduces chemotherapy drug delivery efficiency and promotes tumor metastasis. Microbubbles can not only directly load oxygen (O₂) into their core for targeted delivery to the tumor site, but also produce the mechanical cavitation effect and vascular modulation under ultrasound, thereby further effectively alleviating hypoxia in pancreatic cancer.^52,46,47,48

By encapsulating O₂ within microbubbles and delivering it directly to the tumor site, the hypoxic TME of pancreatic cancer can be effectively mitigated, enhancing the efficacy of various therapies such as chemotherapy and sonodynamic therapy. John F. Callan and colleagues prepared O₂-loaded microbubbles for co-delivery of O₂ with the chemotherapeutic drug gemcitabine (Gem) and the sonosensitizer rose bengal (RB).⁵² O₂-loaded microbubbles deliver O₂ to the TME, directly improving local tumor O₂ partial pressure. This not only provides rich O₂ for sonodynamic therapy (SDT) to generate toxic ROS but also reduces drug resistance induced by hypoxia and enhances the chemosensitivity to Gem. Microbubbles enable targeted release of O₂ and loaded drugs under ultrasound, reducing side effects from systemic administration. The results showed that expression of hypoxia-related genes HIF-1α and VHL was significantly downregulated after treatment, indicating alleviated hypoxia. Simultaneously, tumor growth was delayed by >80% (p < 0.001), with longer duration of efficacy than that of chemotherapy alone.

Regarding the mechanical cavitation effect, microbubbles undergo periodic expansion and contraction under ultrasound, increasing vascular endothelial permeability via “sonoporation” while disrupting the physical barriers of the tumor interstitium, thus improving blood perfusion. Wu et al.⁴⁸ used a clinically available ultrasound system (Power-Doppler) combined with NH002 microbubbles (validated in phase II clinical trials) to study a mouse pancreatic cancer model and found a 1.3–1.9-folds increase in tumor perfusion area 1 to 3 days after treatment, as well as decreased hypoxic areas by 77% (day 1) and 67% (day 3), with effects lasting up to 6 days after treatment. Furthermore, ultrasound combined with microbubbles can promote vascular normalization through stable cavitation, increasing pericyte coverage and endothelial cell junction integrity, reducing vascular leakage, and prolonging the duration of improved blood flow. Ultrasound combined with microbubbles can remodel tumor vascular structure rather than simply increasing vascular density. Researchers also showed that USMB technology can increase vascular maturity in pancreatic cancer, enhance pericyte coverage and endothelial tight junctions, thereby reducing abnormal vascular branching and optimizing blood perfusion efficiency.⁴⁷ Meanwhile, this technology can reduce damage to the vascular endothelium by chemotherapeutic drugs (e.g., doxorubicin, gemcitabine), avoiding the vicious cycle of chemotherapy exacerbating hypoxia. According to their research, combination therapy reduced the hypoxic area in gemcitabine-treated tumors from 33.4% to 7.6%, significantly prolonging mouse survival.

Sonocavitation therapy, based on USMB technology, represents a promising strategy to alleviate hypoxia in pancreatic cancer by directly delivering oxygen to the tumor site, leveraging the mechanical cavitation effect, and promoting vascular normalization et al., as evidenced by the downregulation of HIF-1α and a significant reduction in hypoxic area. It not only boosts the efficacy of chemotherapy and sonodynamic therapy but also reduces the side effects associated with systemic drug delivery, offering a potent combination therapy with strong potential for clinical translation.

Modulating the immunosuppressive state

The pancreatic cancer microenvironment is characterized as a “cold tumor”, featuring defective antigen-presenting cell function, insufficient cytotoxic T cell infiltration, and enrichment of M2-type macrophages, leading to immune escape.^65,66 Sonocavitation therapy can reverse this state through various mechanisms, including immunogenic cell death (ICD), immune cell recruitment, and phenotype regulation.^32,67,68

Regarding immunogenic activation, sonocavitation therapy induces mechanical damage, triggering tumor cells to release damage-associated molecular patterns (DAMPs), such as high-mobility group box 1 (HMGB1). Jordan B. Joiner et al.⁴⁶ found that 2 days after treatment with low-intensity focused ultrasound (LoFU) combined with microbubbles, HMGB1 translocation to the cytoplasm increased in tumor cells, and the number of activated antigen-presenting cells (e.g., macrophages) in draining lymph nodes significantly rose, with an increased proportion of MHC-II positive macrophages, initiating an adaptive immune response. Additionally, this technology reduced the expression of PD-1 and the proportion of PD-1⁺CD8⁺ T cells in draining lymph nodes after 2 days of treatment, suggesting enhanced T cell function. Besides, LoFU + microbubbles also played a role in immune cell regulation and reshaped immune cell composition. After 2 days of treatment, the numbers of CD4⁺ and CD8⁺ T cells in draining lymph nodes significantly increased, but the effect weakened after 15 days, indicating the need for combination with immunotherapy to sustain effects. For macrophages, sonocavitation therapy can promote polarization from M2 to M1 type. Lin et al. showed that USMB treatment increased the expression of the M1 marker CD86 in tumor-associated macrophages (TAMs) by 2-fold while downregulating the M2 marker Arg-1 and reducing pro-tumor factor secretion.⁴⁷ However, tumor size must be considered, as repeated treatments for larger tumors may reduce immune cell recruitment efficiency and even decrease CD8⁺ T cell numbers.

Regarding synergistic enhancement with immunotherapy, sonocavitation therapy can improve the efficacy of immune checkpoint inhibitors. A 2024 study confirmed that this technology increased intratumoral CD8⁺ T cell infiltration by 4-fold.⁴⁸ When combined with anti-PD-L1 antibody, median survival extended from 28 days to 46 days, and tumor growth inhibition rates significantly improved. The mechanism may be related to alleviated hypoxia, increased immune cell infiltration, and reduced immunosuppressive factors, providing a new approach for converting immunological “cold” tumors to “hot”.

In the meantime, the unique pathological features of pancreatic cancer create multiple physical barriers that hinder the delivery and functional performance of microbubbles. Pancreatic tumors possess dense desmoplastic stroma, activated pancreatic stellate cells, high interstitial pressure, and compressed aberrant blood vessels, together with a hypoxic niche enriched with tumor-associated macrophages, myeloid-derived suppressor cells, and regulatory T cells. These features create biological barriers to microbubble perfusion, drug extravasation, oxygen delivery, and impede immune cell infiltration into tumor parenchyma. They also impose design constraints: microbubbles must remain sufficiently stable during circulation, respond to clinically achievable acoustic pressures, and generate bioeffects without damaging fragile or compressed vessels. This pathology explains why treatment efficacy depends not only on the carrier but also on tumor location, blood perfusion status, stromal composition, and acoustic accessibility.

Sonocavitation therapy can effectively remodel pancreatic tumor microenvironments by relieving tumor hypoxia and reversing immunosuppressive status, thereby improving the therapeutic resistance of pancreatic cancer. In addition to modulating the pathological microenvironment, regulating tumor genetic expression is also a fundamental approach to inhibit malignant proliferation and reverse drug resistance.

Gene delivery for pancreatic cancer

Sonocavitation therapy provides an efficient platform for the precise delivery of nucleic acids due to the sonoporation effect, significantly enhancing the cellular internalization efficiency of plasmids, siRNA, miRNA inhibitors, and other nucleic acids, showcasing unique advantages in pancreatic cancer treatment through modulating gene expression or silencing genes. Wang et al. confirmed that in the human pancreatic cancer cell line PANC-1, ultrasound irradiation (1.0 MHz, 1.0 W/cm²) combined with SonoVue microbubbles increased the transfection efficiency of green fluorescent protein (GFP) plasmid to 21.4% ± 3.16%, nearly 7 times higher than that of the plasmid-only group (2.9% ± 0.45%), with cell viability maintained above 91% and no significant toxicity.⁶⁹ In this process, microbubbles act as cavitation nuclei, lowering the ultrasound energy threshold required and enhancing cell membrane permeability, while avoiding the immunogenicity of viral vectors and the tissue damage associated with electroporation.

Recently, this technology has been further integrated with multifunctional nanocarriers for synergistic gene modulation and chemotherapy. Lin et al. constructed a co-delivery system using dendritic gold nanoparticles (Au DENPs) loaded with gemcitabine (Gem) and a miR-21 inhibitor (miR-21i) via electrostatic and hydrophobic interactions.⁴⁹ Results showed that this system increased cellular uptake efficiency of miR-21i by 13 times in SW1990 cells under USMB mediation (0.4 W/cm², 1 MHz), and the IC₅₀ value for combination therapy was 82 times lower than that of free Gem. In vivo, tumor volume significantly shrunk after combination therapy, accompanied by improved blood perfusion and extended median mouse survival to 120 days. The study also found that the combination therapy activated the mitochondrial apoptosis pathway by regulating apoptosis-related genes such as PTEN and Bax, thereby enhancing chemotherapy sensitivity. This study highlighted that optimizing ultrasound parameters is key to improving efficacy. Research suggests that ultrasound with power densities of 0.4–1.0 W/cm² and duty cycles of 10%–20% strikes a balance between stable cavitation and cytotoxicity. Otherwise, too low power (e.g., 0.2 W/cm²) fails to effectively induce sonoporation, and too high power (e.g., 0.6 W/cm²) increases the risk of cell death.

Another 2024 study employed polyethyleneimine (PEI) nanogels as carriers, loading small interfering RNA (siRNA) targeting periostin for gene silencing, while also encapsulating the chemotherapeutic drug Gemcitabine to form a combined “gene modulation+chemotherapy” therapeutic system.⁴³ Sonocavitation therapy, through ultrasound irradiation causing microbubble cavitation, created transient, reversible “pores” in cell membranes, effectively breaking through the dense interstitial barrier of pancreatic cancer and significantly enhancing the tumor cell uptake efficiency of PEI nanogels. Experimental results showed that ultrasound of 0.4 W/cm² increased the uptake rate of FITC-labeled carriers to 19.66% ± 2.64% while ensuring high cell viability (97.38% ± 1.92%). In vitro and in vivo results both confirmed that the USMB-assisted PEI nanogel delivery system efficiently downregulated periostin expression, blocked downstream signaling pathways, and promoted pancreatic cancer cell apoptosis. Combined with the sustained release of Gemcitabine, it significantly inhibited the proliferation of SW1990 and Panc-1 pancreatic cancer cells. In nude mouse xenograft models, tumor volume and weight were significantly reduced compared to the chemotherapy-alone group (p < 0.001). Ultimately, by enhancing drug delivery efficiency and reversing chemotherapy resistance, it laid the foundation for personalized gene modulation treatment strategies for pancreatic cancer.

In summary, gene delivery mediated by sonocavitation therapy has demonstrated safety and targeting in pancreatic cancer therapy. It is particularly suitable for overcoming the tumor stromal barrier, improving the hypoxic microenvironment, and reversing chemotherapy resistance, providing an efficient synergistic strategy for pancreatic cancer treatment. It is also worth noting that research on pancreatic cancer treatment using sonocavitation has been limited to nucleic acid delivery, rather than combining it with gene editors such as CRISPR-Cas systems.⁷⁰ Gene editing technology allows precise changes to an organism’s genome, effectively controlling genetic information, and has made significant progress in treating human genetic diseases and cancer.^71,72 Although related studies have been conducted on models such as breast cancer and lung cancer,^73,74 evidence for true sequence editing requiring programmable editors such as CRISPR-Cas systems in combination with sonocavitation therapy in pancreatic cancer remains limited. In addition, while sonocavitation therapy has a huge role in gene delivery, ultrasound exposure can also introduce risks through excessive inertial cavitation, such as damaging membranes and DNA, altering plasmid integrity, or increasing nonspecific uptake. Accordingly, future studies should quantify strand breaks, plasmid integrity, off-target effects, and editing fidelity using dose-escalation and sham-ultrasound controls before clinical translation.

Sonocavitation-assisted gene delivery achieves precise gene modulation and synergistic chemotherapy, exhibiting great potential in overcoming pancreatic cancer drug resistance and improving therapeutic outcomes. Beyond drug and gene delivery, ultrasound-microbubble technology also exerts versatile anti-tumor effects through multiple expanded therapeutic modalities.

Others

Enhancement of radiotherapy efficacy

Radiotherapy (RT) is a standard treatment for various cancers, killing cancer cells by inducing DNA damage and oxidative stress. However, pancreatic cancer is characterized by a dense stromal compartment and profound hypoxia, which contribute to its radioresistance and consequently lead to unsatisfactory outcomes following RT.^75,76 Sonocavitation therapy has emerged as a promising strategy to enhance radiotherapy by generating ROS, alleviating hypoxia, and disrupting tumor vascular endothelium.^51,77,78 This approach offers a novel direction to overcome radioresistance in pancreatic cancer. The Nakayama group conducted an in vitro study using two pancreatic cancer cell lines, MIAPaCa-2 and PANC-1, in which they combined Sonazoid microbubbles with 150 kVp X-ray irradiation.⁵¹ Additionally, hydrogen peroxide-modified titanium dioxide nanoparticles (TiOxNPs) were introduced to investigate potential synergistic effects. After optimizing ultrasound parameters, a frequency of 4.0 MHz, a mechanical index (MI) of 1.0, and an exposure duration of 3 min were selected, achieving a microbubble cavitation rate exceeding 90%, thereby establishing a foundation for subsequent experiments. Cellular experiments revealed that at a low concentration (0.01% v/v), sonocavitation therapy combined with 6 Gy X-ray irradiation significantly enhanced radiosensitivity in MIAPaCa-2 cells, resulting in reduced cell viability and a 1.24-fold increase in ROS levels. In contrast, PANC-1 cells exhibited only mild sensitization under 3 Gy X-rays combined with 0.01% USMB, with no significant effect observed at 6 Gy. Although a high microbubble concentration (0.4% v/v) reduced basal viability in both cell lines, it failed to produce a radiosensitizing effect, likely due to substantial premature cell death. Notably, the study did not observe synergistic sensitization between USMB and TiOxNPs. It was hypothesized that ineffective coupling between nanoparticles and microbubbles might have caused resealing of USMB-induced cell membrane pores before nanoparticle internalization, thereby limiting sensitization efficacy. These findings indicate that the radiopotentiating effects of sonocavitation therapy are influenced by cell type, microbubble concentration, and radiation dose, and more preclinical studies are still needed to ensure the safety and efficacy of treatment.

Acoustic cluster therapy (ACT)

Conventional ultrasound contrast microbubbles (e.g., SonoVue) exhibit limited drug delivery efficiency due to their small size (2–5 μm) and short vascular contact time. In contrast, ACT utilizes microbubble-microdroplet clusters to form larger composite structures capable of generating enhanced bioeffects and has been demonstrated to significantly improve the delivery and antitumor efficacy of various chemotherapeutic agents in tumors.^79,80,81 A key feature of ACT technology is the formation of clusters composed of negatively charged microbubbles (perfluorobutane, H-EPS shell) and positively charged microdroplets (perfluoromethylcyclopentane, DSPC shell), which are activated via a two-step ultrasound sequence. As shown in Figure 5, first, high-frequency (2–10 MHz) ultrasound induces a phase transition of the microdroplets, resulting in the formation of large bubbles (20–30 μm) that become lodged within the tumor microvasculature. Subsequently, low-frequency (0.5 MHz) ultrasound triggers oscillatory activity in these large bubbles, enhancing vascular permeability. Compared to conventional microbubbles, which are small and flow freely within circulation, the ACT platform offers two major advantages owing to the size and two-step activation process of its bubbles: (1) the large bubbles transiently lodge in tumor microvessels, prolonging vascular contact time to 5–10 min; and (2) the low-frequency-induced oscillations produce stronger biomechanical effects, disrupting the vascular barrier more effectively and promoting deeper penetration of drugs into the tumor parenchyma. Multiple studies have shown that ACT combined with low-dose chemotherapy achieves efficacy equivalent to or superior to that of high-dose standard chemotherapy, while significantly reducing systemic toxicity. For instance, in models of pancreatic, prostate, and breast cancer, ACT combined with agents such as docetaxel resulted in potent tumor growth inhibition.^82,83,84

Figure 5.

The treatment mechanism of ACT involves the use of dual ultrasound irradiation to enhance drug delivery

Reproduced with permission.⁵⁰ Copyright 2022, Elsevier.

Ng Serina and colleagues evaluated ACT in combination with chemotherapy (either nab-paclitaxel + gemcitabine or liposomal irinotecan—both first-line regimens) using a patient-derived xenograft (PDX) mouse model of pancreatic cancer that retained the stromal components of the original tumor and carried a K-RASG12V mutation.⁵⁰ The results demonstrated that ACT increased the probability of complete tumor remission by 7.2-fold (95% CI: 1.622–51.77, p = 0.0195). In terms of tumor growth inhibition, chemotherapy alone yielded an 11% inhibition rate, whereas combining ACT with nab-paclitaxel and Gem resulted in 89% inhibition (q = 0.0075), and with liposomal irinotecan reached 80% inhibition (q = 0.0429), without significant adverse effects such as body weight loss. These findings provide compelling preclinical evidence supporting clinical translation. Currently, a Phase I clinical trial (ACTIVATE study) of ACT for metastatic pancreatic cancer is underway, which aims to validate these results. It is important to note that ACT is not merely an enhancement of conventional microbubble technology, but represents a distinct mechanism and a more potent platform strategy. Through its unique cluster composition and phase-change mechanism, ACT aims to overcome the limitations in delivery efficiency and controllability associated with traditional ultrasound microbubble techniques.

Crosstalk among cavitation, mechanotransduction, and TME remodeling

The therapeutic mechanisms described above are not independent. Stable cavitation can first increase endothelial and cellular permeability, improving delivery of drugs or nanocarriers; subsequent inertial cavitation can produce stronger shear, microjets, mechanical disruption, and ROS. Improved perfusion and stromal relaxation then increase delivery of additional microbubbles, creating a feedforward cascade. In parallel, mechanical injury and immunogenic cell death release tumor antigens and DAMPs, while hypoxia relief and vascular normalization may improve immune-cell access. The resulting synergy is plausible but not yet quantitatively established in pancreatic cancer, because most studies test one mechanism at a time and use different acoustic and biological endpoints. Future experiments should measure delivery, perfusion, stromal stiffness, oxygenation, immune composition, and tumor control within the same protocol.

The above chapters systematically elaborate on the core mechanisms of sonocavitation therapy for pancreatic cancer, including barrier penetration, controlled drug release, tumor microenvironment remodeling, gene delivery modulation, and the intrinsic crosstalk between cavitation mechanics and biological responses. Despite the well-established preclinical theoretical basis and multifunctional therapeutic advantages, the practical clinical value of sonocavitation therapy relies on validating its safety, feasibility, and efficacy in human patients. Therefore, this chapter focuses on summarizing the latest clinical translation progress, existing challenges, and future optimization directions of sonocavitation therapy for pancreatic cancer treatment.

Clinical translation progress of sonocavitation therapy for pancreatic cancer treatment

Pancreatic cancer, characterized by its high malignancy and strong resistance to chemotherapy, has a five-year survival rate of less than 10%, with conventional treatments offering limited efficacy. In recent years, Sonocavitation therapy-enhanced sonoporation has emerged as a non-invasive physical sensitization strategy, demonstrating potential in animal studies to improve drug delivery and modulate the TME. By leveraging ultrasound-triggered microbubble cavitation to increase vascular permeability and enhance drug penetration, this approach has achieved progressive advances in clinical translation (Table 3).

Table 3.

Clinical trials of Sonocavitation therapy for pancreatic cancer treatment

Patients	Drug&Microbubbles	US device	US parameters	Efficacy	Adverse events	Reference
5 Stage II/III or Stage IV unresectable pancreatic cancer patients	Gemcitabine (Gemzar®) SonoVue®	GE LOGIQ 9 + 4C curved array	1.9 MHz; MI 0.2; DC 1%; 31.5 min per session; 0.5 mL bolus every 3.5 min	tumor size shrinks or growth slows; mean chemotherapy cycles: 16 vs. historical control average of 9 cycles.	no additional toxicities	Kotopoulis et al.85
10 Stage III or Stage IV unresectable pancreatic cancer patients	Gemcitabine (Gemzar®) SonoVue®	GE LOGIQ 9 + 4C curved array	1.9 MHz; MI 0.2; DC 0.3%; 31.5 min per session; 0.5 mL bolus every 3.5 min	tumor size reduction observed in 5/10 patients; MST: 17.6 months vs. historical control 8.9 months (p = 0.011); mean chemotherapy cycles: 13.8 ± 5.6 vs. historical control of 8.3 ± 6.0 (p = 0.008).	no additional toxicity or increased side effects	Dimcevski et al.86; NCT01674556
1 T4NIM1 pancreatic cancer	Gemcitabine+S-1 SonoVue®	GE Logiq E9 + C1-5 curved array	2∼4 MHz; MI 0.8; 30 min per session; 1 mL bolus every 4 min	PD with 36 days PFS; no new progressive lesions	only grade 1–2 AEs (fever, myelosuppression);	Wang et al.87; NCT02233205
78 Stage III/IV pancreatic cancer patients	Nab-paclitaxel+Gemcitabine; Nab-paclitaxel+S-1; FOLFIRINOX; SonoVue®	VINNO70 (modified) + S1-8C curved array	1.7 MHz; MI 0.2–0.3; DC 0.67%; 8.5 cycle pulses; 30 min per session; 1 mL bolus every 5 min	MST: EG 9.10 months vs. CG 6.10 months (p = 0.037); DCR: EG 73.68% vs. CG 42.50% (p = 0.005); duration of ECOG status <2: EG 7.20 months vs. CG 5.00 months (p = 0.029)	comparable AE rates between groups; only mild diarrhea (13.16%) higher in EG group; no serious new AEs.	Han et al.88; ChiCTR; 2100044721
19 Stage III or IV unresectable pancreatic cancer patients	Nab-paclitaxel+Gemcitabine; Gemcitabine FOLFIRINOX; SonoVue®	custom dual-frequency probe (4.5 MHz imaging/0.35 MHz therapy) + Verasonics Vantage 256	therapy 0.35 MHz; MI 0.5; pulse length 2.9 ms (1000 cycles); DC 0.04%; 35 min per session; 1 mL bolus every 3.5 min	MST: EG 356 days vs. CG 299 days (NS)	no additional serious AEs; no skin or adjacent tissue injury.	Haram et al.89; NCT04146441
21 Stage III or IV advanced pancreatic cancer patients	Nab-paclitaxel+ Gemcitabine; Nab-paclitaxel+ S-1 FOLFIRINOX; SonoVue®	VINNO70 + S1-8C curved array	1.7 MHz; MI 0.25–0.3; DC 2 s on/1 s off; 8.5 cycle pulses; PRF 2000 Hz; 30 min per session; 4 mL diluted bolus (1/4) every 5 min	Potential tumor perfusion enhancement	no ≥ grade 3 AEs; all events grade 1–2 and self-limiting or manageable.	Deng et al.90

EG, experimental group; CG, control group; MST, median survival time; PD, progressive disease; PFS, progression-free survival; DCR, disease control rate; AE, adverse event; NS, no statistical significance; MI, mechanical index; DC, duty cycle.

In 2013, Professor Kotopoulis and co-authors first reported a case series of five patients with unresectable pancreatic cancer treated with Gem in combination with SonoVue microbubbles under ultrasound exposure.⁸⁵ The results showed that the average number of chemotherapy cycles increased from 9 (in a historical control group) to 16 per patient. Tumor diameter transiently decreased to 80% of the original size in two patients and permanently reduced to 70% in one, without additional toxicity. This pilot study not only confirmed the safety of using diagnostic ultrasound (mechanical index, MI = 0.2) combined with microbubbles but also suggested a trend toward improved ECOG (Eastern Cooperative Oncology Group) performance status and survival benefit. In 2016, the same team conducted a Phase I clinical trial involving 10 patients, applying the same treatment protocol as in 2013.⁸⁶ The study further verified that the combination of ultrasound-stimulated microbubbles and Gem did not increase adverse reactions (such as myelosuppression or abdominal pain). Notably, the median overall survival (OS) was extended from 8.9 months in the control group to 17.6 months (p = 0.011), supporting the clinical safety and preliminary efficacy of this strategy.

Optimization of treatment parameters is crucial for clinical application. In 2018, Wang et al. from Peking University Cancer Hospital conducted a study including 12 patients with gastrointestinal malignancies, among whom one had pancreatic cancer, and another had pancreatic cancer with liver metastasis.⁸⁷ By adjusting the mechanical index (ranging from MI 0.4 to 1.0), they found that an MI ≤ 1.0 was associated with a favorable safety profile. Although the patient with pancreatic cancer (treated at MI = 0.8) was evaluated with progressive disease (PD) after the first treatment cycle without achieving partial response (PR) or stable disease (SD), the progression was confined to the primary site without new metastases. No adverse events directly attributable to the USMB treatment were observed; all reported toxicities were grade 1–2 chemotherapy-related events (e.g., nausea, fatigue), which were manageable with supportive care. In contrast, the patient with liver metastasis from pancreatic cancer (treated at MI = 0.4) achieved a partial response, with a progression-free survival of 168 days, suggesting that an MI of 0.4 may serve as a reference for future studies.

Sonoporation mediated by sonocavitation therapy enhances therapeutic efficacy primarily through stable cavitation-generated shear forces, which increase tumor vascular permeability and ameliorate the hypoxic microenvironment in pancreatic cancer. In 2024, Liu and colleagues⁹⁰ analyzed quantitative contrast-enhanced ultrasound time-intensity curves before and after treatment and observed an increasing trend in peak intensity (PI) and area under the curve (AUC), which are parameters reflective of tumor perfusion, suggesting improved conditions for drug penetration. Regarding survival benefits, the same team conducted a multicenter, open-label, randomized controlled Phase II trial, enrolling 82 patients with stage III/IV pancreatic cancer (4 excluded), comparing sonochemotherapy (USMB combined with chemotherapy) versus chemotherapy alone.⁸⁸ The results demonstrated that the sonochemotherapy group exhibited significantly improved median survival time (MST) from 6.1 to 9.1 months (HR = 0.61, p = 0.037), disease control rate (DCR) from 42.5% to 73.7% (p = 0.005), and prolonged duration of preserved performance status (ECOG score <2) by 2.2 months (p = 0.029). Moreover, 50% of patients achieved tumor shrinkage, with some becoming eligible for surgery or radiotherapy, thereby further extending survival or enhancing treatment efficacy. The study utilized a modified diagnostic ultrasound system (Vflash mode, MI 0.2–0.3) with a weakly focused region of interest (ROI) covering the entire tumor and a peripheral 5–10 mm margin, enabling three-dimensional tumor coverage and ensuring comprehensive energy delivery. This approach provides a valuable reference for future clinical trials.

Despite these encouraging results, current research still faces controversies and challenges. In 2025, Haram et al. conducted a prospective randomized pilot trial to investigate the safety and feasibility of focused ultrasound (FUS) combined with microbubbles (MBs) for enhancing chemotherapy in inoperable patients with pancreatic cancer.⁸⁹ The trial enrolled 20 patients (1 excluded), randomized into either FUS+MBs combined with chemotherapy (n = 9) or chemotherapy alone (n = 10). Unfortunately, no significant differences were observed in tumor volume or overall survival (11.7 vs. 9.8 months, p > 0.05). This neutral result should not be interpreted simply as evidence that the approach is ineffective: the study was small, included heterogeneous advanced-stage patients, and may have been affected by treatment timing, tumor accessibility, acoustic shadowing, and differences in ultrasound frequency, intensity, duty cycle, exposure duration, and microbubble dose. Fibrotic stroma and variable perfusion can further reduce microbubble delivery, while stage, tumor location, prior therapy, and baseline performance status can dilute a treatment effect. The contrast with the positive multicenter trial therefore highlights the need for protocol harmonization and adequately powered stratified trials rather than supporting a single efficacy conclusion.

Taken together, the clinical evidence has progressed from proof-of-concept case studies to randomized evaluation, but it remains heterogeneous rather than uniformly conclusive. The positive phase II findings and the neutral pilot trial should be interpreted in light of differences in enrollment, disease stage, tumor accessibility, chemotherapy backbone, and acoustic parameters. Sonocavitation therapy is therefore best regarded as a promising adjunct for overcoming chemotherapy resistance in PDAC. Future efforts should focus on large-scale, standardized Phase III trials to consolidate its clinical value and explore biomarker-guided precision treatment paradigms.

Challenges and perspectives

Limitations in preclinical models

Systematic biases in model selection during the preclinical development of therapeutic agents for pancreatic cancer pose a major obstacle to clinical translation. Currently, over 60% of studies still utilize subcutaneous heterotopic xenograft models (Tables 1 and 2), which significantly diverge from the unique clinicopathological features of pancreatic cancer. The native pancreatic tumor microenvironment is characterized by extremely dense desmoplasia—a physical and biochemical barrier formed primarily by activated pancreatic stellate cells. This barrier severely impedes drug delivery and fosters an immunosuppressive milieu. However, subcutaneous models fail to recapitulate this complex ecosystem, leading to substantial overestimation of the efficacy of candidate drugs, particularly chemotherapeutic and targeted agents. This discrepancy undoubtedly contributes to the high failure rate of many drugs in subsequent clinical trials.

An even more critical challenge is the notable lack of metastasis models—only about 5% of studies incorporate such models. Pancreatic cancer is highly aggressive, with up to 80% of patients presenting with local invasion or distant metastases (mainly liver metastases) at diagnosis. Mortality is predominantly driven by metastatic disease rather than the primary tumor. The scarcity of relevant models means that most drug candidates are not evaluated in contexts that reflect the most lethal aspects of the disease. Moreover, approximately 78.9% use immunodeficient mice (Tables 1 and 2), thereby overlooking the highly immunosuppressive microenvironment typical of pancreatic cancer. Key cellular components such as myeloid-derived suppressor cells (MDSCs) and tumor-associated macrophages (TAMs), which play crucial roles in promoting tumor progression and immune evasion, are absent in these models. “Effective” outcomes derived from immunocompromised hosts fail to indicate whether a therapeutic agent can modulate the tumor immune microenvironment or reverse immunosuppression. These model deficiencies contribute to a significant disconnect between preclinical efficacy data and actual patient response rates.

Development and translational hurdles of formulation strategies

In translational research on ultrasound microbubble-mediated drug delivery for pancreatic cancer, a major obstacle lies in the disconnect between formulation strategies and clinical practicality. Currently, up to 42.8% of preclinical studies favor the use of drug-loaded microbubbles, where chemotherapeutic agents such as gemcitabine or paclitaxel are physically adsorbed or chemically conjugated onto microbubble surfaces (Tables 1 and 2). This approach allows localized drug release upon ultrasound-triggered microbubble destruction, demonstrating promising efficacy amplification and toxicity reduction in experimental settings. However, the clinical translation of such advanced combination products faces a frequently underestimated yet critical challenge: the requirement for comprehensive additional toxicological evaluations. Since the drug-microbubble complex constitutes a new entity, regulatory agencies (e.g., FDA, NMPA) mandate full non-clinical safety assessments, including evaluations of its novel pharmacokinetic properties, biodistribution, systemic and local toxicity, and immunogenicity. This process is not only costly and time-consuming but also significantly heightens the uncertainty and barriers to clinical application.

In sharp contrast, all current clinical trials have adopted a pragmatic co-administration strategy of simultaneous injection of commercially available microbubbles (e.g., SonoVue) and standard chemotherapy to circumvent these regulatory complexities. This method relies on the transient coexistence of both agents in circulation and simultaneous ultrasound application at the tumor site to achieve synergistic effects. While this workaround avoids regulatory hurdles, it introduces a critical scientific limitation: inherent pharmacokinetic (PK) mismatch. Preclinically optimized drug-loaded microbubbles ensure synchronized arrival, activation, and release of the drug at the target site. In contrast, co-injected formulations, each with distinct molecular sizes, structures, and surface properties, exhibit significantly different PK profiles, including blood circulation half-lives, volume of distribution, and clearance rates. For instance, microbubbles have a very short half-life (on the order of seconds to minutes), while small-molecule chemotherapeutic agents circulate much longer. This temporal disparity results in a suboptimal microbubble-to-drug concentration ratio at the moment of ultrasound triggering, undermining the translational relevance of promising preclinical results obtained with loaded microbubbles.

Technical standardization issues

In the field of ultrasound microbubble-mediated drug delivery for pancreatic cancer treatment, significant heterogeneity in ultrasound irradiation parameters and inadequate reporting have emerged as critical bottlenecks hindering technical reproducibility, data comparison, and ultimately, clinical translation. The severity of this issue is evident at both preclinical and clinical levels. Over 71.4% of preclinical studies utilize a wide variety of laboratory-based ultrasound devices rather than standardized clinical or preclinical machines (Tables 1 and 2). The calibration methods and output performance of these devices vary considerably, leading to striking heterogeneity in core ultrasound parameters. This variability makes it difficult to compare results across different studies. Moreover, “positive” outcomes obtained under different parameter combinations may arise from entirely distinct biological mechanisms and safety profiles, rendering any attempts at meta-analysis or conclusive synthesis largely uninterpretable.

A more detrimental aspect is the propagation of this technical ambiguity into clinical research. Many published clinical trial protocols or reports suffer from insufficient description of crucial parameters, such as the ultrasound device model, transducer frequency, calibrated output of acoustic power/pressure, irradiation duration, duty cycle, and real-time monitoring methods. This lack of transparency makes it nearly impossible for other research centers to precisely replicate the treatment protocols. Consequently, the translation of promising preclinical findings into reliable clinical regimens is severely hampered, and comparative effectiveness assessments across different trials become unfeasible. This significantly delays the standardization and clinical adoption of the technology in pancreatic cancer treatment.

Future directions: Precision medicine and combination strategies

To advance sonocavitation therapy into clinical practice for pancreatic cancer, future research must achieve breakthroughs across multiple dimensions, including model systems, technical standardization, mechanistic investigation, and combination strategies. Firstly, it is imperative to systematically validate the generality and safety of treatment protocols in more predictive models, such as patient-derived xenograft (PDX) models and large animal models (e.g., porcine). PDX models better preserve the heterogeneity, fibrotic stroma, and molecular characteristics of human pancreatic cancer, while large animal models provide an indispensable platform for optimizing ultrasound irradiation parameters under anatomical and physiological conditions closer to those of humans. A central task is the standardization of ultrasound parameters. Future parameter settings should move away from a “one-size-fits-all” approach toward personalized adjustment based on imaging assessments of individual tumor fibrosis (e.g., ultrasound elastography or MRI). Concurrently, quantitative monitoring of intratumoral drug concentration via techniques such as microdialysis or in vivo imaging must be implemented to establish a clear concentration-efficacy relationship, thereby providing a solid foundation for precise “acoustic dosing”.

Secondly, the core issue of PK mismatch in clinical co-administration protocols must be addressed. The key resides in optimizing the timing of ultrasound application. By leveraging the unique PK profiles of chemotherapeutic agents (e.g., gemcitabine, with a plasma half-life of only 5–20 min), the time window for ultrasound triggering can be guided to synchronize with peak drug concentration, thereby maximizing drug capture at the target site. Technologically, efforts should accelerate the development and application of 3D ultrasound probes or MRI-guided systems to achieve precise stereotactic targeting, real-time energy monitoring, and treatment evaluation for deep-seated organs such as the pancreas, ensuring controllability and reproducibility.

Finally, overcoming the formidable physical and immune barriers in pancreatic cancer will inevitably rely on combination therapies. Future studies should actively explore synergies with immune checkpoint inhibitors (e.g., PD-L1 inhibitors). Ultrasound cavitation can induce ICD in situ and remodel the immunosuppressive microenvironment, turning “cold” tumors “hot” and thereby enhancing immunotherapeutic efficacy. In parallel, combination strategies with stroma-modulating agents (e.g., PEGPH20 hyaluronidase) should be rigorously evaluated. By degrading physical barriers and synergistically enhancing drug penetration, such approaches may offer promising comprehensive solutions to overcome therapeutic resistance in pancreatic cancer.

Conclusion

Sonocavitation therapy offers a mechanistically attractive approach to address several central barriers in pancreatic cancer, including poor perfusion, dense stroma, hypoxia, and limited intratumoral drug exposure. However, the current evidence does not yet establish a definitive clinical benefit: preclinical results are often model-dependent, clinical studies remain small and heterogeneous, and acoustic parameters, microbubble formulations, treatment timing, and response assessments are not standardized. The field also lacks validated biomarkers to identify patients most likely to benefit and sufficiently long follow-up to define delayed toxicity or durable disease control. Future studies should therefore combine clinically representative orthotopic, metastatic, PDX, organoid, and immunocompetent models with harmonized acoustic dosimetry, real-time treatment monitoring, pharmacokinetic and biodistribution analyses, and transparent reporting of safety endpoints. Rational combinations with chemotherapy, radiotherapy, immune checkpoint blockade, gene therapy, or stromal-modulating agents should be tested in biomarker-defined populations rather than empirically. Ultimately, adequately powered multicenter randomized trials, supported by reproducible manufacturing and regulatory strategies, will be required to determine whether sonocavitation therapy can improve survival and quality of life beyond current standards of care.

Acknowledgments

This work was financially supported by the Beijing Natural Science Foundation (7244525) and the National Natural Science Foundation of China (82402290).

Declaration of interests

The authors declare no conflict of interests.