Abstract
Ischemic stroke and thrombotic vascular occlusion remain leading causes of mortality and disability worldwide, yet current therapies operate largely in an open-loop paradigm, delivering thrombolytics or mechanical intervention without adaptive feedback on clot state or treatment response. This review reframes thrombolysis as a dynamic, controllable biological process and introduces a systems-engineering perspective built around the Sense → Target → Lyse → Report framework for closed-loop nanomedicine. We synthesize advances across nanotechnology, thrombosis biology, and bioresponsive materials to examine how next-generation nanosystems can detect clot-specific biochemical and biomechanical cues, localize to thrombi under physiological flow, actuate controlled lytic activity, and provide real-time reporting of therapeutic progress. We first characterize the thrombus as a heterogeneous, evolving immunothrombotic structure. It comprises platelet-rich shells, red blood cell (RBC)-dense cores, and microdomains enriched in neutrophil extracellular traps (NETs), each presenting distinct molecular and mechanical signatures that can serve as sensing handles. We then analyze emerging stimulus-responsive nanoplatforms activated by thrombin, reactive oxygen species (ROS), shear, or pH, alongside biomimetic and flow-optimized targeting strategies designed to overcome washout and penetration barriers. Particular emphasis is placed on theranostic systems that integrate imaging and therapy, laying the foundation for feedback-guided intervention. To organize this rapidly evolving field, we propose Closed-Loop Readiness Levels (CLRL) as a translational framework that classifies thrombolytic technologies along a five-tier scale. The scale runs from open-loop systems with no feedback (CLRL-0) to fully autonomous, self-regulating nanosystems that adapt therapy in real time and terminate it once reperfusion is achieved (CLRL-4). Across experimental models, closed-loop concepts show promise in improving spatial precision, reducing systemic exposure, and adapting lytic intensity to clot resistance. However, key barriers remain, including hemodynamic complexity, protein corona effects, sensor specificity, and integration of reporting with autonomous control. Intermediate levels capture the progressive integration of stimulus-triggered activation, thrombus targeting, therapeutic actuation, and reporting before full autonomous control is reached. By unifying disparate advances under a control-systems paradigm, this review positions closed-loop thrombolysis as a transformative direction in stroke therapy, with the potential to shift treatment from static dosing toward responsive, intelligent, and patient-specific vascular intervention.
Graphical Abstract

This graphical abstract outlines the limitations of current ischemic stroke treatments (IV tPA and mechanical thrombectomy), introduces shell-core nanovesicles that respond to clot microenvironmental signals (pH, thrombin, ROS, stiffness), describes their "sense-target-lyse-report" closed-loop strategy, and highlights key translational considerations for clinical application
Supplementary Information
The online version contains supplementary material available at 10.1186/s12951-026-04774-8.
Keywords: Stroke, Ischemic; Thrombolytic therapy; Nanoparticles; Drug delivery systems; Thrombosis; Theranostic nanomedicine
Introduction
Ischemic stroke remains a devastating disease and a leading cause of death and long-term disability worldwide [1]. Each year, on the order of ten million people suffer an ischemic stroke, and thrombosis-related events, including stroke and myocardial infarction, account for roughly one in four deaths globally [2–4]. Survivors often face severe neurological impairment, nearly 40% of stroke patients are left functionally-dependent six months post-stroke [4], underscoring the immense personal and societal burden. The economic costs are staggering, approaching an estimated $890 billion per year (0.7% of global GDP) in stroke-related healthcare and lost productivity [1, 5]. Despite advances in prevention and acute care, the overall stroke burden has actually grown over the past few decades, particularly in low- and middle-income countries, due to population aging and inadequate access to therapies [5]. This reality highlights the necessity for innovative approaches that can surmount the limitations of current treatments and better address the complexity of thrombotic disease pathology.
A key challenge is the dynamic pathophysiology of thrombus formation in stroke. Ischemic stroke thrombi are heterogeneous, evolving structures with platelet/fibrin-rich, red-cell-rich, and neutrophil extracellular trap-enriched regions. This heterogeneity creates spatially variable resistance to lysis and provides potential sensing handles [6–8].Retrieved stroke clots often show a layered shell-core architecture, with platelet/fibrin-rich and red-cell-rich regions that differ in permeability and response to lysis (See next sections for detailed description). Clot composition also varies by stroke etiology, and continuing platelet activation or fibrin deposition during attempted thrombolysis can contribute to early re-occlusion. These dynamic features help explain why uniform thrombolytic strategies often fail.
The standard of care for acute ischemic stroke focuses on rapid revascularization of the occluded artery to salvage ischemic brain tissue [9]. Intravenous thrombolysis with tissue plasminogen activator (tPA) remains the only globally approved pharmacological treatment for acute stroke [10]. Administered within the first few hours of symptom onset, tPA enzymatically cleaves fibrin, dissolving the clot and restoring blood flow [10]. In parallel, for large artery occlusions, mechanical thrombectomy using endovascular devices has become a cornerstone therapy in the last decade, physically extracting the clot and dramatically improving outcomes in eligible patients [11]. Adjunctive medical management includes antithrombotic drugs: antiplatelet agents (such as aspirin) to prevent further platelet aggregation, and anticoagulants (such as heparin or direct oral anticoagulants) especially for cardioembolic stroke prevention [12]. These interventions, used in combination, have saved many lives and improved functional recovery for countless stroke patients. However, they come with significant limitations that leave a large proportion of patients without benefit.
The most prominent limitation of tPA thrombolysis is its narrow therapeutic time window. To be effective and relatively safe, IV tPA must be given within 4.5 h of stroke onset, a requirement that most stroke victims do not meet [12]. Many patients either do not recognize stroke symptoms or cannot access stroke centers in time. Consequently, only a small minority of all ischemic stroke patients receive thrombolytic therapy. Even in high-performing stroke systems, it is estimated that barely 7–10% of acute stroke patients are treated with tPA, with the rest arriving too late or having contraindications [13]. The strict inclusion criteria also reflect another major limitation: bleeding risk. By design, tPA promotes fibrinolysis systemically, which can precipitate hemorrhagic complications. Approximately 5% of tPA-treated stroke patients suffer a symptomatic intracerebral hemorrhage as a direct side effect of thrombolysis [14]. This risk is low enough to be outweighed by tPA’s benefits on average, but it instills substantial caution in clinical practice [14]. Fear of intracerebral hemorrhage contributes to underutilization of tPA and motivates the search for safer fibrinolytic strategies [14].
Beyond safety and timing issues, efficacy limitations of tPA are well recognized. Large, fibrin-rich clots, such as proximal internal carotid or proximal middle cerebral artery occlusions, often do not respond to IV thrombolysis: tPA’s recanalization rates for such large clots are relatively poor [15]. Dense platelet-rich thrombi are also less susceptible to fibrinolysis [16]. Even when tPA succeeds in partially dissolving a clot, it may not completely reopen the artery, or the artery may re-occlude shortly afterward due to ongoing thrombogenic stimuli. Mechanical thrombectomy has circumvented some of these issues by physically removing large clots and extending the treatment time window out to 24 h in selected cases. Yet thrombectomy is an option only for a subset of large occlusions in accessible arteries, and it still faces challenges. In about 20% of attempts the artery cannot be fully recanalized. Even successful reperfusion does not guarantee a good outcome if downstream microthrombi or reperfusion injury occur [17–19]. Moreover, thrombectomy is an invasive procedure available only at specialized centers; many patients globally have no access to it in time [20].
Finally, all current therapies are relatively imprecise in their mechanism of action. IV tPA is given systemically at a fixed dose, unable to distinguish between dissolving a pathologic occlusion and, for example, disrupting a beneficial hemostatic clot elsewhere [20]. Mechanical thrombectomy, while targeted to the clot’s location, relies on the operator’s skill and can cause endothelial trauma or distal embolization [4, 20]. In essence, today’s stroke treatments operate in an “open-loop” fashion: we administer drug or device-based therapy without real-time feedback control, hoping it will work, and we often cannot intervene adaptively if the clot’s response is suboptimal. These limitations set the stage for a new therapeutic paradigm that can address the spatiotemporal precision that is lacking in current open-loop approaches.
In recent years, researchers have increasingly turned to nanomedicine and bioengineering to develop smarter, more targeted thrombolytic therapies [21–25]. The goal is to deliver clot-busting treatment with precision: directly to the thrombus and only when needed, thereby maximizing efficacy and minimizing collateral damage. A variety of innovative nanoparticle-based and physical intervention strategies for thrombolysis have been explored. One major approach is the use of targeted nanocarriers to ferry thrombolytic drugs to the clot site [6]. These nanocarriers tend to accumulate at the clot and prolong the drug’s circulation half-life, thereby increasing the fraction of tPA that reaches the thrombus while reducing systemic exposure [6]. Preclinical studies have shown that such fibrin-targeted nanoparticles can achieve effective clot lysis with lower dosing compared to free tPA [26–28]. Platelet-biomimetic carriers extend this principle: annexin V-modified and cryo-shocked platelet nanomedicines exploit native platelet adhesion to home to thrombi and concentrate therapy at the occlusion [29, 30].
Beyond biochemical targeting, studies have been conducted to design carriers that respond to specific stimuli in the clot microenvironment or external triggers. These so-called “smart” or stimuli-responsive systems remain inert during circulation and then activate under defined conditions. For instance, nanoparticles have been engineered to release their thrombolytic payload only in the presence of high local concentrations of clot-associated enzymes [31–33] or under the acidic pH that develops in ischemic tissue [34, 35]. External physical triggers are also being harnessed: ultrasound-sensitive microbubbles and nanodroplets can be ruptured at the clot by an applied ultrasound field, resulting in a focused burst of thrombolytic drug and mechanical cavitation that enhances clot dissolution. Similarly, magnetically guided nanoparticles loaded with lytic agents have been developed, which can be steered to the thrombus using external magnetic fields [36–40]. These magnetic nanocarriers not only concentrate drug at the target site but in some designs can produce local heating or forces (so-called magnetomechanical disruption) to physically erode the clot when the magnetic field is applied.
In parallel, a new class of thrombolytic interventions blurs the line between drug delivery and mechanical therapy, leveraging advanced materials and microdevices. Light-activated nanomaterials (photothrombolysis) release thermal or photochemical energy on near-infrared irradiation to break down clots [41]. There are also experimental electrothrombectomy approaches, where microelectrical currents or plasmas delivered via microcatheters cause localized clot ablation, essentially a nano- or micro-scale version of mechanical thrombectomy. Perhaps most futuristic, teams are developing tiny microrobots and nanorobots that can navigate the bloodstream to a clot and mechanically attack it. Examples include magnetically actuated helical nanomachines (“nanopropellers”) that drill into fibrin networks, and self-propelled microbots that generate local fluid dynamics to fragment thrombi. These devices are in early stages but have shown an ability to penetrate clots and enhance reperfusion in animal models.
Many of the above strategies also lend themselves to combination with imaging functions. The concept of theranostic nanoparticles, single agents that provide both therapy and diagnostics, is emerging in thrombosis research. For example, iron oxide nanoparticles can serve as thrombolytic drug carriers and simultaneously act as MRI contrast agents, allowing real-time visualization of the clot and the drug delivery process [42–44]. Similarly, microbubbles can be dual-purpose: delivering tPA and enhancing ultrasound imaging of the occlusion. Such theranostic or “self-reporting” systems foreshadow a treatment that could continuously inform the clinician of its own progress.
These advances clearly illustrate the field’s momentum toward more focal and controllable thrombolysis. Yet, in surveying this burgeoning landscape, one notices that most studies tend to address each innovation in isolation. Typical reviews will enumerate various nanoparticles and biochemical targeting ligands, or catalog different device-based strategies as separate categories. What is often missing is a unifying framework that ties all these pieces together as parts of a single therapeutic system. In current research, a given nanoparticle might deliver a drug, or a microrobot might physically disrupt a clot, but rarely is the approach integrated with real-time feedback or adaptive control. In other words, nearly all existing thrombolytic technologies, even the most advanced nanoplatforms, still operate in an open-loop manner. The nanoparticle releases its drug cargo according to some preset trigger or degradation rate, but it does not “know” if the clot has dissolved. A microrobot may drill into a clot, but it does not adjust its activity based on clot composition changes or local biochemical signals. Conventional nanocarriers often suffer issues like premature drug leakage or sticking only to the clot surface, reflecting the difficulty of controlling drug deployment in space and time. Thus, despite incremental improvements, we have not yet departed from the paradigm of delivering a treatment and passively hoping for a favorable outcome.
The perspective of this review
In this review, we propose a new perspective that reframes thrombolysis as an interactive, closed-loop therapeutic process rather than a one-off event. We term this the “Sense → Target → Lyse → Report” paradigm of closed-loop nanomedicine (Fig. 1). The core idea is to build feedback control into thrombolytic therapy. The treatment can then autonomously sense the pathological environment, respond by delivering therapy in a targeted way, and report on the outcome, creating a continuous loop of assessment and action. In practical terms, a closed-loop nanosystem for stroke would work as follows: (1) Sense: The system can detect signature biochemical or biophysical cues associated with an active thrombus. (2) Target: Guided by these cues (or by external direction such as a magnetic field or focused ultrasound), the therapeutic nanoparticles or microdevices home to the thrombus and accumulate in the thrombus microenvironment. This targeting ensures that subsequent therapeutic actions occur chiefly at the clot, sparing normal vessels and tissues. (3) Lyse: Once engaged with the thrombus, the system activates the clot-dissolving therapy in a controlled, on-demand fashion. Drug-loaded nanocarriers could release thrombolytic enzymes only when bound within the clot or when a threshold level of a trigger is detected. Alternatively, a micro-scale device might start mechanically vibrating or heating to disrupt the clot only after confirming its presence in the occlusion. The key is that dosing or activity is automatically adjusted to the local conditions, a stark contrast to the fixed dosing of standard tPA. (4) Report: Concurrently, the closed-loop system provides a real-time readout of the treatment progress, essentially self-reporting its performance. This could be achieved by incorporating imaging agents that light up as the clot is lysed or sensors that measure parameters like clot size, flow restoration, or downstream biomarkers.
Fig. 1.

Schematic figure of the design of a DNA-origami based nanodevice embodying the Sense → Target → Lyse → Report concept. In the figure, adopted with permission from Yin et al. (2024), a flat DNA nanosheet is functionalized with multiple tPA enzymes (the thrombolytic “payload”) attached via special DNA triplex “locks.” These locks are engineered to sense the presence of thrombin, a clot-associated enzyme: when thrombin levels at the target site rise above a set threshold, the DNA locks open. Construction and characterization of the tPA–DNA nanodevice. a, Schematic of device assembly: (i) tPA–ssDNA conjugates were anchored onto a rectangular DNA nanosheet (90 nm × 60 nm × 2 nm) via hybridization with surface capture strands, with three triangularly arranged capture strands forming one binding site; (ii) addition of thrombin aptamer crosslinking strands as locking elements triggered nanosheet rolling into a tubular structure. b, AFM images of DNA nanosheets loaded with 0, 2, 4, 6, 8, 9, and 12 tPA molecules (i–vii), and the resulting tubular nanodevice with 8 tPA molecules (viii). Scale bar, 50 nm. c, Agarose gel electrophoresis (1.5%) of nanosheets with varying tPA loadings. d, Assembly yield of nanodevices with different tPA loadings quantified from AFM images (n = 3). Data are mean ± s.d. e, Dimensional analysis of nanosheet size and inter-tPA spacing along vertical (red) and horizontal (blue) directions from AFM images. Scale bar, 50 nm. Data in c and e are from one representative experiment of three. DOI: 10.1038/s41563-024-01826-y
A closed-loop thrombolytic system can be thought of as an “intelligent” or cyber-physical therapeutic. Rather than acting without guidance, it uses sensor input to deploy therapy selectively and then monitors the result, adjusting if necessary. This represents a fundamentally new engineering paradigm in stroke treatment. It draws inspiration from other areas of medicine: for example, closed-loop insulin pumps in diabetes use continuous glucose sensing to adjust insulin delivery in real time. We envision a similar concept applied to clot dissolution. Notably, early examples of this approach are already appearing at the preclinical level. One striking demonstration was a recent DNA-origami nanodevice programmed for precision thrombolysis [45]. This rationally designed DNA nanostructure carried multiple tPA molecules “locked” by DNA aptamer-based fasteners that sense thrombin. Upon encountering high thrombin levels in a clot, the DNA locks automatically opened, releasing tPA precisely where thrombin (and by extension, the thrombus) was abundant [45] In stroke and pulmonary embolism models, this intelligent nanodevice achieved recanalization with less off-target bleeding, essentially performing on-demand dosing of tPA.
Another example is the development of self-regulating microbubble systems that not only deliver thrombolytics under ultrasound guidance but also carry imaging tracers to indicate how much of the clot has lysed [46]. These pioneering studies, while early, provide preclinical proof-of-concept for feedback-gated thrombolysis and illustrate how progressively more complete closed-loop control might be engineered.
The thrombus as a dynamic biological system
Spatial and molecular heterogeneity of thrombi
Occlusive thrombi in acute ischemic stroke are structurally heterogeneous, comprising discrete regions with distinct cellular and molecular composition (Table 1) [47]. Histological studies of retrieved stroke clots consistently reveal an outer platelet- and fibrin-rich zone and an inner RBC-dominated core [48–50]. The platelet-dominated regions contain densely crosslinked fibrin networks interwoven with platelets, abundant von Willebrand factor (vWF) multimers, and infiltrating leukocytes with extracellular DNA from (Neutrophil Extracellular Traps (NETs) [47, 48]. By contrast, RBC-rich regionsare structurally simpler. They are packed with polyhedral erythrocytes (polyhedrocytes) compressed by platelet contraction, with only a sparse, thin fibrin meshwork filling the gaps between cells and minimal presence of vWF or nucleated cells (Fig. 2) [51]. In essence, platelet-rich “white” areas are compositionally complex (platelets, fibrin, vWF, leukocyte DNA), whereas RBC-rich “red” areas consist mainly of tightly packed erythrocytes in a flimsy fibrin scaffold.
Table 1.
Mechanistic compartmentalization of stroke thrombi and its implications for closed-loop sensing. For each thrombus compartment, the table lists the dominant cell types, key molecular and structural features, the local microenvironmental cue, the primary handle available for molecular sensing, and the resulting implication for sensor design
| Thrombus compartment | Dominant cell types | Key molecular/structural features | Microenvironment cue | Primary sensing handle | Implication for sensing strategies |
|---|---|---|---|---|---|
| RBC-rich core | • Erythrocytes (polyhedrocytes) |
• Tightly packed RBCs • Thin fibrin fibers cementing cells • Low vWF or platelet content |
• Limited oxygen/perfusion • Hypoxic, acidic micro-pocket |
• Hypoxia/pH changes • RBC markers |
• Biochemically relatively inert but forms a physical diffusion barrier • Sensors may need to localize these regions and support strategies to breach the packed core |
| Platelet-/fibrin-rich shell |
• Activated platelets • Leukocytes (neutrophils, monocytes) |
• Dense cross-linked fibrin network with abundant vWF • Activated platelet integrins (GPIIb/IIIa) and P-selectin • Platelet granule releases (ADP, TxA₂) • Antifibrinolytic serpins (PAI-1, PN-1) concentrated in shell • Embedded NET components |
• Highly pro-thrombotic, lysis-resistant zone |
• Fibrin/vWF fibers • P-selectin • GPIIb/IIIa • Local enzyme inhibitors |
• Rich in potential targets for sensors • Sensing these markers could trigger the nanodevice to boost fibrinolysis exactly where needed to erode the shell |
| Neutrophil/NET-rich regions |
• Neutrophils (extracellular traps and intact cells) • Platelets entrapped in NETs |
• Webs of DNA and histones coating clot fibers • Neutrophil proteases (elastase, MPO) bound to NET scaffold • Inflammatory mediators like HMGB1 within NETs • Co-localized platelets and vWF on NET fibers |
• Inflammatory, thrombolysis-resistant niche |
• NET biomarkers • NET-associated enzymes |
• Can be sensed to estimate NET burden • A surge in NET signals could prompt adjunctive treatments • Adjustment of thrombolytic dosing since NET-rich clots need more aggressive lysis • Monitoring NET-associated enzymes might also warn of impending tissue damage |
RBC, red blood cell; vWF, von Willebrand factor; NET, neutrophil extracellular trap; GPIIb/IIIa, glycoprotein IIb/IIIa; ADP, adenosine diphosphate; TxA₂, thromboxane A₂; PAI-1, plasminogen activator inhibitor-1; PN-1, protease nexin-1; MPO, myeloperoxidase; HMGB1, high-mobility group box 1; tPA, tissue plasminogen activator
Fig. 2.

Heterogeneous composition of an ischemic stroke thrombus. Top: RBC-rich region with tightly packed polyhedral erythrocytes (red) cemented by a sparse fibrin network (thin gray fibers). Neutrophils (purple) are present but infrequent. Bottom: Platelet-rich region containing abundant platelets (small gray cells) embedded in thick bundles of fibrin (thick gray fibers) and von Willebrand factor strings (green). This zone also harbors neutrophils releasing NETs (extracellular DNA strands in blue) and other leukocytes. Adopted with permission from Xu & Ariëns (2020). DOI: 10.3324/haematol.2019.238816
Most thrombi contain both RBC-rich and platelet-rich segments in varying proportions that may be intermingled or layered [51]. In many large artery occlusions, a well-defined interface is observed: platelet-rich shell regions often encase a central RBC-rich core [47]. This shell–core architecture is especially pronounced in fibrin-rich clots, where a compact outer shell of fibrin and platelets surrounds an inner core dominated by polyhedrocyte RBCs [48–50]. Electron microscopy shows the shell as a dense, matted mesh of fibrin/platelets lacking clear cell borders, whereas the core retains visible intact cells and fibrillar strands [52]. Functionally, the compacted shell exhibits low porosity and forms a barrier to diffusion, impeding penetration of lytic enzymes into the core [49]. This contributes to therapy resistance: platelet- and fibrin-rich thrombi (with robust shells) tend to resist recombinant tPA fibrinolysis, whereas RBC-dominant clots are often softer and more prone to fragmentation [53].
Biochemical signals in the clot microenvironment
Beyond gross anatomy, thrombi also generate complex biochemical microenvironments defined by local signaling molecules and gradients. In platelet-rich zones, activated platelets release high concentrations of secondary agonists, including ADP and thromboxane A₂, that diffuse through the clot and amplify further platelet recruitment [52]. These regions are also sources of pro-thrombotic and anti-fibrinolytic mediators. Notably, the thrombus shell is enriched in plasminogen activator inhibitor-1 (PAI-1) and protease nexin-1 (PN-1), serpin inhibitors that locally neutralize tPA and plasmin [49, 54]. Such factors create a biochemical shield around the clot, confining fibrinolysis to the periphery until the shell is partially breached.
Meanwhile, RBC-rich interiors, due to stagnant flow, can develop gradients of oxygen and pH. Entrapped erythrocytes consume oxygen and produce metabolic acids under ischemic conditions, leading to relative hypoxia and acidosis in the clot core [55]. This hypoxic, low-pH microenvironment triggers platelet and endothelial stress responses: for example, ischemia-induced oxidative burst (ROS generation) and the release of damage-associated molecular patterns like HMGB1 from dying cells [56]. HMGB1 and other DAMPs diffuse into the thrombus and activate nearby neutrophils and platelets [56]. Thus, the ischemic clot is biochemically active, with platelet-derived agonists, coagulation modifiers, and inflammatory signals all co-localized in a confined space (Fig. 3).
Fig. 3.

Mechanistic landscape of NET-induced thrombus stabilization. In ischemic stroke, the ischemic microenvironment (panel A) triggers neutrophils to release NETs via hypoxia/reperfusion-induced ROS, HMGB1 and other DAMPs from injured tissue, and direct platelet-neutrophil interactions. NET formation (panel B) can occur through a lytic pathway (NADPH-oxidase and protease driven cell lysis) or a non-lytic PAD4-driven pathway, resulting in chromatin expulsion. Once released, NETs (panel C) exert pro-thrombotic effects: the DNA-histone scaffold traps platelets and RBCs and activates coagulation, yielding larger, more stable thrombi that resist fibrinolysis. NET proteases and histones concurrently degrade endothelial tight junctions and activate inflammatory cells, worsening blood–brain barrier disruption and propagating neuroinflammation. Figure 3 was created by the authors using a generative AI image tool and subsequently verified and edited for scientific accuracy
A striking biochemical feature of stroke thrombi is the pervasive involvement of neutrophils and their NETs, which profoundly alter the clot’s microenvironment. NETs amplify thrombosis: the DNA-histone scaffold of NETs provides a charged surface that binds platelets and coagulation factors, accelerating thrombin generation and stabilizing the clot [57, 58]. NET components such as neutrophil elastase and myeloperoxidase can modify fibrin and vWF, while cationic histones augment local platelet activation and endothelium damage [59]. NET-rich areas of the thrombus, identified by citrullinated histones and DNA, co-localize with intense platelet and vWF staining [51], underscoring that NETs integrate into the platelet-fibrin matrix rather than being separate entities. By acting as pro-thrombotic scaffolds and as inflammatory stimuli (histones and DNA are potent damage-associated molecular patterns), NETs make the clot microenvironment highly dynamic and hostile to fibrinolysis [60].
Consistent with this, enzymatic removal of NETs greatly improves thrombolysis: treating thrombi with DNase (to digest NET DNA) accelerates tPA-mediated clot lysis in ex vivo studies [61, 62].
Mechanical forces and hemodynamic gradients
Thrombi are not only biochemical but also biophysical entities that experience and respond to mechanical forces. As platelets aggregate during clot formation, their actomyosin machinery actively contracts the nascent thrombus. This platelet-driven contraction compacts the fibrin network and draws the thrombus tight [63–68]. Mechanically, platelet-rich areas become centers of contractile force, which dramatically increases clot density and stiffness. Direct measurements have shown that integrin-mediated platelet contraction induces a prestress in fibrin fibers, raising the bulk elastic modulus of the clot [69]. In fact, thrombi with high fibrin/platelet content are significantly stiffer and less deformable than those dominated by RBCs [53]. By contrast, RBC-rich clots (with fewer platelets and more hydrated volume) tend to be softer and more viscoelastic in response to strain [53].
Platelet contraction also redistributes components: as the fibrin and platelet meshwork is pulled inward and toward the thrombus periphery, trapped RBCs are compressed into the core. This is the origin of the polyhedral RBC morphology: platelet contraction packs the RBCs into a tighter space. The net effect is the formation of a tense outer shell and a compacted core. A well-developed fibrin–platelet shell endows the thrombus with mechanical resilience against shear and fragmentation, but at the cost of permeability. The polyhedrocyte-rich core and dense shell together create a near-impermeable barrier to flow, severely limiting intrathrombus fluid exchange [70]. Consequently, neither blood plasma nor administered lytic drugs penetrate easily into a contracted clot’s interior. This mechanical barrier reinforces biochemical resistance, as fibrinolytic enzymes remain mostly confined to the clot surface [70]. From a therapeutic standpoint, it highlights that simply delivering a drug to the clot may be insufficient: the drug must also reach the clot’s interior, or the clot’s physical structure must be disrupted. Closed-loop thrombolytic systems may therefore require actuators to overcome the biomechanical integrity of a well-compacted thrombus.
In situ, an occlusive thrombus is subjected to steep hemodynamic gradients. Blood pressure is high upstream of the clot and near-zero downstream, creating a strong pressure differential across the thrombus. This drives flow around or through any channels in the clot and generates shear forces along the thrombus surface. High shear stress at the thrombus front (proximal face) tends to promote further platelet deposition: under arterial shear rates, vWF unfolds and rapidly tethers circulating platelets, reinforcing the platelet-rich shell at the blood–clot interface [71]. Thus, flow can actively grow the thrombus shell by recruiting platelets and fibrin under shear.
Conversely, the low-shear or stagnant zone within and distal to the clot favors RBC pooling and fibrin deposition without strong platelet recruitment, contributing to the RBC-rich core. The mechanical stability of different thrombus regions also diverges under flow. Fibrin- and platelet-rich clots, with their taut polymer networks, exhibit higher resistance to deformation or extraction, whereas RBC-rich clots are more fragile and prone to shear-induced fragmentation. Empirical evidence from thrombectomy cases shows that “white” (platelet-rich) thrombi often require multiple retrieval attempts and resist device capture, whereas “red” (RBC-rich) thrombi more frequently fragment and distalize. The thrombus composition also affects how strongly it adheres to the vessel wall or to interventional devices. For example, clots rich in vWF and platelets form tough adhesive interactions with stent-retriever struts, whereas RBC-heavy clots tend to slip or break apart.
What does “closed-loop” mean in thrombolysis nanomedicine?
Definitions and basics
An open-loop system delivers a fixed or pre-programmed intervention with no measurement of the clot or vessel state and no adjustment during treatment. A stimulus-responsive system releases its payload when one local cue (for example thrombin, ROS, pH, or shear) crosses a set point; this is a single event-triggered action, not continuous control. A theranostic system co-delivers therapy and an imaging or sensing readout, but the readout need not govern the therapy. A closed-loop system measures a state variable, compares it to a target, and continuously modulates therapy on the basis of that comparison. We further distinguish semi-autonomous (human-in-the-loop) control, in which a clinician interprets the readout and adjusts therapy, from fully autonomous control, in which the device adjusts and terminates therapy itself. These distinctions map directly onto the Closed-Loop Readiness Levels introduced below.
From open-loop stroke therapies to closed-loop control
Open-loop therapies for ischemic stroke deliver interventions without adjusting to patient or clot feedback. Intravenous tPA is given as a fixed-dose infusion over time with no sensing of clot status, which can lead to overtreatment and off-target fibrinolysis [71]. Clinical data show standard-dose tPA (0.9 mg/kg) achieves recanalization in only about 25% of large arterial occlusions, yet it significantly raises intracranial hemorrhage risk compared to lower doses [72, 73]. Likewise, mechanical thrombectomy lacks real-time biochemical feedback: operators rely on angiography rather than any automated clot-state sensor [74–76]. The result is that current “open-loop” approaches often trade efficacy for safety, as seen in higher bleeding rates with aggressive thrombolysis and incomplete clot removal when fixed protocols face resistant thrombi. Nanocarriers developed to deliver tPA have largely been open-loop as well, releasing drug payloads passively or upon a predetermined trigger without ongoing adjustment. For example, fibrin-targeted liposomes or microbubbles release tPA upon reaching the clot but cannot halt release if flow is restored, nor can they amplify dosing if the clot proves refractory. This leads to either unnecessary systemic exposure or suboptimal lysis [77–79].
In contrast, a closed-loop therapy continuously monitors the thrombotic state and modulates treatment in response, in the manner of an engineering feedback control system. In thrombolysis nanomedicine, this requires four elements. A sensor reports the clot or hemodynamic state; a decision logic, or controller, compares that state to a desired setpoint; an actuator, such as drug release or mechanical action, is adjusted on the basis of the feedback; and a reporting mechanism confirms the outcome. Key control concepts include the state variable and a threshold/setpoint that triggers action, as well as managing time delays to maintain stability. A closed-loop system aims for timely, on-demand thrombus dissolution: for instance, releasing thrombolytic only when a clot-specific signal exceeds a threshold and stopping release once recanalization is achieved (Fig. 1) [31]. This paradigm has been articulated in recent stroke care frameworks calling for “smart” thrombosis management, integrating real-time monitoring and automated adjustments to optimize efficacy and minimize harm [80]. In essence, closed-loop control replaces the one-size-fits-all dosing of open-loop therapy with responsive, patient-tailored intervention guided by continuous feedback.
Many so-called “stimuli-responsive” thrombolytic nanodevices fall short of true closed-loop control (Table 2). These systems often react to a single biochemical trigger, such pH, enzyme, or temperature by releasing their drug payload, an event-triggered response, but do not continue to sense or adapt after that one-off activation. For example, a recent thrombin-cleavable nanoparticle loaded with plasmin was designed to release its enzyme only in the presence of high thrombin concentrations. In vitro and zebrafish models this targeting improved fibrinolysis by localizing plasmin generation to clots, but the nanoparticle provided no feedback once activated and offered no information on whether the clot was lysed [32].
Table 2.
Experimental Studies of Stimuli-Responsive and Theranostic Thrombolysis Nanoplatforms in Ischemic Stroke and Thrombosis Models
| Platform | Study | Platform type | Payload | Trigger/Imaging | Model | Key point |
|---|---|---|---|---|---|---|
| Stimuli-Responsive Thrombolytic Systems | Sun et al., 2023 [103] | Liposomal NP | Plasmin | Thrombin | In vitro clot; zebrafish | Thrombin-triggered release |
| Zhang et al., 2024 [146] | Polydopamine NP | Urokinase + NO | Thrombin + NIR | Arterial thrombosis (mouse) | Dual thrombin/NIR triggers | |
| Zhao et al., 2020 [187] | RBC–coated NP | Tirofiban | H₂O₂ (ROS) | FeCl₃ carotid thrombosis (mouse) | H₂O₂-responsive release | |
| Zhang et al., 2025 [136] | Mesoporous silica NP | Urokinase + TI | Thrombin | Arterial thrombosis (mouse) | Sequential dual release | |
| Zhang et al., 2021 [33] | MnO₂ nanocarrier | Urokinase | Thrombin | FeCl₃ carotid thrombosis (mouse) | Thrombin-triggered release; ROS-scavenging | |
| Zhang et al., 2021 [188] | Core–shell NP | Urokinase + TI | Shear (blood flow) | Arterial thrombosis (mouse) | Shear-triggered release | |
| Yin et al., 2024 [31] | DNA origami device | tPA | Thrombin (threshold) | Ischemic stroke (mouse) | Thrombin-activated (on/off) | |
| Guo et al., 2018 [110] | “Shrinkable” polymer NP | Glyburide | Thrombin + MMP-9 | Ischemic stroke (mouse) | Protease-triggered size change | |
| Zhang et al., 2025 [110] | Liposome (“IcpLipo”) | Urokinase | NIR (808 nm) | Embolic occlusion (mouse) | NIR-triggered drug release | |
| Zhang et al., 2025 [189] | PVA nanocapsule | Urokinase | NIR (1064 nm) | Venous thrombosis (rat) | NIR-triggered capsule opening | |
| Theranostic/Reporting Thrombolysis Platforms | Wang et al., 2020 [77] | Porous magnetic microbubble | tPA | Ultrasound (US) + magnetic | Venous thrombosis (mouse) | Magnetically guided, US-activated |
| Chen et al., 2023 [190] | Targeted microbubbles | None | Ultrasound (low-power) | Deep vein thrombosis (rabbit) | US-assisted thrombolysis (anti-inflammatory) | |
| Ma et al., 2025 [191] | Nanobubble (PFP core) | Urokinase + CeO₂/MnOₓ | Ultrasound (real-time PA/US) | Arterial thrombosis (mouse) | US-visible, ROS-scavenging | |
| Kwon et al., 2018 [192] | Silica-coated gold NP (TAP-SiO₂@AuNPs) | None (imaging probe) | Thrombin-activatable NIRF + micro-CT | In situ thrombotic mouse model | Thrombin-cleavable peptide enables “switch-on” fluorescence at thrombus, while Au core provides CT contrast for dual-modality thrombus visualisation. | |
| Xing et al., 2025 [193] | NIR-II gold nanorods (GNRs) | None (photothermal) | Photoacoustic imaging (real-time) + NIR-II laser (1064 nm) | Carotid artery thrombosis (mouse) | PA imaging tracks GNR accumulation at the thrombus to time irradiation, enabling localised heating and clot reduction with reduced reliance on systemic thrombolytics. |
NP = Nanoparticle; tPA = Tissue Plasminogen Activator; NO = Nitric Oxide; TI = Thrombin Inhibitor; ROS = Reactive Oxygen Species; RBC = Red Blood Cell; NIR = Near-Infrared; US = Ultrasound; MRI = Magnetic Resonance Imaging; PA = Photoacoustic; CT = Computed Tomography; GNR = Gold Nanorods; PFP = Perfluoropentane; MMP-9 = Matrix Metalloproteinase-9; PLGA = Poly(lactic-co-glycolic acid); Gd = Gadolinium; CeO₂ = Cerium Oxide; MnOₓ = Manganese Oxide; PVA = Polyvinyl Alcohol; LIFU = Low-Intensity Focused Ultrasound
Conversely, theranostic systems include a reporting function (imaging the clot) without feedback control over drug release (Table 2). An example is an iron-oxide nanoconjugate of tPA, which could be magnetically guided to clots and visualized by MRI. While this allowed investigators to confirm nanoparticle accumulation and reduce the tPA dose by 80%, the release of tPA from the carrier was not conditional on a sensor. It was effectively still “always on,” just better targeted [81]. Thus, one system (thrombin-responsive NP) had a biochemical feedback trigger but no reporting, and the other (MRI-visible tPA nanocarrier) had reporting but no autonomous control of release. Neither achieves a fully closed loop. A true closed-loop thrombolytic nano-system would require both: a sensor to trigger or modulate therapy and a coupled reporter to indicate when the therapeutic goal (recanalization) is reached, enabling on-the-fly adjustment.
Control architecture for sense → target → lyse → report
Sense
Closed-loop nanomedicine begins with sensing a relevant clot signal. Experimental thrombolytic nanoparticles have been engineered to sense biochemical cues like thrombin, fibrin, or pathological microenvironments. A notable example is a DNA-origami nanodevice that uses a thrombin-binding aptamer as a molecular sensor. It remains inert until local thrombin concentration rises above a pathological threshold, at which point the DNA “nanolock” springs open to release tPA [31]. In rodent stroke models, this thrombin-responsive device distinguished between high-thrombin clots and low-thrombin hemostatic plugs, effectively localizing drug action to sites of thrombosis [31]. In contrast, Zhao et al. (2021) developed a platelet-membrane-cloaked nanoparticle that senses the inflammatory environment of clots via high levels of hydrogen peroxide (H₂O₂), releasing the anticoagulant argatroban when encountering oxidative stress [82]. In a FeCl3 carotid thrombosis model, this H₂O₂-responsive NP efficiently inhibited clot formation and also scavenged reactive oxygen species, reducing local inflammation [82].
The two sensors, the thrombin aptamer and the H₂O₂ trigger, illustrate a trade-off between specificity and breadth. Thrombin is a specific marker of active coagulation, making thrombin-triggered release highly selective to clots, with Yin et al. reporting 3.7-fold higher lysis in stroke models and negligible effect on normal clots. H₂O₂, on the other hand, captures the prothrombotic inflammatory milieu, which is useful since oxidative burst accompanies thrombosis, but H₂O₂ can also be elevated in unrelated inflammation. Thus, H₂O₂-responsive therapeutics risk off-target activation in tissues with inflammation but no thrombus.
This highlights the importance of signal “focus”. In Zhao’s study, the platelet cloak did confer some thrombus-specific homing to the H₂O₂-NP (colocalizing predominantly in occluded arteries), but one must consider that any systemic inflammatory condition could theoretically trigger argatroban release. By contrast, thrombin-aptamer systems might miss a clot that is forming under strong platelet activation but before thrombin accumulates (false-negative sensing). Achieving robust yet specific sensing may require multi-signal logic, an area just beginning development (AND-gate designs), to avoid signal cross-talk and false triggers. Most current systems rely on one biochemical cue, so assessing their sensitivity and specificity in biologically complex settings (blood flow, multiple clotting factors) is an active area of research.
Target
Effective targeting under flow is essential for any closed-loop thrombolytic system [83]. The carrier must reach and persist at the clot to sense and act [83, 84]. Nanoparticle design strongly influences margination and adhesion to thrombi. Liu et al. (2025) vividly demonstrated this by comparing spherical, rod-shaped, and discoidal PLGA nanoparticles conjugated with tPA under arterial flow conditions [85]. Discoidal nanoparticles showed superior margination and binding to thrombus surfaces at high shear, leading to faster thrombolysis and cerebral perfusion restoration in an embolic stroke model [85]. Spheres, in contrast, largely stayed in the flow stream and had minimal wall interactions, requiring much higher concentrations to achieve comparable clot lysis [85]. The improved performance of discoids was attributed to their larger contact area and tumbling dynamics, which increase frequency of vessel wall collision under shear flow. This highlights that in dynamic settings, physical attributes can matter as much as molecular targeting ligands.
By comparison, many earlier targeted nanocarriers were tested only under static or slow-flow conditions [30, 86, 87]. For instance, in an endothelial-flow model, immunoliposome targeting efficiency was 5-fold lower under flow at 4 dyn/cm² (± 40% hematocrit) than under static conditions, consistent with convective washout and endothelial barriers limiting near-wall residence time [88]. Likewise, clot-binding liposomal nanoparticles have been evaluated under high-shear microfluidics at 25 dyn/cm² (2800 s⁻¹; 20–22 mL/min) where binding is confined to a near-clot “reaction zone” with ongoing dissociation back into bulk flow. This underscores that static binding assays can overestimate delivery in hemodynamic settings [89]. Indeed, a consensus in recent studies is that traditional “lock-and-key” targeting is not enough in large, fast-flowing arteries [90–93]. Flow-based testing reveals whether a carrier can actually localize to the thrombus before being cleared.
Another factor is the protein corona: serum proteins that adsorb nanocarriers in blood, which can mask targeting ligands and alter adhesive properties [94]. For instance, Mirshafiee et al. reported that coronas reduce active targeting yield by masking surface-displayed ligands [95]. Lin et al. reported that tPA covalently attached to chitosan-coated magnetic nanoparticles achieved reperfusion in a rat embolic model with an 80% reduction in tPA dosewhen an external magnet was applied. This compensated for transport and recognition losses by forcing high local nanoparticle accumulation at the occlusion [96]. Even in such systems, however, non-specific adsorption of plasma proteins remains a credible failure mode because it can reduce fibrin-recognition efficiency and shift adhesion towards off-target cellular or immune interactions. This reinforces the broader point that closed-loop thrombolytic carriers must be engineered for the coupled hemodynamic and biochemical complexity of flowing blood.
Lyse
The actuation step (“Lyse”) encompasses the actual clot disruption mechanisms, which in closed-loop systems should be controllable in timing or intensity. Traditional thrombolysis relies on enzymatic digestion of fibrin in an open-loop fashion. Nanomedicine broadens the arsenal to include controlled drug release kinetics and externally triggered physical energies. Each modality has distinct performance metrics and limitations, and direct comparisons are starting to emerge. For example, Lin et al. (2024) introduced a sonodynamic and mechanical thrombolysis platform: perfluorocarbon-loaded nanoparticles that, upon ultrasound exposure, produce reactive oxygen species and mechanical cavitation to attack the fibrin matrix [72]. Under real-time ultrasound imaging, this closed-loop theranostic system achieved > 90% recanalization in a rat model of tPA-resistant carotid occlusion, with restoration of flow in under 30 min and no re-thrombosis observed in the month after treatment [72]. Crucially, the ultrasound could be turned on or off in response to the imaging feedback, effectively modulating the lytic activity in real time.
In contrast, a magnetic hyperthermia approach by Cabrera et al. (2022) physically warms and permeabilises clots via superparamagnetic iron oxide nanoparticles in an alternating magnetic field [38]. The heating softens the fibrin network, making it more susceptible to enzymatic breakdown. In human plasma clots ex vivo, this method significantly increased the penetration of tPA into contracted thrombi and accelerated fibrinolysis [38]. However, translating this to in vivo arterial clots has challenges: ensuring sufficient nanoparticle accumulation in the clot, avoiding tissue heating, and managing immune clearance of the particles. The magnetic approach does allow an external “switch” (the field on/off) but it currently lacks a feedback criterion for when to stop heating [38].
Another modality is microrobotic mechanical lysis. Yang et al. (2023) developed magnetically powered nanorobots (300 nm) that swarm and drill into clots under a rotating magnetic field, delivering attached tPA deep into the thrombus [97]. In rat femoral vein thrombosis, these swarms achieved near-complete clot dissolution and restored flow to baseline in 40 min, with no hemorrhagic complications reported [97]. This mechanical–enzymatic synergy is essentially an “on-demand” lysis: the bots actuate only when and where the magnet is applied and could be halted or re-targeted as needed. Yet here the feedback loop is human-controlled (via imaging to guide magnets).
When comparing modalities, trade-offs become clear. Enzymatic release systems excel in biochemical specificity but may be slower to clear large clots or hampered by enzyme inhibitors at the core of thrombi. Physical actuation like ultrasound or magnetic drilling can attack the clot’s structure directly and help distribute thrombolytic agents throughout the clot mass, but they require hardware and careful calibration to avoid collateral damage [98]. Notably, ultrasound has already seen clinical use in an open-loop way and has shown improved clot resolution and faster lysis [98]. However, trials in acute ischemic stroke (CLOTBUST-ER) yielded mixed results, partly because of inconsistent insonation parameters and lack of real-time control, leading to variable efficacy [99]. This underscores that even with potent actuation methods, incorporating feedback is key to success.
Report
The final element “Report” closes the loop by informing whether the therapeutic goal (clot dissolution and reperfusion) has been met, and by providing a signal that can be used to modulate further therapy. In current research, “reporting” typically takes the form of imaging or sensing outputs: for example, a fluorescent signal, MRI contrast, or ultrasound echo that correlates with nanoparticle accumulation or fibrin presence. A number of theranostic nanoplatforms have been tested in thrombosis models, though not all use their reporting function for feedback control (Table 2). One class of reporters are fibrin-targeted imaging agents: nanoparticles carrying near-infrared fluorophores or iron oxide that light up the clot. These can delineate the clot’s location and size via fluorescence or MRI, confirming that the therapeutic is at the right place [100]. For instance, an MRI-visible tPA-iron oxide nanocarrier allowed real-time tracking of delivery to the thrombus in rabbits [81]. Investigators could observe the thrombus shrinking on MRI as the nanoparticle released tPA, but importantly, the system itself did not alter its behavior in response; the reporting was for the operator’s benefit only.
In another example, Shi et al. developed nitric oxide-releasing microbubbles that not only enhanced thrombolysis when burst by ultrasound, but also acted as ultrasound contrast agents to report the recanalization progress in real time [101]. Under Doppler ultrasound, the restoration of flow could be visualized as the bubbles fragmented the clot, and this informed when to stop ultrasound exposure [101]. Here the reporting and actuation were intimately linked: the same agent (microbubble) provided both therapy and an imaging signal. Yet even this is a semi-closed loop: a clinician interprets the ultrasound images and decides when to cease sonication once flow is adequate, rather than the system regulating itself [101].
To achieve a self-regulating closed-loop, the reporting signal must be quantitatively fed back. For example, one could envision nanoparticles that fluoresce in proportion to fibrin degradation products, ora pressure sensor that detects resumed downstream flow. Once these readouts hit a predefined target, the nanodevice could turn off further tPA release. In practice, such fully integrated reporters are rare. Most studies measure outcomes like clot mass after treatment or use external imaging to infer lysis.
A limitation seen in current theranostics is decoupling of reporting from action. A fluorescent probe might indicate the nanoparticle is at the clot surface, but not whether the clot’s core has lysed. As a case in point, a study attached a fluorescence quencher to tPA so that it would light up when tPA was released from a nanoparticle, thereby reporting drug release occurred [101]. However, release does not guarantee effective lysis; it is a proxy signal. Consequently, that system still needed imaging of flow or clot size to truly assess success. The most informative reporting for thrombolysis would directly measure reperfusion. Early strides include MRI sequences that differentiate older versus fresh clot components and could indicate ongoing lysis, or photoacoustic imaging of oxygenation to signal tissue reperfusion [31, 102].
Closed-loop readiness levels (clrl) for thrombolytic nano-systems
To categorise the maturity of thrombolytic nanomedicine in adopting closed-loop control, we propose a framework called Closed-Loop Readiness Levels (CLRL), from 0 (none) to 4 (fully integrated) (Fig. 4; Table 3). This mirrors technology readiness thinking but focuses on feedback-control capability. CLRL-0 corresponds to purely open-loop approaches with no feedback elements, essentially current standard treatments like fixed-dose systemic tPA or device-only thrombectomy. CLRL-1 systems introduce a single feedback-triggered module, typically a stimulus-responsive drug release in a nano-formulation but remain otherwise open-loop. An example would be a nanoparticle that releases drug in response to one biochemical cue (Table 2) but does not incorporate targeting or reporting; it performs a one-time logical decision (“sense and release”) in a controlled setting.
Fig. 4.

Control architecture and Closed-Loop Readiness Levels (CLRL) for thrombolytic nano-systems. A systems-engineering framework that maps the progressive integration of sensing, targeting, actuation, and reporting functions in thrombolytic nanomedicine onto Closed-Loop Readiness Levels (CLRL-0 to CLRL-4). CLRL-0 represents open-loop thrombolysis, including fixed-dose systemic pharmacotherapy or device-only interventions, with no feedback or adaptive control. CLRL-1 introduces a single feedback-triggered module, typically stimulus-responsive drug release based on one biochemical cue, implementing a one-time sense-and-release logic without targeting or reporting. CLRL-2 reflects integrated smart carriers that combine sensing, active or flow-assisted targeting, and triggered lysis (Sense → Target → Lyse), but lack outcome reporting to close the loop. CLRL-3 adds a reporting function, yielding theranostic systems capable of monitoring localisation or thrombolysis progress; however, feedback is interpreted and acted upon by a human operator (human-in-the-loop control). CLRL-4 denotes the aspirational fully autonomous closed-loop nanosystem, in which sensing, decision logic, actuation, and reporting are internally integrated, enabling real-time self-regulation toward a predefined thrombolytic end-state without overshoot
Table 3.
Closed-Loop Readiness Levels (CLRL) for Thrombolytic Nano-Systems
| CLRL Level | Required Modules & Capabilities | Evidence Required (Models) | Typical Failure Modes | Translational Implications |
|---|---|---|---|---|
| CLRL-0: Open-Loop (Status quo) |
• No feedback integration: fixed dosing or manual control only • Nanocarriers (if used) release drug unconditionally |
• Efficacy/safety from clinical or standard preclinical studies |
• Off-target effects due to inability to adjust • Unnecessary drug use in absence of feedback • Clot heterogeneity unaddressed: therapy may fail in resistant thrombi. |
• Well-characterized path (drug or device) • Regulatory approval focuses on dose-response and safety margins • Lacks personalization; “one-size-fits-all” limitations |
| CLRL-1: Triggered Release (Single feedback mechanism) |
• Sense→Lyse (basic): Device has one sensor trigger controlling drug release. • No targeting beyond passive accumulation; no reporting function |
• In vitro static demos of responsiveness • Possibly ex vivo clot tests. • Proof that feedback trigger works in principle |
• False-positive activation. • False negatives if trigger threshold too high. • Timing delays: Lag between sensing and release onset. |
• Added complexity but relatively low, essentially a “smart drug” • Manufacturing similar to drug-device combo • Needs validation that trigger improves safety/efficacy |
| CLRL-2: Targeted & Triggered (Integrated smart carrier) |
• Sense+Target→Lyse: Nanocarrier actively targets thrombus (affinity or shear-based) AND releases payload via feedback trigger • Still no real-time reporting output |
• In vitro flow models or microfluidics to show targeting under shear + triggered release • In vivo (small animal) showing improved clot lysis or infarct outcomes vs. open-loop |
• Delivery failure: if targeting underperforms, device may never sense trigger (no drug at clot). • Partial clot lysis: one-time release might not fully clear thrombus • No feedback if clot fragments remain |
• More complex design (multi-functional NP) • Regulatory: must prove each module (targeting ligand, trigger) adds benefit • Potentially viewed as combination product. • Translation requires consistent NP fabrication and safety |
| CLRL-3: Theranostic Feedback (Monitoring capable) |
• Sense+Target+Lyse+Report (human-in-the-loop): All modules present. • Device transmits a signal (imaging or biochemical) reflecting its localization or thrombolysis progress • Feedback is used by clinicians or external systems to adjust therapy (not fully autonomous) |
• In vivo efficacy with concurrent imaging/monitoring. • Quantitative improvement in outcomes with feedback-guided intervention |
• Signal ambiguity: reporter indicates NP presence, not clot resolution (risk of stopping too early or continuing too long) • Operator dependency: efficacy still depends on human interpretation of feedback • Sensor/contrast toxicity: e.g. gadolinium or radiation exposure if used for imaging |
• High complexity (diagnostic + therapeutic) • Regulatory: treated as combination device (therapeutic and diagnostic) • Needs validation that feedback improves safety/efficacy in animals • Manufacturing must ensure imaging agent and drug co-delivery consistency • Clinically, requires training to interpret feedback signals |
| CLRL-4: Fully Closed-Loop (Autonomous control) |
• Sense+Target+Lyse+Report (self-regulating): Device autonomously adjusts therapy in real-time based on sensor input, using reporting loop to reach setpoint • Minimal human input after deployment |
• Advanced preclinical (large animal) trials or pilot clinical trials. Demonstration that system achieves thrombolysis goal without manual dose tuning • Evidence of automatic shutdown or attenuation of lysis upon goal achievement, and prevention of overshoot |
• Control instability: oscillation or overshoot if feedback loop not well-tuned • Sensor drift or failure: could lead to incorrect dosing • Individual variability: patient-to-patient differences in clot biology might confound a fixed control algorithm |
• Highest complexity – effectively a “medical robot”. • Regulatory pathway most involved: likely requires demonstrating algorithm safety and reliability of all components • Clinical use would involve close monitoring in initial deployments. If proven, offers maximal efficacy/safety – e.g. infusion that stops itself when clot is gone • Manufacturing and QA must meet both drug and device standards • Long-term, could transform stroke therapy workflow (less need for ICU monitoring if device self-regulates), but liability and malfunction concerns must be addressed |
CLRL-2 indicates a more integrated design with multiple modules working together. For instance, a targeted nanocarrier senses a clot signal and actuates drug release (Sense + Target + Lyse) yet lacks a reporting function to complete the loop. Many current “smart” thrombolytics fall in this category: they home to the clot and unleash therapy upon a trigger, but they do not report back on the outcome. CLRL-3 systems add reporting/monitoring feedback to the above, yielding a theranostic capability, though the feedback might still be interpreted and acted on by a human operator rather than automatically by the device. The DNA origami nanodevice is on the cusp between 2 and 3: it senses and targets with a threshold logic (CLRL-2) and was shown to track circulating microemboli via an imaging tag, providing a form of report. One could therefore argue that it demonstrates CLRL-3 functionality in preclinical models.
Finally, CLRL-4 represents the ideal closed-loop nanosystem: it seamlessly integrates sensing, decision, actuation, and self-contained reporting such that the therapy autonomously adapts in real-time to achieve a defined end-state without overshoot. No current thrombolytic nanotherapy has fully reached CLRL-4, but the concept provides a target for researchers. Each level sets a higher bar for in vivo demonstration. CLRL-1 might be proven by an in vitro benchtop assay showing triggerable drug release, whereas CLRL-3 requires live animal studies with concurrent imaging of clot dissolution and controlled intervention.
Table 3 presents the Closed-Loop Readiness Levels (CLRL) framework as a systems-engineering rubric, defining the modules required at each level from open-loop thrombolysis (CLRL-0) through single-trigger release (CLRL-1), integrated targeting and actuation under flow (CLRL-2), theranostic monitoring with clinician-in-the-loop control (CLRL-3), and the aspirational fully autonomous, self-regulating nanosystem (CLRL-4), alongside the minimum evidence standards, common failure modes, and translational and regulatory implications that differentiate proof-of-concept responsiveness from true closed-loop performance in vivo.
The four modules of closed-loop thrombolysis: evidence-driven dissection
Sensing the thrombus microenvironment
Enzyme-responsive nanosystems achieve on-demand activation by sensing clot-associated proteases such as thrombin. Sun et al. designed clot-targeted thrombin-cleavable nanoparticles (CTNPs) that bind fibrin and platelets and release plasmin only on thrombin-mediated cleavage (Fig. 5) [103]. In microfluidic and zebrafish models, the thrombin-triggered plasmin release from CTNPs achieved clot dissolution comparable to free plasmin while avoiding systemic fibrinogenolysis [103]. The DNA-origami nanodevice of Yin et al. applies the same thrombin-gated logic, exposing tPA only above a local thrombin threshold and so confining activity to clots [45]. These enzyme-sensing platforms leverage the abundance of thrombin in nascent thrombi for specificity but may under-perform in regions where thrombin is depleted or inhibitor-rich. Heterogeneous thrombus composition can leave enzyme triggers unevenly distributed, limiting deep clot activation.
Fig. 5.

Design and mechanism of thrombin-cleavable clot-targeted nanoparticles (CTNPs) for enzyme-responsive thrombolysis. The schematic adopted from Sun et al. (2023) shows CTNPs coated with peptide linkers (purple) that bind fibrin and platelets in a thrombus and are cleaved by thrombin. When exposed to high local thrombin, the nanoparticle shell destabilizes, releasing the encapsulated plasmin (green) to digest fibrin. This represents the Sensing module mechanistically: the nanocarrier “senses” the thrombus milieu (thrombin upregulation) and in response activates its therapeutic payload. The figure illustrates how an enzymatic trigger (thrombin) can achieve site-specific drug release, a core principle of closed-loop thrombolysis. DOI: 10.1016/j.jtha.2022.11.037
(Fig. 5)
Beyond coagulation enzymes, sensing fibrinolytic activity or byproducts can indicate clot lysis progress. Some theranostic systems integrate real-time feedback: Lin et al. developed an ultrasound-responsive nanoplatform that not only releases thrombolytics under ultrasound but also produces an ultrasound imaging signal proportional to clot dissolution [104]. In a tPA-resistant thrombus model, this system achieved > 90% recanalization and allowed noninvasive monitoring to “stop” therapy once flow was restored [104]. Other work has used fibrin degradation biomarkers, such as D-dimer as surrogates for lysis; for instance, thrombin-triggered CTNPs showed a surge in D-dimer release correlating with clot breakdown [103]. In principle, such signals could serve as a shutdown cue to terminate drug release once a clot is cleared. Indeed, enzyme-responsive designs inherently self-limit, when thrombin or fibrin substrata are exhausted, the nanocarrier ceases activation. However, most demonstrations still rely on the operator’s interpretation rather than autonomous control. No truly self-terminating thrombolytic feedback loop has been validated; current “progress sensors” stop therapy only indirectly, risking overtreatment if sensor readouts lag actual fibrinolysis.
Chemical features of the thrombus microenvironment, such as oxidative stress, pH, or hypoxia, have been harnessed to trigger drug release [105–110]. The H₂O₂-responsive, platelet-membrane-cloaked argatroban nanoparticle introduced above exemplifies this chemical-trigger strategy [33]. The polymer core was engineered to degrade and release argatroban upon H₂O₂ exposure, achieving localized anticoagulation that suppressed arterial thrombus growth in rats while sparing systemic coagulation [33]. In ischemic stroke models, pH-sensitive vesicles have been used for “on-demand” thrombolysis. Ren et al. reported protein vesicles that remain intact at physiologic pH 7.4 but disassemble in the acidic ischemic penumbra to release urokinase and an antioxidant, thereby extending the treatment window for fibrinolysis [111]. Chemical cue-driven systems can act autonomously and simultaneously address thromboinflammatory conditions. A limitation, however, is specificity: pathological pH or ROS elevations are not unique to clots, so off-target activation and diffusion gradients between blood and static tissue complicate precise control.
Cells and structural components within thrombi provide another layer of sensing targets. Platelet-rich clots expose markers like P-selectin, integrin GPIIb/IIIa, and activated vWF that nanoparticle sensors can recognize [80]. One strategy is platelet-mimetic nanocarriers: Ma et al. coated polymeric nanoparticles with actual platelet membranes and conjugated tPA to them, creating “nanoplatelets” that home to thrombi via multivalent platelet adhesion mechanisms [112]. In mouse arterial and stroke models, these platelet-mimetic particles accumulated on thrombi and locally released tPA, yielding recanalization equivalent to free tPA with lower hemorrhagic risk [112].
A further approach targets the NET scaffold that drives lysis resistance (Sect. 2). Li et al. developed acid-responsive polymersomes that release DNase I in the post-ischemic acidic milieu, degrading NET DNA in cerebral thrombi, reducing platelet-neutrophil aggregates, and rendering clots more susceptible to tPA [113]. By sensing NETs (indirectly via pH or by targeting DNA), such systems tackle the “hardening” factors of aged thrombi [113]. Still, translating cellular sensing into controlled therapy is nascent. Platelet-mimetic vectors bind indiscriminately to any activated platelets (raising off-target concerns in hemostatic clots), and NET-targeted strategies must balance the dissolution of pathological clots against the preservation of protective ones. These cellular cues are robust but not inherently specific to pathological thrombosis.
Mechanical and hemodynamic cues, from shear flow patterns to clot stiffness, represent a more macroscopic sensing modality. Thrombi in flowing blood create shear gradients that can be exploited for targeted delivery. For instance, discoidal nanoparticles tend to marginate toward vessel walls under flow better than spheres [85]. Liu et al. showed that discoid PLGA nanoparticles bearing fibrinolytic cargo preferentially accumulate in the high-shear zones around thrombi, achieving faster recanalization in a mouse stroke model than conventional particles [85]. Such flow-guided margination essentially “senses” the altered hemodynamics near an occlusion to concentrate therapeutics.
Mechanical activation has also been appliedin the form of micro/nanorobots. Yang et al. (2023) deployed magnetically driven helical nanorobots (300 nm) that drill into clots under an external magnetic field, enhancing penetration of attached tPA [36]. In a rat venous thrombosis, these swarming nanorobots achieved near-complete clot clearance with restored flow in 40 min and could be halted or redirected by modulating the magnetic field. The promise of mechanical sensing lies in its ability to distinguish the physical presence of an obstruction and to actively navigate within the clot matrix [36]. However, current implementations require external control (ultrasound transducers, magnets) and lack autonomous feedback. Most “mechanical” systems respond to operator-applied stimuli rather than the clot itself, risking superficial or uneven treatment unless combined with real-time imaging and adaptive control under physiological flow and hematocrit conditions.
Targeting the clot: navigating the vascular battlefield
Ligand-directed nanocarriers have been developed to home therapeutics to thrombi with high affinity. Many exploit fibrin or platelet-specific ligands to anchor in static clots, but under flow their performance can vary. For instance, polymer nanoparticles functionalized with fibrin-binding peptides or anti-fibrin antibodies show strong adhesion in stasis. Yet in shear flow, single-ligand systems often struggle against washout. To address this, multi-valent targeting has been used: Ma et al.’s platelet-mimicking tPA nanoparticles (PNP-PA) combined GPIIb/IIIa-binding and platelet membrane adhesion for robust attachment under arterial shear (Fig. 6) [112]. In mouse models these particles retained on the thrombus and lysed it, whereas untargeted tPA washed away [112].
Fig. 6.

Engineered “nanoplatelet” for targeted thrombolysis via platelet membrane cloaking. (A) Preparation: a polymeric nanoparticle (grey) is coated with a membrane derived from donor platelets (pink), and recombinant tPA enzymes (red) are conjugated onto the surface via linkers. (B) Mechanism in circulation: the platelet-coated NP mimics native platelets, adhering to von Willebrand factor and activated platelets within the thrombus. Once accumulated, the particle’s surface-bound tPA directly converts plasminogen (orange) to plasmin at the clot, dissolving fibrin (green mesh). This conceptual schematic adopted from Ma et al. (2021) represents the Targeting module: by presenting authentic platelet receptors, the nanocarrier achieves multi-point adhesion under flow and delivers thrombolytic activity into the clot. It exemplifies biomimetic targeting – the nanoparticle is guided to the thrombus by natural platelet interactions, enhancing localized fibrinolysis while reducing washout. DOI: 10.1080/10717544.2021.1879315
Another study decorated liposomes with both fibrin- and P-selectin–binding peptides, creating a heteromultivalent surface that “grabbed” clots at multiple points, achieving higher retention under flow than single-ligand controls [114]. These examples underscore that molecular targeting can work in vivo, but only if designed for flow conditions. High shear can strip weakly bound particles. One group noted that a GPIIb/IIIa-targeted nanocarrier that bound well under static conditions lost much of its adhesion when tested in a flow chamber, as shear forces overcame the single-point binding [114]. Thus, ligand targeting alone faces limitations: under physiological shear, targeted carriers may be outpaced by blood flow, limiting deep penetration into the thrombus, especially if binding affinity isn’t high enough to resist drag.
Biomimetic strategies have emerged to improve targeting by camouflaging nanoparticles with cell membranes. Platelet membrane-coated nanoparticles, for example, present the whole array of platelet surface proteins, enabling multi-faceted adhesion to thrombi [115, 116]. These “nanoplatelets” can bind not only fibrin but also activated endothelium and other platelets, increasing their accumulation at the injury site [115, 117, 118]. In vitro, platelet-coated particles showed minimal binding to healthy vessel walls yet efficiently embedded into platelet-rich clots [119]. In vivo, they enhanced thrombus localization of loaded drugs and reduced off-target deposition compared to bare NPs [119, 120].
(Fig. 6)
Another biomimetic approach uses red-blood-cell (RBC) membranes to prolong circulation and leverage natural margination. RBC-coated thrombolytic nanocapsules (carrying urokinase) circulate longer (half-life 3.3 h) and preferentially accumulate in occluded vessels due to size and flow dynamics (Fig. 7) [121]. Once at the clot, these RBC-camouflaged nanocapsules can be externally triggered to actuate drug release [121]. Biomimicry clearly boosts in vivo targeting, but it can introduce trade-offs: an RBC-coated nanoparticle, for instance, might also interact with the glycocalyx of normal endothelium or get sequestered by the spleen as “aged” RBCs. Moreover, cloaked particles indiscriminately stick to any high-shear region or activated cell, improving uptake but sometimes at the expense of specificity. Ensuring that biomimetic carriers deposit in the thrombus proper rather than upstream or on surrounding tissue remains a key challenge, especially as prolonged circulation increases background interactions.
Fig. 7.

Erythrocyte-membrane-camouflaged magnetic nanocapsules (USIO/UK@EM) with dual photothermal and magnetothermal actuation for thrombolysis. (a) Synthesis and structure: ultrafine superparamagnetic iron oxide (USIO) nanoparticles and urokinase (UK) are co-encapsulated in a polymer nanocapsule, then coated with a natural RBC membrane (red shell). (b) Venous clot removal by photothermal heating: upon NIR laser irradiation, the USIO produces localized heat (yellow) that thins and permeabilizes the fibrin clot, while UK is concurrently released to enzymatically degrade fibrin. (c) Arterial clot removal by magnetothermal effect: an alternating magnetic field (AMF) penetrates the vessel, causing the USIO to generate heat and mechanical vibration inside the clot. This schematic adopted from Zhu et al. (2024) represents the Smart Lysis/Actuation module; the nanocapsules remain inert during circulation, then externally applied energy (laser or AMF) triggers a focused physical + biochemical attack on the clot. DOI: 10.1002/adhm.202400127Digital
(Fig. 7)
Smart lysis: controlled actuation at the clot site
Carrier systems for enzymatic thrombolytics seek to concentrate lytic activity at the clot and control its spatial distribution. A perennial challenge is that free thrombolytics diffuse quickly and primarily act on the clot surface. Nanocarriers can deliver these enzymes into the thrombus bulk. For example, polymer nanoparticles and liposomes have been loaded with tPA: liposomal tPA extended circulation 4–5-fold and, once bound to fibrin, produced effective local fibrinolysis in vitro [114]. Similarly, fibrin-targeted nanogels have been shown to carry tPA into fibrin-rich microclots and enhance lysis in disseminated intravascular coagulation models [89, 122]. A key design aspect is ensuring the enzymatic drug actually penetrates the clot: Sun’s thrombin-cleavable plasmin nanoconjugates released plasmin throughout the clot, achieving uniform fibrin degradation comparable to free plasmin [89]. In contrast, a control nanoparticle that did not respond to thrombin largely remained intact and only eroded fibrin at the exterior. This highlights that smart lysis requires triggers to unleash the drug at the target; simply carrying tPA to the clot is not enough if it stays sequestered in the carrier. Achieving penetration into platelet-rich cores remains an unmet need for purely enzymatic systems. They excel in softer fibrin-rich clots but struggle in calcified or platelet-dense thrombi.
Triggered-release strategies add spatial or temporal control to actuation. Besides biochemical triggers, external energy can be used. Ultrasound-triggered thrombolysis (sonothrombolysis) is already in clinical trials: microbubbles or nanodroplets carrying tPA can be ruptured at the clot site by focused ultrasound, releasing drug in a burst and mechanically cavitating the fibrin network [123, 124]. Lin et al.’s theranostic platform combined perfluorocarbon nanodroplets with a photosensitizer. Upon ultrasound, the droplets vaporized and the photosensitizer generated reactive oxygen, softening the clot and visualizing it via ultrasound imaging. 104 This yielded timely clot clearance in vivo with the ability to turn off ultrasound once recanalization was observed.
Another example used Near-Infrared (NIR) light: gold nanorods were attached to fibrin-targeted peptides so that, after accumulating in the clot, a NIR laser pulse would heat the rods, locally disrupting fibrin structure and accelerating tPA penetration [125]. The trade-off in triggered systems is between precision and responsiveness: an external trigger can be applied at a specific moment and location, but it may take time to deploy and requires imaging guidance. Internal chemical triggers are continuous and autonomous but could respond slower or with delay as the signal accumulates. Conversely, ultrasound can be turned on immediately at treatment, but one must calibrate the duration and intensity carefully. In practice, the “activation latency” and dose control of triggers vary: physical triggers allow instant on/off but involve complex equipment, whereas biochemical triggers are self-contained but offer less instantaneous control.
Reporting and theranostics: seeing treatment in real time
Imaging-capable thrombolytic nanoplatforms (theranostics) are increasingly designed to do more than “find the thrombus”, they aim to co-localize therapy and measurement so that localization, activation, and treatment response can be followed during the same intervention. The most convincing preclinical reports are those in which the imaging readout is explicitly linked to a functional endpoint such as vessel patency or reperfusion, rather than simply demonstrating nanoparticle accumulation at the occlusion.
MRI-visible strategies illustrate this split clearly. In a genuinely theranostic implementation, iron-oxide/polydopamine microparticles conjugated with recombinant tPA (IO@PDA@tPA) enabled molecular MRI detection of stroke-associated microthrombi while simultaneously degrading them. Serial T2*-weighted sequences and magnetic resonance angiography (MRA) tracked the microthrombi signal and patency over time in mouse thromboembolic stroke models [126]. In contrast, some “MRI progress-tracking” systems are not drug carriers but still matter for closed-loop design because they extend the temporal window for measurement. PEGylated ultrasmall iron oxide nanoparticles (PUSIONPs) acted as long-circulating MRA contrast agents, enabling real-time thrombolysis monitoring for up to 4 h after a single injection in rabbit models. In effect, they turn vascular imaging into a continuous state readout rather than a single snapshot [127].
A complementary line of work highlights why “macro-recanalization” can be a misleading success signal: PHySIOMIC (polydopamine iron oxide clusters) were engineered to unmask downstream microthrombi on T2*-weighted MRI after thromboembolic stroke. This explicitly targets the no-reflow problem, in which tissue-level perfusion lags behind angiographic reopening [128]. Even when imaging is purely diagnostic, it can still be functionally informative if it reports thrombus state rather than location. For example, dual-MRI-contrast KCREKA-functionalized nanoparticles were developed to differentiate thrombus age, providing a mechanistic proxy for fibrinolytic susceptibility (fresh versus organized clot), which is more actionable than a binary “present/absent” localization signal [129].
Ultrasound-linked platforms push theranostics further towards intervention because the same modality can image and actuate. Nanoparticle-shelled microbubbles were used as a clot-penetrating delivery strategy in which ultrasound drives deeper transport of thrombolytic payload into the thrombus architecture, accelerating lysis compared with more superficial deposition [130]. A more “systems” version explicitly the failure mode of biochemical resistance. An ultrasound-responsive theranostic platform (hmSi-CREKA-RB-PFH) combined ultrasonic imaging guidance with sonodynamic and mechanical components designed around the NET and fibrin-crosslinking features of tPA-resistant thrombi. It achieved a reported recanalization > 90% in a rat resistant-occlusion model and extending evaluation into larger-animal and graft settings [72]. The contrast between these ultrasound systems is instructive for closed-loop control. Microbubble-style carriers often provide strong contrast that confirms agent presence at the occlusion. The Lin et al. approach, by contrast, ties imaging guidance to a defined mechanistic “keychain” of resistance and then validates a patency outcome, which is closer to a true efficacy-linked feedback variable (Fig. 8).
Fig. 8.

Ultrasound-responsive theranostic nanoparticle for real-time monitoring and enhanced thrombolysis. The schematic adopted from Lin et al. (2024) depicts hollow mesoporous silica nanoparticles (hmSi) loaded with a sonosensitizer (Rose Bengal, RB) and perfluorohexane (PFH), and decorated with a fibrin-targeting CREKA peptide. Upon accumulating in a thrombus, application of a clinical ultrasound beam (US) induces the PFH to vaporize into microbubbles (blue) and RB to produce singlet oxygen (¹O₂), effecting simultaneous ultrasound imaging and clot disruption. The microbubble generation provides contrast for ultrasound imaging – allowing the clot to be visualized in real time, while the ¹O₂ and mechanical bubble action break down fibrin and augment tPA thrombolysis. DOI:10.1038/s41467-024-50741-y
(Fig. 8)
Optical and photoacoustic probes can add depth-resolved thrombus sensing and, in some designs, composition sensitivity, but they expose the core pitfall that “bright signal” is not the same as “open vessel”. Fibrin-specific homopolymer nanoparticles were engineered to integrate NIR-II photoacoustic thrombosis detection with photo-triggered synergistic anti-thrombotic activity, allowing longitudinal tracking of thrombus burden while simultaneously applying an externally gated therapeutic actuation [131]. In a different translational setting, thrombus-targeted nano-agents were used for NIR-II fluorescence imaging-guided therapy across multiple thromboembolism models, emphasizing practical, image-guided decision-making in anatomically constrained tissues [132]. These studies collectively show that optical contrast can quantify localization and apparent mass reduction, but calibration remains difficult. Intensity can drift with tissue depth, local hemorrhage, or probe concentration, so the imaging variable must be interpreted against independent patency measures if it is to serve as a control input [133].
Self-reporting thrombolytic systems attempt to close the semantic gap by reporting activation, not just location. A thrombin-responsive DNA origami nanodevice integrated into a threshold controller that exposes active tPA only when thrombin exceeds a set level. The design also incorporated state switching suitable for fluorescence “turn-on” reporting of device opening, providing a molecular acknowledgement that the drug has been deployed in a pro-thrombotic microenvironment [45].However, this also highlights the key limitation: drug-release reporting can confirm mechanistic activation but still does not guarantee hemodynamic success if the thrombus remains structurally intact or if microvascular obstruction persists downstream [128]. Multimodal designs can mitigate single-signal ambiguity; for instance, an AuIONP platform combined fluorescence/CT/MRI visibility with photothermal thrombolysis, enabling cross-verification of localization by multiple modalities while also delivering an energy-driven lytic mechanism [134].
The experimental literature is converging on a pragmatic lesson for closed-loop thrombolysis: localization signals (MRI contrast, ultrasound contrast, fluorescence uptake) are necessary to validate “on-target delivery”, but they are insufficient as terminal success metrics. Approaches that explicitly measure or infer patency and microcirculatory obstruction, then link that to therapeutic state (dose, actuation timing, or shutdown), are the ones that meaningfully approach an actionable “Report” module rather than a descriptive imaging overlay.
Regulatory and translational challenges
Fig. 9.

Regulatory and translational challenges in closed-loop thrombolytic nanomedicine. Schematic overview of the major barriers limiting clinical translation of closed-loop nano-systems for thrombosis and ischemic stroke. The framework highlights three interacting constraint domains: (1) safety-centered system design, including threshold-gated release, feedback-controlled actuation, and the concept of a therapeutic “safety governor” to prevent hemorrhagic overshoot; (2) manufacturing and regulatory complexity, encompassing multi-component architectures, scale-up reproducibility, stability, product classification uncertainty, and combination product oversight; and (3) clinical workflow and translational constraints, including narrow stroke time windows, integration with emergency care pathways, imaging dependency, and variability between experimental models and human pathology
Safety as a design constraint
Various experimental strategies have tried to overcome the inherent hemorrhage potential of antithrombotic nanomedicine. These include threshold-triggered release systems that deploy drug only in a high-thrombin environment [32],[135–138], and clot-localized activation using mechanical triggers (shear or ultrasound) [139–142] to concentrate lysis at the clot site. Targeted NPs allow dose reduction by delivering drug directly to thrombi, improving thrombolysis efficiency and lowering bleeding in animal models [33, 118, 143, 144]. Bi et al. (2009) conjugated urokinase to magnetic carriers and showed enhanced thrombolysis under a magnetic field, with little prolongation of bleeding time and no plasma fibrinogen depletion in a rat thrombosis model [145]. However, these safety features are intrinsic and passive, without real-time control once the NP is administered. Limited active control over drug release leaves residual hemorrhage risk as a critical limitation.
Safety governor
A “safety governor” is an engineered control layer that actively constrains drug release to safe levels. The safety governor concept borrows from closed-loop control, aiming to limit overshoot in drug delivery [24]. For instance, NPs can be designed with dual gating (requiring two signals such as thrombin and an externally applied cue) so that clot lysis is triggered only when appropriate [77, 146]. Alternatively, drug release can be made self-limiting, diminishing as the clot dissolves [103, 136, 147]. A thrombin-responsive DNA nanodevice is one example: it remains “locked” until local enzyme levels exceed a threshold, then opens to release tPA, thereby preventing premature drug spillover [31]. Such mechanisms resemble a feedback governor that throttles therapy output in real time. Integrating threshold gating and feedback control offers a path to active safety management but adds complexity to an already intricate system.
Conversely, delayed shutdown or sensor drift might cause the system to keep delivering drug despite the clot resolving or to misread signals, leading to overshoot [148–150]. Multi-sensor cross-checks, such as combining chemical and imaging feedback have been suggested to validate clot dissolution and avert spurious activation [130, 151, 152]. Yet each added sensor or logic layer must itself be fail-safe. Without robust safety governors, closed-loop nanodevices risk uncontrolled drug activity, a fundamental limitation to be addressed in design.
Future studies should therefore report not only recanalization, but also endothelial integrity, distal embolization burden, blood-brain barrier disruption, hemorrhagic transformation, biodistribution, clearance, and persistence of off-target particles in reticuloendothelial organs.
Manufacturing complexity and scale-up constraints
Many closed-loop thrombolytic nanomedicines are multi-component systems, far more complex than single-drug formulations. For example, an intelligent DNA-origami nanodevice loaded with tPA and aptamer “locks” demonstrated precise thrombolysis in rodents, and platelet membrane–coated nanoparticles showed enhanced clot targeting and reduced bleeding in mice [45]. These platforms incorporate multiple modules, which complicates manufacturing compared to a conventional drug. Indeed, theranostic nanomaterials’ translation has been limited by complex synthesis and inconsistent reproducibility at scale [153–156]. As a result, experiments often remain at laboratory scale, and design complexity hinders the ability to produce these nanosystems under Good Manufacturing Practice (GMP) conditions. Elaborate multi-module architecture thus poses a significant limitation for real-world production.
Batch-to-batch reproducibility and scale-up
Even when a nanotherapy design is established, batch-to-batch reproducibility and scale-up present major challenges [157–159]. Factors like nanoparticle size distribution, surface functionalization density, and drug loading efficiency can vary with subtle changes in synthesis conditions [160, 161]. Studies have noted that large-scale nanoparticle synthesis suffers from insufficient batch reproducibility, variable physicochemical properties, and low yields [153, 159, 160, 162, 163]. For instance, consistently coating particles with cell membranes or DNA constructs in an identical manner for each batch is difficult, leading to lot-to-lot variability in targeting or release performance. Such variability can undermine translational reliability and complicate regulatory approval. Engineering responses focus on process standardization. However, achieving scalable, high-yield production of complex nanodevices remains an unmet need.
Stability and distribution
Practical manufacturing also involves stability and distribution issues. Many nanocarriers are sensitive to storage conditions: they may aggregate or degrade, necessitating cold-chain handling or lyophilization [164–167]. This raises costs and logistical hurdles for widespread use. Academic prototypes rarely account for long-term shelf-life. Additionally, complex nanosystems often require specialized analytics for quality control, further complicating scale-up. From manufacturing throughput to stability, the leap from bench to bedside exposes severe limitations in current nanoformulation production processes.
Regulatory uncertainty and product classification
Closed-loop thrombolytic nanomedicines blur the line between drug and device, creating ambiguity in regulatory classification [168–170]. Traditional frameworks force products into one category based on primary mode of action (chemical = drug, mechanical = device, biological = biologic). The Food and Drug Administration (FDA)’s Office of Combination Products generally classifies such hybrids as combination products, assigning a lead review center but supplementing device/biologic requirements [170, 171]. In practice, this process can be opaque and inconsistent, as novel integrated technologies don’t fit neatly into existing definitions. A clot-lysing nanorobot that autonomously senses and responds could be argued as a device (since it “acts” mechanically in response to sensors) or a drug (since it delivers a pharmacologic agent). This ambiguity creates uncertainty for developers in determining the approval pathway, required studies, and standards to meet.
Emerging closed-loop systems that incorporate AI guidance or automated feedback control face additional regulatory scrutiny [148, 172]. Agencies will demand rigorous algorithm validation and safety verification akin to high-risk medical devices [173]. Lessons from closed-loop insulin pumps and neurostimulators are instructive: those devices underwent extensive testing to prove that their sensors, control algorithms, and fail-safes protect patients under all conditions [174–177]. This requires in silico modeling, animal studies, and possibly first-in-human trials under close monitoring of the closed-loop behavior [174–177]. Regulators have begun issuing guidance on Physiological Closed-Loop Control (PCLC) devices, emphasizing risk management for sensor errors, communication failure, and “automation bias” in users.
Finally, product classification influences post-market controls and liability. If deemed a drug, the nanotherapy might require pharmacy dispensing and pharmacovigilance for adverse drug reactions; if a device, it might fall under hospital device management and incident reporting systems. A combination product could carry dual obligations. This uncertainty can dissuade investment and complicate how the therapy is integrated into clinical practice. Engaging regulators early (e.g. pre-IND or Q-Submission meetings) can mitigate some uncertainty by defining the primary mode of action and applicable guidelines.
Stroke workflow and time window constraints
Time-sensitivity of stroke management
Acute ischemic stroke treatment is extremely time-sensitive, and any new therapy must work within narrow time windows [178]. Intravenous tPA is most effective when administered within 3–4.5 h of symptom onset, and mechanical thrombectomy for large vessel occlusions is typically indicated up to 6–24 h in select cases [179]. A closed-loop nanomedicine that requires additional steps before restoring perfusion risks losing precious time (“time is brain”). Some nanotherapeutic strategies involve delayed activation: for instance, waiting for an external trigger (like focusing an ultrasound beam or NIR laser on the clot). If such a therapy significantly prolongs time-to-reperfusion, it could reduce clinical benefit despite its advanced control features. Comparisons across experimental platforms show variability in activation latency. A shear-activated NP can begin releasing drug on encountering high shear, virtually immediately in circulation, whereas an enzyme-responsive NP might not fully release tPA until sufficient enzyme accumulates, which could take several minutes. In stroke models, even 10–20 min of treatment delay can worsen outcomes. Thus, a translational constraint is ensuring the closed-loop system does not introduce untenable delays, rapid action is mandatory, and any complexity must be streamlined for emergency use. Slow or complex activation is a critical limitation in the hyper-acute stroke setting.
Workflow integration in emergency care
Another challenge is workflow integration in emergency care. Stroke thrombolysis is typically initiated under intense time pressure in the emergency department or stroke unit. Standard practice is to obtain brain imaging (non-contrast CT or MRI) quickly to confirm an ischemic stroke with no hemorrhage, then start tPA bolus and infusion immediately. A novel nanotherapy that requires extra imaging modalities or device setup could be hard to fit into this streamlined workflow. Imaging dependency is especially problematic: many stroke centers rely on CT, not advanced MRI or sophisticated point-of-care imaging, for initial triage. A therapy requiring MRI-based activation would face logistical delays, since MRI is not routine in the ultra-acute phase in most hospitals. Even ultrasound-based sonothrombolysis, which has shown promise in trials when adding ultrasound and microbubbles to tPA [180], encountered challenges because not all centers have personnel and machines ready to perform continuous transcranial Doppler during thrombolysis.
For thrombectomy-capable strokes, any infusion must also coordinate with endovascular procedures without delaying door-to-puncture times. For instance, researchers combined a shear-activated nanodrug with a temporary stent to restore partial flow without slowing the thrombectomy workflow [181]: such hybrid approaches highlight the need to complement, not complicate, existing stroke protocols.
In practical terms, a closed-loop nanotherapeutic is unlikely to replace intravenous thrombolysis or mechanical thrombectomy in the near term. Its most realistic near-term role is adjunctive. It would be given alongside or immediately after standard reperfusion to address residual microthrombi, resistant clot, and reperfusion injury, rather than serving as a stand-alone substitute. For this role to be credible, a nanosystem must meet three conditions. It must integrate into the existing emergency pathway without extending door-to-needle or door-to-puncture times. It must be deployable with the imaging and personnel already present in the angiography suite. And it must show that it shortens, or at least does not prolong, time to recanalization. Any requirement that interrupts the standard stroke workflow or extends the door-to-treatment interval is a serious limitation for clinical adoption.
A realistic early clinical deployment pathway would therefore test closed-loop nanotherapeutics first as adjuncts in patients already undergoing thrombectomy, where arterial access, angiographic monitoring, and specialist staff are already available. Initial endpoints should not be replacement of reperfusion therapy, but improved distal reperfusion, reduced residual microvascular obstruction, lower thrombolytic exposure, and no prolongation of door-to-puncture or reperfusion time. Only after safety and workflow neutrality are demonstrated should earlier intravenous or prehospital use be considered.
Reproducibility and model validity
Gap between experimental models and human pathology
Translating closed-loop thrombolytic nanomedicines from bench to bedside is hampered by the gap between experimental models and human pathology. In vitro and microfluidic models of blood clots, while useful for initial testing, do not fully recapitulate the composition and dynamics of clots in human arteries. Simple in vitro clots often lack the heterogeneity of real thrombi. Nanodevices that show fast thrombolysis in a microfluidic channel might underperform in vivo where blood flow, vessel elasticity, and endogenous inhibitors come into play. Indeed, performance gaps are commonly reported: one study might show < 15 min clot lysis in vitro with a novel NP, yet in animal embolism models the same NP takes significantly longer or fails to completely recanalize [153]. Reproducibility between labs is also an issue: different groups use different clot models, making it hard to directly compare efficacy results. This lack of standardized testing means a positive result in one model does not guarantee success in another. The variability and oversimplification of preclinical clot models are a limitation that clouds the true efficacy of these therapies.
Species difference
Moreover, species differences in clotting biology and fibrinolysis can mislead translational expectations. Rodents have higher fibrin turnover and different platelet reactivity compared to humans [182–186]. A nanoparticle dose optimal in mice may not scale linearly to humans, either in efficacy or safety. Conversely, a device might dissolve murine carotid thrombi easily, but human clots in large arteries are bigger and more resistant. Flow conditions also differ: blood velocity and shear in a mouse artery are not the same as in an adult human middle cerebral artery. Some mechanical activation mechanisms need validation under human-scale flows; a device tuned to mouse shear levels might release too early or late in human vessels. Large animal models can bridge this gap somewhat, but they are expensive and not widely used, leading to limited data.
Lack of standardized evaluation framework
Another aspect is the lack of standardized evaluation frameworks for closed-loop performance. While there are guidelines for testing pharmacologic thrombolytics, there is no consensus on how to test an adaptive system that might behave differently case by case. Developers might each choose different endpoints. This inconsistency makes it hard to compare approaches or to ensure that an apparently successful strategy is truly superior. The community has called for more rigorous multicenter preclinical trials (as done in some neuroprotection studies) to verify findings in a robust way. Until such standards are widely adopted, reproducibility will remain a concern.
Future insights and research trajectories
Integrating sensing, targeting, actuation, and reporting is essential to achieve fully autonomous (CLRL-4) thrombolysis. A truly closed-loop nanosystem would not rely on incremental nanoparticle tweaks but on seamless system integration and control logic. In other words, the next breakthroughs will come from uniting components, smart triggers, feedback signals, and mechanistic actuators, rather than simply adding more drug carriers. The following six insights serve as a minimum roadmap to autonomy, outlining the key mechanisms needed to reach CLRL-4. Limited cross-disciplinary integration is currently a major barrier to CLRL-4 progress.
Insight 1: multi-signal logic is the only scalable route to specificity
Single-cue thrombolytic systems often suffer off-target activation and signal cross-talk in complex blood environments. Multi-signal logic (e.g. biochemical AND-gates and multi-threshold triggers) offers a path to high specificity. For example, Yin et al. created a DNA-origami nanodevice with a thrombin-sensitive triplex “lock” that releases tPA only when local clotting enzyme levels exceed a set threshold [31]. Likewise, Sun et al. developed clot-targeted liposomes bearing platelet and fibrin ligands, which remain inert until thrombin cleaves a peptide shell to unleash plasmin [103]. In contrast, standard single trigger tPA infusions cannot distinguish pathological clots from necessary hemostatic plugs. Combinatorial sensing markedly improves activation precision and enables safety cut-offs, at the cost of greater design complexity and validation burden. Current prototypes remain mostly single-trigger, risking false activation in vivo.
Insight 2: functional reporting must replace localization reporting
Theranostic systems need to report therapeutic efficacy (clot dissolution or reperfusion) rather than just nanoparticle location. Traditional probes that flag nanoparticle accumulation provide no guarantee that a vessel is recanalized. Alonso-Alonso et al.’s fibrin-targeted MRI nanoparticles, for instance, can visualize clot composition but cannot confirm that blood flow has been restored [15]. By contrast, Ye et al. demonstrated a pH-responsive colloidosome that co-delivers a thrombolytic (uPA) and a NIR-II tracer, disintegrating in acidic clot microenvironments to release both drug and a real-time imaging signal [34]. This enabled live tracking of clot clearance and reperfusion in deep vein thrombosis models, directly linking nanoparticle signal to functional outcome. Simply put, “NP present” does not equal “clot resolved.” Actionable reporting means sensing metrics like flow, fibrin degradation, or oxygenation that indicate successful lysis and can trigger therapy shut-off. Most current thrombolytic probes only indicate nanoparticle presence, not treatment success.
Insight 3: actuation must become penetrative, not merely surface-active
Thrombolytic actuation needs to reach the core of clots, not just erode their surface. Enzyme-only approaches such as free tPA or enzyme-loaded nanoparticles often stall against dense fibrin and platelet networks and contracted clot shells. Adding a physical penetration mechanism can overcome this barrier. Cabrera et al. showed that magnetic hyperthermia from iron oxide nanoparticles loosened the fibrin matrix and roughly doubled tPA-mediated lysis of refractory clots. In a more extreme paradigm, Zhang et al. developed self-propelled nanomotors that drilled through occlusions, achieving clot breakdown without any pharmacological agent. These strategies confirm that clot contraction and low porosity limit purely biochemical lysis, and that disrupting thrombus architecture by localized heating, cavitation, or mechanical boring can enable deeper enzyme penetration. The same mechanical activity, however, carries in vivo risks that must be characterized before translation. Forces or heat sufficient to fragment a clot can also injure the endothelium and subendothelial matrix at the lesion, promoting re-thrombosis or vasospasm; uncontrolled fragmentation can shed emboli that occlude distal micro vessels and convert a single occlusion into several inaccessible ones; magnetothermal and photothermal actuation can raise local temperature beyond the narrow tolerance of perivascular brain tissue; and aggressive disruption of a pathological thrombus risks destabilizing protective hemostatic plugs elsewhere, increasing bleeding risk. Off-target sequestration of magnetic or photoactive particles in the liver, spleen, and wider reticuloendothelial system is a further concern. These hazards argue for actuation that is spatially confined, power-limited, and gated by real-time imaging, and they reinforce the need for the safety governor described in Insight 4.
Insight 4: the safety governor should be treated as a first-class module
A fully autonomous thrombolytic system requires an internal safety governor: a fail-safe mechanism that prevents over-activation and off-target damage. Experimental platforms that localize drug action have begun to play this role. Sun et al.’s clot-anchored, thrombin-responsive nanoparticle confined plasmin generation to the thrombus, avoiding systemic fibrinogen degradation (a major cause of tPA-related bleeding). In another approach, Zhao et al. designed a metal–phenolic–protein nanoparticle that targets fibrin and releases tPA only at low pH; in rodent stroke models it achieved effective lysis with significantly reduced hemorrhage risk. Such built-in governors act as dose ceilings or “kill-switches,” automatically halting therapy when clot signals dissipate or if payload thresholds are reached. To translate this, the closed-loop system must default to an off state in the absence of a clot cue and demonstrate fail-safe shutdown in vivo under worst-case scenarios. No thrombolytic nanodevice yet has proven an automatic bleed-prevention shutoff in human studies.
Insight 5: closed-loop thrombolysis will be hybrid with thrombectomy
Rather than supplanting mechanical thrombectomy, future nanothrombolytics will complement it. Thrombectomy excels at removing bulk occlusions, yet patients often suffer “no-reflow” due to distal microthrombi and perfusion deficits even after a large clot is extracted. Adjunctive nano-actuators can target these inaccessible microclots and also dampen the reperfusion injury that arises from thromboinflammatory cascades. Horie et al. observed that successful recanalization did not always restore microcirculatory flow, implicating platelet-rich microemboli in capillaries. As a potential remedy, Zhang et al. developed a thrombin-triggered nanocarrier that both dissolves residual fibrin and delivers anti-inflammatory agents to the vessel wall, aiming to protect the endothelium post-recanalization. In practice, a hybrid approach could involve administering a closed-loop nanotherapeutic during or immediately after thrombectomy to clean up what the stent retriever cannot reach. This concept tackles secondary ischemia (microinfarcts, no-reflow, inflammation) and could improve outcomes beyond what mechanical recanalization alone achieves. However, incorporating an autonomous nanosystem into an acute stroke workflow raises challenges in timing, delivery route, and regulatory approval. Proving added benefit and safety of a combined nano–thrombectomy therapy is required for clinical adoption.
Insight 6: standardised flow-validated benchmarks will decide winners
Translation to CLRL-4 will hinge on rigorous performance benchmarks in flow conditions. Many thrombolytic nanotechnologies show promise in static in vitro assays but falter under physiological flow. For example, Loyau et al. demonstrated in a microfluidic arterial thrombosis model that tPA-induced fibrinolysis is markedly slower and less complete underflow than in static clot incubations, and that platelet activity modulates lysis efficacy. Similarly, Whyte et al. reported that plasminogen loading on activated platelets accelerates clot breakdown under flow, a mechanism absent in quiescent plasma. Such studies reveal that shear forces, blood elements, and hemodynamics substantially affect thrombolytic performance. It follows that any candidate closed-loop thrombolytic system should meet minimum efficacy criteria in a flow simulation (e.g. microfluidic clot-on-a-chip or an ex vivo perfusion loop) before animal testing. Establishing standard flow-based assays, including realistic geometry, shear rates, and even blood viscoelasticity, will be pivotal for objective head-to-head comparisons of platforms. In fact, progression from CLRL-2 to CLRL-3 may be gated by achieving defined targets in such flow models. The field currently lacks consensus flow benchmarks, impeding transparent evaluation of emerging systems.
Conclusions
Thrombolysis has traditionally been approached as a pharmacologic or mechanical event applied to a biologically complex but passively treated target. This review advances a different perspective: thrombi are dynamic, heterogeneous systems embedded within evolving hemodynamic and inflammatory environments, and therefore require therapies capable of sensing, adapting, and responding in real time. The convergence of nanotechnology, bioresponsive materials, and theranostic imaging now makes it conceivable to transition from open-loop intervention toward closed-loop vascular medicine.
Across literature, individual components of this paradigm are already emerging. Stimuli-responsive nanocarriers demonstrate selective activation in thrombin-rich or oxidative environments. Biomimetic and shape-engineered particles improve localization under flow. Physical actuators such as ultrasound, magnetic fields, and microrobotics offer controllable disruption of clot structure. Imaging-enabled platforms provide early forms of therapeutic reporting. Yet these innovations largely exist in isolation. Few systems integrate sensing, targeting, actuation, and feedback into a unified architecture capable of adaptive control.
The Closed-Loop Readiness Levels (CLRL) framework proposed here highlights both the progress and the gap. Most current technologies remain at early readiness stages, demonstrating single-trigger responsiveness or theranostic capability without autonomous regulation. The transition to fully closed-loop thrombolysis will require advances in multi-signal sensing, flow-robust targeting, quantitative reporting of reperfusion, and safe, stable control logic that can operate within the biological variability of patients.
Looking ahead, the greatest impact may arise not from incremental improvements in drug delivery alone, but from integrating engineering principles with vascular biology. Intelligent nanosystems that dynamically adjust lytic intensity, penetrate resistant clot domains, and terminate activity once reperfusion is achieved could improve efficacy while reducing hemorrhagic risk. Such platforms may ultimately redefine thrombolysis from a fixed intervention into a responsive therapeutic process.
In this context, closed-loop nanomedicine represents more than a technological evolution. It offers a conceptual shift toward precision thrombus management, where therapy is guided continuously by the biological state of the clot itself. Achieving this vision will require interdisciplinary collaboration across materials science, vascular medicine, control engineering, regulatory science, and translational biology. If realised, the paradigm could complement existing reperfusion care by making thrombolysis more spatially precise, measurable, and adaptable.
Supplementary Information
Acknowledgements
Not application.
Author contributions
Yizi Wang.Haishan Zhang.: Conceptualization, Methodology, Writing – original draft; Writing – review & editing.Tuan Wang.Shasha Yu and Yun Wang: Data curation, Formal analysis; Writing – review & editing.Xiaohan Qu.Liang Guo.: Supervision, Writing – review & editing. All authors reviewed and approved the final manuscript.
Funding
Not application.
Data availability
The data could be obtained from responding author.
Declarations
Human ethics and consent to participate.
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Yizi Wang, Tuan Wang and Xiaohan Qu contributed equally to this work.
Contributor Information
Yun Wang, Email: wyida_1021@126.com.
Shasha Yu, Email: ysscmu1h@163.com.
Haishan Zhang, Email: zhanghaishan99@sohu.com.
References
- 1.Feigin VL, et al. World Stroke Organization: global stroke fact sheet 2025. Int J Stroke. 2025;20:132–44. 10.1177/17474930241308142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Global. regional, and national burden of stroke and its risk factors, 1990–2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet Neurol. 2024;23:973–1003. 10.1016/s1474-4422(24)00369-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wendelboe AM, Raskob GE. Global Burden of Thrombosis. Circul Res. 2016;118:1340–7. 10.1161/CIRCRESAHA.115.306841. [DOI] [PubMed] [Google Scholar]
- 4.Fan BE, et al. Emerging thrombolysis technologies in vascular thrombosis. J Clin Med. 2025;14:7758. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Feigin VL, Owolabi MO. Pragmatic solutions to reduce the global burden of stroke: a World Stroke Organization-Lancet Neurology Commission. Lancet Neurol. 2023;22:1160–206. 10.1016/s1474-4422(23)00277-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ma H, Jiang Z, Xu J, Liu J, Guo ZN. Targeted nano-delivery strategies for facilitating thrombolysis treatment in ischemic stroke. Drug Deliv. 2021;28:357–71. 10.1080/10717544.2021.1879315. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Henderson SJ, Weitz JI, Kim PY. Fibrinolysis: strategies to enhance the treatment of acute ischemic stroke. J Thromb Haemost. 2018;16:1932–40. 10.1111/jth.14215. [DOI] [PubMed] [Google Scholar]
- 8.Dohle E, Ashok AH, Bhakta S, Induruwa I, Evans NR. Thrombus composition in ischaemic stroke: histological and radiological evaluation, and implications for acute clinical management. J Thromb Thrombolysis. 2025;58:355–69. 10.1007/s11239-025-03074-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bhatia K, et al. Current Endovascular Treatment of Acute Ischemic Stroke. Mo Med. 2020;117:480–9. [PMC free article] [PubMed] [Google Scholar]
- 10.Zhang B, et al. 30 years of thrombolysis for ischaemic stroke: expanded agents, widened windows and accelerated delivery. Stroke Vasc Neurol. 2026:svn-2025-004987. 10.1136/svn-2025-004987. [DOI] [PubMed]
- 11.Raha O, et al. Advances in mechanical thrombectomy for acute ischaemic stroke. BMJ Med. 2023;2:e000407. 10.1136/bmjmed-2022-000407. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kleindorfer DO, et al. 2021 guideline for the prevention of stroke in patients with stroke and transient ischemic attack: a guideline from the American Heart Association/American Stroke Association. Stroke. 2021;52:e364–467. 10.1161/STR.0000000000000375. [DOI] [PubMed] [Google Scholar]
- 13.Jun-O’Connell AH. Opinion: can we bust the fear of symptomatic intracerebral hemorrhage due to tPA? Front Neurol. 2024;15:1428726. 10.3389/fneur.2024.1428726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Merkler AE, et al. Safety Outcomes After Thrombolysis for Acute Ischemic Stroke in Patients With Recent Stroke. Stroke. 2017;48:2282–4. 10.1161/STROKEAHA.117.018119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Alonso-Alonso ML, et al. Fibrin-targeted nanoparticles for finding, visualizing and characterizing blood clots in acute ischemic stroke. Pharmaceutics. 2022. 10.3390/pharmaceutics14102156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kanji R, Gue YX, Memtsas V, Gorog DA. Fibrinolysis in platelet thrombi. Int J Mol Sci. 2021. 10.3390/ijms22105135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Wassélius J, Hall E, Ramgren B, Andersson T, Ullberg T. Procedural factors associated with successful recanalization in patients with acute ischemic stroke treated with endovascular thrombectomy-a nationwide register-based observational study. Interv Neuroradiol. 2024:15910199241248268. 10.1177/15910199241248268. [DOI] [PMC free article] [PubMed]
- 18.Horie N, et al. Recanalization does not always equate to reperfusion: no-reflow phenomenon after successful thrombectomy. Stroke. 2025;56:183–9. 10.1161/strokeaha.124.048994. [DOI] [PubMed]
- 19.Zhou Y, et al. Reperfusion Injury Is Associated With Poor Outcome in Patients With Recanalization After Thrombectomy. Stroke. 2023;54:96–104. 10.1161/strokeaha.122.039337. [DOI] [PubMed] [Google Scholar]
- 20.Ospel JM, et al. Challenges to widespread implementation of stroke thrombectomy. Stroke. 2024;55:2173–83. 10.1161/STROKEAHA.124.045889. [DOI] [PubMed] [Google Scholar]
- 21.Ren T, et al. Advances in nano-functional materials in targeted thrombolytic drug delivery. Molecules. 2024. 10.3390/molecules29102325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Zhan Y, et al. Application of stimuli-responsive nanomedicines for the treatment of ischemic stroke. Front Bioeng Biotechnol. 2023;11:1329959. 10.3389/fbioe.2023.1329959. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Liu C-H, Rethi L, Weng P-W, Nguyen T, H., Chuang AE. Y. Cutting-edge advances in nano/biomedicine: A review on transforming thrombolytic therapy. Biochem Pharmacol. 2024;229:116523. 10.1016/j.bcp.2024.116523. [DOI] [PubMed] [Google Scholar]
- 24.Ge X, et al. Surface engineering of thrombus-targeting nanocarriers for cardiovascular diseases: a short review. Adv Mater Interfaces. 2025;12:2500276. 10.1002/admi.202500276. [Google Scholar]
- 25.Li X, et al. Recent advances in targeted nanoparticle drug delivery systems for ischaemic stroke. Mater Adv. 2023;4:5003–17. 10.1039/D3MA00472D. [Google Scholar]
- 26.Chen HA, Ma YH, Hsu TY, Chen JP. Preparation of peptide and recombinant tissue plasminogen activator conjugated poly(lactic-co-glycolic acid) (PLGA) magnetic nanoparticles for dual targeted thrombolytic therapy. Int J Mol Sci. 2020. 10.3390/ijms21082690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Osmond MJ et al. Harnessing micrometer-scale tPA beads for high plasmin generation and accelerated fibrinolysis. bioRxiv (2024). 10.1101/2024.11.06.621942
- 28.Wang B, et al. tPA-anchored nanorobots for in vivo arterial recanalization at submillimeter-scale segments. Sci Adv. 2024;10:eadk8970. 10.1126/sciadv.adk8970. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Quan X, et al. Annexin V-Modified Platelet-Biomimetic Nanomedicine for Targeted Therapy of Acute Ischemic Stroke. Adv Healthc Mater. 2022;11:e2200416. 10.1002/adhm.202200416. [DOI] [PubMed] [Google Scholar]
- 30.Quan X, et al. Cryo-Shocked Platelet Coupled with ROS-Responsive Nanomedicine for Targeted Treatment of Thromboembolic Disease. ACS Nano. 2023;17:6519–33. 10.1021/acsnano.2c11865. [DOI] [PubMed] [Google Scholar]
- 31.Yin J, et al. An intelligent DNA nanodevice for precision thrombolysis. Nat Mater. 2024;23:854–62. 10.1038/s41563-024-01826-y. [DOI] [PubMed] [Google Scholar]
- 32.Sun M, et al. Direct delivery of plasmin using clot-anchoring thrombin-responsive nanoparticles for targeted fibrinolytic therapy. J Thromb Haemost. 2023;21:983–94. 10.1016/j.jtha.2022.11.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Zhang H, et al. Thrombus-targeted nanoparticles for thrombin-triggered thrombolysis and local inflammatory microenvironment regulation. J Controlled Release. 2021;339:195–207. 10.1016/j.jconrel.2021.06.043. [DOI] [PubMed] [Google Scholar]
- 34.Ye Y, et al. pH-Responsive Theranostic Colloidosome Drug Carriers Enable Real-Time Imaging of Targeted Thrombolytic Process with Near-Infrared-II for Deep Venous Thrombosis. Res (Wash D C). 2024;7:0388. 10.34133/research.0388. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Zhao X, et al. A Novel Metal-Phenolic‐Protein Hybrid Nanoparticles With Precise Delivery and Enhanced Penetration Effect for tPA Targeted Thrombolysis and Reduced Hemorrhagic Risk. Adv Healthc Mater. 2025. 10.1002/adhm.202503684. [DOI] [PubMed] [Google Scholar]
- 36.Yang M, et al. Swarming magnetic nanorobots bio-interfaced by heparinoid-polymer brushes for in vivo safe synergistic thrombolysis. Sci Adv. 2023;9:eadk7251. 10.1126/sciadv.adk7251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Pontius MHH, et al. Magnetically powered microwheel thrombolysis of occlusive thrombi in zebrafish. Proc Natl Acad Sci U S A. 2024;121:e2315083121. 10.1073/pnas.2315083121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Cabrera D, Eizadi Sharifabad M, Ranjbar JA, Telling ND, Harper AGS. Clot-targeted magnetic hyperthermia permeabilizes blood clots to make them more susceptible to thrombolysis. J Thromb Haemost. 2022;20:2556–70. 10.1111/jth.15846. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Yang Y, et al. Magnetic force enhances pharmacologic thrombolysis by unlocking the “knob-hole” structure of fibrin. Adv Healthc Mater. 2025;14:e01819. 10.1002/adhm.202501819. [DOI] [PubMed] [Google Scholar]
- 40.Zakharzhevskii M, et al. Shape anisotropic magnetic thrombolytic actuators: synthesis and systematic behavior study. J Mater Chem B. 2021. 10.1039/D1TB00783A. [DOI] [PubMed] [Google Scholar]
- 41.Jheng P-R, et al. Cold atmospheric plasma-enabled platelet vesicle incorporated iron oxide nano-propellers for thrombolysis. Mater Today Bio. 2023;23:100876. 10.1016/j.mtbio.2023.100876. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Zhang Y, et al. Targeted thrombolytic therapy with metal–organic-framework-derived carbon based platforms with multimodal capabilities. ACS Appl Mater Interfaces. 2021;13:24453–62. 10.1021/acsami.1c03134. [DOI] [PubMed] [Google Scholar]
- 43.Choi W, et al. Targeted thrombolysis by magnetoacoustic particles in photothrombotic stroke model. Biomater Res. 2022;26:58. 10.1186/s40824-022-00298-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Tang X, et al. Synergistic integration and pharmacomechanical function of enzyme-magnetite nanoparticle swarms for low-dose fast thrombolysis. Small. 2022;18:2202848. 10.1002/smll.202202848. [DOI] [PubMed] [Google Scholar]
- 45.Yin J, et al. An intelligent DNA nanodevice for precision thrombolysis. Nat Mater. 2024;23:854–62. 10.1038/s41563-024-01826-y. [DOI] [PubMed] [Google Scholar]
- 46.Zhang H, et al. Molecularly self-fueled nano-penetrator for nonpharmaceutical treatment of thrombosis and ischemic stroke. Nat Commun. 2023;14:255. 10.1038/s41467-023-35895-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Xu RG, Ariëns RAS. Insights into the composition of stroke thrombi: heterogeneity and distinct clot areas impact treatment. Haematologica. 2020;105:257–9. 10.3324/haematol.2019.238816. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Staessens S, et al. Structural analysis of ischemic stroke thrombi: histological indications for therapy resistance. Haematologica. 2020;105:498–507. 10.3324/haematol.2019.219881. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Di Meglio L, et al. Acute ischemic stroke thrombi have an outer shell that impairs fibrinolysis. Neurology. 2019;93:e1686–98. 10.1212/wnl.0000000000008395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Khismatullin RR, et al. Quantitative Morphology of Cerebral Thrombi Related to Intravital Contraction and Clinical Features of Ischemic Stroke. Stroke. 2020;51:3640–50. 10.1161/strokeaha.120.031559. [DOI] [PubMed] [Google Scholar]
- 51.Pir GJ, et al. Platelet-neutrophil association in NETs-rich areas in the retrieved AIS patient thrombi. Int J Mol Sci. 2022. 10.3390/ijms232214477. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Mereuta OM, et al. Shell Versus Core Architecture and Biology of Thrombi in Acute Ischemic Stroke: A Systematic Review. Clin Appl Thromb Hemost. 2023;29:10760296231213632. 10.1177/10760296231213632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Cahalane R, et al. A review on the association of thrombus composition with mechanical and radiological imaging characteristics in acute ischemic stroke. J Biomech. 2021;129:110816. 10.1016/j.jbiomech.2021.110816. [DOI] [PubMed] [Google Scholar]
- 54.Ahn SH, et al. Histologic features of acute thrombi retrieved from stroke patients during mechanical reperfusion therapy. Int J Stroke. 2016;11:1036–44. 10.1177/1747493016641965. [DOI] [PubMed] [Google Scholar]
- 55.He W, Wu Z, Liu Y, Ye Z. Neutrophil extracellular traps in ischemic stroke: mechanisms, clinical implications, and therapeutic potential. Front Neurol. 2025;16:1641985. [DOI] [PMC free article] [PubMed]
- 56.Park S, Shim JR, Goh RY, Kim DH, Han JY. Evaluating histone H3.1 as a biomarker for acute ischemic stroke: insights into NETs and stroke pathophysiology. Blood Res. 2024;59:40. 10.1007/s44313-024-00047-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Wienkamp AK, Erpenbeck L, Rossaint J. Platelets in the NETworks interweaving inflammation and thrombosis. Front Immunol. 2022;13:953129. 10.3389/fimmu.2022.953129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Kumar R, Patil G, Dayal S. NLRP3-induced NETosis: a potential therapeutic target for ischemic thrombotic diseases? Cells. 2023. 10.3390/cells12232709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Hashemi P, et al. NETosis in ischemic/reperfusion injuries: an organ-based review. Life Sci. 2022;290:120158. 10.1016/j.lfs.2021.120158. [DOI] [PubMed] [Google Scholar]
- 60.Denorme F, et al. Neutrophil extracellular traps regulate ischemic stroke brain injury. J Clin Invest. 2022. 10.1172/jci154225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61.Ducroux C, et al. Thrombus Neutrophil Extracellular Traps Content Impair tPA-Induced Thrombolysis in Acute Ischemic Stroke. Stroke. 2018;49:754–7. 10.1161/strokeaha.117.019896. [DOI] [PubMed] [Google Scholar]
- 62.Laridan E, et al. Neutrophil extracellular traps in ischemic stroke thrombi. Ann Neurol. 2017;82:223–32. 10.1002/ana.24993. [DOI] [PubMed] [Google Scholar]
- 63.Muthard RW, Diamond SL. Blood clots are rapidly assembled hemodynamic sensors: flow arrest triggers intraluminal thrombus contraction. Arterioscler Thromb Vasc Biol. 2012;32:2938–45. 10.1161/atvbaha.112.300312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 64.Kim OV, Litvinov RI, Alber MS, Weisel JW. Quantitative structural mechanobiology of platelet-driven blood clot contraction. Nat Commun. 2017;8:1274. 10.1038/s41467-017-00885-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 65.Tutwiler V, et al. Blood clot contraction differentially modulates internal and external fibrinolysis. J Thromb Haemost. 2019;17:361–70. 10.1111/jth.14370. [DOI] [PubMed] [Google Scholar]
- 66.Pathare SJ, Eng W, Lee S-J, Ramasubramanian AK. Fibrin prestress due to platelet aggregation and contraction increases clot stiffness. Biophys Rep. 2021;1:100022. 10.1016/j.bpr.2021.100022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 67.Michael C, et al. Combined computational modeling and experimental study of the biomechanical mechanisms of platelet-driven contraction of fibrin clots. Commun Biol. 2023;6:869. 10.1038/s42003-023-05240-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 68.Calaminus SDJ, et al. MyosinIIa contractility is required for maintenance of platelet structure during spreading on collagen and contributes to thrombus stability. J Thromb Haemost. 2007;5:2136–45. 10.1111/j.1538-7836.2007.02696.x. [DOI] [PubMed] [Google Scholar]
- 69.Pathare SJ, Eng W, Lee SJ, Ramasubramanian AK. Fibrin prestress due to platelet aggregation and contraction increases clot stiffness. Biophys Rep. 2021;1:100022. 10.1016/j.bpr.2021.100022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 70.Zalewski J, et al. Plasma fibrin clot phenotype independently affects intracoronary thrombus ultrastructure in patients with acute myocardial infarction. Thromb Haemost. 2015;113:1258–69. 10.1160/th14-09-0801. [DOI] [PubMed] [Google Scholar]
- 71.Du J, Aspray E, Fogelson A. Computational investigation of platelet thrombus mechanics and stability in stenotic channels. J Biomech. 2021;122:110398. 10.1016/j.jbiomech.2021.110398. [DOI] [PubMed] [Google Scholar]
- 72.Lin L, et al. Ultrasound-responsive theranostic platform for the timely monitoring and efficient thrombolysis in thrombi of tPA resistance. Nat Commun. 2024;15:6610. 10.1038/s41467-024-50741-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 73.Rahimi F, Zarvani A, Fayedeh F, Sharifi-Razavi A. Low-dose vs. standard-dose alteplase for acute ischemic stroke: a retrospective cohort study of safety and efficacy in an Iranian population. BMC Neurol. 2025;25:395. 10.1186/s12883-025-04413-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 74.Munoz A, et al. A review of mechanical thrombectomy techniques for acute ischemic stroke. Interv Neuroradiol. 2023;29:450–8. 10.1177/15910199221084481. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 75.Patil S, et al. Characterising acute ischaemic stroke thrombi: insights from histology, imaging and emerging impedance-based technologies. Stroke Vasc Neurol. 2022;7:353–63. 10.1136/svn-2021-001038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 76.Messina P, et al. Impedance-based sensors discriminate among different types of blood thrombi with very high specificity and sensitivity. J Neurointerv Surg. 2023;15:526–30. 10.1136/neurintsurg-2021-018631. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 77.Wang S, et al. Accelerating thrombolysis using a precision and clot-penetrating drug delivery strategy by nanoparticle-shelled microbubbles. Sci Adv. 2020;6:eaaz8204. 10.1126/sciadv.aaz8204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 78.Fournier L, et al. rtPA-loaded fucoidan polymer microbubbles for the targeted treatment of stroke. Biomaterials. 2023;303:122385. 10.1016/j.biomaterials.2023.122385. [DOI] [PubMed] [Google Scholar]
- 79.Ahmaditabar P, Mahmoodi M, Taheri RA, Asefnejad A. Preparation and in vitro evaluation of tissue plasminogen activator-loaded nanoliposomes with anticoagulant coating. Biochimica et Biophysica Acta (BBA). 2024;1868:130704. 10.1016/j.bbagen.2024.130704. [DOI] [PubMed] [Google Scholar]
- 80.Wu J, et al. Smart Thrombosis Care: The Rise of Closed-Loop Diagnosis-to-Treatment Nano Systems. Int J Nanomed. 2025;20:7851–68. 10.2147/ijn.S530884. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 81.Kujiraoka H, Tsuchiyama T, Inagaki D, Yoshida K, Fukamizu S. Comparison of the efficacy of excimer laser coronary angioplasty for ST-segment elevation myocardial infarction with onset-to-balloon time. Lasers Med Sci. 2023;38:126. 10.1007/s10103-023-03789-z. [DOI] [PubMed] [Google Scholar]
- 82.Zhao Y, et al. Hydrogen peroxide-responsive platelet membrane-coated nanoparticles for thrombus therapy. Biomater Sci. 2021;9:2696–708. 10.1039/d0bm02125c. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 83.Lovegrove JT, et al. Margination of 2D Platelet Microparticles in Blood. ACS Macro Lett. 2023;12:344–9. 10.1021/acsmacrolett.2c00718. [DOI] [PubMed] [Google Scholar]
- 84.Blanco E, Shen H, Ferrari M. Principles of nanoparticle design for overcoming biological barriers to drug delivery. Nat Biotechnol. 2015;33:941–51. 10.1038/nbt.3330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 85.Liu C, et al. Discoidal nanoparticles exploit thrombus-induced shear gradients to enhance site-specific thrombolysis in stroke. Acta Biomater. 2025;201:471–84. 10.1016/j.actbio.2025.06.027. [DOI] [PubMed] [Google Scholar]
- 86.Nie X et al. Futile Recanalization after Endovascular Therapy in Acute Ischemic Stroke. Biomed Res Int 2018, 5879548 (2018). 10.1155/2018/5879548 [DOI] [PMC free article] [PubMed]
- 87.Quan X, et al. Neutrophil-mimetic upconversion photosynthetic nanosystem derived from microalgae for targeted treatment of thromboembolic stroke. ACS Nano. 2024;18:30307–20. 10.1021/acsnano.4c06247. [DOI] [PubMed] [Google Scholar]
- 88.Kheradmandi M, Ackers I, Burdick MM, Malgor R, Farnoud AM. Targeting dysfunctional vascular endothelial cells using immunoliposomes under flow conditions. Cell Mol Bioeng. 2020;13:189–99. 10.1007/s12195-020-00616-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 89.Sun M, et al. Combination targeting of ‘platelets + fibrin’ enhances clot anchorage efficiency of nanoparticles for vascular drug delivery. Nanoscale. 2020;12:21255–70. 10.1039/d0nr03633a. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 90.Mitchell MJ, et al. Engineering precision nanoparticles for drug delivery. Nat Rev Drug Discov. 2021;20:101–24. 10.1038/s41573-020-0090-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 91.Lane LA. Physics in nanomedicine: phenomena governing the in vivo performance of nanoparticles. Appl Phys Rev. 2020;7:011316. 10.1063/1.5052455. [Google Scholar]
- 92.Naskar A, Kilari S, Baranwal G, Kane J, Misra S. Nanoparticle-based drug delivery for vascular applications. Bioengineering (Basel). 2024. 10.3390/bioengineering11121222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 93.Zhang H, et al. Shear stress regulation of nanoparticle uptake in vascular endothelial cells. Regen Biomater. 2023;10:rbad047. 10.1093/rb/rbad047. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 94.Jiang K, et al. Protein corona on brain targeted nanocarriers: Challenges and prospects. Adv Drug Deliv Rev. 2023;202:115114. 10.1016/j.addr.2023.115114. [DOI] [PubMed] [Google Scholar]
- 95.Mirshafiee V, Mahmoudi M, Lou K, Cheng J, Kraft ML. Protein corona significantly reduces active targeting yield. Chem Commun (Camb). 2013;49:2557–9. 10.1039/c3cc37307j. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 96.Lin ML, et al. Targeted thrombolysis with magnetic nanotherapeutics: a translational assessment. Pharmaceutics. 2024. 10.3390/pharmaceutics16050596. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 97.Yang M, et al. Swarming magnetic nanorobots bio-interfaced by heparinoid-polymer brushes for in vivo safe synergistic thrombolysis. Sci Adv. 2023;9:eadk7251. 10.1126/sciadv.adk7251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 98.Sterling KM, et al. Prospective multicenter international registry of ultrasound-facilitated catheter-directed thrombolysis in intermediate-high and high-risk pulmonary embolism (KNOCOUT PE). Circ Cardiovasc Interv. 2024;17:e013448. 10.1161/CIRCINTERVENTIONS.123.013448. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 99.Alexandrov AV, et al. Safety and efficacy of sonothrombolysis for acute ischaemic stroke: a multicentre, double-blind, phase 3, randomised controlled trial. Lancet Neurol. 2019;18:338–47. 10.1016/S1474-4422(19)30026-2. [DOI] [PubMed] [Google Scholar]
- 100.Qureshi A, et al. MRI-based modelling of left atrial flow and coagulation to predict risk of thrombogenesis in atrial fibrillation. Med Image Anal. 2025;101:103475. 10.1016/j.media.2025.103475. [DOI] [PubMed] [Google Scholar]
- 101.Shi B, et al. Accelerating thrombolysis of arterial thrombus with NO-MBs UTMD therapy. Eur J Pharm Biopharm. 2024;205:114566. 10.1016/j.ejpb.2024.114566. [DOI] [PubMed] [Google Scholar]
- 102.Zeng Z, Christodoulides A, Alves NJ. Real-time tracking of fibrinolysis under constant wall shear and various pulsatile flows in an in-vitro thrombolysis model. Bioeng Transl Med. 2023;8:e10511. 10.1002/btm2.10511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 103.Sun M, et al. Direct delivery of plasmin using clot-anchoring thrombin-responsive nanoparticles for targeted fibrinolytic therapy. J Thromb Haemost. 2023;21:983–94. 10.1016/j.jtha.2022.11.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 104.Lin L, et al. Ultrasound-responsive theranostic platform for the timely monitoring and efficient thrombolysis in thrombi of tPA resistance. Nat Commun. 2024. 10.1038/s41467-024-50741-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 105.Wang Y, et al. Nanotechnology-donated ischemic stroke therapeutics: evolving strategies from the basic to the cutting-edge. Theranostics. 2026;16:3173–205. 10.7150/thno.127504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 106.Paljärvi L, Rehncrona S, Söderfeldt B, Olsson Y, Kalimo H. Brain lactic acidosis and ischemic cell damage: quantitative ultrastructural changes in capillaries of rat cerebral cortex. Acta Neuropathol. 1983;60:232–40. 10.1007/bf00691871. [DOI] [PubMed] [Google Scholar]
- 107.Cui W, et al. pH gradient difference around ischemic brain tissue can serve as a trigger for delivering polyethylene glycol-conjugated urokinase nanogels. J Control Release. 2016;225:53–63. 10.1016/j.jconrel.2016.01.028. [DOI] [PubMed] [Google Scholar]
- 108.Lv W, et al. Bioengineered boronic ester modified dextran polymer nanoparticles as reactive oxygen species responsive nanocarrier for ischemic stroke treatment. ACS Nano. 2018;12:5417–26. 10.1021/acsnano.8b00477. [DOI] [PubMed] [Google Scholar]
- 109.Lu Y, et al. Microthrombus-Targeting Micelles for Neurovascular Remodeling and Enhanced Microcirculatory Perfusion in Acute Ischemic Stroke. Adv Mater. 2019;31:e1808361. 10.1002/adma.201808361. [DOI] [PubMed] [Google Scholar]
- 110.Guo X, et al. Thrombin-Responsive, Brain-Targeting Nanoparticles for Improved Stroke Therapy. ACS Nano. 2018;12:8723–32. 10.1021/acsnano.8b04787. [DOI] [PubMed] [Google Scholar]
- 111.Ren Y, et al. Synthesis of protein vesicles for extending time window of ischemic stroke treatment through microcirculatory thrombolysis. Chem Eng J. 2023;455:140705. 10.1016/j.cej.2022.140705. [Google Scholar]
- 112.Ma H, Jiang Z, Xu J, Liu J, Guo Z-N. Targeted nano-delivery strategies for facilitating thrombolysis treatment in ischemic stroke. Drug Delivery. 2021;28:357–71. 10.1080/10717544.2021.1879315. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 113.Li R, et al. Relieving thrombo-inflammation with acid-triggered polymersomes toward ischemic stroke therapy. Nano Today. 2024;54:102114. 10.1016/j.nantod.2023.102114. [Google Scholar]
- 114.Raghunathan S, Rayes J, Sen Gupta A. Platelet-inspired nanomedicine in hemostasis thrombosis and thromboinflammation. J Thromb Haemost. 2022;20:1535–49. 10.1111/jth.15734. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 115.Wang S, et al. Drug Targeting via Platelet Membrane–Coated Nanoparticles. Small Struct. 2020;1:2000018. 10.1002/sstr.202000018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 116.Han H, Bártolo R, Li J, Shahbazi M-A, Santos HA. Biomimetic platelet membrane-coated nanoparticles for targeted therapy. Eur J Pharm Biopharm. 2022;172:1–15. 10.1016/j.ejpb.2022.01.004. [DOI] [PubMed] [Google Scholar]
- 117.Wei X, et al. Nanoparticle functionalization with platelet membrane enables multifactored biological targeting and detection of atherosclerosis. ACS Nano. 2018;12:109–16. 10.1021/acsnano.7b07720. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 118.Xu J, et al. Engineered nanoplatelets for targeted delivery of plasminogen activators to reverse thrombus in multiple mouse thrombosis models. Adv Mater. 2020;32:1905145. 10.1002/adma.201905145. [DOI] [PubMed] [Google Scholar]
- 119.Wang S, et al. Platelet membrane-functionalized nanoparticles with improved targeting ability and lower hemorrhagic risk for thrombolysis therapy. J Controlled Release. 2020;328:78–86. 10.1016/j.jconrel.2020.08.030. [DOI] [PubMed] [Google Scholar]
- 120.Wang S, et al. Drug Targeting via Platelet Membrane-Coated Nanoparticles. Small Struct. 2020;1. 10.1002/sstr.202000018. [DOI] [PMC free article] [PubMed]
- 121.Zhu L, et al. Erythrocyte-Membrane‐Camouflaged Magnetic Nanocapsules With Photothermal/Magnetothermal Effects for Thrombolysis. Adv Healthc Mater. 2024;13:e2400127. 10.1002/adhm.202400127. [DOI] [PubMed] [Google Scholar]
- 122.Mihalko EP, et al. Fibrin-modulating nanogels for treatment of disseminated intravascular coagulation. Blood Adv. 2021;5:613–27. 10.1182/bloodadvances.2020003046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 123.Tsivgoulis G, Culp WC, Alexandrov AV. Ultrasound enhanced thrombolysis in acute arterial ischemia. Ultrasonics. 2008;48:303–11. 10.1016/j.ultras.2007.11.008. [DOI] [PubMed] [Google Scholar]
- 124.Rubiera M, Alexandrov AV. Sonothrombolysis in the management of acute ischemic stroke. Am J Cardiovasc Drugs. 2010;10:5–10. 10.2165/11316850-000000000-00000. [DOI] [PubMed] [Google Scholar]
- 125.Sarfati P, Lalatonne Y, Chauvierre C. Drug-free thrombolysis mediated by physically activated micro/nanoparticles. Adv Funct Mater. 2026;e28440. 10.1002/adfm.202528440. [Google Scholar]
- 126.Picot A, et al. Engineered nanoconjugates for simultaneous detection and degradation of stroke-associated microthrombi. Theranostics. 2026;16:1082–102. 10.7150/thno.119705. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 127.Lu K, et al. PEGylated Ultrasmall Iron Oxide Nanoparticles as MRI Contrast Agents for Vascular Imaging and Real-Time Monitoring. ACS Nano. 2025;19:3519–30. 10.1021/acsnano.4c13356. [DOI] [PubMed] [Google Scholar]
- 128.Jacqmarcq C, et al. MRI-based microthrombi detection in stroke with polydopamine iron oxide. Nat Commun. 2024;15:5070. 10.1038/s41467-024-49480-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 129.Alonso-Alonso ML, et al. Fibrin-targeted nanoparticles for finding, visualizing and characterizing blood clots in acute ischemic stroke. Pharmaceutics. 2022;14:2156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 130.Wang S, et al. Accelerating thrombolysis using a precision and clot-penetrating drug delivery strategy by nanoparticle-shelled microbubbles. Sci Adv. 2020;6:eaaz8204. 10.1126/sciadv.aaz8204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 131.Song J, et al. Near-infrared-II photoacoustic imaging and photo-triggered synergistic treatment of thrombosis via fibrin-specific homopolymer nanoparticles. Nat Commun. 2023;14:6881. 10.1038/s41467-023-42691-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 132.Cao Z, et al. Thrombus-targeted nano-agents for NIR-II diagnostic fluorescence imaging-guided flap thromboembolism multi-model therapy. J Nanobiotechnol. 2022;20:447. 10.1186/s12951-022-01649-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 133.Song J, et al. Near-infrared-II photoacoustic imaging and photo-triggered synergistic treatment of thrombosis via fibrin-specific homopolymer nanoparticles. Nat Commun. 2023;14:6881. 10.1038/s41467-023-42691-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 134.Fithri NA, et al. Gold-iron oxide nanoparticle: a unique multimodal theranostic approach for thrombosis. Appl Mater Today. 2023;31:101750. 10.1016/j.apmt.2023.101750. [Google Scholar]
- 135.Zhang H, et al. Thrombus-targeted nanoparticles for thrombin-triggered thrombolysis and local inflammatory microenvironment regulation. J Control Release. 2021;339:195–207. 10.1016/j.jconrel.2021.06.043. [DOI] [PubMed] [Google Scholar]
- 136.Zhang H, et al. Thrombin-Responsive and Sequential Targeted Nanoplatform for Synergistic Thrombolysis Therapy. ACS Appl Mater Interfaces. 2025;17:16696–707. 10.1021/acsami.5c01658. [DOI] [PubMed] [Google Scholar]
- 137.Yu W, et al. Rescuing ischemic stroke by biomimetic nanovesicles through accelerated thrombolysis and sequential ischemia-reperfusion protection. Acta Biomater. 2022;140:625–40. 10.1016/j.actbio.2021.12.009. [DOI] [PubMed] [Google Scholar]
- 138.Xu J, et al. Sequentially Site-Specific Delivery of Thrombolytics and Neuroprotectant for Enhanced Treatment of Ischemic Stroke. ACS Nano. 2019;13:8577–88. 10.1021/acsnano.9b01798. [DOI] [PubMed] [Google Scholar]
- 139.Korin N, et al. Shear-activated nanotherapeutics for drug targeting to obstructed blood vessels. Science. 2012;337:738–42. 10.1126/science.1217815. [DOI] [PubMed] [Google Scholar]
- 140.Vasquez Giuliano L, et al. Response of shear-activated nanotherapeutic particles in a clot-obstructed blood vessel by CFD-DEM simulations. Can J Chem Eng. 2022;100:3562–74. 10.1002/cjce.24502. [Google Scholar]
- 141.Mohammed SA, et al. Fibrin-targeted phase shift microbubbles for the treatment of microvascular obstruction. Nanotheranostics. 2024;8:33–47. 10.7150/ntno.85092. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 142.Zhang W, et al. Low-intensity focused ultrasound-responsive nanobubbles enhance thrombus targeting and penetration for highly effective thrombolytic therapy. Small Struct. 2025;6:2400541. 10.1002/sstr.202400541. [Google Scholar]
- 143.Wang S, et al. Platelet membrane-functionalized nanoparticles with improved targeting ability and lower hemorrhagic risk for thrombolysis therapy. J Control Release. 2020;328:78–86. 10.1016/j.jconrel.2020.08.030. [DOI] [PubMed] [Google Scholar]
- 144.Juenet M, et al. Thrombolytic therapy based on fucoidan-functionalized polymer nanoparticles targeting P-selectin. Biomaterials. 2018;156:204–16. 10.1016/j.biomaterials.2017.11.047. [DOI] [PubMed] [Google Scholar]
- 145.Bi F, Zhang J, Su Y, Tang YC, Liu JN. Chemical conjugation of urokinase to magnetic nanoparticles for targeted thrombolysis. Biomaterials. 2009;30:5125–30. 10.1016/j.biomaterials.2009.06.006. [DOI] [PubMed] [Google Scholar]
- 146.Zhang H, Cao K, Wang C, Tian Y, Hou L. Thrombin and NIR dual-responsive system driven by heat-gas for thrombus targeted therapy. Chem Eng J. 2024;498:155054. 10.1016/j.cej.2024.155054. [Google Scholar]
- 147.Farnoudian-Habibi A, et al. Development a novel nano-platform for thrombolysis acceleration by thrombin sensitive polymer-peptide hybrid nancapsules. Int J Pharm. 2024;663:124561. 10.1016/j.ijpharm.2024.124561. [DOI] [PubMed] [Google Scholar]
- 148.U.S. Food and Drug Administration. Marketing Submission Recommendations for a Predetermined Change Control Plan for Artificial Intelligence-Enabled Device Software Functions (Center for Devices and Radiological Health, 2023).
- 149.Parvinian B, Scully C, Wiyor H, Kumar A, Weininger S. Regulatory considerations for physiological closed-loop controlled medical devices used for automated critical care: Food and Drug Administration workshop discussion topics. Anesth Analg. 2018;126:1916–25. 10.1213/ane.0000000000002329. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 150.Brogi E, Cyr S, Kazan R, Giunta F, Hemmerling TM. Clinical performance and safety of closed-loop systems: a systematic review and meta-analysis of randomized controlled trials. Anesth Analg. 2017;124:446–55. 10.1213/ane.0000000000001372. [DOI] [PubMed] [Google Scholar]
- 151.Lin L, et al. Ultrasound-responsive theranostic platform for the timely monitoring and efficient thrombolysis in thrombi of tPA resistance. Nat Commun. 2024;15:6610. 10.1038/s41467-024-50741-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 152.Han M, et al. Rapid synergistic thrombolysis of ischemic stroke guided by high-resolution and high-speed photoacoustic cerebrovascular imaging. Photoacoustics. 2025;43:100722. 10.1016/j.pacs.2025.100722. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 153.Singh D, Dilnawaz F, Sahoo SK. Challenges of Moving Theranostic Nanomedicine Into The Clinic. Nanomedicine. 2020;15:111–4. 10.2217/nnm-2019-0401. [DOI] [PubMed] [Google Scholar]
- 154.Gawne PJ, Ferreira M, Papaluca M, Grimm J, Decuzzi P. New opportunities and old challenges in the clinical translation of nanotheranostics. Nat Rev Mater. 2023;8:783–98. 10.1038/s41578-023-00581-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 155.Đorđević S, et al. Current hurdles to the translation of nanomedicines from bench to the clinic. Drug Deliv Transl Res. 2022;12:500–25. 10.1007/s13346-021-01024-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 156.Agrahari V, Agrahari V. Facilitating the translation of nanomedicines to a clinical product: challenges and opportunities. Drug Discov Today. 2018;23:974–91. 10.1016/j.drudis.2018.01.047. [DOI] [PubMed] [Google Scholar]
- 157.U.S. Food and Drug Administration. Physiologic Closed-Loop Controlled Devices: Guidance for Industry and Food and Drug Administration Staff (Center for Drug Evaluation and Research, 2018).
- 158.Operti MC, et al. PLGA-based nanomedicines manufacturing: technologies overview and challenges in industrial scale-up. Int J Pharm. 2021;605:120807. 10.1016/j.ijpharm.2021.120807. [DOI] [PubMed] [Google Scholar]
- 159.Herdiana Y, Wathoni N, Shamsuddin S, Muchtaridi M. Scale-up polymeric-based nanoparticles drug delivery systems: development and challenges. OpenNano. 2022;7:100048. 10.1016/j.onano.2022.100048. [Google Scholar]
- 160.160 Mülhopt S, et al. Characterization of Nanoparticle Batch-To-Batch Variability. Nanomaterials (Basel). 2018;8. 10.3390/nano8050311. [DOI] [PMC free article] [PubMed]
- 161.Halamoda-Kenzaoui B, et al. Methodological needs in the quality and safety characterisation of nanotechnology-based health products: Priorities for method development and standardisation. J Controlled Release. 2021;336:192–206. 10.1016/j.jconrel.2021.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 162.Sivasubramanian M, Lin L-J, Wang Y-C, Yang C-S, Lo L-W. Industrialization’s eye view on theranostic nanomedicine. Front Chem. 2022;10:918715. [DOI] [PMC free article] [PubMed]
- 163.Lim J-M, et al. Ultra-High Throughput Synthesis of Nanoparticles with Homogeneous Size Distribution Using a Coaxial Turbulent Jet Mixer. ACS Nano. 2014;8:6056–65. 10.1021/nn501371n. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 164.Franzé S, Selmin F, Samaritani E, Minghetti P, Cilurzo F. Lyophilization of liposomal formulations: still necessary, still challenging. Pharmaceutics. 2018. 10.3390/pharmaceutics10030139. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 165.Andreana I, et al. Freeze drying of polymer nanoparticles and liposomes exploiting different saccharide-based approaches. Materials. 2023. 10.3390/ma16031212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 166.Luo W-C, et al. Impact of formulation on the quality and stability of freeze-dried nanoparticles. Eur J Pharm Biopharm. 2021;169:256–67. 10.1016/j.ejpb.2021.10.014. [DOI] [PubMed] [Google Scholar]
- 167.Wang Y, Grainger DW. Lyophilized liposome-based parenteral drug development: Reviewing complex product design strategies and current regulatory environments. Adv Drug Deliv Rev. 2019;151–152:56–71. 10.1016/j.addr.2019.03.003. [DOI] [PubMed] [Google Scholar]
- 168.Ma X, et al. Nanotechnology in healthcare, and its safety and environmental risks. J Nanobiotechnol. 2024;22:715. 10.1186/s12951-024-02901-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 169.Wu Y, et al. Recent Advances in the Development of Theranostic Nanoparticles for Cardiovascular Diseases. Nanotheranostics. 2021;5:499–514. 10.7150/ntno.62730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 170.Paradise J. Regulating Nanomedicine at the Food and Drug Administration. AMA J Ethics. 2019;21:E347–355. 10.1001/amajethics.2019.347. [DOI] [PubMed] [Google Scholar]
- 171.Goble JA. The Potential Effect of the 21st Century Cures Act on Drug Development. J Manag Care Spec Pharm. 2018;24:677–81. 10.18553/jmcp.2018.24.7.677. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 172.Mennella C, Maniscalco U, De Pietro G, Esposito M. Ethical and regulatory challenges of AI technologies in healthcare: a narrative review. Heliyon. 2024;10:e26297. 10.1016/j.heliyon.2024.e26297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 173.Shetty DK, et al. Analyzing AI regulation through literature and current trends. J Open Innov Technol Mark Complex. 2025;11:100508. 10.1016/j.joitmc.2025.100508. [Google Scholar]
- 174.Visentin R, Dalla Man C, Kovatchev B, Cobelli C. The University of Virginia/Padova type 1 diabetes simulator matches the glucose traces of a clinical trial. Diabetes Technol Ther. 2014;16:428–34. 10.1089/dia.2013.0377. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 175.Cobelli C, Kovatchev B. Developing the UVA/Padova type 1 diabetes simulator: modeling, validation, refinements, and utility. J Diabetes Sci Technol. 2023;17:1493–505. 10.1177/19322968231195081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 176.Toffanin C, Kozak M, Sumnik Z, Cobelli C, Petruzelkova L. In silico trials of an open-source Android-based artificial pancreas: a new paradigm to test safety and efficacy of do-it-yourself systems. Diabetes Technol Ther. 2020;22:112–20. 10.1089/dia.2019.0375. [DOI] [PubMed] [Google Scholar]
- 177.T Sun F, J Morrell M. Closed-loop Neurostimulation: The Clinical Experience. Neurotherapeutics. 2014;11:553–63. 10.1007/s13311-014-0280-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 178.Spanò R, et al. Enhancing Thrombolysis Safety in Post-Acute Ischemic Stroke with Tissue Plasminogen Activator-Associated Microparticles. ACS Nano. 2025;19:22882–99. 10.1021/acsnano.5c01499. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 179.Leng T, Xiong ZG. Treatment for ischemic stroke: From thrombolysis to thrombectomy and remaining challenges. Brain Circ. 2019;5:8–11. 10.4103/bc.bc_36_18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 180.Lu Y, et al. Microbubble-Mediated Sonothrombolysis Improves Outcome After Thrombotic Microembolism-Induced Acute Ischemic Stroke. Stroke. 2016;47:1344–53. 10.1161/STROKEAHA.115.012056. [DOI] [PubMed] [Google Scholar]
- 181.Marosfoi MG, et al. Shear-Activated Nanoparticle Aggregates Combined With Temporary Endovascular Bypass to Treat Large Vessel Occlusion. Stroke. 2015;46:3507–13. 10.1161/strokeaha.115.011063. [DOI] [PubMed] [Google Scholar]
- 182.Ruckdeschel JC, Peters T, Lee KT. Fibrinogen catabolism in rats fed thrombogenic diets. Atherosclerosis. 1972;16:277–85. 10.1016/0021-9150(72)90078-0. [DOI] [PubMed] [Google Scholar]
- 183.Vilar R, Fish RJ, Casini A, Neerman-Arbez M. Fibrin(ogen) in human disease: both friend and foe. Haematologica. 2020;105:284–96. 10.3324/haematol.2019.236901. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 184.Balkenhol J, et al. Comparison of the central human and mouse platelet signaling cascade by systems biological analysis. BMC Genomics. 2020;21:897. 10.1186/s12864-020-07215-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 185.Renna SA, McKenzie SE, Michael JV. Species differences in platelet protease-activated receptors. Int J Mol Sci. 2023. 10.3390/ijms24098298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 186.Jirouskova M, Shet AS, Johnson GJ. A guide to murine platelet structure, function, assays, and genetic alterations. J Thromb Haemost. 2007;5:661–9. 10.1111/j.1538-7836.2007.02407.x. [DOI] [PubMed] [Google Scholar]
- 187.Zhao Y, et al. Biomimetic fibrin-targeted and H2O2-responsive nanocarriers for thrombus therapy. Nano Today. 2020. 10.1016/j.nantod.2020.100986. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 188.Zhang H, et al. Shear force responsive and fixed-point separated system for targeted treatment of arterial thrombus. Nano Today. 2021;38:101186. 10.1016/j.nantod.2021.101186. [Google Scholar]
- 189.Zhang H, et al. Thrombin-responsive and sequential targeted nanoplatform for synergistic thrombolysis therapy. ACS Appl Mater Interfaces. 2025. 10.1021/acsami.5c01658. [DOI] [PubMed] [Google Scholar]
- 190.Chen J, et al. Targeted microbubbles combined with low-power focused ultrasound promote the thrombolysis of acute deep vein thrombosis. Front Bioeng Biotechnol. 2023;11:1163405. 10.3389/fbioe.2023.1163405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 191.Ma Z, et al. Ultrasound Trigger Ultrasmall Multifunction Nanobubbles for Thrombus Visual Treatment. Adv Healthc Mater. 2025;14:e2405075. 10.1002/adhm.202405075. [DOI] [PubMed] [Google Scholar]
- 192.Kwon S-P, et al. Thrombin-activatable fluorescent peptide incorporated gold nanoparticles for dual optical/computed tomography thrombus imaging. Biomaterials. 2018;150:125–36. 10.1016/j.biomaterials.2017.10.017. [DOI] [PubMed] [Google Scholar]
- 193.Xing Y, et al. Precision Photothermal Therapy for Thrombolysis: Utilizing NIR-II Gold Nanorods for Photoacoustic Imaging-Guided Thrombolysis. Mol Pharm. 2025;22:7550–60. 10.1021/acs.molpharmaceut.5c01153. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data could be obtained from responding author.
