Neutrophils are the first innate immune cells to infiltrate ischemic tissue within hours following AMI. The dual role of neutrophils, encompassing both detrimental effects (e.g., ROS release, NETosis) and beneficial contributions, is schematically illustrated in Figure 2. Shortly after AMI onset (within hours), neutrophils rapidly transmigrate across the endothelium via interactions between surface integrins and endothelial adhesion molecules. Their infiltration into the ischemic myocardium is orchestrated by DAMPs and alarmins (46–49). This recruitment is driven by a dual signaling mechanism: necrotic cardiomyocyte-derived DAMPs are sensed by PRRs on resident macrophages and endothelial cells, which in turn release chemokines that establish a gradient guiding neutrophils to the infarct zone (48, 50). Although neutrophils contribute to clearing necrotic debris, they also induce secondary myocardial injury through the release of reactive oxygen species (ROS), proteolytic enzymes, and inflammatory mediators, reflecting their dual role (51).

At the molecular level, neutrophil-derived ROS generated via NADPH oxidase pathways contribute to structural damage by oxidizing cellular components such as proteins and lipids (52, 53), and further amplify local inflammation by promoting pro-inflammatory cytokine secretion (e.g., IL-6, IL-1β) (54). Concurrently, degranulation products-including myeloperoxidase (MPO), elastase, and matrix metalloproteinases (MMPs)-aggravate cardiac remodeling by inducing cardiomyocyte apoptosis, degrading extracellular matrix (ECM), and triggering the release of additional cytokines and chemokines (e.g., TNF-α, IL-1β, CXCL-1-8) (49, 55). These factors may also impair cardiomyocyte contractility through disruption of calcium homeostasis (56), collectively sustaining acute phase injury (7).

Activated neutrophils release NETs, composed of decondensed chromatin and granular proteins, which contribute to AMI pathophysiology through multiple pathways: (1) Promoting microvascular thrombosis (57, 58); (2) Activates the Toll-like receptor 4 (TLR4)/NLRP3/IL-1β signaling axis to stimulate hematopoietic stem/progenitor cells, leading to increased neutrophil production (7, 59); and (3) mediating further neutrophil recruitment via IL-1R, forming a self-amplifying inflammatory circuit that exacerbates myocardial edema and fibrosis (60, 61).

As the pathology evolves (days 1–7 post-AMI), neutrophils undergo phenotypic switching (N1 to N2), which is implicated in repair regulation. In the acute phase (days 1–3), N1 neutrophils promote M1 macrophage polarization via IL-1β and TNF-α secretion. During the repair phase (days 4–7), N2 neutrophils support M2 macrophage polarization through upregulation of CD206, TGF-β, and IL-10 (62, 63). This transition involves Annexin A1, which, upon activation, interacts with formyl peptide receptors (FPRs) to suppress excessive inflammation and promote a pro-angiogenic macrophage phenotype, thereby facilitating tissue repair (64–67). Annexin A1 additionally modulates neutrophil activity through anti-inflammatory, pro-apoptotic, and pro-resolving mechanisms (68, 69).

Notably, S100A8/A9—primarily released from NETs—further modulates inflammatory and reparative processes post-MI. It activates the TLR4/NLRP3/IL-1β signaling axis, stimulating hematopoietic stem/progenitor cells to amplify neutrophil production (7, 59). Moreover, S100A9 exerts a unique time-dependent bidirectional regulatory effect: short-term inhibition (within 3 days) attenuates inflammation and improves outcomes (70), whereas long-term blockade (up to 21 days) impairs the generation and efferocytic function of Ly6CloMerTKhi macrophages via the Nur77 signaling pathway, ultimately worsening cardiac injury (71). This temporal specificity aligns with pathological findings from neutrophil depletion experiments (72), highlighting the need for spatiotemporally precise therapeutic strategies that target neutrophil-mediated responses—such as suppressing excessive acute-phase activation while promoting phenotypic switching during repair—to balance injury containment and tissue regeneration. Thus, targeting S100A8/A9 represents a promising therapeutic approach under active investigation.

The monocyte-macrophage system, part of the myeloid lineage, exhibits spatiotemporal dynamics in composition and function. Monocytes originate from bone marrow and extramedullary hematopoietic sites (e.g., spleen). After maturation, they enter the bloodstream and migrate to peripheral tissues, where they differentiate into macrophages or dendritic cells and participate in immune defense, inflammation regulation, and tissue repair (73). In ischemic heart disease, macrophages play a dual role: they can exacerbate injury through pro-inflammatory responses while also promoting repair via anti-inflammatory mechanisms (74–76), as illustrated in Figure 3 (74–76).

Following AMI, the population of cardiac-resident macrophages declines sharply, while circulating monocytes are extensively recruited to the infarct zone (77). These recruited monocytes originate not only from the spleen but also from extramedullary sources under the regulation of IL-1β (78). Within 24 h post-AMI, classical monocytes preferentially infiltrate the infarct area. Driven by cytokines such as TNF-α and IFN-γ released from injured cardiomyocytes and immune cells, they polarize predominantly toward the classically activated (M1) phenotype in the early phase (78–80). M1 macrophages highly express pro-inflammatory cytokines (e.g., TNF-α, IL-6), cytotoxic mediators (e.g., NO, ROS), and phagocytosis-associated proteins (81). Although they help clear necrotic debris and degrade extracellular matrix, excessive M1 activation may exacerbate the inflammatory microenvironment. In addition, pro-inflammatory exosomes (e.g., containing miR-155) released by M1 macrophages can inhibit angiogenesis, contributing to secondary myocardial injury (19, 82). Upregulation of interleukin-1 receptor-associated kinase-M (IRAK-M) in the infarct microenvironment has been shown to suppress M1 macrophage overactivation, thereby attenuating adverse cardiac remodeling (81).

As the pathology progresses (approximately days 3–7), a phenotypic shift occurs from pro-inflammatory M1 to anti-inflammatory M2 macrophages, which begin to dominate and help protect the heart from adverse outcomes (83). M2 macrophages secrete mediators such as IL-10, CCL17, VEGF, and TGF-β, which stimulate fibroblast activation, extracellular matrix synthesis, and angiogenesis, thereby promoting tissue repair (84, 85). This polarization process is influenced by several mechanisms, including efferocytosis, cell-cell contact signals, anti-inflammatory factors released by neutrophils (e.g., NGAL) (80), and extracellular vesicles carrying anti-inflammatory miRNAs. M2 macrophages specifically express the Mer tyrosine kinase (MerTK) receptor, which enables efficient clearance of necrotic cardiomyocytes through recognition of phosphatidylserine on apoptotic cells. Loss of MerTK disrupts phagocytic function and impedes repair (86). Importantly, the balance between M1 and M2 macrophages is essential for myocardial healing-persistent M1 activation can prolong inflammation, while impaired M2 function may suppress fibrosis resolution and angiogenesis, ultimately worsening cardiac function (87).

Macrophage origin, abundance, and phenotypic heterogeneity further influence the course of post-AMI injury and repair (83). Based on CCR2 expression, cardiac macrophages are classified into two main subsets: CCR2+ macrophages, which are derived from circulating monocytes and recruited to the infarct via the MyD88-dependent pathway. They exhibit M1-like characteristics, express pro-inflammatory cytokines such as CCL2, drive sustained monocyte infiltration, and promote adverse remodeling (88, 89). In contrast, CCR2−macrophages are primarily cardiac-resident cells of embryonic origin, maintained by self-proliferation. They help suppress monocyte recruitment (90), support coronary development, and facilitate cardiac regeneration (91).

Different macrophage subpopulations cooperate during cardiac injury and repair. Recent studies reveal that small extracellular vesicles derived from M2 macrophages can reduce CCR2⁺ macrophage abundance, limit monocyte recruitment to the infarct, and promote M1-to-M2 phenotypic conversion, thereby enhancing angiogenesis and improving myocardial repair (81). These findings suggest that targeting of macrophage function and phenotypic switching may hold potential for reducing adverse remodeling and supporting cardiac repair after AMI (92).

T cells play a dual role in the post-AMI inflammatory response, involving a dynamic balance between pro- and anti-inflammatory mechanisms. During ischemia-reperfusion injury, CD4⁺ T cells-particularly the Th1 subset-can influence infarct size by releasing IFN-γ and IL-17, cytokines associated with cardiomyocyte death and fibroblast proliferation (93, 94). CD8⁺ T cells show a more complex role: although impaired CD8⁺ T cell function has been linked to better initial cardiac recovery, their absence delays necrotic tissue clearance, impairing scar formation and increasing the risk of cardiac rupture (95).

Regulatory T cells (Tregs) generally support cardiac repair (45, 96). They help modulate the post-AMI immune environment by suppressing CD8⁺ T cell activity (97) and influencing monocyte/macrophage differentiation (96). Studies in animal models indicate that increasing Treg numbers through exogenous administration improves cardiomyocyte survival, cardiac function, and repair outcomes, partly by reducing pro-inflammatory monocytes/macrophages and encouraging a reparative macrophage phenotype (98).

Dendritic cells (DCs), as key antigen-presenting cells, also contribute to immune regulation and repair after AMI. Tolerogenic DCs (tDCs), a specific subset, help activate Tregs and modulate macrophage polarization. In DC-depleted mice, post-infarction ventricular remodeling is more severe, with increased inflammatory monocytes, macrophages, and cytokines, as well as higher rupture risk, supporting a protective role for DCs (40, 99, 100). By promoting Treg activation, tDCs facilitate a shift from M1 to M2 macrophages, thereby improving cardiacfunction (101). In mouse models, tDC administration after AMI reduces infarct size, improves systolic function, and enhances survival. Imaging and molecular analyses show that tDC treatment promotes Treg infiltration, elevates anti-inflammatory cytokines (e.g., IL-4, IL-10) and VEGF, and accelerates the M1-to-M2 transition, contributing to reduced inflammation and better tissue repair (40, 102, 103). Overall, tDCs help attenuate inflammation, support tissue repair, and improve cardiac outcomes after AMI through Treg activation, macrophage polarization, and stimulation of angiogenesis. These findings indicate that targeting DC subsets may offer therapeutic potential in myocardial infarction and heart failure.

B lymphocytes contribute to immune responses through antibody production, antigen presentation, and cytokine secretion. Certain B cell subsets expand in pericardial adipose tissue and accumulate in the infarcted heart after AMI, where they may exert anti-inflammatory effects via IL-10, potentially helping to resolve inflammation (104). In a clinical study of 14 MI patients, higher B cell levels after PCI correlated with improved LVEF. Mouse studies further showed that empagliflozin enhanced cardiac repair by restoring naïve B cell number and function, and infusion of B cells improved cardiac function and reduced infarct size, supporting a protective role (105). However, the exact mechanisms by which B cells regulate post-AMI immunity require further investigation.

A variety of nanoparticle (NP) classes, including organic systems such as liposomes, micelles, dendrimers, and polymeric NPs, as well as inorganic carriers like gold and silica NPs, provide various approaches for targeted drug and gene delivery to the heart. To visually organize the diverse array of NPs discussed in this section, Figure 5 provides a schematic overview of the major NP classes (organic, inorganic, biomimetic) and their immunomodulatory mechanisms targeting different immune cells in MI. Emerging biomimetic nanoplatforms are further expanding possibilities in this area.

Organic NPs, which include liposomes, micelles, and dendrimers, generally exhibit tunable biocompatibility, reduced toxicity, and improved drug delivery efficiency through adjustment of their chemical and structural properties (118). Their nanoscale size (1–100 nm) allows size-dependent effects and customizable surface chemistry, enabling advances in gene delivery, medical imaging, and targeted drug transport (106, 119).

Research on organic NP-based drug delivery systems has largely centered on liposomes, micelles, polymeric NPs, and dendrimers (113). Micelles are colloidal particles formed by the self-assembly of amphiphilic molecules, with a hydrophobic core that encapsulates poorly soluble drugs and a hydrophilic shell that improves solubility. Their small size (typically under 80 nm) promotes better penetration into ischemic myocardial tissue compared to larger NPs (120). By incorporating targeting groups such as antibodies or ROS-sensitive peptides, micelles can be engineered to recognize specific components in atherosclerotic plaques, improving targeting accuracy (121). Preclinical studies suggest that targeted micelles carrying anti-inflammatory or anti-angiogenic drugs can extend circulation half-life, lower pro-inflammatory cytokine levels, and reduce plaque area (120, 122).

Polymers are widely used as NP materials due to their low toxicity and versatility for chemical modification (123). Both natural polymers (e.g., starch, cellulose) and synthetic ones such as PLA, PGA, and PLGA have been utilized (124). PLGA, in particular, is known for its controllable degradation and is often used for sustained drug release. For example, PLGA NPs prepared by emulsion solvent diffusion have been used to deliver glutathione or heparin to the heart within 2 h in models of myocardial ischemia-reperfusion injury (MIRI) (125, 126).

Dendrimers are highly branched, well-defined macromolecules whose size can be controlled by the number of synthetic generations. Their multifunctional surfaces allow efficient loading of drugs, imaging agents, and targeting molecules such as folic acid or antibodies (127). With optimized conjugation methods, dendrimers represent potentially useful platforms for cardiovascular disease diagnosis and targeted treatment (128).

Liposomes are self-assembled phospholipid bilayer vesicles that mimic cell membranes and can carry both hydrophilic and hydrophobic drugs. They are especially useful for co-delivering genes and drugs, and they offer prospects for scalable production (129). However, issues such as drug leakage and particle aggregation may affect release kinetics, indicating a need for further formulation improvement (130).

Inorganic nanocarriers, including silica-, carbon-, and metal-based NPs (e.g., AuNPs), often show high physicochemical stability and strong drug-loading capacity, supporting precise delivery (131, 132). Gold NPs (AuNPs) are easily synthesized, exhibit low toxicity and minimal immunogenicity, and have been used to deliver cardioprotective drugs such as Simdax. In heart failure models, AuNP-conjugated Simdax showed better efficacy than the free drug, likely due to improved tissue targeting (133). Silica NPs provide high surface area and a mesoporous structure that can be functionalized for efficient drug or gene delivery, as seen in adenosine delivery to MIRI-affected heart tissue (134). However, the relatively low biodegradability of inorganic NPs compared to organic ones has raised concerns about long-term accumulation and potential toxicity. Complicated surface modification processes also present challenges, highlighting the importance of developing controlled degradation and functionalization methods to improve biosafety and efficacy (131, 135).

Although traditional nanocarriers are widely used, emerging biomimetic nano-delivery systems combine nanotechnology with biomimetic principles to provide alternative strategies for treating AMI and MIRI (136–138). These systems employ nanoscale carriers (1–1,000 nm) designed to imitate biological structures. Examples include cell membrane-coated nanoparticles, nano-sized extracellular vesicles such as exosomes, and nanozymes—nanomaterials that mimic the catalytic activity of natural enzymes.

Biomimetic systems allow efficient encapsulation and targeted delivery of therapeutics while leveraging natural biological mechanisms to avoid immune clearance, extend circulation time, and increase accumulation in ischemic heart tissue (139, 140) For instance, nanoparticles coated with autologous cell membranes or exosome-like vesicles inherit the targeting ability and biocompatibility of the source cells, functioning similarly to liposomes but with enhanced bio-specificity (141). Biomimetic nanozymes can reduce oxidative stress by scavenging excess ROS (142, 143). In comparison with conventional nanocarriers, biomimetic systems often improve the solubility and stability of poorly soluble drugs and allow stimuli-responsive release in response to microenvironmental signals such as pH changes or enzyme activity in ischemic regions. This can help minimize off-target effects and related side effects (144–146). One example is a neutrophil membrane-camouflaged delivery system loaded with siRNA. Modified with integrins to improve targeting and hemagglutinin to promote endosomal escape, this system enhanced siRNA delivery in MIRI models, reduced neutrophil infiltration and microthrombus formation, limited infarct size, and improved cardiac function (147).

Emerging evidence indicates that the initial response to myocardial tissue damage triggers intense neutrophil-dominated inflammation, exacerbating injury and potentiating ventricular remodeling. Recent advances leverage nanomedicine to precisely modulate neutrophil functions, offering cardioprotective effects through extended drug efficacy and enhanced targeting.

Tregs within ischemic myocardial tissue can exert protective effects, including anti-apoptotic, anti-inflammatory, and antioxidant actions, which may help reduce left ventricular remodeling (45). The accumulation of Tregs in the ischemic heart is considered an important factor for supporting cardiac repair. Systemic delivery of exogenous Tregs to increase their circulating levels after AMI has been associated with improved cardiomyocyte survival, cardiac function, and overall repair outcomes (98).

In addition to direct Treg infusion, some strategies aim to expand Treg populations locally. Fangyuan Li et al. developed platelet membrane-coated NPs (CsA@PPTK) that target ischemic myocardium. CsA@PPTK can scavenge ROS and increase the Treg population and the M2/M1 macrophage ratio. In high-ROS environments, the PTK component degrades to release CsA, which inhibits mPTP over-opening and may help reduce cardiomyocyte apoptosis, inflammation, and fibrosis (157).

In another approach, Kwon et al. designed liposomal NPs (L-Ag/R) co-loaded with myocardial infarction-associated antigens and rapamycin. After intravenous administration and uptake by DCs, L-Ag/R can induce tolerogenic dendritic cells and promote the generation of antigen-specific Tregs. These Tregs migrate to the infarcted myocardium, potentially enabling more targeted immune tolerance with reduced risk of non-specific systemic immunosuppression compared to polyclonal Tregs. They appear to suppress pro-inflammatory M1 macrophage activity and promote M2 macrophage polarization, which may help mitigate local inflammation, reduce cardiomyocyte apoptosis, and limit fibrosis (165).

Current research in this area often focuses on engineered cell therapies, such as CAR-Tregs, or small-molecule immunomodulators like rapamycin, which can directly expand or activate Tregs (96). In contrast, nanoparticle-mediated targeting of Tregs remains less explored. This may be due to challenges in achieving specificity, as Tregs share surface markers with effector T cells (e.g., CD25, CTLA-4), combined with the generally low infiltration of Tregs into inflammatory MI zones. These factors can limit efficient nanoparticle targeting and accumulation, indicating a need for further investigation (97) (Table 3).

CAR-Treg-NP therapy is being explored for application in cardiovascular diseases. In the context of cardiac fibrosis, lipid NPs encapsulating mRNA encoding FAP-targeting chimeric antigen receptors have been employed to generate transient CAR-T cells in vivo through a single intravenous injection. These cells selectively clear activated cardiac fibroblasts, which has been shown to reduce fibrosis and improve cardiac function in mouse models (166). This CAR-Treg-NP strategy represents a potential direction for the precise modulation of immune-mediated cardiovascular pathologies.

NP-NDDSs represent a promising approach for targeted therapy in AMI. Different classes of NPs offer distinct advantages and limitations. Organic NPs, such as liposomes, polymeric NPs (e.g., PLGA), and micelles, generally exhibit high biocompatibility, tunable drug release profiles, and ease of surface functionalization. However, they may present challenges related to stability, potential drug leakage, and local acidification caused by degradation products (121, 122, 126, 127, 129, 130). Inorganic NPs, including gold (AuNPs) and silica nanoparticles, often provide high physicochemical stability and substantial drug-loading capacity. A potential limitation is their relatively limited biodegradability, which raises considerations about long-term accumulation and possible toxicity (134, 135, 143, 144). Biomimetic NPs, such as cell membrane-coated or exosome-based systems, offer advantages in immune evasion, targeting ability, and biocompatibility. Their development, however, can face challenges related to scalable production and batch-to-batch consistency (18, 20, 142, 148). Overall, organic and biomimetic NPs may have greater clinical potential due to their biodegradable nature and bioinspired properties, whereas inorganic NPs may require further development to address safety profiles. A systematic comparison of key parameters—including drug loading capacity, biocompatibility, targeting accuracy, and translational feasibility—is provided in Table 4 to aid in the rational selection of NP-NDDSs for AMI therapy.

Most nanotherapeutics for MI remain at the preclinical stage, having been tested primarily in small animal models such as mice and rats. Among the studies reviewed here, only one included a small-scale clinical trial (151), while all others are still in the preclinical phase. This heavy reliance on animal models represents a major limitation in the development of NP-NDDS for AMI.

Although clinical research on NP-NDDS for AMI is limited, some studies in broader cardiovascular fields have begun to explore clinical applications. For example, Lu Yang et al. conducted a randomized controlled trial using a liposomal NP-NDDS to co-deliver low-dose clopidogrel and aspirin. The study involved 270 patients with coronary artery disease and reported that the liposomal formulation reduced the incidence of major adverse cardiovascular events with a favorable safety profile compared to conventional dual antiplatelet therapy (169).

However, not all clinical trials have yielded positive outcomes. In a study by Fleur M. van der Valk et al., 30 patients with atherosclerosis received liposomal prednisolone. Although the nanoparticles reached macrophages within atherosclerotic plaques, no significant anti-inflammatory effect was observed—possibly due to insufficient drug dosage or the small sample size (170). Another planned trial led by Yan Fang et al. aims to enroll 200 participants to evaluate a self-assembling nanoprobe for detecting fibroblast growth factors (FGFs) for cardiovascular disease screening and treatment assessment (ChiCTR2400089047). Despite considerable preclinical progress, the translation of NP-NDDS into clinical practice remains limited.

The challenges in clinical translation are partly attributable to anatomical, physiological, and pathological differences between species. For instance, the heart rate of mice (500–700 bpm) is much higher than that of humans (60–100 bpm), which affects hemodynamic shear stress and may alter NP circulation time, protein corona formation, and targeting efficiency. Pathologically, post-AMI inflammation in rodents is relatively short, typically resolving within 3–7 days, with fibrotic remodeling often complete within a week. In humans, however, the inflammatory phase can persist for 2–4 weeks, and fibrotic remodeling is more prolonged and complex.

Heavy reliance on small animal models means that key parameters such as targeting efficiency and safety observed in rodents may not reliably predict outcomes in humans. Additionally, clinical patients often exhibit significant individual variability—unlike standardized animal models. Factors such as immune status, age, and comorbidities can further influence NP efficacy and targeting. Therefore, the use of large animal models such as porcine MI models, which more closely mimic human physiology and pathology, is essential. These models allow for better evaluation of NP targeting efficiency, long-term safety (e.g., tracking cardiac function over ≥6 months), and immunogenicity (e.g., anti-NP antibody generation), thereby improving predictive validity.

Another challenge lies in the choice of administration route and the dynamic nature of the inflammatory response following AMI. The administration method significantly affects NP accumulation in the infarcted area, as well as the duration of therapeutic action and overall safety. Direct intramyocardial injection can achieve high local drug concentrations but may cause mechanical injury to vulnerable cardiac tissue and increase the risk of myocardial rupture. Intravenous injection is less invasive but often results in rapid systemic clearance and delayed delivery to the target site. As shown in the study by van der Valk et al., even when NPs reach the lesion, treatment effects may be limited by factors such as insufficient dosing (151).

Moreover, the post-AMI inflammatory response evolves dynamically over time, involving phenotypic shifts in immune cells such as N1/N2 neutrophils and M1/M2 macrophages. This underscores the need for NP systems capable of spatiotemporally controlled drug release. Future research could focus on developing “phase-specific” NPs that adapt to the changing inflammatory microenvironment—from acute injury to repair phases. Strategies may include dual-targeting ligands that respond to both microenvironmental cues and specific cell phenotypes, or biomimetic approaches that leverage the homing behavior of immune cells, such as neutrophil- or macrophage-mimetic NPs.

With advances in computational methods, artificial intelligence (AI)-assisted design has emerged as a potential tool for developing NPs with spatiotemporally controlled release properties. Machine learning (ML) models can analyze high-throughput experimental and computational data to predict NP behavior in complex inflammatory microenvironments, supporting more rational NP design (171). For example, ML can help predict protein corona composition based on protein sequences and NP physicochemical properties (172, 173), which may guide the design of surface ligands to enhance targeting toward specific cell phenotypes (e.g., M1/M2 macrophages) while reducing non-specific uptake. In ligand optimization, AI may aid in identifying dual- or multi-targeting ligands that respond to microenvironmental signals (e.g., pH, ROS) and cell surface receptors, enabling NPs to adapt to stage-specific inflammatory cues. ML models may also analyze relationships between immune cell homing behavior and membrane protein expression, informing the design of biomimetic nanocarriers (174). By integrating multi-omics data with computational modeling, AI-assisted approaches could help design NPs that dynamically adjust their function during disease progression, improve targeting accuracy, and reduce off-target effects.

While nanoparticles offer potential advantages in medical applications, they also present considerable immunogenic risks. Owing to their surface properties and foreign nature, nanoparticles can activate the immune system, potentially triggering hypersensitivity or inflammatory responses. For instance, gold nanoparticles (AuNPs) used in diagnostic imaging have been associated with allergic reactions such as rashes, swelling, and respiratory distress in some patients (175). Dendrimers and polymeric nanoparticles may activate immune cells while enhancing drug delivery, which could lead to inflammation or allergic reactions (176). Even typically biocompatible carriers such as liposomes have been reported to induce immune responses under high-dose or long-term administration (177, 178). These observations highlight the importance of prioritizing endogenous degradable materials (e.g., hyaluronic acid, chitosan) or biomimetic materials (e.g., exosomes, cell membrane vesicles) to help minimize long-term toxicity risks. It is also considered essential to expand long-term safety assessments to include large animal models.

The metabolic pathways, accumulation effects, and potential chronic health impacts of nanoparticles in humans remain inadequately understood, partly due to a lack of large-scale clinical studies (179). Although two clinical studies in cardiovascular disease using lipid nanoparticles reported no major safety concerns, their sample sizes were limited (169, 170). Animal studies have indicated that certain inorganic nanoparticles can accumulate in organs such as the liver, lungs, and kidneys, where they may induce oxidative stress, inflammation, or other adverse effects (180). Additionally, the environmental behavior of nanoparticles warrants attention: they may enter soil and water systems through medical waste or emissions, with the potential to bioaccumulate and move through the food chain, posing possible risks to ecosystems and public health (179, 181). Therefore, advancing the development of biodegradable nanomaterials and strengthening regulatory oversight of NPs production and use are important priorities.

These scientific uncertainties further complicate the regulatory evaluation of nanomedicine products. Regulatory agencies such as the U.S. FDA and the European EMA currently rely largely on frameworks designed for conventional drugs, which may not fully address the specific complexities of nanomaterials, including batch-to-batch variability, bioequivalence assessment, and dynamic release behaviors (182–184). The manufacturing of nanomedicines requires precise control over critical parameters such as particle size distribution, surface charge, drug encapsulation efficiency, and release kinetics, which can vary between batches and pose challenges to product consistency and quality (185). Adherence to Good Manufacturing Practice (GMP) and the establishment of a comprehensive quality control system—covering raw materials, production processes, and final products—are essential to ensure reproducible quality during scale-up from laboratory research to commercial production (185). Some nanomedicines, including Doxil and Abraxane, have encountered delays or obstacles in approval related to stability, toxicity, or manufacturing consistency issues (184). Moreover, the absence of harmonized international regulatory standards may limit the global accessibility and equitable distribution of nanomedicines (186, 187). To support the responsible development of nanomedicine, there is a need to establish transnational collaborative mechanisms and develop unified global evaluation standards specifically addressing the safety, efficacy, and environmental impact of nanomaterials.