Enhancing the Safety of Nanoparticles in Medicine: Highlights of Progress and Critical Objectives.

Introduction

Nanoparticles have become one of the most influential enabling technologies in modern medicine because they allow therapeutic and diagnostic agents to be formulated, protected, transported, and presented to biological systems in ways that are not achievable with conventional molecular dosage forms alone. 1–4 By controlling size, shape, composition, deformability, surface chemistry, and interfacial behavior, nanoparticle-based systems can modify solubility, circulation half-life, tissue distribution, intracellular trafficking, and release kinetics, thereby widening the therapeutic index of potent but otherwise limited payloads. 1 , 2 These advantages are not merely conceptual. Over the past three decades, nanomedicine has generated clinically used liposomes, lipid nanoparticles, polymer conjugates, protein nanoparticles, iron-carbohydrate nanocolloids, and nanocrystal formulations that have altered standards of care in oncology, infectious disease, metabolic disorders, vaccination, imaging, and supportive care. 2 , 5 , 6 The field therefore occupies a distinctive translational position: unlike many platform technologies that remain largely aspirational, nanoparticle systems have already demonstrated that nanoscale engineering can produce regulatory approvals, industrial adoption, and measurable patient benefit. 2 , 3 , 5 , 6

The clinical promise of nanoparticles in medicine arises from a convergence of pharmacologic and biologic principles. First, nanoparticles can rescue otherwise unmanageable active pharmaceutical ingredients by increasing apparent solubility, improving chemical stability, reducing premature metabolism, and limiting nonspecific exposure. 1 , 2 Second, they can decouple systemic exposure from target exposure by altering biodistribution and organ partitioning, which is particularly relevant for cytotoxic drugs, nucleic acids, immunomodulators, radionuclides, and imaging agents. 1–4 Third, they permit multifunctionality: the same carrier can integrate stealth properties, targeting ligands, triggered-release mechanisms, imaging labels, and immune-modulating features. Fourth, nanoparticles can be designed to engage biological transport processes at several scales, including vascular extravasation, lymphatic trafficking, endocytosis, endosomal escape, and interstitial penetration. 3 , 4 , 7 , 8 In principle, these capabilities support a transition from passive formulation science toward programmable therapeutic systems that are matched to the pathophysiology of specific diseases and even to the heterogeneity of individual patients. 3 , 4 , 7

This promise has been especially visible in cancer. Nanoparticles were embraced in oncology because they offered solutions to several chronic limitations of anticancer pharmacology: poor aqueous solubility, dose-limiting systemic toxicity, low tumor selectivity, and the need to combine multiple agents with incompatible physicochemical properties. 3 , 4 , 7 , 8 Clinically approved systems such as PEGylated liposomal doxorubicin and albumin-bound paclitaxel showed that nanoparticle reformulation could reduce selected toxicities, change infusion behavior, improve tolerability, and create viable commercial products. 3 , 4 , 7 More recently, the success of lipid nanoparticle-enabled nucleic acid delivery, including globally deployed mRNA vaccines, has expanded the perceived scope of nanomedicine beyond oncology and has demonstrated that nanoparticle systems can support rapid manufacturing scale-up, cold-chain distribution, and worldwide clinical deployment. 2 , 6 This broadening of indication space matters for safety because it shifts the field from relatively controlled specialist settings toward larger, more heterogeneous patient populations, repeated dosing paradigms, preventive applications, and combinations with biologics, devices, and immunotherapies. 2 , 6 , 9

Yet the same features that confer therapeutic opportunity also create safety complexity. Nanoparticles are not simply small versions of conventional formulations. Their clinical behavior is governed by emergent properties produced by the coupling of material composition, nanoscale architecture, colloidal stability, interfacial adsorption, dynamic restructuring in biological fluids, and interaction with immune and clearance systems. 1 , 6 , 9–14 Accordingly, nanoparticle safety cannot be reduced to the toxicology of the free drug or the bulk material alone. The relevant hazard profile is a system-level property of the entire nanoscale construct, including its excipients, degradation products, adsorbed biomolecular corona, process-related impurities, storage-dependent transformations, and dose- and route-specific biological interactions. 6 , 9–11 This principle is now well recognized in translational nanomedicine, but it remains imperfectly implemented in discovery pipelines, where efficacy-oriented optimization often still precedes robust mechanistic safety evaluation. 5 , 6 , 10

Safety has therefore become the central barrier to translation not because nanoparticle systems are uniformly unsafe, but because their risk profiles are frequently difficult to predict, compare, standardize, and regulate across platforms. 5 , 6 , 10 The problem is compounded by the historical mismatch between the sophistication of nanoparticle engineering and the maturity of the corresponding safety science. A formulation can be exquisitely optimized for encapsulation efficiency, targeting ligand density, or stimulus-responsive release, yet still fail clinically because of unanticipated complement activation, infusion reactions, altered pharmacokinetics after repeat dosing, off-target accumulation in mononuclear phagocyte system organs, unexpected degradation behavior, assay interference, manufacturing drift, or poor reproducibility across scales. 5 , 6 , 9–14 Translation is further constrained by the reality that preclinical models often overestimate delivery performance and underestimate safety liabilities, particularly when tumor biology, vascular permeability, immune competence, protein composition, and sex- or ancestry-related host variables diverge from those in the intended human population. 7–9 , 11 , 12 , 15

A major conceptual advance in the field has been the recognition that nanoparticle safety and efficacy are inseparable. The traditional translational paradigm treated safety primarily as a downstream gatekeeping exercise that follows proof-of-concept efficacy. In nanomedicine, however, the material features that determine efficacy often also determine toxicity. For example, increasing circulation time through PEGylation can improve exposure and enable tumor accumulation or extrahepatic delivery, but it can also influence anti-PEG immunogenicity, complement activation, and accelerated blood clearance on repeat administration. 13 , 14 , 16 Enhancing cellular uptake by increasing cationic character may improve transfection or endosomal escape, yet at the same time increase membrane disruption, protein adsorption, hemolytic potential, and innate immune activation. 9 , 10 Optimizing size for tissue penetration may change renal clearance thresholds, hepatosplenic sequestration, or placental transfer. Similarly, pursuing multifunctionality through ligand decoration, image-enabling components, or responsive coatings can intensify compositional complexity, increase batch-to-batch variability, and complicate analytical comparability, thereby widening both regulatory uncertainty and toxicologic ambiguity. 5 , 6 , 8 , 10 , 11

This tight coupling is especially important in cancer nanomedicine, where early enthusiasm was strongly shaped by the enhanced permeability and retention (EPR) effect. The EPR framework provided an appealing mechanistic basis for selective tumor delivery, and it helped justify extensive investment in passively targeted nanoparticles. 3 , 4 , 8 , 15 However, subsequent work has shown that EPR is highly heterogeneous across tumor types, anatomical sites, disease stages, and individual patients; moreover, the magnitude and clinical utility of passive accumulation in human solid tumors are often much less predictable than in rodent xenograft models. 8 , 15 This does not negate the value of nanoparticles in oncology, but it changes the safety discussion in a fundamental way. When tumor delivery is variable and often modest, the balance between intended and unintended exposure becomes more dependent on systemic disposition, immune recognition, stromal barriers, vascular physiology, and patient selection. In other words, safety cannot be considered independently of delivery realism. A nanoparticle that performs well in a permissive animal model may expose patients primarily through liver, spleen, blood, or infusion-related pathways rather than through therapeutically meaningful tumor deposition. 7 , 8 , 15

A second major realization is that the biological identity of a nanoparticle is not fixed at the time of manufacture. Once introduced into blood or other biological fluids, nanoparticles rapidly acquire a biomolecular corona composed of proteins, lipids, metabolites, and other macromolecules. 11 , 12 This adsorbed layer can profoundly alter colloidal stability, opsonization, immune recognition, cell-surface interactions, trafficking, and clearance. As a result, the in vivo behavior of a nominally identical formulation may vary across species, disease states, and patient subpopulations as a function of the surrounding biological milieu. 11 , 12 The translational implications are substantial. Protein corona composition is influenced by methodological factors such as sample preparation and analytical workflow, but also by patient-specific variables including inflammation, sex, ancestry, hormonal state, metabolic status, and comorbidities. 11 , 12 Safety assessment that ignores these dynamic interfaces risks both false reassurance and poor reproducibility. The growing emphasis on corona standardization and patient-dependent corona biology represents one of the clearest examples of the field moving from descriptive nanotoxicology toward mechanistically informed translational safety science. 11 , 12

Immunological interactions remain among the most consequential and incompletely resolved safety issues in nanomedicine. Nanoparticles can engage complement pathways, coagulation cascades, platelets, monocytes, macrophages, dendritic cells, mast cells, and pattern-recognition systems in ways that depend on composition, curvature, surface charge, PEG density, impurity burden, and aggregation state. 9 , 13 , 14 These responses can manifest as cytokine release, altered innate and adaptive immunity, changes in biodistribution, or acute infusion reactions. Complement activation-related pseudoallergy (CARPA) is a paradigmatic concern for liposomes and certain other lipid-based or colloidal products, because it demonstrates how a nanoparticle can trigger clinically relevant reactions that are neither classical IgE-mediated allergy nor easily predicted by conventional toxicology alone. 13 The persistence of this issue across decades of liposomal development underscores a broader lesson: in nanomedicine, safety barriers are not always exotic or late-stage; some are recurrent, mechanistically understandable, and therefore potentially preventable if incorporated early into platform design and translational testing. 9 , 13 , 14

Repeat dosing introduces an additional layer of complexity. Many therapeutic contexts require chronic or cyclical administration, yet repeated exposure can change nanoparticle disposition through adaptive immune responses, altered corona formation, saturation or priming of clearance pathways, and anti-carrier antibodies. 14 , 16 Anti-PEG immunity has become the most widely discussed example, reflecting the extensive use of PEG in liposomes, lipid nanoparticles, protein conjugates, and other long-circulating systems. 14 , 16 Although PEG remains an important technology, its benefits are now understood to coexist with liabilities related to pre-existing antibodies, de novo antibody induction, accelerated blood clearance, and potential hypersensitivity. 14 , 16 This tension is highly instructive for the broader field. It illustrates that excipients once considered largely inert may become critical determinants of clinical performance and safety when deployed repeatedly, at scale, and in immunologically diverse populations. Thus, contemporary nanoparticle safety evaluation must consider not only acute tolerability but also immunologic memory, sequence effects, and the consequences of prior exposure to structurally related materials. 9 , 14 , 16

In parallel with biologic complexity, the chemistry, manufacturing, and controls (CMC) dimension of nanoparticle safety has assumed increasing importance. For many nanomedicines, critical quality attributes extend beyond the conventional identity-purity-potency triad to include size distribution, morphology, lamellarity, drug localization, surface composition, corona susceptibility, aggregation tendency, residual solvents, endotoxin burden, free versus encapsulated payload, and time-dependent instability during storage or handling. 6 , 10 Minor shifts in these variables can alter biodistribution, immunoreactivity, or release kinetics enough to change clinical safety. As nanomedicine products become more complex, developers increasingly rely on advanced analytical technologies, orthogonal characterization, and platform-specific comparability strategies. 6 , 10 The importance of this evolution cannot be overstated. Translation failures are often attributed to biology, but a substantial fraction of safety uncertainty arises upstream from insufficiently controlled material definition or from the inability to show that the material used in pivotal studies is analytically and functionally comparable to the material manufactured at larger scale. 6 , 10 In this sense, nanoparticle safety is not only a toxicology problem; it is also a materials science, process engineering, and regulatory science problem. 5 , 6 , 10

Encouragingly, the field has made real progress. First, there is now broad consensus that nanosafety must be integrated into design rather than appended after optimization. 5 , 6 , 9 ,10 Second, clinically informative case studies, especially from liposomes, albumin nanoparticles, LNPs, and iron-based nanocolloids, have clarified which properties are most consequential for tolerability and regulatory acceptability. 2 , 5 , 6 , 13 Third, translational frameworks have shifted toward more realistic disease models, improved bioanalytical methods, and stronger emphasis on patient stratification, imaging-based pharmacology, and mechanism-aware trial design. 3 , 5 , 7 , 8 Fourth, reproducibility efforts are beginning to address long-standing heterogeneity in corona analysis, material characterization, and interlaboratory comparability. 6 , 11 Finally, the recent industrial and regulatory maturation of nucleic acid nanomedicines has strengthened the notion that scalable, clinically acceptable nanoparticle platforms can be developed when platform engineering, process control, analytical depth, and risk assessment evolve together. 2 , 6 , 9 The central challenge now is to convert these lessons from isolated successes into generalizable translational principles.

The next phase of the field therefore depends less on proving that nanoparticles can work and more on demonstrating when, why, and in whom they can work safely. This requires moving beyond generalized claims of “biocompatibility” toward mechanistic, indication-specific, and patient-aware safety definitions. It also requires distinguishing platform-intrinsic hazards from modifiable design liabilities; separating transient, manageable on-target effects from unpredictable off-target reactions; and recognizing that successful translation often rests on selecting the right clinical problem for a given nanoparticle architecture rather than forcing a versatile platform into biologically incompatible settings. 3 , 5 , 7 , 8 Particularly in oncology, future progress will likely depend on aligning carrier design with realistic delivery constraints, measurable biomarkers of accumulation or response, and combination strategies that justify the added complexity of nanoscale formulations. 3 , 4 , 7 , 8 , 15 More broadly, the field must address chronic exposure, special populations, immunogenic memory, degradation fate, and long-term organ retention with the same rigor that it has historically applied to short-term efficacy.

Against this background, the present review examines the safety of nanoparticles in medicine from a translational perspective. Its purpose is not to restate the generic advantages of nanomedicine, but to critically assess what has been learned from approved products, clinical candidates, mechanistic safety studies, and emerging analytical frameworks. The review is organized around three overarching questions. First, what genuine progress has been made in understanding and mitigating the safety liabilities of medical nanoparticles? Second, which limitations remain unresolved despite technological and regulatory advances? Third, what critical objectives should guide the next generation of safer, more clinically reliable nanomedicines? By focusing on the interface between physicochemical design, biological identity, immunological reactivity, manufacturing control, and patient heterogeneity, this review argues that nanoparticle safety should be treated not as a late-stage hurdle, but as the organizing principle of clinically meaningful nanomedicine development. 5 , 6 , 9–14

More specifically, the review highlights progress in platform characterization, standardization of nanosafety-relevant assays, improved recognition of immune-mediated adverse events, deeper understanding of protein corona biology, and growing appreciation of the mismatch between permissive preclinical models and heterogeneous human disease. 6 , 8–12 , 15 At the same time, it addresses persistent limitations: insufficient mechanistic comparability between formulations, incomplete prediction of infusion and repeat-dose toxicities, inadequate representation of susceptible populations, underdeveloped strategies for long-term fate assessment, and the continuing tendency to frame translational failure primarily as a delivery problem rather than as a coupled delivery-safety problem. 5 , 6 , 8–16 The future priorities emerging from this analysis include safe-by-design formulation paradigms, clinically relevant exposure models, patient-stratified evaluation, integrated CMC-toxicology frameworks, and prospective use of biomarkers and imaging to identify the patients most likely to achieve benefit without disproportionate risk. 3 , 5–8 , 11 , 12 , 15 In this sense, enhancing nanoparticle safety is not a defensive or restrictive agenda. On the contrary, it is the most direct route to making nanomedicine more precise, more reproducible, and more clinically transformative. To preserve a translational rather than purely descriptive focus, the literature considered in this narrative synthesis was prioritized according to clinical relevance, mechanistic contribution to nanosafety, regulatory or manufacturing significance, and the ability to clarify reproducibility, patient heterogeneity, long-term fate, or clinically important failure modes. Recent peer-reviewed studies were emphasized while landmark earlier work was retained when it remains foundational. The distinctive contribution of the present review is therefore the integration of physicochemical design, biological identity, host response, preclinical evidence, CMC, regulation, and patient context into a single safety-by-design continuum rather than the treatment of these domains as independent topics. 6 , 10 , 11 , 17–19

Nanoparticles in Medicine: Classes and Biomedical Uses

Nanoparticles used in medicine comprise a structurally diverse family of materials whose biomedical behavior is determined not only by core composition but also by size, shape, surface chemistry, colloidal stability, protein adsorption, degradation kinetics, and cargo-release behavior. From a translational perspective, it is useful to organize them into broad material classes—lipid-based, polymeric, metallic and metal oxide, silica-based, carbon-based, and hybrid nanomaterials—because each class brings a distinct combination of advantages, limitations, manufacturing constraints, and safety considerations. Although these categories often overlap in practice, especially in multifunctional theranostic platforms, they remain conceptually valuable for understanding how nanomaterial design governs biomedical use in drug delivery, imaging, vaccines, and oncology. 2 , 3 , 20 The current clinical and preclinical landscape further shows that no single nanoparticle class is universally optimal; instead, platform selection must be aligned with the physicochemical requirements of the payload, the biological barriers of the target tissue, the intended route of administration, and the level of multifunctionality required for the therapeutic objective. 2 , 21 , 22

Lipid-Based Nanoparticles

Lipid-based nanoparticles are the most clinically mature and commercially visible class of nanomedicines. Their central importance reflects a favorable convergence of biocompatibility, formulation flexibility, scalable manufacturing, and the capacity to encapsulate a wide spectrum of payloads ranging from hydrophobic small molecules to fragile nucleic acids. This category includes classical liposomes, solid lipid nanoparticles (SLNs), nanostructured lipid carriers (NLCs), nanoemulsions, niosomes in broader colloidal discussions, and, most prominently in current translational medicine, lipid nanoparticles (LNPs) designed for RNA and other nucleic acid delivery. 2 , 20 , 23 , 24 Across these subclasses, lipids serve not merely as inert carriers but as functional structural elements that determine membrane fusion behavior, endosomal escape, circulation half-life, storage stability, and cell-type selectivity. 20 , 23

The biomedical utility of lipid-based nanoparticles arises from their amphiphilic organization. Liposomes and related vesicular systems can accommodate hydrophilic cargo in aqueous cores and hydrophobic compounds within bilayers, enabling unusually broad formulation versatility. SLNs and NLCs, by contrast, exploit solid or mixed solid–liquid lipid matrices to stabilize loaded agents and modulate release kinetics. Contemporary LNPs for nucleic acid delivery rely on multicomponent formulations that typically include an ionizable lipid, a phospholipid, cholesterol, and a PEG-lipid. The ionizable lipid is especially important because it permits low-to-neutral surface charge under physiological conditions while acquiring positive charge in acidic endosomes, thereby facilitating endosomal escape and intracellular payload release. 20 , 23 , 25 This structure–function logic has become foundational in modern nanomedicine, particularly after the successful deployment of LNP-enabled mRNA vaccines and the broader expansion of RNA therapeutics. 20 , 25 , 26

A major strength of lipid-based nanoparticles is their capacity to improve the pharmacology of otherwise problematic drugs. Lipid encapsulation can increase the apparent solubility of poorly water-soluble compounds, shield labile cargos from hydrolysis or enzymatic degradation, reduce premature systemic exposure, and alter tissue distribution in ways that improve tolerability or efficacy. In oncology, these effects have been exploited to reformulate cytotoxic agents with the goal of reducing off-target toxicity and modifying biodistribution. In nucleic acid therapeutics, lipid nanocarriers protect RNA from extracellular nuclease degradation and promote intracellular delivery, thereby converting a highly unstable macromolecule into a clinically viable platform. 2 , 23 , 25 , 26 The translational significance of this class is therefore not limited to one disease area; instead, lipid-based systems have become enabling technologies across infectious disease, oncology, inflammatory disease, and regenerative medicine. 2 , 20 , 26

Liposomes remain the archetypal lipid nanocarrier and continue to be highly relevant because they illustrate the central design principles of lipid-based drug delivery. Their bilayer architecture permits passive loading, active loading, ligand functionalization, and tuning of rigidity or fusogenicity through lipid composition. The historical success of liposomal formulations demonstrated that nanoscale lipid engineering could translate into approved medicinal products, but the platform has also revealed enduring constraints such as limited loading for some cargo classes, leakage during storage, uptake by the mononuclear phagocyte system, complement activation, and modest target-site delivery in many solid tumors. 2 , 3 , 27 These issues have not displaced liposomes from the field; rather, they have driven increasingly sophisticated lipid selection, PEG-lipid optimization, surface decoration, and trigger-responsive designs.

SLNs and NLCs were developed partly to address some of the limitations associated with conventional emulsions and liposomes, especially in relation to physical stability and sustained release. Their solid or partially solid lipid matrices can enhance drug retention and improve formulation robustness, while their use of physiologically acceptable lipids supports broad interest in oral, topical, pulmonary, and parenteral applications. Recent literature continues to emphasize their suitability for antibacterial agents, chemotherapeutics, nucleic acids, and phytoconstituents, as well as their relative manufacturing simplicity and biocompatibility. 24 Nevertheless, SLNs and related systems are not free of constraints. Drug expulsion during storage, polymorphic lipid transitions, limited loading of certain molecules, and scale-sensitive process variability remain important formulation concerns. 24

The most transformative recent development in lipid nanotechnology has been the rise of LNPs for nucleic acid delivery. These systems differ from classical liposomes in both compositional philosophy and functional purpose. Rather than serving mainly as passive vesicles, LNPs are typically engineered as dynamic condensed assemblies in which the ionizable lipid plays a decisive role in nucleic acid complexation, intracellular trafficking, and endosomal disruption. Their success in clinically deployed mRNA vaccines and siRNA therapeutics has established LNPs as one of the clearest demonstrations that nanoparticle platforms can achieve industrial-scale manufacturing, regulatory acceptance, and global medical impact. 20 , 23 , 25 , 26 At the same time, this success has intensified scrutiny of their immunological and toxicological profiles, including the influence of ionizable lipid chemistry, PEG-lipid immunogenicity, reactogenicity, biodistribution to liver and spleen, and the consequences of repeated administration. 25 , 28

Beyond vaccines and RNA therapeutics, lipid-based nanoparticles are being intensively developed for cancer immunotherapy. Their ability to deliver mRNA, immunostimulatory molecules, tumor antigens, adjuvants, and combination payloads makes them attractive tools for engineering the cancer-immunity cycle. They can be designed to target antigen-presenting cells, modulate tumor-associated myeloid compartments, or support in situ vaccination strategies. In this context, lipid nanoparticles are not simply drug carriers; they are immunopharmacologic devices whose composition can shape innate sensing, antigen expression kinetics, and the spatial pattern of immune activation. 26 , 29 This is one reason why lipid-based nanocarriers occupy a central position in both prophylactic and therapeutic vaccine design.

Lipid-based nanoparticles also contribute substantially to theranostics and imaging, although they are often less intrinsically image-active than metallic or silica-based systems. Their value in this domain comes from their ability to co-load imaging probes and therapeutic agents or to encapsulate radionuclides, fluorophores, MRI-active species, or contrast enhancers together with drugs or nucleic acids. Such combinations enable image-guided drug delivery, biodistribution tracking, and pharmacodynamic readouts. In cancer, lipid-based theranostic systems are being developed to merge molecular imaging with targeted or immune-mediated intervention, though translational success still depends on solving the same core issues of circulation behavior, tumor penetration, and reproducible manufacturability that affect therapeutic-only systems. 17 , 28

Overall, lipid-based nanoparticles represent the most advanced and versatile nanomedicine class currently available. Their success derives from a balance of biological compatibility and engineering adaptability that is difficult to match. However, their maturation has also clarified an essential lesson for the broader field: clinical usability depends not only on elegant nanoparticle design but also on deep control over lipid composition, interfacial behavior, immune compatibility, storage stability, and scalable process reproducibility. 2 , 20 , 23 , 25 , 28

Polymeric Nanoparticles

Polymeric nanoparticles form one of the broadest and most adaptable classes of nanocarriers in medicine. Unlike lipid systems, whose behavior is largely dictated by supramolecular assembly of amphiphilic molecules, polymeric nanoparticles are built from covalent macromolecular frameworks that can be tuned with exceptional precision in relation to molecular weight, block architecture, degradability, hydrophobicity, charge density, glass transition behavior, and functional group distribution. As a result, polymeric platforms occupy a central place in nanomedicine for applications requiring controlled release, structural durability, ligand-mediated targeting, co-delivery of multiple payload classes, or stimulus-responsive behavior. 21 , 30

This category includes nanospheres, nanocapsules, polymeric micelles, dendritic structures, polymersomes, polyplexes, nanogels, and a wide range of biodegradable or semi-degradable constructs derived from materials such as PLGA, PLA, PCL, chitosan, alginate, PEG-containing copolymers, poly(amino acid)s, and synthetic cationic polymers. Their most important translational advantage is modularity. By altering polymer composition and architecture, investigators can engineer drug-loading mode, release kinetics, degradation rate, surface charge, hydrodynamic size, and interactions with serum proteins and cellular membranes. 21 , 30 This tunability makes polymeric nanoparticles particularly attractive for difficult biomedical tasks such as intracellular macromolecule delivery, local depot formation, BBB-oriented transport strategies, and sequential or combination therapy.

A defining biomedical strength of polymeric nanoparticles is their capacity for sustained and programmable release. Encapsulated drugs can be dispersed throughout polymer matrices, trapped in cores, conjugated by cleavable linkers, or incorporated into responsive compartments that degrade or swell in specific microenvironments. Compared with many lipid formulations, polymeric systems often provide stronger control over release over timescales ranging from hours to weeks. This is especially valuable for anticancer drugs, anti-inflammatory agents, antimicrobials, and growth factors whose therapeutic activity depends on maintaining a defined exposure profile rather than achieving a sharp transient peak. 21 , 30 Polymeric nanocarriers have therefore become major candidates for improving therapeutic index through spatiotemporal control of drug exposure.

The diversity of polymeric nanoparticle behavior also supports extensive surface bioengineering. Ligands, peptides, antibodies, cell-membrane mimics, stealth coatings, and environmentally responsive motifs can be introduced either during polymer synthesis or after nanoparticle formation. Such modification strategies are widely used to promote receptor-mediated uptake, evade rapid clearance, enhance mucosal penetration, cross biological barriers, or create disease-microenvironment sensitivity. 30 This is particularly relevant in oncology, where polymeric nanoparticles are often designed not only to carry cytotoxic agents but also to navigate a series of biological barriers including blood circulation, stromal transport, tumor cell binding, endosomal escape, and intracellular organelle access.

Polymeric nanoparticles are also highly relevant for nucleic acid delivery. Although LNPs currently dominate the clinical landscape of RNA therapeutics, polymer-based platforms remain scientifically important because they offer alternative mechanisms of cargo condensation, protection, endosomal escape, and sustained intracellular release. Cationic polymers and block copolymers can complex DNA, mRNA, siRNA, miRNA, or gene-editing cargos into polyplexes or hybrid assemblies. In principle, these systems can reduce some limitations of lipid carriers and provide prolonged expression or tissue-selective targeting. In practice, however, polymeric nucleic acid delivery faces substantial translational hurdles, especially regarding toxicity associated with cationic charge, variable unpackaging, and manufacturing reproducibility. 21 , 30 Even so, the field continues to develop lower-toxicity biodegradable cationic polymers and highly engineered poly(amino acid) systems that may become important in future generations of precision nanomedicine.

In cancer therapy, polymeric nanoparticles are especially attractive for combination regimens. Their matrix-based or micellar organization can support co-encapsulation of hydrophobic drugs, hydrophilic adjuvants, immunomodulators, photosensitizers, and imaging probes, allowing rational synchronization of release or hierarchical delivery. Such designs are conceptually appealing because they align with the increasingly multidimensional character of oncology, where effective therapy often requires simultaneous modulation of tumor cells, vasculature, stroma, and immune compartments. 21 , 29 , 30 Polymeric nanocarriers can therefore function not only as passive drug depots but as integrated treatment systems capable of coordinating several therapeutic mechanisms.

Another important area of application is vaccination and immunoengineering. Polymeric nanoparticles can protect antigens and nucleic acids, enhance uptake by antigen-presenting cells, and permit controlled release of adjuvants or immune modulators. Relative to lipids, some polymeric platforms offer stronger structural stability and longer retention in lymphoid tissues or mucosal environments. They also enable elaborate design of degradation-triggered antigen presentation and cytokine programming. These features have supported growing interest in polymer-based nanovaccines for infectious disease and cancer, as well as hybrid lipid–polymer strategies intended to combine the immunological and manufacturing strengths of both classes. 27

Despite their versatility, polymeric nanoparticles face persistent translational limitations. Their synthesis is often more chemically complex than that of lipid-based systems, and this complexity can create burdens in characterization, batch reproducibility, impurity control, and regulatory comparability. Residual solvents, monomers, catalysts, and broad molecular-weight distributions may influence safety or performance. Furthermore, while cationic and highly functionalized polymers may excel in vitro, they frequently encounter in vivo obstacles related to protein adsorption, hepatic sequestration, complement activation, limited penetration, or unexpected toxicity. 17 , 21 , 30 These issues have not diminished the scientific value of polymeric nanomedicine, but they have highlighted the need for platform choices that are realistic from both a biological and CMC standpoint.

From a biomedical perspective, polymeric nanoparticles are most compelling when a therapeutic problem truly requires what polymers uniquely offer: sustained release, structural persistence, advanced responsiveness, or multistage functionality. Their future importance will likely remain high in local delivery, implant-associated systems, immunomodulation, combination therapy, and disease settings where kinetic control is more important than simple encapsulation. In this sense, polymeric nanoparticles complement rather than compete directly with lipid systems, broadening the design space of nanomedicine by offering a macromolecular toolkit for precision control over nanoparticle fate and payload behavior. 21 , 27 , 30 The major nanoparticle classes and their medical applications are summarized in Figure 1 .

Schematic representation of the principal nanoparticle classes currently used or investigated in medicine, including lipid-based, polymeric, metallic and metal oxide, silica-based, carbon-based, and hybrid or biomimetic nanoparticles. These platforms support a broad range of biomedical applications, including drug delivery, molecular imaging, theranostics, vaccines, nucleic acid delivery, oncology, inflammatory disease treatment, and anti-infective strategies. The figure highlights the structural diversity of nanomedicine platforms and the overlap between material class and therapeutic or diagnostic use. The illustration is conceptual and not drawn to scale. This is an original schematic synthesis prepared specifically for the present review and is not reproduced from a previously published figure.

Metallic and Metal Oxide Nanoparticles

Metallic and metal oxide nanoparticles are distinguished from organic nanocarriers by their intrinsic physicochemical functionalities. Their appeal in medicine is not primarily based on cargo encapsulation alone, but on properties such as plasmonic behavior, magnetism, catalytic activity, X-ray attenuation, redox reactivity, and high electron density. These features make them uniquely attractive for imaging, externally triggered therapy, photothermal therapy, magnetic targeting, radiosensitization, antimicrobial applications, and multimodal theranostics. 22 , 31 , 32 Gold, silver, iron oxide, titanium dioxide, zinc oxide, copper oxide, cerium oxide, and other related materials represent the best-studied members of this class.

Gold nanoparticles are among the most extensively investigated metallic systems because of their relative chemical inertness, facile surface functionalization, tunable shape-dependent optical properties, and compatibility with photothermal and imaging applications. Nanorods, nanospheres, nanoshells, and other anisotropic gold structures can absorb or scatter light in ways that support optical imaging, photothermal ablation, and image-guided therapy. In oncology, this has motivated development of gold-based systems for tumor visualization, thermal destruction, radiosensitization, and combination treatment strategies. 33 Their surface can also be readily conjugated with drugs, peptides, antibodies, nucleic acids, and diagnostic labels, allowing flexible construction of multifunctional theranostic systems.

Silver nanoparticles, although historically prominent in antimicrobial applications, have also attracted biomedical attention in wound care, anti-infective coatings, and selected anticancer strategies. However, their strong bioactivity is coupled to concerns about oxidative stress, ion release, and off-target toxicity, which have limited their direct translational momentum relative to gold and iron oxide systems. 31 More broadly, the biomedical role of metallic nanoparticles cannot be separated from metal-specific toxicology. Surface oxidation, ion dissolution, catalytic ROS production, and interactions with proteins and membranes may all drive adverse effects. As a result, the same properties that make metallic nanoparticles attractive for therapy often complicate their biocompatibility profile. 31 , 34

Metal oxide nanoparticles broaden the field further by introducing functionalities associated with magnetism, semiconductor behavior, catalytic activity, or redox buffering. Among them, iron oxide nanoparticles are the most clinically and translationally important. Superparamagnetic iron oxide nanoparticles (SPIONs) and related systems have been intensively studied for MRI contrast, magnetic hyperthermia, magnetic guidance, drug delivery, cell tracking, and multimodal imaging. Their external-field responsiveness offers a distinctive route to spatial control: they can be visualized by MRI, heated under alternating magnetic fields, or directed by magnetic gradients under selected conditions. 32 , 35 These features position iron oxide systems at the intersection of imaging and therapy more clearly than most organic nanoparticles.

The promise of iron oxide nanoparticles in medicine lies in their multifunctionality. They can act as imaging agents by altering local magnetic relaxation, as carriers by adsorbing or conjugating drugs, and as therapeutic effectors through hyperthermia or magnetically assisted localization. In cancer and cardiovascular applications, they have been explored as platforms for lesion imaging, local heating, plaque targeting, and cell trafficking studies. 32 Their biomedical translation, however, depends heavily on core size, crystallinity, aggregation state, surface coating, and degradation behavior. Surface functionalization with polymers, silica, lipids, or biomolecules is often necessary to improve colloidal stability, reduce aggregation, and modulate protein corona formation. 32 , 35

At the same time, iron oxide nanomedicine provides a strong example of the inseparability of utility and safety. Although iron is a biologically familiar element, iron oxide nanoparticles are not inherently benign in every formulation context. Their biodistribution, lysosomal processing, oxidative stress potential, macrophage uptake, and possible long-term tissue persistence must be evaluated carefully. Recent analyses emphasize that cytotoxicity and systemic effects are highly dependent on dose, coating chemistry, particle size, aggregation, and degradation kinetics rather than on composition alone. 35 This principle also applies across the broader family of metallic and metal oxide systems.

A major biomedical use of metallic and metal oxide nanoparticles is theranostics. Because many of these materials are intrinsically image-active or energy-responsive, they are unusually well suited to platforms that combine diagnosis and therapy in one construct. Gold nanoparticles can support optical or CT imaging while enabling photothermal treatment; iron oxide nanoparticles can function in MRI and hyperthermia; titanium dioxide and related systems may be used in photodynamic or catalytic strategies; and radiolabeled metallic cores can support SPECT or PET imaging together with drug or radionuclide therapy. 22 , 28 , 31–33 In this class, theranostics is not an added layer but often a native consequence of material choice.

Another area of strong interest is radiosensitization and image-guided oncology. High-atomic-number metallic nanoparticles can amplify radiation effects or enhance imaging contrast, thereby supporting strategies that aim to improve locoregional tumor control while reducing exposure to surrounding tissue. Such concepts remain attractive, but they also illustrate the gap between mechanistic promise and clinical translation. Achieving sufficient, reproducible intratumoral accumulation in patients remains difficult, and the systemic safety profile of repeated or high-dose administration can be challenging to define. 17 , 22 , 31

In summary, metallic and metal oxide nanoparticles occupy a critical but more safety-sensitive branch of nanomedicine. Their greatest value lies where intrinsic material functionality is essential—magnetic imaging, hyperthermia, plasmonic treatment, catalytic therapy, high-contrast imaging, or multimodal theranostics. Where simpler carrier functions suffice, they may offer less favorable risk–benefit balance than lipid or polymeric systems. Their future clinical role will therefore depend on disciplined indication matching, robust surface engineering, and precise control of degradation, retention, and host response. 22 , 31 , 34 , 35

Silica-Based, Carbon-Based, and Hybrid Nanomaterials

Silica-based nanomaterials, especially mesoporous silica nanoparticles (MSNs), occupy a distinctive niche in nanomedicine because they combine inorganic structural stability with highly tunable porosity and surface chemistry. Their large surface area, adjustable pore size, high loading capacity, and amenability to functionalization make them attractive for drug delivery, gene delivery, imaging, and theranostics. Unlike many dense inorganic nanoparticles, MSNs are especially valued as cargo-hosting frameworks, since their pores can be engineered to load therapeutic agents while external surfaces are modified for targeting, stealth, or stimulus-responsive gating. 36 , 37 This combination of structural regularity and chemical adaptability has made MSNs one of the most sophisticated scaffold platforms in nanomedicine research.

The biomedical logic of silica nanoparticles is closely linked to controlled release. Drugs can be physically entrapped in mesopores, linked through cleavable chemistries, or retained by molecular gates that open in response to pH, redox state, enzymes, heat, or light. As a result, silica systems are often discussed as model platforms for precision release engineering. Their utility extends to chemotherapy, antimicrobial delivery, gene transfer, tissue engineering, and regenerative medicine. 36 In oncology, MSNs have been especially attractive because they allow co-delivery of multiple agents, high drug loading, and the possibility of integrating imaging elements into the silica matrix or surface.

Surface and matrix modification are central to the medical use of silica nanoparticles. Functional groups, polymers, biomolecules, dyes, radionuclides, and metal species can be introduced either onto the surface or directly into the silica framework, producing particles with enhanced stability, bioactivity, degradability, or imaging capability. Recent work on in situ modified MSNs highlights how such hybridization can generate theranostic systems that preserve drug-loading capacity while adding diagnostic or responsive functions. 37 These strategies also reveal one of the broader strengths of silica: it is an unusually versatile “bridge material” that can combine the precision of inorganic synthesis with the functionality of organic or metallic add-ons.

Nevertheless, silica-based nanomedicine also faces well-recognized translational obstacles. Biocompatibility and biodegradation are not uniform across all silica formulations. Toxicity depends on particle size, porosity, shape, silanol density, dose, route, and surface modification. Colloidal stability and aggregation behavior may change dramatically in biological fluids, and organ retention remains a relevant concern for some formulations. 36 Thus, although silica nanoparticles are often presented as highly adaptable, their translational success depends on careful optimization of degradability and surface reactivity alongside therapeutic performance.

Carbon-based nanomaterials represent another broad and functionally rich class that includes carbon nanotubes, graphene and graphene oxide, fullerenes, nanodiamonds, carbon dots, graphene quantum dots, and carbon nano-onions. Their biomedical appeal arises from high surface area, unique mechanical and electrical properties, strong π–π interactions, tunable optical behavior, and, for some subclasses, photothermal or fluorescent characteristics. 38–40 Carbon nanomaterials have been investigated for drug delivery, biosensing, tissue engineering, imaging, gene delivery, photothermal therapy, and cancer theranostics.

One of the most important biomedical advantages of carbon nanomaterials is their exceptional surface engineering potential. Aromatic drugs can adsorb through π–π stacking, polymers can improve dispersibility, and ligands can be attached for targeting or immune modulation. Nanodiamonds and carbon dots, in particular, have drawn attention for relatively favorable biocompatibility profiles compared with some earlier carbon nanostructures, while graphene derivatives and nanotubes continue to offer powerful optical and thermal functionalities. 38–40 In cancer, carbon-based systems are especially attractive for photothermal therapy, fluorescence-enabled imaging, and multifunctional delivery constructs.

At the same time, carbon nanomedicine has historically been restrained by concerns about biodistribution, persistence, inflammation, oxidative stress, and structure-dependent toxicity. These concerns are particularly pronounced for fibrous or poorly degradable forms such as certain carbon nanotubes, where aspect ratio, residual catalysts, aggregation, and surface defects can profoundly influence biological response. 38 , 39 Surface functionalization can reduce some of these liabilities, improve aqueous dispersibility, and enhance biocompatibility, but it does not eliminate the need for rigorous, structure-specific safety evaluation. As a result, carbon-based nanomaterials remain scientifically powerful but less clinically mature than lipid systems.

Hybrid nanomaterials are increasingly important because they reflect the recognition that no single material class perfectly satisfies all biomedical requirements. Hybridization can combine the encapsulation efficiency and biocompatibility of lipids, the kinetic control of polymers, and the imaging or magnetic functionality of inorganic cores. Examples include lipid–polymer hybrid nanoparticles, silica–metal composites, polymer-coated iron oxide systems, metal nanoparticle–hydrogel constructs, and other composite architectures. 27 , 37 , 41 The rationale is straightforward: by integrating complementary materials, one can seek simultaneous control over loading, stability, targeting, imaging, and triggered release.

In cancer theranostics, hybrid nanoparticles are especially attractive because they enable concurrent treatment and visualization. A lipid or polymer shell may improve circulation and payload capacity, while an inorganic core provides imaging contrast or energy conversion. Similarly, mesoporous silica frameworks may be combined with magnetic or fluorescent elements to create image-guided delivery systems. 28 , 37 , 41 Hybrid systems can therefore be viewed as convergence platforms, bringing together otherwise separate functionalities into single nanoscale constructs.

However, hybridization also increases complexity. Each added component can alter manufacturing reproducibility, interfacial behavior, degradation pathways, regulatory classification, and safety uncertainty. For this reason, hybrid nanomaterials exemplify both the promise and the translational burden of advanced nanomedicine. Their future clinical relevance will depend on showing that the added complexity yields a meaningful biomedical advantage rather than merely an attractive materials-science concept. 17 , 27 , 41 Biomimetic and cell-membrane-coated nanoparticles deserve separate translational consideration because their biological interface is intentionally derived from cells or extracellular structures rather than from a wholly synthetic coating. Membranes from erythrocytes, platelets, leukocytes, tumor cells, and other sources can confer immune camouflage or disease-relevant adhesion properties, but these advantages introduce distinct safety and CMC liabilities, including source variability, membrane-protein heterogeneity, residual biological material, pathogen-control requirements, difficult purification, and scale-dependent reproducibility. Consequently, biomimicry should not be equated automatically with improved safety; clinically meaningful benefit depends on demonstrating that the biological coating remains compositionally controlled and functionally reproducible across batches. 42–44

Current Applications in Drug Delivery, Imaging, Theranostics, Vaccines, Oncology, and Inflammation

Across material classes, drug delivery remains the dominant biomedical application of nanoparticles. The most consistent and clinically relevant contribution of nanocarriers is their ability to modify the disposition of active agents in ways that conventional formulations often cannot achieve. By increasing apparent solubility, protecting labile cargos from chemical or enzymatic degradation, reducing premature clearance, altering tissue exposure, and enabling controlled or triggered release, nanoparticles can substantially reshape pharmacokinetics and pharmacodynamics. 2 , 20–22 , 36 Lipid-based systems are currently the strongest clinically in this area, particularly for nucleic acids and reformulated small molecules, whereas polymeric carriers remain especially valuable when sustained release, matrix-controlled delivery, or environmentally responsive behavior is required. 20 , 21 , 27 Inorganic and silica-based systems contribute where drug carriage can be integrated with imaging, catalytic activity, or externally triggered functions. 22 , 28 , 33 , 36 , 37 The practical implication is that “drug delivery” should not be treated as a single therapeutic category, but rather as a set of distinct functions requiring different nanoparticle solutions depending on payload chemistry, route of administration, target tissue architecture, and the biological barriers that must be overcome.

Imaging is another major use domain, particularly for inorganic, metal oxide, and hybrid platforms. Iron oxide nanoparticles have been widely studied and used in MRI-related applications; gold and other high-atomic-number materials support CT, optical, and photoacoustic strategies; radiolabeled nanoparticles enable nuclear imaging; and silica-, lipid-, or polymer-based carriers can be engineered to transport fluorophores, contrast agents, or radionuclides. 28 , 32 , 33 , 35 , 37 The attraction of nanoparticle-based imaging lies not only in signal enhancement, but also in the possibility of modulating circulation time, improving lesion-to-background ratios, and enabling multimodal readouts within a single platform. 17 , 28 , 37 However, successful translation in imaging depends on more than the generation of contrast. It requires reproducible biodistribution, predictable clearance, low nonspecific retention, and a clear diagnostic advantage over established imaging agents and workflows. 17 , 28 In this sense, imageability alone is insufficient; the platform must produce clinically useful information without creating disproportionate complexity or retention-related safety concerns.

Theranostics has emerged as one of the defining conceptual themes of nanomedicine because it seeks to unite therapeutic and diagnostic functions within the same nanoscale construct. Metallic, metal oxide, silica-based, and hybrid nanoparticles are especially attractive for theranostic applications because they can intrinsically support imaging, local heating, catalytic action, magnetic responsiveness, or radionuclide incorporation while simultaneously carrying drugs, nucleic acids, or immune-active agents. 17 , 22 , 28 , 37 , 41 In oncology, theranostic nanoparticles are often framed as tools for precision medicine because they offer the possibility of noninvasively identifying patients most likely to benefit, guiding intervention, and monitoring treatment response in real time. 3 , 17 , 28 , 37 Similar logic is increasingly being extended to inflammatory disease, where theranostic platforms may help visualize inflammatory burden, identify active lesions, and coordinate local treatment delivery. 22 , 37 Nevertheless, the translational bar for theranostics remains high. The diagnostic component must provide information that is genuinely actionable, and the therapeutic component must justify the platform’s added material, regulatory, and manufacturing complexity. 17 , 28 , 37 , 41

Vaccines constitute one of the clearest modern success stories of nanomedicine. Nanoparticles can protect antigens or nucleic acids, improve uptake by antigen-presenting cells, shape innate immune activation, and facilitate trafficking to lymphoid tissues. 26 , 27 , 29 Lipid nanoparticles have become the leading example because they transformed mRNA from a fragile biological intermediate into a clinically deployable vaccine modality with global public-health impact. 26 , 27 At the same time, polymeric nanoparticles and lipid–polymer hybrid systems are receiving growing attention in cancer vaccine design because they can co-deliver antigens, adjuvants, cytokine regulators, and immune-modulating payloads in a synchronized manner. 26–28 In these settings, nanoparticle composition does not merely determine pharmacokinetics; it becomes part of the immunological architecture of the vaccine itself, influencing antigen presentation, innate sensing, and the quality and magnitude of adaptive immune responses. 26–28 This dual role—as both carrier and immunological modulator—helps explain why nanoparticle vaccines are among the most translationally successful and biologically sophisticated applications of nanomedicine.

Oncology remains the largest and most conceptually fertile field for nanoparticle application. Nearly every major nanoparticle class has been investigated in cancer drug delivery, molecular imaging, immunotherapy, photothermal therapy, hyperthermia, radiotherapy enhancement, gene delivery, or multimodal theranostics. 3 , 17 , 22 , 29 , 33 This dominance reflects both need and opportunity. Cancer therapy frequently relies on highly potent agents with narrow therapeutic windows, while tumors themselves are biologically heterogeneous, anatomically diverse, and difficult to access selectively. 3 , 17 Nanoparticles are therefore attractive because they promise to improve tolerability, concentrate treatment at disease sites, and enable rational combinations of imaging, cytotoxicity, and immune modulation. 3 , 17 , 22 , 29 However, oncology has also exposed the limits of nanomedicine more clearly than any other field. The heterogeneity of nanoparticle delivery in human tumors, the variability of the EPR effect, and the persistent gap between impressive preclinical accumulation and modest clinical benefit have all demonstrated that tumor targeting is often less predictable than originally anticipated. 3 , 17 Accordingly, the value of cancer nanomedicine increasingly rests not on idealized assumptions of selective tumor homing, but on whether the platform can meaningfully improve the therapeutic index in real clinical contexts.

Inflammation has emerged as another major and increasingly important application area for nanomedicine. Inflammatory diseases, including inflammatory bowel disease, rheumatoid arthritis, atherosclerosis, psoriasis, neuroinflammatory disorders, and inflammation-associated tissue injury, are attractive targets for nanoparticles because they often involve dysregulated vascular permeability, activated immune-cell populations, oxidative stress, altered pH, enzyme-rich microenvironments, and tissue-specific inflammatory mediators. 22 , 23 , 25 , 36 , 37 These disease-associated features create opportunities for selective accumulation, triggered release, and cell-specific immune modulation that are difficult to achieve with conventional systemic therapies. 22 , 23 , 25 Nanoparticles are especially valuable in this context because many anti-inflammatory drugs are limited by poor tissue selectivity, off-target immunosuppression, metabolic toxicity, or the need for frequent administration. By improving local retention, protecting sensitive biologics or nucleic acids, and enabling delivery to macrophages, dendritic cells, endothelial cells, or inflamed epithelium, nanoparticle systems may reduce systemic exposure while enhancing therapeutic activity at the site of disease. 22 , 23 , 25 , 36 , 37

Within inflammatory medicine, nanoparticle applications are highly diverse. In inflammatory bowel disease, nanoparticles have been developed to cross mucus barriers, accumulate in inflamed intestinal tissue, and release corticosteroids, biologics, antioxidants, or RNA-based therapeutics in response to pH, ROS, or enzymatic triggers. 25 In atherosclerosis, nanocarriers are being investigated for targeting vascular inflammation, macrophage-rich plaques, and endothelial dysfunction, with the goal of reducing systemic drug exposure while stabilizing high-risk lesions. 22 , 23 In autoimmune and chronic inflammatory disorders, polymeric, lipidic, and hybrid nanoparticles are being explored as vehicles for tolerogenic therapies, cytokine modulation, antigen-specific immune reprogramming, and selective suppression of pathogenic immune-cell populations. 22 , 23 , 36 These applications are particularly important because they broaden the role of nanomedicine beyond cancer and infectious disease into chronic immunological conditions, where long-term safety, repeat-dose compatibility, and immune-specific precision are especially critical.

At the same time, inflammatory applications illustrate a fundamental challenge in nanomedicine: the same inflammatory microenvironments that can be exploited for targeting may also alter nanoparticle behavior in unpredictable ways. Changes in plasma proteins, cytokine profiles, endothelial permeability, phagocyte activation, and tissue architecture can modify protein corona formation, organ distribution, immune recognition, and clearance. 22 , 23 Thus, inflammation is not simply a therapeutic target; it is also a biological state that can reshape the safety and efficacy of the nanoparticle itself. This makes inflammatory nanomedicine particularly important from a translational and safety perspective, because it requires the field to account for disease-conditioned biology rather than assuming that nanoparticle behavior in healthy or simplified preclinical models will translate directly to inflamed human tissues. 22 , 23 , 25

The present state of nanoparticle medicine is therefore characterized by marked asymmetry. Some applications—notably lipid nanoparticle delivery of nucleic acids, vaccine technologies, and selected reformulations of established drugs—have demonstrated clear clinical impact. 2 , 20 , 26 , 27 Others, including many tumor-targeting, theranostic, and multifunctional hybrid constructs, remain scientifically compelling but clinically less established. 3 , 17 , 22 , 28 , 37 , 41 This unevenness should not be interpreted as failure. Rather, it indicates that different biomedical applications impose very different translational demands. Platforms that succeed in vaccines may not be optimal for solid tumor penetration; systems that excel in imaging may not provide the safest chronic exposure profile; and hybrid constructs that appear ideal in mechanistic studies may prove unnecessarily complex for routine manufacturing, regulatory evaluation, or long-term clinical use. 2 , 3 , 17 , 22 , 28 The same principle applies to inflammatory disease, where a platform that successfully exploits local inflammatory cues must still demonstrate acceptable repeat-dose tolerability and minimal unintended immune perturbation. 22 , 23 , 25

The major nanoparticle classes, their principal translational advantages, and their dominant safety liabilities are summarized in Table 1 .

What “Safety” Means in Nanomedicine

In nanomedicine, the term safety cannot be reduced to the simple absence of overt toxicity. For conventional small-molecule drugs, safety assessment is already multidimensional, encompassing dose-limiting toxicity, organ-specific adverse effects, tolerability, exposure margins, patient susceptibility, and manufacturing control. In nanomedicine, these considerations become more complex because the product is not only a pharmacologically active ingredient but also a structured material system whose clinical behavior depends on emergent properties such as size distribution, morphology, surface chemistry, corona formation, colloidal stability, degradability, impurity profile, and process-dependent heterogeneity. 6 , 10 , 46 , 47 As a consequence, the safety of a nanomedicine is not fully captured by the toxicology of the encapsulated drug, the bulk composition of the carrier, or even the nominal formulation recipe. Rather, safety is a composite property that arises from the interaction between material design, biological context, patient susceptibility, and manufacturing consistency. 6 , 10 , 45–47

This broader framing has become increasingly important as nanomedicine has matured from a field centered on proof-of-concept delivery systems to one that includes marketed products, complex generic formulations, nucleic acid carriers, combination theranostics, and increasingly patient-specific platforms. Regulatory and translational literature now recognizes that nanoscale products require integrated evaluation of physicochemical characterization, biological interactions, immune effects, biodistribution, degradation fate, and critical quality attributes because relatively modest shifts in these parameters can alter both efficacy and toxicity. 6 , 46–48 The meaning of safety in nanomedicine is therefore necessarily layered. It includes whether a product is biocompatible enough to perform its intended function without unacceptable host disturbance, whether it is biosafe across relevant exposure scenarios and populations, whether it is tolerable at clinically useful dose schedules, whether its acute and chronic risks are understood, whether local and systemic toxicities are acceptably controlled, and whether product quality can be manufactured reproducibly enough that the clinical safety profile remains stable across batches and over time. 6 , 10 , 45–48

Distinction Between Biocompatibility, Biosafety, and Tolerability

A persistent problem in the nanomedicine literature is the interchangeable use of the terms biocompatibility, biosafety, and tolerability, even though they are not synonymous. The distinction matters because conceptual imprecision often leads to weak claims about clinical readiness. In many preclinical studies, nanoparticles are described as “biocompatible” merely because they preserve short-term cell viability in one or two in vitro assays. However, such evidence is insufficient to support broader claims of safety, much less clinical suitability. 9 , 10 , 45

Biocompatibility is best understood as a context-dependent property describing the ability of a material system to perform its intended medical function with an acceptable host response in a specific application. 9 , 45 This definition is inherently relational. A nanomaterial is not universally biocompatible in the abstract; it may be acceptably biocompatible for one route, tissue, exposure duration, or indication and not for another. A formulation that is acceptable as a locally administered imaging probe may not be acceptable as a chronically dosed intravenous therapeutic. Likewise, a degree of transient immune stimulation that is useful in a vaccine may be undesirable in gene replacement therapy. In nanomedicine, biocompatibility therefore includes not only the absence of gross cytotoxicity, but also the compatibility of the nanoparticle with blood components, immune cells, complement pathways, endothelial interfaces, reticuloendothelial clearance systems, and relevant target tissues. 9 , 13 , 45

Biosafety is broader. It refers to the overall safety profile of a nanomaterial or nanomedicine across relevant biological systems, exposure conditions, and time scales, including hazard identification, dose-response behavior, organ disposition, degradation fate, immunotoxicity, genotoxicity where relevant, reproductive considerations where relevant, and environmental or occupational implications when appropriate. 10 , 45 , 49 In the biomedical context, biosafety extends beyond whether the product functions without immediate damage; it asks whether the full range of foreseeable biological risks has been adequately characterized and controlled. For nanomedicines, this broader view is particularly necessary because nanoscale materials may undergo dynamic transformations in vivo, including aggregation, protein corona remodeling, dissolution, biodegradation, or redistribution after repeated administration. 10 , 45 , 47 A product may appear biocompatible in a narrow assay but still raise biosafety concerns if it accumulates chronically in the liver or spleen, alters innate immunity, or produces toxic degradation products after prolonged exposure. 10 , 35 , 45 , 49

Tolerability, by contrast, is a clinical term centered on the degree to which adverse effects can be endured by patients at therapeutically relevant exposures. 50 Tolerability does not imply absence of biologic perturbation. A nanomedicine can be nontrivially reactogenic, immunostimulatory, or organ-active yet still be considered clinically tolerable if adverse effects are transient, manageable, reversible, dose-dependent, and justified by benefit. This distinction is especially important in oncology and vaccinology. Liposomal doxorubicin, for example, may improve tolerability relative to conventional doxorubicin for selected toxicities without eliminating toxicity altogether; similarly, lipid nanoparticle-enabled mRNA vaccines may be clinically acceptable despite transient local and systemic reactogenicity. 50 Tolerability therefore operates at the interface of safety and therapeutic utility. It is inseparable from dose, schedule, route, indication, alternatives, and patient preference.

These three concepts overlap but should be kept distinct in a rigorous review. Biocompatibility concerns the appropriateness of the host response in a given use context. Biosafety concerns the broader hazard and risk profile across biologic systems and exposure scenarios. Tolerability concerns the practical clinical acceptability of adverse effects at useful doses. The distinction is not semantic. It has methodological consequences. Biocompatibility is often probed through hemocompatibility, cytocompatibility, complement activation, protein corona behavior, cell uptake, and tissue-response studies. 9 , 13 , 45 Biosafety requires a larger framework that may include biodistribution, repeated-dose toxicology, degradation studies, immunotoxicity, chronic retention analysis, and quality-related risk assessment. 10 , 35 , 47 , 49 Tolerability ultimately depends on clinical data or clinically meaningful translational proxies, because it asks whether patients can receive the product under realistic treatment conditions. 50

In nanomedicine, failures often arise precisely because these domains are conflated. A nanoparticle platform may look favorable in vitro and be labeled “biocompatible,” yet later prove poorly biosafe because of complement activation-related pseudoallergy, anti-PEG immune responses, or accumulation-driven chronic toxicity. 13 , 51 , 52 Alternatively, a platform may have acceptable biosafety margins but poor tolerability if infusion reactions, fever, fatigue, or local inflammatory effects prevent repeated administration at therapeutic doses. A careful review must therefore treat safety terminology as mechanistic and translationally consequential, not merely descriptive.

Acute versus Chronic Toxicity

Another essential dimension of nanomedicine safety is the distinction between acute and chronic toxicity. This distinction is familiar in toxicology, but it is especially important for nanoparticles because many nanomaterials produce early and late adverse effects through different mechanisms. Acute toxicity typically refers to harmful effects occurring shortly after administration or exposure, often ranging from minutes to days. In nanomedicine, acute toxicity may include infusion reactions, complement activation, cytokine release, vascular effects, hemolysis, thrombogenicity, acute hepatic or renal injury, neurobehavioral effects in selected systems, or local tissue irritation at the administration site. 11 , 13 , 51 , 52 Such effects may be strongly influenced by the particle’s surface properties, aggregation state, endotoxin burden, residual solvents, injection rate, and first-contact interactions with blood and immune systems. 10 , 13 , 47

Among the best-recognized acute safety phenomena in nanomedicine is complement activation-related pseudoallergy (CARPA), which has been described for liposomes and other intravenously administered nanoparticulate systems. 13 , 52 CARPA illustrates why nanomedicine safety cannot rely only on conventional assumptions about excipient inertness or classical IgE-mediated hypersensitivity. Complement activation can trigger cardiopulmonary and systemic symptoms that are acute, clinically significant, and not readily predicted by standard cytotoxicity assays. 13 , 52 This phenomenon is particularly instructive because it demonstrates that a nanomedicine may show acceptable cellular viability in vitro yet still produce clinically meaningful acute intolerance through blood-facing immune mechanisms.

Chronic toxicity, by contrast, refers to adverse effects that emerge or persist after repeated administration or long-term retention, often over weeks to months or longer. In nanomedicine, chronic toxicity is especially relevant for inorganic nanoparticles, slowly degradable systems, repeated-dose lipid or polymer carriers, depot formulations, and products intended for long-term diseases rather than one-time or short-course use. 10 , 35 , 49 The mechanisms involved may include persistent organ accumulation, macrophage overloading, lysosomal stress, ongoing oxidative injury, prolonged inflammatory signaling, fibrosis, adaptive immune sensitization, endocrine or reproductive interference in selected materials, and long-term changes in organ function. 35 , 49 Importantly, chronic risk is not limited to overt pathology. Even moderate but persistent perturbation of immune homeostasis, hepatic processing, or splenic architecture may matter if repeated administration is required over months or years.

The distinction between acute and chronic toxicity has direct implications for study design. Acute studies emphasize early clinical chemistry, infusion reaction monitoring, blood compatibility, innate immune readouts, and immediate histopathology. Chronic studies require repeated-dose protocols, recovery cohorts, toxicokinetic interpretation, delayed histopathology, long-term biodistribution analysis, and often more detailed material fate characterization. 10 , 35 , 49 For nanomedicines, chronic studies are especially challenging because the material itself may change over time in vivo. Particles may dissolve, shed coatings, exchange corona components, degrade into small molecules or ions, or redistribute among tissues. Thus, chronic toxicity assessment must consider not only the original nanoparticle but also its transformation products and evolving biological identity. 10 , 45 , 49

The acute–chronic distinction is also clinically important because the same product can be acceptable in one context and unacceptable in another depending on dosing pattern. A nanomedicine with manageable acute infusion reactions may still be clinically useful for short-course cancer therapy, whereas the same liability could be problematic for repeated administration in chronic inflammatory disease. Conversely, a formulation with minimal acute reactogenicity may still prove unsuitable if chronic hepatic retention, splenic accumulation, or anti-carrier antibody formation limits long-term use. 35 , 51 Anti-PEG responses and accelerated blood clearance exemplify this issue: they may not dominate first-dose safety yet become highly relevant under repeated exposure. 51

In a rigorous safety framework, acute and chronic toxicity should therefore be treated as related but distinct domains. Acute toxicity concerns immediate biological incompatibilities and first-exposure hazards. Chronic toxicity concerns long-term burden, repeated-dose adaptation, retention, and delayed injury. Nanomedicine development frequently underemphasizes the latter, particularly when early proof-of-concept efficacy is prioritized. Yet for many advanced nanoparticle systems, chronic safety may be the more decisive translational barrier.

Local versus Systemic Adverse Effects

Nanomedicine safety must also distinguish between local and systemic adverse effects, because the route of administration and spatial distribution of nanoparticles strongly shape the risk profile. This distinction is not merely anatomical. It reflects different mechanisms of injury, different biomarkers, and different clinical management strategies.

Local adverse effects occur at or near the site of administration or deposition. Depending on route, these may include injection-site inflammation, tissue necrosis, endothelial irritation, pain, edema, local complement activation, pulmonary irritation after inhalation, mucosal damage after intranasal or oral exposure, ocular irritation, or local neural injury in central nervous system applications. 10 , 45 , 50 For implant-associated or locally retained nanoparticle systems, local safety may additionally involve granulomatous reactions, foreign-body responses, fibrosis, altered wound healing, and chronic inflammatory infiltration. In many cases, local effects are driven by high regional concentration, surface reactivity, osmotic or surfactant-related effects, pH shifts, or direct membrane interactions. A nanoparticle may therefore be systemically acceptable yet locally problematic if it creates irritancy or persistent inflammatory signaling at the administration site. 45 , 50

Local safety has special importance in nanomedicine because nanoparticles are often deliberately designed for site-specific retention. Intratumoral, intranasal, pulmonary, ocular, transdermal, intra-articular, and implantable nanoformulations all exploit regional delivery to improve efficacy or reduce systemic exposure. However, reducing systemic exposure does not guarantee safety if the price is concentrated local toxicity. For example, cationic or highly adhesive nanocarriers may enhance mucosal retention yet damage epithelial barriers; photothermal or catalytic nanoparticles may achieve local tumor destruction yet also injure surrounding healthy tissue if energy deposition is poorly controlled. 45 , 50 Thus, local safety must be assessed in relation to the intended tissue microenvironment and the concentration-time profile actually experienced at that site.

Systemic adverse effects involve effects beyond the administration site, arising through circulation, organ distribution, immune activation, or secondary mediator release. In nanomedicine, systemic toxicity often centers on blood–material interactions, hepatosplenic sequestration, renal handling where applicable, immune activation, endocrine perturbation in selected cases, or off-target accumulation in organs such as liver, spleen, lung, kidney, heart, or brain. 10 , 35 , 45 , 49 Systemic effects may be pharmacologic, carrier-related, or interaction-mediated. A formulation may cause systemic toxicity because the encapsulated drug is released too quickly, because the nanoparticle activates complement or cytokine pathways, because degradation products are toxic, or because long-circulating particles accumulate in immune or filtration organs.

The local-versus-systemic distinction also helps clarify why route matters so strongly in nanomedicine. Intravenous products face immediate blood compatibility and systemic distribution challenges, making complement activation, hemolysis, coagulation effects, and organ uptake central concerns. 13 , 52 Inhaled nanoparticles require careful evaluation of local pulmonary inflammation but may also produce systemic immunologic or cardiovascular effects if translocation occurs. Intramuscular vaccine nanoparticles are often judged partly by local reactogenicity, yet systemic fever, malaise, or cytokine-mediated symptoms are also central to tolerability. 50 Oral nanoparticles confront local gastrointestinal barrier interactions and microbiome-related issues, but systemic absorption and hepatic first-pass exposure remain relevant for those that are bioavailable.

In practice, local and systemic effects frequently interact. Local tissue injury can generate systemic inflammatory mediators; conversely, systemic immune priming can intensify local reactions on repeat dosing. This is particularly relevant for nanomedicines that are immunologically active by design or by consequence. A review article on nanosafety should therefore avoid treating local and systemic toxicity as mutually exclusive compartments. They are analytically distinct but biologically coupled.

For translational evaluation, local adverse effects require route-specific models, tissue histology, barrier-function assays, and exposure-aware regional pharmacology. Systemic adverse effects require biodistribution analysis, organ toxicology, hematology, complement and cytokine profiling, repeated-dose studies, and mechanistic understanding of nanoparticle fate. 10 , 13 , 35 , 45 , 47 , 49 A robust safety framework must integrate both scales, because clinical failure can result from either one.

Patient-Level, Product-Level, and Manufacturing-Level Safety Considerations

Perhaps the most important modern insight in nanomedicine safety is that risk cannot be understood solely at the level of the material. Safety is produced jointly by the patient, the product, and the manufacturing process. These three levels are analytically separable but clinically inseparable.

Patient-Level Safety

At the patient level, nanomedicine safety depends on host biology. Protein corona formation, complement activity, macrophage function, prior immune exposure, vascular permeability, inflammatory status, sex, age, ancestry, metabolic disease, organ impairment, co-medications, and disease-specific pathophysiology may all alter nanoparticle disposition and toxicity. 11 , 45 This makes nanomedicine fundamentally different from any simplistic idea of one-size-fits-all biomaterial compatibility. A formulation may behave differently in healthy volunteers, cancer patients, immunocompromised individuals, or patients with chronic liver disease because the biological environment that the nanoparticle encounters is different. 11 , 45

Protein corona research has been especially important in demonstrating this point. Once introduced into biological fluids, nanoparticles acquire adsorbed biomolecular layers that alter their effective biological identity. Recent multicenter work has emphasized the need for standardization in corona characterization precisely because small methodological and biological differences can alter interpretation and reproducibility. 11 From a patient-level safety perspective, the protein corona is not just a mechanistic curiosity; it is one route by which individual physiology can reshape clearance, immune recognition, cell targeting, and toxicity. Similarly, immunotoxicity reviews have emphasized that susceptible populations may respond differently to the same nanomaterial, particularly when pre-existing inflammatory or immune dysregulation is present. 9

Patient-level safety therefore includes questions such as these: Which patients are at increased risk of infusion reactions? Does prior exposure to PEG-containing products modify risk? Will liver dysfunction change nanoparticle clearance? Does tumor-associated inflammation alter complement activation or macrophage sequestration? Are pediatric, geriatric, or pregnant populations expected to handle the material differently? These are not peripheral considerations. They are central to precision safety assessment in nanomedicine.

Product-Level Safety

At the product level, safety depends on the complete nanoparticle construct rather than any single ingredient. Relevant properties include composition, size distribution, morphology, surface charge, ligand density, encapsulation efficiency, release kinetics, aggregation tendency, endotoxin burden, degradation profile, and biological interaction signatures such as hemocompatibility or complement activation. 6 , 46 , 47 These are often treated as critical quality attributes (CQAs) because they influence in vivo exposure, efficacy, and toxicity. 6 , 46 In nanomedicine, product-level safety is thus deeply intertwined with analytical characterization.

The literature on therapeutic nanomedicines has repeatedly emphasized that inadequate definition of CQAs undermines both translational development and safety evaluation. 6 , 46 , 47 A nanoparticle is not safe merely because its intended composition is acceptable; it must also have the right distribution of actual properties, and those properties must remain stable during storage, handling, and administration. Batch heterogeneity matters because a small fraction of larger aggregates, free drug, unstable particles, or impurity-rich species may disproportionately affect safety. Likewise, release behavior is not simply a performance parameter; it is a safety parameter because rapid leakage or altered dissolution can increase off-target exposure. 6

Product-level safety also includes immunological interface properties that are often missed in simplistic screening workflows. Hemolysis, platelet interaction, complement activation, cytokine induction, and uptake by mononuclear phagocyte system cells may all be product-specific consequences of nanoscale design. 9 , 13 , 52 This is why quality control in nanomedicine cannot stop at size and zeta potential; it must aim to identify clinically consequential material behaviors.

Manufacturing-Level Safety

At the manufacturing level, safety depends on whether the process reproducibly produces material with the intended CQAs and impurity profile. This level is sometimes underestimated in academic nanomedicine, yet it is central to clinical translation. Manufacturing variables such as raw material purity, mixing conditions, solvent removal, sterilization method, filtration, lyophilization, container closure, and storage conditions can alter size, aggregation, encapsulation, free-drug fraction, oxidation status, endotoxin contamination, and long-term stability. 6 , 46–48 These changes may not be obvious from nominal composition alone, but they can materially alter safety.

Regulatory and industry-oriented analyses now emphasize that nanomedicines demand especially strong integration of CMC, preclinical characterization, and risk-based development because formulation-process interactions are often more consequential than for conventional small molecules. 6 , 47 , 48 This is particularly evident for complex products and follow-on nanomedicines, where demonstrating sameness or bioequivalence can be difficult precisely because clinically relevant CQAs may be incompletely defined or hard to measure. 46 , 48 Manufacturing-level safety is therefore not only about preventing contamination or gross batch failure. It is about ensuring that the product reaching patients is sufficiently comparable, predictable, and stable that prior safety knowledge remains valid.

An important implication follows: manufacturing inconsistency can become a safety issue even when the nominal formulation is well designed. A clinically tolerable nanoparticle platform can become unsafe if process drift changes particle size distribution, free payload levels, complement-activating impurities, or degradation behavior. Thus, manufacturing control is not downstream of safety; it is one of its determinants.

An Integrated Definition of Safety in Nanomedicine

Taken together, these distinctions show that safety in nanomedicine is best understood as a multiscale translational property. It includes:

The appropriateness of the host response to the material in a defined use context, which relates to biocompatibility; 9 , 45

The broader hazard and risk profile across exposure conditions, which relates to biosafety; 10 , 45 , 49

The practical clinical acceptability of adverse effects at useful doses, which relates to tolerability; 50

The separation of immediate first-exposure hazards from repeated-dose and long-term risks, which relates to acute versus chronic toxicity; 13 , 35 , 49 , 51 , 52

The distinction between administration-site injury and body-wide adverse outcomes, which relates to local versus systemic effects; 10 , 45 , 50

and the interaction among patient biology, product attributes, and manufacturing reproducibility, which collectively determine real-world safety. 6 , 11 , 45–48

Determinants of Nanoparticle Toxicity

Nanoparticle toxicity is not a fixed property of a material name or platform label. It is an emergent consequence of physicochemical design, biological transformation, dose context, and manufacturing-dependent variability. In nanomedicine, toxic responses arise not only from the intrinsic chemistry of a nanoparticle core, but also from how particle dimensions influence cellular uptake, how surface chemistry reshapes protein adsorption, how colloidal instability changes effective dose, how degradation products behave in tissues, and how route and schedule determine which biological barriers and clearance systems are engaged. For this reason, the toxicology of nanoparticles must be understood as determinant-based rather than composition-only. The key determinants consistently identified across translational nanomedicine and nanosafety literature include size, shape and aspect ratio, surface charge and surface chemistry, core composition, biomolecular corona formation, aggregation and colloidal stability, degradation behavior, and the exposure variables of dose, route, and administration schedule. 6 , 10 , 20 , 31 , 53 These determinants do not operate independently. Instead, they interact to govern the most clinically relevant mechanisms of nanosafety failure, including hemocompatibility problems, complement activation, inflammatory signaling, oxidative stress, lysosomal injury, organ accumulation, chronic retention, and changes in biodistribution after repeated exposure. 9 , 10 , 20 , 31 , 53

A central implication follows from this framework: the same nanoparticle class may range from clinically useful to toxic depending on how these determinants are configured. A lipid nanoparticle that is near-neutral, biodegradable, and well-controlled in size may be clinically acceptable, whereas a closely related formulation with altered ionizable lipid chemistry, broader size distribution, or greater aggregate burden may display a substantially different inflammatory or biodistribution profile. 6 , 20 Likewise, a metallic nanoparticle may be useful when surface-engineered for imaging or externally triggered therapy, yet become problematic if dissolution, persistence, or aspect ratio lead to oxidative or chronic tissue injury. 31 , 49 The purpose of this section is therefore not merely to list determinants, but to explain how each variable shapes toxicity in translationally meaningful ways. An additional mechanistic layer is that these safety-relevant attributes are themselves products of synthesis and formulation. Precursor ratio, solvent and surfactant selection, nucleation or self-assembly conditions, polymer molecular weight, lipid composition, mixing and shear, purification, surface-conjugation chemistry, and storage conditions can alter size distribution, morphology, surface charge, ligand density, residual impurities, free-payload fraction, aggregation tendency, and degradation kinetics. Thus, synthesis is not upstream of toxicology in a purely technical sense; it is an upstream determinant of the material state that the biological system actually encounters. This explains why nanoparticles with the same nominal composition can produce different biological outcomes when manufactured by different routes or under different process conditions. 21 , 31 , 37 , 46 , 54 The coupled physicochemical determinants that shape nanoparticle safety are summarized in Figure 2 .

Schematic overview of the major physicochemical variables that influence nanoparticle safety, including size, shape and aspect ratio, surface charge, surface chemistry, core composition, aggregation state, colloidal stability, degradation behavior, and exposure conditions such as dose, route, and administration schedule. These determinants shape subsequent biological interactions, including protein corona formation, blood compatibility, immune recognition, cellular uptake, biodistribution, and clearance, which together define the overall toxicological profile of a nanoparticle system. The figure emphasizes that nanoparticle safety is a systems-level property emerging from the interaction between material design and biological context. Mechanistically, these variables are coupled: decreasing size can increase surface-area-driven reactivity and alter clearance; cationic charge can enhance membrane interaction and hemocompatibility risk; hydrophobic or reactive surfaces can intensify biomolecular adsorption; aggregation changes the effective hydrodynamic size and delivered dose; and slow degradation can convert short-term exposure into prolonged tissue burden. 10 , 20 , 31 , 53 , 55 This is an original schematic synthesis prepared specifically for the present review.

Size, Shape, and Aspect Ratio

Particle size is one of the most influential determinants of nanoparticle toxicity because it affects nearly every stage of biological interaction, including diffusion, vascular behavior, cellular uptake, endocytic route, intracellular trafficking, renal filtration, reticuloendothelial sequestration, and degradation kinetics. Smaller nanoparticles generally possess a higher surface-area-to-volume ratio, which increases interfacial reactivity and can amplify adsorption of biomolecules, catalytic activity, membrane interaction, and generation of reactive oxygen species in susceptible material systems. 34 , 53 This does not mean that smaller particles are always more toxic, but it does mean that decreasing size often changes the effective biological dose and the mode of tissue interaction. In medicinal systems, small changes within the nanometer range can shift whether particles circulate, extravasate, accumulate in liver or spleen, enter cells efficiently, or are cleared more rapidly. 34 , 35 , 53

Size also affects dose normalization, a frequently underappreciated toxicologic issue. Nanoparticles can be dosed by mass, particle number, surface area, drug content, or active ingredient concentration, and these measures are not equivalent. Two formulations with the same mass concentration but different sizes may present very different total surface area and particle number to cells or organs, thereby producing different biological effects even when nominal dose appears identical. This is one reason why in vitro-to-in vivo comparisons and cross-study toxicity claims are often difficult to interpret in the nanomedicine literature. 53 In practice, smaller particles may penetrate more effectively into tissues or intracellular compartments, while larger particles may remain more confined to vascular or phagocytic pathways. Toxicity can therefore shift from diffusely cellular to predominantly organ-sequestration-driven depending on size regime. 35 , 53

Shape and aspect ratio are equally important but historically less standardized than size in preclinical safety workflows. Spherical particles are often used as default systems because they are easier to produce and characterize, but many nanomaterials used in medicine or investigated for biomedical purposes are rods, tubes, disks, plates, worm-like micelles, or other anisotropic structures. Such geometries can profoundly alter margination in blood flow, membrane wrapping dynamics, phagocytic recognition, endosomal processing, and mechanical interaction with tissues. 53 , 56 High-aspect-ratio nanomaterials are especially important toxicologically because elongated geometries may prolong cellular interaction, frustrate phagocytosis, or increase local membrane stress. These concerns are most strongly associated with fibrous or rigid inorganic materials, but the principle extends more broadly: geometry influences biological handling, and biological handling influences toxicity. 53 , 56

In translational nanomedicine, the toxicologic significance of shape is often mediated through uptake kinetics and intracellular fate. Cells do not internalize rods, spheres, and plate-like particles identically. Differences in membrane wrapping energy and contact orientation can alter uptake rate and subcellular trafficking, which in turn can change lysosomal burden, autophagic stress, and inflammatory signaling. For immune cells, geometry may also alter the balance between rapid clearance and prolonged surface interaction, affecting cytokine release and clearance organ burden. 53 , 56 Shape therefore influences not only whether a nanoparticle enters cells, but how it is sensed and processed once internalized.

Aspect ratio also matters at the organ level. Long or rigid particles may be retained differently in pulmonary or reticuloendothelial tissues than compact isotropic nanoparticles, and they may degrade more slowly if enzymatic or hydrolytic access is geometry-limited. This distinction is especially relevant when evaluating chronic toxicity or retention of inorganic systems. Accordingly, size and shape must be considered together rather than separately. A 50 nm sphere and a 50 nm × 500 nm rod may share one nominal dimension, but their biological exposure signatures and toxicologic consequences are likely to be very different. 35 , 53 , 56

From a safety perspective, the key principle is that particle dimensions determine not only access to therapeutic targets, but also access to vulnerable biological interfaces. Optimizing size and shape for efficacy without parallel safety evaluation is therefore inadequate. In nanomedicine, dimensional design is always a toxicologic decision as well as a delivery decision. 10 , 53

Surface Charge and Surface Chemistry

Surface charge is one of the most direct determinants of nanoparticle interaction with biological systems because it governs electrostatic attraction and repulsion with membranes, proteins, glycocalyx-rich surfaces, extracellular matrix components, and blood constituents. Highly cationic nanoparticles often show efficient cellular uptake and endosomal escape in vitro because mammalian cell membranes are generally net negative. However, this apparent delivery advantage is frequently accompanied by increased nonspecific protein adsorption, membrane disruption, hemolysis, platelet interaction, complement activation, and inflammatory signaling in vivo. 9 , 20 , 57 For this reason, strongly cationic surface character is commonly associated with greater acute toxicity risk, particularly after intravenous administration. 9 , 20

Near-neutral or modestly negative surfaces are often better tolerated in systemic delivery because they reduce some electrostatic injury and may slow opsonization, but these effects are context-dependent rather than absolute. A near-neutral nanoparticle can still be immunoreactive if its interfacial chemistry, hydrophobicity, ligand density, or impurity profile promotes complement activation or unfavorable protein corona formation. 9 , 20 , 58 This is especially relevant to lipid nanoparticles, where the distinction between permanently cationic and ionizable lipids has become translationally decisive. Ionizable lipids allow particles to remain less charged at physiologic pH yet acquire positive charge in acidic endosomal compartments, thus separating systemic tolerability from intracellular delivery function more effectively than older permanently cationic systems. 20

Surface chemistry extends beyond net charge and includes hydrophobicity, steric shielding, functional group identity, ligand presentation, polymer coatings, roughness, and chemical reactivity at the interface. These factors shape the first-contact biology of nanoparticles by determining how plasma proteins adsorb, whether opsonins dominate the surface, how complement pathways are engaged, and how cells interpret the material. 20 , 58 , 59 Poly(ethylene glycol), zwitterionic coatings, polysaccharides, cell membrane-derived layers, and biomimetic ligands have all been used to reduce rapid clearance or mitigate immune recognition. Yet no coating is universally protective. PEGylation, for example, can extend circulation and improve colloidal stability, but anti-PEG antibodies and accelerated blood clearance have become well-recognized liabilities in some contexts, particularly under repeated exposure. 14 , 51

Surface chemistry also determines whether the nanoparticle is bioinert, biointeractive, or bioactive. A surface bearing targeting ligands, catalytic motifs, redox-active groups, or pH-responsive functionalities may be advantageous for efficacy, but each added function can also increase toxicologic uncertainty. Ligands may alter off-target binding; responsive groups may destabilize unexpectedly in blood; hydrophobic patches may intensify protein adsorption; and reactive groups may disturb membranes or activate innate immune pathways. 10 , 58 In this sense, increasing surface sophistication often increases the burden of safety proof.

An important translational point is that surface properties are not static after administration. Adsorption of proteins, lipids, and metabolites can partially or completely mask the designed interface, producing a new biological surface that may differ sharply from the manufactured one. Thus, the toxicity implications of surface charge and chemistry must be evaluated not only in formulation buffer but also in biologically relevant media and, ideally, across patient-relevant conditions. 11 , 58 , 59 Surface engineering remains one of the most powerful tools for reducing nanoparticle toxicity, but it is only effective when assessed under realistic exposure conditions.

Composition and Core Material

Core composition remains a foundational determinant of nanoparticle toxicity because it governs intrinsic chemical reactivity, degradability, persistence, dissolution products, density, magnetic or optical properties, and the likelihood of chronic retention. Organic carriers such as lipid-based and many biodegradable polymeric nanoparticles are often favored clinically because they are more readily metabolized or cleared and are generally associated with more familiar pharmaceutical excipients. Even within these classes, however, toxicity depends on the specific lipid, polymer, linker, stabilizer, and impurity profile rather than on broad category labels alone. 6 , 10 , 20 For example, the inflammatory profile of lipid nanoparticles depends strongly on ionizable lipid chemistry, helper lipid composition, PEG-lipid content, and degradation pathways, while polymeric systems may vary substantially depending on cationic charge density, hydrolysis rate, and residual synthetic byproducts. 6 , 20

Inorganic cores introduce different toxicologic considerations. Metallic and metal oxide nanoparticles are often attractive because of their intrinsic optical, magnetic, catalytic, or radiologic properties, but these same features may increase risk of oxidative stress, ion release, or prolonged organ retention. Gold nanoparticles are often considered relatively inert chemically, yet their toxicology still depends on size, shape, coating, and clearance behavior. Silver nanoparticles may release biologically active ions and generate oxidative injury under some conditions. Iron oxide systems offer magnetic functionality and can be clinically useful, but chronic tissue handling, macrophage processing, and dose-dependent oxidative effects remain relevant safety considerations. 31 , 35 , 49 Metal oxides such as titanium dioxide, zinc oxide, cerium oxide, and copper oxide can be particularly sensitive to redox context, dissolution, and photo- or catalytically mediated reactivity, all of which affect their toxicity profile. 31 , 49

Silica-based and carbon-based materials illustrate that “nonmetallic” does not mean toxicologically simple. Mesoporous silica nanoparticles are attractive because of their pore architecture and drug-loading capacity, but their safety depends on particle size, porosity, silanol density, degradation kinetics, and surface modification. Recent translational reviews emphasize that biosafety and clinical potential of mesoporous silica systems are strongly tied to how degradable and surface-stabilized they are in relevant biological environments. 45 Carbon-based materials, meanwhile, present a wide toxicologic spectrum ranging from comparatively favorable profiles for some carbon dots or nanodiamonds to more difficult concerns for rigid or fibrous forms such as certain nanotubes. In these systems, aspect ratio, residual catalysts, surface defects, and persistence can dominate toxicologic behavior. 31

Composition is also inseparable from degradation fate. A nanoparticle may appear acceptable initially but become problematic if its breakdown products are bioactive, acidic, oxidizing, poorly excretable, or immunogenic. Conversely, a material with moderate short-term reactivity may still be clinically acceptable if it degrades rapidly into well-tolerated metabolites. The crucial issue is not only what the nanoparticle is made of, but what it becomes after circulation, uptake, lysosomal processing, oxidation, hydrolysis, or enzymatic attack. 10 , 31 , 45 , 49 This is especially important in chronic dosing or depot settings, where accumulated transformation products can contribute to long-term toxicity.

Thus, composition and core material should be treated as necessary but incomplete descriptors of nanosafety. They set the boundaries of possible behavior, but actual toxicity emerges only when composition is interpreted together with interface properties, colloidal behavior, and exposure context. 10 , 20 , 31 , 53

Protein Corona Formation

Protein corona formation is one of the most consequential determinants of nanoparticle toxicity because it alters the biological identity of nanoparticles immediately after exposure to blood or other biological fluids. Once a nanoparticle enters a biological environment, proteins and other biomolecules adsorb to its surface, creating a dynamic corona that can reshape cellular uptake, immune recognition, complement activation, colloidal stability, organ distribution, and degradation. 11 , 58 , 59 In practical terms, the cells and tissues of the body do not “see” the pristine manufactured nanoparticle; they encounter a biologically transformed object whose interface reflects both material design and the surrounding biological milieu. 11 , 58

The toxicologic importance of the corona lies in its capacity to alter both hazard and dose delivery. Adsorption of specific plasma proteins may promote opsonization and macrophage uptake, thereby increasing liver and spleen burden while reducing access to intended target tissues. Alternatively, a corona may partially shield reactive surfaces and reduce direct membrane damage. In some cases, targeting ligands are masked by the corona, lowering efficacy without necessarily reducing systemic exposure, thereby shifting the benefit-risk balance unfavorably. 11 , 58 Corona composition can also affect complement activation and cytokine responses, making it directly relevant to infusion reactions and immunotoxicity. 9 , 58

A crucial translational development has been the recognition that corona formation is not only particle-specific but also patient-specific. Recent work emphasizes that sex, inflammatory status, disease state, metabolic phenotype, and other host variables can influence corona composition and therefore alter nanoparticle behavior among patient populations. 11 , 59 This has major implications for safety: a formulation that performs acceptably in healthy donor serum may behave differently in oncology, inflammatory disease, or critical care settings where the circulating proteome is altered. Protein corona formation is therefore one of the clearest mechanisms by which patient-level biology modifies product-level toxicity.

The corona is also methodologically challenging. Its composition depends on incubation conditions, dilution, shear, separation method, analytical workflow, and time scale. A recent multicenter study specifically addressed the need to standardize protein corona characterization in nanomedicine in order to improve reproducibility and data homogeneity across laboratories. 11 This need for standardization is not merely academic. Without robust and comparable corona analysis, toxicologic interpretation remains vulnerable to artifacts and poor cross-study comparability. In other words, protein corona science is now central to both mechanism and method in nanosafety.

From a toxicology perspective, protein corona formation should not be viewed as a secondary modifier but as a primary determinant of biological response. It links material properties to immune recognition, biodistribution, and patient heterogeneity, and it is therefore essential to any credible account of why apparently similar nanoparticles can show different safety profiles in practice. 11 , 58 , 59

Aggregation, Colloidal Stability, and Degradation Behavior

Aggregation and colloidal instability are major determinants of nanoparticle toxicity because they change the effective size, surface presentation, dose distribution, and clearance behavior of the administered material. Nanoparticles that aggregate in formulation, during storage, upon dilution, or after contact with biological fluids may no longer behave as intended. Aggregates can alter vascular passage, increase local deposition, intensify macrophage uptake, change sedimentation in in vitro assays, and modify the extent of protein adsorption or complement activation. 10 , 20 , 55 In extreme cases, aggregation can transform an otherwise acceptable nanosystem into one that produces embolic risk, infusion reactions, or highly variable organ exposure. 10 , 55

Colloidal stability is therefore both a pharmaceutical quality issue and a toxicologic issue. Stable dispersions maintain a more predictable relationship between administered dose and biological exposure, whereas unstable dispersions introduce time-dependent heterogeneity. This matters especially for intravenous products, where even short-lived changes after injection can alter first-pass interactions with blood, lung, liver, and spleen. In vitro, instability may also generate misleading toxicity readouts because sedimented or aggregated particles can expose adherent cells to far higher local concentrations than intended. 55 Thus, colloidal behavior must be interpreted under the relevant physicochemical environment, including ionic strength, protein content, pH, and shear conditions.

Degradation behavior is closely linked to colloidal stability but conceptually distinct. A nanoparticle may remain colloidally dispersed while undergoing chemical or structural degradation, or it may aggregate as a consequence of degradation-related surface changes. Degradation can be beneficial when it leads to clearance of the material and controlled release of payload, but it can also introduce new safety liabilities if intermediate or final products are reactive, acidic, inflammatory, or poorly excreted. 31 , 45 For biodegradable polymers, degradation rate influences local microenvironment and release kinetics. For lipid nanoparticles, hydrolysis or oxidation can alter potency, immunogenicity, and membrane interaction. For inorganic particles, dissolution may reduce persistence but also generate ions with distinct biological activity. 31 , 35 , 49

Chronic toxicity is often shaped by the balance between persistence and degradability. Reviews of inorganic nanoparticle delivery systems emphasize that subchronic and chronic toxicity depend heavily on how long particles or their products remain in tissues and whether clearance pathways are efficient. 35 Similarly, analyses of inorganic nanomaterials and mesoporous silica systems identify stability, biodegradation, and long-term fate as central translational bottlenecks because retention in liver, spleen, or other organs may convert short-term tolerability into long-term concern. 45 , 60 Thus, neither “stable” nor “degradable” is inherently safer; what matters is whether the time course and products of degradation are compatible with the intended dose and exposure duration.

A robust nanosafety framework therefore evaluates aggregation and degradation not as isolated physicochemical descriptors but as in vivo fate determinants. They determine what form of the material the body actually experiences over time, and that form may differ substantially from the product originally characterized in formulation buffer. 10 , 20 , 45 , 55

Dose, Exposure Route, and Administration Schedule

Dose remains a central determinant of toxicity in nanomedicine, but it is more complex than in many conventional pharmacologic settings because the relevant dose metric may differ depending on mechanism. Mass dose alone may not capture biologically relevant exposure when particle number, surface area, or free versus encapsulated fraction change between formulations. 53 For some toxic effects, such as oxidative stress or membrane interaction, surface area may correlate more closely with biological burden than mass. For others, particle number or total delivered active ingredient may be more informative. This complexity does not abolish dose-response relationships; it means that nanomedicine toxicology must choose dose metrics carefully and interpret them mechanistically. 53

The route of exposure or administration fundamentally shapes which toxicities are most likely. Intravenous administration immediately exposes nanoparticles to blood components, making hemocompatibility, complement activation, coagulation effects, and organ sequestration especially important. 9 , 51 Inhaled nanoparticles must be evaluated for pulmonary deposition, epithelial injury, macrophage burden, and possible systemic translocation. Oral exposure introduces gastrointestinal barriers, enzymatic conditions, and microbiome interactions, while intramuscular or subcutaneous administration raises questions about local reactogenicity, lymphatic trafficking, and depot persistence. 10 , 53 , 61 Thus, a nanoparticle cannot be classified as safe or unsafe independent of route; the route determines which interfaces are challenged first and which organs define early exposure.

Administration schedule adds another layer. Single-dose toxicity may differ substantially from repeated-dose toxicity because of accumulation, adaptive immune responses, changing protein corona patterns, or altered clearance after prior exposure. This is particularly relevant for PEGylated or other long-circulating systems, where repeated administration can provoke anti-carrier antibodies and accelerated blood clearance, thereby modifying both exposure and safety over time. 14 , 51 Repeated dosing may also overwhelm clearance pathways or create chronic low-grade inflammation even when first-dose tolerability appears acceptable. 35 Accordingly, schedule should be treated as a toxicity determinant in its own right rather than merely a dosing convenience.

Infusion rate and administration method can also alter acute toxicity. Complement-mediated reactions or local concentration spikes may be influenced by how rapidly a formulation enters systemic circulation. Likewise, bolus versus slow infusion can change peak nanoparticle concentration at blood-facing interfaces, potentially modifying acute cardiopulmonary or immune effects. 9 , 51 These practical considerations are especially important in clinically deployed nanomedicines, where administration protocols are part of the safety profile rather than external to it.

Taken together, dose, route, and schedule define the exposure architecture of a nanomedicine. They determine not just how much material reaches the body, but how fast, in what form, through which barrier, and how often. Because nanoparticle toxicity is often interface-dependent and time-dependent, these exposure variables are as important as the nanoparticle’s intrinsic physicochemical properties. 10 , 35 , 53 , 61

Integrative View: Toxicity as a Systems Property

The determinants discussed above are often presented in separate categories, but translationally they operate as a system. Size influences corona formation; corona formation reshapes surface charge effects; surface chemistry alters aggregation; aggregation changes effective size and clearance; composition determines degradation products; route and schedule determine which of these properties matter most in vivo. For example, a cationic, small, biodegradable nanoparticle may show potent cellular uptake but also strong protein adsorption and acute inflammatory reactivity after intravenous dosing. The same core material, reformulated to alter surface chemistry and schedule, may become substantially safer without any change in therapeutic payload. 9 , 10 , 20 , 53 Likewise, an inorganic nanoparticle with useful imaging properties may be clinically acceptable at low dose for intermittent imaging but not for chronic therapeutic use if retention and long-term tissue burden become limiting. 31 , 35 , 49

This systems view is increasingly reflected in modern nanomedicine development. Recent translational analyses emphasize that safety cannot be inferred from one property at a time, nor from simplified in vitro screens alone. Instead, determinants must be evaluated in combination and under realistic biological conditions, including attention to patient-dependent corona biology, immune profiling, critical quality attributes, and repeated-dose fate. 6 , 10 , 11 , 20 The practical consequence is that nanoparticle toxicity should be managed through safe-by-design integration: selecting size, geometry, chemistry, and exposure parameters not only for efficacy but also for predictable interaction with blood, tissues, and clearance systems.

In this sense, determinants of toxicity are not merely risks to be catalogued after formulation. They are design variables that define whether a nanoparticle platform is likely to become a viable medicine. Nanomedicine succeeds when these variables are engineered as a coherent biological system rather than as isolated material features. 6 , 10 , 20 , 53

Biological Interactions That Govern Safety

The safety of nanomedicines is determined not only by their nominal physicochemical properties but by the biological interactions that occur immediately after administration and continue throughout circulation, cellular uptake, tissue deposition, degradation, and clearance. 34 , 52 , 62–70 In practical terms, the body does not respond to a nanoparticle as a static material. It responds to a dynamic entity that encounters blood proteins, immune cells, endothelial barriers, phagocytic organs, intracellular trafficking machinery, oxidative pathways, and tissue-specific clearance systems. 34 , 52 , 62–70 These interactions govern whether a nanoparticle remains clinically tolerable, triggers acute immune or hemolytic toxicity, accumulates silently in the mononuclear phagocyte system, or produces delayed organ injury after repeated dosing. 34 , 52 , 62–70 For this reason, nanosafety must be understood as a consequence of nano–bio interactions rather than as a simple extension of conventional drug toxicology. 34 , 52 , 62–70 This section reviews the key biological interactions that most directly shape the safety profile of medical nanoparticles: blood compatibility, immune recognition and complement activation, mononuclear phagocyte system uptake, cellular internalization and intracellular trafficking, oxidative stress and downstream injury pathways, and organ accumulation with long-term retention. 34 , 52 , 62–70 The causal sequence linking biointerface formation, immune recognition, biodistribution, intracellular injury, clearance, and persistence is summarized in Figure 3 .

Schematic representation of the principal biological pathways that govern nanoparticle safety after administration. Following systemic or local exposure, nanoparticles interact with blood components and plasma proteins, undergo protein corona formation, and may activate coagulation and complement pathways. These early interactions influence immune recognition, uptake by the mononuclear phagocyte system, tissue biodistribution, and cellular internalization. Intracellular trafficking may then lead to lysosomal stress, oxidative stress, inflammatory signaling, genotoxicity, mitochondrial dysfunction, and apoptosis. In parallel, nanoparticles may be biodegraded and cleared or may persist in organs such as liver and spleen, contributing to long-term retention and chronic toxicity. The figure highlights the dynamic relationship between early biointerface events, organ distribution, cellular injury mechanisms, and eventual clearance or persistence. The causal sequence represented by the schematic is exposure → biointerface formation → immune and cellular recognition → organ distribution and intracellular trafficking → oxidative, mitochondrial, lysosomal, inflammatory, genotoxic, apoptotic, clearance, or persistence outcomes. This is an original schematic synthesis prepared specifically for the present review.

Blood Compatibility and Hemocompatibility

Blood compatibility is one of the earliest and most consequential determinants of nanomedicine safety because blood is either the first compartment encountered after intravenous administration or the gateway through which nanoparticles administered by other routes eventually reach systemic tissues. 63 , 64 Once in blood, nanoparticles interact with erythrocytes, platelets, leukocytes, plasma proteins, endothelial surfaces, and the coagulation and complement systems. 52 , 63–65 These interactions may alter membrane integrity, induce hemolysis, activate platelets, perturb coagulation, trigger leukocyte activation, or reshape the protein corona in ways that influence subsequent biodistribution and immune clearance. 52 , 63–65 Reviews of nanomedicine hemocompatibility consistently emphasize that these effects are not peripheral screening concerns but central translational determinants, especially for systems intended for intravenous use. 63 , 64

Hemocompatibility is broader than hemolysis alone. 63 , 64 A nanoparticle can be nonhemolytic yet still be incompatible with blood if it promotes platelet aggregation, shortens coagulation times, activates leukocytes, or alters red-cell deformability. 63 , 64 Likewise, normal results in a single blood assay do not establish global compatibility because different blood constituents respond through distinct mechanisms. 63 , 64 The most clinically relevant evaluation therefore includes erythrocyte integrity, platelet behavior, coagulation pathways, leukocyte activation, plasma protein adsorption, and, where relevant, endothelial interaction. 63 , 64 This breadth is essential because nanomedicines may fail clinically through acute blood-mediated events long before classic organ toxicities emerge. 52 , 63 , 64

The protein corona is also central to hemocompatibility. 64–66 Plasma proteins adsorb to nanoparticles within seconds and can either buffer or amplify blood-facing toxicity. 65 , 66 Corona formation may mask reactive surfaces and reduce direct membrane damage, but it may also enrich opsonins or complement proteins that increase leukocyte activation and clearance. 52 , 65 , 66 Accordingly, blood compatibility is not simply an intrinsic property of the pristine formulation; it is a property of the nanoparticle as transformed in blood. 64–66 This is one reason why thorough hemocompatibility testing under biologically relevant conditions is so strongly emphasized in translational nanosafety literature. 52 , 63–66

Immune Recognition and Complement Activation

Immune recognition is one of the defining biological challenges of nanomedicine. 52 , 65 , 66 Even when nanoparticles are engineered to be long-circulating or “stealth,” they are still subject to surveillance by innate immune systems that interpret particulate structures, adsorbed proteins, repeated surface motifs, and damaged-self signals. 52 , 65 , 66 Complement activation is especially important because it can occur rapidly in blood and produce clinically meaningful infusion reactions, inflammatory symptoms, vascular effects, and accelerated clearance. 52 , 65 Complement proteins that bind nanoparticles may also function as opsonins, thereby linking acute immune activation to downstream phagocytic uptake. 52 , 65 , 67

Complement activation-related pseudoallergy, or CARPA, remains one of the best-characterized immune safety phenomena in nanomedicine. 52 It is particularly relevant to liposomes and other intravenously administered nanoparticulate systems and illustrates that clinically significant hypersensitivity-like reactions can arise through non-IgE-mediated innate pathways. 52 This has two major implications. First, conventional assumptions about excipient inertness are insufficient for nanomedicine 47 . Second, standard cytotoxicity assays are poor predictors of some of the most clinically relevant acute reactions. 52 CARPA therefore serves as a model for why nanosafety must include mechanistic immune assessment rather than relying primarily on general toxicity screening. 52

Immune recognition also extends beyond complement. 52 , 65 , 66 Protein corona composition can alter how nanoparticles are perceived by monocytes, macrophages, dendritic cells, and other innate cells. 65 , 66 Protocol-focused reviews note that complement proteins in the corona can generate both opsonins such as C3b and inflammatory mediators such as C3a and C5a, thereby coupling surface adsorption to phagocytosis and inflammatory signaling. 65 More recent work has also shown that inflammatory disease states can reshape nanoparticle corona composition and shift innate immune activation profiles through pathways including TLR4/MyD88/NF-κB signaling. 65 , 66 This is highly relevant to translational safety because it means that immune recognition is not only product-dependent but also disease-dependent. 65 , 66

Repeated exposure introduces additional immune complexity. 52 , 65 PEGylated systems can prolong circulation and reduce some early opsonization, but anti-PEG antibodies may arise from prior environmental or pharmaceutical exposure and can accelerate blood clearance or increase immune reactivity after repeat dosing. 52 This phenomenon is clinically important because it changes both safety and efficacy over time. 52 A nanomedicine may be acceptable on first administration but behave differently after immunologic priming. 52 In this sense, immune recognition is not a static barrier but a dynamic interaction shaped by exposure history. 52 , 65 , 66

Mononuclear Phagocyte System Uptake

The mononuclear phagocyte system, or MPS, is a dominant determinant of nanoparticle safety and disposition after systemic administration. 66–68 It includes resident macrophages in organs such as liver and spleen, circulating monocytes, dendritic cells, and related phagocytic populations. 66 , 67 Rapid recognition and uptake by the MPS limits nanoparticle circulation time and strongly influences the distribution of both carrier and cargo. 66–68 This is often discussed mainly as a delivery barrier, but it is equally a safety determinant because MPS uptake governs which organs receive the highest off-target exposure and where chronic retention or inflammatory burden may develop. 66–68

Recent analyses have summarized this issue clearly: rapid uptake by the MPS significantly hampers nanoparticle efficacy, many particles have circulation half-lives of only minutes, and less than 1% of the injected dose reaches intended target tissue in many settings. 67 , 68 The MPS is a primary barrier after intravenous administration, and serum protein coating drives recognition by MPS cells through multiple receptors. 65–68 From a safety perspective, these findings matter because liver and spleen become the major recipients of systemic nanoparticle burden. 67 , 68 Quantitative meta-analysis of biodistribution datasets likewise shows that, after intravenous administration, nanoparticles typically accumulate most strongly in liver and spleen, with much lower deposition in most other tissues. 68

This MPS bias has several safety consequences. 66–68 It can protect other organs from exposure, but it can also create chronic burdens in phagocytic tissues, alter macrophage function, and increase the importance of degradation fate within lysosomal compartments. 66–68 For nanoparticles intended for repeated administration, MPS uptake may also produce cumulative exposure patterns that are not evident from first-dose studies. 66–68 As a result, MPS interaction should be viewed not merely as an efficacy obstacle but as a core axis of nanosafety, especially for inorganic or slowly degradable materials. 66–68

Cellular Internalization and Intracellular Trafficking

Cellular internalization is a prerequisite for the activity of many nanomedicines, but it is also a major determinant of toxicity. 62 , 69 Nanoparticles generally do not diffuse freely across plasma membranes and instead enter cells through endocytic pathways. 69 The uptake route, uptake rate, and ultimate intracellular destination affect whether the material is recycled, degraded, trapped in endosomes or lysosomes, released into the cytosol, or delivered to sensitive organelles such as mitochondria or nuclei. 62 , 69 Major reviews on intracellular trafficking emphasize that the entry pathway and intracellular localization of nanoparticles are key determinants of both biomedical function and cytotoxicity. 62 , 69

This relationship matters because intracellular trafficking governs exposure at the subcellular level. 62 , 69 Endosomal and lysosomal sequestration can be beneficial when degradative release is intended, but it can also create local acidification-related instability, lysosomal membrane stress, and prolonged intracellular retention. 62 , 69 Cytosolic escape may be essential for nucleic acid therapeutics, yet the membrane-disruptive strategies used to achieve it can increase toxicity. 62 , 69 Similarly, trafficking to mitochondria or prolonged association with endomembrane systems can intensify reactive oxygen species generation or apoptotic signaling. 62 , 69 Thus, the question is not simply whether cells take up nanoparticles, but how they process them after uptake. 62 , 69

Physicochemical features such as size, shape, charge, and surface chemistry strongly affect uptake pathways and downstream safety. 62 , 69 The intracellular-trafficking literature notes that these variables influence internalization mechanism and biological function; therefore, optimizing intracellular delivery without considering toxicity is not feasible. 62 , 69 In nanomedicine, a delivery gain achieved through more aggressive uptake may be offset by greater membrane damage, lysosomal overload, or inflammatory signaling. 62 , 69 Intracellular trafficking is therefore one of the clearest sites where efficacy and safety are mechanistically inseparable. 62 , 69 Mechanistically, nanoparticle internalization can involve clathrin-mediated or caveolae-associated endocytosis, macropinocytosis, phagocytosis, and receptor-mediated uptake, with the dominant route depending on particle properties, cell phenotype, ligand presentation, and the adsorbed biomolecular corona. Receptor-mediated uptake may increase selectivity when ligand density and receptor accessibility are appropriate, whereas highly cationic or membrane-active designs can increase nonspecific adsorption and uptake at the cost of membrane injury. Because endocytic route influences endosomal maturation, lysosomal exposure, recycling, and the probability of cytosolic escape, apparently similar uptake levels can lead to very different efficacy and toxicity profiles. 62 , 69

Oxidative Stress, Inflammation, Genotoxicity, and Apoptosis

Oxidative stress and inflammation are among the most common convergent mechanisms of nanoparticle-induced injury. 34 , 62 , 70 Reviews of metal nanoparticles note that interactions with DNA, proteins, mitochondria, and other cellular structures can trigger reactive oxygen species production, mitochondrial dysfunction, altered ATP levels, lysosomal changes, membrane impairment, and apoptosis. 34 Although these findings are often discussed in the context of metallic nanoparticles, the broader principle extends across nanomedicine: ROS generation and inflammatory signaling are frequent downstream consequences of unfavorable nano–bio interactions. 34 , 62 , 70

Inflammation can emerge through several routes, including complement activation, inflammasome engagement, macrophage activation, damage-associated molecular signaling, and oxidative injury. 34 , 52 , 62 , 65 Once initiated, inflammatory pathways may amplify local tissue toxicity, alter vascular permeability, recruit immune cells, and reshape corona composition or phagocytic uptake. 34 , 62 , 65 , 66 This creates feedback loops in which early oxidative or membrane stress leads to broader inflammatory injury. 34 , 62 , 70 Such mechanisms are especially important for particles that persist intracellularly, undergo redox cycling, or accumulate in phagocytic organs. 34 , 62 , 68 , 70

Genotoxicity is a more specialized but important safety endpoint. 62 , 70 It may arise directly through interaction of nanoparticles or released ions with DNA-associated structures, or indirectly through oxidative stress, inflammation, and interference with chromatin regulation. 34 , 62 , 70 Updated reviews on silica nanoparticles conclude that these materials can damage genetic material at the base, DNA-molecular, and chromosomal levels and that oxidative stress is a major mechanism underlying these effects. 70 The same work also emphasizes epigenetic alterations, including changes in DNA methylation, miRNA expression, histone modification, and chromatin remodeling. 70 These findings matter translationally because persistent sublethal genomic stress can be relevant even when overt cytotoxicity is limited. 62 , 70 A robust genotoxicity evaluation should therefore combine complementary endpoints rather than infer DNA injury from oxidative stress alone. Depending on the material and intended exposure, relevant approaches include the alkaline comet assay for DNA strand breaks, micronucleus testing for chromosomal damage or loss, gamma-H2AX-based assessment of DNA double-strand-break signaling, and chromosome-aberration or related cytogenetic assays. Nanoparticle-specific controls remain essential because adsorption, optical interference, sedimentation, or secondary inflammatory effects can complicate interpretation, particularly when a single assay is used in isolation. 62 , 70 , 71

Apoptosis often appears as a downstream outcome of oxidative stress, organelle injury, and inflammatory signaling. 34 , 62 , 70 In nanomedicine this may be therapeutically desirable in tumor cells but problematic in healthy tissues. 34 , 62 The safety challenge is therefore context-dependent: the same pathway can be beneficial when restricted to target cells and harmful when triggered systemically or in clearance organs. 34 , 62 , 70 This duality reinforces a broader principle of nanosafety: a biologic mechanism cannot be labeled simply as good or bad without considering where, when, and in which cells it occurs. 34 , 62 , 70

Organ Accumulation and Long-Term Retention

Organ accumulation and long-term retention are critical determinants of chronic nanoparticle safety. 62 , 66–68 Quantitative biodistribution analysis indicates that, after intravenous administration, liver and spleen are typically the principal sites of nanoparticle accumulation, with kidneys, lungs, and tumors generally receiving much smaller fractions on average. 67 , 68 This is not only a pharmacokinetic observation; it is a safety map. 67 , 68 It identifies the organs most likely to experience chronic exposure, macrophage burden, degradation product release, and delayed inflammatory or fibrotic effects. 62 , 67 , 68

Long-term safety depends on whether accumulated particles are efficiently degraded and cleared or instead persist in tissues. 62 , 67 , 68 Reviews of inorganic nanoparticle toxicity emphasize that subchronic and chronic outcomes are strongly shaped by retention time, organ-specific processing, and the biological activity of degradation products. 62 For magnetic iron oxide nanoparticles, recent reviews highlight that toxicity is highly dependent on coating chemistry, aggregation, dose, and degradation kinetics rather than on core composition alone. 62 This illustrates a general rule: chronic safety is determined by the fate of the particle within organs, not just by the identity of the starting material. 62 , 67 , 68

Long-term retention also has immunological implications. 62 , 66–68 Persistent particles in liver, spleen, or other tissues may sustain low-grade inflammation, alter macrophage phenotype, or create organ-specific stress that is not obvious in short-term studies. 62 , 66–68 This is one reason why acute tolerability cannot be taken as evidence of chronic safety. 62 , 67 , 68 In nanomedicine, clearance kinetics and tissue persistence are as important as immediate blood compatibility. 52 , 62–64 A formulation that is well tolerated on first administration may still be unsuitable for chronic use if it accumulates progressively in phagocytic organs or generates slowly cleared transformation products. 62 , 67 , 68

Integrative Perspective

These biological interactions are best understood as a connected safety network rather than independent checkpoints. 34 , 52 , 62–70 Blood compatibility affects corona formation and complement activation. 52 , 63–66 Complement activation and opsonization accelerate MPS uptake. 52 , 65–68 MPS uptake determines organ accumulation and intracellular lysosomal burden. 62 , 66–68 Intracellular trafficking influences oxidative stress, inflammation, apoptosis, and potentially genotoxicity. 34 , 62 , 69 , 70 Long-term retention feeds back into chronic immune and tissue responses. 62 , 67 , 68 The safety of a nanoparticle therefore emerges from the sequence and coupling of these interactions across time and biological scale. 34 , 52 , 62–70

For translational nanomedicine, the practical conclusion is clear: safety cannot be inferred from one assay, one organ, or one time point. 34 , 52 , 62–70 It must be evaluated through an integrated framework that links blood-facing behavior, immune recognition, phagocytic clearance, intracellular fate, stress-response pathways, and long-term tissue disposition. 34 , 52 , 62–70 This is the level at which “biocompatibility” becomes clinically meaningful. 34 , 52 , 62–70

Organ-Specific Safety Concerns

Organ-specific toxicology is a decisive component of nanomedicine safety because nanoparticles rarely distribute uniformly after administration. 67 , 68 , 72–82 Instead, their disposition is shaped by vascular architecture, endothelial permeability, protein corona composition, macrophage surveillance, renal filtration constraints, and tissue-specific clearance pathways. 67 , 68 , 75–77 As a result, particular organs repeatedly emerge as dominant sinks or vulnerable interfaces, especially the liver, spleen, kidney, lung, brain, cardiovascular system, and reproductive tissues. 67 , 68 , 72–82 This matters translationally because a nanomedicine may appear systemically tolerable while still imposing clinically relevant burdens on one or more organ systems through sequestration, prolonged retention, local inflammatory signaling, endothelial injury, or organ-specific accumulation of carrier degradation products. 67 , 68 , 72–82 Modern biodistribution analyses show that after intravenous administration, nanoparticles accumulate predominantly in liver and spleen, with far lower fractions generally reaching most other tissues, reinforcing that organ safety is inseparable from disposition biology. 67 , 68

Liver and Spleen Sequestration

The liver and spleen are the principal organs of nanoparticle sequestration after systemic administration because they contain dense populations of phagocytic cells and highly specialized vascular structures that favor particle capture. 67 , 68 In the liver, Kupffer cells, sinusoidal endothelial cells, and, in some settings, hepatocytes participate in nanoparticle uptake. 67 , 68 In the spleen, macrophage-rich red pulp and filtration architecture contribute to retention of circulating particulate material. 67 , 68 This pattern is so consistent that it is best regarded as a default biodistribution outcome for many intravenously administered nanomedicines rather than as an exception. 67 , 68 Recent analyses emphasize that rapid uptake by the mononuclear phagocyte system remains a primary barrier after intravenous administration, while quantitative biodistribution studies identify liver and spleen as the dominant recipients of systemic nanoparticle dose. 67 , 68

From a safety perspective, hepatosplenic sequestration is not merely an efficacy limitation. 67 , 68 It is also the main determinant of chronic off-target exposure for many nanomedicines. 67 , 68 Accumulation in Kupffer cells and splenic macrophages can alter innate immune tone, sustain low-grade inflammatory signaling, and create organ-specific burdens that become more pronounced under repeated dosing. 67 , 68 , 82 For biodegradable platforms, the liver and spleen are often where degradation and metabolite release primarily occur. 67 , 68 For inorganic or slowly degradable particles, these organs may become sites of long-term persistence, lysosomal storage stress, oxidative injury, or fibrosis-related concern. 67 , 68 , 82 Reviews of inorganic nanoparticle toxicity consistently identify liver and spleen burden as central to subchronic and chronic safety evaluation. 67 , 68 , 82

The translational meaning of hepatosplenic uptake is context-dependent. 67 , 68 In some applications, such as liver-directed RNA therapeutics, hepatic accumulation is partly advantageous. 67 , 68 However, even then, the distinction between therapeutic hepatocyte delivery and nontarget sequestration by Kupffer cells or splenic macrophages remains crucial. 67 , 68 A formulation that reaches the liver is not necessarily reaching the intended hepatic cell type in a safe or efficacious way. 67 , 68 Therefore, liver and spleen safety assessment should include not only bulk organ accumulation but also cell-specific localization, inflammatory readouts, histopathology, and the fate of degradation products over time. 67 , 68 , 82

Renal Clearance and Nephrotoxicity

The kidney occupies a dual role in nanomedicine safety: it is both a potential clearance route and a potential toxicity target. 75–77 Renal elimination is generally desirable for nanoparticles or degradation products that are sufficiently small and appropriately surface-engineered, because efficient urinary excretion can reduce long-term accumulation in liver and spleen. 75 , 76 Classical work on nanoparticle renal filtration showed that ultrasmall nanostructures can undergo rapid urinary clearance when hydrodynamic properties fall within the renal filtration window, and subsequent reviews have emphasized that size, shape, surface charge, and protein adsorption all influence renal passage. 75 , 76

However, renal clearance is not synonymous with renal safety. 75–77 Nanoparticles or released components that enter the kidney may interact with glomerular structures, proximal tubular cells, podocytes, mesangial cells, or renal endothelial compartments. 75 , 77 Comprehensive toxicology reviews conclude that nanoparticles can produce nephrotoxic effects at both tubular and glomerular levels through mechanisms including oxidative stress, mitochondrial injury, cytoskeletal disruption, and inflammatory signaling. 75 , 77 These risks are especially relevant when particles accumulate intracellularly in renal epithelia or when repeated dosing is used in patients with impaired kidney function, where clearance may be reduced and retention increased. 75–77

The kidney also introduces a safety paradox in nanomedicine design. 75–77 Features that favor renal clearance, such as very small size or specific surface properties, may reduce chronic systemic retention but can simultaneously increase direct renal exposure. 75–77 Conversely, larger particles that avoid renal filtration may spare the kidney acutely while increasing hepatosplenic burden. 67 , 68 , 75–77 Thus, kidney safety is not determined solely by whether a particle is renally cleared, but by how much material reaches renal microanatomical compartments, how long it remains there, and whether disease-related reductions in glomerular filtration or tubular handling alter its fate. 75–77 For translational assessment, this requires integration of biodistribution, urinalysis, serum chemistry, renal histology, and disease-state pharmacokinetics. 75–77

Pulmonary Toxicity

The lung is a major safety concern for nanoparticles because it can be exposed both directly, through inhalation or inhaled formulations, and indirectly, through systemic circulation, pulmonary first-pass interactions, immune-cell trafficking, or embolic deposition of unstable aggregates. 78 , 79 Pulmonary toxicology is governed by particle deposition pattern, aerodynamic and hydrodynamic behavior, epithelial interaction, alveolar macrophage uptake, mucociliary clearance, and inflammatory amplification within the delicate gas-exchange architecture. 78 , 79 Recent reviews emphasize that nanoparticle toxicity in the respiratory tract is shaped by particle properties, exposure conditions, and deposition behavior, with oxidative stress and inflammation as recurrent mechanisms. 78

In pulmonary delivery, the lung offers attractive therapeutic opportunities because of its large surface area and accessible epithelial interface, but those same features create vulnerability. 78 , 79 Excessive local deposition, reactive surfaces, or poorly degradable materials can damage alveolar epithelium, activate macrophages, alter surfactant function, and impair respiratory mechanics. 78 , 79 These risks are not limited to environmental nanoparticles; they are directly relevant to inhaled nanomedicines, particularly when formulation stability, aerodynamic sizing, or mucosal interaction is suboptimal. 78 , 79 Reviews of nanoparticle-based inhaled drug delivery highlight the translational promise of respiratory nanomedicine while underscoring the need to control local toxicity, retention, and inflammatory responses. 79

Even for systemically administered nanoparticles, pulmonary safety matters. 78 , 82 Complement-mediated reactions, aggregate trapping, endothelial interaction, and inflammatory cross-talk can produce lung involvement without primary inhalation exposure. 78 , 82 Accordingly, pulmonary safety assessment should not be restricted to inhaled formulations alone. 78 , 79 It should also be considered in intravenous nanomedicine development, particularly for formulations with aggregation liability, strong complement activation potential, or pronounced leukocyte interaction. 78 , 82

Neurotoxicity and Blood–Brain Barrier Considerations

The central nervous system presents a distinctive safety challenge because the blood–brain barrier, or BBB, strongly restricts entry of most circulating materials, yet some nanoparticles can cross or perturb this barrier depending on size, surface chemistry, ligand decoration, inflammation, or disease-induced BBB disruption. 80 , 81 This has obvious therapeutic importance for CNS drug delivery, but it also raises neurotoxicologic concerns. 80 , 81 Recent reviews note that nanocarriers can cross the BBB and may produce neurotoxic effects through mechanisms including oxidative stress, DNA damage, neuroinflammation, accumulation, and persistence in brain tissue. 80

Neurotoxicity in nanomedicine may arise through several distinct mechanisms. 80 , 81 Nanoparticles may directly enter neural tissue and interact with neurons, astrocytes, microglia, or cerebrovascular endothelium. 80 , 81 They may indirectly compromise CNS homeostasis by disrupting BBB integrity, altering endothelial tight junctions, or provoking systemic inflammatory signals that secondarily affect the brain. 80 , 81 In addition, disease-targeting strategies that intentionally enhance BBB crossing may narrow the margin between therapeutic access and unwanted CNS exposure. 80 , 81 Reviews of BBB-targeting nanoparticles emphasize both the promise of biomaterial engineering for brain delivery and the importance of understanding BBB models, pathological barrier alteration, and nanoparticle permeability determinants. 80 , 81

These issues make BBB-oriented nanomedicine unusually safety-sensitive. 80 , 81 A carrier designed to overcome the BBB is, by definition, designed to reach one of the body’s most protected organs. 80 , 81 Therefore, neurotoxicity assessment should move beyond general cytotoxicity and address BBB integrity, microglial activation, behavioral outcomes where relevant, persistence in CNS compartments, and the long-term fate of materials or degradation products in neural tissue. 80 , 81 This is especially important for chronic neurological indications, repeated dosing, and platforms with uncertain CNS clearance. 80 , 81 The principal strategies for BBB access also carry different safety trade-offs. Receptor-mediated transcytosis can improve endothelial uptake and transcellular transport but may be limited by receptor saturation, ligand density, endosomal routing, and loss of targeting after protein-corona formation. Adsorptive-mediated transport can exploit electrostatic interactions but generally provides less specificity and may increase membrane perturbation. Carrier-mediated strategies rely on endogenous nutrient-transport systems and therefore face competition and capacity constraints, whereas biomimetic or cell-mediated approaches may improve biological recognition at the cost of greater compositional and manufacturing complexity. These approaches should therefore be compared according to transport efficiency, specificity, barrier integrity, intracellular fate, repeat-dose feasibility, and manufacturability rather than by BBB crossing alone. 80 , 81

Cardiovascular Effects

The cardiovascular system is both the transport network for systemically administered nanoparticles and a direct toxicity target. 68 , 82 Nanoparticles can interact with vascular endothelium, circulating blood cells, coagulation factors, autonomic signaling pathways, and cardiac tissues. 68 , 82 Recent reviews summarize that nanoparticle-associated cardiovascular adverse events include arrhythmia, cardiac hypertrophy, atherosclerosis, thrombosis, endothelial dysfunction, oxidative stress, and multiple forms of programmed cell death. 82 These effects are mechanistically plausible because the cardiovascular system experiences immediate nanoparticle exposure after intravenous administration and secondary exposure after redistribution from other routes. 68 , 82

Cardiovascular toxicity may manifest acutely as hemodynamic instability, altered vasoreactivity, complement-mediated cardiopulmonary symptoms, procoagulant activity, endothelial permeability changes, or platelet activation. 68 , 82 It may also emerge subacutely or chronically through vascular inflammation, oxidative injury, atherogenic signaling, myocardial stress, or thrombogenic remodeling. 68 , 82 Importantly, some of these effects may be driven more by surface properties, aggregation, and immune activation than by core chemistry alone. 68 , 82 This means that a nanomedicine with acceptable general tolerability can still carry cardiovascular risk if it perturbs endothelial function or coagulation. 68 , 82

From a translational standpoint, cardiovascular safety deserves special attention in oncology nanomedicine, repeated-dose intravenous regimens, and formulations designed for long circulation. 68 , 82 Many of these products are administered to patients who already have cardiovascular vulnerability due to age, inflammation, prior chemotherapy, or metabolic disease. 68 , 82 Thus, cardiovascular assessment should include not only general blood compatibility but also vascular reactivity, thrombogenicity, endothelial integrity, cardiac biomarkers where appropriate, and interaction with preexisting cardiovascular risk states. 68 , 82

Reproductive and Developmental Safety Issues

Reproductive and developmental safety is one of the least resolved but most important domains in nanoparticle toxicology. 72–74 Recent reviews emphasize that nanoparticles may affect fertility, gonadal function, endocrine balance, gamete quality, embryonic development, placental biology, and potentially later-life offspring outcomes, although the magnitude and generalizability of these risks remain incompletely defined across nanoparticle classes. 72–74 Recent toxicology reviews conclude that the effects of different nanoparticles on reproductive function are not yet fully determined and that further work is required to clarify reproductive toxicity across in vitro and in vivo systems. 72

Several mechanistic concerns recur in this literature. 72–74 Nanoparticles may cross biological barriers relevant to reproduction, including the blood-testis barrier and, in some settings, the placental interface. 72–74 They may also disrupt reproductive tissues indirectly through oxidative stress, inflammation, endocrine perturbation, mitochondrial dysfunction, or epigenetic alteration. 72–74 The concern is not limited to overt infertility. 72–74 Developmental toxicology also encompasses altered embryo-fetal growth, placental dysfunction, congenital abnormalities, developmental programming, and possible transgenerational effects. 72–74 Reviews on nanoparticle-mediated embryotoxicity underscore the importance of understanding how nanoparticle properties contribute to developmental injury and maternal-fetal exposure risk. 73

For nanomedicine, these issues are especially relevant in chronic indications, products likely to be used in women of reproductive age, and platforms with prolonged systemic persistence. 72–74 They also matter for oncology and inflammatory disease, where treatment may coincide with reproductive vulnerability or pregnancy. 72–74 Reproductive safety assessment should therefore not be treated as an afterthought or a generic toxicology appendix. 72–74 It requires targeted evaluation of gonadal distribution, placental transfer where relevant, endocrine and gamete endpoints, embryo-fetal development, and the possibility that disease states may modify maternal-fetal nanoparticle handling. 72–74

Integrative Perspective

These organ-specific safety domains are interconnected rather than isolated. 67 , 68 , 72–82 A nanoparticle that avoids renal clearance may increase hepatosplenic retention. 67 , 68 , 75–77 A formulation optimized for BBB crossing may alter cardiovascular or immune safety. 80–82 Pulmonary inflammation may be driven by systemic immune activation rather than only local deposition. 78 , 82 Reproductive risk may depend on the same determinants that shape chronic persistence in liver, spleen, or placenta. 72–74 Organ-specific safety in nanomedicine is therefore best approached as a systems problem linking biodistribution, barrier biology, degradation fate, disease state, and dosing schedule. 67 , 68 , 72–82 This is why organ-resolved toxicokinetics and long-term fate studies are indispensable for clinical translation. 67 , 68 , 72–82

Major Advances That Have Improved Nanoparticle Safety

The improvement of nanoparticle safety in medicine has not resulted from a single technological breakthrough, but from the progressive convergence of materials science, colloid chemistry, immunology, pharmaceutical engineering, and translational manufacturing. 6 , 42 , 83 Early generations of nanomedicines often focused primarily on payload loading, circulation time, and target accumulation. 6 , 83 Over time, however, it became clear that clinically meaningful progress required deeper control over immune recognition, blood compatibility, degradability, tissue persistence, and product reproducibility. 6 , 24 , 42 , 83–85 This shift has reshaped the field from one centered largely on delivery efficiency toward one increasingly guided by safe-by-design principles. 42 , 83 Contemporary advances that have materially improved nanoparticle safety include stealth strategies such as PEGylation, biomimetic coatings, biodegradable and bioresponsive materials, safer lipid nanoparticle architectures, surface engineering approaches to reduce immunotoxicity, controlled-release and site-specific delivery systems, and cleaner manufacturing approaches that reduce impurities and process-related hazards. 6 , 24 , 43 , 44 , 83–99

PEGylation and Stealth Strategies

PEGylation was one of the earliest and most influential advances in improving nanoparticle safety because it addressed a central problem in systemic nanomedicine: rapid recognition by proteins and phagocytic cells. 83 , 97 , 98 By grafting poly(ethylene glycol) chains onto nanoparticle surfaces, developers created sterically shielded interfaces that reduce protein adsorption, diminish opsonization, improve colloidal stability, and often prolong circulation time. 83 , 97 , 98 These effects can lower first-pass uptake by the mononuclear phagocyte system and reduce some acute blood-facing interactions, which in turn may improve tolerability and biodistribution for selected products. 83 , 97 , 98 Recent reviews continue to identify PEGylation as a foundational stealth strategy in nanomedicine and a major contributor to the clinical usability of liposomes, polymeric nanoparticles, and lipid nanoparticles. 83 , 97 , 98

The safety value of PEGylation lies not merely in longer circulation, but in the fact that prolonged circulation can sometimes be achieved while reducing direct membrane interactions and early immune recognition. 83 , 97 , 98 In practical terms, PEG chains create a hydrated steric barrier that helps prevent aggregation and moderates the immediate protein-layer composition formed in blood. 83 , 97 , 98 This can reduce nonspecific uptake and dampen some forms of acute reactivity. 83 , 97 , 98 For many formulations, PEGylation was therefore a decisive step in moving from highly interactive, rapidly cleared colloids toward more clinically tolerable systemic nanocarriers. 83 , 97 , 98

At the same time, the field has learned that PEGylation is an advance with limitations rather than a universal solution. 83 , 97 , 98 Anti-PEG antibodies, accelerated blood clearance, and PEG-associated hypersensitivity have become well-recognized concerns, particularly under repeated exposure. 83 , 97 , 98 This has not negated the importance of PEGylation, but it has refined its role. 83 , 97 , 98 The major safety advance has been twofold: first, PEGylation itself improved the safety profile of many early nanocarriers; second, recognition of the “PEG dilemma” drove the development of cleavable PEG, alternative stealth polymers, and more nuanced exposure strategies. 83 , 97 , 98

Biomimetic and Cell Membrane-Coated Nanoparticles

A major conceptual advance beyond classical stealth chemistry has been the emergence of biomimetic nanoparticles, especially cell membrane-coated systems. These platforms use membranes derived from erythrocytes, platelets, leukocytes, cancer cells, stem cells, bacteria, or hybrid membrane sources to cloak synthetic nanoparticle cores with biologically derived surface features. The safety rationale is compelling: instead of merely reducing interaction with the immune system through an inert polymer shell, biomimetic coatings attempt to present a more physiologically recognizable interface that can evade clearance, reduce nonspecific immune activation, and sometimes confer tissue-homing behavior. Recent reviews describe cell membrane-coated nanoparticles as a major frontier in biomimetic delivery precisely because they combine immune-evasive and targeting functions in a single surface design strategy. 43 , 44 , 86 , 87

From a safety perspective, one of the most important advantages of membrane-coated nanoparticles is their capacity to reduce abrupt blood-facing incompatibility. Erythrocyte membrane coatings, for example, may help prolong circulation and mitigate rapid phagocytic clearance, while platelet and leukocyte membrane systems can be designed to exploit native adhesion or inflammatory trafficking pathways without exposing the underlying synthetic core directly to plasma components. This type of camouflage may reduce complement-triggering surface motifs and reshape the protein corona toward a more biologically compatible identity. 43 , 44 , 86 , 87

The major safety advance here is not that biomimetic systems are automatically safe, but that they offer a more sophisticated way to engineer the nano–bio interface. By using natural membrane proteins and lipids, these systems may reduce some of the trade-offs associated with strongly synthetic stealth coatings. At the same time, they introduce new safety considerations such as membrane source selection, pathogen safety, batch variability, membrane integrity, immunologic compatibility, and manufacturing complexity. Thus, biomimetic coating represents an important advance because it shifts safety engineering from steric masking alone toward biologically informed interfacial design. 43 , 44 , 86 , 87

Biodegradable and Bioresponsive Materials

The move toward biodegradable nanomaterials has been one of the clearest long-term improvements in nanosafety. A major historical concern in nanomedicine has been chronic retention, especially in liver, spleen, and other phagocyte-rich tissues. Biodegradable platforms reduce this risk by allowing nanoparticles to break down into smaller, more excretable or metabolizable components after fulfilling their delivery function. Reviews on biodegradable nanomaterials and biomaterial degradation assessment emphasize that degradability is central to reducing long-term tissue burden and improving translational suitability. 84 , 88

This advance is especially important for polymeric and lipid-based systems. 84 , 88 , 89 Biodegradable polyesters, polypeptides, polycarbonates, natural polymers, and hydrolytically or enzymatically labile linkers have made it increasingly possible to design nanoparticles whose persistence is limited by intent rather than by accident. 84 , 88 In safety terms, this reduces the likelihood that repeated dosing will lead to progressive organ storage, lysosomal overload, or long-term inflammatory signaling. The improvement is not merely theoretical; it changes what regulators and developers can reasonably expect regarding chronic exposure and recovery after treatment cessation. 84 , 88

Bioresponsive materials build on this by coupling degradation or structural change to the disease microenvironment or intracellular conditions. pH-sensitive, redox-sensitive, enzyme-responsive, ROS-responsive, and externally triggered systems can be designed to remain relatively stable during circulation but disassemble or release payload only after reaching a target site or intracellular compartment. The safety benefit is obvious: reducing premature release and minimizing nonspecific exposure can lower systemic toxicity while preserving efficacy. Although bioresponsive systems vary widely in maturity, their key contribution to safety is the replacement of indiscriminate exposure with conditional activation. 88 , 94

Safer Lipid Nanoparticle Platforms

The recent evolution of lipid nanoparticles has been one of the most important safety advances in all of nanomedicine. 6 , 24 , 89 Modern lipid nanoparticle platforms, particularly those for nucleic acid delivery, differ substantially from earlier cationic liposome systems in that they increasingly rely on ionizable, rather than permanently cationic, lipids. This change has major safety implications. Ionizable lipids can remain close to neutral charge at physiological pH, reducing nonspecific blood and membrane interactions, while becoming protonated in acidic endosomal compartments to support intracellular release. Recent reviews emphasize that this transition has been central to the improved tolerability and clinical viability of LNP-mediated mRNA and siRNA delivery. 24 , 89

In addition to ionizable lipid design, safer LNP platforms have benefited from better control over helper lipid composition, cholesterol content, PEG-lipid type, particle size distribution, biodegradability of lipid tails and linkers, and analytical characterization of potency and reactogenicity. 6 , 24 , 89 These refinements have reduced some of the acute toxicity and inflammatory burden associated with earlier delivery systems while enabling the large-scale clinical success of mRNA vaccines and other nucleic acid medicines. 6 , 24 , 42 , 89 The safety advance here is particularly significant because it demonstrates that rational chemical redesign of a carrier class can turn a historically difficult modality into a clinically deployable one. 24 , 89

Current work on safer LNPs also reflects a broader shift in nanosafety philosophy. Rather than treating toxicity as an unavoidable trade-off of intracellular delivery, developers increasingly optimize lipid chemistry specifically to reduce reactogenicity, improve biodegradation, minimize repeat-dose immune complications, and direct delivery toward desired cell types more selectively. Thus, safer LNP platforms are both a specific technological advance and a model for how safety-guided carrier design can reshape an entire therapeutic class. 24 , 89

Surface Engineering to Reduce Immunotoxicity

Advanced surface engineering has become one of the most effective ways to reduce nanoparticle immunotoxicity because it directly governs how the material is first interpreted by blood proteins, immune cells, and clearance organs. Recent reviews on surface decoration and surface engineering emphasize improvements in colloidal stability, reduced toxicity, enhanced targeting, and better adaptation to complex biological environments. These strategies include PEGylation, zwitterionic coatings, polysaccharides, ligand-density optimization, charge modulation, biomimetic ligands, protein pre-coating, and hybrid surface architectures. 85 , 90 , 91

The safety benefit of surface engineering lies in its ability to reduce unfavorable early events such as aggregation, complement activation, nonspecific membrane disruption, and uncontrolled protein corona formation. For example, shifting a particle surface from strongly cationic toward near-neutral or zwitterionic character may reduce hemolysis and inflammatory signaling while preserving sufficient delivery function. Likewise, judicious ligand presentation can improve selectivity without exposing highly reactive surface motifs. In many cases, the greatest safety gain is not complete immune invisibility, which is often unrealistic, but a more controlled and less inflammatory pattern of immune engagement. 85 , 90 , 91

This area represents a major advance because it reframes nanoparticle surfaces as programmable biological interfaces rather than passive outer layers. Modern surface engineering seeks not only to enhance delivery, but also to shape the protein corona, reduce innate immune activation, and lower organ-specific off-target burden. In this respect, it has become one of the most powerful operational tools for implementing safe-by-design nanomedicine. 85 , 90 , 91

Controlled-Release and Site-Specific Delivery Approaches

A major source of nanoparticle-related toxicity is premature or nonspecific exposure of healthy tissues to either the carrier or the payload. Advances in controlled-release and site-specific delivery have improved safety by reducing this unwanted exposure. These approaches include diffusion-controlled matrices, triggerable release systems, stimuli-responsive nanocarriers, prodrug-like nanoconstructs, local depot systems, heat- or pH-triggered carriers, and active-targeting strategies that bias uptake toward particular tissues or cell populations. Reviews on targeted and site-specific nanocarrier systems consistently note that one of their principal safety advantages is the reduction of systemic toxicity through spatiotemporal control of drug exposure. 94–96

Controlled release improves safety because it can flatten harmful peak concentrations, prevent burst release, and keep free drug levels lower outside the intended site of action. Site-specific delivery adds another protective layer by limiting exposure in vulnerable organs. In oncology, for example, nanocarriers that localize drugs more effectively to tumors or release them preferentially within tumor-associated conditions can reduce damage to bone marrow, heart, or gastrointestinal tissues. In inflammatory or infectious disease, localized delivery may similarly decrease the systemic burden of potent agents. 94–96

Importantly, the safety gain from these approaches depends on actual delivery realism. The field has moved beyond simplistic assumptions that all nanoparticles automatically target diseased tissue. The major advance has been the creation of more conditional and mechanistically grounded release strategies, not merely more targeting labels. When successful, these systems reduce the mismatch between where a nanoparticle circulates and where its payload becomes biologically active. 94–96

Green Synthesis and Cleaner Manufacturing Approaches

Although green synthesis is often discussed mainly in environmental terms, it also has direct relevance to medical nanoparticle safety. Cleaner synthesis and manufacturing approaches aim to reduce toxic solvents, residual reagents, harsh reducing agents, heavy-metal contaminants, and unstable byproducts that can compromise the safety profile of a final nanomedicine. Recent reviews of green nanoparticle synthesis and eco-friendly bio-nanocomposites describe these approaches as environmentally responsible and potentially safer because biologically derived stabilizers and milder process conditions can reduce hazardous residues and improve biocompatibility-related attributes. 92 , 93

In pharmaceutical nanotechnology, the most important safety implication of cleaner manufacturing is not simply that the synthesis is “green,” but that it can produce materials with lower impurity burden, fewer toxic residuals, and potentially more benign surface chemistries. 6 , 92 , 93 This is particularly relevant for metallic and inorganic nanoparticles, where reductants, surfactants, and postsynthetic modifiers can strongly influence toxicity if not adequately removed. 92 , 93 Cleaner manufacturing can therefore support both environmental sustainability and better clinical-grade quality. 6 , 92 , 93

However, this area also illustrates the need for translational discipline. 6 , 92 , 93 Green synthesis does not automatically guarantee batch reproducibility, pharmaceutical purity, or regulatory readiness. The true advance lies in integrating cleaner chemistries with robust process control, scalable purification, and analytical verification. When those conditions are met, greener manufacturing becomes part of a broader movement toward safer nanomedicine production rather than a separate sustainability agenda. 6 , 42 , 92 , 93

Integrated Perspective

The evidentiary maturity of safety-enhancing strategies is not equivalent. PEGylation and ionizable-lipid formulations have extensive clinical validation but also well-defined liabilities, including anti-PEG immunity, infusion reactions, and repeat-dose effects. Biodegradable materials and controlled-release approaches have platform-specific clinical precedent, whereas cell-membrane coatings, green-synthesis strategies, and highly responsive or transient constructs remain largely preclinical and require stronger comparative toxicology, CMC standardization, and scale-up evidence before generalized claims of improved safety are justified. 6 , 42–44 , 83–99

Taken together, these advances show that nanoparticle safety has improved most when the field has intervened at the interface between material and biology. 6 , 24 , 42–44 , 83–98 PEGylation reduced rapid immune recognition. 83 , 97 , 98 Biomimetic coatings made that interface more physiologic. 43 , 44 , 86 , 87 Biodegradable and bioresponsive materials reduced long-term burden and premature exposure. 84 , 88 , 94 Safer LNP chemistry improved intracellular delivery without requiring permanently aggressive surface charge. 24 , 89 Surface engineering reduced immunotoxicity more broadly. 85 , 90 , 91 Controlled-release and site-specific systems reduced off-target exposure. 94–96 Cleaner manufacturing reduced impurities and process-related hazards. 6 , 42 , 92 , 93 These are not isolated innovations. They are mutually reinforcing developments that have shifted nanomedicine toward more predictable, biologically informed, and clinically defensible design. 6 , 24 , 42–44 , 83–98

The most important lesson is that safety improvements have rarely come from making nanoparticles completely inert. 42 , 83–85 Instead, progress has come from controlling how nanoparticles interact with blood, immune systems, tissues, and manufacturing environments. 6 , 24 , 42–44 , 83–98 In this sense, the major advances described in this section represent the maturation of nanomedicine from formulation ingenuity to translational engineering. 6 , 42 , 83 The major technological and conceptual advances that have improved nanoparticle tolerability and translational safety are outlined in Table 2 .

Progress in Preclinical Safety Assessment

Preclinical safety assessment in nanomedicine has advanced substantially over the past decade, moving from relatively simple viability assays and gross organ histology toward a more integrated framework that combines physicochemical characterization, mechanistic cell biology, biodistribution analysis, imaging-based tracking, computational modeling, and systems-level toxicology. This evolution reflects a central lesson from translational nanomedicine: conventional toxicology workflows developed for small molecules are often insufficient for nanoparticles because nanoscale materials interact dynamically with proteins, cells, blood components, tissues, and clearance systems in ways that depend on size, surface chemistry, aggregation state, and exposure context. Modern preclinical assessment therefore seeks not only to determine whether a nanoparticle is toxic, but also to understand where it goes, how it transforms, which cells it perturbs, and which molecular pathways link early biological interaction to later adverse outcomes. 71 , 100–102

A second major change is that preclinical safety is increasingly viewed as a tiered evidence system rather than a single experimental domain. In vitro assays remain indispensable for throughput and mechanism discovery, but they are now complemented by 3D cultures, organoids, organ-on-chip systems, advanced in vivo biodistribution studies, physiologically based pharmacokinetic modeling, and multi-omics methods that can reveal early molecular signatures of toxicity. The field has therefore progressed not by abandoning traditional methods, but by embedding them within a more mechanistically informed and translationally realistic safety-testing architecture. 68 , 71 , 100–103

In vitro Models: Strengths and Limitations

Conventional in vitro models remain the foundation of early nanosafety screening because they are rapid, comparatively inexpensive, experimentally controlled, and compatible with systematic variation of nanoparticle properties such as size, charge, coating, and dose. They are especially useful for identifying acute cellular effects including cytotoxicity, oxidative stress, membrane damage, inflammatory signaling, barrier perturbation, endocytic uptake, and assay-specific interactions with blood or immune cells. 71 , 100 Their greatest strength is mechanistic tractability. A well-designed cell-based study can isolate how a specific formulation variable changes cellular uptake, lysosomal burden, mitochondrial function, or cytokine release without the confounding complexity of whole-animal physiology. 71 , 100

The in vitro field has also become more sophisticated in how exposure is modeled. For airborne and inhaled nanomaterials, for example, recent reviews have emphasized the importance of exposure configuration, comparing submerged exposure with air–liquid interface systems and static versus dynamic exposure modes. Jaber et al noted that modern in vitro approaches now encompass primary cells, cell lines, organoids, spheroids, reconstituted tissues, and organ-on-chip systems, together with intermittent, continuous, static, or dynamic exposure methods. 101 This is an important advance because nanoparticle dose delivery to cells is highly sensitive to sedimentation, diffusion, agglomeration, protein content, and exposure geometry. Preclinical assessment has therefore progressed from asking only “what concentration was added?” to asking “what biologically effective dose actually reached the cells?”. 71 , 101

Despite these strengths, the limitations of conventional in vitro models remain substantial. Monocultures often fail to reproduce tissue architecture, multicellular signaling, flow-dependent transport, macrophage surveillance, organ-level metabolism, and long-term fate. Many standard toxicity assays are also vulnerable to nanoparticle interference, particularly optical readouts, dye-binding assays, and colorimetric metabolic assays. Earlier methodological work and subsequent reviews have repeatedly cautioned that nanomaterials may absorb, quench, catalyze, scatter, or otherwise perturb assay reagents, leading to false-positive or false-negative toxicity findings if assay compatibility is not verified. 71 , 100 Furthermore, cellular uptake and toxicity in vitro may not predict in vivo disposition if protein corona formation, tissue barriers, blood interactions, or clearance pathways differ substantially from the cell-culture environment. 71 , 100 , 101

Thus, the progress in this area is not that traditional in vitro systems have become sufficient on their own. Rather, the advance lies in understanding more clearly what they can and cannot answer. They are now used more deliberately for screening, rank-ordering, mechanistic hypothesis generation, and safe-by-design optimization, while more complex models are used to address questions of tissue realism and translational predictivity. 71 , 100 , 101

3D Models, Organoids, and Organ-on-Chip Systems

One of the most important developments in preclinical nanosafety assessment has been the expansion of 3D in vitro models, including spheroids, organoids, reconstructed tissues, and organ-on-chip platforms. These systems were developed to address the poor physiological fidelity of standard monolayer cultures. By incorporating multicellular architecture, extracellular matrix, gradients of oxygen and nutrients, disease-relevant microenvironments, and in some cases microfluidic flow, they offer a more realistic framework for evaluating nanoparticle penetration, retention, toxicity, and target selectivity. 101 , 102

The safety significance of these models is considerable. Three-dimensional systems can reveal diffusion limitations, barrier effects, and cell-type-specific toxicities that remain hidden in 2D assays. In tumor spheroids and organoids, for example, nanoparticles may show strong uptake in peripheral cell layers yet poor penetration into the inner tissue, producing very different efficacy–toxicity tradeoffs than those suggested by monolayer culture. Likewise, liver, kidney, placental, gut, lung, and blood–brain barrier organoids can provide more realistic readouts of tissue stress, barrier disruption, and intracellular accumulation than isolated cell lines. Jaber et al explicitly identified organoids, spheroids, reconstituted tissues, and organs-on-chip as part of the modern in vitro toolkit for nanotoxicity characterization. 101

Among advanced in vitro platforms, organ-on-chip systems have generated especially strong interest because they combine tissue architecture with controlled flow and, in some designs, immune, endothelial, or multicompartment interactions. A 2025 review by Sampaio et al concluded that organ-on-chip platforms provide a powerful alternative to traditional in vitro and in vivo models for evaluating nanoparticle toxicity and therapeutic efficacy, helping to identify negative effects and mechanisms of action under more physiologic conditions, while also noting important barriers related to scalability and regulatory integration. 102 This is a meaningful advance for safety science because many nanoparticle toxicities are flow-dependent, barrier-dependent, or multicellular in nature. Endothelialized chips, vascularized liver chips, renal tubule chips, placental chips, and BBB chips can therefore probe safety domains that conventional static cultures capture poorly. 102

Nevertheless, these systems also have limitations. They are more complex, less standardized, lower throughput, and often more difficult to characterize analytically than standard cultures. Their biological relevance depends strongly on cell source, matrix composition, fluid shear, and device architecture, and interlaboratory reproducibility remains a challenge. Regulatory acceptance is still evolving. Thus, the progress here lies not in replacing all other models, but in strategically deploying 3D and microphysiological systems when the safety question specifically depends on barrier transport, multicellular interaction, or tissue architecture. 101 , 102

In vivo Models and Biodistribution Studies

In vivo studies remain indispensable in nanomedicine safety assessment because only whole-animal models can integrate circulation, protein corona evolution, immune surveillance, organ-level sequestration, metabolism, degradation, repeated-dose adaptation, and long-term clearance. Even the most sophisticated in vitro systems cannot fully reproduce the dynamic redistribution of nanoparticles across liver, spleen, kidney, lung, blood, and target tissues, nor can they capture the full coupling between biodistribution and toxicity. 68 , 103 , 104

A major advance in the past few years has been the more quantitative treatment of in vivo transport and tissue distribution. Xu et al reviewed size-dependent in vivo transport of nanoparticles and emphasized how size influences delivery, targeting, and clearance. 103 Related work on nanoparticle biodistribution coefficients has attempted to quantify tissue distribution across studies and platforms, highlighting how strongly organ exposure depends on formulation variables and how consistently liver and spleen dominate systemic disposition in many intravenous settings. 68 This quantitative turn matters for safety because organ toxicity is fundamentally exposure-driven. Better biodistribution analysis improves the ability to predict which tissues experience the highest cumulative nanoparticle burden and which design changes are most likely to reduce off-target risk. 68 , 103 The scale of the available evidence is increasingly sufficient for cross-platform quantitative synthesis. Kumar et al analyzed 2,018 published pharmacokinetic and biodistribution datasets spanning graphene oxide, lipid, polymeric, silica, iron-oxide, and gold nanoparticles across plasma, tumor, and 13 tissues, demonstrating both strong tissue-dependent partitioning and substantial formulation-dependent variability. 68 Such datasets are valuable because they move nanosafety away from isolated organ-percentage observations toward comparative exposure metrics that can support PBPK modeling, rational dose selection, and identification of organs in which cumulative burden is most likely to become safety-limiting.

Another important development is the closer integration of biodistribution data with pharmacokinetic modeling. A 2024 ACS Pharmacology and Translational Science review highlighted recent advances in physiologically based pharmacokinetic (PBPK) modeling for nanoparticles, emphasizing that PBPK frameworks can help simulate absorption, distribution, metabolism, and elimination for nanomaterials whose disposition differs substantially from that of conventional small molecules. 104 In translational safety terms, PBPK modeling is valuable because it can connect animal biodistribution data, route-specific exposure, and organ kinetics in ways that improve dose selection and interspecies interpretation. 104

At the same time, the limitations of in vivo models remain well recognized. Species differences in protein composition, immune reactivity, phagocyte function, tumor microenvironment, vascular permeability, and clearance pathways can make animal data difficult to extrapolate directly to humans. Rodent models often overestimate delivery performance and may underestimate infusion-related or repeat-dose liabilities. The progress in this domain, therefore, is not that animal studies have become fully predictive, but that they are increasingly interpreted together with quantitative imaging, PBPK modeling, and mechanistic biomarkers rather than as stand-alone evidence. 68 , 103 , 104

Radiolabeling and Imaging-Based Tracking of Nanoparticles

One of the most consequential advances in preclinical nanosafety has been the use of radiolabeling and imaging-based tracking to determine where nanoparticles go in vivo and how long they persist there. Traditional biodistribution methods based on tissue digestion and elemental analysis remain important, but they provide limited spatial and temporal resolution. By contrast, radiolabeling and multimodal imaging permit longitudinal, often quantitative assessment of nanoparticle localization, clearance, target uptake, and off-target retention. 28 , 105

This progress is especially visible in radionuclide-based nanoparticle studies. Reilly et al described radiation nanomedicines as nanoparticles labeled with radionuclides that emit therapeutic radiation and reviewed the broader landscape of such systems, while also underscoring the need to understand local versus systemic distribution and the safety consequences of radionuclide release or off-target retention. 105 More broadly, recent reviews of radiolabeled nanoparticles for multimodality imaging and theranostic applications emphasize that PET, SPECT, CT, MRI, fluorescence, and hybrid approaches can be used not only to guide therapy but also to resolve key safety questions such as RES uptake, renal clearance, tumor-to-organ ratios, and long-term retention. 28

The safety value of imaging is straightforward but profound. Nanomedicine toxicology depends on exposure, and exposure is difficult to interpret without tracking the intact carrier, not just the released drug or a terminal elemental signal. Imaging-based approaches can reveal whether a formulation remains stable in circulation, whether a label dissociates, whether liver and spleen burden declines over time, and whether repeated administration changes organ uptake patterns. These methods therefore strengthen the causal link between formulation design, biodistribution, and organ-specific safety findings. 28 , 105

Limitations remain. Labels may detach or redistribute independently of the nanoparticle, radiochemistry can alter formulation behavior, and not all imaging modalities quantify the same biological entity with equal fidelity. Nonetheless, radiolabeling and advanced imaging have clearly improved preclinical safety assessment by making nanoparticle fate visible, serially measurable, and more directly linked to toxicologic interpretation. 28 , 105

Multi-Omics and Mechanistic Toxicology Approaches

A major conceptual advance in nanosafety is the shift from endpoint-only toxicology toward mechanistic toxicology, increasingly supported by transcriptomics, proteomics, metabolomics, lipidomics, and integrated multi-omics analysis. Traditional assays often capture late manifestations of injury such as cell death, enzyme leakage, or histologic damage. Omics approaches, by contrast, can detect early pathway perturbations and reveal coordinated changes in stress response, inflammation, mitochondrial function, metabolism, DNA damage, proteostasis, and immune signaling long before overt toxicity becomes visible. 106–108

A 2023 review in Advanced Drug Delivery Reviews specifically discussed transcriptomics, proteomics, metabolomics, lipidomics, and multi-omics in the assessment of biological responses to nanoparticles, emphasizing their value for characterizing whole-system responses rather than isolated endpoints. 106 More recently, Li et al reviewed omics approaches for assessing toxicity and biological interactions of inorganic nanoparticles, arguing that these methods improve mechanistic understanding and support safer nanoparticle design. 107 Complementing this, a 2024 review on bioinformatics and machine learning in nanotoxicology highlighted how omics data can support grouping, read-across, and mode-of-action analysis for nanomaterials, helping move the field away from one-nanomaterial–one-endpoint logic toward more predictive hazard frameworks. 108

The translational importance of multi-omics lies in its ability to connect formulation properties to biologic mechanisms. For example, two nanoparticles may appear similarly cytotoxic at high dose, yet omics analysis may show that one primarily induces oxidative stress and mitochondrial dysfunction while the other activates inflammatory and DNA-damage pathways. This distinction matters for safe-by-design redesign, chronic-risk prediction, and the interpretation of susceptible populations. Omics also helps bridge in vitro and in vivo toxicology by identifying pathway-level signatures that can be compared across models and, eventually, linked to clinical biomarkers. 106–108

At the same time, multi-omics introduces challenges of data complexity, batch effects, platform comparability, and biological interpretation. Large datasets are only useful if integrated with careful nanoparticle characterization, exposure dosimetry, and orthogonal validation. The field’s progress, therefore, is not simply the accumulation of more molecular data, but the increasing use of omics as part of a mechanistically anchored toxicology strategy. 106–108

Integrated Perspective

Taken together, progress in preclinical safety assessment has been driven by a move from isolated assays to integrated evidence chains. Conventional in vitro systems remain essential for throughput and hypothesis generation. Three-dimensional models and organ-on-chip systems improve tissue relevance. In vivo studies and PBPK modeling define whole-body exposure. Radiolabeling and imaging visualize nanoparticle fate directly. Multi-omics identifies molecular mechanisms and early hazard signatures. The strongest current preclinical programs are those that connect these layers rather than treating them as substitutes for one another. 28 , 68 , 71 , 100–108

This integrated approach has already improved nanosafety science, but the remaining challenge is standardization and translation. Preclinical models still vary widely in exposure conditions, assay compatibility, imaging strategy, and omics interpretation. Yet the field is clearly moving toward a more rigorous and predictive framework in which mechanistic insight, biodistribution realism, and advanced model systems jointly inform the safety profile of nanoparticle medicines before they enter the clinic. 101 , 102 , 104 , 106–108

The principal in vitro and in vivo approaches currently used for nanoparticle safety assessment, together with their major strengths and limitations, are summarized in Table 3 .

Analytical and Methodological Challenges

Analytical and methodological limitations remain among the most persistent barriers to reliable nanoparticle safety assessment and, more broadly, to the clinical translation of nanomedicine. 11 , 46 , 71 , 111 , 112 Unlike conventional small-molecule drugs, nanoparticle products are defined not only by chemical identity but also by multivariate physicochemical attributes such as size distribution, shape, surface chemistry, aggregation state, cargo localization, impurity burden, and biological transformation. 46 , 111 As a result, analytical uncertainty does not simply reduce scientific precision; it can change the interpretation of efficacy, toxicity, comparability, and product quality. 46 , 111 , 112 Recent standards-oriented and translational reviews have argued that the absence of harmonized characterization workflows, incomplete reporting of critical quality attributes, and inconsistent experimental design continue to undermine the reliability of the nanomedicine literature. 11 , 46 , 111 , 112

A central consequence of this problem is that apparently similar nanoparticles are often not truly comparable across laboratories, studies, or manufacturing runs. 46 , 54 , 111 , 112 Differences in raw materials, synthesis route, purification, storage, dispersion protocol, protein-containing media, and dose normalization can substantially alter biological behavior. 46 , 54 , 71 , 111 In nanomedicine, therefore, methodological rigor is not a secondary issue. It is part of the safety profile itself, because toxicology findings are only as meaningful as the material definition and exposure conditions on which they are based. 11 , 46 , 71 , 111 , 112

Lack of Standardized Characterization Protocols

One of the most important challenges in nanosafety is the lack of universally implemented characterization protocols that are sufficiently robust for medical translation. 46 , 111 Nanoparticles are often described using a limited set of measurements, typically mean hydrodynamic diameter, polydispersity index, and zeta potential, but these descriptors rarely capture the full set of properties that govern biological interaction. 46 , 111 Reviews focused on analytical standardization have emphasized that reproducible nanomedicine development requires broader and context-aware characterization, including morphology, composition, aggregation behavior, surface chemistry, protein corona susceptibility, drug loading, release behavior, and stability before, during, and after biological testing. 46 , 111

Recent standardization-focused analyses have argued that the lack of a universal reporting system and decentralized characterization methodologies have contributed to low reliability and reproducibility in the nanomedicine literature, and specifically called for use of existing standards tailored to intended medical use. 111 Similarly, work on critical quality attributes has framed them as essential to therapeutic nanomedicine development, emphasizing that inadequate analytical definition compromises comparability and translation. 46 These observations reflect a broader field-wide transition: characterization is no longer viewed as a descriptive appendix, but as a foundational element of product identity and safety interpretation. 46 , 111

This challenge is amplified by the fact that the “relevant” characterization parameters depend on the product class and test context. For example, the key analytical variables for a radiolabeled liposome, an ionizable lipid nanoparticle, and a mesoporous silica carrier are not identical. Standardization in nanomedicine therefore cannot mean one inflexible assay panel for all products. Instead, it requires harmonized frameworks that define which measurements are essential, how they should be performed, and under which biologically relevant conditions they should be reported. 46 , 111

Batch-to-Batch Variability

Batch-to-batch variability remains a major obstacle because nanoparticle systems are highly sensitive to small changes in synthesis and processing. Parameters such as precursor purity, mixing rate, solvent removal, temperature, scale-up conditions, and postformulation handling can change size distribution, surface properties, agglomeration tendency, free payload fraction, and impurity levels. These changes may be modest analytically yet significant biologically. 46 , 54 A multicenter study spanning 14 European laboratories and 46 batches of selected nanomaterials highlighted how substantial batch-dependent variation can be even for materials that are nominally the same. 54

From a safety perspective, batch variability is especially problematic because toxicologic effects may be driven by minor subpopulations rather than by the mean formulation. A small increase in large aggregates, free drug, endotoxin contamination, or reactive surface impurities may alter blood compatibility, biodistribution, or inflammatory response without being obvious from a simplified certificate of analysis. This is why batch consistency in nanomedicine is not merely a manufacturing issue. It is a toxicology issue and, ultimately, a clinical safety issue. 46 , 54

The translational literature has increasingly linked batch variability to the need for better control of critical quality attributes and for orthogonal analytical methods capable of detecting subtle but clinically meaningful drift. 46 , 54 , 111 Without such control, safety findings from one batch may not be transferable to later preclinical or clinical material. 46 , 54 In this sense, batch-to-batch variability directly undermines both internal validity and regulatory confidence. 46 , 54 , 111

Interference of Nanoparticles with Toxicological Assays

Nanoparticle interference with toxicological assays is one of the most widely recognized methodological problems in nanotoxicology. 71 , 109 , 110 Because nanoparticles can absorb, scatter, quench, fluoresce, catalyze redox reactions, bind assay reagents, or sediment onto cells in nonuniform ways, they can distort results from common viability, oxidative stress, fluorescence, enzyme activity, and colorimetric assays. Classical work demonstrated that engineered nanoparticles can interfere with in vitro toxicity assays in concentration-, particle-, and assay-specific ways, while later studies showed that assay interference is widespread enough to threaten the accuracy of nanotoxicity testing more generally. 109 , 110

This challenge has practical consequences. A nanoparticle may appear cytotoxic because it quenches a fluorescent readout, or appear innocuous because it binds or inactivates an assay reagent. Similarly, assay artifacts may vary between formulations that differ only slightly in size or coating, leading investigators to misattribute methodological noise to biologic mechanism. Reviews and commentaries on assay interference have therefore stressed the need for interference controls, orthogonal assays, cell-free control experiments, and validation of readout compatibility before drawing toxicologic conclusions. 109 , 110

This issue also helps explain why apparent contradictions persist across the nanoparticle safety literature. Some inconsistencies likely reflect genuine biological differences, but others arise because different laboratories use different assays, media conditions, dispersion methods, and control structures. Progress in nanosafety therefore depends not only on inventing better assays, but also on recognizing when existing assays are being used outside their reliable operating range for particulate materials. 71 , 109 , 110

Dose Metrics: Mass, Particle Number, and Surface Area

Dose selection and dose normalization are unusually difficult in nanoparticle toxicology because mass concentration alone often fails to capture the biologically relevant exposure. 71 , 113 , 114 Nanoparticles of the same mass can differ profoundly in particle number, total surface area, and interface reactivity, which means that equivalent mass doses may not correspond to equivalent biological burden. Foundational work argued that surface area is often a more appropriate dose metric than particle mass or particle number for nanoparticle dose–response assessment, and subsequent studies showed that surface area explained acute pulmonary toxicity better than the other common metrics for spherical nanoparticles in the systems analyzed. 113 , 114

This does not mean that surface area is always the correct metric. Rather, the relevant metric depends on mechanism. Surface area may better reflect oxidative or interfacial injury, particle number may be more relevant for certain uptake-driven or fiber-like effects, and mass may remain useful in pharmacologic contexts where payload quantity dominates. The methodological challenge is that many studies still report only mass concentration, making cross-study interpretation difficult and sometimes misleading. 71 , 113 , 114 Reviews on in vitro nanomaterial hazard assessment have specifically recommended the use of comparable and realistic dose metrics and test conditions. 71

This issue becomes even more important when translating between in vitro and in vivo studies. Administered dose, delivered cellular dose, retained tissue dose, and biologically effective dose are not interchangeable. 71 , 113 , 114 Sedimentation, agglomeration, and protein adsorption can all change the actual dose that cells experience. 71 , 109 , 110 As a result, rigorous nanosafety studies increasingly distinguish nominal concentration from effective exposure and justify the selected metric in relation to the hypothesized mechanism of injury. 71 , 113 , 114

Reproducibility and Comparability Across Studies

Reproducibility remains a field-wide challenge in nanomedicine, extending beyond isolated assay issues to encompass material reporting, experimental design, data analysis, and model selection. 11 , 71 , 111 , 112 In a widely cited perspective, it was argued that transparency and reproducibility are central unresolved issues in nanomedicine and that better reporting, deeper characterization, and clearer justification of biological models are urgently needed. This concern was not merely about academic rigor. It was that weak reproducibility directly impairs downstream translation by making it difficult to distinguish robust platform effects from laboratory-specific artifacts. 112

Comparability across studies is especially difficult because nanomedicine experiments often differ in dispersion protocol, serum content, exposure duration, cell type, endpoint selection, and characterization depth. 11 , 71 , 111 , 112 Even when the same nanoparticle name is used, the tested materials may not be analytically equivalent. 11 , 46 , 54 , 111 , 112 Recent multicenter work on protein corona characterization showed that standardization measurably improved data homogeneity across facilities, illustrating that at least part of the reproducibility problem can be reduced through harmonized methodology. This finding is highly relevant beyond corona research, because it suggests that systematic protocol alignment can improve comparability in other nanosafety domains as well. 11

The broader implication is that reproducibility in nanomedicine is not achieved only by repeating experiments more often. It requires standardized material description, predefined analytical criteria, validated assay controls, appropriate dose metrics, and transparent reporting of all conditions that could alter nano–bio interactions. Without that infrastructure, the literature remains difficult to synthesize, and safety conclusions remain less portable than they need to be for regulatory or clinical use. 11 , 71 , 111 , 112

Integrated Perspective

The analytical and methodological challenges described in this section are tightly interconnected. 11 , 46 , 54 , 71 , 109–114 Lack of standardized characterization makes batch variability harder to detect. 46 , 54 , 111 Batch variability complicates assay interpretation. 54 , 109 , 110 Assay interference obscures true dose–response relationships. 109 , 110 Inadequate dose metrics hinder comparison across studies. 71 , 113 , 114 All of these factors together erode reproducibility and weaken the evidence base for safe-by-design development. 11 , 71 , 111 , 112

For translational nanomedicine, the central lesson is that methodological rigor is not ancillary to safety. It is part of safety. 11 , 46 , 71 , 111 , 112 A nanoparticle cannot be judged safe, unsafe, or equivalent unless it is sufficiently characterized, reproducibly manufactured, appropriately dosed, and evaluated with assays that the material does not invalidate. 11 , 46 , 54 , 71 , 109–114 The more complex the nanomedicine, the more critical this principle becomes. 11 , 46 , 111 , 112

Regulatory Progress and Translational Considerations

Regulatory progress in nanomedicine has been substantial, but it has not taken the form of a single unified “nanomedicine law”. 6 , 18 , 19 , 42 , 115 Instead, both the United States and the European Union continue to regulate nanomedicines under existing medicinal-product frameworks, supplemented by product-specific guidance, reflection papers, quality expectations, and case-by-case scientific advice. 18 , 19 The FDA guidance for drug products containing nanomaterials explicitly states that the agency does not regard nanomaterial-containing products as inherently benign or harmful; rather, regulation is based on the specific characteristics of the product and its safety and effectiveness for the intended use. 18 In parallel, EMA’s 2025 horizon-scanning report states that there is still no dedicated EU legal framework for nanomedicines, and that these products are regulated under existing medicines legislation, complemented by EMA reflection papers and scientific-regulatory support mechanisms. 19

This regulatory position is important because it reflects both progress and constraint. On the one hand, nanomedicines are no longer treated as regulatory outliers; there is now accumulated experience with liposomes, iron-carbohydrate colloids, albumin-bound nanoparticles, lipid nanoparticle vaccines, micelles, and related complex products. On the other hand, the absence of a globally harmonized, nanomedicine-specific framework means that developers still face uncertainty regarding product classification, critical quality attributes, comparability requirements, and the level of evidence needed to support manufacturing changes or follow-on products. 6 , 18 , 19 , 42 , 115 EMA’s recent report identifies quality and manufacturing as key challenge areas, particularly robust characterization of particle properties, critical quality attributes, and control of scale-up, stability, and impurities. 19

Current Regulatory Expectations for Nanomedicines

Current regulatory expectations increasingly center on a risk-based, product-specific approach. 6 , 18 , 19 , 42 The FDA guidance for drug products containing nanomaterials includes specific recommendations spanning pharmaceutical quality/CMC, nonclinical studies, bioanalytical considerations, in vitro tests with human biomaterials, immunogenicity, and postmarket submissions. 18 This is an important sign of progress: nanoparticle-containing products are now expected to be supported by integrated evidence linking material properties, manufacturing controls, biological interaction, and clinical relevance rather than by generic claims of “nanotechnology-enabled” benefit. 6 , 18 , 42

In practical terms, regulators now expect developers to define the nanomaterial within the finished product, justify which physicochemical properties are critical to performance, and show how those properties relate to safety, identity, strength, quality, and purity. 6 , 18 , 19 , 42 FDA specifically highlights the need to characterize attributes such as particle size distribution, structural features, surface properties, in vitro release, impurities, sterility, endotoxin levels, and pyrogenicity. 18 This expectation reflects a major maturation of the field: nanoparticle products are no longer viewed simply as formulations with a small size range, but as complex products whose clinically relevant behavior depends on multivariate material attributes. 6 , 18 , 19 , 42

The EU position is broadly similar, though expressed through a combination of legislation, reflection papers, and scientific advice pathways rather than one single nanomedicine guidance document. 19 , 42 EMA’s 2025 report emphasizes that nanomedicines are regulated under existing medicinal product rules, with support available through the Innovation Task Force, Scientific Advice Working Party, Quality Innovation Group, PRIME, and international collaborations such as the IPRP Nanomedicines Working Group. 19 This indicates meaningful institutional progress: although nanomedicine-specific legislation is lacking, regulatory infrastructure for early dialogue and product-specific assessment has clearly expanded. 19 , 42

Safety-by-Design Approaches

One of the most important translational developments is the growing regulatory relevance of safety-by-design. 18 , 42 , 115 Historically, safety was often treated as a downstream hurdle that followed proof-of-concept efficacy. 42 , 115 In contemporary nanomedicine, this sequence is increasingly viewed as inadequate, because the same properties that drive efficacy also shape blood compatibility, immune recognition, biodistribution, degradation, and chronic retention. 18 , 42 , 115 Recent safe(r)-by-design guidance efforts for the nanotechnology industry reflect this shift by advocating earlier and more systematic incorporation of hazard reduction into material selection, product design, and lifecycle decision-making. 115

Regulatory-science literature in nanomedicine now aligns with this philosophy. 6 , 18 , 42 , 115 The FDA guidance repeatedly ties product quality to safety and effectiveness, while industry-focused reviews emphasize that nanomedicine development requires early identification of critical quality attributes and risk factors that may affect clinical performance. 6 , 18 , 42 In other words, safety-by-design in nanomedicine is no longer just an academic slogan; it is increasingly consistent with what regulators and manufacturers need in order to justify translation. 6 , 18 , 42 , 115

The practical meaning of safety-by-design is not complete elimination of biological interaction, which is often impossible and sometimes undesirable, but rational control of that interaction. This includes selecting degradable materials where long-term retention is a concern, minimizing complement-triggering surfaces, reducing free-drug leakage, designing appropriate release kinetics, and incorporating in vitro and in vivo assays that interrogate likely failure modes early enough to influence formulation decisions. 18 , 42 , 115 That perspective is now central to clinically serious nanomedicine development. 6 , 18 , 42 , 115 Safety-by-design also has practical industrial limits that should be acknowledged. Nanomedicine optimization is intrinsically multi-objective: reducing one liability can worsen another, as illustrated by strategies that prolong circulation but increase immune recognition after repeated exposure, or by cationic designs that improve endosomal escape while increasing membrane toxicity. In addition, formulation changes made during scale-up can shift critical quality attributes enough to invalidate an apparently optimized preclinical safety profile. For industry, safety-by-design is therefore most useful when it is implemented as a risk-based, iterative process linked to manufacturability, analytical comparability, and clinically meaningful performance rather than as an expectation that every theoretical hazard can be eliminated before development advances. 6 , 18 , 42 , 46 , 115

Chemistry, Manufacturing, and Controls (CMC) Challenges

CMC remains one of the most difficult and decisive areas in nanomedicine translation. 6 , 18 , 19 , 42 , 115 , 116 The FDA guidance explicitly devotes major sections to nanomaterial quality attributes, structural characterization, physicochemical characterization methods, dissolution or release testing, manufacturing process and in-process controls, excipient function and safety, stability, and postmarket CMC changes. 18 This structure alone reflects how heavily nanomedicine regulation now depends on quality science. 6 , 18 , 19 , 42

The core challenge is that nanomedicine quality is inherently multidimensional. 6 , 18 , 42 FDA notes that nanomaterial-containing products can be highly sensitive to process conditions and production scale, and that robust control strategies and process validation are essential because variability may be unpredictable and scale effects may be unknown. 18 The same guidance also notes that different shape or size distributions can become batch-consistency issues if they affect quality, safety, or efficacy, and that empty carriers or incompletely coated carriers may constitute impurities that need quantification. 18 These are unusually demanding expectations compared with many conventional products, but they are appropriate because nanomedicine performance is often tightly coupled to subtle structural heterogeneity. 6 , 18 , 42

Recent industry and regulatory reviews echo this point. 6 , 42 The 2024 industry perspective in Journal of Pharmaceutical Sciences emphasizes the current complexity of nanomedicine drug products and the importance of deep analytical characterization and platform-specific understanding. 6 Similarly, regulatory reviews highlight the complexity of concerns for nanomedicines and gene-vaccine products, including product characterization and quality control. 42 Together, these sources show that CMC is not peripheral to nanomedicine translation; it is one of the principal determinants of whether preclinical and clinical safety evidence remains interpretable across development stages. 6 , 18 , 42

Scale-up, Sterilization, Storage, and Stability

Scale-up remains a major translational bottleneck because nanoparticle properties are often process-dependent. 6 , 18 , 19 , 116 Mixing rate, shear environment, solvent removal, temperature, filtration, and fill-finish steps can all change size distribution, colloidal stability, surface presentation, free payload fraction, and impurity burden. 6 , 18 , 116 FDA explicitly states that changes in analytical methods, manufacturing process, scale, and manufacturing sites may make bridging from early development lots to commercial-scale lots difficult, and recommends retaining sufficient material from all batches to allow future analysis with updated or complementary methods. 18

Sterilization and microbial control are also unusually important in nanomedicine because the product itself may be sensitive to processes that would be routine for conventional dosage forms. 6 , 18 , 19 FDA specifically identifies sterility, endotoxin levels, and pyrogenicity among relevant quality attributes for nanomaterial-containing drug products. 18 This matters because sterilization methods such as filtration, heat, radiation, or aseptic processing may alter nanoparticle size, aggregation behavior, surface chemistry, or cargo integrity. Thus, for nanomedicines, sterilization is not merely a terminal processing requirement; it is a potential source of product transformation that must be built into formulation and process design. 6 , 18 , 19

Storage and stability are likewise more complex than in conventional small-molecule formulations. 6 , 18 , 19 , 116 FDA states that container closure suitability, storage conditions, shelf life, and in-use conditions must be supported by both chemical and physical stability data, and specifically highlights risks such as changes in size and size distribution, morphology, aggregation, surface charge, release rate, drug leakage, ligand loss, container interactions, and reconstitution behavior. 18 Recent review literature reinforces this point, emphasizing that nanocarrier stability directly affects biocompatibility, targeted delivery, therapeutic efficacy, and shelf life. 6 , 116 The translational implication is clear: stability is not just a logistics issue; it is a safety and performance issue. 6 , 18 , 19 , 116

Post-Marketing Surveillance Considerations

Post-marketing surveillance is especially important for nanomedicines because pre-approval studies may not fully capture repeat-dose effects, rare immune reactions, long-term organ retention, manufacturing drift, or population-specific responses. 18 , 19 , 117 , 118 FDA’s nanomaterial guidance explicitly includes postmarket submissions within scope, and the agency separately states that postmarketing safety reporting requirements apply to marketed drug and biological products, including serious and unexpected adverse experiences from all sources. 18 , 118 EMA likewise states that monitoring the safety of a medicine once it is on the market is the cornerstone of pharmacovigilance and maintains extensive post-authorisation pharmacovigilance guidance. 19 , 117

For nanomedicine, post-marketing surveillance has several special roles. 18 , 19 , 117 , 118 First, it can detect adverse events that are strongly context-dependent, such as infusion reactions, accelerated blood clearance after repeated exposure, rare complement-mediated events, or long-latency organ effects. 18 , 117 , 118 Second, it can reveal whether real-world manufacturing changes, site transfers, or supply-chain adjustments alter the product’s safety profile over time. 6 , 18 , 19 , 117 Third, it is particularly important for complex or platform-like products, where prior exposure to related excipients or carriers may modify response. 18 , 117 , 118 In this sense, pharmacovigilance for nanomedicine is not only about routine adverse-event collection; it is also about lifecycle control of a complex material system. 18 , 19 , 117 , 118

The broader translational lesson is that nanomedicine regulation increasingly extends beyond approval into lifecycle management. Quality changes, formulation drift, stability problems, and rare safety signals may only become evident after wider clinical use. 6 , 18 , 19 , 117 , 118 Accordingly, post-authorisation pharmacovigilance, CMC comparability management, and real-world safety monitoring should be viewed as integral parts of nanomedicine development rather than as purely downstream obligations. 18 , 19 , 117 , 118

Integrative Perspective

Regulatory progress in nanomedicine is therefore real but uneven. 6 , 18 , 19 , 42 , 115–118 Agencies now recognize that nanomedicines require deeper quality characterization, product-specific nonclinical evaluation, and close attention to process sensitivity, stability, and postmarket change management. 6 , 18 , 19 , 42 , 115 Early scientific advice pathways and risk-based evaluation frameworks have expanded. 18 , 19 , 42 At the same time, major challenges remain: no universal legal definition or dedicated framework exists across jurisdictions, CMC burdens are high, scale-up remains difficult, and post-approval surveillance remains especially important because clinically relevant liabilities may emerge only with broader use. 6 , 18 , 19 , 42 , 115–118

From a translational standpoint, the most important shift is conceptual. The field has moved away from viewing nanomedicine regulation as a novelty problem and toward treating it as a complex product quality-and-safety problem. That is a major advance. 6 , 18 , 19 , 42 , 115 It means the questions regulators now ask are more clinically relevant: Can the product be defined robustly? Can it be manufactured reproducibly? Are the critical attributes linked to safety and performance understood? Can the product remain stable and sterile across its lifecycle? Can postmarketing systems detect safety signals that preapproval studies may miss? These are the right questions for a mature nanomedicine field. 6 , 18 , 19 , 42 , 115–118 Although the FDA and European regulatory systems use different procedural architectures, their scientific expectations converge on the same core principles: product-specific characterization, identification of critical quality attributes, justified nonclinical testing, control of impurities and microbiological risk, stability across the product lifecycle, and evidence that manufacturing changes do not compromise safety or performance. The FDA expresses many of these expectations through dedicated nanomaterial-containing drug guidance, whereas the European framework relies more heavily on existing medicinal-product legislation, product-specific scientific advice, and specialized innovation pathways. This convergence supports international harmonization at the scientific level even in the absence of a single globally uniform legal definition of a nanomedicine. 18 , 19 , 42 Representative clinically established nanomedicines and the safety lessons that emerged from their use are summarized in Table 4 .

Special Focus: Safety in Oncology and Inflammatory Nanomedicine

Disease context strongly conditions what “safe enough” means for a nanomedicine. A nanoparticle platform that is clinically acceptable in one indication may be unsuitable in another because the relevant exposure pattern, target biology, therapeutic window, and tolerance for adverse effects are different. 3 , 15 , 17 , 123 This principle is especially important in oncology, which has historically dominated nanomedicine development, and in inflammatory disease, which is emerging as another major application space for targeted and stimuli-responsive nanoparticles. 3 , 17 , 28 , 124–128 In both settings, nanoparticles are attractive because they may improve biodistribution, reduce free-drug exposure, and enable combination or image-guided treatment; however, in both settings, the same properties that create therapeutic opportunity can also create disease-specific safety liabilities. For this reason, oncology and inflammation deserve focused consideration as distinct translational ecosystems rather than being treated as generic examples of nanoparticle drug delivery. 3 , 15 , 17 , 28 , 123–128

Why Oncology Dominates Nanomedicine Development

Oncology has remained the dominant arena for nanomedicine development because cancer therapy presents a convergence of problems that nanoparticles are, at least in principle, well suited to address: poor solubility of cytotoxic drugs, narrow therapeutic windows, severe off-target toxicity, the need for combination therapy, and the biological premium placed on spatial selectivity. 3 , 17 , 28 , 123 Cancer also offers a regulatory and clinical context in which some toxicity is acceptable if it is offset by meaningful gains in efficacy or reductions in severe conventional toxicities. 3 , 17 , 123 These features made oncology the earliest and most persistent proving ground for liposomes, albumin nanoparticles, polymeric carriers, inorganic theranostics, and, more recently, immuno-nanomedicines. 3 , 17 , 28 , 123

This dominance is visible in the clinical landscape. Recent reviews of clinical cancer nanomedicines emphasize that cancer remains the most mature therapeutic area for approved and late-stage nanoparticle products, and that many clinically successful nanomedicines in practice operate by altering pharmacokinetics and biodistribution rather than by achieving exquisitely selective tumor homing. 17 This is an important safety point. The clinical value of cancer nanomedicine has often come from reducing peak normal-tissue exposure, changing tissue partitioning, or enabling formulations that would otherwise be unusable, rather than from absolute elimination of off-target exposure. 3 , 17 , 123 Oncology therefore dominates nanomedicine not only because tumors are biologically interesting targets, but because the clinical need for safer handling of potent drugs is especially acute in cancer. 3 , 17 , 28 , 123

At the same time, oncology has exposed the limits of nanomedicine more clearly than any other field. Because cancer was the major translational test case, it also became the area in which the mismatch between preclinical targeting narratives and clinical delivery reality was most visible. Thus, oncology dominates nanomedicine development both because it offers the strongest rationale for nanoparticle use and because it has provided the clearest evidence that safety and efficacy must be assessed together rather than sequentially. 3 , 15 , 17 , 123

Tumor Targeting versus Off-Target Toxicity

The core promise of oncology nanomedicine is the possibility of increasing tumor exposure while reducing normal-tissue toxicity. However, this balance is rarely absolute. In practice, most systemically administered nanomedicines still distribute predominantly to clearance organs such as liver and spleen, while tumor deposition is often modest and heterogeneous. Consequently, the relevant safety question is not whether nanoparticles can eliminate off-target exposure, but whether they can improve the therapeutic index enough to justify their complexity and residual organ burden. 3 , 15 , 17 , 123

This issue is particularly important because tumor targeting and off-target toxicity are mechanistically coupled. Features that prolong circulation may increase opportunities for tumor accumulation but also extend exposure of blood-facing immune pathways and clearance organs. Surface chemistries that enhance cellular uptake may promote tumor-cell internalization but can also increase membrane disruption, protein adsorption, and inflammatory signaling in healthy tissues. 3 , 15 , 123 Similarly, active targeting ligands may increase binding to selected tumor-associated receptors, but they can also increase nonspecific uptake in organs expressing related epitopes or intensify immunogenic complexity. 3 , 15 , 17 , 123

Recent translational frameworks emphasize that clinically relevant cancer nanomedicines must be judged by whether the nanoparticle produces a meaningful net safety gain, not by whether it simply reaches tumors more than free drug does in preclinical models. 15 , 17 This distinction is crucial. A tumor-targeting claim is insufficient if the same formulation also produces chronic hepatosplenic burden, infusion-related reactions, or unpredictable repeat-dose clearance. In oncology, where background toxicity of conventional therapy is often already high, nanomedicines are valuable when they reduce the most dose-limiting or life-altering adverse effects without introducing new liabilities that offset the gain. 3 , 15 , 17 , 123

EPR Effect Limitations and Safety Implications

Few concepts have shaped cancer nanomedicine more deeply than the enhanced permeability and retention (EPR) effect. The EPR framework provided a persuasive mechanistic basis for passive tumor targeting and justified decades of work on long-circulating nanoparticles. However, the contemporary literature is unequivocal that EPR is highly heterogeneous across tumor types, anatomical sites, disease stages, and species, and that its magnitude in human cancers is often overestimated relative to rodent models. 3 , 15 , 123 Reviews on the tumor EPR effect now emphasize its limitations, heterogeneity, and the need for alternatives or patient-selection strategies rather than universal reliance on passive accumulation. 15

This has important safety implications. If tumor accumulation is inconsistent or modest, then the ratio between intended and unintended exposure becomes less favorable. In other words, weak or heterogeneous EPR does not merely reduce efficacy; it can actively worsen the benefit-risk balance by leaving substantial normal-organ exposure in place while failing to generate sufficient tumor benefit. 3 , 17 , 123 A nanomedicine that accumulates mainly in liver and spleen but only minimally in tumor is not simply inefficient. It may represent avoidable off-target burden without commensurate therapeutic value. 3 , 15 , 17 , 123

The limitations of EPR also affect clinical trial design and patient selection. 3 , 15 , 17 , 123 Broad enrollment of heterogeneous tumor populations without regard to vascular phenotype, stromal barrier properties, inflammatory status, or metastatic site may dilute both efficacy and safety signals. 3 , 15 , 123 This is one reason why recent translational perspectives in cancer nanomedicine increasingly advocate imaging-based pharmacology, biomarker-guided stratification, and more realistic definitions of target engagement. 15 , 17 Safety in this context cannot be divorced from delivery realism: a nanoparticle is safer in a clinically meaningful sense when it reduces normal-tissue harm while achieving enough target-site activity to justify exposure. 3 , 15 , 17 , 123 Quantitative interpretation of tumor delivery also illustrates why apparently contradictory nanomedicine findings must be examined carefully. A widely cited analysis of 117 preclinical studies estimated that a median of approximately 0.7% of the injected nanoparticle dose reached solid tumors. 129 A subsequent pharmacokinetic reanalysis showed that conclusions can change when delivery is evaluated using conventional exposure metrics rather than percentage of injected dose alone, emphasizing that absolute tumor deposition, tumor-to-plasma exposure, and pharmacokinetic benefit are related but non-equivalent measures. 130 These studies should not be treated as mutually exclusive; together they demonstrate that low absolute tumor accumulation can coexist with pharmacologically meaningful exposure in selected settings, while also reinforcing that passive targeting is not a universal mechanism of clinical selectivity.

Combination Nanotherapies and Theranostics in Oncology

Cancer is also the disease area in which combination nanotherapies and theranostic nanomedicines have been pursued most intensively. The rationale is compelling. Tumors are biologically heterogeneous and often require multimodal intervention, whether chemotherapy plus chemosensitization, immunotherapy plus local tumor modulation, or treatment plus real-time imaging. 7 , 17 , 28 , 131 Nanoparticles can, in principle, synchronize pharmacokinetics, co-localize agents that would otherwise distribute differently, and integrate imaging and treatment into a single platform. 7 , 17 , 131

Recent analysis of multi-drug cancer nanomedicine indicates that co-formulated combination nanotherapies can provide therapeutic advantages in preclinical models, particularly when two agents are delivered in the same formulation rather than in separate nanoformulations. 131 From a safety perspective, this can be beneficial because coordinated delivery may reduce the need for higher systemic doses of each agent individually and may flatten unfavorable exposure peaks. 17 , 131 However, combination nanotherapies also increase complexity. They make it harder to control release hierarchy, interpret toxicity attribution, define critical quality attributes, and establish comparability across batches. Thus, while combination nanomedicine may improve efficacy and sometimes safety, it also raises the threshold for robust translational evaluation. 15 , 17 , 131

The same is true for theranostic nanoparticles. Radiolabeled and multimodal nanoparticles can support image-guided treatment, biodistribution monitoring, and patient stratification, all of which may improve safety by identifying poor accumulators early and avoiding futile or harmful exposure. Yet theranostic systems also add material and regulatory complexity, and the imaging component must provide clinically actionable value rather than simply technological appeal. 7 , 17 In oncology, therefore, the safety case for combination and theranostic nanomedicines depends on showing that added complexity produces a real gain in therapeutic precision or toxicity reduction, not merely a richer platform design. 7 , 17 , 131

Balancing Efficacy and Safety in Cancer Applications

The central challenge in oncology nanomedicine is to balance efficacy and safety under conditions where neither can be considered in isolation. 3 , 15 , 17 , 123 Cancer patients may accept substantial toxicity if benefit is strong, but this does not reduce the need for rigorous nanosafety science. Rather, it increases the importance of identifying which toxicities are acceptable, predictable, and manageable, and which represent avoidable platform liabilities. A clinically meaningful cancer nanomedicine is therefore not one that is “nontoxic,” but one that improves the overall therapeutic index relative to available alternatives. 3 , 17 , 123

This balance is especially delicate in contemporary oncology, where nanomedicines increasingly intersect with immunotherapy, radiotherapy, and precision oncology. 3 , 7 , 17 A carrier that is acceptable in short-course cytotoxic therapy may not be acceptable when repeated chronically as part of immune modulation. Likewise, a formulation with favorable tumor imaging properties may still be problematic if it accumulates persistently in liver or spleen or perturbs innate immune function in ways that compromise combination regimens. 3 , 7 , 15 , 17 , 123 This is why recent translational frameworks emphasize early integration of manufacturability, stability, biodistribution realism, efficacy, and toxicity rather than sequential optimization. 17

The most defensible future direction for oncology nanomedicine is therefore selective ambition. 3 , 15 , 17 , 123 Nanoparticles should be pursued where they solve clinically meaningful problems that conventional formulations do not solve well—such as reducing severe normal-tissue toxicity, enabling otherwise impossible combination regimens, or providing image-guided selection of likely responders. 3 , 7 , 15 , 17 , 123 , 131 When such gains are marginal or uncertain, nanoparticle complexity may not be justified. In oncology more than any other indication, the safety value of nanomedicine must be judged by the degree to which it produces a better clinical bargain, not merely a more sophisticated formulation. 3 , 15 , 17 , 123

Special Focus: Safety in Inflammatory Nanomedicine

Inflammatory diseases are becoming an increasingly important frontier for nanomedicine because they share several features that favor targeted drug delivery: localized vascular dysfunction, activated immune-cell populations, diseased tissue microenvironments with altered pH or redox state, and major limitations of conventional systemic therapy such as broad immunosuppression, poor tissue selectivity, and cumulative toxicity. 28 , 124–128 Recent reviews emphasize that nanotechnology is opening new possibilities for the treatment and diagnosis of inflammatory diseases, including inflammatory bowel disease, autoimmune disorders, atherosclerosis, and refractory inflammation-related conditions. From a safety perspective, this shift is highly significant because inflammatory disorders often require chronic or recurrent treatment, making long-term tolerability, immune compatibility, and persistence more important than in many short-course oncologic applications. 28 , 124–128

Why Inflammatory Disease Is an Attractive but Safety-Sensitive Target

Inflammatory disease is attractive for nanomedicine because conventional anti-inflammatory and immunomodulatory drugs often have limited site selectivity and substantial systemic adverse effects. Nanoparticles offer the possibility of concentrating therapy at inflamed tissues, improving local bioavailability, protecting fragile payloads, and exploiting disease-associated cues such as reactive oxygen species, enzyme activity, altered mucus layers, or local permeability changes. 28 , 124–128 This is particularly evident in inflammatory bowel disease, where nanoparticle systems have been developed to target inflamed intestinal epithelium, inflammatory immune cells, and locally altered luminal or mucosal microenvironments. 126

However, inflammation is also a uniquely safety-sensitive target because the disease itself modifies the biological environment that nanoparticles encounter. Inflammatory states alter plasma proteins, vascular permeability, macrophage activation, cytokine tone, and tissue barrier properties, all of which can change protein corona formation, immune recognition, cellular uptake, and organ distribution. As a result, a nanoparticle that appears benign in healthy models may behave differently in inflammatory disease because the host biology is altered by the indication itself. 124 , 125 , 127 , 128 This means that safety in inflammatory nanomedicine must be assessed in disease-relevant contexts rather than inferred from healthy systems alone. 28 , 124 , 125 , 127 , 128

Inflamed-Tissue Targeting versus Off-Target Immunosuppression

The therapeutic logic of inflammatory nanomedicine is to increase local activity while decreasing systemic immunosuppression or organ toxicity. In principle, targeted nanocarriers can deliver corticosteroids, biologics, siRNA, antioxidants, or other immune-modulating agents preferentially to diseased tissues, thereby lowering total systemic exposure. This can be a major safety advantage over conventional formulations, especially in diseases where long-term systemic therapy causes infection risk, metabolic toxicity, endocrine complications, or generalized immune suppression. 28 , 124–128

Yet the same logic creates specific risks. In inflammatory disease, off-target effects do not only mean accumulation in liver or spleen; they also mean unintended modulation of immune function in tissues not meant to be treated. 28 , 124–128 A nanoparticle carrying an anti-inflammatory payload may reduce disease-site inflammation while also suppressing protective host responses elsewhere, or it may alter macrophage and dendritic-cell behavior in ways that change infection susceptibility. 124 , 125 , 127 , 128 Thus, “targeting inflammation” is not automatically equivalent to improving safety. It improves safety only when the formulation materially reduces nonspecific immune perturbation relative to standard therapy. 28 , 124–128

This balance is particularly important in chronic inflammatory disease, where repeated dosing is common. 124 , 126–128 Even a modest amount of unintended immune activation, anti-carrier immunogenicity, or organ retention may become clinically relevant over time. For this reason, inflammatory nanomedicine requires especially careful assessment of repeat-dose tolerance, cytokine perturbation, macrophage interaction, and long-term clearance. 28 , 124–128

Disease-Associated Permeability in Inflammation: Opportunity and Limitation

Inflamed tissues often exhibit increased vascular permeability, altered epithelial barriers, and local retention phenomena that conceptually resemble, but are not identical to, the EPR effect discussed in oncology. 124 , 126–128 In diseases such as inflammatory bowel disease or inflamed joints, these barrier alterations can be exploited to improve nanoparticle localization. 124 , 126 This creates an important therapeutic opportunity: nanoparticles may accumulate preferentially in diseased tissue even without highly elaborate targeting ligands. 124 , 126 , 128

However, the safety lesson from oncology applies here as well. Disease-associated permeability is heterogeneous, dynamic, and often unpredictable. 124 , 126–128 If inflammatory targeting is weak or variable, then nanoparticles may still distribute primarily to liver, spleen, or other nontarget tissues, leaving patients exposed to both carrier burden and payload-related toxicity without sufficient site-specific benefit. 28 , 124–128 In other words, the existence of an inflamed microenvironment does not guarantee safe or efficient targeting. It simply creates a potential advantage that must be demonstrated under realistic disease conditions. 124 , 126–128

Furthermore, in inflammatory disease, increased permeability may coexist with tissue fragility. 124 , 126–128 A nanoparticle that penetrates inflamed tissue effectively may also aggravate barrier dysfunction, oxidative stress, or local immune-cell activation if its surface chemistry or degradation behavior is unfavorable. 124 , 127 , 128 This is especially relevant in mucosal diseases such as IBD, where a formulation intended to improve local therapy may also interact directly with compromised epithelial barriers and microbiota-associated environments. 126 Thus, disease-associated targeting in inflammation is simultaneously a delivery opportunity and a safety challenge. 28 , 124–128

Combination Nanotherapies and Theranostics in Inflammation

Combination approaches are also relevant in inflammatory nanomedicine, though they are often less emphasized than in oncology. Chronic inflammatory diseases frequently involve overlapping pathways—oxidative stress, macrophage activation, cytokine dysregulation, endothelial activation, and tissue remodeling—so there is a strong rationale for carriers that deliver more than one therapeutic modality or that integrate treatment with diagnostic readouts. 124 , 125 , 127 , 128 Nanotheranostic strategies for inflammation are being developed to combine imaging of inflammatory lesions with local drug release or responsive treatment activation. 28 , 127

From a safety standpoint, these approaches offer the possibility of better patient selection and response monitoring, which may reduce unnecessary exposure. For example, imaging-guided identification of highly inflamed lesions or nanoprobes that report inflammatory status could help determine when and where a therapeutic nanoparticle is likely to be beneficial. 28 , 127 At the same time, adding imaging or multiple payloads increases formulation complexity and may introduce new sources of immunogenicity, stability problems, or off-target accumulation. As in oncology, the safety benefit of combination and theranostic inflammation nanomedicine therefore depends on whether the added complexity produces actionable precision rather than technological ornamentation. 28 , 127

Balancing Efficacy and Safety in Inflammatory Applications

The safety calculus in inflammatory nanomedicine differs from oncology because many inflammatory diseases require long-term management rather than short-course intervention. 28 , 124–128 Consequently, tolerability thresholds are often stricter. A formulation that would be acceptable in advanced cancer may not be acceptable in chronic inflammatory bowel disease, rheumatoid arthritis, or atherosclerosis if it causes persistent organ retention, cumulative immunogenicity, or repeated infusion-related risk. 124 , 126–128 This means that the bar for safety in inflammatory nanomedicine is, in some respects, higher than in oncology even when the payload itself is less intrinsically toxic. 28 , 124–128

For inflammatory diseases, the most promising nanomedicines are likely to be those that combine local or cell-selective delivery with clear advantages in long-term safety—such as reduced systemic steroid burden, less generalized immunosuppression, lower dosing frequency, or better disease-site retention with rapid post-treatment clearance. 124 , 126–128 Recent inflammation-focused reviews also emphasize the importance of designing nanoparticles to minimize unintended inflammatory responses, which is especially relevant when the target indication is itself defined by immune dysregulation. 127 In this context, the best inflammatory nanomedicine is not merely one that reaches an inflamed tissue, but one that does so without amplifying unwanted immune activation or creating chronic carrier burden. 28 , 124–128

Integrative Perspective

Taken together, oncology and inflammatory disease illustrate two complementary truths about nanomedicine safety. 3 , 7 , 15 , 17 , 28 , 123–128 , 131 In oncology, nanoparticles are justified when they materially improve the therapeutic index of potent therapies, even if some toxicity remains. 3 , 7 , 15 , 17 , 123 , 131 In inflammatory disease, nanoparticles are attractive when they reduce systemic exposure and chronic toxicity relative to long-term conventional treatment, but the tolerability bar is often higher because treatment may be prolonged. 28 , 124–128 In both cases, the central challenge is the same: target-site delivery must be sufficient to justify residual off-target exposure. 3 , 7 , 15 , 17 , 28 , 123–128 , 131

These indications also demonstrate that safety is not a platform-intrinsic label. It is a disease-specific, exposure-specific, and patient-specific outcome. A nanocarrier that is acceptably safe in cancer may be too reactogenic or persistent for chronic inflammatory disease, while a formulation optimized for long-term inflammatory treatment may not deliver sufficient intensity for oncologic use. For this reason, disease-focused nanosafety should become a standard feature of translational nanomedicine development rather than a late interpretive exercise. The future of safer nanomedicine will depend not only on better carriers, but on better alignment between platform behavior and disease-specific safety requirements. 3 , 7 , 15 , 17 , 28 , 123–128 , 131

Emerging Strategies to Enhance Safety

The next phase of nanosafety is increasingly being shaped by strategies that do not merely mitigate toxicity after formulation, but attempt to predict, personalize, and design out risk earlier in development. This shift reflects a broader maturation of nanomedicine: safety is no longer viewed only as a downstream toxicology requirement, but as a design variable linked to patient heterogeneity, formulation architecture, computational prediction, and lifecycle control. Emerging approaches with particular relevance to safety include personalized nanomedicine and patient stratification, AI-assisted nanomaterial design, predictive toxicology and digital-twin frameworks, safer-by-design methodologies, transient or stimuli-responsive nanoplatforms, and exosome-inspired or other biologically derived nanoparticles. 115 , 132–138

Personalized Nanomedicine and Patient Stratification

Personalized nanomedicine has emerged from the recognition that nanoparticle behavior is not uniform across patients. Tumor heterogeneity, inflammatory status, vascular permeability, immune tone, organ function, and biomarker expression can all change nanoparticle accumulation, intracellular uptake, and adverse-event profiles. Recent reviews of precision nanomedicine in oncology emphasize that nanoplatforms are increasingly being linked to molecular diagnostics, imaging, and biomarker-informed treatment selection, with the goal of improving efficacy while reducing unnecessary exposure in patients unlikely to benefit. 132 From a safety perspective, this is important because many nanomedicines fail not because the platform is universally unsafe, but because the right patient population has not been identified in advance.

Patient stratification is therefore becoming a safety tool as much as an efficacy tool. If biomarkers, imaging phenotypes, or omics readouts can identify patients with favorable delivery conditions or lower risk of immune and organ-specific toxicity, nanomedicine can move away from empiric “one-platform-for-all” dosing and toward more selective use. This strategy is still early in clinical implementation, but the conceptual shift is already clear: safer nanomedicine will likely depend not only on better materials, but also on better matching of materials to biological subpopulations. 132

AI-Assisted Nanomaterial Design

Artificial intelligence and machine learning are increasingly being used to optimize nanoparticle composition, size, charge, ligand density, and formulation process parameters. Recent reviews specifically focused on AI and nanomedicine describe this convergence as a way to connect nanoparticle design variables with biological performance, toxicological datasets, and patient-level information. 133 In parallel, work in nanotoxicology has emphasized that AI can help identify patterns in large, heterogeneous datasets that are difficult to resolve through conventional statistical methods alone, thereby improving prediction of biocompatibility and hazard profiles. 135 , 136

The safety relevance of AI-assisted design is substantial. Instead of synthesizing and testing large numbers of candidates experimentally, AI models can help prioritize formulations less likely to aggregate, trigger immune activation, or exhibit undesirable biodistribution. Machine learning models are also being developed specifically for nanoparticle toxicity prediction, using physicochemical descriptors and biological data to estimate cellular or dynamic toxicity outcomes. 134 Although model generalizability remains limited by data quality and standardization, the field is clearly moving toward computational preselection of safer formulation space rather than purely trial-and-error development. 133–136 Current limitations remain substantial, however. Many available datasets are small, heterogeneous, incompletely standardized, and biased toward positive experimental results; descriptors are not always measured under biologically comparable conditions; external validation is uncommon; and models trained on one nanoparticle family may not extrapolate reliably to another. For translational use, predictive performance must therefore be accompanied by transparent feature definition, uncertainty estimation, prospective validation, and experimentally testable mechanistic interpretation. 108 , 133–136

Predictive Toxicology and Digital Twins

Predictive toxicology is evolving from relatively simple quantitative structure–activity approaches toward multiscale computational frameworks that integrate physicochemical data, mechanistic biology, and exposure modeling. In nanotoxicology, recent analyses argue that AI can reduce dependence on extensive in vitro and in vivo testing by improving pattern recognition, mechanistic inference, and early identification of hazardous nanomaterials. 135 This does not eliminate experimental toxicology, but it can make it more focused and hypothesis-driven.

Digital twins are a more ambitious extension of this logic. In healthcare broadly, digital twins are being developed as virtual, data-updated representations of patients, organs, or disease processes that support prediction and personalization. 137 In nanomedicine, the concept remains largely early-stage and partly aspirational, but recent reviews explicitly identify digital twins as a future route to simulate nanoparticle behavior in patient-specific biological environments, including treatment response and delivery performance. 133 , 135 , 137 For safety, the long-term promise is considerable: digital twins could eventually help forecast organ exposure, repeat-dose behavior, or patient-specific adverse-event risk before actual administration. At present, however, this remains more a strategic direction than a clinically established capability in nanomedicine. 133 , 135 , 137

Safer-by-Design Frameworks

Safer-by-design has become one of the most influential emerging frameworks because it formalizes the idea that hazard reduction should begin at the earliest design stages. The NanoImpact guidelines developed from the NanoReg2 experience describe safe(r)-by-design as an iterative process that balances functionality, human health and safety, environmental impact, and industrial feasibility from the start of innovation rather than after product definition. 115 The framework is built around three pillars: safer materials and products, safer use and end-of-life, and safer industrial production. 115

For medical nanoparticles, the importance of safer-by-design lies in its operational effect. It encourages developers to treat degradability, immune compatibility, aggregation liability, impurity control, and process safety as core design parameters rather than as downstream validation problems. This approach also aligns with the broader movement in nanotechnology policy toward early integration of safety considerations into research and development strategies. 115 In practice, its success will depend on whether predictive assays, standardized datasets, and translational decision criteria become strong enough to guide real formulation choices. Even so, safer-by-design has already changed the way advanced nanomedicines are conceptually developed. 115 , 135

Stimuli-Responsive and Transient Nanoplatforms

Stimuli-responsive and transient nanoplatforms are emerging as safety-enhancing strategies because they aim to reduce nonspecific exposure in both space and time. Instead of remaining indefinitely stable or releasing payload constitutively, these systems are designed to respond to endogenous signals such as pH, redox gradients, enzymes, or reactive oxygen species, or to exogenous triggers such as light, ultrasound, heat, or magnetic fields. The safety logic is straightforward: if a nanoparticle remains relatively inert during circulation but disassembles, degrades, or activates only in the target microenvironment, off-target toxicity and long-term retention may be reduced. 138

Recent work on biodegradable, stimuli-responsive silica nanoparticles illustrates this trend clearly. That review emphasizes that poor degradability and nonspecific accumulation have limited broader clinical translation of silica-based systems and argues that multi-stimuli-responsive biodegradable designs can improve in vivo degradation and clearance under pathological conditions. 138 More broadly, transient nanoplatforms represent an important conceptual shift away from persistent multifunctional carriers toward time-limited therapeutic systems whose clinical value depends partly on controlled disappearance after use. 138

Exosome-Inspired and Biologically Derived Nanoparticles

Biologically derived nanoparticles, especially extracellular vesicles and exosome-inspired systems, have attracted intense interest because they offer an alternative to fully synthetic carriers at the nano–bio interface. Recent high-impact reviews describe exosomes as promising next-generation delivery systems with advantages that may include tissue selectivity, barrier crossing, and favorable safety-related properties compared with some synthetic nanocarriers, while also emphasizing that manufacturing, purity, cargo heterogeneity, and standardization remain major obstacles. 139

Plant-derived exosome-like nanoparticles have further expanded this area. A 2024 review in the International Journal of Molecular Sciences describes PELNs as membranous vesicles containing proteins, lipids, and nucleic acids, and highlights their abundant sources, high yields, and reportedly low risk of immunogenicity in vivo as reasons for growing biomedical interest. 140 These properties make biologically derived particles attractive as potentially lower-immunogenic and more biodegradable delivery systems, although the field still faces substantial challenges in isolation, characterization, compositional consistency, and mechanism-of-action definition. 139 , 140 Thus, exosome-inspired platforms are promising for safety enhancement, but they are not yet a solved translational problem; their main current value lies in offering a biologically grounded template for the next generation of safer delivery systems. 139 , 140

Integrative Perspective

Taken together, these emerging strategies indicate that the future of nanoparticle safety will likely depend on integration rather than on any single material innovation. Personalized nanomedicine links safety to patient biology. AI-assisted design and predictive toxicology shift hazard recognition upstream. Digital twins point toward virtual pretesting of patient-specific exposure and response. Safer-by-design embeds risk reduction into innovation itself. Stimuli-responsive and transient platforms reduce unnecessary persistence and premature activity. Biologically derived nanoparticles attempt to solve safety at the interface level by learning from natural vesicle systems. 115 , 132–140

The common direction across all of these approaches is clear: safer nanomedicine will increasingly be predictive, adaptive, and biologically informed. The main challenge now is to convert these emerging strategies from promising conceptual advances into standardized, manufacturable, and clinically validated safety gains.

Critical Objectives for the Next Decade

If the first generation of nanomedicine was defined by proof-of-concept delivery and the second by early clinical translation, the next decade must be defined by predictability, comparability, and clinical relevance. The principal obstacles to safer nanomedicine are no longer merely conceptual. They are now well recognized: fragmented testing pipelines, incomplete long-term fate data, insufficiently predictive models, limited mechanistic resolution of immunotoxicity, poor reporting consistency, late separation of efficacy from safety, and underdeveloped attention to equity, ethics, and environmental consequences. 9 , 19 , 141 , 142 Recent translational, standards, and nanosafety frameworks increasingly converge on the view that future progress will depend less on inventing ever more complex nanoplatforms and more on building a disciplined evidence infrastructure that can determine which nanomedicines are truly safe, for whom, and under what conditions. 9 , 19 , 141 , 142 The resulting iterative safety-by-design lifecycle is summarized in Figure 4 .

Proposed iterative framework for developing safer and more translatable nanomedicines. The process begins with definition of the clinical problem, target patient population, and acceptable safety margin, followed by rational platform selection and early physicochemical characterization. Early screening should integrate hemocompatibility, immunotoxicity, protein corona analysis, assay compatibility, and intracellular fate, rather than evaluating safety only after efficacy optimization. Advanced translational evaluation should incorporate disease-relevant in vitro and in vivo models, biodistribution studies, imaging-based tracking, repeated-dose assessment, and long-term fate analysis. Chemistry, manufacturing, and controls (CMC), including scale-up, sterilization, storage, stability, and batch reproducibility, should be integrated throughout development. Patient stratification, regulatory alignment, and post-marketing surveillance complete the framework, emphasizing that safe nanomedicine development is iterative and depends on continuous co-optimization of safety, efficacy, and product quality. This is an original schematic synthesis prepared specifically for the present review; the cross-cutting role of CMC and lifecycle quality management is developed in Regulatory Progress and Translational Considerations .

Harmonized Safety Testing Pipelines

A primary objective for the next decade is the development of harmonized safety testing pipelines that link nanoparticle characterization, exposure definition, in vitro screening, advanced non-animal models, in vivo confirmation where needed, and mechanism-aware data integration. 141–143 Current nanosafety assessment remains too fragmented across laboratories, product classes, and jurisdictions. 141 , 143 Recent standardization-focused reviews explicitly call for global efforts to overcome inconsistency in nanotoxicology workflows, while broader healthcare nanotechnology reviews identify the lack of harmonized safety and risk-assessment frameworks as a continuing translational bottleneck. 141 , 143

What is needed is not a single rigid test battery for all nanomedicines, but a tiered, harmonized logic. 141 , 143 Such pipelines should define which core assays are required for all platforms, which are route- or material-specific, how dose and exposure should be normalized, and what physicochemical descriptors must accompany every safety result. 141 , 143 Harmonization should also include assay interference controls, reference materials where feasible, standardized biological media conditions, and predefined criteria for escalating from simple in vitro testing to more complex models. 141–143 Without this kind of structured pipeline, nanosafety data will remain difficult to compare across studies and difficult to use in regulatory decision-making. 141 , 143

Better Long-Term Fate and Clearance Studies

A second critical objective is much stronger evidence on long-term fate, degradation, and clearance. Short-term tolerability is not enough for many nanomedicines, especially those intended for repeated dosing, chronic diseases, imaging follow-up, or containing inorganic or slowly degradable components. 19 , 141 Recent healthcare nanotechnology reviews continue to emphasize environmental and biological persistence as major unresolved safety issues, and EMA’s 2025 horizon-scanning report highlights the importance of understanding product quality, biodistribution, and long-term behavior for nanotechnology-based medicinal products. 19 , 141

The field needs longitudinal studies that distinguish the fate of the intact nanoparticle from that of released drug, degradation fragments, dissolved ions, or shed coatings. 19 , 107 , 141 It also needs organ-resolved persistence studies in liver, spleen, kidney, lung, brain, placenta where relevant, and clearance compartments under repeated-dose conditions. 19 , 141 Ideally, these studies should combine imaging, bioanalytical chemistry, histopathology, and mechanistic biomarkers rather than relying on terminal bulk accumulation alone. 19 , 107 , 141 The central goal is to move from static biodistribution snapshots to true fate mapping across clinically relevant timescales. 19 , 141

Clinically Relevant Animal and Non-Animal Models

A third priority is the development and validation of clinically relevant animal and non-animal models. 19 , 142 , 144 Preclinical nanomedicine has long been constrained by model systems that overestimate delivery success or fail to represent human immune diversity, vascular physiology, barrier function, and chronic disease states. 17 , 142 At the same time, the field is under growing scientific and ethical pressure to reduce unnecessary animal use and adopt more predictive non-animal methodologies. 142 , 144 EFSA’s 2024 review of new approach methodologies for nanomaterials highlights nano-specific frameworks for non-animal testing, while WHO’s draft 3Rs-oriented guidance on phasing out certain animal-derived quality tests reflects a wider movement toward scientifically and ethically improved methods. 142 , 144

The next decade should not frame animal and non-animal approaches as adversarial alternatives. The real objective is a fit-for-purpose model ecosystem. This includes disease-relevant rodent and non-rodent models when whole-body pharmacology is essential, but also barrier chips, immune-competent organoids, microphysiological systems, and exposure-aware 3D platforms for questions involving BBB transport, placental passage, pulmonary deposition, or immune-cell cross-talk. 142 , 144 The highest priority should be given to models that improve human predictivity, not merely experimental convenience. 142 , 144

Greater Mechanistic Understanding of Immunotoxicity

Among all unresolved safety domains, immunotoxicity remains one of the most urgent. Nanoparticles interact with complement, antibodies, coagulation pathways, macrophages, dendritic cells, mast cells, inflammasomes, and innate pattern-recognition systems in ways that are often context-dependent and still incompletely predictable. 9 , 145 , 146 A 2025 review on mechanistic paradigms of nanoparticle immunotoxicity and a 2025 Biochemical Pharmacology article on charting new frontiers in nanoparticle immunotoxicity both underscore that current assessment still relies too heavily on conventional frameworks that were not designed around the distinctive immune interactions of nanoscale materials. 145 , 146 Earlier immunotoxicity reviews similarly emphasize the importance of susceptible populations and disease-state effects. 9

The key objective is not just to catalog immune adverse events, but to build mechanistic maps that link particle attributes to specific immune pathways and then to clinically relevant outcomes such as infusion reactions, accelerated blood clearance, cytokine patterns, macrophage polarization, or chronic inflammatory burden. 9 , 145 , 146 This requires better immune-competent model systems, more standardized complement and cytokine testing, repeated-exposure study designs, and greater use of systems immunology. It also requires recognizing that immunotoxicity is shaped by host biology, including inflammation, prior exposure history, and patient-specific protein corona formation. 9 , 145 , 146

Standard Reporting Guidelines

A fifth objective is the establishment and enforcement of standard reporting guidelines. The nanomedicine literature still suffers from incomplete reporting of synthesis conditions, physicochemical characterization, biological media composition, dose metrics, assay interference controls, and batch identity. 141 , 143 This undermines reproducibility, meta-analysis, regulatory interpretation, and model building. Standardization-oriented reviews in nanotoxicology explicitly identify this gap, and the broader reproducibility literature in nanomedicine has made the same point repeatedly. 141 , 143

The next decade should produce community-accepted minimum reporting standards covering at least the following: critical material attributes, dispersion protocol, storage and age of material, exposure conditions, dose metric rationale, interference controls, biological model description, statistical treatment, and batch identifiers. 141 , 143 Reporting standards should also accommodate advanced datasets such as omics, imaging, and AI-ready descriptors. 107 , 141 , 143 Without such infrastructure, it will remain difficult to distinguish true biological variability from avoidable methodological noise. 141 , 143

Integration of Efficacy and Safety Early in Design

Another essential objective is the early integration of efficacy and safety in nanoparticle design. 9 , 17 , 147 Too much nanomedicine development still optimizes payload loading, targeting, or in vitro potency first and then evaluates safety later. 17 , 147 This sequence is often misleading because the very same attributes that increase efficacy may also increase hemolysis, complement activation, off-target accumulation, or chronic organ burden. 9 , 17 , 147 The Nature Nanotechnology “DELIVER” framework emphasizes that manufacturability, stability, release behavior, reproducibility, and therapeutic-index improvement should be assessed early enough to determine whether development should proceed. 17

In the next decade, the field should normalize dual-optimization workflows in which efficacy and safety are treated as co-primary design outcomes. That means designing screening cascades in which biodistribution realism, blood compatibility, degradation fate, and immune compatibility are evaluated alongside potency and target engagement from the earliest stages. 9 , 17 , 147 It also means abandoning platforms whose safety liabilities are not justified by a meaningful clinical advantage, even if they perform well in permissive proof-of-concept models. This is the practical meaning of mature translational nanomedicine. 17 , 147

Equity, Ethics, and Environmental Considerations

The next decade must also expand the definition of nanosafety beyond individual-patient toxicology to include equity, ethics, and environmental responsibility. 141 , 148–150 The 2024 US National Nanotechnology Initiative EHS research strategy explicitly links nanotechnology research with environmental justice and health equity, and more recent ethical analyses emphasize social justice, public health, data governance, and responsible innovation as integral to nanosafety ecosystems. 148–150

For medical nanoparticles, equity means asking whether safer nanomedicines will be accessible only in highly resourced settings, whether patient stratification tools will work across diverse populations, and whether data used to train AI or predictive models adequately represent the populations who will receive these products. 148–150 Ethics includes informed consent in complex theranostic settings, privacy in data-rich personalized nanomedicine, and responsible communication of benefit versus uncertainty. 148 , 149 Environmental considerations include occupational exposure during manufacturing, lifecycle release of nanomaterials, disposal, persistence, and the broader ecological implications of scaled medical use. 141 , 150 A nanomedicine that is clinically effective but environmentally irresponsible or distributively unjust cannot be considered a fully successful safety achievement. 141 , 148–150

Integrative Perspective

Taken together, these priorities define a clear agenda for the field. The next decade should aim for harmonized testing pipelines, deeper long-term fate studies, model systems with greater clinical relevance, mechanistically richer immunotoxicity science, enforceable reporting standards, early co-optimization of efficacy and safety, and explicit incorporation of equity, ethics, and environmental stewardship. None of these objectives is peripheral. They are the infrastructure required to turn nanosafety from a reactive screening exercise into a predictive translational discipline. 9 , 17 , 19 , 107 , 141–150

The most important shift is strategic rather than merely technical. Nanomedicine no longer needs only better particles; it needs better systems for deciding which particles deserve to become medicines. That is the central safety objective for the coming decade. 17 , 147

The principal strategic objectives for the coming decade, together with recommended actions for the field, are summarized in Table 5

Conclusion

Nanomedicine has reached a decisive stage in its development. The field has moved well beyond the era in which nanoparticles were viewed mainly as elegant delivery concepts or highly flexible experimental platforms. It has now generated approved products, shaped important areas of modern therapeutics, and demonstrated that nanoscale engineering can solve clinically meaningful problems. At the same time, this progress has made one point unmistakably clear: the future of nanomedicine will depend less on how complex nanoparticles can become and more on how safely, reproducibly, and rationally they can be translated into real medical use.

One of the most important lessons to emerge from the field is that safety cannot be treated as a late checkpoint added after efficacy has already been optimized. In nanomedicine, safety is inseparable from function. The same attributes that influence circulation, targeting, uptake, intracellular delivery, and release also shape immune recognition, blood compatibility, off-target accumulation, degradation, and long-term tolerance. This means that safety is not a peripheral constraint on innovation. It is the framework that determines whether innovation becomes clinically meaningful. In many ways, the maturation of nanomedicine can be understood as a shift from designing nanoparticles that can work to designing nanoparticles that can work reliably, safely, and for the right patients.

There has been genuine and important progress. The field now has a far better understanding of how physicochemical properties govern biological behavior. It has developed safer lipid nanoparticle systems, more sophisticated stealth and biomimetic strategies, more biodegradable and responsive materials, and more advanced tools for preclinical characterization and biodistribution analysis. Regulatory thinking has also become more mature, with greater emphasis on critical quality attributes, manufacturing control, stability, and lifecycle management. These advances show that the safety of nanoparticles is not an unsolved abstraction. It is a problem that can be progressively reduced through careful design, rigorous characterization, and translational discipline.

And yet, major challenges remain. Human biodistribution is still difficult to predict with confidence. Long-term fate and clearance are still insufficiently resolved for many platforms. Immunotoxicity remains only partly understood in mechanistic terms, especially under repeated dosing and in heterogeneous patient populations. Reproducibility across studies and batch-to-batch consistency in complex products still represent major obstacles. Perhaps most importantly, the field still needs to become more selective in deciding which nanoparticle systems truly justify clinical development. Not every technically impressive nanoplatform offers a meaningful therapeutic advantage, and not every increase in sophistication leads to better medicine.

For that reason, the next decade should be guided by a more deliberate philosophy. Safer nanomedicine will not come only from inventing new materials. It will come from integrating efficacy, safety, manufacturing, and patient relevance from the earliest stages of design. It will come from better long-term studies, more predictive biological models, stronger mechanistic toxicology, clearer reporting standards, and greater willingness to abandon platforms whose complexity exceeds their clinical value. It will also come from recognizing that different diseases impose different safety demands, and that truly successful nanomedicine must be adapted not only to a target, but to a therapeutic context.

The field has therefore moved beyond the question of whether nanoparticles can contribute meaningfully to medicine. The more consequential translational question is which nanoparticle systems should advance, in which clinical settings, and with what degree of confidence in their long-term safety, manufacturing reproducibility, and patient-specific benefit-risk profile. Addressing this question requires increasingly mechanistic, quantitative, and clinically grounded evidence rather than generalized assumptions of platform biocompatibility.

Ultimately, enhancing the safety of nanoparticles in medicine should not be seen as a defensive exercise or a regulatory burden. It is the most direct route toward making nanomedicine more credible, more precise, and more useful to patients. The strongest future for the field will belong not to the most complicated systems, but to the ones whose design, biological behavior, manufacturability, and safety profile are aligned in a coherent and clinically justified way. If nanomedicine can sustain that level of rigor, it will not simply remain promising. It will become one of the most durable and transformative areas of modern therapeutic science.

Accordingly, the most consequential priorities for the next decade are harmonized characterization and reporting, mechanistic immunotoxicity and genotoxicity assessment, longitudinal fate and clearance studies, clinically relevant experimental models, quantitative exposure-based toxicology, tighter CMC-toxicology integration, and patient stratification supported by biomarkers and imaging. AI, multi-omics, PBPK modeling, microphysiological systems, and digital-twin approaches can strengthen this framework, but only when their predictions are externally validated and linked to experimentally measurable mechanisms. Progress in these areas should accelerate regulatory confidence and clinical adoption while reducing the probability that increased nanoparticle complexity produces risk without proportional therapeutic value.

Funding Statement

This study was funded by the Carlos Chagas Filho Foundation for Research Support of the State of Rio de Janeiro - FAPERJ - through the grants Cientista do Nosso Estado - E-26/200.815/2021, Rede NanoSaude - E-26/010.000981/2019, Doenças Raras - E-26/210.010/2026, and Pensa Rio - E-26/210.728/2025. This work was also supported by the National Council for Scientific and Technological Development - CNPq - through the grants Productivity Fellowship 1B - 301069/2018-2, International Cooperation - 440852/2023-4, Universal Call - 401692/2023-0, and Precision Medicine - 444089/2023-3 awarded to Ralph Santos-Oliveira. The funders had no role in the literature selection, interpretation, preparation of the manuscript, or decision to submit the work for publication.

Data Sharing Statement

No new datasets were generated or analyzed for this narrative review. The evidence discussed in the manuscript is available from the cited literature. Additional information regarding the review synthesis is available from the corresponding author, Dr. Ralph Santos-Oliveira, upon reasonable request.

Ethics Approval

This is a narrative review article and did not involve new research with human participants, human materials, human data, or animals.

Disclosure

The authors declare that they have no financial or non-financial competing interests related to this work.

References

1. Doane TL, Burda C. The unique role of nanoparticles in nanomedicine: imaging, drug delivery and therapy. Chem Soc Rev . 2012;41(7):2885–70. doi: 10.1039/C2CS15260F

2. Shan X, Gong X, Li J, Wen J, Li Y, Zhang Z. Current approaches of nanomedicines in the market and various stage of clinical translation. Acta Pharm Sin B . 2022;12(7):3028–3048. doi: 10.1016/j.apsb.2022.02.025

3. Bhatia SN, Chen X, Dobrovolskaia MA, Lammers T. Cancer nanomedicine. Nat Rev Cancer . 2022;22(10):550–556. doi: 10.1038/S41568-022-00496-9

4. Fan D, Cao Y, Cao M, Wang Y, Cao Y, Gong T. Nanomedicine in cancer therapy. Signal Transduct Target Ther . 2023;8(1):293. doi: 10.1038/s41392-023-01536-y

5. Younis MA, Tawfeek HM, Abdellatif AAH, Abdel-Aleem JA, Harashima H. Clinical translation of nanomedicines: challenges, opportunities, and keys. Adv Drug Deliv Rev . 2022;181. doi: 10.1016/j.addr.2021.114083

6. Clogston JD, Foss W, Harris D, et al. Current state of nanomedicine drug products: an industry perspective. J Pharm Sci . 2024;113(12):3395–3405. doi: 10.1016/j.xphs.2024.09.005

7. Zhang P, Xiao Y, Sun X, et al. Cancer nanomedicine toward clinical translation: obstacles, opportunities, and future prospects. Med . 2023;4(3):147–167. doi: 10.1016/j.medj.2022.12.001

8. Lammers T. Nanomedicine Tumor Targeting. Adv Mater . 2024;36(26). doi: 10.1002/ADMA.202312169

9. Hofer S, Hofstätter N, Punz B, Hasenkopf I, Johnson L, Himly M. Immunotoxicity of nanomaterials in health and disease: current challenges and emerging approaches for identifying immune modifiers in susceptible populations. Wiley Interdiscip Rev Nanomed Nanobiotechnol . 2022;14(6):e1804. doi: 10.1002/WNAN.1804

10. Anchordoquy T, Artzi N, Balyasnikova IV, et al. Mechanisms and barriers in nanomedicine: progress in the field and future directions. ACS Nano . 2024;18(22):13983–13999. doi: 10.1021/ACSNANO.4C00182

11. Ashkarran AA, Gharibi H, Modaresi SM, Saei AA, Mahmoudi M. Standardizing protein corona characterization in nanomedicine: a multicenter study to enhance reproducibility and data homogeneity. Nano Lett . 2024;24(32):9874–9881. doi: 10.1021/ACS.NANOLETT.4C02076

12. Cisneros EP, Morse BA, Savk A, Malik K, Peppas NA, Lanier OL. The role of patient-specific variables in protein Corona formation and therapeutic efficacy in nanomedicine. J Nanobiotechnology . 2024;22(1). doi: 10.1186/S12951-024-02954-Y

13. Szebeni J, Muggia F, Gabizon A, Barenholz Y. Activation of complement by therapeutic liposomes and other lipid excipient-based therapeutic products: prediction and prevention. Adv Drug Deliv Rev . 2011;63(12):1020–1030. doi: 10.1016/j.addr.2011.06.017

14. Yang Q, Lai SK. Anti-PEG immunity: emergence, characteristics, and unaddressed questions. Wiley Interdiscip Rev Nanomed Nanobiotechnol . 2015;7(5):655–677. doi: 10.1002/WNAN.1339

15. Sun R, Xiang J, Zhou Q, et al. The tumor EPR effect for cancer drug delivery: current status, limitations, and alternatives. Adv Drug Deliv Rev . 2022:191. doi: 10.1016/j.addr.2022.114614

16. Golombek SK, May JN, Theek B, et al. Tumor targeting via EPR: strategies to enhance patient responses. Adv Drug Deliv Rev . 2018;130:17–38. doi: 10.1016/j.addr.2018.07.007

17. Joyce P, Allen CJ, Alonso MJ, et al. A translational framework to DELIVER nanomedicines to the clinic. Nat Nanotechnol . 2024;19(11):1597–1611. doi: 10.1038/S41565-024-01754-7

18. Products D. Including biological products, that contain nanomaterials - guidance for industry | FDA. 2022. Available from: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/drug-products-including-biological-products-contain-nanomaterials-guidance-industry . Accessed April 12, 2026.

19. Medicines agency - heads of medicines agencies E. Nanotechnology-based medicinal products for human use - EU-IN horizon scanning report. 2025. Available from: https://www.ema.europa.eu/en/documents/report/nanotechnology-based-medicinal-products-human-use-eu-horizon-scanning-report_en.pdf . Accessed April 12, 2026.

20. Wu S, Lin L, Shi L, Liu S. An overview of lipid constituents in lipid nanoparticle mRNA delivery systems. Wiley Interdiscip Rev Nanomed Nanobiotechnol . 2024;16(4). doi: 10.1002/WNAN.1978

21. Beach MA, Nayanathara U, Gao Y, et al. Polymeric nanoparticles for drug delivery. Chem Rev . 2024;124(9):5505–5616. doi: 10.1021/ACS.CHEMREV.3C00705

22. Rahimkhoei V, Alzaidy AH, Abed MJ, Rashki S, Salavati-Niasari M. Advances in inorganic nanoparticles-based drug delivery in targeted breast cancer theranostics. Adv Colloid Interface Sci . 2024;329. doi: 10.1016/j.cis.2024.103204

23. Arabestani MR, Bigham A, Kamarehei F, et al. Solid lipid nanoparticles and their application in the treatment of bacterial infectious diseases. Biomed Pharmacother . 2024:174. doi: 10.1016/j.biopha.2024.116433

24. Wang J, Ding Y, Chong K, et al. Recent advances in lipid nanoparticles and their safety concerns for mRNA delivery. Vaccines . 2024;12(10):1148. doi: 10.3390/VACCINES12101148

25. Jacob EM, Huang J, Chen M. Lipid nanoparticle-based mRNA vaccines: a new frontier in precision oncology. Precis Clin Med . 2024;7(3):pbae017. doi: 10.1093/PCMEDI/PBAE017

26. Estapé Senti M, Del Valle L G, Schiffelers RM. mRNA delivery systems for cancer immunotherapy: lipid nanoparticles and beyond. Adv Drug Deliv Rev . 2024;206. doi: 10.1016/j.addr.2024.115190

27. Sunoqrot S, Abdel Gaber SA, Abujaber R, Al-Majawleh M, Talhouni S. Lipid- and polymer-based nanocarrier platforms for cancer vaccine delivery. ACS Appl Bio Mater . 2024;7(8):4998–5019. doi: 10.1021/ACSABM.3C00843

28. Najdian A, Beiki D, Abbasi M, et al. Exploring innovative strides in radiolabeled nanoparticle progress for multimodality cancer imaging and theranostic applications. Cancer Imaging . 2024;24(1). doi: 10.1186/S40644-024-00762-Z

29. Tong F, Wang Y, Gao H. Progress and challenges in the translation of cancer nanomedicines. Curr Opin Biotechnol . 2024;85. doi: 10.1016/j.copbio.2023.103045

30. Li Y, He J, Liu J, Um W, Ding J. Challenges and opportunities of poly(amino acid) nanomedicines in cancer therapy. Nanomedicine . 2024;19(29):2495. doi: 10.1080/17435889.2024.2402677

31. Shahalaei M, Azad AK, Sulaiman WMAW, et al. A review of metallic nanoparticles: present issues and prospects focused on the preparation methods, characterization techniques, and their theranostic applications. Front Chem . 2024:12. doi: 10.3389/FCHEM.2024.1398979

32. Vasić K, Knez Ž, Leitgeb M. Multifunctional iron oxide nanoparticles as promising magnetic biomaterials in drug delivery: a review. J Funct Biomater . 2024;15(8):227. doi: 10.3390/JFB15080227

33. Fernandes DA. Multifunctional gold nanoparticles for cancer theranostics. 3 Biotech . 2024;14(11). doi: 10.1007/S13205-024-04086-4

34. Wang YL, Lee YH, Chou CL, Chang YS, Liu WC, Chiu HW. Oxidative stress and potential effects of metal nanoparticles: a review of biocompatibility and toxicity concerns. Environ Pollut . 2024;346. doi: 10.1016/j.envpol.2024.123617

35. Nowak-Jary J, Machnicka B. Comprehensive analysis of the potential toxicity of magnetic iron oxide nanoparticles for medical applications: cellular mechanisms and systemic effects. Int J Mol Sci . 2024;25(22):12013. doi: 10.3390/IJMS252212013

36. Zhang W, Liu H, Qiu X, Zuo F, Wang B. Mesoporous silica nanoparticles as a drug delivery mechanism. Open Life Sci . 2024;19(1). doi: 10.1515/BIOL-2022-0867

37. Trayford C, van Rijt S. In situ modified mesoporous silica nanoparticles: synthesis, properties and theranostic applications. Biomater Sci . 2024;12(21):5450–5467. doi: 10.1039/D4BM00094C

38. Rahchamandi SYR, Mirhadi E, Gheybi F, et al. Engineering carbon-based nanomaterials for the delivery of platinum compounds: an innovative cancer disarming frontier. Environ Res . 2024;262(Pt 2):119933. doi: 10.1016/j.envres.2024.119933

39. Parveen A, Chatterjee A, Karak P. Biomedical applications of carbon-based nanomaterials: exploring recent advances in therapeutics, diagnostics, and tissue engineering. Adv Pharm Bull . 2025;15(2):232–247. doi: 10.34172/APB.025.44083

40. Khan MZ, Tahir D, Asim M, Israr M, Haider A, Xu DD. Revolutionizing cancer care: advances in carbon-based materials for diagnosis and treatment. Cureus . 2024;16(1). doi: 10.7759/CUREUS.52511

41. Wang Y, Zhang M, Yan Z, Ji S, Xiao S, Gao J. Metal nanoparticle hybrid hydrogels: the state-of-the-art of combining hard and soft materials to promote wound healing. Theranostics . 2024;14(4):1534–1560. doi: 10.7150/THNO.91829

42. Souto EB, Blanco-Llamero C, Krambeck K, et al. Regulatory insights into nanomedicine and gene vaccine innovation: safety assessment, challenges, and regulatory perspectives. Acta Biomater . 2024;180:1–17. doi: 10.1016/j.actbio.2024.04.010

43. Jiang T, Zhan Y, Ding J, et al. Biomimetic cell membrane-coated nanoparticles for cancer theranostics. ChemMedChem . 2024;19(22). doi: 10.1002/CMDC.202400410

44. Zhang S, Zhang X, Gao H, et al. Cell membrane-coated biomimetic nanoparticles in cancer treatment. Pharmaceutics . 2024;16(4). doi: 10.3390/PHARMACEUTICS16040531

45. Lérida-Viso A, Estepa-Fernández A, García-Fernández A, Martí-Centelles V, Martínez-Máñez R. Biosafety of mesoporous silica nanoparticles; towards clinical translation. Adv Drug Deliv Rev . 2023;201. doi: 10.1016/j.addr.2023.115049

46. Taha MS, Padmakumar S, Singh A, Amiji MM. Critical quality attributes in the development of therapeutic nanomedicines toward clinical translation. Drug Deliv Transl Res . 2020;10(3):766–790. doi: 10.1007/S13346-020-00744-1

47. Csóka I, Ismail R, Jójárt-Laczkovich O, Pallagi E. Regulatory considerations, challenges and risk-based approach in nanomedicine development. Curr Med Chem . 2021;28(36):7461–7476. doi: 10.2174/0929867328666210406115529

48. Halamoda-Kenzaoui B, Baconnier S, Bastogne T, et al. Bridging communities in the field of nanomedicine. Regul Toxicol Pharmacol . 2019;106:187–196. doi: 10.1016/j.yrtph.2019.04.011

49. Mohammadpour R, Dobrovolskaia MA, Cheney DL, Greish KF, Ghandehari H. Subchronic and chronic toxicity evaluation of inorganic nanoparticles for delivery applications. Adv Drug Deliv Rev . 2019;144:112–132. doi: 10.1016/j.addr.2019.07.006

50. Sorrentino C, Ciummo SL, Fieni C, Di Carlo E. Nanomedicine for cancer patient-centered care. MedComm . 2024;5(11). doi: 10.1002/MCO2.767

51. Kozma GT, Shimizu T, Ishida T, Szebeni J. Anti-PEG antibodies: properties, formation, testing and role in adverse immune reactions to PEGylated nano-biopharmaceuticals. Adv Drug Deliv Rev . 2020;154-155:163–175. doi: 10.1016/j.addr.2020.07.024

52. Szebeni J. Complement activation-related pseudoallergy: a stress reaction in blood triggered by nanomedicines and biologicals. Mol Immunol . 2014;61(2):163–173. doi: 10.1016/J.MOLIMM.2014.06.038

53. Nyström AM, Fadeel B. Safety assessment of nanomaterials: implications for nanomedicine. J Control Release . 2012;161(2):403–408. doi: 10.1016/j.jconrel.2012.01.027

54. Mülhopt S, Diabaté S, Dilger M, et al. Characterization of nanoparticle batch-to-batch variability. Nanomater . 2018;8(5):311. doi: 10.3390/NANO8050311

55. Dobrovolskaia MA, Clogston JD, Neun BW, Hall JB, Patri AK, McNeil SE. Method for analysis of nanoparticle hemolytic properties in vitro. Nano Lett . 2008;8(8):2180–2187. doi: 10.1021/NL0805615

56. Bayda S, Adeel M, Tuccinardi T, Cordani M, Rizzolio F. The history of nanoscience and nanotechnology: from chemical–physical applications to nanomedicine. Molecules . 2020;25(1):112. doi: 10.3390/MOLECULES25010112

57. Dobrovolskaia MA, McNeil SE. Understanding the correlation between in vitro and in vivo immunotoxicity tests for nanomedicines. J Control Release . 2013;172(2):456–466. doi: 10.1016/j.jconrel.2013.05.025

58. Cai R, Chen C. The crown and the scepter: roles of the protein corona in nanomedicine. Adv Mater . 2019;31(45). doi: 10.1002/ADMA.201805740

59. Soliman MG, Martinez-Serra A, Antonello G, et al. Understanding the role of biomolecular coronas in human exposure to nanomaterials. Environ Sci Nano . 2024;11(11):4421–4448. doi: 10.1039/D4EN00488D

60. Paris JL, Baeza A, Vallet-Regí M. Overcoming the stability, toxicity, and biodegradation challenges of tumor stimuli-responsive inorganic nanoparticles for delivery of cancer therapeutics. Expert Opin Drug Deliv . 2019;16(10):1095–1112. doi: 10.1080/17425247.2019.1662786

61. Saadh MJ, Mustafa MA, Kumar A, et al. Stealth nanocarriers in cancer therapy: a comprehensive review of design, functionality, and clinical applications. AAPS Pharm Sci Tech . 2024;25(6). doi: 10.1208/S12249-024-02843-5

62. Havelikar U, Ghorpade KB, Kumar A, et al. Comprehensive insights into mechanism of nanotoxicity, assessment methods and regulatory challenges of nanomedicines. Discov nano . 2024;19(1). doi: 10.1186/S11671-024-04118-1

63. Urbán P, Liptrott NJ, Bremer S. Overview of the blood compatibility of nanomedicines: a trend analysis of in vitro and in vivo studies. Wiley Interdiscip Rev Nanomed Nanobiotechnol . 2019;11(3). doi: 10.1002/WNAN.1546

64. de la Harpe Km, Kondiah PPD, Choonara YE, Marimuthu T, Du Toit LC, Pillay V. The hemocompatibility of nanoparticles: a review of cell–nanoparticle interactions and hemostasis. Cells . 2019;8(10):1209. doi: 10.3390/CELLS8101209

65. Morbidelli M, Papini E, Tavano R. Essential protocols for decoding the composition and the functional effects of the nanoparticle protein Corona. Front Nanotechnol . 2024;6:1500567. doi: 10.3389/FNANO.2024.1500567/FULL

66. Shaw JR, Caprio N, Truong N, et al. Inflammatory disease progression shapes nanoparticle biomolecular Corona-mediated immune activation profiles. Nat Commun . 2025;16(1). doi: 10.1038/S41467-025-56210-4

67. Zelepukin IV, Shevchenko KG, Deyev SM. Rediscovery of mononuclear phagocyte system blockade for nanoparticle drug delivery. Nat Commun . 2024;15(1). doi: 10.1038/S41467-024-48838-5

68. Kumar M, Kulkarni P, Liu S, Chemuturi N, Shah DK. Nanoparticle biodistribution coefficients: a quantitative approach for understanding the tissue distribution of nanoparticles. Adv Drug Deliv Rev . 2023;194. doi: 10.1016/j.addr.2023.114708

69. Foroozandeh P, Aziz AA. Insight into cellular uptake and intracellular trafficking of nanoparticles. Nanoscale Res Lett . 2018;13(1). doi: 10.1186/S11671-018-2728-6

70. Zheng M, Chen Z, Xie J, et al. The genetic and epigenetic toxicity of silica nanoparticles: an updated review. Int J Nanomed . 2024;19:13901–13923. doi: 10.2147/IJN.S486858

71. Drasler B, Sayre P, Steinhäuser KG, Petri-Fink A, Rothen-Rutishauser B. In vitro approaches to assess the hazard of nanomaterials. NanoImpact . 2017;8:99–116. doi: 10.1016/J.IMPACT.2017.08.002

72. Kavya Sree B, Kumar N, Singh S. Reproductive toxicity perspectives of nanoparticles: an update. Toxicol Res . 2024;13(3). doi: 10.1093/TOXRES/TFAE077

73. Acharya B, Behera A, Moharana S, Prajapati BG, Behera S. Nanoparticle-mediated embryotoxicity: mechanisms of chemical toxicity and implications for biological development. Chem Res Toxicol . 2025;38(4):521–541. doi: 10.1021/ACS.CHEMRESTOX.4C00472

74. Dugershaw-Kurzer B, Bossart J, Buljan M, et al. Nanoparticles dysregulate the human placental secretome with consequences on angiogenesis and vascularization. Adv Sci . 2024;11(28):2401060. doi: 10.1002/ADVS.202401060

75. Iavicoli I, Fontana L, Nordberg G. The effects of nanoparticles on the renal system. Crit Rev Toxicol . 2016;46(6):490–560. doi: 10.1080/10408444.2016.1181047

76. Soo Choi H, Liu W, Misra P, et al. Renal clearance of quantum dots. Nat Biotechnol . 2007;25(10):1165–1170. doi: 10.1038/NBT1340

77. Hauser PV, Chang HM, Yanagawa N, Hamon MN. Nanomedicine, and the Kidney. Appl Sci . 2021;11(16):7187. doi: 10.3390/APP11167187

78. Khadanga V, Mishra PC. A review on toxicity mechanism and risk factors of nanoparticles in respiratory tract. Toxicology . 2024;504. doi: 10.1016/j.tox.2024.153781

79. Cojocaru E, Petriș OR, Cojocaru C. Nanoparticle-based drug delivery systems in inhaled therapy: improving respiratory medicine. Pharmaceuticals . 2024;17(8):1059. doi: 10.3390/PH17081059

80. Teixeira MI, Lopes CM, Amaral MH, Costa PC. Navigating neurotoxicity and safety assessment of nanocarriers for brain delivery: strategies and insights. Acta Biomater . 2024;189:25–56. doi: 10.1016/j.actbio.2024.09.027

81. Asimakidou E, Tan JKS, Zeng J, Lo CH. Blood-brain barrier-targeting nanoparticles: biomaterial properties and biomedical applications in translational neuroscience. Pharmaceuticals . 2024;17(5):612. doi: 10.3390/PH17050612

82. Huang C, Liu X, Wu Q, et al. Cardiovascular toxic effects of nanoparticles and corresponding molecular mechanisms. Environ Pollut . 2024:356. doi: 10.1016/j.envpol.2024.124360

83. Zalba S, ten Hagen TLM, Burgui C, Garrido MJ. Stealth nanoparticles in oncology: facing the PEG dilemma. J Control Release . 2022;351:22–36. doi: 10.1016/j.jconrel.2022.09.002

84. Su S, Kang PM. Systemic review of biodegradable nanomaterials in nanomedicine. Nanomater . 2020;10(4):656. doi: 10.3390/NANO10040656

85. Ly PD, Ly KN, Phan HL, Nguyen HHT, Duong VA, V NH. Recent advances in surface decoration of nanoparticles in drug delivery. Front Nanotechnol . 2024;6:1456939. doi: 10.3389/FNANO.2024.1456939/FULL

86. Zhong Z, Deng W, Wu J, et al. Cell membrane coated nanoparticles as a biomimetic drug delivery platform for enhancing cancer immunotherapy. Nanoscale . 2024;16(18):8708–8738. doi: 10.1039/D4NR00284A

87. Guo H, Guo M, Xia Z, Shao Z. Membrane-coated nanoparticles as a biomimetic targeted delivery system for tumour therapy. Biomater Transl . 2024;5(1):33–45. doi: 10.12336/BIOMATERTRANSL.2024.01.004

88. Mndlovu H, Kumar P, du Toit LC, Choonara YE. A review of biomaterial degradation assessment approaches employed in the biomedical field. npj Mater Degrad . 2024;8(1):66. doi: 10.1038/s41529-024-00487-1

89. Kumar R, Kommineni N, Aadil KR, et al. Lipid nanoparticle-based mRNA therapeutics for infectious diseases. Int J Pharm . 2026;687:126420. doi: 10.1016/J.IJPHARM.2025.126420

90. Zheng J, Sun Y, Shen Y, Zhou Z. Surface engineering of nanoparticles for precision medicine. Precis Med Eng . 2025;2(3):100037. doi: 10.1016/J.PREME.2025.100037

91. Haidar LL, Bilek M, Akhavan B. Surface bio-engineered polymeric nanoparticles. Small . 2024;20(21). doi: 10.1002/SMLL.202310876

92. Nandhini J, Karthikeyan E, Rajeshkumar S. Eco-friendly bio-nanocomposites: pioneering sustainable biomedical advancements in engineering. Discov Nano . 2024;19(1). doi: 10.1186/S11671-024-04007-7

93. Saxena R, Kotnala S, Bhatt SC, et al. A review on green synthesis of nanoparticles toward sustainable environment. Sustain Chem Clim Action . 2025:6. doi: 10.1016/J.SCCA.2025.100071

94. Sunayana SRV, Pathade P, Tambe BR. Stimuli-responsive nanocarriers: revolutionizing site-specific drug release. Med Pharm J . 2025;4(2):79–99. doi: 10.55940/MEDPHAR2025118

95. Islam S, Ahmed MMS, Islam MA, Hossain N, Chowdhury MA. Advances in nanoparticles in targeted drug delivery–A review. Results Surfaces Interfaces . 2025;19. doi: 10.1016/J.RSURFI.2025.100529

96. Shameli K, Kalali B, Moeini H, Kartouzian A. Lipid-based colloidal nanocarriers for site-specific drug delivery. Colloids Interfaces . 2026;10(1):7. doi: 10.3390/COLLOIDS10010007

97. Krishnan J, Poomalai P, Ravichandran A, Reddy A, Sureshkumar R. A concise review on effect of pegylation on the properties of lipid-based nanoparticles. Assay Drug Dev Technol . 2024;22(5):246–264. doi: 10.1089/ADT.2024.015

98. Makharadze D, Del Valle LJ, Katsarava R, Puiggalí J. The art of pegylation: from simple polymer to sophisticated drug delivery system. Int J Mol Sci . 2025;26(7):3102. doi: 10.3390/IJMS26073102

99. Tabish TA, Pranjol MZI, Jabeen F, et al. Investigation into the toxic effects of graphene nanopores on lung cancer cells and biological tissues. Appl Mater Today . 2018;12:389–401. doi: 10.1016/j.apmt.2018.07.005

100. Savage DT, Hilt JZ, Dziubla TD. In vitro methods for assessing nanoparticle toxicity. Methods Mol Biol . 2019;1894:1–29. doi: 10.1007/978-1-4939-8916-4_1

101. Jaber N, Billet S. How to use an in vitro approach to characterize the toxicity of airborne compounds. Toxicol Vitr . 2024;94. doi: 10.1016/j.tiv.2023.105718

102. Sampaio AR, Maia RF, Ciardulli MC, Santos HA, Sarmento B. Organ-on-chip platforms for nanoparticle toxicity and efficacy assessment: advancing beyond traditional in vitro and in vivo models. Mater Today Bio . 2025;33:102053. doi: 10.1016/J.MTBIO.2025.102053

103. Xu M, Qi Y, Liu G, Song Y, Jiang X, Du B. Size-dependent in vivo transport of nanoparticles: implications for delivery, targeting, and clearance. ACS Nano . 2023;17(21):20825–20849. doi: 10.1021/ACSNANO.3C05853

104. Ozbek O, Genc DE, Ulgen K. Advances in Physiologically Based Pharmacokinetic (PBPK) modeling of nanomaterials. ACS Pharmacol Transl Sci . 2024;7(8):2251–2279. doi: 10.1021/ACSPTSCI.4C00250

105. Reilly RM, Georgiou CJ, Brown MK, Cai Z. Radiation nanomedicines for cancer treatment: a scientific journey and view of the landscape. EJNMMI Radiopharm Chem . 2024;9(1). doi: 10.1186/S41181-024-00266-Y

106. Abdelkader Y, Perez-Davalos L, LeDuc R, Zahedi RP, Labouta HI. Omics approaches for the assessment of biological responses to nanoparticles. Adv Drug Deliv Rev . 2023;200. doi: 10.1016/j.addr.2023.114992

107. Li Y, Vulpe C, Lammers T, Pallares RM. Assessing inorganic nanoparticle toxicity through omics approaches. Nanoscale . 2024;16(34):15928–15945. doi: 10.1039/D4NR02328E

108. Bahl A, Halappanavar S, Wohlleben W, et al. Bioinformatics and machine learning to support nanomaterial grouping. Nanotoxicology . 2024;18(4):373–400. doi: 10.1080/17435390.2024.2368005

109. Kroll A, Pillukat MH, Hahn D, Schnekenburger J. Interference of engineered nanoparticles with in vitro toxicity assays. Arch Toxicol . 2012;86(7):1123–1136. doi: 10.1007/S00204-012-0837-Z

110. Ong KJ, MacCormack TJ, Clark RJ, et al. Widespread nanoparticle-assay interference: implications for nanotoxicity testing. PLoS One . 2014;9(3):e90650. doi: 10.1371/JOURNAL.PONE.0090650

111. Sharifi S, Mahmoud NN, Voke E, Landry MP, Mahmoudi M. Importance of standardizing analytical characterization methodology for improved reliability of the nanomedicine literature. Nano-micro Lett . 2022;14(1). doi: 10.1007/S40820-022-00922-5

112. Leong HS, Butler KS, Brinker CJ, et al. Publisher correction: on the issue of transparency and reproducibility in nanomedicine. Nat Nanotechnol . 2019;14(8):811. doi: 10.1038/s41565-019-0523-x

113. Oberdörster G, Oberdörster E, Oberdörster J. Concepts of nanoparticle dose metric and response metric. Environ Health Perspect . 2007;115(6):A290–A290. doi: 10.1289/EHP.115-A290A

114. Schmid O, Stoeger T. Surface area is the biologically most effective dose metric for acute nanoparticle toxicity in the lung. J Aerosol Sci . 2016;99:133–143. doi: 10.1016/J.JAEROSCI.2015.12.006

115. Sánchez Jiménez A, Puelles R, Perez-Fernandez M, et al. Safe(r) by design guidelines for the nanotechnology industry. NanoImpact . 2022:25. doi: 10.1016/j.impact.2022.100385

116. Mahajan K, Bhattacharya S. The advancement and obstacles in improving the stability of nanocarriers for precision drug delivery in the field of nanomedicine. Curr Top Med Chem . 2024;24(8):686–721. doi: 10.2174/0115680266287101240214071718

117. Pharmacovigilance: post-authorisation | European Medicines Agency (EMA) [homepage on the Internet]. Available from: https://www.ema.europa.eu/en/human-regulatory-overview/post-authorisation/pharmacovigilance-post-authorisation . Accessed April 12, 2026.

118. Postmarketing safety reporting for human drug and biological products including vaccines | FDA. 2001. Available from: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/postmarketing-safety-reporting-human-drug-and-biological-products-including-vaccines . Accessed April 12, 2026.

119. U.S. Food and Drug Administration. DOXIL (doxorubicin hydrochloride liposome injection) prescribing information. Initial U.S. approval: 1995. Available from: https://www.accessdata.fda.gov/drugsatfda_docs/label/2019/050718s055lbl.pdf . Accessed August 11, 2026.

120. U.S. Food and Drug Administration. Abraxane (paclitaxel protein-bound particles for injectable suspension) prescribing information. Initial U.S. approval: 2005. Available from: https://www.accessdata.fda.gov/Drugsatfda_docs/Label/2020/021660s047lbl.Pdf . Accessed August 11, 2026.

121. U.S. Food and Drug Administration. ONPATTRO (patisiran) lipid complex injection prescribing information. Initial U.S. approval: 2018. Available from: https://www.accessdata.fda.gov/drugsatfda_docs/label/2022/210922s013lbl.pdf . Accessed August 11, 2026.

122. U.S. Food and Drug Administration. VYXEOS (daunorubicin and cytarabine) liposome for injection prescribing information. Initial U.S. approval: 2017. Available from: https://www.accessdata.fda.gov/drugsatfda_docs/label/2022/209401s011lbl.pdf . Accessed August 11, 2026.

123. Pallares RM, Barmin RA, Wang A, Kiessling F, Lammers T. Clinical cancer nanomedicines. J Control Release . 2025;377:385. doi: 10.1016/j.jconrel.2025.113991

124. Liu Y, Lin Z, Wang Y, Chen L, Wang Y, Luo C. Nanotechnology in inflammation: cutting-edge advances in diagnostics, therapeutics and theranostics. Theranostics . 2024;14(6):2490–2525. doi: 10.7150/THNO.91394

125. Huang Y, Guo X, Wu Y, et al. Nanotechnology’s frontier in combatting infectious and inflammatory diseases: prevention and treatment. Signal Transduct Target Ther . 2024;9(1). doi: 10.1038/S41392-024-01745-Z

126. Gao J, Li J, Luo Z, Wang H, Ma Z. Nanoparticle-based drug delivery systems for inflammatory bowel disease treatment. Drug Des Devel Ther . 2024;18:2921–2949. doi: 10.2147/DDDT.S461977

127. Yuan F, Da LZ, Li Q, Zeng Y, Zhang G, Li Y. Designing nanoparticles to minimize unintended inflammatory responses: a step toward safer and more effective precision nanomedicine. Nanomedicine . 2025;20(11):1213–1217. doi: 10.1080/17435889.2025.2476377

128. Wang X, Song X, Feng W, Chang M, Yang J, Chen Y. Advanced nanomedicines for treating refractory inflammation-related diseases. Nano-micro Lett . 2025;17(1). doi: 10.1007/S40820-025-01829-7

129. Wilhelm S, Tavares AJ, Dai Q, et al. Analysis of nanoparticle delivery to tumours. Nature Rev Mater . 2016;1(5). doi: 10.1038/natrevmats.2016.14

130. Price LSL, Stern ST, Deal AM, Kabanov AV, Zamboni WC. A reanalysis of nanoparticle tumor delivery using classical pharmacokinetic metrics. Sci Adv . 2020;6(29):eaay9249. doi: 10.1126/sciadv.aay9249

131. Benderski K, Lammers T, Sofias AM. Analysis of multi-drug cancer nanomedicine. Nat Nanotechnol . 2025;20(8):1163–1172. doi: 10.1038/S41565-025-01932-1

132. Mao Y, Xie J, Yang F, Luo Y, Du J, Xiang H. Advances and prospects of precision nanomedicine in personalized tumor theranostics. Front Cell Dev Biol . 2024;12. doi: 10.3389/FCELL.2024.1514399

133. Chou WC, Canchola A, Zhang F, Lin Z. Machine learning and artificial intelligence in nanomedicine. Wiley Interdiscip Rev Nanomed Nanobiotechnol . 2025;17(4). doi: 10.1002/WNAN.70027

134. Khokhlov I, Legashev L, Bolodurina I, Shukhman A, Shoshin D, Kolesnik S. Prediction of dynamic toxicity of nanoparticles using machine learning. Toxics . 2024;12(10):750. doi: 10.3390/TOXICS12100750

135. Campagnolo L, Lacconi V, Filippi J, Martinelli E. Twenty years of in vitro nanotoxicology: how AI could make the difference. Front Toxicol . 2024;6. doi: 10.3389/FTOX.2024.1470439

136. Ahmadi M, Ayyoubzadeh SM, Ghorbani-Bidkorpeh F. Toxicity prediction of nanoparticles using machine learning approaches. Toxicology . 2024;501. doi: 10.1016/j.tox.2023.153697

137. Katsoulakis E, Wang Q, Wu H, et al. Digital twins for health: a scoping review. NPJ Digit Med . 2024;7(1). doi: 10.1038/S41746-024-01073-0

138. Qi Q, Shen Q, Geng J, et al. Stimuli-responsive biodegradable silica nanoparticles: from native structure designs to biological applications. Adv Colloid Interface Sci . 2024:324. doi: 10.1016/j.cis.2024.103087

139. Kim HI, Park J, Zhu Y, Wang X, Han Y, Zhang D. Recent advances in extracellular vesicles for therapeutic cargo delivery. Exp Mol Med . 2024;56(4):836–849. doi: 10.1038/S12276-024-01201-6

140. Sha A, Luo Y, Xiao W, et al. Plant-Derived Exosome-like Nanoparticles: a Comprehensive Overview of Their Composition, Biogenesis, Isolation, and Biological Applications. Int J Mol Sci . 2024;25(22):12092. doi: 10.3390/IJMS252212092

141. Ma X, Tian Y, Yang R, et al. Nanotechnology in healthcare, and its safety and environmental risks. J Nanobiotechnology . 2024;22(1). doi: 10.1186/S12951-024-02901-X

142. Usmani SM, Bremer-Hoffmann S, Cheyns K, et al. Review of new approach methodologies for application in risk assessment of nanoparticles in the food and feed sector: status and challenges. EFSA Support Publ . 2024;21(9):8826E. doi: 10.2903/SP.EFSA.2024.EN-8826

143. Zayerzadeh E, Koohi MK. A review on the importance of standardization in nanotoxicology for promoting safe and sustainable nanotechnology: benefits, challenges, and solutions. Nanomedicine Res J . 2024;9(4):339–347. doi: 10.22034/NMRJ.2024.04.001

144. WHO. Guidelines on the phasing out of animal tests for the quality control of biological products. 2024. Available from: https://cdn.who.int/media/docs/default-source/biologicals/call-for-comments/who-guideline-on-3rs-draft-1-version-29-nov-2024-pc_clean.pdf . Accessed April 27, 2026.

145. Rana SVS. Mechanistic paradigms of immunotoxicity, triggered by nanoparticles - a review. Toxicol Mech Methods . 2025;35(3):262–278. doi: 10.1080/15376516.2024.2431687

146. Dobrovolskaia MA. Charting new frontiers in nanoparticle immunotoxicity: a perspective on current, emerging, and future approaches. Biochem Biophys Res Commun . 2025;777. doi: 10.1016/j.bbrc.2025.152280

147. Prasad R, Ghosh A, Patel V, et al. Voices of nanomedicine: blueprint guidelines for collaboration in addressing global unmet medical needs. ACS Nano . 2025;19(3):2979–2991. doi: 10.1021/ACSNANO.4C13513

148. Tien J, Kenny K, Broom A, Motion A, Bartlett S. Ensuring responsible nanotechnology innovation. NanoEthics . 2025;19(2):14. doi: 10.1007/S11569-025-00478-9

149. Malsch I, Willighagen E, Sánchez Burmester CF, Mody C. Ethical impact assessment of sharing nanosafety data. Open Res Eur . 2024;4:237. doi: 10.12688/OPENRESEUROPE.18345.1/DOI

150. National Nanotechnology Initiative Environmental. Health, and safety research strategy: 2024 update - National Nanotechnology Coordination Office (NNCO). 2024. Available from: https://www.nano.gov/ehsstrategy2024update/ . Accessed April 27, 2026.

Associated Data

Data Availability Statement

No new datasets were generated or analyzed for this narrative review. The evidence discussed in the manuscript is available from the cited literature. Additional information regarding the review synthesis is available from the corresponding author, Dr. Ralph Santos-Oliveira, upon reasonable request.