Macrophage-Based Nanoplatforms for Tumor-Targeted Drug Delivery and Cancer Immunomodulation: Extracellular Vesicles, Membrane-Coated Nanoparticles, and Live Cells

Abstract

Macrophages are highly plastic immune cells with tumor-homing capacity and immunomodulatory functions, making macrophage-based nanoplatforms attractive for cancer immunotherapy. Macrophage-based nanoplatforms include macrophage-derived extracellular vesicles (EVs), macrophage membrane-coated nanoparticles, and live macrophage carriers. The platforms combine the biological properties of macrophages with the advantages offered by nanocarriers to achieve targeted delivery and immune regulation. Beyond serving as delivery vehicles, macrophages can be therapeutically reprogrammed through strategies such as CD47-SIRPα blockade and TLR7/MyD88 pathway activation to enhance antitumor immunity. Despite promising preclinical outcomes, clinical translation of macrophage-based Biomimetic nano-drug delivery systems (BNDDS) remains challenging due to cytokine release syndrome risk, scalable manufacturing limitations, and species differences between animal models and human immunity. The review provides insights into the development and clinical translation of macrophage-based BNDDS for cancer therapy.

Introduction

As a major disease posing a serious threat to human health worldwide, cancer has long been a central focus of medical research, particularly with respect to the mechanisms underlying tumor initiation and progression, as well as the development of effective therapeutic strategies.¹ The TME plays a critical role in tumor development and progression, especially through immune regulation.² As a core component of the innate immune system, macrophages profoundly shape the immune status of the TME through specialized immunomodulatory functions, which has driven macrophage targeting to emerge as a major research focus in contemporary tumor immunotherapy strategies.^3,4

As essential immune cells, macrophages play a central regulatory role within the TME.⁵ Macrophages interact extensively with tumor cells, stromal cells, and other immune cells, allowing them to sense local changes and coordinate communication among different cellular components of the TME.^5–7 Through these interactions, macrophages connect local tumor-associated signals with broader immune responses, making them an important link between the TME and systemic immunity and providing a biological basis for macrophage-targeted cancer immunotherapy.⁵ Of note, macrophages display substantial plasticity and heterogeneity, giving rise to a spectrum of functional states in response to microenvironmental signals, and the M1/M2 polarization model has historically served as a simplified conceptual framework to describe such functional divergence.⁸ M1 macrophages exhibit anti-tumor properties, secreting pro-inflammatory cytokines and toxic substances that enhance immune response to kill tumor cells.⁹ In contrast, M2 macrophages exhibit pro-tumor properties, promoting tumor growth, angiogenesis, metastasis, and immune evasion.¹⁰

Tumor-associated macrophages (TAMs) represent the most abundant immune cell population within TME.¹¹ TAMs primarily originate from circulating monocytes that are recruited into tumor tissues and subsequently differentiate.^12,13 TAMs are activated by Toll-like receptor (TLR) ligands and A2 adenosine receptor agonists.^12,13 Phenotypically and functionally, TAMs predominantly resemble M2-type macrophages.¹⁴ Consequently, TAMs secrete a broad range of growth factors, including epidermal growth factor (EGF) and hepatocyte growth factor (HGF), which directly promote tumor cell proliferation. High-level expression of vascular endothelial growth factor (VEGF) by TAMs effectively induces tumor angiogenesis.^15–17 TAMs further promote tumor invasion and metastasis through the release of matrix metalloproteinases (MMPs) and cathepsins, which degrade cell-cell junctions and extracellular matrix components.^18–20 TAMs secrete anti-inflammatory mediators, such as interleukin-10 (IL-10) and transforming growth factor-β (TGF-β), leading to suppression of T-cell-mediated immune responses.²¹ TAMs also produce immunosuppressive cytokines, including interleukin-6 (IL-6) and IL-10, contributing to tumor cell resistance to apoptosis and reduced sensitivity to chemotherapeutic agents.²² The quantity, phenotype, and functional status of TAMs are closely associated with tumor initiation, progression, and clinical prognosis, highlighting TAMs as promising therapeutic targets in cancer therapy.^23,24 Accordingly, intervention strategies targeting TAM recruitment inhibition and phenotypic reprogramming have emerged as major research directions in tumor immunotherapy.^23,24

Recent advances in anticancer drug development and manufacturing, including synthetic biology-enabled production of etoposide and teniposide precursors in engineered yeast, have improved the sustainability and security of drug supply.²⁵ However, poor selectivity, limited tumor accumulation, systemic toxicity, and drug resistance remain major limitations of conventional anticancer therapy.²⁶ Nanotechnology can partially address these limitations through drug encapsulation, controlled release, and passive targeting.²⁷ However, the effectiveness of EPR-based delivery in human tumors is highly variable and limited by tumor heterogeneity and biological barriers, including dense extracellular matrix, elevated interstitial fluid pressure, and abnormal tumor vasculature. Moreover, the EPR effect observed in preclinical models may be overestimated compared with that in human tumors due to species differences and the complexity of the tumor microenvironment.^27,28 To address the limitations of conventional synthetic nanocarriers, particularly immune clearance and insufficient tumor accumulation, BNDDS have emerged as promising alternatives by combining functional nanomaterials with the biological functions of natural cells or cell-derived components.^26,29,30 Natural tumor-homing, phagocytic activity, and the capacity to cross biological barriers make macrophages attractive biomimetic delivery vehicles for transporting therapeutic agents to tumor sites and regulating the TME.³¹ Macrophage phenotype is an important determinant of delivery performance and therapeutic function.³² In cancer therapy, M1 macrophages are generally more suitable for delivery systems designed to enhance antitumor immune responses because of their tumor-homing ability, efficient nanoparticle uptake, and tumor penetration.^31,33,34 In contrast, M2-like macrophages, particularly TAMs, have immunosuppressive and tumor-promoting functions and are therefore more commonly used as therapeutic targets for repolarization in cancer treatment.³¹

Macrophage-based delivery platforms have shown increasing potential for cancer therapy. Miao et al³⁵ developed M1 macrophage-derived EVs loaded with CX3CR1 siRNA, which inhibited pancreatic tumor growth, supporting the potential of macrophage-derived EVs for nucleic acid delivery. Hua et al³⁶ constructed M1 macrophage membrane-coated photodynamic nanoparticles that enhanced tumor accumulation and induced ferroptosis in non-small cell lung cancer models. Engineered live macrophages have progressed further toward clinical translation. In a Phase 1 trial, the anti-HER2 CAR-macrophage product CT-0508 showed preliminary safety and tumor trafficking in patients with advanced HER2-overexpressing solid tumors,³⁷ while mesothelin-targeting CAR macrophages also showed favorable preliminary safety in ovarian cancer.³⁸ Overall, EVs- and membrane-based platforms remain largely preclinical, whereas engineered live macrophages have entered early clinical evaluation.

Although recent reviews have addressed individual aspects of macrophage biology and therapeutic applications, a comprehensive cancer-focused framework integrating macrophage ontogeny, phenotypic heterogeneity, polarization regulation, and biomimetic delivery platforms remains lacking.^{31,32,39–42} To address the gap, the present review summarizes macrophage ontogeny, phenotypes, plasticity, and polarization regulation, followed by a systematic comparison of three major platforms: macrophage-derived EVs, macrophage membrane-coated nanoparticles, and live macrophage carriers (Figure 1). The review highlights the construction strategies, advantages, and translational challenges of the three platforms. By linking macrophage immunobiology with biomimetic delivery, the present review provides a framework for the design and clinical translation of macrophage-based BNDDS for cancer therapy.

Figure 1.

The article is organized into four main sections: macrophage origins, macrophage phenotypes, strategies for regulating phenotypic polarization, and macrophage-based BNDDS. Created with BioRender.com.

Source of Macrophages

Macrophages are widely distributed across various tissues and are recognized as among the earliest immune cells to emerge during evolution.⁴³ A prevailing view holds that macrophages are derived from circulating monocytes originating from myeloid progenitor cells in the bone marrow.⁴⁴ Under conditions such as tissue injury, pathogenic infection, or carcinogenesis, circulating monocytes are rapidly recruited to affected sites and subsequently undergo extensive differentiation into macrophages.⁴⁵ In recent years, accumulating evidence has revealed an additional developmental pathway involving embryonic progenitor cells originating from the yolk sac or fetal liver.^46–48 For instance, microglia within the central nervous system, alveolar macrophages within the lungs, and Kupffer cells within the liver, are tissue-resident macrophages that primarily originate from progenitor cells in the yolk sac or fetal liver during the embryonic stage.⁴⁸

Macrophages from different sources exhibit distinct biological characteristics. In tumors, embryonically derived tissue-resident macrophages (TRMs) and monocyte-derived tumor-associated macrophages (MoTAMs) differ in origin, plasticity, and therapeutic response.^45,49,50 TRMs can self-renew locally and retain tissue-specific programs, whereas MoTAMs arise from recruited circulating monocytes and show greater phenotypic plasticity in response to the TME.^50,51 Macrophage source can influence cellular responses and therapeutic sensitivity.⁵² TRMs and monocyte-derived macrophages respond differently to particle uptake and anti-CSF1R treatment.^49,53 Therefore, macrophage source should be considered when selecting cells and designing therapeutic strategies for BNDDS.^54,55

Primary Macrophages

Primary macrophages are widely applied in tumor research and therapeutic development and constitute an indispensable experimental model for investigating the biological functions of TAMs.^56,57 By culturing primary macrophages in vitro under TME-mimicking conditions, key signaling pathways governing macrophage polarization can be systematically elucidated, thereby facilitating the identification of molecular targets for suppressing protumor activities of TAMs.^56,57 In addition, primary macrophages serve as an important screening platform for the development and evaluation of cancer immunotherapeutic agents.^58,59 For instance, primary macrophage models are widely employed to assess TAM reprogramming strategies, allowing detailed evaluation of drug-induced alterations in macrophage phenotypes and cytokine secretion profiles.^58,59

Commonly used primary macrophages include BMDMs, peritoneal macrophages (PMs), splenic macrophages (SPMs), and alveolar macrophages (AMs).⁶⁰ Owing to low intrinsic immunogenicity and favorable biocompatibility, primary macrophages have emerged as highly promising carriers for drug delivery applications.⁶⁰ Unlike tissue-resident primary macrophages such as PMs, SPMs, and AMs, BMDMs are generated ex vivo from bone marrow progenitors under defined differentiation conditions.⁶¹ The phenotype and functions of BMDMs can be experimentally regulated by specific stimuli, including changes in polarization, cytokine secretion, migration, and phagocytic activity.⁵⁷ This experimental controllability makes BMDMs particularly suitable for mechanistic studies of macrophage polarization and therapeutic engineering.⁶¹ Shields et al⁶² employed BMDMs as carriers to anchor interferon-γ-loaded polymer “backpacks” onto BMDMs via adhesion-mediated interactions, achieving targeted breast cancer therapy. Building on this delivery capability, BMDMs have also been explored for deep intratumoral transport of therapeutic payloads while maintaining antitumor functions. Evans and co-workers⁶³ further demonstrated that BMDMs could function as “cellular Trojan horses” to deliver poly(lactic-co-glycolic acid) (PLGA) microspheres deep into tumor tissues. Nevertheless, instability associated with genotypic modification and proliferative limitations remains a major challenge for the application of BMDMs.⁶¹

Increasing attention has also been directed toward PMs as nanoparticle carriers in recent years.⁶⁴ For instance, co-incubation of PMs loaded with SN38 (7-ethyl-10-hydroxycamptothecin) nanoparticles with lung cancer cells could promote the apoptosis of lung cancer cells.⁶⁵ Murine peritoneal macrophages are commonly used to load doxorubicin (DOX)-encapsulated liposomes. After intravenous administration, drug-loaded macrophages undergo tumor homing and accumulate in tumor tissues under the guidance of tumor-secreted chemokine gradients like CCL2, enabling DOX to be detected in tumor tissues within 24 h.⁶⁶ Such findings also indicate robust in vivo migratory capacity of murine peritoneal macrophages. Lv et al⁶⁷ constructed a macrophage-mediated biomimetic delivery system by co-incubating DOX-loaded natural egg yolk lipid-derived nanovectors with purified PMs. Similarly, macrophage-based delivery platforms were further advanced by integrating genetic engineering strategies to enhance tumor specificity and therapeutic efficacy. Dong et al⁶⁸ designed an innovative chimeric antigen receptor macrophage made from genetically modified human PMs, which was made from genetically modified human PMs and loaded with oxaliplatin for use in gastric cancer treatment research. Rat AMs were also shown to encapsulate gold-silica nanoshells and combined with photothermal therapy (PTT) for the treatment of glioblastoma.⁶⁹ Furthermore, Nguyen et al⁷⁰ further employed primary mouse SPMs as carriers for citric acid-modified superparamagnetic nanoparticles (CA-MNPs) and thermosensitive liposomes encapsulating doxorubicin (DOX-TSLPs). By leveraging the intrinsic chemotactic properties of SPMs along with the dual-targeting effect of an external magnetic field, the tumor targeting efficiency of the nanoparticles was enhanced. In brief, in the preclinical studies on various tumor models, primary macrophages achieved precise drug delivery and improved therapeutic efficacy by loading a variety of functional nanomaterials and combining targeting and controlled release technologies.

Macrophage Cell Line

THP-1 is a widely used human monocytic cell line that can be differentiated into macrophage-like cells and polarized into distinct functional phenotypes, with advantages of easy culture, reproducibility, and scalability.^71,72 THP-1-derived macrophages have been applied for the generation of macrophage-derived EVs as natural drug delivery carriers.⁷³ For example, THP-1-derived macrophage EVs were engineered with surface A15 expression and used for co-delivery of doxorubicin and cholesterol-modified miR-159, achieving enhanced uptake and synergistic antitumor effects in triple-negative breast cancer models.⁷³ U937 is a less commonly used human monocytic cell line that can also differentiate into macrophage-like cells upon stimulation and is mainly used to study monocyte differentiation and macrophage function.⁷⁴

Commonly used murine macrophage cell lines include J774A.1 and RAW264.7.⁶⁰ The murine macrophage cell line RAW264.7 is widely used to investigate macrophage functions, mechanisms, and signaling pathways.⁷⁵ Compared with primary macrophages, RAW264.7 offers several advantages, including higher biosafety, faster proliferation rates, lower phenotypic variability, and easier acquisition and cultivation.⁷⁶ In addition, RAW264.7 cells are generally easier to obtain and maintain than THP-1 cells.⁷⁶ The RAW264.7 cell line is a murine leukemia monocyte-macrophage cell line that was first isolated and established in 1978 by Raschke et al from male BALB/c mice.⁷⁷

The RAW264.7 cell line is not only a widely used macrophage model but also serves as a potential carrier for targeted delivery of anti-tumor drugs.^60,76 Zhang et al⁷⁸ demonstrated that RAW264.7 macrophages could be loaded with DOX, reaching a maximum drug content of 16.6 pg per cell. Building on the demonstrated drug-loading capacity of RAW264.7 macrophages, factors influencing drug-loading efficiency have also been systematically investigated. Li et al⁷⁹ systematically compared the phagocytic behavior of RAW264.7 macrophages toward different drug formulations and revealed that lipophilic agents and nanoparticle-encapsulated drugs were internalized to a significantly greater extent than hydrophilic small-molecule drugs. Elucidation of formulation-dependent uptake behavior further promoted the development of macrophage-based biomimetic nanocarriers for tumor-targeted therapy. Qiang et al⁸⁰ employed RAW264.7 macrophages as delivery vehicles for DOX-loaded reduced graphene oxide (rGO) nanosheets targeting murine prostate cancer cells. The macrophage-associated rGO nanocomplexes exhibited pronounced tumor-homing capability, highlighting the intrinsic tumor tropism of macrophages and demonstrating the feasibility of integrating nanomaterials with cell-mediated delivery strategies. Physicochemical optimization of drug formulations, together with the inherent tumor-homing capacity of macrophages, collectively highlights the significant potential of macrophage-based platforms in the development of advanced targeted drug delivery systems.

However, the translational limitations of RAW264.7 cells should be considered.⁷⁷ As a murine macrophage-like tumor cell line, RAW264.7 cells may exhibit immunological behaviors that differ from those of primary human macrophages. Species-specific differences in immune signaling, receptor expression, and tumor–immune interactions may further limit the extrapolation of murine findings to humans.^77,81 These differences may also affect nanoparticle recognition and uptake in BNDDS.⁸² Therefore, findings obtained using RAW264.7 cells should be validated in primary human macrophages or other clinically relevant models before clinical translation.^77,83 The characteristics and translational potential of different macrophage sources and models for BNDDS are summarized in Table 1.

Table 1.

Advantages, Limitations, and Clinical Translational Potential of Different Macrophage Sources for Macrophage-Based BNDDS

Macrophage Sources	Key Advantages	Major Limitations	Clinical Translational Potential	Ref.
Bone marrow-derived macrophages	Nanoparticle uptake capacity; migration preservation; ex vivo engineering feasibility	Invasive acquisition; cell heterogeneity; limited clinical validation	Engineered cell carriers; cargo delivery applications; limited accessibility	[84]
Peripheral blood monocyte-derived macrophages	Easy accessibility; high cell yield	Isolation requirement; ex vivo differentiation; limited tissue specificity	Patient-derived BNDDS; personalized therapy potential; clinical feasibility	[53]
Peritoneal macrophages	High accessibility; phagocytic activity; engineering feasibility	Limited proliferation; disease-dependent availability; restricted application	Peritoneal tumor therapy; patient-derived macrophage therapy; local delivery potential	[67]
Splenic macrophages	Phagocytic capacity; tumor-homing ability; nanoparticle loading	Difficult acquisition; limited human evidence; uncertain translation	Preclinical therapy; tumor-targeted delivery	[69]
Alveolar macrophages	Particle uptake; lung localization; phenotype plasticity	Environmental sensitivity; functional heterogeneity; limited validation	Lung-targeted BNDDS; respiratory disease applications; need for further validation	[68]
RAW264.7 macrophages	Easy culture; rapid proliferation; genetic reproducibility	Murine origin; immunogenicity concerns; limited human relevance	Mechanistic studies; drug-delivery evaluation; unsuitable for direct therapy	[75, 76, 78]

Strategies for Regulating the Phenotype and Polarization of Macrophages

Macrophage phenotype and polarization are central determinants of macrophage function in the TME and provide an important biological basis for macrophage-centered cancer therapy.^42,85 The marked heterogeneity and plasticity of TAMs enable dynamic responses to local microenvironmental signals and influence the therapeutic regulation of macrophage functions.^86,87 The following discussion summarizes macrophage phenotypic heterogeneity and plasticity in the TME and discusses representative strategies for regulating macrophage polarization, including phagocytic particles, surface modification, and genetic modification.

The Phenotype of Macrophages

Macrophages comprise a highly heterogeneous and plastic immune cell population whose phenotypes and functions are dynamically shaped by cytokine networks, metabolic cues, and microenvironmental signals within tumors.^88,89 During tumor development and progression, TAMs can arise from either tissue-resident macrophages or circulating monocytes and acquire diverse functional characteristics in response to local environmental signals.^85,88 To facilitate functional description, macrophage activation has historically been summarized within the M1 and M2 polarization framework, in which M1 macrophages are generally associated with anti-tumor activity, whereas M2 macrophages are more frequently linked to immunosuppression, tissue remodeling, and tumor progression.^85,90

Non-polarized macrophages (M0) polarize toward the M1 phenotype in response to pro-inflammatory cues, including lipopolysaccharide (LPS), interferon-γ (IFN-γ), and tumor necrosis factor-α (TNF-α).⁹¹ M1 macrophages exhibit well-recognized antitumor functions.⁹² Firstly, in the TME, M1 macrophages can eliminate malignant cells via direct phagocytosis.⁹³ Secondly, M1 macrophages generate reactive oxygen species (ROS) and upregulate inducible nitric oxide synthase (iNOS), which drives nitric oxide (NO) production. Both ROS and NO can directly exert cytotoxic effects to kill tumor cells.^94,95 M1 macrophages also secrete interleukin-12 (IL-12), promoting activation of T cells and natural killer (NK) cells and strengthening antitumor immune responses.⁹⁵ Additionally, M1 macrophages facilitate NF-κB pathway activation by suppressing expression of the NF-κB p50 subunit, thereby increasing the release of immunostimulatory cytokines such as TNF-α and IL-1β.⁹⁶ Such cytokines promote reprogramming of M2 macrophages toward an M1 state, further reinforcing T cell-mediated antitumor immunity. Finally, M1 macrophages promote differentiation of naive T cells into T helper 1 (Th1) cells and cytotoxic T lymphocytes (CTLs), augmenting Th1-skewed immunity and cytotoxic effector function and contributing to tumor control.⁹⁷

M2 macrophages arise from non-polarized macrophages (M0) in response to interleukin-4 (IL-4), interleukin-10 (IL-10), and interleukin-13 (IL-13).⁹⁸ Within the conventional M2 framework, M2 macrophages can be further classified into M2a, M2b, M2c, and M2d subtypes according to their activating stimuli and functional characteristics, with M2d macrophages being particularly associated with tumors.⁹⁹ Within the TME, M2 macrophages function as central orchestrators of tumor progression by reinforcing immunosuppression, promoting angiogenesis, and driving stromal remodeling.¹⁰⁰ Signal regulatory protein α (SIRPα) on macrophages engages CD47, which is frequently overexpressed by tumor cells, thereby strengthening the “don’t eat me” signal and suppressing macrophage-mediated phagocytosis.^101,102 Additional immune-evasion checkpoints, including the CD24/sialic acid-binding Ig-like lectin 10 (Siglec-10) axis and the programmed cell death protein 1 (PD-1)/programmed death ligand 1 (PD-L1) axis, suppress macrophage phagocytic capacity and facilitate tumor immune escape.¹⁰³ M2 macrophages exhibit impaired antigen-presenting function, attenuating T-cell recognition of tumor antigens and constraining activation of adaptive antitumor immunity.¹⁰⁰ M2 macrophages also produce inhibitory mediators, including IL-10 and transforming growth factor-β (TGF-β), which suppress effector programs in CD8⁺ T cells and natural killer (NK) cells and thereby weaken cytotoxic elimination of malignant cells.^103,104 Moreover, M2 macrophage-derived C-C motif chemokine ligand 22 (CCL22) preferentially recruits immunosuppressive regulatory T cells into the TME, further intensifying local immune suppression.¹⁰⁵ In parallel, M2 macrophages stimulate vascular endothelial cell expansion through secretion of pro-angiogenic factors like vascular endothelial growth factor (VEGF), promoting neovascularization.¹⁰⁶ M2 macrophages additionally release proteolytic and matrix-remodeling factors, including cathepsin B, TNF-α, and MMPs, which degrade basement membrane components and peritumoral collagen, compromise structural tissue constraints, and facilitate tumor invasion.^105,107

Macrophage polarization is closely linked to metabolic reprogramming.^108,109 M1 macrophages generally rely more on aerobic glycolysis to support pro-inflammatory functions, whereas M2 macrophages preferentially use mitochondrial oxidative phosphorylation (OXPHOS) and fatty acid oxidation (FAO).¹⁰⁸ IL-4-induced M2 polarization is also associated with enhanced TCA cycle activity, glutamine metabolism, and OXPHOS, while inhibition of glutamine metabolism or mitochondrial OXPHOS can attenuate M2 polarization.¹⁰⁸ In the TME, however, these metabolic patterns are not absolute, as glycolysis and lactate accumulation can also promote M2-like immunosuppressive states.¹⁰⁹ Therefore, macrophage immunometabolism represents an important layer of plasticity linking microenvironmental signals to macrophage polarization and function.¹⁰⁹

Notably, increasing in vivo evidence indicates that macrophage states within the TME are highly dynamic and cannot be adequately described by the traditional M1/M2 classification.^84,99,110 Single-cell transcriptomic studies have shown that individual TAMs can co-express canonical M1- and M2-like genes, indicating that the traditional M1/M2 classification does not fully capture TAM heterogeneity.⁸⁷ For example, SPP1⁺ TAMs are associated with angiogenesis, matrix remodeling, and tumor metastasis, whereas C1Q⁺ TAMs are mainly involved in immune regulation and immunosuppression.¹¹¹ Therefore, M1 and M2 are better regarded as representative states within a broader functional spectrum rather than strictly separated macrophage subtypes.^90,99 Because TAM gene signatures can overlap, their classification should also consider functional characteristics and spatial context.⁸⁷ Spatial transcriptomic studies have further shown that TAM phenotype and function vary according to their location within tumors. TAMs in hypoxic or necrotic regions, perivascular regions, and invasive margins exhibit distinct molecular and functional characteristics.¹¹² For example, spatial analysis of colon cancer showed that IL4I1⁺ macrophages were associated with tumor cell phagocytosis and favorable prognosis, whereas SPP1⁺ macrophages were preferentially localized in hypoxic and necrotic tumor regions and were associated with poor prognosis.¹¹² These spatial differences may also influence BNDDS performance because vascular accessibility, hypoxia, and local microenvironmental conditions can affect nanoparticle delivery, uptake, and macrophage reprogramming.^113,114 Therefore, both TAM subtype and spatial localization should be considered when designing macrophage-targeted BNDDS to improve therapeutic efficacy.

Strategies for the Regulation of Macrophage Phenotypic Polarization

Monocyte-derived macrophages have the potential to be activated upon stimulation, a process known as polarization.¹¹⁵ Diverse microenvironmental stimuli drive macrophage polarization toward multiple phenotypic states, enabling specialized roles in physiological and pathological contexts.⁸⁶ Macrophage polarization is highly dynamic, and macrophage phenotypes can interconvert in response to local environmental signals.^116,117 Notably, macrophage activation states form a dynamic continuum rather than discrete categories, and M2-like TAMs can be reprogrammed toward an M1-like state without necessarily passing through an M0 intermediate state in response to altered microenvironmental or therapeutic signals.⁸⁷

Within the TME, directing macrophage polarization toward antitumor states and preventing polarization toward protumor states represents a feasible and increasingly important strategy in tumor immunotherapy.^118,119 The conversion of TAMs from the M2 phenotype to the M1 phenotype may also trigger anti-tumor immunity and inhibit tumor metastasis.¹²⁰ From the perspective of tumor angiogenesis, inhibiting the polarization of M2-type macrophages can block the formation of new blood vessels around tumors, thereby contributing to tumor treatment.^121,122 Different phenotypes of macrophages in the TME exerted varying impacts on tumor progression.¹⁰³ The polarization balance between the M1 and M2 phenotypes of macrophages plays a pivotal role in tumor development and treatment.¹⁰³ Consequently, induction of macrophage programs with optimal antitumor activity provides a key entry point for macrophage-centered cancer therapy. The following section summarizes three representative approaches for modulating macrophage phenotypes: phagocytic particles, surface modification, and genetic modification. A comparative summary of these strategies in terms of representative approaches, therapeutic efficacy, safety, and translational potential is provided in Table 2.

Table 2.

Comparison of Macrophage Polarization Strategies in Terms of Therapeutic Efficacy, Safety, and Translational Potential

Strategies	Representative Approaches	Therapeutic Efficacy	Safety Considerations	Translational Potential
Phagocytic particles	Inorganic, polymeric, hybrid, and biomimetic particles carrying polarization regulators	M2-to-M1-like repolarization; TME remodeling	Targeted delivery; reduced systemic toxicity; off-target accumulation risk	Predominantly preclinical development; long-term safety assessment; manufacturing optimization
Surface modification	CSF1R, TREM2, integrin β3, CD40 modulation; macrophage surface engineering	Polarization regulation; enhanced antitumor macrophage activity; immune activation	Systemic receptor modulation; normal macrophage disruption; potential adverse effects	Clinical evaluation of receptor-targeting approaches; preclinical surface engineering
Genetic modification	siRNA, plasmid DNA, mRNA, CRISPR-Cas9, and CAR-M engineering	Precise macrophage reprogramming; sustained functional modulation; enhanced antitumor responses	Off-target effects; immune responses; delivery-related risks	Preclinical gene-editing approaches; emerging CAR-M clinical evaluation

Phagocytic Particles

Macrophages possess intrinsic phagocytic activity, which provides a biological basis for the use of phagocytic particles to regulate macrophage polarization.¹²³ Phagocytic Particles are designed to exploit macrophage uptake for the delivery of polarization-regulating agents, thereby promoting TAM polarization toward antitumor phenotypes.^118,123 Based on material composition and structural design, phagocytic particles used for macrophage polarization can be broadly categorized into inorganic particles, organic or polymeric particles, and hybrid or biomimetic particles.

Inorganic particles can regulate macrophage polarization through stimulus-responsive cargo release or modulation of the TME. For example, Sui et al¹²⁴ developed calcium carbonate-coated quercetin-manganese nanoparticles, which degraded and released active components under the stimulation of the acidic TME and high glutathione, thereby reprograming TAMs from the M2 phenotype to the M1 phenotype. Oxygen-generating nanoparticles represented by manganese dioxide (MnO₂) regulated the polarization of TAMs by ameliorating the hypoxic TME, which served as an effective strategy for achieving synergistic tumor immunotherapy.¹²⁵ For example, manganese dioxide nanoparticles modified with mannan or hyaluronic acid could specifically target M2-type TAMs, induced the reprogramming of M2-type TAMs toward the anti-tumor M1 phenotype via activating TLR2 or TLR4 signaling pathways, thereby enhancing the synergistic effect of tumor immunotherapy and chemotherapy.^125,126

Vesicular and polymeric particles primarily serve as carriers for polarization-regulating agents and facilitate intracellular delivery following macrophage uptake. Liposomes and polymeric vesicles also serve as commonly used vehicles for transporting cytokines, chemokines, and Toll-like receptor (TLR) agonists to modulate TAM polarization.¹²⁷ For example, Rodriguez-Perdigon and colleagues¹²⁷ reported that polymer vesicles encapsulating a colony-stimulating factor 1 receptor inhibitor were efficiently internalized by M2-like macrophages and drove repolarization toward an M1-like phenotype by sustained suppression of the CSF1/CSF1R signaling axis.

Targeted regulation of TAMs into the M1 phenotype often contributes to the improvement of the tumor immune microenvironment.¹²⁸ Han et al¹²⁸ developed an M2-type TAMs-targeting nano-complex and achieved selective recognition and delivery to M2-like TAMs by dual-peptide functionalization of the nanoparticle surface with M2pep and an SR-B1-targeting α-peptide. The nano-complex could trigger the disintegration of polydopamine in an acidic lysosomal environment to release baicalin, Hgp antigen, and CpG, thereby inducing the polarization of M2-type TAMs to M1-type and remodeling the TME. Beyond receptor-mediated targeting of M2-like TAMs, other delivery strategies have also exploited the intrinsic phagocytic properties of TAMs to promote macrophage reprogramming. Zhou ‘s group¹²⁹ encapsulated imiquimod using a hydrolyzable fibrous matrix. The rod-like particles with a length less than 3 μm formed after degradation of the matrix could be selectively phagocytosed by TAMs, and then released imiquimod to stimulate the toll-like receptor 7 (TLR7) pathway, thereby achieving the reprogramming of M2-like TAMs into the M1 phenotype and the remodeling of the tumor immune microenvironment. Flow cytometry and immunofluorescence staining revealed an elevation in the proportions of CD8⁺ CTL cells and NK cells, confirming the significant enhancement of immune effector cell infiltration and activation, including T cells and NK cells, in the TME.^128,129

Hybrid or biomimetic particles integrate multiple functional components, such as targeting ligands, therapeutic agents, or cell membrane coatings, to enhance selective uptake by TAMs and promote macrophage repolarization. Macrophages are highly sensitive to material interfaces, where everything from surface charge to particle shape can influence the signaling related to macrophage polarization.^123,130 Zhang and co-workers¹³¹ constructed engineered microparticles (MPs) modified with M2 macrophage-targeting peptide (M2pep). After specific modification, the engineered MPs were easily captured and engulfed by M2 TAMs. The engineered MPs reprogrammed immunosuppressive M2 TAMs into M1 phenotype through the TLR7/MyD88-dependent signaling pathway.

Modified nanoparticles can synergistically drive macrophage polarization from the M2 to the M1 phenotype, while also offering additional antitumor benefits beyond repolarization.^132,133 Gong et al¹³² designed a dual-functional nanoparticle loaded with the photosensitizer IR820 and the SHP2 inhibitor SHP099. After targeting TAMs, the nanoparticle reprogrammed the TAMs from the M2 phenotype to the M1 phenotype by generating intracellular ROS under laser irradiation via IR820, while restoring the ability of TAMs to phagocytose tumor cells by inhibiting SHP2 to block the CD47-SIRPa pathway. Beyond photoresponsive drug delivery, macrophage reprogramming can also be integrated with CD47 blockade through genetically engineered membrane-coated nanoplatforms. Rao and colleagues¹³³ developed the genetically edited cell membrane-coated magnetic nanoparticles. The outer layer of the aforementioned nanoparticles could block CD47 on cancer cells, thereby activating macrophages to phagocytose cancer cells. After being phagocytosed by TAMs, the inner core of the nanoparticles could induce the repolarization of TAMs from the M2 type to the M1 type by mediating Fenton reactions to generate ROS. In addition to ROS-mediated repolarization and CD47 blockade, selective targeting of M2-type TAMs can be further enhanced through peptide modification and biomimetic membrane coating. For example, Wang ‘s group¹³⁴ employed targeting peptides and chitosan for the co-modification of curcumin nanoparticles, and successfully reprogrammed M2-type TAMs into the M1 type by regulating the COX-2 gene and modulating the TNF and IL-17 signaling pathways. Such a dual-targeted modification enhanced the specificity and efficiency of the nanoparticles being phagocytosed by M2-type TAMs. Han and co-workers¹³⁵ not only utilized M2pep peptide for targeted modification but also employed liver cancer cell membrane-coated nanoparticles. After being preferentially taken up by M2 TAMs, the liver cancer cell membrane-coated nanoparticles induced M2 to M1 polarization by regulating the intracellular PI3K/Akt and NF-κB pathways, while simultaneously remodeling the immunosuppressive TME.

Phagocytic particles can also be used to deliver metabolic modulators to regulate macrophage immunometabolism. Lu et al¹³⁶ developed succinate-loaded tumor cell-derived microparticles that reprogrammed TAM metabolism and promoted M1-like polarization through protein succinylation. Similarly, Luo et al¹³⁷ designed a macrophage membrane-coated “Spark-Relay” nanoinitiator that sequentially increased ROS and NO levels in TAMs, shifting their metabolism from OXPHOS toward glycolysis and promoting an antitumor phenotype. This metabolic reprogramming was accompanied by increased CD8⁺ T-cell infiltration and enhanced antitumor activity. These studies demonstrate that BNDDS can couple metabolic reprogramming with TAM repolarization to enhance antitumor responses.^136,137

Surface Modification

Surface modification can regulate macrophage polarization by altering the expression or activity of receptors on the macrophage membrane and by interfering with receptor-ligand interactions.¹³⁸ Based on the mode of surface intervention, surface modification strategies can be broadly categorized into modulation of surface receptor expression, blockade or inhibition of surface receptor signaling, and activation of immunostimulatory surface receptors.

Modulation of surface receptor expression can directly alter macrophage polarization. Zhang et al¹³⁹ demonstrated that C1qbp knockdown in macrophages co-cultured with tumor cells promoted a shift toward an M1-like phenotype. Recent studies have also identified TREM2 as an important surface receptor associated with immunosuppressive macrophage programs. TREM2 loss of function in tumor-associated myeloid cells enhanced proinflammatory polarization and antitumor activity, indicating that modulation of surface receptor expression can promote macrophage reprogramming.¹⁴⁰

Blockade or inhibition of surface receptor signaling represents another major strategy for suppressing M2-like polarization and protumor macrophage functions. For example, CSF1R is a key receptor involved in the survival and polarization of M2-type TAMs. The CSF1R inhibitor PLX3397 can block the CSF1R signaling pathway, thereby inhibiting the polarization, survival, and chemotaxis of M2-type TAMs and reducing the number of M2-type TAMs in the TME.¹⁴¹ Beyond CSF1R-mediated signaling, other polarization-associated surface receptors have also been identified as regulators of M2 TAM differentiation and function. Integrin β3 is highly expressed in M2 TAMs and can mediate M2 polarization by regulating the expression and activation of peroxisome proliferator-activated receptor-γ (PPARγ).¹⁴² Integrin β3 inhibitors such as triptolide can downregulate PPARγ expression and transcriptional activity, thereby inhibiting TAM polarization toward the M2 phenotype.¹⁴² Dendritic cell-associated C-type lectin-1 (Dectin-1) on the macrophage surface can specifically recognize and bind galectin-9, thereby promoting macrophage polarization toward the M2 phenotype. Therefore, blockade of the Dectin-1/galectin-9 axis can suppress M2 polarization and contribute to improvement of the immunosuppressive TME.¹⁴³

Activation of immunostimulatory surface receptors can promote macrophage reprogramming toward antitumor phenotypes. In pancreatic tumor models, CD40 agonists reduced the proportion of immunosuppressive CD206⁺PD-L1⁺ macrophages and increased the proportion of proinflammatory IA/IE⁺F4/80⁺ macrophages. CD40 agonists also upregulated genes associated with antigen presentation and proinflammatory responses and promoted TAM reprogramming toward an M1-like phenotype.¹⁴⁴ In addition to receptor modulation, direct engineering of the macrophage surface can also influence macrophage phenotype. Wang et al¹⁴⁵ conjugated functional nanoparticles onto macrophage surfaces through bioorthogonal chemistry, and the engineered macrophages maintained an M1-like phenotype in the TME. Metabolic glycoengineering of the macrophage surface has also been used to remodel macrophage surface properties, resulting in sustained macrophage activation and enhanced phagocytic and tumoricidal functions.¹⁴⁶

Genetic Modification

Genetic modification enables precise macrophage reprogramming by regulating key genes involved in polarization, immune activation, and tumor-associated functions. Compared with transient stimulation, genetic strategies allow targeted and durable modulation of macrophage phenotypes. Recent advances in genome editing, nucleic acid delivery, and engineered macrophage technologies have expanded their potential for cancer immunotherapy and biomimetic delivery applications.^147–153

Macrophages can be reprogrammed via genetic modification to restore or activate specific functions associated with anti-tumor activity.¹⁵⁴ For example, the polarization of TAMs towards the M1 phenotype could be achieved via CRISPR/Cas9-mediated genome editing technology.¹⁵⁵ The polarization of TAMs toward the M1 phenotype is often accompanied by extensive transcriptional reprogramming. Such a process is mainly driven by specific signaling pathways and transcription factors, among which the activation of STAT1 is of crucial importance.^156,157 Li and colleagues⁷⁵ found that AZD5153 reset TAMs from M2-type to M1-type macrophages by inhibiting the expression of the transcription factor MAF.

Macrophages respond to environmental signals such as hypoxia and cytokines, which are modulated by intricate signaling pathways, including transcriptional regulatory mechanisms.^158,159 For example, the activation of the TLR3 signaling pathway could promote the polarization of TAMs towards the M1 type.^158,159 MicroRNAs, DNA methylation, and histone modifications played significant roles in macrophage polarization by regulating gene expression in response to environmental signals.^160–162 As a class of small non-coding RNAs that negatively regulate transcription through sequence-dependent mechanisms, microRNAs such as miR-155 and miR-127 could affect macrophage phenotypes by modulating gene expression.^163–165 Sun et al¹⁶⁶ designed a cyclodextrin nanoparticle carrying CSF-1 siRNA that targeted M2-type macrophages, suppressed expression of M2-related genes, and reprogrammed M2 macrophages into the M1 phenotype. Additionally, DNA methyltransferase DNMT3b was upregulated in M1 macrophages and was thought to mediate the silencing of M2 macrophage-associated genes.¹⁵⁷ The acetylation modification of histones or transcription factors played a dynamic regulatory role in the phenotype regulation of TAMs. For example, Ubiquitin-Specific Peptidase 24 enhanced the acetylation level of histone H3 in M2-type TAMs by stabilizing E1A-Binding Protein P300 and β-Transducin Repeat-Containing Protein, thereby promoting the transcription of IL-6. IL-6 not only consolidated the M2 phenotype of TAMs but also induced more monocytes to differentiate into M2-like TAMs.¹⁶⁷ Enzymes involved in the gene transcription process are also important targets for regulating macrophage phenotypes.¹⁶⁰ Depletion of Dicer 1, Ribonuclease III in macrophages could promote the reprogramming of TAMs into the M1 type.¹⁶⁴

Regulation of the phenotype and function of macrophages can also be achieved by introducing exogenous genes.¹⁴⁷ For example, Lee et al¹⁴⁷ reported that delivery of cytokine-encoding plasmid DNA into macrophage nuclei enabled genetic manipulation and induced macrophage polarization. In addition to ROS-mediated repolarization and CD47 blockade, selective targeting of M2-type TAMs can be further enhanced through peptide modification and biomimetic membrane coating. Wang and co-workers¹⁴⁸ found that bispecific single-domain antibody targeting CCL2 and CCL5 inhibitor messenger ribonucleic acid (BisCCL2/5i mRNA) could function as a gene expression template. By encoding the BisCCL2/5i protein, BisCCL2/5i mRNA achieved bispecific targeted blocking of CCL2 and CCL5, and ultimately induced the polarization of TAMs from the M2 type to the M1 type. Beyond transient gene expression–mediated reprogramming, emerging strategies further extended macrophage polarization control to permanent genome-level editing approaches, enabling more durable phenotypic modulation of tumor-associated macrophages. Zhao et al¹⁴⁹ developed a delivery system based on nanovesicles derived from escherichia coli protoplasts, which could deliver the CRISPR-Cas9 tools to perform gene editing on the Pik3cg of TAMs, thereby inducing macrophages to polarize toward the M1 phenotype (Figure 2). Additionally, the chimeric antigen receptor-modified macrophages (CAR-Ms) technology enabled ex vivo genetic modification of macrophages. The CAR-Ms were used as “Trojan horses” and infused back into patients.¹⁵⁰ CAR-Ms could convert M2-type macrophages to the M1 phenotype, thereby promoting T cell infiltration and enhancing antigen presentation function.^151,152 Han and colleagues developed an in situ CAR-macrophage therapeutic strategy based on phosphatidylserine lipid nanoparticles for the co-delivery of CAR mRNA and a STING agonist, which directly reprogrammed TAMs within the TME and enhanced antitumor activity, achieving significant tumor suppression and systemic immune activation in a murine melanoma model.¹⁵³

Figure 2.

Schematic diagram of the mechanism by which the delivery system based on escherichia coli protoplast-derived nanovesicles reprograms TAMs from the M2 phenotype to the M1 phenotype. The -| lines indicate inhibition of PI3Kγ signaling. Reproduced from ref.149 with permission from Springer Nature, copyright 2026.

BNDDS Based on Macrophages for Cancer Treatment

The BNDDS enable precise drug accumulation at target sites by mimicking the structure and function of natural biological carriers, thereby improving therapeutic efficacy and minimizing toxic side effects.^168,169 Among various BNDDS, macrophage-based carriers offer notable advantages in tumor therapy due to natural targeting ability and immunomodulatory function of macrophages.^168–170 Based on the different application forms of macrophages in constructing BNDDS, macrophage-based BNDDS can be mainly divided into three categories: macrophage-derived exosome delivery systems, macrophage membrane delivery systems, and live macrophage delivery systems. Representative preclinical studies of these three macrophage-based BNDDS and their translational characteristics are summarized in Table 3. The following text will provide an elaboration on the unique advantages and recent applications of the three types of BNDDS in cancer treatment.

Table 3.

Summary and Translational Comparison of Representative Preclinical Studies Utilizing Macrophage-Based BNDDS for Cancer Therapy

BNDDS Type	Therapeutic Payload	Target Tumor Model	Key Mechanistic Finding	Ref.
Macrophage-derived EVs	Oxaliplatin; retinoic acid; Libidibia ferrea	CT26/MC38 colorectal cancer	STAT3/NF-κB/AKT pathway suppression; M2-to-M1 repolarization	[171]
Macrophage-derived EVs	PTX; DOX	Orthotopic T11 and MDA-MB-231 TNBC	Enhanced tumor accumulation; growth inhibition	[35]
Macrophage-derived EVs	HOTTIP (HOXA transcript at the distal tip)	Hep-2 HNSCC xenograft	TLR5/NF-κB activation; tumor suppression	[172]
Macrophage-derived EVs membrane	Suberoylanilide hydroxamic acid	LLC lung cancer xenograft	Integrin α4β1-mediated tumor homing; HDAC inhibition	[33]
Macrophage membrane	Gemcitabine	Breast cancer	Lymphocyte infiltration enhancement; anti-PD-L1 efficacy improvement	[173]
Macrophage membrane	DOX	Orthotopic osteosarcoma; patient-derived xenograft	IL-11-mediated targeting; ROS-triggered DOX release	[174]
Macrophage membrane	SN38; ICG	Orthotopic GL261 glioblastoma	BBB targeting; NIR-triggered drug release	[175]
M1 macrophage membrane	Cisplatin	Peritoneal metastatic forestomach carcinoma	Nanomotor-driven penetration; STING activation	[176]
M1 macrophage membrane	Anti-PD-L1 siRNA; pyropheophorbide a	Panc02 pancreatic cancer	NO-mediated ECM remodeling; enhanced immunotherapy	[177]
Live macrophage	DOX-loaded nanosponges	B16F10 metastatic melanoma	Macrophage homing; acidic TME-responsive drug release	[178]
Live macrophage	DOX-loaded PLGA nanoparticles	Orthotopic 4T1 breast cancer	Surface loading; macrophage function preservation; drug release promotion	[179]
Live macrophage	Pexidartinib-loaded ZIF-8	K7M2 osteosarcoma	Tumor targeting enhancement; sustained TME modulation	[180]
Live macrophage	Ferritin-monomethyl auristatin E conjugate	Orthotopic glioblastoma	Contact-dependent drug transfer; immune activation	[181]
Live macrophage	β-Elemene-loaded GeS nanosheets	4T1 breast cancer	Macrophage hitchhiking; ultrasound-induced immune remodeling	[182]

Abbreviations: AKT, protein kinase B; BBB, blood-brain barrier; DOX, doxorubicin; ECM, extracellular matrix; EVs, extracellular vesicles; GeS, germanium sulfide; HDAC, histone deacetylase; HNSCC, head and neck squamous cell carcinoma; HOTTIP, HOXA transcript at the distal tip; ICG, indocyanine green; IL-11, interleukin-11; NF-κB, nuclear factor kappa B; NIR, near-infrared; NO, nitric oxide; PD-L1, programmed death-ligand 1; PLGA, poly(lactic-co-glycolic acid); PTX, paclitaxel; ROS, reactive oxygen species; siRNA, small interfering RNA; SN38, 7-ethyl-10-hydroxycamptothecin; STAT3, signal transducer and activator of transcription 3; STING, stimulator of interferon genes; TLR5, Toll-like receptor 5; TNBC, triple-negative breast cancer; ZIF-8, zeolitic imidazolate framework-8.

BNDDS Based on Macrophage-Derived EVs

According to the Minimal Information for Studies of Extracellular Vesicles (MISEV) guidelines, EVs represent a broad category of membrane-bound vesicles, and the term EVs is preferred over exosomes when biogenetic origin cannot be experimentally confirmed.¹⁸³ As natural intercellular communication mediators, macrophage-derived EVs inherit key biological properties of parental macrophages, including tumor homing capability, immune compatibility, and immunomodulatory functions.¹⁸⁴ Through the transfer of bioactive molecules, such as proteins, nucleic acids, cytokines, and antigen-presenting molecules, macrophage-derived EVs participate in the regulation of immune responses, tumor progression, and host defense.^41,185,186 These intrinsic biological characteristics, together with excellent biocompatibility and natural targeting capacity, make macrophage-derived EVs attractive candidates for biomimetic nanodelivery systems.^41,185,186 Currently, macrophage-derived vesicular BNDDS can be broadly categorized into naturally secreted macrophage-derived EVs and macrophage-derived exosome-mimetic nanovesicles. Naturally secreted macrophage-derived EVs preserve the intrinsic biological functions of macrophage-derived vesicles, whereas macrophage-derived exosome-mimetic nanovesicles offer improved scalability, controllable fabrication, and functional customization. The following sections summarize recent advances in these two macrophage-derived vesicular delivery strategies.

BNDDS Based on Naturally Secreted Macrophage-Derived EVs

Macrophage-derived EVs inherit specific molecules on the surface of macrophages and can actively migrate to tumor sites by virtue of chemotactic signals induced by chronic inflammation in the TME.^187,188 As vesicular carriers derived from macrophages, macrophage-derived EVs can mimic the biocompatibility of macrophages, effectively evade the clearance of the mononuclear phagocyte system, and thereby prolong the blood circulation time of drugs in the body.¹⁸⁸ Triple-negative breast cancer (TNBC) is highly invasive, prone to metastasis, and lacks effective targeted therapies, resulting in suboptimal overall treatment outcomes currently.^103,189 To address the clinical challenges of TNBC, Haney et al¹⁹⁰ developed macrophage-derived EVs as biomimetic carriers for paclitaxel (PTX) and DOX in TNBC. Drug loading was enhanced through systematic optimization of pH, temperature, and sonication protocols, generating high-loading DOX-loaded and PTX-loaded exosomal formulations. Macrophage-derived exosome membranes displayed LFA-1, supporting LFA-1/ICAM-1-mediated recognition of ICAM-1-overexpressing TNBC cells and enabling preferential tumor-cell association. Membrane-cell interactions facilitated cytosolic drug delivery via membrane fusion and reduced endosomal sequestration. In MDA-MB-231 cells, exosome-mediated delivery increased intracellular drug accumulation by ~150-fold for DOX and ~90-fold for PTX relative to liposomal formulations, accompanied by a pronounced anti-proliferative response. Overall, macrophage-derived exosome-based formulations offer a promising route toward precision chemotherapy for TNBC.

Macrophage-derived EVs can not only express immune recognition proteins including CD47 to achieve long-circulating delivery of chemotherapeutic drugs but also actively regulate the TME through loading immunopromoting drugs.^143,191,192 Carvalho and co-workers¹⁷¹ used EVs derived from M1-polarized RAW264.7 macrophages as nanocarriers and formulated oxaliplatin-loaded EVs alone or combined with retinoic acid, Libidibia ferrea, or retinoic acid plus Libidibia ferrea for evaluation in CT-26 and MC-38 colorectal cancer cells and in allograft and peritoneal models. Exosome characterization identified vesicle-associated markers including LAMP-1, HSP70, actin, and integrin α6, and CT-26 cells showed efficient uptake of M1 macrophage-derived EVs within 24 h, supporting intracellular payload delivery. Beyond drug delivery, M1 macrophage-derived exosome-based systems suppressed STAT3, NF-κB, and AKT signaling and reduced immunosuppressive cues within the TME. Tumor cells exhibited increased FADD and APAF-1 expression and decreased MDR1 and survivin expression after treatment with the M1 macrophage-derived exosome-based formulations. The aforementioned finding indicated that EVs derived from M1-type macrophages functioned not only as efficient drug delivery carriers but also as modulators of the TME, enhancing anti-tumor activity through the synergistic effects of chemotherapy and immune regulation.

M1 macrophage-derived EVs can effectively enhance anti-tumor immunity and inhibit tumor progression by promoting the release of pro-inflammatory factors and being taken up by antigen-presenting cells.^193,194 M1 macrophage-derived EVs can interact specifically with M2-type macrophages and exchange information, thereby promoting the polarization of macrophages from the M2 phenotype to the M1 phenotype.^195,196 Jiang and colleagues¹⁷² found that HOTTIP, a molecule carried by EVs derived from M1-polarized macrophages, promoted the release of pro-inflammatory factors such as TNF-α and suppressed the expression of the anti-inflammatory cytokine IL-10 in circulating monocytes, thereby generating a microenvironment conducive to M1 polarization. In vitro and in vivo experiments further demonstrated that EVs derived from M1-polarized macrophages competitively bound to miR-19a-3p and miR-19b-3p through HOTTIP. Such interactions upregulated the TLR5/NF-κB signaling pathway and consequently inhibited the proliferation and invasion of head and neck squamous cell carcinoma cells.

Macrophage-derived EVs retained key adhesion and recognition properties associated with macrophage translocation, including integrin-adhesion molecule interactions and glycosylation-mediated recognition that promoted binding to brain microvascular endothelial cells.^197,198 Leveraging such specific adhesion and recognition properties, macrophage-derived exosome-based BNDDS enhanced the targeted transport of chemotherapeutic agents to intracranial tumor cells located behind the blood-brain barrier (BBB), providing a feasible strategy for brain tumor therapy.¹⁹⁸ Parada et al¹⁹⁹ reported that RAW274.6-derived macrophage EVs were co-loaded with superparamagnetic iron oxide nanoparticles and curcumin by electroporation, followed by sulfonyl-azide-assisted click cycloaddition to conjugate a neuropilin-1-targeting RGE peptide onto the exosomal membrane, yielding a glioma-targeted engineered exosome platform with imaging and therapeutic functions. In addition, click conjugation preserved key exosome properties, and surface modification remained stable for four weeks during storage at −80 °C. Based on a similar engineering strategy, some studies focused on structural stability and platform controllability, whereas others emphasized in vivo BBB penetration, receptor-dependent recognition, and imaging–therapy synergy. Jia and co-workers²⁰⁰ loaded curcumin and superparamagnetic iron oxide nanoparticles into EVs by electroporation and introduced a glioma-targeting peptide via click chemistry to generate RGE-Exo-SPION/Cur for enhanced brain-tumor delivery and therapeutic performance. In an orthotopic glioma model, intravenous administration enabled delivery to intracranial tumors, and the engineered exosome platform was reported to traverse the BBB for in vivo imaging and therapy, highlighting macrophage-derived EVs as advantageous carriers for BBB-penetrating delivery. Orthotopic glioma magnetic resonance imaging showed clearer lesion contrast and sharper tumor margins at 2 h after administration in the RGE-Exo-related group, consistent with improved intracranial tumor localization after BBB transit. Western blot analysis confirmed high neuropilin-1 expression in U251 cells, and competitive blockade with free RGE reduced exosome uptake signals, supporting RGE-neuropilin-1 receptor-dependent recognition in glioma-targeted enrichment.

Functional modification of macrophage-derived EVs via physicochemical methods or gene editing technologies significantly enhances the tumor-targeting specificity of macrophage-derived EVs, providing a new approach for precise drug delivery.¹⁸⁷ Hybrid exosome-based vesicles constructed by fusing macrophage-derived EVs with liposomes exhibit not only better colloidal stability, stronger drug-loading capacity, and pH-sensitive drug release properties but also differential targeting and cytotoxicity toward normal cells and cancer cells.^201–203 Yang and colleagues²⁰⁴ pointed out that when SIRP-α-modified macrophage-derived EVs were used as drug carriers, macrophage-derived EVs could specifically bind to vitronectin highly expressed on the surface of tumor cells, by virtue of molecules such as integrins expressed on the membrane surfaces. Macrophage-derived EVs actively attached to the surface of tumor cells, and SIRP-α further increased the enrichment of macrophage-derived EVs at the tumor site. The transferrin receptor naturally expressed on the membrane surface of macrophage-derived EVs specifically bound to the corresponding receptors on brain microvascular endothelial cells. SIRP-α modification enhanced the affinity between macrophage-derived EVs and endothelial cells, after which macrophage-derived EVs completed trans-barrier transport through transcytosis.

Despite these advantages, macrophage-derived EVs still face major manufacturing challenges that limit clinical translation. Low production yields, donor-cell variability, and processing conditions can affect EVs composition and batch consistency, complicating large-scale production.^197,198 Purification remains another bottleneck because scalable methods that achieve high recovery, purity, and preservation of EVs integrity are still lacking.^197,205 Common isolation methods have distinct limitations. Ultracentrifugation may cause EVs loss and structural damage, size-exclusion chromatography has limited processing capacity and may co-isolate similarly sized particles, and polymer precipitation can introduce contaminants and reduce EVs purity.²⁰⁴ Electroporation can induce nucleic acid aggregation and alter EVs size, surface charge, and surface-protein profiles, whereas sonication may damage EVs surface proteins like CD63 and promote the loss of endogenous cargo.^190,206,207 Therefore, standardized quality control of EVs identity, purity, cargo loading, stability, and biological activity is essential for reproducible manufacturing and clinical translation.^183,208

BNDDS Based on Macrophage-Derived Exosome-Mimetic Nanovesicles

Distinct from naturally secreted macrophage-derived EVs, macrophage-derived exosome-mimetic nanovesicles are engineered to enable scalable production and structural tunability, while retaining immunological properties and enhancing tumor accumulation and immune synergy. For instance, Baek et al²⁰⁹ generated exosome-mimetic nanovesicles from LPS/IFN-γ-polarized Raw 264.7 M1 macrophages by serial cell extrusion and introduced DSPE-PEG2000 to obtain PEG-MNVs, improving systemic circulation behavior. In CT26 tumor-bearing BALB/c mice, DiR fluorescence imaging showed markedly enhanced tumor accumulation for PEG-MNVs, and the tumor-to-liver signal ratio increased by approximately sevenfold relative to unmodified MNVs, supporting reduced reticuloendothelial sequestration and improved tumor enrichment after PEGylation. In addition, PEG-MNVs exerted M1-associated immunomodulatory activity in an IL-4-induced M2 macrophage setting, as reflected by favorable cellular responses in viability assays, suggesting a potential contribution to M2-to-M1-like repolarization and reinforcing a combined strategy that integrates carrier optimization with immune regulation in tumor therapy.

Macrophage-derived exosomal membranes can be used as biomimetic materials to modify nanoparticles, constructing biomimetic nanoplatforms for tumor therapy.³³ Li and co-workers³³ used exosome membranes (EMS) derived from M1-type macrophages to coat upconversion nanoparticles loaded with histone deacetylase inhibitors (SAHA), and constructed an artificial exosome-based BNDDS for lung cancer treatment. The exosome membranes derived from M1-type macrophages not only endowed the system with extremely high drug-loading capacity, but also could efficiently encapsulate hydrophobic SAHA to solve the problem of low solubility of SAHA. The loading efficiency of SAHA reached approximately 70% when the concentration of SAHA was 1.6 mg/mL. Antibody blocking experiments showed that pretreatment of LLC cells with anti-VCAM-1 antibody reduced the cellular uptake of EMS by approximately 50%, which strongly confirmed the ability of EMS to mediate targeted homing of BNDDS to tumor tissues (Figure 3). At the same time, the exosome membranes could mimic the properties of natural EVs, avoid immune recognition, prolong blood circulation time, further increase the accumulation of drugs at tumor sites, and enhance the therapeutic effect on lung cancer. Although macrophage-derived EVs and macrophage-derived exosome-mimetic nanovesicles both exploit macrophage-associated biological features for biomimetic delivery, these two vesicular systems differ in fabrication, drug-loading capacity, targeting properties, and translational potential. Table 4 summarizes the key characteristics, advantages, limitations, and potential applications of macrophage-derived EVs and macrophage-derived exosome-mimetic nanovesicles in cancer therapy.

Figure 3.

(A) Preparation flow chart of EMS. (B) Schematic diagram of the in vivo anti-tumor mechanism of EMS. The -| lines indicate inhibition of HDAC activity. Reproduced from ref.33 with permission from Springer Nature, copyright 2026.

Table 4.

Comparison of BNDDS Based on Macrophage-Derived EVs and Exosome-Mimetic Nanovesicles

Types	Preparation Process	Drug Loading Efficiency	Targeting Specificity	Immunogenicity	Limitations
Macrophage EVs	Native biogenesis; preserved membrane integrity; source-inherited bioactive components	Efficient cargo encapsulation; cytosolic payload transfer	Inflammation- and tumor-directed targeting; adhesion molecule-mediated uptake	Immune privilege; high biocompatibility; reduced phagocytic clearance	Low yield; batch heterogeneity; limited scalability; loading-induced damage
Exosome-mimetic nanovesicles	Biomimetic fabrication; scalable production; structural customization	Increased production yield; enhanced drug loading capacity	Engineered tumor targeting; surface protein-mediated recognition	Reduced immune stimulation; membrane-camouflaged immune evasion	Altered membrane properties; limited standardization; insufficient clinical validation

BNDDS Based on Macrophage Membranes

Macrophage membranes have emerged as promising biomimetic materials for nanodelivery systems due to their biocompatibility, immune-evasive properties, and tumor-targeting ability.^210–212 By preserving functional membrane proteins, such as CD47 and integrins, macrophage membrane-coated nanoparticles can mimic key macrophage behaviors, including prolonged circulation, reduced immune clearance, and enhanced tumor accumulation.^{174,213–215} These unique characteristics have enabled macrophage membrane-based BNDDS for diverse cancer therapeutic applications, including chemotherapy, immunotherapy, and combination treatments.^174,177,213

Ni et al²¹³ fabricated ZID@RM by cloaking ZID nanoparticles co-loaded with DOX and indocyanine green with RAW264.7 macrophage membranes, establishing a combinational platform for the treatment of Hepa1–6 hepatocellular carcinoma. Membrane-displayed CD47 bound signal regulatory protein SIRPα on host macrophages, delivering a “don’t eat me” signal and attenuating mononuclear phagocyte system clearance, thereby prolonging blood circulation and improving systemic stability. In parallel, membrane proteins such as integrin α4β1 engaged vascular cell adhesion molecule (VCAM-1) associated with tumor vasculature, promoted active enrichment of ZID@RM within inflammatory and tumor microenvironments, and enhanced uptake by Hepa1–6 cells, providing a targeting basis for subsequent near-infrared-triggered photothermal-photodynamic chemo-combination treatment. Building on the intrinsic targeting properties of macrophage membranes, membrane surface engineering can further introduce tumor-specific ligands to enhance receptor-mediated recognition and targeted delivery. Jiang and co-workers¹⁷⁴ used engineered macrophage cell membranes to wrap ROS-responsive nanoparticles loaded with DOX. IL-11 expressed on the surface of the engineered macrophage cell membranes could specifically interact with IL-11 receptor α (IL-11Rα) that was highly expressed on the surface of osteosarcoma (OS) cells. Through this interaction, the engineered macrophage cell membranes could selectively recognize IL-11Rα-expressing OS cells, thereby enhancing nanoparticle accumulation in OS tissues and facilitating intracellular uptake by OS cells. In the 143B orthotopic osteosarcoma model, no visible tumors were observed after the treatment of nanodrugs, and the median survival time of mice was prolonged. The results highlighted the key role of engineered macrophage membranes in enhancing the targeted delivery of nanodrugs, and improving the efficacy of chemotherapy.

The BNDDS prepared by coating nanoparticles with macrophage membranes possesses an immune evasion function.²¹⁴ For example, Chen and colleagues²¹⁵ constructed a core-shell structured nanocomposite and coated the surface of the nanoformulation with J774A.1 macrophage cell membranes for tumor treatment. Compared with nanoparticles without macrophage cell membrane coating, cell imaging and flow cytometry analysis indicated that near-infrared persistent luminescence intensity emitted by biomimetic nanoformulations was significantly weaker in J774A.1 macrophages. Weaker luminescence intensity demonstrated that phagocytosis of biomimetic nanoformulations by J774A.1 macrophages was significantly reduced. Transwell assay results showed that under simulated inflammatory conditions, after addition of biomimetic nanoformulations into upper chambers, SCC-7 cells in lower chambers exhibited strong red luminescence. No obvious luminescence was observed in all control groups. The above experimental results confirmed that macrophage cell membrane coating could help nanoformulations cross the vascular endothelial barrier more effectively and accumulate in tumor cells. Such enhanced tumor accumulation and barrier traversal were fundamentally attributed to the immune-evasive properties conferred by macrophage membrane camouflage, which reduced recognition and phagocytic clearance by immune cells. Wang ‘s group²¹¹ developed a macrophage membrane–camouflaged mesoporous silica nanorod formulation bearing a cationic polymer layer, denoted FA-MSNR/LMDI@MP/MPCM, for breast cancer therapy. In pharmacokinetic studies using tumor-bearing nude mice, FA-MSNR/LMDI@MP/MPCM exhibited a prolonged elimination-phase half-life, a higher peak plasma concentration, and a lower clearance rate relative to free DOX and the non-camouflaged control formulation FA-MSNR/LMDI, indicating improved systemic exposure and slower elimination conferred by macrophage membrane camouflage (Figure 4). Macrophage uptake was further evaluated by confocal laser scanning microscopy and quantified by flow cytometry across different nanocarriers. Macrophage phagocytosis toward the membrane-camouflaged formulations, FA-MSNR/LMDI@MP/MPCM and FA-MSNR/LMDI@MPCM, was markedly reduced. Reduced uptake supported an immune-camouflage effect of macrophage membrane coating, which attenuated mononuclear phagocyte system-mediated clearance and enhanced in vivo stability of the nanoformulations.

Figure 4.

(A) Core composition and basic mechanism of action of the BNDDS. (B) The responsive process of the BNDDS in the TME and the near-infrared (NIR) light-regulated synergistic therapeutic mechanism. Reproduced from ref.211 with permission from Springer Nature, copyright 2026.

BNDDS based on macrophage cell membranes possess the advantages of prolonged circulation and low toxicity.^216–218 Liu and colleagues²¹⁹ developed a rapamycin-loaded TPGS/F127 micelle wrapped with macrophage membranes, and the micelle was specifically designed for the treatment of lung cancer complicated with pulmonary fibrosis. Pharmacokinetic studies showed that after intratracheal instillation administration at a dose of 0.4 mg/kg to male SD rats, the apparent elimination half-life of the biomimetic nanomicelles was 139.139±36.916 h, which was 1.7 times that of the free rapamycin group. The area under the plasma concentration-time curve (AUC_(0-t)) of the biomimetic nanomicelles reached 2063.819±35.023 mg/L h, which was higher than that of the free rapamycin group. The results fully confirmed that macrophage membranes could effectively prolong the in vivo circulation time of drugs and increase plasma drug concentration levels. In addition to improving pharmacokinetic performance and prolonging systemic circulation, macrophage membrane camouflage also contributed to enhanced biosafety and reduced cytotoxicity toward normal cells. Liang et al²²⁰ encapsulated DOX hydrochloride in quantum dot-loaded liposomes via self-assembly and subsequently coated the liposomal surface with purified macrophage membranes, yielding DOX-QDs-Lip@M for anti-metastasis therapy of breast cancer. An MTT assay in normal L02 cells showed cell viabilities above 84.01% across the tested DOX-QDs-Lip@M concentrations, with no evident downward trend, supporting reduced cytotoxicity toward normal liver cells after macrophage-membrane camouflage.

The biomimetic modification strategy based on macrophage cell membranes can prevent the enzymatic degradation of drugs, and simultaneously significantly enhance the distribution and accumulation of drugs at the tumor site under the chemotaxis mediated by inflammatory chemokines.^221–223 For example, Rao and co-workers²²⁴ developed macrophage membrane-wrapped upconversion nanoparticles for tumor targeting and imaging. The core design of the system lay in reconstructing natural macrophage cell membranes into vesicles, and then coating the vesicles onto the surface of synthetic upconversion nanoparticles through a mechanical extrusion method. The process completely retained the natural membrane proteins on macrophage cell membranes, enabling the nanoparticles to inherit the unique intercellular adhesion ability of macrophages. After intravenous injection of nanoparticles into BALB/c nude mice bearing MCF-7 breast cancer xenografts, in vivo upconversion luminescence imaging at 48 hours showed that upconversion luminescence signals at tumor sites of mice in the biomimetic nanoparticle group were significantly stronger than those in the unmodified nanoparticle group. After dissection, inductively coupled plasma atomic emission spectrometry results of tumors and major organs showed that the cumulative amount of biomimetic nanoparticles in tumor tissues was significantly higher than that in the control group. Results indicated that the macrophage membrane-modified delivery system achieved dual barrier breakthrough under sequential ultrasound stimulation through the mediating effect between ligands on the membrane surface and receptors on tumor blood vessels, leading to enrichment of nanoparticles in tumor tissues.

In the field of tumor therapy, BNDDS constructed based on macrophage cell membranes can effectively break through various physiological barriers in the TME and achieve deep penetration of drugs into tumor regions.²²² For example, She et al¹⁷⁶ developed an oxygen-driven heterojunction nanomotor coated with M1-type macrophage cell membranes, which was specifically used for the treatment of peritoneal metastatic tumors. In terms of tumor penetration function, the coating of the M1-type macrophage cell membrane structure made the nanomotors more likely to break through the physical barrier on the tumor surface. CCK-8 assays demonstrated that macrophage membrane-coated nanomotors exhibited enhanced cytotoxicity against MFC gastric cancer cells, increasing the inhibition rate from 67% for uncoated nanomotors to 74%. The increase in anti-tumor efficacy indirectly reflected enhanced ability to break through tumor surface barriers with the help of membrane structures. Meanwhile, driven by oxygen bubbles generated through the catalytic reaction between MnO₂ within the nanomotors and H₂O₂ in the TME, the nanomotors penetrated deeper into tumor tissues. During this penetration process, M1-type macrophage cell membranes functioned as a “protective shell” and a “penetration aid” for the nanomotors. Compared with approaches relying solely on the intrinsic biomimetic properties of macrophage membranes, ligand engineering further enhanced target recognition and translocation across complex physiological barriers. Li and colleagues¹⁷⁵ developed a bionic nano-platform modified with Angiopep-2 and coated with macrophage cell membranes. During the drug delivery process, the receptors on the surface of macrophage cell membranes could guide the nanoparticles to approach the BBB endothelial cells accurately, laying a targeting foundation for crossing the barrier. When Angiopep-2 bound to low-density lipoprotein receptor-related protein on brain microvascular endothelial cells to enhance the interaction between the nanoparticles and the endothelial cells, the macrophage cell membranes reduced the rejection of the endothelial cells by virtue of biocompatibility and provided an “affinity environment” for endocytosis. The macrophage cell membranes also assisted in maintaining the structural stability of the nanoparticles to ensure the continuous effect of Angiopep-2. The two components synergistically promoted the nanoparticles to penetrate the BBB efficiently and treated glioblastoma effectively (Figure 5).

Figure 5.

Preparation flow chart and mechanism diagram of the angiopep-2-modified, macrophage membrane-coated biomimetic nanoplatform. The red lightning bolt indicates near-infrared laser irradiation, and the red upwards arrows indicate photothermal-induced temperature elevation. Reproduced from ref.175 with permission from the Taylor and Francis Group, copyright 2026.

Nanoparticles with biomimetic modification using M1 macrophage-derived cell membranes can reprogram M2 macrophages into M1 macrophages during the tumor treatment process, thereby further enhancing the anti-tumor therapeutic effect.¹⁷⁷ Liang ‘s group¹⁷⁷ fabricated M1@PAP by electrostatically adsorbing anti-PD-L1 small interfering RNA onto NanoPA, a self-assembled nanoparticle derived from the type I photosensitizer pyropheophorbide a conjugated with poly-L-arginine Arg9, followed by coating with M1-polarized macrophage membranes for hypoxia-tolerant therapy of pancreatic cancers. Flow-cytometric profiling of tumor-associated macrophages showed that M1@PAP markedly shifted macrophage polarization from an immunosuppressive M2-like phenotype toward an antitumor M1-like phenotype, as evidenced by CD86 upregulation and CD206 downregulation, supporting a role of M1 macrophage membrane-associated cues in TAM reprogramming. Following 650 nm laser irradiation, macrophage repolarization was further enhanced. The M1-type macrophages formed by reprogramming could improve the tumor immunosuppressive microenvironment by upregulating the secretion of pro-inflammatory cytokines such as TNF-α and IL-12, and downregulating the expression of anti-inflammatory cytokines such as IL-10, and laid a foundation for enhancing the anti-tumor immune response.²²⁵

A growing number of researchers fuse macrophage cell membranes with other cell membranes to construct BNDDS for tumor therapy.²²⁶ Compared with nanoparticles coated with a single type of cell membrane, the nanoparticles camouflaged with mixed cell membranes exhibit more excellent tumor-targeting ability and BBB penetration performance.^214,227 Gong and colleagues²²⁸ developed DOX-loaded PLGA nanoparticles cloaked with a RAW264.7 macrophage and 4T1 breast cancer hybrid membrane, denoted DPLGA@[RAW-4T1] NPs, for anti-metastatic therapy in a 4T1 breast cancer lung metastasis setting. Western blotting combined with immunogold transmission electron microscopy verified the simultaneous retention of α4 integrin and VCAM-1 on membrane-coated nanoparticles, supporting α4 integrin-VCAM-1-mediated recognition. At the cellular level, an Annexin V-APC flow-cytometric apoptosis assay showed that, at a DOX concentration of 1 μg/mL, DPLGA@[RAW-4T1] NPs induced 58.15% apoptosis in 4T1 cells, exceeding apoptosis levels observed for free DOX and uncoated nanoparticle controls. At the animal level, an experimental breast cancer lung metastasis model was established, and IVIS Spectrum bioluminescence imaging confirmed the formation of pulmonary metastatic foci. Following intravenous administration of DiR-labeled formulations, ex vivo fluorescence imaging and quantitative organ analysis demonstrated a time-dependent increase in lung fluorescence for DiR-PLGA@[RAW-4T1] NPs, with a 5.14-fold higher lung signal than DiR-PLGA NPs. Overall, inflammatory chemotaxis and targeting conferred by macrophage membranes, together with homotypic recognition derived from 4T1 membranes, enhanced delivery and suppression of lung metastatic lesions.

Despite these advantages, macrophage membrane-coated nanoparticles still face stability and manufacturing challenges.⁶⁰ Membrane isolation and coating remain technically complex and are largely confined to laboratory-scale production. Low membrane yield, batch variability, coating instability, and storage limitations further hinder scale-up.²²⁴ Membrane-protein orientation is another concern, as mechanical disruption and extrusion may not ensure uniform right-side-out orientation of functional proteins such as CD47 and integrins, potentially weakening immune-evasion and targeting functions.^60,229 In addition, the membrane shell may hinder nanoparticle uptake and intracellular drug release after tumor accumulation.²¹¹ Stimuli-responsive strategies, including pH-responsive membrane detachment and MMP-sensitive systems, have therefore been developed to remove or disrupt the membrane barrier within the TME and facilitate cargo release.^173,230

BNDDS Based on Live Macrophages

Unlike macrophage-derived EVs and membrane-based systems, live macrophage-based BNDDS exploit intact cellular functions to achieve active tumor homing, biological barrier penetration, and therapeutic cargo delivery.^231–233 The intrinsic phagocytic capacity and intracellular space of macrophages enable efficient loading of nanomedicines, while chemokine–receptor interactions guide macrophage migration toward tumor tissues.^184,197,234 Therapeutic cargos can be loaded through intracellular uptake or surface anchoring using approaches such as incubation, electroporation, and surface adhesion.^40,41 Because macrophage source influences drug-loading capacity, migratory behavior, and immunomodulatory function, live macrophage-based BNDDS can be broadly considered according to primary macrophages and macrophage cell lines, with their major characteristics summarized in Table 5.

Table 5.

Core Performance Comparison of BNDDS Based on Macrophages from Different Sources

Macrophage Sources	Drug Loading Efficiency	In vivo Migratory Capacity	Immunomodulatory Function
PMs	High drug-loading capacity; low-toxicity cargo encapsulation	Inflammation-guided homing; lesion accumulation	Potent immunomodulatory activity; microenvironment remodeling
SPMs	High phagocytic loading capacity; cargo biocompatibility	Inflammation-guided homing; lesion infiltration	Low immunogenicity; preserved immune bioactivity
BMDMs	Intracellular payload tolerance; flexible engineering capability	Tumor-homing capability; delivery to primary and metastatic lesions	Antitumor immune activation; immunosuppressive microenvironment reprogramming
RAW264.7	Nanoparticle uptake capacity; reproducible cargo loading	Tumor-tropic migration; hypoxic-region infiltration	Immunotherapeutic potential; microenvironment modulation
J774A.1	Nanoparticle co-loading capacity; trigger-controlled cargo release	Cytokine-guided tissue homing; barrier-crossing delivery potential	Inflammation-like activation; preserved phenotypic plasticity

BNDDS Based on Primary Macrophages

Primary macrophages, including BMDMs and monocyte-derived macrophages, possess intrinsic migratory, immune-regulatory, and tumor-homing capabilities, making them promising living carriers for BNDDS.^235–237 These cellular platforms enable active tumor targeting and therapeutic cargo delivery through interactions with the tumor microenvironment.²³⁸

BMDMs served as biomimetic drug delivery carriers that evaded host immune clearance, actively homed to tumor tissues in response to tumor-associated signals, and thereby markedly prolonged systemic circulation and improved the pharmacokinetic performance of loaded therapeutics.^235,239 Huang and colleagues²³⁵ used BMDMs as carriers to load zinc phthalocyanine-encapsulated oxaliplatin (IV) prodrug nanoparticles (Oxa (IV)@ZnPc@M) and applied the delivery system to the targeted treatment of primary breast cancer lesions and bone metastatic lesions. Pharmacokinetic results experiments conducted in SD rats confirmed that when administered at an equivalent oxaliplatin dose of 1 mg/kg, compared with free oxaliplatin (Oxa) and Oxa (IV)@ZnPc, Oxa (IV)@ZnPc@M exhibited a longer blood circulation time. Among these groups, the blood clearance half-life of Oxa (IV)@ZnPc@M was approximately 3.9 times that of Oxa (IV)@ZnPc, and the area under the concentration-time curve (AUC) of Oxa (IV)@ZnPc@M was 17.5 times that of Oxa (IV)@ZnPc. In addition, Oxa(IV)@ZnPc loading resulted in CD80⁺CD86⁺ expression in BMDMs comparable to LPS-induced M1 macrophages, suggesting a shift toward a pro-inflammatory, antitumor M1 phenotype and the presence of immunostimulatory potential in engineered BMDMs.

The BBB severely limits the penetration of many chemotherapeutics into brain parenchyma, thereby compromising therapeutic efficacy against malignancies such as glioblastoma.^236,240,241 The BBB endothelium expresses multiple transport and targeting receptors, and macrophages can cross the BBB via receptor-ligand interactions with endothelial cells.²³⁸ Consequently, macrophage-based BNDDS offer a promising strategy for BBB crossing and targeted therapy of intracranial tumors.^236,240 For example, Sun et al¹⁸¹ developed a macrophage-based cellular drug delivery platform in which heavy chain ferritin (HFt) was conjugated to the cytotoxic payload monomethyl auristatin E via a biodegradable linker to generate HFt-735. The resulting conjugate was subsequently internalized by BMDMs to obtain the corresponding cell-based formulation. In vitro coculture with glioma cells showed that macrophages underwent cytoskeletal rearrangement to form immune synapse-like interfaces, accompanied by vesicular trafficking. Through the transfer of the iron-binding proteins pathway, macrophages enabled contact-dependent and highly efficient transfer of the conjugates, leading to HFt signals in >90% of glioma cells within 24 h and superior tumor cell killing compared with free-drug controls. In an orthotopic glioma model, intratumoral administration promoted deep macrophage infiltration into tumor tissue and localized drug accumulation within tumors, while markedly reducing systemic exposure and toxicity, highlighting the potential of a macrophage-centered “Trojan horse” delivery strategy for immunosuppressive solid tumors. Building on macrophage-based delivery strategies using BMDMs, macrophages were derived from murine bone marrow through in vitro differentiation and further polarized to the M1 phenotype by IFN-γ/LPS stimulation for the construction of cell- BNDDS capable of tumor-directed migration and barrier traversal in glioma treatment. Wang ‘s group²⁴² adopted M1-type macrophages to load poly PLGA nanoparticles for efficient glioma therapy. By constructing an in vitro endothelial barrier model, M1-type macrophage-mediated nanoparticles (M1-NPs) could significantly improve the penetration efficiency. The fluorescence intensity of nanoparticle uptake by U87 cells at the bottom was 1.6 times higher than that of free nanoparticles, which confirmed that M1-type macrophages could promote nanoparticles to penetrate BBB-like structures. In therapeutic evaluation, the M1-NPs regimen extended the median survival of tumor-bearing mice to 38.5 days. Moreover, tumor sections from the M1-NPs group exhibited the highest caspase-3 expression, alongside a pronounced increase in apoptotic cells. Collectively, M1 macrophage-loaded nanoparticles suppressed glioma progression after traversing the BBB-like barrier. The study utilized the core roles of M1-type macrophages in multiple aspects, including targeted delivery, barrier penetration, and inherent anti-tumor activity, and provided a new strategy for solving the problem of drug delivery across the BBB.

Similarly, in vitro differentiated bone marrow derived M1 macrophages can be further engineered at the cell surface to enhance tumor recognition and intratumoral retention, as exemplified by peptide modified M1 macrophage systems for targeted delivery. Dong and co-workers¹⁸⁰ prepared P@ZIF/M1-KTP by membrane-anchoring a K7M2-targeting peptide, KTP, through palmitic acid lipidation to generate M1-KTP, followed by macrophage loading of pexidartinib-encapsulated ZIF-8 nanoparticles, P@ZIF-8. KTP modification enhanced M1 macrophage affinity toward K7M2 osteosarcoma cells and improved tumor recognition. In vivo IVIS imaging showed higher tumor accumulation for PA-KTP-related formulations. Flow cytometry further indicated increased intratumoral CFSE-positive macrophages: 0.20% in the M1 group, 1.13% in the M1-KTP group, and 1.48% in the P@ZIF/M1-KTP group, with tumor-associated macrophage levels peaking at approximately 24 h. Enhanced tumor homing and prolonged intratumoral retention supported sustained intratumoral delivery of pexidartinib by P@ZIF/M1-KTP and continuous suppression of tumor cells.

Beyond intracellular encapsulation, macrophage-based BNDDS have also leveraged the physicochemical properties of macrophage membranes to enable nanoparticle anchoring at the cell surface.^178,243 Owing to the complex composition of the macrophage membrane, including lipids, polysaccharides, and membrane proteins, the cell surface presents diverse functional groups and surface affinities.^3,244 Nanoparticles can therefore be immobilized on macrophage surfaces through rapid physical adsorption or coordination-driven chemical interactions, enabling efficient surface loading while preserving cellular migratory functionality.^3,244 Liao and colleagues²⁴³ engineered macrophages by decorating the outer membrane with DOX-loaded porous nanosponges sealed by a Fe³⁺-tannic acid metal-polyphenol gatekeeper, yielding the macrophage-driven automated homing of metal-phenolic-gated nanosponges (MAGN) platform for metastatic melanoma therapy. In B16F10 tumor–bearing mice, in vivo imaging at 24 h after intravenous administration showed an approximately threefold higher tumor fluorescence signal in the MAGN group than in the free chemotherapeutic drug group, supporting macrophage-driven tumor homing and enhanced drug enrichment at tumor sites. TUNEL and Ki67 staining of tumor sections further showed the strongest TUNEL-positive and weakest Ki67-positive signals in the MAGN group, consistent with augmented apoptosis and suppressed proliferation (Figure 6). The strategy that deeply combined cell surface engineering with the tumor-homing property of macrophages gave full play to the core roles of macrophages in targeted transport and TME penetration, and provided a multifunctional method for the treatment of advanced metastatic cancer.

Figure 6.

(A) Schematic illustration of drug loading, metal–phenolic gating, and macrophage assembly of pH-responsive nanosponges. (B) Schematic illustration of the pH-responsive drug release mechanism. The downward arrow indicates the transition from the “OFF” state at pH 7.4 to the “ON” state at pH 6.0. Reproduced from ref.243 with permission from the John Wiley and Sons, copyright 2026.

Inflammatory monocytes isolated from mouse peritoneal lavage are differentiated into inflammatory macrophages under GM-CSF induction, and nanomedicine delivery systems constructed using such macrophages exhibit markedly enhanced in vivo migratory capacity.²³⁷ For instance, Ren et al²³⁷ constructed Res/ICG-R8-Lip@MP by co-loading resveratrol (Res) and indocyanine green (ICG) into octaarginine (R8)-modified liposomes, followed by macrophage uptake. Tumor cells and tumor-associated stromal cells secreted C-C motif chemokine ligand 2 (CCL2), which chemotactically recruited and guided macrophages to migrate toward the tumor, thereby endowing the macrophage-carried formulation Res/ICG-R8-Lip@MP with active homing capability and rendering tumor targeting not solely dependent on the EPR effect. Under near-infrared irradiation, ICG-mediated photothermal heating disintegrated macrophage carriers at the tumor site and markedly accelerated local release of Res/ICG-R8-Lip@MP. Building on the inflammation-guided homing of macrophage carriers demonstrated in postoperative recurrence models, further evidence showed that R8-modified liposome-loaded macrophages preserved chemotaxis-related protein expression and efficiently transferred liposomal “backpacks” to tumor cells within an inflammatory TME. Qiu and co-workers²⁴⁵ loaded R8-modified liposomes into macrophages and delivered the R8-modified liposomes to recurrent tumor sites. In vitro studies demonstrated that macrophages could respond to inflammatory factor gradients and efficiently deliver the loaded liposomes to the vicinity of 4T1 tumor cells. Results from Western blot assays and transwell assays showed that drug-loaded macrophages could maintain the expression of inflammation chemotaxis-related proteins such as Ly-6C and F4/80 and retain effective migration ability, thereby ensuring targeted delivery to the inflammatory TME. In addition, macrophages could directly transfer the loaded liposomes to adjacent 4T1 tumor cells through membrane fusion, under the observation of a confocal laser scanning microscope. Thus, in postoperative recurrence settings characterized by surgically remodeled vasculature and an unstable EPR effect, the role of macrophages extended beyond passive drug loading. Macrophages enabled inflammation- and tumor-signal-driven active homing and further facilitated efficient drug transport and release at local sites, thereby enhancing delivery reliability and broadening the therapeutic window.

BNDDS Based on Macrophage Cell Line

Macrophage cell lines, such as RAW264.7 and J774A.1, provide accessible and reproducible sources of living carriers for BNDDS. Compared with primary macrophages, macrophage cell lines offer advantages in scalability and engineering flexibility while maintaining macrophage-associated migratory and tumor-homing capabilities, thereby facilitating nanoparticle loading and targeted tumor delivery.^246–248

The BNDDS constructed based on the tumor-homing property of macrophages could optimize drug distribution and lesion retention by virtue of the directional migration of macrophages to tumors, thus improving delivery efficiency.^246,247 For example, Liu et al²⁴⁹ constructed USIP@M by co-incubating RAW264.7 macrophages with photoresponsive micelles co-loaded with UCNPs, sorafenib, and IMD-0354 (USIP) (Figure 7). To evaluate in vivo tumor homing, H22 tumor-bearing mice received intravenous administration, followed by longitudinal IVIS imaging at multiple time points in three groups: free DiR, DiR-loaded micelles (DP), and macrophage-loaded DiR micelles (DP@M). DP@M produced the strongest fluorescence signal at the tumor site at 12 h, significantly exceeding signals observed in the DP and free DiR groups. Additionally, Western blot analysis confirmed the presence of the chemotactic targeting protein CCR2 in macrophage-related formulations, providing mechanistic evidence that CCR2 contributed to preferential tumor homing and increased intratumoral accumulation of DP@M/USIP@M.

Figure 7.

(A) Preparation of USIP@M and intravenous administration. (B) In vivo delivery and antitumor mechanism of USIP@M. i–iii indicate light-triggered release, macrophage phenotypic switch, and tumor microenvironment regulation, respectively. Red arrows indicate near-infrared laser irradiation-induced activation of USIP@M; purple arrows indicate light-triggered release and transport of SI-exosomes; blue arrows indicate macrophage phenotypic switching and cytokine-mediated regulation; green arrows indicate SI-exosome-mediated tumor cell apoptosis; and black arrows indicate cellular interactions within the tumor microenvironment. Reproduced from ref.249 with permission from Elsevier, copyright 2026.

Hypoxia is a prominent characteristic of solid tumors.²⁵⁰ Hypoxic regions in tumors secrete high concentrations of chemokines, such as the hypoxia-inducible factor 1 (HIF-1), which recruit macrophages to the hypoxic areas.^251,252 Meanwhile, the hypoxic environment significantly inhibits the motility of macrophages, resulting in more macrophages being trapped in the hypoxic regions.^251,252 Therefore, compared with traditional nanoparticles, macrophage-based BNDDS can penetrate deeply into the hypoxic regions of tumors without being affected by the interstitial fluid pressure.^250,253 An and co-workers²⁵⁴ cocultured small anionic gold nanorods (AuNRs) with RAW264.7 macrophages to generate the macrophage-mediated formulation anionic-AuNRs@RAW (Figure 8). After intravenous administration in 4T1 tumor-bearing mice, photoacoustic imaging revealed that anionic-AuNRs@RAW became detectable in the tumor region as early as 1 h and reached a peak tumor signal at 8 h. Moreover, the photoacoustic signal of anionic-AuNRs@RAW showed strong colocalization with the deoxyhemoglobin signal, a marker of hypoxic status, suggesting preferential macrophage-mediated delivery to hypoxic tumor regions. In contrast, free anionic-AuNRs exhibited limited overlap with deoxyhemoglobin, consistent with predominantly EPR-driven passive accumulation. Histological analysis further demonstrated that apoptotic regions after anionic-AuNRs@RAW treatment nearly coincided with HIF-1α-positive severely hypoxic areas, corroborating that RAW264.7 macrophages conferred active targeting of hypoxic tumor niches to anionic-AuNRs@RAW.

Figure 8.

Schematic illustration of macrophage uptake of AuNRs with different surface charges and macrophage-mediated delivery for hypoxia-targeted photothermal therapy of tumors. The red cross indicates high cytotoxicity. The black arrows in the upper panel indicate the endocytosis process of AuNRs with different surface charges by macrophages. The red arrows in the lower panel indicate the migration of AuNR-loaded macrophages toward tumor tissues. Reproduced from ref.254 with permission from American Chemical Society, copyright 2026.

As with murine primary BMDMs, the cell surface of RAW264.7 macrophages can also be exploited for nanomedicine loading, enabling macrophage-mediated drug delivery through surface engineering.¹⁷⁸ For instance, Zhu et al¹⁷⁹ exploited coordination between divalent metal ions and tannic acid to construct a metal-phenolic network, enabling rapid adsorption of DOX-loaded nanoparticles onto RAW264.7 macrophage surfaces within approximately 2 min under ice-bath conditions, while minimizing cellular internalization and preserving migration-associated bioactivity, thereby yielding DOX-NP@Mϕ. In vivo biodistribution analysis showed that, at 2 h after intravenous administration in tumor-bearing mice, intratumoral DOX accumulation from DOX-NP@Mϕ was 3.7-fold higher than that achieved by PEGylated DOX nanoparticles, consistent with macrophage responsiveness to signal gradients in the tumor microenvironment and enhanced delivery. Upon near-infrared irradiation, photothermal conversion by the metal–phenolic network induced carrier-cell disruption and release of nanoparticle-associated vesicles, markedly accelerating local drug release and strengthening tumor-cell uptake.

Technologies such as genetic engineering, surface modification of macrophages, and optimization of nanoparticle loading can effectively enhance the phagocytic function of macrophages and the tumor-targeting efficiency of macrophages.^248,255 Desai and colleagues²⁵⁶ used stepwise loading followed by flow sorting to generate co-loaded macrophage carriers. J774A.1 macrophages first internalized drug-loaded nanoparticles and, after washing, subsequently internalized superparamagnetic iron oxide nanoparticles. Side-scatter-guided flow cytometric sorting then enriched the co-loaded cell population, improving loading efficiency and uniformity. Regarding loading performance, single loading of mesoporous silica nanoparticles into J774A.1 macrophages reached 99%, whereas co-loading of mesoporous silica nanoparticles with superparamagnetic iron oxide nanoparticles exceeded 70%. Functionally, an alternating magnetic field induced SPION-mediated magnetic hyperthermia, enabling thermally triggered activation of carrier macrophages and providing a means to regulate payload release at lesion sites. Phenotype profiling further indicated that nanoparticle-loaded macrophages retained migration-associated functional features under appropriate conditions, supporting macrophage-based delivery as a versatile strategy.

Noncovalent interactions, including hydrophobic interactions, electrostatic interactions, and host-guest recognition, have been broadly exploited for in vitro assembly of macrophage-based BNDDS.²⁵⁷ One representative strategy applied cationic polymers to impart a positively charged macrophage surface, thereby enabling electrostatic anchoring of negatively charged nanoparticles and generating a nanoparticle–macrophage complex with high DOX loading efficiency.²⁵⁷ Xie ‘s group¹⁸² reduced the disulfide bonds on the surface of macrophages to provide specific binding sites for the anchoring of nanotherapeutics. Subsequently, the highly specific reaction between maleimide and sulfhydryl groups was utilized to stably conjugate the nanotherapeutics to the surface of macrophages, and constructed a macrophage-mediated BNDDS. Compared with free drugs, the fluorescence intensity of the macrophage-mediated BNDDS at the tumor site was significantly enhanced, and the accumulation peak of the system at the tumor site was reached at 4 hours after intravenous injection. Macrophages acted as “living carriers”, could respond to the gradient of inflammatory factors in the tumor area, migrate actively to the tumor site, further transport the nanotherapeutics conjugated on the surface of macrophages to the tumor area efficiently.

A major limitation of live macrophage delivery systems is phenotypic instability after administration, as ex vivo M1-polarized macrophages may be reprogrammed toward M2-like states by the immunosuppressive TME.^42,180,258 Strategies to stabilize the antitumor phenotype include targeting M2-associated regulators like STAT6, CRISPR-Cas9-based reprogramming, and sustained TLR7/8 stimulation.^258–260 However, durable stabilization of the M1 phenotype remains to be established.^258,260 Live macrophage systems also face challenges in large-scale manufacturing, storage, and regulatory translation, requiring standardized cell processing, engineering, drug loading, quality control, and cryopreservation under GMP conditions.^60,257 Although the phase 1 CT-0508 trial demonstrated the feasibility of clinical-scale manufacturing and cryopreservation, further standardization of production, release criteria, product stability, and long-term safety is required.^37,257,261 The major advantages, limitations, and clinical readiness of the three macrophage-based BNDDS are summarized in Table 6.

Table 6.

Comparison of the Major Advantages, Limitations, and Clinical Readiness of Macrophage-Based BNDDS for Cancer Therapy

BNDDS Type	Major Advantage	Limitation	Clinical Readiness
Macrophage-derived EVs	Tumor tropism; biocompatibility; immunomodulatory capacity	Low yield; purification difficulty; batch heterogeneity; variable drug loading	Preclinical development; clinical translation potential
Macrophage membrane	Immune evasion; tumor targeting; nanocarrier compatibility	Complex membrane preparation; coating consistency; storage instability	Preclinical applications; multimodal therapeutic potential
Live macrophage	Active tumor homing; tissue penetration; cargo adaptability; immunomodulatory activity	Loading-induced functional impairment; phenotypic instability; manufacturing complexity; regulatory challenges	Early clinical evaluation; manufacturing feasibility

Conclusion and Future Perspectives

Macrophage-based BNDDS combine tumor-targeted delivery with immune regulation through three complementary platforms: macrophage-derived EVs, macrophage membrane-coated nanoparticles, and live macrophage carriers. By exploiting macrophage-associated tumor tropism, immune evasion, tissue penetration, and immunomodulatory activity, these platforms offer distinct opportunities for cancer therapy. However, most EV- and membrane-based systems remain preclinical, whereas engineered live macrophages have only recently entered early clinical evaluation.^36,171,181

Important biological and technical barriers remain. Macrophage-derived EVs are limited by low yield, purification difficulties, batch heterogeneity, and cargo-loading-related changes in EVs integrity.^{183,205–209} Membrane-coated nanoparticles face challenges in membrane preparation, coating consistency, right-side-out protein orientation, storage stability, and controlled removal of the membrane barrier after tumor accumulation.^{173,224,229,230} Live macrophage carriers may undergo loading-induced functional impairment or phenotypic reprogramming within the immunosuppressive TME.^258–260 Therefore, future optimization should consider not only drug-loading efficiency and tumor accumulation, but also macrophage viability, migration, phenotype stability, immune function, and controlled payload release.

Clinical translation requires standardized and scalable GMP-compliant manufacturing. Defined cell sources, reproducible engineering and drug-loading procedures, validated quality-control assays, cryopreservation, and product stability are essential.^206,207 The transition from operator-dependent laboratory procedures to automated and closed manufacturing systems remains a major translational bottleneck for macrophage-based BNDDS.^257,261 Although the phase 1 CT-0508 trial demonstrated the feasibility of clinical-scale manufacturing and cryopreservation of engineered macrophages,³⁷ broader translation will require reduced batch variability, standardized release criteria, long-term safety evaluation, and clearer regulatory pathways. Manufacturing cost and scalability should also be considered for future commercialization.^32,37

Future research should prioritize controllability and translational relevance rather than simply increasing platform complexity. Firstly, intelligent and stimuli-responsive macrophage-based BNDDS are already under investigation, which may improve controllability through engineered response mechanisms, including pH-responsive membrane detachment, matrix metalloproteinase-responsive release, and near-infrared-triggered release.^179,237,256 Secondly, establishing renewable and standardized macrophage sources is essential for clinical translation. Renewable human macrophage sources, particularly induced pluripotent stem cell-derived macrophages, may help reduce donor variability and facilitate scalable production.²⁶² Thirdly, the development of clinically relevant evaluation systems will be important for accurately assessing the safety and efficacy of macrophage-based BNDDS. Patient-derived organoids, tumor-on-chip models, multi-omics approaches, and quantitative biodistribution studies may improve evaluation of macrophage heterogeneity, species differences, and therapeutic performance.^3,31,263 Finally, the integration of macrophage-based BNDDS with established cancer therapies, particularly chemotherapy and immune checkpoint blockade, may enhance synergistic therapeutic effects.^264,265 Macrophage-derived EVs-mimetic nanovesicles can enhance immune checkpoint blockade by activating antitumor immunity, while macrophage membrane-coated nano-gemcitabine combines chemotherapy with anti-PD-L1 therapy by promoting lymphocyte infiltration and restoring lymphocyte-mediated tumor killing.^264,265 Live macrophage-based therapies may also be combined with adoptive cell therapies such as CAR-T by remodeling the TME and promoting T-cell recruitment and activation.³⁷ Overall, successful translation of macrophage-based BNDDS will depend on precise phenotype control, controlled drug release, reproducible manufacturing, clinically relevant validation, and integration with established cancer therapies.

Funding Statement

This work was supported by National Natural Science Foundation of China (82274113), the Shenyang Science and Technology Talent Project (RC230343), the high-quality development science and technology funding project of China Medical University (2023JH2/20200044), the Liaoning Revitalization Talents Program (XLYC2503120), the Natural Science Foundation of Liaoning Province (2026-MS-186), the Multidisciplinary Interdisciplinary Team Project for the Innovation of Traditional Chinese Medicine Science and Technology of Liaoning University of Traditional Chinese Medicine (2025-Dxkjc-02-01; 2025-DxkjcU45-23-19) and the Open Fund of the Key Laboratory of Ministry of Education for TCM Viscera-State Theory and Applications, Liaoning University of Traditional Chinese Medicine (zyzx2409).

Abbreviations

AMs, Alveolar macrophages; BBB, Blood-brain barrier; BMDMs, Bone marrow-derived macrophages; BDDS, Biomimetic drug delivery systems; CAR-Ms, Chimeric antigen receptor macrophages; CCL, Chemokine (c-c motif) ligand; CD, Cluster of differentiation; CTLs, Cytotoxic t lymphocytes; Dectin-1, Dendritic cell-associated c-type lectin 1; DOX, Doxorubicin; EGF, Epidermal growth factor; EPR, Enhanced permeability and retention; EVs, Extracellular vesicles; HFt, Heavy chain ferritin; HGF, Hepatocyte growth factor; HIF-1, Hypoxia-inducible factor 1; ICG, Indocyanine green; IFN-γ, Interferon-gamma; IL, Interleukin; iNOS, Inducible nitric oxide synthase; LPS, Lipopolysaccharide; MMPs, Matrix metalloproteinases; MnO2, Manganese dioxide; NIR, Near-infrared; NK, Natural killer (cells); NO, Nitric oxide; OS, Osteosarcoma; Oxa, Oxaliplatin; PD-1, Programmed cell death protein 1; PD-L1, Programmed death-ligand 1; PLGA, Poly(lactic-co-glycolic acid); PMs, Peritoneal macrophages; PPARγ, Peroxisome proliferator-activated receptor gamma; PTT, Photothermal therapy; PTX, Paclitaxel; Res, Resveratrol; rGO, Reduced graphene oxide; ROS, Reactive oxygen species; Siglec-10, Sialic acid-binding immunoglobulin-like lectin 10; SIRPα, Signal-regulatory protein alpha; SPMs, Splenic macrophages; TAMs, Tumor-associated macrophages; TGF-β, Transforming growth factor-beta; Th1, T helper 1; TLR, Toll-like receptor; TME, Tumor microenvironment; TNBC, Triple-negative breast cancer; TNF-α, Tumor necrosis factor-alpha; VCAM-1, Vascular cell adhesion molecule 1; VEGF, Vascular endothelial growth factor.

Data Sharing Statement

No new data were generated or analyzed in this study. Data sharing is not applicable to this article.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declared no potential conflicts of interest with respect to the research, authorship, and publication of this article.