Exosome-mimetic nanocarriers in cancer diagnosis and drug delivery: engineering strategies, functional biomimicry, safety considerations, and clinical translation

Abstract

Cancer remains one of the leading causes of mortality worldwide, driving the development of advanced drug delivery systems to improve therapeutic selectivity and overcome the complex defense mechanisms of malignant cells. Exosome-mimetic nanocarriers (EMNs) have emerged as an advanced biomimetic platform for cancer diagnosis and targeted drug delivery, combining the biological functionality of natural exosomes with the manufacturing flexibility and scalability of synthetic nanocarriers. This review analyzes the composition, design, and architecture of EMNs, as well as their applications in cancer drug delivery, drawing on fundamental concepts of pharmaceutical technology to provide a translational perspective. It also includes a dedicated section on cancer diagnosis and theranostic platforms, as well as a critical analysis of recent technological advancements in exosome-mimetic systems. Although the clinical translation of natural exosomes remains limited, emerging evidence suggests that engineered EMNs offer improved scalability, reproducibility, and therapeutic versatility. Recent studies highlight their potential to overcome key limitations of natural vesicles, positioning them as promising candidates for future clinical translation and commercialization.

Graphical Abstract

Introduction

Cancer remains one of the principal causes of morbidity and mortality worldwide, representing a major global health challenge [1]. Even with advances in prevention, early detection, and treatment, cancer therapy is still limited by therapeutic efficacy, systemic toxicity, and drug resistance, thus reducing patient survival and quality of life. These limitations underscore the urgent need for innovative strategies that lead to more precise diagnoses, effective treatments, and personalized approaches in oncology [2]. In this context, exosomes have emerged as a promising area of study, opening new avenues for both understanding and controlling cancer. Exosomes are tiny extracellular vesicles, measuring about 30 to 150 nm in diameter, that originate from the endosomal pathway and are secreted by almost all cell types. These vesicles carry a diverse array of bioactive molecules, including proteins, lipids, DNA, messenger RNAs (mRNAs), and non-coding RNAs such as microRNAs (miRNAs), long non-coding RNAs (lncRNAs), and circular RNAs (circRNAs), which accurately reflect the physiological or pathological state of their parent cells [3]. Exosomes play crucial roles in intercellular communication, immune regulation, and disease progression. Thanks to their natural biocompatibility and low immunogenicity, they have become promising tools for biomarker discovery and therapeutic delivery in various diseases, notably cancer, minimizing systemic toxicity [4].

Natural exosomes and exosome-mimetic nanocarriers (EMNs) are structurally related but functionally distinct. The main differences lie in their origin, fabrication, and production scale. Despite their potential, natural exosomes face significant limitations, including low production yields, heterogeneity, and challenges in controlling loading, which hinder their clinical application. To address these issues, researchers have developed EMNs, engineered nanoscale vesicles designed to mimic the structure and biological functions of natural exosomes. These can be produced via top-down methods such as mechanical cell extrusion, microfluidics, liposomal engineering, or bottom-up synthetic assembly, offering improved scalability, consistent production, and precise drug loading [5]. Additionally, their surfaces can be functionalized with targeting ligands or exosomal markers and loaded with high concentrations of therapeutic agents, thereby enhancing tissue-specific delivery and improving therapeutic effectiveness [4, 6, 7].

The progress of cancer nanomedicine has underscored the need for highly specific, biocompatible, and efficient drug-delivery platforms. As naturally derived vesicles, exosomes possess biological properties that make them ideal for both diagnostic and therapeutic applications in oncology. Their ability to encapsulate and safeguard diverse biomolecules allows them to carry tumor-specific signatures, making them valuable for liquid biopsy diagnostics [8]. Moreover, exosomes exhibit natural targeting abilities and can cross biological barriers, including the blood-brain barrier (BBB), a critical feature for treating brain tumors such as glioblastoma [9]. Preclinical studies, including both in vitro and in vivo models, have demonstrated the efficacy of exosome-mimetic platforms in delivering drugs such as doxorubicin, paclitaxel, and siRNAs across various cancer models, resulting in tumor growth inhibition, reduced systemic toxicity, and improved survival in animal models [6]. Furthermore, these systems support personalized medicine by tailoring their composition and surface features to patient-specific tumor markers or genetic profiles, although their clinical validation remains limited.

Previous reviews have described the biological properties of natural exosomes and the engineering aspects of synthetic nanocarriers; however, an integrative framework addressing both systems remains insufficient. Moreover, little attention has been paid to biomimetic design principles and their translation into functional, clinically relevant nanocarrier systems. This review addresses this gap by proposing a conceptual framework that situates EMNs within a rational design strategy guided by physicochemical parameters, surface architecture, and biological interactions, drawing on existing knowledge of exosomes and natural exosomes. A key contribution of this work is the critical integration of structural, functional, and translational dimensions, enabling a direct comparison between natural exosomes and EMNs, beyond conventional descriptive approaches.

Review methodology

In this narrative review, we analyze the current state of knowledge regarding the prospects and limitations of EMNs for cancer diagnosis and drug delivery. The manuscript is organized into six main sections: (1) Structure and Properties of Exosome-Mimetic Nanocarriers, (2) Applications in Drug Delivery, (3) Diagnostic Applications, (4) Clinical Applications and Ongoing Trials, (5) Comparative Advantages of Exosome-Mimetic Nanocarriers Over Traditional Nanoparticles in Drug Delivery, and (6) Challenges and Innovations. We conducted a comprehensive literature search across PubMed, Scopus, and Web of Science, and compiled articles published over the last 15 years. The search strategy included using MeSH descriptors and keywords such as “exosomes”, “exosome-mimetic nanocarriers”, “extracellular vesicles”, “drug delivery”, “cancer diagnosis”, and “synthetic exosomes”. Boolean operators were adapted to each database to optimize search sensitivity and specificity. Peer-reviewed articles in English were included if they addressed the diagnostic or therapeutic applications of exosome-mimetic nanocarriers in oncology, including experimental and clinical research, as well as high-quality reviews discussing structure, biocompatibility, production techniques, and scalability. Conference abstracts, editorials, and brief communications were excluded.

Structure and properties of exosome-mimetic nanocarriers

Natural exosomes are extracellular vesicles of endosomal origin, typically 30–150 nm in diameter, and are secreted by nearly all cell types, playing a pivotal role in intercellular communication [4]. Their lipid bilayer, enriched with specific proteins and functional nucleic acids, makes them highly effective biological vectors for the transfer of molecular information. Inspired by these properties, EMNs have emerged as biomimetic delivery systems that replicate both the structural and functional traits of natural exosomes. Unlike natural exosomes, which are directly isolated from cells via ultracentrifugation or immunoaffinity techniques, EMNs are synthetic or semi-synthetic constructs engineered to reproduce the architecture and biological behavior of their natural counterparts [10]. The structural biomimicry aims not only to enhance biocompatibility and stability in biological fluids but also to optimize parameters such as biodistribution, immune evasion, and cellular targeting, all of which are closely linked to their physicochemical and topological features. Table 1 compares the main features of natural exosomes and EMNs, providing a clear framework for recognizing their similarities and distinctions, ranging from origin and size range to cargo loading and target specificity. Understanding these differences is crucial, as it enables researchers to select the most suitable platform for specific applications in cancer diagnosis and therapy. Moreover, such comparisons can also highlight existing limitations, thereby providing opportunities to improve the design of nanocarrier systems.

Table 1.

Comparative features of natural exosomes and exosome-mimetic nanocarriers

Comparative Feature	Natural exosomes	Exosome-Mimetic Nanocarriers	References
Origin/System	Naturally secreted by cells via the endosomal pathway	Artificially produced or bioengineered	[4, 6]
Size Range	30–150 nm	50–200 nm (tunable during synthesis)	[4, 7]
Composition	Natural proteins, lipids, RNA, and DNA reflecting the cell of origin	Designed components such as synthetic lipids, targeting ligands, and loaded therapeutic cargo	[6, 11]
Biological Function	Involved in intercellular communication, immune signaling, and disease progression	Therapeutic delivery and functional mimicry of natural exosomes	[4, 7]
Surface Markers	CD9, CD63, CD81, TSG101, Alix, among others	Functionalized with exosomal markers or targeting moieties	[6, 11]
Production	Low yield, batch variability, complex isolation methods such as ultracentrifugation	Scalable, reproducible production via engineering techniques like extrusion or microfluidics	[5–7]
Loading Capacity & Strategy	Limited control, mainly natural cargo, or indirect modification	High loading efficiency with tunable cargo such as drugs, siRNA, or miRNA	[6, 7]
Immunogenicity	Generally low, especially from autologous sources	Variable, depending on the materials used and surface modifications	[4, 7]
Stability and Shelf-life	Variable, sensitive to storage conditions	Generally, more stable and can be optimized for storage and circulation	[5, 7]
Targeting Specificity	Natural tropism related to the parent cell type	Customizable targeting via ligand conjugation or surface engineering	[5–7]

The EMN structural features are critical for determining their in vivo behavior, including circulation time and biodistribution. For example, following systemic administration, exosomes tend to accumulate predominantly in organs such as the liver, spleen, kidney, and lung, and are rapidly cleared from the bloodstream. This distribution is usually far from random. Instead, it reflects the cell they come from and the specific makeup of their membrane (proteins, lipids, and glycans), which actively guide their affinity for certain tissues [12].

The structural core of EMNs is a lipid bilayer that mimics the membrane composition of native exosomes (Fig. 1A). The bilayer not only serves as a protective barrier against enzymatic degradation but also enables the encapsulation of both hydrophilic and lipophilic biomolecules [13]. Commonly used lipids include phosphatidylcholine, cholesterol, sphingomyelin, and anionic lipids, which mimic the asymmetric, organized composition of natural exosomes secreted by stem cells, tumor cells, or immune cells [14]. Morphologically, EMNs are characterized by a spherical structure with a well-controlled size range typically between 80 and 150 nm. Nanoscale architecture is crucial for efficient cellular uptake, as it promotes internalization through mechanisms such as clathrin-mediated endocytosis and macropinocytosis [15]. Additionally, their uniform spherical shape and minimal surface roughness enhance colloidal stability, reducing the likelihood of aggregation in suspension. These structural features contribute to prolonged in vivo circulation times, a key factor in achieving effective biodistribution and therapeutic impact under physiologically relevant conditions [16].

Fig. 1.

(A) Physical architecture of extracellular vesicles; (B) Pathways involved in extracellular vesicle production

A crucial element of EMNs is the integration of membrane proteins or peptides that mimic surface markers of native exosomes. These include tetraspanins such as CD9, CD63, and CD81, as well as proteins involved in the Endosomal Sorting Complex Required for Transport (ESCRT), like Alix and TSG101 [17, 18]. Their presence not only enhances structural resemblance but also mediates specific interactions with cellular receptors, enabling selective recognition by target cells. Furthermore, EMNs can be strategically engineered to include particular glycoproteins or surface-bound ligands that closely mimic the complex glycosylation profiles naturally present on exosomal membranes. These glycan structures play a central role in modulating the interactions between vesicles and their biological environment, particularly with immune cells. Although frequently overlooked in nanocarrier design, glycosylation significantly impacts intracellular routing, endosomal escape, and vesicles’ ability to evade phagocytic clearance. By reproducing these patterns, EMNs acquire greater functional resemblance to native exosomes, thereby improving circulation time and promoting selective uptake by target cells [10].

Internally, EMNs can encapsulate a diverse array of bioactive compounds, including mRNA, miRNA, proteins, and other regulatory molecules, closely mimicking the cargo typically found in native exosomes [19]. This functional mimicry enhances their potential for therapeutic delivery and intercellular communication. The chosen fabrication technique significantly impacts the efficiency and stability of this encapsulation process. Methods such as high-pressure extrusion, ultrasonic sonication, and microfluidic processing not only determine the loading capacity of EMNs but also influence the structural integrity and biological performance of the resulting nanocarriers.

Another key feature emulated by EMNs is their slightly negative zeta potential (typically −10 to −30 mV), which contributes to colloidal stability, reduces aggregation, and extends circulation half-life (Fig. 2). This electrostatic characteristic is fundamental for maintaining colloidal stability, as it minimizes particle aggregation and prolongs the nanocarrier’s circulation time in biological fluids [20]. When combined with the intrinsic membrane elasticity that closely resembles that of native exosomes, this property enhances the deformability of EMNs. As a result, they exhibit improved ability to traverse biological barriers, such as endothelial or epithelial layers, and demonstrate increased retention within target tissues [18].

Fig. 2.

Architectural features of EMNs

Physicochemical parameters and functional mimicry

The physicochemical properties of EMNs are fundamental to their ability to emulate native exosomes. These parameters not only determine their behavior in biological environments but also influence delivery efficiency and targeting specificity. One of the most critical parameters is the zeta potential, which reflects the surface charge in colloidal media. Native exosomes typically display a mildly negative zeta potential, ranging from −10 to −30 mV, which contributes to their suspension stability and prevents aggregation [20]. To replicate this, EMNs are formulated with anionic lipids, such as phosphatidylserine or phosphatidic acid, thereby optimizing their electrostatic behavior to avoid rapid clearance by the reticuloendothelial system [10]. This adjustment enhances circulation time and bioavailability compared to traditional cationic nanoparticles.

Membrane elasticity is another feature accurately reproduced in EMNs, which is essential for crossing biological barriers, such as the intestinal epithelium or the BBB [19]. Natural exosomes possess optimal deformability due to their fluid bilayer and lack of rigid protein coatings. EMNs replicate this by incorporating low-phase-transition lipids, such as DOPE (dioleoyl phosphatidylethanolamine), and using assembly techniques that prevent crystalline domain formation [10]. This flexibility not only facilitates cellular uptake but also reduces immune activation.

Colloidal stability represents a fundamental parameter in the design of EMNs, as it directly influences their behavior in physiological environments. Native exosomes exhibit inherent stability in biological media due to the delicate balance between hydrophilic and hydrophobic membrane components. EMNs replicate this characteristic by incorporating stabilizing agents, most notably polyethylene glycol (PEG), which form a protective corona around the nanoparticle surface. This modification significantly reduces nonspecific protein adsorption, prevents nanoparticle aggregation, and limits recognition by the mononuclear phagocyte system [14]. As a result, PEGylated EMNs display improved pharmacokinetic profiles and reduced immunogenic clearance.

In terms of uptake mechanisms, EMNs effectively mimic exosomal entry routes, including clathrin- and caveolin-mediated endocytosis and macropinocytosis [15]. This ability is mainly due to their nanoscale morphology, smooth surface, and membrane proteins that function as cell-targeting ligands. The inclusion of tetraspanins, integrins, or ligands such as folate and transferrin enhances internalization efficiency and recapitulates the natural tropism of native exosomes [18]. Moreover, EMNs are designed to replicate the inherent structural heterogeneity of native exosomes by precisely regulating their polydispersity and size distribution. Advanced fabrication techniques, including membrane extrusion and microfluidic systems, enable the generation of EMNs with narrow size ranges and low polydispersity indices (PDI < 0.2), closely matching the physicochemical profile of exosomes derived from biological fluids [16, 21]. This controlled uniformity not only improves batch-to-batch reproducibility but also contributes to more predictable in vivo biodistribution, circulation kinetics, and cellular uptake dynamics.

The surface charge characteristics and lipid composition of EMNs are meticulously optimized to replicate the fusion capacity observed in natural exosomes. By incorporating key lipid components such as sphingolipids, cholesterol, and phosphatidylserine in biologically relevant ratios, EMNs can closely mimic the membrane properties necessary for direct interaction with target cell membranes. This lipid arrangement promotes fusion with both the plasma membrane and endosomal compartments, facilitating efficient release of encapsulated cargo into the cytoplasm [18]. Additionally, this mechanism helps minimize lysosomal degradation, thereby improving the bioavailability and functional delivery of therapeutic agents Table 2.

Table 2.

Physicochemical properties of natural exosomes and exosome-mimetic nanocarriers

Parameter	Natural exosomes	EMNs	Functional Relevance in Drug Delivery	Reference
Size	30–150 nm vesicles	50–200 nm (commonly 80–150 nm)	Facilitates cellular uptake and tumor penetration; supports enhanced permeability and retention (EPR) effect	[9, 13, 18]
Polydispersity Index (PDI)	Narrow biological distribution	< 0.2 (controlled systems)	Ensures reproducibility and predictable pharmacokinetics	[19]
Zeta Potential	Slightly negative surface charge	−10 to −30 mV	Improves colloidal stability and prolongs circulation	[22]
Membrane Composition	Lipid bilayer of cellular origin	Phospholipids, cholesterol, sphingomyelin	Enables encapsulation of diverse cargo and improves stability	[16, 17, 22]
Membrane Fluidity/Elasticity	Naturally fluid bilayer	Tunable (e.g., DOPE incorporation)	Facilitates membrane fusion, uptake, and barrier crossing	[22, 23]
Surface Functionalization	CD9, CD63, CD81, integrins	Ligands, peptides, antibodies, tetraspanins	Enables receptor-mediated targeting and uptake	[20, 21]
Loading Capacity	Natural RNA/protein cargo	High, controlled encapsulation	Enables delivery of drugs, siRNA, miRNA, proteins	[23, 24]
Colloidal Stability	Intrinsic membrane stability	Enhanced via PEGylation or lipid design	Prevents aggregation and improves circulation time	[17]
Fusion Capability	Natural membrane fusion	Promoted by lipid composition	Enables cytoplasmic delivery and reduces lysosomal degradation	[21]
Surface Glycosylation	Complex natural glycome	Partial synthetic reproduction	Influences immune evasion and cell recognition	[22]

Engineering and assembly of EMNs

The structural assembly of EMNs represents a pivotal phase in replicating the morphological and functional characteristics of native exosomes with high fidelity. This stage involves a series of carefully controlled physicochemical techniques, such as high-pressure extrusion, ultrasonic sonication, and microfluidic mixing, that enable precise regulation of critical parameters. These include particle size, PDI, bilayer symmetry, and the spatial orientation of functional membrane proteins [22]. By fine-tuning these variables, researchers can generate EMNs with consistent physicochemical profiles, thereby enhancing reproducibility, biological performance, and clinical translation potential.

High-pressure extrusion is widely used to produce EMNs with uniform diameters of 80–150 nm. This method forces lipid suspensions through membranes with defined pore sizes, resulting in unilamellar vesicles with low polydispersity [10]. Sonication, on the other hand, fragments larger vesicles and promotes the formation of spherical EMNs. Microfluidics offers even greater control, regulating the ratios of aqueous and lipid phases, flow rates, and mixing gradients to yield stable and reproducible architectures [19].

Replicating the lipid bilayer of native exosomes is a fundamental aspect of EMN design. These nanostructures are formulated with specific lipids, such as cholesterol, sphingomyelin, and phosphatidylserine, in ratios that mirror the asymmetric distribution and dynamic fluidity characteristic of natural exosomal membranes [10]. This biomimetic composition contributes to membrane stability and functionality under physiological conditions. Additionally, the bilayer can be engineered to incorporate targeted ligands or exogenous membrane proteins, effectively recreating the surface recognition motifs of natural exosomes. Such modifications enhance selective interaction with receptor-expressing cells, thereby improving targeting efficiency [17].

Surface markers, such as tetraspanins (CD9, CD63, CD81), can be engineered into EMNs to reproduce exosome hallmark features. These proteins contribute to adhesion and signaling and regulate intracellular trafficking, thereby facilitating the targeted delivery of bioactive cargo [17]. Membrane engineering techniques enable the selective insertion of these proteins into the EMN bilayer, thereby recapitulating their functional activity. PDI control is another essential structural element. A PDI below 0.2 is considered ideal to mimic the homogeneity of biological exosomes, ensuring predictable biodistribution and uniform cell interactions [16, 21]. Microfluidics excels in this regard by enabling continuous production of EMNs with tightly controlled size distributions, improving batch-to-batch reproducibility.

Although EMNs cannot spontaneously reproduce the full proteomic complexity of natural exosomes, they can be engineered to incorporate selected proteins that emulate key functional elements. This incorporation is achieved through strategies such as co-incubation with donor membranes, membrane fusion techniques, or transient transfection during the fabrication process [23]. These approaches enable the integration of specific surface- or intraluminal proteins, thereby enhancing the nanocarriers’ biological activity and target specificity. As a result, EMNs can facilitate the controlled and efficient delivery of bioactive molecules to designated cells or tissues.

Surface glycosylation patterns, known to be crucial for immune evasion and precise cellular recognition, can be partially reproduced in EMNs. Although replicating the entire glycome profile of native exosomes remains technically challenging, promising advances have been made through the incorporation of synthetic glycans and the retention of glycoproteins derived from natural cell membranes [25]. These strategies help recreate essential surface features that influence vesicle biodistribution, receptor binding, and immunological invisibility. By mimicking these glycosylation signatures, EMNs enhance their biomimetic fidelity and potential for effective therapeutic delivery in vivo.

From a functional standpoint, EMNs have consistently demonstrated the capacity to reproduce the biological behavior characteristic of native exosomes. Their unilamellar spherical structure, combined with a fluid lipid bilayer and the strategic inclusion of membrane-associated markers, enables them to interact effectively with target cells. These features facilitate key cellular entry mechanisms, such as receptor-mediated endocytosis and, in some cases, direct fusion with the plasma membrane [4]. These pathways are crucial for ensuring the efficient intracellular delivery of encapsulated cargo, thereby promoting the therapeutic efficacy of EMNs in drug delivery and regenerative medicine applications.

Applications in drug delivery

EMNs are emerging systems of great interest for the loading and release of active agents due to their excellent properties, which position them as the next generation of nanocarriers [26]. Some characteristics of exosomes are their high biocompatibility, low immunogenicity, stability in biological media, efficient tissue penetration, and targeting capacity. Their similarity to natural carriers allows them to blend into the biological environment and act without alerting the biological system, thereby reducing the likelihood of triggering an immune response [27]. As delivery systems, natural exosomes and EMNs protect therapeutic agents that degrade in biological environments, enabling more effective administration. Therefore, natural exosomes and EMNs are ideal for delivering sensitive drugs or biological agents, such as proteins, which may lose their activity in adverse environments [28]. EMNs can travel through the bloodstream and even cross the BBB due to their nanometric size (30–150 nm), enabling access to hard-to-reach targets. Another notable feature of these systems is their targeting capacity. The natural exosomal membrane contains surface proteins and ligands that enable binding and uptake by specific tissues, resulting in more effective, localized delivery of therapeutic agents while reducing side effects that may arise from their use. While EMNs try to mimic this behavior by incorporating ligands into their structure [27, 29].

Although exosomes are often described as efficient tumor-targeting carriers, their actual penetration into tumor tissues remains a complex, incompletely resolved process. Although their small size favors passive accumulation via processes such as enhanced permeability and retention, their distribution within a tumor is far from uniform. In practice, it is strongly shaped by the tumor microenvironment’s biological features. Factors such as irregular vascularization, extracellular matrix density, and differential uptake by different cell types can hinder exosome movement, making it difficult for exosomes to reach and penetrate deeper regions of the tumor [30]. Furthermore, although it has been proposed that tumor-derived exosomes preferentially return to their tumor of origin, this behavior is not consistently observed in in vivo models. In practice, their ability to target and penetrate tumor tissue varies considerably and depends largely on the biological context in which they are found [12]. These limitations demonstrate that the success of drug delivery depends not only on drug accumulation in the tumor, but also on effective distribution within the tumor.

Several methods have been developed for loading active agents into exosomes; the ideal method will depend on the desired release profile and the type and properties of the active agent [28]. The two main types are passive and active loading methods (Fig. 3). The passive method applies only to cell-derived natural exosomes; in this case, cells are cultured in the presence of the active agent, which naturally incorporates into the exosomes during their production. This method is the simplest; however, the loading efficiency is low (< 10%), and the active agents must not affect the cells. One passive method application is the direct loading of genetically engineered substances, such as proteins or RNA [31].

Fig. 3.

Passive and active loading methods for the incorporation of active agents in exosomes

On the other hand, the active method applies to natural exosomes of both cellular and synthetic origin. In this method, exosomes are first extracted and purified. Then, they are loaded with the active agent using techniques such as sonication, electroporation, or incubation with detergents or saponins, which increase permeability and facilitate their incorporation. This form of loading is more efficient, with loading percentages of up to 40%, and allows greater control of the process. Finally, the exosomes are purified to remove the unloaded agent by techniques such as ultracentrifugation, chromatography, and ultrafiltration. The main limitations of the active method are cost, long processing times, potential stress on exosomes, and instability [24, 28]. However, active loading is ideal for synthetic exosomes and for active ingredients that can be harmful to cells.

Beyond the choice of loading strategy, the therapeutic performance of exosome-based delivery systems is strongly influenced by cargo loading efficiency, encapsulation stability, and release kinetics. These parameters are particularly relevant because they determine how much active agent is effectively incorporated, how well it is retained during storage and circulation, and whether it is released in a controlled manner at the target site [24, 28]. Generally, passive loading strategies are gentler on the membrane and better preserve its integrity, but this often results in a relatively low loading rate, which in many systems does not exceed 10%. In contrast, when active loading methods are used, the amount of encapsulated material can increase significantly, reaching 20–40% or higher, depending on the type of load and the methodology employed [24, 28]. For example, electroporation has shown encapsulation efficiencies of around 20% for doxorubicin; however, its performance can vary significantly with nucleic acids. In some miRNA-based systems, efficiencies as low as 1.68% have been reported. In contrast, with other biomolecules, such as insulin or certain recombinant proteins, and under carefully controlled conditions, efficiencies approaching 50% have been achieved [24, 28]. Similar techniques, such as sonication and extrusion, generally achieve more efficient loading than simple passive incubation or freeze-thaw cycles, especially for proteins or hydrophobic compounds. However, these methods are not without drawbacks, as they can also alter the vesicle membrane structure if not carefully adjusted [18, 23].

Encapsulation stability is another critical but inconsistently reported parameter. In practical terms, it refers to whether the therapeutic cargo remains associated with the vesicle long enough to retain its biological activity and prevent premature release. This issue is particularly relevant for unstable biomolecules such as siRNA, miRNA, proteins, or enzyme-based therapeutics, which may rapidly lose functionality if not properly protected [10]. Several studies have shown that exosome-based systems can protect their cargo from enzymatic degradation and environmental stress while maintaining structural integrity under physiological conditions [18]. In parallel, release kinetics also play a decisive role in therapeutic efficacy. Rather than simply maximizing loading, an ideal platform should ensure sustained or stimulus-responsive release. In this regard, several exosome-mimetic systems have shown prolonged release over 48–72 h and enhanced drug release at acidic pH 5.5, which is particularly relevant for tumor-targeted delivery [32, 33]. However, these parameters are not uniformly quantified across studies, and direct comparisons remain difficult due to differences in isolation methods, vesicle source, loading protocols, and analytical readouts [28].

EMNs aim to mimic the properties of natural exosomes, such as size, surface charge, and expression of specific ligands, while facilitating more precise control over their manufacture. These synthetic nanovesicles can be derived from liposomes, biodegradable polymers, or hybrids and are designed to improve stability, facilitate functionalization, and adapt to clinical requirements [34]. Furthermore, their production is scalable, and the system design allows for optimized loading conditions while minimizing structural damage. Additionally, the absence of residual cellular components reduces the risk of unwanted effects and improves the reproducibility of the process. Natural exosomes exhibit high selectivity within cells because they carry specific markers that interact with the system. However, the use of exosomes from cells has limitations, including low isolation yields, complex processing, inefficient drug payloads, and safety issues due to limited understanding of their functionality. The alternative is nanoparticles that mimic exosome characteristics. These nanovesicles allow targeting the weak points of natural exosomes and, because they are more controlled, are easier to approve for clinical trials [35].

Natural exosomes and EMNs can carry a wide variety of active agents, especially those that tend to lose their activity when exposed to the biological environment. These include biological molecules such as proteins and RNA for gene therapy or replacement therapy, photosensitizing agents and radionuclides, as well as molecules for chemotherapy and those with moderate toxicity to reduce side effects and facilitate more efficient delivery. Simultaneous administration of multiple agents is also possible [31, 35]. The current goal is to develop surface-functionalized exosomes to achieve targeted delivery to specific tissues. These systems will protect the active agents, reduce side effects, and increase the bioavailability and efficacy of the treatment [36].

Although cancer is one of the most studied diseases, effective treatments are still lacking, prompting the search for therapies that can selectively target tumors and minimize drug side effects. Trials using EMNs for cancer treatment are promising. Exosomes and EMNs are organotropic materials, so naturally derived exosomes from cancer cells strongly target tumors and metastatic cells, resulting in ultra-specific targeting. Additionally, their high biocompatibility and low immunogenicity make them ideal for cancer patients with weakened immune systems. However, more understanding and monitoring of their biological functions are needed before they can reach the clinical stage [35]. Currently, plant-based or synthetic exosomes with controlled features are the most favored options. Vázquez-Ríos et al. developed nanosystems that simulate the functionality and structure of tumor-derived exosomes, but with a controlled composition containing tumor-specific protein markers and a simple synthesis process under mild conditions. These systems aim to deliver therapeutic oligonucleotides specifically to lung adenocarcinoma cells, demonstrating high specificity and the capacity to transport therapeutic agents [37].

The administration of chemotherapeutic molecules is one of the most promising fields for exosomes. The administration of doxorubicin via natural exosomes derived from breast and ovarian cancer cell lines has been studied, increasing doxorubicin efficacy by reducing its toxicity, thereby enabling greater tolerability to high doses in female mouse tests [38, 39]. The use of exosomes derived from bovine colostrum and modified with folic acid as a targeting ligand for the selective administration of paclitaxel has also been proven; a chemotherapy drug that, due to its low selectivity, can cause damage to healthy cells, producing severe systemic toxicity with severe side effects, as a potential system for the treatment of lung cancer. These EMNs exhibited a significant increase in antiproliferative activity against A549 cells and in tumor xenograft inhibition relative to direct paclitaxel. In addition, in in vivo studies, animals treated with exosomes did not show toxicity at any scale, making them a viable option for clinical applications [40]. Another active agent studied for delivery via exosomes is curcumin, a highly valued anti-inflammatory and antioxidant, but difficult to administer due to its high hydrophobicity. In vivo test results displayed that curcumin was more stable and had greater bioavailability in exosomes [41].

One of the most promising features of exosomes for cancer treatments is their immunogenicity and molecular transfer capacity, which enable them to deliver active ingredients to target cells and modulate their function. Exosomes and their mimetic analogs offer an effective means of delivering therapeutic agents locally, with promising applications in cancer treatment. Improving their composition and targeting capabilities could yield safer, more efficient drug-delivery systems. Additionally, developing biomimetic nanocarriers offers a solution to the limitations of cell-derived exosomes, opening the way for reproducible, scalable, and safe systems [42]. Despite the results in preclinical models, most studies evaluating natural exosomes and EMNs for drug delivery are limited by experimental heterogeneity and insufficient standardization. In vivo studies using small-animal models often lack assessments of long-term toxicity, immunogenicity, and pharmacokinetics, making it difficult to attribute observed effects specifically to exosomes Table 3.

Table 3.

Representative exosome-based and exosome-mimetic drug delivery approaches in cancer therapy

Reference /Study ID	System type	Preparation/Fabrication Method	Source Material	Cargo Loading Strategy	Targeting Ligand/Surface Feature	Cancer Model	Key Therapeutic Outcome
[39, 40]	Natural exosomes	-	-	Doxorubicin loading	Native exosomal surface properties	Breast and ovarian cancer mouse models	Improved antitumor performance relative to conventional administration
[41]	Natural exosomes	Isolation from bovine colostrum-derived exosomes	Bovine colostrum-derived exosomes	Paclitaxel loading	Folic acid	A549 lung cancer cells and tumor xenograft model	Increased antiproliferative activity and inhibition of tumor growth.
[42]	Natural exosomes	-	-	Curcumin loading	Native exosomal membrane properties	In vivo cancer-related delivery context	Improved curcumin stability and greater bioavailability
NCT03608631	Natural/engineered exosomes	-	Blood-derived exosome therapeutic platform	siRNA loading	Molecular targeting of the KRASG12D pathway	Pancreatic cancer	Translation toward targeted gene silencing strategy in metastatic pancreatic cancer
NCT01159288	Natural exosomes/immunotherapeutic exosome platform	-	Dendritic cell-derived exosomes	Tumor antigen loading	Immune-targeted exosomal presentation	NSCLC	Improved progression-free survival and favorable safety profile
NCT01294072	Natural exosomes	-	Plant-derived exosomes	Curcumin loading	Native vesicle properties	Colon cancer	Entered Phase I safety evaluation, supporting translational feasibility
NCT05559177	Engineered/chimeric exosomal system	Bioengineered exosomal vaccine platform	Blood-associated exosomal therapeutic platform	Vaccine-associated therapeutic payload	Chimeric tumor-associated surface engineering	Metastatic bladder cancer	Early clinical evaluation as an immunotherapeutic exosome-based strategy
[27, 29]	EMNs	Mechanical extrusion, microfluidics, liposomal engineering, bottom-up synthetic assembly	Synthetic lipids and/or bioengineered membrane-mimetic materials	Passive or active loading (active loading includes sonication, electroporation, and detergent or saponin-assisted loading)	Surface functionalization with targeting ligands, glycoproteins, or exosomal markers	Multiple preclinical cancer models	Improved scalability, reproducibility, tunable cargo loading, targeted delivery, and reduced off-target effects

-: Not specified; Exosome-mimetic nanocarriers: EMNs

Diagnostic applications

Exosomes have revolutionized the concept of liquid biopsy and have emerged as pivotal mediators in tumor biology and cancer diagnostics. Secreted by nearly all cell types, including malignant cells, they carry a complex molecular cargo reflective of their cell of origin [43]. Multiple studies have shown that exosomes found in biological fluids (e.g., blood, urine, saliva, pleural fluid) provide a minimally invasive window into tumor biology [44]. As illustrated in Fig. 4, exosomes are widely distributed across biological fluids and can encapsulate a diverse range of cargo. This wide availability enables the collection of tumor information without needing invasive procedures, providing a significant diagnostic advantage and linking these molecules to cancer presence, progression, and aggressiveness. Therefore, exosomes obtained through liquid biopsies have become a promising tool for early diagnosis, patient stratification, and dynamic cancer monitoring [45].

Fig. 4.

Exosomes distribution across biological fluids

Among all the features considered, the most relevant for this work is drug loading, which leads to another classification based on the type of biomolecule contained. This aspect will be discussed in the following sections, focusing on four major categories: metabolites, proteins, nucleic acids, and lipids. As illustrated in Fig. 5, exosomes can encapsulate a wide variety of biomolecules. This figure also provides representative examples of each category, serving as a visual framework that complements the detailed discussion.

Fig. 5.

Overview of clinically relevant and candidate exosomal biomarkers across different types of cancer. The schematic representation illustrates the diverse molecular cargo of exosomes, including proteins, miRNAs, lncRNAs, circRNAs, lipids, and metabolites associated with tumor development and progression. Biomolecules are organized according to cancer type, highlighting their potential diagnostic relevance. Filled shapes represent clinically validated biomarkers, whereas outlined shapes indicate biomolecules associated with cancer-related processes but lack robust clinical validation. The plus (+) symbol denotes biomarkers commonly used in combination panels to improve diagnostic performance. This figure emphasizes the heterogeneity of exosomal cargo and its implications for liquid biopsy and precision oncology

Variations in isolation methods, sample processing, and analytical platforms lead to significant inconsistencies in the biomarker profiles reported across studies. In addition, many biomarkers reported in the literature have not been validated in large, independent patient populations, limiting the confidence in their specificity and reproducibility. The lack of standardized normalization approaches and reliable internal controls further complicates cross-study comparisons, ultimately restricting their applicability in diagnostic settings.

Exosomal metabolites as biomarkers

In cellular metabolism, nutrients are converted into intermediate metabolites that participate in cell signaling processes. However, the importance of metabolite production goes beyond serving as fuel sources or building blocks, as mutations in genes associated with metabolism and metabolite production have been shown to cause diseases ranging from metabolic disorders to tumor development [46]. There is a correlation between elevated levels of specific metabolites and different types of cancer; therefore, the detection and quantification of metabolites in exosomes from liquid biopsies can serve as a diagnostic and prognostic tool.

Within the tumor microenvironment, which is often characterized by hypoxia, cancer cells rely on alternative metabolic pathways for energy production. In this context, exosomes play a crucial role in metabolic reprogramming by transporting essential metabolites that serve as carbon and energy sources, as in macropinocytosis in recipient cells, thereby promoting cancer cell survival and proliferation [47]. In cancer cells, it is well established that their metabolism is significantly altered compared to that of healthy cells. A characteristic example of this is the “Warburg effect,” in which these cells prefer aerobic glycolysis even in the presence of abundant oxygen. This results in the production of less ATP per glucose molecule but at a faster rate, allowing for accelerated growth [48]. The metabolites contained in exosomes are transported from one cell to another and, in addition to promoting the Warburg effect in tumors, they contribute to the complex carbon metabolism of cancer cells.

Despite these findings, the clinical application of exosomal metabolites as biomarkers remains limited. Most available studies are exploratory, and standardized platforms for their quantification are still lacking. Although a wide range of metabolic signatures provides relevant insights into tumor biology (as illustrated in Fig. 5), only a limited number have been validated in large clinical cohorts.

Exosomal proteins as biomarkers

Proteins are abundant components of exosomes, and their composition can provide valuable information about cellular origin and pathological state. In cancer, many exosomal proteins are either tumor-specific or present at altered levels, reflecting the molecular signature of malignancy and aiding in diagnosis, prognosis, and therapeutic monitoring. Numerous exosomal proteins have been associated with cancer development and progression; however, only a limited subset has undergone robust clinical validation. Those evaluated in clinical settings have demonstrated promising performance in cancer diagnosis, prognosis, and therapeutic monitoring.

Among exosomal protein biomarkers, glypican-1 (GPC-1) is among the most extensively studied and clinically validated candidates, particularly in pancreatic cancer [49]. GPC-1–positive exosomes have been shown to distinguish cancer patients from healthy individuals with high sensitivity and specificity, highlighting their potential for early detection. Subsequent studies have reported diagnostic performance with an area under the curve (AUC) exceeding 0.90, along with sensitivity above 90% and specificities around 80%, supporting its robustness as a biomarker [50]. Moreover, circulating levels of GPC-1–positive exosomes have been reported to correlate with tumor burden and disease progression, suggesting a role not only in diagnosis but also in prognosis and disease monitoring [50]. However, despite these promising findings, subsequent studies have reported variability in diagnostic performance, emphasizing the need for further validation in large, independent cohorts before routine clinical implementation [49].

In breast cancer, for example, exosomes often contain proteins like HER2 and EGFR, both well-established markers of this cancer type. HER2 in plasma-derived exosomes has been proposed as a promising biomarker, showing a sensitivity of 45.1% and a specificity of 97.22%. These values highlight its potential diagnostic utility; however, given its limited sensitivity, its use in combination with other biomarkers in multi-marker panels is recommended [51]. The presence of these proteins in patient-derived exosomes underscores the utility of these vesicles as a reflection of tumor biology [52]. Similarly, in colorectal cancer, elevated expression of EpCAM, an epithelial marker, in exosomes highlights their epithelial origin and their potential for detecting tumor-derived vesicles in circulation [53]. It is important to highlight that exosomal proteins not only act as passive biomarkers but also as active mediators of tumor progression. For instance, exosomes carrying EGFR or PD-L1 can modulate the immune response by promoting regulatory T cell activity and suppressing cytotoxic T cell function, thereby facilitating immune evasion [54]. These immunosuppressive cells dampen the antitumor activity of CD8 + T cells, thereby facilitating cancer growth, angiogenesis, and metastasis [55].

PD-L1 is considered an emerging biomarker, as available studies report variable results across different populations and methodologies, preventing the standardization of diagnostic parameters such as sensitivity and specificity. For instance, in pancreatic cancer, a sensitivity of 14% and a specificity of 94% have been reported, indicating high confirmatory value but limited utility as a standalone diagnostic biomarker. In this context, its use in combination with other biomarkers within multi-marker panels is more appropriate.

Melanoma-derived exosomes reveal another layer of complexity. They are known to contain PD-L1, a protein that is not only diagnostically relevant but also predictive of immunotherapy response. Alongside PD-L1, calcium-binding proteins such as S100B have also been detected, offering additional biomarkers with diagnostic and prognostic value [56, 57].

Together, these findings highlight the dual role of exosomal proteins as both molecular fingerprints of distinct cancer types and active mediators of tumor progression, immune modulation, and therapeutic resistance.

Exosomal nucleic acids as biomarkers

RNAs encapsulated in exosomes, especially non-coding RNAs, are of great interest due to their stability and regulatory role in cancer progression [58]. The following non-coding RNAs, along with the types of cancer they can be detected through the analysis of exosomes derived from liquid biopsies, are discussed below.

miRNAs

miRNAs are key post-transcriptional regulators of gene expression. They can act as oncogenes or tumor suppressors by binding complementary sequences in mRNAs, inducing degradation or inhibiting translation, thereby modulating processes such as proliferation, cell differentiation, and apoptosis [58]. An overview of the most relevant exosomal miRNAs associated with different cancer types is presented in Fig. 6, and clinically relevant exosomal miRNAs are summarized in Table 4.

Fig. 6.

Exosomal miRNAs associated with cancer diagnosis. The schematic summarizes miRNAs identified across multiple cancer types. Filled shapes indicate clinically validated miRNAs, outlined shapes represent non-validated candidates, and the plus (+) symbol denotes their use in biomarker panels. The figure highlights the diversity and clinical potential of exosomal miRNAs in liquid biopsy

Table 4.

Clinically validated exosomal molecular biomarkers and their diagnostic performance metrics (sensitivity, specificity, and AUC)

Exosomal Biomarker	Cancer Type	Sensitivity (%)	Specificity (%)	AUC	Ref.
Proteins
GPC-1	Pancreatic	90	80	0.90	[50]
HER2	Breast cancer	45	97.22	-	[51]
PD-L1	Pancreatic	14	94	-	[59]
miRNAs
miR-21	Pancreatic	85	84	0.91	[60]
	Gastric	65	77	-
	Breast	76	85	-
	NSCLC	70	78	0.85
miR-141/miR-375	Prostate	-	-	~ 0.90
miR-1246/miR-150-5p	Ovarian	100	91.7	0.964	[61]
miR-1246, miR-663b, miR-4730, miR-642a-3p, miR-658, miR-486-3p, miR-1207-5p, miR-4419b, miR-6124	Ovarian	82	91	0.86	[62]
miR-200b/miR-200c	Ovarian	-	-	0.784	[63]
miR-1246	Hepatocellular carcinoma	-	-	0.97	[64]
	Breast	-	-	0.967	[65]
miR-150-5p	Prostate	-	-	0.817	[66]
	Colorectal	-	-	0.87	[67]

Parameters such as sensitivity, specificity, and the AUC are essential for validating diagnostic biomarkers, as they reflect their discriminative ability and provide a quantitative assessment of diagnostic performance across different clinical settings. Among these, miR-21 is one of the most extensively studied exosomal biomarkers, with substantial clinical evidence across multiple malignancies, including pancreatic, gastric, breast cancer, and non-small cell lung cancer (NSCLC). In these contexts, it has shown diagnostic AUC ranging from 0.65 in gastric cancer to up to 0.85 in pancreatic cancer. However, despite its relatively high sensitivity, its clinical utility is limited by low specificity, as it is broadly overexpressed across diverse tumor types [60]. Niemira et al. developed a model based on two miRNA expression datasets. This model discriminated between patients with ovarian cancer and lung cancer (AUC 0.868, sensitivity 90%, and specificity 83.5%) or colon cancer (AUC 0.864, sensitivity 89.8%, and specificity 83.0%) [61].

miR-141 and miR-375 have been consistently associated with prostate cancer progression and metastatic disease. While individual diagnostic performance of these miRNAs remains moderate, with reported AUC values typically ranging from 0.70 to 0.85, their combined use in biomarker panels significantly improves diagnostic accuracy, reaching AUC values close to 0.90 in certain cohorts. These findings support their potential utility as part of multi-marker diagnostic strategies rather than as standalone biomarkers [68].

In the case of ovarian cancer, an extensive panel of elevated or altered miRNAs has been identified in exosomes. miR-200c, miR-200b, and miR-200a are known for their role in epithelial-mesenchymal transition (EMT), and their deregulation in exosomes may indicate tumor progression and chemotherapy resistance. At the same time, the miR-200 family serves as a diagnostic marker when detected at high levels in blood [58]. Both miR-205 and miR-141-3p have been associated with EMT and the promotion of vascularization in endothelial cells, and have been used as diagnostic and prognostic markers [69].

In colorectal cancer, miR-200c-3p has been shown to inhibit tumor cell invasion and migration when expressed at elevated levels [58]. miR-4323, miR-4284, miR-1290, and miR-1246 are serum-derived miRNAs also used in diagnosis, as they are present only in colorectal cancer patients [70]. miR-106b-3p, extracted from serum exosomes, is used for diagnosis and prognosis, as it is closely linked to metastasis-promoting activity through downregulation of DLC-1 [71].

Lung cancer research has increasingly focused on non-invasive methods such as sputum analysis, since invasive procedures like biopsies carry risks such as tumor collapse [72]. miR-200b, miR-200c, miR-141, and miR-375 are elevated in patients with lung cancer [72]. miR-1246 is linked to lymph node metastasis and TNM staging [73]. miR-96 from plasma exosomes associates with diagnosis and prognosis of radioresistance in NSCLC [74].

In breast cancer, miR-21-5p, miR-423, miR-424, let-7i, and miR-660, isolated from urinary exosomes, exhibit tumor-suppressor effects; therefore, the panel can serve as a prognostic marker [75, 76].

Cervical cancer has also been linked to specific miRNA profiles [77]. miR-21 and miR-146a are usually found at high levels in vaginal lavage samples [78]. miR-200c inhibits invasion, migration, and cell proliferation by targeting MAP4K4 [58].

Prostate cancer shows a strong connection with miR-21, which promotes cell proliferation, EMT, invasion, and metastasis while inhibiting apoptosis, making it a useful marker of tumor aggressiveness [79]. miR-375 plays a role in invasion and metastasis and is also linked to cell growth, but may have a dual role, depending on androgen-mediated regulation and cellular context [80]. Both miR-125a-5p and miR-141-5p act as tumor suppressors; their levels in plasma exosomes aid in cancer diagnosis and show consistent associations across multiple cohorts. Nevertheless, despite these promising findings, standardized diagnostic performance metrics for these miRNAs remain limited [81]. miR-375 and miR-451a in urinary exosomes permit the diagnosis and prediction of prostate cancer progression [82].

In oral squamous cell carcinoma, miR-155, miR-21, and miR-126 are used for diagnosis and prognosis due to their association with PTEN suppression and Bcl-6 expression, both of which are tumor suppressors [83].

Finally, in thyroid cancer, miR-200c presents a more ambiguous profile. Some studies suggest that it induces apoptosis and inhibits proliferation, while others indicate that it downregulates PTEN, leading to increased proliferation and invasion. Therefore, its use in thyroid cancer diagnosis should be complemented with monitoring other exosomal biomarkers [58]. Although these findings are promising, many of these biomarkers remain in early-stage validation and have not yet been fully implemented in routine clinical practice.

Long non-coding RNAs (lncRNAs)

lncRNAs are RNA molecules longer than 200 nucleotides that do not code for proteins. They regulate gene expression, and their deregulation is implicated in the onset and progression of various cancers, positioning them as critical molecular players in carcinogenesis [58]. A considerable number of exosomal lncRNAs have been associated with multiple cancer types and have been proposed as potential biomarkers. However, their clinical validation remains limited. Most studies are based on small patient cohorts and lack standardized methodologies, which restricts their current applicability in clinical settings.

In gastric cancer, one notable lncRNA, lncRNA-GC1, is overexpressed and is linked to cell proliferation, migration, and invasion. It often affects the transcription of tumor suppressor genes [84]. Bladder cancer research has identified ELNAT1 as an lncRNA present in urinary exosomes, with observational studies in patients evaluating its utility for preoperative prediction of lymph node metastasis [85]. In prostate cancer, HOXD-AS1 plays a significant role in determining if patients will develop castration resistance, a condition in which the cancer continues to progress despite androgen deprivation therapy. This lncRNA helps clinicians anticipate treatment failure in patients receiving therapy aimed at suppressing androgen hormones, which are known to fuel tumor growth [86]. Colorectal cancer is another area where several lncRNAs have been identified as potential diagnostic markers. Several lncRNAs, including LNCV6116, LNCV698390, and LNCV638772, have been reported to be highly expressed in colorectal cancer patients. Although these molecules have not yet achieved robust clinical validation, they have shown potential as future diagnostic biomarkers [87].

In lung cancer, HAGLR (HOXD Antisense Growth-Associated Long Non-Coding RNA) is of particular interest due to its prognostic value. Reduced expression of HAGLR has been associated with poor outcomes, suggesting a protective or stabilizing role in tumor biology. This effect seems to be mediated by its interaction with the miR-320d/E2F1 axis, highlighting the complex regulatory networks in which ncRNAs participate and their capacity to drive malignancy when dysregulated [88].

Circular RNAs (circRNAs)

circRNAs are a class of non-coding RNAs characterized by a closed-loop structure. They are abundant and stable in exosomes, and several candidates, such as circ0109046 and hsa_circ_0002577, have been associated with endometrial cancer. However, most available evidence is derived from small, exploratory studies, and their diagnostic performance has not yet been consistently established in large patient cohorts [72].

Exosomal lipids as biomarkers and challenges in detection

Although less studied than proteins and nucleic acids, exosomal lipids may also reflect changes in the membrane composition of tumor-origin cells and contribute to intercellular communication. Lipids such as ceramide, phosphatidylserine, phosphatidylethanolamine, and cholesterol are key components of the exosomal lipid bilayer, and altered proportions could indicate pathological states. More research is needed to establish specific lipid profiles as diagnostic biomarkers for particular cancer types [89].

Exosome-based diagnostic platforms have ushered in a paradigm shift toward precision oncology, offering unprecedented opportunities for early detection, prognosis, and personalized treatment strategies. However, despite their immense potential, these platforms face several critical technical challenges that must be addressed to facilitate widespread clinical adoption [90]. One of the foremost obstacles is the lack of standardized protocols for exosome isolation. Current methods vary significantly in efficiency and purity, resulting in inconsistent yields and inconsistent biomarker detection. Moreover, normalizing sample volume and concentration remains difficult, complicating comparisons across studies and patient samples. The identification and validation of appropriate internal controls are also essential for ensuring assay reliability and reproducibility [90].

In addition to technical variability, large-scale clinical validation and inter-laboratory reproducibility have yet to be fully established. Without robust, multi-center studies confirming the sensitivity and specificity of exosome-based biomarkers, regulatory approval and clinical confidence remain limited [91]. Another key challenge lies in the debate over the use of universal biomarkers versus tumor-specific signatures. While universal biomarkers could streamline diagnostics, tumor-specific markers promise higher precision but require extensive characterization across diverse cancer types [92].

To address these issues, the development of automated, high-throughput platforms with enhanced sensitivity is imperative. Such systems would not only improve standardization but also accelerate the translation of exosome diagnostics into routine clinical workflows.

Parallel to advances in diagnostics, therapeutic applications of exosomes are evolving rapidly. Natural exosomes, due to their intrinsic biocompatibility and ability to transport complex molecular cargoes like miRNAs, mRNAs, and proteins, are ideal candidates for targeted drug delivery [37]. However, their clinical utility is limited by challenges in large-scale isolation, batch-to-batch variability, and difficulties in efficiently and selectively loading cargo.

To circumvent these limitations, researchers have developed exosome-mimetic nanocarriers, synthetic constructs that emulate the beneficial properties of natural exosomes while offering enhanced scalability and customizability. These engineered nanocarriers can be precisely loaded with therapeutic agents and functionalized for targeted delivery, improving treatment specificity and minimizing off-target effects [37]. Looking ahead, significant research efforts must focus on optimizing the production and purification of both natural and synthetic exosomes, refining cargo loading techniques, and improving targeting accuracy. Advances in microfluidics, nanotechnology, and molecular engineering will be critical in overcoming current bottlenecks. Ultimately, integrating exosome-based diagnostics with therapeutic delivery platforms promises a future where cancer care is truly personalized, enabling earlier diagnosis, tailored treatment regimens, and improved patient outcomes. Achieving this vision will require collaborative efforts spanning basic research, clinical trials, regulatory frameworks, and industrial manufacturing [93].

Clinical applications and ongoing trials

Exosomes have gained momentum in clinical translation, particularly in oncology, with over 100 registered clinical trials exploring their diagnostic and therapeutic potential. Several ongoing trials evaluate exosomal components as non-invasive biomarkers for early cancer detection, treatment monitoring, relapse prediction, and targeted therapy guidance. In diagnostics, exosomal nucleic acids (miRNAs, mRNAs, lncRNAs, and DNA) are being extensively studied across various cancers (Table 3). Numerous trials focus on NSCLC, pancreatic, and prostate cancers, highlighting the diverse sources of natural exosomes (plasma, serum, urine) and biomarker types. Specifically, NSCLC is the subject of multiple studies examining exosomal proteins, mRNA, lncRNA, and DNA for early detection (e.g., NCT02890849, NCT03830619, NCT03228277, NCT04529915, and NCT03542253). Pancreatic cancer trials (NCT03821909, NCT02393703) analyze exosomal mRNAs and proteins from portal venous blood to enhance early detection. Prostate cancer research includes large validation trials of urinary exosomal RNA panels correlating with Gleason grade (NCT02702856).

Other cancers under investigation include bladder cancer (NCT06193941, NCT05559177, NCT05270174), utilizing urinary exosomal lncRNAs such as ELNAT1, exosomal RNA, and Chimeric exosomal Tumor Vaccines, respectively, to assess nodal metastasis risk.

Beyond diagnostics, exosomes show promise as therapeutic platforms. Immunotherapeutic exosome vaccines are being tested in NSCLC, with Phase II trials (NCT01159288) reporting improved progression-free survival and favorable safety profiles. siRNA-loaded exosomes are being evaluated for targeted gene silencing in metastatic pancreatic cancer (NCT03608631), with a focus on KRAS G12D mutations. Additionally, plant-derived natural exosomes loaded with curcumin are undergoing Phase I safety trials in colon cancer patients (NCT01294072). Cutting-edge chimeric exosomal vaccines that combine antigen-presenting and tumor-derived exosomes are also entering clinical evaluation in bladder cancer (NCT05559177).

Together, these diverse clinical investigations, as summarized in Table 3, underscore the growing role of exosomes as versatile biomarkers and innovative therapeutic vehicles, offering new avenues for precision oncology and personalized medicine. By highlighting ongoing and completed clinical trials, Table 5 provides critical insight into how exosome-based strategies are advancing from the bench to the bedside. Continued advancements in production, characterization, and targeted delivery will be key to fully harnessing their potential in routine clinical practice.

Table 5.

Clinical trials elucidating the value of the use of exosomes in cancer clinics

Study ID	Cancer Type	Exosome Source	Biomarker Type/Intervention	Phase	Study Type
NCT02890849	NSCLC	Plasma	Exosomal mRNA	NS	Interventional
NCT03830619	NSCLC	Serum	Exosomal lncRNA	NS	Observational
NCT03228277	NSCLC	bronchoalveolar lavage fluid	Exosomal DNA	Phase 2	Interventional
NCT04529915	NSCLC	Blood Plasma	Various exosomal proteins	NS	Observational
NCT03542253	Early Lung Cancer	Blood	Exosomal miRNAs	NS	Observational
NCT03821909	Pancreatic Cancer	Portal venous blood	Exosomal mRNAs	NS	Observational
NCT02393703	Pancreatic Cancer	Blood, pancreatic fluid, and pancreas tissue	Exosomal RNA/protein biomarkers	NS	Observational
NCT02702856	Prostate Cancer	Urine	Exosomal RNA panel (correlated with Gleason grade)	NS	Observational
NCT03108677	Lung Metastasis/Osteosarcoma	Blood	Exosomal RNA	NS	Observational
NCT05270174	Lymphatic Metastasis /Bladder Cancer	Urine	Exosomal lncRNAs (ElNAT1)	NS	Observational
NCT06193941	Bladder Cancer	Urine	Exosomal RNA	NS	Observational
NCT01159288	NSCLC	Dendritic cell-derived exosomes	Tumor Antigen-loaded Dendritic Cell-derived Exosomes	Phase 2	Interventional
NCT03608631	Pancreatic Cancer	Blood	KRASG12D siRNA-loaded exosomes	Phase 1	Interventional
NCT01294072	Colon Cancer	Plant-derived exosomes	Curcumin-loaded exosomes	Phase 1	Interventional
NCT05559177	Metastatic Bladder Cancer	Blood	Chimeric exosomal Tumor Vaccines	Phase 1	Interventional

The data were taken from ClinicalTrials.gov

NS Not Specified

Despite promising preclinical outcomes, the clinical translation of exosome-based and exosome-mimetic nanocarriers remains limited by several critical challenges. One of the primary barriers is the lack of standardized large-scale production methods that ensure batch-to-batch reproducibility and purity [94]. The inherent heterogeneity of extracellular vesicles, influenced by cell source and isolation techniques, complicates their characterization and regulatory approval. Additionally, issues related to scalability, storage stability, and cost-effectiveness continue to hinder industrial translation [42].

From a regulatory standpoint, the classification of exosome-based systems remains ambiguous, as they may be considered biologics, drug delivery systems, or advanced therapy medicinal products depending on their composition and intended use. This lack of clear regulatory frameworks delays clinical approval and complicates quality control requirements [95].

Furthermore, although several clinical trials have explored the therapeutic value of natural exosomes, their efficacy has been modest, with challenges including rapid clearance, limited targeting efficiency, and variability in cargo loading [96]. Safety concerns such as immune activation, complement-related pseudoallergy, and potential off-target effects must also be carefully addressed [18]. These limitations highlight the need for more robust engineering strategies and standardized evaluation protocols to facilitate successful clinical translation.

Most studies remain in the early stages of clinical development, with limited progression toward late-phase trials or regulatory approval. Furthermore, heterogeneity in trial design, endpoints, and biomarker selection complicates the interpretation and comparison of clinical outcomes. Importantly, many of the advances described for EMNs are still primarily based on in vitro studies and animal models, in which these systems have demonstrated promising drug-delivery capabilities, tumor-targeting, and reduced toxicity. To date, only a limited number of early-phase clinical trials have evaluated extracellular vesicle-based therapies, focusing mainly on safety and feasibility rather than therapeutic efficacy. This substantial gap between preclinical success and clinical validation underscores the urgent need for standardized evaluation frameworks, harmonized clinical guidelines, and large-scale translational studies.

Comparative advantages of exosome-mimetic nanocarriers over traditional nanoparticles in drug delivery

EMNs have emerged as one of the most promising next-generation platforms for drug delivery by combining the unique biological features of natural extracellular vesicles (EVs), including immune evasion, cell-specific targeting, and biocompatibility, with the scalability, tunability, and reproducibility of synthetic nanotechnological systems [97–99].

In contrast, traditional nanocarriers such as polymeric nanoparticles, liposomes, and lipid-based nanoparticles have been extensively studied for cancer drug delivery and imaging applications [100–102]. However, these systems often face limitations, including rapid immune clearance, nonspecific biodistribution, limited ability to cross biological barriers, and potential off-target toxicities [103].

By utilizing bioinspired design, EMNs can overcome many of these hurdles, offering improved pharmacokinetics, intrinsic targeting, and more efficient drug delivery. This section discusses their comparative advantages in the key domains, including immune evasion and circulation time, biocompatibility and safety, natural targeting and uptake, drug loading and release profiles, and clinical translation potential (Fig. 7).

Immune evasion and extended circulation

One significant limitation of systemic nanoparticle delivery is rapid clearance by the mononuclear phagocyte system, particularly in the liver and spleen. Following intravenous administration, nanoparticles undergo opsonization, in which serum proteins such as complement components, immunoglobulins, and fibrinogen bind to their surfaces, marking them for phagocytosis [62], leading to short circulation half-lives and reduced delivery to target tissues.

To overcome this limitation, synthetic nanocarriers have traditionally been modified with PEG, which sterically hinders opsonin adsorption and prolongs circulation time [104]. PEGylation offers several advantages, including enhanced blood half-life, better exploitation of the enhanced permeability and retention (EPR) effect, partial ability to cross the BBB, and reduced nonspecific immune interactions. These properties have enabled PEGylated formulations like Doxil^® and Onivyde^® to achieve clinical success.

While PEGylation offers advantages, it also has significant drawbacks. Repeated administration can induce accelerated blood clearance (ABC), mediated by anti-PEG IgM and IgG antibodies, which promote rapid uptake by macrophages [64, 105]. Interestingly, one-third of the human population has pre-existing anti-PEG antibodies from prior exposure to pharmaceuticals and cosmetics, potentially limiting efficacy from the first dose onward [105]. In cases of pre-existing anti-PEG antibodies, immune complexes can lead to hypersensitivity reactions, including rare but severe anaphylaxis. At the same time, PEG itself is non-biodegradable and may accumulate in tissues [67, 106]. Alternative stealth coatings, such as poly (glycerols), poly (oxazolines), poly (amino acids), and sugar-based polymers, are under investigation, but none have received FDA approval yet [64].

Exosome-mimetic nanocarriers bypass many of these issues by incorporating endogenous membrane proteins that provide immune-evasive signals. For example, CD47 acts as a “don’t eat me” signal to macrophages, reducing phagocytosis; high levels of CD47 trigger the CD47-SIRPalpha interaction, enabling immune evasion [107, 108]. Additionally, tetraspanins (CD9, CD63, CD81) and integrins form a natural protein corona that prevents complement activation and nonspecific immune recognition [109]. Collectively, these properties enable EMNs to achieve prolonged systemic circulation and enhanced tumor accumulation compared to liposomes or polymeric nanoparticles.

Although exosomes can partially evade immune recognition, they are cleared from the circulation quite rapidly in practice, which remains a significant challenge for their therapeutic use [12]. After intravenous administration, their presence in the bloodstream is typically short-lived, as they are quickly taken up by phagocytic cells such as macrophages and neutrophils. As a result, instead of remaining in circulation long enough to reach their intended targets, they tend to accumulate in organs associated with the mononuclear phagocyte system, particularly the liver and spleen, where they can be retained for extended periods. While various surface modification strategies have been explored to reduce this uptake, the mechanisms governing their clearance remain poorly understood and continue to affect both their bioavailability and overall therapeutic performance [12]. In this respect, simply extending circulation time is not enough; it is equally important to understand how these clearance pathways influence their distribution throughout the body and their ability to reach target tissues effectively.

Although natural exosomes are also rapidly cleared and tend to accumulate in the liver, subsets enriched with CD47 or specific integrins show markedly improved pharmacokinetics, with measurable circulation even 24 h post-injection [108, 110]. Importantly, EMNs can be engineered to replicate and even amplify these immune-evasive properties, offering a level of control that is not possible with unmodified exosomes.

Biocompatibility and safety

Biocompatibility is essential for any clinically translatable nanocarrier. Polymeric nanoparticles, such as PLGA-based systems, are generally regarded as safe, but their acidic degradation products (lactic and glycolic acids) may cause localized inflammation or cytotoxicity at high doses [111]. In vitro and in vivo studies have shown that PLGA nanoparticle toxicity is influenced by surface charge and the stabilizer used. For instance, negatively charged particles are associated with a higher pro-inflammatory cytokine response, likely due to increased cellular uptake [112].

Similarly, dendrimers are highly versatile carriers but exhibit dose-dependent cytotoxicity that depends on their generation, surface charge, and terminal group chemistry. Cationic dendrimers show higher toxicity, which has been mitigated through PEGylation, acetylation, or conjugation with carbohydrates to shield surface charges [113].

Liposomes, while considered relatively safe, can trigger complement activation-related pseudoallergy (CARPA) and infusion-related reactions (IRs). Briefly, complement activation leads to the release of C3a and C5a anaphylatoxins, provoking mast cell and basophil degranulation and resulting in symptoms ranging from flushing and fever to bronchoconstriction and hypotension [114–117]. PEGylated liposomes, such as the approved Doxil^®, are especially prone to CARPA due to their elongated morphology and high cholesterol content, which enhance complement activation [117].

In contrast, EMNs are inherently biocompatible; their lipid composition and protein cargo are derived from natural membranes, minimizing the risk of complement activation and avoiding the need for synthetic surfactants [118, 119]. Also, they can efficiently encapsulate hydrophilic and hydrophobic drugs, proteins, and nucleic acids without altering their physicochemical stability [120, 121].

In particular, immune-derived exosome mimetics (IDEMs) offer a balance between safety and scalability. Compared with natural exosomes, IDEMs have been shown to yield 2.5× more particles, achieve higher drug encapsulation efficiency (28% vs. 17%), and deliver comparable or superior therapeutic effects at lower drug doses, thereby potentially reducing systemic side effects [122–124].

Natural targeting and cellular uptake

Traditional nanoparticles typically require surface functionalization with antibodies, peptides, or aptamers to achieve targeted delivery, which increases production complexity and costs while still leaving off-target interactions [125–128]. EMNs, on the other hand, inherit the natural tropism of their parental cells.

For example, tumor-derived EMNs preferentially accumulate in the tumor microenvironment through integrin–extracellular matrix interactions and specific receptor–ligand recognition [129], and exploit endogenous internalization pathways such as clathrin-mediated endocytosis, caveolae-mediated uptake, and macropinocytosis, resulting in more efficient cellular entry.

Moreover, functionalized EMNs enhance their inherent properties. In a recent study, functionalized EMNs loaded with miR-145 and decorated with integrin α6β4 achieved a 5-fold increase in gene expression compared to unmodified EMNs and an 800-fold increase compared to free miR-145. Also, confocal microscopy confirmed superior cellular internalization, while in vivo studies demonstrated efficacy similar to that of tumor-derived exosomes, with improved safety and production yield [37].

The intrinsic targeting capability reduces the need for artificial ligand conjugation and can lower off-target toxicity. Nonetheless, there are still some challenges regarding reproducibility, as EMN properties depend on the cell source, and there is ongoing concern over transferring unwanted biomolecules with tumor-derived material [130, 131].

Drug loading, stability, and release profiles

An ideal drug delivery platform should allow high loading efficiency, preserve cargo stability, and provide controlled release. Liposomes are excellent for encapsulating hydrophilic drugs but often require surface modification to load hydrophobic molecules efficiently. Polymeric nanoparticles can encapsulate a broad range of molecules but frequently display a burst release profile, potentially leading to suboptimal pharmacodynamics [132, 133].

EMNs combine the strengths of both systems, efficiently encapsulating diverse payloads, including small molecules, nucleic acids, and proteins, while shielding them from enzymatic degradation [37]. Additionally, their lipid bilayers facilitate controlled and stimulus-responsive release. For instance, EMNs derived from mesenchymal stem cells exhibit pH-sensitive drug release, with enhanced release in acidic tumor microenvironments (pH 5.5) compared to physiological conditions (pH 7.4) [134].

Fig. 7.

Comparative advantages of exosome-mimetic nanocarriers over traditional nanocarriers. Schematic comparison of EMNs, liposomes, and polymeric nanoparticles (PNPs) across five critical parameters: (A) Immune evasion: EMNs express CD47/tetraspanins; liposomes rely on PEGylation (risk of ABC); PNPs require stealth coatings. (B) Targeting: EMNs inherit parental tropism; liposomes/PNPs need ligand conjugation. (C) Biocompatibility: EMNs show an excellent safety profile; liposomes may cause CARPA; PNPs can trigger inflammation at high doses. (D) Drug loading and release: EMNs carry diverse cargos with controlled release; liposomes favor hydrophilic drugs; PNPs risk burst release. (E) Clinical translation: EMNs face scalability limits; liposomes are FDA-approved; PNPs are in clinical trials with established regulations

Several studies have reported that EMNs can sustain drug release for 48–72 h, minimizing burst release and maintaining therapeutic concentrations for longer periods [134]. The controlled release profile is attributed to their biomimetic lipid structure, which undergoes gradual destabilization in response to biological cues such as pH, enzymatic activity, or membrane fusion events [6, 37, 135].

Comparative summary

Comparing EMNs with more established drug delivery systems is crucial to understanding where they truly add value. When examining parameters such as immune evasion, targeting, biocompatibility, circulation time, and drug loading, we realize that EMNs combine the advantages of two worlds: the biological sophistication of natural exosomes and the scalability of synthetic nanoparticles. The “hybrid” nature helps them address many of the limitations seen with liposomes and polymeric systems, giving them a unique edge as a next-generation drug delivery platform. Table 6 brings these features together and highlights why EMNs are considered a bridge between biology-inspired and fully synthetic systems.

Table 6.

Comparison of exosome-mimetic nanocarriers, liposomes, and polymeric nanoparticles across key drug delivery features

Feature	Exosome-Mimetic nanocarriers	Liposomes	Polymeric nanoparticles
Origin	Biologically inspired, mimic exosome membrane protein/lipids	Synthetic lipid bilayers, sometimes PEGylated	Synthetic polymers (PLGA, PEG-PLA)
Immune evasion	High-natural surface markers reduce clearance	Moderate-require PEGylation	Low-often recognized by the immune system
Targeting ability	Intrinsic, via exosome proteins and ligands	Limited, requires functionalization	Limited, requires ligand conjugation.
Biocompatibility	Excellent, low toxicity	Good, but may trigger complement activation	Variable, depends on polymer and dose
Circulation half-life	Long, due to immune evasion	Moderate (PEG-dependent)	Short-moderate
Drug loading	High versatility: hydrophilic, hydrophobic proteins, nucleic acids	Primarily hydrophilic, some hydrophobic	Broad versatility but prone to burst release
Clinical translation	Promising, but scale-up remains a challenge	Multiple FDA-approved	Several in trials, but limited approvals

Upon closer examination, the differences are not just a matter of composition. Traditional systems such as liposomes and polymeric nanoparticles often require extensive additional modifications to function as intended. Liposomes, for example, usually require PEGylation to remain in circulation for longer periods, but this comes with drawbacks, including immune-related effects such as complement activation or accelerated blood clearance upon repeated administration [120–123]. Polymeric nanoparticles offer significant flexibility in their design and what they can carry; however, issues such as burst release or unintended inflammatory responses are not uncommon, particularly depending on the materials used [136, 137].

EMNs, on the other hand, approach the problem from a different angle. Because they are built around a biomimetic membrane, they already contain proteins and lipids that naturally interact with biological environments. This design can affect how the body recognizes them, how they enter cells, and how they release their cargo, often without requiring as many additional surface modifications as might be expected [6]. Even so, this does not mean they solve everything. There are still important hurdles to consider, especially when it comes to producing them consistently at scale and ensuring reproducibility between batches, which are key aspects for any real clinical application [138, 139].

Taken together, these comparisons make it clear that EMNs are among the most promising tools for targeted drug delivery; however, their journey from bench to bedside will depend on solving practical hurdles, such as scaling up production and meeting stringent quality standards. In the next section, we explore how the field is addressing these challenges and where EMNs currently stand in terms of clinical development.

Clinical translation perspective

While liposomal formulations such as Doxil^® and Onivyde^®, as well as several polymeric nanomedicines, have already received clinical approval and commercialization [136, 140], EMNs remain largely at the preclinical stage, with only limited progress toward early clinical evaluation [6, 37, 135]. This contrast highlights the significant translational gap that still exists between biologically inspired nanocarriers and clinically established synthetic platforms.

Recent preclinical advances illustrate the therapeutic potential of EMNs and the growing convergence of nanotechnology with exosome-inspired design. For instance, macrophage-derived EMNs loaded with doxorubicin demonstrated enhanced tumor accumulation and reduced systemic toxicity [6]. EMNs delivering miR-145 showed significant antitumor effects in lung adenocarcinoma models [137], while bispecific EM-LNP systems co-delivering KRAS and TP53 siRNAs induced significant tumor regression in pancreatic cancer models [108, 126, 137]. Additional biohybrid strategies, such as exosome-camouflaged porous silicon nanoparticles for chemo-photothermal therapy, also achieved improved tumor targeting and ablation [141]. Collectively, these studies provide strong proof-of-concept value of EMNs in oncology; however, most evidence remains limited to in vitro or animal studies.

Despite these encouraging findings, several critical barriers continue to hinder clinical translation. Large-scale manufacturing remains challenging, particularly in terms of production yield, purification efficiency, sterility assurance, and batch-to-batch reproducibility. The biological complexity of membrane-derived systems also complicates quality control, characterization, and long-term storage stability [142]. Furthermore, regulatory classification remains uncertain, as EMNs may be considered drug delivery systems, biologics, combination products, or advanced therapeutic platforms depending on their composition and intended use [94]. Such ambiguity may delay approval pathways and increase Chemistry, Manufacturing, and Controls requirements [42].

Importantly, although natural extracellular vesicles have entered several early-phase clinical trials, these studies have primarily focused on safety and feasibility rather than clear therapeutic efficacy. Comparable clinical evidence for EMNs remains scarce. Therefore, future progress will require not only technological optimization but also robust comparative studies, predictive translational models, harmonized manufacturing standards, and well-designed clinical trials capable of demonstrating meaningful clinical benefit [28].

Overall, EMNs represent a promising next-generation platform that combines the scalability of synthetic nanocarriers with selected biological functionalities of natural vesicles. However, their ultimate success will depend on overcoming current manufacturing, regulatory, and clinical validation challenges before widespread medical adoption can be achieved [32, 37].

Challenges and innovations

Exosomes have been gaining prominence as new therapies for various diseases due to their ability to transport genetic or bioactive molecules to recipient cells, thereby mediating intercellular communication [138]. However, there are obstacles to overcome, the most notable being the transition from the laboratory to the clinic while maintaining product quality, functionality, and safety, and simultaneously achieving industrial-scale production [143].

First, the low natural exosome yield can pose a critical challenge. For example, most current protocols yield only micrograms of exosomes, which are insufficient for clinical applications that require doses of 10^8–10^12 exosomes per treatment [144, 145]. Even with extensive two-dimensional (2D) plate cultures, cell expansion is limited, and resources are intensive [1]. The move towards three-dimensional (3D) systems, such as bioreactors with microcarriers or hollow fibers, has been shown to increase exosome yield significantly. In recent studies, the combined use of 3D culture and tangential flow filtration (TFF) increased the yield 140-fold compared with conventional methods [138, 145].

The biology underlying exosome biogenesis is also a target for engineering to improve secretion. Genetic modifications that overexpress tetraspanins, such as CD9, or proteins such as STEAP3 and syndecan-4 have increased exosome release without compromising their functional profile [146]. Moreover, non-genetic strategies, such as small-molecule stimulation (e.g., Methyl Dopamine and norepinephrine), have been shown to induce up to 3-fold increases in exosome production or in the production of more mesenchymal stem cell derivatives, while maintaining their therapeutic properties [145, 147].

Another key challenge is the large-scale isolation and purification of exosomes. Traditional methods, such as ultracentrifugation, are not viable at the industrial level due to low recovery rates, high variability, and a risk of structural damage to exosomes [147]. Other methods, such as TFF and anion exchange chromatography (AEX), have shown better performance and reproducibility [148, 149]. TFF, for example, offers efficient concentration, removes contaminants, and allows for consistency between batches. The combination of TFF with chromatographic techniques, such as size-exclusion chromatography (SEC), has proven useful for obtaining high-purity exosomes suitable for clinical applications [138].

It is also important to note the lack of standardized regulatory guidelines on quality and characterization, as consensus on manufacturing guidelines for therapeutic exosomes often requires close communication between regulators and companies [138]. Standardized protocols or guidelines for the quality attributes of exosomes are needed, regardless of the originating cell line and culture conditions. This quality control should encompass purity, protein content, and residual cellular DNA (host cell proteins (HCP)/host cell DNA (HCD)), as well as biological activity [138, 150].

One of the main obstacles to advancing natural exosomes and EMNs toward clinical use is the challenge of producing them in large quantities while maintaining a quality that does not readily degrade [151]. This challenge can be particularly difficult because, unlike fully synthetic nanoplatforms, these systems are inherently affected by biological variability. Their final characteristics tend to depend not only on their formulation but also on the conditions under which they are produced [151, 152]. Parameters such as the donor or cell source, the number of passages, the composition of the culture medium, oxygen conditions, and the bioreactor configuration can affect vesicle secretion and, in turn, alter their molecular cargo, membrane composition, and biological activity [151, 152].

This variability can be further complicated by the methods used for vesicle recovery and purification. Commonly used methods, such as ultracentrifugation, tangential flow filtration, size-exclusion chromatography, and precipitation isolation, often differ in terms of the vesicle subpopulations they recover, as well as in terms of yield and purity. Therefore, it can be said that preparations obtained using different procedures may not be comparable, even when they are described within the same general category [151, 152].

Consistency between batches is also often another unresolved issue, as small variations during production can lead to noticeable changes in particle concentration, encapsulation efficiency, surface marker expression, or yield, directly affecting reproducibility, comparability, and regulatory acceptance [151, 153, 154]. For this reason, the industry should be moving toward more rigorous reporting standards, the use of complementary characterization tools, validated reference materials, and potency assays designed in accordance with the intended therapeutic application [153, 154].

In addition to the challenges associated with scale-up, heterogeneity represents one of the main barriers in exosome research and application. This phenomenon has been extensively studied, revealing that what has traditionally been considered a homogeneous population of extracellular vesicles is, in fact, a mixture of vesicular and non-vesicular components with diverse molecular properties and functions [155]. One of the main findings in this regard stems from the use of advanced techniques, such as density gradient fractionation and direct immunoaffinity capture (DIC), which have enabled the accurate differentiation between true exosomes and other extracellular elements. It has been shown that proteins commonly associated with exosomes, such as Argonaute 1–4, GAPDH, HSP90, and even histones, are absent in DIC-purified exosomes and are instead enriched in non-vesicular fractions [155, 156]. This issue would imply establishing, for future research, a more effective isolation method to accurately characterize exosome composition.

Proteomic and transcriptomic analysis also reveals differential distribution patterns between sEVs (small extracellular vesicles) and NV (non-vesicular) fractions [155]. For example, while systenin-1 and ALIX proteins are highly enriched in purified exosomes, proteins such as GAPDH, PKM, ENO1, and tubulins predominate in NV fractions. At the RNA level, miRNAs and other small non-coding RNAs show distinct profiles between exosomes and non-vesicular compartments, and many of the most abundant miRNAs are preferentially localized in NV fractions [156, 157].

Heterogeneity also extends to mimetic exosomes, such as cell-derived nanometric vesicles (CNVs), which are mechanically generated from HEK293 cells. Although these vesicles mimic some physical characteristics of exosomes, their cellular origin is remarkably diverse [158, 159]. Using cell lines and transported with organelle-specific fluorescent markers, Lee et al. (2020) found that while the majority of CNVs originate from the plasma membrane, a considerable proportion originates from other organelles such as lysosomes, mitochondria, the nucleus, the Golgi apparatus, and the endoplasmic reticulum. Variability in origin has significant functional implications, as CNVs derived from mitochondria or lysosomes can induce cytotoxic effects in recipient cells due to their intrinsic proapoptotic or degradative properties [159].

In addition, analyses using Western blot tracking of fluorescent particles (fluorescence NTA) confirmed the presence of multiple organelle-specific markers in all CNV fractions. The finding corroborates that mechanical manufacturing processes, such as air-spraying, indiscriminately fragment diverse subcellular membranes, resulting in a heterogeneous population of vesicles [159].

One of the main innovations is synthetic exosomes, which have emerged to overcome the limitations of natural exosomes, particularly in terms of large-scale production, uniformity, and functionalization [160]. Three main approaches have been developed through non-biotechnology technologies: top-down strategies to generate nanovesicles, bottom-up strategies to form mimetic exosomes from synthetic materials, and hybrid strategies to combine natural exosomes with liposomes, generating hybrid exosomes [160, 161].

Nanovesicles generated by methods such as sequential extrusion of cells through polycarbonate filters maintain the membrane and protein composition of the parental cells. They effectively mimic the biological complexity of natural exosomes, but with lower heterogeneity and a 100-fold higher yield [160, 162]. On the other hand, hybrid strategies, such as the fusion of synthetic liposomes with exosomes, allow the incorporation of specific features without compromising essential biological properties [160, 163].

The development of these synthetic platforms has enabled more precise engineering. Surface modification techniques have been applied, such as the addition of aptamers, peptides, or biotinylated molecules, and the controlled incorporation of therapeutic payloads via chemical or photo-dependent strategies, including the blue-light-activated protein system CRY2/CIB1. Similarly, hybrid exosomes are effective vectors for delivering gene editing systems, such as CRISPR-Cas9, thereby broadening their applicability in advanced therapies [163]. These multifunctional properties position exosome-mimetic nanocarriers as highly promising theranostic platforms, capable of integrating disease diagnosis, molecular imaging, real-time monitoring, and targeted therapy within a single nanosystem.

Although numerous technological advances have been proposed to overcome current limitations, many are still in the proof-of-concept phase and lack validation under clinically relevant conditions. Despite the promising therapeutic potential of these systems, several regulatory and quality-related challenges must be addressed to enable their clinical translation. As highlighted in recent regulatory-focused reviews, extracellular vesicle-based and exosome-mimetic products may fall under different regulatory categories depending on their biological source, degree of engineering, cargo incorporation, and intended therapeutic use, including biologics, drug delivery systems, combination products, or advanced therapy medicinal products [94, 95]. This regulatory heterogeneity underscores the need for clear product classification early in development.

For clinical translation, exosome-mimetic systems must comply with stringent Good Manufacturing Practice (GMP) standards and require robust Chemistry, Manufacturing, and Controls documentation. Manufacturing workflows should ensure traceability of raw materials, validated production processes, aseptic handling, scalability, and batch-to-batch reproducibility. In addition, donor cell origin or membrane source must be carefully controlled to minimize variability and ensure biosafety.

Critical quality attributes include particle size distribution, PDI, zeta potential, membrane composition, cargo consistency, encapsulation efficiency, sterility, endotoxin levels, residual impurities, storage stability, and biological potency. These parameters are essential to guarantee identity, purity, safety, and functional performance of the final product.

Advanced analytical techniques, including nanoparticle tracking analysis (NTA), dynamic light scattering (DLS), transmission electron microscopy (TEM), flow cytometry, Western blotting of vesicle-associated markers, and omics-based profiling, are recommended to comprehensively characterize these systems. Functional assays evaluating uptake efficiency, target engagement, and biological activity are also necessary, as physicochemical characterization alone may not predict therapeutic performance.

Overall, future regulatory approval of clinical-grade exosome-mimetic systems will depend on harmonized manufacturing standards, validated release criteria, and internationally accepted quality control frameworks tailored to these hybrid nanobiological products.

Conclusion and prospects

Cancer research has driven the development of novel drugs seeking greater efficacy, addressing unresolved challenges, and enabling targeted, stealthy delivery systems with high efficacy in in vivo models and excellent biocompatibility. Natural exosomes have emerged as an attractive alternative for drug delivery systems due to their unique natural composition, nanoscale size, and drug-carrying capacity. However, limitations such as reproducibility and nanoscale size control have hindered their technological advancement and pharmacokinetic performance. EMNs can utilize a synthetic core with a sophisticated surface architecture composed of membrane proteins or peptides that mimic the surface markers of native exosomes and proteins involved in the endosomal sorting complex required for transport. However, the development of nanomaterials is still in its early stages and is not yet advanced enough to fully understand the impact of this new technology, although progress is significant. The coupling of surface ligands requires additional technological control, which, like any chemical reaction, produces a degree of “yield”. This yield must be highly reproducible because it will determine therapeutic efficacy. Additionally, as with various nanosystems, biocompatibility aspects require exhaustive study to demonstrate short-, medium-, and especially long-term effects. Undesirable effects over prolonged periods are one reason some nanosystems have halted development. Others have faced the significant challenge of pharmacokinetic parameters, which are highly dependent on the conventional pharmaceutical form and route of administration. Some nanosystems travel long distances, losing efficacy due to their small size and a highly reactive biological environment. Prospects could include expanding preclinical studies of cell proliferation and viability, mechanisms of interaction in cancer cell lines, and animal model trials, with a particular focus on biodistribution, metabolism, and elimination, as well as clinical trials. Depending on the type and location of the cancer, the pharmaceutical form and route of administration should be adjusted accordingly.

Future efforts in biomimetic exosome-based nanocarriers should prioritize the development of reproducible, scalable, and clinically translatable systems rather than increasingly complex proof-of-concept designs [142, 164]. Although preclinical studies have yielded encouraging results, their translation to clinical use remains constrained by challenges in large-scale manufacturing, standardization, and regulatory approval. Techniques such as microfluidic-assisted fabrication and tangential flow filtration have shown promise in enhancing production efficiency and batch uniformity under controlled settings. Nonetheless, these approaches must be rigorously validated in accordance with GMP standards, particularly with respect to process reliability, contamination control, and overall cost. Equally important will be the integration of pharmacokinetic, biodistribution, and toxicity assessments under conditions that closely resemble real clinical scenarios. Comprehensive evaluation in more advanced models, including humanized systems and standardized tumor models, will be essential to confirm both therapeutic efficacy and safety. Effective collaboration between academic researchers, industry stakeholders, and regulatory authorities will be key to establishing clear frameworks and facilitating successful clinical translation. Addressing these challenges will be essential to unlock the full potential of EMNs in precision oncology.

Abbreviations

ABC

Accelerated blood clearance

APC

Antigen–presenting cell

ATP

Adenosine triphosphate

BBB

Blood–brain barrier

CA125

Cancer antigen 125

CD

Cluster of differentiation

CNV

Cell–derived nanovesicles

CRISPR

Clustered regularly interspaced short palindromic repeats

CT

Computed tomography

CTC

Circulating tumor cell

DLS

Dynamic light scattering

DNA

Deoxyribonucleic acid

DOX

Doxorubicin

EGFR

Epidermal growth factor receptor

EGFRvIII

Epidermal growth factor receptor variant III

ELISA

Enzyme–linked immunosorbent assay

EM

Lnp–exosome–mimetic lipid nanoparticle

EMN

Exosome–mimetic nanocarrier

EpCAM

Epithelial cell adhesion molecule

EPR

Enhanced permeability and retention

ESCRT

Endosomal sorting complex required for transport

EVs

Extracellular vesicles

FDA

Food and drug administration

GMP

Good manufacturing practice

HEK293

Human embryonic kidney 293 cells

HPLC

High–performance liquid chromatography

HSP

Heat shock protein

KRAS

Kirsten rat sarcoma viral oncogene homolog

lncRNA

Long non–coding RNA

LNP

Lipid nanoparticle

miRNA

Micro rna

MS

Mass spectrometry

MVB

Multivesicular body

NSCLC

Non–small cell lung cancer

NTA

Nanoparticle tracking analysis

NV

Non–vesicular fraction

PBS

Phosphate–buffered saline

PDI

Polydispersity index

PD

L1–programmed death ligand 1

PEG

Polyethylene glycol

PET

Positron emission tomography

PLGA

Poly(lactic–co–glycolic acid)

PSA

Prostate–specific antigen

RNA

Ribonucleic acid

ROS

Reactive oxygen species

sEVs

Small extracellular vesicles

siRNA

Small interfering RNA

TCA

Tricarboxylic acid cycle

TEM

Transmission electron microscopy

TFF

Tangential flow filtration

TF

Tissue factor

TME

Tumor microenvironment

TNM

Tumor, node, metastasis classification

TSG101

Tumor susceptibility gene 101

UPLC

Ultra–performance liquid chromatography

Author contributions

All authors contributed significantly to the work reported, whether in the conception, study design, execution, acquisition of data, analysis, and interpretation, or all these areas. That is revising or critically reviewing the article, giving final approval of the version to be published, agreeing on the journal to which the article has been submitted, and confirming accountabilities for all aspects of the work.

Funding

Not Applicable.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not Applicable.

Consent for publication

Not Applicable.

Competing interests

The authors declare no competing interests.