Nanomedicine novel strategies: deciphering the EV-metabolic axis as a natural nanocarrier network in lung cancer progression and cachexia

Abstract

The systemic progression of lung cancer involves a complex interplay between local tumor microenvironment (TME) dynamics and host-level metabolic decline, culminating in cachexia. Extracellular vesicles (EVs), have emerged as critical mediators in this process. This review constructs a comprehensive model of the “EV-metabolic axis” in lung cancer, framing EVs as natural nanocarriers within a systemic communication network that orchestrates a dual pathological process. Locally, EVs remodel the TME to support tumor growth, metastasis, and therapeutic resistance by transferringdiverse metabolic cargoes. Systemically, they transmit catabolic signals to distant adipose and muscle tissues, driving the severe tissue wasting characteristic of cachexia. This integrated perspective reveals the EV-metabolic axis as a central, targetable node in lung cancer pathology. From a nanomedicine perspective, targeting EV biogenesis, cargo loading, or uptake offers a novel, multifaceted therapeutic strategy to simultaneously inhibit tumor growth and mitigate cachexia, heralding a paradigm shift in future lung cancer treatment Scheme 1. This schematic illustrates the tripartite “EV-Metabolic Axis” framework linking local tumor metabolism, systemic EV trafficking, and cachexia development in lung cancer. In the Local Metabolic Axis, primary tumors and stromal cells (CAFs, TAMs, BMSCs) secrete extracellular vesicles (EVs) that reprogram glucose, lipid, and amino acid metabolism via cargoes such as miRNAs, metabolic enzymes, and cytokines — promoting glycolysis, glutamine addiction, ferroptosis resistance, and epithelial-mesenchymal transition (EMT). In the Circulatory System EV Transport Axis, EVs (40–150 nm exosomes, 50–1000 nm ectosomes) traverse biological barriers via membrane fusion, receptor-mediated endocytosis, or ligand-receptor binding, acting as natural nano-carriers. In the Systemic Cachexia Axis, circulating EVs deliver catabolic signals (e.g., miR-21, IL-6, HSP70/90, TGF-β, PTHrP) to distant organs — triggering adipose tissue browning, lipolysis, myofibrillar atrophy, and mitochondrial dysfunction — culminating in cancer-associated cachexia. This integrated axis positions EVs as both biomarkers and therapeutic targets across the nano-bio interface.

Graphical Abstract

Background

Lung cancer remains the leading cause of cancer-related mortality worldwide, with over 2 million new cases and approximately 1.8 million deaths annually [1]. Histologically, lung cancer is primarily classified into Non-Small Cell Lung Cancer (NSCLC) and Small Cell Lung Cancer (SCLC) [2]. NSCLC accounts for 80–85% of all lung cancer cases, while SCLC, a highly malignant and rapidly progressing neuroendocrine (NE) tumor, comprises about 15% [3]. The high mortality rate of lung cancer is associated with its insidious early clinical symptoms [2]; over 80% of patients are diagnosed at an advanced or locally advanced stage, precluding surgical intervention [4, 5]. The advent of molecular targeted therapy and immunotherapy has significantly improved the prognosis for patients with advanced lung cancer, but drug resistance and treatment-related adverse effects result in a still-unfavorable prognosis, with an overall 5-year survival rate of only about 18% [6–8].

In the course of lung cancer, cachexia is a common and fatal complication [9]. It is a complex metabolic syndrome characterized by progressive, involuntary weight loss, primarily due to the continuous wasting of skeletal muscle and adipose tissue, which is difficult to reverse with conventional nutritional support [10]. Approximately 50% of patients with advanced cancer experience cachexia [11], which is particularly prevalent among lung cancer patients and is the direct cause of death in about 20% of cancer patients [12, 13]. Cachexia not only severely impairs patients’ quality of life but also significantly reduces their tolerance to anti-cancer treatments [14, 15]. Therefore, elucidating its underlying mechanisms and identifying effective therapeutic targets is of paramount importance.

In recent years, the study of EVs has shifted from being overlooked “cellular debris” to a hotspot in life sciences [16]. EVs are naturally released, lipid bilayer-enclosed nanoscale biological entities that cannot self-replicate [24]. Initially regarded as “cellular debris” for waste disposal, EVs are now recognized as key “information messengers” in intercellular communication, encapsulating a rich cargo of bioactive molecules such as proteins, lipids, and nucleic acids [17]. With the development of isolation and identification techniques like ultracentrifugation, size-exclusion chromatography, and immunocapture, our understanding of EV biogenesis, cargo sorting, and their functions in physiological and pathological processes has deepened [18, 19]. In the complex pathological network of lung cancer, the metabolic remodeling of the TME and the cachectic wasting of distant organs may seem independent, but growing evidence suggests that EVs play a crucial “bridging” role [20, 21]. EVs serve as dual-function mediators: locally, they reprogram the TME metabolism; systemically, they disseminate catabolic cues to adipose and muscle tissues, precipitating cachexia [21–23]. This systemic communication, enabled by endogenous nanomedicine platforms, offers a new perspective for understanding and intervening in lung cancer.

Therefore, this review aims to systematically organize and integrate recent research on the role of EVs in the metabolic remodeling of the lung cancer microenvironment and in cachexia from a nanomedicine perspective. The core objective is to explore and substantiate their importance and feasibility as metabolic therapeutic targets in lung cancer. To ensure precision and focus, this review will specifically concentrate on EVs originating from primary lung cancer cells and their microenvironment, discussing them as a core component of the “EV-metabolic axis” that links local tumor progression with systemic metabolic disorders.

The discussion will cover several key stages of this process, including, but not limited to, the metabolic preparatory role in forming the pre-metastatic niche in distant organs and the metabolic adaptations of tumor cells after successful colonization of metastatic sites. Consequently, metabolic phenomena induced by metastases from other cancer types to the lung will not be discussed. Building on this foundation, this paper will first provide an overview of the biological basis of EVs, then delve into how they mediate the metabolic reprogramming of the lung cancer TME, and further elucidate their molecular mechanisms in inducing cachexia. Finally, the paper will summarize the potential of EVs as diagnostic biomarkers and therapeutic targets for lung cancer and offer a perspective on future research directions, aiming to provide a theoretical basis and new insights for understanding the systemic impact of lung cancer and developing novel therapeutic strategies.

A systematic search of PubMed and Web of Science identified literature on the EV-metabolic axis in lung cancer using key terms: “extracellular vesicles (including their subtypes)”, “lung cancer”, “metabolic reprogramming”, and “cachexia”. Literature from foundational to recent studies (up to 2026) was selected based on relevance to the review’s mechanistic focus. Included were English-language, peer-reviewed original research and authoritative reviews on EVs’ role in lung cancer metabolism or cachexia. Excluded were studies on non-lung cancer pulmonary metastases, publications without explicit metabolic mechanistic discussion, and conference abstracts. This structured approach ensured methodological rigor and transparency.

Overview of EVs

According to the latest guidelines from the International Society for Extracellular Vesicles (ISEV), EVs are particles naturally released from cells, defined by a lipid bilayer, and incapable of self-replication [24]. Their unique lipid membrane structure provides exceptional stability, protecting them from enzymatic degradation in bodily fluids and allowing for wide distribution in various fluids and tissues, including blood [25], urine [26], milk [27], bronchoalveolar lavage fluid [28], and adipose tissue [29], thereby enabling "liquid biopsies" [30]. Critically, EVs possess targeting capabilities, recognizing specific target tissues or cells via surface molecules (e.g., integrins), making them natural drug delivery vehicles [31]. Furthermore, EVs can cross biological barriers such as the blood-brain barrier [32], placental barrier [33], and air-blood barrier [34], delivering information to regions inaccessible to traditional signaling molecules, thus opening new avenues for the diagnosis and treatment of diseases like lung cancer.

Classification and biogenesis of EVs

The process of EV secretion, once considered a basic mechanism for cellular waste disposal, is now recognized as a vital mechanism for intercellular molecular exchange and signal transduction [35, 36]. Based on their subcellular origin, EVs can be classified into various subtypes that differ in biogenesis, size, markers, and function (Table 1). The main subtypes include exosomes and ectosomes.

Table 1.

Classification and biological characteristics of EV subtypes

EV subpopulations	Mechanism	Diameter	Biomarkers	Ref
Exosomes	MVBs with intraluminal vesicles that fuse with the plasma membrane for release (dependent or independent of ESCRT)	40–150 nm	CD9, CD63 etc.	[37, 38]
Ecotosomes (microvesicles and oncosomes)	Outward budding of the plasma membrane, scission/pinching off from membrane protrusions	50–10,000 nm	Annexin A1, ARF6	[39–41]
Apoptotic bodies	Cytoplasmic fragmentation of apoptotic cells	50–4000 nm	AnnexinV, thrombospondin, and C3b	[42–44]
Migrasomes	Retraction fibers with vesicles break.	1–3 μm	Tetraspanin-4 (TSPAN4)	[45]
Exomeres	Cleavage of large cytoplasmic extensions from cell body	< 50 nm	TGFBI, ENO1 and GPC1	[46]
Supermeres	Unknown	~35 nm	TGFBI, ACE2 and PCSK9	[47]
Blebbisomes	Plasma membrane vesicle retraction	10–20 μm	Syntenin-1, TSG101, Annexin A1 and A2	[48]

Exosomes are formed through the endocytic pathway and are approximately 40–150 nm in diameter. This process involves the inward budding of the endosomal membrane to form intraluminal vesicles (ILVs). After the multivesicular bodies (MVBs) containing these ILVs fuses with the plasma membrane, the ILVs are released as exosomes (Fig. 1a) [37]. Exosomes are enriched in tetraspanins (e.g., CD9, CD63, CD81), heat shock proteins (HSPs) (e.g., HSP70, HSP90), and endosomal markers like ALIX and TSG101 [38]. Another major subtype of EV is ectosomes, which form via direct outward budding and shedding from the plasma membrane and exhibit a diameter range of 50–10,000 nm. They exhibit a broad size range (from < 100 nm to several µm) and encompass various types, including oncosomes (Fig. 1b) [39–41]. Detailed information on other subtypes like apoptotic bodies and migrasomes can be found in Table 1 [42–48]. Although different types of EVs have distinct biological characteristics (Table 1), they all serve as carriers for intercellular communication [38]. However, current isolation techniques often limit the precise differentiation of EVs subtypes, and many studies do not investigate their specific biogenesis [49]. Therefore, to maintain scientific rigor and clarity, this review will uniformly use the term “EVs” unless the original literature explicitly specifies the vesicle’s biogenesis (e.g., “exosomes derived from multivesicular bodies”).

Fig. 1.

Biogenesis and classification of EVs. This schematic illustrates the primary formation mechanisms of EVs. (a) Exosomes are formed via the endocytic pathway, originating as ILVs within MVBs, and are released upon fusion of MVBs with the plasma membrane. (b) Ectosomes, including oncosomes, are formed by direct budding and shedding from the plasma membrane. (c) EVs interact with recipient cells through various mechanisms, including endocytosis, receptor-ligand binding, or direct membrane fusion, to deliver their molecular EV-cargo (e.g., proteins, miRNAs) and modulate cellular functions. Other EVs subtypes are detailed in Table 1

Role of EVs in intercellular communication and disease

EVs primarily interact with recipient cells in three ways: direct fusion to release their contents, surface ligand-receptor binding to activate signaling pathways, and endocytic uptake followed by regulation of gene expression, participating in physiological processes such as immune regulation and tissue repair (Fig. 1c) [50, 51]. In tumors, EVs mediate complex bidirectional communication: on one hand, tumor cells release EVs containing pro-tumorigenic factors (e.g., pro-angiogenic factors, miRNAs) to “educate” microenvironmental cells to support tumor growth [22, 52, 53]; on the other hand, microenvironmental cells also transmit signals via EVs to influence tumor metabolism and behavior [35]. EVs are deeply involved in all stages of tumor progression, including promoting proliferation, epithelial-mesenchymal transition, angiogenesis, immunosuppression (e.g., by transferring PD-L1 to inhibit anti-tumor immunity), and mediating therapeutic resistance [54, 55]. Additionally, EVs from specific sources also show therapeutic potential; for example, exosomes derived from dendritic cells (DCs) can act as tumor vaccines to activate anti-tumor immunity [56].

Methodological standardization and EV heterogeneity

The rigorous study of EVs necessitates strict adherence to standardized methodologies, as outlined by the Minimal Information for Studies of Extracellular Vesicles (MISEV2023) guidelines [24]. Current isolation techniques, such as ultracentrifugation, size-exclusion chromatography (SEC), and immunoaffinity capture, often yield highly heterogeneous populations [36]. Proper characterization requires a multi-parametric approach, including Nanoparticle Tracking Analysis (NTA) for size distribution and concentration, Transmission Electron Microscopy (TEM) for morphological validation, and Western blotting for specific positive markers (e.g., CD9, CD63, TSG101) and negative non-vesicular markers (e.g., Calnexin, Albumin) [57]. Furthermore, intrinsic EV heterogeneity—variations in size, density, and molecular cargo even from a single clonal cell source—profoundly impacts metabolic interpretations. For instance, small EVs (< 100 nm) may be differentially enriched with specific metabolic enzymes or regulatory RNAs compared to larger microvesicles (> 200 nm) [58, 59], leading to distinct functional outcomes in recipient cells. Failure to account for this subpopulation heterogeneity can confound the attribution of specific metabolic effects, blurring the lines between targeted EV-mediated reprogramming and general cellular stress responses [60].

The EV-mediated metabolic remodeling of the lung cancer microenvironment: from homeostatic imbalance to systemic malignant cycle

This chapter establishes EVs as the pivotal orchestrators in lung cancer progression [61]. It details how EVs function as critical nanoscale messengers within the TME. By transferring specific metabolic instructions between tumor and stromal cells (e.g., CAFs: Cancer-Associated Fibroblasts, TAMs: Tumor-Associated Macrophages), EVs precisely coordinate and amplify a state of metabolic reprogramming [62, 63]. This EV-driven metabolic rewiring is not merely supportive but is the primary driver of key malignant processes, including angiogenesis, immune evasion, resistance to ferroptosis, and the establishment of distant metastases. Ultimately, this creates a self-sustaining, pro-tumorigenic cycle that fuels comprehensive lung cancer advancement. To ensure conceptual rigor, the evidence discussed herein is explicitly classified into in vitro (cell culture), in vivo (animal models), and clinical data, highlighting the translational maturity of these findings.

Disruption of homeostasis and the formation of the lung cancer microenvironment

Normal lung tissue relies on its unique cellular composition and coordinated metabolic network to maintain homeostasis [64]. As shown in Fig. 2a (Homeostatic State), the alveolar microenvironment primarily consists of type I alveolar epithelial cells (AEC I), which relying on oxidative phosphorylation (OXPHOS) for gas exchange, type II alveolar epithelial cells (AEC II) (relying on glycolysis for pulmonary surfactant synthesis), as well as capillary endothelial cells, fibroblasts, and alveolar macrophages, among others [65–67]. These cells maintain functional equilibrium through precise metabolic division of labor and signaling communication, including the exchange of physiological EVs.

Fig. 2.

Comparative illustration of the normal alveolar microenvironment and the lung cancer microenvironment. In the homeostatic state (left), the normal alveolar architecture—including intact bronchial epithelium, basement membrane, surfactant layer, and distinct alveolar epithelial cells (AEC I and AEC II)—is shown alongside resident immune cells (e.g., alveolar macrophages) and stromal fibroblasts, which operate synergistically within an organized extracellular matrix (ECM) to maintain tissue homeostasis and physiological gas exchange. The cancer pathological state (right) depicts pathological progression, characterized by basement membrane disruption, infiltration of diverse immune and stromal cells (e.g., M2-polarized macrophages, CAFs, TANs), and metabolic reprogramming that supports tumor adaptation. A key illustrated process is the intravasation of tumor cells, including CSCs, into the vasculature, emphasizing its dependence on metabolic adaptations essential for metastasis and distant growth. TANs, Tumor-Associated Neutrophils; CSC, Cancer Stem Cells

When subjected to persistent assaults by carcinogenic factors such as genetic mutations, this homeostasis is disrupted [68]. The activation of key proto-oncogenes (e.g., KRAS, EGFR) or the inactivation of tumor suppressor genes (e.g., p53) not only drives the malignant transformation of cells but also releases signals that remodel the microenvironment [69–71]. As depicted in Fig. 2b (Cancer Pathological State), the original orderly structure is replaced by the lung cancer TME, composed of a massive proliferation of tumor cells, activated CAFs, and polarized TAMs, among others [65]. A core feature of this transformation is the fundamental reprogramming of cellular metabolic patterns, which provides the material and energy foundation for rapid tumor growth and adaptation [61].

Systemic coordinators and effector amplifiers of metabolic reprogramming

Within the lung cancer TME, metabolic reprogramming is not an isolated event but a systemic process involving the participation of both tumor and stromal cells, facilitated by efficient “dialogue” and “division of labor” mediated by EVs [61]. Acting as natural nanocarriers, EVs encapsulate and deliver a rich cargo of proteins, metabolic enzymes, lipids, and RNAs, making them key mediators for transmitting metabolic instructions and reshaping the metabolic phenotypes of recipient cells [72].

As integrated in the complex network shown in Fig. 3, metabolic reprogramming in the lung cancer TME encompasses glucose, lipid, and amino acid metabolism, along with key signaling pathways such as PI3K-AKT-mTOR [73]. EVs derived from tumor cells often carry pro-metabolic cargo to directly “educate” recipient cells. For instance, in vitro studies have demonstrated that irradiated lung cancer cells can release EV-(ALDOA, ALDH3A1: aldehyde dehydrogenase 3A1), directly transferring glycolytic enzymes to recipient cells to accelerate their glycolysis [74]. Both in vitro and in vivo animal models have shown that EV-(METTL3: methyltransferase like 3) or EVs-(LINC01614) released by CAFs can provide glutamine metabolic support for cancer cells, meeting their biosynthetic demands [53, 75]; EVs secreted by M2-type TAMs have been shown in clinical samples and xenograft models toenhance glycolysis in cancer cells through axes such as LINC01001/METTL3 [76]. This EV-mediated, bidirectional, or even multidirectional exchange of metabolic signals leads to the formation of complex metabolic symbiotic networks within the TME, such as the “reverse Warburg effect”, where stromal cells perform glycolysis to produce lactate, which is then taken up by tumor cells for oxidative phosphorylation to efficiently generate adenosine triphosphate (ATP). This metabolic collaboration is a crucial foundation for the maintenance and expansion of the lung cancer TME.

Fig. 3.

EV-mediated metabolic regulatory network in the lung cancer microenvironment. This schematic illustrates the multidimensional regulatory roles of EVs derived from tumor cells, immune cells, and stromal cells in driving the metabolic reprogramming that defines the TME. These EVs function as pivotal intercellular messengers, transferring a heterogeneous cargo of bioactive molecules—including miRNAs, lncRNAs, circRNAs, functional proteins, and metabolites—to recipient cells within the TME. By delivering this molecular cargo, EVs intricately modulate core metabolic pathways essential for tumor survival and progression, such as glycolysis (regulating glucose uptake via GLUT1/GLUT4, HK activity, and lactate secretion through LDHA), lipid metabolism (spanning fatty acid synthesis, β-oxidation, and cholesterol/lipogenesis pathways orchestrated by SREBP and GLS1), and amino acid metabolism (with glutamine utilization via SLC7A5/SLC1A5 and c-Myc-dependent regulation as central nodes). Beyond these core pathways, the figure also delineates EV-mediated regulation of complementary cellular processes: pH homeostasis (via CA and the sodium-hydrogen exchanger NHE1), redox balance (through modulation of the NADPH/NADH ratio by NAMPT and nicotinamide), and autophagy (governed by mTOR, ULK1, and the LKB1-AMPK signaling axis). Visually, the schematic utilizes distinct elements-color-coded molecular categories (e.g., RNA, protein, metabolite), directional arrows denoting regulatory directionality, and annotations for canonical signaling cascades (e.g., Ras-PI3K-AKT, LKB1-AMPK, and c-Myc-driven transcriptional programs)-to clarify the intricate crosstalk between EVs cargo, recipient cell metabolic rewiring, and TME remodeling. Together, this representation highlights how EV-mediated molecular transfer reshapes metabolic networks to sustain tumor growth, evade immune detection, and adapt to the nutrient-deprived, hypoxic conditions of the TME, thereby underscoring the therapeutic potential of targeting EV-driven metabolic crosstalk. GLUT1/4, Glucose Transporter 1/4; HK, Hexokinase; LDHA, Lactate Dehydrogenase A; SREBP, Sterol Regulatory Element-Binding Protein; GLS1, Glutaminase 1; SLC7A5/1A5, Solute Carrier Family 7 Member 5/1 Member 5; c-Myc, Cellular Myelocytomatosis Oncogene; CA, Carbonic Anhydrase; NHE1, Sodium-Hydrogen Exchanger 1; NADPH/NADH, Nicotinamide Adenine Dinucleotide Phosphate/Nicotinamide Adenine Dinucleotide; NAMPT, Nicotinamide Phosphoribosyltransferase; ULK1, Unc-51 Like Autophagy Activating Kinase 1; LKB1-AMPK, Liver Kinase B1-AMP-activated Protein Kinase

EV-mediated metabolic reprogramming: a core driver of lung cancer progression

EVs serve as pivotal mediators that orchestrate a comprehensive metabolic rewiring within the lung cancer microenvironment, directly fueling tumor growth, adaptation, and dissemination [61]. This reprogramming spans glucose, lipid, and amino acid metabolism, and extends to the regulation of cell death pathways such as ferroptosis. The complex interplay of these pathways is summarized in Fig. 3, which illustrates how EV-transported cargoes modulate key transporters, enzymes, and signaling cascades (e.g., PI3K-AKT-mTOR) to alter the metabolic landscape of recipient cells. The following sections detail how EVs induce these metabolic alterations to promote lung cancer initiation, proliferation, metastasis, and cancer-related cachexia.

EVs in glucose metabolic reprogramming: fueling the warburg effect

To meet the heightened energy and biosynthetic demands of rapid proliferation, lung cancer cells and associated stromal cells utilize EVs to enforce a glycolytic shift, even under normoxic conditions—a hallmark known as the Warburg effect [77]. Tumor-derived EVs can directly transfer glycolytic enzymes. For instance, FRK-knockout lung cancer cells secrete exosomal miR-9-3p, which specifically targets NQO2 to enhance mitochondrial OXPHOS in both in vitro and in vivo models [78]; In vitro experiments demonstrate that irradiated Lewis lung carcinoma cells (LCCs) release EVs that carry the proteins ALDOA and ALDH3A1, which are taken up by recipient cells to directly enhance their glycolytic flux, promoting survival after radiation stress [74]. Furthermore, CAF-derived Evs contribute to this process by delivering specific miRNAs [63]. M2 macrophage-derived EVs containing LINC01001 have been shown in both in vitro assays and in vivo nude mouse models to promote glycolytic metabolism in lung adenocarcinoma cells through the METTL3 pathway, enhancing tumor growth [76]. This EV-mediated enhancement of glycolysis not only provides ATP but also generates metabolic intermediates for anabolic processes, supporting uncontrolled proliferation.

EVs in lipid metabolic reprogramming: building blocks and signaling molecules

Lipid metabolism is crucial for membrane synthesis, energy storage, and signal transduction [79]. EVs play a significant role in modulating lipid metabolism to support tumorigenesis [22]. For example, in lipid uptake and synthesis, tumor-derived EVs can carry fatty acid transport proteins or regulatory RNAs that upregulate receptors like CD36 (as implied in Fig. 3), enhancing fatty acid uptake. In more advanced or resistant states, in vitro and in vivo evidence indicates that EVs from low-metastatic lung cancer cells carrying lncRNA ROLLCSC have been found to promote lipid metabolism and metastatic capacity in high-metastatic counterparts, acting as competitive endogenous RNAs [80]. This reprogramming provides essential lipids for membrane biogenesis in rapidly dividing cells and for energy production through β-oxidation.

EVs in amino acid metabolic reprogramming: supporting anabolism and redox balance

Amino acids, particularly glutamine, serve as nitrogen sources for nucleotide and amino acid synthesis and are involved in maintaining redox homeostasis [81, 82]. CAF-derived EVs are instrumental in providing metabolic support by supplying key enzymes. For instance, preclinical in vivo models have shown that EVs from CAFs carrying METTL3 promote glutamine metabolism in lung cancer cells, supporting their growth under nutrient stress [53]. Similarly, CAF-derived EVs containing LINC01614 enhance glutaminolysis in cancer cells, fueling their anabolic needs [75]. This metabolic crosstalk ensures a steady supply of building blocks for proteins and nucleic acids, facilitating tumor expansion.

EVs in regulating ferroptosis: a metabolic vulnerability and defense

Ferroptosis, an iron-dependent form of regulated cell death driven by lipid peroxidation, represents a metabolic vulnerability that tumors must evade [83]. The core regulatory axis of ferroptosis, involving lipid peroxidation promotion through pathways such as Acyl-CoA Synthetase Long Chain Family Member 4 (ACSL4) and Lysophosphatidylcholine Acyltransferase 3 (LPCAT3), and antioxidant defense mediated by factors including Solute Carrier Family 7 Member 11 (SLC7A11) and Glutathione Peroxidase 4 (GPX4), is depicted in Fig. 4, titled “Regulatory mechanisms of ferroptosis”. EVs provide a mechanism for cells within the TME to confer resistance to this process. Notably, in vitro and xenograft studies demonstrate that CAF-derived EVs carrying the lncRNA ROR1-AS1 can stabilize SLC7A11 mRNA in recipient lung cancer cells. This enhances cystine uptake and glutathione (GSH) synthesis, thereby strengthening the cellular antioxidant defense system and increasing resistance to ferroptosis, which may contribute to therapy evasion and survival [84]. This highlights how EVs can manipulate a metabolism-linked cell death pathway to promote tumor resilience.

Fig. 4.

Regulatory mechanisms of ferroptosis. This schematic illustrates the core molecular pathways of ferroptosis, a form of regulated cell death. Key regulatory modules include: Redox homeostasis and cystine metabolism: cells uptake cystine via transporters (e.g., SLC7A11), which is reduced to cysteine for GSH synthesis. GSH acts as a cofactor for GPX4 to inhibit lipid peroxidation. Lipid metabolism and iron-driven oxidative stress: PUFAs are incorporated into cell membranes via ACSL4 and LPCAT3. Iron mediates Fenton reactions with these PUFAs, inducing ROS production and lipid peroxidation. Antioxidant defense network: In addition to GPX4, complementary pathways (e.g., FSP1-mediated CoQH₂ generation, DHODH-dependent CoQH₂ synthesis) mitigate ROS accumulation. Exogenous FAs are taken up via CD36 and FATPs, influencing membrane lipid profiles. Mitochondrial lipid peroxidation further exacerbates oxidative damage. Overall, ferroptosis is orchestrated by the crosstalk of iron metabolism, lipid metabolism, and redox homeostasis, where a disruption in the balance between pro-oxidant and anti-oxidant signals leads to lethal membrane lipid peroxidation. PUFAs, polyunsaturated fatty acids; FSP1, ferroptosis suppressor protein 1; CoQH2, ubiquinol; DHODH, dihydroorotate dehydrogenase; FAs, fatty acids; CD36, cluster of differentiation 36; FATPs, fatty acid transport proteins

EVs in metabolic reprogramming of the PMN

Beyond the primary site, EVs are crucial for preparing distant organs for metastasis by reprogramming the metabolism of resident stromal cells, forming the PMN. For instance, in vivo tracking studies show that EVs from chemotherapy-persistent lung cancer cells carrying IGF-2/IGFBP2 can activate the IGF-1R pathway in bone marrow mesenchymal stem cells (BMSCs) [85]. This induces a glycolytic shift in BMSCs, creating a nutrient-rich, supportive microenvironment for future metastatic seeding. In the lung as a metastatic site, tumor-derived EVs can polarize macrophages via the TLR2/NF-κB axis, skewing their metabolism toward glycolysis and lactate production. This metabolic reprogramming fosters an immunosuppressive PMN conducive to tumor cell invasion [86]. These examples, among others detailed in Table 2, demonstrate how EV-mediated metabolic instruction of distant stromal cells is a fundamental step in establishing metastasis.

Table 2.

Summary of EV-mediated metabolic regulation in lung cancer

Metabolic pathway	Producing cells	Recipient cells	Isolation method	EV type/ key cargo	Mechanism	Experiment type	Phenotype	Ref
Glucose metabolism	FRK-knockout LLCs	LCCs	ExoQuick kit	Exosomes/ miR-9-3p	Targets NQO2 to regulate mitochondrial function	in vitro and in vivo	Enhance OXPHOS	[78]
	Irradiated LLCs	LLCs	Differential ultracentrifugation	Exosomes/ ALDOA, ALDH3A1	Transfers glycolytic enzymes, promotes lactate production	in vitro	Enhance glycolysis	[74]
	High-metastatic LLCs	Normal fibroblasts	ExoQuick Kit	Exosomes/ miR-1290	Targets MT1G to activate AKT pathway, induces CAF differentiation	in vitro and in vivo	Induces CAF differentiation, confers oxidative stress resistance	[63]
	M2 macrophages	LCCs	ExoQuick Kit	Exosomes/LINC01001	Interacts with METTL3, regulates NASP methylation to enhance glycolysis	in vitro and in vivo	Enhances glycolysis	[76]
	M2 macrophages	LCCs	VEX Exosome Isolation Reagent	Exosomes/ miR-3679-5p	Targets NEDD4L to stabilize c-Myc, enhances aerobic glycolysis	in vitro and in vivo	Enhances glycolysis, induces cisplatin resistance	[35]
	H292 cells (treated with SAHA)	Immune system cells	Differential ultracentrifugation	Exosomes/HSP60	SAHA induces ROS overproduction and mitochondrial dysfunction, leading to HSP60 nitration	in vitro	Induces cell death and immune interaction	[87]
	Gefitinib-sensitive LCCs	Gefitinib-resistant LCCs	Exosome Isolation Kit	Exosomes/ circKIF20B	Sponges miR-615-3p to regulate MEF2A	in vitro and in vivo and clinical study	Reverses gefitinib resistance, inhibits OXPHOS	[88]
	High-metastatic LCCs	Low-metastatic LCCs, CAFs	Differential ultracentrifugation	Exosomes/ Pathogenic mtDNA	Horizontal transfer of mtDNA via small EVs	in vitro and in vivo	Spreads pathogenic mtDNA, enhances tumor progression	[89]
	Chemoresistant Kras (MT) LLCs	LCCs-DR	Differential ultracentrifugation	Exosomes/ BACH2, GATA-3	Regulates miR-146/miR-210 via RIP-3 dependent necroptosis, shifts PKM2-mediated metabolism	in vitro and in vivo and clinical study	Maintain tumor cell metabolic chemoresistance	[90]
	LCCs	autocrine/ paracrine	Differential centrifugationExoQuick precipitation kit	Exosomes/PKM2	Phosphorylates SNAP-23 at Ser95 to promote SNARE complex formation	in vitro	Links aerobic glycolysis with exosome secretion	[91]
	Hypoxic LCCs	M2 macrophages	Differential ultracentrifugation	Exosomes/PKM2	Induces M2 polarization via AMPK/p38 signaling pathway	in vitro and in vivo	Induces M2 macrophage polarization	[92]
	Hypoxic LCCs	M2 macrophages	Differential ultracentrifugation	Exosomes /miR-let-7a	Inhibits insulin-Akt-mTOR pathway, enhances OXPHOS	in vitro and in vivo	Induces M2 macrophage polarization	[73]
	LCCs	M0 macrophages	Differential ultracentrifugation	Exosomes /unspecified	p53-independent M2 polarization, modulates mitochondrial metabolism	in vitro and in vivo	Induces M2 polarization, increases oxygen consumption rate	[93]
Lipid metabolism	Tumor cells	T cells	Differential ultracentrifugation	Exosomes /PD-L1	Activates CREB/STAT pathways, reprograms lipid metabolism	in vitro and in vivo	Induces T-cell senescence, activates lipid metabolism	[22]
	LLCs (A549, H1299, LLC)	Macrophages	Total Exosome Isolation Reagent	Exosomes/ TRIM59	Ubiquitinates/degrades ABHD5, induces lipid metabolism reprogramming	in vitro and in vivo	Induces metabolic reprogramming, NLRP3 activation	[94]
	Lung cancer stem cells (LLC-SD)	Low-metastatic LCCs	Differential ultracentrifugation	Exosomes/ lncRNA ROLLCSC	Acts as ceRNA, targeting miR-5623-3p and miR-217-5p, upregulates lipid metabolism	in vitro and in vivo	Stimulates lipid metabolism, enhances metastasis	[80]
	LCCs (with PtNPs)	Autocrine/ paracrine	Differential centrifugationExoQuick kit	Exosomes/ unspecified	Induces oxidative stress, activates ceramide pathway	in vitro	Increases sEVs release	[95]
Amino acid metabolism	CAFs	NSCLC cells	ExoQuick kit	Exosomes/ METTL3	Induces m⁶A modification of SLC7A5 mRNA to stabilize its expression	in vitro and in vivo and clinical study	Promotes proliferation, invasion, and glutamine metabolism	[53]
	CAFs	LUAD cells	ExoQuick kit and differential ultracentrifugation	Exosomes/ LINC01614	Interacts with ANXA2/p65 to activate NF-κB, upregulates SLC38A2/SLC7A5	in vitro and in vivo and clinical study	Enhances glutamine uptake	[75]
	TKI-resistant LCCs	TKI-sensitive LCCs, NFs, epithelial cells	Exosome isolation kit	EVs/ SLC1A5, SLC25A5	Enhances glutamine uptake via SLC1A5, activates STAT3 pathway	in vitro and in vivo and clinical study	Induces drug resistance, transforms NFs to CAFs, enhances invasion	[96]
Regulatory mechanisms of ferroptosis	LUAD cells	LUAD cells	Differential ultracentrifugation	Exosomes/ circRNA_101093 (cir93)	Interacts with FABP3 to reduce AA levels and lipid peroxidation via NAT-TLX2 axis	in vitro and in vivo and clinical study	Reduces lipid peroxidation, desensitizes to ferroptosis	[97]
	Cancer-associated fibroblasts	Lung cancer cells	Differential ultracentrifugation	Exosomes/ ROR1-AS1	Interacts with IGF2BP1 to stabilize SLC7A11 mRNA, increases GSH levels	in vitro and in vivo and clinical study	Inhibits ferroptosis, promotes cell survival	[84]
SCLC	Non-NE SCLC cells	NE SCLC cells	Differential ultracentrifugation	sEVs/ ECM proteins, Integrins	Promotes adhesion and survival of NE SCLC cells	in vitro and proteomics and clinical study	Promotes NE cell adhesion and survival	[98]
PMN and metastasis	Cisplatin-induced dormant LCCs	BMSCs	ExoQuick kit	Exosomes/ IGF-2, IGFBP2	Activates IGF-1R signaling to enhance glycolysis	in vitro and in vivo	Enhances glycolysis in BMSCs, promotes cancer cell growth	[85]
	LCCs	Non-alveolar inflammatory macrophages	/	Microparticles Oncosomes/unspecified	Activates mTORC1 signaling to enhance OXPHOS	in vitro and in vivo	Enhances mitochondrial function and OXPHOS	[99]
	LLCs	Macrophages	Differential ultracentrifugation and SEC	Exosomes/HMGB-1	Activates TLR2/NF-κB pathway, induces glycolytic reprogramming	in vitro and in vivo and clinical study	Induces immunosuppressive phenotype, facilitates premetastatic niche formation	[86]
	HBMECs	SCLC cells	Differential ultracentrifugation	Exosomes/ S100A16	Upregulates PHB-1 to maintain mitochondrial membrane potential	in vitro and in vivo and clinical study	Promotes survival, inhibits apoptosis under oxidative stress	[100]

In Table 2, LCCs represent cells derived from non-small cell lung cancer, while LUAD stands as the abbreviation for lung adenocarcinoma. Current isolation techniques, such as ultracentrifugation, co-precipitate multiple vesicle types alongside non-vesicular contaminants. Although size-exclusion chromatography provides higher purity, it risks losing specific EV subpopulations. This technical limitation significantly hinders the precise attribution of distinct biological functions to individual EV subtypes

In summary, EVs are master regulators of a pan-metabolic reprogramming in lung cancer. By shuttling specific proteins, enzymes, and regulatory RNAs between tumor and stromal cells, they modulate glucose, lipid, and amino acid metabolism, influence cell death decisions like ferroptosis, prepare distant sites for metastasis, and ultimately contribute to systemic metabolic decline in cachexia. This integrated view, supported by Figs. 3 and 4 and detailed examples in Table 2, positions EV-mediated metabolic communication as a central axis driving lung cancer progression from local growth to systemic disease.

The role of EVs in lung cancer cachexia

Definition, staging, and irreversibility of cachexia

Cancer cachexia is a multifactorial syndrome driven by systemic inflammation, metabolic disturbances, and anorexia. Its pro-inflammatory cytokines and catabolic factors often originate from reprogrammed cells within the TME, such as M2 macrophages and CAFs (Fig. 3) [53, 62]. According to international consensus, it can be divided into three stages: pre-cachexia, cachexia, and refractory cachexia [101]. The irreversibility of cachexia lies in its complex molecular basis: once systemic inflammation and catabolic pathways are activated by tumor signals including EVs, a vicious cycle is established that is difficult to break. Conventional nutritional support cannot effectively reverse the wasting of muscle and fat [102].

EV-mediated metabolic reprogramming of adipose tissue

Physiological framework of adipocyte thermogenesis and browning

The generation of beige adipocytes (browning) and their thermogenic function are precisely regulated by a highly coordinated cascade of extracellular signals and intracellular events, as depicted in Fig. 5. This process begins with the binding of various extracellular signaling molecules to specific receptors on the adipocyte membrane, leading to the activation of key intracellular signaling pathways, including cyclic adenosine monophosphate (cAMP)/protein kinase A (PKA) and p38 mitogen-activated protein kinase (p38 MAPK). These signals ultimately converge in the nucleus to activate core transcription factors such as Activating Transcription Factor 2 (ATF-2), Peroxisome Proliferator-Activated Receptor Gamma Coactivator 1-Alpha (PGC1α), and Peroxisome Proliferator-Activated Receptor Gamma (PPARγ), which in turn upregulate the expression of a series of thermogenic genes, including uncoupling protein 1 (UCP1)(103]. UCP1, located on the inner mitochondrial membrane, activates uncoupling oxidative phosphorylation, releasing energy as heat.

Fig. 5.

Regulatory mechanisms of adipocyte thermogenesis and beige adipocyte biogenesis. This schematic depicts the multi-layered regulatory network governing adaptive thermogenesis, integrating extracellular stimuli and intracellular signaling cascades to drive lipid droplet dynamics and mitochondrial heat production. Extracellular signals (e.g., ACTH, NE, secretin, IGF-1) bind to cell surface receptors, activating intracellular pathways (cAMP/PKA, p38 MAPK, ERK1/2, PKA, PKC, etc.) that converge on LD dynamics. Key lipolytic enzymes—including HSL, ATGL, and MGL—are regulated by cofactors (e.g., CGI-58, PLIN1) to hydrolyze stored triglycerides, releasing FFAs. These FFAs fuel mitochondrial thermogenesis via UCP1-mediated proton leak, dissipating energy as heat. Concurrently, transcriptional regulation in the nucleus (driven by PGC-1α, PPARγ, PRDM16, and other co-activators/receptors) coordinates the expression of key genes for mitochondrial biogenesis, brown/beige adipocyte identity, and thermogenic capacity (e.g., UCP1, CIDEA, PRDM16). The diagram further highlights the structural transition of white to beige adipocytes, underscoring the coordinated control of lipid metabolism and energy dissipation at both cellular and molecular levels. ACTH, Adrenocorticotropic Hormone; NE, Norepinephrine; IGF-1, Insulin-like Growth Factor 1; PKC, Protein Kinase C; ATGL, Adipose Triglyceride Lipase; MGL, Monoacylglycerol Lipase; CGI-58, Comparative Gene Identification-58; PLIN1, Perilipin 1; FFAs, Free Fatty Acids; CIDEA, Cell Death Inducing DFFA Like Effector A; PRDM16, PR/SET Domain 16

EVs suppress adipogenesis: blocking the source of energy storage

Lung cancer-derived EVs inhibit the formation of new adipocytes by delivering specific molecular cargo. In vitro studies have demonstrated that EV TGF-β suppresses the activity of the master adipogenic transcription factor PPARγ by activating the SMAD2/4 pathway [104]. Concurrently, both in vitro and in vivo experiments show that EV miR-425-3p silences the expression of key adipogenic factors such as GATA2 and C/EBPα, co-inhibiting the proliferation and differentiation of preadipocytes [105]. These actions directly counteract the normal storage function of adipose tissue.

EVs activate lipolysis and drive thermogenesis: mobilizing stored energy and inducing ineffective dissipation

While blocking energy storage, lung cancer-derived EVs potently activate the breakdown of stored fat and induce its transformation into a thermogenic phenotype, i.e., pathological “browning”. This process precisely “hijacks” the physiological thermogenic pathway depicted in Fig. 5. First, in vivo murine models demonstrate that EV PTHrP acts as the core initiating signal, mimicking the action of hormones like ACTH in Fig. 5. It activates the cAMP/PKA pathway within adipocytes, which in turn phosphorylates and activates HSL, directly driving triglyceride hydrolysis to mobilize stored energy [106]. Activated PKA is not only a key kinase for lipolysis but also an upstream signal for inducing the expression of thermogenic genes like PGC1α and UCP1, thereby coordinately initiating the lipolysis and thermogenesis programs [105]. To amplify this catabolic signal, EV-delivered EIF5A further enhances cAMP signaling via the GPBAR1/cAMP/PKA/CREB pathway [107], while EV-carried IL-6 synergistically promotes catabolism through the STAT3 pathway [21]. These signals interact with the network in Fig. 5, collectively reinforcing the catabolic state. In the terminal execution step of energy dissipation, in vitro evidence indicates that EV GRP75 directly targets the final executor of thermogenesis—the mitochondria. It promotes the formation of a functional complex between adenine nucleotide translocase 2 (ANT2) and UCP1 on the inner mitochondrial membrane, directly corresponding to the core module of mitochondrial thermogenesis on the right side of Fig. 5. This greatly enhances proton leak and uncoupling, leading to futile energy dissipation as heat [108]. Furthermore, to maintain this high-energy-consuming state, EV-anchored HSP70/90 sustains adipocytes in a chronic, basal state of lipolysis by regulating molecules such as vacuolar H⁺-ATPase [109].

It remains crucial to contextualize EV-mediated effects within the classical pathways driving skeletal muscle atrophy. Traditionally, cachexia is propelled by soluble inflammatory mediators (e.g., TNF-α, free IL-6) that activate the NF-κB and STAT3 pathways, subsequently upregulating the ubiquitin-proteasome system (UPS) via E3 ligases like MuRF1 and MAFbx, while simultaneously triggering the autophagy-lysosome pathway [110, 111]. Tumor-derived EVs do not supplant these classical pathways; instead, they function as potent upstream amplifiers. For instance, EV-encapsulated IL-6 or HSP70/90 gain protection from systemic degradation, shielding them from degradation and enabling selective internalization by myocytes. This leads to a far more sustained and concentrated activation of the TLR4/NF-κB and STAT3 axes compared to their soluble counterparts [21, 23]. Thus, the distinct contribution of EVs lies in their remarkable capacity to precisely deliver concentrated catabolic payloads directly into the muscle cell’s intracellular space, thereby synergizing powerfully with classical systemic inflammation to dramatically accelerate UPS-mediated protein degradation.

Mechanism integration and pathological effects

Lung cancer-derived EVs systematically deliver specific molecular cargo, including TGF-β, miR-425-3p, PTHrP, HSP70/90, and GRP75, precisely interfering with the inherent metabolic regulatory network of host adipocytes (Fig. 5) [104]. While clinical data show an association between these EV cargoes and fat loss in patients, in vivo studies provide evidence for causation. These actions collectively constitute a “bidirectional dysregulation” metabolic attack: on one hand, blocking adipogenesis by inhibiting PPARγ and adipogenic transcription factors [104]; on the other hand, promoting lipolysis and energy dissipation by activating the PKA/HSL lipolytic pathway and the ANT2/UCP1 thermogenic module [105, 106]. This strategy of simultaneously suppressing energy storage while intensifying breakdown and futile thermogenesis is the core molecular mechanism leading to the rapid and irreversible loss of adipose tissue in cancer cachexia.

EV-mediated mechanisms of skeletal muscle atrophy in lung cancer cachexia

Physiological and pathological molecular regulatory network of skeletal muscle atrophy

Skeletal muscle, constituting approximately 30–40% of total body mass, serves as the primary reservoir for amino acids and a major site for glucose uptake and storage, playing a crucial role in maintaining muscle function and systemic energy-protein homeostasis [112–114]. In cachexia, skeletal muscle atrophy is characterized by a severe imbalance between suppressed protein synthesis and increased degradation [115].

Consequently, the maintenance of skeletal muscle mass depends on a precise balance between anabolism and catabolism [110, 111]. As shown in Fig. 6, the molecular core of this process involves a complex interplay between EV-mediated signals and classical skeletal muscle atrophy pathways [116]. While classical pathways such as the ubiquitin-proteasome system (UPS), autophagy-lysosome pathway, and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) signaling are the ultimate executioners of muscle degradation, tumor-derived EVs act as critical upstream modulators. They do not replace these classical pathways but rather hijack and hyperactivate them. On one hand, catabolic signals delivered by EVs (e.g., IL-6, HSP70/90) directly feed into the signal transducer and activator of transcription 3 (STAT3), NF-κB, and mothers against decapentaplegic homolog 2/3 (SMAD2/3) axes. These transcription factors subsequently upregulate the expression of muscle-specific E3 ubiquitin ligases, namely Muscle RING Finger Protein 1 (MuRF1) and muscle atrophy F-box (MAFbx/Atrogin-1), thereby driving UPS-mediated protein degradation [21, 109]. On the other hand, anabolic signals (e.g., IGF-1: insulin-like growth factor 1) that promote protein synthesis via the PI3K/AKT/mTOR pathway are often suppressed in the cachectic state. The relative contribution of EVs lies in their ability to act as systemic, concentrated delivery vehicles that persistently trigger these classical local catabolic networks, tipping the balance irreversibly towards atrophy [117].

Fig. 6.

Molecular regulatory mechanisms of skeletal muscle atrophy. This schematic illustrates the dynamic balance between catabolic and anabolic signaling pathways that control muscle mass homeostasis. Catabolic stimuli (e.g., IL-6, myostatin, activin A) activate pathways converging on transcription factors (NF-κB, SMAD2/3, FOXO), which upregulate E3 ubiquitin ligases (MuRF1, MAFbx/atrogin-1) to drive proteasomal and autophagic-lysosomal protein degradation. Conversely, anabolic signals IGF-1 activate AKT/mTOR signaling, which promotes protein synthesis while inhibiting FOXO-mediated catabolism. The interplay between these opposing pathways determines net muscle atrophy or hypertrophy, with catabolic signal dominance leading to myofibrillar protein loss and functional decline. Tumor-derived EVs act as upstream modulators, delivering cargo that hyperactivates these classical catabolic pathways. IL-6, Interleukin-6; FOXO, Forkhead Box O

EVs induce muscle cell apoptosis

EVs can directly cause myocyte death. In vitro and in vivo studies demonstrate that EV miR-21, once internalized by muscle fibers, binds to Toll-like receptor 7/8 (TLR7/8). This binding triggers activation of the JNK-dependent apoptotic pathway, culminating in myonuclear depletion [118]. This demonstrates a direct causal link between specific EV cargo and muscle cell apoptosis in experimental models. This pathway corresponds to the activation of the “Apoptosis” module in Fig. 6. EV-mediated pro-apoptotic signals and their specific molecular mechanisms are summarized in Table 3.

Table 3.

Lung cancer cells promote cachexia through secreting EVs

Characteristic	Recipient cell	Isolation method	EV type/ carried molecular	Mechanisms demonstrated	Evidence type	Phenotype	Ref
Reduced adipogenesis	hAD-MSCs	Differential ultracentrifugation	Exosomes/ TGF-β	SMAD2/SMAD4/PPARγ pathway	in vitro	Anti-adipogenic	[104]
Reduced adipogenesis/lipidolysis	Preadipocytes/Mature adipocytes	Differential ultracentrifugation	Exosomes/ miR-425-3p	Targets GATA2/ IGFBP4/ MMP15/ CEBPA; Downregulates PDE4B, activating cAMP/PKA/lipophagy	in vitro	Inhibits preadipocytes proliferation/differentiation; Promotes adipocyte lipolysis and white adipose tissue browning	[105]
Lipidolysis	3T3-L1 adipocytes, inguinal/epididymal WAT	Differential ultracentrifugation	EVs/ PTHR	PTHrP/PKA/HSL pathway	in vitro and in vivo	Adipocyte lipolysis, white adipose tissue browning, fat mass loss	[106]
Lipidolysis	3T3-L1 adipocytes, primary adipocytes, inguinal/epididymal WAT	Differential ultracentrifugation	EVs/ GRP-75	GRP75-ANT2-UCP1	in vitro and in vivo	WAT Browning	[108]
Muscle atrophy /lipidolysis	C2C12 myotubes, skeletal muscle, subcutaneous/epididymal adipose tissue	ExoQuick precipitation kit; AchE activity assay	EVs/ HSP70 HSP90	HSP70/HSP90-TLR4-p38 MAPK axis activates UPP/ALP; omeprazole inhibits EV release via V-H⁺-ATPase/Rab27b	in vitro and in vivo	Loss of epididymal adipose tissue	[109]
Lipidolysis	3T3-L1 adipocytes, HIB1B adipocytes, iWAT/eWAT/BAT	Differential ultracentrifugation	EVs/ EIF5A	Binds GPBAR1 mRNA to promote translation, activates CREB pathway	in vitro and in vivo	Adipocyte lipolysis, WAT browning, and fat mass loss are reversed by EIF5A silencing/GC7	[107]
Muscle atrophy/lipidolysis	C2C12 myotubes, 3T3-L1 adipocytes, skeletal muscle, iWAT/eWAT/BAT	Differential ultracentrifugation	EVs/ IL-6	IL-6 binds IL-6R/gp130, activates STAT3; upregulates Atrogin1 (muscle) and phosphorylates HSL (adipocytes)	in vitro and in vivo and clinical study	Induces myotube atrophy, adipocyte lipolysis, WAT browning; IL-6 high expression correlates with poor prognosis; reversed by anti-IL-6/S3I-201	[21]
Lipidolysis	3T3-L1 adipocytes, inguinal/epididymal WAT	Qiagen EV Isolation Kit	EVs/ IL-8	IL-8 binds CXCR2, activates NF-κB (p65 phosphorylation), upregulates PGC1α/UCP1	in vitro and in vivo	Adipocyte lipolysis, WAT browning, fat mass loss; reversed by anti-IL-8/SB225002/ BAY11-7082	[119]
Lipidolysis	3T3-L1 adipocytes, inguinal/epididymal WAT (iWAT/eWAT)	Differential ultracentrifugation	Exosomes/Unspecified	LLC-exosomes induce lipolysis/WAT browning; GW4869 inhibits exosome release to block lipolysis	in vitro and in vivo	Adipocyte lipolysis, WAT browning, fat mass loss; reversed by GW4869	[120]
Muscle atrophy	C2C12 myoblasts, primary myoblasts, Pax7⁺ muscle satellite cells	Differential ultracentrifugation	MVs/ miR-21	Binds TLR7, activates JNK signaling to induce myoblast apoptosis	in vitro and in vivo	Induces myoblast/satellite cell apoptosis, reduces myofiber nuclei, muscle mass loss	[118]
Muscle Atrophy	C2C12 myotubes, primary rat myotubes, skeletal muscle, muscle satellite cells	ExoQuick kit	EVs/ HSP70/90	Binds TLR4, activates p38β MAPK-C/EBPβ pathway, upregulates UPP/ALP	in vitro and in vivo and clinical correlation	Induces myotube atrophy, muscle protein degradation, muscle mass loss; reversed by anti-HSP70/90/ Rab27a/b silencing/ TLR4 KO	[23]
Muscle atrophy	C2C12 myoblasts/myotubes, skeletal muscle	Differential ultracentrifugation	MVs/ miRNAs, snoRNA	1. Delays myogenesis via targeting Myf5/Myog/ Myh; 2. Impairs mitochondria	in vitro and in vivo	Delays myoblast differentiation, reduces mitochondrial respiration, increases lactate production; Induces muscle mass loss	[116]
Muscle atrophy/therapeutic intervention	C2C12 myoblasts, human primary myoblasts, skeletal muscle	Differential ultracentrifugation	miR-21, miR-29a	miR-21/29a-TLR7/8-JNK axis induces apoptosis; IMO-8503 competitively binds TLR7/8 to block axis	in vitro, in vivo and clinical specimen studies	Induces myoblast apoptosis, Pax7 upregulation, and lean mass loss; reversed by IMO-8503 without tumor growth effect.	[121]

EVs activate the ubiquitin-proteasome degradation pathway

Driving UPS-mediated proteolysis is a core mechanism by which EVs cause muscle atrophy. In vitro experiments demonstrate that EV IL-6, upon delivery to muscle cells, induces sustained activation of the STAT3 signaling pathway; this subsequently triggers a direct and precise upregulation in the transcription of MuRF1 and Atrogin-1 [21]. As shown in Fig. 6, the upregulation of these two E3 ligases is a key step in initiating muscle protein degradation. Additionally, in animal models demonstrate that EV HSP70/90 can synergistically enhance catabolic signaling by activating the TLR4/p38 MAPK/C/EBPβ pathway [23]. Further evidence regarding EVs activating UPS via pathways such as IL-6/STAT3 can be found in Table 3.

EVs impair muscle regeneration and function maintenance

EVs also damage muscle repair and regeneration capacity. In vitro studies show that tumor-derived EVs can carry factors that inhibit myoblast differentiation, delaying or blocking the muscle regeneration process by interfering with the function of key regulators such as MyoD and myogenin, thereby weakening the muscle’s recovery potential after injury [116]. The inhibitory effects of EVs on myogenesis and regeneration and the related molecules are detailed separately in Table 3.

Mechanism integration and therapeutic implications

In summary, lung cancer-derived EVs launch a multi-dimensional attack on skeletal muscle by delivering various cargoes such as miR-21, IL-6, and HSP70/90, precisely targeting key nodes of the network depicted in Fig. 6: inducing apoptosis, activating protein degradation, amplifying catabolic signals, and impairing regenerative capacity. This combined effect of “promoting death, promoting decomposition, and inhibiting regeneration” ultimately leads to severe muscle atrophy. Similar to adipose tissue, while animal models establish causality, clinical observations primarily highlight the association between EV profiles and muscle wasting. This understanding has important therapeutic implications: targeting EV biogenesis, secretion, or neutralizing their specific toxic cargo has emerged as a potential strategy to alleviate muscle wasting in cancer cachexia. These EV-mediated mechanisms of muscle atrophy have been systematically summarized in Table 3.

Integrated model: the EV-driven systemic metabolic axis in lung cancer

This review synthesizes evidence establishing EVs as the core mediators linking local tumor progression with systemic host depletion in lung cancer, forming an “EV-metabolic axis”. Locally, EVs reprogram TME metabolism to support tumor growth. More critically, these EVs enter the circulation, functioning as systemic nanoscale transport carriers that precisely deliver their specific metabolic instructions to distant metabolic organs. In adipose tissue, EV cargo drives fat consumption; in skeletal muscle, it leads to muscle atrophy. This coordinated attack on the two major energy reservoir organs, mediated by EVs from the same source, constitutes the core mechanism of systemic tissue wasting in cancer cachexia. Figure 7, as the central integrative model, systematically illustrates the pivotal role of EVs as systemic nanoscale transporters in coordinating local TME metabolic reprogramming and driving fatal distal tissue consumption.

Fig. 7.

EVs as systemic nanoscale transporters coordinating the metabolic axis in lung cancer progression and cachexia. This integrative schematic depicts the dual role of lung cancer-derived EVs as both local metabolic orchestrators and systemic pathogenic messengers, forming a coherent “EV-metabolic axis”. Left panel: Local metabolic reprogramming in the TME: Within the primary tumor, EVs secreted by cancer cells and activated stromal components (CAFs, TAMs) transfer a diverse cargo (e.g., metabolic enzymes, regulatory RNAs, signaling proteins) to recipient cells. This exchange reprograms cellular metabolism, enhancing glycolysis, modulating nutrient utilization, and fostering an immunosuppressive niche that supports tumor growth and adaptation, as detailed in chap. 3. Central panel: Systemic dissemination via circulation, EVs from the TME enter the bloodstream, highlighting their function as systemic nanoscale transporters. This panel emphasizes their nanoscale size, lipid bilayer structure, and the encapsulated molecular payload (proteins, nucleic acids, lipids) that are protected and transported to distant sites. Right panel: Systemic tissue wasting in cachexia circulating EVs home to and are taken up by adipose tissue and skeletal muscle, executing tissue-specific catabolic programs. In adipose tissue, EV-(PTHrP, GRP75) activates lipolysis and thermogenesis while inhibiting adipogenesis, leading to fat loss. In skeletal muscle, EV cargo (e.g., IL-6, miR-21, HSP70/90) activates proteolysis via the ubiquitin-proteasome system (upregulating MuRF1/Atrogin-1), triggers apoptosis, and amplifies catabolic signaling, collectively causing muscle atrophy, as detailed in chap. 4

EV-based nanomedicine applications

The critical role of EV in the metabolic remodeling of the lung cancer microenvironment and the development of cachexia has opened new avenues for their use as clinical diagnostic markers and therapeutic targets [122]. As natural nanocarriers, the specific metabolism-related molecules carried by EV reflect the metabolic state of the tumor, making them ideal liquid biopsy markers [22]. Concurrently, the bidirectional regulatory capacity of EV in metabolic remodeling has inspired the design of novel therapeutic strategies, including pharmacological interventions targeting EV biogenesis and the use of engineered EV to deliver metabolic modulators. This integration of basic research and clinical application is poised to advance precision medicine in lung cancer [123, 124].

EVs as biomarkers for lung cancer diagnosis

Compared to traditional markers such as cytokeratin 19 fragment (CYFRA 21 − 1), carcinoembryonic antigen (CEA), and neuron-specific enolase (NSE), which have limited sensitivity (42%) and specificity (83%), EV-related markers show unique advantages in the early diagnosis and non-invasive monitoring of lung cancer [125]. EVs isolated from various bodily fluid samples carry a wealth of molecular information, as detailed in Table 3. Examples include: miR-619-5p in cell culture supernatants [126]; metabolites like Kanzonol Z and ceramidyl-carnitine in urine (with an AUC: Area Under the Curve value up to 96%) [123]; Thomsen-Friedenreich Antigen-alpha (TF-Ag-α) in serum (diagnostic accuracy exceeding 95%) [127]; and cachexia-related factors like EV-(PTHrP), EV-(IL-6), and EV-(miR-21) [21, 106, 121]. See Table 4 for other aspects of metabolic regulation. These findings provide strong support for the development of novel, highly sensitive, and specific diagnostic methods for lung cancer.

Table 4.

EV as biomarkers for diagnosis of lung cancer

Sample Source (Producing cell)	EV Type	Isolation method	Key molecular cargo	Clinical observation (Levels)	Associated pathological/ metabolic process	Ref
A549 cell supernatant	Exosomes	Differential ultracentrifugation and ultrafiltration and exoEasy Maxi kit	miR-619-5p, miR-122-5p, SPP1, CD44	High miR-619-5p expression and CD44/SPP1 association with NSCLC progression	Metabolic reprogramming, EMT, angiogenesis	[126]
Urine from lung cancer patients/healthy controls	EVs	EXODUS device purification	Metabolites: 4 diagnostic markers	Diagnostic AUC = 0.84–1.00.84.00; distinguishes early-stage lung cancer	Purine metabolism, beta-Alanine metabolism etc.	[123]
NSCLC tissues/ patient serum/ NSCLC cell lines	Exosomes	ExoQuick kit	circ-MEMO1	High expression associated with advanced stage and lymph node metastasis; diagnostic AUC = 0.76	Aerobic glycolysis, cell proliferation, G1/S transition	[128]
LLC cell supernatant/ serum/3T3-L1 adipocytes	EVs	Differential ultracentrifugation	PTHrP	High expression associated with poor prognosis; induces lipolysis	Adipocyte lipolysis, WAT browning, cachexia	[106]
LLC cell supernatant/lung cancer patient serum/tissues from tumor-bearing mice	EVs	Differential ultracentrifugation	IL-6	High expression associated with poor prognosis; induces muscle atrophy/lipolysis	Skeletal muscle atrophy, lipolysis, cachexia (IL-6-STAT3 axis)	[21]
Supernatant of cachexia-inducing tumor cells/serum from tumor-bearing mice	EVs	ExoQuick kit/Differential ultracentrifugation/CD9 immunoprecipitation	HSP70, HSP90	Serum levels increased 3–4 fold; associated with muscle atrophy	Skeletal muscle atrophy, protein degradation, cachexia	[109]
Tumor cell supernatant/healthy human PBMC/mouse tumor tissues	EVs	Differential ultracentrifugation	PD-L1	Induces T cell senescence; no induction of exhaustion	T cell senescence, lipid metabolism reprogramming, immunosuppression	[22]
Peripheral blood from SCLC patients/healthy controls	Large EVs	Stepwise centrifugation	Lipid metabolites: TG, Cer, CE families	Diagnostic AUC = 0.84–0.90; high expression associated with poor OS	Lipid metabolism reprogramming, sphingolipid metabolism, necroptosis	[59]
Supernatant of TKI-sensitive/resistant cells/serum from NSCLC patients	EVs	Differential ultracentrifugation/isolation reagent	SLC1A5, SLC25A5, ALDH1A1	High SLC1A5 associated with short PFS; induces TKI resistance	TKI resistance, glutamine metabolism/OXPHOS, TME remodeling	[96]
M0/M2 macrophages/ NSCLC cell lines/ patient tissues	Exosomes	ExoQuick kit	LINC01001	High expression in M2-derived EVs; correlated with METTL3	Glycolysis, NASP methylation, CD8⁺ T cell inhibition	[76]
Serum from LUAD patients/ healthy controls/ LUAD cell lines	Exosomes	ExoQuick kit	LINCLUCAT1	Diagnostic AUC = 0.852; associated with advanced stage and poor OS	Tumor metastasis, glycolysis, VEGFA-mediated angiogenesis	[129]
NSCLC tissues/ NSCLC cell lines/ cell-derived exosomes	Exosomes	Ribobio exosome isolation kit	LINCHOTAIRM1	High expression; negatively correlated with miR-498	Tumor proliferation /metastasis, glycolysis, apoptosis inhibition	[130]

EVs as nanobiotechnological tools and targets

As an emerging biological carrier, the core value of EVs in lung cancer therapy lies in their immense potential as metabolic therapeutic targets. This includes directly targeting the generation and release of EVs, as well as utilizing them as carriers to deliver metabolic modulators.

Drug Delivery vesicles: EVs possess excellent biocompatibility and low immunogenicity, enabling the effective delivery of anti-cancer drugs. For example, doxorubicin (DOX) and paclitaxel (PTX) can be targetedly delivered to lung cancer cells via EVs [131, 132]. A strategy proposed by Zheng et al.., involving the aerosol inhalation of PTX-loaded chimeric antigen receptor-EVs (CAR-EVs), achieved in-situ targeting of lung cancer and demonstrated superior efficacy over free drugs [122]. In thetyrosine kinase inhibitors (TKIs) like Sotorasib have shown efficacy against KRAS G12C-mutated NSCLC, acquired resistance remains a major hurdle. Key resistance mechanisms involve metabolic reprogramming and the development of an immune-tolerant TME. EV-based delivery systems are being explored to co-deliver KRAS inhibitors with metabolic or immunomodulatory agents to overcome these challenges.

Immunotherapy: EVs derived from LLC cells, rich in tumor antigens, can stimulate anti-tumor immunity. For instance, engineered EV-like nanovesicles can induce a cytotoxic T lymphocyte (CTL) immune response (157). EVs from LLC cells, rich in tumor antigens, can stimulate anti-tumor immunity. For instance, EV-like nanovesicles engineered with fibroblast activation protein (FAP) (eNVs-FAP) can induce a cytotoxic T lymphocyte (CTL) immune response and remodel the immunosuppressive TME [133]. A novel approach in immunotherapy involves EVs loaded with IL-12 mRNA. Preclinical studies show that this strategy enables targeted delivery to the lung TME, significantly reducing the systemic toxicity associated with recombinant IL-12. This localized therapy induces a robust anti-tumor immune response, characterized by enhanced T-cell infiltration and the formation of immune memory, demonstrating strong potential for treating both primary and metastatic lung umors [134]. Furthermore, DC-derived EVs coupled with IL-12 and anti-cytotoxic T-lymphocyte-associated antigen 4 (aCTLA-4) can effectively inhibit tumor growth in lung cancer-bearing mice [135].

Metabolic Regulation and Cachexia Treatment: This is the most promising direction for EVs as therapeutic targets. Shikonin was found to increase cisplatin sensitivity by inhibiting the secretion of PKM2 in EVs, thereby affecting the glycolytic metabolic pathway [136]. A team from Huazhong University of Science and Technology developed a technology to encapsulate succinate in tumor EVs (stable succinate-encapsulated microparticles, SMPs), which can effectively polarize the metabolic pattern of tumor-associated macrophages and inhibit tumor progression [137]. In cachexia treatment, downregulating or neutralizing HSP70/90 in LLC cell-derived EVs can inhibit muscle catabolic signaling pathways and alleviate muscle atrophy [23]. The emergence of Physiactisome, the first patented EVs drug for treating muscle atrophy, also brings hope for overcoming cancer cachexia [138].

Translational status and limitations of EV-targeted strategies

As endogenous nanocarriers, the therapeutic potential of EVs in nanobiotechnology is vast, primarily focusing on EV engineering, drug delivery, and targeting EV biological processes [139]. However, when proposing the inhibition of EV biogenesis or uptake as a therapeutic strategy, it is vital to distinguish between robust preclinical validation and conceptual stages. Strategies like using omeprazole to inhibit EV release have shown promise in in vivo preclinical models of cachexia [109]. Similarly, inhibiting the production of pro-cancer EVs using pharmacological agents (e.g., amiloride, GW4869) can block their pathological effects at the source [120, 140]. Despite these advances, many approaches remain conceptual or in early in vitro stages. Significant limitations hinder immediate clinical translation, including off-target toxicity (as physiological EVs are essential for normal homeostasis), unpredictable biodistribution, and the lack of highly specific inhibitors for tumor-derived EVs versus normal EVs. EV Engineering aims to overcome some limitations, such as low drug-loading efficiency and insufficient targeting [141]. Displaying targeting ligands on the EV surface through genetic engineering significantly improves their specificity for tumor cells, which is a core concept in “nano-bio interface” design [142, 143].

Clinical applications

The translation of EV-based therapies from the laboratory to the clinic is an area of intense research. Currently, numerous clinical trials are underway to evaluate the safety and efficacy of EVs as drug delivery vehicles and therapeutic agents for various diseases, including cancer (data from ClinicalTrials.gov). For instance, trials have explored using plant-derived exosomes to deliver chemotherapeutics like curcumin and paclitaxel (NCT01294072, NCT01668849). In lung cancer, clinical studies are investigating EVs as vaccines, drug carriers, and metabolic intervention tools. The development of Physiactisome, the first patented EV drug for treating muscle atrophy, also brings hope for overcoming cancer cachexia [138]. Despite these advances, challenges related to large-scale production, quality control, and storage efficiency must be addressed to realize the full clinical potential of EV-based nanomedicine.

Conclusion and perspectives

This review synthesizes a substantial body of evidence to systematically elaborate on and validate the EV-metabolic axis as a pivotal, systemic regulatory network that fundamentally drives both the local progression of lung cancer and the development of cancer cachexia. Our analysis establishes that EVs, functioning as “natural nanocarriers” and endogenous nanomedicine platforms, are active commanders rather than passive bystanders. They functionally bridge the localized TME with distant host tissues, orchestrating a dual pathological cascade.

By serving as versatile vectors for a diverse molecular cargo-(including metabolic enzymes (e.g., ALDOA, PKM2), regulatory RNAs (e.g., miR-425-3p, circRNA_101093, LINC01001), and signaling molecules (e.g., HSP70/90, IL-6)—EVs reprogram the metabolic landscape both locally and systemically [74, 76, 91, 97, 105, 109, 144]. Within the local TME, this cargo exchange fosters a nutrient-rich, immunosuppressive niche that supports tumor growth, invasion, and therapy resistance. Simultaneously, these same EVs enter the systemic circulation to disseminate catabolic instructions. They precisely target distant metabolic organs: in adipose tissue, EV cargo (e.g., PTHrP, GRP75) drives lipolysis and pathological “browning”; in skeletal muscle, cargo (e.g., IL-6, miR-21) activates proteolysis and apoptosis [145]. This coordinated attack on the body’s principal energy reservoirs catalyzes the severe tissue wasting characteristic of cachexia.

The paradigm shift offered by the EV-metabolic axis model lies in its identification of a unified, targetable biological system responsible for the co-evolution of tumor sustenance and host depletion. It moves beyond earlier explanations that attributed cachexia to a diffuse mix of inflammatory cytokines, by pinpointing EVs as the specific, information-rich vectors of this fatal metabolic crosstalk-. This integrated perspective fundamentally reframes advanced lung cancer from a localized disease to a systemic metabolic disorder, with EV-mediated nanoscale communication at its core.

Clinical Implications and Future Perspectives

The clinical implications of this axis are profound. Diagnostically, EVs present a treasure trove of biomarkers for liquid biopsy, enabling non-invasive monitoring of tumor dynamics and cachectic progression. Therapeutically, the axis presents a unique “two-pronged” intervention strategy. Targeting EV biogenesis, secretion, or uptake can simultaneously cripple the tumor’s local support network and shield the host from metabolic collapse. Strategies such as neutralizing specific pathogenic EV cargo (e.g., HSP70/90) or inhibiting their release (e.g., via drugs like omeprazole) have shown promise in preclinical models for alleviating cachexia [109, 120]. Furthermore, engineered EVs present a next-generation platform for targeted drug delivery and immunotherapy, holding immense promise for combination therapies [142].

However, the translation of this knowledge faces significant challenges. Even when following the MISEV2023 guidelines for rigorous characterization, the inherent heterogeneity of EVs and limitations in isolation techniques still complicate the attribution of specific functions to distinct subtypes [36]. The molecular mechanisms governing selective cargo sorting remain incompletely elucidated. Most evidence is derived from preclinical models, necessitating rigorous validation in well-designed clinical trials with standardized protocols.

To fully harness the therapeutic potential of the EV-metabolic axis, future research must prioritize:

Advanced Characterization: Employing single-vesicle analysis technologies and multi-omics approaches (proteomics, lipidomics, RNA-seq) to deconvolute EV heterogeneity and definitively link specific subpopulations and cargo profiles to distinct functional outcomes in metabolism and cachexia [36].

In Vivo Dynamics: Conducting in-depth in vivo tracking studies using labeled EV to map their biodistribution, identify specific recipient cells in different tissues, and validate the causal relationships proposed in vitro [146, 147].

Mechanistic Deep Dive: Investigating the molecular machinery governing selective cargo loading into EV and the specific receptors mediating their uptake in target adipose and muscle tissues.

Clinical Translation: Prioritizing the development of robust, standardized EV-based biomarker panels and moving them into clinical validation cohorts. Simultaneously, designing innovative clinical trials to evaluate the safety and efficacy of EV-targeted therapies, ensuring they incorporate biomarker-driven patient stratification [60].

Limitations of the “EV-metabolic axis” Conceptual Framework: While the “EV-metabolic axis” provides a compelling model, it is not without limitations. First, the framework heavily relies on data from highly controlled in vitro and animal models, which may not fully recapitulate the complex, heterogeneous nature of human lung cancer and cachexia [102, 148–150]. Second, the current model often attributes specific metabolic shifts to isolated EV cargoes, potentially oversimplifying the synergistic or antagonistic effects of the myriad of molecules within a single EV [151–153]. Finally, distinguishing the relative contribution of EVs from other systemic factors (e.g., soluble cytokines, metabolites) in driving cachexia remains technically challenging in clinical settings. Addressing these limitations requires advanced in vivo tracking and multi-omics approaches in human cohorts [24, 154–156].

In conclusion, the EV-metabolic axis provides a powerful and holistic conceptual framework that deepens our understanding of lung cancer as a systemic disease. This field, situated at the intersection of EVs biology, cancer metabolism, and nanomedicine, inherently fosters the interdisciplinary collaboration essential for tackling complex diseases. By continuing to decipher the intricate language of EV-mediated communication, we can advance towards the development of holistic therapeutic strategies. These strategies will not only aim to eradicate the tumor but also to protect the metabolic integrity of the patient, offering a tangible path to mitigate the debilitating syndrome of cachexia and improve overall survival and quality of life.

Author contributions

Author Contributions: XXZ, DML, MW and QYG contributed equally to this work. XXZ contributed to conceptualization, drafting, writing the original draft, and revision. DML contributed to drafting and revision. MW and QYG participated in the revision of the manuscript. YDG was responsible for graphical and textual revision. PPW and WXF contributed to conceptualization, supervision, and graphical and textual revision. WLZ contributed to conceptualization, supervision, and graphical and textual revision. All authors read and approved the final manuscript.

Funding

This study was supported by grants from the National Natural Science Foundation of China (No.82303775); the Natural Science Foundation of Anhui Province (2308085QH273).

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.