There are various theories regarding the origin of CAFs, but their exact origins are still not fully understood. It is currently believed that CAFs mainly arise from resident fibroblasts, bone marrow mesenchymal stem cells (BMSCs), adipocytes, pericytes, endothelial cells, epithelial cells, and smooth muscle cells (28, 50, 51) ( Figure 1 ). Through the influence of growth factors, cytokines, chemokines, and epigenetic modifications, different signaling pathways in tissue resident normal fibroblasts can be activated and transformed into CAFs (52, 53). Metabolic reprogramming, induced by metabolic products from cancer cells, aging fibroblasts, inflammatory cells, and mechanical forces within the ECM, is another factor that contributes to the conversion of normal fibroblasts into CAFs. In various types of cancer tissues, resident fibroblasts present in the surrounding areas are an essential source of CAFs within the TME. Growth factors, cytokines, and chemokines can also recruit and activate BMSCs and stationary stellate cells, transforming them into CAFs (54–56). TGF-β1 has been found to induce endothelial cells to undergo a phenotypic transformation into fibroblasts. Additionally, adipocytes can promote fibroblast differentiation by directly interacting with tumor cells through adipose stem cells. Furthermore, pericytes, epithelial cells, and smooth muscle cells can also undergo conversion into CAFs (57, 58).

The heterogeneity of CAFs can be attributed to the various pathways through which they originate. CAFs express different markers such as α-smooth muscle actin (α-SMA), fibroblast activating protein (FAP), tendon protein C, periosteal proteins, NG2 chondroitin sulfate proteoglycans, and platelet-derived growth factor α (PDGF-α) (59–62). FAP is commonly used in tumor therapy and diagnostic imaging and is involved in regulating the extracellular matrix, exhibiting pro-tumorigenic activities. α-SMA is used to evaluate the therapeutic function of targeted CAFs and promotes tumor angiogenesis by regulating cell proliferation and the transport of pro-angiogenic factors. Inhibition of α-SMA transport attenuates angiogenic capacity and thus inhibiting tumor progression. Other markers such as waveform protein, proline 4 hydroxylase, fibronectin, type I collagen, fibroblast-specific protein-1, and fibroblast surface proteins are used to characterize mesenchymal stromal cells, including CAFs. The existence of heterogeneity in CAFs is primarily due to differences in cellular phenotypes, which exhibit two characteristics: temporal (associated with tumor development stage) and spatial (distinct phenotypes in different areas of cancer tissues) (63, 64). CAFs can be categorized into reactive CAFs (rCAFs), myofibroblast CAFs (myCAFs), inflammatory CAFs (iCAFs), and antigen-presenting CAFs (apCAFs). The combined effect of myCAFs, iCAFs, and apCAFs obviously promote cancer cell proliferation, migration, invasion, metastasis, and drug resistance, ultimately contributing to cancer development (65, 66) ( Figure 2 ). The role of CAFs in various aspects of cancer is well established. CAFs can promote cancer progression through different mechanisms. myCAFs can promote ECM remodeling by synthesizing collagen and regulating mechanical conduction. ICAFs can regulate the immune process of the body by altering their secretion characteristics. ApCAFs can activate CD4+T cells in an antigen-specific manner. The combined effect of myCAFs, iCAFs, and apCAFs can ultimately significantly promote the proliferation, migration, invasion, metastasis, and drug resistance of cancer cells, thereby promoting the development of cancer.

The function of CAFs in different aspects of cancers has been well established, and they could accelerate cancer progression through a variety of mechanisms ( Figure 3 ). Most of the current studies support that CAFs primarily take a tumor-promoting function, and it is believed that targeting CAFs may be a future-proof tactics for cancer therapy. First, CAFs can promote tumor growth. CAFs are widely distributed in the tumor stroma and are in close contact with tumor cells, and it can directly mutual effect with tumor cells to accelerate tumor growth (67, 68). Besides, CAFs are the main source of growth factors, cytokines and exosomes in the TME, which can indirectly promote tumor growth (69–71). ITGB2-mediated metabolic switching of CAFs promotes oral squamous cell carcinoma (OSCC) proliferation through the oxidation of NADH in the mitochondrial oxidative phosphorylation system (72). Cancer metastasis is a multilevel procedure. Firstly, cancer cells break through the substrate membrane to migrate and invade into the ambient tissues. After endocytosis, dissemination and exocytosis, several surviving cancer cells will ectopically colonize other tissues, and ultimately form metastatic foci that are visible to the naked eye. CAFs have a crucial promotional function in the process of tumor metastasis (73, 74). Exosomes derived from CAFs miR-18b could accelerate breast cancer invasion and metastasis by regulating TCEAL7 (75). Resistance to cancer therapies usually result in tumor progression, and many studies have identified a function for stromal CAFs in tumor drug resistance. PDPN-positive CAFs can accelerate resistance to trastuzumab in HER2-positive breast cancer by holding back antibody-dependent NK cell-mediated cytotoxicity (76). Factors secreted by CAFs also affect other components of the TME, and they could do it according to diversity of immune cells, such as CD8⁺ T cells, regulatory T cells (Tregs), and macrophages, with immunomodulatory effects (77, 78). Most of the current studies suggest that the main role of CAFs is immunosuppression such as the immunosuppressive molecules IL-10, TGF-β and certain CXCL secreted by CAFs can significantly inhibit the proliferation and activation of T cells.

EVs are particles made up of lipid bilayer membranes that encapsulate the cytoplasmic sol compartment. They can be formed by outward sprouting of the plasma membrane or through the intracellular endocytic transport pathway, where multiple vesicles fuse with the plasma membrane in late-stage endocytic compartments ( Figure 4 ). EVs can be categorized into exosomes, microvesicles, and apoptotic bodies based on their biochemical characteristics. Exosomes are generated through endosomal sorting complexes required for transport (ESCRT)-dependent or non-ESCRT-dependent pathways, which sort specific contents into exosomes (91, 92). Exosomes can fuse with the target cell membrane and deliver their cargo to the recipient cell through receptor-mediated endocytosis, megalocytosis, or phagocytosis (93, 94). Exosomes can also directly interact with target cells through receptor-ligand interactions for intercellular communication (95). Furthermore, the composition and biogenesis of exosomes can affect the condition of secretory cells and regulate the intra- and extracellular environment (96, 97).

Exosomes carry out diverse biological functions that depend on their cellular origin. The most crucial function of exosomes is signal transduction, as they transport molecules that regulate various physiological and pathological processes (98). As carriers of cellular communication, exosomes transport their substance to neighboring or distant cells, thereby regulating multifarious physiological and pathological procedure. Exosomes can also regulate the inflammatory vesicle activation. Stem cell-derived exosomes inhibit inflammatory vesicle activation, yet immune cell-derived exosomes trigger inflammatory vesicle activation, advising that exosomes maybe act as a therapeutic tool for diseases associated with inflammatory responses (99). Immune cell-derived exosomes also play a role in modulating immune responses, depending on the cellular environment in which they are produced (100, 101). Besides, exosomes have been implicated in neurovascular regeneration by promoting nerve regeneration and functional recovery through regulation of axon growth (102). They also serve as important biomarkers for disease diagnosis, particularly in the early detection of cancer (103). With the increasing research on exosomes, the role of them in tumor pathology has been continuously revealed. Tumor cell-secreted exosomes can contribute to the formation of a TME that make tumor cells evade immune surveillance and facilitate tumor growth (104). In addition, exosomes secreted by tumor cells can induce neoangiogenesis, which ensures access to nutrients and contributes to the continued proliferation of tumor cells (105).

Exosomes are dispersed in complex human body fluid specimens in a non-uniform membranous vesicle structure. Therefore, the separation and enrichment of exosomes is a key step in their detection technology. The purity and activity of the obtained exosomes can directly affect subsequent identification and functional analysis. At present, commonly used exosomes separation and enrichment techniques are mainly based on the physical and immunological characteristics of exosomes (106, 107). The centrifugation method separates exosomes by gradually increasing centrifugation force or time based on the differences in sedimentation coefficients of vesicles, cells, cell fragments, and protein molecules in solution (108). It mainly includes differential centrifugation and density gradient centrifugation. The precipitation method separates and purifies exosomes by co precipitation or reverse screening based on the physicochemical properties of compounds and exosomes (109). It mainly includes polymer precipitation method and organic solvent precipitation method. The particle size separation method separates exosomes based on the difference in particle size between them and other biomolecules, mainly including ultrafiltration and size exclusion chromatography (SEC) (110). Immunoaffinity methods use antigen-antibody interactions to specifically bind antibodies to exosomes (111). In addition, researchers have attempted to combine traditional techniques with microfluidic techniques based on the physicochemical characteristics of exosomes and developed new methods for exosomes separation, including capture microfluidic technology, filtration microfluidic technology, magnetic separation microfluidic technology, acoustic separation microfluidic technology, and dielectric electrophoresis microfluidic technology (112). Furthermore, the magnetic separation method based on aptamers involves modifying the surface of magnetic beads with aptamers, using them as recognition capture agents to specifically bind with exosomes surface proteins for exosomes capture (113). At present, there are more and more nucleic acid aptamer sensors for isolating exosomes, which have been successfully applied in the extraction of exosomes from samples such as urine, serum, and plasma.

The identification of exosomes involves analyzing their morphological characteristics, concentration, particle size, protein molecules, and cumulative components (114). Techniques like scanning electron microscopy, transmission electron microscopy, atomic force microscopy, and random optical reconstruction microscopy can be used to study the morphological features of exosomes (115, 116). Nanoparticle tracking analysis (NTA), dynamic light scattering (DLS), and tunable resistance pulse sensing technology provide real-time concentration and size analysis of exosomes (117). Flow cytometry is also applicable for particle number analysis of exosomes (118). Protein molecule characterization and concentration can be obtained through Western blot, ELISA, and flow cytometry (119). Topological identification focuses on exosomal membrane proteins and vesicular proteins and involves treatments, separations, proteomic analysis, verification, and analysis using techniques like protease treatment, biotinylation, liquid chromatography tandem mass spectrometry, mass spectrometry, flow cytometry, fluorescence microscopy, and bioinformatics.

CCA is a lethal tumor originating from epithelial cells in the bile ducts (120). CCA often presents at advanced stages, limiting surgical options (121, 122). Traditional radiotherapy is not very effective, and alternative treatments like molecular targeting and immunization require further exploration (123, 124). Studies have shown that CCA cell-derived exosomes can accelerate the malignant development of CCA by constraining the conversion of NFs to CAFs and thereby promoting the lethal progression of CCA. Qin et al (125), demonstrated that miR-34c expression was reduced in CCA cell exosomes, and exosomal miR-34c could be transferred to fibroblasts and mediate their activation. Mechanistic experiments indicate that miR-34c inhibits CCA progression by targeting WNT1, a gene involved in tumor growth. Similarly, decreased levels of miR-206 in intrahepatic CCA (iCCA) contribute to cell proliferation, migration, and invasion. Co-culturing CCA cells with NFs reduces miR-206 expression in NFs and promotes their transformation into CAFs. This interaction leads to increased aggression and resistance to gemcitabine in CCA cells. Overexpressing miR-206 either in CAFs or CCA cells suppresses this reciprocal promotion and inhibits tumor progression (126). However, further research is needed to fully understand the interaction between extracellular vesicles and CAFs in CCA, as well as explore potential therapeutic strategies targeting this interaction.

Surgical resection, liver transplantation, ablation, hepatic artery embolization and systemic chemotherapy, targeted drugs and immunotherapy are only effective for long-term survival in a limited number of patients (127, 128). Researches have shown that CAFs could restrain the biological processes of HCC cells through the secretion of exosomes ( Table 1 ). Zhang et al (129), discovered that the expression level of miR-320a in CAFs-derived exosomes was obviously decreased in HCC and could be transported to HCC cells. Functional experiments revealed that exosomal miR-320a repressed the proliferation, migration and metastasis of HCC cells. Mechanistic experiments revealed that miR-320a could inhibit the activation of the MAPK pathway by targeting PBX3, and further inhibited the process of epithelial-mesenchymal transition (EMT), also the expression of cell-cycle protein-dependent kinase 2 (CDK2) and MMP2, thereby inhibiting cell proliferation and metastasis. Ultimately, this inhibits cell proliferation and metastasis. Therefore, CAFs delivering miR-320a to HCC cells could potentially be a therapeutic target for HCC. Yugawa et al (130), discovered that co-culturing HCC cells with conditioned medium from safranin-treated CAFs inhibits cell viability and invasiveness, while promoting apoptosis. Besides, the expression level of miR-150-3p was closely associated with the clinical characteristics of HCC patients. Lv et al (131), demonstrated that Safranin treatment also blocks the secretion of a circular RNA called circCCT3 in CAFs-derived exosomes. Besides, the expression of circCCT3 was significantly elevated in HCC, and treatment with safranin blocked the secretion of circCCT3 in the exosomes of CAFs. It is found that circCCT3 regulates glucose metabolism in HCC by affecting the expression of HK2. Liu et al (132), demonstrated the mechanism behind this involves the direct inhibition of DNA methyltransferase 3b (DNMT3b) expression by miR-29b, along with the up-regulation of metastasis suppressor 1 (MTSS1) expression, ultimately inhibiting the progression of HCC.

Researchers have found that exosomes derived from CAFs played a role in promoting aggressive behaviors in HCC cells ( Table 1 ). These exosomes contain specific molecules that can accelerate cell migration, invasion, and metabolic reprogramming. For example, exosomal TUG1 identified by Lu et al., enhanced HCC progression by binding to miR-524-5p and upregulating SIX1 expression (133). Another study by Qin et al,. showed that exosomal Gremlin-1 from CAFs reduced the sensitivity of HCC cells to sorafenib and promotes epithelial-mesenchymal transition (EMT) by inhibiting the Wnt/β-catenin and BMP signaling pathways (134). Furthermore, Gremlin-1 was enriched in plasma exosomes of HCC patients and could predict the resistance to sorafenib in HCC patients. CAFs derived exosomes could also dramatically elevate the rates of intrahepatic and pulmonary metastasis of HCC. Jin et al (138), demonstrated that exosomes separated from CAFs could be absorbed by HCC cells and accelerated their invasive phenotype compared with para-cancerous fibroblasts (PAFs), and mechanistic experiments revealed that CAFs could offer multiple miRNAs (miR-329-3p, miR-380-3p, miR-410-5p, miR-431-5p) were delivered to HCC cells and increased their expression levels. Additionally, downregulation of LIMA1 expression was associated with poor OS and recurrence-free survival (RFS) in HCC patients. LIMA1 acted as a suppressor of HCC progression by inhibiting the Wnt/β-catenin signaling pathway. Treatment with miR-20a-5p-overexpressing CAFs exosomes inhibited LIMA1 expression and promoted aggressive tumor behavior in HCC (135). In addition, treatment of HCC cells with miR-20a-5p overexpressing CAFs exosomes inhibited LIMA1 expression and promoted lethal tumor progression in vitro and in vivo. Zhou et al (136), demonstrated that circZFR manifestation was boosted in cisplatin (DDP)-resistant HCC cells as well as in CAFs and CAFs-derived exosomes. Functional experiments revealed that circZFR could be obtained from CAFs cells to HCC cells via exosomes and accelerate tumor development and DDP resistance. Mechanistic experiments showed that circZFR could constrain the STAT3/NF-κB pathway and thus accelerate the progress of HCC and chemoresistance.

HCC cell-derived exosomes could also translocate into fibroblasts to promote a pro-tumorigenic microenvironment ( Table 1 ). Fang et al (137), revealed that the exosome miR-1247-3p secreted by high metastatic HCC cells directly targetd B4GALT3, thereby activating β1-integrin-NF-κB signaling in fibroblasts. Activated CAFs further accelerated cancer progression by secreting pro-inflammatory cytokines, including IL-6 and IL-8. Clinical data showed that high levels of the serum exosome miR-1247-3p were associated with lung metastasis in HCC patients. Zhou et al (139), demonstrated that HCC cells had a strong ability to deliver common hematopoietic stem cells into CAFs. Similarly, exosomal miRNA-21 from HCC cells could be delivered to hematopoietic stem cells, which then transform into CAFs through the activation of the PDK1/AKT signaling pathway. These CAFs contributed to cancer progression through the secretion of angiogenic cytokines.

In terms of therapeutic applications, exosomes can be utilized as vehicles to deliver anticancer drugs for targeted treatment of HCC. However, more research is needed to fully understand the molecular mechanisms and targets involved in exosome-mediated drug delivery. Overall, further investigation is warranted to advance the translation of exosome-based technologies into clinical applications for HCC.

CRC is a deadly tumor that affects the GI tract and is a leading cause of cancer-related deaths worldwide. Currently, the primary examination modality for the primary diagnosis, screening and detection of CRC is colonoscopy (140). The main treatment modality for primary CRC is laparoscopic surgical treatment, and the main treatment modality for metastatic CRC is surgical resection combined with radiotherapy treatment (141). Despite advancements in screening and treatment, metastasis occurs in about 20% of patients with CRC and recurrence happens in about 40% of patients even with systemic treatment (142). Consequently, there has been growing interest in studying the interaction between CRC cells and other stromal cells in the TME.

Crosstalk between ESCC cells and CAFs has been observed to play a significant role in the development and progression of ESCC (171). Even after undergoing extended radical resection, the 5-year OS rate of patients with clinically high-level ESCC remains poor (6). Studies have revealed that there are also mutual efficacy between ESCC cells and CAFs. Zhao et al (172), confirmed that the manifestation levels of SHH were importantly elevated in CAFs lysate, conditioned medium for culturing CAFs (CAF-CM) and CAFs-derived exosomes. Hedgehog signaling pathway has been found to be promoted in ESCC cells when treated with CAF-derived factors such as conditioned medium and exosomes. This stimulation of the Hedgehog pathway leads to increased cell growth and migration. Jin et al (173), discovered that the expression level of hsa-miR-3656 has been found to be higher in CAFs compared to NFs. Overexpression of miR-3656 in CRC cells also enhances cell proliferation, migration, and invasion. Exosomal miR-3656 can be transferred from CAFs to ESCC cells, where it activates the PI3K/AKT and Wnt/β-catenin signaling pathways by down-regulating ACAP2. This activation ultimately promotes ESCC growth and metastasis. Shi et al (174), confirmed that LINC01410 was enriched in CAFs and could be transferred to ESCC cells and increase their LINC01410 levels. This transfer of LINC01410 promoted ESCC metastasis and EMT by binding to miR-122-5p, resulting in increased PKM2 levels. Cui et al (175),. demonstrated that CAFs-derived exosomes might accelerate cell proliferation and regulate apoptosis by RIG-I/IFN-β signaling and influence the chemosensitivity of ESCC to DDP, and that targeted inhibition of the RIG-I/IFN-β signaling axis in the exosomes of CAFs maybe an underlying therapeutic tactics for ESCC. Furthermore, the presence of CAFs and micro-lymphatic vessels in ESCC was associated with a poor prognosis (176). Exosomes derived from CAFs promote invasion, proliferation, migration, and tube formation of tumor-associated lymphatic endothelial cells (TLECs), while miR-100-5p exerts the opposite effect. Mechanistically, miR-100-5p inhibited ESCC lymphatic metastasis by targeting the IGF1R/PI3K/AKT regulatory axis and suppressing lymphangiogenesis.

In addition, ESCC cell-derived exosomes also regulated the conversion of NFs to CAFs. Interestingly, ESCC cell-derived exosomes have also been found to play a role in the conversion of NFs into CAFs (177). The transfer of lncRNA POU3F3 from ESCC cells to NFs leads to their transformation into CAFs. These activated fibroblasts further enhance ESCC cell proliferation and resistance to cisplatin by secreting interleukin 6 (IL-6). Moreover, high levels of plasma exosomal lncRNA POU3F3 are associated with poor treatment response and survival outcomes in ESCC patients. While it has been established that CAFs and ESCC cells communicate and regulate tumor development, progression, and drug resistance through exosomes, the specific mechanisms involved require further exploration and research due to limited current understanding.

Crosstalk between GC cells and CAFs has been a focus of research due to the challenges in GC occurrence, progression, and treatment (178, 179). The median survival of patients with metastatic or unresectable GC is still less than two years (180). Finding the difficulties in GC occurrence and progression as well as clinical treatment has been the direction of research in the last few years. Studies have shown that CAFs could regulate GC progression through exosomes ( Table 3 ). Miki et al (188), certified that exosomes carrying CD9 were taken up by GC cells, promoting their migration and invasion. Patients with CD9-positive GC were associated with worse clinical outcomes. In vitro cellular assays showed that exosomes from CAFs promoted the migration and invasion of OCUM-12 and NUGC-3 cells, which was regulated by anti-CD9 antibody or CD9-siRNA. In addition, CD9-positive GC patients were significantly associated with smooth muscle-type GC, lymph node metastasis and venous invasion, and had a meaningfully lower 5-year survival rate. Xu et al (181), confirmed that MMP11 was enriched in CAFs and could be separated to GC cells thereby promoting the facilitation of GC cell migration. Besides, MMP11 was significantly overexpression in exosomes refined from both plasma and tumor tissues of GC patients and was strongly relevant to poor prognosis of patients. The consequence of mechanistic experiments revealed that exosomal miR-139 could regulate GC growth and metastasis by decreasing the expression of MMP11. Zhang et al (182), demonstrated that exosomal miR-522 was mainly separated from CAFs of GCs, and USP7 could mediate the entry of miR-522 into exosomes by deubiquitinating and stabilizing hnRNPA1. In addition, they identified that DDP and paclitaxel could accelerate miR-522 secretion from CAFs by activating the USP7/hnRNPA1 axis and reduce the gross of lipid-ROS in GC cells by restraining ALOX15, which finally reduced chemosensitivity in vitro and in vivo. Overall, the exosomal miR-522 secreted by CAFs could inhibit iron apoptosis in GC cells by targeting ALOX15 and blocking lipid-ROS accumulation. Shi et al (183), discovered that CAFs could deliver functional circ_0088300 to GC cells via exosomes and promote cell migration, proliferation and invasive capacity. Mechanistically, KHDRBS3 could transport circ_0088300 packaging into exosomes, and circ_0088300 could act as a sponge to target miR-1305 and inhibit the JAK/STAT signaling pathway to accelerate the proliferation, migration and intrusion of GC cells. Wang et al (184), identified that miR-199a-5p could be transferred into GC cells via CAFs-derived EVs and promoted the malignant phenotypes and gastric tumorigenesis in vitro and in vivo. Mechanistically, miR-199a-5p could increase the phosphorylation level of AKT1 by down-regulating FKBP5 thereby promoting the lethal phenotype of AGS cells through mTORC1. Qu et al (185), demonstrated that DACT3-AS1 expression was reduced in GC and associated with low prognosis in GC patients, and the results of in vitro and in vivo experiments revealed that DACT3-AS1 inhibited cell proliferation, migration and invasion by targeting the miR-181a-5p/sirtuin 1 (SIRT1) axis. In addition, DACT3-AS1 was mainly delivered from CAFs to GC cells via exosomes and could attenuate the growth of xenograft tumors. Furthermore, DACT3-AS1 could sensitize cancer cells to oxaliplatin via SIRT1-mediated iron oxidation.

In addition, GC cells also function to regulate CAFs ( Table 3 ). Wang et al (186), discovered that miR-27a enticed programming coding of fibroblasts into CAFs. Besides, they demonstrated that miR-27a was enriched in the exosomes of GC cells and could promote the proliferation and motility of GC in vitro and in vivo by targeting CSRP2. Xia et al (187), confirmed that the expression of LINC00691 in serum exosomes of GC patients was higher above the fine subjects and patients with benign gastric disease and correlated with the clinicopathology of GC patients. Treatment of NFs with GC exosomes increased their LINC00691 expression levels and enabled them to acquire the properties of CAFs. Mechanistic experiments date revealed that LINC00691 could inhibit the JAK2/STAT3 signaling pathway to convert NFs into CAFs. The value of exosomes in the clinical diagnosis, treatment, and prediction of recurrence and metastasis of GC needs further confirmation to truly translate into clinical practice.

PC is a highly lethal GI tumor with an adding incidence, clinically characterized by difficulty in early diagnosis, poor rate of surgical resection, high rate of postoperative recurrence, and low prognosis, with the global five-year survival rate of patients being only about 10% (189, 190). Research finding have revealed that CAFs could be separated to PC cells via exosomes to inhibit cell pullulation and metastasis ( Table 4 ). Zhao et al (196), demonstrated that miR-320a expression was significantly increased in exosomes derived from CAFs compared to those from NFs. These miR-320a-enriched exosomes could transfer miR-320a to macrophages and promote M2 polarization, which, in turn, accelerated PC cell proliferation and invasion. Mechanistic experiments revealed that miR-320a regulated the PTEN/PI3Kc signaling pathway to promote the lethal progression of PC. Zhou et al (191), demonstrated that treatment of PC cells with CAF exosomes increased the expression of intracellular miR-421, leading to enhanced cell migration, proliferation, and invasion. Mechanistic experiments revealed that miR-421 targeted SIRT3 and regulated H3K9Ac, resulting in increased expression of HIF-1α. These findings suggested that miR-421 could be transferred from CAFs to PC cells via exosomes, promoting the development of PC. Raghavan et al (197), confirmed that secreted EVs derived from a specific subtype of CAFs called NetG1⁺ CAFs were found to stimulate Akt-mediated survival in nutrient-poor PDAC cells, protecting them from apoptosis. The expression of NetG1 in CAFs was identified as necessary for the pro-survival properties of EVs. Guo et al (192), demonstrated that miR-125b-5p expression was significantly boosted in PC cell lines and tissues. Besides, miR-125b-5p was enriched in the exosomes of CAFs and could be separate to PDAC cells to accelerate PDAC growth, invasion and metastasis by targeting APCs.

In addition, CAFs-derived exosomes have a vital regulatory function in the clinical management of PC ( Table 4 ). Richards et al (198), identified CAFs exposed to chemotherapy could promote PC cell survival and amplification. Besides, CAFs exposed to GEM had increased exosomal release and accelerated cell amplification and drug resistance through delivery to passaged snail in receptor epithelial cells. Kong et al (193), isolated CAFs from primary fibroblasts of PC patients and discovered that CAFs were congenitally oppose to GEM by assay. Treatment of CAFs or CAFs-exosomes using GEM could promote the effect of PC cell proliferation. The results of mechanistic experiments discovered that miR-106b could be directly moved from CAFs to PC cells via exosomes and promoted the opposition of PC cells to GEM by directly targeting TP53INP1. Fang et al (194), found that miR-21, miR-221, miR-181a, miR-222, and miR-92a expression were elevated in CAF exosomes secreted during gemcitabine treatment and could promote cell amplification and chemoresistance by inhibiting PTEN. Treatment with the inhibitor GW4869 could block the inhibition of PTEN in vivo and thus improve the chemotherapeutic efficacy of PDAC. Qi et al (195), confirmed that CAFs would be separated to PDAC cells via secreted exosomes and accelerate chemoresistance after GEM treatment. Automatically, miR-3173-5p in CAF exosomes targets ACSL4 and suppresses iron mutations in PDAC. In conclusion, the exosomal miR-3173-5p/ACSL4 pathway might be a promising target for the treatment of GEM-resistant PC. Exosomes, as carriers and messengers of intercellular cross talk, play a regulatory role in the occurrence, development, drug resistance and tumor immunity of PC. Although the underlying mechanisms are not yet fully understood, the abundant biological information contained in exosomes provides opportunities for uncovering the pathogenesis of PC and developing targeted therapeutic interventions.