MSCs/EVs-based therapy targeting DAD: research progress and future perspectives from ARDS and COVID-19 to RP-ILD Authors: Xiangzhen Meng, Luo Wang, Xinlun Tian Published: 2026 ABSTRACT AbstractDiffuse alveolar damage (DAD) is the core pathological process of acute lung injury, characterized by dual injury to alveolar epithelial and capillary endothelial cells, initiation of an inflammatory storm, and subsequent fibrotic remodeling, with persistently high clinical mortality. This pathological change is not restricted to a single disease but widely exists in critical pulmonary disorders such as acute respiratory distress syndrome (ARDS), severe COVID-19, and rapidly progressive interstitial lung disease (RP-ILD), acting as a key driver of disease progression. In recent years, mesenchymal stromal cells (MSCs) and their derived extracellular vesicles (EVs) have become research hotspots for their superior immunomodulatory, anti-apoptotic, tissue-repair, and anti-fibrotic properties. Based on the DAD pathological framework, this article systematically correlates epithelial barrier injury, inflammatory cascade amplification and pulmonary fibrotic remodeling with the therapeutic mechanisms of MSCs and their derived EVs in ARDS, COVID-19 and RP-ILD. It aims to provide a solid basis for the clinical translation of MSCs- and EVs-based therapies and promote their development into a mechanism-driven therapeutic paradigm targeting DAD, with great clinical value and academic innovation.Keywords: Mesenchymal stromal(stem) cell, Extracellular vesicles, Diffuse alveolar damage, Acute respiratory distress syndrome, Coronavirus disease 2019, Rapidly progressive interstitial lung disease, Immunomodulation, Tissue repair Introduction Diffuse alveolar damage (DAD) is a hallmark pathological lesion of acute lung injury (ALI) [1]. It is characterized by impaired alveolar-capillary barrier integrity, dysregulated immune inflammation, and aberrant tissue repair [2], posing major diagnostic and therapeutic challenge and making it a pivotal research focus in critical respiratory medicine. DAD is the typical pathological substrate of acute respiratory distress syndrome (ARDS) [3], the pathogenic driver of fulminant lung injury in severe coronavirus disease 2019 (COVID-19) pneumonia [4], and the underlying pathological basis for acute and rapid lung tissue destruction in rapidly progressive interstitial lung disease (RP-ILD) [5]. It directly induces progressive hypoxemia and respiratory failure [6], thereby representing a key determinant of high mortality in these diseases. The international LUNG-SAFE study involved 459 intensive care units (ICUs) from 50 countries. It reported an overall in‑hospital mortality of 40% in patients with Berlin‑defined ARDS, with respective rates of 34.9%, 40.0%, and 46.1% for mild, moderate, severe ARDS [7]. A multicenter prospective cohort study (COVID-ICU) conducted in 138 hospitals across France, Belgium, and Switzerland, involving 4244 critically ill adults with COVID-19, reported an overall 90-day mortality of 31%. Stratified by ARDS severity, 90-day mortality reached 30%, 34%, and 50% for mild, moderate, and severe disease, respectively [8]. Furthermore, a subset of survivors continues to suffer from long COVID-related complications following hospital discharge [9]. In patients with RP-ILD, a European myositis cohort study demonstrated that 40%–60% required ICU admission, with an in-hospital mortality of 45%–51% [10]. Another cohort study of 272 patients with anti-melanoma differentiation-associated gene 5‑positive (MDA5+) dermatomyositis showed a 6-month all-cause mortality rate as high as 47.8% among those complicated with RP-ILD [11]. Current therapies for DAD-related diseases mainly include anti-infective treatment, glucocorticoids, immunosuppressants, and mechanical ventilation [2, 12, 13]. Although these strategies improve prognosis to some extent, they fail to fundamentally repair injured lung tissue or modulate dysregulated immune-inflammatory responses. Moreover, they are associated with substantial adverse effects, higher risk of secondary infection, and poor long-term efficacy. Thus, there is an urgent need to explore novel, safe, effective therapeutic strategies that precisely target pathological mechanisms of DAD. Mesenchymal stromal/stem cells (MSCs), derived from the mesoderm, have self-renewal and multi-lineage differentiation capacities, and are widely distributed in bone marrow, adipose tissue, umbilical cord, and placenta [14, 15]. Their biological functions (immune regulation, anti-inflammation, tissue repair, anti-fibrosis) are mediated by direct cellular engraftment/differentiation and paracrine effects. Extracellular vesicles (EVs) are nanoscale membrane-bound vesicles secreted by MSCs. By encapsulating and transferring bioactive cargoes (e.g., proteins, nucleic acids), EVs mediate the paracrine effects and biological functions of MSCs [16], rendering them a prominent research focus. Focusing on the pathophysiological mechanisms underlying DAD, this review systematically summarizes the latest advances of MSCs and EVs in ALI/ARDS, COVID-19, and RP-ILD, including preclinical mechanistic studies and clinical trial data. It analyzes commonalities and disparities in MSCs/EVs-mediated interventions for DAD, and discusses current challenges and limitations in clinical application. Future research directions and clinical translation potential are proposed, aiming to provide a theoretical basis for precise treatment of DAD-related diseases. Pathological features and mechanisms of DAD Numerous clinical autopsy and biopsy studies have confirmed that DAD is the common histopathological basis for ARDS, COVID-19 pneumonia, and RP-ILD [17–19]. An open-lung biopsy study showed that more than 56% of ARDS patients had DAD pathology, with an in-hospital mortality rate of 71.9%, significantly higher than that of non-DAD patients (45.5%) [20]. Autopsies of patients with severe COVID-19 pneumonia revealed predominant DAD-type lung injury [21], morphologically consistent with DAD from other etiologies [22]. Similarly, a study on fatal RP-ILD related to juvenile dermatomyositis found that 7 of 10 patients died of disease progression within 1–4 months after diagnosis, and DAD was confirmed by autopsy in all cases [23]. The pathological process of DAD is time-dependent, encompassing acute exudative, proliferative, and chronic fibrotic phases. Its core pathogenesis involves alveolar-capillary barrier impairment, excessive inflammatory response, and abnormal tissue repair, as shown in Fig. 1. Pathological features and mechanisms of DAD. In normal alveoli, the alveolar-capillary barrier remains intact, with type I/II alveolar epithelial cells, alveolar macrophages, and fibroblasts maintaining tissue homeostasis. An intact alveolar-capillary barrier, composed of type I/II alveolar epithelial cells, resident macrophages, and quiescent fibroblasts, enables efficient O₂/CO₂ exchange and maintains tissue homeostasis. In DAD, barrier disruption causes endothelial damage and protein-rich fluid leakage, activating the NLRP3 inflammasome and recruiting neutrophils that release NETs. Macrophages polarize to the proinflammatory M1 phenotype, while platelet aggregation and T-cell infiltration amplify inflammation. Fibroblast-to-myofibroblast differentiation drives excessive ECM deposition, collectively impairing gas exchange and initiating fibrotic remodeling. This figure was created with BioRender.com Alveolar-capillary barrier impairment The alveolar-capillary barrier serves as a core structure for gas exchange, comprising thin-layer alveolar epithelial cells, basement membrane, and capillary endothelial cells. Specifically, the alveolar epithelium is composed of tightly aligned flattened type I alveolar epithelial cells (AEC I) and scattered type II alveolar epithelial cells (AEC II) [24, 25]. The capillary endothelium forms a single-layer structure covered by a basement membrane, which acts as a physical barrier to separate circulating blood cells and plasma from the pulmonary interstitium and airway. Injury to alveolar and vascular endothelium triggers a cascade of pathological changes. Cellular injury elevates barrier permeability, prompting protein-rich fluid to leak from blood vessels into the airway. This process induces pulmonary edema and hyaline membrane formation along the inner alveolar walls. Consequently, ventilation-perfusion mismatch and intrapulmonary shunting occur, ultimately resulting in severe impairment of pulmonary gas exchange [26]. Impairment of type II alveolar epithelial cells may compromise the synthesis of pulmonary surfactant. Meanwhile, alveolar edema concurrently inactivates pulmonary surfactant. These changes collectively result in atelectasis, end-expiratory alveolar collapse, and reduced lung compliance [27]. Additionally, pulmonary vascular endothelial injury activates procoagulant pathways, thereby initiating pulmonary microvascular thrombosis, increasing dead space ventilation, and further exacerbating gas exchange dysfunction [17]. Desquamation of alveolar epithelial and endothelial cells characterizes the early exudative phase of DAD. Excessive inflammatory response Pulmonary immunity relies primarily on innate and adaptive immunity [28, 29]. Necrosis of lung epithelial cells releases damage-associated molecular patterns (DAMPs), which trigger innate immune responses and activate M1 alveolar macrophages. Activated macrophages secrete proinflammatory cytokines (IL-6, TNF-α, IL-1β) and neutrophil chemokines (IL-8), which recruit neutrophils and initiate adaptive immunity. Neutrophils release neutrophil extracellular traps (NETs) to activate the NLRP3 inflammasome, amplifying the inflammatory response. Their released inflammatory mediators (elastase, prostaglandins, defensins, reactive oxygen species) induce apoptosis and necrosis of alveolar epithelial cells [30]. Sustained epithelial injury re-release DAMPs, forming a “damage-inflammation” vicious cycle, a core link in the initiation and progression of DAD [31]. Immune cells such as macrophages and lymphocytes gradually accumulate with DAD progression [32, 33]. Tissue repair and fibrosis Effective lung repair after injury relies on the clearance of alveolar edema, inflammatory cells, and necrotic tissue, as well as the reconstruction of the alveolar epithelial barrier. Disordered repair processes ultimately lead to fibrosis and persistent pulmonary dysfunction [34]. M2 macrophages phagocytose necrotic tissue and inflammatory cells while secreting anti-inflammatory cytokines (IL-10, TGF-β) [35]. AEC II promote barrier repair through proliferation and differentiation into AEC I [36, 37]. Cytokines in the inflammatory microenvironment (TGF-β, PDGF) stimulate fibroblast proliferation [38, 39]. Uncontrolled inflammation and overactivated repair signals induce massive secretion of collagen and fibronectin by myofibroblasts [40]. Additionally, the imbalance between matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs) suppresses extracellular matrix degradation and causes excessive accumulation. This disrupts normal alveolar architecture and eventually contributes to irreversible pulmonary fibrosis. MSCs and EVs Basic characteristics of MSCs and EVs MSCs are a class of adult multipotent stromal cells of mesodermal origin, characterized by adherent growth, high expression of CD73/CD90/CD105, lack of CD45/CD34/HLA-DR expression, and trilineage differentiation potential into osteoblasts, chondrocytes, and adipocytes [41]. MSCs can be isolated from multiple tissues, including bone marrow (BM-MSCs), adipose tissue (AD-MSCs), umbilical cord (UC-MSCs), and placenta (PL-MSCs) [42]. Compared with other stem cell populations, MSCs possess distinct advantages: wide availability, relative ease of acquisition, robust in vitro expansion capacity [43], and low immunogenicity [44]. The immunomodulatory effects of MSCs are also mediated through the paracrine secretion of EVs [45]. EVs are a heterogeneous group of phospholipid bilayer-enclosed vesicles that mediate intercellular communication by delivering bioactive molecules such as proteins, RNA, and/or DNA. Endowed with most core properties and immunomodulatory functions of parental MSCs, EVs are non-oncogenic and non-mutagenic [46], making them a promising candidate for cell-free therapy. EVs are primarily categorized into three subgroups: exosomes (40–150 nm in diameter, originating from the inward budding of multivesicular bodies), micro vesicles (100–1000 nm in diameter, generated by outward budding and fission of the plasma membrane), and apoptotic bodies (up to 5000 nm in diameter, derived from direct budding of the cell membrane during late-stage apoptosis) [47]. The International Society for Extracellular Vesicles (ISEV) recommends using the term “EV” as a collective designation for micro vesicles and exosomes [41]. Owing to their nanoscale size, EVs can cross physiological barriers such as the pulmonary barrier and blood-brain barrier to exert biological effects. Compared with MSCs, EVs are more amenable to storage and scalable production [48]. Nevertheless, heterogeneity in EV formulation remains a critical barrier to reliable research interpretation and clinical translation. To standardize relevant assessments, the ISEV MISEV2023 guidelines clarify core requirements for EV characterization and reporting: quantify particle count, size distribution and polydispersity index via nanoparticle tracking analysis; determine total protein content and particle-to-protein ratio via the bicinchoninic acid assay. Detect at least two positive markers (e.g., CD9, CD63, CD81) and negative markers (e.g., Calnexin) to confirm EV origin and purity. Verify EV morphology and bilayer membrane structure by cryogenic transmission electron microscopy (TEM) or negative-staining TEM. Evaluate biological activity and mechanisms using in vitro cell and in vivo animal models. Report traceable information (cell culture conditions, extraction/purification parameters, batch-to-batch quality control, storage stability) to minimize heterogeneity‑associated bias and improve result reproducibility and comparability. Mechanisms of MSCs and EVs in lung protection against DAD The lung is an ideal target organ for MSC- and EV-based therapies. The “pulmonary first-pass effect”, which is a limitation when MSCs are utilized for treating other organs, is converted into an advantage in pulmonary interventions. Following intravenous infusion, MSCs are filtered by the pulmonary vasculature, with only a small proportion entering the systemic circulation [49]. Due to their relatively large size, MSCs are trapped in the branches of pulmonary arterioles, where they secrete EVs to exert paracrine effects [16]. Non-invasive administration routes, such as nebulized inhalation and intratracheal instillation, have further expanded therapeutic options [46]. Additionally, MSCs can sense signals (e.g., inflammatory cytokines and chemokines) in the microenvironment of injured tissues, enabling targeted homing and engraftment at lesion sites [50], which lays a foundation for localized therapeutic efficacy. The core mechanisms by which MSCs and their EVs exert protective effects against DAD primarily encompass three aspects, as elaborated below, as shown in Fig.2. Multi - dimensional therapeutic mechanisms of MSCs and their EVs in DAD. (1) Anti-inflammatory: Inhibition of TLR/NF-κB signaling, activation of the COX2/PGE2 axis to induce macrophage PD-L1 expression, and suppression of NLRP3 inflammasome assembly via miRNAs collectively suppress T-cell hyperactivation and promote M2 macrophage polarization. (2) Cell protection & barrier repair: Suppress apoptosis, activate Nrf2-mediated antioxidation, secrete trophic factors, transfer mitochondria, and upregulate tight junction/surfactant proteins to preserve alveolar-endothelial barrier integrity. (3) Antifibrotic: Downregulate TGF-β/PDGF, inhibit Wnt/β-catenin/ERK1/2 pathways, and block EMT via miRNAs, maintaining MMP/TIMP balance to reduce ECM deposition and delay DAD fibrosis. This figure was created with BioRender.com Regulation of inflammatory responses and suppression of inflammatory storms MSCs can directly downregulate key inflammatory signaling pathways, including Toll-like receptors (TLRs) [51] and nuclear factor-κB (NF-κB) [52, 53]. Specifically, they inhibit the phosphorylation of inhibitor of κBα and subsequent activation of NF-κB. This blocks the inflammatory cascade at the upstream level. Through the cyclooxygenase-2/prostaglandin E2 (COX2/PGE2) axis [54, 55], MSCs induce macrophages to upregulate programmed death-ligand 1 (PD-L1). PD-L1 then binds to programmed death-1 (PD-1) on T cells to activate PD-1/PD-L1 signaling pathway. This interaction suppresses aberrant T cell activation and reduces the infiltration of Ly6C⁺CD8⁺ T cells into lung tissues [56]. Additionally, MSCs promote M2 macrophages polarization via the chemokine (C-C motif) ligand 2/chemokine (C-C motif) receptor 2 (CCL2/CCR2) signaling axis. This facilitation leads to the formation of regulatory CCR2⁺CD4⁺ T cells [57] and attenuated inflammatory responses. In contrast, EVs are enriched with functional microRNAs (miRNAs), such as miR-182-5p, miR-23a-3p, miR-377-3p, miR-193b-5p, and hsa-let-7i-5p [58–61]. These miRNAs directly target genes including Ikbkb, Usp5, and NLRP3. By targeting these genes, EVs can inhibit NF-κB pathway activation and NLRP3 inflammasome assembly [62]. These effects reduce neutrophil infiltration and neutrophil extracellular trap (NET) formation [63], and promote the transition of macrophages toward the M2 anti-inflammatory phenotype [56]. Furthermore, they downregulate the secretion of pro-inflammatory cytokines, such as TNF-α, IL-1β and IL-6. Modulation of cellular protection and barrier repair MSCs suppress the expression of apoptosis-related proteins (e.g., caspase-3/9) [59], thus mitigating excessive apoptosis of lung parenchymal cells. Meanwhile, MSCs activate the nuclear factor erythroid 2-related factor 2 (Nrf2) antioxidant pathway. This upregulates ferroptosis-regulatory proteins including glutathione peroxidase 4 (GPX4) and solute carrier family 7 member 11 (SLC7A11) [64], relieving iron overload and oxidative lung injury. Additionally, MSCs exert paracrine effects via secreting multiple growth factors, including hepatocyte growth factor (HGF) [65], keratinocyte growth factor (KGF) [66], prostaglandin E2 (PGE2) [54] and angiopoietin-1 (Ang-1) [67], which promote the proliferation and differentiation of AEC II into AEC I, thereby accelerating the reconstruction of the alveolar epithelial barrier. MSCs/EVs upregulate surfactant proteins (SP-A, SP-D) in AEC II, a key process for maintaining alveolar stability and preventing collapse. Through mitochondrial transfer, MSCs/EVs enhance the tricarboxylic acid (TCA) cycle and fatty acid synthesis. They stimulate endothelial cell proliferation and pro-angiogenic factors release, while reducing vascular endothelial cell apoptosis [68, 69]. Furthermore, MSCs and EVs upregulate the expression of endothelial junction proteins, such as cadherins and zonula occludens-1 (ZO-1) [70–72], lowering vascular permeability and ameliorating pulmonary edema. Intervention in the anti-fibrotic process MSCs and their EVs downregulate the expression of profibrotic factors (e.g., TGF-β [73], PDGF [70]). This inhibits fibroblast proliferation and reduces the excessive extracellular matrix (ECM) deposition, including collagen accumulation. Moreover, EVs-enriched miRNAs (e.g., miR-20 [74] and miR-22 [59]) suppress epithelial-mesenchymal transition (EMT) and fibrosis-related signaling pathways [70]. Mechanistically, MSCs and EVs block hyperactivated Wnt/β‑catenin signaling to restrain fibroblast proliferation and activation [71]. They also suppress ERK1/2 pathway-mediated CCL2 expression [75], thereby limiting monocyte-macrophage migration recruitment to injured lung tissue. Collectively, these synergistic effects maintain the balance between MMPs and TIMPs, which ultimately delays the progression from DAD to chronic pulmonary fibrotic. Evidence for MSCs and EVs therapy in ARDS and severe COVID-19 pneumonia Preclinical studies To date, published studies encompassing both in vitro cellular assays and in vivo animal models have validated the therapeutic potential of MSCs and EVs from multiple sources, including bone marrow [76], umbilical cord [77], placenta [78], adipose tissue [79], and menstrual blood [80]. These agents exert protective effects against diverse lung injury triggers, such as lipopolysaccharide (LPS)-induced damage [81], sepsis [72], SARS-CoV-2 infection [82], paraquat poisoning [53], and burns combined with smoke inhalation [83]. Primary administration routes include local delivery and intravenous infusion. The therapeutic efficacy can be further enhanced through approaches like inflammatory preconditioning and genetic modification [84]. In vitro, MSCs reduce apoptosis in LPS-stimulated murine lung epithelial MLE-12 cells. MSCs also reverse EMT, characterized by upregulated E-cadherin and downregulated α-smooth muscle actin (α-SMA) as well as collagen Ⅰ/Ⅲ, and inhibit the NF-κB signaling pathway [70]. Adipose-derived stem cell-EVs (ASC-EVs) exert a dose-dependent attenuation of SARS-CoV-2-induced cytopathic effect in Vero E6 cells, accompanied by reduced plaque formation, decreased viral titers, and downregulated expression of viral genes (E, RdRP, and N) [82]. Small extracellular vesicles (sEVs) derived from Chorionic Basal Plate (CBP-MSCs), placental MSCs (PL-MSCs), and Wharton’s jelly MSCs (WJ-MSCs) can all enhance the viability of A549 and THP-1 cells, suppress apoptosis, and reduce the secretion of pro-inflammatory cytokines such as interleukin-6 (IL-6) [81]. Animal models have further corroborated the in vivo therapeutic efficacy. In LPS-induced ARDS mouse models, treatment with BM-MSCs, AD-MSCs, or UC-MSCs exerted therapeutic effects versus controls. These treatments significantly improved chest computed tomography findings, increased survival rates, attenuated epithelial cell apoptosis, and reduced lung wet weight. They also downregulated pro-inflammatory cytokine mRNA levels (TNF-α, IL-6) and upregulated anti-inflammatory cytokine IL-10 [85]. Treatment with MSC-conditioned medium in LPS‑induced ALI mice downregulated miR-193b-5p and elevated occludin mRNA and protein levels, thereby maintaining cellular barrier integrity [72]. In a porcine ARDS model, bedside infusion of fresh autologous BM-MSCs delayed disease onset, downregulated pro-inflammatory cytokines, attenuated pulmonary inflammatory infiltration and alveolar edema, and reduced 72-hour mortality (10% vs. 50% in allogeneic BM-MSC and control group) [83]. These findings support their clinical feasibility. Other studies in vitro and in vivo have shown that BM-MSC-conditioned medium (BM-MSC-CM) preconditioned with IL-6-rich serum from hyperinflammatory ARDS patients, as well as AD-MSCs-derived EVs preconditioned with inflammatory cytokines, can significantly enhance immunomodulatory and tissue repair capacities, reduce pulmonary permeability and promote alveolar epithelial barrier repair [86, 87]. These findings suggest that MSCs and EVs exert more potent effects on ARDS under hyperinflammatory conditions. Clinical trials Clinical trials have primarily focused on MSCs and EVs therapy for ARDS and severe COVID-19, encompassing Phase Ⅰ, Phase Ⅱ, and long-term follow-up studies. Consistently, these trials have verified the favorable safety and tolerability of MSCs and EVs, with some demonstrating significant clinical efficacy: improved oxygenation, shortened mechanical ventilation and ICU stay, reduced mortality, alleviated pulmonary hyperinflammation, and mitigated DAD-associated lung injury [88–92], as detailed in Table 1. Table 1A Summary of Published MSCs and EVs Therapy Studies for ARDS/COVID-19 Over the Past 5 YearsFist author(Year)Target disease(N)Trial TypeInterventionagentAdministrationrouteDosage and frequencyEfficacy findingsSafety findingsHashemian SR (2021) [95]COVID-19 (11)Single armUC-MSCs/PL-MSCsIV infusion200 × 10⁶ cellsevery other Day3 times totalSix patients survived.; Significant chest CT lesion resolution with fibrosis excluded; significant reductions in TNF-α, CRP, IL-8, IL-6 and IFN-γ.No SAEs reported within24–48 h of infusions or 60-daypost-hospitalization follow-up.Lanzoni G (2021) [96]COVID19-associated ARDS (24)RCT(12 vs. 12)UC-MSCsIV infusionMSCs (n = 12) 1 × 108 cells/kg2 dosesDay 0, 331-day patient survival, significantly improved; IL-6, TNF-α, PDGF and IFN-γ were significantly decreased in UC-MSC group.No SAEs were observed relatedto UC-MSC infusions within 6 or24 h.Monsel A (2022) [97]COVID-19-associated ARDS (45)RCT(15 vs. 30)UC-MSCsIV infusionMSCs group (n = 15)1 × 106/kgD1, D3 ± 1 and D5 ± 1No significant difference was observed in the change of PaO₂/FiO₂ from day 0 to day 7.No UC-MSC-related SAEsreported within 7 days.Bukreieva T (2023) [98]COVID19-associated ARDS (28)RCT(13 vs. 15)UC-MSCsIV infusionMSCs (n = 13),3 timesDosage unknownThe total CT score in the MSC group was 12-fold lower than the control group at week 48. IL-6, TNF-α, Neu count, and ESR were significantly decreased, while IL-10, Lym count, and SPD were markedly increased.No UC-MSC-related SAEs wereobserved after 3 infusions .Morteza Zarrabi (2023) [99]COVID-19-associated ARDS (43)RCT(11 vs. 8 vs. 24)PT-MSCs/EVsIV infusion/ nebulizationMSCs group (n = 11):2 doses of 100 × 10⁶ cells (IV, q48h);MSC + EVs group(n = 8): 1 dose of 100 × 10⁶cells (IV) + 200 × 10⁶cells EVs (nebulization q48h)IL-6, TNF-α, IFN-γ and CRP decreased in the MSC group, with a more marked reduction in the MSC + EV groupNo SAEs were observed linkedto intervention in 28 days.Lightner AL (2023) [76]ARDS (102)RCT(34 vs. 34 vs. 34)BM-MSC-EVsIV infusion15 ml IP group (n = 34)10 ml IP group (n = 34)(1 mL IP: 8 × 1010 EVs), 2 doses, Day 1,4No difference in 60-day mortality; In 18–65 years subgroup, 15 ml BM-MSC-EVs group: significant improvement in ventilation-free daysNo AEs or SAEs were observedin any cohort within the first72 h.Pochon Cécile (2023) [93]ARDS (30)RCT(15 vs. 15)WJ-MSCsIV infusionMSC:1 × 10⁶ cells/kg within 48 h, followed by 2 infusions of 0.5 × 106/kg over 5 days.No significant differences were observed in PaO₂/FiO₂ ratio, 90-day mortality rate, and levels of IL-10,IiFN-γ, TNF-α, IL-1α.No SAEs were observed within90 days.Sitbon A (2024) [100]COVID-19-associated ARDS (47)RCT(22 vs. 27)UC-MSCsIV infusionMSC(n = 22): 1 × 106, cells/kg3times, q48hNo differences were found in mortality, lung CT morphology at 6 and 12 months post-discharge.Long-term safety (12 months)of intravenous UC-MSCs wasconfirmed.Zamanian M.H. (2024) [101]COVID-19-associated ARDS(45)RCT(21 vs. 24)hPMSCs-EVsIV infusionhPMSC-EVs (n = 21): 1.5-2 × 109/kgtwo times, qdHPMSC-sEVs could significantly decrease the mortality ratio and the mean time to death.No AEs related to hPMSC‑sEVsinfusion were noted at 24‑hpost‑intervention.Yuan MQ (2025) [102]COVID-19 (101)RCT(66 vs. 35)UC-MSCsIV infusionMSC (n = 66): 4 × 107 cells3 doses, Day: 0, 3, 6Over 3 years, the MSC group exhibited superior lung CT normalization and better 6‑MWD outcomes.No new complications or tumorrisks were identified at 3‑yearfollow‑up.Tonelli R (2025) [103]COVID-19(17)RCT(MSC = 5, CTRL = 2, mCTRL = 10)AD-MSCs /UC-MSCsIV infusionMSC (n = 5)Dosage unknowntwo timesIL-6, IL-18, and IL-1RA in the MSC group were significantly lower than the control group.No AEs related to MSC infusionoccurred.ARDS: Acute Respiratory Distress Syndrome; UC-MSCs: Umbilical Cord-derived Mesenchymal stromal/Stem Cells; PL-MSCs: Placenta-derived Mesenchymal stromal/Stem Cells; PL-MSCs: perinatal tissue-derived Mesenchymal stromal/Stem Cells; WJ-MSCs: Wharton’s Jelly-derived Mesenchymal stromal/Stem Cells; BM-MSC-EVs: Bone Marrow-derived Mesenchymal stromal/Stem Cell-derived Extracellular Vesicles; AD-MSCs: Adipose -derived Mesenchymal stromal/Stem Cells; hPMSC-EVs: human placental mesenchymal stromal cell-derived small extracellular vesicles; IP: investigational product; Lym: lymphocyte; Neu: neutrophils; SPD: surfactant protein D; SF-36: Short Form 36 Health Survey Questionnaire; qd: once a day; q48h: every 48 h; AEs: adverse events ; SAEs: serious adverse events A Summary of Published MSCs and EVs Therapy Studies for ARDS/COVID-19 Over the Past 5 Years UC-MSCs /PL-MSCs 200 × 10⁶ cells every other Day 3 times total No SAEs reported within 24–48 h of infusions or 60-day post-hospitalization follow-up. RCT (12 vs. 12) MSCs (n = 12) 1 × 108 cells/kg 2 doses Day 0, 3 No SAEs were observed related to UC-MSC infusions within 6 or 24 h. RCT (15 vs. 30) MSCs group (n = 15) 1 × 106/kg D1, D3 ± 1 and D5 ± 1 No UC-MSC-related SAEs reported within 7 days. RCT (13 vs. 15) MSCs (n = 13), 3 times Dosage unknown No UC-MSC-related SAEs were observed after 3 infusions . RCT (11 vs. 8 vs. 24) MSCs group (n = 11): 2 doses of 100 × 10⁶ cells (IV, q48h); MSC + EVs group(n = 8): 1 dose of 100 × 10⁶cells (IV) + 200 × 10⁶cells EVs (nebulization q48h) No SAEs were observed linked to intervention in 28 days. RCT (34 vs. 34 vs. 34) 15 ml IP group (n = 34) 10 ml IP group (n = 34) (1 mL IP: 8 × 1010 EVs), 2 doses, Day 1,4 No AEs or SAEs were observed in any cohort within the first 72 h. RCT (15 vs. 15) No SAEs were observed within 90 days. RCT (22 vs. 27) MSC(n = 22): 1 × 106, cells/kg 3times, q48h Long-term safety (12 months) of intravenous UC-MSCs was confirmed. RCT (21 vs. 24) hPMSC-EVs (n = 21): 1.5-2 × 109/kg two times, qd No AEs related to hPMSC‑sEVs infusion were noted at 24‑h post‑intervention. RCT (66 vs. 35) MSC (n = 66): 4 × 107 cells 3 doses, Day: 0, 3, 6 No new complications or tumor risks were identified at 3‑year follow‑up. RCT (MSC = 5, CTRL = 2, mCTRL = 10) MSC (n = 5) Dosage unknown two times No AEs related to MSC infusion occurred. ARDS: Acute Respiratory Distress Syndrome; UC-MSCs: Umbilical Cord-derived Mesenchymal stromal/Stem Cells; PL-MSCs: Placenta-derived Mesenchymal stromal/Stem Cells; PL-MSCs: perinatal tissue-derived Mesenchymal stromal/Stem Cells; WJ-MSCs: Wharton’s Jelly-derived Mesenchymal stromal/Stem Cells; BM-MSC-EVs: Bone Marrow-derived Mesenchymal stromal/Stem Cell-derived Extracellular Vesicles; AD-MSCs: Adipose -derived Mesenchymal stromal/Stem Cells; hPMSC-EVs: human placental mesenchymal stromal cell-derived small extracellular vesicles; IP: investigational product; Lym: lymphocyte; Neu: neutrophils; SPD: surfactant protein D; SF-36: Short Form 36 Health Survey Questionnaire; qd: once a day; q48h: every 48 h; AEs: adverse events ; SAEs: serious adverse events However, current clinical trials still have limitations, and several studies failed to meet their primary endpoints [76, 93]. Pochon et al. evaluated the efficacy of WJ-MSCs in adult patients with COVID-19-related moderate-to-severe ARDS receiving invasive mechanical ventilation. While the primary endpoint (proportion of patients with PaO₂/FiO₂ > 200 mmHg on day 10) was not met, the treatment group exhibited trends toward shorter mechanical ventilation and ICU stay. Another study assessed BM MSC-EV therapy for severe COVID-19 patients complicated by moderate-to-severe ARDS. Although 60-day all-cause mortality showed no significant between-group difference in the primary endpoint, high-dose intervention significantly lowered 60-day mortality relative to placebo, with an obvious dose-effect relationship. Notably, COVID-19-associated ARDS exhibits phenotypic heterogeneity of hyperinflammation and hypoinflammation, which affects the efficacy of immunomodulatory therapy. In addition, factors such as viral strains, MSC source and preparation process, dosage and frequency of administration, severity of ARDS in patients, and differences in study endpoint settings all make direct comparisons between studies difficult. These variables require further exploration and clarification in subsequent research to facilitate the optimization and advancement of therapeutic protocols. Application prospect: MSCs/EVs in RP-ILD treatment Given that RP-ILD shares similar pathological alterations and biomarker changes with ARDS and COVID-19, the therapeutic application of MSCs and EVs is underpinned by cross-disease pathological and molecular mechanisms. Currently, clinical evidence supporting MSC-based therapy for RP-ILD is limited to case reports. In 2025, Zhou RL et al. published a case report on HUC-MSC therapy for anti-MDA5 antibody-positive DM complicated with RP-ILD. The report indicated that the patient had failed to respond to conventional treatments, including methylprednisolone (160 mg/day), cyclophosphamide (400 mg every two weeks), tacrolimus (3 mg/day), and mechanical ventilation, over a 45-day period. Subsequently, researchers added intravenous infusion of 1 × 10⁸ HUC-MSCs to the original treatment regimen, administered in four cycles with a 28-day interval between cycles. No adverse events were observed during the infusion period or follow-up. Following treatment, the patient’s clinical condition gradually improved: computed tomography scans revealed marked resolution of lesions, serum anti-MDA5 antibodies seroconverted to negative, and ferritin levels returned to normal [94]. Nevertheless, large-sample prospective studies are warranted to clarify the safety and efficacy of HUC-MSCs in the treatment of anti-MDA5 antibody-positive DM complicated with RP-ILD. In 2025, a Phase I clinical trial (NCT06919380) targeting anti-MDA5 antibody-positive RP-ILD was initiated in China. This single-center, single-arm study employs nebulized inhalation of human amniotic mesenchymal stem cell-derived exosomes (HAMSC-Exos) as the intervention and plans to enroll 10 patients. The final outcomes of this trial are highly anticipated. Challenges and future directions Currently, preclinical and clinical research on RP-ILD remains scarce. Nevertheless, based on previous studies focusing on the pathological basis of DAD, the established findings and methodologies can be extended to inform the design of RP-ILD research. Notably, significant heterogeneity exists in the design of existing clinical studies (Table 2). For instance, variations in MSC sources (bone marrow, umbilical cord, adipose tissue, etc.), donor individual differences, and cell culture passages can lead to inconsistencies in cytokine secretion and EV components. Additionally, unified quality control standards are lacking (e.g., EV particle size, purity, and bioactivity detection indicators). Existing studies are Phase Ⅰ, small-sample, single-arm, single-center trials, and substantial differences exist in dosage, frequency, and regimens across different studies, making it challenging to develop unified treatment guidelines. Table 2Global Ongoing Clinical Projects of MSCs/EVs for ARDS and RP-ILD in the Past Three YearsRegistration NumberdiseaseIntervention AgentAdministration RouteSample SizeTrial PhaseTrial TypecountryRegistration DateJPRN-jRCT1051240234ARDSUC-MSCs supernatantN/AN/AN/AN/AJapan2023ChiCTR2400091226ARDSUC-MSCs-ExosNI12ISingle-arm StudyChina2024ChiCTR2400081032ARDSUC-MSCsIV12I/IISingle-arm StudyChina2024ChiCTR2400088213ARDSMSCsIV80I/IIControlled TrialChina2024NCT06002841COVID-19MSCs-EVsIV15I/IIControlled TrialBrazil2024CTIS2024-518160-13-00ARDSMSC-CMIV7IN/ASweden2024NCT06505941ARDSHUC-MSCs9I/IIControlled TrialChina2024NCT05491681ARDSHUC-MSCsIV60I/IIControlled TrialChina2024NCT06919380RP-ILDHA-MSC-ExosNI10ⅠSingle-arm StudyChina2025IV: Intravenous; NI: Nebulization Inhalation; N/A: Not Available Global Ongoing Clinical Projects of MSCs/EVs for ARDS and RP-ILD in the Past Three Years IV: Intravenous; NI: Nebulization Inhalation; N/A: Not Available In the future, technically, it is necessary to optimize research design schemes, realize process standardization, and improve delivery efficiency through targeted modification and upgrading of administration methods. Clinically, future research should focus on precision treatment, screen responsive patient populations, and explore combined treatment strategies. Large-scale with long-term follow-up trials are required to validate long-term safety. Meanwhile, identifying key bioactive components of EVs will facilitate the development of safer cell-free therapies. Conclusion As a common core pathological target in ARDS, COVID-19, and RP-ILD, DAD provides a mechanistic foundation for the cross-disease application of MSCs and EVs. Supported by robust preclinical and clinical evidence in ARDS and COVID-19, MSCs/EVs exhibit considerable therapeutic potential by regulating inflammation, repairing epithelial‑endothelial barrier, and inhibiting fibrotic -key pathological processes underlying DAD. Extending this therapeutic strategy to RP-ILD has the potential to address the unmet clinical need for this condition. Moving forward, the integration of multidisciplinary technologies, optimization of formulations and delivery strategies, and the conduct of high-quality clinical trials with precise patient stratification are anticipated to establish MSCs/EVs as a novel cross-disease therapeutic modality targeting DAD. Acknowledgements Not available. Abbreviations Diffuse alveolar damage Acute lung injury Acute respiratory distress syndrome Coronavirus disease 2019 Rapidly progressive interstitial lung Melanoma differentiation-associated gene 5 Mesenchymal stromal/stem cells Extracellular vesicles Bone marrow-derived mesenchymal stromal/stem cell Adipose-derived mesenchymal stromal/stem cells Umbilical cord mesenchymal stromal/stem cells Human umbilical cord mesenchymal stromal/stem cells Placenta-derived mesenchymal stromal/stem cells Wharton’s jelly-derived mesenchymal stromal/stem cells Perinatal tissue-derived Mesenchymal stromal/stem Cells Human placental mesenchymal stromal/stem cell-derived small extracellular vesicles Bone marrow mesenchymal stromal/stem cell-conditioned medium Type I alveolar epithelial cells Type II alveolar epithelial cells Transforming growth factor-beta Tumor necrosis factor-α Dermatomyositis Interferon-gamma Interleukin-6 Interleukin-10 C-reactive protein Nuclear factor-kappa B Lipopolysaccharide Epithelial-mesenchymal transition Neutrophil extracellular traps Toll-like receptors Extracellular matrix NOD-like receptor pyrin domain-containing 3 Matrix metalloproteinases Tissue inhibitors of metalloproteinases Surfactant protein D Platelet-derived growth factor Intravenous Adverse events Serious adverse events Transmission electron microscopy Author contributions Xiangzhen Meng and Luo Wang conceptualized the manuscript. Xiangzhen Meng wrote it and prepared all the figures, and Xinlun Tian edited it. All authors reviewed and approved the final version of the manuscript. Funding This study was supported by the National High Level Hospital Clinical Research Funding (Grant No. 2026-PUMCH-A-027). 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. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement No datasets were generated or analysed during the current study.