Lung organoids for preclinical evaluation of stem cell therapies: Opportunities, evidence, and translational challenges

Abstract

Pulmonary diseases impose a substantial global burden and remain major causes of morbidity and mortality. Despite advances in disease-modifying therapies, effective regenerative treatments remain limited for several chronic lung diseases, including idiopathic pulmonary fibrosis and chronic obstructive pulmonary disease. Stem cell-based therapies have shown promise in preclinical studies and early-phase clinical trials for lung tissue repair; however, substantial obstacles remain before clinical translation, including incomplete understanding of therapeutic mechanisms, safety concerns regarding tumorigenicity and immunogenicity, and a lack of standardized evaluation platforms. Lung organoids are three-dimensional, in vitro tissue constructs derived from pluripotent or adult stem cells that may bridge basic stem cell research and clinical application. This review introduces the sources and isolation methods of stem cells for lung organoid generation, describes preparation processes, and evaluates the utility of lung organoids in building pulmonary disease models. Importantly, we distinguish between current validated applications and proposed future capabilities. Organoid-based evidence offers new directions for investigating pathological mechanisms in stem cell treatment and suggests the exploratory potential of organoids in drug screening and early toxicity testing. However, direct evidence for the validation of stem cell therapies using human lung organoids is limited; most supporting data come from small-scale animal studies or proof-of-concept experiments. In addition, much of the evidence remains preclinical and model-dependent, with current organoid systems lacking vascularization, innervation, systemic circulation, and a complete immune context. Finally, we summarize the validated applications of lung organoids in stem cell therapy assessment (mechanistic elucidation and donor-specific screening), identify the remaining experimental gaps (clinical predictive validity, long-term safety surrogates), and propose the evidence needed to address these limitations.

Key Words: Stem cell therapy; Lung organoids; Pulmonary fibrosis; Chronic obstructive pulmonary disease; Extracellular vesicles; Disease modeling

Core Tip: Lung organoids are emerging human-relevant ex vivo models that may complement conventional preclinical systems in studying lung disease mechanisms and evaluating selected responses to stem cell-based therapeutic products. Current evidence does not support their use as clinically validated predictors of transplantation outcomes or as independent tools for comprehensive assessment of local safety-related responses. Standardized protocols and prospective studies linking organoid responses with animal and clinical outcomes are needed to establish their translational value.

INTRODUCTION

Lung diseases impose a substantial public health burden worldwide, such as idiopathic pulmonary fibrosis (IPF), chronic obstructive pulmonary disease (COPD) and acute respiratory distress syndrome (ARDS). Current treatments cannot resolve damage to the lung tissue. COPD is one of the main causes of death worldwide, causing 3.7 million deaths globally in 2021[1]. IPF is a slow-progressing, incurable interstitial lung disease with a median survival historically estimated at 2.5-3.5 years from diagnosis[2]. ARDS still has a high rate of hospital death, 30%-45%[3]. Although advances have been made in anti-fibrotic drugs and biologics, as well as critical care treatments, they only delay the progression of the disease and do not repair and rebuild the structure and function of the lungs[4]. Multiple cell types have been explored as candidate sources for lung regenerative medicine.

Recently, stem cell-based therapies have been investigated as potential treatments for lung damage, with preliminary studies suggesting possible benefits in quality of life and pulmonary function[5]. Mesenchymal stromal cells (MSCs) derived from bone marrow, adipose tissue, or umbilical cord, as well as induced pluripotent stem cells (iPSCs) and endogenous adult lung stem cells - including alveolar type II (AT2) cells and airway basal cells - have all been investigated in preclinical and early-phase clinical studies for their capacity to suppress inflammation, promote alveolar regeneration, and improve tissue repair[6-8]. Recent research has shown that MSC-based therapies are relatively safe and have shown promising therapeutic effects for IPF, COPD, and ARDS. Regenerative medicine based on airway basal stem cells has also offered new treatment options for chronic respiratory diseases[9-11]. However, there are still many problems with the clinical application of stem cell therapy at present. The repair function of stem cells and their derivatives in the body is not fully understood, and further research is needed to verify their long-term safety, potential for tumorigenicity and immune responses. At the same time, there is no humanised evaluation system to assess the effects of various cell sources, different doses and routes of administration, as well as combinations of treatments reliably. Traditionally, two-dimensional (2D) cell cultures have failed to reproduce the complex 3D structure and intercellular interaction environment of the lung in vitro, and animal models are also limited by species differences and cannot accurately represent human clinical conditions[12,13].

These 3D lung organoids, which originate from adult or iPSCs, have been able to replicate the cellular arrangement, spatial structure and some physiological functions of alveolar and airway tissues in vitro[14-17]. Lung organoids have demonstrated superior capabilities in replicating the human lung microenvironment and are now used to investigate the effects, efficiencies, and safety of stem cell therapy in terms of repair mechanisms at the basic research level; they also offer new references for clinical applications[18]. Basic epithelial organoids have evolved into more advanced co-culture systems of lung organoids incorporating immune, endothelial and mesenchymal cells. Organ-on-a-chip systems and spatial omics may further increase the physiological and molecular resolution of lung organoid models, but their clinical predictive value has not been validated[19-25]. This narrative review reports on the new applications of lung organoids as auxiliary in vitro models for testing some stem cell-based therapy products, such as MSCs, MSC-derived extracellular vesicles (MSC-EVs), iPSCs-derived epithelial cells and gene-edited cells. We will examine the possible functions of these in disease-related phenotype evaluation and investigate their mechanisms and therapeutic effects. We also discuss the deficiencies of organoid-based assessment of local safety-related responses, particularly in terms of systemic immunogenicity, biodistribution, thromboembolic risk and long-term tumorigenicity.

METHODS OF LITERATURE IDENTIFICATION

This narrative review was informed by literature searches of PubMed/MEDLINE, Web of Science, and EMBASE conducted from January 1, 2010 to August 31, 2026. The search concepts combined terms related to lung organoids, alveolar organoids, airway organoids, stem cell therapy, mesenchymal stromal cells, extracellular vesicles, induced pluripotent stem cells, gene-edited cells, organ-on-a-chip systems, and pulmonary diseases, including idiopathic pulmonary fibrosis, chronic obstructive pulmonary disease, infection, and lung cancer. Representative search terms included “lung organoid”, “alveolar organoid”, “airway organoid”, “stem cell therapy”, “mesenchymal stromal cell”, “extracellular vesicle”, “induced pluripotent stem cell”, and “lung regeneration”. Only English-language full-text articles were formally included. We prioritized peer-reviewed original studies and authoritative reviews; classic studies were retained to establish foundational technologies and differentiation protocols, whereas recent studies (primarily from 2020 onward) were prioritized for evidence on stem cell therapy validation, immune co-culture systems, and organ-on-a-chip integration. Evidence from organoid studies, conventional cell culture, animal models, and clinical studies was identified separately and interpreted according to its level of directness. Conference abstracts and preprints were considered only when relevant peer-reviewed evidence was unavailable and were explicitly labeled as preliminary evidence. Because this was a narrative review, no formal meta-analysis or risk-of-bias assessment was performed; therefore, selection and citation bias cannot be excluded. Notably, in the final reference list of this review, no preprints or conference abstracts were included; all cited references are peer-reviewed articles, with the exception of two official regulatory documents (EMA guideline and FDA news release), which are clearly identified as such. Articles published ahead of print were included if they had been formally accepted for publication and assigned a DOI. In the main text, evidence from different experimental systems is identified by the terms used to describe each model (e.g., ‘murine organoid’, ‘human iPSC-derived organoid’, ‘mouse organoid-macrophage co-culture’). This distinction is further summarized in Table 1, where evidence levels are explicitly categorized as direct (organoid-based), animal model (in vivo), or theoretical (proof-of-principle).

Table 1 The lung organoid platform as feasibility evidence for preclinical stem cell validation.

Assessment dimension	Study type	Key findings	Evidence level	Limitations	Ref.
Immunomodulatory mechanism	Murine lung organoid-macrophage co-culture	Bone marrow-derived MSCs suppressed LPS-induced proinflammatory cytokines and NF-κB activation in alveolar and interstitial macrophages	Direct (murine organoid)	Not validated in human organoids; culture medium incompatibility limits generalizability	[18]
Long-term safety monitoring (tumorigenicity, genetic stability)	Theoretical framework	Genomic approaches (WGS, karyotyping, scRNA-seq) proposed for assessing oncogenic risk of candidate cells	Theoretical (proof-of-principle)	No prospective study	[65,81,82]

Evidence level definitions: Direct evidence - stem cell or extracellular vesicle therapies directly validated within a specified organotypic system; theoretical evidence - proposed applications for which no directly validated studies have been conducted. MSCs: Mesenchymal stromal cells; LPS: Lipopolysaccharide; NF-κB: Nuclear factor kappa B; WGS: Whole genome sequencing; scRNA-seq: Single-cell RNA sequencing.

DEVELOPMENT OF LUNG ORGANOIDS

The development of lung organoids relies on a suitable stem cell source, which is the foundational prerequisite for establishing this biological platform. As the stem or progenitor cells have the ability to self-reproduce and differentiate, they are essential for the construction of lung organoids, affecting the cellular structure, maturity, and application of these organoids. Recently, continuous improvements have been made in iPSCs-directed differentiation technology, biomaterials and 3D culture systems, and lung organoids have become an essential humanized model for studying the lung development and lung diseases and stem cell therapy[26] (Figure 1).

Figure 1 Stem cell sources for lung organoid generation and the role of mesenchymal stromal cells as therapeutic modulators. This figure presents a conceptual schematic distinguishing two distinct categories of cells relevant to lung organoid research. Left side: Major stem cell sources used for lung organoid generation - including adult stem cells, induced pluripotent stem cells, and embryonic stem cells. Right side: Mesenchymal stromal cells are not used as starting cells for epithelial lung organoid formation. This is a conceptual framework proposed by the authors, not an evidence-based algorithm. ASC: Adult stem cell; iPSC: Induced pluripotent stem cell; ESC: Embryonic stem cell; MSC: Mesenchymal stromal cell; EVs: Extracellular vesicles.

Origin of adult stem cells

AT2 cells, airway basal cells and progenitor cells for regeneration are all groups of stem and progenitor cells residing in adult lung tissue in the distal part of the lung. 3D matrices, such as Matrigel, can be used to form alveolar or airway organoids of these cells in vitro that retain the genetic background and disease features of the original donor. Patient-derived lung organoids can be used to study specific aspects of tissue repair and drug response in vitro, but they have not yet been validated as pre-transplantation biomarkers. Accordingly, lung organoids have been widely used for disease modeling, developmental studies, and the ex vivo assessment of selected regenerative responses[27,28]. Now, patient-derived organoids (PDOs) are often used to study pulmonary regeneration. Studies that have used HTII-280+ AT2 cells isolated from the lung tissue of IPF patients to build alveolar organoid models have shown that as the disease progresses, both the number of AT2 cells and their capacity for clone formation significantly decline; at the same time, organoid formation efficiency decreases, SFTPC expression is reduced, and abnormal epithelial remodeling occurs[29,30]. Thus, patient-derived lung organoids can retain the characteristics of the disease and are suitable for investigating endogenous lung stem cell regenerative capacity and individual variations in response to regenerative therapy. A proof-of-concept validation paradigm has been proposed to evaluate the repair capacity of PDOs by comparing their responses to stem cell therapy in healthy and diseased organoids. No research has yet prospectively linked these in vitro responses to clinical outcomes in the same patient group.

Origin of iPSCs

iPSCs can be obtained by reprogramming somatic cells (such as fibroblasts or peripheral blood mononuclear cells) to simulate the embryonic lung development process in vitro, and then gradually differentiate through stages, including the definitive endoderm and NKX2-1+ lung precursor cells, to form alveolar or airway organoids[14,31]. iPSC-derived organoids are not organoids of adult stem cells (ASCs), so they are easy to scale up and patient-specific, and thus highly reliable resources for personalised disease modelling and assessment of cell therapy products. With the recent progress of bioreactors and matrix-free culture technology, large-scale production of lung organoids from iPSCs has also became available. The team led by Budeus et al[32] has introduced a method for large-scale production of lung organoids based on iPSCs. The team developed an extracellular matrix-free culture protocol and, in a stirred bioreactor, effectively derived branched lung organoids from embryonic bodies. Improving the uniformity and reproducibility of organoids through large-scale standardized culture can provide a foundation for a standardized platform for evaluating future drugs and cell therapies[32]. Lung organoids that are derived from the same patient’s iPSCs can be combined with cell therapy products of that individual to explore the efficacy and mechanisms of treatment in a proof-of-concept setting[33].

Source of embryonic stem cells

Embryonic stem cells (ESCs) are an essential cell source for the early studies of lung organoids due to their diverse differentiation capabilities. Several research groups have used ESCs to promote the formation of lung precursor cells and alveolar-like structures, and thus have explored the regulatory mechanisms of various signaling pathways, including fibroblast growth factor (FGF), Wnt, bone morphogenic protein, and retinoic acid, in lung development using these models[14]. Although there are ethical and regulatory constraints on the use of ESCs in clinical practice, they are still required for research on pulmonary development and improvement of iPSCs differentiation methods.

MSCs and their exosomes are related to lung organoids

Although MSCs are not epithelial progenitors and do not serve as the conventional starting cells for lung organoid formation, they can regulate the growth, differentiation and repair of lung organoids through paracrine effects and extracellular vesicles (EVs)[34]. MSC-EVs have become a promising cell-free regenerative therapy due to their low immunogenicity, biocompatibility and engineering potential[35]. According to studies, MSC-EVs inhibit fibrosis by suppressing the transforming growth factor (TGF)-β/Smad signaling pathway and regulating the Wnt/β-catenin signaling pathway; therefore, it can help repair alveolar epithelial cells and reduce fibroblast activation[36]. MSC-EVs can also make macrophages switch to the M2 type, help increase regulatory T cells, and thus reduce excessive inflammation in the immune microenvironment after lung injury[37]. Lung organoids are employed to assess the effectiveness of MSC-EVs, and single-cell transcriptomics, spatialomics and high-content imaging are also used to conduct in-depth mechanistic research for the application of these organoids in medicine.

Horizontal comparison and safety evaluation of different stem cell therapies

Preclinical validation based on a lung organoid platform requires selecting an appropriate stem cell type according to the therapeutic objective. ASCs, iPSCs, MSCs, and gene-edited cells exhibit significant differences in their therapeutic mechanisms, applicable disease types, and clinical translation potential, reflecting variations in cellular plasticity, immunogenicity, and genetic stability. ASCs retain tissue-specific regional characteristics and epigenetic traits, demonstrating direct physiological relevance for alveolar regeneration and airway epithelial reconstruction[11]. However, its therapeutic application is limited by the accessibility of donor tissues; patients with advanced fibrosis or severe COPD often struggle to obtain sufficient quantities of viable cells, while ASCs cannot differentiate into non-epithelial lineages, thereby hindering their ability to exert therapeutic effects through systemic immunomodulation[38,39]. Additionally, patient-derived ASCs may carry disease-associated genetic defects or senescent characteristics, which can impair their regenerative potential[40]. iPSCs overcome donor limitations through reprogramming, possess multipotent differentiation potential, and can differentiate into alveolar epithelial cells, airway basal cells, or even mesenchymal cells, making them suitable for autologous transplantation following gene correction[41]. However, as therapeutic products, iPSCs may generate copy number variations and single nucleotide polymorphisms during the reprogramming process, carry a tumorigenic risk, exhibit a prolonged differentiation cycle, and demonstrate limited cellular maturity and discrepancies compared with adult lung function[42]. MSCs represent the most well-established strategy in current clinical trials for pulmonary diseases; their core therapeutic mechanism does not involve cell replacement, but rather relies on paracrine-mediated remodeling of the immune microenvironment[36,43]. This transient yet sustained therapeutic profile implies that clinical efficacy depends on the temporal dynamics of the secretome rather than long-term colonization by the cells themselves. The immunosuppressive effects of systemic MSC infusion may increase the risk of infection, necessitating cautious evaluation when applied in cases of infectious lung disease. It is noteworthy that gene-edited cells have demonstrated the potential to correct pathogenic mutations in both in vitro and preclinical models, offering a theoretical basis for functional cure. Therefore, lung organoid platforms should not be utilized merely as screening platforms for various stem cell types; instead, they should support the development of a differentiated assessment of local safety-related responses framework tailored to specific therapeutic strategies.

Comparison of characteristics among lung organoids derived from different sources

Lung organoids derived from ASCs, iPSCs, and ESCs exhibit significant differences in terms of maturity, physiological fidelity, and technical feasibility (Table 2); understanding these differences is crucial for selecting the appropriate experimental platforms. ASC-derived lung organoids retain the epigenetic information and tissue-specific characteristics of the donor to replicate disease-related pathological features effectively; for example, airway organoids derived from COPD patients have maintained pathological features such as goblet cell hyperplasia and ciliary dysfunction[15,44,45]. ASC-derived lung organoids advantages lie in its short cultivation cycle, high maturity level, and the fact that it does not require the complex processes of reprogramming or directed differentiation[15]. However, ASC-derived organoids exhibit limited cellular diversity and difficulty in spontaneously forming endothelial or mesenchymal components, thereby restricting their application in studies of multicellular interactions. In contrast, iPSCs-derived lung organoids are generated by mimicking the embryonic lung development process; they possess unlimited expansion potential and the highest potential for cellular diversity, enabling them to generate multi-lineage structures[14]. However, these organoids exhibit a prolonged differentiation cycle, and significant variability in maturity across different batches, and may contain undifferentiated precursor cells, thereby compromising the reliability of functional assessments[41]. ESC-derived lung organoids exhibit multi-directional differentiation potential comparable to that of iPSCs, serving as a vital tool for early-stage lung development research and having been successfully employed to investigate the regulatory mechanisms of key signaling pathways such as FGF and Wnt[46,47]. However, due to ethical constraints, ESCs cannot be obtained from patient-specific sources, limiting their application in clinical translation[48]. In summary, when validating stem cell therapies, the choice of organoid source should be based on validation requirements rather than therapeutic needs: ASC-derived sources are suitable for patient-specific rapid validation; iPSCs-derived sources are appropriate for large-scale standardized screening and research into gene editing mechanisms; whereas ESC-derived sources are primarily used for basic developmental research.

Table 2 Comparison of characteristics among lung organoids derived from different stem cell sources.

	ASCs	iPSCs	ESCs
Main sources	Autologous lung tissue from patients (AT2 cells, airway basal cells, etc.)	Somatic cell reprogramming (e.g., from fibroblasts, peripheral blood mononuclear cells, etc.)	Inner cell mass of blastocysts (allogeneic)
Applicable to pulmonary disease scenarios	Airway epithelial reconstruction; local tissue repair	Modeling of developmental disorders; autologous transplantation following gene correction	Modeling of early lung development (e.g., congenital anomalies); investigation of key signaling pathways (FGF, Wnt)
Cost	Due to limitations in biopsy tissue accessibility and amplification capacity, the cost of personalized sample acquisition is high	Reprogramming and directed differentiation techniques are complex, with high costs associated with reagents and quality control materials	High, associated with acquisition, specialized maintenance, and ethical compliance procedures
Advantage	Short cultivation cycle, high maturity; preserves tissue specificity and epigenetic characteristics	Nearly unlimited amplification capacity; multi-lineage differentiation potential	Nearly unlimited amplification capacity; multi-lineage differentiation potential; serves as a vital tool for early-stage lung development research
Limitations	Cannot differentiate into non-epithelial lineages; may harbor disease-associated genetic alterations	Reprogramming carries a tumorigenic risk; there are significant inter-batch variations in cell maturity	Ethical constraints; immune rejection risk in allogeneic settings; teratoma formation potential
Focus of organoid validation	Assessment of local epithelial regenerative potential; functional validation prior to autologous transplantation	Differentiation efficiency and maturity monitoring; long-term genomic stability tracking	Monitoring of differentiation efficiency and maturity; evaluation of teratoma risk; basic developmental research

The comparative assessment in this table is a qualitative synthesis of published literature and expert interpretation, rather than a validated consensus ranking or meta-analysis conclusion. ASC: Adult stem cell; iPSC: Induced pluripotent stem cell; ESC: Embryonic stem cell; AT2: Alveolar type 2; FGF: Fibroblast growth factor.

LUNG ORGANOIDS MODEL SYSTEMS USED TO ASSESS THE EFFECTS AND REASONS FOR STEM CELL THERAPY

Pulmonary diseases, such as IPF, COPD, acute lung injury and lung cancer, are serious problems around the world, but effective treatments to reverse the damage in these tissues have yet to be found[4,49,50]. At present, most drugs are used to reduce the severity of diseases by preventing further damage to the lungs, and the exact reasons for changes at the level of cells and molecules in various pulmonary diseases have not been identified yet[51,52]. Clinical trials on stem cell therapy for respiratory system diseases have now been conducted to some extent; however, the quality of these trials is relatively low at present, and although safety has been preliminarily confirmed, conclusive data on treatment efficacy (especially improvements in pulmonary function and prolonged benefits) have yet to be obtained from large-scale randomised controlled trials[53-55]. In recent years, with the development of humanised lung organoid systems, new paths for disease modelling have opened up. Lung organoids derived from pluripotent stem cells or patient-derived tissues have been used to emulate the pathological features of pulmonary fibrosis and COPD, such as epithelial cell senescence, fibroblast activation, and abnormal AT2 cell states[56,57]. These models are theoretically suitable for therapeutic intervention trials, but studies directly applying these models to evaluate stem cell or EV-based therapies are extremely rare. In models of cigarette smoke-induced pulmonary injury, some studies have used COPD organoid models to evaluate the regenerative effects of stem cell therapies that EVs derived from human umbilical cord mesenchymal stem cells were reported to restore the AT2/AT1 Lineage balance, reduce inflammatory infiltration, and inhibit collagen deposition in a murine organoid model (n = 3-5 mice per group). However, the magnitude of these effects and their reproducibility across different experimental conditions have not been systematically characterized[58]. In an lipopolysaccharide (LPS)-induced acute lung injury model, bone marrow-derived MSCs reduce damage to lung organoids by inhibiting the pro-inflammatory effect of macrophages in a mouse lung organoid-macrophage co-culture system[18]. At the same time, a proof-of-concept study using iPSC-derived lung and lung cancer organoids investigated the therapeutic effect of cisplatin-loaded EVs derived from iPSC-MSCs; however, cytotoxic effects induced by the drug-free EVs alone (i.e., the EV carrier without cisplatin) were observed in both healthy and cancerous organoids, and no selective targeting effect was demonstrated[33]. Notably, cisplatin-loaded EVs at the tested concentration failed to induce cytotoxicity in either organoid model, while empty EVs alone exhibited non-specific cytotoxic effects, highlighting challenges in achieving a therapeutic window with this approach[33]. Although the application of human organoids in the study of the pathogenesis of pulmonary fibrosis and drug screening has been expanding gradually, no research has yet explored the anti-fibrotic effects of MSCs or MSC-EVs on organoids derived from patients with IPF or on human alveolar organoids stimulated by TGF-β1; furthermore, no study has been conducted to prospectively examine the correlation between organoid-based stem cell responses and clinical outcomes in the same patient group (Table 3, Figure 2).

Figure 2 Strategies for lung organoid construction and disease modeling. This figure presents a proposed conceptual workflow for generating lung organoids from different stem cell sources and their application in modeling pulmonary diseases. 3D: Three-dimensional; AT2: Alveolar type 2; αSMA: Alpha-smooth muscle actin; COL1A1: Collagen type 1 alpha 1; COPD: Chronic obstructive pulmonary disease.

Table 3 Applications of lung organoids in evaluating stem cell therapy.

Disease type	Source/organoid type	Stem cell/EV therapy	Evaluation metrics	Key findings	Study design/sample size	Ref.
Pulmonary fibrosis	-	-	-	No published studies have been identified that directly evaluate MSC- or MSC-EV-based therapy in pulmonary fibrosis lung organoid models; current evidence is primarily derived from animal models and 2D culture experiments	-	-
COPD/emphysema	Mouse-derived/Lung epithelial progenitor cell-derived organoids	hUC-MSC-EVs; hUC-MSCs	Number/size of organoids; AT2/AT1 marker balance; collagen deposition; inflammatory infiltration	hUC-MSC-EVs reduce inflammatory infiltration and collagen deposition, restore the normal number and size of organoids, and rebalance the AT2/AT1 ratio	Murine organoids; n = 3-5 mice per group; single exposure model	[58]
Acute lung injury	Mouse-derived/Lung organoid-macrophage co-culture model	MSCs (bone marrow origin)	Macrophage pro-inflammatory function; organoid morphology	MSCs mitigate LPS-induced acute lung injury by inhibiting the pro-inflammatory function of macrophages	Murine organoid-macrophage co-culture; n = 3 independent experiments; 2 organoid donors + ≥ 3 MSC donors	[18]
Lung cancer	Human-derived/iPSCs-derived bronchial organoids (BLO); patient-derived LCOs	PSC-MSC-derived EVs (loaded with cisplatin)	LDH release; CCK8 metabolic activity; apoptosis-related genes (e.g., P53)	The empty EVs themselves exhibit cytotoxicity toward both LCO and BLO, suggesting that MSC-EVs may exert non-specific effects	Human iPSCs-derived organoids; n = 3-4 technical replicates; 2 independent differentiations	[33]

Except for the pulmonary fibrosis row (for which no published studies were identified), all other therapies involving stem cells or stem cell-derived extracellular vesicles listed in this table have been directly validated within lung organoid systems. EV: Extracellular vesicle; MSC: Mesenchymal stromal cell; 2D: Two-dimensional; COPD: Chronic obstructive pulmonary disease; hUC-MSC-EVs: Human umbilical cord mesenchymal stem cell-derived extracellular vesicles; hUC-MSCs: Human umbilical cord mesenchymal stem cell; AT2: Alveolar type 2; LPS: Lipopolysaccharide; iPSCs: Induced pluripotent stem cells; BLO: Branched lung organoid; LCO: Lung cancer organoid; PSC: Pluripotent stem cell; LDH: Lactate dehydrogenase.

Several structural limitations currently prevent direct verification of stem cell therapy outcomes in lung organoid systems. First, the current lung organoid system lacks a blood vessel network, functional immune cells, and physiological mechanical stress; therefore, it cannot replicate the safety end points required by regulations for the approval of cell therapy products and thus fails to demonstrate the systemic biodistribution, embolism risk, or long-term tumourigenicity of administered stem cells. Second, the development path of pulmonary regenerative medicine is driven by clinical demands. Stem cell-related phase I/II clinical trials for respiratory diseases entered a period of rapid expansion around 2014, and lung organoids only began their rapid development phase during the pandemic; therefore, clinical trials are often started before strict, humanised preclinical validation, creating a time lag[59,60]. Furthermore, due to the heterogeneity of stem cell products, standard organoid testing methods that can be widely applied across the entire field have not been established[61,62]. Crucially, the predictive value of organoid-based assays for the results of clinical stem cell therapy has not been fully verified, and no study has shown that the therapeutic effect observed in the lung organoid system can be replicated in patients treated with these organoids (Table 3, Figure 2).

LUNG ORGANOIDS AS PRECLINICAL PLATFORMS FOR SAFETY AND IMMUNOCOMPATIBILITY ASSESSMENT OF STEM CELL THERAPIES

Therapeutic methods for pulmonary diseases based on stem cells have not been widely applied in clinical practice due to unresolved safety problems, such as the risk of abnormal proliferation of transplanted cells, ectopic differentiation, genetic instability and tumourigenicity[42,43,63-65]. Lung organoids have the ability to be maintained for an extended period and have been proposed as a humanised research platform that can evaluate the effectiveness of stem cell-based products[66] (Table 1). Theoretically, extended co-culture of stem cells or their differentiated products with lung organoids can be used to observe continuous cell proliferation, lineage stability and changes in tissue structure. At the same time, an early abnormal phenomenon in the non-target differentiation process can be observed morphologically and analyzed by transcriptomics[65]. Genomic approaches are used to examine the genetic stability and tumourigenicity of the potential therapeutic cells by means of whole-genome sequencing, chromosome karyotyping and single-cell sequencing[67,68]. However, it should be pointed out that these applications are still in the stage of theory or principle verification at present; no study has explicitly confirmed whether the safety findings observed in the organoid co-culture system can predict tumourigenicity or long-term genetic stability in vivo for patients. In addition to safety, immunocompatibility is another critical factor influencing the efficacy of cell-based therapies. In recent years, models integrating immune cells with lung organoids have advanced rapidly, providing a pivotal platform for evaluating the interaction between stem cells and the immune system[69] (Figure 3).

Figure 3 Proposed framework for lung organoids in combination therapy and safety assessment. This figure presents a conceptual model for how lung organoids may be applied to evaluate combination and key safety domains. EV: Extracellular vesicle; MSC: Mesenchymal stromal cell.

Immune co-culture is not merely a simple mixture of cells; the choice of technical strategy directly determines the reliability of elucidating stem cell immunomodulatory mechanisms. Depending on variations in the introduction method of immune cells, spatial configuration, and dynamic microenvironmental regulation, five primary technical approaches have currently emerged, each with its own advantages and limitations. These approaches must be designed based on the specific scientific questions to be addressed during the validation of stem cell therapies[69,70]. Culturing strategies for immune co-culture lung organoids each exhibit distinct characteristics in terms of physiological fidelity, mechanistic resolution, and technical accessibility, making them suitable for addressing different stem cell validation challenges. Currently, most experiments opt to directly co-embed immune cells and epithelial cells within a matrix, enabling direct contact between the two components within a 3D matrix, this method enables the assessment of contact-dependent immunomodulatory effects. In a mouse lung organoid-macrophage co-culture model, studies have demonstrated that bone marrow-derived MSCs can mitigate LPS-induced acute lung injury by inhibiting the production of pro-inflammatory cytokines [e.g., C-C motif chemokine ligand 3 (CCL3), CCL4, interleukin (IL)-1β, IL-6] in alveolar and interstitial macrophages, as well as by blocking the activation of the nuclear factor kappa B pathway[18]. This study has provided direct evidence that the organoid-immune cell co-culture system can reproduce MSC-mediated immunomodulatory effects, and that this system is suitable for mechanistic investigation studies. There are some serious technical limitations in the generalizability of these results. Incompatibility among organoid maintenance factors (e.g., epidermal growth factor) and the requirements for immune cell activation (e.g., IL-2 required for T cells) in the culture medium, uneven distribution of immune cells within Matrigel, and a lack of standardised operating procedures for inoculation ratios and co-culture duration all lead to significant batch-to-batch variation[69]. Current organoid systems are also unable to simulate the systemic immune surveillance, biological distribution patterns and complex chemokine gradients in the body that regulate the recruitment of immune cells. Indirect co-culture approaches frequently employ microplate-based systems, which separate organoids from immune cells via a semi-permeable membrane, permits only the exchange of soluble factors, thereby enabling investigation of the underlying mechanisms. It is noteworthy that the dynamic co-culture system implemented in microfluidic organ-on-a-chip platforms can utilize continuous perfusion to simulate blood shear stress and establish a stable chemokine gradient, which has the potential to recapitulate immune cell extravasation and tissue infiltration processes[71,72] (Figure 3).

In stem cell research, this technique can be utilized to observe in real-time distribution of MSC-EVs in dynamic fluids and its regulation of the chemotactic behavior of immune cells; however, these applications are still in their initial stages; at present, no studies have connected the immunomodulatory effects observed in organ-on-a-chip assays to clinical outcomes following stem cell transplantation[72]. Different methods for co-culture each possess distinct advantages: Indirect co-culture is employed for screening key paracrine factors; direct co-culture is used to validate contact-dependent mechanisms; air-liquid interface preserves the original immune ecosystem; and chip-based systems integrate multiple physical factors for systematic validation. These methods progress stepwise in their implementation. Whether using a single type of immune cell or multiple types thereof, numerous challenges arise during the co-culture process, particularly the lack of standardized protocols regarding the seeding ratio of immune cells, the duration of co-culture, and the functional validation metrics, which further hinder the comparability of findings across different studies. In the future, it is imperative to establish integrated standardized operating procedures for co-culture, clearly defining the optimal seeding ratios and co-culture durations for various immune cell combinations (Figure 3).

It is worth noting that stem cell-derived airway cells, rather than lung organoids themselves, are increasingly being viewed as candidates for cell-based therapies[73]. Studies in animals have shown that lung progenitor-derived organoids exhibit strong engraftment, survival and epithelial regeneration after injury in the lung, and can promote alveolar regeneration and inflammation resolution[73-75]. Herriges et al[74] demonstrated durable alveolar engraftment of human pluripotent stem cell-derived lung epithelial cells in immunocompetent mice, with grafts persisting for months and contributing to epithelial repair after injury. This study provided preclinical evidence that pluripotent stem cell (PSC)-derived lung epithelial progenitors can integrate into host alveolar structures and maintain functional phenotypes in vivo. Ma et al[73] described a method for reconstructing airway stem cells by transplanting primary or PSC-derived basal cells, thereby establishing a long-term functional regenerative model of the airway epithelium in mice. It is important to note that the above findings stem from in vivo animal transplantation models, rather than from the validation of stem cell products using organoid technology. In addition, co-culture of lung epithelial progenitor cells and endothelial cells produced vascularized lung organoids that could be transplanted into a host for engraftment in the circulation, this approach thereby overcomes the limitations associated with nutrient supply and metabolic waste removal in organoid transplantation[76]. It should be noted that these results are not directly applicable to the validation of the efficacy of stem cell products in organoid technology, as they are all derived from in vivo animal transplantation models. At present, the predictive value of organoid co-culture data for the success rate of in vivo colonisation has not been fully verified. To transform organoid technology from a disease model into a regulatory-approved platform for validating stem cell therapy, prospective studies need to be conducted; that is to say, correlations between safety and immunocompatibility parameters measured in organoids and clinical outcomes of the same patients should be analyzed, but such validation has not yet been carried out.

CHALLENGES AND PROSPECTS

Although lung organoids are promising complementary preclinical models for stem cell-based therapies, they have not yet fully reproduced the complex systems of the human body. Moreover, most of the organoids lack vascular networks, nerve cells and full immune-cell systems; thus, their impact on the immune regulation of the body by stem cells cannot be fully assessed. There are varying methodologies for constructing organoids from different sources, and the lack of unified standards regarding culture conditions and evaluation criteria further restricts cross-sectional comparisons among studies and compromises their reliability in clinical translation. As the next frontier in organoid research, lung organoids should follow a progressive translational roadmap characterized by “increased structural complexity → standardized platforms → personalized applications → clinical validation”, thereby facilitating the evolution of lung organoids from disease modeling tools to clinical stem cell therapy validation platforms. From a structural perspective, it is essential to reconstruct the cellular interaction network within the lung tissue microenvironment. The efficacy of stem cell-based therapies for pulmonary diseases depends not only on their ability to directly differentiate and replace damaged cells but also on their paracrine communication with other cells within the local microenvironment. Similarly, lung epithelial cells in vivo are subjected to unique physical stimuli, including cyclic respiratory stretching, fluid shear stress, and gas-liquid interface interactions, which cannot be accurately replicated by static-culture-derived organoids. Organoid-on-a-chip technology, on the other hand, can also create dynamic mechanical and fluid-dynamic microenvironments to some extent; thus, it is suitable for the verification of stem cell therapy. MSC-EVs can be distributed in real time under physiologically relevant fluid shear stress in a microfluidic lung-on-a-chip platform; this stress directly affects the membrane integrity, content retention and uptake by target epithelial cells of the EVs. Such platforms can also simulate the alveolar-capillary barrier to determine whether MSC-EVs accumulate preferentially in damaged epithelial areas over healthy epithelial areas, and this parameter cannot be assessed in static Matrigel culture systems. Dynamically perfuse the system and create a stable chemokine gradient that mimics the process of immune cell extravasation. Such chip-based systems can be used in the field of stem cell research to observe the distribution of MSCs or MSC-EVs after systemic administration under physiological shear stress, as well as their regulatory role in the chemotactic behaviour of immune cells, in principle in real time[77]. The detection methods of these chips have also revealed some mechanisms for stem cell products. However, it should be pointed out that no previous studies have performed an association analysis between immune modulation-related biomarkers measured by organoid-on-chip technology and clinical efficacy in patients who have received stem cell therapy. On the platform, efforts should be made to replace animal-derived substrates with chemically defined culture media, establish cross-laboratory standard operating procedures, and develop unified validation benchmarks. At the same time, an automated liquid-handling system should be integrated with artificial intelligence (AI)-based live-cell imaging analysis technology; deep-learning algorithms for dynamic analysis of bright-field images can be employed to increase detection throughput and improve the reproducibility of result assessment. It should not be assumed that because the AI-assisted analysis technology is good at pattern recognition and feature extraction, it can independently determine the clinical predictive value of organoid response. Large-scale, diverse training datasets and strict external validation are required for organoid research deep learning models to prevent overfitting and demographic bias[13]. To ensure that the research platform for lung organoids can support regulatory decision-making, the efficacy and safety prediction results of AI should be prospectively validated in controlled clinical trials on patients, and they should not be regarded as surrogate endpoints based solely on computer simulation results. At the level of implementation, the resource and regulatory challenges for the stem cell validation platform based on lung organoids in clinical translation are considerable. Construction of PDO biobanks is limited by the risk and quality issues in obtaining biopsy tissue samples. Based on an analysis of about 12000 surgical lung biopsy cases for interstitial lung disease from the National Inpatient Sample Database (2000-2011), the in-hospital death rate for elective surgical operations was 1.7%, but it reached a high of 16.0% in the case of emergency surgery[78]. In addition, organoid cultures derived from patients with advanced pulmonary fibrosis or severe COPD have shown that as the disease progresses, the number of progenitor cells available for organoid construction and their clonogenic ability are both reduced; thus, organoid formation efficiency has declined significantly[29-38]. In addition, the grant provided by the National Center for Advancing Translational Sciences at the United States In July 2024, the National Institutes of Health established a Microphysiological Systems Translation Research Center and incurred significant costs for GMP facility compliance, annual maintenance, environmental monitoring and personnel training; as a result, this technology may not be widely available in the short run and may be limited to specialised academic institutions or commercial enterprises. Regulations for advanced therapeutic medicinal products are continuously being updated; at present, the data from organoids have not yet been fully approved. Although the EMA’s 2025 guidance suggests adding data from microphysiological systems to the non-clinical submission materials, it also requires parallel in vivo verification and warns that an immature quality development may harm the application for marketing authorization[79]. In the United States, in April 2025, the FDA announced a phased introduction of organ-on-a-chip data to reduce animal experiments; at present, this policy primarily covers the safety assessments of monoclonal antibodies, and for cell- and gene-therapy products, in vivo studies on biodistribution, persistence and tumourigenicity are still required[80]. Therefore, the lung organoid platform should be used as a support for evidence generation in areas such as mechanistic exploration, assessment of donor heterogeneity and large-scale screening, but not replace animal experiments, manufacturing quality control or randomised clinical trials (Table 4).

Table 4 Comparison of the advantages and limitations among lung organoids, lung-on-a-chip systems, animal models, and two-dimensional cell cultures.

Comparison dimensions	Animal model	2D cell culture	Lung organoids	Lung-on-a-chip
Human-derived physiological fidelity	Moderate	Low	High	Medium to high
Cost	Low	Low	Medium to high	Moderate
Throughput	Low	High	High	Moderate
Level of operational difficulty	Medium to high	Low	Low	Low
Reproducibility and standardization	Moderate	High	Moderate	Moderate
Applicability of stem cell therapy	Moderate	Low	High	High
Core limitations	Significant species differences; ethical controversies	Lack of a physiological gradient; absence of multicellular interactions or an immune microenvironment	No physiological mechanical stress; lacks functional vascular and nervous systems	High technical complexity; incomplete vascularization
Classic applications	Research on systemic disease mechanisms; in vivo analysis; systemic toxicity assessment	Early drug screening; basic research on cellular signaling pathways; large-scale cytotoxicity testing	Disease modeling; PDO drug sensitivity testing; research on stem cell differentiation and repair mechanisms; establishment of a precision medicine biobank	The relationship between mechanics and disease; organ-organ interactions; assessment of inhaled nanoparticle toxicity

The comparative assessment results presented in this table reflect the authors’ qualitative synthesis of existing literature and expert interpretations, rather than representing a validated consensus ranking or the conclusions of a meta-analysis. The ratings (‘Low’, ‘Moderate’, ‘High’) are descriptive generalizations based on the volume and directness of available evidence for each platform and should be interpreted as relative indicators rather than absolute performance metrics. These ratings may vary depending on specific experimental contexts, disease models, and research questions. They should be regarded as indicative references, not as quantitative benchmarks. PDO: Patient-derived organoid.

Furthermore, a standardized, high-throughput, and automated evaluation system should be established. In terms of application, patient-derived autologous iPSCs or biopsy tissues should be utilized to construct lung organoids, which, combined with gene-edited stem cells, can predict their survival, differentiation, and repair potential within specific patient microenvironments, thereby enabling the precise screening of patient cohorts most likely to benefit from these therapies. A personalized evaluation platform integrating patients, diseases, and treatments should be established. This approach should be validated through clinical controlled trials, establishing a correlation between in vitro efficacy predictions and actual clinical outcomes in patients, while clarifying the sensitivity and specificity of the predictive model, thus paving the way for the clinical translation of organoids. In the long term, it is necessary to develop a “lung-immune-metabolism” multi-organ chip system, transplant the vascularized lung construct into an immunodeficient animal model to validate its in vivo vascularization, stem cell engraftment, and long-term functional reconstruction capabilities, thereby completing a closed-loop validation process spanning from in vivo to in vitro settings. In summary, the development of next-generation engineered lung constructs should not be confined to the optimization of individual technologies alone, and multidisciplinary integration of vascularized constructs, immune co-culture systems, organ-on-a-chip platforms, AI-driven analysis, and bioreactor technologies is required to transform lung organoids into a standardized, high-throughput, and personalized stem cell therapy preclinical validation platform, thereby advancing the transition of regenerative medicine approaches for pulmonary diseases from the laboratory to clinical practice.

CONCLUSION

Lung organoids provide increasingly sophisticated human-derived systems for modeling selected features of lung disease and for exploring the activity of stem cell-based therapeutic products. Current evidence most strongly supports their use in disease modeling, mechanistic studies, and comparative ex vivo evaluation of candidate interventions. However, evidence that organoid responses predict clinical outcomes remains limited, and prospective clinical correlation studies are lacking. Organoids also cannot independently assess systemic biodistribution, thromboembolic risk, whole-body immunogenicity, host-pathogen interactions, or long-term tumorigenicity. Future translational development will require standardized protocols, multicenter reproducibility studies, quantitative potency assays, integration with vascularized and immune-competent models, and prospective comparison with animal and clinical outcomes. Thus, lung organoids should be regarded as complementary components of a broader preclinical framework rather than replacements for animal studies, manufacturing quality control, or clinical trials.