Organoids for disease modeling and treatment: state-of-the-art.

3D printing technologies are widely democratized for customized processing and tissue engineering, including cell-laden scaffold generation and multicellular constructs for nerve-bone crosstalk [ 112 , 113 ]. Differing from conventional 3D organoid protocols with the inherent defects in high variability of self-organizing growth, 3D bioprinting technologies have enabled personalized and architecturally engineered organoid fabrication, and thus fulfill the stimulation of physiological organogenesis and disease progression [ 114 , 115 ] (Fig.  2 C, Table  2 , Supplementary Table S1). By utilizing the NovoGen MMX bioprinter, cell pellets were resuspended for the standard organoids bioprinted as patches onto Transwell membranes [ 116 ]. Recently, custom-built multi-head extrusion 3D printers were employed for cell-laden spinal cord organoid scaffold construction in a layer-by-layer manner with cell-laden ink comprised of Matrigel, neural medium with growth factors, and regionally specific spinal neural progenitor cells (sNPCs) derived from hiPSCs [ 117 ]. Notably, Li et al. recently summarized the encouraging advantages as well as ongoing challenges of the three categorized bioprinted organoids for personalized healthcare and biomedical science [ 114 ]. Collectively, 3D bioprinting serves as an alternative strategy for improving graft-host communications post-transplantation, yet the application for regionally specific organoid generation remains nascent [ 117 ]. Compared to the aforementioned conventional 2D or 3D approaches, the 3D bioprinted biomimetic organoid models are more promising for the fabrication of multicellular and reproducible structures, which also reveal reduced experimental costs but enhanced high-throughput screening [ 118 ] (Table  3 ).

In recent years, both the sources of health donors and patients have been employed for multitudinous organoid preparation in clinical trials. According to the ClinicalTrials.gov website ( https://clinicaltrials.gov/ ) of National Institutes of Health (NIH), a total number of 190 organoid-based clinical trials were registered, including 84 interventional studies and 106 observational ones (Fig.  4 A, Supplementary Table S2). Of them, 49.47% of total trials were under recruiting and only 5.79% were completed, whereas a considerable proportion of the projects are in a state of unknown (20.53%) or not recruiting (21.05%) (Fig.  4 B, Supplementary Table S2). Meanwhile, as shown by the histogram (excluded 2 studies in the withdraw status ( NCT04254705 , NCT05378048 )), China (with 59 trials), Italy (with 28 trials) and France (with 26 trials) ranked in the top three of the 17 countries in the number of clinical trial locations (Fig.  4 C, Supplementary Table S3). These data also intuitively revealed the rapid progress of Asia and Europe in the translational trajectory of organoid-based clinical trials for disease modeling and intervention. Fig. 4 The overview of organoid-based clinical trials. A. The categories of 190 organoid-based clinical trials according to ClincialTrial.gov database of NIH, including 84 items of interventional trials and 106 items of observational trials up to July 31, 2025. B. The status of the registered clinical trials, including withdraw (2 items), unknown (39 items), terminated (4 items), recruiting (94 items), not recruiting (40 items), and completed (11 items). C. The locations of the 188 registered organoid-based clinical trials (excluded the 2 withdraw items) in 17 countries (including 173 items) and those unknown (including 15 items). D–E. The distribution of 188 registered clinical trials of organoid-based cancer research ( D ) and non-cancer research ( E ) The overview of organoid-based clinical trials. A. The categories of 190 organoid-based clinical trials according to ClincialTrial.gov database of NIH, including 84 items of interventional trials and 106 items of observational trials up to July 31, 2025. B. The status of the registered clinical trials, including withdraw (2 items), unknown (39 items), terminated (4 items), recruiting (94 items), not recruiting (40 items), and completed (11 items). C. The locations of the 188 registered organoid-based clinical trials (excluded the 2 withdraw items) in 17 countries (including 173 items) and those unknown (including 15 items). D–E. The distribution of 188 registered clinical trials of organoid-based cancer research ( D ) and non-cancer research ( E ) To further dissect the detailed information of organoid-based trials, we grouped the residual 188 studies into cancer research and non-cancer research, and generated the statistical charts with clinical trial numbers (“conditions” in Tables 5 and 6 for details) and the corresponding enrollment. As shown in Fig.  4 D and Table  5 , we intuitively observed that breast cancer, pancreatic cancer and lung cancer occupied the top three both in registered trial number (shown as Conditions (Log 2 N11)) and enrollment (shown as Enrollment (Log 10 N12)). Simultaneously, other organoid-based trials were widely involved in 39 kinds of cancers among digestive system (e.g., colorectal cancer, gastric cancer, hepatocellular carcinoma, cholangiocarcinoma, esophageal cancer), reproductive system (e.g., ovarian cancer, cervical cancer, endometrial cancer, gynecologic cancer), nervous system (e.g., glioblastoma/glioma, brain tumor, vestibular schwannoma), urinary system (e.g., prostate cancer, bladder cancer, kidney cancer, urothelial carcinoma), skeletal system (e.g., osteosarcoma, sarcoma), endocrine system (e.g., thyroid cancer, pancreatic ductal adenocarcinoma), circulatory system (e.g., hematologic malignancy), and head and neck cancer (Fig.  4 D, Table  5 ). Table 5 The registered clinical trials of organoid-based cancer research Cancer Types Conditions (N11) Conditions (log 2 N11 ) Enrollment (N12) Enrollment (Log 10 N12 ) Breast Cancer 22 4.46 1959 3.29 Pancreatic Cancer 17 4.09 2418 3.38 Lung Cancer 15 3.91 1543 3.19 Colorectal Cancer 13 3.70 864 2.94 Ovarian Cancer 12 3.58 802 2.90 Glioblastoma/Glioma 11 3.46 720 2.86 Gastric Cancer 7 2.81 576 2.76 Head and Neck Cancer 7 2.81 567 2.75 Hepatocellular Carcinoma 7 2.81 522 2.72 Pancreatic Ductal Adenocarcinoma 5 2.32 1128 3.05 Prostate Cancer 5 2.32 766 2.88 Gastrointestinal Cancer 4 2.00 193 2.29 Solid Tumor 4 2.00 476 2.68 Bladder Cancer 3 1.58 355 2.55 Colon Cancer 3 1.58 640 2.81 Brain Tumor 2 1.00 511 2.71 Cervical Cancer 2 1.00 64 1.81 Cholangiocarcinoma 2 1.00 65 1.81 Endometrial Cancer 2 1.00 56 1.75 Esophageal Cancer 2 1.00 130 2.11 Kidney Cancer 2 1.00 31 1.49 Osteosarcoma 2 1.00 60 1.78 Rectal Cancer 2 1.00 100 2.00 Sarcoma 2 1.00 61 1.79 Thyroid Cancer 2 1.00 95 1.98 Biliary Tract Cancer 1 0.00 14 1.15 Esophageal Squamous Cell Carcinoma 1 0.00 192 2.28 Gynecologic Cancer 1 0.00 200 2.30 Hematologic Malignancy 1 0.00 70 1.85 Intrahepatic Cholangiocarcinoma 1 0.00 40 1.60 Medulloblastoma 1 0.00 60 1.78 Melanoma 1 0.00 50 1.70 Meningioma 1 0.00 20 1.30 Neuroblastoma 1 0.00 11 1.04 Neuroendocrine Neoplasm 1 0.00 200 2.30 Rare Tumour 1 0.00 200 2.30 Salivary Gland Cancers 1 0.00 40 1.60 Urothelial Carcinoma 1 0.00 200 2.30 Vestibular Schwannoma 1 0.00 100 2.00 Table 6 The registered clinical trials of organoid-based non-cancer research Non-cancer Types Conditions (N21) Enrollment (N22) Enrollment (Log 2 N22 ) Allergy 1 100 6.6 Amyotrophic Lateral Sclerosis 1 30 4.9 Multiple Sclerosis 1 30 4.9 Primary Sclerosing Cholangitis 2 380 8.6 Inflammatory Bowel Disease 5 370 8.5 Gut Inflammation 1 41 5.4 Intestine Disease 1 375 8.6 Enterocolitis 1 18 4.2 Irritable Bowel Syndrome 1 17 4.1 Pediatric Intestinal Pseudo-obstruction 1 4 2.0 NAFLD 1 60 5.9 Cystic Fibrosis 2 72 6.2 Diabetes Mellitus 2 34 5.1 Food Hypersensitivity 1 115 6.8 Endometriosis 1 92 6.5 Endometrial Receptivity 1 60 5.9 Asherman Syndrome 1 6 2.6 Infertility 2 105 6.7 Preeclampsia 1 20 4.3 Klinefelter Syndrome 1 20 4.3 Alzheimer Disease 1 14 3.8 Bipolar Disorder (BD) 1 10 3.3 Chronic Obstructive Pulmonary Disease 1 30 4.9 Spondyloarthritis 1 30 4.9 Eye Abnormalities 1 20 4.3 Staphylococcal Infections 1 500 9.0 Rare Diseases 1 100 6.6 The registered clinical trials of organoid-based cancer research The registered clinical trials of organoid-based non-cancer research As to non-cancer research, statistical charts were implemented with registered trial number (shown as Conditions (N21)) and enrollment (shown as Enrollment (Log 2 N22)) (Fig.  4 E, Table  6 ). These organoid-based clinical trials were mainly catalogued in immuo-dysfunctional disorders (e.g., allergy, amyotrophic lateral sclerosis, multiple sclerosis, primary sclerosing cholangitis), gut diseases (e.g., inflammatory bowel disease, gut inflammation, intestine disease, enterocolitis, irritable bowel syndrome, pediatric intestinal pseudo-obstruction), liver diseases (e.g., nonalcoholic fatty liver disease, cystic fibrosis, diabetes mellitus, food hypersensitivity), reproductive diseases (e.g., endometriosis, endometrial receptivity, Asherman syndrome, infertility, preeclampsia, Klinefelter syndrome), neurological disorders (e.g., Alzheimer disease, bipolar disorder, chronic obstructive pulmonary disease, spondyloarthritis), and relative diseases (e.g., eye abnormalities, staphylococcal infections, rare diseases). Among them, inflammatory bowel disease and staphylococcal infections ranked first of 27 diseases in registered trial number and in enrollment, respectively (Fig.  4 E, Table  6 ). Collectively, the landscape of registered clinical trials showed the state-of-the-art updates in the field and indicated the concerns of clinicians upon organoid-based precise diagnosis and individualized treatment. In particular, the predominance of oncology trials suggests organoids are most mature in cancer drug testing, whereas applications in neurological disorders remain underdeveloped despite strong preclinical promise.

To date, enormous types of miniature organoids with self-organizing properties have been reported by pioneering investigators in the respective field, which collectively arise groundbreaking insights into precision medicine and boost the course of human development biology and disease research (Fig.  1 B, Supplementary Table S1, Supplementary Information: Supplementary References for Supplementary Table S1). These organoids can be roughly catalogued into endoderm, mesoderm, and ectoderm derivatives according to the origins of organ development (Fig.  1 C, Table  1 ). In recent years, human organoids have been derived from diverse cell sources such as human induced pluripotent stem cells (hiPSCs) [ 26 , 27 ], human embryonic stem cells (hESCs) [ 27 , 28 ], mesenchymal stem/stromal cells (MSCs) [ 4 , 29 , 30 ], patient tissue-derived adult stem cells (ASCs) [ 31 , 32 ], adult progenitor cells [ 33 ], somatic cells [ 17 , 31 ], primary cells enriched from fetal tissues (e.g., fetal kidney tissue) and placental tissues (e.g., villous cytotrophoblasts (vCTBs)) [ 34 , 35 ], and commercially available cell lines [ 36 , 37 ] (Fig.  1 B, Tables 1 and 2 , Supplementary Table S1). Generally, the majority of the conventional organoids are in vitro models except for a very few in vivo ones, including in vivo BM-MSCs-based osteo-organoids and bone organoids [ 29 , 30 , 38 ]. Table 1 Representative organoids from hPSC-derived germ layers Classification Subtypes of organoids The applications References Endoderm Human intestinal organoids (HIOs) • To benchmark hPSC-derived HIOs under multiple culture conditions; • To recapitulate the transient states and genomic characteristics during physiological and pathological organogenesis; • To illuminate the key genes during regionalization of intestinal epithelium and mesenchyme Refs. [ 10 , 104 ] Hepatic organoids (HOs) • To achieve the structural complexity and functional maturity of multi-lineage liver organoids (mLOs); • Applications in modeling liver pathologies; • To uncover the underlying mechanism including signaling pathways Ref. [ 248 ] Mesoderm Vascular organoids (VOs) • VOs enable the efficient co-differentiation of endothelial cells and mural cells; • VOs reflect the vascular heterogeneity as well as temporal regulation of transcription factor expression; • To study vascular development, vascular disease modeling and treatment Refs. [ 87 , 188 ] Kidney organoids • Long-term clonal expansion of primary nephron progenitor cells in vitro; • To facilitate genome-wide CRISPR screening for kidney development and disease modeling; • To uncover plasticity in human podocyte reprogramming Refs. [ 57 , 116 , 249 , 250 ] Ectoderm Cerebral organoids • For large-scale generation of brain region-specific organoids for regenerative medicine; • To explore brain development in forebrain, midbrain and hypothalamus; • To fulfill the demands for compound testing in brain-related disorders Refs. [ 11 , 23 , 95 , 251 ] Striatal organoids (SOs) • To illuminate the mitochondrial dysfunction and neurodegeneration; • To verify the regulatory mechanism of medium spiny neurons (MSNs); • To underscore the potential therapeutic options for Huntington's disease (HD) Ref. [ 55 ] Table 2 Organoid model induction from tissue-enriched stem cells and the derivatives Classification Methods Cell Source Culture Medium and ECM Signal Transducer Culture period References In vivo osteo-organoids In NSG mice BM-MSCs α-MEM, Matrigel-equivalent matrix Daily PTH subcutaneous injection for 28 days 8–10 weeks Reinisch, et al. [ 29 ] In vivo hematopoietic organoids/BM organoids 3D culture with bioreactors and microfluidics, In NSG mice BM-MSCs Hypertophyc medium/Chondrogenic medium/Osteogenic medium, FBS, Matrigel/hydrogels/silicate structures BMP2, BMP7, parathyroid hormone (PTH), TGF-β 3–28 days Pievani, et al. [ 48 ] In vivo miniature bone/marrow organoids In SCID/beige mice Bone marrow stromal cell (BMSC), blood-borne fibroblasts (CB-BFs) Chondrogenic differentiation medium (DMEM-high glucose, ITS premix, sodium pyruvate), RPMI medium, Matrigel TGF-β1, Dexamethasone 8 weeks Pievani, et al. [ 252 ] Intestinal organoids 2D culture, 3D culture Isolated crypts DMEM/F12, DMEM medium, B27 supplement, Matrigel/Collagen EGF, R-Spondin 1, Noggin 7 days Meng, et al. [ 32 ] Colonic organoids 3D culture Crypts from human intestinal samples Advanced DMEM/F12, RPMI 1640, Smooth Muscle Cell Growth Medium-2, FBS, B27 supplement, Matrigel, Type I Collagen Solution, Cryomatrix Rspondin-1, Noggin, FGF2, IGF-1, IGF-2, EGF, NRG1, A83-01, Y27632, M-CSF 7–8 days Mitrofanova, et al. [ 127 ] Fetal kidney organoids 3D culture hFK cells from fetal kidney tissue hNPSR medium, AKO medium, DMEM/F12, B-27 minus vitamin A, Cultrex BME, Matrigel BMP-7, CHIR99021, A83-01, LDN-193189, FGF-2, R-Spondin 1, FGF-10 2–3 weeks Namestnikov, et al. [ 34 ] Osteosarcoma organoids 2D culture, On organ chip HUVECs, patient tissue derived cells Endothelial cell medium, DMEM, DMEM/F-12, FBS Thrombin, Aprotinin 8–20 days Du, et al. [ 71 ] Soft tissue and bone sarcoma (STBS) organoids Minced tissues for suspension culture Tumor tissue ADMEM/F12 medium, B-27 without vitamin A, N-acetylcysteine, Nicotinamide Wnt3a, R-Spondin 1, Noggin, EGF, A83-01, Gastrin, SB-202190 2 weeks Ma, et al. [ 17 ] Ovarian cancer organoids 3D culture Cancer cells from patient samples Advanced DMEM/F12, Ultimatrix, B27 supplement Noggin, R-Spondin 1, EGF, Heregulin beta-1, A83-1, forskolin, hydrocortisone, Y27632 7–14 days Nikeghbal, et al. [ 253 ] Testicular organoids 3D culture SSC, Leydig cells, Sertoli cells Complete StemPro-34 medium, FBS, human testis ECM RA, SCF, follicle-stimulating hormone 23 days Pendergraft, et al. [ 70 ] Trophoblast organoids 3D culture villous cytotrophoblasts (vCTBs) Advanced DMEM/F12, N2B27 supplement, Matrigel A83-01, Noggin, EGF, CHIR99021, PGE2, HGF, R-spondin 2–3 weeks Haider, et al. [ 75 ] Trophoblast organoids 3D culture Placental villous stromal cells Advanced DMEM/F12 medium, FBS, ITS-X supplement, KSR, GFR-Matrigel EGF, FGF2, CHIR99021, A83-01, R-Spondin 1, Y-27632, PGE2 10–14 days Turco, et al. [ 35 ] Representative organoids from hPSC-derived germ layers • To benchmark hPSC-derived HIOs under multiple culture conditions; • To recapitulate the transient states and genomic characteristics during physiological and pathological organogenesis; • To illuminate the key genes during regionalization of intestinal epithelium and mesenchyme • To achieve the structural complexity and functional maturity of multi-lineage liver organoids (mLOs); • Applications in modeling liver pathologies; • To uncover the underlying mechanism including signaling pathways • VOs enable the efficient co-differentiation of endothelial cells and mural cells; • VOs reflect the vascular heterogeneity as well as temporal regulation of transcription factor expression; • To study vascular development, vascular disease modeling and treatment • Long-term clonal expansion of primary nephron progenitor cells in vitro; • To facilitate genome-wide CRISPR screening for kidney development and disease modeling; • To uncover plasticity in human podocyte reprogramming • For large-scale generation of brain region-specific organoids for regenerative medicine; • To explore brain development in forebrain, midbrain and hypothalamus; • To fulfill the demands for compound testing in brain-related disorders • To illuminate the mitochondrial dysfunction and neurodegeneration; • To verify the regulatory mechanism of medium spiny neurons (MSNs); • To underscore the potential therapeutic options for Huntington's disease (HD) Organoid model induction from tissue-enriched stem cells and the derivatives

Precise prediction of drug toxicity is a long-standing challenges for therapeutic development [ 198 ]. Longitudinal studies have highlighted diverse invitro organoid-based models in drug efficacy-, metabolism- and toxicity- relevant pharmaceutical research at both preclinical and clinical stages, including liver organoids [ 199 – 201 ], lung organoids and lung-on-a-chip [ 202 ], brain organoids [ 203 , 204 ], cardiac organoids [ 45 ], head and neck tumor organoids [ 205 ], placental organoids [ 206 ], kidney organoids [ 207 ], breast cancer organoid [ 208 ], primary testicular cell-derived organoids [ 209 ] (Fig.  3 ). Differing from animal models with major limitations in species differences, organoids eliminate the discrepancy in drug-mediated hepatotoxicity, metabolism and toxicological outcomes [ 210 ]. For instance, hiPSC-derived testicular organoids were constructed for mimicking the tissue structure and the sensitivity to compounds like semaglutide were further verified, while the urothelial carcinoma organoids with high resemblance emerged as novel platforms for prospective pharmacotyping [ 211 , 212 ]. As to human cerebral organoids, researchers demonstrated the toxicological effects of polypropylene nanoplastic (PP-NP) exposure upon neurobehavioral outcomes and fetal brain development during pregnancy by targeting CYSLTR1 and PTH1R [ 204 ]. By conducting liver organoids and machine learning, the hepatotoxicity of 6-PPDQ (a novel pollutant) exposure related to liver injury and chronic liver diseases was uncovered [ 213 ]. Similarly, angiosarcoma-derived organoids (denoted as “sarconoids”) served as accurate models and held far-reaching implications for advancing the pharmacological development integrated with the drug library [ 214 ]. Instead, a novel microfluidic organoid-slice-on-a-chip (OSOC) platform was introduced for simultaneously evaluating the anti-cholangiocarcinoma drug efficacy as well as hepatorenal toxicity by integrating microfluidic multi-organ chip with cholangiocarcinoma organoids (CCOs) [ 215 ]. In consequence, organoids serve as a physiologically pertinent platform for organ-specific toxicological studies and monitoring cellular responses due to the high similarities with real organ in structure and function [ 216 ]. The involvement of multidisciplinary technologies is inevitable for fulfilling the accuracy and sensitivity of organoids for drug toxicity assessment.

As an exemplary ex vivo experimental system, organoids represent innovative approach for high throughput screening (HTS) of diverse disease-associated first-in-class drugs and therapeutic compounds [ 24 ]. For example, Karim et al. took advantage of the lung organoids for broad-spectrum antiviral drug screening, and identified the small molecule inhibitor RMC-113 for replication suppression of multiple RNA viruses (e.g., SARS-CoV-2), which collectively indicated the pharmaceutical screening value of organoid as a candidate strategy for HTS purposes [ 192 ]. Instead, organoid platform were applied for cancer nanomedicine and nanoparticle drug development, while the SNX10 deficiency- and phosphoinositide 3-kinase (PI3K) inhibitor-associated anti-HER2 ADCs resistance were respectively identified by integrating the transcriptome data and patient-derived breast organoids [ 22 , 193 , 194 ]. Of note, the multidrug-drug administration regimens and their synergistic effect by combing machine learning and organoids were reported, which would accelerating the advanced cancer treatment [ 195 ]. Interestingly, investigators reported the directed maturation of hPSC-derived cardiac organoids (hCOs) for complex disease modeling (e.g., cardiomyopathy) and desmoplakin-related drug screening (e.g., INCB054329) [ 196 ]. Additionally, the updates upon microfluidic devices with liver organoids were described for simulating liver microenvironment and facilitating high-throughput drug screening for diverse liver diseases (e.g., viral liver diseases, liver fibrosis, hepatitis, and monogenic diseases) [ 197 ]. Taken together, the integration of organoids with multidisciplinary technologies such as gene sequencing and organ chip provides new avenues for fulfill the high throughput pharmaceutical screening for personalized medicine in future (Fig.  3 ).

During the early stages of embryonic development orchestrated by complex and robust regulatory mechanisms, germ layer specification occurs from the gastrulation stage, and followed by the initial fate determination towards endoderm, mesoderm, and ectoderm derivatives [ 39 – 41 ]. Therewith, a variety of organoids have been created for simulating the corresponding organs such as those of endoderm derivatives (e.g., intestinal organoids [ 42 ], esophageal organoids [ 43 ], airway organoids [ 44 ]), mesoderm derivatives (e.g., cardiac organoids [ 45 ], Jawbone-like organoids [ 46 ], testicular organoids [ 47 ], hematopoietic organoids [ 48 ]), and ectoderm derivatives (e.g., striatal organoids [ 49 ], hair follicle organoids [ 50 ], spinal cord organoids [ 51 ]) (Fig.  1 B, Table  1 ). HPSCs, including hESCs and hiPSCs, are cell populations with self-renewal and multi-lineage differentiation potential [ 41 , 52 ]. The first cell line of hESCs was prominently reported in 1998 [ 53 ], while the emergence of hiPSCs allowed patient- and disease-specific organoid-based research [ 54 , 55 ]. Nowadays, hPSCs are recognized as advantaged cell sources for ex vivo simulation of early embryonic development and organoid generation owing to the versatile germ specialization and the wide range of derivatives [ 7 , 56 ]. Longitudinal studies have indicated the involvement of hPSCs for diverse kinds of organoid construction, including kidney organoids [ 57 ], heart-forming organoids (HFOs) [ 58 ], spinal cord organoids (SCOs) [ 51 ], skeletal muscle organoids [ 59 ], skin organoids [ 60 ], lung and gut organoids [ 26 ], cardiac and hepatic organoids [ 7 ], inner ear organoids [ 61 , 62 ], and neuromuscular organoids (NMOs) [ 63 ]. As to human organoids of mesoderm derivatives, researchers reported the establishment of blood-generating hPSC-derived HFOs encompassing hematopoietic progenitor cells (HPCs) and multiple hematopoietic derivatives, which reflected multidimensional characteristics of primitive and definitive haematopoiesis [ 58 ]. As to human organoids of endoderm derivatives, the induction of human intestinal organoids (HIOs) from hPSCs was achieved within 6–8 weeks, which recapitulated the multi-endodermal organ atlas and reconstructed the molecular dynamics of intestinal mesenchyme and epithelium regionalization [ 10 ]. As to human organoids of ectoderm derivatives, extensive literatures have shown the generation and assembly of brain region-specific organoids from hPSCs, which recapitulated the interactions of GABAergic and glutamatergic neurons in subdomain-specific forebrain [ 64 ]. Collectively, these findings demonstrated the broad differentiation spectrum of hPSCs for organoid-based precision medicine and regenerative medicine (Fig.  1 B, C, Table  1 and Supplementary Table S1). Notably, current literatures have further put forward the feasibility of concurrent generation of assembled organoids from hPSCs of the same origin. For instance, Miao et al. reported the co-development of endoderm and mesoderm from iPSCs, which enabled the organotypic vascularization in both lung and gut organoids [ 26 ]. Very recently, both cardiac vascularized organoids (cVOs) and hepatic vascularized organoids (hVOs) by geometric micropatterning of hPSCs were generated, which were adequate to mimic the first 3 weeks of human development and thus beneficial for addressing the de novo organ vascularization [ 7 ]. Distinct from body tissue-derived cell sources, hPSCs possess remarkable advantages in mimicking the features of early embryogenesis and fulfilling the generation of diverse organoids of the same origin via programmed differentiation whereas with apparent deficiency of tumor microenvironment (TME) and the tumor-forming risk in vivo [ 65 – 67 ].

The 3D organoid systems indicate a paradigm shift from 2D culture and PDX models in cancer modeling attribute to the remarkable immunocompatibility and the resultant faithful recapitulation of the in vivo tumor heterogeneity [ 105 , 151 ]. In detail, PDTOs can mimic the ultrastructure and physiological functions of the in vivo tumor tissues to some extent, and thus attract the attention of investigators in recent years [ 152 ]. For decades, hematological malignancy and metastatic solid tumors have become a heavy burden with high morbidity and mortality worldwide, and the personalized therapeutic regimen remains hindered by the tumor heterogeneity and dynamic TME [ 16 , 22 , 153 ]. Differing from the patient-derived tumor xenograft (PDTX) model with notoriously inherent defects [ 19 , 154 ], PDTOs faithfully replicate the multifaceted characteristics of primary tumors and TME, and thus provide innovative platforms for investigating tumor-immune cell interactions [ 16 , 152 , 155 ]. Longitudinal studies have introduced a variety of PDTOs for cancer research such as soft tissue and bone sarcoma (STBS) organoids [ 17 ], PDAC organoids [ 145 ], primary liver cancer organoids [ 155 ], lung cancer organoids [ 156 , 157 ], pancreatic cancer organoids [ 158 ], ovarian cancer organoids [ 159 ], bladder cancer organoids [ 160 ], and thyroid cancer organoids [ 161 ] (Table  4 ). For instance, investigators in the field described the generation of patient-derived sarcoma organoids that stably recapitulated the parental tumor tissues including molecular characteristics and sarcoma heterogeneity, and resulted in new personalized treatments [ 17 ]. Besides the establishment of PDO-invariant natural killer T cell (PDO-iNKT) platform for personalized cancer immunotherapy, tumor-derived organoids from primary liver cancers were highlighted with encouraging advances [ 155 , 162 ]. By utilizing the PDOs from endometrial cancer (EC), Zhang et al. conducted high-throughput drug repurposing screening and identified the chlorpromazine derivative JX24120 as a prospective compound for EC therapy [ 163 ]. With the aid of PDOs, Shan et al. identified Sarcosine as a ferroptosis inducer for sensitizing lung adenocarcinoma to chemotherapy via orchestrating PDK4/PDHA1 signaling [ 164 ]. Instead, PDTOs and organ chip were reported for evaluating tumor metastasis, which offered valuable insights for personalized decision-making in osteosarcoma patients [ 71 ]. Additionally, the latest updates of innovative organ chip platforms integrated with AI for mimicking the TME of head and neck cancer (HNC) and individualized treatment strategies were introduced as well [ 165 ]. Collectively, PDTOs have offered new opportunity for cancer research and anti-cancer therapy, and organoid biobank is essential and promising for uncovering tumor initiation and progression [ 16 , 166 ]. Additionally, it’s of extreme importance to further verify the gene expression profiling and the genetic mutation spectrum of tumor initiating genes by assembling PDTOs with high-throughput sequencing and machine learning algorithm.

The high-efficiency induction of organoids with faithful recapitulation of the parental organs largely attributes to microenvironment simulation including culture medium with variegated components. To date, diverse commercialized culture products have been developed for feeder-free, serum-free hPSCs maintenance and early differentiation (e.g., Essential 8 ™ medium, STEMium ™ , TeSR ™ -E8, mTeSR ™ 1 and mTeSR ™ Plus medium, TeSR ™ -E6 medium, and KnockOut ™ DMEM with KnockOut ™ Serum Replacement). The basal medium and the concomitant components for organoid construction are vary considerably according to the subtypes of organoids and the stepwise induction (Table  2 , Supplementary Table S1). For vascularized lung organoid generation, hPSCs were dissociated into single cells and seeded in aggregation media (consist of knockout DMEM/F12, knockout serum replacement) for embryonic body formation, and followed by meso-endodermal spheroid induction in N2/B27 basal medium supplemented with the indicated components (e.g., Activin A, CHIR99021, BMP4, PIK90, SB-431542, FBS), and the AT1-specification N2/B27 media with the indicated components (e.g., FGF10, LATS-IN-1, Dexamethasone, cAMP, IBMX, VEGFA, ANG1) for lung organoids as previous reported [ 26 , 98 ]. A serum-free and chemically defined medium was reported for the prolonged culture of hFKOs, and the orchestrating effect of Notch signaling upon proximal and distal tubule formation was further identified [ 34 , 67 ]. Collectively, diverse culture mediums with the nutritional supplements constitute unique microenvironment for initiating stem cell spheroid specification and organoid generation. In consequence, organoid-based precision medicine will extensively benefit from the further development of good manufacturing practice (GMP)-grade, serum-free, component-defined culture medium.

Last but not least, ethical and regulatory considerations should be intensively investigated and receive sufficient attention by the organoid community [ 74 , 242 – 244 ]. For example, a variety of challenges with regard to organoid models generated from hPSCs and authentic human tissues have been extensively reported, including the moral value, informed consent procedures, ownership and commercialization, clinical guidelines, ontology, and legal status [ 20 , 245 ]. As to cell sources for organoid formation, hESCs are currently controversial and subjected to the ethical and political agenda worldwide largely attributes to the origins of human embryos [ 245 , 246 ]. Despite ethics necessitate in vitro investigations, yet the diversity and variability of current organoid systems have hindered pathogenic and mechanistic insights to some extent [ 247 ]. For example, to overcome the ethical risk of organoid symmetry breaking during early embryonic research, investigators turned to a framework of bioengineering and machine learning, and discovered an excitable system of human axial elongation derived by WNT/FGF signaling [ 247 ]. Nowadays, it’s delightful to note that a series of ethical guidelines and standards are under drafting and formulation process by academic organizations and industry associations. Collectively, organoids with complexity and modelling capability constitute a tremendous improvement in the territory of model systems. Organoid technologies have revolutionary prospects in biomedical research and precision medicine owing to their versatility of stimulating the corresponding in vivo conditions including immune-tumor interaction for tissue development and disease modeling as well as functional integrity and immunotherapy. By integrating with the multidisciplinary advances in the technology, organoid-based models are capable of re-creating the architecture and function of human organs, which will provide more powerful and efficient platforms for understanding human development in the context of physiological and pathological conditions.

To date, numerous stimulators and the concomitant signaling pathways have been involved for organoid construction and maturity (Table  2 , Supplementary Table S1). Fibroblast growth factor (FGF), bone morphogenetic protein (BMP), transforming growth factor beta (TGF-β), Notch and Wnt/β-catenin, have been demonstrated to facilitate the expansion and differentiation of oral organoids [ 93 ]. Recently, BMP and Notch signaling are verified crucially for vascularization in cardiac and hepatic organoids [ 7 ], while the signaling pathways such as Notch, PDGFB, and TGF-β are indispensable for vascular maturation in VOs depending on ECs-MCs interactions [ 87 ]. FGF8 signaling can modulate regional identity within a single telencephalic organoid for mimicking early brain development and malformations [ 94 ]. In multi-regional cerebral organoids, FGF8 increases cellular heterogeneity and impacts regional patterning, including the anteroposterior and dorsoventral identity as well as the balance between glutamatergic and GABAergic neurons [ 95 ]. Instead, highly efficient neural conversion of hPSCs was achieved by utilizing two inhibitors of SMAD signaling (Noggin and SB431542) [ 96 ]. With the aid of proteomic analysis, the key islet promoting factors (e.g., canonical WNT, interferon-γ) during islet organoid development and maturation from iPSCs were identified, which highlighted the crucial role of cytokine cocktails in the differentiation microenvironments [ 88 , 89 ]. Generally, the cytokine and chemical compound cocktails vary considerably according to the types of organoids and the stepwise differentiation stages. For example, Bi et al. detailed described the five-stage serum-free induction of islet-like organoids that harboring the major pancreatic endocrine cell types (e.g., α, β, δ) and pancreatic polypeptide cells by stage-specific cytokine and compound cocktails from hiPSCs within 28 days [ 88 ]. According to the research by Shi et al., human ureteric bud (UB) organoids with functional collecting duct (CD) cell types were induced from H9 hESCs in BJFF.6 hiPSCs by stage-specific stimulations, including hPSC maintenance (ROCK inhibitor Y27632 in mTeSR), mesendoderm induction (Activin A, BMP4, CHIR99021, FGF2 in Advanced RPMI 1640), pronephric intermediate mesoderm (IM) induction (A83-01, FGF2, LDN193189, RA in Advanced RPMI 1640), 3D nephric duct (ND) spheres (FGF9/RA, GDNF/RA, GDNF/FGF10/CHIR99021/LDN/A83–01/RA/U0126 in basic differentiation medium), CD organoids (AVP and aldosterone in basic differentiation medium), or UB organoids (CHIR99021, Activin A, FGF9 in basic differentiation medium) [ 97 ]. Therefore, the generation of organoids with functional maturity are practicable and proof-of-concept by spatiotemporally orchestrating the stage-specific signaling cascades. Compared to the expensive cytokines derived from engineered microorganisms, chemical compounds are cost-effective and thus have preferable scalability for large-scale preparation of clinical grade organoids.

Organoids, also known as “miniature organs in a dish” [ 8 , 151 ], have emerged as a promising tool for ex vivo disease remodeling, and organoid techniques have provided a robust and unique platform for recapitulating organogenesis and personalized next-generation pharmacotherapy [ 25 , 38 ]. As to malignant diseases including metastatic tumors, treatment resistance linked to TME will hinder the effectiveness of targeted therapy by promoting tumor growth and suppressing immune function. In detail, the intricate constituents in TME (e.g., multitudinous stromal cells, heterogeneous cancer cells, secreted cytokines, and ECM components) are involved in multidrug resistance via orchestrating cellular, metabolic, and molecular networks [ 222 – 224 ]. However, the majority of the traditional organoid models have fatal deficiency in immune system components, and the integration of organoids with immune cells is of great urgency for drug screening and cancer immunotherapy [ 225 , 226 ]. Therefore, it’s noteworthy that a certain number of immunocompetent organoids have been newly developed for dissecting immune-tumor interactions and advancing personalized medicine including tumor immunotherapy [ 151 ].

Organoids have effectively complemented the current model systems and extended the biomedical research into more physiologically and pathologically relevant human setting [ 20 , 231 , 232 ]. However, the concomitant technical challenges of organoid optimization for fulfilling the requirement are still intractable, which largely attributes to its infancy compared to the well-established disease models (e.g., model organisms, cell lines) [ 20 , 65 ]. Among the pressing issues, the variability of the system (e.g., person-to-person variability, contrasting methods, and quality control) remains the major challenge in organoid technology [ 65 ]. Meanwhile, the complex components (e.g., culture medium and supplements) in personalized organoid systems further deteriorate the degree of complexity for organoid establishment. Hence, for the purpose of increasing the reliability of organoids, researchers should determine the optimal system for organoid construction according to the pros and cons of the complexity and the research objectives. Of note, it’s beneficial to harness the application prospect of organoid technology by accelerating the emerging interdisciplinary integrations, including machine learning, AI, scRNA-SEQ, 3D bioprinting and microfluidics, and intracellular tracing technology (e.g., luminescence, fluorescent protein, immunofluorescence, surface-enhanced Raman scattering, nanoprobes, HaloTag) [ 67 , 138 , 233 – 235 ]. For instance, a Ca 2+ -sensitive HaloTag ligand was infused into hiPSCs for the non-invasive monitoring of PIEZO1 localization and activity in derived neural organoids for investigating mechanotransduction and targeted drug screening [ 234 ]. At the meantime, the TRACER platform (short for tissue roll for analysis of cellular environment and response) was exploited for location-specific isolation of cell populations and scRNA-SEQ analysis of organoids at specific spatial locations for dissecting the impact of cell-generated oxygen gradients upon PDAC organoid heterogeneity [ 236 , 237 ]. Additionally, the innovation of more precise and advanced 3D bioprinter for the personalized fabrication of organoids in terms of tissue mimicry, structural precision and functional fidelity is also extremely necessary [ 114 ].

Novel organoid model system should be developed to improve maturity and complexity for fulfilling the in vivo engraftment and functionality, and particularly the initiating hPSCs and the immaturity of the derivatives [ 87 ]. For instance, Bagley et al. introduced the cerebral organoid co-culture “fusion” paradigm for facilitating independently patterned organoids into single tissues, and the cells revealed high similarity in migratory dynamics and molecular taxonomy as that of cortical interneurons [ 24 ]. Similarly, ureteric bud organoids derived from hESCs and hiPSCs revealed complex morphological development (e.g., uretic bud structures, terminal bifurcation, stalk elongation), together with differentiation capacity into functional collecting duct tissues [ 97 , 227 ]. Of note, the coordinated differentiation of HIOs from hPSCs with functional enteric neurons and vasculature was achieved, which underwent peristaltic-like contractions indicative anastomosed with host vasculature and extended the translational potential [ 104 ]. These findings revealed the multifaceted improvement in the induction of mature organoids from hPSCs. Meanwhile, longitudinal literatures have repeatedly highlighted the vascularization for the biostructure and functional maturation of organoid models. For example, Gong et al. turned to Dox-inducible and modRNA technology for rapid generation of VOs with mature and functional vessel formation within ischemic and transplant model [ 87 ]. However, the further establishment of versatile and functional mature organoids by non-gene-editing strategies such as cytokine- or chemical compound- mediated programming will further fulfill disease modeling and regenerative medicine in future. For example, Jun and the colleagues developed the engineered vasculature in hPSC-derived islet organoids with functional maturation property via modulating the BMP2/4-BMPR2 signaling [ 228 ]. Very recently, parental tumor tissues that were dissociated into pieces (100 μm diameter) for 3D culture of spherical PDTOs (named STBS organoids), which were different from those approaches by dissociating tissues into single cells and thus were competent for rapid and personalized drug screening [ 17 ]. Additionally, the development of more compatible or decellularized ECM substitutes (e.g., decellularized skeletal muscle, tissue-specific microparticles) for faithful recapitulation of the in vivo microenvironment and organoid-based clinical therapeutics is worthy of attention [ 63 , 229 , 230 ]. Taken together, the structural and functional maturity of organoids depend on the continuous progress of the crucial steps, including the regulatory network for developmental patterning, culture medium and construction formulations for proper terminal cell differentiation, and cell aggregation embedded in ECM for 3D structure formation [ 20 ].

Overall, organoids with its unique advantages has profoundly revolutionized our understanding of human development and disease mechanisms, bringing about a paradigm shift in drug discovery and clinical diagnosis (especially in oncology and precision medicine). Despite there’s still a need for continuous breakthroughs in maturity, standardization, cost, and ethical regulation, its potential is enormous and promising. With the deep integration of multidisciplinary technologies (e.g., bioengineering, microfluidics, gene editing, single-cell genomics, machine learning and AI), the next generation of more functional and humanized organoid models will vastly accelerate the progress of basic research and translational medicine, and ultimately promoting the transition to a "patient-centered" future medical model.

For decades, a variety of organoids have been exploited for congenital disease diagnosis and treatment, including neuromuscular organoids and skeletal muscle organoids (SMOs) for Duchenne muscular dystrophy (DMD) [ 63 , 175 ], liver organoids for congenital metabolic liver disease [ 176 ], striatal organoids (SOs) for Huntington’s disease (HD) [ 49 ], cerebrocortical organoids (COs) for familial Alzheimer's disease (AD) [ 177 ], intestinal organoids for cystic fibrosis (CF) [ 178 ], ROs for Stargardt disease (STGD) [ 179 ], duodenal organoids for familial adenomatous polyposis (FAP) [ 180 ], kidney organoids for autosomal recessive polycystic kidney disease (ARPKD) [ 57 ], testicular organoids for adult Klinefelter syndrome [ 181 ], colonic organoids for Hirschsprung disease (HSCR) [ 182 ], and microglia-sufficient brain organoids for hereditary diffuse leukoencephalopathy with axonal spheroids (HDLS) [ 183 ] (Table  4 ). DMD is a severe muscle-wasting disease caused by mutations in dystrophin, which eventually leads to assisted ventilation and premature death [ 175 , 184 ]. Auletta et al. created tissue engineered-neuromuscular organoids (t-NMOs) from DMD patient-specific iPSCs, and demonstrated the feasibility for mimicking the cellular phenotype upon neurotransmitter stimulation, including calcium dysregulation and reduced muscle contraction [ 63 ]. To verify the pathogenesis of duplication 15q (dup15q) syndrome, the dup15q patient-derived hiPSCs were employed for cortical organoid induction to illustrate disease-associated modules involved in metabolic dysregulation, synaptic dysfunction, altered neuron projection, and neuron hyperexcitability [ 185 ]. Currently, the dysfunction of HSF1 and NEDD4L proteins in hiPSC-derived striatal organoids were verified, which disrupted mitochondrial function during lipoic acid metabolism in Huntington’s disease (HD), respectively [ 49 , 55 ]. As to familial Alzheimer's disease (AD), diverse heterozygous familial AD mutations (e.g., PSEN1 ΔE9/WT , PSEN1 M146V/WT , APP swe/WT ) in hiPSC-derived AD cerebrocortical organoid (CO) model were identified, which suggested the novel mTOR inhibitor-independent drug candidate for drug testing [ 177 ]. Taken together, these data emphasize the robust prospective of PDOs for congenital disease modeling and personalized therapeutic regimen formulation. Meanwhile, with the advent of lineage tracing and genetic-engineering strategies, pathogenic genes and mutations in inheritable genetic disorders can be directly verified and corrected in organoids, which thus enables the mechanistic basis and therapeutic reagent discovering of genetic diseases including congenital diseases.

Over the years, the miniature organoid models by natural cultivation or artificial construction have garnered significant attention for personalized healthcare purposes from bench to bedside. Considering the physiological and pathological variations in tissue-specific microenvironment, the translational trajectory of organoid-based biotechnology have been partially constrained by insufficient recapitulation of the tissue architecture and functionality [ 76 ]. Herein, we outline the key constituents in ecological niche mimicry for organoid construction from the aspects of ECM, cytokine- and compound-mediated signal transduction, culture medium, and biophysical stimulation (Table  2 , Supplementary Table S1).

Infectious diseases are caused by pathogens (e.g., bacteria, viruses, and other microorganisms) that multiply within the host and have far-reaching effects upon pathogen-host interaction and physical health [ 167 ]. Current advances in organoids provide groundbreaking alternatives for infectious disease modeling and mimicking the complex host–pathogen relationships such as tissue- or hiPSC-derived lung organoids for tuberculosis [ 168 ], liver organoids for hepatitis B virus infection [ 167 ], kidney organoids for SARS-CoV-2 infection [ 116 ], gastric organoids for helicobacter pylori infection [ 169 ], skin organoids for enterovirus 71 (EV-A71) infection [ 170 ], and human thymus and spleen organoids for human immunodeficiency virus (HIV) infection [ 171 ]. For instance, hPSC-derived immuno-lung organoids were developed for mimicking macrophage-mediated tissue damage and THBS1-mediated lung cell senescence during SARS-CoV-2 infection [ 172 ]. Meanwhile, investigators also developed a kidney organoid model with enhanced metanephric specification to proximal tubules as well as mature nephron patterning and segmentation, which enabled toxicity screening for infectious diseases like SARS-CoV-2 [ 116 ] (Table  4 ). With the aid of the brain-region-specific forebrain organoid platform, researchers modelled Zika virus (ZIKV) exposure and demonstrated the productive, preferential infection of neural progenitors, which were confirmed by neurospheres and brain organoids [ 11 , 173 ]. Interestingly, current advances also advocated the involvement of organoids for examining co-infections caused by various pathogens (e.g., bacteria, parasites, and viruses) [ 174 ]. As exemplified by the aforementioned infectious diseases, organoids are preferable to animal models owing to the narrow species and tissue tropism of specific pathogens [ 20 ]. Collectively, the assembly of organoid infection models and numerous microbiome strains are valuable for simulating pathogen-host interaction and disease pathogenesis of infectious diseases and advancing personalized therapeutics. Considering the outstanding advantages of organoid-based models and the sudden occurrence of epidemics and pandemics, it’s absolutely necessary to develop comprehensive strategies by integrating with the microfluidic device (e.g., OoC) and multi-omics for therapeutic target prediction and high-throughput pharmaceutical drug testing [ 20 ].

State-of-the-art literatures have highlighted the limitations of static organoids in faithfully recapitulating disease pathology attribute to the requisite biophysical microenvironment during disease progression [ 99 ]. To mimic the dynamic mechanical forces in the intestinal microenvironments, cyclic stretch was applied to intestinal organoid cultures, and the remarkable expansion of SOX9 + intestinal progenitors and activation of Wnt/β-Catenin signaling were further demonstrated [ 32 ]. Instead, with the aid of microfluidic system, two mechanosensing molecules for suppressing cyst formation as well as 229 signal pathways differing from those in the static models were illuminated [ 99 ]. Nowadays, multidimensional integration of biophysical and biochemical signals have been extensively adopted for modulating organoid generation and cell-fate determination. For the generation of the nonrenewable and specialized mechanosensitive hair cells with inner ear-like accessory structures, a rotary cell culture system (RCCS) was employed for deriving vestibular tissue-like organoids from hESCs (e.g., H3, H9) and hiPSCs (e.g., 007-5), which highlighted the advantages of microgravity-derived organoids over those in static organoid culture systems [ 85 ]. Similarly, neural organoids were induced from hESCs based on the utilization of the RCCS, which showed the bias to the midbrain-hindbrain fate [ 100 ]. Furthermore, the rescue effect of simulated microgravity intervention upon neuropathological deficits in brain organoids was identified by modulating the YAP/BMP/ID1 axis [ 101 ]. Taken together, these findings confirmed the incorporation of mechanical stretch and biochemical stimulation for optimizing the organoid induction system, and paved the way for understanding the cross-talk between diseases and mechanical factors [ 32 ]. Moreover, the biophysical and biochemical stimulation could collaborate with culture supplements in the ecological niche mimicry, and thus further accelerates the formation of physiological structure and functional maturation of organoids.

Bioreactors and RCCS possess multidimensional preponderances for long-term cell culture and large-scale organoid preparation [ 106 ] (Fig.  2 B, Table  2 , and Supplementary Table S1). To date, a variety of stem cell-derived organoids have been manufactured by utilizing bioreactors such as microspheric skin organoids [ 107 ], blood vessel organoids [ 108 ], brain organoids [ 109 ], hepatoblast organoids [ 110 ], epithelial organoids [ 111 ], and inner ear organoids [ 85 ]. For example, a simple stirred bioreactor (SBR) platform by 3D-printing was employed for facilitating retinal organoid production via improving physiological oxygen concentration and nutritional exposure [ 90 ]. For the induction of forebrain-specific organoids, a miniaturized spinning bioreactor (SpinΩ) was developed, and the generated cerebral organoids recapitulated the key characteristics of human cortical development [ 11 ]. Recently, investigators turned to an orbital shaker (CS-LR; TAITEC) for the continuous culture of hiPSC-derived cell aggregates for HOX-positive neural crest cell (NCC) and mandibular prominence (mdEM) induction in Jawbone-like organoids [ 46 ]. As described in the literature, inner ear organoids were derived from hPSCs by the dynamic 3D RCCS, which resembled fetal inner ear hair cells and accessory otoconia-like structures [ 85 ]. 3D bioprinting by cutting jigs and arraying method was employed for fabricating four classes of organoid from hESCs in mini-spin bioreactors [ 106 ]. For the non-invasive quality control of brain organoid culture, a mesofluidic bioreactor device was induced, which supplied reproducible culture standards for long-term and robust organoid preparation [ 109 ]. Therefore, differing from those static culture-based methodologies, bioreactors and the rotating culture system serve as effective platforms for large-scale, around-the-clock, automated preparation of organoids and the derivatives (Table  3 ).

Currently, a variety of technologies have been developed for the construction of organoids, including the 2D monolayer cell culture, the embryoid body (EB), rotating culture in mini-bioreactor, 3D bioprinting-based tissue engineering, organ-on-a-chip systems and microfluidics, which have been designed as a multiplex platform for automating the culture of individual organoids in distinct microenvironments [ 102 , 103 ]. Meanwhile, the integration of organoids and multidisciplinary cutting-edge technologies has significantly advanced the methodology of organoid construction such as gene editing and lineage tracing, multi-omics technology, machine learning and artificial intelligence (AI) technology (Fig.  2 A–D, Table  2 , Supplementary Table S1). Fig. 2 Methodology and multidisciplinary technologies for organoids. A-D The illustration shows the construction methods for organoids, including 2D monolayer culture ( A ), the static and rotary 3D suspension culture ( B ), 3D bioprinting ( C ), and organ-on-a-chip ( D ), integration of multidisciplinary technologies (e.g., gene editing, lineage tracing, multi-omics, machine learning) ( D ) Methodology and multidisciplinary technologies for organoids. A-D The illustration shows the construction methods for organoids, including 2D monolayer culture ( A ), the static and rotary 3D suspension culture ( B ), 3D bioprinting ( C ), and organ-on-a-chip ( D ), integration of multidisciplinary technologies (e.g., gene editing, lineage tracing, multi-omics, machine learning) ( D )

Organoids are self-organized and 3D in vitro cell aggregates that highly simulate the biological structure and molecular characteristics of the corresponding organs, and thereby play a pivotal role in developmental biology and disease conceivable [ 1 , 2 ]. Since the pioneering attempt at dissociated sponge cells for in vitro organism regeneration by Wilson in 1907, investigators have been devoted to simulating human organs, and groundbreaking discoveries have been achieved by utilizing the 3D organoids [ 3 , 4 ]. To date, a variety of organoids have been developed for multiple aspects of bioresearch and regenerative medicine, including morphogenesis, metabolism, pharmacology, tumorigenesis and immunology, and tissue engineering (e.g., vascular remodeling) [ 5 – 7 ]. Unlike the traditional two-dimensional (2D) cell culture models, the 3D organoids recapture the multitudinous cellular architecture and resemble embryonic organs both at the cellular and molecular levels, which thus supply novel strategies for simulating organogenesis and developmental disorders [ 8 ]. For example, Qian and the colleagues highlighted the current advances in brain organoid methodologies and compared with embryonic human brain in recapitulating multifaceted key features of the early brain development [ 9 ]. As a remarkable branch of tissue engineering technology, organoid technologies leverage the self-organizing capacity of stem cells and adult tissue derivatives to assemble in a 3D culture environment “in a dish” [ 5 , 8 , 10 ]. Nowadays, organoid technology have provided a versatile and accessible platform for precision medicine and regenerative medicine purposes. For example, organoid technology is extensively applied in modeling human early organ development and disease progression, first-in-class drug discovery, potential compound testing, and preparation of diverse kinds of functional cells (e.g., cardiomyocytes, islet cells and retinal cells) [ 7 , 11 – 14 ]. Although animal models are extensively employed for validating drug efficacy and uncovering underlying mechanisms, yet the inherent limitations in genetic heterogeneity with humans in terms of the mimicry of organ microenvironment further hinder their application [ 15 ]. For example, patient-derived xenograft (PDX) models are capable of partially recapitulating the tumor environment, yet cannot fulfill the large-scale drug screening due to the inherent defects in immunocompatibility with human-specific immune components, time-consuming and high maintenance costs [ 16 – 19 ]. Differing from the model organisms with diverse limitations (e.g., ethical restrictions, technical difficulty, time/money-consuming, and variations in predictive value), stem cell-derived organoids enable the scalable and reproducible modeling of key aspects in human embryos, including gastrulation, germ layer formation, and organ-specific cell type creation [ 7 ]. While PDX models remain valuable for immune-tumor interaction studies, patient-derived tumor organoids (PDTOs) offer advantages in scalability and patient-specific drug testing, albeit with limitations in systemic immune modeling. Meanwhile, current literatures have also pointed out diverse identified biological processes, which are specific to human body and thus cannot be mimicked in animal models including PDX models [ 20 ]. Therefore, organoid technologies have made it attainable to study the physiological and pathological organogenesis in vitro exempt from model organisms (e.g., PDX models) [ 21 , 22 ]. For example, the patient-specific cerebral organoids have been proven competent for recapitulating the complexity of brain development and brain disorders [ 23 ]. With the aid of organoid fusion technology, Bagley and the colleagues revealed the complex interactions between diverse dorsal–ventral brain regions [ 24 ]. Similarly, the faithful recapitulation of the development of human medial ganglionic eminence (MGE) and cortex domains via fusion of regionally specified brain organoids from human pluripotent stem cells (hPSCs) was also accessible [ 25 ]. In this review article, we mainly outline the key advances in organoid-guided organogenesis, including cell sources, ecological niche mimicry in static and bioreactor-based rotating culture, and construction strategies integrated with novel technologies (e.g., organoid-on-a-chip, gene editing and lineage tracing, multi-omics technology, machine learning and artificial intelligence). After that, we highlighted the application of organoids in precision medicine such as disease modeling (e.g., cancer research, infectious diseases, congenital diseases, and organ-specific diseases), drug innovation (e.g., high throughput screening, drug toxicity assessment, individualized target treatment), and the concomitant registered clinical trials. Furthermore, we highlight the diverse opportunities and challenges for organoid-based investigations such as immune-tumor interactions, cancer immunotherapy, reproducibility and standardization, ethical considerations and the emerging interdisciplinary integrations (Fig.  1 A). Collectively, the 3D organoids with favorable architecture and microenvironment have become a promising area of biomedicine for personalized therapeutics. Fig. 1 The workflow and cell sources for organoid construction. A. The illustration shows the workflow. B. Representative cell sources for generating organoids, including hESCs, hiPSCs, hASCs, patient tissue, normal tissue derivatives, fetal tissue, and cell lines. C. The illustration shows the germ layer specification of hPSCs for organoid construction The workflow and cell sources for organoid construction. A. The illustration shows the workflow. B. Representative cell sources for generating organoids, including hESCs, hiPSCs, hASCs, patient tissue, normal tissue derivatives, fetal tissue, and cell lines. C. The illustration shows the germ layer specification of hPSCs for organoid construction

ECM serves as the microenvironment for the generation of 3D spheroids and the following maintenance and maturation of organoids [ 29 ]. On account of the delicate structure and small size, organoids are commonly embedded into supporting hydrogels, which are catalogued into natural hydrogels (e.g., polysaccharide hydrogels, protein hydrogels, and decellularized ECM hydrogels) and synthetic hydrogels (e.g., polyethylene glycol (PEG), poly lactic-co-glycolic acid (PLGA), RADA 16, and polycaprolactone (PCL)) [ 37 , 77 – 79 ]. During the past decades, diverse types of ECM have been explored for organoid generation, including Matrigels for hindgut spheroids [ 80 ], synthetic hydrogels for HIOs [ 81 , 82 ], alginate for small intestinal organoids [ 83 ], PEG hydrogel for intestinal epithelial organoids (IEOs) [ 84 ], collagen matrix and chitosan matrix for female reproductive tract organoids [ 37 ], Cultrex BME for hFKOs [ 34 ], and filamentous matrix for vestibular tissue-like organoids [ 85 ]. For instance, Matrigel was used as basement membrane ECM for the subsequent culture and maturation of 3D hindgut spheroids, while hydrogels with an embossed surface and the all-in-one platform were employed for adipose-derived stem cell spheroid culture and production [ 80 , 83 ]. Of note, both Matrigel and Cultrex BME have inherent limitation, which derive from mouse Engelbreth-Holm-Swarm (EHS) sarcoma and suffer from their animal origin and batch-to-batch variability. For optimized HIOs induction, directed differentiation protocols with defined alternatives to Matrigel have also been involved such as alginate and polyethylene glycol hydrogels [ 83 , 86 ]. Aiming to benchmark PSC-derived organoids, investigators turned to the commercial basement membrane extract Cultrex BME and Matrigel for long-term culture of hFKOs [ 34 ]. Meanwhile, ECM exposure was adopted for VOs maturation and the generation of larger and structured vessels in immunodeficient mice, while in a loose filamentous matrix was employed for mimicking otoconial membrane and otoliths during mature inner ear organoid induction [ 85 , 87 ]. For programming iPSCs into islet-like organoids, type V collagen (ColV) and collagen type II (COL2) in the decellularized ECM hydrogel were introduced as natural and biological niche for enhancing glucose responsive secretions, respectively [ 88 , 89 ]. As to retinal organoid (RO) induction, the polyHEMA and BD GFR Matrigel were employed for EB formation and OV generation, repsectively [ 90 ]. Interestingly, Pendergraft and the colleagues introduced decellularization of human testis tissues for ECM extraction and the generation of testicular organoids within 23 days [ 70 ]. Additionally, fibrinogen-encapsulated patient tissue-derived cells were reported for vessel formation in vascular osteosarcoma organoids, while decellularized skeletal muscles (dSKMs) were implemented as scaffolds for generating tissue-engineered neuromuscular organoids from hiPSCs [ 63 , 71 ]. However, the distribution of the embedded organoids in Matrigel or relative biomaterials is random and uncontrollable. To address this issue, the acoustic micromanipulation platform with a polyethylene glycol diacrylate (PEGDA)-gelatine hydrogel to levitate organoids was introduced to a prescribed positions [ 77 ]. It’s noteworthy that the in vivo organoids are also available for the mass preparation of therapeutic cells for regenerative medicine purposes. For example, the biomaterial-assisted in vivo osteo-organoids in ossicle-bearing NSG mice for the expansion and differentiation of hematopoietic stem/progenitor cells (HSPCs) were outlined, which provided a novel platform for modeling physiological hematopoiesis and hematological diseases [ 29 , 91 ]. Interestingly, investigators further underlined the potential of integrating organoids with 4D alginate hydrogel systems for the tissue engineering innovation [ 92 ]. Collectively, these in vivo organoid technologies supply relatively independent microenvironment for the acquisition of lineage-specific primitive stem cells and differentiated functional cells. Matrigel of animal origin and relative commercial biomaterials have been vastly applied for fundamental research as ECM for organoid establishment, while the decellularized tissues with organic compatibility are cost-effective and could faithfully simulate the in vivo microenvironment including TME.

Organoid models with multiple cell types face particular challenges with reproducibility due to the complexity in unstandardized generation procedures and validation methodology (e.g., cell lineages, culture protocols and environments) [ 147 , 238 ]. Despite the robust progress in PDTO-based precision medicine, yet the challenges in relation to standardizing diverse organoid models for clinical implementation should be urgently resolved [ 155 ]. In consequence, there’s an urgent need for the setup of large-scale organoid biobanks [ 22 , 239 ]. For instance, researchers reported the establishment of STBS organoid biobank representing eight subtypes from patient tumor tissues. These organoids in the biobank faithfully recapitulated multiple aspects of parental tumors and enabled the personalized sarcoma treatments on the basis of the inter- and intra-tumoral heterogeneity [ 17 ]. The large-scale organoid biobanks for nanoparticle drug development and the long-standing in tumor modeling were further highlighted [ 22 ]. Meanwhile, the integration of organoid technologies with the relevant microphysiological systems (MPS) will robustly facilitate the standardization and reproducibility of organoid production for precision medicine [ 238 , 240 ]. For instance, Mohapatra et al. put forward the proof-of-conception guidance for good in vitro reporting standards (GIVReSt) for various cell culture methods, including stem cells and organoids [ 147 ]. Simultaneously, researchers further emphasized the involvement of engineering principles for increasing reproducibility of organoids. For example, organ chip for dynamic parameter monitoring in different layers of organoids and the design of matrices for spatiotemporal shape-guided morphogenesis, respectively [ 115 , 241 ]. Additionally, organoid-derived extracellular vesicles (OEVs) are promising communication medium for stimulating the function and structure of corresponding organs on the basis of their significant physiological effects [ 1 ]. For example, Liu et al. verified the significant physiological effects of intestinal OEVs for gut and bone communication [ 1 ].

The patient tissue-specific organoids have greatly accelerated the identification of biomarkers, individualized target treatment and the concomitant precision medicine (Fig.  3 ). For instance, FOXG1-initiated gene network modules are correlated with symptom severity of severe idiopathic ASD by orchestrating the shift toward GABA/Glutamate neuron fate in proband-derived telencephalic organoids [ 217 ]. With the aid of hiPSC-derived ROs and scRNA-SEQ analyses, the principal histological hallmarks (e.g., c.380G > A mutation in exon 4, IVS5-1G > C transversion in intron 5) and adeno-associated virus 9 (AAV9)-mediated tripeptidyl peptidase 1 (TPP1) gene therapy for neuronal ceroid lipofuscinosis type 2 (CLN2) were demonstrated [ 218 ]. Similarly, the generation of ROs from inherited retinal dystrophies (IRDs) patient-derived hiPSCs with the heterozygous missense variant (c.209G > A) in the ARL3 gene was described [ 219 ]. Aiming to illuminate the causative gene for congenital ichthyoses, investigators derived skin organoids from hESCs with pathogenic variants on KLF4, and identified the heterozygous missense variants, including c.1323 T > A (p.Asp441Glu) and c.1322A > G (p.Asp441Gly) [ 220 ]. As to gastric cancer (GC), the application of GC organoids for investigating immune evasion in TEM and facilitating novel drug screening for individuals was highlighted [ 221 ]. Therefore, by combining with gene-editing technologies, organoid-based platforms and the derivatives hold the permission for individualized target treatment. Even though the promising prospects for personalized medicine, yet the concomitant safety and ethical risks of organoid-based targeted therapy should be treated with caution.

PDOs are advantageous ex vivo disease models composed of diverse cell types for traversing the stepwise stem cell-to-enterocyte differentiation trajectory and transient morbid states during organogenesis and clinical therapeutics [ 10 , 163 ]. For instance, Ariño et al. reported the construction of biopsy-derived liver organoids (b-Orgs) for recapitulating liver epithelial heterogeneity in alcohol-associated liver disease (ALD) [ 186 ]. Instead, combined with cell reprogramming technology, the fused dorsal–ventral cerebral organoids offered the feasibility to dissect the complex neurodevelopmental defects of neurological disease patients [ 24 ]. The pancreas functions indispensable in proper nutrient metabolism, and the dysfunctions caused by mutations in pancreatic beta cells usually result in metabolic disorder such as mature-onset diabetes of the young (MODY) and neonatal diabetes (ND) [ 2 , 56 ]. Differing from model organisms, pancreatic organoids serves as preferable in vitro model for understanding the mechanisms governing the 3D morphogenesis and organogenesis of pancreas [ 187 ]. Gut organoids closely replicate the properties of the gut, and act as powerful tool for intestinal disease modeling, personalized medicine and microenvironmental simulation [ 4 ]. For instance, the well-established patient tissue-derived HIOs reveals high consistency and maturity for modeling diverse intestinal diseases, including Crohn's disease (CD), ulcerative colitis (UC) and intestinal cancers [ 6 ]. By employing the kidney organoids as recently reported [ 67 ], ARPKD organoids were performed for the dissection of PTH1R/cAMP-pPKA-pCREB axis-mediated pathomechanisms during developmental polycystic kidney disease [ 57 ]. Additionally, investigators in the field further showed the latest updates of hPSC-derived inner ear organoids for investigating the causes of hearing loss and auditory regeneration [ 61 , 85 ]. Vasculopathy has contributed to diverse disorders such as hemangioma, cardiovascular diseases, cerebral ischemia, and diabetic complication (e.g., lower limb ischemia) [ 188 ]. VOs have offered new avenues for dissecting vascular abnormalities and vascular regenerative medicine [ 188 ]. For example, VOs have been reported with efficient co-differentiation of endothelial and mural lineages, and can further mature into perfused vasculature for revascularization in animal models including pancreatic islet transplantation and hind limb ischemia [ 87 ]. Instead, arteriovenous malformation (AVM) blood vessel organoids from patient-specific hiPSCs were established for AVM modeling and pharmacological testing with thalidomide, U0126, and rapamycin [ 189 ]. In brief, numerous organ-specific organoids have emerged as powerful platforms for disease modeling and patient-specific pathogenesis (Table  4 ). Differing from the traditional ex vivo models and transgenic animal models for disease modeling, organoid models bridge the remaining gap between human organs and animal models owing to the fast and robust outcomes, high success rate, and scalability in microbiota and pathogen composition.

Adult tissue-derived stem cells (ASCs) have been extensively explored for patient-specific organoid generation, which are responsible for disease remodeling [ 68 ] (Fig.  1 B, Table  2 ). For instance, adult intestinal stem cells possess the remarkable features of the intestinal epithelial lining, and the derived intestinal organoids are widely applied for investigating the regeneration, homeostasis and cancers in the intestine [ 32 ]. Compared with the traditional models, intestinal organoids that assembled MSCs and intestinal epithelial cells could accurately simulate the dynamic differentiation and development bioprocesses of the intestinal crypts [ 69 ]. MSCs are also introduced for hematopoietic organoid construction via self-organization and mapping the early steps of bone repair within bone organoids [ 30 , 48 ]. Similarly, the isolated human spermatogonial stem cell (SSC), Leydig cells, and Sertoli cells were synchronously resuspended and embedded in decellularized human testis extracellular matrix (ECM) for the induction of testis organoids [ 70 ]. Meanwhile, the biopsies and surgically resected tumor tissue are available for preparing organoids. For example, sarcoma organoids were generated from minced tumor tissues cultured in advanced DMEM/F12 (aDMEM/F12) medium supplemented with cytokine cocktails for 2 weeks dispense with cell dissociation or exogenous ECM [ 17 ]. Additionally, viable patient-derived cells (PDCs) via tissue digestion and centrifugal resuspension have also been reported for generating PDTOs such as human embryonic lung fibroblasts (e.g., IMR90), endothelial cells (ECs) and osteosarcoma cells (e.g., MNNG/HOS) [ 71 ]. Additionally, diverse commercially available cell lines isolated from diverse body tissues are reported for diverse organoid induction such as BTS11 and BTS5 for trophoblast organoids, LIM1863 for colon cancer organoids, and BT-474 for breast cancer organoids [ 37 , 72 , 73 ]. Simultaneously, the fetal tissues and the extraembryonic placental tissues are also advantageous sources for organoid preparation (Fig.  1 B, Table  2 ). The establishment of human fetal kidney-derived organoids (hFKOs) from 15 to 20 weeks’ fetal kidney tissue-dissociated cells confirmed the fetal tissue-specific cells for organoid generation [ 34 ]. Meanwhile, placental tissues have also been considered as advantageous sources for long-term trophoblast organoid generation, which were adequate to differentiate into both extravillous trophoblast and syncytiotrophoblast [ 35 , 74 ]. Similarly, Haider et al. reported the self-renewing trophoblast organoids from first-trimester cytotrophoblasts (CTBs), and recapitulated the formation of syncytiotrophoblast (STB), trophoblast progenitors and differentiated subtypes, and invasive extravillous trophoblast (EVT) [ 75 ]. Overall, these findings consistently confirm the feasibility and convenience of utilizing diverse tissue-enrich stem cells and the derivatives in the original tissues for organ-specific organoid formation (Fig.  1 B, Table  2 ). Of note, ASCs and fetal tissue-enriched stem cells with the key features of stem cells can largely fulfill the demands for patient-derived organoid (PDO) and PDTO formation, while the tissue derivatives reveal preferable safety and plasticity for therapeutic purposes in clinical practice.

In general, 2D monolayer culture serves as the most common protocol for initiating organoid induction from hPSCs or ASCs such as IEOs and retinal organoids (ROs) [ 83 , 90 ]. Subsequently, suspension culture protocols are common for EB formation and the classic 3D organoid generation [ 26 ] (Fig.  2 A, Table  2 , and Supplementary Table S1). For example, the induction of IEOs with hydrogel scaffolds by utilizing the 2D monolayer culture was achieved, which remained preferable organization for tissue engineering over the 3D spheroidal architecture [ 83 ]. For directed differentiation into HIOs, hiPSCs were incubated in 2D culture for self-organization and 3D spheroids followed by embedding in Matrigel for described differentiation paradigm [ 104 ]. As to telencephalic organoid induction, HMGU1 cells were first seeded in the 2D Matrigel-coated 24 well-plates for 7 days, followed by dissociation and centrifugation in 96 U-bottom well plates for the 3D EB formation in Matrigel [ 95 ]. In general, static culture is appropriate for adherent monolayer cells and the early differentiation initiating stage of the subsequent organoids. The 2D culture models are instrumental in exploring fundamental molecular mechanisms. However, these organoid models possess inherent defects in accurately replicating the intricate 3D architecture and dynamic microenvironment characteristics [ 105 ] (Table  3 ). Table 3 The pros and cons of different platforms for organoid induction Items 2D cell culture 3D static model Bioreactors and RCCS 3D bioprinting Organoid-on-a-chip Ease of model establishment  + +   + + +   + + +   + + +   + + +  Ease of maintenance  +   + +   + +   + +   + +  Duration of experiments  + + +   + +   + +   +   + +  Recapitulation of bioprocess  + +   + + +   + + +   + + +   + + +  Ease of genetic manipulation  + + +   + + +   + + +   + + +   + + +  Complexity of structure  + +   + +   + +   + + +   + + +  High-throughput drug screening  +   +   + + +   + + +   + + +  The cost of the model  + + +   + + +   + +   + +   + +  (Note: +, low degree; + +, middle degree; + + +, high degree) The pros and cons of different platforms for organoid induction (Note: +, low degree; + +, middle degree; + + +, high degree)

The well-established 3D cell culture is extensively employed for cultivate cells or tissues, which provides possibilities and ex vivo models for mimicking the authentic tissue architecture and biofunctionality [ 31 ]. In recent years, diverse 3D culture methods and direct differentiation protocols have been developed and implemented for high-efficient organoid induction [ 83 , 104 ] (Fig.  2 B, Table  2 , Supplementary Table S1). Of them, the 3D organoid culture technology enables the development of complex, organ-like tissues recapitulating many aspects of in vivo organogenesis [ 24 ]. Distinct from the 2D monolayer arrangements on flat surfaces, the 3D organoids are floating self-organizing cell aggregates and thus could better simulate the tissue architecture and cellular diversity of the corresponding organs [ 83 ]. To date, EB models have been widely used for spontaneous differentiation and organoid induction despite with limited control over their lineages and cellular composition [ 87 ]. For instance, cerebral organoids with a cerebral cortex and mature cortical neuron subtypes were induced from hPSCs by utilizing the 3D organoid culture system [ 23 ]. Of note, a novel construction of dorsal–ventral cerebral organoids was reported by infusion of the separately and individually patterned EB derivatives in a single Matrigel droplet based on the dorsal CycA , dorsal Unt , or ventral protocols [ 24 ]. Collectively, the 3D EB suspension culture has served as the default method of the main organoid models, whereas the potential obstacles for tissue-engineering purposes due to the 3D spheroidal architecture should not be neglected [ 83 ]. Additionally, compared those by 3D bioprinting, the spontaneous EB-based 3D organoid models reveal considerable inhomogeneity, randomization and uncontrollability (Table  3 ).

Organoid-on-a-chip technology, also known as microphysiological system or organ chip, has offered novel avenues for exploring organogenesis and disease modeling [ 119 , 120 ]. The organ chip microfluidic devices embed into living cells are adequate to recapitulate the whole-body inter-organ physiology or pathophysiology with high fidelity, which thus enable the spatiotemporal surveillance of mechano-physiological parameters and improve accessibility of functional readouts in organoids [ 115 , 121 , 122 ] (Fig.  2 D, Table  2 , Supplementary Table S1). OcO models have been vastly involved in diverse organoids for substance exchange and signal visualization, including vascularized organoids [ 123 ], brain organoids [ 124 ], tubular biliary organoids [ 125 ], lung organoids [ 126 ], intestinal organoids [ 42 ], colon organoids [ 127 ], taste organoids [ 128 ], and diverse tumor organoids [ 129 , 130 ]. For instance, a brain OoC platform was introduced for dissecting the signaling gradients in multiple domains of topographic forebrain organoids [ 119 ]. Instead, an innovative vascularized tumor organoid-on-a-chip platform was developed for exploring tumor-vascular dynamics and the concomitant anti-vascular drug efficacy, which revealed hierarchical, tumor-specific microvasculature [ 71 ]. Interestingly, two molecules (RAC1 and FOS) of distal nephron dilatation were identified by utilizing the PKHD1-mutant organoids-on-a-chip [ 99 ]. In recent years, an advanced “min-colons” organoid model by integrating OoC technology and colon organoids were reported, which revealed in vivo-like patterning and diversity of differentiated cell types and incorporated microenvironmental components [ 127 ]. Taken together, the bioengineered organoids with organ chip technology provide precise platforms for systematically exploring human organ physiology and pathology. Differing from the 3D bioprinting organoids with bioethical and legal issues as well as deficiency in novel biomaterial innovation, OoC technology acts as a next-generation platform that can bypass the indicated constraints and allow the establishment of multiple biomimetic organ models for novel drug pharmacokinetics [ 118 ] (Table  3 ).

Additional file 1. Additional file 2. Additional file 3. Additional file 4. Additional file 1. Additional file 2. Additional file 3. Additional file 4.