The stomach, a crucial organ for food digestion, is a complex system consisting of four parts: cardia, fundus, corpus, and antrum (pylorus) (1). The cardia connect stomach and esophagus which allows food flow from esophagus to stomach followed by the fundus. The corpus, the main site for the secretion of acid and digestive enzymes, is vital for proper digestion. Meanwhile, the antrum plays a crucial role in mucus and hormone secretion. In mice, the stomach has an additional forestomach that functions to store and mechanically dissociate food (2, 3). Since the stomach is exposed to various harmful substances, such as toxins, and bacteria, it generates a hostile environment for stomach (4). Therefore, maintaining the mucosa’s integrity and functionality is crucial, necessitating continuous epithelium self-renewal (3, 5, 6).

The stomach’s mucosa comprises gland-organized epithelial layers divided into four regions: base, neck, isthmus, and pit (7). The corpus epithelium is composed of long glands with short pits that habour mucus-secreting pit and neck cells, parietal cells, endocrine cells, and chief cells that secrete digestive enzymes. The antral glands, in contrast, are shorter and have larger pit regions. They contain fewer chief cells and no parietal cells but have mucus-secreting pit and neck cells and endocrine cells. The isthmus region, known to host gastric epithelial stem cells, contains proliferating cells. Mucus pit cells are exclusively found in the pit region, and mucus neck cells are located in the neck region, while chief cells are present in the base region only. The endocrine cells found in both regions of the stomach secrete hormones (Fig. 1) (8).

Figure 1. The epithelial architecture of the stomach. It is illustrating the intricate cellular arrangement and organization within the corpus epithelium, emphasizing the presence of distinct layers and cell types. ECL: enterochromaffin-like.

Organoids are three-dimensional in vitro models that mimic the in vivo characteristics of diverse living organs (9). These in vitro models are developed from stem cells or disease tissues extracted from patients, and are cultivated in a specific growth factor combination which mimics in vivo signalings regulating stem cell homeostasis and disease condition (10). Organoids can be stably maintained for a long period of time, making them a promising tool for wide range of biology from development to adult stem cell (AdSC) homeostasis, and disease modeling. They are commonly derived from various types of origin, including embryonic stem cells (ESCs), induced pluripotent stem cells (iPSCs), AdSCs, and patient-derived tissue (10). Notably, cancer patient-derived organoids (PDOs) have been successfully constructed from various types of cancer, including breast, ovarian, lung, gastric, prostate, bladder, liver, esophageal, pancreatic, neuroblastoma, glioblastoma, and colorectal cancer (11). Tumor organoids have become commonly used tools in oncology research, including studies on cancer initiation and progression, preclinical models, explorative and personalized therapies, and drug screening. Moreover, recent announcement for a new legislation replacing the current use of animal model for the safety test of drug by the US Food and Drug Administration which eliminates the requirement for animal testing before clinical trials support the value of organoid for drug development (Fig. 2) (12).

Figure 2. Overview of gastric organoid establishment and its applications. Adult stem cell-derived organoid established from healthy individual biopsy or patients. Pluripotent stem cell-derived organoids utilize embryonic stem cell (ESC) or induced pluripotent stem cell (iPSC) for organoid establishment which follow the stomach development process. The successfully established organoids are maintained in the extracellular matrix with organ-specific growth factor supplements and stored at an organoid biobank. The organoid biobank has diverse applications, such as gene editing, which can be further utilized for patient organoid therapy, drug screening and development for personalized medicine, disease modelling, basic research, etc.

Therefore, organoids are promising tools in stem cell biology, disease research and drug discovery, and their potential applications are enormous (13). The establishment and maintenance of PDOs requires attention in multiple aspects. For the PDOs establishment, it requires attention on tumor purification, genetic profiles, culture medium and matrix. To maintain PDOs, morphological characteristics, passging, and storage. Co-culture systems, drug sensitivity test, and high-throughput drug screening can be considers for application of PDOs. The success of organoid technology in mimicking the characteristics of real organs in vivo, coupled with the recent legislation allowing for alternatives to animal testing, provides exciting opportunities for the advancement of cancer research and drug discovery.

Organoids, remarkable three-dimensional structures, can be derived from two distinct sources of stem cells: AdSCs and pluripotent stem cells (PSCs), encompassing iPSCs and ESCs (9). AdSC-derived organoids have the ability to establish and preserve an intact stem cell niche and niche factors are supplemented in the culture medium (14). Stem cells within organoid cultures maintain self-renewal capacity and often possess unlimited proliferative potential. Depending on the supplement of growth factors in the culture medium, cells within AdSC-derived organoids can differentiate into diverse cell types within the epithelium of the tissue from which they originate with different cellular composition (15, 16). Conversely, PSC-derived organoids comprise cells representing different germ layers and are generated from PSCs by mimicking the sequential signaling interactions that occur during in vivo development. Notably, organoids faithfully recapitulate the physiological functions of the derived organs. The key distinguishing feature between AdSC-derived and PSC-derived organoids lies in the presence of mesenchymal cells within the PSC-derived organoid culture. While AdSCs can generate specific cells corresponding to their tissue of origin, PSCs possess inherent plasticity and the capacity to differentiate into any cell type. Consequently, the derivation of PSC-derived organoids necessitates a stepwise differentiation protocol to guide the PSCs towards their desired target tissue identity, whereas AdSC-derived gastric organoids can be initiated using a growth factor-enriched medium.

The pioneering work of murine AdSC-derived stomach organoid culture initiated with pyloric glands harboring Lgr5＋ stem cells (17). Barker et al. (17) identified that Lgr5 expressing pyloric gland cells are self-renew and multipotent. Moreover Lgr5＋ cells are suitable to generate long-lived gastric organoid. The initial culture condition was based on the intestinal organoid culture condition, the modified gastric organoid culture protocol involved the addition of fibroblast growth factor 10 (FGF10) and the hormone gastrin (Table 1). Notably, the presence of gastric specific gene expression such as chief cell markers, PGC, and mucus neck cell markers, MUC6, were observed within this initial gastric organoid culture. Modulation of the WNT concentration yielded differentiation towards the mucous pit and endocrine cell lineages, while parietal cells were not detected. Subsequently, the same conditions were applied to murine corpus organoids originating from Troy＋ stem cells, which have cell type specific markers such as the expression of proliferating cell, chief cell and mucus neck cell (18). In addition, different compositions of culture medium, withdrawal of WNT and supplementation with Noggin and FGF10, resulted in the differentiation of pit cells, while enteroendocrine or parietal cells were not observed in this context.

Table 1 . Gastric organoid culture medium composition

Basic growth medium components for gastric organoid culture (ENRGFW medium)
Basal medium: advanced DMEM/F-12＋＋＋ (＋P/S, ＋HEPES, and ＋GlutaMax)
Component	Working mechanism
B-27	Increasing survival
N-acetylcysteine	Scavenging reactive oxygen species
Wnt3A or surrogate Wnt	Activating Wnt signaling
EGF	Activating RAS/RAF/MEK/ERK signaling
Noggin	Inhibiting BMP signlaing
R-spondin 1	Augmenting Wnt signaling
FGF10	Activating FGF signaling
Gastrin	Activating Ihh signaling

EGF: epidermal growth factor, FGF: fibroblast growth factor.

Later, human antral organoids can be successfully established by adapting the mouse protocol to human systems by Gifford et al. (19). The authors identified that Notch signaling is critical for mouse and human gastric organoid establishment. Inhibition of Notch signaling decreases the size of gastric organoid for both from mouse and human. On the other hand, the establishment of human corpus organoids requires the inhibition of the transforming growth factor-β (TGF-β) signaling pathway using A83-01, an inhibitor of activin receptor-like kinase 5, to ensure long-term growth (20).

The differentiation of PSCs into organoids offers a promising approach for generating gastric organoids comprising both epithelial and mesenchymal cell populations. McCracken et al. (21, 22) described the initial differentiation protocol for generating human PSC-derived gastric organoids. The differentiation process commenced by directing human PSCs toward the endodermal lineage through the addition of Activin A and BMP4. Activin A activation stimulated Nodal signaling, a crucial pathway involved in foregut formation. Subsequent addition of FGF4 and either WNT or CHIR99021, a glycogen synthase kinase 3β inhibitor promoting WNT pathway activation, facilitated posterior foregut formation. To generate the foregut region from which the stomach originates, Noggin was additionally applied to inhibit BMP signaling. Embedding these cells within an extracellular matrix facilitated the formation of three-dimensional foregut spheroids. Differentiation into antral organoids was achieved by treatment with retinoic acid and epidermal growth factor (EGF). The complete differentiation process took approximately 34 days and resulted in antral organoids containing pit, mucus neck, and enteroendocrine cells (21, 23). To guide the differentiation toward the corpus region, the organoids were supplemented with CHIR99021, EGF, and FGF10. The subsequent addition of BMP4 and the MEK inhibitor PD032590 stimulated the production of parietal cells (22). The fully differentiated corpus organoids comprised pit, mucus neck, endocrine, chief, and parietal cells (22, 23). In addition, Noguchi et al. (24) demonstrated the generation of organoids from murine PSCs using a stepwise differentiation protocol. The PSCs were cultured as embryoid bodies and exposed to the sonic hedgehog (SHH), the WNT antagonist dickkopf 1, and Noggin. Activation of SHH signaling and inhibition of WNT signaling promoted the formation of tube-like structures, resembling early stomach-like structures. Embedding the spheroids into an extracellular matrix supplemented with FGF10, Noggin, WNT, and Rspo led to the formation of corpus glands after approximately 60 days. Analogous to human PSC-derived corpus organoids, the murine PSC-derived organoids consisted of pit, mucus neck, endocrine, chief, and parietal cells (Table 2).

Table 2 . Specific medium composition and outcomes

Composition (to growth medium)	Outcome
From AdSC
−FGF10 −Noggin −Wnt	Pit cell differentiation
＋Notch inhibitor (DAPT, etc.)	Reduced growth, increasing differentiation
＋Nicotinamide ＋TGF-β inhibitor	To ensure long-term expression
From human PSC (step by step)
＋Activin A	1. PSC to endoderm induction
＋WNT ＋FGF4 ＋Noggin ＋RA	2. Endoderm to posterior foregut spheroid formation
＋RA ＋Noggin ＋EGF	3. Posterior foregut spheroid to antral epithelium
＋EGF	4. Antral epithelium to gastric organoid (containing pit, mucus neck, and enteroendocrine cells)
From murine PSC-2 (step by step)
＋SHH ＋DKK1 ＋Noggin	1. Formation of early stomach-like structures
＋FGF10 ＋Noggin ＋Wnt ＋Rspo	2. Formation of corpus glands (containing pit, mucus neck, endocrine, chief, and parietal cells)
From human foregut progenitor spheroid (step by step)
＋CHIR ＋FGF4 ＋Noggin ＋RA	1. Foregut patterning
＋EGF ＋Noggin ＋RA (＋CHIR)	2. Foregut to gastric specification
＋EGF (＋CHIR)	3. Gastric pattern maintenance
＋EGF (＋CHIR ＋FGF10)	4. Growth morphogenesis
＋EGF (＋BMP4 ＋MEK inhibitor)	5. Antrum (production of parietal cells)

AdSC: adult stem cell, FGF: fibroblast growth factor, TGF-β: transforming growth factor-β, PSC: pluripotent stem cell, RA: retinoic acid, EGF: epidermal growth factor, SHH: sonic hedgehog, DKK1: dickkopf 1.

To address the limitation of AdSC-derived organoids, which are primarily composed of epithelial cells, a co-cultivation protocol was established for murine AdSC-derived organoids in conjunction with mesenchymal cells (25). The inclusion of mesenchymal niche cells resulted in the generation of the entire repertoire of stomach epithelial cells, including parietal cells (albeit for a limited time). Moreover, recent advances on microfluidic device referred as organoid-on-a-chip system considered as a model system which provide in vivo environment including gastric organoid-on-a-chip model (26). However, organoid-on-a-chip is in emerging stage, so it still have technical limitations such as designing a new microfluidic devices (27). Overall, normal gastric organoids serve as an excellent model system for addressing a wide range of scientific inquiries from fundamental research to translational clinical studies.

PDOs are a remarkable tool in research that hold great promise in understanding and treating various forms of the disease such as cancer (9). Similar to the first generation of AdSC-derived organoids from Lgr5＋ mouse intestinal epithelial stem cells, cancer organoids can now be generated from patient-derived tissue samples (28). Culture medium and matrix play crucial roles in constructing cancer organoids same as normal organoid, as they are responsible for regulating multiple signaling pathways and promoting self-renewal, proliferation, and tissue-specific differentiation of cancerous cells. These extra components are required for supporting the growth of cancer organoids.

PDOs maintain characteristics of patient’s disease, which can be widely used for patient such as understanding disease, predict therapeutic responses to certain drugs, etc (29, 30). Up to date, multiple groups reported the generation of gastric PDOs in their own way, which were obtained from diverse sources including surgical resection specimens, endoscopic biopsy, etc. Seidlitz et al. (31) established a biobank of 20 different human gastric cancer organoids with an in-depth molecular analysis of four lines. Vlachogiannis et al. (32) reported a PDO biobank including cancers of different gastrointestinal origin, including four gastroesophageal cancers. Nanki et al. (33) characterized 37 gastric cancer organoid lines so that demonstrated a correlation between mutational signatures and independency from the addition of growth factor in the culture medium. A large-scale gastric cancer biobank was generated by Yan et al. (34), consisting of 46 molecularly characterized gastric cancer organoid lines in addition to 17 normal gastric organoid lines. Importantly, Yan et al. (34) took multiple biopsies from patient which allowed the analysis of subclones within the primary cancer. In addition, Song et al. (35) established small size of 5 gastric cancer organoid lines with matched normal gastric cancer organoid which demonstrated patient-derived gastric cancer organoid resembles the characteristics of primary tumor.

One major challenge in cancer organoid culture is the purity of tumor cells (36). Tumor biopsy always has normal cell contamination, and the overgrowth of normal cells can affect the growth of cancerous cells. Therefore, it is important to only allow growing cancer cells to gain reliable results. Various approaches have been explored to overcome this issue. For example, Wallaschek et al. (37) suggested approaches of purifying cancer organoids for gastric cancers and proposed three methods. The first approach combines prior knowledge from mutational assays. For example, if a tumor exhibits p53 mutations, then Nutlin-3, a MDM2 inhibitor, can be used to disrupt the binding of p53 with MDM2, leading to growth retardation of normal organoids. The second approach is removing normal organoids from cultures based on morphological phenotype by manual selection. The third method is to use flow cytometry-based cell sorting to collect single cells from cancer organoids. However, purifying tumor organoids may lead to a loss of cellular heterogeneity relative to their initial cultures.

The genomes of cancerous cells are a veritable minefield of mutated sequences that exert a profound influence on the onset, progression, and outcome of diverse human malignancies, including gastric cancer (38, 39). Genetic aberrations that accumulate in tumor suppressor genes and oncogenes contribute to the formation and growth of cancerous tumors (40). To probe the impact of specific mutations on tumor progression, investigators turn to tumor organoids, which preserve the histological features and mutation spectra of their parent tumors (30, 41). Genome editing techniques such as clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 and RNAi are utilized to introduce driver mutations into either normal or tumor organoids, thereby facilitating the study of gene function and drug resistance (42). CRISPR/Cas9 represents the most efficient and effective genome editing tool and can introduce DNA double-strand breaks at the targeted locus through RNA-guided endonucleases, enabling gene knock-in and knockout (43). Besides CRISPR/Cas9-mediated genetic engineering for gastric cancer modeling, traditional genetically engineered mouse model-derived organoids are also widely used to comprehend the correlation between genetic alterations and cancer phenotype (44-47).

Recent investigations demonstrate that CRISPR/Cas9 technology can bring about ARID1A knockout in primary TP53^−/− human gastric organoids, leading to tumorigenicity, morphologic dysplasia, and mucinous differentiation (48). The introduction of genetic Wnt/β-catenin activation rescued mucinous differentiation but not hyperproliferation, suggesting alternative pathways of ARID1A knockout-mediated transformation. ARID1A mutation also induced transcriptional regulatory modules characteristic of microsatellite instability and Epstein–Barr virus–associated subtypes of human gastric cancer, such as FOXM1-associated mitotic genes and BIRC5/survivin. In addition, Seidlitz et al. (44) established inducible stomach-specific cancer mouse lines to model different molecular subtypes of human gastric cancer. The authors generated stomach specific Anxa10-CreER^T2;Kras^G12D/＋;Tp53^R172H/＋;Smad4^fl/fl, Anxa10-CreER^T2;Cdh1^fl/fl;Kras^G12D/＋;Smad4^fl/fl, and Anxa10-CreER^T2;Cdh1^fl/fl;Kras^G12D/＋;Apc^fl/fl conditional mice lines and derived organoids from those mice lines. The organoid lines from genomically stable type tumors were more resistant to docetaxel, whereas organoids from chromosome instability type tumors were more resistant to trametinib. Genetic engineering has been employed in cancer modeling and is henceforth widely embraced in other cancerous conditions. For instance, Takeda et al. (49) demonstrated that activin signaling is having an important role in the advancement of colorectal cancer through the use of CRISPR/Cas9-mediated genetic engineering.

It is widely known in our research for diseases including cancer that organoids are a suitable system for testing in vitro sensitivity to different kinds of compounds and simulateing therapeutics efficiency for clinical applications since organoid keep the characteristics of organ of origin (36). The most advantageous feature of organoid-based drug screening is that the test utilizes organoid directed from patient, so it can reflect characteristics of individual patient. In addition, it can be performed within a few months so that patients do not need to wait long time, sparing them from the pain of prolonged illness. Functional tests are more easily performed in organoids to assess therapeutic responses, such as the detection of the secretion of functional proteins and the swelling assay to assess responsiveness to cystic fibrosis drugs, helping to identify the cancerous cells (Fig. 3) (50, 51).

Figure 3. Expectation of organoid-based drug development. The conventional drug development process begins with disease information followed by target selection altered in the disease condition. Once the target is selected, primary in vitro screening will be performed to search candidate compounds for pre-clinical trials. Then, in vitro hit validation with primary candidates finalize candidates for pre-clinical trials with animal models. Successful candidates after the pre-clinical trial will be transferred to a clinical trial. However, organoid-mediated drug development is initiated with a group of patients with the disease. Patient-derived organoids (PDOs) will be established and stored at an organoid biobank. Drug screening using the PDO library will discover effective drugs for the disease, and successful drug candidates will be introduced for clinical trials.

A most commonly used method of drug screening is to assess organoid cell viability under drug-treated condition with a wide range of concentration (52). However, if different experimental conditions such as different reagents, combination of drugs, machines and data analysis approaches are used, data with some differences may be obtained, leading to confusion and data misreading. The assessment methods of drug screening include half maximal inhibitory concentration or area under the curve, growth curve and so on. Until now, there are no standardized protocols in the field of organoids culture and drug screening, which makes it difficult to compare screening results among experiments, organoid cultures and even laboratories and different researchers. Therefore, there are efforts to standardize methodology of organoid culture and screening (53-55). However, details of the protocols used for drug screening by different groups are overall similar. Generally, PDOs are cultured with growth factor cocktails and passaged for some time. Subsequently, organoids are collected, filtered, counted and plated at the desired density and then exposed to the therapy of choice such as different concentrations of small molecular inhibitors, radiotherapy, etc., hoping for a cure.

Recently, Toshimitsu et al. (56) established a robust drug screening platform with suspension culture for better organoid scalability with better organoid culture condition, and performed drug screening using patient-derived colorectal cancer organoid. In the study, the authors established suspension culture condition which allows 10⁶∼10⁷ cells without perturbing organoid growth with 23 colorectal cancer PDO lines, and applied those organoid lines for single cell seeding followed by robust highthroughput drug screening 4 days after plating with a set of drug pannels. Moreover, the authors applied image-based pharmacotyping which allows automated data collection and analysis. In the end, the authors demonstrated that bromodomain and extra-terminal bromodomain protein inhibitor can be a cancer-selective growth suppressor which targets colorectal cancer activated genes. This knowledge can be used to create new therapies, allowing us to fight back against the deadly cells that threaten us every day.

In addition to organoid drug screening, microfluidic and organ-on-chip technologies have been recently applied to organoid drug screening, like a tumor that has spread throug-hout the body (57-59). Microfluidics is the system which manipulating in vivo fluids in tiny channels, allowing researchers to create a complex tumor microenvironment to test more types of drugs. The human organ-on-a-chip is an improved technique that simulates intracellular relevance and organ interactions, thus providing the possibility of in vitro testing of pharmacodynamics and toxicodynamics of drugs. Thus, the progress of basic technology can improve the effect and expand applications for organoids drug screening, providing us with hope for a brighter future.

In the era of advanced technology, intricate signaling networks and diverse cellular states have been unveiled in model systems. CRISPR/Cas9, a genetic engineering tool, has revolutionized genome-wide, targeted loss-of-function pooled screens in human and mouse cells, and has facilitated the identification of functional gene sets in various conditions, including homeostasis and disease (59). Researchers widely use CRISPR/Cas9 system due to its ease of use, precise gene editing, and other advantages. Several groups developed genomewide CRISPR screening in mouse and human system (60-62). Shalem et al. (61) reported the genome-scale CRISPR screening in diverse human cell types including cancer cell and PSC. In this study, the authors applied genome-scale CRISPR/Cas9 knock-out library which cover more than 18,080 genes with 64,751 unique single guide RNAs (sgRNAs). The genome-wide CRISPR/Cas9 screening identified novel genes involved in mutant BRAF inhibition such as NF2, CUL3, TADA2B, and TADA1. These genome-wide screening techniques have been applied to several cell lines to unravel the mechanism driving malignancy, stem cell maintenance, development, etc. However, due to lack of cellular heterogeneity in two-dimensional cell line, researchers tried to perform CRISPR screening in organoid since organoid maintain cellular heterogeneity.

Recently, several groups have reported genome-wide CRISPR screening in various organs, including the colon, kidney, and stomach (63-66). Ringel et al. (63) established a genome-wide CRISPR screening method in human intestinal organoids to dissect oncogenic signaling pathways. By generating a CRISPR library targeting 19,114 genes with 4 sgRNAs per gene, the authors identified a set of genes involved in the resistance to TGF-β signaling-mediated growth restriction. Similarly, Michels et al. (66) reported a study related to TGF-β signaling, identifying optimal conditions and sgRNA requirements for organoid genetic screening with a smaller set of 100 genes with 20 sgRNAs per gene. Ungricht et al. (65) utilized iPSCs for initial transduction to generate kidney organoids and identified the cis-inhibitory effect of Jag1 in tubular cells. Lastly, Murakami et al. (64) performed CRISPR screening using mouse gastric organoids and confirmed that knockout of Alk, Bclaf3, or Prkra supports the Wnt independent stem cell self-renewal of gastric epithelium. The progress in organoid screening has opened new avenues for understan-ding organ development, diseases, and therapies.

Gastric organoids culture established from AdSCs and PSCs have emerged as a powerful technique for studying the maintenance of gastric stem cells and disease development. Both AdSC- and PSC-derived organoids highly mimic the cellular composition and physiological functions of the stomach, and PSC-derived organoids have a benefit of generating both epithelial and mesenchymal cell populations. These organoids have been successfully established from diverse mammalian species, including mice and humans, providing valuable insights into the development and differentiation of gastric tissues from gastric stem cells. Additionally, patient-derived gastric organoids have revolutionized cancer research by firmly recapitulating and maintaining the characteristics of individual patients’ tumours. The PDO has the potential to serve as a personalized model for understanding disease progression, predicting therapeutic responses, and developing a novel treatment against the tumour. Furthermore, genetic engineering techniques, such as CRISPR/Cas9 mediated gene editing, introduce specific mutations at the desired locus in organoids, allowing researchers to understand gene function and related drug resistance/response in the context of gastric diseases, including cancer. Overall, gastric organoids hold immense promise for advancing our understanding of gastric biology, disease, and personalized medicine (Table 3).

Table 3 . Summary of gastric organoid applications and challenges

Aim	Example of application
Extablishment organoid	AdSC- and PSC-derived gastric organoid
To mimic in vivo microenvironment	Organoid-on-a-chip, co-culture system
To understand pathophysiology	Patient-derived gastric organoid
To unravel the mechanism of cancer driver genes	Gastric cancer modeling by genetic engineering
Precision medicine	Drug screening and development
To unveil signaling networks and understand regeneration	Genetic screening for gastric cancer and organoid
Future perspective	Challenge
Establishment improved organoid	Incomplete niche configuration. Maturation
Forecasting individualized treatment reactions	Partial recapitulating and maintaining characteristics of individual patients’ tissue
CRISPR/cas9 mediatedtreatment to overcome resistance	Lack of understanding gastric biology
Large-scale patient-derived organoid biobank	Insufficient sample size
Recapitulating gastric microenvironment	Incomplete co-culture system
Automated large-scale screening platform	Scale of library and culture system

AdSC: adult stem cell, PSC: pluripotent stem cell, CRISPR: clustered regularly interspaced short palindromic repeats.

Although gastric organoid research is continuously evolving, there are remaining spaces to overcome. First, further optimization and standardization of organoid culture protocols will be crucial to enhance the reproducibility and scalability of organoids. It includes fine-tuning the culture medium composition and matrix components which provide a closer in vivo microenvironment of the stomach. Additionally, advances in microfluidic organoid-on-a-chip systems hold promise for creating more physiologically relevant and complex models that can better mimic the dynamic nature of the stomach and facilitate high-thro-ughput drug screening. Establishing large-scale PDO biobanks is also significant point for covering diverse diseases and cancer heterogeneity. Current PDO biobank mainly focuses on cancer organoid biobank. However, numerous hundreds of thousands of diseases still require more attention to develop therapeutics. A wider variety of organoid biobanks will be a good solution for discovering therapeutics. In addition, it is possible that the cancer heterogeneity can be altered while selecting a pure cancer population out of PDOs. Therefore, large-scale organoid biobanks with various organoid selection methods could cover cancer heterogeneity.

For the disease modelling, introducing additional genetic alterations and exploring genotype-phenotype correlation in gastric organoids will allow researchers to better understand the complex interplay between genetic alteration and their outcome in gastric disease initiation and progression. Furthermore, developing co-culture systems, such as immune cells such as macrophage and organ specific microbiota co-cultures, will provide a more comprehensive representation of the gastric microenvironment and enable studying immune responses and host-microbe interactions in gastric diseases.

Lastly, integrating gastric organoids with emerging technologies, such as organoid genetic screening and automated large-scale organoid-based drug screening platforms, will facilitate the translation of basic research findings into clinical applications. These advancements have the potential to support drug discovery and development, enabling the identification of novel therapeutic components and the establishing personalized medicine. In the end, the continued advancement of gastric organoid research will contribute to a deeper understanding of gastric biology, disease, and personalized medicine which provide better opportunities in gastric disorders.

Fig. 2 was created with BioRender.com under publication and licensing right agreement number XK26L04MSD.

There is no potential conflict of interest to declare.

Conceptualization: JK. Funding acquisition: JK. Visualization: SL, JHC, SYP. Writing – original draft: SL. Writing – review and editing: JK, SYP.

This work was supported by funding program of the National Research Foundation of Korea Grant funded by the Korean Government (2021K1A4A7A02097757).