Organoids: approaches and utility in cancer research

Abstract

Organoids are three-dimensional cellular structures with self-organizing and self-differentiation capacities. They faithfully recapitulate structures and functions of in vivo organs as represented by functionality and microstructural definitions. Heterogeneity in in vitro disease modeling is one of the main reasons for anti-cancer therapy failures. Establishing a powerful model to represent tumor heterogeneity is crucial for elucidating tumor biology and developing effective therapeutic strategies. Tumor organoids can retain the original tumor heterogeneity and are commonly used to mimic the cancer microenvironment when co-cultured with fibroblasts and immune cells; therefore, considerable effort has been made recently to promote the use of this new technology from basic research to clinical studies in tumors. In combination with gene editing technology and microfluidic chip systems, engineered tumor organoids show promising abilities to recapitulate tumorigenesis and metastasis. In many studies, the responses of tumor organoids to various drugs have shown a positive correlation with patient responses. Owing to these consistent responses and personalized characteristics with patient data, tumor organoids show excellent potential for preclinical research. Here, we summarize the properties of different tumor models and review their current state and progress in tumor organoids. We further discuss the substantial challenges and prospects in the rapidly developing tumor organoid field.

Introduction

Organoids are complex three-dimensional (3D) cellular structures that capture the architectural and functional aspects of organs in vivo. Usually, organoids can be generated by directed differentiation [Figure 1] from pluripotent stem cells (PSCs) combined with 3D culture system. Over the past few decades, organoids have provided novel models for studying organ and tissue biology and have fulfilled their promise in advancing basic and translational research. Since the seminal discovery by Hans Clevers and Yoshiki Sasai, this field has blossomed. A growing number of scientists have developed protocols to generate various types of organoids. Clevers et al^[1] constructed intestinal organoids derived from a single Lgr5⁺ intestinal stem cell. Sasai et al recapitulated the development of cortical tissue^[2] and optic cup^[3]in vitro by using PSCs. Lancaster et al^[4] generated cerebral organoids representing different brain regions in 2013. Muguruma et al^[5] developed a protocol to generate the first trimester cerebellum with a polarized structure in vitro. In addition to ectoderm organoids, scientists have generated various endoderm organoids from PSCs by mimicking primitive gut tube development, including the stomach,^[6] lung and thyroid,^[7] and liver.^[8] Self-renewing pancreatic and liver organoids were constructed from adult primary tissues with long-term expansion abilities.^[9] Protocols have also been established to grow mesodermal organoids. In 2005, Yuasa et al^[10] induced the differentiation of mouse embryonic stem cells into cardiomyocytes by inhibiting bone morphogenetic protein (BMP) signaling. Later, robust human cardiomyocyte differentiation protocols were developed via modulation of the canonical WNT pathway.^[11] Recently, 3D heart organoids were generated to recapitulate early heart development.^[12–14] Single-tissue organoids and multiple-tissue organoids are emerging. Engineered hepato-biliary-pancreatic organoids^[15] that exhibit connections between neighboring organs have been developed.

Figure 1.

Directed differentiation of ESCs/iPSCs. Snapshot of ESCs/iPSCs directed differentiation routes: (1) ESCs/iPSCs can be driven to ectoderm by regulating WNT 3A and ACTIVIN A. Ectoderm continues to differentiate into neural progenitors under EGF and bFGF simulation. Retinal cells can be obtained by using DKK, Noggin, IGF-1, bFGF, ACTIVIN A, and Notch regulators. (2) ESCs/iPSCs can differentiate into mesoderm. Then cardiomyocytes and nephrons can be generated by applying WNT with insulin and FGF with WNT, respectively. (3) Endoderm can generate from ESCs/iPSCs through regulating ACTIVIN A. Lung/airway epithelial derived from endoderm by regulating TGF, BMP4, FGF, WNT, and KGF pathways. Hepatocytes can differentiate from endoderm through stimulating HGF, EGF, FGF, OSM, and DEX. bFGF: Basic fibrobast growth factor; BMP4: Bone morphogenetic protein; DEX: Dexamethasone; DKK: Dickkopf; EGF: Epidermal growth factor; ESC: Embryonic stem cells; FGF: Fibrobast growth factor; HGF: Hepatocyte growth factor; IGF-1: Insulin-like growth factor 1; iPSC: Induced pluripotent stem cells; KGF: Keratinocyte growth factor; OSM: Oncostatin M; TGF: Transforming growth factor.

In addition to recapitulating embryonic development, organoids have also been used as tumor models. Endowed by several advantages that have not been achieved using classical 2D cell culture techniques, tumor organoids will accelerate the development of tumor research and drug discovery. A large number of studies on tumor organoids have been published over the past decade. Most tumor organoids were generated from primary carcinoma samples using adult stem cells (ASCs) derived organoid culture system [Figure 2]. Moreover, organoid studies in the field of cancer are accelerated by new technological developments. Clustered regularly interspaced short palindromic repeats (CRISPR) gene editing technology has been widely used in PSC-derived organoids to generate tumor organoids [Figure 2]. The current progress in microfluidic culture systems combined with organoid technology applied in cancer research contributes to the study of cancer mechanisms and drug screening in a precisely controlled manner.

Figure 2.

Generation of tumor organoid. Summary of two main methods of tumor organoid generation: (1) primary tumor cells obtained from patients using 3D culture system to establish tumor organoid and (2) organoids derived from PSCs combined with gene editing technology to generate tumor organoid. PSCs: Pluripotent stem cells.

Current Tumor Models

The lack of effective tumor models is an important factor restricting tumor research. This is because the tumor microenvironment (TME) and cellular milieu influence tumor development. Hence, the lack of these elements hinders realistic in vitro modeling. Therefore, the establishment of an ideal tumor model has always been the focus of tumor research. The commonly used tumor models are described in Table 1.

Table 1.

Comparison of current tumor models.

Model type	Advantages	Disadvantages	References
Cell model
Tumor cell lines (2D)	Low cost and easy construction; high-throughput toxicity testing and drug screening; easy for genetic manipulation.	Lack native TME; lack tumor phenotyes and heterogeneity; lack certain types of tumor.	[16,17]
Animal tumor models
Spontaneous mouse tumor model	Recapitulate the TME; allow systemic evaluation of therapies.	Tumor incidence rates vary greatly; limited tumor model types; low throughput and time-consuming.	[18,19]
Induced mouse tumor model	Have the advantages of spontaneous mouse tumor models; have a higher tumor induction rate.	Time-consuming; Tumor occurrence time and development speed vary greatly.	[20,21]
Genetically engineered mouse tumor model	Accurately reflect human tumors; available for studying the relationship between specific genes and tumors.	High costs, low throughput, and time-consuming; difficult to control the TME and monitor the tumors.	[22,23]
Tumor xenograft models
Human tumor CDX model	Human source of cancer cells; easy to operate; convenient for gene manipulation.	Require immune-deficient hosts; cannot reproduce the TME; low throughput.	[24–26]
PDX model	Maintain the heterogeneity of the original tumor; high consistent rate of drug response.	Require immune-deficient hosts; low tumor implantation rates; low throughput.	[26–28]
Hu PDX model	Provide TME similar to the human body; simulate the interaction between tumors and the immune system.	Low success rate and high cost; limited disease models.	[19,29–31]
PDOX model	Mimic primary tumor environment; available for studying tumor invasion and metastasis.	Operation technology is complex; time-consuming and high cost.	[32–34]
3D tumor culture models
MCTS model	Low cost and easy genetic manipulation; high-throughput toxicity testing and drug screening.	Lack of native TME; poor consistency in size and uniformity; different sensitivities to drug tests.	[17,35]
Tumor organoid model	3D spatial organization; preservation of genetic and epigenetic signature of derived tumor tissue; easy for genetic manipulation; high-throughput in toxicity testing and drug screening; genetic stable after long-term passage; simulate the TME.	Limited size and uniformity; lack of vascular elements; varied reproducibility; high costs; side effects of animal products such as Matrigel, BME, and FBS.	[17,36,37]

BME: Basement membrane extract; CDX: Cell-derived xenograft; FBS: Fetal bovine serum; Hu PDX: Humanized patient-derived xenograft; MCTS: Multicellular tumor spheroids; PDOX: Patient-derived orthotopic xenograft; PDX: Patient-derived xenograft; TME: Tumor microenvironment.

Tumor cell line

The tumor cell line is an immortalized cell model established in a two-dimensional (2D) cell culture system that takes advantage of the unrestricted proliferation of tumor cells from a patient's tumor tissue. Since the first human immortalized cell line, the HeLa cervical cancer cell line, was established in the 1950s^[38]; this cell line has become the most widely used tumor model because of its advantages of easy construction, short culture cycle, and high survival rate.^[16] Thousands of tumor cell lines from multiple systems and tissues have been established (https://portals.broadinstitute.org/ccle).

Animal tumor models

Animal tumors appear in nature or can be induced in mouse models by human intervention. Generally, mouse tumor models are the most widely used models, and >46,000 mouse tumor models (http://tumor.informatics.jax.org) have been generated by different approaches so far. For example, a spontaneous mouse tumor model was developed under natural conditions. Since there is no interference from human factors, its tumorigenesis initiation mechanism and cytological characteristics are similar to human tumors.^[18] An induced mouse tumor model was generated under experimental conditions using carcinogens. Carcinogenic factors include physical factors (ultraviolet rays and radiation), chemical factors (carcinogenic compounds), and biological factors (viruses and hormones). Different carcinogenic factors can induce tumorigenesis in different organs and systems. As the inducing factors and conditions can be controlled artificially, the tumor induction rate is much higher than the natural incidence rate.^[39,40] This approach is commonly employed to understand the mechanism of a known carcinogen. Recently, genetically engineered mouse tumor models have been established using gene editing. Tumorigenesis can be caused by activation of proto-oncogenes, overexpression of oncogenes,^[41] or knockout of tumor suppressor genes.^[42] In addition, modification of immune system-related or stromal-related genes can drive or promote tumorigenesis.^[43]

Tumor xenograft models

A tumor xenograft model refers to the transplantation of tumor tissue or tumor cells into experimental animals (mainly mice) to study different tumor processes and drug responses. Three major tumor xenograft models are available at present, including the cell-derived xenograft (CDX), patient-derived xenograft (PDX), and patient-derived orthotopic xenograft (PDOX). The CDX model was established by culturing human tumor cells in 2D culture conditions in vitro to form stable tumor cell lines, which were then injected into immunodeficient mice.^[24] This model is easy to operate, is convenient for gene manipulation, and is commonly used in preclinical drug development.^[25] PDX models are established by surgically transplanting tumor tissues (orthotopic or ectopic) into immunodeficient animals.^[44] This model preserves the TME and retains the characteristics of the original tumor.^[27] They are currently widely used in preclinical drug and clinical prognosis evaluations, biomarker identification, tumor biology research, and personalized medicine.^[45,46] The PDX model established using humanized animal models is called a humanized patient-derived xenograft.^[29] Since humanized mice recapitulate many features of the human immune system and immune response, they have broad application prospects in immunotherapy and immunology research of various cancers^[30,31] and help bridge species-specific differences in carcinogenesis and drug response. A PDOX model was established by transplanting human tumor tissues into the corresponding or adjacent tissues of immunodeficient mice.^[32] Orthotopic transplantation can provide an environment closer to the primary tumor. Therefore, PDOX models have been widely used in tumor metastasis research and preclinical drug testing.^[33,34]

3D tumor culture models

The advent of 3D culture systems has improved the authenticity of in vitro biological systems. When a 3D culture system is used in a tumor culture model, tumor cells form tumor spheroids that better simulate cell–cell and cell–matrix interactions.^[47] 3D tumor models can be generated from multicellular tumor spheroids (MCTS) or patient-derived tumor organoids (PDTOs).^[35] Tumor tissues or cells isolated from patients were digested with enzymes to obtain a single-cell suspension and then cultured in a 3D environment to form MCTS.^[35] MCTS can mimic the main characteristics of solid tumors in vivo, such as the metabolic rate, growth kinetics, resistance to radiation, and chemotherapy.^[48] Organoids derived from PSCs or ASCs in a 3D culture environment in vitro can recapitulate the characteristics of tissues or organs in vivo.^[49] PDTOs are derived from tumor-specific stem cells, such as cells from surgical resection or biopsy^[50] or circulating tumor cells^[51]; therefore, their biological properties are similar to those of the original tumor tissue. PDTOs can be passaged, expanded long-term in vitro, or genetically manipulated. They are genetically stable and can maintain the heterogeneity of their original tumors after long-term passage. When co-cultured with other relevant cells (e.g., immune cells and mesenchymal cells), the tumor organoid models can mimic the TME in a Petri dish, providing an ideal tumor research platform in vitro.^[36,37] In addition, PDTOs can be transplanted into immunodeficient mice to establish a PDX model as needed.^[52–54] It exhibits tumorigenic ability similar to that of the original tumor cells in immunodeficient mice and can reproduce the clinical characteristics of original tumor invasion and metastasis.^[52–54]

The tumor models introduced above have their own advantages and disadvantages and are widely used in different tumor studies. At present, there is no single model that can meet the needs of all types of tumor research. Recent advances in in vitro 3D culture technology (especially organoid technology) have attracted the attention of many oncology researchers and are expected to open new avenues for developing superior human tumor models.

Brief History of Tumor Organoids

The Clevers’ lab first established an intestinal organoid system in 2009.^[1] They cultured adult mouse intestinal stem cells expressing Lgr5 in Matrigel.^[1] These stem cells were self-organized and differentiate into intestinal crypt-villus units without a non-epithelial cellular niche.^[1] In 2011, tumor organoids derived from mouse adenoma and human colorectal cancer (CRC) cells were successfully established.^[55] Since then, various tumor organoids such as those of prostate,^[56] pancreatic,^[57] endometrial,^[58] liver,^[59] breast,^[60] lung,^[61,62] gastric,^[63] esophageal,^[64] bladder,^[52] and ovarian^[65] cancers have also been developed to model various cancers [Table 2]. Many tumor organoid biobanks have also been established to capture tumor subtype heterogeneity and enable therapeutic screening.^[52,61,63] The tumor organoids in these biobanks have been derived from tumors of different subtypes and pathological grades. Most, if not all, of the tumor organoids established earlier have been derived from epithelial tumor tissue cells. In recent years, with the development of organoid culture techniques, some organoids from non-epithelial tumors, such as kidney tumors^[54] and glioblastomas,^[66] have been generated, providing a reference for the establishment of other non-epithelial tumor organoids. In addition, tumor organoids driven by the introduction of oncogene activating mutations into primary normal organoids using gene-editing technology have also been successfully established,^[67–69] further expanding the source of tumor organoids and increasing our understanding of human tumorigenesis caused by suspected or known gene mutations.

Table 2.

Organoids (or biobanks) of various tumors.

First report year	Cancer types	Description	Reference
2011	CRC	Sato et al developed an organoid culture technology that can be used to study gastrointestinal tract tumors. Using this technology, they successfully established organoids from mouse adenomas and human CRC cells. This was the first report of human tumor organoids.	[55]
2014	Prostate cancer	Gao et al have successfully cultured prostate cancer organoids for a long time. These patient-derived prostate cancer organoids can be cultured on a large scale and used for genetic and pharmacological studies.	[56]
2015	Pancreatic cancer	Huang et al generated human PDAC organoids, which maintained the pathological properties of the primary tumor and patient-specific physiologic changes. These tumor organoids can be used to model PDAC and test personalized therapies.	[57]
2016	Glioblastoma	Hubert et al established a 3D organoids culture system that supports generation of GBOs derived from glioblastoma specimens. Orthotopic transplantation of these patient-derived GBOs resulted in tumors exhibiting histological features of the original tumor.	[66]
2017	Endometrial cancer	Turco et al derived 3D cultured organoids from human normal, decidualized, and malignant endometrium. These organoids are helpful for studying uterine diseases such as endometriosis and endometrial cancer.	[58]
2017	Liver cancer	Broutier et al established PDTOs cultures from patients with PLC, including three tumor tissue types (HCC, CC, and HCC/CC). These organoids retained the histological architecture and genetic features of the original tumor and can be used to understand liver cancer biology and develop personalized medicine.	[59]
2018	BC	Sachs et al generated >100 primary and metastatic BC organoids from different types of patients with BC. The morphology, histopathology, and hormone receptor status of these BC organoids matched those of the original tumor tissue, and these organoids can be used in cancer research and personalized medicine development.	[60]
2018	Lung cancer	Clevers’ group cultured lung cancer organoids from surgically resected tissues and needle biopsy tissues. These organoids can retain tumor histopathology and gene mutations, are suitable for large-scale drug screening, and are ideal models for studying pulmonary diseases.	[61,62]
2018	GC	Vlachogiannis et al generated a biobank of PDTOs from patients with colorectal cancer, esophageal cancer, and GC. These PDTOs highly preserved phenotypic and genotypic profiling of the original tumor, and can predict drug responses.	[63]
2018	Esophageal cancer	Li et al generated EAC organoids that reproduce the morphological, genomic, and transcriptomic features of the primary tumor. These EAC organoids provide preclinical tools for the precision therapy.	[64]
2018	OC	Hill et al established 33 HGSC organoids from 22 patients with HGSC and performed functional profiling of DNA repair in these HGSC organoids. These HGSC organoids are genetically and functionally matched to the original tumor and are expected to predict the clinical response of patients with HGSC.	[65]
2018	Bladder cancer	Lee et al established a biobank of human bladder cancer organoids derived from patient biopsies. These organoids retained original tumor heterogeneity and can be used for drug screening.	[52]
2019	Renal carcinoma	Grassi et al established renal cell carcinoma organoids derived from renal tumor tissues. These tumor organoids retain primary tumor-specific markers and can be orthotopically transplanted into immunocompromised mice to develop neoplastic masses. It is expected to become a new model for kidney replacement and for studying renal diseases.	[54]
2020	Thyroid cancer	Sondorp et al established long-term cultured PTC organoids from 13 patients’ PTC tissues. They also established tumor organoids from patients with three I131-resistant PTC. These organoids can be used for pre-treatment diagnosis of patients with I131-resistant PTC.	[70]

BC: Breast cancer; CC: Cholangiocarcinoma; CRC: Colorectal cancer; EAC: Esophageal adenocarcinoma; GBOs: Glioblastoma organoids; GC: Gastric cancer; HCC: Hepatic cellular cancer; HGSC: High-grade serous ovarian cancer; I¹³¹: Radioactive iodine; OC: Ovarian cancer; PDAC: Pancreatic ductal adenocarcinoma; PDTOs: Patient-derived tumor organoids; PLC: Primary liver cancer; PTC: Papillary thyroid cancer.

Application of Tumor Organoids in Tumor Research

Owing to the maintenance of histological properties and genomic features of the original tumor tissue, tumor organoids have been widely used in tumor research [Figure 3].

Figure 3.

Tumor organoid application. Tumor organoids can be established from diverse cancer subtypes and grades to generate tumor organoid biobanks. Multiple omics analysis at genome, transcriptome, proteome, and metabolites levels can expand tumor knowledge in basic studies. Meanwhile, tumor organoids can also be used to perform drug screening to identify effective anti-cancer drugs on an individualized level.

Tumor organoids for basic research

Tumors are caused by the gradual accumulation of genetic alterations driven by various tumorigenic factors,^[71] and an in-depth study of these genetic alterations is critical for understanding tumorigenesis. PDTOs can be used to simulate and study the occurrence and progression of tumors in specific organs.^[72] For example, recent studies have successfully generated CRC organoid models by introducing a series of combinations of CRC driver mutations into healthy human intestinal organoids through CRISPR–Cas9 genome editing.^[67–69] These organoids can be used to identify novel cancer drivers or tumor suppressor genes and to screen for patient-specific drug targets.^[68,69] Other studies have established adenomatous polyposis coli gene knockout (APC^KO) esophageal organoids by CRISPR–Cas9 genome editing.^[73] These APC^KO organoids showed higher proliferative and replicative activity, histological atypia, and reduced apoptosis, confirming the critical role of abnormal Wnt/β-catenin signaling activation in Barrett esophagus tumor transformation.^[73] In addition, CRISPR/Cas9-mediated homology-independent organoid transgenesis (CRISPR-HOT) has been successfully applied to generate human liver cancer organoids,^[74] which have been utilized to confirm the role of p53 mutation-induced aneuploidy in promoting liver cancer.^[74] Taken together, these studies highlight organoids as powerful research tools for studying the relationship between tumorigenesis and genetic alterations.

The occurrence of many tumors has been confirmed to be related to pathogenic microbial infections, such as gastric tumors and Helicobacter pylori, cervical cancer and human papilloma virus,^[75] liver cancer and hepatitis virus, nasopharyngeal cancer and Epstein–Barr virus, and liver cancer and Aspergillus flavus.^[75] However, the mechanism by which these pathogenic infections cause tumorigenesis remains unclear. Co-culture of organoids with different pathogens has recently been used to study the possible mechanisms of pathogen-driven tumorigenesis and malignant transformation. For example, after long-term exposure of oncogenic Escherichia coli carrying pathogenicity island pks to healthy human intestinal organoids, researchers discovered a mutation associated with human CRC development in organoids.^[76] The relationship between hepatitis B virus ^[77,78] or hepatitis C virus ^[79] infection and hepatocellular carcinogenesis has also been explored in liver organoids in recent studies. Gastric organoids have also been used to study the connection between chronic Helicobacter pylori infection and gastric cancer (GC) development.^[80] These studies suggested that organoids are powerful tools for studying pathogenic infection-induced tumorigenesis. As technology advances, organoids are expected to help researchers unravel the detailed mechanisms underlying pathogenic microbial infections and tumorigenesis.

Moreover, PDTOs have been used to study the process and mechanism of tumor invasion or metastasis. For example, Huang et al^[81] applied pancreatic ductal adenocarcinoma (PDAC) organoids to characterize molecular alterations critical for tumor invasion and revealed that SMAD4 can mediate collective invasion of PDAC organoids through non-canonical transforming growth factor β signaling. Another study generated paired organoids from primary tumors and matched liver metastases to model CRC metastases. The results showed that SOX2 is associated with the invasion and proliferation of liver metastases in CRC.^[82] Similarly, by using tumor organoids, researchers found that rho-associated protein kinase 2 inhibition can induce the collective invasion of CRC cells.^[83] These studies showed the great potential of organoids in basic research on tumorigenesis, progression, invasion, and metastasis.

Living tumor organoid biobanks

Owing to their 3D structures, PDTOs can maintain the heterogeneity of original tumor tissues for a long time in vitro and display higher clinical relevance to their original tumors than the immortal tumor cell lines, which is critical for the establishment of a biobank. The utility of tumor organoid biobanks is especially important for rare tumor subtypes that have difficulty in producing stable tumor cell lines. These PDTOs biobanks could serve as valuable biological resources for basic and clinical tumor research, expanding the repertoire of patient tumor samples that can be studied in the laboratory. At present, many biobanks for various types of tumor organoids have been established. Organoids in these biobanks are derived from different subtypes and grades of tumors and can be passaged and cryopreserved, similar to tumor cell lines. For example, organoid biobanks have been established from different types of lung cancer tissues, and paired non-neoplastic airway tissues.^[84–87] Because these organoids are self-renewing, the biobanks can be screened and tested on a large scale to develop effective anti-tumor drugs and personalized treatment regimens. In addition, these biobanks typically include PDTOs and their matching healthy tissue organoids, which greatly facilitate genetic association studies and multiomics analyses. Indeed, these tumor organoids have been successfully used to profile the cancer genome,^[48] transcriptome,^[88] proteome,^[89,90] and metabolome.^[91] These studies highlight the potential of tumor organoid biobanks for studying inter- and intra-tumor heterogeneity, genetic mutations, tumorigenesis and progression, and clinical drug resistance. Since the majority of existing tumor organoid cultures have been developed for epithelial tumors in biobanks, more organoids generated from non-epithelial cell fractions, matching immune cells, and rare populations within tumors are needed to increase the realism of tumor organoids. Additionally, unified standards for organoid manipulation or production in biobanks are urgently required to control organoid quality and improve reproducibility and comparability.

Tumor organoids in drug development

PDTOs can faithfully demonstrate the response of the original tumor to drugs in vitro compared to classical 2D models, and thus can be used for high-throughput drug screening to develop effective anti-tumor drugs. Many types of tumor organoids have been used in anti-tumor drug development and have achieved exciting results. Saito et al^[92] established biliary tract carcinoma organoids from patients with neuroendocrine carcinoma of the ampulla of Vater, gallbladder carcinoma, and intrahepatic cholangiocarcinoma. This organoid was used to screen 339 drugs, 22 of which were confirmed to significantly inhibit the growth of tumor organoids.^[92] Other researchers have established a scalable, high-throughput CRC organoid platform with 8-well strips and 8 × 9 microwells per strip^[93] for effective drug screening and clinical outcome prediction.^[93] By analyzing the gene expression profile of tumor organoids at bulk, single-cell, or spatial resolution, several drug candidates have been identified recently. For example, the assessment of the gene expression profiles of GC organoids (GCOs) from oxaliplatin (L-OHP)-resistant GC revealed that myoferlin is highly associated with L-OHP resistance and tumor progression in these GCOs, indicating that myoferlin may be a potential therapeutic target for L-OHP-resistant GC cases.^[94] In another study, by combining tumor organoids and targeted RNA-seq, researchers developed a high-throughput and high-content drug discovery platform (TORNADO-seq) to identify effective anti-cancer drugs.^[95] As mentioned above, PDTOs biobanks can be used for large-scale drug screening, with great potential for personalized cancer therapy.

Tumor organoids in immunotherapy

Currently, drug resistance in neoplastic tissues is one of the main obstacles restricting the development of anti-cancer drugs. In recent years, immunotherapy, which stimulates the immune system of patients to kill tumor cells, has attracted significant research interest. However, the immunogenicity of tumor antigens is not sufficiently strong to elicit an appropriate immune response, leading to immunotherapy resistance in most patients. The tumor organoids can be used as an in vitro test platform for tumor immunotherapy and show great potential for anti-tumor immunotherapy. The introduction of an immune component by the co-culture of tumor organoids with peripheral blood mononuclear cells (PBMCs) can generate patient-specific tumor-reactive T cells.^[96] For example, a study of co-cultured patient-derived CRC organoids and non-small-cell lung cancer organoids with lymphocytes from autologous peripheral blood^[61] revealed that patient-specific tumor-reactive cytotoxic T cells were enriched, enabling efficient killing of the matched tumor organoids.^[61] In another study, researchers established human GCOs from biopsy or excised tissues and co-cultured them with cytotoxic T lymphocytes and myelogenous suppressor cells.^[97] They found that co-expression of human epidermal growth factor receptor 2 and programmed death-ligand 1 may be involved in the immune evasion of tumor cells, and co-culture of autologous organoids and immune cells can be used to screen effective anti-tumor immunotherapeutic combinations.^[97] Overall, the application of tumor organoid cultures in immunotherapy studies could greatly facilitate the discovery of the responsiveness of immune cells to tumors and predict the sensitivity of tumors to immunotherapy.

Tumor organoids in personalized cancer treatment

Personalized medicine, also known as precision medicine, refers to designing the best treatment plan for patients based on personal genome information combined with individual differences. The ability of PDTOs to capture a patient's tumor-specific gene expression and histopathology provides an opportunity for personalized medicine for patients with cancer. The application of PDTOs in drug screening and immunotherapy suggests that PDTOs have great potential for identifying feasible and optimized therapeutic strategies for individual patients. Many PDTOs have been applied in personalized medicine, including ovarian cancer (OC) and CRC organoids.^[98] In OC organoid studies, heterogeneity and responsiveness to chemotherapy and targeted drugs were evaluated.^[98] The results showed that PDTOs exhibit inter-and intra-patient drug response heterogeneity to these anti-cancer drugs, and 88% of patients are hyper-responsive to at least one drug.^[98] This study suggests that PDTOs are a valuable preclinical model that can provide insights into individualized treatment of patients with OC.^[98] In another study, researchers established organoids from 54 patients with CRC and assessed tumor organoids by whole-exome sequencing.^[99] An “organoid scoring” system was developed to evaluate and compare drug responses of matched patients in vitro.^[99] The study confirmed that the “organoid scoring” system helped predict anti-cancer treatment effects.^[99] In conclusion, PDTOs can faithfully reproduce the phenotype, genotype, and physiological and pathological changes of original tumors in vitro. By testing PDTOs, valuable insights into the choice of treatment, optimization of efficacy, and reduction of side effects in individual patients can be gained.

Tumor Organoid Chip

Applications of “organ chip” technology in the tumor field promote preclinical development of cancer therapeutics. Organ chips are a novel microfluidic cell culture system that allows cells to mimic physiology and pathophysiology at the tissue/organ-level in vitro. These devices exhibit multicellular architectures and functionalities, and incorporate physical parameters into the system, including tunable flow rates and different biomaterial matrices. Organ chips are more sophisticated than 3D organoid technologies and offer a more physiological approach to investigating cancer biology. These systems can also incorporate imaging and nanosensor modules, making it easier to evaluate real-time high-resolution images. These microfluidic chip devices can be rationally designed to introduce and monitor multiple responses to various chemicals and microbes.^[100] Numerous microfluidic organ chips have been generated, such as those for the blood-brain barrier,^[101] liver,^[102] small intestine,^[103] retina,^[104] and kidneys.^[105]

Microfluid chip systems have been applied to recapitulate specific events such as tumor growth, expansion, and angiogenesis. Tumor chips have also been used to study tumorigenesis progression, including epithelial-mesenchymal transition (EMT), tumor invasion, and metastasis. Lung chip research indicated that microenvironmental factors secreted by lung endothelial cells and normal lung alveolar epithelial cells support the growth of non-small cell lung cancer cells.^[106] Angiogenesis is an essential process that plays an indispensable role in the development of precursor lesions. Many microfluidic chip systems have been constructed to represent vessel formation in vitro.^[107,108] Miller et al^[109] developed a chip model with primary human clear cell renal cell carcinoma cells and human endothelial cells to study angiogenesis process. A 3D OC model using a microfluidic device was used to investigate gene expression patterns and cell morphology during the EMT process.^[110] Wang et al^[111] developed a microfluidic culture system using human lung cancer cells embedded in an ECM gel to study tumor cell invasion. This advanced microfluidic culture system can precisely control the microenvironment of tumor cells. Tumor cell morphology and cell-cell interactions can be studied in a real-time, high-resolution manner. Undoubtedly, tumor chip technology will strengthen cancer research and improve our understanding of the biological characteristics of tumors.

Challenges and Future

This review briefly discussed the latest advances in the development and application of tumor organoids. However, several challenges must be addressed in this field. First, current tumor organoid protocols are expensive and time-consuming, limiting their further application of tumor organoids. Second, the low culture success rate is a major bottleneck. Although Sachs's group optimized the culture medium to improve the culture success rate of breast cancer organoid,^[61] it remains unclear whether this method can meet the criteria for other types of tumor organoids. Third, although the tremendous potential of tumor organoids has been demonstrated in drug development, the lack of unified culture standards of tumor organoids still hinders the application of tumor organoids in drug screening. Additional restrictions on tumor organoids present an ethical issue. Although tumor organoids are derived from cells or tissues obtained from patients, they should not be considered a morally neutral choice. Therefore, empirical studies and ethical discussions are required.

Modeling human tumorigenesis and cancer heterogeneity and testing the efficiency of anti-cancer therapeutics have been a challenging frontier; this has been significantly limited in traditional 2D cancer cell lines and animal models. Promising tumor organoid models combined with novel biology and synthetic biology technologies, such as PSCs, gene editing, and microfluidic chip systems, will accelerate the discovery of the mechanism of all phases in human cancer progression in vitro [Figure 3]. Tumor studies can be further explored at the genome, transcriptome, proteome, and metabolite levels using tumor organoid platforms [Figure 3]. Moreover, tumor organoids can be used to high-throughput anti-cancer drug screening. Collectively, tumor organoid models provide a powerful platform for fundamental cancer studies and clinical research.

Funding

This work was supported by the National Natural Science Foundation of China (No. 82100821), Key Research & Development Program of Shanxi Province (International Cooperation, No. 201903D421023), the Central Guidance on Local Science and Technology Development Fund for Shanxi Province (No. YDZJSX2021B008), Shanxi Basic Research Program (No. 20210302124406), and Science Research Start-up Fund for Doctor of Shanxi Medical University (Nos. SD2012, XD2019).

Conflicts of interest

None.