Applications of organoid technology to brain tumors

Abstract Lacking appropriate model impedes basic and preclinical researches of brain tumors. Organoids technology applying on brain tumors enables great recapitulation of the original tumors. Here, we compared brain tumor organoids (BTOs) with common models including cell lines, tumor spheroids, and patient‐derived xenografts. Different BTOs can be customized to research objectives and particular brain tumor features. We systematically introduce the establishments and strengths of four different BTOs. BTOs derived from patient somatic cells are suitable for mimicking brain tumors caused by germline mutations and abnormal neurodevelopment, such as the tuberous sclerosis complex. BTOs derived from human pluripotent stem cells with genetic manipulations endow for identifying and understanding the roles of oncogenes and processes of oncogenesis. Brain tumoroids are the most clinically applicable BTOs, which could be generated within clinically relevant timescale and applied for drug screening, immunotherapy testing, biobanking, and investigating brain tumor mechanisms, such as cancer stem cells and therapy resistance. Brain organoids co‐cultured with brain tumors (BO‐BTs) own the greatest recapitulation of brain tumors. Tumor invasion and interactions between tumor cells and brain components could be greatly explored in this model. BO‐BTs also offer a humanized platform for testing the therapeutic efficacy and side effects on neurons in preclinical trials. We also introduce the BTOs establishment fused with other advanced techniques, such as 3D bioprinting. So far, over 11 brain tumor types of BTOs have been established, especially for glioblastoma. We conclude BTOs could be a reliable model to understand brain tumors and develop targeted therapies.

the median survival period of glioblastoma (GBM) patients has only been extended by 3.7 months on average compared to the 1980s, despite advances in neurosurgical resection, chemotherapy, and radiotherapy. [3][4][5] Moreover, multiple novel therapies for brain tumors have recently been developed, including targeted therapy, immunotherapy, tumor vaccines, and oncolytic viruses. However, the limited success of these therapeutics has restricted their clinical application prospect. 6 The key obstacle is the lack of an appropriate model to comprehensively mimic the characteristics of brain tumors, which hampers the investigation of tumor biology and the development of novel therapies and drug screening for precision treatment. 4 Two-dimensional cell culture is convenient and represents accurate molecular signatures in the early generations. 7 However, subsequent generations may show genetic and transcriptional changes owing to spontaneous variations and selection of cells with rapid proliferation. 8,9 It also loses three-dimensional functional cell-cell interactions, which further reduces its applicability under in vivo conditions. Tumor spheroids retain the three-dimensional architecture and physical cell interactions 10 but consist of tumor cells with limited intra-tumor heterogeneity and lack a tumor microenvironment (TME). 11,12 The TME, where stromal interactions, immune responses, and extracellular matrix (ECM) generation occur, plays an important role in tumorigenesis and therapeutic resistance. Brain tumor TMEs include specific cell types, such as neurons, astrocytes, microglia, macrophages, tumorinfiltrating lymphocytes, vascular cells, and fibroblasts.
Compared with two-dimensional cultures and spheroids, patient-derived xenografts (PDXs) can maintain TME. PDXs are generated from surgical tissues transplanted into immunosuppressed rodents and consistently maintain primary tumor phenotypes and heterogeneity. 13,14 However, species differences at the gross neuroanatomical, cellular, and molecular levels have led to varied results 15 ; additionally, low success rate, prolonged latency, and high cost impede the broad application of PDXs. Tumor organotypic explant cultures are established from patient tumors mechanically and preserve the cellular composition and TME as present in situ. 16 While this model has been applied to investigate tumor invasion and drug responses, it showed short-term survival and poor expandability. [16][17][18] The manipulative complexity and the subsequent cellular reaction after mechanical slicing also impeded the application. 19 Altogether, none of the current models for brain tumors are optimal and technical innovations are required (Table 1).
Organoids are three-dimensional cellular self-aggregates that precisely mimic the source tissue and are commonly derived from human pluripotent stem cells (hPSCs) or cancer stem cells (CSCs).
The first brain organoids and brain tumor organoids (BTOs) were reported in 2013 and 2016, respectively. 20,21 Organoids can maintain multiple cellular lineages and preserve complex cell-cell communications. 22,23 Importantly, this model recapitulates the genotype and phenotype, including the heterogeneity of the parental tumor. 20,24 Furthermore, organoids provide a humanized TME to investigate brain tumors. Currently, BTOs can be generated within 1-2 weeks with success rates far higher than those of PDX. They can be cultured for long-term biobank application. 24 Different forms of BTOs can be customized to research objectives and brain tumor features (Table 2). These advantages of BTOs have attracted attention in preclinical research. In this comprehensive review, we introduce the different forms of established BTOs and their characteristics, including their strengths and applications in the study of CSCs, therapy resistance, drug testing, and preclinical research.

| NORMAL B R AIN ORG ANOIDS
In. 2013, Lancaster et al 21 established brain organoids derived from embryonic stem cells (ESCs), recapitulating the three-dimensional structural organization with neural identity and differentiation. The procedure entailed: (1) inducing hPSCs to generate embryoid bodies (EBs); (2) feeding EBs and initiation of germ cells; (3) induction of the neural ectoderm; (4) transfer of neuroepithelial tissues to Matrigel droplets and neuroepithelial bud expansion; (5) brain tissue growth and expansion. 21,25,26 Using this technique, normal brain organoids TA B L E 1 Characteristics of the three mainstream preclinical cancer models.

Tumor organotypic explant BTOs
Basic criteria for preclinical model can be generated within a month and maintained for more than a year, exhibiting the spatial topography identified by region-specific markers ( Figure 1A). 21,25,26 To date, different region-specific brain organoids have been established, including the forebrain, midbrain, hindbrain, choroid plexus, cerebellum, hypothalamus, and pituitary. 27-33 Normal brain organoids, which are differentiated and self-aggregated from hPSCs, also preserve multiple cell types including neuronal and astrocytic sublineages. Oligodendrocytes, 34 vascular endothelium, 35 and microglial cells 36 can be derived using a modified protocol. These features enable normal brain organoids to mimic the human brain to a high degree, and extensively model neural diseases.

| B R AIN TUMOR ORG ANOIDS DERIVED FROM PATIENT SOMATIC CELL S
Some types of brain tumors resulting from specific germline mutations emerge and grow during neurodevelopment, but lack an appropriate in vivo or in vitro model. 37,38 Several studies have collected somatic cells (blood mononuclear cells and fibroblasts) from patients with these tumors, reprogrammed these cells to induced pluripotent stem cells (iPSCs), and generated brain organoids from reprogrammed iPSCs with intrinsic genetic defects.
During the growth of this form of brain organoids, brain tumors initiate at a specific developmental point and proliferate within the brain organoids, exhibiting morphological progression, biological behavior, and signaling mimicking human disease ( Figure 1B) Figure 1C). This form of BTOs can be used to investigate the role of genetic mutations in tumorigenesis and tumor development (Table 3). EGFRvIII is a common mutation in meeting the needs of precision oncology. 44 Brain tumors have been suggested to arise from or be driven by neural stem-like cells. [47][48][49][50][51] Recurrent mutations in brain tumors also affect neurodevelopment. 52,53 Similarly, some perturbed signaling pathways in neurodevelopment lead to the initiation and proliferation of brain tumors. 54,55 Therefore, tumorigenesis may be closely associated with neurodevelopment. Brain organoid growth mimics neurodevelopment and contains multiple cellular lineages in the human brain. By performing genetic manipulation at different stages of organoid establishment, brain organoids can be developed as an optimal model to study tumorigenesis, especially for pediatric brain tumors that appear during active neurodevelopment. For example, atypical teratoid rhabdoid tumors (ATRTs) are challenging pediatric brain cancers caused by the inactivation of SMARCB1 during neurodevelopment. During neuronal differentiation in brain organoids, SMARCB1 was knocked down using CRISPR/Cas9. The

| B R AIN TUMOROIDS
Based on the tumoral property of infinite proliferation, the models of brain tumors can be generated from brain tumor specimens, such Metabolism is a key facet of glioma growth and metastasis, and cannot be accurately represented outside the influence of the TME.
The diverse microenvironment in Hubert GBOs also triggers metabolic alterations that can serve as therapeutic targets. For example, hypoxia can induce lipid droplet biogenesis to protect cells from oxidative damage and provide energy. 58 Lipid enrichment has been identified in hypoxic GBOs. Because GSCs are mainly distributed in the outer zone, they have low lipid droplet accumulation compared to non-GSCs. Deeper lipidomic analysis showed that GSCs contained reduced levels of major classes of neutral lipids, but displayed higher polyunsaturated fatty acid production compared to non-CSCs, due to high expression of fatty acid desaturase (FADS1/2).
Upon knocking down FADS1/2, the viability and self-renewal ability of GSCs are damaged, indicating a therapeutic target. 59 Therefore, Hubert GBOs provide a platform to identify the abnormal metabolisms and target them.

TA B L E 3
Summary of tumorigenesis studies using BTOs.

Stage of mutations occurring Findings References
Glioma EGFRvIII OE hESCs Excessive gliogenesis at the expense of neurogenesis 138 1.  (Table 4). 20 By using the techniques of cellular tracking and sorting, specific subpopulation of GBM cells could be identified from GBOs to investigate. 60,61 For example, by isolating quiescent GBM cells, bioinformatics analyses and functional assays showed F I G U R E 2 Establishment of brain tumoroids and their applications. Hubert brain tumoroids are generated within matrigel as BOs, composed of multiple TME due to gradient exposure to nutrients and oxygen. Jacob brain tumoroids could be generated within 2 weeks without cell-cell dissociation and matrigel, which are the most clinically applicable BTOs. The diagram was created with BioRe nder.com.
that hypoxia and TGFβ signaling may drive the identity of quiescent GBM cells, providing a potential mechanism to ameliorate therapy resistance. 60 While targeting populations of resistant cells is promising, complicated cell-cell crosstalk within GBOs could also be a potential mechanism supporting therapy resistance. Tunneling with TNTs to participate in therapy resistance. 65 Mitochondrial transfer between tumor cells can provide metabolic support and rescue aerobic respiration for recipient tumor cells in response to treatment-related stress, revealing a partial mechanism of therapy resistance in patients. 64 Therefore, glioblastoma cells in GBOs may overcome therapy through cooperation in the TME, aided by complicated cellular connections, which may be closer to the responses observed in parental tumors and an appropriate model to study therapy resistance.
In 2020, Jacob et al reported a revolutionary method that could generate GBM tumors directly from resected GBM samples rather than via dissociation, retaining native cell-cell interactions ( Figure 2).
By optimizing a chemically defined medium, they cultured tumoroids with few exogenous growth factors and no Matrigel to minimize clonal selection and decrease potential treatment confounders.
Besides, growing without matrigel, which is a kind of undefined and complex ECM, could also avoid the unstability and matrigel-specific effects. 66 Tumoroids derived from patients without dissociation also retained a heterogeneous cellular composition, including immune and endothelial cells. In this model, immune and endothelial cells can persist for more than 8 months and gradually decrease over time.
Moreover, this method generated GBM organoids approximately 1-2 weeks after initial surgical resection with high fidelity and an overall success rate of 91.4%. 24,67 Jacob GBOs depend on the gradient exposure of growth-supporting materials and precisely recapitulate the intra-and inter-tumoral heterogeneity from genotype to phenotype. Profiling of somatic variants and copy number variants (CNV) in GBOs is largely similar to tumors derived from different patients, indicating inter-tumoral heterogeneity. The GBOs derived using this method from different subregions in the same patient also showed subregion-specific mutations. 24 Specifically, Jacob GBOs preserved EGFR mutation, a driver in GBM, which was rapidly lost in two-dimensional culture. 67 GBOs also showed high similarity to parental GBM samples at the transcriptome level for over 12 weeks. Even for the macrophage/ microgial-related genes, the expression was comparable between GBOs and parental tumors for 2 weeks. Due to the disability of replication and immortality for non-tumor cells, the most differentially downregulated genes between GBOs and parental tumors were immune-and blood-related genes, indicating incomplete retention of immune cells and blood cells over a long period of time relative to in vivo conditions. Furthermore, scRNA-seq analysis showed that cellular and molecular signatures in GBOs were highly similar to those of the parental tumor, maintaining cell-type heterogeneity and molecular properties. 24 Finally, the GBOs preserved similar morphology compared to parental tumors and could be transplanted into the mouse brain intact, displaying not only invasiveness, but also angiogenesis. TA B L E 4 Different therapy responses of preclinical in vitro models.
The omics revolution has led to the identification of various targets and a more comprehensive view of the molecular signaling underlying brain tumors through the integration of genomic, epigenomic, transcriptomic, metabolomic, and proteomics data. 68 confirming that mutation analysis alone without functional testing is insufficient to predict response to treatment. GBOs could also provide a platform to study the biological mechanisms of novel effective drugs that have not been reported in GBM. For example, the proteasome inhibitor carfilzomib was identified as a targeted drug from high-throughput screening of 320 drugs combined with proteomic and bioinformatic analyses and a series of functional assays in GBOs. 80 GBOs can also be used to predict the effects of combination therapy, which can improve outcomes in patients with malignancies. 81 The responses of GBOs to combination therapy showed greater effects than those of monotherapy. 82 While immunotherapy has achieved great success in several types of cancer, its efficacy on brain tumors is limited. 83  Oncoprotein IDH enzymes were observed pervasively using immunohistochemistry. More importantly, 2HG accumulation was observed in LGG tumors, comparable to parental tumors, using liquid chromatography-mass spectrometry analysis. 86 Meningioma is the most common primary tumor of the brain and is derived from the neural crest. 1 Meningiomas have a high proportion of interstitial matrix; therefore, dissociating the original tumor samples using enzymatic methods is hard to perform without disrupting cell viability. Most meningiomas are benign and proliferate slowly. This has led to a lack of models for this disease. 87 Organoid techniques can be used to establish meningioma tumoroids by embedding the dissociated meningioma cells into matrigel with the supplementation of growth factors similar to the generation of cerebral organoids. 88 The establishment was within 2 weeks with 100% success rates, recapitulating multiple characteristics of the parental tumors. 88

| B R AIN ORG ANOIDS CO -CULTURED WITH B R AIN TUMOR S (BO -BT )
To study tumor invasion and cell-cell interactions between tumors and normal cells, brain organoids can be co-cultured with brain tumor cells/spheres ( Figure 3). CSCs are a group of cells that most recapitulate tumors molecularly and phenotypically and are most commonly cocultured with organoids. 45,91,92 Depending on the tissue-clearing method and microscopy technique used, invasive protrusions, and microtube networks formed in brain tumor cells can be observed and measured as surrogates of invasive ability. Reporter genes, such as luciferase, can also be ectopically expressed in brain tumor cells for real-time live imaging. Three types of co-culture patterns were established. First, co-cultured brain tumor cells with iPSCs and then induced brain tumor organoids. In this pattern, recurrent GBM stem

cells (GSCs) exhibited enhanced invasiveness compared to primary
GSCs at an early stage. However, both recurrent and primary GSCs stopped growing after day 10 and survived for up to day 20. Second, brain tumor cells were implanted into the established brain organoids.
Distinguishing invasiveness between recurrent and primary GSCs was also significant. The invasive protrusions and microtube-like structures of surgical GBM tumor specimens resembled GSCs formed in this pattern. Third, brain tumorspheres were co-cultured with established brain organoids. In this system, individual GSCs invade brain organoids, showing a profound tropism of GSCs to the brain tissue. While GSCs were compact in spheres, invasive protrusions, and microtubes could not be quantified.

oids (GLICOs) often takes over a month, which is longer than that for
GBOs. However, GLICOs are reported to be the most accurate models when compared with the two-dimensional, GBOs, and PDX models. Among these models, GLICOs exhibited the highest correlation with parental tumors at many levels, such as similar transcriptomes, diversity of cellular states, and strong stemness and invasiveness signatures. 91 GSCs in GLICOs also preserved key genetic and signaling components of the parental tumors, 93  tumor microtubes compared to GSCs in GBO, which may also be due to the TME. The interactions of brain tumor cells with TME components and how they affect tumor growth and behavior are now gradually being revealed, although the specific mechanisms remain unknown. 94 Brain organoids provide a platform for studying the mechanism of interaction between the brain and brain tumors. Compared with PDX, which also contains a TME, brain organoids are humanized, manipulable, and fast for the establishment with a higher success rate, allowing real-time imaging. scRNA-seq analysis of GBM cells before and after co-culture with brain organoids showed that GBM cells could sense the neuron once co-cultured and upregulate the gene expression related to dispersion and ligand-receptor interaction between GBM and organoid cells. Therefore, targeting and breaking the cell-cell connections could be a novel therapeutic strategy. 95 TMs and TNTs have been observed in GLICOs. Their connections with normal cells in brain organoids may be the reason for the enhancement because tumor growth in the brain has been shown to require neighboring cellular activity. 96 TMs can form synapses with neurons and astrocytes and drive tumor progression in primary brain tumors 97,9 or brain metastases. 99 In GLICOs, TMs are found in an interconnected network that can effectively propagate calcium signals for cellular communication. They deeply penetrate the brain organoids and provide potential routes for invasion, proliferation, and interconnection over long distances. 100 In brain metastases co-cultured with brain organoids, astrocytes also form GAP junctions with metastatic lung cancer cells, which promotes tumor growth. 101  Invasiveness is linked to cancer-TME crosstalk. By changing the characteristics of brain organoids, the consequent invasive capacity can be altered, thus offering an opportunity for a deeper understanding of the invasive process of brain tumors and discovery of potential therapeutic targets. For example, GSCs exhibited faster and deeper invasion in mature brain organoids compared to younger ones; this was linked with the synaptic protein Neuroligin-3, which is generated by mature neurons only. By blocking Neuroligin-3 function, the invasiveness of GSCs was markedly reduced, indicating a potential target. More variants can be changed in brain organoids to further study the interactions between the brain and brain tumors, such as different brain regions or brain organoids derived from syngeneic and non-syngeneic iPSCs.
Patient-derived meningioma cells also exhibited phenocopy invasiveness when co-cultured with brain organoids. In brain organoids, higher grades of meningioma cells exhibited an invasive phenotype, and lower grades of meningioma cells only formed tumorspheres at the surface of the brain organoids. Meningioma cells co-cultured with brain organoids showed the greatest number of overlapping genes with parental tumors when compared to two-and threedimensional monocultures. CDH2 and PTPRZ1 have been identified as oncogenes driving the tumorigenesis of meningiomas in brain organoids, indicating potential targets. 103 With the incorporation of brain components in organoids, GSCs exhibit resistance to chemotherapy and radiotherapy compared with the cells in two-dimensional culture. 93 This means that the BO-BT model can also serve as a model for studying therapy resistance and is even more suitable than brain tumors because of the preservation of TME. The tumor cells in BO-BT are more sensitive to therapy than cell lines, including GBM treated with TMZ 79 and NSCLC brain metastases treated with Gefitinib. 103 However, the reason remains unknown, and though tumor cells in two-dimensional culture are speculated to be restricted in terms of growth and malignant behavior and thus protected from chemotherapy or radiation therapy, further investigation is required. 79 GLICOs can be used to evaluate the efficacy of novel therapies in preclinical trials, and have been widely used for this purpose, including the evaluation of cytotoxicity, invasion inhibition, and radiother- in GLICOs showed decreased tumor growth and stronger radiotherapy effects, which could not be modeled in PDX. 104 Oncolytic viruses are emerging antitumor therapies that selectively target, internalize, and kill tumor cells while sparing normal cells. 105 The Zika virus can enter the brain, and this viral infection can lead to neonatal microcephaly and other neurodevelopmental defects; infected adults are often asymptomatic. 106 Recently, Zika virus was engineered as an oncolytic virus for patients with brain tumors by targeting SOX2 cells. [107][108][109] Normally, SOX2 is a transcription factor that is expressed at high levels during human neurodevelopment and contributes to the induction of pluripotency. 110 SOX2 is also highly expressed in many brain tumor stem cells (BTSCs) such as GBM and medulloblastoma. 111 Using humanized BT-BO, the effects of oncolytic viruses on both tumor cells and normal cells could be evaluated. In BT-BO, Zika virus preferentially and effectively infected and killed BTSCs, including GBM, ATRTs, and medulloblastoma, but had limited effects on mature brain organoid size. [107][108][109] In the future, more oncoloytic viruses that can enter the brain may be engineered to target brain tumor cells, and BTOs represent a major opportunity for preclinical studies of this emerging treatment modality.
The BO-BT models can also be used to evaluate the safety and tolerance of novel therapies. The neurological impairment caused by tumors and associated therapies can have serious lifelong consequences on daily function and deeply influence the quality of life, including fatigue, memory loss, emotional distress, and sleep disorders. Minimizing the side effects of therapies is as equally important as inhibiting the tumor. The identification of the effective and tolerable range of dosage and therapeutic intensity is important in preclinical trials. 112,113 Brain organoids as a "mini brain" can be used as a surrogate to evaluate the side effects of antitumor therapy and provide valuable information for clinical decisions. For example, TTFields showed inhibitory effects on GBM cell proliferation at both 75% and 100% duty cycles; the neurotoxicity of brain organoids at 75% was less prominent than at 100%, indicating that 75% may be a better choice. 114 The targeted drug UM-002 employed in GLICO showed that higher concentrations (>500 nM) reduced GBM cell proliferation but also induced toxicity in normal brain organoids. In a dose-response study, 100 nM was found to not only be cytotoxic for GBM cells, but also safe for brain organoids. 115 The neural side effects of radiation 116,117 and Zika virus 62,63 have also been evaluated in brain organoids. In addition, normal organoids of other organs that are frequently impaired in systemic therapy, such as the heart, liver, and stomach, can be used to test toxicity. 118,119 Notably, multi-organ organoids with tumoroids have been constructed into a connected system with circulation using F I G U R E 3 Establishment of brain organoids co-cultured with brain tumors and their applications. BO-BTs are the most similar BTOs as original tumors, which could model humanized interactions between tumor cells and brain components in vitro. The diagram was created with BioRe nder.com.
microfluidic techniques to synchronously test toxicity and treatment efficacy. 119

| BTO E S TAB LIS HMENT US ING OTHER ADVAN CED TECHNIQUE S
Internal hypoxia and cell death due to insufficient diffusion of culture media and oxygen are prominent causes for current brain organoid culture methods generating insufficient numbers of mature neurons. However, organotypic slices can bypass diffusion limitations to prevent cell death and enhance neuronal maturation and viability. 120,121 Sliced brain organoids co-cultured with brain tumor cells/spheres have been identified as a feasible method for assessing how mature neurons interact with brain tumors. For example, network structures comparable to those of synapses between neurons and GSCs have been observed in this system. 92  bioprinting, 3D bioprinting is combined with smart materials that respond to stimuli, and this has been used to form 4D organoid arrays.
This technique not only allows high-throughput drug screening, but also reduces manual operation, thus simplifying the process and increasing reproducibility. 82 However, the ability of these bioprinted organoids to recapitulate cellular heterogeneity and organization comparable to that of parental tumors remains uncertain. In other words, presently, it is more appropriate to regard them as "biofabricated spheroids" until further characterization studies prove their ability to recapitulate their source tissue. 125,126

| CHALLENG E S AND FUTURE PROS PEC TS
BTOs are an important new platform for understanding tumor development and developing precision oncology for brain tumors.
Below, we detail the current limitations and future prospects of this technology: 1. Accurate recapitulation of brain cellular architecture Brain tumors mostly occur in adults. 1 Although the current brain organoids are remarkably similar to the fetal brain, the mature components in brain organoids are insufficient, and neural functions in brain organoids differ from those in the adult brain. 127,128 Some define "mature" brain organoids as at least 6 months old-a time period in which most NPCs differentiate into neurons and astrocytes and express mature markers. 128,129 A more extended culture period (>9 months) was proposed to facilitate greater functional maturity, including the formation of dendritic spines and active neuronal networks. 130 The main reasons for immaturity in BOs include (1) diffusion limitation of culture media, (2) non-physiological ECM, (3) and missing cell types, such as microglia. 131  press hETV2, leading to the acquisition of several blood-brain barrier characteristics and enhanced functional maturation of BOs. 35 Intracerebral implantation of BOs into immunodeficient mice also generated blood vessels in the BOs. 134 However, to date, these BOs have not been used in conjunction with BTs. The lack of persistent immune cells is also a defect in BTOs. Co-culturing could be a solution, as was discussed in this review.

Expanding clinical relevance for BOs and BTOs
BTOs could potentially be used to guide personalized therapy

Standardization and automatization of BO and BTO techniques
As emerging state-of-the-art models, techniques for BOs and BTOs are constantly being optimized. No acknowledged standard protocol exists for all BOs or BTOs. Inter-and intra-batch variability are common across studies because BO and BTO generation largely depends on self-patterning and self-organization of PSCs/BTSCs without guided differentiation. 21,25,130,135 The complexity of manual processes in culture is also a major source of variability and error that hampers large-scale production. Developing standardized protocols and automatized devices will be helpful in ensuring authenticity and expanding the application of precision oncology. Furthermore, organoid factories can be used for high-throughput drug screening and target investigations. Three-dimensional bioprinting, computational automatic techniques, and microfluidic techniques can help achieve this goal.

Identification of pathogenic factor driving brain tumors
Although the pathogenesis of brain tumors is mainly related to genetic mutations, microenvironmental factors and their relationship with susceptibility are also important factors leading to brain tumors. Ionizing radiation (hazardous factor) and history of allergies (protective factors) are well-documented risk factors for brain tumors. Other possible risk factors have also been reported by analyzing large clinical databases that require further validation. 136 Brain organoids can be used to confirm the relationship between tumorigenesis and exposure to risk factors in future studies.
For example, exposing brain organoids to hormonal contraception could help determine if associations exist between maternal hormonal contraception use and central nervous system tumors. 137 Through the genetic manipulation of brain organoids, the relationship between risk factors and genetic susceptibility may also be revealed.

| CON CLUS IONS
Since the BOs and BTOs emerged in 2013 and 2016, the worldwide application of organoid technology has resulted in remarkable advances in the study of precision oncology for brain tumors.
In this review, we described the current literature on the establishment of several forms of BTOs and how precisely they modeled different types of brain tumors. Additionally, the promoting effects of BTOs for deeper biological understanding and personalizing therapy for brain tumors are also described. In summary, even though current BTOs are facing some challenges and required op-

CO N FLI C T O F I NTE R E S T S TATE M E NT
None.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that supports the findings of this study are available in the supplementary material of this article.