TECHNOLOGY EXPLAINER
Organoids and Spheroids
— taking cells from flat to three-dimensional, where diffusion sets the size limit
Cells spread flat on the bottom of a culture dish live in a very different environment from cells inside the body. Spheroids, in which cells are gathered into round clumps, and organoids, in which stem cells self-organise into small organ-like structures, are the leading forms of 3D culture that close that gap. In 2025 the US FDA set out a policy to promote the use of organoids and similar models in drug safety testing. Yet they face variable size, a dying core and poor reproducibility, and many of those problems are, at root, problems of materials and mass transfer.
- Organoids and spheroids (the short version)
- How 2D and 3D culture differ
- How spheroids and organoids differ
- Milestones — from spheroids in 1971 to the FDA in 2025
- How they are made — low-attachment plates, hanging drops, stirring and Matrigel embedding
- A materials engineer's view (1): a non-stick surface and a U-shaped bottom set performance
- Our calculation: at what size does the core die?
- A materials engineer's view (2): diffusion sets the size limit
- Uses in drug discovery and toxicity testing
- Challenges — size uniformity, necrotic cores and reproducibility
- A materials engineer's view (3): much of the reproducibility problem is a scaffold problem
- Glossary / References / Claim-to-source audit
Sourced = stated in published material or a paper (link given)
Our calculation = a figure this article derived, with the assumptions spelled out
Not yet confirmed = a plan, outlook or research-stage result with no confirmed track record in practice
Structural summaries and readings about materials design are marked separately as Commentary.
This article explains technology and is not medical advice. Results from animal studies or research-stage work do not show therapeutic effect in humans.
1. Organoids and spheroids (the short version)
- Spheroid: cells gathered into a round clump (a sphere). Clumps of cancer cells, "multicellular tumour spheroids", have long been used as models that mimic normal tissue or solid tumour tissueSourced
- Organoid: a three-dimensional tissue that self-organises from stem cells (pluripotent stem cells, or fetal or adult stem cells) and mimics the key functions, structure and biological complexity of an organ. The definition and scope are, however, said to be debatedSourced
- Why the interest: they let human cells build structures close to human tissue, which makes them usable as models of development and disease and for testing drugs. In April 2025 the FDA announced that it would promote the use of human "organoids" and organ-on-a-chip systems in drug safety testingSourced
The moment cells go 3D, they become subject to the limits of diffusion. Oxygen can only soak in from the outside of the clump, and in one colorectal cancer cell spheroid the depth oxygen reaches (the diffusion limit) was measured at about 232 µmSourced. That means beyond a diameter of about 460 µm, oxygen no longer reaches the centreOur calculation. Many of the challenges of organoids and spheroids come down to this mass-transfer problem.
2. How 2D and 3D culture differ
Vinci and colleagues wrote in 2012 that "the evidence that three-dimensional tumour cell cultures better reflect the complex in vivo microenvironment than simple two-dimensional monolayers is overwhelming, including gene expression, signalling pathway activity and drug sensitivity"Sourced. They also pointed out that many 3D techniques are slow or lack reproducibility, so standardised, rapid protocols are urgently neededSourced.
A review by Duval and colleagues says that 3D culture still faces challenges in tissue-to-tissue interfaces, the mechanical microenvironment, and the spatial and temporal distribution of oxygen, nutrients and metabolic wasteSourced. Hirschhaeuser and colleagues note that, unlike monolayer culture, spheroids reflect the three-dimensional cellular relationships of tumours in vivo and the pathophysiological gradients that matter for therapySourced. The advantage of going 3D (there are gradients) and its drawback (there are gradients) are two sides of the same phenomenon (our commentary).
3. How spheroids and organoids differ
| Point | Spheroid | Organoid |
|---|---|---|
| Starting point | Cell lines (such as cancer cells) or primary cells. Spheroids of primary human hepatocytes are one example | Pluripotent stem cells (ES and iPS cells), or fetal or adult tissue stem cells |
| How it forms | Cells gather into a clump (aggregation) | Stem cells multiply and, while differentiating, self-organise into an organ-like structure |
| Structure | A simple sphere. As it grows, layers form: an outer rim, a hypoxic zone and a necrotic core | Epithelium, a lumen, and structures such as crypts and villi. Contains several cell types |
| Typical methods | Low-attachment U-bottom plates, hanging drops, stirred or rotating culture | Embedding in an extracellular matrix such as Matrigel with added growth factors. From pluripotent stem cells, stepwise induction that mimics development |
| Main uses | Testing anticancer and other drugs, assessing liver toxicity, tissue models | Models of development and disease, drug testing, patient-derived tumour organoids |
| Weak points | A large one develops a necrotic core. Depending on the method, sizes vary | Heterogeneity, variable cellular complexity, difficulty controlling self-organisation. Often lacks stromal, vascular and immune components |
The content of each cell is Sourced (Zhao et al. [Ref. 2], Lancaster and Knoblich [Ref. 1], Kelm et al. [Ref. 5], Vinci et al. [Ref. 6], Bell et al. [Ref. 9], Spence et al. [Ref. 10], Sato et al. [Ref. 3]). Splitting them into two contrasting columns is, however, this article's own framing, and where the boundary lies varies between papers.
The review by Lancaster and Knoblich recalls that classic experiments half a century ago showed that vertebrate cells, once fully dissociated, could reaggregate and rebuild the structure of the original organ, and explains that 3D cultures that exploit this property, made from tissue stem cells or embryonic stem cells, are called organoidsSourced. Zhao and colleagues note that organoid research has been intertwined with 3D culture, stem cells and tissue engineering for more than a century, and that there has been much debate about its definition and scopeSourced.
4. Milestones — from spheroids in 1971 to the FDA in 2025
| Year | What happened |
|---|---|
| 1971 | Sutherland and colleagues report "growth of multicell spheroids in tissue culture as a model of nodular carcinomas"Sourced |
| 2003 | Kelm and colleagues report the hanging-drop method, making one spheroid per drop in droplets hanging from a microtitre plate, with close to 100% efficiencySourced |
| 2009 | Toshiro Sato and colleagues (Hubrecht Institute, the Netherlands) report that a single sorted Lgr5-positive stem cell can build an intestinal organoid with crypt-villus structures. They conclude that "intestinal crypt-villus units are self-organising structures, which can be built from a single stem cell in the absence of a non-epithelial cellular niche"Sourced |
| 2011 | Mototsugu Eiraku and colleagues at the RIKEN Center for Developmental Biology report that an optic cup (the rudiment of the retina) forms autonomously from aggregates of mouse ES cellsSourced. The same year, Spence and colleagues report a method of making intestinal organoids from human pluripotent stem cells by switching growth factors in a timed sequenceSourced |
| 2013 | Lancaster and colleagues make "cerebral organoids" with several brain regions from human pluripotent stem cells and model microcephaly, which had been hard to reproduce in miceSourced. The same year, Takanori Takebe and colleagues at Yokohama City University make iPS cell-derived liver buds and report that, once transplanted, they connected to host vessels within 48 hours to form functional vasculatureSourced |
| 2015 | van de Wetering and colleagues report a "living biobank" of tumour organoids established from 20 consecutive colorectal cancer patients, usable for high-throughput drug screeningSourced |
| 2016 | Gjorevski and colleagues culture intestinal stem cells and organoids in synthetic hydrogels, without animal-derived matricesSourced |
| 10 April 2025 | The FDA publishes a plan and roadmap for phasing down animal testing for monoclonal antibodies and other drugs, saying it will promote the use of human "organoids" and organ-on-a-chip systemsSourced |
Sources: Sutherland et al. [Ref. 4], Kelm et al. [Ref. 5], Sato et al. [Ref. 3], Eiraku et al. [Ref. 11], Spence et al. [Ref. 10], Lancaster et al. [Ref. 12], Takebe et al. [Ref. 13], van de Wetering et al. [Ref. 14], Gjorevski et al. [Ref. 15], FDA [Refs. 18 and 19]. Affiliations follow the author information in each paper.
The Takebe study reports that transplanting liver buds rescued animals in a model of drug-induced lethal liver failure, but the authors themselves say that "further efforts will be needed to translate this into treatment for patients"Sourced. These are results from animal experiments, and this article does not present any of these studies as a treatment in humans.
5. How they are made — low-attachment plates, hanging drops, stirring and Matrigel embedding
| Method | What the primary sources say |
|---|---|
| 1 Ultra-low-attachment (ULA) U-bottom plates | Vinci and colleagues describe using ultra-low-attachment 96-well round-bottom plates that need no coating to prevent cell attachment, in which a tumour cell suspension forms a 3D structure within 24 to 48 hours, with the well shape producing a single spheroid in the centreSourced. For comparison, they also show spheroids made in U-bottom ULA plates, agar-coated flat-bottom plates, poly-HEMA-coated 24-well plates and a rotating culture deviceSourced |
| 2 Hanging drop | Kelm and colleagues say that the conventional methods (growth on non-adherent surfaces, suspension, scaffolds) have drawbacks such as needing manual selection to get a uniform population and using non-physiological matrices, whereas the hanging-drop method gives one spheroid per drop with close to 100% efficiency. For HepG2 cells, the coefficient of variation (CV) of size was 10 to 15%, narrower than 40 to 60% for growth on non-adherent surfacesSourced |
| 3 Stirring and rotation | Zhao and colleagues say that necrosis in the core of brain organoids can be partly solved by orbital shaking, rotating bioreactors and continuously stirred bioreactors, among othersSourced |
| 4 Matrigel embedding | Sato and colleagues' intestinal organoids were grown in an extracellular matrix with a growth factor cocktail of EGF, Noggin and R-spondin (according to a review by Clevers and colleagues)Sourced. Zhao and colleagues describe dissociating tissue enzymatically and then seeding it into MatrigelSourced |
| Others | Zhao and colleagues cite controlling aggregation with microwell arrays, and using droplet microfluidics to make one organoid per wellSourced |
6. A materials engineer's view (1): a non-stick surface and a U-shaped bottom set performance
Lay the methods in Section 5 out again with a materials eye, and most spheroid-making techniques turn out to be techniques for keeping cells from sticking to a surface.
- Vinci and colleagues used round-bottom plates with an ultra-low-attachment surface to which cells do not attach even without coating. For comparison they also show methods that prevent attachment by coating with agar or poly-HEMASourced
- Two things are at work here: surface chemistry (cells do not attach) and geometry (a round bottom gathers the cells in one place). Vinci and colleagues write that "the shape of the well promotes the formation of a single spheroid in the centre"Sourced
And Kelm's figures show that differences in method become differences in quality directly. The size CV was 40 to 60% for aggregation on non-adherent surfaces and 10 to 15% for hanging dropsSourced. The variation is about a third to a sixthOur calculation (40 ÷ 15 ≈ 2.7; 60 ÷ 10 = 6).
Non-fouling coatings made from hydrophilic polymers, suppression of protein adsorption on the surface, and moulding of fine concave shapes — these are exactly the technologies of medical plastics, surface modification and precision moulding. High-throughput spheroid testing calls for hundreds of "clumps of the same size" on a single plate. A surface and moulding precision that guarantee "every well, equally non-stick" translate directly into the reproducibility of the test (our commentary).
7. Our calculation: at what size does the core die?
Vinci and colleagues write that tumour spheroids larger than 500 µm in diameter often have a hypoxic zone and a necrotic coreSourced. Grimes and colleagues determined experimentally that in spheroids of the colorectal cancer cell line DLD1 the depth oxygen reaches (the diffusion limit) is 232 ± 22 µmSourced. They also assumed an oxygen diffusion coefficient close to that of water, 2 × 10⁻⁹ m²/s, and estimated the radius of a DLD1 cell at 7.42 ± 0.47 µmSourced. Let us use these two numbers to check the orders of magnitude.
- Critical diameter: the 232 µm diffusion limit is the depth oxygen reaches from the outside. Taking it as the radius, diameter = 2 × 232 = about 464 µm (about 420 to 508 µm with ± 44 µm)Our calculation
- Assumptions: each cell is a sphere of radius 7.42 µm, and cells fill the spheroid at a volume fraction of 0.64 (a rule of thumb for random close packing of spheres)
- Volume of one cell: 4/3 × π × (7.42 µm)³ = about 1,711 µm³
- Cells in a spheroid of critical diameter: 4/3 × π × (232 µm)³ × 0.64 ÷ 1,711 µm³ = about 20,000Our calculation
- Working back from cell count to diameter: 1,500 cells → about 197 µm; 5,000 → about 294 µm; 10,000 → about 371 µm; 20,000 → about 467 µmOur calculation
Assumptions and limits: the diffusion limit changes with cell type, oxygen consumption rate and the surrounding oxygen partial pressure. The Grimes value was measured for DLD1 cells at a surrounding oxygen partial pressure of 100 mmHgSourced. The packing fraction of 0.64 is our assumption; real tissue contains extracellular matrix and gaps. "About 464 µm" is a calculation to check that it agrees in order of magnitude with Vinci's statement about spheroids "larger than 500 µm".
Bell and colleagues seeded 1,500 primary human hepatocytes per well to make spheroids and reported no obvious necrosis even after five weeks in cultureSourced. Hepatocytes differ in size from DLD1 cells, so the numbers in Fig. 6 do not apply directly, but this is consistent with a design that keeps clumps well below the critical diameter (our commentary).
8. A materials engineer's view (2): diffusion sets the size limit
The Grimes model treats a spheroid as a sphere into which oxygen diffuses from the surface and in which the cells consume it at a constant rate, and finds the radius at which both the oxygen concentration and its gradient fall to zero at the centreSourced. This has the same form as the problem of a reactant running out inside a catalyst particle, described by the effectiveness factor and the Thiele modulus (our commentary).
Seen this way, the ways to avoid necrosis fall into three groups (this article's framing).
- Keep them small: make clumps smaller than the critical diameter. This is the thinking behind choosing how many cells to seed per spheroid. Zhao and colleagues say organoids need to be split periodically into small cell clusters and replatedSourced
- Raise the outside concentration, thin the boundary layer: keep the surrounding liquid moving by shaking, rotating bioreactors or continuous stirring. Zhao and colleagues say necrosis in the core of millimetre-sized brain organoids can be "partly" solved this waySourced
- Build channels inside: create flow paths that stand in for blood vessels. Zhao and colleagues cite an example in which kidney organoids matured and became vascularised inside a perfused organ-on-a-chipSourced
The third is a question of the materials that form the channels, the materials of containers and membranes that let oxygen through, and the devices that create flow. That Takebe's liver buds "matured once transplanted and connected to host vessels"Sourced can be read the other way round: in the dish, the lack of blood vessels is a barrier to maturation (our commentary). Materials and device technologies that control mass transfer could be the key to raising the "size ceiling" of organoidsNot yet confirmed.
9. Uses in drug discovery and toxicity testing
(1) Anticancer drug testing: drugs act differently in 2D and 3D
Vinci and colleagues report establishing a set of assays that analyse tumour spheroids made in U-bottom ULA plates by automated imaging, and finding that sensitivity to molecularly targeted drugs differs between 2D and 3D cultureSourced. Hirschhaeuser and colleagues say spheroids could help weed out unpromising candidates before animal and clinical studies, or find promising drugs that 2D assays would missSourced.
(2) Liver toxicity testing: long-lived spheroids
Bell and colleagues note that monolayer cultures of primary human hepatocytes are hard to use because they dedifferentiate rapidly, and report that spheroids in chemically defined, serum-free conditions kept their morphology, viability and hepatocyte-specific functions for at least five weeksSourced. Chronic exposure greatly increased sensitivity, and they say they could detect the toxicity of a series of hepatotoxic compounds at clinically relevant concentrations. As an example, repeated dosing reproduced the chronic toxicity of fialuridine, which conventional in vitro systems had failed to detectSourced.
(3) Patient-derived tumour organoids
van de Wetering and colleagues report that tumour organoids established from 20 colorectal cancer patients closely recapitulated several properties of the original tumours, and that high-throughput drug screening could detect gene-drug associationsSourced. The authors say organoid technology may fill the gap between cancer genetics and patient trials and allow personalised treatment to be designed, but this is an outlookNot yet confirmed.
(4) Regulatory moves: the FDA roadmap
- At the end of 2022 the US Congress passed the FDA Modernization Act 2.0, which explicitly allowed non-animal alternatives, such as cell-based assays and computer models, to support investigational new drug (IND) applicationsSourced
- The first target is monoclonal antibodiesSourced
- It describes organoids as "self-organizing cell cultures that model the architecture and function of native tissues (e.g., liver organoids, intestinal organoids)"Sourced
- It aims, over the long term (three to five years), to make animal testing "the exception rather than the norm" in preclinical safety and toxicity testingNot yet confirmed
In its announcement of 10 April 2025, the FDA says it will promote the use of human "organoids" and organ-on-a-chip systems that mimic the liver, heart, immune organs and othersSourced. The three-to-five-year timeframe is the FDA's target, and how far animal testing will actually be replaced could not be confirmed at the time of writing.
10. Challenges — size uniformity, necrotic cores and reproducibility
(1) Size uniformity
As Sections 5 and 6 showed, the spread in size changes greatly with the method (CV of 10 to 15% versus 40 to 60%)Sourced. In the calculation behind Fig. 6, diameter scales with the cube root of cell number, so doubling the cell count makes the diameter only about 1.26 times largerOur calculation. Put the other way, variation in diameter looks smaller than variation in the number of cells seeded, so judging uniformity by diameter alone could miss variation in cell number (our commentary).
(2) Necrotic cores
As the calculation in Section 7 shows, the oxygen diffusion limit sets an upper limit on size. Zhao and colleagues say that in millimetre-sized brain organoids, necrosis of the core is often seen because nutrients cannot reach itSourced.
(3) Reproducibility
Zhao and colleagues sum up organoid cultures as "showing considerable heterogeneity and variable cellular complexity, sometimes with insufficiently controlled morphogenesis during self-organisation, and often lacking stromal, vascular and immune components"Sourced. They also note that the choice of starting cells affects the variability and heterogeneity of the resulting structures, and that recombinant natural matrices and synthetic hydrogels are being explored to deal with lot-to-lot variation in MatrigelSourced.
11. A materials engineer's view (3): much of the reproducibility problem is a scaffold problem
Organoid variability has causes on the cell side (differences in starting cells, the stochastic nature of self-organisation) and on the environment side. The biggest on the environment side is the matrix that surrounds the cells.
A review by Aisenbrey and Murphy cites an example in which Matrigel's elastic modulus varied between lots by a factor of about two, 400 to 420 Pa versus 840 Pa, and states that "intra- and inter-lot variability in the mechanical and biochemical properties of Matrigel has contributed to uncertainty and lack of reproducibility in cell culture experiments"Sourced.
Gjorevski and colleagues, meanwhile, varied the stiffness and the adhesive ligands of a synthetic gel independently, and found that intestinal stem cells need stiffness while they expand, but softness and laminin while they differentiate and form organoidsSourced.
In other words, replacing a variable natural product with a synthetic material whose specification can be fixed not only reduces variability but also reveals what is actually doing the work. On this reading, the opening for materials makers lies not in "making the same thing as Matrigel" but in making "a material with only the properties needed, without lot-to-lot differences, that can be changed from stage to stage" (our commentary). For the design variables of hydrogels, see our explainer on hydrogels in this series.
- A spheroid is an aggregate of cells; an organoid is an organ-like structure that self-organises from stem cells. The boundary and definitions are, however, debatedSourced
- 3D culture is closer to the body, but it creates gradients of oxygen and nutrients. In DLD1 spheroids the oxygen diffusion limit was about 232 µmSourced
- By our calculation, oxygen stops reaching the centre at a diameter of about 464 µm, or about 20,000 cellsOur calculation
- The method greatly changes the spread in size. The CV was 10 to 15% for hanging drops and 40 to 60% for aggregation on non-adherent surfacesSourced
- Spheroids of primary human hepatocytes are reported to keep their function for at least five weeks and to detect chronic liver toxicitySourced
- In 2025 the FDA said it would promote organoids and similar models and aims to make animal testing the exception within three to five yearsNot yet confirmed
- Surface chemistry, moulding, mass transfer and scaffold materials — many of the challenges can be written in the language of materials technology (our commentary)
12. Glossary
- Spheroid
- Cells gathered into a round clump. Multicellular tumour spheroids are used as cancer models.
- Organoid
- A 3D tissue that self-organises from stem cells and mimics the structure and function of an organ.
- 2D culture
- Growing cells attached to the flat bottom of a culture dish.
- 3D culture
- Growing cells three-dimensionally, as clumps or within a gel.
- Self-organisation
- A structure forming through interactions between cells, without shape being imposed from outside.
- Ultra-low-attachment (ULA) plate
- A culture plate with a surface cells do not readily attach to. With a round bottom, one clump forms per well.
- Hanging drop
- A method of forming clumps by gathering cells at the bottom of a droplet hanging from a plate.
- Matrigel
- A basement membrane extract from a mouse sarcoma, widely used for embedding organoids.
- Diffusion limit
- The depth oxygen reaches from the surface of a clump. Deeper regions become anoxic.
- Necrosis
- Cell death from a lack of oxygen, nutrients and so on. It occurs at the centre of large clumps.
- Coefficient of variation (CV)
- The standard deviation divided by the mean; a measure of spread.
- Lgr5
- A marker protein found on tissue stem cells in the intestine and elsewhere.
- Organ-on-a-chip
- A system that cultures human cells inside a small device with flow channels to mimic organ function.
- NAMs
- New approach methodologies: in vitro assays, computer models and the like that supplement or replace animal testing.
13. References
- Lancaster MA, Knoblich JA. "Organogenesis in a dish: modeling development and disease using organoid technologies", Science 345:1247125 (2014) — doi.org
- Zhao Z, et al. "Organoids", Nat Rev Methods Primers 2:94 (2022) (PMC) — pmc.ncbi.nlm.nih.gov
- Sato T, et al. "Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche", Nature 459:262 (2009) — doi.org
- Sutherland RM, McCredie JA, Inch WR. "Growth of multicell spheroids in tissue culture as a model of nodular carcinomas", J Natl Cancer Inst 46:113 (1971) (PubMed) — pubmed.ncbi.nlm.nih.gov
- Kelm JM, et al. "Method for generation of homogeneous multicellular tumor spheroids applicable to a wide variety of cell types", Biotechnol Bioeng 83:173 (2003) — doi.org
- Vinci M, et al. "Advances in establishment and analysis of three-dimensional tumor spheroid-based functional assays for target validation and drug evaluation", BMC Biol 10:29 (2012) (PMC) — pmc.ncbi.nlm.nih.gov
- Grimes DR, et al. "A method for estimating the oxygen consumption rate in multicellular tumour spheroids", J R Soc Interface 11:20131124 (2014) (PMC) — pmc.ncbi.nlm.nih.gov
- Hirschhaeuser F, et al. "Multicellular tumor spheroids: an underestimated tool is catching up again", J Biotechnol 148:3 (2010) — doi.org
- Bell CC, et al. "Characterization of primary human hepatocyte spheroids as a model system for drug-induced liver injury, liver function and disease", Sci Rep 6:25187 (2016) (PMC) — pmc.ncbi.nlm.nih.gov
- Spence JR, et al. "Directed differentiation of human pluripotent stem cells into intestinal tissue in vitro", Nature 470:105 (2011) (PMC) — pmc.ncbi.nlm.nih.gov
- Eiraku M, et al. (RIKEN) "Self-organizing optic-cup morphogenesis in three-dimensional culture", Nature 472:51 (2011) — doi.org
- Lancaster MA, et al. "Cerebral organoids model human brain development and microcephaly", Nature 501:373 (2013) (PMC) — pmc.ncbi.nlm.nih.gov
- Takebe T, et al. (Yokohama City University) "Vascularized and functional human liver from an iPSC-derived organ bud transplant", Nature 499:481 (2013) — doi.org
- van de Wetering M, et al. "Prospective derivation of a living organoid biobank of colorectal cancer patients", Cell 161:933 (2015) (PMC) — pmc.ncbi.nlm.nih.gov
- Gjorevski N, et al. "Designer matrices for intestinal stem cell and organoid culture", Nature 539:560 (2016) — doi.org
- Aisenbrey EA, Murphy WL. "Synthetic alternatives to Matrigel", Nat Rev Mater 5:539 (2020) (PMC) — pmc.ncbi.nlm.nih.gov
- Duval K, et al. "Modeling physiological events in 2D vs. 3D cell culture", Physiology 32:266 (2017) (PMC) — pmc.ncbi.nlm.nih.gov
- U.S. Food and Drug Administration (FDA) "Roadmap to Reducing Animal Testing in Preclinical Safety Studies" (PDF) — fda.gov
- U.S. Food and Drug Administration (FDA) "FDA Announces Plan to Phase Out Animal Testing Requirement for Monoclonal Antibodies and Other Drugs", 10 April 2025 — fda.gov
- Boonekamp KE, Dayton TL, Clevers H. "Intestinal organoids as tools for enriching and studying specific and rare cell types", J Mol Cell Biol 12:562 (2020) (PMC) — pmc.ncbi.nlm.nih.gov
14. Claim-to-source audit
| Claim in the text | Basis | Label |
|---|---|---|
| That experiments half a century ago showed fully dissociated vertebrate cells could reaggregate and rebuild organ structure; that 3D cultures made from tissue stem cells or embryonic stem cells are called organoids; and their possible use as models of development and disease and for drug testing | Reference 1 https://doi.org/10.1126/science.1247125 | Sourced |
| The definition of an organoid (derived from stem cells, self-organising, mimicking the key functions, structure and complexity of an organ) and the debate over its definition and scope; more than a century of intertwined research; heterogeneity, variable cellular composition, difficulty controlling morphogenesis, and the lack of stromal, vascular and immune components; that starting cells affect variability; the procedure of enzymatic dissociation and seeding into Matrigel; necrosis in the core of millimetre-sized brain organoids and its partial solution by shaking, rotation and stirring; the need to split and replate periodically; microwell arrays, hanging drops and droplet microfluidics; maturation and vascularisation of kidney organoids in organ-on-a-chip; replacing Matrigel with recombinant collagen, fibrin or synthetic gels; and standards for data quality and reproducibility | Reference 2 https://pmc.ncbi.nlm.nih.gov/articles/PMC10270325/ | Sourced |
| That a single sorted Lgr5-positive stem cell builds an organoid with crypt-villus structures, self-organising without a non-epithelial niche; and that the authors were at the Hubrecht Institute (2009) | Reference 3 https://doi.org/10.1038/nature07935 | Sourced |
| That Sutherland and colleagues reported growth of multicellular spheroids as a model of nodular carcinomas in 1971 | Reference 4 https://pubmed.ncbi.nlm.nih.gov/5101993/ | Sourced |
| That multicellular tumour spheroids are used as models of normal and solid tumour tissue; the drawbacks of conventional methods; one spheroid per hanging drop at close to 100% efficiency; for HepG2, a CV of 10 to 15% versus 40 to 60% on non-adherent surfaces | Reference 5 https://doi.org/10.1002/bit.10655 | Sourced |
| That the evidence that 3D culture is closer to in vivo in gene expression, signalling and drug sensitivity is overwhelming; that many 3D techniques are slow or lack reproducibility; ULA 96-well round-bottom plates needing no coating, formation in 24 to 48 hours, one spheroid centred by the well shape; comparison with agar, poly-HEMA and rotating culture; that spheroids larger than 500 µm often have hypoxia and a necrotic core; and that sensitivity to targeted drugs differs between 2D and 3D | Reference 6 https://pmc.ncbi.nlm.nih.gov/articles/PMC3349530/ | Sourced |
| The diffusion limit of 232 ± 22 µm in DLD1 spheroids and the oxygen consumption rate; the assumed oxygen diffusion coefficient of 2 × 10⁻⁹ m²/s; the estimated DLD1 cell radius of 7.42 ± 0.47 µm; the surrounding oxygen partial pressure of 100 mmHg; and the model with a necrotic core, hypoxic zone and proliferating rim | Reference 7 https://pmc.ncbi.nlm.nih.gov/articles/PMC3899881/ | Sourced |
| That spheroids reflect the 3D relationships of in vivo tumours and pathophysiological gradients; and that they could help weed out unpromising candidates before animal and clinical studies, or find drugs missed in 2D | Reference 8 https://doi.org/10.1016/j.jbiotec.2010.01.012 | Sourced |
| That monolayers of primary human hepatocytes dedifferentiate rapidly; that chemically defined, serum-free spheroids kept function for at least five weeks; that 1,500 cells per well were seeded into ULA 96-well plates; that there was no obvious necrosis after five weeks; and that chronic exposure detected toxicity at clinically relevant concentrations and reproduced the chronic toxicity of fialuridine | Reference 9 https://pmc.ncbi.nlm.nih.gov/articles/PMC4855186/ | Sourced |
| That intestinal organoids were made from human pluripotent stem cells by temporal manipulation of growth factors | Reference 10 https://pmc.ncbi.nlm.nih.gov/articles/PMC3033971/ | Sourced |
| That an optic cup formed autonomously from aggregates of mouse ES cells; and that the authors were at the RIKEN Center for Developmental Biology | Reference 11 https://doi.org/10.1038/nature09941 | Sourced |
| That cerebral organoids were made from human pluripotent stem cells and used to model microcephaly | Reference 12 https://pmc.ncbi.nlm.nih.gov/articles/PMC3817409/ | Sourced |
| That transplanted iPS cell-derived liver buds connected to host vessels within 48 hours and matured; the rescue effect in an animal model of drug-induced lethal liver failure; the authors' statement that further efforts are needed to translate this into treatment for patients; and that the authors were at Yokohama City University | Reference 13 https://doi.org/10.1038/nature12271 | Sourced |
| That tumour organoids established from 20 colorectal cancer patients recapitulated properties of the original tumours and that high-throughput drug screening detected gene-drug associations | Reference 14 https://pmc.ncbi.nlm.nih.gov/articles/PMC6428276/ | Sourced |
| The outlook that organoid technology may allow personalised treatment to be designed | An outlook from van de Wetering and colleagues, not a track record https://pmc.ncbi.nlm.nih.gov/articles/PMC6428276/ | Not yet confirmed |
| That intestinal stem cells were cultured in synthetic gels, needing stiffness to expand and softness plus laminin to differentiate and form organoids, without animal-derived matrices | Reference 15 https://doi.org/10.1038/nature20168 | Sourced |
| Lot-to-lot differences in Matrigel's elastic modulus (400 to 420 Pa versus 840 Pa), and that variability has contributed to a lack of reproducibility | Reference 16 https://pmc.ncbi.nlm.nih.gov/articles/PMC7500703/ | Sourced |
| That 3D culture still faces challenges in tissue interfaces, the mechanical environment, and the spatiotemporal distribution of oxygen, nutrients and waste | Reference 17 https://pmc.ncbi.nlm.nih.gov/articles/PMC5545611/ | Sourced |
| That the FDA Modernization Act 2.0 at the end of 2022 explicitly allowed non-animal alternatives for INDs; that the first target is monoclonal antibodies; and that organoids are described as self-organising cell cultures | Reference 18 https://www.fda.gov/files/newsroom/published/roadmap_to_reducing_animal_testing_in_preclinical_safety_studies.pdf | Sourced |
| The goal of making animal testing the exception over the long term (three to five years) | An FDA plan and target, an outlook rather than a track record https://www.fda.gov/files/newsroom/published/roadmap_to_reducing_animal_testing_in_preclinical_safety_studies.pdf | Not yet confirmed |
| That the announcement of 10 April 2025 said the FDA would promote the use of human organoids and organ-on-a-chip systems mimicking the liver, heart, immune organs and others | Reference 19 https://www.fda.gov/news-events/press-announcements/fda-announces-plan-phase-out-animal-testing-requirement-monoclonal-antibodies-and-other-drugs | Sourced |
| That the culture of Sato and colleagues (2009) used EGF, Noggin and R-spondin in an extracellular matrix | Reference 20 https://pmc.ncbi.nlm.nih.gov/articles/PMC7683021/ | Sourced |
| Critical diameter of about 464 µm (about 420 to 508 µm); volume of one cell about 1,711 µm³; about 20,000 cells at the critical diameter; diameters by cell count (about 197, 248, 294, 371 and 467 µm); diameter about 1.26 times larger for twice the cells; and the ratio of CVs (about a third to a sixth) | Our calculation. Treating the diffusion limit as the radius, the packing fraction of 0.64 and cells as spheres of radius 7.42 µm are assumptions set by this article. Results change greatly with cell type and oxygen environment | Our calculation |
| That materials and devices controlling mass transfer could be the key to raising the size limit of organoids; and the spread of synthetic alternatives | This article's outlook; no track record could be confirmed within its research | Not yet confirmed |
| The figure setting 2D and 3D side by side; the spheroid-organoid contrast; the choice of milestones; the matching of challenges to approaches; the three-way grouping of necrosis countermeasures; the analogy with the catalyst effectiveness factor and Thiele modulus; the reading that surface chemistry and geometry decide quality; the reading that judging uniformity by diameter alone could miss variation in cell number; the opening for materials makers; and the general additions "weeks" and "sizes vary with stirring" | This article's summary and commentary based on published content. Not views expressed by the authors or institutions | Commentary |
| That Figs. 1, 2, 3, 4 and 7 are explanatory drawings; that Fig. 6 is a drawing that includes our calculation; and that the hero image and Fig. 5 are AI-generated images | Our note | Commentary |
Last updated 23 September 2026. Sources are limited to primary material (original papers, peer-reviewed reviews and published FDA material). No market size or share estimates are used. Because the article includes structural summaries and readings about materials design, those are marked as Commentary and kept separate from sourced fact. How far the FDA's three-to-five-year target will be met, how widely synthetic matrices have spread, and any track record of treatment using organoids could not be confirmed in published primary sources, so they are treated as outlook or not stated. Results from animal studies and research-stage work do not show therapeutic effect in humans. All figures are explanatory. Figs. 1, 2, 3, 4 and 7 are concept diagrams, Fig. 6 is a drawing that includes our calculation, and the hero image and Fig. 5 are AI-generated images; none of them is a real micrograph, cross-section or product.