In vitro liver disease models go 3D: Organoids and chips reshape translational

What's new across major liver injury paradigms

1) MASLD/MASH: from lipid droplets to multicellular crosstalk

A practical take-home message is that inducing steatosis is “easy,” but modelling lipotoxic stress biology is harder. Many in vitro MASLD models rely on free fatty acid (FFA) mixtures (often palmitate + oleate) to drive triglyceride accumulation and metabolic stress readouts (eg, Oil Red O staining, inflammatory markers, oxidative stress/mitochondrial effects). However, progression-relevant features—immune recruitment/activation, stromal responses, ECM remodelling—require co-culture, organoids, or liver-on-a-chip platforms that enable dynamic cytokine exchange and cell–cell interactions. The review’s schematic explicitly links increasing model complexity to increasing ability to recapitulate inflammation and tissue-level interactions.

2) Fibrogenesis: mechanics and microenvironment matter

For fibrosis, the review emphasises hepatic stellate cell (HSC) activation as the mechanistic core, while warning that common in vitro approaches can artefactually accelerate activation timelines. Primary human HSCs are high-fidelity but difficult to obtain and can activate “artificially” in culture; cell lines are useful for screening but may represent terminal activation states. The authors discuss strategies to better preserve in vivo-like phenotypes (e.g., co-culture with Kupffer cells; multicellular organoids), and elevate ECM mechanics as a major advantage of engineered in vitro systems: tunable stiffness and scaffold-based constructs can interrogate mechanoresponsive pathways and better reflect the 3D nature of fibrosis.

3) Liver cancer: rebuilding the tumour microenvironment

2D liver cancer cell lines remain ubiquitous and useful for proliferation/migration/drug response, but often lack differentiation and microenvironmental context. The review summarises how 3D spheroids/organoids can better maintain heterogeneity, enrich stem-like features, and recreate drug-resistance phenotypes, while noting that immune and stromal reprogramming (especially macrophage state diversity) remains challenging to simulate. More advanced “cancer-on-a-chip” efforts aim to integrate tumour–immune–stroma interactions with spatial organisation.

4) Cholestatic injury: bile acids, barrier function, and duct models

Cholestatic injury is framed around hydrophobic bile acid accumulation as a primary cytotoxic driver, affecting both hepatocytes and cholangiocytes through membrane disruption, mitochondrial dysfunction, ROS, and apoptosis. The review highlights bile-acid exposure models (e.g., lithocholic acid, glycochenodeoxycholic acid) and points to organoid-derived cholangiocyte monolayers and impedance-style barrier measurements as ways to quantify epithelial stress and dysfunction. It also introduces bile-duct-on-a-chip constructs for chronic bile acid exposure studies without mechanical obstruction.

5) DILI: regulatory momentum meets microphysiology

DILI (drug-induced liver injury) is presented as a major clinical and regulatory driver for improved human-relevant testing. The review notes poor concordance between animal and human toxicity findings in some contexts and describes a broader push toward non-animal “New Approach Methodologies.” Acetaminophen is described as a classic intrinsic DILI inducer, with injury assessed via functional biomarkers (e.g., AST/ALT) and cellular stress/death indicators. The key argument: organoids and microphysiological (chip) systems can sustain higher metabolic activity and stability than conventional 2D cultures, improving predictive value for hepatotoxicity screening.

Technology watch: three platforms pushing the field forward

1) Precision-cut liver slices (PCLSs): native architecture, short window

PCLSs preserve in situ cell arrangement, ECM, polarity, and regional heterogeneity, making them powerful for short-term functional and drug-response studies. The limitation is durability: viability typically spans days, with early stress responses and inflammatory marker upregulation that can confound longer experiments, particularly in fibrosis research.

2) Multilineage organoids: self-organisation with growing realism

Organoids offer 3D tissue-like structure and can be built from iPSCs/ESCs or primary sources. The review spotlights two frontiers: (1) adding immune components (e.g., Kupffer-like cells) to capture inflammatory dynamics, and (2) creating zonated organoids to mimic periportal–pericentral functional patterning. It also addresses a persistent bottleneck—insufficient maturity of iPSC-derived hepatocyte-like cells—and summarises maturation strategies (media cues, oxygen modulation, developmental signalling mimics).

3) Liver-on-a-chip: engineering flow, gradients, and decision-ready readouts

Chips are framed as “design meets function”: microfluidic perfusion and spatial co-culture aim to reproduce lobule-like organisation, oxygen/nutrient gradients, and physiologic shear. The review describes how perfusion influences polarity, transporter activity, and endothelial integrity, and provides examples and a table summarising chip materials, cell sources, and applications (page 9). The authors also note expanding designs—region-specific bile-duct chips and multi-organ interactions—reflecting how questions are increasingly microenvironment-specific.

Practical takeaways for researchers

• Start with the question, then pick the platform. The review explicitly argues there is no single “best” model—only the most appropriate one for the mechanism and readout you need.
• Complexity should be justified, not automatic. 2D systems remain useful for accessible, reproducible screening; organoids/chips add fidelity but cost time and resources.
• Aim for outputs that support decisions. The authors’ closing point is blunt: a useful in vitro model should guide mechanistic inference and therapeutic prioritisation, not simply resemble liver tissue.

See the article: 

Rao Q, Wang L, Tacke F, et al. Engineering liver disease models in vitro: emerging trends and innovations. eGastroenterology 2025;3:e100279. doi:10.1136/egastro-2025-100279

About eGastroenterology

eGastroenterology, a BMJ journal partnered with Gut and launched by leading scientists in gastroenterology and hepatology, has been indexed in the Web of Science Core Collection (ESCI), PubMed, DOAJ, Scopus, CAS, ROAD, and many other major international databases within just two years of its launch. The journal is expecting to receive its first Impact Factor in June 2026.

For more information, please visit: egastroenterology.bmj.com and follow us on Twitter (@eGastro_BMJ).

Sign-up to Email Alerts for eGastroenterology: https://emails.bmj.com/k/Bmj/jausu/egastroenterology