
Comprehensive study with Genentech and AstraZeneca validates superiority of Human Liver Microtissues
A landmark Archives of Toxicology article, “Utility of spherical human liver microtissues for prediction of clinical drug-induced liver injury DILI” by our collaborators Will Proctor, Alison Foster, Simon Messner, Dominic Williams and several other notable experts in the field, highlights the advantages of using 3D InSight™ Human Liver Microtissues (3D hLiMT) for DILI prediction.
Study Objective
Systematic validation of 3D InSight ™ Human Liver Microtissues vs. 2D primary human hepatocytes for their predictive value in discriminating between known hepatotoxicants and clinically safe drugs
Study Design
The study tested a panel of 110 clinically known drugs on both primary human hepatocytes and 3D InSight™ Human Liver Microtissues, utilizing the same hepatocyte lot, compound concentrations and endpoint (ATP-content)
Key Findings
- 2-fold more sensitive in identifying known hepatotoxicants in comparison to 2D PHH (sensitivity of e.g. 60.9% vs. 33.3%)
- Specificity for prediction of non-DILI drugs remained very high, even after 14 days of compound exposure
- Assessment of novel hepatic injury biomarkers, such as miRNA122, HMGB1 and α-GST
About the Data
- AstraZeneca and Genentech chose the validation compound set to be representative of a “real-world” situation. In order to conduct a truly independent assessment of the model, all 3D studies were performed at InSphero under conditions where the identities of the 110 compounds were blinded, and all 2D studies were conducted independently by Genentech.
- The compound set included some with limited solubility, as well as many weak DILI compounds with known rare and late DILI onset. For example, out of the 21 DILI compounds where no IC50 value was determined, 14 compounds were only tested at less than 20-fold of the patients’ plasma concentrations due to solubility issues, thus preventing calculations of margin-of-safety.
- Four drugs (Stavudine, Methotrexate, Dantrolene, Nifedipine) were only toxic in vivo after several months of administration, and another drug (Trovafloxacin) is known to cause DILI in an idiosyncratic reaction, i.e., in the presence of inflammation, thus requiring different stimuli. Trovafloxacin toxicity was, however, correctly detected in the presence of an inflammatory stimulus.
About the Model
Organotypic 3D InSight™ Human Liver Microtissues feature:
- Polarized hepatocytes with active hepatobiliary bile-salt secretion (BSEP and MRP2) and presence of bile canaliculi
- Preserved cytochrome P450 activity over the 14-day assay period
- 6-fold increase in mitochondrial spare respiratory capacity compared to 2D PHH
- Incorporation of inflammation-sensing Kupffer cells for detection of indirect effects on liver function
Displaying > 28-day in vitro longevity and more physiologically relevant biology, 3D InSight™ Human Liver Microtissues enable cumulative assessment of the most important factors driving DILI in vivo:
- Long-term drug exposure
- Active drug metabolism
- BSEP inhibition
- Mitochondrial impairment
Read the Paper
Genentech and AstraZeneca put 3D InSight™ Human Liver Microtissues to the test against traditional PHH assays, using a set of 110 clinically-defined DILI compounds. The results speak for themselves.
Learn how you can improve sensitivity for DILI prediction by employing long-term drug exposure
Guaranteed viability for 28 days in most of our liver models allows highly reproducible assessment of both chronic and short-term drug exposure without the use of animals.
Arch Toxicol. | 2017 Jun 13. doi: 10.1007/s00204-017-2002-1
DILI Prediction Applications

Tier 1: Short-Term
(Max. 7 days)
Screen quickly with high specificity for potential DILI drugs

Tier 2: Long-Term
(14 days)
Ensure highest sensitivity for DILI drugs with long-term exposure

Translational Toxicity
Compare susceptibility of different pre-clinical species to hepatotoxicity