DishTrial4LiverTox™

Our service for DILI prediction

A seamless four-step workflow for predictive DILI assessment

STEP 1

STEP 2

STEP 3

STEP 4

Design
Your
Human Cohort

Define the
population
you want to evaluate

Select the characteristics of your study cohort based on your development needs. Choose the number of donors and customize the population by sex, age, and physiological status (e.g., healthy or disease-specific, if applicable)

Evaluate
Your
Compound

Submit your investigational compound for confidential testing

Using our proprietary PredictCan-MPS™ platform, we perform a population-representative cohort « trial-in-a-dish », exposing personalized phenotypic liver models to your compound under standardized experimental conditions.

Throughout the study, your intellectual property remains fully protected under strict confidentiality agreements.

Generation Of Individual-Centric Data

Receive a detailed comprehensive experimental dataset

Depending on your study design, the dataset may include:

Cell viability
Liver function endpoints measurements ( bilirubin, ALT, AST, albumin…etc.)
Individual donor responses

The complete dataset is delivered to your team for independent analysis or downstream integration.

Data Analysis Using
PredictCan-MIND™

Turn Data
into
Confident Decisions

Personalized data generated by PredictCan-MPS are subsequently integrated into our proprietary PredictCan-MIND algorithm to provide a comprehensive assessment of the compound’s overall DILI risk, combining both severity and incidence metrics. By benchmarking the compound’s profile against a panel of well-characterized DILI reference drugs, PredictCan-MIND contextualizes the observed risk within an established clinical and pharmacological landscape. This comparative approach transforms complex experimental findings into an interpretable risk profile, enabling a more informed assessment of how a newly developed drug may compare with known hepatotoxic compounds and supporting earlier, evidence-based decisions on its safety and development potential.

From Scientific Validation to Your Drug Development

DishTrial4LiverTox™ builds on years of research and experimental validation of PredictCan’s individual-centric approach. Our scientific studies demonstrate the ability of our technology to reproduce interindividual variability, distinguish different levels of DILI risk, and identify individuals with increased susceptibility to DILI.

This scientific foundation gives our team the expertise to support your compound from toxicity assessment to deeper mechanistic investigation, turning complex individual responses into meaningful insights for drug development.

Capturing Interindividual Variability in CYP Induction and Activity

We capture interindividual variability in hepatic drug response by generating donor-specific liver spheroids conditioned with human serum from individual donors.

By incorporating donor-specific serum into our Liver hSES™ model, we drive a more personalized and functional hepatocyte-like phenotype that reflects the biological diversity observed across individuals. This approach enables the characterization of donor-dependent differences in phase 1 and phase 2 drug-metabolizing enzyme expression and CYP activity, providing a more physiologically relevant view of hepatic drug response.

Rather than relying on a single standardized liver model, our platform reveals how the same compound can elicit distinct metabolic responses across individuals. This comparative approach helps identify variability in CYP induction and activity that may influence drug exposure, metabolism, and ultimately susceptibility to drug-induced liver injury (DILI).

The result is a patient-relevant assessment of metabolic variability, enabling the identification of potentially susceptible populations, the characterization of interindividual differences in drug response, and a more informed evaluation of DILI risk.

Roux et al., 2024

Controlled growth and long-term stability of liver spheroids

Controlled Growth and Long-Term Stability Enable Reliable Liver Safety Studies

Long-term safety assessment requires tight control of model growth and structural stability.

Our Liver hSES™ model rapidly forms compact and organized multicellular spheroids within three days and subsequently maintains a stable architecture and controlled proliferative state for nearly two weeks, without uncontrolled growth or expansion.

This sustained stability provides a robust foundation for long-term and repeated-dose toxicity studies, allowing observed changes in cellular and molecular endpoints to be attributed more confidently to compound exposure rather than spontaneous changes in the model itself.

By combining rapid spheroid formation with prolonged structural and proliferative stability, our platform supports reproducible, longitudinal assessment of hepatic responses under repeated exposure conditions.

Cherradi et al., 2026

Reproducibility of the PredictCan liver model

High Reproducibility Across Independent Experiments and Users

A predictive DILI platform must deliver consistent and reproducible results across independent experiments and users.

Our Liver hSES™ model demonstrates highly consistent results across independent experiments performed by different experimenters on different days, using the same compounds, serum, and cell batches. This consistency highlights the robustness of the platform and the reliability of its biological readouts.

By demonstrating reproducibility across independent experimental runs, our platform provides confidence that its measurements can be reliably reproduced and compared over time.

The result is a robust, reproducible, and decision-ready DILI platform, providing the reliability needed to support confident compound characterization and data-driven decisions throughout drug development.

Cherradi et al., 2026

Benchmarking Against Reference Compounds Enables DILI Risk Classification

We place your molecule within a clinically relevant DILI risk framework by benchmarking its profile against reference compounds.

By positioning your drug on a reference DILI map, we contextualize its observed biological profile relative to compounds spanning different levels of clinical DILI risk. This comparative approach helps determine whether your molecule behaves more like compounds associated with a lower, intermediate, or higher DILI concern.

Rather than interpreting experimental findings in isolation, reference-compound benchmarking connects your molecule’s in vitro safety profile to patterns observed across clinically characterized drugs.

The result is a contextualized and actionable DILI risk classification, helping identify potential liabilities early, prioritize follow-up investigations, and support data-driven development decisions.

Cherradi et al., 2026

Personalized Liver Spheroids Enable Patient-Specific DILI Risk Prediction

DILI risk scoring in personalized liver spheroids provides a patient-specific approach to identifying clinically relevant liver toxicity.

Using a small blood serum sample collected from patients before initiation of clinical treatment, we generate personalized liver spheroids that recapitulate key patient-specific characteristics. These models can then be exposed to the drug of interest in vitro, enabling assessment of the patient’s individual susceptibility to drug-induced liver injury.

Importantly, the resulting in vitro DILI risk profiles show a striking concordance with liver toxicity observed clinically. Patients whose personalized spheroids display a high-risk response correspond to patients experiencing clinical liver toxicity, while lower-risk profiles are consistent with better hepatic tolerance.

This patient-matched approach creates a direct bridge between in vitro liver response and real-world clinical outcome, supporting the identification of susceptible patients and more personalized prediction of DILI risk.

ALT and bilirubin hepatotoxicity biomarkers

Comprehensive Measurement of Hepatotoxicity Biomarkers

A multidimensional assessment of liver toxicity through complementary cellular injury and functional biomarkers.

The platform enables measurement of key hepatotoxicity biomarkers, including ALT release and total bilirubin. ALT provides a sensitive readout of hepatocellular injury, while total bilirubin provides complementary information on hepatic functional impairment.

These parameters can be quantified in individual Tox-positive and Tox-negative subjects, providing an additional layer of biological characterization alongside the overall DILI risk score.

Rather than relying on a single endpoint, the platform captures distinct and complementary manifestations of hepatotoxicity, enabling richer characterization of drug-induced liver responses and subtle differences in toxicity profiles between individuals.