Image from left to right: Stéphanie Billiald and Sylvain Fouliard.


  • Sylvain Fouliard
  • Stéphanie Billiald
  • Nicolas Sajot

What are your main research interests?
Within Translational Medicine, the non-Clinical Safety and the Quantitative Pharmacology groups aim at supporting the development of new medications and treatment modalities to improve patients’ lives. Our mission in Translational Medicine is to effectively bridge discovery, preclinical and clinical phases thereby improving the overall success rates of our drug candidates for the right patients through an agile selection process, a biomarker driven decision-making, a data integration, as well as a companion diagnostic approach.

PK/PD & QSP modeling for drug pharmacokinetics and efficacy is routinely and successfully integrated into our processes during preclinical and clinical development and informs the translation for first-in-human studies. Quantitative models integrating drug exposure in systemic and tissue compartments, mechanistic data for drug-induced toxicity, as well as organ physiology and sensitive, monitorable biomarkers associated with organ injury are a valuable tool in this setting.

The organ systems considered in TransQST are relevant for our research interests. Representatives from different functions have been contributing to the consortium across several work packages with their scientific knowledge and drug development experience, providing relevant data for the development of models within the consortium, especially in the cardiovascular field.

How has participating in TransQST impacted your company?
Joining TransQST has allowed our organization to participate to the evolution of novel computational approaches which are key enablers for a more efficient drug discovery and development. Although the general framework of QST modelling is still relatively new to our organization, TransQST science and proof-of-concept quantitative systems toxicology (QST) models helped to illustrate the potential utility of QST models and applicability to safety data sets.

TransQST proof-of-concept models for drug-induced kidney injury DIKI and drug-induced liver injury (DILI) allowed for fruitful discussions between quantitative scientists and experimentalists leading to convergence in data generation for integration in mechanistic models at early and late stages of development. This is today a first stone to increase model-informed in vitro and in vivo study designs and inform compound selection. Working through retrospective and current portfolio cases, as well as data for key reference compounds, helps to characterize the utility and limitations of QST models in internal practice.

What challenges did you face while supporting the project? What are your main accomplishements in TransQST?
The major hurdle that was identified throughout the project was the generalisation of the approaches and models developed and discussed for implementation in an industrial pipeline and workflow and the turnaround time for model development/refinement leading to model-based decisions.

Our main achievements in TransQST are contributions to the advances in the set-up of two computational trials, one based on a novel human Purkinje model and the other on a CVS (cardiovascular System Model)-contractility. The development of these models has been the subject of scientific communications.

Early prediction of cardiotoxicity is critical for drug development. Current animal models raise ethical and translational questions and have limited accuracy in clinical risk prediction. Human-based computer models constitute a fast, cheap and potentially effective alternative to experimental assays, also facilitating translation to human.

One of the main objectives of the new CiPA initiative promoted by the pharmaceutical industries, the FDA, HESI and CSRC is to integrate in vitro ion channel data into a computational model of human ventricular myocytes based on O’Hara-Rudy model and to identify a mechanistic metric that can quantify the relative risk of inducing EAD/TdP.

In the line of the CiPA paradigm, Passini et al. from Oxford University have developed an interesting computational tool using human ventricular model populations, called Virtual Assay®, capable to predict the risk of drug-induced adverse cardiac events, based on ion channel information. Servier has implemented Virtual Assay® in the safety pharmacology strategy which is a cheap complement to experimental methods following cardiac ion channel screening (hERG, NaV1.5 and soon CaV1.2) done in routine. Virtual Assay® can strengthen the confidence we have in predicting the cardiotoxicity of our compounds.

The productive integration of in silico trials for drug testing requires their evaluation against currently used experiments. Trovato et al. from Oxford University evaluated the consistency between in silico simulations using a novel human Purkinje model. For this evaluation, Servier provided experimental data obtained in rabbit Purkinje fibres. This preparation which allows to test effects of compounds on all cardiac channel currents and thus, to test mitigation effects, is often used for preclinical drug evaluation.

Concerning in vivo cardiovascular effects, in Safety Pharmacology, the interpretation of safety studies to determine the drug mode-of-action and concentration-effect relationship for specific hemodynamic variables can be challenging due to underlying complex homeostasis and feedback interrelationships. The use of quantitative systems models to characterize the dynamics and PK/PD relationships of hemodynamic variables after administration of drugs was proposed to address this challenge. Thus, a novel CVS (cardiovascular System Model)-contractility was developed. Our contribution, with GSK and Astra Zeneca, was to provide cardiovascular data to develop and validate this model.

Therefore, in silico drug trials are likely to play soon a major role in drug development, identifying drug cardiotoxicity in the pre-clinical phase, thus improving the quality of new candidate drugs and reducing drug failure at later stages.

Which do you consider as the principal project outcomes?
TransQST was able to move forward several innovative modelling approaches for different organ toxicities, which illustrate how quantitative models will ultimately be successfully developed for toxicity and safety pharmacology. Thanks to the consortium, Virtual Assay tool was implemented as part of our safety pharmacology workflow. The need for the convergence between robust data generation, hypothesis set-up and mechanistic QST models was highlighted as a key step for the full implementation of the approach in a drug discovery workflow. The discussions within the consortium also stressed the need for adequate human-relevant input data to perform benchside-to-bed translation and the need for robust PBPK models to predict drug exposure at target site.

How do believe TransQST has contributed to advance the field?
Toxicity assessments in pharmaceutical practice dominantly depend on preclinical animal toxicity studies. TransQST approaches can allow to initiate a vision where integrating data from pre-clinical studies and species-specific tissue physiology, together with PBPK approaches, can facilitate translation to a human setting based on dynamical and quantitative model simulations. The scientific robustness of the model, together with adequate qualification and validation needs to be thoroughly assessed before leading to in silico replacing in vivo. The path is still long, and the science is rapidly growing. TransQST significantly helped the field.

*The provided commentaries are based on the interviewee’s personal opinions towards the consortium which do not reflect the views from Servier as an organization.