Day 1: Toxicology

09:00 - 09:30 Delegate Arrival: Welcome & Networking

09:30 - 09:45 Opening Speech

09:45 - 10:45 Plenary: Toxicology Prediction - The present, the future, the challenges
                          Professor Alan R Boobis,
                          Department of Medicine, Imperial College London
Abstract
In general, current toxicity testing is relatively effective in protecting human health. However, toxicology is undergoing a profound paradigm shift. There are too many chemicals to test using conventional approaches, animal models may not be predictive of all endpoints, combined exposures need to be evaluated and biological knowledge has advanced markedly, all leading to the need for new approaches to toxicity testing.  This will require a new generation of in silico and in vitro methods, where perturbations of toxicity pathways are identified and effects in vivo are predicted using physiologically-based models. Conventional validation will not be appropriate and hence there will need to be agreement on how the tests will be assessed for reliability.  Toxicity testing should be proportionate to the potential degree of concern. More consideration will need to be given to margin of exposure, taking into account relevant human exposure, mode of action and uncertainty.

10:45 - 11:30 The Next Generation of Toxicology Risk Assessment
                          Dr David J Dix,
                          Deputy Director, US Environmental Protection Agency
Abstract
Coming soon.

11:30 - 12:00 Coffee

12:00 - 12:45 New approaches to the prediction of metabolites from the cytochromes                          P450                       
                         Professor Patrik Rydberg,
                         University of Copenhagen
Abstract
The cytochromes P450s are involved in the metabolism of ~90% of all drugs. They are a family of promiscuous enzymes which can metabolize many different compounds leading to multiple products for each drug compound. Prediction of this metabolism has been shown to be tricky at best, however, recent approaches have shown that we can improve the state-of-the-art significantly, using reactivities which have been pre-computed at a very advanced computational level.
No matter how good a method is, the medicinal chemist wants a prediction to make sense, and if a model is interpretable modifications to chemical structure is easier to perform. This talk will introduce a successive model of determining both which cytochrome P450 isoforms contribute to the metabolism of a drug, and what the metabolites will be, all built from chemically sensible information that can be easily understood by the non-expert.

12:45 - 13:45 Buffet Lunch & Poster Session

13:45 - 14:30 Bridging Computational Modelling and Wet-bench Toxicology: Inter-disciplinary                           Approach to Move the Field Forward
                          Professor Ivan Rusyn,
                          UNC Gillings School of Publich Health, University of North Carolina
Abstract
Quantitative structure-activity relationship (QSAR) models are widely used for in silico prediction of in vivo toxicity endpoints. The ability to virtually screen tens of thousands of chemical structures for their potential activity or toxicity adds value to the process of candidate selection in drug development or in a search for replacing chemicals in commerce with less hazardous substances. Numerical descriptors representing the chemical structure can be easily calculated for any number of molecules and they have been traditionally used as multi-dimensional data matrix for QSAR model development and application.

While experimentally-derived toxicity data has been difficult to obtain on a large number of chemicals in the past, recent efforts by the Tox21 consortium of the US Federal agencies, and the academic laboratories are generating quantitative in vitro toxicity screening data on hundreds of environmental chemicals in thousands of in vitro experimental systems.

In addition, publicly accessible toxicogenomics data on hundreds of chemicals provides another dimension of the molecular information that is potentially useful for modeling. We posit that a combination of chemical structural information, in vitro screening, and/or toxicogenomics data can be used to generate “hybrid” quantitative models to predict human toxicity and carcinogenicity. Using several case-studies, we illustrate the benefits of a “hybrid” modeling approach, namely improvements in the accuracy of models, enhanced interpretation of the most predictive features, and expanded applicability domain for wider chemical space coverage.

14:30 - 15:15 Modelling of the Toxicity of Small Molecules as an Aid in Designing New Drugs
                          Professor Hans Westerhoff,
                          Director, Manchester Centre for Integrative Systems Biology (MCISB)
Abstract
Coming soon.

15:15 - 15:45 Coffee

15:45 - 16:30 Application of Stem Cells in ADME and Toxicity Testing
                          Dr Kyle Kolaja
                          Director and Global Head of Predictive Toxicology Screens and Investigative Safety,                           Hoffmann-La Roche
Abstract
Coming soon.

16:30 - 17:15 Application of Transgenic Models in Metabolism and Toxicology
                          Professor Roland Wolf
                          University of Dundee
Abstract
Coming soon.

17:15 - 17:30 Closing remarks & Thanks
                          David Watson
                          CEO, Lhasa Limited

19:00 - 22:00 Dinner & After Dinner Speech
                          Toxicity - A View From Outside
                          Professor Hugo Kubyini
                          Retired Professor, BASF AG and University of Heidelberg

Abstract

“Toxicity” results from a multitude of different physiological mechanisms, drug toxicity may either come from a direct effect on a certain biological target (e.g. hERG channel inhibition), from the chemical reactivity of the xenobiotic or its various metabolites, from CYP inhibition or CYP induction, or from any other drug-drug or food-drug interactions. In addition, genetic variability may be a reason for unexpected toxic effects. This results in various problems in the prediction of toxicity and the extrapolation of human toxicity from animal studies also poses unexpected surprises. In the past, most drug failures in (late) clinical studies resulted from a lack of efficacy or toxicity – two reasons which are to some extent interrelated: if doses are too low, there may be no effect, if doses are too high, toxicity is observed.

  Day 2: Systems Approaches

The prediction of the extent of metabolically-based drug-drug interactions from in vitro data has become a significant issue in drug development. Retrospective studies often question the reliability of such extrapolation. However, the quality of the output always reflects the quality of the input, and it is essential that the complexities associated with the exercise are acknowledged fully and incorporated, where possible, into the algorithm. For example, factors that are not always allowed for include – the role of the intestine in ‘first-pass’ metabolism; the different concentrations encountered by enzymes on ‘first-pass’ and re-circulation; non-specific microsomal binding; correct numbers for microsomal protein per gram of liver and hepatocellularity; the interplay between transporters and enzymes; mechanism-based inhibition; simultaneous enzyme inhibition and induction; inhibitory metabolites; and important features of the experimental design of the in vivo study that is being simulated. In addition, it is vital to predict outcomes in virtual populations, not just the non-existent ‘average patient’, if individual patients at the extreme of risk are to be identified. By simulating a relevant patient population, incorporating a broad range of demographic, physiological, genetic, enzyme abundance values etc, the exercise can provide early warning of the complex mix of patient characteristics predisposing to risk.

Signaling and metabolic pathways acting within and between cells in an organism may be considered as integrated circuits working to maintain cell growth or homeostasis based on inputs from external influences and other, connected pathways. The components of these circuits comprise DNA, RNA, proteins, enzymatic activities, small molecules, ions etc., and in order to fully understand the circuit, or to “reconstruct the system” it is necessary to consider all of it’s components as well as the connections (interactions) between them. These concepts are the basis of the systems biology approach to biological pathway analysis and network reconstruction. In recent years this approach has been successfully used to integrate data of different types, and from different sources, on multiple types of circuit component – genes, mRNA, proteins, metabolites, microRNA, to obtain a more precise and comprehensive understanding of how disruptions to the circuitry can lead to disease or toxicity. Gross damage to the circuitry by removal or functional impairment of certain components may impact the functioning of the circuit. Individual differences in the system, arising through sequence variations in the DNA template for individual components, also may affect the response of the circuit to normal stimuli or to xenobiotic interference. These variations manifest as different susceptibilities to disease, responses to therapeutic intervention or risks of chemical toxicity. The integration of data on the functional consequences of sequence variation to the systems biology circuit model will be critical to complete understanding of system malfunction or failure in disease or toxicity, and to the successful implementation of personalized medicine.