Archive - 2010 Agenda
Day 1: Current Approaches & Initiatives in Toxicity Prediction
10:00 - 11:00 Plenary: Current Issues in Toxicity Prediction
Dr Thomas Hartung
Johns Hopkins University, Center for Alternatives to Animal Testing (CAAT), Doerenkamp-Zbinden Chair for Evidence-based Toxicology, Baltimore, Maryland USA, and Professor of pharmacology and toxicology, CAAT-EU, University of Konstanz, Germany
Abstract
Europe and the U.S. further the development of new toxicological tools in very different ways. While the replacement of animal tests has been promoted strongly in Europe over the last decades (following the 3Rs principles - reduce, replace, refine), in the U.S. the vision for toxicology in the 21st century (Tox-21c), which was prompted by the National Research Council document only three years ago, dominates the discussion. In both cases, there is significant political support. However, in Europe the 1986 horizontal animal welfare legislation (which urges the use of 3Rs methods wherever possible) is currently under revision and cosmetics and chemical legislation are the primary drivers. In the U.S. it is mainly federal agencies, most prominently the Environmental Protection Agency (EPA), that made the implementation of the NRC report their toxicity testing strategy only in 2009. This pre-empts such likely legislative measures as the reauthorization of the Toxic Substances Control Act (TSCA) in the U.S. The European implementation is characterized by substantial broad funding programs to develop 3Rs methods and can be termed a "bottom-up" approach; the Tox-21 program, in contrast, represents a "top-down" approach, where programmed research is carried out and commissioned. The two approaches are two sides of the same coin, they can instruct and complement each other. More importantly, however, if brought together they can result in a Human Toxicology Project and a real revolution in regulatory toxicology.
11:00 - 11:40 Good practices for Building Robust QSAR Models for Chemical Toxicity
Professor Alexander Tropsha
Professor and Chair of Medicinal Chemistry and Natural Products,
UNC School of Pharmacy, NC, USA
Abstract
A wealth of available data on chemical effects on biological systems requires new computational approaches to link chemical structure, in vitro data, and potential adverse health effects. We have advanced a predictive Quantitative Structure-Activity Relationships (QSAR) modeling workflow that relies on effective statistical model validation routines and implements both chemical and biological (i.e., in vitro assay results) descriptors of molecules to develop in vivo chemical toxicity models. We have developed two distinct methodologies for in vivo toxicity prediction utilizing both chemical and biological descriptors. In the first approach, we employ biological descriptors directly in combination with chemical descriptors to build models. Our second modeling approach employs the explicit relationship between in vitro and in vivo data as part of the two-step hierarchical modeling strategy where a special binary QSAR model using chemical descriptors only is built to partition compounds into classes defined by patterns of in vitro - in vivo relationship. In the second step, class specific conventional QSAR models are built, also using chemical descriptors only. Thus, this hierarchical strategy affords external predictions using chemical descriptors only. We will present the results of applying both strategies to several datasets including ToxCast Phase I data; the ZEBET dataset including in vitro IC50 cytotoxicity values and in vivo rodent LD50 values for more than 300 chemicals; and a dataset with known in vitro cytotoxicity and in vivo carcinogenicity and mutagenicity data available in the Berkley Carcinogenic Potency Database. All studies suggest that utilizing in vitro assay results as biological descriptors afford prediction accuracy that is superior to both the conventional QSAR modeling that utilizes chemical descriptors only or in vivo effect classifiers based on in vitro biological response only. We will discuss how our models can be used to prioritize compound selection for experimental chemical toxicology studies.
11:40 - 12:10 Coffee
12:10 - 12:40 ChemSpider - A Platform to Gather, Host and Integrate Structure Based Data Across
the Web
Dr Antony Williams
Vice President of Strategic Development,
ChemSpider at Royal Society of Chemistry, UK
Abstract
ChemSpider was developed with the intention of aggregating and indexing available sources of chemical structures and their associated information into a single searchable repository and making it available to everybody, at no charge. There are many tens of chemical structure databases such as literature data, chemical vendor catalogs, molecular properties, environmental data, toxicity data, analytical data etc. and no single way to search across them. Despite the diversity of databases available online their inherent quality, accuracy and completeness is lacking in many regards. ChemSpider was established to provide a platform whereby the chemistry community could contribute to cleaning up the data, improving the quality of data online and expanding the information available to include data such as reaction syntheses, analytical data and experimental properties. ChemSpider has now grown into a database of well over 20 million chemical substances integrated with over 300 disparate data sources, many of these directly supporting the Life Sciences. This presentation will provide an overview of our efforts to improve the quality of data online, to provide a foundation for the semantic web for chemistry and to provide access to a set online tools and services to support access to these data. I will also discuss how ChemSpider is being used to enhance Semantic Publishing in Chemistry at RSC.
12:40 - 13:20 Implementation of the Category Approach in the OECD (Q)SAR Application Toolbox
Dr Sabcho Dimitrov,
Deputy Head Laboratory of Mathematical Chemistry,
Laboratory of Mathematical Chemistry,
Bourgas "Prof. As. Zlatarov" University
Abstract
To reduce resources and animal welfare, it is important to limit the number of tests to be conducted, where this is scientifically justifiable. One approach is to consider similar chemicals in a chemical category, rather than as individual chemicals. In the category approach, the data for chemicals and endpoints that have been tested are used to estimate the corresponding properties for the untested chemicals and endpoints. The categories could be defined by the structural (empirical or statistical), parametric, mechanistic and metabolic boundaries specifying its domain as well as the category members. Once mechanistic/metabolism based grouping tools are used the category is becoming endpoint specific. The categories could be characterized by their structural and mechanistic consistency which defines the category robustness. The structural and mechanistic grouping of chemicals should be considered as distinct phases of categorization process rather than two categorization alternatives. The missing data in a chemical category could be filled in by one of the following approaches: read-across, trend analysis or (Q)SAR models. Read-across involves identification of the closest analogues according to appropriate measure for similarity and the assumption that the analogues and the target chemical behave similarly. Trend analysis involves ordering of analogues according to molecular parameters associated with bioavailability, such as molecular weight, chain-length, water solubility, partitioning coefficients, etc., and interpolation or extrapolation of missing values via linear or quadratic regression. (Q)SAR models belongs to the estimation methods used to predict the behaviour of a chemical in a biological or environmental system based on qualitative or quantitative relationship between an endpoint (activity) and one or more molecular descriptors. In this respect, the trend analysis could be regarded as a simplified version of QSAR model based on data availability for analogues of the target chemical. The categorization and data gap filling approaches available in OECD (Q)SAR Application Toolbox will be illustrated by well selected case studies within the regulatory context and REACH legislation.
13:20 - 14:20 Buffet Lunch & Poster Session
14:20 - 15:00 REACH - Implementation, Application & Impact
Tom Holmes, Agchem Project Consulting (APC) Ltd
Abstract
REACH - Implementation, Application & Impact
REACH (Regulation 2006/1907) was introduced to address short comings of an old chemicals regulation system. The previous EU frame work for chemicals was a patchwork of many different directives accreted over time and was deemed to be no longer fit for purpose. The majority of chemicals existing before 1981, ca. 100,000, had been "Grandfathered-in" and many had little or no safety data. In comparison, chemicals registered to more modern standards since 1981, numbered only ca. 4000. It was public authorities, rather than producers, that produced the risk assessments and there was no consideration of downstream uses.
The implementation of REACH has been relatively rapid and utilises IT to the maximum extent possible. In the 4 short years since REACH entered into force, ECHA has been set up and staffed, extensive guidance has been published, and the IT set-up and dissemination has confounded pessimistic expectations. Submissions are already being made and registrations have already been issued. The list of candidate SVHCs ('Substances of Very High Concern' i.e. those to be listed in Annex XIV) is being populated.
The implementation of REACH cannot be seen in isolation and is part of a family of regulations and directives fully cascading and extending the principles of the animal welfare directive. For REACH, Classification and Labelling, and other recent regulations, in vivo testing is to be the "last resort". This has been seized on as an opportunity to reduce the laboratory burden as far as possible through the use of, "intelligent" testing strategies, readacross (analogues, groups), SARs, QSARs, in vitro methodologies, exposure based waiving and use of increased uncertainty factors. The fact that the anticipated problems in terms of lab capacity for remedial testing have not materialised, is testament in large part, to the "last resort" approach.
Although no waiver or estimation of toxicity based on QSAR or readacross can be assigned a Klimisch score of more than 2 under REACH, REACH goes further than any other previous legislation in allowing in silico methods to substitute for laboratory data. In other sectors, (Q)SARs have in the main, only been used as an initial screening tool and allowed to implicate guilt (and prompt further specific investigations), but not to demonstrate innocence. Critical to bolstering the reliability of the in silico model predictions is an understanding and incorporation of mechanistic considerations. Where (Q)SARs outputs are used, extensive justification and validation is required with their submission, which is daunting for the uninitiated.
Over the next few years ECHA will accumulate a wealth of data and experience which will hopefully be used aid the accelerated development, and greater acceptance of in silico techniques in hazard prediction, not just in chemicals, but also in other sectors.
15:00 - 15:40 Data Integration and Predictive Systems Biology
Dr William B. Mattes
PharmPoint Consulting,
Former Executive Director, Predictive Safety Testing Consortium
Abstract
In a fashion analogous to that of structure-activity relationships, where the biological effects of a molecule are predicted from its component substructures, predictive systems toxicology strives to anticipate the spectrum of responses to a biological perturbation based upon the effect of the perturbation with one component of a system and the connections of that component with others in the system. To quote Denis Noble, "Systems biology...is about putting together rather than taking apart, integration rather than reduction." Systems toxicology extends this to predictions of adverse biological outcomes. The most ambitious model envisions integration of signaling pathways, protein-protein interactions, metabolic pathways, transcriptional control networks, inter-tissue hormonal signaling, immune system interactions, genotypic variation, and epigenetic control with multi-omic pathology and clinical data. The goal is to create a system that can take data from a particular stimulus in an in vitro or in vivo model and predict the responses in different species, individuals, and environments. In addition, the model should suggest specific endpoints (ie, biomarkers) to be measured to confirm the prediction. Currently we see steps where some components of the more ambitious global model are integrated. This talk will examine those steps, review recent data, and suggest paths for future development.
15:40 - 16:10 Coffee
16:10 - 16:50 Metabolism Mediated Toxicity Prediction
Dr Tomas A. Baillie
University of Washington
Abstract
It is now widely appreciated that the two major causes of attrition in the drug development process are: (i) preclinical toxicity, and (ii) lack of adequate efficacy in human trials. Failures due to poor efficacy reflect the reality that in vitro systems and in vivo animal models do not always predict the pharmacological behavior of new chemical entities (NCEs) in man, although the incorporation into drug development programs of newer tools, such as efficacy biomarkers and imaging techniques, should help in predicting clinical outcomes prior to the initiation of large scale clinical trials. Predicting toxicity in animals (and subsequently in humans) that results from unanticipated off-target effects of an NCE is more problematic, given our limited understanding of the molecular events that lead to cellular insult. However, consideration of the various mechanisms of drug-induced toxicity indicate that several types of serious adverse reactions appear to involve the formation of a chemically-reactive metabolite(s) of the parent compound; as a result, strategies to minimize the generation of such electrophilic intermediates during the lead optimization stage of drug discovery have been put in place by many pharmaceutical companies. Combining early drug metabolism studies with appropriate in silico evaluations, and taking into account the potential role of "toxicophores" in drug toxicity, represents a viable approach to optimizing NCEs for successful development. Examples will be presented of this integrated approach to lead optimization, where the primary goal is to minimize the likelihood of toxicity due to chemically reactive metabolites.
S. Kumar, K. Mitra, K. Kassahun, and T. A. Baillie, "Approaches for Minimizing Metabolic Activation of New Drug Candidates in Drug Discovery", in Adverse Drug Reactions, Handbook of Experimental Pharmacology 196, J. Uetrecht (ed.), Springer-Verlag, Berlin, 2010, pp. 511-544.
16:50 - 17:00 Closing Presentation & Thanks
David Watson
CEO,
Lhasa Limited
Day 2: New Directions & Challenges in Toxicity Prediction
09:00 - 10:00 Plenary: Emerging Areas in Toxicity Prediction
Dr Jim Bridges
Head of the European Institute of Health and Medical Sciences,
University of Surrey, UK
Abstract
Assessment of the hazardous properties has been a key element in the regulation of medicines, pesticides, industrial chemicals, food additives and contaminants, personal care products and other consumer products for several decades. During this period the thrust of new technical developments has been to standardise existing animal tests and to address apparent deficiencies in coverage by adding new tests particularly in vitro tests.
Despite the generally acclaimed success of current methods of hazard assessment in public health there are strong drivers for fundamental changes to be made within the next decade. These include:
-major advances in technologies. These will enable the many changes caused by individual chemicals at the molecular level to be readily identified and classified.
-continuing public/political pressures to reduce/ abolish the use of laboratory animals for toxicity testing.
-the large number of chemicals, whose hazardous properties have either not been assessed, or assessed inadequately. With current procedures there are high resource demands and progress to remedy this situation is inevitably slow. In part this aspect is likely to be addressed through a shift towards exposure driven risk assessments using approaches such as thresholds of toxicological concern.
In deciding on the changes that need to be made we must not lose sight of the key objective namely to ensure that hazard assessment is much more predictive of the effects that may occur in the human population. It also needs to be appreciated that to improve human health risk assessment major changes are needed to in exposure assessment.
In my view the core of a new strategy must be the integration of hazard assessment with modes of action. This will also provide a sounder scientific rationale for distinguishing significant and insignificant changes for human health.
Among the requirements for the successful implementation of the desired changes will be:
-improved interfaces between toxicology, molecular biology and systems biology
-greater flexibility in assessment protocols
-establishment/modification of hazard data bases to ensure that they are properly validated, kept up to date and are accessible to legitimate investigators.
-close collaboration across the use sectors and legislative regimens
-a phased approach to the introduction of new techniques and the phasing out of traditional tests in order to validate the benefits and limitations of proposed changes. Vested interests should not be in a position to dictate this process.
10:00 - 10:40 Modulation of Drug Response by Transporter-Mediated Membrane Transport
Professor Hiroyuki Kusuhara
Professor & Chairman, Dept. of Molecular Pharmacokinetics,
Graduate School of Pharmaceutical Sciences, Japan
Abstract
Active transporters are capable of concentrating drugs inside the cells, or lowering intracellular drug concentrations and, thereby, modulating their effects. In the body, transporters characterized by broad substrate specificities play significant roles in the elimination of drugs from the systemic circulation in the liver and kidney, and they also act as a tissue defense system, such as the blood-brain barrier. Therefore, variations in the activities of transporters in the clearance organs alters the systemic exposure of drugs to the peripheral organs, whereas variations in the transporter activity in the barriers of the CNS affects the brain concentrations even although systemic exposure is not affected. Drug-drug interactions and genetic variations cause such variations in transporter activities and modulate the pharmacological effect. This presentation focuses mainly on animal studies in relation to clinical reports.
An organic cation transporter, Oct1, mediates the hepatic uptake of small hydrophilic cationic drugs. In Oct1(-/-) mice, the hepatic uptake of metformin is reduced, resulting in attenuation of the effect of metformin on plasma lactate. Recently, we demonstrated that Mate1, a proton/organic cation exchanger, is involved in the canalicular efflux of metfomin in mice. Administration of pyrimethamin, a potent Mate1 inhibitor, inhibits the canalicular efflux of metformin, resulting in a significant increase in the hepatic concentration and plasma level of lactate in mice.
Cholesterol-lowering drugs, so-called statins, are mainly eliminated from the liver where an organic anion transporter, OATP1B1, plays a predominant role in the hepatic uptake of anionic statins. Comparison of the uptake clearance with the intrinsic hepatic clearance showed that the hepatic uptake is rate-determining in the overall hepatic elimination. Under such circumstances, variations in the uptake markedly affect the systemic exposure, but have a much weaker effect on the liver concentration as far as the renal elimination is negligible, and vice versa for variations in metabolism and canalicular efflux. For example, variations in OATP1B1 activities may have a small and a large impact on the therapeutic efficacy and side effect (myopathy), whereas those involving canalicular efflux and metabolism may have opposite impacts (i.e., large and small impacts on the therapeutic efficacy and side effect).
We also examined the importance of Bcrp and Mrp4 with regard to the blood-brain barrier. Knockout of Bcrp and Mrp4 enhanced the brain accumulation of their substrate drugs. It was found that Bcrp shows overlapping substrate specificity with P-gp and thus, simultaneous dysfunction of P-gp and Bcrp markedly increased the concentrations of their common substrates in the brain.
10:40 - 11:10 Coffee
11:10 - 11:50 Toxicogenomics in Predictive Toxicology
Dr Eric Blomme
Project Leader, Cellular & Mollecular Toxicology,
Abbott Laboratories, USA
Abstract
Toxicity represents an important cause of failure in the late stages of discovery and preclinical development. Therefore, early identification of the toxic liabilities of experimental compounds represents one of the most promising alternatives to decrease overall R&D costs. In this presentation, we will review how toxicogenomics can be used to predict and characterize the toxicologic profile of compounds at an early stage with a particular emphasis on the strengths and limitations of the technology. Using specific examples, we will illustrate how toxicogenomics can be successfully implemented in a discovery or preclinical organization.
11:50 - 12:30 Metabolomics in Toxicity Assessment
Professor Elaine Holmes
Chemical Biology,
Imperial College London, UK
Abstract
Metabolic phenotypes defined using high-resolution spectroscopic tools in combination with mathematical modeling can provide a window on gene-environment interactions and a roadmap for identifying physiological and pathological mechanisms. The application of metabolic profiling in toxicology has allowed the characterization of lesions based on site and/or mechanism of toxicity and in some cases can even provide a tool for prognostic screening of toxic response. Toxicological stimuli influence metabolite profiles in a characteristic and consistent manner, which partially results from large scale changes in gene expression and protein activity, as the organism strives to maintain homeostatic equilibrium by adjusting the composition of the fluids in cells and their surrounding tissues and organs. The perturbations in the composition of the biological samples are reflected in the spectral profiles and can be rapidly be identified by using appropriate mathematical and pattern recognition procedures. The metabolic response can subsequently be integrated with proteomic or transcriptomic data using the same statistical tools in order to achieve a more holistic vision of the toxic process. Moreover, since the metabolic profiling operates on non- or minimally-invasive samples such as blood or urine, it is possible to obtain temporal data and to map the dynamic profile of a toxic lesion. Here the strategy for applying metabolic profiling to toxicological studies is discussed using case studies to illustrate the approach. Species transferability, idiosyncratic toxicity, environmental toxicity and biomarker detection and validation will also be addressed.
12:30 - 13:10 Computational Systems Biology & Dose Response
Dr Melvin Anderson
Director, Division of Translational and Computational Biology,
The Hamner Institutes for Health Sciences, NC, USA
Abstract
The 2007 NAS report, "Toxicity Testing in the 21st Century - A Vision and A Strategy", outlined initiatives for using in vitro methods to evaluate toxicity pathway responses, for applying computational systems models for assessing dose response of pathway perturbations, and for grounding test-tube to human extrapolations with in vitro-in vivo dosimetry. The key recommendation of this report is that in vitro test systems using human cells or human tissue surrogates are not alternatives to current whole animal approaches but should be regarded as the preferred tools for toxicity testing and risk assessment. This technology transformation will require equally as much attention to new computational methods for assessing dose response behaviors of in vitro response pathways as in the development of assays to test for activity of compounds in perturbing pathway function. This author envisions a future in which each in vitro test in the risk assessment arsenal is validated by mapping the circuitry of the response pathway and by computational systems biology modeling of the expected dose response patterns for thresholds and for both adaptive and excessive perturbations. This talk focuses on the ancillary computational tools and biological models that will be necessary to move in vitro assays from the realm of secondary assays for risk assessment to center stage in human health safety assessments and on the manner in which these tools will be integrated into new testing methods.
13:10 - 14:20 Lunch (+ Time to View Royal Armouries Museum)
14:20 - 15:00 TK/TD Relationships: Exploring Dose, Concentration & Time Relationships in Drug Toxicity
Dr Dennis Smith
Vice President - Pharmacokinetics, Dynamics & Metabolism,
Pfizer Global Research & Development, UK
Abstract
Plasma concentrations are now routinely part of safety assessment. There are two prime reasons for determining plasma concentrations in pre-clinical studies
1. To allow interpretation of dose response within a study. Due to the processes of dissolution, absorption and metabolism plasma concentrations may not be linear with dose resulting in possible shallow or pronounced dose response curves. For instance low solubility can result in sub-proportional increases in plasma concentrations, whilst saturation of metabolism can result in super-proportional increases in plasma concentrations
2. To allow comparison across species to understand dose response relationships and to allow some prediction of possible therapeutic windows in human.
Sampling of pre-clinical studies is usually sparse and not easily amenable to true PK/ PD exploration. Moreover the dynamic measurements show considerable variation in effects between individuals.
Data is often condensed into a measure of the peak exposure (Cmax) as an indicator the acute stress on the system a drug may impose and chronic stress (normally AUC). Both measures have interdependency but have different sensitivities depending on the pharmacokinetic properties of the compound. Improvements in TK/TD analysis are provided by converting AUC into Cav (average concentration over specified time) values.
Cav units are identical to Cmax and allow immediate comparison with this value and thus numerically describe a plasma concentration curve. Moreover these values allow easy direct comparison with the potency and selectivity of the drug obtained during in vitro testing. Candidate drugs are now screened against a large panel of proteins (receptors, ion channels, enzymes etc.) alongside their intended target. This data can be exploited to its full by the use of Cmax and Cav comparisons in TK/TD analysis.
Circulating metabolites may contribute to TK/TD relationships and a much clearer picture is now available of the form of metabolites likely to contribute to TK/TD outcomes in terms of physicochemistry, structure and concentration. Information about metabolites is again best conveyed by the use of Cmax and Cav values
15:00 - 15:40 Human Adult Stem Cells and Cell-Cell Communication As Targets for
Epigenetic Toxicity: A New Strategy for Toxicity Testing
James E. Trosko, Ph.D.
Director of the Division of Genetic & Epigenetic Toxicology, Human Adult Stem Cell
Biology and Oncology
Center for Integrative Toxicology, Michigan State University, USA
Abstract
The NRC Report, "Toxicity Testing in the 21st Century: A Vision and a Strategy", has called for a renewed examination for a better assessment of potential risks to human health. This presentation provides a major paradigm challenge to current concepts of how chemicals induce toxicities and how mechanisms of toxicities contribute to the pathogenesis of human diseases. In concordance with the NRC Report to take "...advantage of the on-going revolution in biology and biotechnology", this presentation supports the use of human adult stem cells, grown in vitro under simulated in vivo niche conditions and the rejection of the idea that toxic chemicals work via mutagenic mechanisms [ 1,2]. Currently, in vitro toxicity assays (mutagenesis, cytotoxicity, epigenetic modulation), performed on 2-dimensional primary, immortalized or tumorigenic rodent or human cell lines, do not represent normal human cells in vivo. Clearly, basic science studies must be done to elucidate the mechanisms of toxicity of chemicals with assays that truly represent genomic mutagenic, cytotoxic and altered gene expression ( epigenetic toxicity) events in the human in vivo target cells.. Furthermore, the precise mechanisms of pathogenesis of specific human diseases (birth defects, cancer, neuro-toxicities, etc.) must be determined. Only then, can the mechanisms of toxicity of a chemical be integrated into the pathogenesis of a given disease state. The ultimate potential for in vitro testing of human stem cells will be to try to mimic a 3-D in vitro micro-environment on multiple organ-specific and multiple genotypic/gender adult stem cells. The role of stem cells in many chronic diseases, such as cancer, birth defects, and possibly adult diseases after pre-natal and early post-natal exposures (Barker hypothesis), as well as for regenerative therapy, demands toxicity studies of stem cells. While qualitative alteration of gene expression ("toxico-epigenomics") is a legitimate endpoint of these toxicity studies, quantitative alteration of stem cells during development must be serious considered. If the future utility of human stem cells proves to be valid, the elimination of less relevant, expensive and time-consuming rodent and 2-D human in vitro assays will be eliminated.
1. Trosko, J.E. Chang, C.C. Factors to consider in the use of stem cells for
pharmaceutic drug development and for chemical safety assessment. Toxicology
270: 18-34, 2010..
2. Trosko, J.E., and Upham, B.L., Commentary on "Toxicology testing in the 21st
Century: A vision and a strategy": Stem Cells and cell-cell communication as
fundamental targets in assessing the potential toxicity of chemicals. Human & Exper.
Toxicology 29: 21-29, 2010
15:40 - 16:00 Closing Presentation and Thanks
David Watson
CEO,
Lhasa Limited
Royal Society of Chemistry
Venue Partner
Downing College, Cambridge, UK
Additional Information
3rd International Symposium (2012)
Meet the Lhasa Limited team:
Coming soon.Press Releases:
Coming soon. 2nd International Symposium (2010)
Archives:
Coming soon.