Edited byMartin Bertau, Erik Mosekilde,and Hans V. Westerhoff

Biosimulation in Drug Development

InnodataFile Attachment9783527622689.jpg

Biosimulation in DrugDevelopment

Edited byMartin Bertau, Erik Mosekilde,and Hans V. Westerhoff

Related Titles

U. Marx, V. Sandig (Eds.)

Drug Testing in vitroBreakthroughs and Trends inCell Culture Technology

2007ISBN 978-3-527-31488-1

M. Reddy, R. S. Yang, M. E. Andersen,H. J. Clewell III

Physiologically BasedPharmacokinetic ModelingScience and Applications

2005ISBN 978-0-471-47814-0

E. Klipp, R. Herwig, A. Kowald,C. Wierling, H. Lehrach

Systems Biology in PracticeConcepts, Implementationand Application

2005ISBN 978-3-527-31078-4

T. I. Oprea (Ed.)

Chemoinformatics inDrug Discovery

2005ISBN 978-3-527-30753-1

P. M. Carroll, K. Fitzgerald (Eds.)

Model Organisms inDrug Discovery

2003ISBN 978-0-470-84893-7

A. Avdeef

Absorption and DrugDevelopmentSolubility, Permeability, and Charge State

2003ISBN 978-0-471-42365-2

H. van de Waterbeemd, H. Lennernäs,P. Artursson (Eds.)

Drug BioavailabilityEstimation of Solubility, Permeability,Absorption and Bioavailability

2003ISBN 978-3-527-30438-7

M. Lutz, T. Kenakin

Quantitative MolecularPharmacology and Informaticsin Drug Discovery

1999ISBN 978-0-471-98861-8

Edited byMartin Bertau, Erik Mosekilde,and Hans V. Westerhoff

Biosimulation in Drug Development

The Editors

Prof. Dr. Martin BertauFreiberg University of Mining andTechnologyInstitute of Technical ChemistryLeipziger Strasse 2909599 FreibergGermany

Prof. Dr. Erik MosekildeTechnical University of DenmarkDepartment of PhysicsSystems Biology GroupFysikvej 3092800 Kgs. LyngbyDenmark

Prof. Dr. Hans V. WesterhoffVrije UniversiteitFaculty of Earth and Life SciencesMolecular Cell PhysiologyDe Boelelaan 10871081 HV AmsterdamThe Netherlands

All books published by Wiley-VCH arecarefully produced. Nevertheless, authors,editors, and publisher do not warrant theinformation contained in these books,including this book, to be free of errors.Readers are advised to keep in mind thatstatements, data, illustrations, proceduraldetails or other items may inadvertently beinaccurate.

Library of Congress Card No.:applied for

British Library Cataloguing-in-Publication DataA catalogue record for this book is availablefrom the British Library.

Bibliographic information published bythe Deutsche NationalbibliothekDie Deutsche Nationalbibliothek lists thispublication in the DeutscheNationalbibliografie; detailed bibliographicdata are available in the Internet at.

© 2008 WILEY-VCH Verlag GmbH & Co.KGaA, Weinheim

All rights reserved (including those oftranslation into other languages). No part ofthis book may be reproduced in any form – byphotoprinting, microfilm, or any othermeans – nor transmitted or translated into amachine language without writtenpermission from the publishers. Registerednames, trademarks, etc. used in this book,even when not specifically marked as such,are not to be considered unprotected by law.

Typesetting VTEX, Vilnius, LithuaniaPrinting betz-druck GmbH, DarmstadtBinding Litges & Dopf GmbH, HeppenheimCover Design WMX-Design, Bruno Winkler,Heidelberg

Printed in the Federal Republic of GermanyPrinted on acid-free paper

ISBN: 978-3-527-31699-1

V

Contents

Preface XVIIList of Contributors XXIII

Part I Introduction

1 Simulation in Clinical Drug Development 3J. J. Perez-Ruixo, F. De Ridder, H. Kimko, M. Samtani, E. Cox,S. Mohanty and A. Vermeulen

1.1 Introduction 31.2 Models for Simulations 71.3 Simulations in Clinical Drug Development: Practical Examples 81.3.1 Predicting the Outcome of Phase I Studies of Erythropoietin

Receptor Agonists 81.3.2 Simulations for Antimicrobial Dose Selection 101.3.3 Optimizing the Design of Phase II Dose Finding Studies 141.3.4 Predicting the Outcome of Phase III Trials Using Phase II Data 191.4 Conclusions 23

2 Modeling of Complex Biomedical Systems 27E. Mosekilde, C. Knudsen and J. L. Laugesen

2.1 Introduction 272.2 Pulsatile Secretion of Insulin 312.3 Subcutaneous Absorption of Insulin 382.4 Bursting Pancreatic β-Cells 432.5 Conclusions 52

3 Biosimulation of Drug Metabolism 59M. Bertau, L. Brusch and U. Kummer

3.1 Introduction 593.2 Experimental Approaches 61

VI Contents

3.2.1 Animal Test Models 613.2.2 Microbial Models 613.3 The Biosimulation Approach 633.4 Ethical Issues 633.5 PharmBiosim – a Computer Model of Drug Metabolism in Yeast 643.5.1 General Concept 643.5.1.1 Chemical Abstraction 643.5.1.2 Biological Abstraction 663.5.2 Initial Steps – Experimental Results 673.5.2.1 Dehalogenation (Pathways II and III) 703.5.2.2 Retro-Claisen Condensation (Pathway IV) 703.5.2.3 Ester Hydrolysis (Pathway VI) 703.5.2.4 Competing Pathways and Stereoselectivity 723.6 Computational Modeling 723.6.1 Selection of the Modeling Software 723.6.2 SBML-compatible Software 733.6.2.1 Cellware 733.6.2.2 Copasi 733.6.2.3 Ecell 733.6.2.4 JigCell 733.6.2.5 JSim 733.6.2.6 Systems Biology Workbench 743.6.2.7 Virtual Cell 743.6.2.8 XPPAUT 743.6.3 CellML-compatible Software 743.6.4 Kinetic Model 753.6.4.1 Methods 753.6.4.2 Model Derivation 763.6.4.3 Results 773.6.5 Stoichiometric Model 783.6.5.1 Methods 783.6.5.2 Model Derivation 783.6.5.3 Results 793.7 Application of the Model to Predict Drug Metabolism 793.8 Conclusions 80

Part II Simulating Cells and Tissues

4 Correlation Between In Vitro, In Situ, and In Vivo Models 89I. González-Álvarez, V. Casabó, V. Merino and M. Bermejo

4.1 Introduction 894.2 Biophysical Models of Intestinal Absorption 914.2.1 Colon 92

Contents VII

4.2.2 Small Intestine 924.2.3 Stomach 924.3 Influence of Surfactants on Intestinal Permeability 934.3.1 Absorption Experiments in Presence of Surfactants 944.3.1.1 Colon 944.3.1.2 Intestine 964.3.1.3 Stomach 964.4 Modeling and Predicting Fraction Absorbed from Permeability

Values 994.4.1 Mass Balance, Time-independent Models 994.4.2 Prediction of the Fraction of Dose Absorbed from In Vitro and

In Situ Data 1014.4.3 Prediction from In Situ Absorption Rate Constant Determined with

Closed Loop Techniques 1014.4.4 Prediction from Permeabilities Through Caco-2 Cell Lines 1024.4.5 Prediction from the PAMPA In Vitro System 1044.5 Characterization of Active Transport Parameters 1074.5.1 In Situ Parameter Estimation 1074.5.2 In Vitro–In Situ Correlation 109

5 Core-Box Modeling in the Biosimulation of Drug Action 115G. Cedersund, P. Strålfors and M. Jirstrand

5.1 Introduction 1165.2 Core-Box Modeling 1175.2.1 Shortcomings of Gray-Box and Minimal Modeling 1175.2.1.1 Full-Scale Mechanistic Gray-Box Modeling 1175.2.1.2 Minimal Modeling Using Hypothesis Testing 1195.2.2 Outline of the Framework 1215.2.3 Model Reduction to an Identifiable Core Model 1215.2.3.1 Identifiability Analysis 1235.2.3.2 Model Reduction 1245.2.4 System Identification of the Core Model 1265.2.4.1 Parameter Estimation 1265.2.4.2 Model Quality Analysis 1285.2.5 Back-Translation to a Core-Box Model 1295.3 A Core-Box Model for Insulin Receptor Phosphorylation and

Internalization in Adipocytes 1325.4 Discussion 1355.5 Summary 137

VIII Contents

6 The Glucose–Insulin Control System 141C. E. Hallgreen, T. V. Korsgaard, R. N. Hansen andM. Colding-Jørgensen

6.1 Introduction 1426.1.1 Glucose and Insulin 1426.1.2 Diabetes Mellitus 1436.1.3 Biosimulation and Drug Development 1446.1.4 The Glucose–Insulin Control System 1456.2 Biological Control Systems 1466.2.1 Features of Biological Control 1466.2.2 The Control System 1476.2.3 Simple Control Types 1486.2.3.1 Proportional Control 1486.2.3.2 Integral Control 1496.2.3.3 Differential Control 1496.2.3.4 PID Control 1506.2.3.5 Predictive Control 1506.3 Glucose Sensing 1516.3.1 Glucokinase 1516.3.1.1 Glucose Phosphorylation 1516.3.1.2 Translocation of Glucokinase 1526.3.1.3 PI Control 1546.3.2 The Beta Cell 1556.3.2.1 First Phase of Insulin Secretion 1566.3.3 The Liver Cell 1586.3.4 The Hepato-portal Sensor 1596.3.5 The Intestine 1606.3.6 The CNS 1606.3.7 Conclusion 1616.4 Glucose Handling 1616.4.1 Glucose Intake 1616.4.1.1 Plasma Glucose 1626.4.2 Glucose Uptake 1626.4.3 Brain 1636.4.4 Liver 1646.4.4.1 Gluconeogenesis 1646.4.4.2 Hepatic Glucose Output and Gluconeogenesis 1656.4.5 Muscle 1696.4.5.1 Glucose Transport 1696.4.5.2 Glucose Phosphorylation 1706.4.5.3 The Fate of Glucose-6-phosphate 1736.4.6 Adipocytes 1776.4.6.1 Triglyceride and Free Fatty Acid 1776.4.6.2 De Novo Lipogenesis 1796.4.7 Conclusion 181

Contents IX

6.5 The Control System at Large 1826.5.1 The Fasting State 1826.5.1.1 Futile Cycles 1836.5.2 Normal Meals 1856.5.3 Glucose Tolerance Tests 1856.5.3.1 Intravenous Glucose Tolerance Test 1856.5.3.2 OGTT and MTT 1876.5.4 The Glucose Clamp 1876.5.4.1 Glucose Infusion Rate 1876.5.4.2 Nutrition During Clamp 1886.5.4.3 Clamp Level 1886.6 Conclusions 1906.6.1 Biosimulation 1906.6.2 The Control System 1916.6.3 Diabetes 1916.6.4 Models and Medicines 192

7 Biological Rhythms in Mental Disorders 197H. A. Braun, S. Postnova, B. Wollweber, H. Schneider, M. Belke,K. Voigt, H. Murck, U. Hemmeter and M. T. Huber

7.1 Introduction: Mental Disorders as Multi-scale and Multiple-systemDiseases 197

7.2 The Time Course of Recurrent Mood Disorders: Periodic, Noisyand Chaotic Disease Patterns 200

7.2.1 Transition Between Different Episode Patterns: The ConceptualApproach 202

7.2.2 A Computer Model of Disease Patterns in Affective Disorders 2037.2.3 Computer Simulations of Episode Sensitization with Autonomous

Disease Progression 2057.3 Mood Related Disturbances of Circadian Rhythms:

Sleep–Wake Cycles and HPA Axis 2077.3.1 The HPA Axis and its Disturbances 2077.3.2 Sleep EEG in Depression 2097.3.3 Neurotransmitters and Hormones Controlling Sleep Pattern and

Mood 2097.3.4 A Nonlinear Feedback Model of the HPA Axis with Circadian

Cortisol Peaks 2107.4 Neuronal Rhythms: Oscillations and Synchronization 2147.4.1 The Model Neuron: Structure and Equations 2157.4.2 Single Neuron Impulse Patterns and Tonic-to-Bursting

Transitions 2177.4.3 Network Synchronization in Tonic, Chaotic and Bursting

Regimes 219

X Contents

7.4.4 Synchronization Between Neurons at Different Dynamic States 2207.5 Summary and Conclusions: The Fractal Dimensions of

Function 222

8 Energy Metabolism in Conformational Diseases 233J. Ovádi and F. Orosz

8.1 What is the Major Energy Source of the Brain? 2338.2 Unfolded/Misfolded Proteins Impair Energy Metabolism 2388.3 Interactions of Glycolytic Enzymes with “Neurodegenerative

Proteins” 2398.4 Post-translational Modifications of Glycolytic Enzymes 2428.5 Triosephosphate Isomerase Deficiency, a Unique Glycolytic

Enzymopathy 2448.6 Microcompartmentation in Energy Metabolism 2478.7 Concluding Remarks 250

9 Heart Simulation, Arrhythmia, and the Actions of Drugs 259D. Noble

9.1 The Problem 2599.2 Origin of the Problem 2619.3 Avoiding the Problem 2649.4 Multiple Cellular Mechanisms of Arrhythmia 2659.5 Linking Levels: Building the Virtual Heart 268

Part III Technologies for Simulating Drug Action and Effect

10 Optimizing Temporal Patterns of Anticancer Drug Delivery bySimulations of a Cell Cycle Automaton 275A. Altinok, F. Lévi and A. Goldbeter

10.1 Introduction 27510.2 An Automaton Model for the Cell Cycle 27710.2.1 Rules of the Cell Cycle Automaton 27710.2.2 Distribution of Cell Cycle Phases 27910.2.3 Coupling the Cell Cycle Automaton to the Circadian Clock 28110.2.4 The Cell Cycle Automaton Model: Relation with Other Types of

Cellular Automata 28210.3 Assessing the Efficacy of Circadian Delivery of the Anticancer Drug

5-FU 28310.3.1 Mode of Action of 5-FU 28310.3.2 Circadian Versus Continuous Administration of 5-FU 28310.3.3 Circadian 5-FU Administration: Effect of Time of Peak Drug

Delivery 284

Contents XI

10.3.4 Effect of Variability of Cell Cycle Phase Durations 28810.4 Discussion 289

11 Probability of Exocytosis in Pancreatic β-Cells: Dependence on Ca2+

Sensing Latency Times, Ca2+ Channel Kinetic Parameters, andChannel Clustering 299J. Galvanovskis, P. Rorsman and B. Söderberg

11.1 Introduction 29911.2 Theory 30111.3 Mathematical Model 30111.4 Dwell Time Distributions 30211.5 Waiting Time Distribution 30411.6 Average Waiting Time 30511.7 Cases N = 1, 2, and 3 30611.8 Numerical Simulations 30711.9 Discussion 30911.10 Conclusions 311

12 Modeling Kidney Pressure and Flow Regulation 313O. V. Sosnovtseva, E. Mosekilde and N.-H. Holstein-Rathlou

12.1 Introduction 31312.2 Experimental Background 31712.3 Single-nephron Model 32012.4 Simulation Results 32612.5 Intra-nephron Synchronization Phenomena 33312.6 Modeling of Coupled Nephrons 33612.7 Experimental Evidence for Synchronization 34112.8 Conclusion and Perspectives 343

13 Toward a Computational Model of Deep Brain Stimulation inParkinson’s Disease 349A. Beuter and J. Modolo

13.1 Introduction 34913.2 Background 35213.2.1 DBS Numbers, Stimulation Parameters and Effects 35213.2.2 DBS and Basal Ganglia Circuitry 35413.2.3 DBS: The Preferred Target Today 35513.2.4 DBS: Paradox and Mechanisms 35613.3 Population Density Based Model 35713.3.1 Modeling Approach: Multiscaling 35713.3.2 Model Equations 35913.3.3 Synapses and the Population Density Approach 362

XII Contents

13.3.4 Solving the Conservation Equation 36413.3.5 Results and Simulations 36413.4 Perspectives 36713.5 Conclusion 368

14 Constructing a Virtual Proteasome 373A. Zaikin, F. Luciani and J. Kurths

14.1 Experiment and Modeling 37414.2 Finding the Cleavage Pattern 37614.3 Possible Translocation Mechanism 37714.4 Transport Model and Influence of Transport Rates on the Protein

Degradation 38114.4.1 The Transport Model 38114.4.2 Analytics – Distribution of Peptide Lengths 38214.4.2.1 One Cleavage Center 38214.4.2.2 Two Cleavage Centers 38314.4.2.3 Maximum in Peptide Length Distribution 38414.5 Comparison with Numerical Results 38514.5.1 Monotonously Decreasing Transport Rates 38614.5.2 Nonmonotonous Transport Rates 38714.6 Kinetic Model of the Proteasome 38914.6.1 The model 38914.6.2 Kinetics 39214.6.3 Length Distribution of the Fragments 39414.7 Discussion 39514.7.1 Development of Modeling 39514.7.2 Kinetics Models and Neurodegenerative Associated Proteasome

Degradation 397

Part IV Applications of Biosimulation

15 Silicon Cell Models: Construction, Analysis, and Reduction 403F. J. Bruggeman, H. M. Härdin, J. H. van Schuppen andH. V. Westerhoff

15.1 Introduction 40315.2 Kinetic Models in Cell Biology: Purpose and Practice 40615.3 Silicon Cell Models 40815.4 Model Reduction by Balanced Truncation 40915.5 Balanced Truncation in Practice 41115.6 Balanced Truncation in Action: Reduction of a Silicon Cell Model

of Glycolysis in Yeast 41415.7 Conclusions 418

Contents XIII

16 Building Virtual Human Populations: Assessing the Propagation ofGenetic Variability in Drug Metabolism to Pharmacokinetics andPharmacodynamics 425G. L. Dickinson and A. Rostami-Hodjegan

16.1 Introduction 42516.2 ADME and Pharmacokinetics in Drug Development 42616.2.1 Absorption 42716.2.2 Distribution 42716.2.3 Drug Metabolism 42816.2.4 Excretion 42816.3 Sources of Interindividual Variability in ADME 42816.3.1 Pharmacokinetic Variability 43016.3.1.1 Variability in Absorption 43016.3.1.2 Variability in Distribution 43116.3.1.3 Variability in Metabolism 43116.3.1.4 Variability in Excretion 43216.3.2 Pharmacodynamic Variability 43216.3.3 Other Sources of Variability in Drug Response 43316.4 Modeling and Simulation of ADME in Virtual Human

Population 43316.4.1 The Need for More Efficient Clinical Trials 43516.4.2 Current Clinical Trial Simulation in Drug Development 43516.4.3 Incorporation of In Vitro Preclinical Data Into CTS 43616.4.3.1 Prediction of Absorption 43616.4.3.2 Prediction of Metabolism 43616.4.3.3 Prediction of Efficacy/Toxicity 43816.5 The Use of Virtual Human Populations for Simulating ADME 43816.5.1 Assessing the Interindividual Variability of In Vivo Drug Clearance

from In Vitro Data 43816.5.2 Prediction of Clearance and its Variability in Neonates, Infants, and

Children 43916.5.2.1 Incorporating Information on Population Variability into

Mechanistic DM-PK-PD Modeling to Assess the Power ofPharmacogenetic Studies 442

16.6 Conclusions 443

17 Biosimulation in Clinical Drug Development 447T. Lehr, A. Staab and H. G. Schäfer

17.1 Introduction 44717.2 Models in Clinical Development 44817.2.1 Model Types 44917.2.1.1 Complexity of Models 45017.3 Clinical Drug Development 45217.3.1 Phase I 452

XIV Contents

17.3.2 Phase II 45317.3.3 Phase III 45417.3.4 Phase IV 45417.4 Modeling Technique: Population Approach 45517.4.1 Model Structure 45617.4.1.1 Structural Model 45617.4.1.2 Statistical Model 45617.4.1.3 Covariate Model 45817.4.1.4 Population Model 45917.4.2 Parameter Estimation 45917.4.3 Building Population Models 46017.5 Pharmacokinetic Models 46117.5.1 Empirical Pharmacokinetic Models 46217.5.1.1 Example: NS2330 (Tesofensine) 46217.5.1.2 Results 46317.5.2 Mechanism-based Pharmacokinetic Models 46517.5.2.1 System Parameters 46717.5.2.2 Drug-dependent Parameters 46717.5.2.3 Examples 46817.6 Pharmacodynamic Models 46817.6.1 Empirical Pharmacodynamic Models 46817.6.1.1 Linking Pharmacokinetics and Pharmacodynamics 46917.6.2 Mechanism-based Pharmacodynamic Models 47017.6.2.1 Examples 47217.6.3 Semi-mechanistic Models 47317.6.3.1 Example: BIBN4096 47317.6.3.2 Results 47417.7 Disease Progression Models 47517.8 Patient Models 47617.8.1 Covariate Distribution Model 47717.8.2 Compliance Model 47717.8.3 Drop-out Model 47817.9 Outlook/Future Trends 47817.10 Software 47917.10.1 General Simulation Packages 47917.10.2 PBPK Software 48017.10.3 Population Approach Software 48017.10.4 Clinical Trial Simulators 481

18 Biosimulation and Its Contribution to the Three Rs 485H. Gürtler

18.1 Ethical Considerations in Drug Development 48518.2 The Three Rs – An Ethical Approach to Animal Experimentation 48618.3 The Three Rs Alternatives 487

Contents XV

18.3.1 Replacement Alternatives 48718.3.2 Reduction Alternatives 48718.3.3 Refinement Alternatives 48818.4 The EU and the Three Rs 48818.4.1 European Partnership 48818.4.2 European Centre for the Validation of Alternatives 48918.4.3 European Consensus Platform for Alternatives 49018.5 Applying the Three Rs to Human Experimentation 49018.6 Biosimulation and its Contribution to the Three Rs 49118.6.1 Biosimulation – A New Tool in Drug Development 49118.6.2 The Challenges in Drug Development 49118.6.3 Biosimulation’s Contribution to Drug Development 49318.6.4 Biosimulation’s Contribution to the Three Rs 494

Index 497

XVII

Preface

Drug development is far from a straightforward endeavor. It starts with the identifi-cation of a pharmaceutically promising substance, the potential of which is furtherinvestigated in near-exhausting physiological analyses (Chapter 1). If it is foundeffective, this does not necessarily mean that we know all the molecular conse-quences when administered. Each patient is an individual, with unique features,which may mean patient treatment at the level of the individual; so-called personalmedicine. This illustrates the complexity of drug development and discovery.

Does it get any simpler when we consider the physiological and molecular eventsupon drug ingestion? A perorally administered drug first of all enters the gastro-intestinal tract where it encounters the biotransformatory activity of the intestinalmicroflora. Our knowledge of which compounds finally enter the blood stream isfragmentary at best, not to mention the interactions of the drug with the differentorgans and the resulting organ–organ interplays (Chapter 3). And this is not theonly question arising. How do drugs interact with human physiology? Whether wecan simulate drug effects then becomes very important (Chapters 6 and 8). Fur-ther, how do the organs interact with each other upon the ingestion of a drug?Can we achieve a whole-body simulation? Or virtual populations? Can we, and towhat extent, treat individuality and gender in drug administration? What advan-tages lie behind technologies that can simulate drug action and effect (Chapters 11and 14)? If we can do all this, how can we use these simulations to our advantage indrug discovery and development? To what are undesired side-effects attributable?What do we know of drug–drug interactions, drug–drug metabolite interactions,and their outcomes on organs or on the interplay between different organs? Canwe, by applying novel methodologies, shorten package inserts, even make themsuperfluous?

In other words, drug development is increasingly characterized by the require-ment to understand highly complex biological processes and to exploit the rapidlygrowing amount of biological information. The methods that are currently appliedin the development of new medicines require new and effective means to evaluate,integrate, and accumulate this biological knowledge. Essential improvements willresult from the use of computational models that can provide a dynamic and morequantitative description of the relevant biological, pathological, and pharmacoki-

XVIII Preface

netic processes. These ambitious goals will in fact be achieved by the integration ofa new methodology, biosimulation, into the drug development process.

Why biosimulation? The high development costs and long lead times of newdrugs are associated with numerous clinical trials that a drug must undergo todocument its function and show that adverse side-effects do not occur. The use ofprofessional simulation models can simplify the drug development process enor-mously by exploiting the information available in each individual trial much moreeffectively. Besides, this approach is used in many industries where computer sim-ulation allows new concepts and designs to be tested and optimized long beforethe first example of the product is manufactured. Consequently the simulationapproach is strongly recommended by the United States Food and Drug Admin-istration, both because of its potential for reducing the development costs, and be-cause of the associated reduction in the use of laboratory animals and test persons.Through the establishment of “virtual populations”, as information is gradually ac-cumulated, variations in the efficacy and possible side-effects of a new drug can bepredicted on a gender- and age-specific basis and/or for patients with specific genemodifications (Chapter 16).

The establishment of a more quantitative description of biomedical systems asthe foundation for a disease-driven drug development process requires an unusu-ally broad range of insights and skills in biological as well as in technical and math-ematical realms. This approach is highly interdisciplinary and commands out-standing expertise in bioinformatics, biochemistry, cellular biology, electrophysiol-ogy, intercellular communication, physiology, endocrinology, neurology, nephrol-ogy, pharmacology, pharmacokinetics, systems biology, complex systems theory,and in silico modeling and simulation techniques (Chapters 2 and 15). It also re-quires close collaboration between experimental and theoretical partners. Modelingis the translation of (often) experimentally derived scientific hypotheses into a for-mal mathematical framework. In order to improve, the models must continuouslybe confronted with new experimental data; and precisely this process of generatingnew hypotheses and formulating critical experiments represents the most effectiveway of expanding our knowledge about drug function.

The use of biosimulation and mathematical models in drug development isbased on the circumstance that, contrary to the conventional assumption of home-ostasis, many biological systems have unstable equilibrium points and operate ina pulsatile or oscillatory mode. This is the case, for instance, for the release of pitu-itary hormones that typically occurs at more or less regular intervals of two hours,or for the optimal point-of-time of drug administration (Chapters 10 and 12). Or,in several cases it has been reported that the cellular response to an oscillatorysignal is stronger than the response to a constant signal of the same average mag-nitude, suggesting that the oscillatory dynamics play a role in the regulatory effectof the hormone. Rhythmic and pulsatile signals are also encountered in intracel-lular processes as well as in communication between cells. Many nerve cells areexcitable and respond in an unusual fashion to small external stimuli. Other cellsdisplay complicated patterns of spikes and bursts (Chapters 7 and 13).

Preface XIX

Biosimulation can also contribute to the development of a methodology for theprediction of drug-likelihood. Neural networks can be used to prescreen drug can-didates and to predict absorption rates, binding affinities, metabolic rate constants,etc., from knowledge about previously examined compounds. This allows the sub-sequent biochemical screening to be performed on a reduced set of candidates thathave a significantly increased likelihood of possessing the desired functionality.In this context reports from the European Federation for Pharmaceutical Sciences(EUFEPS) and from other European organizations have repeatedly stressed theneed for a targeted effort to speed up the development of new and safe drugs.In fact, the introduction of new technologies such as high-throughput screening(HTS), computational chemistry, and combinatorial and automated chemistry hasmade research and discovery in the pharmaceutical industry significantly moreeffective. These technologies allow the screening of compound libraries for po-tential drugs at rates that are hundreds of times faster than even the best skilledchemist can do. Thus, a major improvement in drug development will require amore knowledge-oriented process that builds on a detailed and quantitative under-standing of the biological and pathological processes associated with the function-ing of the drug.

Specifically the biosimulation approach to drug discovery provides efficientmeans to: (1) define and control the conditions under which experiments and clin-ical tests are performed, (2) extract the information available in the individualtrial and validate it in terms of current knowledge in cell biology, medicine, etc.,(3) accumulate information from trial to trial and redesign the trial procedure tobecome an adaptive process where information acquired in one trial is immedi-ately used to improve the process, (4) extrapolate results obtained from experimentson cell cultures and from animal experiments to applications for human patients,(5) predict the variation of drug efficacy and the occurrence of side-effects, takingaccount of genetic modifications, and gender, age, or weight characteristics, and(6) predict the likelihood that a particular chemical compound will function as adrug on the basis of knowledge about related compounds.

How can we implement biosimulation within the drug development process?For these purposes, let us first take a look at what information can be taken fromsimulation models. They describe the temporal variation of a system in terms ofthe processes and interactions that are presumed to be at work. If the informationobtained from these models is viewed in connection with the development of adrug, then the model will combine a pharmacokinetic description of the absorp-tion, distribution, metabolism, and excretion of the drug with a detailed represen-tation of the mechanisms responsible for its function and for the development ofside-effects and possible synergetic interactions with other drugs (Chapters 3 and4). To the extent that they are important, this description will include a represen-tation of the drug’s interaction with cellular receptors of the intracellular reactioncascades, and of the drug’s effects on the intercellular communication or cellularenergy metabolism (Chapters 8, 9, 11 and 14). It may also be important to examineinteractions with specific organs (heart, liver, kidney, etc.) as well as with hormonaland immune regulations (Chapters 6, 9 and 12). These combined efforts will finally

XX Preface

result in a simulation model that so to speak translates our knowledge about thebiological system into mathematical equations. In the initial stage of the drug de-velopment process, one can use the simulation model to test any hypothesis onemight have regarding the function of the drug vis-à-vis the established biologicalunderstanding. With approaches from bioinformation, one can change the productin order to optimize its function, and even before the first molecule is produced onecan estimate the likelihood that a given agent will function as a drug (Chapter 17).

During the subsequent trial phase, the simulation model can be used as a vehicleto define an effective test protocol. The model can be used to check that the infor-mation obtained from the tests is consistent. Provided that it represents a properrepresentation of the underlying pharmacokinetic and biological mechanisms, themodel can be used to predict the metabolism of a drug compound or its effectoutside its normal physiological regime and under conditions not previously expe-rienced (Chapter 2). To the extent that these predictions are substantiated, our un-derstanding of the drug’s action is gradually extended. If the predictions are falsewe must examine both the hypotheses and the experimental procedures. The ad-vantage is here that any discrepancy between hypotheses and experimental resultsappear right away and in a clear manner.

At the present very few models exist that can describe a disease longitudinally,i.e. from its very start to the final cure. The advantage of simulation models is thatthey can represent available information about the relevant biological and patho-logical processes on any time scale. Hence, it is possible to describe how a diseasedevelops over years with respect to the performed treatment (the so-called diseaselife-cycle). This is of particular interest in connection with modern life-style dis-eases where active collaboration and compliance by the patient plays an impor-tant role. By means of a simulation model health care providers can examine thelong-term consequences of possible adjustments in the treatment, and from sim-plified versions of the models the patients may learn to appreciate the significanceof following a prescribed treatment and/or keeping a specific diet. Insights gainedfrom such models can also be used to tailor health care policies, public campaigns,or educational programs. Examples could include the models developed some 15years ago to design policies against the spread of AIDS or the treatment of diabetes(Chapter 6).

The 3R Declaration of Bologna on Reduction, Refinement and Replacement Alter-natives and Laboratory Animal Procedures adopted by the Third World Congresson Alternatives and Animal Use in the Life Sciences (31 August 1999) calls fora reduced use of laboratory animals in cosmetic, chemical, and medical researchthrough the use of alternative methods and through the use of methods that canproduce the same information with fewer animal experiments. The United StatesFood and Drug Administration strongly advocates the use of computer simula-tion models, and except for the replacement of animal experiments by human ex-periments, it is unlikely that any other technique than biosimulation can accom-plish the goals of the 3R Declaration. Application of computer simulations in thedrug development process can reduce the use of laboratory animals significantlythrough a more rational exploitation of the information acquired in each test and

Preface XXI

through a better planning of the experiments. As information is accumulated us-ing in silico models, the use of laboratory animals can gradually be replaced bycomputer models. However, establishing the information needed in the biologicalsimulation models cannot go without a minimum number of independent animalexperiments. But their number will be significantly smaller than the number ofanimal experiments required by standard test procedures. In this way the biosimu-lation can contribute to the fulfillment of the goals of the 3R Declaration. Preciselythe same arguments for the use of a more rational test procedure also apply toexperiments on volunteers. Biosimulation may therefore contribute towards theestablishment of a 3R Declaration pertaining to volunteer experiments.

All these issues are addressed by activities which are currently being conductedin part by the BioSim Network (www.biosim-network.net) which is a Network ofExcellence under the Life Sciences, Genomics and Biotechnology for Health ThematicPriority Area of the Sixth European Framework Programme. The Network was ini-tiated on 1 December 2005 through a grant from the European Commission. Theimmediate goal of the network is to develop a modeling and simulation approachthat can render the drug development process more rational to the benefit of thepatient. In a broader perspective the network aims at contributing in an essentialmanner to the development of an integrated and quantitative understanding ofbiological processes in accordance with the 3R principles.

BioSim includes a total of 40 partners, out of which 26 are academic researchgroups from various universities throughout Europe and nine are industrial part-ners. The network also includes the regulatory agencies from Denmark, Holland,Sweden, and Spain, since the introduction of simulation models in the drug devel-opment process will require significant changes in the way the regulatory authori-ties evaluate new drug candidates, and the regulatory authorities need to establishtheir own expertise in the use of simulation models. Industrial partners are in-volved in a variety of different activities from the development of new simulationsoftware to the test of drug candidates. Half of the academic groups are primarilyexperimentally oriented, and the other half cover many forms of modeling exper-tise. The network also involves hospital departments that perform experimentaltreatments of depression, various forms of tremor, and cancer. In this way a plat-form for fruitful interdisciplinary collaboration has been established.

In this way biosimulation aims at developing insights, concepts, and illustrativeexamples that can support drug development by an approach in which mathemat-ical modeling and simulation are used as essential tools. The goal is to introducethis methodology as a natural tool in the initial selection of new drug candidates, inthe planning of in vitro and in vivo trials, and in the interpretation of the obtainedresults. Correctly applied, this approach should allow pointless lines of research tobe exposed – and ineffective or toxic drug candidates to be abandoned – at an earlystage in the process.

With this collection of 17 contributions this book reflects the broad variety ofmethods, technologies, and fields of applications that biosimulation adopts withinthe drug development process. What is further illustrated by these chapters isthe high complexity of biological systems. This book also shows that enormous

XXII Preface

progress has already been achieved now in successfully converting the languageof biology, biochemistry, and physiology into terms of mathematics, i.e. the trans-lation of in vivo into in silico. Yet, even with this research already begun, it hasbecome clear that everything that can be summed up today as a quantum leap inthe mathematical treatment of biological systems can be only the starting pointfor much more complex research which cannot be realized except by more inten-sive interdisciplinary approaches than ever to understand how living systems work.With respect to drug development this book demonstrates that both much has beendone and much remains to be done to improve drug development processes. Allthe approaches and strategies depicted here point clearly in one direction: towardthe benefit of the patient.

It is, above all, especially this feature that characterizes the envisaged impact ofbiosimulation on drug development. The diverse research activities in neighboringscientific areas have too long been conducted in parallel. In addition each disciplinehas accumulated a wealth of knowledge, the combination of which with others ismuch more a must than an interesting option for patients’ benefit. This was alsothe impetus for us to combine the biosimulation activities of different internationalresearch groups into one book. As will be obvious to the reader, biosimulation is afascinating multi-faceted research field with a direct link to application. For this rea-son it is not only academic groups speaking in this book, but also industry whosejob it is to translate scientific results in terms of commercially available drugs.

The editors of this book would like to thank all contributors for the professionalmanner in which they have demonstrated the usefulness of biological modelingand simulation in drug development. It is their commitment and their investmentof effort and time we are indebted to, without which this book never would havebeen accomplished. Also we gratefully acknowledge the integrating activities of theEuropean Commission who realized the potential of biosimulation in drug devel-opment with regard to the benefit of the patient and who invested additional effortin providing the research framework BioSim under the roof of which internationalactivities in the field of biosimulation in drug development are being combined.Special thanks also go to Dr. Waltraud Wüst and Dr. Frank Weinreich from thepublishing house Wiley-VCH who from the very beginning greatly supported ouractivities in preparing this book and who always had an open mind for our ideasand special wishes along the publishing process. Last but not least we would liketo thank the assistance of Dr. Frank Bruggeman and Dr. Olga Sosnovtseva.

It is our concern to spread our enthusiasm for the fascinating and facet-rich fieldof biosimulation to a broad scientific community and share it with them. If this hasbeen accomplished, then the goal of this book will have achieved its purpose.

Freiberg, Lyngby, Amsterdam Martin BertauSeptember 2007 Erik Mosekilde

Hans Westerhoff

XXIII

List of Contributors

Atilla AltinokUniversité Libre de BruxellesUnité de Chronobiologie ThéoriqueFaculté des SciencesCampus Plaine, C.P. 2311050 BrusselsBelgium

Marcus BelkePhilipps University of MarburgInstitute of PhysiologyDeutschhausstrasse 235037 MarburgGermany

Martin BertauFreiberg University of Mining andTechnologyInstitute of Technical ChemistryLeipziger Strasse 2909599 FreibergGermany

Marival BermejoUniversidad de ValenciaFacultad de FarmaciaVicente Andrés Estellés s/n46100 Burjassot (Valencia)Spain

Anne BeuterUniversité Bordeaux 2Institut de Cognitique146, rue Léo Saignat33076 Bordeaux CedexFrance

Hans A. BraunPhilipps University of MarburgInstitute of PhysiologyDeutschhausstrasse 235037 MarburgGermany

Frank J. BruggemanCentrum voor Wiskunde enInformatica (CWI)Multiscale Modelling and NonlinearDynamicsKruislaan 4131098 SJ AmsterdamThe NetherlandsandVrije UniversiteitFaculty of Earth and Life SciencesMolecular Cell PhysiologyDe Boelelaan 10851081 HV AmsterdamThe Netherlands

XXIV List of Contributors

Lutz BruschDresden University of TechnologyCentre of Information Services andHigh Performance Computing01062 DresdenGermany

Vicente CasabóUniversidad de ValenciaFacultad de FarmaciaVicente Andrés Estellés s/n46100 Burjassot (Valencia)Spain

Gunnar CedersundUniversity of LinköpingDepartment of Cell Biology58185 LinköpingSweden

Morten Colding-JørgensenNovo Nordisk A/SDevelopment Projects ManagementNovo Allé2880 BagsværdDenmark

Eugene CoxJohnson & Johnson PharmaceuticalAdvanced PK/PD Modeling &SimulationClinical Pharmacology2340 BeerseBelgium

Filip De RidderJohnson & Johnson PharmaceuticalResearch & DevelopmentAdvanced Modeling & SimulationBiometrics2340 BeerseBelgium

Gemma L. DickinsonUniversity of SheffieldSchool of Pharmacy andPharmaceutical Sciences3.119 Stopford BuildingOxford RoadManchester M13 9PTUnited Kingdom

Christine Erikstrup HallgreenNovo Nordisk A/SDevelopment Projects ManagementNovo Allé2880 BagsværdDenmark

Juris GalvanovskisUniversity of OxfordOCDEM, Department of PhysiologyChurchill Hospital, Old RoadOxford OX3 7LJUnited Kingdom

Albert GoldbeterUniversité Libre de BruxellesUnité de Chronobiologie ThéoriqueFaculté des SciencesCampus Plaine, C.P. 2311050 BrusselsBelgium

Isabel González-ÁlvarezUniversidad de ValenciaFacultad de FarmaciaVicente Andrés Estellés s/n46100 Burjassot (Valencia)Spain

Hanne GürtlerConsultant to Novo Nordisk A/SLiving United ConsultKongevejen 23450 AllerødDenmark

List of Contributors XXV

René Normann HansenNovo Nordisk A/SDevelopment Projects ManagementNovo Allé2880 BagsværdDenmark

Hanna M. HärdinVrije UniversiteitFaculty of Earth and Life SciencesMolecular Cell PhysiologyDe Boelelaan 10851081 HV AmsterdamThe NetherlandsandCentrum voor Wiskunde enInformatica (CWI)Scientific Computing and ControlTheoryP.O. Box 940791090 GB AmsterdamThe Netherlands

Ulrich HemmeterUniversity of MarburgDepartment of Psychiatry andPsychotherapyRudolf-Bultmannstrasse 4335033 MarburgGermany

Niels-Henrik Holstein-RathlouDepartment of Medical PhysiologyPanum InstituteUniversity of Denmark2200 Copenhagen NDenmark

Martin T. HuberUniversity of MarburgDepartment of Psychiatry andPsychotherapyRudolf-Bultmannstrasse 4335033 MarburgGermany

Mats JirstrandFraunhofer-Chalmers Research Centrefor Industrial MathematicsChalmers Teknikpark41288 GothenburgSweden

Hui KimkoJohnson & Johnson PharmaceuticalResearch & DevelopmentAdvanced PK/PD Modeling &SimulationClinical PharmacologyRaritan, NJ 08869USA

Carsten KnudsenThe Technical University of DenmarkDepartment of PhysicsSystems Biology GroupFysikvej 3092800 Kongens LyngbyDenmark

Thomas Vagn KorsgaardNovo Nordisk A/SDevelopment Projects ManagementNovo Allé2880 BagsværdDenmark

Ursula KummerDepartment of Modelling ofBiological ProcessesInstitute for Zoology/BIOQUANTINF 26769120 HeidelbergGermany

Jürgen KurthsUniversity of PotsdamInstitute of PhysicsAm Neuen Palais 1014469 PotsdamGermany

XXVI List of Contributors

Jakob Lund LaugesenThe Technical University of DenmarkDepartment of PhysicsSystems Biology GroupFysikvej 3092800 Kgs. LyngbyDenmark

Thorsten LehrBoehringer Ingelheim PharmaGmbH & Co. KGDepartment of Drug Metabolismand PharmacokineticsBirkendorfer Strasse 6588397 Biberach an der RissGermany

Francis LéviINSERM U776Université Paris Sud XIRythmes Biologiques et CancersHôpital Paul Brousse94800 VillejuifFrance

Fabio LucianiUniversity of New South WalesSchool of Biotechnology andBiomolecular Sciences2026 SydneyAustralia

Virginia MerinoUniversidad de ValenciaFacultad de FarmaciaVicente Andrés Estellés s/n46100 Burjassot (Valencia)Spain

Julien ModoloUniversité Bordeaux 2Institut de Cognitique146, rue Léo Saignat33076 Bordeaux CedexFrance

Surya MohantyJohnson & Johnson PharmaceuticalResearch & DevelopmentAdvanced Modeling & SimulationBiometrics1125 Trenton Harbourton RdTitusville, NJ 08560-1504USA

Erik MosekildeThe Technical University of DenmarkDepartment of PhysicsSystems Biology GroupFysikvej 3092800 Kgs. LyngbyDenmark

Harald MurckNovartis Pharmaceuticals CorporationUS Clinical Development andMedical AffairsOne Health Plaza, Building 701, 642BEast Hanover, NJ 07936-1080USA

Denis NobleDepartment of PhysiologyAnatomy and GeneticsParks RoadOxford, OX1 3PTUnited Kingdom

Ferenc OroszHungarian Academy of SciencesInstitute of EnzymologyBiological Research CenterP.O. Box 71518 BudapestHungary

Judit OvádiHungarian Academy of SciencesInstitute of EnzymologyBiological Research CenterP.O. Box 71518 BudapestHungary

List of Contributors XXVII

Juan José Perez-RuixoJohnson & Johnson PharmaceuticalResearch & DevelopmentAdvanced Modeling & SimulationClinical Pharmacology2340 BeerseBelgium

Svetlana PostnovaPhilipps University of MarburgInstitute of PhysiologyDeutschhausstrasse 235037 MarburgGermany

Patrik RorsmanUniversity of OxfordOCDEM, Department of PhysiologyChurchill Hospital, Old RoadOxford OX3 7LJUnited Kingdom

Amin Rostami-HodjeganUniversity of SheffieldAcademic Unit of ClinicalPharmacologyM Floor Medicine and PharmacologyRoyal Hallamshire HospitalSheffield S10 2JFUnited Kingdom

Mahesh SamtaniJohnson & Johnson PharmaceuticalResearch & DevelopmentAdvanced PK/PD Modeling &SimulationClinical PharmacologyRaritan, NJ 08869USA

Hans Günter SchäferBoehringer Ingelheim PharmaGmbH & Co. KGDepartment of Drug Metabolismand PharmacokineticsBirkendorfer Strasse 6588397 Biberach an der RissGermany

Horst SchneiderPhilipps University of MarburgInstitute of PhysiologyDeutschhausstrasse 235037 MarburgGermany

Bo SöderbergComplex Systems DivisionDepartment of Theoretical PhysicsLund University22834 LundSweden

Olga V. SosnovtsevaSystems Biology GroupDepartment of PhysicsThe Technical University of DenmarkFysikvej 3092800 LyngbyDenmark

Alexander StaabBoehringer Ingelheim PharmaGmbH & Co. KGDepartment of Drug Metabolismand PharmacokineticsBirkendorfer Strasse 6588397 Biberach an der RissGermany

Peter StrålforsUniversity of LinköpingDepartment of Cell Biology58185 LinköpingSweden

Jan H. van SchuppenCentrum voor Wiskunde enInformatica (CWI)Scientific Computing and ControlTheoryP.O. Box 940791090 GB AmsterdamThe Netherlands

XXVIII List of Contributors

An VermeulenJohnson & Johnson PharmaceuticalResearch & DevelopmentAdvanced PK/PD Modeling &SimulationClinical Pharmacology2340 BeerseBelgium

Karlheinz VoigtPhilipps University of MarburgInstitute of PhysiologyDeutschhausstrasse 235037 MarburgGermany

Hans V. WesterhoffVrije UniversiteitFaculty of Earth and Life SciencesMolecular Cell PhysiologyDe Boelelaan 10851081 HV AmsterdamThe Netherlands

andManchester Interdisciplinary Biocentre(MIB)131 Proncess StreetManchester M1 7NDUnited Kingdom

Bastian WollweberPhilipps University of MarburgInstitute of PhysiologyDeutschhausstrasse 235037 MarburgGermany

Alexey ZaikinUniversity of PotsdamInstitute of PhysicsAm Neuen Palais 1014469 PotsdamGermany