Simulation for designing clinical trials _ a pharmacokinetic-pharmacodynamic modeling perspective.pdf

Simulation for Designing Clinical Trials A Pharmacokinetic-Pharmacodynamic Modeling Perspective Hui C.Kimko Johnson&Johnson Pharmaceutical Research&Development Raritan,New Jersey,U.S.A.Stephen B.Duffull University of Queensland Brisbane,Queensland,Australia M A R C E L MARCEL DEKKER,INC.NEW YORK EASEL D-E K Y E R Copyright 2003 Marcel Dekker,Inc.Library of Congress Cataloging-in-Publication DataA catalog record for this book is available from the Library of Congress.ISBN:0-8247-0862-8This book is printed on acid-free paper.HeadquartersMarcel Dekker,Inc.270 Madison Avenue,New York,NY 10016tel:212-696-9000;fax:212-685-4540Eastern Hemisphere DistributionMarcel Dekker AGHutgasse 4,Postfach 812,CH-4001 Basel,Switzerlandtel:41-61-260-6300;fax:41-61-260-6333World Wide Webhttp:/The publisher offers discounts on this book when ordered in bulk quantities.For moreinformation,write to Special Sales/Professional Marketing at the headquarters addressabove.Copyright 2003 by Marcel Dekker,Inc.All Rights Reserved.Neither this book nor any part may be reproduced or transmitted in any form or by anymeans,electronic or mechanical,including photocopying,microfilming,and recording,or by any information storage and retrieval system,without permission in writing fromthe publisher.Current printing(last digit):10 9 8 7 6 5 4 3 2 1PRINTED IN THE UNITED STATES OF AMERICACopyright 2003 Marcel Dekker,Inc.To Dennis,Edwin,Jason&Aynsley,and Ella&HollyCopyright 2003 Marcel Dekker,Inc.ForewordComputer simulations of clinical trials,employing realistic virtual subjects andtypical trial conditions,based on both experimentally informed disease progressand drug intervention models,originated within the last decade.Previously,clini-cal trials were designed using ad hoc empirical approaches,unaided by a system-atic clinical pharmacology orientation or a quantitative pharmacokinetic-pharma-codynamic framework,leading to highly inefficient drug development programs.Stimulated by research and educational contributions from academia,and encour-agement from regulatory agencies,expert simulation teams have recently de-ployed hundreds of clinical trial simulation projects.The advent of modern clini-cal trial simulation is transforming clinical drug development from na veempiricism to a mechanistic scientific discipline.The editors and contributors have provided more than just a comprehensivehistory,critical vocabulary,insightful compilation of motivations,and clear ex-planation of the state of the art of modern clinical trial simulation.This bookadvances a rigorous framework for employing simulation as an experiment,ac-cording to a predefined simulation plan that reflects good simulation practices.*We describe attributes of the multidisciplinary simulation team that position itto achieve benefits of enhanced communication and collaboration during the de-velopment and employment of the simulation.*Holford NHG,Hale M,Ko,HC,Steimer J-L,Sheiner LB,Peck CC.Simulation in Drug Develop-ment:Good Practices,1999.http:/cdds.georgetown.edu/SDDGP.html.Copyright 2003 Marcel Dekker,Inc.While the far future of scientific drug development is difficult to predict,successful advancement and integration of clinical trial simulation lead to a dar-ing prediction:in the not so distant future,most clinical trials will be virtualonly a few actual trials will be undertaken.These few human trials will be de-signed to inform simulation models and to confirm model predictions.The aca-demic,pharmaceutical,and regulatory scientists who have articulated the stateof the art of clinical trial simulations in this book provide the first comprehensivedescription of a breakthrough technology that is enabling this bold departure frominefficient past practices.Carl Peck,M.D.Center for Drug Development ScienceGeorgetown University,Washington D.C.,U.S.A.Copyright 2003 Marcel Dekker,Inc.PrefaceSimulation has been used widely in various disciplines such as engineering,phys-ics,and economics to support the development and testing of the performanceof systems.Some of the most notable examples arise from the wealth of experi-ence in the aerospace industry,in which over the past 30 years it has becomeroutine practice to simulate the performance of aircraft before production andlaunch.The use of simulations in these industries has been shown to reduce costsand shorten development time.Some of the experience gained from these highlytechnical systems has pervaded the highly stochastic area of biological systems.Within the pharmaceutical industry,this has culminated in the growth of model-ing and simulation in the drug development process,where computer simulationis gaining popularity as a tool for the design of clinical trials.The integration of modeling and simulation into drug developmenthas beena gradual process in spite of the long-term use of stochastic simulations by bio-statisticians for exploration and analyses of data and deterministic simulations bypharmacologists for descriptive purposes.However,simulation has considerablepotential for design of trials as evidenced by the rapid growth in discussiongroups,conferences,and publications.The basis of the use of simulation lies inthe argument that if,in theory,a virtual trial could be constructed that incorpo-rated all relevant influences(controllable and uncontrollable,and deterministicand stochastic)and their related outcomes,the researcher could then explore theinfluence of changes in the design on the performance of the trial.If this theoreti-cal construct were a reality,then it is equally conceivable that trial designs couldbe selected based on their probability for success.While this seems a straightfor-Copyright 2003 Marcel Dekker,Inc.ward task,albeit computationally intensive,its introduction has been at the mercyof the availability of powerful computing.Since fast computing machines arenow available on almost every office desk,it is no surprise that both the designof trials and the necessary knowledge and understanding of the time course ofdrug effects that underpins the design process have shown a dramatic upsurge.The parallel development of more complex and mechanistic models for drugeffects and design of trials is not coincidental,since an understanding of theeffects of drugs is paramount for design of trials and the more complex modelsthemselves rely heavily on computational methods for their solution.This book describes the background and lays the foundation for simulationas a tool for the design of clinical trials.The target audience is any researcher orpractitioner who is involved in the design,implementation,analysis,or regulatorydecisions concerning clinical trials.This book does not embrace all aspects oftrial design,nor is it intended as a recipe for using computers to design trials.Rather,it is a source of information that enables the reader to gain a better under-standing of the theoretical background and knowledge of the practical applica-tions of simulation for design.It is assumed that the reader has a working under-standing of pharmacokinetics and pharmacodynamics,modeling,and the drugdevelopment process.In addition,some knowledge of types and practicalities ofdesigns commonly used for clinical trials is assumed.The book is divided into parts that describe model development,modelevaluation,execution ofsimulation,choice of design,and applications.It is usefulto partition the simulation model into submodels(e.g.,input-output model,covar-iate distribution model,execution model)in order to describe specific aspects ofthe process.The input-output model(Chapter 2)describes the relationship be-tween dosing schedule and response in an explanatory manner for any givenpatient.This model itself usually comprises a number of submodels:the phar-macokinetic and pharmacodynamic models,disease progression models,and(patho)physiological models of homeostatic systems in the body.The covariatedistribution model(Chapter 3)describes the characteristics of the virtual patientthat affect the input-output models and execution models.The execution model(Chapter 4)describes the deviation from the nominal design,thereby mimickingthe actual trial conduct.Details of model evaluation methods are provided in Chapter 5.The me-chanics required for simulation of a trial,including replications,random numbergeneration,and the implementation of numerical integration are outlined in Chap-ter 6.Analysis of replicates of the subsequent virtual trial is discussed in Chapter7.Chapter 8 addresses the important issue of the sensitivity of the trial designto assumptions in the development of the models that underpin the response ofthe virtual patient.Finally,in this section discussion is raised about how a suffi-cient design might be selected from all possible designs(Chapter 9).While simulation as an investigation tool has proven conceptually to beCopyright 2003 Marcel Dekker,Inc.straightforward,complete acceptance by regulatory authorities and the pharma-ceutical industry remains elusive.Details of perspectives by regulatory authori-ties,academia,and the pharmaceutical industry are provided by Chapters 10,11,and 12,respectively.In addition to these perspectives,an overview and historyof mechanism-based model development for physiological/pharmacological pro-cesses are presented in Chapter 13.We have also included a part devoted to applications of simulation for trialdesign and evaluation(Chapters 14 to18).These include a wide range of practicalapplications,including optimization of sampling strategies,dose selection,inte-gration of optimal design with simulation,prediction of efficacy,and side effects.We accept that our current knowledge of predicting clinical responses forthe individual patient pales beside that imagined by science fiction writers.Ourway of exploring the nature of drug activity is limited to conducting clinical trialsin the hope of learning about clinical responses and confirming these findingsusing methods that are often empirical.Since it is recognized that the analysisof data is dependent on the quality of the data,and the quality of the data depen-dent on the quality of the study design,we are dependent on adequately designedclinical trials to pave the way for better treatments.In due course,it is expectedthat methods that promote more informative and rigorous designs such as thosebased on modeling and simulation will provide the same benefits for drug devel-opment that have been seen in other industries.We thank the authors of the chapters and Marcel Dekker,Inc.,for providingthe opportunity to publish this book.Hui C.KimkoStephen B.DuffullCopyright 2003 Marcel Dekker,Inc.ContentsForewordPrefaceContributors1.Introduction to Simulation for Design of Clinical TrialsHui C.Kimko and Stephen B.DuffullI.Models for Simulation2.Input-Output ModelsNicholas H.G.Holford3.Defining Covariate Distribution Models for Clinical TrialSimulationDiane R.Mould4.Protocol Deviations and Execution ModelsHelen Kastrissios and Pascal Girard5.Determination of Model AppropriatenessPaul J.Williams and Ene I.EtteCopyright 2003 Marcel Dekker,Inc.II.Computation in Simulation6.Computational Considerations in Clinical Trial SimulationsJohn C.Pezzullo and Hui C.KimkoIII.Analyses of the Simulations7.Analysis of Simulated Clinical TrialsEne I.Ette,Christopher J.Godfrey,Stephan Ogenstad,andPaul J.Williams8.Sensitivity Analysis of Pharmacokinetic and PharmacodynamicModels in Clinical Trial Simulation and DesignIvan A.Nestorov9.Choice of Best DesignJonathan P.R.Monteleone and Stephen B.DuffullIV.Perspectives on Clinical Trial Simulation10.Clinical Trial Simulation(CTS):A Regulatory ClinicalPharmacology PerspectivePeter I.D.Lee and Lawrence J.Lesko11.Academic Perspective:Modeling and Simulation as a TeachingToolNicholas H.G.Holford and Mats O.Karlsson12.Modeling and Simulation of Clinical Trials:An IndustryPerspectiveTimothy Goggin,Ronald Gieschke,Goonaseelan(Colin)Pillai,Ba rbel Fotteler,Paul Jordan,and Jean-Louis Steimer13.History-Informed Perspectives on the Modeling and Simulationof Therapeutic Drug ActionsJohn UrquhartV.Applications14.Evaluation of Random Sparse Sampling Designs for aPopulation Pharmacokinetic Study:Assessment of Power andBias Using SimulationMatthew M.Hutmacher and Kenneth G.Kowalski15.Use of Modeling and Simulation to Optimize Dose-FindingStrategiesJaap W.Mandema and Wenping WangCopyright 2003 Marcel Dekker,Inc.16.Model-Based Integration for Clinical Trial Simulation andDesign:A Phase II Case Study for NaratriptanIvan A.Nestorov,Gordon Graham,Stephen B.Duffull,andLeon Aarons17.Prediction of Hemodynamic Responses in Hypertensive andElderly Subjects from Healthy Volunteer Data:The Example ofIntravenous NicardipinePatrice Francheteau,Henri Merdjan,Madeleine Guerret,Ste phane Kirkesseli,Paolo Sassano,and Jean-Louis Steimer18.Assessment of QTc Interval Prolongation in a Phase I StudyUsing Monte Carlo SimulationPeter L.Bonate19.Optimizing a Bayesian Dose-Adjustment Scheme for a PediatricTrial:A Simulation StudyMarc R.Gastonguay,Ekaterina Gibiansky,Leonid Gibiansky,and Jeffrey S.BarrettCopyright 2003 Marcel Dekker,Inc.ContributorsLeon AaronsSchool of Pharmacy and Pharmaceutical Sciences,University ofManchester,Manchester,United KingdomJeffrey S.BarrettAventis Pharmaceuticals,Bridgewater,New Jersey,U.S.A.Peter L.Bonate*Quintiles TransnationalCorp.,KansasCity,Missouri,U.S.A.Stephen B.DuffullSchool of Pharmacy,University of Queensland,Brisbane,Queensland,AustraliaEne I.EtteVertex Pharmaceuticals,Inc.,Cambridge,Massachusetts,U.S.A.Ba rbel FottelerModeling and Simulation Team,Clinical Pharmacology andBiostatistics,Pharma Development,F.Hoffmann-La Roche,Basel,SwitzerlandPatrice FrancheteauSandoz Pharmaceuticals,Rueil-Malmaison,FranceMarc R.GastonguaySchool of Pharmacy,University of Connecticut,Farm-ington,Connecticut,U.S.A.*Current affilation:ILEX Oncology,San Antonio,Texas,U.S.A.Retired.Copyright 2003 Marcel Dekker,Inc.Ekaterina Gibiansky*GloboMax LLC,Hanover,Maryland,U.S.A.Leonid GibianskyThe Emmes Corporation,Rockville,Maryland,U.S.A.Ronald GieschkeModeling and Simulation Team,Clinical Pharmacology andBiostatistics,Pharma Development,F.Hoffmann-La Roche,Basel,SwitzerlandPascal GirardPharsight,Lyon,FranceChristopher J.GodfreyVertex Pharmaceuticals,Inc.,Cambridge,Massachu-setts,U.S.A.Timothy GogginModeling and Simulation Team,Clinical Pharmacology andBiostatistics,Pharma Development,F.Hoffmann-La Roche,Basel,SwitzerlandGordon GrahamCentre for Applied Pharmacokinetic Research,School ofPharmacy and Pharmaceutical Sciences,University of Manchester,Manchester,United KingdomMadeleine GuerretSandoz Pharmaceuticals,Rueil-Malmaison,FranceNicholas H.G.HolfordDivision of Pharmacology and Clinical Pharmacol-ogy,University of Auckland,Auckland,New ZealandMatthew M.HutmacherPharmacia Corp,Skokie,Illinois,U.S.A.Paul JordanModeling and Simulation Team,Clinical Pharmacology and Bio-statistics,Pharma Development,F.Hoffmann-La Roche,Basel,SwitzerlandMats O.KarlssonDivision of Pharmacokinetics a