An Introduction to Statistics in early phase trial.pdf

An Introduction to Statistics inEarly Phase TrialsAn Introduction to Statistics in Early Phase Trials Steven A.Julious,Say Beng Tan and David Machin 2010 John Wiley&Sons Ltd ISBN:978-0-470-05985-2AN INTRODUCTIONTO STATISTICS INEARLY PHASETRIALSSTEVEN A.JULIOUSMedical Statistics Group,School of Health and Related Research,University of Sheffield,UKSAY BENG TANSingapore Clinical Research Institute and Duke-NUS Graduate Medical School,SingaporeDAVID MACHINUniversity of Leicester and University of Sheffield,UKAJohn Wiley&Sons,Ltd.,PublicationThis edition first published 2010?2010 John Wiley&Sons LtdWiley-Blackwell is an imprint of John Wiley&Sons,formed by the merger of Wileys global Scientific,Technical and Medicalbusiness with Blackwell Publishing.Registered officeJohn Wiley&Sons Ltd,The Atrium,Southern Gate,Chichester,West Sussex,PO19 8SQ,UKOther Editorial Offices9600 Garsington Road,Oxford,OX4 2DQ,UK111 River Street,Hoboken,NJ 07030-5774,USAFor details of our global editorial offices,for customer services and for information about how to apply for permission to reuse thecopyright material in this book please see our website at right of the author to be identified as the author of this work has been asserted in accordance with the Copyright,Designs andPatents Act 1988.All rights reserved.No part of this publication may be reproduced,stored in a retrieval system,or transmitted,in any form or by anymeans,electronic,mechanical,photocopying,recording or otherwise,except as permitted by the UK Copyright,Designs and PatentsAct 1988,without the prior permission of the publisher.Wiley also publishes its books in a variety of electronic formats.Some content that appears in print may not be available inelectronic books.Designations used by companies to distinguish their products are often claimed as trademarks.All brand names and product namesused in this book are trade names,service marks,trademarks or registered trademarks of their respective owners.The publisher isnot associated with any product or vendor mentioned in this book.This publication is designed to provide accurate and authoritativeinformation in regard to the subject matter covered.It is sold on the understanding that the publisher is not engaged in renderingprofessional services.If professional advice orother expert assistanceisrequired,the servicesofacompetentprofessionalshould besought.The contents of this work are intended to further general scientific research,understanding,and discussion only and are notintended and should not be relied upon as recommending or promoting a specific method,diagnosis,or treatment by physiciansfor any particular patient.The publisher and the author make no representations or warranties with respect to the accuracy orcompleteness of the contents of this work and specifically disclaim all warranties,including without limitation any impliedwarranties of fitness for a particular purpose.In view of ongoing research,equipment modifications,changes in governmentalregulations,and the constant flow of information relating to the use of medicines,equipment,and devices,the reader is urged toreview and evaluate the information provided in the package insert or instructions for each medicine,equipment,or device for,among other things,any changes in the instructions or indication of usage and for added warnings and precautions.Readersshould consult with a specialist where appropriate.The fact that an organization or Website is referred to in this work as a citationand/or a potential source of further information does not mean that the author or the publisher endorses the information theorganization or Website may provide or recommendations it may make.Further,readers should be aware that Internet Websiteslisted in this work may have changed or disappeared between when this work was written and when it is read.No warranty maybe created or extended by any promotional statements for this work.Neither the publisher nor the author shall be liable for anydamages arising herefrom.Library of Congress Cataloguing-in-Publication DataJulious,Steven A.An introduction to statistics in early phase trials/Steven Julious,David Machin,Say Beng Tan.p.;cm.Includes bibliographical references and index.ISBN 978-0-470-05985-21.DrugsTestingStatistical methods.2.Clinical trialsStatistical methods.I.Tan,Say Beng.II.Machin,David,1939III.Title.DNLM:1.Clinical Trials,Phase I as Topicmethods.2.Biometrymethods.3.Clinical Trials,Phase II as Topicmethods.4.Models,Statistical.5.Statistics as Topicmethods.WA 950 J94i 2010RM301.27.J85 20106150.19012dc222009033763ISBN:978-0-470-05985-2A catalogue record for this book is available from the British Library.Set in 10.5/12.5 Times by Integra Software Services Pvt.Ltd,Pondicherry,India.Printed in Singapore by Markono Print Media Pte LtdFirst Impression2010ContentsChapter 1Early phase trials1Chapter 2Introduction to pharmacokinetics13Chapter 3Sample size calculations for clinical trials37Chapter 4Crossover trial basics55Chapter 5Multi-period crossover trials71Chapter 6First time into man87Chapter 7Bayesian and frequentist methods113Chapter 8First-time-into-new-population studies125Chapter 9Bioequivalence studies139Chapter 10Other Phase I trials169Chapter 11Phase II trials:general issues187Chapter 12Doseresponse studies197Chapter 13Phase II trials with toxic therapies211Chapter 14Interpreting and applying early phase trial results223Chapter 15Go/No-Go criteria231Appendix245References251Index2571Early Phase Trials1.1INTRODUCTIONTrials conducted in the early phases of the molecule-to-marketplace clinical developmentparadigm take compounds from first time into man through to the start of the pivotalclinical trial programme.These early phases could be considered to be the learning andexplaining phases.They are learning as by definition there is no experience of thecompound in man when starting these studies and many important factors relevant to acompounds development will need to be quantified.They are explaining as early trialshelp to describe the properties of the compound,inputting to its rationale.Although it is only in late-phase development that definitive proof can be obtained,early phase studies are importantas they informsome of the most importantdecisions inaclinical programme,suchasthemostappropriatedosetocarryforward,and theposology.They also contribute to important factors such the inclusion/exclusion criteria for acompound with respect to late-phase protocols populations(or labelling).For the inclu-sion/exclusion criteria,without sufficient enabling studies these may be so tight as tomake recruitment rates impracticably slow.By definition in early drug development there is little information available whendesigning trials,and resource is often constrained both financially and in terms ofpopulations available to recruit.These factors can have a major impact on the designand conduct of the trials,with innovative and adaptive designs often being applied as away of overcoming the restrictions.However,in early phase trials,what is lacking inresource can,in part,be made up for by increased control,for example the trialiststhemselves control the rate of recruitment in a healthy volunteer study.In addition,early trials often are more tightly controlled with respect to more restricted populationsfrom smaller pools of specialist centres.This can positively benefit statistical variability,enabling smaller signals to be detected from these smaller studies.1.2DEFINITIONS OF THE PHASES OF EARLY DEVELOPMENTThe naive view of the phases of clinical development is that they follow a chronologicalordering of the form described in Figure 1.1a.In this ordering the distinct phases are likebatons in a relay race.Initially a compound is picked up in a preclinical setting byAn Introduction to Statistics in Early Phase Trials Steven A.Julious,Say Beng Tan and David Machin 2010 John Wiley&Sons Ltd ISBN:978-0-470-05985-2researchers,where workis undertaken until a point is reached when it can be handed on toPhase I.Phase I researchers then take on the baton and run with it until they can hand it onto Phase II and so on.The ordering of the phases is better described in Figure 1.1b.Here what actuallyhappens in clinical development is that the minimum body of work is done in eachphase before the compound progresses to the next phase.The consequence is that muchof what would be considered early phase trials actually takes place when a compound islate in development.This minimum body of work is usually referred to as critical-pathactivities.Hence,where Figure1.1a can be considered to be accurate is that it figurativelydescribes these critical-path studies.The International Conference on Harmonisation of Technical Requirements forRegistration of Pharmaceuticals for Human Use(ICH)Topic E8 gives a detaileddescription of the phases of development(ICH,1998a),while the Food and DrugAdministration(FDA,2006a)defines the critical path and what we learn along theway as follows:At the start of the Critical Path,developers form hypotheses about performance characteristicssuchassafety,biologicalormechanicalaction,andbiocompatibility.Theythenseektoevaluateand confirm these hypotheses using in vitro,animal,and human testing.Once uncertainty aboutbenefits and risks of a product has been reduced to an acceptable level,the product may beapproved for marketingif the benefits outweigh the risks.The great challenge in developmentlies in predicting a potential products performance as early as possible with the greatest degreeof certainty.With this definition,in terms of studies necessary for filing registration,allstudies fall along the critical path.However,it can be argued that critical-pathstudies are just those that drive the development timelines and must be done prior tothe start of subsequent activities.Studies that can be done in parallel with otheractivities,that is,that must be done but not before some pre-prescribed critical pathtask,are termed non-critical path.As an aside,in this context given that muchpreclinical work happens when a compound is actually in clinical development thereis an argument that phase 0 as opposed to preclinical may be a more accuratenomenclature.(a)(b)PreclinicalPhase IPhase IIPhase IIIPhase IVPhase IPreclinicalPhase IIPhase IIIPhase IVFigure 1.1Perception of the phases of drug development.2EARLY PHASE TRIALSFor a given compound,therefore,activities in a clinical plan may look like Figure 1.2,moving between learning and confirming,and where critical-path activities preclinicallymaybetriggeredbyactivitiesinPhaseIandviceversa.Acompoundmayalsomovealongthe different phases according to different indications or populations for which it may bebeing targeted.Due to the wide variety of work undertaken in early phase development,genericdefinition of the early phases is not really possible.Simply it can be said that Phase Iis performed in healthy volunteers and is the phase where safety and tolerability areassessed,while Phase II is undertaken in patients and is the phase where the firstassessment of efficacy is made.However,things are not as simple as these defini-tions imply.In therapeutic areas such as oncology,Phase I may be undertaken inpatients,while for certain areas,the size and scale of Phase II may resemble PhaseIII for other areas.In truth what can be stated is that a compound must go first time into man(FTIM),andthat some time after this study the pivotal Phase III programme must start.From the pointof the start of FTIM to the start of the pivotal Phase III programme we can define as earlyphase development(in addition of course to the early phase work being undertaken inparallel to Phase III).Subsequent chapters will detail the different types of early phase trials.1.3CLINICAL DEVELOPMENT PLANSWhen taking a compound from molecule to marketplace it could be argued that the vastmajority of studies would fall under the banner of early phase.This greater number ofstudies equates toagreater variety oftypes of studiesthat canbe undertaken,fromroutinebleed them and feed them type studies for example studies to assess drug or foodinteractions through to innovative designs.The trials themselves are undertaken in aPreclinicalLearningConfirmingPhase IPhase IIPhase IIIPhase IVFigure 1.2Interweaving of the phases of drug development.CLINICAL DEVELOPMENT PLANS3number of populations.Initially these trial populations will most likely be of healthyvolunteersbeforemovingintopatientpopulationsandmaybesubpopulationswithinthesepatient populations.To coordinate all these activities a clinical development plan(CDP)needs to bedrafted.This document will be needed to help coordinate the activities,critical path andnon-critical path,required for development.The focus initially will be on the critical-path activities,as defined in Table 1.1,as these will determine timelines,whilst non-critical-path activities could be completed in parallel.Examples of critical-path studies are the first-time-in-man study(which obviouslymust be done),maybe followed by the repeat-dose study for a chronic intervention.Obviously a compound must be shown to be safe and tolerable in a single-dose studybefore it can be given in a repeat-dose study.Other studies may follow in sequentialorder to these.Examples of non-critical-path activities could be studies to investigate possible phar-macokinetic interaction or pharmacodynamic interactions.These may not fall on thecritical path but would need to be done prior to the start of the pivotal programme,aswithout them the protocol inclusion criteria would be affected and as a consequence therate of recruitment.Figure 1.3 gives an illustration of how a clinical development planmay be summarized.Table 1.1Definition of critical-path and non-critical-path activitiesCritical pathNon-critical pathStudies that must be undertaken prior to the start ofsubsequent activitiesStudies that need to be done but not beforesome pre-prescribed critical path taskDrive the development timelinesCan be done in parallelExamples:first time into man;repeat dosestudyExamples:pharmacokinetic-orpharmacodynamic-interaction studiesPhaseIIIFigure 1.3Typical clinical development plan from early stages to start of Phase III.4EARLY PHASE TRIALS1.4PHARMACOKINETICSAmainaimofmanyearlyphasestudiesistoassessthepharmacokinetic(PK)activityofagiven compound.This is often referred to as analysing the effect of the body on the drug.This is achieved by the derivation of a concentrationtime profile for each individualgiven the compound.Figure 1.4 gives an illustrative example of a hypothetical extra-vascular pharmacokinetic profile(to be discussed in more detail in Chapter 2).In truth an analysis of the full pharmacokinetic profile is not usually undertaken butinstead appropriate summary statistics are used to assess the pharmacokinetics.Thisassessment is usually performed by determining the extent and rate of absorption,usingthe AUC(area under the concentration curve)to assess the exte