Clinical Trial Optimization Using R 1st Edition

Visit the link below to download the full version of this book:

https://medipdf.com/product/clinical-trial-optimization-using-r-1st-edition/

Click Download Now

Published Titles

Noninferiority Testing in Clinical Trials: Statistical Design and Analysis of Clinical

Issues and Challenges Trials: Principles and Methods
Tie-Hua Ng Weichung Joe Shih and Joseph Aisner
Optimal Design for Nonlinear Response Statistical Design and Analysis of
Models Stability Studies
Valerii V. Fedorov and Sergei L. Leonov Shein-Chung Chow
Patient-Reported Outcomes: Statistical Evaluation of Diagnostic
Measurement, Implementation and Performance: Topics in ROC Analysis
Interpretation Kelly H. Zou, Aiyi Liu, Andriy Bandos,
Joseph C. Cappelleri, Kelly H. Zou, Lucila Ohno-Machado, and Howard Rockette
Andrew G. Bushmakin, Jose Ma. J. Alvir, Statistical Methods for Clinical Trials
Demissie Alemayehu, and Tara Symonds Mark X. Norleans
Quantitative Evaluation of Safety in Drug Statistical Methods for Drug Safety
Development: Design, Analysis and Robert D. Gibbons and Anup K. Amatya
Reporting
Statistical Methods for Healthcare
Qi Jiang and H. Amy Xia
Performance Monitoring
Quantitative Methods for Traditional Alex Bottle and Paul Aylin
Chinese Medicine Development
Statistical Methods for Immunogenicity
Shein-Chung Chow
Assessment
Randomized Clinical Trials of Harry Yang, Jianchun Zhang, Binbing Yu,
Nonpharmacological Treatments and Wei Zhao
Isabelle Boutron, Philippe Ravaud,
Statistical Methods in Drug Combination
and David Moher
Studies
Randomized Phase II Cancer Wei Zhao and Harry Yang
Clinical Trials
Statistical Testing Strategies in the
Sin-Ho Jung
Health Sciences
Repeated Measures Design with Albert Vexler, Alan D. Hutson,
Generalized Linear Mixed Models for and Xiwei Chen
Randomized Controlled Trials
Statistics in Drug Research:
Toshiro Tango
Methodologies and Recent
Sample Size Calculations for Clustered Developments
and Longitudinal Outcomes in Clinical Shein-Chung Chow and Jun Shao
Research
Statistics in the Pharmaceutical Industry,
Chul Ahn, Moonseong Heo,
Third Edition
and Song Zhang
Ralph Buncher and Jia-Yeong Tsay
Sample Size Calculations in Clinical
Survival Analysis in Medicine and
Research, Second Edition
Genetics
Shein-Chung Chow, Jun Shao,
Jialiang Li and Shuangge Ma
and Hansheng Wang
Theory of Drug Development
Statistical Analysis of Human Growth
Eric B. Holmgren
and Development
Yin Bun Cheung Translational Medicine: Strategies and
Statistical Methods
Dennis Cosmatos and Shein-Chung Chow
Clinical Trial
Optimization
Using R

Edited by
Alex Dmitrienko
Erik Pulkstenis

Harry Yang
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 2017 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Printed on acid-free paper

Version Date: 20170504

International Standard Book Number-13: 978-1-4987-3507-0 (Hardback)

This book contains information obtained from authentic and highly regarded sources. Reasonable
efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication
and apologize to copyright holders if permission to publish in this form has not been obtained. If any
copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information
storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access
www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc.
(CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization
that provides licenses and registration for a variety of users. For organizations that have been granted
a photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and
are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging‑in‑Publication Data

Names: Dmitrienko, Alex. | Pulkstenis, Erik.

Title: Clinical trial optimization using R / Alex Dmitrienko, Erik Pulkstenis.
Description: Boca Raton : CRC Press, 2017. | Includes bibliographical
references and index.
Identifiers: LCCN 2016057709 | ISBN 9781498735070 (hardback : alk. paper)
Subjects: LCSH: Clinical trials--Statistical methods. | R (Computer program language)
Classification: LCC R853.C55 D6525 2017 | DDC 610.72/7--dc23
LC record available at https://lccn.loc.gov/2016057709

Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Contents

Preface xiii

List of Contributors xvii

1 Clinical Scenario Evaluation and Clinical Trial Optimization 1

Alex Dmitrienko and Gautier Paux
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Clinical Scenario Evaluation . . . . . . . . . . . . . . . . . . 2
1.2.1 Components of Clinical Scenario Evaluation . . . . . . 2
1.2.2 Software implementation . . . . . . . . . . . . . . . . 4
1.2.3 Case study 1.1: Clinical trial with a normally distributed
endpoint . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.2.4 Case study 1.2: Clinical trial with two time-to-event
endpoints . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.3 Clinical trial optimization . . . . . . . . . . . . . . . . . . . . 28
1.3.1 Optimization strategies . . . . . . . . . . . . . . . . . 30
1.3.2 Optimization algorithm . . . . . . . . . . . . . . . . . 33
1.3.3 Sensitivity assessments . . . . . . . . . . . . . . . . . . 34
1.4 Direct optimization . . . . . . . . . . . . . . . . . . . . . . . 38
1.4.1 Case study 1.3: Clinical trial with two patient popula-
tions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.4.2 Qualitative sensitivity assessment . . . . . . . . . . . . 43
1.4.3 Quantitative sensitivity assessment . . . . . . . . . . . 44
1.4.4 Optimal selection of the target parameter . . . . . . . 53
1.5 Tradeoff-based optimization . . . . . . . . . . . . . . . . . . 59
1.5.1 Case study 1.4: Clinical trial with an adaptive design 59
1.5.2 Optimal selection of the target parameter . . . . . . . 67

2 Clinical Trials with Multiple Objectives 71

Alex Dmitrienko and Gautier Paux
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.2 Clinical Scenario Evaluation framework . . . . . . . . . . . . 73
2.2.1 Data models . . . . . . . . . . . . . . . . . . . . . . . 74
2.2.2 Analysis models . . . . . . . . . . . . . . . . . . . . . 74
2.2.3 Evaluation models . . . . . . . . . . . . . . . . . . . . 85
2.3 Case study 2.1: Optimal selection of a multiplicity adjustment 91

ix
x Contents

2.3.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 92

2.3.2 Qualitative sensitivity assessment . . . . . . . . . . . . 99
2.3.3 Quantitative sensitivity assessment . . . . . . . . . . . 107
2.3.4 Software implementation . . . . . . . . . . . . . . . . 114
2.3.5 Conclusions and extensions . . . . . . . . . . . . . . . 120
2.4 Case study 2.2: Direct selection of optimal procedure parame-
ters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
2.4.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 121
2.4.2 Optimal selection of the target parameter in
Procedure B1 . . . . . . . . . . . . . . . . . . . . . . . 133
2.4.3 Optimal selection of the target parameters in Proce-
dure B2 . . . . . . . . . . . . . . . . . . . . . . . . . . 140
2.4.4 Sensitivity assessments . . . . . . . . . . . . . . . . . . 145
2.4.5 Software implementation . . . . . . . . . . . . . . . . 149
2.4.6 Conclusions and extensions . . . . . . . . . . . . . . . 154
2.5 Case study 2.3: Tradeoff-based selection of optimal procedure
parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
2.5.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 156
2.5.2 Optimal selection of the target parameter in
Procedure H . . . . . . . . . . . . . . . . . . . . . . . 161
2.5.3 Software implementation . . . . . . . . . . . . . . . . 168
2.5.4 Conclusions and extensions . . . . . . . . . . . . . . . 172

3 Subgroup Analysis in Clinical Trials 173

Alex Dmitrienko and Gautier Paux
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
3.2 Clinical Scenario Evaluation in confirmatory subgroup analysis 175
3.2.1 Clinical Scenario Evaluation framework . . . . . . . . 175
3.2.2 Multiplicity adjustments . . . . . . . . . . . . . . . . . 181
3.2.3 Decision-making framework . . . . . . . . . . . . . . . 184
3.3 Case study 3.1: Optimal selection of a multiplicity adjustment 188
3.3.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 189
3.3.2 Direct optimization based on disjunctive power . . . . 194
3.3.3 Direct optimization based on weighted power . . . . . 196
3.3.4 Qualitative sensitivity assessment . . . . . . . . . . . . 199
3.3.5 Quantitative sensitivity assessment . . . . . . . . . . . 203
3.3.6 Software implementation . . . . . . . . . . . . . . . . 208
3.3.7 Conclusions and extensions . . . . . . . . . . . . . . . 213
3.4 Case study 3.2: Optimal selection of decision rules to support
two potential claims . . . . . . . . . . . . . . . . . . . . . . . 215
3.4.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 215
3.4.2 Influence condition . . . . . . . . . . . . . . . . . . . . 215
3.4.3 Optimal selection of the influence threshold . . . . . . 219
3.4.4 Software implementation . . . . . . . . . . . . . . . . 225
3.4.5 Conclusions and extensions . . . . . . . . . . . . . . . 228
Contents xi

3.5 Case study 3.3: Optimal selection of decision rules to support

three potential claims . . . . . . . . . . . . . . . . . . . . . . 228
3.5.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 228
3.5.2 Interaction condition . . . . . . . . . . . . . . . . . . . 233
3.5.3 Optimal selection of the influence and interaction thresh-
olds . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
3.5.4 Software implementation . . . . . . . . . . . . . . . . 244
3.5.5 Conclusions and extensions . . . . . . . . . . . . . . . 249

4 Decision Making in Clinical Development 251

Kaushik Patra, Ming-Dauh Wang, Jianliang Zhang, Aaron Dane, Paul
Metcalfe, Paul Frewer, and Erik Pulkstenis
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
4.2 Clinical Scenario Evaluation in Go/No-Go decision making and
determination of probability of success . . . . . . . . . . . . 253
4.2.1 Clinical Scenario Evaluation approach . . . . . . . . . 253
4.2.2 Go/No-Go decision criteria . . . . . . . . . . . . . . . 255
4.2.3 Probability of success . . . . . . . . . . . . . . . . . . 258
4.2.4 Probability of success applications . . . . . . . . . . . 262
4.3 Motivating example . . . . . . . . . . . . . . . . . . . . . . . 264
4.3.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 265
4.3.2 Software implementation . . . . . . . . . . . . . . . . 268
4.4 Case study 4.1: Bayesian Go/No-Go decision criteria . . . . 269
4.4.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 269
4.4.2 General sensitivity assessments . . . . . . . . . . . . . 271
4.4.3 Bayesian Go/No-Go evaluation using informative priors 274
4.4.4 Sample size considerations . . . . . . . . . . . . . . . . 276
4.4.5 Software implementation . . . . . . . . . . . . . . . . 279
4.4.6 Conclusions and extensions . . . . . . . . . . . . . . . 282
4.5 Case study 4.2: Bayesian Go/No-Go evaluation using an alter-
native decision criterion . . . . . . . . . . . . . . . . . . . . . 283
4.5.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 283
4.5.2 Software implementation . . . . . . . . . . . . . . . . 285
4.5.3 Conclusions and extensions . . . . . . . . . . . . . . . 286
4.6 Case study 4.3: Bayesian Go/No-Go evaluation in a trial with
an interim analysis . . . . . . . . . . . . . . . . . . . . . . . . 286
4.6.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 287
4.6.2 Software implementation . . . . . . . . . . . . . . . . 289
4.6.3 Conclusions and extensions . . . . . . . . . . . . . . . 290
4.7 Case study 4.4: Decision criteria in Phase II trials based on
Probability of Success . . . . . . . . . . . . . . . . . . . . . . 290
4.7.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 290
4.7.2 Software implementation . . . . . . . . . . . . . . . . 292
4.7.3 Conclusions and extensions . . . . . . . . . . . . . . . 293
xii Contents

4.8 Case study 4.5: Updating POS using interim or external infor-
mation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
4.8.1 Clinical trial . . . . . . . . . . . . . . . . . . . . . . . 295
4.8.2 Software implementation . . . . . . . . . . . . . . . . 298

Bibliography 301

Index 309
Preface

Clinical development is a strategic and operational, scientific process that

involves a significant level of multidimensional decision making, often in the
presence of incomplete information or considerable uncertainty. Along the
way, several key questions must be answered including selection of the right
dose and schedule, selection of the right patient population and selection of the
right analytical strategy to provide optimal treatment options for patients with
the highest probability of success. Against this backdrop, more fundamental
questions loom, e.g., whether the drug is even safe or effective at all and
whether the final risk-benefit profile will be acceptable.
Adding to the research burden is the desire to make correct decisions as
quickly and efficiently as possible. The rise in adaptive design science over
the past decade speaks to the increasing desire to learn in real time, and im-
mediately incorporate that learning into drug development, in an attempt to
eliminate or reduce the white space present in a traditional drug development
paradigm. Advances in multiplicity theory support the growing pressure to
definitively establish as many hypotheses/claims as possible within the con-
straints of Type I error rate control. At the same time clinical trial programs
are changing with the lines becoming increasingly blurred between traditional
phases of drug development, or individual trials that are increasing in com-
plexity to ask more questions (e.g., platform trials). External pressures also
exist, driven by the changes in reimbursement and increased expectations that
medications demonstrate their value to the patient and clinical community in
a more comprehensive way. The goal is simply no longer to conduct two large
Phase III trials with statistically significant p-values because the bar has been
raised.
Despite significant methodological advances provided by the statistical and
clinical community and associated computational advances, it remains clear
that development risk remains high, and that decision making has not yet been
universally optimized within clinical development. This can be seen by the
accumulating evidence that Phase II results often do not translate to Phase III
success (Brutti, De Santis, and Gubbiotti, 2008; Kirby et al., 2012; Kola and
Landis, 2004; Chuang-Stein and Kirby, 2014). One review of 253 published
Phase III oncology trials found that 62% of the trials failed to establish a
statistically significant treatment effect, mostly due to the fact that observed
treatment effects were simply smaller than expected based on previous data
(Gan et al., 2012).
This combination of increasing clinical trial complexity/options, coupled

xiii
xiv Preface

with increasing scrutiny on the risk-benefit profile that new treatments bring
to the table, against a backdrop of lackluster industry development perfor-
mance introduces both financial constraints, and a heightened level of urgency
around strategic drug development decision making in order to stop develop-
ment of inferior compounds early and accelerate development of promising
compounds (Arrowsmith and Miller, 2013; Paul et al., 2010). As a result,
clinical trial optimization is an absolute necessity in support of these objec-
tives, though quantitative methods to this end are frequently not part of the
drug development process.
Clinical trial optimization can be thought of at the trial level, or the de-
velopment plan level as clinical research designed to most efficiently and with
least risk answer the most important research questions to the developer. His-
torically the process has been fairly empiric, though the increasing availability
of computational resources and methods, along with the generally low success
rates, is driving an opportunity to marry clinical trial modeling and simulation
with decision making in a more comprehensive and holistic fashion, resulting
in evidence-based development which examines the operating characteristics
of development decisions themselves. In this book, we explore a promising
approach known as the Clinical Scenario Evaluation framework (Benda et al.,
2010) which endeavors to optimize clinical development considering a set of
objectives, design and analysis alternatives, underlying assumptions and, fi-
nally, quantitative metrics to facilitate decision making with better line of
sight into the decision space one is dealing with. We use specific common
clinical trial problems to elucidate the methodology in a case study setting. R
code is provided to both demonstrate common methods while providing some
preliminary tools for the practitioner. Examples include optimally spending
the Type I error rate across trial objectives or patient subgroups in the pres-
ence of multiple desired claims and optimally selecting associated decision
rules or analytical methods in Chapters 2 and 3. In addition, we present an
evaluation of Go/No-Go decision making at the proof-of-concept stage as well
as considerations for probability of success based on Bayesian principles in
Chapter 4. These case studies serve to scratch the surface regarding the po-
tential utility of modeling and simulation to optimize decision making within
a complex and highly dimensional development decision space.
The Clinical Scenario Evaluation paradigm is broadly flexible and applica-
ble to any scenario a clinical trial or researcher can envision and, as a result,
may impact the overall quality of the drug development process. It is a valu-
able tool available to the researcher and enables a move from empirical decision
making around myriads of options, to a more disciplined and evidence-based
approach to how one designs clinical trials and clinical trial programs.
Preface xv

Acknowledgments
We would like to thank the reviewers who have provided valuable comments
on selected chapters in the book: Thomas Brechenmacher (Novartis), Michael
Lee (Johnson and Johnson), Christoph Muysers (Bayer), Xin Wang (AbbVie).
We would also like to thank the book’s acquisitions editor, David Grubbs,
for his support and his work on this book publishing project.

Alex Dmitrienko, Mediana Inc.

Erik Pulkstenis, MedImmune.
List of Contributors

This book is based on a collaborative effort of nine statisticians from the

pharmaceutical industry:
Aaron Dane, Therapeutic Area Head for Infection, Biometrics and Infor-
mation Sciences, AstraZeneca; Independent Consultant, DaneStat Consulting
Limited.
Alex Dmitrienko, President, Mediana Inc.
Paul Frewer, Statistical Science Director, Biometrics and Information Sci-
ences, AstraZeneca.
Paul Metcalfe, Scientific Computing Director, Advanced Analytics Centre,
Biometrics and Information Sciences, AstraZeneca.
Kaushik Patra, Science Director, Biostatistics, MedImmune; Director, Bio-
statistics, Alexion.
Gautier Paux, Senior Statistician, Clinical Biostatistics Unit, Institut de
Recherches Internationales Servier.
Erik Pulkstenis, Vice President, Clinical Biostatistics and Data Manage-
ment, MedImmune.
Ming-Dauh Wang, Principal Research Scientist, Global Statistical Sciences,
Eli Lilly and Company.
Jianliang Zhang, Senior Director, Statistical Methods, MedImmune.

xvii