Classical and Adaptive Clinical Trial Designs Using ExpDesign Studio E-Book

Mark Chang

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 339

Veröffentlichungsjahr: 2010

Beschreibung

ExpDesign Studio facilitates more efficient clinical trial design This book introduces pharmaceutical statisticians, scientists, researchers, and others to ExpDesign Studio software for classical and adaptive designs of clinical trials. It includes the Professional Version 5.0 of ExpDesign Studio software that frees pharmaceutical professionals to focus on drug development and related challenges while the software handles the essential calculations and computations. After a hands-on introduction to the software and an overview of clinical trial designs encompassing numerous variations, Classical and Adaptive Clinical Trial Designs Using ExpDesign Studio: * Covers both classical and adaptive clinical trial designs, monitoring, and analyses * Explains various classical and adaptive designs including groupsequential, sample-size reestimation, dropping-loser, biomarker-adaptive, and response-adaptive randomization designs * Includes instructions for over 100 design methods that have been implemented in ExpDesign Studio and step-by-step demos as well as real-world examples * Emphasizes applications, yet covers key mathematical formulations * Introduces readers to additional toolkits in ExpDesign Studio that help in designing, monitoring, and analyzing trials, such as the adaptive monitor, graphical calculator, the probability calculator, the confidence interval calculator, and more * Presents comprehensive technique notes for sample-size calculation methods, grouped by the number of arms, the trial endpoint, and the analysis basis Written with practitioners in mind, this is an ideal self-study guide for not only statisticians, but also scientists, researchers, and professionals in the pharmaceutical industry, contract research organizations (CROs), and regulatory bodies. It's also a go-to reference for biostatisticians, pharmacokinetic specialists, and principal investigators involved in clinical trials. ERRATUM Classical and Adaptive Clinical Trial Designs Using ExpDesign Studio By Mark Chang The license for the ExpDesign Studio software on the CD included with this book is good for one-year after installation of the software. Prior to the expiration of this period, the software will generate a reminder about renewal for the license. The user should contact CTriSoft International (the owners of ExpDesign Studio) at www.CTriSoft.net or by email at [email protected], about renewal for the license. This should have been made clear in the first printing of this book. We apologize for this error.

Leseprobe

Contents

Cover

Half Title page

Title page

Copyright page

Preface

Self-Study and Practice Guide

Chapter 1: ExpDesign Studio

1.1 Introduction

1.2 How to Design a Trial Using ExpDesign Studio

1.3 ExpDesign Menus

Chapter 2: Clinical Trial Design

2.1 Introduction

2.2 Classical Clinical Trial Design

2.3 Selection of a Trial Design

2.4 Adaptive Clinical Trial Design

Chapter 3: Classical Trial Design

3.1 Introduction

3.2 How to Calculate Sample Size Using ExpDesign

3.3 Mathematical Notes on Classical Design

Chapter 4: Group Sequential Trial Design

4.1 Introduction

4.2 Basics of Group Sequential Design

4.3 How to Design Sequential Trials Using ExpDesign

4.4 How to Monitor a Group Sequential Trial Using ExpDesign

4.5 Mathematical Notes on Sequential Trial Design

Chapter 5: Adaptive Trial Design

5.1 Introduction

5.2 Basics of Adaptive Design Methods

5.3 How to Design a Sample-Size Reestima, Tion Trial Using ExpDesign

5.4 How to Design a Drop-Loser Trial Using ExpDesign

5.5 How to Design a Trial Using a Classifier Biomarker

5.6 How to Design a Play-the-Winner Trial Using ExpDesign

Chapter 6: Adaptive Trial Monitoring

6.1 Introduction

6.2 Error-Spending Approach

6.3 How to Recalculate Stopping Boundaries Using ExpDesign

6.4 Conditional Power and the Futility Index

6.5 How to Reestimate Sample Size Using ExpDesign

6.6 Trial Examples

Chapter 7: Oncology Adaptive Trial Design

7.1 Multistage Trial Design

7.2 Dose-Escalation Trial Design

7.3 Dose-Escalation Trial Monitoring Using CRM

7.4 Mathematical Notes on Multistage Design

7.5 Mathematical Notes on The CRM

Chapter 8: Adaptive Trial Simulator

8.1 Adjusting The Critical Region Method

8.2 Classical Design With Two Parallel Treatment Groups

8.3 Flexible Design with Sample-Size Reestimation

8.4 Design with Random-Play-The-Winner Randomization

8.5 Group Sequential Design with One Interim Analysis

8.6 Design Permitting Early Stopping and Sample-Size Reestimation

8.7 Classical Design with Multiple Treatment Groups

8.8 Multigroup Trial with Response-Adaptive Randomization

8.9 Adaptive Design Featuring Dropping Losers

8.10 Dose–Response Trial Design

8.11 Dose-Escalation Design for an Oncology Trial

Chapter 9: Further Assistance From ExpDesign Studio

9.1 ExpDesign Probability Functions

9.2 Virtual Trial Data Generation Using ExpDesign Randomizor

9.3 ExpDesign Toolkits

Chapter 10: Classical Design Method Reference

10.1 Single-Group Design

10.2 Two-Group Design

10.3 Multigroup Trial Design

Appendix A: Validation of ExpDesign Studio

A.1 Validation Process for ExpDesign Studio

A.2 Validation of the Classical Design Module

A.3 Validation of the Group Sequential Design Module

A.4 Validation of the Adaptive Design Module

A.5 Validation of the Multistage Design Module

A.6 Validation of the Traditional Dose-Escalation Design Module

A.7 Validation of the Trial Simulation Module

A.8 Validation of the Randomizor

A.9 Validation of the ExpDesign Toolkits

A.10 Computer Programs for Validations

Appendix B: Sample-Size Calculation Methods: Classical Design

References

Index

System Requirements, Software Installation, and Software License Agreement

ERRATUM

Classical and Adaptive Clinical Trial Designs Using ExpDesign StudioBy Mark Chang

© 2008 John Wiley & Sons, Inc.    ISBN 978-0-470-27612-9

The license for the ExpDesign Studio software on the CD included with this book is good for one-year after installation of the software. Prior to the expiration of this period, the software will generate a reminder about renewal for the license. The user should contact CTriSoft International (the owners of ExpDesign Studio) at www.CTriSoft.net or by email at [email protected], about renewal for the license.

This should have been made clear in the first printing of this book. We apologize for this error.

CLASSICAL AND ADAPTIVE CLINICAL TRIAL DESIGNS USING EXPDESIGN STUDIO™

Copyright © 2008 by John Wiley & Sons, Inc. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, [phone], fax [phone], or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, [phone], fax [phone], or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and srategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at [phone], outside the United States at [phone] or fax [phone].

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data:

Chang, Mark.Classical and adaptive clinical trial designs using ExpDesign Studio™ / Mark Chang.p. ; cm.Includes bibliographical references and index.ISBN 978-0-470-27612-9 (cloth/cd)1. Drugs–Testing–Computer simulation. 2. Adaptive sampling (Statistics) 3. Clinical trials–Data processing. I. Title.[DNLM: 1. Drugs, Investigational. 2. Research Design. 3. Clinical Trials as Topic-methods. 4. Software. QV 771 C4565c 2008]RM301.27.C47 2008615’.190113—dc222008001358

PREFACE

Drug development is shifting from the classical approaches to more dynamic or adaptive approaches. The pharmaceutical industry and the U.S. Food and Drug Administration (FDA) has been seeking efficient methods of drug development as indicated in the FDA’s critical path document. Many people believe that the innovative approach of adaptive design is a major pathway to success in drug development in today’s challenging drug development environment.

In a book that I co-authored, Adaptive Design Methods in Clinical Trials (Chow and Chang, 2006), various adaptive design methods were introduced. Six months later I authored a second book, Adaptive Design Theory and Implementation Using SAS and R (Chang, 2007a), which provided in-depth and unified theory regarding adaptive designs and implementations, with many trial examples. These two books require a strong statistical background and clinical trial experience.

However, based on feedback from recent adaptive design workshops and conferences, I realize that there are many practitioners who are very good at strategic thinking and solution of practical problems but little interested in or lacking time to study the theory. Although I have kept the SAS and R program units as small as possible, with a clear logic flow from my previous books, there are still minimal requirements for knowledge of SAS or R. Also, many statisticians who are familiar with SAS would prefer to have software with a graphic user interface that can provide user-friendly tools for both classical and adaptive designs and monitoring. Among other options, I believe that ExpDesign Studio® fits the practical needs and provides a one-stop-shopping experience (CTriSoft, www.CTriSoft.net). This book, which avoids dealing with theory, is complementary to the two books mentioned earlier. Readers can jump-start to adaptive design without difficulty if they have one or two years of clinical trial design experience. However, for readers interested in the mathematical details, the mathematical notes at the end of each chapter will provide the key formulations for each method, or they can review Adaptive Design Theory and Implementation Using SAS and R (Chang, 2007a) for an in-depth understanding of the theory and algorithms for computer implementation.

ExpDesign is commercial software used by major pharmaceutical companies, universities, and research institutes worldwide. With ExpDesign you can design a classical or adaptive design literately in two minutes if you have the parameters ready. The ExpDesign enterprise version can also generate SAS and R code for an adaptive design.

The book has been written with practitioners in mind. It is not intended to teach adaptive design theory nor to function as a simple software user manual. The objective of the book is to demonstrate the use of ExpDesign in trial design, to assist strategic decision making, and to help solve issues related to classical and adaptive trials. It is written as a tutorial, a self-learning textbook (see the Self-Study and Practice Guide following the preface). Readers are expected to master the basic adaptive trial techniques in about one week. The book, together with the software, makes learning easy and fun. The accompanying software is a fully professional version of ExpDesign Studio 5.0, not a typical trial version. The book and the software, covering both classical and adaptive designs, can be used to leverage drug development in such a way that statisticians and other parties have more freedom and time to focus on the real issues, not the calculation or theory. The book is organized as follows:

In Chapter 1 we present an overview of the software ExpDesign Studio, provide a feeling for what it can do in trial designs, demonstrate simple design examples from classical, group sequential, adaptive, and other trials with ExpDesign Studio, and explain the basic operation of the software.

Chapter 2 provides an overview of a variety of clinical trial designs, their advantages and disadvantages, and when different classical and adaptive designs can be used.

Chapter 3 focuses on classical designs. After a discussion as to how sample size should be determined and on the variety of factors that affect the decision as to what sample size to use in a trial, examples are given on how to utilize ExpDesign to calculate sample size. Among nearly 150 sample-size calculation methods available in ExpDesign, the examples are carefully chosen to include a variety of designs, types of endpoints, and phases of clinical trials.

In Chapter 4 we discuss group sequential design (GSD), a commonly used and probably the simplest adaptive design. Starting with an overview of group sequential design, how to design and monitor a GSD trial using ExpDesign Studio is discussed. Finally, the key mathematic formulations for GSD are summarized for those interested in the mathematical details.

In Chapter 5 we discuss adaptive trial designs and introduce the stagewise test statistic and stopping rules. Interim analysis and trial monitor tools such as conditional power are described. We also discuss how to use ExpDesign Studio to design sample-size reestimation, drop-loser, biomarker-adaptive, response-adaptive randomization, and adaptive dose-finding trials. The mathematic formulations are summarized in the final section.

In Chapter 6 we discuss the specific design of early-phase oncology trials, because of its uniqueness. It includes multiple-stage single-arm design and dose-escalation design for maximum tolerated dose and show how to use ExpDesign to design oncology trials and how to compare and evaluate different designs based on their operating characteristics.

In Chapter 7 we focus on adaptive trial monitoring. The importance of trial monitoring and mathematic tools for monitoring is discussed, and how to use the trial monitor in ExpDesign to monitor an adaptive trial is described in detail using real-world examples.

In Chapter 8 we present a computer simulation approach in which the test statistic is the same as the classical design. The simulation module in ExpDesign allows for any combinations of the following adaptations: early futility and/or efficacy stopping, sample-size reestimation, drop-loser, and response-adaptive randomization based on the dose–response relationship. Step-by-step instructions are presented with trial examples.

In Chapter 9 we discuss how to get further assistance from ExpDesign. ExpDesign provides many toolkits for design, monitoring, and analysis of trials: the graphical calculator, which allows you to plot complicated mathematical expressions, the probability calculator for probability and percentile calculations, and the confidence interval calculator for exact confidence interval calculations. For advanced users, we also discuss how to use ExpDesign to generate univariate and multivariate data that can be used for various purposes of trial design, monitoring, and risk assessment.

In Chapter 10 we present notes on technique for nearly 100 methods for sample-size calculation, grouped by the number of arms, the trial endpoint, and the analysis basis. We describe the purpose of each method, information about the methods, such as when and how to use each one, the formula and/or references, and the assumptions or limitations of the methods.

Appendix A is about validation of ExpDesign. Several reviewers have indicated the importance of software validation and suggested including the validation information in the book. The validation document is also meant to support pharmaceutical end users to meet FDA 21 CFR part 11 requirements.

Installation instructions for the software CD and the license agreement appear at the end of the pages.

MARK CHANG

Lexington, MassachusettsWinter 2007www.Statisticians.org

SELF-STUDY AND PRACTICE GUIDE

Day 1

ExpDesign Studio 5.0 Installation (10 minutes)

Chapter 1: ExpDesign Studio (30 minutes of reading and practice)

Chapter 2: Clinical Trial Design (3 hours of reading)

Chapter 3: Classical Trial Design (4 hours of reading and practice)

Chapter 10: Classical Design Method Reference (15 minutes of reading)

Appendix A: Validation of ExpDesign Studio (15 minutes of reading)

Day 2

Chapter 4: Group Sequential Trial Design (8 hours of reading and practice) The classical group sequential design and simplest adaptive design are discussed. Make sure you understand the basic concepts of group sequential design, such as the notion of early stopping, error inflation due to multiple looks, different types of stopping boundaries, and different scales for stopping boundaries. Go through all the trial examples using ExpDesign; it helps you get “hands-on” experience. Trial monitoring requires your effort, which will give you the feeling of running an actual group sequential trial.

Days 3

Chapter 5: Adaptive Trial Design (8 hours of reading and practice)

You will learn various adaptive designs. Make sure that you understand the three commonly used statistical methods. Again, go through the trial practice using ExpDesign for hands-on experiences. The practices are straightforward and should take no more than 20 minutes each.

Day 4

Chapter 6: Adaptive Trial Monitoring (8 hours of reading and practice) Adaptive trial monitoring can be considered as the most challenging part of this book. It is about how you make actual adaptations for an ongoing trial based on the design without undermining the validity and integrity of the trial. Play around with the trial examples using ExpDesign, and spend extra time if needed.

Day 5

Chapter 7: Oncology Adaptive Trial Design (5 hours of reading and practice)

Chapter 8: Adaptive Trial Simulator (2 hours of optional reading and practice)

Chapter 9: Further Assistance with ExpDesign Studio (1 hour of reading and practice)

The mathematical notes in Chapters 3, 4, and 7 are not meant to be studied in your first reading; rather, they are for future reference. Similarly, Chapter 10 and Appendix A can be read as needed.

Chapter 1

ExpDesign Studio

1.1 INTRODUCTION

ExpDesign Studio (ExpDesign) is an integrated environment for designing experiments or clinical trials. It is a powerful and user-friendly statistical software product that has seven integrated main components: classical design (CD), sequential design (SD), multistage design (MSD), dose-escalation design (DED), adaptive design (AD), adaptive trial monitoring (ATM), and dose-escalation trial monitoring (DTM) modules. In addition, the ExpDesign randomizor can generate random variates from a variety of distributions. The ExpDesign toolkit provides features for distributional calculation, confidence intervals, and function and data plotting (Figure 1.1).

Figure 1.1 ExpDesign integrated environment.

Classical trials are the most commonly used in practice. ExpDesign provides nearly 150 methods for sample-size calculations in CD for different trial designs. It includes methods for single-, two-, and multiple-group designs, and for superiority, noninferiority, and equivalence designs with various endpoints. See the list of classical design methods in Appendix B.

Group sequential trials are advanced designs with multiple analyses. A group sequential trial is usually a cost-effective design compared to a classical design. SD covers a broad range of sequential trials with different endpoints and different types of stopping boundaries.

A multistage design is an exact method for group sequential trials with a binary response, whereas group sequential design uses an asymptotic approach. MSD provides three optimal designs among others: MinMax, MinExp, and MaxUtility, which minimize the maximum sample size, minimize the expected sample size, and maximize the utility index, respectively.