Market Report Source

Overview

Report overview Key findings
The costs associated with developing drugs have risen dramatically over the past decade and fewer drugs are obtaining regulatory approval. The pharmaceutical industry is continually exploring new ways of improving drug developments and one area of focus is adaptive clinical trial designs. These innovative clinical trial designs use accumulating data to guide potential modifications to the study as it progresses, without undermining the validity and integrity of the trial. The advantages of such designs include the reduced length and cost of clinical trials, lower patient numbers and the ability to stop a trial early where a drug has not shown efficacy.

Innovative Clinical Trial Design and Management is a new report published by Business Insights that explores the major types of adaptive design and their role in dose-finding. The report investigates seamless Phase 2/3 trials and adaptive trials in pharmacogenomics, assesses the logistical implications of adaptive trial implementations and reviews the current regulatory standpoints of the FDA and EMEA. Detailed case studies of recent adaptive clinical trials are provided and the companies offering statistical expertise in this area are profiled. This report also includes a breakdown of the potential cost and time savings that innovative trial designs can offer throughout the clinical development process.

Use detailed case studies to explore recent adaptive trial implementations, identify the companies pioneering and supporting innovative designs and understand the most effective planning and logistics strategies.

Key findings

• Major pharma companies are implementing adaptive trials to improve dose-finding in the Phase 2 setting. The use of adaptive clinical trials will increase across the industry over the next 2-3 years.
• Adaptive clinical trial designs are more effective than traditional designs in cases where there is uncertainty surrounding the dose, effect size and variability, clinical endpoint or patient populations.
• The planning and execution of adaptive designs is more complex than the traditional approach. Successful implementations require teams of statisticians, data managers, clinicians and drug supply and logistics managers to work together as early as possible.
• Predictive biomarkers have been found to require detailed prospective analysis far earlier in the clinical development process, and with the same clarity as traditional drug approvals. Post-hoc correlations were previously thought to be good enough for identifying the biomarkers used to predict the patients most likely to respond well to a new treatment.
• Regulatory authorities are supportive of adaptive trials, particularly in the Phase 2 setting. However, there are concerns over the confidentiality of data and companies have been asked to demonstrate that the parties involved in running the study will remain unaware of ongoing adaptations.

Key questions answered

• How can adaptive trials improve the success rates of clinical drug projects?
• How are pharma companies implementing adaptive trials and what major hurdles can prevent such implementations?
• What is the position of the FDA and EMEA in regards to different types of adaptive trial?
• How can logistical and strategic planning be managed most effectively?Which companies are offering services to support adaptive clinical trials?
• Which companies are co-developing drugs and diagnostic products?

Key issues examined by this report

• Adaptive trial implementations. The aim of adaptive trials is to improve the information value of clinical trials whilst maintaining their integrity and validity. The use of adaptive trials in the early phases of drug development should yield better information and lead to the earlier termination of unsuccessful compounds.
• Dose finding improvements. The availability of new Bayesian study designs that acknowledge prior information and allow for the testing of a wider range of doses has enabled more accurate dose-finding. This may have important consequences for the success of future Phase 3 clinical programs
• Seamless trial speed. Major pharma companies are interested in the prospect of combining drug development phases into ‘seamless trials’, with the potential to reduce the length of clinical development programs in the Phase 2b/3 setting
• Regulatory issues. Engaging with the FDA/EMEA during the protocol design stages of an adaptive trial is important, especially for studies intended for use in packages of pivotal clinical data. The EMEA’s current position on adaptive clinical trial design is summarized within a reflection paper published in October 2007, while draft FDA guidance is expected in 2008.

Table of Contents

Executive Summary 10
Introduction 10
Adaptive clinical trials 11
Adaptive dose-finding studies 12
Seamless adaptive trials 13
Adaptive trials in drug-diagnostic co-development 14
Simulation, logistics and technology in adaptive trials 15
Regulating adaptive trials 16
Outlook for adaptive trials in the pharma industry 17

Chapter 1 Introduction 20
Summary 20
State of the industry 21
Rising development costs 22
Longer development times 23
Increasing complexity 24
Higher attrition 25
Increasing innovation 26
Innovation in drug discovery and development 28
Adaptive clinical trials – an introduction 29
Report outline 32

Chapter 2 Adaptive clinical trials 36
Summary 36
Introduction 37
Adaptive designs 37
Flexibility in adaptive trials 40
Statistics for adaptive designs 41
Group Sequential Designs 43
Stopping early for benefit – controversy 44
Case study: RALES – a Searle Phase 3 Group Sequential Design 45
Case study: CAPTURE – a Centocor study in refractory unstable angina 47
Case study: Proof-of-concept study in neuropathic pain – GlaxoSmithKline 47
Case study: PURSUIT – a Millennium Pharmaceuticals study in unstable angina 48
Sample size re-estimation 48
Case study: a group sequential study for sample size re-estimation 51
Case study: GlaxoSmithKline pivotal study for Advair Diskus 52
Response adaptive randomization 52
Case study: Troxacitabine in acute myeloid leukemia 54
Case study: Eli Lilly out-patient study in depression 55
Case study: Three dosing schedules of decitabine in myelodysplastic
syndrome 56
Case studies: A study of sorafenib in different types of cancer 57
Bayesian methods in First in Man studies 58
Conclusions 58

Chapter 3 Adaptive dose-finding studies 62
Summary 62
Introduction 63
Case study: ASTIN 65
Case study: Phase 2 Pfizer dose-ranging study in neuropathic pain 68
Case study: Phase 2 Merck dose-finding study of anti-migraine compound 68
Case study: Phase 3 Napo Pharmaceuticals study in HIV patients with diarrhea 69
The PhRMA Adaptive Dose-Ranging Studies Working Group 69
Conclusions 71

Chapter 4 Seamless adaptive trials 74
Summary 74
Introduction 75
Pros, cons and controversy 77
Speeding up clinical development 78
Efficiency gains in seamless trials 79
When are seamless designs appropriate? 80
Regulatory view of seamless Phase 2b/3 studies 81
Case study: Phase 1/2 trials in oncology 82
Case study: Phase 2b/3 Novartis study in a chronic disease indication 83
Case study: HORIZON III an AstraZeneca study in oncology 86
Conclusions 86

Chapter 5 Adaptive trials in drug-diagnostic co-development projects 90
Summary 90
Introduction 91
Biomarkers and classifiers 93
Co-development of drugs and companion diagnostics 94
Pivotal Phase 3 trial designs using predictive biomarkers 95
Enrichment designs 97
Patient stratification 98
Sample size calculation for enrichment and stratification designs 101
Adaptive signature designs 102
Biomarker adaptive threshold determination 104
Case study: identifying a classifier of response to Velcade 105
Conclusions 106

Chapter 6 Simulation, logistics and technology in adaptive trials 110
Summary 110
Introduction 111
Preplanning and simulation 112
Case study: A study design and simulation using Cytel’s East® simulation package 113
Logistical issues 117
Real-time data collection 118
Electronic data capture (EDC) 119
Electronic patient reported outcomes (ePRO) 119
Interactive voice recognition and interactive web response 120
Companies involved in real-time data collection 120
Maintaining data confidentiality 122
Minimizing operational bias and assuring consistency between study stages 124
Clinical supply management 126
Case study: simulating drug supply 128
Summary – logistical considerations for adaptive trials 130
Conclusions 131

Chapter 7 Regulating adaptive trials 134
Summary 134
Introduction 135
FDA guidance on adaptive designs 136
EMEA guidance on adaptive designs 136
PhRMA working groups 137
Conclusions 138

Chapter 8 Outlook for adaptive trials in the pharma industry 140
Summary 140
Introduction 141
Adaptive trials and success rates in clinical drug development 141
Costs in adaptive clinical trials 145
Time means money 145
Costs in a dose-finding clinical trial 146
Cost savings in a seamless Phase 2b/3 trial 147
Cost savings in a random adaptive allocation trial 148
Costs in a sample size re-estimation study 149
Costs of clinical trials in pharmacogenomics 150
Pharma’s uptake of adaptive designs 151
Outlook for adaptive trials 154

Chapter 9 Appendix 156
Primary research methodology 156
Index 156
Glossary 158
Glossary 158

List of Figures
Figure 1.1: New drug approvals versus R&D costs, 1995-2006 23
Figure 1.2: Trend towards longer drug development times, 1997 to 2005 24
Figure 1.3: Increasing complexity of clinical trial 25
Figure 1.4: Change in attrition rates, 1994 to 2000 26
Figure 1.5: Increasing innovation in drug development 27
Figure 1.6: Attrition rates and development times by drug novelty status 28
Figure 2.1: Summary of motivations for using adaptive trials 38
Figure 2.2: Examples of motivations for adapting clinical trials 39
Figure 2.3: Advantages and disadvantages of increasing the flexibility of adaptive trial designs 41
Figure 2.4: Bayes Theorem explained 42
Figure 2.5: How Bayesian statistics work 43
Figure 2.6: Data from interim monitoring visits during the RALES study 47
Figure 2.7: The random play-the-winner trial design 53
Figure 2.8: Response adaptive randomization – case study with troxacitabine 55
Figure 3.1: Advantages of adaptive dose-finding studies over traditional methods 64
Figure 3.2: Structure of the ASTIN study 66
Figure 3.3: Data from the ASTIN study 67
Figure 4.1: Structure of an adaptive seamless clinical trial 75
Figure 4.2: Key aims of each phase of drug development 76
Figure 4.3: Seamless trials in Phases 1, 2 and 3 77
Figure 4.4: When to use seamless adaptive Phase 2b/3 designs 81
Figure 4.5: Novartis’ ongoing seamless adaptive Phase 2b/3 study design 84
Figure 5.1: Examples of targeted treatments and their companion diagnostics 92
Figure 5.2: Definitions of predictive and prognostic biomarkers 93
Figure 5.3: Co-development of a drug and companion diagnostic 95
Figure 5.4: Types of study design for use in drug-diagnostic co-development 96
Figure 5.5: The enrichment design 97
Figure 5.6: Including classifier positive and negative patients: stratification 99
Figure 5.7: Marker-based strategy for patient stratification 101
Figure 5.8: Adaptive signature designs 103
Figure 5.9: Discovery and testing of a predictive classifier for Velcade 106
Figure 6.1: Case study: stopping probabilities calculated using Cytel’s East software 116
Figure 6.2: Integrative electronic systems in adaptive trial designs 118
Figure 6.3: Working Group proposal for sponsor involvement in decision making 124
Figure 6.4: Working Group proposal for selection decisions 125
Figure 6.5: Choosing an adaptive design: infrastructure and process requirements 130
Figure 8.1: Attrition rates and costs in drug development 142
Figure 8.2: Adaptive trials may help to restore success rates 143
Figure 8.3: Adaptive trials may help to restore success rates 144
Figure 8.4: Cost savings in a hypothetical seamless Phase 2b/3 adaptive trial 148

List of Tables
Table 1.1: Advantages and disadvantages of using adaptive clinical trial designs 31
Table 1.2: How adaptive designs can fight attrition 32
Table 2.1: Group sequential study design may save time 46
Table 2.2: Consequences of incorrect planning- treatment difference and/or standard deviation 49
Table 2.3: Comparison of simple randomization and response-adaptive randomization 54
Table 2.4: Results of a response adaptive randomization study in depression 56
Table 3.1: Bayesian designs can investigate more doses 68
Table 4.1: Advantages and disadvantages of seamless adaptive designs 78
Table 5.1: Efficiency of enrichment study designs 98
Table 5.2: Comparison of number of patients needed for enrichment designs: gefitinib example 98
Table 5.3: Free software available for calculating sample sizes in pharmacogenomic studies 102
Table 6.1: Integrity and validity in adaptive clinical trials 111
Table 6.2: Companies offering statistical expertise for adaptive trials 113
Table 6.3: Providers of software, tools and services for adaptive trials 122
Table 6.4: Clinical trial supply simulation software providers 128
Table 8.1: Cost savings for various options in COPD trial 146
Table 8.2: Retrospective analysis identifies benefits of an adaptive design 149
Table 8.3: Effect of HER2 testing on the development of Herceptin 150
Table 8.4: Potential additional sales for a drug targeted to 25% of patients tested 151
Table 8.5: Phase distribution of case studies, PhRMA Adaptive Designs Working Group 152
Table 8.6: Indications in which Bayesian adaptive designs have been used 154