Adaptive and Flexible Clinical Trials

 Content: Background Introduction Definition and History of Traditional Clinical Trials Definition of Adaptive Clinical Trial Precursors to Modern Adaptive Clinical Trials New Enabling Technologies and Other Requirements for Adaptive Trials Rationale for Adaptive Clinical Trials Learn and Confirm Classification and Terminology of Adaptive Clinical Studies Non-Adaptive Study Designs Limitations of Adaptive Clinical Trials Performance Criteria for Well Designed Clinical Trials Evolving Regulatory Environment for Adaptive Clinical Trials Regulatory Guidance from the FDA References  Background: Conventional Statistics Basic Statistics Statistical Schools Frequentist Method Bayesian Method Likelihood Method Other Schools Descriptive Statistics Inferential Statistics Comparative Statistics Hypothesis Testing Null Hypothesis and Standard of Proof in Clinical Trials Example of Inference and Hypothesis Testing Statistical Tests and Choice of Statistical Test Examples of Statistical Tests Fundamental Statistical Assumptions  Statistics Used in Adaptive Clinical Trials Introduction Preserving the Alpha What Is Alpha? Misconception about p Values Splitting the Alpha Methodologies for Allocating Alpha Evolution of Adaptive Analytic Methods: Interim Analysis Adaptive Methods Limitations of Adaptive Statistical Techniques Bayesian Approach Simulations and Modeling The FDA\'s Stance on Adaptive Techniques References  Specific Requirements for Adaptive Trials Requirements for Endpoints in Adaptive Studies Surrogates and Biomarkers Practical Requirements  Adaptive Randomization and Allocation Traditional Fixed Allocation Simple Randomization Restricted and Blocked Randomization Stratified, Nested, and Similar Randomization Balancing (Covariate) Adaptive Randomization Response (Outcome) Adaptive Randomization Combination and Multimodal Randomization Bayesian Randomization Adaptation of Inclusion and Exclusion Criteria Based on Blinded Data Patient Enrichment Adaptations  Sample Size Reestimation Background Sample Size Reestimation Based on Blinded Data Sample Size Reestimation Based on Unblinded Data Adjustment in Follow-Up Time Internal Pilot Studies Additional Rules References  Traditional Dosing Introduction Definitions and Objectives of Dose Selection Issues Pharmacokinetics Factors Affecting Pharmacokinetics Dose-Response Curves Types of Dosing Traditional Dose-Escalation and Dose-Ranging Studies  Adaptive Dosing Adaptive Dose Finding Phase I Studies 3 + 3 and Related Designs Continual Reassessment Method Dose Escalation with Overdose Control Stochastic Approximation Methods Summary of Single-Parameter Models Additional Models and Methods Dose Adaptation in a Pivotal Study Changes in Concomitant Medications and Procedures References  Interim Analysis and Adaptive Termination of Study and Study Arms Overview Data and Safety Monitoring Boards Stopping Rules Individual Sequential Designs Group Sequential Designs References  Adaptive Changes in Study Design and Decision Rules Overview Changes to Follow-Up Period Flexible Designs Changing the Endpoints and Hypothesis Changes to Test Statistic or Analysis References  Seamless Designs and Adaptive Clinical Trial Conduct Seamless Designs Challenges in Adaptive Trials Maintaining the Blind Infrastructure and Operations Adaptive Trial Protocols  Analysis and Interpretation of Results General Issues in Interpretation of Clinical Trial Results Interpretation of Adaptive Trial Results Documentation of Trial Integrity Statistics of Adaptive Trial Analysis Summary References Index
 Abstract:             Adaptive clinical trial designs, unlike traditional fixed clinical trial designs, enable modification of studies in response to the data generated in the course of the trial. This often results in studies that are substantially faster, more efficient, and more powerful. Recent developments in web-based real-time data entry and advances in statistical methods have made adaptive clinical trials much more popular because they have become both more practical and attractive. However, there is paucity of resources that explain the mathematical framework and the practical considerations for adaptive