FAST Congress

 

ADT Header

Day 1 | Day 2 | Day 3 | Download Brochure 

Tuesday, September 14

7:00 am Registration Open

7:30-8:30 Morning Coffee

 

Implementation of Adaptive Clinical Trials

8:25-8:30 Chairperson’s Opening Remarks
Tom Parke, Head of Clinical Trial Solutions, Tessella

8:30-9:00 Adaptive Design in Clinical Trials: Theory and Implementation

Zhenming Shun, Ph.D., Director, U.S. Head, Oncology, Biostatistics and Programming, Sanofi-Aventis

The interim data of an on-going trial can be valuable for the sponsor to validate trial assumptions and, in case the assumptions are not appropriate, to modify the original study design accordingly. However, adequate methods should be carefully selected and certain safe measures should always be considered to maintain the integrity of the study if a study is designed with such an adaptive nature. In this presentation, fundamental statistical principles and practical issues in the implementation of such adaptive designs are discussed. Some commonly used methods in adaptive clinical trials are introduced.

9:00-9:30 Lessons from Traditional Designs

Robert Cuffe, Ph.D., Statistician, Infectious Diseases, Medicine Development Centre, GlaxoSmithKline

A lack of experience with adaptive designs can prevent project teams from exploring the suitability of adaptive designs for their development program. The number of traditionally designed trials still outweigh the number of adaptively designed trials by some margin. It can seem hard to find examples that will speak to the decision that a project team has to make. Fortunately, many traditional trials confront issues very similar to those specific to adaptive designs. This talk will explore the issues of Phase IIb/III trials drawing on lessons learned from adaptive and traditionally designed trials. Drawing on the more readily available experience of such trials will enable project teams make informed decisions about the benefits of adaptive designs and increase their use where appropriate.

Sponsored by
Tessella logo
9:30-10:00 Adaptive Clinical Trials and Biomarkers

Tom Parke, Head of Clinical Trial Solutions, Tessella

Biomarkers and translational medicine promise much, but until they are properly incorporated into clinical trial design their benefits to drug development will be limited. We’ll look at how prognostic biomarkers can be exploited in adaptive trials, not only when they are well characterized in advance, but also when they are characterized during the trial itself. Example trial designs will be discussed and simulation results presented.

10:00-10:30 Implementing Adaptive Clinical Trials: From Concept Development to Final Analysis

J. Kyle Wathen, Ph.D., Research Statistician, University of Texas, MD Anderson Cancer Center

Recently the interest in adaptive clinical trial designs has grown considerably, largely due to flexibility, financial gains and better decision making. However, logistical and technical complexities have hindered the actual conduct of trials utilizing adaptive methodology. In this talk, I present three trials employing Bayesian adaptive designs. I will discuss the initial concept development, implementation process, evaluation of methodology and trial conduct. For each step, I will discuss the skills that are required and the specific interactions that occur. Using the trials as examples, I will demonstrate different approaches that have been successful in application of adaptive designs.

10:30-11:30 Networking Coffee Break with Poster and Exhibit Viewing

11:30-12:00 pm Enrollment and Data Delivery Kinetics and their Impact on Adaptive Designs

Jason Clark, Ph.D., Associate Director, BARDS, Merck Research Labs

Adaptive designs have great potential to reduce the number of patients needed in the conduct of clinical trials. When constructing adaptive designs, simplifying assumptions may be made regarding patient enrollment rates and the time between study entry and delivery of a patients’ primary endpoint data to the sponsor. Patient enrollment is often slower at the beginning of a trial and rates increase over time as additional centers are initiated. Lags between realization of trial endpoints and delivery of such data to a trial sponsor are to be expected in clinical research. These lags exist for a variety of reasons including, but not limited to, data point adjudication and resource constraints around data entry at investigational sites. Variations from initial enrollment assumptions and significant lags in the delivery of patient data may impact and potentially undermine the benefits of such designs. We will consider a case of a hypothetical oncology study using a progression-free survival endpoint and a group-sequential stopping rule for efficacy and futility for the endpoint in question. We will examine through simulation how variable enrollment and data entry lags might impact the expected sample sizes, trial costs, and analysis timings of such a trial. We will conclude by addressing ways statisticians and clinical trial operational personnel may interact to mitigate and address such concerns.

12:00-12:30 Agile Designs for First-in-Human Studies

Itay Perlstein, Ph.D., Quantitative Clinical Pharmacology, Merck & Co., Inc.

First-in-human (FIH) studies are conducted with healthy subjects and escalating doses to discern the initial safety and pharmacokinetics of new drug candidates. The conventional FIH study design explores the same number of subjects for doses associated with no effect, with relevant effect and with adverse events. An alternative study design is suggested that employ adaptive and flexible elements. Simulations demonstrate that such a design is a feasible alternative, with no meaningful loss of relevant information. Utilizing the agile design provides a real opportunity to increase the efficiency of early clinical development.

12:30-2:00 Lunch on Your Own

 

Clinical Trial Design for Personalized Medicine

2:00-2:30 High-Throughput Biomarker Adaptive Design: A Shortcut to Personalized Medicine?

Yu Shyr, Ph.D., Professor and Chief, Division of Cancer Biostatistics, Department of Biostatistics; Director, Cancer Biostatistics Center, Vanderbilt-Ingram Cancer Center; Ingram Professor of Cancer Research, Vanderbilt University School of Medicine

In the last decade, researchers have seen a veritable explosion in high-throughput biomarker data. Today’s technology makes it possible for molecular biologists worldwide to collect increasingly vast quantities of biological information of unprecedented detail. The advancement of personalized medicine depends on clinical trials with high-throughput biomarker endpoints. This talk outlines recent developments in biomarker-adaptive trial design, including adaptive signature design (ASD), biomarker-adaptive threshold design (BATD), and mixed design (ASD + Bayesian). The advantages and limitations of these designs will be highlighted.

Sponsored by
Tessella logo
2:30-3:00 Designing Clinical Trials for Personalized Medicine

Scott Berry Ph.D., President and Senior Statistical Scientist, Berry Consultants

Developing personalized medicines runs the risk of higher development costs for smaller markets. Looking for subgroup effects after a trial completes is a sure-fire way to find false positives. How can we look for possible subgroups effects without requiring huge trial sizes to cope with the multiplicities? Using examples, it will be shown how adaptive trial designs and Bayesian modelling can help.

3:00-3:30 Adapting Randomization Based on Patient Characteristics

J. Kyle Wathen, Ph.D., Research Statistician, University of Texas, MD Anderson Cancer Center

Randomized clinical trials are the gold standard for obtaining scientifically valid treatment comparisons. However, if a physician favors one treatment over another, it may be more ethical for the physician to used the favored treatment rather than randomizing. As a compromise, many adaptively randomized procedures exist to unbalance the randomization in favor of the treatment that, on average, has superior results. Most of these methods assume that patients are homogeneous, and thus are likely to reach incorrect conclusions in the presence of patient heterogeneity. I propose a Bayesian adaptively randomized design that adapts randomization probabilities for each patient by including patient characteristics such as biomarkers, prognostic subgroups or disease subtype. A simulation study is presented and the method is illustrated in the context of two ongoing clinical trials.

3:30-4:30 Networking Refreshment Break with Poster and Exhibit Viewing

 

Regulatory Issues with
Adaptive Trial Designs

4:30-5:00 Developing Internal Regulatory Guidance for Adaptive Trials

Jennifer Dudinak, Ph.D., Global Head, Inflammation, Regulatory Affairs, Roche

This talk addresses key regulatory strategic points to consider for adaptive designs, and focuses primarily on what development teams should be thinking about when moving forward with adaptive trials.

5:00-5:30 Regulatory Issues with Adaptive Trial Designs

Allan Rosen, M.S., Associate Director, Biostatistics, Kendle International, Inc.

Adaptive trial designs are one of the hottest topics of current research and debate in the drug development industry. There has been ongoing debate to highlight, understand, and avoid the real and perceived pitfalls of adaptive designs, partially due to reluctance of regulators to fully embrace the use of adaptive trial designs in all stages of drug development. This presentation will provide an overview of the current landscape of adaptive clinical trials and will discuss current agency perspectives on the use of adaptive trial designs and recent developments in regulatory guidance on this topic. The discussion will include recommendations for successful registration strategies that employ such techniques. Agency concerns about specific knowledge gaps and suggestions for additional research will be discussed.

 

5:30 Close of Day