FAST Congress


ADT Header

Day 1 | Day 2 | Day 3 | Download Brochure 


7:00 am Registration Open

8:00-8:30 Morning Coffee


Adaptive Dose-Finding Trial Designs

8:25-8:30 Chairperson’s Opening Remarks
David H. Manner, Ph.D., Group Leader, Exploratory and Programme Medical Statistics, Eli Lilly

8:30-9:00 Assessing the Value of an Adaptive Design within a Product Development Program

Zoran Antonijevic, Ph.D., Senior Director, Center for Statistics in Drug Development, Innovation, Quintiles

Inclusion of an adaptive design at the dose-finding stage of drug development can greatly improve chances to select an optimal dose for Phase III, and improve program’s probability of success (PoS). This improvement may sometimes be associated with an increased complexity and cost. An assessment of how much is worth investing in a trial can be done by comparing design options and development scenarios based on the expected net present value (NPV) of the product; a measurement that combines PoS, remaining time on patent, and cost. This session will demonstrate how to compare adaptive designs to traditional approached based on the expected NVP, and how to define its parameters such that that the expected NPV is maximized.

9:00-9:30 Adaptive Dose Finding Case Study: Introducing the “Maximizing Procedure”

Kenneth Liu, Ph.D., Senior Biometrician, Merck & Co.

This presentation will compare sample sizes for parallel, cross-over, and adaptive studies, and will evaluate Anastasia Ivanova’s “Maximizing Procedure” in a simulation. I will explain how to apply the “Maximizing Procedure” to an adaptive study, and will discuss the results and lessons learned.

9:30-10:00 Bayesian Adaptive Designs in Phase 2 Drug Development

David H. Manner, Ph.D., Group Leader, Exploratory and Programme Medical Statistics, Eli Lilly

In Phase 2 drug development, researchers are often faced with the challenge of choosing doses to include in a dose-finding study. This challenge can become more difficult if the dose range is wide and limited resources are available (e.g. total number of patients). The researcher must efficiently use every patient in the study to inform the dose-response curve and make decisions about Phase 3 dose(s). Given these challenges, adaptive designs have become more prevalent because they allow the researcher to learn from the accruing information and adapt to the doses that are most efficacious and safe. Bayesian adaptive designs are useful in this stage of drug development because of their ability to update information as it accrues. Bayesian adaptive designs involve careful consideration of many different aspects including the dose-response model and longitudinal model (the prior distributions needed for the parameters in these models), adaptive treatment allocation or arm-dropping, stopping rules for early efficacy or futility, and evaluation of operating characteristics. A case study will be given for a diabetes compound in phase 2 development that uses a Bayesian adaptive design.

10:00-10:30 Implementing the Continuous Reassessment Method (CRM) in a Phase I Oncology Trial

Adam Hamm, Ph.D., Director, Biostatistics, Clinsys Clinical Research

In this discussion, we present an implementation and the results of a dose escalation strategy known as the continual reassessment method (CRM) in a phase I oncology trial. This method provides an alternative to the usual 3+3 method of dose escalation. The goal is to determine the maximum tolerated dose (MTD) by using a working model that is updated according to the toxicity results of an enrolled cohort of patients (2-3 patients). Dose escalation is determined using the results of the cohort patients (dose limiting toxicities) and integrating this information with a prior model to determine the dose for the next cohort. A real example is presented.

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


Adaptive Design for
Proof-of-Concept Studies

11:30-12:00 A Bayesian Multivariate Adaptive Design for Proof-of-Concept

Kay S. Tatsuoka, Ph.D., Manager, Discovery Biometrics/Quantitative Sciences, GlaxoSmithKline

In proof-of-concept (PoC) studies there can be multiple equally important endpoints and hence multiple scenarios involving one endpoint, or a combination of endpoints, that can lead to a commitment to further medicine development. Adaptive designs have proved useful for a variety of purposes in PoC studies, including monitoring safety and increasing efficiency. In this paper a Bayesian multivariate adaptive design is proposed. The method is based on the predictive probability of observing a clinically significant multivariate outcome at the end of the study given the observed data. The proposed method is used to derive stopping rules for success and futility in clinical trials. Within the Bayesian multivariate framework various alternatives to declare a multivariate outcome clinically significant are compared. Advantages and drawbacks of the multivariate Bayesian approach versus univariate adaptive designs are explored. The interim monitoring strategies discussed in the paper are illustrated using examples from clinical trials.

12:00-12:30 Bayesian Adaptive Randomization (BAR) for Proof-of-Concept Oncology Studies

Pantelis Vlachos, Ph.D., Principal Biostatistician, Merck Serono International S.A.

This presentation outlines our use of placebo, low and high dose of experimental treatment to evaluate BAR characteristics in proof-of-concept dose selection studies in oncology. The statistical design linked treatment assignment probabilities to performance of respective arms, and smooth adaptation of assignment probabilities was guaranteed by a tuning parameter controlling how the randomization was influenced by data. A screening design was used to calculate the maximum sample size and to compare operating characteristics of the two methods. We concluded that BAR is a flexible tool which may fit easily to the needs of early development and may be ethically appealing as more subjects are assigned to more active treatment arms. 

12:30-2:00 Lunch on Your Own



2:00-2:30 Fulfilling the Promise of a Sequenced Human Genome

Eric D. Green, M.D., Ph.D., Director, National Human Genome Research Institute, National Institutes of Health

The Human Genome Project’s completion of the human genome sequence in 2003 was a scientific achievement of historic proportions. It also signified a critical transition, as this new foundation of genetic information started to be used in powerful ways by researchers and clinicians to tackle increasingly complex problems in biomedicine. Current efforts in genomics research are focused on using genomic data and technologies to acquire a deeper understanding of biology and to uncover the genetic basis of human disease. Together, these pursuits are moving us down an exciting path towards genomic medicine and fulfilling the promise of a sequenced human genome.

2:30-3:00 Personalized Medicine, Companion Diagnostics and Regulatory Considerations

Živana Težak, Ph.D., Associate Director for Science and Technology, Personalized Medicine/OIVD/CDRH, U.S. Food and Drug Administration

The U.S. FDA evaluates many of the products that will ultimately allow personalized medicine to be successfully implemented. This talk will focus on regulatory and scientific issues in personalized medicine, in particular on the diagnostic part, including companion and novel diagnostic devices. There are a number of approaches for clinical study designs used to evaluate companion diagnostic assays that may include specific diagnostic and therapeutic considerations. FDA faces evolving regulatory challenges for in vitro diagnostic assays, including further development of the regulatory structure for companion diagnostics and clarity on co-development issues. The talk will describe some of FDA efforts to integrate the various medical product regulatory authorities in order to improve clarity and efficiency in regulating personalized medicine products.

3:00-4:00 Networking Refreshment Break with Poster and Exhibit Viewing


Data Management Challenges
in Adaptive Clinical Trials

4:00-4:30 pm Data Management Across the Drug Development Continuum: The Place for Data Governance

Anne Wiles, Advanced Analytics Strategist, Health and Life Sciences, SAS Institute, Inc.

The operational complexity of implementing an adaptive clinical trial is well recognized as is the need for a robust technology solution to support both the design and the execution of an adaptive trial. As organizations scale up in their ability to design and implement adaptive trials it is critical that in addition to the technology and process changes involved, a structure be put in place to govern the data across the drug development continuum. If the data upon which a simulation is based are questionable then so are the outputs. An approach will be discussed that addresses the need to ensure the quality, availability and integrity of the data, assess the reliability of the sources of the data and describe the need for metadata management.

4:30 Close of Day