ASA Biopharmaceutical Section Statistics Workshop

Online Program

Search by Last Name:
Keyword:

Mon, Sep 17
SC1 Short Course 1		8:30 AM - 12:00 PM Salon III



Causal Inference and Its Applications in Clinical Development View Presentation Donald Rubin, Harvard University
Additional Resources View Presentation


SC2 Short Course 2		8:30 AM - 12:00 PM Salon III



Adaptive Designs - Part 1 View Presentation Shein-Chung Chow, Duke University
Adaptive Designs - Part 2 View Presentation


SC3 Short Course 3		1:30 PM - 5:00 PM Salon III



Statistical Methods for Evaluating Tests and Biomarkers in Medicine View Presentation Margaret Pepe, University of Washington and Fred Hutchinson Cancer Research


SC4 Short Course 4		1:30 PM - 5:00 PM Salons H&J



Data Monitoring Committees in Drug Development View Presentation Scott Evans, Harvard School of Public Health; C. George Rochester, U.S. Food & Drug Administration, CDER


SC5 Short Course 5		1:30 PM - 5:00 PM





SC5 Short Course 5		1:30 PM - 5:00 PM Salon I



Bayesian Analysis View Presentation Bradley P. Carlin, University of Minnesota


Tue, Sep 18
GS1 Global Harmonization: The Role of Statisticians from FDA, Industry, and Academia in Effective Partnership		8:30 AM - 10:00 AM Salons I, II & III
Organizer(s): Bruce Binkowitz, Merck; Shein-Chung Chow, Duke University; Lilly Yue, FDA

To accomplish our public health missions, effective partnership among industry, academia and government is essential. Effective partnership for global harmonization requires a close cooperation among multiple parties. Each party brings their own principles and needs to the partnership. To make the partnership effective, each party must look beyond their own local needs and establish joint goals and responsibilities. Such an effective partnership can produce a synergy of outcomes far beyond what any individual party could accomplish alone. FDA, academic and industry statisticians should, can and do play a vital role in such partnerships. These partnerships can be the most effective way of bringing innovation to practice and can occur at a local level, or on a global scale. The speakers in this session will represent the perspectives of industry, government, and academia, about past, existing, and potential joint efforts of FDA, academic and industry statisticians, where such efforts include areas of effective partnership and the resulting positive impacts on the public health.
"Critical Path Initiative - An Opportunity for Partnership in Statistics" View Presentation George YH Chi, J&JPRD
Global Harmonization: The Role of Academia in Effective Partnerships View Presentation Mitchell W. Krucoff MD, Duke University Medical Center
The Global Harmonization Task Force View Presentation Larry Kessler, FDA


GS2 Adaptive Designs		8:30 AM - 10:00 AM Salons I, II & III
Organizer(s): Shein-Chung Chow, Duke University; Ning Li, FDA; Mei-Hsiu Ling, Novartis; Kooros Mahjoob, FDA

Recently, there is an enthusiastic interest to use adaptive designs in clinical development program. This session will focus on regulatory and application issues of the adaptive design. Regulatory presentations will focus on clinical and statistical rationale behind the adaptive trial designs for doable vs. problematic adaptive approaches. Application presentations will focus on the methodology of adaptive design including data analysis and dose selection.
Agenda and Q&A View Presentation
Flexible Sample Size Design in Clinical Trials - a Too for Improvement Lu Cui, Eisai Medical Research Inc.
Stagewise Planning for Clinical Trials View Presentation Sue-Jane Wang, FDA
Combining Hypothesis Testing with Bayesian Analyses for Dose Response and Dose Selection View Presentation David Gruben, Pfizer
Panel Discussion View Presentation Robert O'Neill, FDA; Peter Thomas, Novartis; Bruce Turnbull, Cornell University


R1 2007 FDA Roundtables		11:45 AM - 1:00 PM



Data Analysis for Microarrays SAmir Lababidi, FDA
Logistics of Implementing Adaptive Trial Designs and Adaptive Randomizations Eva Miller, Almac Group
Statistics and Quality by Design Roswitha E. Kelly, FDA
Causal Inference in Clinical Trials Thamban Valappil, FDA
Evaluating the Quality of Trial Methodology Vance Berger, NIH
Industry and FDA Interactions: What to talk about and when Daphne Lin, FDA
Sample Size Re-estimation for Clinical Trials Dennis King, Stat King Consulting, Inc.
Proof of Concept Trials Larry Gould, Merck
Adaptive Design in Global Clinical Trial H.M. James Hung, FDA
CDISC Standards for Clinical Trial Data Chris Holland, FDA
Tc Clinical Trials Yi Tsong, FDA
Bayesian Designs in Clinical Trials for Medical Devices Xuefeng Li, FDA
Statistical Analysis for QT Data Phillip Dinh, FDA
Statistical Computing Environments Susan P. Duke, GlaxoSmithKline
Biomarker Qualification in Clinical Trials Sue-Jane Wang, FDA
Blinded Sample Size Re-estimation Biao Xing, Genentech
Analysis of Non-randomized Clinical Studies in a Regulatory Environment Yunling Xu, FDA
FDA's Statistical Guidance on Reporting Results for Evaluating Diagnostic Tests Kristen Meier, FDA
Normalizing Reporting Ratios in Adverse Event Data Mining Stuart Chirtel, FDA
Statistical Issues in Neurology Drug Trials Kun Jin, FDA
Sample Size Effect in Two-Arm Non-Randomized Trials Shiowjen Lee, FDA
The Case for Improving the Role of the Biostatistician in Clinical Drug Development Jitendra Ganju, Amgen
Design Issues for Long-Term Maintenance Trials When the Comparator is Placebo Yeh-Fong Chen, FDA
Adaptive Designs for Phase II Studies Xiaoming Li, sanofi-aventis
Qualification of Individual & Composite Biomarkers for Clinical Use Karen Kesler, Rho, Inc.
Bayesian Decision Analysis in Clinical Trials Telba Irony, FDA
Single-arm Phase II Studies: Do they have a place in modern clinical research? Rebecca R. Hozak, Eli Lilly & Co.
Study Design Issues in Genetic Diagnostic Devices Kyunghee Song, FDA
Modeling and Simulation in Drug Development Pat Kelley, Astra Zeneca
The Role of Overall Survival in Oncology Trials Yanzhi Hsu, Novartis
Group Sequential Design Yanli Zhao, Merck
Using Baseline Data as a Covariate: Can we do better than ANCOVA? G. Frank Liu, Merck
Project Management Tips & Tools for Statistical Projects Shailaja Suryawanshi, Merck
Guidance Development: An Update for Statisticians Steve Wilson, FDA
Challenges of PKPD Modeling for Statisticians Kuenhi Tsai, Merck
Objective Performance Criteria (OPC) in clinical studies Yao Huang, FDA
The Client/Provider Relationship: How to work with your CRO or with internal “clients” in your company Katherine Monti, Rho, Inc.
Analysis Plans--Lessons Learned Steve Gilbert, Rho, Inc.
Optimizing Sponsor-FDA Interaction During Development Duane Snavely, Merck
One versus Two Adequate Studies for Providing Clinical Evidence of Effectiveness: Issues, concerns and benefits Jianliang Zhang, MedImmune
Top five statistical concepts you wish your clinical partners understood better: Come share your list, tips and experiences Virginia Recta, FDA
Regional Differences in Primary Endpoints and the Statistician's Role in Dealing with Them Lucy Shneyer, Schering Plough Research Institute
Disease Modifying Trial Designs and Statistical Analysis Chunming M. Li, Pfizer
Quantifying Synergy for Combination Drug Studies John Joseph Peterson, GlaxoSmithKline Pharmaceuticals
Primary vs Secondary Endpoints--Are we doing more harm than good using obsolete classifications? Qi Jiang, Amgen
Clinical Trials in India, Africa, Asia Carolyn A. Carroll, Stat Tech, Inc.
Evaluating effectiveness of masking patients Richard Kotz, FDA


PS01 Topics in Bayesian Clinical Trials		1:00 PM - 2:15 PM Salons I&II
Organizer(s): Suman Bhattacharya, Genentech; Telba Irony, FDA; Andrew Mugglin, University of Minnesota

The nascent field of Bayesian Clinical Trials continues to gain visibility and popularity due to several inherent advantages, such as the availability of prior information, flexibility of modeling, and the appeal of adaptive designs based on predictive probabilities. The number of "success stories" continues to grow as more organizations try out these techniques, but the challenges remain numerous, as well. In this session we will focus on several experiences involving the implementation of Bayesian methods in clinical trials, in organizations as diverse as the FDA, industry, and academia.
Using Bayesian Methods to Give Patients a Better Outcome in a Clinical Trial Jay Kadane, Carnegie Mellon University


PS02 Innovative Modeling and Simulation for Efficient Drug Development		1:00 PM - 2:15 PM Salons III & IV
Organizer(s): Brenda Gaydos, Eli Lilly & Company; Alan Hartford, Merck; Sue-Jane Wang, FDA

The FDA Critical Path Initiative white paper describes the urgent need to modernize medical product development. The associated Opportunities List presents specific opportunities, that if implemented, can help achieve this objective. Clinical trial simulation has been identified as an opportunity to streamline clinical trials. Simulations that incorporate biological models such as PK/PD and disease progression models together with design, trial execution, and data analysis models can be used to develop more efficient and informative designs. However, it is recognized that identification of tools and best practices are needed as a first step. In addition, more work is needed in model development. In this session, an overview of model-based drug development will be presented. In addition two different areas of modeling will be discussed: disease progression modeling, and disease modeling from a systems biology approach.
A Model-Based Framework for Quantitative Decision-Making in Drug Development View Presentation Kenneth G. Kowalski, Pfizer, R&D
Statistical Method to Analyze Data from Delayed Start Design in Parkinson’s Disease View Presentation Ohidul I Siddiqui, CDER, FDA
Using Mathematical-Statistical Modeling to Inform the Design of HIV Treatment Strategies and Clinical Trials View Presentation Marie Davidian, North Carolina State University


PS03 Statistical Challenges/Issues in the design and analysis in Therapeutic Areas—Chronic Pain, Inflammation and Drug Abuse		1:00 PM - 2:15 PM Salon V
Organizer(s): Gary Aras, Amgen; David Breiter, Guidant; Ling Chen, FDA; Daphne Lin, FDA

This session is devoted to statistical issues that are specific to Chronic Pain, Inflammation and Drug Abuse therapeutic areas. The first speaker will address issues in chronic pain opioid analgesic trials. The second speaker will discuss pros and cons of currently popular ACR20 endpoint in rheumatology trials and discuss an alternative, CDAI-Ltie, proposed by leading researchers in the area. The third speaker will discuss drug abuse potential studies and propose strategies for implementing a crossover design in such a study.
Design and analysis issues in randomized clinical trials of opioids for chronic pain View Presentation Catherine Munera, Purdue Pharma
Clinical Disease Activity Index (CDAI)-lite as an End Point in Rheumatoid Arthritis(RA) View Presentation Yun Chon, Amgen, Inc
Implementing Crossover Design in Drug Abuse Potential Studies -- Issues and Strategies View Presentation Ling Chen, FDA


PS04 Quality by Design: Gaining Process Understanding in a Risk Based Culture		1:00 PM - 2:15 PM Salon VI
Organizer(s): Dave Christopher, Schering-Plough; Atiar Rahman, FDA; Timothy Schofield, Merck Research Labs; Yi Tsong, FDA

Just as clinical trials establish the efficacy and safety of a new drug, Chemistry, Manufacturing and Controls (CMC) development establishes the process to manufacture the product to meet clinical requirements. Quality by Design, an FDA initiative, is based on designing necessary quality into a manufacturing process instead of “testing” quality into the finished product. This requires developing a mechanistic understanding of how formulation and process factors impact product performance. And, as stated in ICH Q8, “It should be recognized that the level of knowledge gained, and not the volume of data provides the basis for science-based submissions and their regulatory evaluation.” Statistical thought process and methodology is critical for achieving this objective. This session will provide an overview of the QbD initiative and illustrate some statistical approaches relevant to QbD.
Quality by Design: Challenges for CMC Statisticians View Presentation Yi Tsong, FDA
A Bayesian Approach to the ICH Q8 Definition of Design Space View Presentation John Joseph Peterson, GlaxoSmithKline Pharmaceuticals
Improving Quality through Robust Design View Presentation Vijay Nair, University of Michigan


PS05 Approaches to Handling Missing Data in Clinical Trials		2:30 PM - 3:45 PM Salons I&II
Organizer(s): Merrill Birkner, Genentech; Joan Buenconsejo, FDA; Robert Small, AtheroGenics

In modern clinical trials, where large amounts of data are collected on each individual, it is common for at least some of that data to be missing. The reasons for this missing data can include: patient dropout due to adverse events or lack of efficacy, lost lab samples, missed patient visits, mistakes at the clinic, noncompliance, etc. The trend has been to increase the amounts and types of data (i.e. laboratory data), which therefore leads to a greater likelihood of missing data. The missing data can have widely varying effects on the interpretation of the results of the trial –from virtually no effect to making the trial un-interpretable. There are a number of approaches to handling these missing data problems. In this session, the three speakers will describe very different approaches to the missing data problem. Dr. Mallinckrodt will describe the uses of multiple imputation and its use to assess the effects of the missing data. Dr. Scharfstein will describe a completely different concept. His method assesses the sensitivity of the results by parameterizing the effect of the missing data and evaluating the impact by varying this parameter. Lastly, Dr. Kim will discuss different imputation strategies (e.g. continuous responder analysis) in handling missing data in pain trials. These three approaches cover a range of concepts and approaches while being possibly mutually supportive. They also include recent research and more sophisticated methods as compared to some classical methods.
An Analytic Road Map for Incomplete Longitudinal Clinical Trial Data Craig H Mallinckrodt, Eli Lilly
A Global Sensitivity Analysis Paradigm for Analyzing Randomized Clinical Trials with Informative Loss of Follow-up View Presentation Daniel Oscar Scharfstein, Johns Hopkins University
Chronic Pain Trial and Missing Data View Presentation Yongman Kim, FDA


PS06 Biomarkers		2:30 PM - 3:45 PM Salons III & IV
Organizer(s): Alex Bajamonde, Genentech; Shenghui Tang, FDA; William Wang, Merck

With rapid advancements of medical sciences and technologies, many innovative biomarkers become available. These makers can play an important role in the critical path of developing medical products. In this session, we will discuss the design, validation, analysis and application of these biomakers by three experts from academia, FDA and industry.
Predictive and Prognostic Markers: An FDA Perspective View Presentation Estelle Russek-Cohen, OSB/CDRH/FDA
Towards Standards for Design of Definitive Biomarker Accuracy Studies View Presentation Margaret Pepe, University of Washington and Fred Hutchinson Cancer Research
Assessing Immune Markers and Correlates of Protection in Vaccine Studies Ivan S. F. Chan, Merck Research Laboratories


PS07 Statistical Challenges/Issues in the design and analysis in Therapeutic Areas—Cardiovascular and Organ Transplantation		2:30 PM - 3:45 PM Salon V
Organizer(s): Gary Aras, Amgen; David Breiter, Guidant; Ling Chen, FDA; Daphne Lin, FDA

This session is devoted to statistical issues specific to Cardiovascular and organ transplantation therapeutic areas. The first speaker will address issues in drug/device combination trials, specifically addressing drug eluting stent trials.. The second speaker will discuss the critical issues surrounding clinical trials in organ transplantation using examples from published experiences. The third speaker will discuss issues in the design and analysis of cardiovascular device trials with hospitalizations as an endpoint of interest, including issues in the representation of hospitalization analyses.
Trial Design Challenges in Next Generation Drug-Eluting Stents View Presentation Peggy J Pereda, Boston Scientific Corporation
Statistical Challenges in the Design and Analyses of Clinical Trials in Solid Organ Transplantation LaRee Ann Tracy, FDA/CDER/OB/DB4
Presenting and Analyzing Hospitalization Data View Presentation Elizabeth G Galle, Boston Scientific CRM


PS08 Assuring Product Quality through Specifications		2:30 PM - 3:45 PM Salon VI
Organizer(s): Dave Christopher, Schering-Plough; Atiar Rahman, FDA; Timothy Schofield, Merck Research Labs; Yi Tsong, FDA

End product testing helps assure quality and consistency of manufactured material. Specifications are established on critical quality attributes that bridge properties of commercial lots to material used in clinical trials. Properties such as potency, dose uniformity, and dissolution are drug product characteristics that are typically used to monitor quality. Part of a risk based strategy for controlling product includes acknowledgement of risk to the manufacturer and customer alike. Increased risk to manufacturers of unnecessarily rejecting good product may arise from testing paradigms utilizing increased sampling, such as medium and large samples in dose uniformity testing. Likewise there is confusion throughout regulation and the industry, regarding whether a specification on dose or potency refers to the batch average, or individual doses or measurements on a sample. This session will explore some of the risks associated with quality control testing, and illustrate statistical strategies for mitigating risk.
Setting CU Acceptance Criteria for Larger Sample Sizes View Presentation Gregory L Larner, Pfizer
STATISTICAL CONSIDERATIONS ON DOSE CONTENT UNIFORMITY (DCU) TEST WITH SMALL, MEDIUM AND LARGE SAMPLE SIZES View Presentation Meiyu Shen, FDA
The Challenges of Achieving Clinically Relevant CM&C Specifications View Presentation Jeffrey D Hofer, Eli Lilly and Company


PS09 Study Designs and Analyses in Veterinary Medicine: Similarities and Differences to Human Medicine		4:00 PM - 5:15 PM Salon V
Organizer(s): Thomas J Keefe, Colorado State University; Anna Nevius, FDA; David M Petullo, Center for Veterinary Medicine

The FDA Center for Veterinary Medicine is responsible for ensuring that animal drugs and medicated feeds are safe and effective for their intended uses and that food from treated food animals is safe for human consumption. New animal drugs are typically classified as therapeutic or non-therapeutic. Animal species can be generally classified as companion animals or food animals. This diverse range of indications and animal species leads to many different factors that encompass the experimental design and analyses of studies used to demonstrate the effectiveness and safety of new animal drugs. While some issues in new animal drug submissions are similar to the human medicine approval process, some issues are different. This session will discuss these similarities and differences in detail.
Introduction and Overview View Presentation David M Petullo, Center for Veterinary Medicine
Issues in the Design and Analysis for Clinical Studies of Companion Animals: Similarities and Differences to Human Clinical Studies View Presentation Thomas J Keefe, Colorado State University
Study Designs and Analyses in Food-Producing Animal Clinical Studies: Similarities and Differences to Human Clinical Studies View Presentation Zhanglin Lin Cui, Eli Lilly and Company


PS10 Important Design Concepts in the Planning of Pivotal Vaccine Trials		4:00 PM - 5:15 PM Salons III & IV
Organizer(s): Allen Izu, Novartis; Jingyee Kou, FDA; Santosh Sutradhar, Pfizer

Three different aspects of important design concepts for the planning of pivotal vaccine trials are presented, respectively, by speakers from Academia, Industry, and FDA. Academia: When diagnostic tests cannot distinguish between transient and chronic infections, there may be different ways to define the chronic infection endpoint in a vaccine trial. In this work, several definitions of chronic infection that are based on the periodically observed viral statuses are evaluated using a multi-state model, with application to HCV. The type I error and the power of log-rank tests for the vaccine efficacy against chronic infections are assessed in the framework of composite hypotheses Industry: Standard sample size formulas used for designing vaccine lot consistency trials rely on only one component of variation, namely, the variation in antibody titers within lots. The other component, the variation in the means of titers between lots, is unintentionally ignored or is assumed to be equal to zero. Using data from a real lot consistency trial we show that when the variation between lots is only 0.5% of the total variation the sample size increases by nearly 400% when compared to the assumption that there is no variation between lots. We also discuss the increase in the sample size due to correlated comparisons arising from 3 pairs of lots as a function of the between-lot variance. FDA: Lastly, an FDA Reviewer will present a summary of experience and perspective with designs and associated sample size methods used for pivotal trials for licensed vaccine products since 2000.
Evaluation of Chronic Infection Endpoints for HCV Vaccine Trials View Presentation Minhee Kang, Harvard School of Public Health
Sample size for equivalence trials: A case study from a vaccines lot consistency trial View Presentation Jitendra Ganju, Amgen
An Overview of the Vaccine Products licensed by US FDA Since Year 2000 View Presentation Jingyee Kou, FDA


PS11 Design Issues for Diagnostic Devices		4:00 PM - 5:15 PM Salons I&II
Organizer(s): Vicki Petrides, Abbott Laboratories; Lakshmi Vishnuvajjala, FDA

There are many design challenges that must be addressed during the development of diagnostic products. This session will explore some of these aspects including design verification and validation practices, assessing diagnostic accuracy with and without a gold standard, and adaptive randomization in drug/diagnostic co-development.
Estimating Diagnostic Accuracy from Designs with no Gold Standard, Partial Gold Standard, or Imperfect Gold Standard Evaluation Paul S Albert, NCI
Response Adaptive Randomization in Clinical Trials for Predictive Marker Validation Thomas E Gwise, FDA
Mechanistic and Experimental Approaches to Assay Design and Associated Issues with Performance Verification Beimar N Iriarte, Abbott Laboratories


PS12 Method Transfer		4:00 PM - 5:15 PM Salon VI
Organizer(s): Dave Christopher, Schering-Plough; Atiar Rahman, FDA; Yi Tsong, FDA

Analytical methods are used during pharmaceutical development and in quality control to measure critical quality attributes of a drug substance or a drug product. These methods are typically developed in the research laboratories, where they are used to describe the properties of nonclinical and clinical research materials, and to establish commercial product specifications. These may then be transferred to the commercial testing laboratory, where they are used to help guarantee the quality of manufactured product. Strategic method transfer study designs and analyses may be employed to demonstrate acceptable performance of the method in the commercial testing laboratory. The method transfer study should account for the risks of unsatisfactory performance of the method in the commercial laboratory, and should acknowledge the specification range for the critical quality attribute.
Comparison of Different Analysis Approaches to Analytical Method Transfe View Presentation Jinglin Zhong, FDA
Transfer of Methods Supporting Biologics and Vaccines View Presentation Timothy Schofield, Merck Research Labs
Analytical Method Transfer View Presentation Kathy Lee, Food and Drug Administration


Wed, Sep 19
PS13 Pharmacogenomics that Identify Diagnostic (Expression) Biomarkers to Characterize Therapeutic Benefits versus Risks		8:45 AM - 10:00 AM Salons I&II
Organizer(s): Alex Bajamonde, Genentech; James J Chen, NCTR - FDA; Jared Lunceford, Merck Research Labs; Hojin Moon, California State University - Long Beach; Sue-Jane Wang, FDA

Pharmacogenomics is the science of determining how the benefits and adverse effects of a drug vary among a target population of patients based on genomic features of the patient’s germ line and diseased tissue. Many envision that pharmacogenomics may be a pathway to personalized medicine as it may help to identify the optimal treatment to individual patients based on their genomic/genetic profile. What can statisticians contribute to this vision? In this session, we highlight three statistical presentations. A robust statistical classification algorithm to classify patients likely vs. unlikely to response (including efficacy and safety) to therapy based on high dimensional gene expression data will be introduced. A novel statistical pharmacogenomic approach “Co-eXpression ExtrapolatioN” (COXEN) will be illustrated with application to real dataset. The approach can effectively identify concordant genomic chemosensitivity biomarkers between two independent expression profiling data sets with high accuracy. Statistical approaches to identify prognostic/predictive biomarker for binary classification and the bias correction methods will be discussed in the context of drug/device co-development.
High-dimensional biomarkers in personalized medicine via variable importance ranking View Presentation Hojin Moon, California State University - Long Beach
COXEN: A New In Silico Pharmacogenomic Approach to Personalized Chemotherapeutics Jae K Lee, University of Virginia
The identification of Prognostic/Predictive biomarkers using optimal cutpoints for predicting response to therapy View Presentation Nusrat Rabbee, Genentech, Inc.


PS14 Issues and Challenges in Dichotomizing Continuous Variables in Clinical Trials		8:45 AM - 10:00 AM Salons III & IV
Organizer(s): Ivan S. F. Chan, Merck Research Laboratories; H.M. James Hung, FDA; Qi Jiang, Amgen; Yi Tsong, FDA

In clinical trials, dichotomization or categorization of a continuous variable into “responders” and “non-responders” is rather frequent, though it is recognized that such a process results in loss of valuable information. It is less well recognized that this process may not achieve its main goal of ensuring clinical relevance. When the cut-off level involved in the process is somewhat arbitrary, additional issues may arise, such as multiple cut-off levels leading to the multiplicity problem in statistical inference. Is there a good rationale for application of the process? The speakers will share their perspectives on the issues and challenges with this process.
Responder Analyses and the Assessment of a Clinically Relevant Treatment Effect View Presentation Steven Snapinn, Amgen
Not Enough Dichotomies View Presentation Thomas Permutt, FDA
Dichotomisation in clinical trials: issues and concerns Steven Julious, University of Sheffield


PS15 Challenges in Oncology Drug Development		8:45 AM - 10:00 AM Salon V
Organizer(s): Suman Bhattacharya, Genentech; Aloka Chakravarty, FDA; Gene Penello, FDA; Erik Pulkstenis, Human Genome Sciences; Xiaolin Wang, Genentech

Recent advancement of cancer therapies has presented new challenges and triggered resurgence of interests in oncology trial design and analysis. In this session, we will hear from the experts in academia, industry and the regulatory agencies about various issues of clinical trials in oncologic therapeutic area. In particular, we will hear discussions on selection of endpoints, investigator-based PFS endpoint vs independent radiological review (IRR)-based endpoints, attenuation of survival benefit due to cross over, impact of informative censoring when using IRR-based endpoints, and how to combine new therapeutic agent with multiple standard of care (SOC) to name a few. Case studies will be presented to highlight these issues.
Tykerb Experience View Presentation Alakananda J. Preston, GlaxoSmithKline Pharamaceuticals
Challenges in Oncology Drug Development – A Regulatory Perspective View Presentation Rajeshwari Sridhara, CDER, FDA
Panel Discussion View Presentation Elizabeth Garrett-Mayer, Medical University of South Carolina; Mark Rothman, Food & Drug Administration; Rajeshwari Sridhara, CDER, FDA; Grant Williams, GalxoSmithKline


PS16 Statistical Approaches for Early Drug Development		8:45 AM - 10:00 AM Salon VI
Organizer(s): Alan Hartford, Merck; Karl Lin, FDA

Early drug development continues to be a challenging arena for statisticians with problems very different from other areas of drug development. In this session, approaches to a few of these problems will be discussed including designs for formulation studies, formal determination of steady-state, and issues in translational medicine.
Flexible Statistical Methods for Early Clinical Trials View Presentation Yaning Wang, FDA
Issues in the Estimation of Time to Steady State View Presentation Deborah Lorah Panebianco, Merck & Co. Inc.
Power/Sample Size Determination for High Dimensional Data Experiments View Presentation Gengqian Cai, GlaxoSmithKline


PS17 Clinical Significance vs. Statistical Significance		10:20 AM - 11:35 AM Salons I&II
Organizer(s): Paul Gallo, Novartis; John Lawrence, FDA; Gosford Sawyerr, Purdue Pharma

In some areas of medical research, describing the clinical importance of a hypothesized or observed level of effect associated with a treatment can be quite controversial. Statistical significance is generally a more straightforward concept. An understanding of the interplay between these concepts is clearly highly relevant for clinical trial design, as statisticians and clinicians are challenged to design studies to detect effects which must be meaningful clinically; alternately, upon study completion it is natural to address whether a statistically significant finding is clinically important. For example, current regulations for prescription drug advertising preclude using “statistical significance to support a claim that has not been demonstrated to have clinical significance or validity.” In some situations, sponsors are required to address clinical significance, but there is little guidance available on how this is to be done. ICH E9 states that in a “confirmatory trial, it is equally important to estimate with due precision the size of the effects attributable to the treatment of interest and to relate these effects to their clinical significance.” A group of experienced clinicians and statisticians of varied backgrounds will discuss their perspectives on this issue, and provide a forum for healthy interchange of ideas on the topic.
Statistical Significance vs Clinical Significance Is Not A Useful Debate. View Presentation Raymond John Lipicky, LIPICKY, LLC
Clinical vs. Statistical Significance: Necessary and Competitive Allies View Presentation Joel I Veter, Statistics Collaborative
Statistical Significance and Clinical Significance View Presentation Richard Pazdur, FDA-CDER
Relationship of P-value with Observed Treatment Effect Size and Its Clinical Relevancy View Presentation Peiling Yang, Food And Drug Administration


PS18 Patient Reported Outcomes		10:20 AM - 11:35 AM Salon V
Organizer(s): Carmen Mak, Schering-Plough; Tammy Massie, FDA; Kate Meaker, FDA

A PRO (patient-reported outcome) is a measurement of any aspect of a patient’s health that comes directly from a patient. The proper design and use of PROs have presented considerable challenge to clinicians, statisticians and the FDA. Recently the FDA released a guidance document on the use of PRO which has generated a lot of interest and discussion. In this session we will explore issues related to the calibration, validation and analysis of PROs.
Defining Treatment Response as Measured by PRO Instruments View Presentation Laurie Burke, FDA
Evaluating the Responsiveness and Determining Clinical Significance for PRO Measures View Presentation Dennis Revicki, Center for Health Outcomes Research
Expressing PRO Outcomes in Terms of a Responder Analysis View Presentation Davis Gates, Schering-Plough Research Institute


PS19 Safety Data: What Should We be Doing with It?		10:20 AM - 11:35 AM Salons III & IV
Organizer(s): Qian Graves, FDA; Karen Kesler, Rho, Inc.; Barbara Krasnicka, FDA; Edmund Luo, Merck Research Labs

An ever-present concern in performing clinical trials is that we are not exposing patients to a risk of adverse events (AEs) without an appropriate level of benefit. From the statistician’s perspective, adverse events are troublesome to analyze as they can be too rare for the sample size or buried in an avalanche of data. This session examines approaches to this very important type of data including Bayesian approaches to AEs, AE meta-analysis, tracking prespecified AEs throughout the trial, and an overview of the vaccine adverse events reporting system (VAERS).
Post-Market Analysis of Safety Signals: Recent Experience Mark Steven Levenson, CDER/FDA
Computing End-of-Trial Estimates of Risk for Trials that Use Interim Safety Monitoring View Presentation Michael J Dallas, Merck Research Laboratories


PS20 Statistics in Non-Randomized Clinical Studies		10:20 AM - 11:35 AM Salon VI
Organizer(s): Kevin J Anstrom, Duke University Medical Center; Gary Aras, Amgen; Lilly Yue, FDA

Though randomized double-blind trial is the gold standard, they are not always possible or practical due to budgetary, operational or logistical reasons. Non-randomized studies are becoming more popular in pharmaceutical and device industry in recent years. Registries, case-control retrospective studies based on databases such as insurance claims database, and non-randomized prospective cohort studies are now quite popular. In contrast to randomized clinical trials, investigators have no control over the treatment assignment in these studies. Therefore, the treated and non-treated groups may differ significantly on many observed and unobserved covariates. Confounding by indication—sicker patients tend to be on more aggressive treatment--is an important problem that statisticians need to address in their analysis of some of these studies. This session will address these important statistical issues through examples in health industry.
Controlling for Time Dependent Confounding using Marginal Structural Models to Assess Epoetin Alfa Dose Relationship with Mortality View Presentation Ouhong Wang, Amgen, Inc.
Inverse Probability Weighted Methods for Estimating Treatment Effects in Observational Studies Kevin J Anstrom, Duke University Medical Center
Some Challenges in Non-randomized Medical Device Clinical Studies: A Regulatory Perspective Lilly Yue, FDA


GS3 Developing Better Clinical Trial Strategies		1:00 PM - 2:30 PM Salons I, II & III
Organizer(s): Fred Immerman, Wyeth; Richard Zhang, Pfizer

Randomized Clinical Trials (RCT) have been the major pathway for development of new medications in modern times. Today we are facing the challenges of rising R&D costs and an increase in the failure rate of clinical trials during late stage drug development. Recently, there have been significant discussions/seminars/symposia in the pharmaceutical, regulatory, and academic statistical communities on techniques that will enhance clinical trial efficiency, including adaptive and sequential designs. However, one area which has not been addressed sufficiently is the clinical trial development process itself. Before searching for specific solutions, let us step aside and ask ourselves a general question: “Do we need to revise the paradigm for clinical trial development?” The objective of this session is to present ideas and insights of academic, regulatory, and industry leaders on how we may re-define and optimize the process.
Moving from Challenge to Change on the R&D Continuum View Presentation Christopher Milne, Center for the Study of Drug Development
Critical Path Initiative: Update on FDA's Progress View Presentation ShaAvhree Buckman, Office of Translational Sciences
Maximizing Understanding in Clinical Trials by Maximizing the Talents of the Contributors Charles Knirsch, Pfizer Inc
Statistics Role in Creating a Better Learn Environment View Presentation Vladimir Dragalin, Wyeth Research; Cyrus Hoseyni, Wyeth Research


GS4 Seeking a Gold Standard Method for Non-inferiority Trials		2:45 PM - 4:00 PM Salons I, II & III
Organizer(s): Qi Jiang, Amgen; Junfang Li, Sanofi-Aventis; Daphne Lin, FDA; Tristan Massie, FDA

Non-inferiority trial may be appropriate for evaluation of the efficacy of an experimental treatment via a direct comparison with an effective active control when the use of a placebo would be considered unethical. Non-inferiority trials are difficult to design and analyze. Many issues need to be carefully considered, such as how to select the active control, how to address the assay sensitivity and constancy assumptions, how to document the effect size of the active control versus placebo, and how to establish the non-inferiority margin of interest. In this session, some emerging issues and the strength and weaknesses of existing methods will be discussed by regulatory and industry leaders and an academic expert in the field.
Statistical Inference Issues in Non-inferiority Methodology H.M. James Hung, FDA
The Case for the Synthesis Method View Presentation Steven Snapinn, Amgen
Commentary on Issues in Non-Inferiority Trials View Presentation Thomas Fleming, University of Washington

Key Dates

April 30 - May 22, 2013
Invited Abstract Submission Open

June 4, 2013
Online Registration Opens

August 9 - August 23, 2013
Invited Abstract Editing

August 23, 2013
Short Course materials due from Instructors

August 26, 2013
Housing Deadline

September 9, 2013
Cancellation Deadline and Registration Closes @ 11:59 pm EDT

September 16 - September 18, 2013
Marriott Wardman Park, Washington, DC