2006 Workshop Registrant Directory
SC 3 Adaptive Clinical Trials
PhRMA has sponsored a working group to evaluate and promote appropriate use of adaptive design in clinical trials. Another intent is to provide material to open discussions with regulators on the appropriate use of adaptive designs for consideration and future regulatory guidance. This group has now produced white papers and background slides on several topics related to adaptive design. This course will present four subtopics from this effort: 1) definition and classification, 2) dose selection and seamless design, 3) sample size reestimation and group sequential design, and 4) logistic/regulatory issues. Attendees should leave with a good understanding of many considerations related to the application of adaptive design and late stage development.
GS1 Statistics in the FDA and Industry: Past Present, and Future
In the centennial year of the FDA (the original Food and Drugs Act of 1906), this session reflects on the application of regulation in response to public health issues, and how the recognition of new medical and societal needs has influenced the design, conduct, analysis and interpretation of clinical studies. Initially, manufacturers were required only to show the safety of their products in order to obtain FDA approval for marketing. Perhaps the most substantial influence came in 1962 when the Kefauver-Harris Drug Amendments were passed to ensure drug efficacy and greater drug safety. For the first time, drug manufacturers were required to demonstrate the effectiveness of their products before marketing them. Patient package inserts were first required in 1970, and rapid changes occurred during the 1980s and beyond based on the Belmont Report, the Orphan Drug Act, the Childhood Vaccine Act, provisions for expanded access to experimental drugs, accelerated review of products for life-threatening diseases, improved assessments of response as a function of gender, the Pediatric Rule, and the Pediatric Research Equity Act. The evolution of the regulatory environment as FDA responded to the many clinical research issues has provided a very fertile ground for the application of statistics and statistical thinking, and it has created the opportunity for an enormous number of statisticians to specialize in this area. The importance of statistical contributions and innovation will continue to grow in the future as product development becomes more complex and greater emphasis is placed on flexibility and efficiency. The perspectives of statisticians from FDA, industry and academia, as well as the importance and productivity of the interactions among these three groups will be presented and discussed during this session.
MB_PPF_FDA_Industry.ppt (97 KB)
Salsburg_FDA Industry talk.doc (39 KB)
FDA CT Future v4.ppt (156 KB)
Foulkes.ppt (6 MB)
GS2 Flexibility in Clinical Trials: How Do We Deal With It?
The Adaptive Clinical Trial has emerged as a potentially powerful tool to help improve the efficiency and success rate of the drug development process. Recent advances in statistical methodology now permit the modification of specific aspects of the design of an ongoing trial while preserving the overall validity of the analyses. Pharmaceutical companies are planning for the broader implementation of adaptive clinical trials. This session will discuss lessons learned from our experiences implementing these trials. We will review the steps required for statisticians within pharmaceutical companies to design, review, and analyze flexible clinical trials. We will also discuss the need for simulations in the development of adaptive trial protocols, issues and concerns from the FDA in evaluating clinical trials with adaptive designs, as well as problems and their solutions encountered during the adaptive trial process.
GS3 - Surrogate Endpoints and Accelerated Approval
Clinical trials designed to definitively assess improvements in clinically meaningful endpoints will typically require enrolling many patients and then subsequently following them for a long time. Hence, there has been great interest in developing strategies for reducing the time and cost of drug development. One such strategy is the use of surrogate endpoints either in proof-of-concept trials or in label-enabling trials. The validation of surrogate endpoints has been studied relatively intensively in the literature (eg, Prentice 1989; Freedman, Graubard, and Schatzkin 1992; Fleming and de Mets 1996). Efforts have been made to converge to a common framework, encompassing the wide variety of settings one can encounter. Recent developments including variance reduction factors, sample size assessment methodologies, and surrogate threshold effects will be discussed. Examples of drug development programs that incorporate the use of biomarkers and surrogate endpoints to accelerate the evaluation of new drugs will also be discussed. These examples include: (1) use of a biomarker implicated directly in the purported mechanism of action of the candidate drug to form nested hypotheses to demonstrate efficacy, and (2) planned analyses of a surrogate endpoint for accelerated approval (e.g., progression-free survival in certain oncology settings) with a subsequent planned and appropriately powered analysis on a definitive endpoint (e.g., overall survival) for full approval. Technical, logistical/tactical, and regulatory considerations will be elaborated on. Finally, regulatory perspectives on the use of surrogate endpoints will be discussed.
GS4 Interpreting Subgroups for Regulatory Purposes
A single primary analysis population, often the intention-treat-population, is typically pre-specified for label-enabling clinical studies. However, at the time of analysis, many subgroup analyses are performed and the information they provide can impact regulatory decisions. In this session, the speakers will review many issues/controversies surrounding the interpretation of subgroups and review some case studies. The session will close with panel Q&A.
GS5 Data Monitoring Committees: Getting a New Perspective on an Old Issue
Data Monitoring Committees (sometimes known as Data and Safety Monitoring Boards) provide the vital service of objectively representing the patients' interest in clinical trials. However, this responsibility often raises tough ethical and statistical questions that concern industry, regulatory, and academic statisticians. In this session, the speakers will present these issues in detail, drawing on their personal experiences. We will also discuss evolving approaches for dealing with these issues in the area of adaptive designs. In a discussion after the presentations, Susan Ellenberg will provide her views on points made by the speakers.
PS01 The Role of the Statistician in Post-Marketing, Including Surveillance
The statistician's role in clinical development in industry and regulatory settings is generally well understood. However, when it comes to post-marketing surveillance or the medical device industry, it is unclear what role industry and FDA statisticians play. Recent issues with approved pharmaceuticals and medical devices highlight the importance of the statistician's role as part of a multi-disciplinary team in assessing post-market data for signal/noise inference. Risk/benefit and health outcomes data are becoming increasingly important components in determining public health/reimbursement policy. This session will provide some perspectives on the role of statistics in post-marketing pharmaceutical and medical device trials and will challenge statisticians to broaden their sphere of influence in post-marketing surveillance issues.
PS02 - Biomarker Analysis
Biological marker or biomarker, as more commonly known, refers to a variety of physiologic, pathologic, or anatomic measurements that are thought to relate to some aspect of normal or pathological biologic processes (Temple 1995; Lesko and Atkinson 2001). In recent years, it has received wide interest as a crucial element of the FDA Critical Path initiative. The session will cover the statistical basis of biomarker theory - why it is considered in clinical trials, its advantages and pitfalls, and situations in which use of a biomarker should be considered (or not). In particular, validation of a biomarker and its escalation potential to a surrogate marker will be of special interest. Issues such as biomarkers for patient selection as well as endpoint evaluation will be covered. Experiences in pharmaceutical industry as well as regulatory issues will be discussed with case examples. At the conclusion of this session, the attendees are expected to have a broad-based theoretical introduction to biomarker use, practical and logistic knowledge of issues in using a biomarker in a clinical trial, and an orientation to the regulatory implications of biomarker use.
PS03 Bridging Studies, Migration Studies, and Related Topics
In 1998, the ICH published its E5 Guideline for Industry on 'Ethnic Factors in the Acceptability of Foreign Clinical Data.' Since then, about 20 pharmaceutical products were approved in Japan using bridging strategies. In the US, use of bridging has been encountered in the submission of preventive vaccines and the extension of approved products to the pediatric population. For example, in the clinical development of influenza vaccines, a population might not have been included in the clinical endpoint efficacy study. Immunogenicity bridging studies can then be conducted to compare the immune response observed in the clinical endpoint efficacy study to that elicited in other populations. E5 provides a guideline for using the bridging strategy to minimize duplication of clinical data and to allow for extrapolation of foreign clinical data to a new region. In practice, there are successful and failed cases as well as unsettled issues (clinical and statistical) when implementing the bridging strategy. In this session, the speakers will share their experiences, knowledge, and advances in this area.
PS04 - Statistical Issues in Medical Device Trials
Clinical trials that evaluate the safety and effectiveness of medical devices offer unique challenges to the clinical and statistical communities, as well as the medical device industry. These devices include drug eluting (coated) stents, cardiac pacing systems, heart valves, cardioverter defibrillators, cochlear implants, ophthalmic laser systems, total hip replacement systems, robotic surgical systems, left ventricular assist devices, transmyocardial revascularization lasers, and many other applications across all fields of medicine. Topics related to the design, analysis and interpretation of medical device clinical trials will be discussed in this workshop including: pooling of foreign and domestic data, the challenges of historical controls and the use of Objective Performance Criteria (OPC) in non-randomized studies, and the incorporation of prior information (Bayesian approaches). A general question and answer session will follow the presentations.
Breiter 2006 Workshop OPC for distribution.pdf (624 KB)
GHK_FDA_Industry_2006.ppt (115 KB)
Evaluating Poolability of Continuous and Binary Endpoints Across.ppt (115 KB)
Woodworth Slides for Discussant.ppt (42 KB)
PS05 - High Dimensional Expression Data: Consistency Across Platforms and Statistical Prediction Modeling
High-dimensional technologies for measuring gene and protein expression are being widely used in the quest to develop predictive biomarkers. Data quality and cross-platform consistency issues associated with these rapidly evolving technologies impact their ability and utility for predicting ultimate therapeutic responses. This session will address current efforts to characterize cross platform consistency and data quality, and examine their impact on the development of predictive statistical models. The utility will be illustrated by bridging the development from non-clinical discovery to clinical exploration.
PS06 - Advantages and Challenges of Bayesian Clinical Trials
The use of Bayesian statistics has conspicuously increased in clinical trial research because it is ideally suited to adapting to information that accrues during a trial, potentially resulting in smaller, shorter, and more informative trials and for allowing patients to receive better treatment during the course of a trial. In this session we will hear from experts in academia, industry and FDA how to design, analyze, and interpret a Bayesian clinical trial in various settings and how the resulting design can be used in the development of a novel health care product. We will hear the advantages and challenges of using Bayesian methodologies, followed by several case studies. We will also hear about the use of the Bayesian approach in the regulatory setting.
PS07 – Standards and Processes for Effective Communication with the FDA
Effective communication with the FDA is an essential part of the drug development and approval process. This session will present suggestions and examples of how to improve communication through the use of standards. The speakers will focus on improving communication with the FDA during drug development and on streamlining the processes involved in submission and review through the use of standards like pre-NDA meetings, CDISC and controlled computing environments.
PS08 - Diagnostic Medical Imaging
Diagnostic imaging is an increasingly vital tool for identifying and treating disease and, more recently, for monitoring effectiveness of therapy. Examples include digital and three-dimensional tomosynthesis mammography, computer-automated detection and diagnosis (CAD or CADx), and flourescence-based enhanced imaging techniques (e.g., fluorodeoxyyglucose-positron emission tomography or FDG-PET). In general, imaging modalities can present challenges in both study design and analysis. In this session, noted researchers in imaging will discuss approaches to evaluating the diagnostic performance of specific imaging modalities.
PS09 - Guidance and Standards for Diagnostic Devices
There are many aspects of diagnostic assays that need to be evaluated by manufacturers, the FDA, and laboratories. Fortunately, there are also guidance documents and standards that have been published through the joint efforts of these interested parties to assist in evaluating many of these characteristics. This session will address and highlight some of the guidances and standards that are available.
PS10 - FDA's Quality by Design Initiative
FDA's Quality by Design (QbD) initiative emphasizes building quality into a product during development rather than attempting to "test" quality into manufacturing of the product. In this session perspectives from both industry and FDA will be presented, discussing pilot programs, risk assessment, design space and the role of FDA. The QbD approach to pharmaceutical development relies heavily on statistical methodologies, but the roadmap to its implementation is still in progress. Clarification of objectives and chemistry review considerations must be understood to ensure the use of appropriate statistical approaches. The goal of this session is to heighten awareness of the potential benefits to both industry and FDA and to identify important issues for consideration, including Q&A from the audience.
PS11 - Smart Choices: Decision Analysis Approaches to Clinical Trials
Decision Analysis (DA) is methodology for defining and analyzing a decision problem and making optimal decisions. The basic idea is to delineate the decision problem by quantifying uncertainties and preferences, and then to solve the resulting optimality problem by finding the best possible decision. The use of decision analysis approaches has been gaining increasing importance in making clinical trial decisions for novel treatments. The use of this approach is becoming more and more relevant in the process of making stage-gate decisions during drug and devices development program. In this session we will hear from experts in the academia, industry and FDA about the possibilities of using DA in various clinical trial settings and in general how it can be used while developing a novel health care product. Case studies will illustrate the use of the approach. Finally, we will also hear the regulatory perspective on the use of DA.
PS12 - Use of Historical Control Data in the Development of Medical Products
This session will explore the use of historical controls in the development of medical products using case studies. The session will include a case study that uses the doubly robust methods described by Lunceford and Davidian. The second speaker will describe his company’s experience with historical controls in pivotal trials leading to registration. The third speaker will describe of the use of historical controls in a preclinical setting.
PS13 - Classifiers in Combination Rx/Dx Submissions
Recent advances in diagnostic technologies are enabling opportunities to revolutionize treatment decision making by identifying variability among patients in drug response, both in efficacy and in toxicity. Examples of FDA-approved diagnostic tests to indicate the eligibility of patients for particular therapies include the Dako immunohistochemistry (IHC) test for Her-2/neu overexpression to select metastatic breast cancer patients for treatment with Herceptin (trastuzumab), the Dako IHC test for epidermal growth factor receptor expression to select metastatic colorectal cancer patients for treatment with Erbitux (cetuximab), and the Roche Amplichip to classify a patient's rate of metabolism of therapeutics metabolized primarily by the gene products of two cytochrome P450 genes (CYP2D6 or CYP2C19). Pharmaceutical and medical device companies are partnering to submit in parallel new Dx classifiers to determine patient eligibility for new Rx therapies. FDA guidances relevant to the submission of such combination Rx/Dx products include the CDRH guidance Multiplex Tests for Heritable DNA Markers, the CDER guidance Pharmacogenomic Data Submissions, and the inter-center Drug-Diagnostic Co-Development Concept Paper. In this session, participants will discuss validation of Dx classifiers of combination Rx/Dx products in Phase III trials for simultaneously validating both the Rx therapy and the Dx classifier used to select it. Topics that will be discussed include study design challenges, regulatory schemes for analytical and clinical validation of biomarkers and for combination Rx/Dx products, and business costs and benefits to inclusion of putative predictive biomarkers in clinical trials of pharmaceuticals.
PS14 - Case Studies in Modeling and Simulation
Modeling and simulation continue to gain importance as efforts are made to streamline drug development. The areas of drug development incorporating these tools are as varied as the statistical methods needed to accomplish them. Two case studies will be presented to give examples of how modeling and simulation are used in different phases of development. The first has a large scope by dealing with optimizing clinical development as a whole and the second is more focused on a specific aspect of drug development by exploring relationships between pharmacokinetics and pharmacodynamics. A third speaker, yet to be identified, will describe the viewpoint of a reviewer of submissions involving modeling and simulation.
PS15 - Assessing Agreement
In method comparison studies, we are often interested in whether two methods can be used interchangeably or a new method can replace an existing method. Traditionally, method comparison has been evaluated through assessment of agreement between methods. Assessing agreement has been based on indices such as intraclass correlation coefficient (ICC) or concordance correlation coefficient (CCC) for continuous data and kappa index for categorical data. We note that interchangeability between methods is similar to bioequivalence between drugs in bioequivalence studies. In 2001 FDA guidance for industry, FDA adopted an individual equivalence criteria for assessing individual bioequivalence between a test drug and a reference drug. Therefore, it is of great interest to investigate how the traditional agreement indices are related to the individual bioequivalence criteria and whether new coefficients of agreement can be developed that are similar to individual bioequivalence criteria. We propose to assess individual agreement among methods between measurements by different methods. New coefficients of individual agreement (CIA) are proposed to link the individual bioequivalence criteria with existing agreement indices. A unified approach that includes CCC and kappa as special cases will be presented for multiple (k) raters (instrument/methods) each with multiple (m) readings. We hope that these recent advances will stimulate discussion that can lead to future guidance for industry that develops new method, device, assay, or instrument for clinical use.
PS16 - Rare Events Estimation Using Insurance Claims Databases
Insurance claims databases offer the ability to study very large numbers of subjects to assess safety, particularly for rare adverse events, although issues concerning the validity of covariates, confounding by indication and outcome ascertainment must be considered carefully. Related topics such as registries, prospective cohort studies to assess drug safety, will also be discussed.
Bob Ball.ppt (237 KB)
FDAtalk-cwc.ppt (164 KB)
Skovron FDA_Industry_statistics workshop 092906.ppt (274 KB)
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
September 9, 2013
Cancellation Deadline and Registration Closes @ 11:59 pm EDT
September 16 - September 18, 2013
Marriott Wardman Park, Washington, DC