Design, Data Monitoring & Analysis of Clinical Trials with Multiple Outcomes
Instructors: Dr. Toshimitsu Hamasaki, National Cerebral and Cardiovascular Center
Dr. H.M. James Hung, US Food and Drug Administration
Moderator: Alfred Balch, University of Utah
Traditionally, in confirmatory clinical trials, a single clinical outcome is selected as a primary endpoint. This endpoint serves as the basis for the trial designs including sample size determination, interim data monitoring, final analyses and the reporting of the trial results. The primary endpoint should be an outcome which can provide the most clinically relevant measure to address the primary objective of a trial. However, a single primary endpoint may not provide a comprehensive picture of the important effects of the intervention as the diseases are caused by multi-factors such as genetic, environmental, lifestyle and other factors, and the disease may have different and interdependent outcomes. For this reason, clinical trials may still be designed with either a single primary endpoint plus key secondary endpoints or more than one primary endpoint.
A single primary endpoint plus key secondary endpoints are commonly analyzed using a pre-specified multiple comparison procedure to control the study-wise Type I error probability. This is a well-known union-intersection problem. There remain many statistical issues with analysis of a single primary endpoint plus key secondary endpoints. In contrast, multiple primary endpoints may offer an attractive design feature as they could capture a more complete characterization of the effect of an intervention and provide more informative intervention comparisons. However multiple primary endpoints also create challenges in design, interim monitoring, analysis and reporting of clinical trials. When evaluating the interventions’ effects with multiple primary endpoints, we should distinguish decision-making frameworks based upon whether it is desirable to evaluate if there are effects on AT LEAST ONE of the endpoints (multiple primary endpoints: MPE) or whether there are effects on ALL of the endpoints (co-primary endpoints: CPE). Thus, a single primary endpoint plus key secondary endpoints may be a special case of MPE. This decision defines the alternative hypothesis to be tested and provides a framework for approaching trial design. Although CPE is a special case of MPE, it is important to recognize their differences in controlling Type I and Type II errors.
This two-day short course will begin with discussion of some challenging statistical issues with analysis of a single primary endpoint plus key secondary endpoints, and then discuss statistical challenges created by multiple endpoints, especially associated with MPE and MCP. The short course will
- provide an overview of the concepts, principle and procedures of MPE and MCP
- discuss challenging problems with analysis of a single primary endpoint plus key secondary endpoints
- review relevant regulatory guidelines and related documents
- integrate recent methodological developments for design, monitoring and analysis of clinical trials with MPE and MCP
- include serval case studies from actual clinical trials to help attendees quickly understand common methods and apply them to problems in real world.
- demonstrate how to implement the methods using popular statistical software including SAS and R
Session 1: An overview: Statistical issues in clinical trials with multiple endpoints
Session 2: Methods for design and analysis of clinical trials with MPE: Part 1
Session 3: Methods for design and analysis of clinical trials with MPE: Part 2
Session 4: Software implementation: MCP
Session 5: Q and A
Session 6: Methods for design and analysis of clinical trials with CPE: Part 1
Session 7: Methods for design and analysis of clinical trials with CPE: Part 2
Session 8: Software implementation: CPE
Session 9: Further issues
Session 10: Q and A
Toshimitsu Hamasaki is the Director of Data Science at National Cerebral and Cardiovascular Center, Osaka, Japan. He has been involved in biopharmaceutical statistics for over 25 years, and prior to joining NCVC, worked at Shiogoni, Pfizer Japan and Osaka University. He has been actively involved in biostatistical research, and is the author of more than 180 peer-reviewed publications and three textbooks on clinical trials including Group-Sequential Clinical Trials with Multiple Co-Objectives. Dr. Hamasaki was the member of ICH E5 Guideline Implementation Working Group as a representative of Japan Pharmaceutical Manufacturers Association to develop the Q & A document on the guideline. He currently serves as an Associate Editor for Statistics in Biopharmaceutical Research, Journal of Biopharmaceutical Statistics and Japanese Journal of Statistics and Data Science, and Editor for CHANCE. He is an elected member of International Statistical Institute and a Fellow of the American Statistical Association. He is a recipient of the Japanese Society of Computational Statistics Distinguished Article Award and Behaviormetric Society of Japan Hida-Mizuno Prize, and the Poster Competition Winner at the ASA Biopharmaceutical Section Regulatory-Industry Statistics Workshop.
Dr. H.M. James Hung is presently Director of Division of Biometrics I, Office of Biostatistics, Office of Translational Sciences, Center for Drug Evaluation and Research (CDER), U.S. Food and Drug Administration (FDA). The Division provides statistical services supporting Division of Cardiovascular and Renal Products, Division of Neurology Products, Division of Psychiatry Products, and Division of Medical Imaging Products, CDER, FDA.
During his tenure with FDA, Dr. Hung has reviewed many large mortality/morbidity clinical trials in cardiovascular and renal disease areas. He published many peer reviewed articles in Biometrics, Statistics in Medicine, Controlled Clinical Trials, Biometrical Journal, Journal of Biopharmaceutical Statistics, and Pharmaceutical Statistics. His research areas include multi-regional clinical trials, adaptive design/analysis, non-inferiority trials, factorial design clinical trials, and utility of p-value distribution. He delivered a great many of invited talks, lectures or short courses in many countries.
Dr. Hung received two FDA/CDER Scientific Achievement Awards and many other awards for the recognition of his scientific contributions to the US FDA. He is the recipient of 2011 FDA Scientific Achievement Award – Excellence of Analytic Science. He also received awards for recognizing his contributions to the FDA guidance documents on adaptive designs and non-inferiority trial designs.
Dr. Hung served as an Editor-in-Chief for Journal of Pharmaceutical Statistics in 2009-2011. Currently, he serves as an Associate Editor for Statistics in Medicine. He is a Fellow of the American Statistical Association and an elected member of the International Statistical Institute.
Hamasaki, T., Evans, S.R.Asakura, K., and Ochiai, T. Group-Sequential Clinical Trials with Multiple Co-Objectives, 2016, 110 Pages, ISBN: 978-4-431-55900-9 54.99 36
Sozu, T., Sugimoto, T., Hamasaki, T., and Evans, S.R. Sample Size Determination in Clinical Trials with Multiple Endpoints, 2015, 80 Pages, ISBN: 978-3-319-22005-5