Data Monitoring Models and Adaptive Designs: Some Regulatory Experiences Sue-Jane Wang, Ph.D....
-
Upload
heidi-aldous -
Category
Documents
-
view
220 -
download
1
Transcript of Data Monitoring Models and Adaptive Designs: Some Regulatory Experiences Sue-Jane Wang, Ph.D....
Data Monitoring Models and Adaptive Designs: Some Regulatory Experiences
Sue-Jane Wang, Ph.D.
Associate Director for Adaptive Design and Pharmacogenomics, Office of BiostatisticsOffice of Translational Sciences, CDER,U.S. Food and Drug Administration
Presented at DIA China in Data Monitoring and Adaptive Designs: Is China Ready? May 17, 2011 | Beijin, China
Wang SJ, May 17, 2011, DIAChina 2
The views and opinions expressed in the following PowerPoint slides are those of the individual presenter and should not be attributed to Drug Information Association, Inc. (“DIA”), its directors, officers, employees, volunteers, members, chapters, councils, Special Interest Area Communities or affiliates, or any organization with which the presenter is employed or affiliated.
These PowerPoint slides are the intellectual property of the individual presenter and are protected under the copyright laws of the United States of America and other countries. Used by permission only. All rights reserved. The professional views presented here are those of the author and not necessarily those of the U.S. Food and Drug Administration. Drug Information Association, DIA and DIA logo are registered trademarks or trademarks of Drug Information Association Inc. All other trademarks are the property of their respective owners.
Disclaimers
Wang SJ, May 17, 2011, DIAChina 3
OUTLINE
3
• Adaptive Design Consideration
• Monitoring trial conduct / trial integrity
• Monitoring adaptive process
• Concluding Remarks
Wang SJ, May 17, 2011, DIAChina 4
Adaptive Design• Prospectively planned opportunity for modification
of one or more specified aspects of the study design and hypotheses based on analysis of data (usually interim data) from subjects in the study
• Analysis of the accumulating study data are performed at prospectively planned time-points within the study
• Analyses can be performed in a fully blinded manner or in an unblinded manner, and can occur with or without formal statistical hypothesis testing
Wang SJ, May 17, 2011, DIAChina 5
Adaptive Designed Clinical Trials
5
• Early Stages
• Early tolerability
• Early drug activity
• Early safety/tolerability signal
Wang SJ, May 17, 2011, DIAChina 6
Adaptive Designed Clinical Trials
6
• Mid to Later Stages
• Dose response
• Short term trial
• Long term trial
Wang SJ, May 17, 2011, DIAChina 7
Over-arching principles“The well-being of trial participants takes precedence over societal interests” – The World Medical Association Declaration of Helsinki
Wang SJ, May 17, 2011, DIAChina 8
Adaptive Designed Clinical Trials
8
• How to maintain
• Validity of the trial results
• Integrity of the trial being implemented
• Applicable to both fixed design trial and adaptive design trial
• More complex with adaptive design trial
Wang SJ, May 17, 2011, DIAChina 9
Adaptive Design – Validity
– Pre-specification of possible modification of design elements
– Statistical methods that yield a valid false positive error control for regulatory consideration (ICH E9)
– A valid point estimate and CI that is consistent with the proposed test method in assessment of treatment effect
Some principles
Wang SJ, May 17, 2011, DIAChina 10
In addition to overall false positive-error control, forscientifically interpretable AD results, potential sourcesof bias need to be minimized by maintaining trialintegrity on – Trial conduct– Trial logistics– Trial operation
* May consider sub-structure of DMC for AD* Plan double-blinded study whenever possible
Adaptive Design – Integrity
Wang SJ, May 17, 2011, DIAChina 11
Data Monitoring Committee (DMC) Current Practice: A Formal Committee in GSD
Safety monitoring required by a formal committee WHEN Safety concerns are unusually high In general, study is large, of long duration, multi-center
Is DMC review practical in small trials, short term trials ? Will a DMC help assure the scientific validity of trial ?
Who are unblinded parties Who makes changes in an ongoing trial if there is a need Will changes impair the credibility of study results?
Wang SJ, May 17, 2011, DIAChina 12
External changes may prompt an interest in modifying some aspects of the trial as it progresses
Sometimes accumulating data from within the trial may suggest the need for modification
Recommendations to change the inclusion criteria, the trial endpoints, or the size of the trial are best made by those w/o knowledge of the accumulating data
Even aggregate data on safety and efficacy may be informative, these data may be needed for some trial management functions
The DMC Has Been a GSD Framework
“Possibility of Adaptations” by Blinded Party
Wang SJ, May 17, 2011, DIAChina 13
Setup firewalls to ensure those proposing interim protocol changes based on external data are insulated from knowledge of interim comparative results
Regulatory staff generally remain blinded to interim unblined results
Unblinded interim dataInterim Protocol Changes
Wang SJ, May 17, 2011, DIAChina 14
Adaptive Design of Wider Interests Study hypotheses may be a moving target May use biomarker for adaptive selection Interim analysis may be blinded or unblinded Adapt mainly from internal accumulating study
data that are pre-specified Often safety at DMC’s discretion May drop for partial futility or terminate for full
futility Seamless between stages
Wang SJ, May 17, 2011, DIAChina 15
ObjectivityIndependence
Wang et al. (2011, European Neuropsychopharmacology Journal)
Wang SJ, May 17, 2011, DIAChina 16
Key components for a successful Adaptive Design Trial
Wang SJ, May 17, 2011, DIAChina 17
Challenges with Adaptive Monitoring Balancing Adaptive Rules Against Safety Monitoring Adaptive Rules
Allocation, sampling, stopping, decision Pre-specified with limited flexibility in confirmatory
trials Chosen rules may not be ‘medically or scientifically
sound’ This is of less concern in exploratory trials
Safety Monitoring Control of objectivity with adapting multiple (global)
nulls to just a few or single nulls for final analysis When DMC overrules
Concluding Remarks Prospective adaptive scenario planning to mimic the
possible adaptive monitoring scenarios An established infrastructure for adaptive monitoring
and/or adaptive decision making Documentation on “How adaptive designs were
implemented in real time including the resulting trial conduct, interim analyses, changes, etc.” could help address the issues on biases due to adaptation
Legal consequence of confidentiality agreement ? More research and experiences are needed
Wang SJ, May 17, 2011, DIAChina 18