Chair: Robert Barrier, PhD (Parexel)
Speaker: Nikki Freeman, PhD (Duke University)
Title: Aspirations Meet Reporting: Results from a Scoping Review of Sequential Multiple Assignment Randomized Trials (SMARTs) from 2009-2023
Abstract: Sequential
multiple assignment randomized trials (SMARTs) are the “gold standard” for generating data for precision medicine evidence. They are increasingly being used in a wide range of therapeutic areas and care settings. SMART designs can be leveraged for both hypothesis testing and hypothesis generation. Thus, the primary and secondary trial objectives may vary from one SMART to another. This, along with the design’s inherent complexity, complicates how the analysis of these kinds of trials is communicated and reported to the scientific community. In this presentation, we will discuss the questions that SMARTs can answer and why different study teams might pursue different objectives with this type of design. We will also present results from a scoping review of SMARTs, focusing on trial objectives and how results were reported in the literature.
Speaker: Beibo Zhao, PhD (University of North Carolina at Chapel Hill)
Title: Pragmatic Solutions to Precision Medicine Trial Challenges: A Trialist’s Highlights From the Design and Implementation of the BEST Trial, a Multi-Site Sequential Multiple Assignment Randomized Trial (SMART)
Abstract: The Biomarkers for Evaluating Spine Treatments (BEST) trial was a multi-site sequential multiple assignment randomized trial (SMART) investigating four established interventions to inform a precision medicine approach for treating chronic low back pain (cLBP). As the BEST trial team prepares to release the top-line results, we discuss key challenges and lessons learned from the trial’s design and implementation: (1) operationalizing study objectives through estimating and evaluating comprehensive and simplified (but more readily interpretable and clinically implementable) versions of optimal dynamic treatment regimes (DTRs), (2) considerations that guided trial design and analysis population definitions, and (3) reporting framework for aligning scientific questions with operational definitions. Throughout, we share our decision-making process and commitment to rigorous yet pragmatic solutions.
Speaker: Anastasi Ivanova, PhD (University of North Carolina at Chapel Hill)
Title: PrecISE, a Biomarker-Stratified Trial in Severe Asthma
Abstract: PrecISE (Clinical Trials.gov NCT04129931) is an NHLBI sponsored clinical trial to screen five novel interventions that might work in severe asthma. Enrollment in each intervention is stratified by a biomarker-defined subgroup with a potential enrichment after the interim analysis if the biomarker negative group is stopped for futility. We discuss biomarker-stratified trials with enrichment with PrecISE as an example.
Speaker: Inna Perevozskaya, PhD (Bristol Myers-Squibb)
Title: Getting the Right Data for Statistical Designs Involving Biomarkers: A Case Study
Abstract: Recent developments in drug development and reimbursement policy landscapes are calling for more personalized treatments. As a result, there is growing need to incorporate biomarker data into clinical trials to improve personalized treatment recommendations. While many trials routinely collect biomarker data, its often done on an ad-hoc basis without formal statistical considerations (unless biomarker is a primary endpoint or used for stratification). There is a need for more comprehensive systematic approach to biomarker collection, so that the data we collect in clinical trials today can be used to answer the research questions of tomorrow in a statistically rigorous way. In this talk, we present preliminary findings of statistical and science collaborative effort to develop a “smart” strategy for biomarker collection to enable greater utilization of biomarker-driven statistical designs, while keeping cost of additional sampling under control.
