PS48: Innovative Strategies for Longitudinal Data Analysis in Neurodegenerative Disease Trials: Lessons from Alzheimer's for Rare Diseases

Neurodegenerative disease trials face unique challenges due to the differing prevalence of these conditions. Alzheimer's disease (AD), as a common disease, has facilitated the development of robust trial methodologies, while rare diseases such as frontotemporal dementia (FTD), present significant hurdles in terms of recruitment and statistical power. Traditional efficacy evaluations, which compare changes from baseline (CFB) at the final study visit (e.g., 18 or 24 months) using a mixed model for repeated measures (MMRM), may not fully leverage the wealth of information available across earlier time points, unscheduled visits, or multiple endpoints. Recent advancements in AD clinical trials demonstrate consistent proportional treatment effects across well-established endpoints when standardized relative to placebo decline, as well as measurable benefits at earlier visits. These insights highlight opportunities to refine efficacy inference strategies, which could significantly improve trial efficiency and reduce sample size requirements. Such innovations are particularly crucial for rare neurodegenerative diseases like FTD, where smaller, more efficient trial designs are essential to feasibility. This session will discuss cutting-edge statistical methods that build upon lessons learned from AD trials, emphasizing their potential to enhance clinical trial designs for both common and rare neurodegenerative diseases. Topics will include: 1. Incorporating Multiple Visits: Leveraging treatment effects observed at multiple post-baseline time points using enhanced MMRM techniques. We will discuss how to incorporate unscheduled visits with a modified MMRM model. 2. Multivariate Analysis: Modeling multiple endpoints simultaneously to allow for global statistical inference. 3. Alternative Metrics: Exploring proportional treatment effects, reduction rates, and rank-based win probabilities to provide innovative perspectives on efficacy evaluation. The session will feature two speakers and a discussant from the FDA, followed by an audience Q&A. The first speaker, representing industry, will provide an overview of clinical trial results and alternative efficacy inference strategies. The second speaker, from academia, will explore the details of each strategy, including primary endpoints, statistical models, and their respective advantages and limitations. Each approach will be demonstrated using semi-real trial data to illustrate practical applications and implications for study design. The FDA discussant will offer insights and feedback on the presentations, emphasizing their relevance and feasibility. The session will conclude with a discussion on how these strategies can enhance efficiency and reduce sample size requirements, particularly in rare neurodegenerative disease trials like those for FTD.