SAMPLE SIZES FOR 24-MONTH ALZHEIMER’S PREVENTION TRIALS USING BIOMARKER ENDPOINTS IN COGNITIVELY UNIMPAIRED AMYLOID-POSITIVE ADULTS

Eric M. Reiman,Xiaoying Kuang,Ji Luo,Yinghua Chen,Pradeep Thiyyagura,Robert Bauer,Wendy Lee,Hillary Protas,Dhruman D. Goradia,Paul S. Aisen,Keith Johnson,Jessica B. Langbaum,Reisa A. Sperling,Pierre Tariot,William J. Jagust,Michael Weiner,Kewei Chen
DOI: https://doi.org/10.1016/j.jalz.2017.06.2313
2017-01-01
Abstract:In the next few years, several Alzheimer's disease (AD) trials have a chance to clarify which biomarker endpoints are reasonably likely to predict a treatment's clinical benefit, such that those biomarker endpoints could help to find and support the accelerated approval of effective AD prevention therapies in cognitively unimpaired at-risk persons as soon as possible. We used data from the AD Neuroimaging Initiative (ADNI) to estimate the number of unimpaired amyloid-β (Aβ)-positive older adults, including those with or without at least one APOE4 allele, needed to evaluate an AD prevention therapy in 24-month placebo-controlled trials using several biomarker endpoints. Baseline and approximately 24-mo follow-up florbetapir PET, FDG PET and structural MRI data were used to estimate the respective number of cognitively unimpaired 65–80 year-old Aβ-positive trial completers, including those with or without at least one APOE4 allele, needed to detect a 25% treatment effect in 24-mo AD prevention trials with 80% power and two-tailed P=0.05. Aβ positivity was defined as a baseline florbetapir standard uptake value ratio (SUVR) ≥1.18, PiB SUVR ≥ 1.47, or CSF Aβ42 level ≤ 192 pg/mL. We estimate that approximately 159, 183, 273, and 820 Aβ-positive trial completers per group, 94, 128, 254, and 457 Aβ-positive APOE4 carrier completers per group, or 190, 247, 286, and 1,483 Aβ-positive APOE4 non-carrier completers per group are needed for the respective detection of 25% treatment effects on cerebral-to-white matter florbetapir SUVR increases (in the absence of SUVR-reversing effects), Iterative Principal Component Analysis (IPCA)-based whole brain volume declines, FreeSurfer-based hippocampal volumes declines, and FDG PET declines in an AD-related statistical region-of-interest (sROI) in 24-month placebo-controlled AD prevention trials. We have begun to estimate the number of cognitively unimpaired at-risk participants needed to evaluate AD prevention therapies in a 24-month trial using biomarker endpoints. These trials may provide the best chance to find and support the approval and widespread availability of effective AD prevention therapies by 2025. Planning for a potential trial is underway.
What problem does this paper attempt to address?