The Anti-Amyloid Treatment in Asymptomatic Alzheimer’s Disease (A4) and Longitudinal Evaluation of Amyloid Risk and Neurodegeneration (LEARN) studies were large, multicenter investigations designed to characterize preclinical Alzheimer’s disease (AD) and evaluate whether solanezumab could slow cognitive decline in amyloid‑positive, cognitively unimpaired older adults. Both studies implemented a comprehensive, multimodal neuroimaging program aligned with NIH data‑sharing requirements and the principles of the Collaboration for Alzheimer’s Prevention, including staged dissemination of baseline and final longitudinal imaging datasets. Participants underwent longitudinal structural MRI, resting‑state fMRI, florbetapir amyloid PET, and flortaucipir tau PET assessments over follow‑up periods extending to 240 weeks, with additional data collected during an open‑label extension. Imaging acquisition followed standardized, centrally qualified protocols across 68 international sites. MRI data underwent site qualification, automated and expert quality control, and volumetric processing using NeuroQuant. PET imaging was processed by a centralized imaging core using validated, 21 CFR Part 11-compliant pipelines incorporating motion correction, spatial normalization, standardized region‑of‑interest placement, and calculation of SUVs and SUVRs. For data sharing, all imaging data were de‑identified, converted to NIfTI format, organized in alignment with BIDS specifications, and de‑faced using mriᵣeface with expert review. Final datasets were disseminated via the LONI Imaging and Data Archive, Synapse, and a custom platform developed with Gates Ventures (a4studydata. org). Bibliometric analyses across three indexing platforms were conducted to assess scientific impact. The baseline release included 1, 771 MRI scans, 4, 468 amyloid PET scans, and 449 tau PET scans. The final longitudinal dataset comprises 5, 661 amyloid PET scans, 1, 576 tau PET scans, and extensive longitudinal MRI data, making A4/LEARN one of the largest preclinical AD neuroimaging resources available. As of May 2025, more than 2, 100 data access requests from over 60 countries had been submitted, with approval rates exceeding 90 percent and more than 1. 2 million imaging files downloaded. Over 90 publications have referenced A4/LEARN data, advancing biomarker characterization, risk stratification, disease modeling, and clinical trial design. These efforts demonstrate that large‑scale, timely sharing of clinical trial neuroimaging data is both feasible and highly impactful, even when primary trial endpoints are not met.
Jimenez-Maggiora et al. (Tue,) studied this question.