Real-World Challenges in the MRI Detection and Classification of Amyloid-Related Imaging Abnormalities.
TL;DR
Key issues addressed include recognizing technical factors that affect interpretation of susceptibility-weighted and FLAIR images, managing variability in lesion detection across serial examinations, and accurately reporting the temporal evolution of ARIA-related findings.
Abstract
Anti-amyloid monoclonal antibody therapies (AATs) are increasingly used for the treatment of early symptomatic Alzheimer disease. Safe implementation of AATs relies on accurate assessment for exclusionary findings on baseline brain MRI and reliable longitudinal detection of amyloid-related imaging abnormalities (ARIA), including ARIA-H (hemosiderin or hemorrhage) and ARIA-E (edema or effusion). In routine clinical practice, MRI findings encountered during baseline screening and ARIA surveillance sometimes fall into borderline or ambiguous categories that do not clearly conform to trial-defined criteria. For example, subtle or artifactual FLAIR signal abnormality, equivocal microhemorrhage counts, and overlapping features of ARIA-E and ARIA-H can create uncertainty with direct implications for treatment management. Given limited practical guidance addressing these and other gray zones, this imaging-focused review synthesizes common interpretive challenges encountered in a high-volume AAT program and offers practical management-oriented recommendations. Key issues addressed include recognizing technical factors that affect interpretation of susceptibility-weighted and FLAIR images, managing variability in lesion detection across serial examinations, and accurately reporting the temporal evolution of ARIA-related findings. As AAT use expands, recognition of the presented pitfalls can improve diagnostic confidence and help avoid misclassification of findings that may influence therapeutic decisions.