Residualized occipital amyloid burden and plasma biomarker profiles in relation to ARIA-E during anti-amyloid treatment.

INTRODUCTION

Amyloid-related imaging abnormalities with edema (ARIA-E) represent the main safety concern of anti-amyloid monoclonal antibodies. Whether regional amyloid topology refines ARIA-E risk in real-world practice remains unclear.

METHODS

We analyzed 31 treatment courses in early Alzheimer's disease under European Medicines Agency labeling. Residualized occipital amyloid burden was defined as observed occipital standardized uptake value ratio (SUVR) minus occipital SUVR predicted from global composite SUVR in an expanded baseline positron emission tomography derivation sample (n = 53). Associations with ARIA-E and baseline plasma biomarkers were assessed.

RESULTS

ARIA-E occurred only in the elevated occipital amyloid group: 4/13 treatment courses, versus 0/18 low or comparable courses (Fisher exact p = 0.055). All ARIA-E events involved posterior regions. The elevated pattern showed higher baseline glial fibrillary acidic protein (GFAP), phosphorylated tau (p-tau)181, and p-tau217 levels, whereas amyloid beta 42/40 did not differ between groups. A GFAP + p-tau217 model showed exploratory discrimination of the elevated phenotype (area under the curve: 0.76).

DISCUSSION: Disproportionate occipital amyloid burden may identify a biologically vulnerable subgroup requiring validation in larger cohorts.