APOE-guided lecanemab dosing dissociates ARIA-E from ARIA-H: The K-ROAD study.

APOE ε4 allele dose drives the risk of amyloid-related imaging abnormality with edema (ARIA-E) under lecanemab; dose escalation here targets genotype-specific maximum plasma concentrations, whereas ARIA with hemosiderin deposition (ARIA-H) tracks pre-existing cerebrovascular burden.

Five hundred eight patients with early Alzheimer's disease received APOE genotype-guided lecanemab escalation across eight Korean centres, with target maximum plasma concentrations of 118, 308 and 404 µg/mL for ε4/ε4 homozygotes, heterozygotes and non-carriers respectively, under MRI-based ARIA surveillance. ARIA-E incidence was 2.0% in non-carriers and 6.2% in ε4/ε4 homozygotes, an odds ratio of 3.25 versus 8.49 in Clarity AD and 8.32 in Japanese post-marketing surveillance. Any ARIA-H was independently predicted by baseline cerebral microbleeds, white matter hyperintensity, age and ε4 allele dose (adjusted odds ratios 1.23, 1.50, 1.04 and 1.75). The authors state that these observations are compatible with — but do not establish — attenuation of the ε4-driven ARIA-E gradient, and propose a dual-axis safety framework that still requires randomized confirmation.

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

The strength lies less in genotype-guided escalation than in the dissociation of the two ARIA types: one axis driven by genotype and exposure, one vascular axis driven by the state of the parenchyma before treatment. Comparison with Clarity AD and Japanese pharmacovigilance remains indirect, with no internal control arm: 3.25 versus 8.49 is a hypothesis, and the authors are honest enough not to conclude. In practice this mainly argues for genotyping APOE before lecanemab and for reading the baseline MRI — microbleeds and white matter burden — as a decision criterion distinct from genotype.

Analysis by Dr Thibaut Benquey