For anyone tracking Alzheimer's therapeutics, the central unresolved question has never been whether anti-amyloid antibodies can remove plaques — it's whether plaque removal actually translates into preserved neural architecture and slowed disease progression. Postmortem tissue analysis now offers a rare, direct window into what aducanumab actually does inside the human brain after treatment ends.
Examining medial temporal lobe tissue from six aducanumab-treated participants in clinical trials — individuals who died between seven weeks and five years after their final infusion — and comparing them with nine age-, APOE genotype-, and Braak stage-matched untreated Alzheimer's donors, researchers documented a robust reduction in amyloid-beta burden in treated brains. Neuritic phospho-tau also declined in concert with fewer plaques. However, the density of neurofibrillary tangles, marked by PHF-1 and AT8 immunostaining, was statistically indistinguishable from untreated controls. Microglial and astroglial reactivity measures were similarly comparable between groups. A particularly notable finding: amyloid load appeared to increase proportionally with time elapsed since the last dose, consistent with gradual plaque redeposition after cessation of therapy.
This autopsy-based evidence places aducanumab's mechanism in sharper relief and sharpens a critical debate in the field. The drug's accelerated withdrawal from the U.S. market in 2024 — amid controversy over its clinical approval and modest cognitive benefit data — makes these tissue-level findings all the more consequential for interpreting the entire anti-amyloid class, including lecanemab and donanemab. The dissociation between amyloid clearance and tangle burden is telling: if neurofibrillary tau pathology is the more proximate driver of neuronal death and cognitive loss, amyloid removal alone may be insufficient to halt downstream neurodegeneration. The redeposition signal also implies continuous or maintenance dosing may be biologically necessary rather than optional. Limitations are substantial — only six treated cases, no direct cognitive outcome data linked to the tissue, and survival bias inherent to any autopsy cohort. Still, this work is incrementally paradigm-clarifying: it confirms the amyloid hypothesis's mechanistic premise while simultaneously illuminating why plaque clearance has not yet translated into transformative clinical outcomes.