Neuropathological findings after aducanumab treatment in Alzheimer's disease
Baayla D C Boon, MD PhD; Yoav D Piura, MD; Christina M Moloney, PhD; Jessica L Chalk, BS; Sarah J Lincoln, BS; Matthew H Rutledge, PhD; et al.
November 1, 2025
- The Lancet Neurology
- The Lancet Neurology, Volume 24, Issue 11, Pages 931–944
- 10.1016/S1474-4422(25)00313-8
Evidence before this study
We searched PubMed, bioRxiv, and medRxiv for articles published in English from database inception to March 28, 2025, using the terms “neuropathology,” and [“amyloid immunotherapy” OR “monoclonal antibody” OR “anti-amyloid” OR “aducanumab” OR “lecanemab” OR “donanemab” OR “AN1792” OR “gantenerumab”]. We selected original studies that assessed brain tissue of individuals with Alzheimer's disease who had received monoclonal anti-amyloid β (Aβ) therapy. This search showed that although multiple case studies describe neuropathological findings following IgG-based Aβ immunotherapy—including aducanumab, lecanemab, and AN1792—systematic case–control comparisons are rare. Most notably, the AN1792 active immunisation trial included a comparative analysis of immunised versus non-immunised participants. However, no previous study has comprehensively examined Aβ isoforms, plaque morphology, tau pathology, cerebrovascular markers, and neuroinflammation markers including complement, macrophage, and microglial activation across matched treated and untreated participants with Alzheimer's disease using multiplex immunofluorescence and digital pathology. Moreover, previous reports on amyloid-related imaging abnormalities (ARIA)-related neuropathology were limited to isolated case descriptions.
Added value of this study
To our knowledge, this study presents the first systematic case–control neuropathological comparison between aducanumab-treated and Aβ-targeting therapy-naive participants with Alzheimer's disease, including detailed characterisation of Aβ isoform-specific patterns, plaque clearance, and ARIA-associated changes. We show that aducanumab treatment results in disproportionate clearance of Aβaa1–8 and Aβ42 in cortical layer I, with no significant changes in deeper layers. In participants who developed ARIA, we identified localised superficial siderosis, haemosiderin-laden microinfarcts, complement activation, and CD68-positive vessel wall changes along Aβ-laden leptomeningeal and penetrating vessels. Furthermore, we observed a potential discrepancy between Aβ PET and pathology with substantial clearance on PET but persistent postmortem Aβ pathology in deeper layers. These findings advance our understanding of the spatial distribution and downstream effects of monoclonal antibody-mediated Aβ clearance in the human brain.
Implications of all the available evidence
Building on previous studies, our findings suggest that IgG-based anti-Aβ therapies preferentially clear Aβ from superficial cortical layers, likely due to limited antibody penetration beyond subarachnoid and perivascular spaces. This spatially restricted clearance pattern has important implications for both therapeutic efficacy and interpretation of imaging biomarkers. The distinctive neuropathological features of ARIA—such as microinfarcts, haemosiderin deposition, and vascular inflammation—point to a mechanistic overlap with cerebral amyloid angiopathy-related inflammation. As next-generation monoclonal antibodies are engineered for improved blood–brain barrier transport, a detailed understanding of Aβ clearance dynamics and vascular side effects will be essential to optimise safety and clinical outcomes. Our findings underscore the critical role of postmortem analysis in validating and refining in-vivo biomarkers, which currently lack the spatial resolution to detect these cortical layer subtleties. Furthermore, the presence of ARIA-related microinfarcts despite radiological resolution emphasises the need for long-term clinical surveillance in treated individuals.
Source: The Lancet Neurology