A study of postmortem human brain tissue published in Nature Medicine on June 4, 2026 reports that the point at which amyloid-driven inflammation gives way to tau-associated cellular programs is marked by a change in the state of microglia, the brain's resident immune cells, and that people who kept their cognition despite carrying plaques did not follow the same course. The paper, with Ashley Lu as first author and Mark Fiers and Bart De Strooper of the VIB-KU Leuven Center for Brain and Disease Research and the UK Dementia Research Institute among the senior authors, alongside Niels Plath of Muna Therapeutics, is a descriptive study of donated brain tissue. It reports associations, not causes, and it tested no drug in anyone.
The team combined Visium spatial transcriptomics, single-nucleus RNA sequencing and Xenium in situ analysis on the superior frontal cortex of 24 octogenarian brains, matched for age, sex and APOE status and free of major non-Alzheimer's lesions, plus 20 centenarian brains from the Dutch 100-plus Study. According to Alzforum's report of the paper, the octogenarian group comprised eight cognitively normal donors without Alzheimer's pathology, eight cognitively normal donors with substantial amyloid, and eight with dementia and both plaques and tangles. The authors describe six tissue domains forming a pathological continuum, with a sharp switch between the third and fourth from an early inflammatory plaque-induced gene program to a late, antigen-presenting one. Octogenarians without dementia lacked the late program; centenarians showed the late program, but uncoupled from tau accumulation, which the authors read as two different routes to resilience.
The boundaries are wide. Every donor was studied once, after death, so the ordering of events over years is inferred from a static map rather than observed. Expert comment published with Alzforum's report cautioned that centenarian donors are selected for having stayed well, are enriched for protective genetic backgrounds and carry co-pathology such as cerebral amyloid angiopathy and TDP-43, any of which could shape the microglial response. Commercial interest is also in the frame: Muna Therapeutics co-authored and part-funded the work, and in its own announcement of the paper the company said the early program is enriched for TREM2 signalling, the target of its oral candidate MNA-001, which it said began Phase 1 testing in healthy adults in November 2025. The paper itself does not show that altering microglial state changes anyone's cognition.
The finding builds on the plaque-induced gene concept the same Leuven group introduced in 2020, and it is accompanied in Nature Medicine by a research briefing summarising the work for the journal's readers. For memory clinics, the practical question it speaks to is the familiar one of the patient with amyloid on a scan and an intact mind. This work argues that such resilience may be an active cellular program rather than luck, but it offers no test, no marker and no treatment that a clinician can use today.