A tau antibody drove a tangle marker down 90% — in seven people

At this month's Clinical Trials on Alzheimer's Disease meeting, held 1–4 December in San Diego, Kristin Wildsmith of Eisai presented phase 1b/2 data on etalanetug, an antibody targeting tau, and the biomarker numbers were striking [s1].

They were also generated in seven people [s1].

Both facts belong in the same sentence, and the second is the reason this is a report on a mechanism rather than a report on a treatment.

What was measured

The seven participants had mild to moderate dominantly inherited Alzheimer's disease — the rare genetic form in which a mutation makes the disease essentially certain and its timing relatively predictable [s1]. They were followed for two years on an ascending dose schedule, beginning at 750 mg intravenously every four weeks and escalating to 1,500 mg, 3,000 mg and 4,500 mg at weeks 12, 24 and 36 [s1].

The markers of interest were forms of MTBR-tau243, a fragment derived from the microtubule-binding region of tau. In cerebrospinal fluid, MTBR-tau243 fell by about 70% by nine months and was sustained through two years, ending at a 75% reduction [s1]. The extended form, eMTBR-tau243, fell by about 90% by nine months and was likewise sustained across two years [s1].

In plasma, eMTBR-tau243 fell 80% at three months and more than 90% by nine months [s1].

Why this fragment rather than total tau

The choice of marker is the scientifically interesting part. MTBR-tau243 is a fragment of the microtubule-binding region of the tau protein, and the conference coverage frames these species as tangle markers — that is, as readouts of aggregated tau rather than of tau protein generally [s1]. That framing is what makes the size of the reduction meaningful rather than merely large.

A drug that lowers this fragment by 90% and holds it there for two years is producing an effect that is hard to explain as noise, even in seven people. Effects of that magnitude in a small sample are more interpretable than small effects in a large one — the failure mode of tiny studies is spurious small differences, not spurious ninety-percent shifts sustained over 24 months.

What it does not show

That the drug helps anyone.

The trial reports biomarker change. Only three participants had PET imaging data available for assessment of tangle burden [s1], which means even the imaging confirmation of the fluid biomarkers rests on three scans.

There is no control arm described, no randomisation, and no cognitive or functional endpoint reported in the coverage of the presentation. Dominantly inherited Alzheimer's disease is a small population by definition, so phase 1b studies in it are necessarily tiny; that is an explanation for the design, not a repair of its limits.

The context in which this data arrived sharpens the point. At the same meeting, the evoke and evoke+ phase 3 trials in a combined 3,808 people showed that a drug can improve Alzheimer's-related biomarkers by up to approximately 10% and produce no measurable clinical benefit [s2]. The relationship between moving a marker and helping a patient is precisely the thing the field has not established, and a 90% reduction in a different marker in seven people does not establish it either.

What it does establish is target engagement: the antibody reaches the compartment and does the biochemical thing it was designed to do, durably. That is a necessary condition for a tau therapeutic to work. It is not a sufficient one.

The infrastructure story underneath

A quieter development at the same meeting may matter more in the medium term. Sysmex Corporation, Johnson & Johnson and Beckman Coulter each presented work on immunoassays for automated MTBR measurement [s1].

MTBR-tau species are currently measured largely by mass spectrometry, which is accurate, expensive and available in a limited number of specialised laboratories. Converting the measurement to a standard automated immunoassay is what would allow the marker to be used in large trials and, eventually, in clinical settings. Whether these assays reproduce the mass spectrometry results is the question, and it is not answered yet.

What to watch

Whether the etalanetug results replicate in larger cohorts, and whether they hold in sporadic Alzheimer's disease rather than the dominantly inherited form. Whether any tau programme reports a cognitive or functional endpoint alongside biomarker change. And whether the automated MTBR assays under development agree with mass spectrometry closely enough to be used interchangeably.

Nothing here is guidance about treatment. Etalanetug is investigational and not approved for any use.

Sources

• [s1] Alzforum, "After Tau Immunotherapy, Tangle Markers Tank in CSF, Plasma," 22 December 2025. https://www.alzforum.org/news/conference-coverage/after-tau-immunotherapy-tangle-markers-tank-csf-plasma
• [s2] Alzheimer Europe, "Results of EVOKE and EVOKE+ trials show no effect of oral semaglutide on AD progression," 3 December 2025. https://www.alzheimer-europe.org/news/results-evoke-and-evoke-trials-show-no-effect-oral-semaglutide-ad-progression