For the first time in Alzheimer’s research, a drug targeting tau protein has simultaneously reduced brain pathology and slowed cognitive decline in the same clinical trial. That trial officially failed its primary endpoint.

Biogen’s Phase 2 CELIA study tested diranersen, an antisense oligonucleotide designed to block the cellular instructions that produce tau, in 416 patients with mild cognitive impairment or mild dementia due to Alzheimer’s. Across three doses tested over 76 weeks, the drug reduced tau in cerebrospinal fluid and on PET imaging. Patients on treatment declined more slowly on standard cognitive measures than expected.

But CELIA’s primary endpoint was not whether the drug worked overall. It was whether higher doses worked better — a statistical question designed to identify the optimal dose for late-stage trials. They didn’t. The strongest cognitive signal came from the lowest dose, 60 mg administered every 24 weeks, roughly the opposite of what pharmacology would predict.

“Candidly, we don’t know why the dose response here wouldn’t be linear,” Stifel analyst Paul Matteis wrote in a note to clients, calling the finding “definitely an open question that will be key to understanding the strength/realness of this signal.”

Biogen is moving to Phase 3 regardless.

A Genuine First

The statistical miss obscures a result that no tau-targeting drug has achieved before. Tau is one of the two hallmark proteins in Alzheimer’s disease — alongside amyloid — forming neurofibrillary tangles inside neurons that correlate more closely with cognitive decline than amyloid plaques do. Yet every previous attempt to drug tau has ended in failure.

Johnson & Johnson abandoned its antibody posdinemab after a Phase 2b flop in November 2025. Eli Lilly’s zagotenemab and its small-molecule OGA inhibitor both disappointed. Roche and AC Immune’s semorinemab fell short across multiple trials. Biogen itself scrapped an earlier tau antibody, gosuranemab, in 2021 after its own Phase 2 miss.

Diranersen takes a fundamentally different approach. Rather than clearing tau protein after it accumulates, the drug uses antisense technology to silence the MAPT gene’s messenger RNA, reducing production at the source. That means less tau both inside neurons and in the spaces between them. Originally discovered by Ionis Pharmaceuticals, the compound was licensed to Biogen in 2019 and received FDA Fast Track designation in 2025.

Dr. Jeff Cummings, professor of brain sciences at the University of Nevada, Las Vegas, called the CELIA results “an important advance for the field” and the first evidence that reducing tau “may meaningfully impact disease progression.”

The Missing Number

What Biogen has not disclosed is the magnitude of cognitive benefit — and that gap is driving most of the outside caution.

Guggenheim Securities analysts, who called the CELIA readout the “moment of truth for tau,” told investors that a 0.6-point benefit over placebo on the CDR-SB, a standard dementia severity scale, could send Biogen’s stock up 50 percent. A benefit of 0.4 to 0.5 points might leave shares roughly flat. Biogen has not said where its results land.

RBC Capital Markets analyst Brian Abrahams cited “scant detail on the actual effect size” and “oddities around dose dependence” in urging restraint. BMO Capital Markets analysts said the topline data suggest intracellular tau targeting may address some of the drivers of Alzheimer’s disease, but more work in Phase 3 is needed to confirm the benefit.

Biogen shares initially rose roughly 4 percent in premarket trading before falling about 5 percent as the market absorbed the ambiguity.

Dr. Priya Singhal, Biogen’s head of development, described the results as showing “an unprecedented and compelling confluence of efficacy and biomarkers results.” She told CNBC the company had gotten “really getting close to isolating a dose,” which she listed among the three requirements for advancing to Phase 3.

The drug’s safety profile was consistent with earlier studies, though serious adverse events were more common at the highest dose. Diranersen is administered intrathecally — via injection into the fluid surrounding the spinal cord — which analysts noted could pose challenges for commercial viability compared to intravenous alternatives. The ripple effects nonetheless reached other companies: Voyager Therapeutics, which has its own tau-silencing gene therapy, saw shares rise 10 percent in premarket trading.

Waiting on London

Full results will be presented at the Alzheimer’s Association International Conference in London in July. Until then, the field has a drug that does something biologically meaningful to tau, paired with a cognitive signal of unknown strength.

The Alzheimer’s Association called the results “important progress.” Biogen called them unprecedented. Analysts called them mixed.

All three are probably right.

Sources