Biogen's Alzheimer's drug lowers toxic protein, but results were mixed

CHICAGO - The lowest dose of Biogen's anti-tau Alzheimer's drug diranersen, which missed its primary goal in a midstage trial, reduced cognitive decline in five of six Alzheimer's assessment tools and significantly cut levels of the toxic protein in the brain, researchers reported on Tuesday.

The trial is the first to show that lowering tau - a protein closely linked with brain cell death and cognitive decline - can slow progression of Alzheimer's.

"This trial, as far as I'm concerned, is proof of principle. If you change tau, you can change the course of the disease, and we haven't had that before," said Cath Mummery, a professor of clinical neurology at University College London who led the study.

Instead of removing toxic forms of tau with an antibody, an approach used by prior failed drugs, diranersen silences a gene responsible for producing all forms of tau.

The result, presented at the Alzheimer's Association International Conference in London, was a 50% to 65% reduction in tau across all studied doses based on brain imaging and spinal fluid measures.

As reported in May, the drug missed the study's main goal of showing a dose-dependent response - an increasing reduction in cognitive decline in patients who received successively higher doses of the treatment on the Clinical Dementia Rating-Sum of Boxes or CDR-SB, a widely used assessment tool.

Instead, "it was the opposite," said Maria Carrillo, chief science officer of the Alzheimer's Association, noting that the strongest response came from the lowest studied dose. She described the drug as "worth pursuing."

The cumulative findings offered enough proof for Biogen to advance to a large-scale trial, said Dr. Priya Singhal, Biogen's head of development.

She said the company is consulting with numerous Alzheimer's experts and is gearing up for a single, pivotal Phase 3 trial starting in 2027 with data expected in 2030-2031.

UP TO 50% SLOWER DECLINE

The 18-month Phase 2 CELIA trial of 416 patients with early Alzheimer's tested three doses of diranersen injected into the spine: either a 60 mg or 115 mg dose given every six months or a 115 mg dose given every three months.

Details have not been previously reported.

At 60 mg, diranersen slowed cognitive and functional decline by 26% or 0.54 point versus placebo on the CDR-SB, an 18-point scale.

The result is roughly on par with the 25-30% slowing seen in the Phase 3 trials of Biogen and Eisai's Leqembi and Eli Lilly's Kisunla, drugs that target amyloid, another Alzheimer's-related protein.

The 26% slowing was lower than the 30% or greater efficacy JPMorgan analyst Chris Schott was looking for. However, Baird analyst Jack Allen said a 0.4 point difference or greater was the threshold for a meaningful therapeutic effect.

Analysts said the results could impact companies developing other gene-silencing drugs targeting tau including Voyager Therapeutics, Arrowhead Pharmaceuticals and Denali Therapeutics.

Beyond CDR-SB, diranersen showed a benefit on four other assessment tools.

On the ADAS-Cog13, which tracks cognitive decline, the 60 mg dose slowed decline by 42% versus placebo; it slowed decline by 50% on the MMSE, an 11-question test of cognitive function. On modified iADRS, a test of cognition and daily function, there was a 30% slowing, and on ADCOMS, which detects early-stage decline, there was a 23% slowing.

Higher doses also showed some benefit; however, there was no difference at any dose on the ADCS-ADL-MCI, which measures daily functioning. Biogen is continuing to assess patients on this scale for 24 months.

SIDE EFFECTS WERE MILD TO MODERATE

Mummery said the missed endpoint was disappointing but given the consistency across the secondary endpoints, she believes the drug deserves more testing. "It's a very encouraging signal," she said.

The most frequent side effects were mild to moderate and were procedure-related including pain and headache associated with the lumbar puncture and confusion which resolved quickly.

There were no cases of ARIA - a brain-swelling condition associated with Leqembi and Kisunla - drugs that target amyloid.

(Reporting by Julie Steenhuysen; Editing by Caroline Humer and Matthew Lewis)

Copyright Reuters or USA Today Network via Reuters Connect

This story was originally published July 14, 2026 at 9:18 AM.