Recent advances, 2023 to 2026

The first disease-modifying drugs for Alzheimer’s

In July 2023 the FDA granted traditional approval to lecanemab (Leqembi), the first antibody shown in a confirmatory trial to slow cognitive decline in early Alzheimer’s disease by clearing amyloid from the brain. Donanemab (Kisunla) followed in July 2024. The effect is real but modest, roughly a quarter to a third slowing of decline over 18 months, and the drugs carry risks of brain swelling and small bleeds that require regular MRI monitoring. They are available only to people with confirmed amyloid pathology at an early stage. For the field, their importance is less the size of the effect than the proof that changing brain biology can change the course of the disease.

A blood test, at last

On 16 May 2025 the FDA cleared the first blood test to aid diagnosis of Alzheimer’s disease, the Lumipulse G pTau217/β-Amyloid 1-42 plasma ratio. In the validation study, 91.7 percent of people with a positive result had amyloid plaques confirmed by PET scan or spinal fluid, and 97.3 percent of those with a negative result did not. It is intended for people aged 55 and over who already have symptoms and are being evaluated in specialist care, not as a screening test. Its significance is access: a blood draw instead of a PET scan or a lumbar puncture means earlier, cheaper and far more widely available diagnosis.

Opening the blood–brain barrier on purpose

In January 2024, researchers at West Virginia University reported in the New England Journal of Medicine that MRI-guided focused ultrasound with microbubbles could temporarily open the blood–brain barrier in targeted brain regions in three people with Alzheimer’s receiving an anti-amyloid antibody. Amyloid fell about 32 percent more in the treated regions than in matching untreated regions over six months, and the barrier resealed within 24 to 48 hours. Three people is a proof of concept, not a treatment, but it demonstrates that the brain’s most protective structure can be engineered around safely.

A large, careful negative result

In November 2025 Novo Nordisk announced that its two phase 3 trials of oral semaglutide in early Alzheimer’s disease, evoke and evoke+, enrolling 3,808 people across 40 countries, did not slow clinical decline over two years. Full results published in The Lancet in March 2026 showed the drug reduced several Alzheimer’s-related biomarkers in spinal fluid by up to 10 percent and lowered inflammation markers in blood, yet produced no measurable benefit on cognition or function.

This matters for two reasons. It closes off a widely hoped-for repurposing of a drug already taken by millions, and it is the clearest recent demonstration of the gap between moving a biomarker and helping a patient. Any claim, by anyone, that a biomarker change proves clinical benefit should be read with the evoke trials in mind.

Prevention moved from hope to numbers

The 2024 Lancet Commission on dementia added untreated vision loss and high LDL cholesterol to its list of modifiable risk factors, bringing the total to fourteen and the estimated share of potentially preventable dementia to about 45 percent. Separately, the ACHIEVE trial reported that hearing aids slowed cognitive decline by nearly half in older adults with hearing loss who were at elevated risk. These are not laboratory curiosities; they are things health systems can act on now.

What to watch

Blood biomarkers moving toward screening of people without symptoms; antibodies and other approaches against tau and alpha-synuclein; the first senolytic and NAD+ trials with brain outcomes; wider human measurement of the glymphatic system with new MRI methods; and the question of whether the brain’s own maintenance systems can be supported directly. On that last point, several groups including ours are working toward evidence. Nobody has it yet, and this page will say so until someone does.