"After decades without a meaningful breakthrough, Australia has recently approved lecanemab and donanemab, anti-amyloid drugs hailed as the first medicines that might finally slow Alzheimer’s disease rather than simply manage symptoms. The question is the extent of their benefit, especially given the costs. For families watching a loved one lose memory, judgement and independence, the promise has carried enormous hope. But a major Cochrane review, pooling 17 trials involving 20,342 people, delivers a far more sobering verdict: overall, it says, these drugs do not provide a benefit that most people in the early stages of Alzheimer’s would find meaningful, while brain swelling occurred in about 119 per every 1000 on treatment. 

Dementia affects about 57 million people worldwide, and Alzheimer’s is its leading cause. It is a progressive brain disease in which nerve cells and the connections between them are damaged and lost, leading to worsening memory, thinking, language, judgement and day-to-day function. For more than 40 years, one dominant idea has guided the field: that amyloid-beta, a sticky protein fragment, builds up into clumps called plaques in the brain, helps drive this damage, and therefore must be removed if the disease is to be slowed.

Anti-amyloid drugs were developed based on that idea. They are laboratory-made antibody drugs, given by repeated intravenous infusion, designed to bind to amyloid and clear it from the brain. The plaques can be seen on special amyloid PET brain scans, and these drugs are very effective at reducing that plaque signal. Because the drugs can also cause brain swelling and small bleeds in the brain, treatment has to be monitored with repeated MRI scans.

That is why this Cochrane review is so dramatic. It is not a paper based on one company trial but assesses a large number of studies. It is the best independent test of whether a dramatic effect on amyloid PET scans translates into a meaningful benefit for patients. Its answer is sobering: even when plaques are successfully cleared, the hoped-for improvements in memory, thinking, daily function and independence do not meaningfully follow. In plain English, these drugs clear plaques far more convincingly than they preserve memory, thinking and independence.

This Cochrane review forces a blunt question: if these drugs remove the very plaques long thought to drive Alzheimer’s, why are patients not getting a meaningful improvement in memory, thinking, daily function or independence? In sum:

• This Cochrane review says the drugs are very good at clearing amyloid plaques from the brain but overall benefit of the current anti-amyloid drugs is not sufficiently clinically meaningful.
• The numbers show how small the effect was. Across 13 trials over 18 months, the average advantage on the main memory and thinking test, the ADAS-Cog, was just 0.85 points on a scale that runs to about 70 points. Across nine trials over 18 months, the average advantage on the main dementia severity measure, the CDR-SB, was just 0.29 points on an 18-point scale. These are very small changes, unlikely to be truly noticeable to most patients or families in daily life.
• On day-to-day function, the picture was much the same. The average advantage was 1.32 points on the 78-point ADCS-ADL scale, 1.75 points on the 56-point ADCS-iADL scale, and 1.90 points on the 53-point ADCS-ADL-MCI scale. In plain English, any slowing of decline was small at best and far from the kind of change patients and families would recognise as a breakthrough.
• These are average trial results, and individual responses will vary. But the overall signal from more than 20,000 participants is clear: strong plaque removal has not translated into meaningful gains in memory, thinking, daily function or independence.
• The safety signal is notable. Brain swelling was seen in about 119 of every 1000 treated patients, compared with 12 of every 1000 on placebo. Symptomatic brain swelling rose from about 1 per 1000 on placebo to about 30 per 1000 on treatment. The review also suggested more tiny brain bleeds on scans, although results varied across trials.
• The review did not show more overall serious adverse events or more deaths at 18 months, but that is not the same as a clean safety bill. Symptom reporting was inconsistent, the longer-term picture remains unsettled, and these are intensive, expensive treatments that require repeated infusions and repeated MRI monitoring.
• Our wider concern is that the safety story may still be more serious than is often presented. Other work has raised concerns about accelerated brain volume loss and possible treatment-related deaths. Those questions deserve independent scrutiny.
• This review does not prove amyloid has no role in Alzheimer’s, and it does not rule out future amyloid-directed therapies that may yet help patients. But it does show that the current generation of anti-amyloid drugs is not delivering the promise that has surrounded it.

Alzheimer’s is the progressive failure and loss of synapses and nerve cells, which patients experience as failing memory, failing thinking and lost independence. We and our collaborators have been raising concerns in the scientific literature for years about the dominance of the amyloid hypothesis. We have long argued that while amyloid has role in the disease, it alone does not explain Alzheimer’s dementia and a new way forward is needed The way forward is to widen the lens: focus on the molecular and cellular biology of what is going wrong inside vulnerable brain cells and synapses, and develop genuinely individualised treatments judged by meaningful gains in cognition, daily function and independence.

No patient can derive benefit from plaque clearance unless it translates into tangible improvements in quality of life. Until we identify and target root disease causes and achieve profound clinical benefit, no therapy can be called disease modifying. Amyloid may still play a role, and future therapies aimed at it may yet prove useful, but the field now needs a broader, more individualised medicine approach with a sharp focus on preserving cognition, daily function and independence."

"This review provides important context within the evolving clinical landscape of the treatment and prevention of Alzheimer’s disease. While the lack of improvement in clinical outcomes may be disheartening to hear, two key messages are important to keep in mind.

One, the findings should refocus our attention on addressing the 14 modifiable approaches to dementia risk reduction identified by the Lancet commission – particularly early in the disease course. Two, the use of these drugs has always been limited to very specific cases meeting distinct criteria, however, the brain is complex and what this review highlights is a single mechanism is unlikely to offer the magic bullet we seek. One of the reasons multi-domain approaches have shown promise as being potentially more beneficial than singe domain interventions for reducing dementia risk may be due to their ability to target different mechanisms.

For example, exercise may offer neuroprotective effects by reducing inflammation, building cognitive reserve, and reducing depression. Effective pharmacological disease modifying therapies for Alzheimer’s disease may still be a while away. However, what we have now is robust evidence-based approaches for supporting brain health across life which, if broadly implemented, will likely reduce the number of people being diagnosed with Alzheimer’s disease and dementia in the future and help slow disease progression."

"The authors concluded that the effects of these therapies on cognitive function and dementia severity were trivial to small, while increasing the risk of adverse events such as brain swelling and bleeding. The findings concord with the decisions of several international governmental agencies to deny approval based on efficacy, including the Australian PBAC, but differ from others, such as the USA.

Cochrane reviews represent the highest standard of evidence-based healthcare. They have included all major trials published to 7AUG2025. However, the inclusion of older trials known to be negative does skew the findings somewhat, as we know that these agents had less amyloid clearance than newer agents. It would have been interesting to have a pooled analysis of the 3 agents with positive trials, aducanumab, lecanemab, and donanemab.

In addition, their sobering conclusions that “Successful removal of amyloid from the brain does not seem to be associated with clinically meaningful effects … future research … should focus on other mechanisms of action” will be contentious. The Alzheimer’s disease field still contains a solid majority of adherents to the amyloid hypothesis, and these comments will spark intense debate. However, there will be many clinicians who welcome this objective guidance for evidence-based care."