Cochrane review finds little benefit from Alzheimer’s amyloid drugs, drawing expert criticism

For families weighing whether to start an amyloid drug, and for clinicians deciding whether the payoff justifies the burden, the new question is not just whether these medicines work, but whether a sweeping review treated very different drugs as if they were the same.

The Cochrane Collaboration’s systematic review of anti-amyloid monoclonal antibodies examined 17 trials involving 20,342 people with mild cognitive impairment or mild dementia due to Alzheimer’s disease. Its conclusion was stark: after 18 months, the drugs probably made little to no difference in memory, thinking ability or dementia severity, while probably increasing brain swelling and tiny brain bleeds. The review’s authors said the effects were “absent or trivial,” below the minimum clinically important difference.

That conclusion landed squarely in the middle of an expensive and politically fraught market. Anti-amyloid drugs are usually given by infusion every two or four weeks, with monitoring that adds to the cost and complexity of care. A referenced estimate put the annual price of donanemab at about $32,000 in the United States, before administration and surveillance costs. The review also said that successfully removing amyloid from the brain did not translate into clinically meaningful improvement, a finding that will strengthen skepticism among payers and health systems already under pressure to judge whether these therapies are worth covering.

But the methodology drew immediate fire. Critics argued that the review lumped together failed older antibodies with the two newest, and only, amyloid-clearing drugs approved in the United States: Leqembi, or lecanemab, which became available in 2023, and Kisunla, or donanemab, approved in 2024. Francesco Nonino, Edo Richard, Ian Maidment and John Hardy were among the experts cited in reaction to the analysis. Their central complaint was that the review conflated drugs with different mechanisms, including aducanumab, which removes existing plaques, and lecanemab, which mainly binds soluble amyloid to prevent plaque formation.

Supporters of the review described it as a gold-standard synthesis of the evidence. Still, the criticism was not only about mechanism. Six studies were still ongoing and were excluded, and the included trials had relatively few Black or Hispanic/Latino participants, limiting how far the findings can be generalized. That matters because the disease burden is large, the treatment options remain limited, and the evidence base is still being built in real time.

The policy stakes are already visible. In the United Kingdom, the National Institute for Health and Care Excellence previously declined to approve lecanemab and donanemab for National Health Service use because the benefits were judged too small relative to cost, and that decision is now being reassessed under new cost-effectiveness thresholds. For now, the review will not settle the amyloid debate, but it does sharpen it: between a desperate need for better Alzheimer’s treatments and the hard arithmetic of benefits, harms and price.