Cochrane Review Finds Anti-Amyloid Alzheimer's Drugs Offer No Meaningful Cognitive Benefit

A large Cochrane systematic review published in April has handed the anti-amyloid drug class its most rigorous challenge yet, concluding that the treatments approved to slow Alzheimer's disease likely do not deliver benefits that matter to patients - while meaningfully raising the risk of serious brain side effects.

The review, published in the Cochrane Database of Systematic Reviews (Issue 4, 2026, DOI: 10.1002/14651858.CD016297), was led by Francesco Nonino, a neurologist and epidemiologist at the IRCCS Institute of Neurological Sciences of Bologna, Italy, and co-authored by ten additional researchers. The work pooled data from 17 randomized controlled trials enrolling 20,342 participants, all of whom had mild cognitive impairment or early-stage Alzheimer's dementia - exactly the population regulators and drug makers have argued is most likely to benefit from early amyloid clearance.

The headline finding is blunt. According to the review as covered by Applied Clinical Trials Online, pooled data showed that amyloid-targeting antibodies "probably result in little to no difference in cognitive function" on the Alzheimer's Disease Assessment Scale-Cognitive (ADAS-Cog), with a standardized mean difference of -0.11 (95% CI -0.16 to -0.06; moderate certainty). On dementia severity measured by the Clinical Dementia Rating Scale Sum of Boxes, the treatments "may result in little to no difference," with an SMD of -0.12 (95% CI -0.24 to -0.00; low certainty). Effect sizes that small sit well below any threshold clinicians consider detectable by a patient or caregiver in daily life.

The safety signal deserves equal attention. The review, as reported by Applied Clinical Trials Online, found a probable small increase in amyloid-related imaging abnormalities involving edema - a known side effect class called ARIA-E - with an absolute risk difference of 107 more cases per 1,000 participants compared to placebo. That is not a rounding error; it is roughly one in ten trial participants experiencing a brain-swelling event that, in some cases, requires treatment discontinuation or hospitalization.

The drug maker Eisai, which markets lecanemab under the brand name Leqembi, pushed back immediately. According to reporting reviewed by Powers Health, Eisai argued that the meta-analysis was "scientifically deeply flawed by inappropriately combining ineffective antibodies and failed studies with effective, regulatory approved anti-amyloid treatments." The Alzheimer's Association raised a similar objection, noting that lumping first-generation failed compounds with newer, higher-affinity antibodies inflates the noise and suppresses any real signal from the approved drugs.

That methodological dispute sits at the center of how to read these data. Meta-analyses that combine drugs with different mechanisms, dosing regimens, and target-engagement profiles can obscure heterogeneity. The Cochrane team would counter that pooling is precisely the point - it asks whether the class-level hypothesis holds, and the answer, across the broadest available evidence base, appears to be: barely, if at all.

The amyloid hypothesis - that plaque clearance will translate to preserved cognition - has driven more than two decades of drug development and billions of dollars in trial spending. The hypothesis has always rested on a surrogate-to-outcome gap: clearing amyloid can be measured radiographically, but whether that change propagates to meaningful clinical benefit has remained contested. A separate Bayesian meta-analysis published in the Journal of Comparative Effectiveness Research in January 2026, reviewed on PubMed (PMID 41328628), found that while the surrogate relationship between amyloid reduction and clinical function was statistically meaningful across the antibody class as a whole, individual drug-level relationships showed "large uncertainty."

For reporters, clinicians, and patients trying to make sense of this: the Cochrane review does not prove these drugs are inert. It says the current evidence base, taken together and held to the standard of clinical meaningfulness rather than statistical significance, does not yet support the effect sizes their advocates claim. That is a different statement - but it is the correct one to make. Single-study approvals built on surrogate endpoints require exactly this kind of cumulative scrutiny. The Cochrane group has now provided it, and the field will need to answer it with data, not press releases.

Sources cited:
- Cochrane Database of Systematic Reviews, Issue 4, 2026 (via UK Dementia Research Institute) (https://www.ukdri.ac.uk/news-and-events/uk-dementia-research-institute-responds-cochrane-review-suggesting-anti-amyloid)
- Applied Clinical Trials Online (https://www.appliedclinicaltrialsonline.com/view/cochrane-review-anti-amyloid-alzheimerstrials-criticism)
- Powers Health / HealthDay (https://www.powershealth.org/about-us/newsroom/health-library/2026/04/16/new-alzheimers-drugs-provide-no-meaningful-benefit-major-evidence-review-concludes)
- Journal of Comparative Effectiveness Research, Bayesian meta-analysis (PubMed) (https://pubmed.ncbi.nlm.nih.gov/41328628/)