Cochrane Review of 17 Trials Found Anti-Amyloid Antibodies Made No Clinically Meaningful Difference in Early Alzheimer’s

TL;DR: A 2026 systematic review in The Cochrane Database of Systematic Reviews pooled 17 trials of anti-amyloid monoclonal antibodies (N=20,342 people with mild cognitive impairment or mild Alzheimer’s dementia) and concluded that any effect on memory decline or dementia severity was either nonexistent or well below the threshold considered clinically meaningful for patients.

Source: The Cochrane Database of Systematic Reviews (2026) | Nonino et al.

Amyloid-beta is a protein that accumulates in the brains of people with Alzheimer’s disease, often years before symptoms appear.

The dominant pharmaceutical strategy of the past two decades has been to develop monoclonal antibodies that bind amyloid and help clear it from the brain, on the assumption that lowering amyloid will slow cognitive decline.

This Cochrane systematic review tests whether that assumption holds up across the full set of randomized trials in the population the drugs target.

17 Randomized Trials in 20,342 Patients With Early Alzheimer’s

The Cochrane authors pooled randomized controlled trials of amyloid-beta-targeting monoclonal antibodies.

The included population was deliberately narrow:

• Mild cognitive impairment (MCI): Patients with measurable cognitive decline that does not yet meet dementia criteria.
• Mild Alzheimer’s dementia: Patients in the earliest dementia stage, where the prevailing hypothesis predicts amyloid removal should produce its largest benefit.
• Excluded: Moderate-to-severe Alzheimer’s, where amyloid plaques are presumed too entrenched for clearance to help.

Trial-level data covered 17 randomized trials and 20,342 participants — one of the largest pooled samples assembled for this drug class.

Pooled Effect on Memory Decline Was Below the Clinically Meaningful Threshold

The headline conclusion is the gap between statistical and clinical significance.

Across the pooled outcomes, the measured effect on memory decline and dementia severity was either nonexistent or, when measurable, smaller than the threshold the Cochrane authors define as the minimum patients can detect in daily life.

The conclusion was direct:

“The evidence suggests that these drugs make no meaningful difference to patients.”

The point is not that no measurable effect exists. It is that the effect, when present, is too small to translate into a difference patients or their families can feel.

Statistical Significance Stayed Below Patient-Relevant Benefit

Drug trials in dementia routinely report changes on standardized cognitive scales.

A statistically significant slowing of decline can mean very different things depending on size:

• A change of 0.5 to 1 point on a 30-point scale: Often statistically detectable in large trials but clinically invisible to patients and caregivers.
• A change of several points sustained over time: The kind of effect that translates into preserved daily functioning, retained memory of recent events, and slower transition to higher levels of care.
• A reversal of decline: Not what these drugs were designed to produce, and not what the trials show.

The Cochrane review is built around the first scenario. The drugs produce something measurable in trials. They do not produce something patients can feel.

Amyloid-Clearing Antibodies Did Not Produce Meaningful Clinical Benefit

For decades, the dominant model of Alzheimer’s pathophysiology has been the amyloid cascade hypothesis: amyloid-beta accumulation triggers tau pathology, neurodegeneration, and cognitive decline, in roughly that sequence.

That model predicts that clearing amyloid early should slow the downstream cascade.

The Cochrane review puts that prediction to a direct test in the population where it should work best, and finds the predicted clinical benefit does not materialize at clinically meaningful levels.

That does not refute the amyloid cascade hypothesis as a pathophysiological model. It does suggest the therapeutic implication — that pharmacologically lowering amyloid in symptomatic patients meaningfully slows the disease — needs revising.

Earlier-Stage Prevention, Long-Term Outcomes, and Subgroup Responders Stayed Out of Reach

• Earlier-stage prevention: The trials enrolled patients who already had MCI or mild dementia. Whether anti-amyloid treatment in pre-symptomatic at-risk individuals would slow eventual decline is a separate, ongoing research question.
• Long-term outcomes: Most trials lasted 18 to 24 months. Whether a small near-term effect compounds into a clinically meaningful difference over 5 to 10 years is not addressed by the available data.
• Subgroup responders: Pooled estimates may obscure subgroups (specific genotypes, biomarker profiles, disease stages) where effects could be larger. Subgroup-level evidence in the included trials is limited.
• Drug-specific differences: The review groups multiple antibodies together. Whether lecanemab, donanemab, and other agents differ from each other in clinically meaningful ways is partially obscured by pooling.
• Adverse event tradeoffs: The review focuses on efficacy; the safety profile (including amyloid-related imaging abnormalities) is a separate consideration in the risk-benefit calculation.

Clinical and Coverage Decisions Need Patient-Relevant Effect Thresholds

The practical implications follow from the statistical/clinical distinction:

• Approval frameworks should require patient-relevant effect sizes: A drug that produces a statistically detectable but clinically invisible improvement is a poor fit for the kind of patient-functional benefit Alzheimer’s families are seeking.
• Reimbursement and coverage: Payers and health systems weighing the high cost of monoclonal antibodies against their effect size now have a pooled estimate to use, rather than relying on individual-trial significance claims.
• Patient and family counseling: Clinicians describing these drugs should communicate the gap between statistical detectability and felt difference, not just report whether a trial reached significance.
• Research direction: The result raises pressure on alternative therapeutic strategies — tau-directed therapies, anti-inflammatory approaches, lifestyle interventions, and combination regimens — rather than additional refinements of the amyloid-clearance approach in symptomatic patients.