Featured StoryEditor's Pick
Latest Articles
The Anti-Amyloid Alzheimer's Drug Reckoning: What the April 2026 Cochrane Review Actually Found, and Why the Field Is Fighting Over It
On April 16, 2026, Cochrane published a systematic review of 17 clinical trials and 20,342 participants evaluating the class of anti-amyloid drugs — including the approved lecanemab and donanemab — and concluded that their effects on cognitive decline were 'absent or trivial.' The findings prompted immediate pushback from major Alzheimer's research and advocacy organizations. Here is what the Cochrane review actually said, what the primary trials of the approved drugs showed, and why serious experts disagree so sharply about how to read the evidence.
A Reckoning Two Decades in the Making
On April 16, 2026, the Cochrane Collaboration — the international evidence-synthesis organization whose systematic reviews are among the most rigorous in medicine — published a review of the entire class of anti-amyloid monoclonal antibody drugs for Alzheimer's disease [1]. The review, led by neurologist Francesco Nonino at the IRCCS Institute of Neurological Sciences in Bologna, pooled data from 17 clinical trials involving 20,342 participants. Its central conclusion, phrased in Cochrane's characteristically measured language, was blunt: the absolute effects of these drugs on cognitive decline and dementia severity were "absent or trivial," falling well below the established threshold for a minimum clinically important difference. The review also found that these drugs "likely increase" the risk of amyloid-related brain swelling and bleeding.
The response was immediate. Eisai (maker of lecanemab, marketed as Leqembi) pushed back publicly. Eli Lilly (maker of donanemab, marketed as Kisunla) pushed back publicly. The Alzheimer's Association's chief science officer defended the drugs. The Alzheimer's Society, one of the UK's largest patient advocacy organizations, argued that the review's conclusions "make the picture look bleaker than it really is" by pooling failed drug trials with more recent successful ones [4]. Regulatory bodies including the FDA, MHRA (UK), and multiple Asian and Latin American regulators continue to endorse lecanemab and donanemab.
This is a story about how medicine actually works when the evidence is contested — when serious, credentialed experts disagree not about the raw numbers but about what those numbers mean. It is also, potentially, a reckoning for the amyloid hypothesis that has driven Alzheimer's drug development for the past two decades.
What the Approved Drugs Did in Their Pivotal Trials
The two currently approved anti-amyloid antibodies — lecanemab and donanemab — reached the market on the basis of large Phase 3 trials in 2022 and 2023, both of which met their primary endpoints with high statistical significance.
The CLARITY-AD trial of lecanemab, published in the New England Journal of Medicine in January 2023, enrolled 1,795 patients with early Alzheimer's disease and randomized them to lecanemab or placebo for 18 months [2]. The primary endpoint was change on the Clinical Dementia Rating-Sum of Boxes (CDR-SB), an 18-point scale where higher scores indicate greater impairment. Lecanemab produced a 27% relative reduction in the rate of decline, corresponding to an absolute treatment difference of 0.45 points on the CDR-SB scale (P=0.00005). All secondary endpoints also reached statistical significance. Amyloid-related imaging abnormalities with edema (ARIA-E) — the characteristic side effect of this drug class involving brain swelling — occurred in 12.5% of lecanemab-treated patients versus 1.7% of placebo patients [2].
The TRAILBLAZER-ALZ 2 trial of donanemab, published in JAMA in August 2023, enrolled 1,736 patients with early symptomatic Alzheimer's disease across 277 medical centers in 8 countries [3]. The primary endpoint was the integrated Alzheimer's Disease Rating Scale (iADRS), a 144-point composite of cognition and daily function. In the low/medium tau population (the pre-specified primary analysis group), donanemab produced a 35% relative reduction in decline (P<0.001); in the combined population, the reduction was 22%. The absolute treatment difference on the iADRS was 3.25 points in the low/medium tau group. Brain swelling (ARIA-E) occurred in 24% of donanemab-treated patients; brain microbleeds (ARIA-H) in 31.4% versus 13.6% on placebo, with 1.6% of donanemab-treated patients experiencing serious ARIA [3].
Both trials cleared the regulatory bar for statistical significance. Both are the basis for current approval and clinical use.
What "Clinically Meaningful" Actually Means
The Cochrane review's central argument is not that the trial results were faked, mismeasured, or invalid. It is that "statistically significant" and "clinically meaningful" are different things — and that in this class of drugs, the observed treatment effects, while statistically real, are so small in absolute terms that they fall below the threshold at which a patient or a treating clinician would actually notice a difference in daily function.
The concept the review invokes is the "minimum clinically important difference" (MCID) — the smallest change on a rating scale that reliably corresponds to a change patients or families would perceive as meaningful in their real lives. For the CDR-SB scale, published estimates of the MCID for patients with mild cognitive impairment range from about 1.0 to 2.5 points. The 0.45-point treatment effect observed in CLARITY-AD is well below the lower end of that range. For the iADRS in TRAILBLAZER-ALZ 2, the 3.25-point treatment difference similarly sits below common MCID estimates.
Cochrane's Nonino put it directly: "It is common for trials to find statistically significant results that do not translate into a meaningful clinical difference for patients." That is the crux of the review's argument. Not that the trials failed, but that a demonstrable-but-small effect does not equal a therapeutically meaningful one.
What the Cochrane Review Found in Detail
The Cochrane review examined trials of nine different anti-amyloid antibodies: aducanumab, bapineuzumab, crenezumab, donanemab, gantenerumab, lecanemab, ponezumab, remternetug, and solanezumab [1]. Of the 17 trials included, 12 evaluated drugs that failed their primary endpoints. Three assessed aducanumab, which failed one of its two pivotal trials and was subsequently withdrawn from the market. Only two trials evaluated currently approved drugs — one for lecanemab (CLARITY-AD) and one for donanemab (TRAILBLAZER-ALZ 2).
Across the pooled evidence, the review authors concluded that anti-amyloid antibodies "probably result in little to no difference in cognitive function" as measured by standard scales, with effect sizes below the MCID. They also concluded that these drugs "likely increase" the risk of ARIA — the brain swelling and bleeding characteristic of the class — with implications for patient safety and monitoring cost. The authors recommended that Alzheimer's research redirect toward other targets — including inflammation, which some current research suggests may be a more central driver of disease progression than amyloid alone [1].
The Pushback — Why the Field Is Fighting
The response from established Alzheimer's research and advocacy organizations was rapid and largely critical. The core methodological objection was consistent: Cochrane pooled 15 trials of drugs that had already failed with only 2 trials of drugs that had succeeded, and treated the aggregate as if it represented the effect of "anti-amyloid drugs" as a class. Critics argued that this framing obscures the possibility that the two successful drugs may reflect genuine mechanistic improvement over their predecessors — the earlier antibodies targeted less-toxic amyloid forms or bound less selectively — rather than a class-wide failure.
The Alzheimer's Society's UK response, published April 22, 2026, made this argument explicitly: the review's conclusions "make the picture look bleaker than it really is, as authors combined results for a majority of failed drug trials with a small number of more recent successful trials" [4]. Regulatory bodies including the UK MHRA have specifically acknowledged "a small but meaningful benefit for lecanemab and donanemab" as the basis for their approvals. The UK Dementia Research Institute expressed similar concerns about the methodological approach of pooling drugs with meaningfully different mechanisms.
Both sides are, in some sense, correct. The pooled evidence across the entire class does not support a meaningful clinical benefit — that is a defensible reading of the Cochrane analysis. And the specific evidence for lecanemab and donanemab individually does show statistically significant effects that some regulators and clinical societies have judged clinically meaningful, particularly for patients at early disease stages. The disagreement is about what happens when the individual-drug analysis and the class-level analysis point in different directions — a genuinely difficult evidentiary question rather than a simple "who is right."
The Deeper Question — Is the Amyloid Hypothesis in Trouble?
The amyloid hypothesis — the idea that amyloid-beta plaque accumulation is the primary cause of Alzheimer's disease and that removing it should prevent or reverse the disease — has driven the majority of pharmaceutical research on Alzheimer's for roughly two decades. The pattern of trial results across the anti-amyloid drug class raises a serious question: if this many drugs, targeting amyloid in this many different ways, produce this small an effect in patients with amyloid pathology, is amyloid really the driver of the disease it has been assumed to be?
The Cochrane review authors themselves suggested that future research should look at alternative targets, particularly inflammation — a plausible mechanism given that lifestyle factors associated with reduced Alzheimer's risk (physical activity, cardiovascular health, sleep, social engagement) all also reduce systemic inflammation. This does not mean amyloid is irrelevant, but it does suggest amyloid may be part of a larger cascade rather than the sole cause. Research into tau protein, neuroinflammation, vascular contributors, and mitochondrial dysfunction has continued in parallel but has received far less pharmaceutical investment.
What This Means for Patients Right Now
For someone with a family member recently diagnosed with mild cognitive impairment or mild Alzheimer's disease and considering treatment, the current situation is difficult. Lecanemab and donanemab remain FDA-approved. Both are approved in the UK by the MHRA (though not currently reimbursed by the NHS on cost-effectiveness grounds). The European Medicines Agency approved lecanemab in November 2024 with the restriction that patients have no more than one copy of the APOE ε4 allele and cannot be on anticoagulants — a precaution to reduce the risk of ARIA-related bleeding.
The practical questions to discuss with a treating specialist include: How large is the expected benefit for this specific patient (a low-tau early-stage patient is likely to gain more than a high-tau or later-stage patient)? What is the likelihood of clinically significant ARIA, particularly in APOE4 carriers? What is the cost, and what monitoring (regular MRIs, clinical follow-up) is required? And crucially — is a treatment effect of the size demonstrated in the trials worth the risks and burdens of an infusion-based therapy with meaningful monitoring requirements?
There is no single right answer to these questions. Different families, different patients, and different specialists will legitimately reach different conclusions.
A Practical Synthesis
The April 2026 Cochrane review is a genuinely important document. It is also not the final word on the anti-amyloid drug class, and its aggregation of drugs across mechanisms and generations has drawn substantive methodological criticism. The honest reading of the current evidence: lecanemab and donanemab both produced statistically significant benefits in their pivotal trials [2][3], those benefits are small in absolute terms and sit below common thresholds for clinical meaningfulness [1], the class has real ARIA-related safety concerns, and reasonable specialists disagree about whether these drugs represent a meaningful advance or a marginal improvement that comes with substantial cost and risk.
The broader story is that amyloid, as a therapeutic target in Alzheimer's disease, has produced smaller clinical returns than two decades of research investment predicted. Whether that reflects mechanism, patient selection, timing of intervention, drug design, or a fundamental gap in our understanding of the disease is a question the next several years of research will need to address.
For families making treatment decisions right now, the right conversation is with a specialist in Alzheimer's disease who can help weigh the specific benefits, risks, monitoring requirements, and costs for the individual patient. The evidence base is genuinely contested, and thoughtful clinicians on either side of the debate can make defensible cases for their positions. The one framing that is not supported by the evidence — from either the Cochrane review or the industry response — is confident certainty in either direction.
Further Reading
Nonino F., Minozzi S., Sambati L., Del Giovane C., Baldin E., Bassi M.C., De Santis C., Gonzalez-Lorenzo M., Vignatelli L., Filippini G., Richard E. — Amyloid-beta-targeting monoclonal antibodies for people with mild cognitive impairment or mild dementia due to Alzheimer's disease (Cochrane Database of Systematic Reviews, April 2026) — systematic review of 17 clinical trials and 20,342 participants concluding that effects on cognitive decline are 'absent or trivial'
cochranelibrary.com
Van Dyck C.H., Swanson C.J., Aisen P. et al. — Lecanemab in Early Alzheimer's Disease (New England Journal of Medicine, January 2023) — CLARITY-AD phase 3 trial of 1,795 early Alzheimer's patients showing 27% slowing of clinical decline on CDR-SB, with 12.5% ARIA-E incidence
nejm.org
Sims J.R., Zimmer J.A., Evans C.D. et al. — Donanemab in Early Symptomatic Alzheimer Disease: The TRAILBLAZER-ALZ 2 Randomized Clinical Trial (JAMA, August 2023) — phase 3 trial of 1,736 patients showing 35% slowing on iADRS in low/medium tau population, with 24% ARIA-E and 31.4% ARIA-H
jamanetwork.com
New Cochrane review of amyloid targeting Alzheimer's disease treatments (Alzheimer's Society UK response, April 22, 2026) — official UK Alzheimer's Society response to the Cochrane review arguing that aggregating failed and successful drug trials misrepresents the evidence for lecanemab and donanemab
alzheimers.org.uk
📢 Advertisement — In-Article Ad