- Lecanemab (Leqembi) and donanemab (Kisunla) are infused antibodies that clear amyloid from the brain in early Alzheimer's disease. In the 18-month CLARITY-AD trial of 1,795 people, lecanemab slowed decline on the CDR-SB scale by 0.45 points (95% CI 0.23 to 0.67). That trial was funded by Eisai and Biogen (van Dyck et al., NEJM 2023).
- Donanemab slowed decline on the same scale by 0.7 points (95% CI 0.45 to 0.95) in TRAILBLAZER-ALZ 2, a trial of 1,736 people sponsored by Eli Lilly, whose employees are among the authors (Sims et al., JAMA 2023).
- Both numbers are below the thresholds that independent reviewers use for a change patients would notice. A rapid review put the smallest meaningful CDR-SB change in mild cognitive impairment at about 1 point and said recent anti-amyloid effects were lower than published thresholds (Muir et al., 2024). A 2026 Cochrane review of 17 industry-funded trials concluded the effects on cognition and dementia severity are trivial (Nonino et al., Cochrane 2026).
- The main harm is ARIA (amyloid-related imaging abnormalities): brain swelling or small bleeds seen on MRI. On the US label, any ARIA with lecanemab was 45% in people with two copies of ApoE4, 19% with one and 13% with none, against 22%, 9% and 4% on placebo. Both US labels carry a boxed warning that ARIA can be fatal (Leqembi label; Kisunla label).
- Regulators split. The FDA gave lecanemab traditional approval in July 2023 and the MHRA licensed it in August 2024. The EU authorised lecanemab in April 2025 and donanemab in September 2025, both only for people with one or no copy of ApoE4 (EMA Leqembi; EMA Kisunla). Alzheimer's Research UK reports that NICE's final decision (June 2025) was that lecanemab will not be offered on the NHS in England (ARUK).
- Money trail: both pivotal trials were run by the companies that sell the drugs, and all 17 trials in the Cochrane review were industry-funded. Independent cost studies mostly find the drugs poor value at list price (Nguyen et al., 2026).
- Evidence grade: Strong that they remove amyloid. Weak that the slowing is large enough to matter to patients. Insufficient beyond 18 months. Risk for ARIA, especially in ApoE4 carriers.
Independent evidence review · Prescription medicine
Lecanemab and donanemab are the first drugs to show, in large trials, that removing amyloid from the brain slows early Alzheimer's disease by a measurable amount. The amount is small, and the trials that showed it were run and funded by the companies that sell the drugs. Independent reviewers disagree on whether the benefit is large enough to matter to patients, and the drugs carry a real risk of brain swelling and bleeding that is highest in people with two copies of the ApoE4 gene. We have moderate confidence in the size of the measured effect, low confidence that it is clinically meaningful, and high confidence in the ARIA risk. Regulators in the US, UK and EU have approved the drugs with restrictions. NICE has not recommended them for the NHS.
These are prescription-only infusions given in specialist settings. Do not seek them privately without understanding the risks. Get urgent help if someone on treatment develops a new severe headache, confusion, a change in vision, difficulty walking, seizures or any sudden neurological change: these can be signs of symptomatic ARIA, and both US labels warn that ARIA can be fatal. Tell every doctor, dentist and emergency team that the person receives an anti-amyloid antibody, because clot-dissolving drugs and anticoagulants have been linked to serious bleeding (Leqembi label). In an emergency call your local emergency number.
Table of contents
- Evidence summary
- What lecanemab and donanemab are
- How they work
- What they are prescribed for
- What works and what does not
- Benefits by claim
- Risks and all side effects
- All interactions
- Who should avoid these drugs
- Dosage and how they are given
- Follow the money: who makes them and who funded the evidence
- Related research
- Frequently asked questions
- Sources and funding notes
Evidence summary
| Claim | Evidence | Source | Funding / conflict | Strength |
|---|---|---|---|---|
| Lecanemab removes amyloid | In a substudy of 698 people, amyloid fell by 59.1 centiloids more than on placebo (95% CI 55.6 to 62.6). | CLARITY-AD, NEJM 2023 | Funded by Eisai and Biogen (the sellers) | Strong (surrogate marker) |
| Lecanemab slows decline | 1,795 people, 18 months. CDR-SB change 1.21 vs 1.66 on placebo: difference -0.45 (95% CI -0.67 to -0.23). ADAS-cog14 difference -1.44 (-2.27 to -0.61). | CLARITY-AD | Funded by Eisai and Biogen | Moderate |
| Donanemab slows decline | 1,736 people, 76 weeks. iADRS difference 2.92 points (95% CI 1.51 to 4.33); CDR-SB difference -0.7 (-0.95 to -0.45). | TRAILBLAZER-ALZ 2, JAMA 2023 | Sponsored by Eli Lilly; authors include Lilly employees and shareholders | Moderate |
| The slowing is large enough to matter | Cochrane pooled 17 trials (20,342 people): CDR-SB SMD -0.12 (95% CI -0.24 to 0.00), low certainty; ADAS-Cog SMD -0.11 (-0.16 to -0.06). Authors call effects "trivial". A separate meta-analysis found no result above the minimal clinically important difference. | Nonino, Cochrane 2026; Ebell, Ann Fam Med 2024 | Cochrane: Emilia-Romagna region and Italian Ministry of Health; all included trials industry-funded. Ebell: no funding statement seen in the record we opened. | Weak (contested) |
| ARIA brain swelling or bleeding | Lecanemab ARIA-E 13% vs 2% and ARIA-H 17% vs 9% on placebo. Donanemab ARIA-E 24% vs 2%. Cochrane: 107 more per 1,000 people have ARIA-E (95% CI 77 to 148). | Leqembi label; Kisunla label; Cochrane 2026 | Regulator-approved labels built on the makers' trial data; Cochrane as above | Risk |
| Higher ARIA risk in ApoE4 carriers | Lecanemab any ARIA: 45% (two copies), 19% (one), 13% (none) vs 22%, 9%, 4% on placebo. MHRA advice: risk-benefit favourable for non-carriers and heterozygotes, not homozygotes. | MHRA; FDA label | Regulators; data originate from the maker's trial | Risk |
| The drugs cause more deaths | Cochrane: 2 more deaths per 1,000 (95% CI 3 fewer to 11 more), high certainty of little or no difference. Donanemab trial: 3 treatment-related deaths vs 1. Both labels note fatal brain haemorrhage cases. | Cochrane 2026; TRAILBLAZER-ALZ 2 | As above | Insufficient (small numbers) |
| The drugs are cost-effective | Donanemab in Australia: ICER $342,424 per QALY (healthcare perspective). A review found manufacturer-funded lecanemab models below thresholds and others above; no independent study found savings. | Gao, Med J Aust 2026; Nguyen, 2026 | Gao: no funding listed; one author reports honoraria from Lilly. Nguyen: no funding stated. | Weak |
| Benefit lasts beyond 18 months | Both pivotal trials ended at 18 months (76 weeks). Long-term projections are models, including one co-authored by Eisai employees. | Burn et al., 2026 | Authors affiliated with Eisai (the seller) | Insufficient |
Independent evidence and credibility scorecard
Independence tiers: 1 = regulator or government body; 2 = academic, Cochrane or charity source with no declared commercial funding for the work; 3 = independent authors with declared industry ties or a competitor funder; 4 = manufacturer-funded or manufacturer-authored. Credibility runs from A (highest) to D.
| Source | Who funds it | Country | Independence tier (1–4) | Credibility (A–D) | Why it would tell the truth / residual bias |
|---|---|---|---|---|---|
| FDA Leqembi label | US government (the maker, Eisai Inc., writes the label with FDA approval) | US | 1 | A for warnings | Warnings and ARIA tables are legally required and checked. Efficacy text rests on the maker's trial. |
| FDA Kisunla label | US government (Eli Lilly is the labeller) | US | 1 | A for warnings | Same logic. The label reports ARIA by ApoE4 genotype, which the press releases do not. |
| EMA Leqembi and Kisunla pages | EU agency, funded by EU budget and industry fees | EU (Netherlands) | 1 | A | Narrowed both licences to non-carriers and heterozygotes after first refusing both drugs. Fee income from industry is a structural tie. |
| MHRA lecanemab announcement | UK government | UK | 1 | A | Reports ARIA by genotype and says risk-benefit is not favourable in homozygotes. Does not discuss cost. |
| Nonino et al., Cochrane 2026 | Emilia-Romagna region (Italy); Italian Ministry of Health Ricerca Corrente funding | Italy (review team); published by Cochrane (UK) | 2 | A− | Cochrane bars commercial sponsorship. Residual bias: it pooled seven antibodies, including ones that failed, which critics say dilutes the result for the two approved drugs. |
| Ebell et al., Ann Fam Med 2024 | No funding statement seen in the PubMed record | US | 2 | B | Family-medicine researchers with no stated product stake who judged results against minimal clinically important differences. We could not check the full-text disclosures. |
| Muir et al., Alzheimer's & Dementia 2024 | No funding section in the record; two of four authors report Lilly, Eisai or Roche consulting or research contracts | Canada | 3 | B | A systematic search of published thresholds, so the numbers are not the authors' opinion. Some authors have ties to the sellers and their rivals. |
| Nguyen et al., 2026 | Funding not stated; authors declare no conflicts | US | 2 | B | A narrative review of 16 cost studies. It makes visible which models were manufacturer-funded. |
| Gao et al., Med J Aust 2026 | No funding listed; one author reports honoraria from Eli Lilly and Novo Nordisk | Australia | 3 | B | A transparent donanemab model, but the discount to list price is uncertain and one author has declared Lilly honoraria. |
| Alzheimer's Research UK (NICE report) | Donations from the public; the page does not state sources or amounts | UK | 2 | B | A research charity needs credibility with donors and patients. Its mission is to speed Alzheimer's treatments, which could tilt it towards optimism. Secondary source for NICE. |
| CLARITY-AD (van Dyck), NEJM 2023 | Eisai and Biogen | Multinational; sponsor Japan and US | 4 | C | A peer-reviewed phase 3 trial with pre-set endpoints, so cheating is hard. But the sellers designed it and benefit from the result. |
| TRAILBLAZER-ALZ 2 (Sims), JAMA 2023 | Eli Lilly; the only grant listed in the record is an NIH grant (S10 OD025214) | US (277 sites, 8 countries) | 4 | C | Same logic. The drug is sold by the sponsor, and sponsor employees are authors. |
Sources cluster in the US and UK, and the two pivotal trials are the only evidence for the approved dosing. The Cochrane and Ebell analyses, which have no commercial funding, drew on those same industry trials, so they check the interpretation but not the underlying data.
What lecanemab and donanemab are
- Lecanemab (brand Leqembi) is a humanised monoclonal antibody given as an intravenous infusion. The US label approves it for treatment that begins at the mild cognitive impairment or mild dementia stage, and later added a subcutaneous autoinjector (Leqembi label). It was developed jointly by Eisai, Biogen and BioArctic (Wikipedia, secondary).
- Donanemab (brand Kisunla) is an intravenous antibody made by Eli Lilly. It targets a modified form of amyloid found in established plaques (Kisunla label).
- US: the FDA gave lecanemab accelerated approval in January 2023 and converted it to traditional approval on 6 July 2023, after an advisory committee voted unanimously on 9 June 2023 that the trial verified clinical benefit (FDA). Donanemab was approved in July 2024 (FDA).
- UK: the MHRA licensed lecanemab on 22 August 2024 for adults with early Alzheimer's disease who have zero or one copy of ApoE4 (MHRA). Donanemab was licensed in October 2024 according to a secondary source; we could not open an MHRA page for it (Wikipedia).
- EU: lecanemab was authorised on 15 April 2025 and donanemab on 24 September 2025, each for people with one or no copy of ApoE4 and confirmed amyloid, under a controlled access programme (EMA Leqembi; EMA Kisunla). The EMA committee first recommended refusing lecanemab in July 2024 and donanemab in March 2025, then reversed both after re-examination (Bobbins et al., 2025; Wikipedia for the donanemab dates).
How they work
Both drugs bind amyloid-beta, a protein that builds up in the brain in Alzheimer's disease, and prompt the immune system to clear it. Lecanemab has high affinity for soluble amyloid protofibrils (CLARITY-AD). Clearance is proven in humans: lecanemab lowered PET amyloid by 59.1 centiloids more than placebo. What is not proven is that clearing amyloid is what slows symptoms. The Cochrane authors argue that clearance "did not translate" into clinically meaningful benefit, and that future work should look at other mechanisms (Nonino 2026). Defenders reply that the two approved drugs work differently from earlier failed antibodies and that Cochrane's pooling diluted this (Bernath et al., 2026). Those authors report industry ties including Eli Lilly, Eisai and Biogen, and also a US National Institute on Aging grant.
ARIA is thought to arise because amyloid also sits in the walls of small brain blood vessels, so removing it can leave vessels leaky. The sources we opened describe ARIA as a class effect of the antibodies and do not settle the mechanism, so we treat that as a hypothesis.
What they are prescribed for
Only early Alzheimer's disease: mild cognitive impairment or mild dementia with confirmed amyloid on PET or spinal fluid. The labels say treatment should start at this stage because it is the population studied (Leqembi label). The UK licence stops treatment if disease progresses to moderate (MHRA). Neither drug is shown to help moderate or severe dementia or to prevent Alzheimer's in people without symptoms.
NICE. A 2025 pharmacovigilance review stated that NICE had not recommended lecanemab for early Alzheimer's disease because of cost and uncertainty about whether the trial effect is clinically meaningful (Bobbins et al., BJCP 2025). Alzheimer's Research UK updated its page on 19 June 2025 to say that NICE and the Scottish Medicines Consortium have both announced lecanemab will not be available on the NHS in England and Scotland respectively, quoting NICE that the benefits "do not outweigh the significant financial costs" (ARUK). We could not open NICE's own documents (they blocked automated access) and could not verify NICE's position on donanemab or whether any appeal or reconsideration followed. Check the NICE website for the current position.
What works and what does not
| Claimed benefit | Verdict | Evidence | Key caveat |
|---|---|---|---|
| Clearing amyloid plaques | WORKS | -59.1 centiloids vs placebo (CLARITY-AD). | A marker, not a patient outcome. Maker-funded. |
| Slowing decline over 18 months | MIXED | CDR-SB -0.45 (lecanemab) and -0.7 (donanemab) vs placebo. | Statistically significant in maker-funded trials; below the usual meaningful-change thresholds. |
| Improving memory or restoring lost function | NOT SHOWN | All outcomes measured slower decline, not improvement. | Participants still declined on treatment. |
| Delaying the move to moderate dementia for years | INSUFFICIENT | Model projections only (Burn 2026). | Authors include Eisai employees; models assume constant progression rates. |
| Better quality of life or longer survival | INSUFFICIENT | Cochrane found no mortality increase (high certainty) but trial follow-up is short. | No independent long-term data. |
| Safe in ApoE4 carriers | NOT FOR HOMOZYGOTES | Highest ARIA rates in two-copy carriers. | UK and EU licences exclude them. |
Benefits by claim
How big is the effect?
CLARITY-AD randomised 1,795 people aged 50 to 90 to lecanemab 10 mg/kg every two weeks or placebo. On the 18-point CDR-SB scale, the placebo group worsened by 1.66 points and the lecanemab group by 1.21, a difference of 0.45 (95% CI 0.23 to 0.67). That is about a 27% relative slowing, which is the figure used in the sellers' publicity (our arithmetic: 0.45 divided by 1.66). Other differences were 1.44 points on ADAS-cog14 and 2.0 points on a daily-living scale (van Dyck 2023).
TRAILBLAZER-ALZ 2 randomised 1,736 people to donanemab or placebo every four weeks. On the 144-point iADRS, the combined-population difference was 2.92 points (95% CI 1.51 to 4.33), or 3.25 in people with low or medium tau. On CDR-SB it was -0.7 (-0.95 to -0.45) (Sims 2023). The FDA also reports an ADAS-Cog13 difference of -1.33 (FDA). The two trials used different populations and designs, so we do not compare the drugs head to head.
Effect size against the minimal clinically important difference
A statistically significant change is not necessarily a noticeable one. A rapid review of published thresholds found that in mild cognitive impairment the smallest meaningful change was about +1 CDR-SB point, -5 on the iADRS and +2 to +3 on ADAS-Cog; in mild Alzheimer's it was +2, -9 and +3. It concluded that the average effects in recent anti-amyloid trials were lower than previously published thresholds (Muir 2024). In plain terms: 0.45 CDR-SB is under half of the 1-point threshold, and 2.92 iADRS is about 58% of the 5-point threshold. A separate analysis funded by Roche found that a 12-month worsening of 1 to 2.5 CDR-SB points reflects minimal-to-moderate deterioration, which again places the observed differences at the very low end (Lansdall et al., 2023; funded by Roche, which has its own Alzheimer's programme).
Ebell and colleagues pooled 19 publications on eight antibodies (23,202 people) and judged each result against the threshold. The pooled CDR-SB difference was -0.18 points (95% CI -0.34 to -0.03), and no result exceeded the threshold, including for lecanemab and donanemab (Ebell 2024).
Independent critiques and the counter-argument
- Cochrane 2026 (17 industry-funded trials, 20,342 people): cognition SMD -0.11 (95% CI -0.16 to -0.06, moderate certainty), dementia severity SMD -0.12 (-0.24 to 0.00, low certainty), with efficacy outcomes at some risk of bias because side effects can reveal who received the drug (Nonino 2026). The authors conclude that effects are trivial.
- Rebuttal: a commentary argues the Cochrane pooling of seven antibodies dilutes the effect of the two plaque-targeting drugs and that the benefit is clinically meaningful (Bernath 2026). A neuroscientist's comment notes that the Cochrane critique was mostly about method and asks what "modest" benefit means in clinical practice. That author, funded by the Canadian Institutes of Health Research, is co-founder of a cognitive-testing company (Owen 2026).
- Safety viewpoint: a 2026 viewpoint argues ARIA, accelerated brain-volume loss and therapy-related deaths deserve independent scrutiny (Høilund-Carlsen et al., 2026). It is opinion, not new trial data.
- European specialists: 52 members of the European Alzheimer's Disease Consortium published a comment criticising the EMA committee's first negative opinion on donanemab. Most reported payments from Lilly, Eisai and other firms (Jessen et al., 2025). We could not read its arguments, only its title and disclosures.
Risks and all side effects
Both US labels carry a boxed warning that these antibodies can cause ARIA, that it can be fatal, and that serious brain haemorrhages larger than 1 cm have occurred. ApoE4 homozygotes have a higher incidence. Genetic testing is recommended before treatment (Leqembi label; Kisunla label).
ARIA and ApoE4
| Measure | Lecanemab | Donanemab |
|---|---|---|
| ARIA-E (brain swelling), overall | 13% vs 2% placebo | 24% vs 2% |
| ARIA-H (small bleeds) | 17% vs 9% | 31% overall |
| Any ARIA, two copies of ApoE4 | 45% vs 22% | 55% vs 22% |
| Any ARIA, one copy | 19% vs 9% | 36% vs 13% |
| Any ARIA, no copies | 13% vs 4% | 25% vs 12% |
| Brain haemorrhage over 1 cm | 0.7% (6 of 898) vs 0.1% (1 of 897) | 0.5% vs 0.2% |
| Infusion-related reactions | 26% vs 7% | 9% vs 0.5% |
All numbers come from the US labels (Leqembi; Kisunla), which reflect the makers' trials. The two trials differ in design, so the columns are not a fair head-to-head. On the lecanemab label, symptomatic ARIA-E was 9% in two-copy carriers, 2% in one-copy and 1% in non-carriers. Donanemab's symptomatic ARIA-E was 8%, 7% and 4%. Cochrane estimated 107 extra ARIA-E cases per 1,000 treated (95% CI 77 to 148) and 29 extra symptomatic cases per 1,000 (22 to 38) (Nonino 2026).
Deaths reported
- The donanemab trial reported 3 treatment-related deaths on donanemab and 1 on placebo (Sims 2023).
- Both labels state that fatal intracerebral haemorrhages have been observed. The Kisunla label describes one fatal haemorrhage in each of its two studies, both with ARIA and clot-dissolving treatment. The Leqembi label notes a fatal haemorrhage with an anti-amyloid antibody in a patient with ARIA symptoms who also received a thrombolytic.
- Cochrane found no increase in all-cause mortality: 2 more per 1,000 (95% CI 3 fewer to 11 more; high certainty), and no clear rise in serious adverse events (Nonino 2026).
- Alzheimer's Research UK states that four trial deaths were linked to lecanemab. We could not trace the primary source or timing of that figure (ARUK).
Other effects
Lecanemab: headache (11% vs 8%), rash, nausea or vomiting (6% vs 4%) and superficial siderosis (6% vs 3%). Donanemab: headache (13% vs 10%) and superficial siderosis (15% vs 3%). Both labels contraindicate serious hypersensitivity, including anaphylaxis and angioedema. Treatment burden is a harm too: frequent infusions and repeated MRI scans.
All interactions
| Drug or condition | Why it matters | Seriousness | Source |
|---|---|---|---|
| Anticoagulants (warfarin and similar) | Contraindicated in the UK licence; EU contraindicates uncontrolled bleeding disorders or anticoagulant treatment. The FDA notes more brain haemorrhages with anticoagulant use. | High | MHRA; EMA; FDA |
| Thrombolytics (clot-busting drugs for stroke) | Fatal haemorrhages occurred when given to people with ARIA symptoms. | High | Kisunla label |
| Other bleeding-risk drugs (aspirin, other antiplatelets) | The pages we opened do not give figures for these. Prescribers weigh them case by case. | Could not verify | No direct source opened |
We did not find interaction tables of the kind found for older medicines in the pages we opened. Cholinesterase inhibitors such as donepezil are often continued alongside these drugs, but we did not open a source that tests the combination.
Who should avoid these drugs
- People with two copies of ApoE4: the UK and EU licences exclude them. The MHRA's advisers said risk-benefit is not favourable for them (MHRA).
- People on anticoagulants or with an MRI showing cerebral amyloid angiopathy: contraindicated in the UK. The EU also lists previous brain bleeds, more than 4 microbleeds, superficial siderosis or vasogenic oedema (EMA).
- Moderate or severe dementia: not studied; treatment is stopped if disease reaches the moderate stage in the UK and EU.
- Those who cannot have regular MRI or who have a serious allergy to the drug or its ingredients.
- Pregnancy, breastfeeding and children: these drugs are for older adults with Alzheimer's disease. We did not open a source describing pregnancy data.
Dosage and how they are given
The treating specialist sets the plan. The figures below are US label doses for context only.
| Drug | Label dosing | Monitoring | Source |
|---|---|---|---|
| Lecanemab | 10 mg/kg by infusion every 2 weeks. After 18 months, the label allows continuing that dose or moving to 10 mg/kg every 4 weeks, or a weekly subcutaneous maintenance dose. | Baseline MRI; MRI at 1, 2, 3 and 6 months. The EU requires MRI before the 5th, 7th and 14th doses. | FDA label; EMA |
| Donanemab | Recommended step-up regimen: 350 mg, then 700 mg, then 1,050 mg, then 1,400 mg every 4 weeks. The label says to consider stopping when amyloid PET shows minimal levels. | Baseline MRI, then MRI before the 2nd, 3rd, 4th and 7th infusions. | FDA label |
The UK licence requires ApoE4 testing before treatment, a central registration system and a controlled access programme, with a post-authorisation safety study (MHRA). The EU requires a controlled access programme, an ARIA guide and an alert card for patients (EMA Kisunla).
Cost
- Private UK price: Alzheimer's Research UK says private clinics advertise lecanemab at £60,000 to £80,000 a year, and that NHS use would add costs for infusions, MRI and monitoring (ARUK).
- US list prices (secondary source): a Wikipedia entry gives US$26,500 a year for lecanemab and cites a US$32,000 list price for a year of donanemab. We could not open the primary company releases (lecanemab; donanemab).
- Value: an Australian model put donanemab at $342,424 per QALY from the healthcare perspective (Gao 2026). A review of 16 studies found no independent study that showed cost savings, with manufacturer-funded lecanemab models falling below thresholds and others above (Nguyen 2026).
Follow the money: who makes them and who funded the evidence
Who makes and sells them
- Lecanemab: developed jointly by Eisai, Biogen and BioArctic; Eisai Inc. is the US labeller. Eisai Co., Ltd. is based in Tokyo and Eisai Inc. in Nutley, New Jersey (Wikipedia). The EU authorisation holder is Eisai GmbH (EMA). We could not verify the profit-sharing terms between the three companies.
- Donanemab: Eli Lilly and Company (US); the EU holder is Eli Lilly Nederland B.V. (EMA).
Who funded the evidence
- CLARITY-AD: "Funded by Eisai and Biogen" (NEJM record).
- TRAILBLAZER-ALZ 2: sponsored by Eli Lilly. Authors Sims, Zimmer, Evans, Lu, Ardayfio, Shcherbinin, Monkul Nery, Collins, Brooks, Mintun and Skovronsky are employees or shareholders; two have pending patents related to the research (JAMA record).
- The Cochrane review: all 17 included trials were industry-funded; the review itself was publicly funded in Italy (Nonino 2026).
- The threshold study (Lansdall 2023): funded by Roche and written by Roche and RTI Health Solutions staff (record). Roche has its own amyloid-antibody programme, so it is an interested party, not a neutral one.
- Long-term modelling: the lecanemab projection by Burn and colleagues lists Eisai Inc. and Eisai Europe among the author affiliations (record).
Funding is not proof that a result is wrong. These trials were randomised, double-blind and peer-reviewed, and regulators saw the raw data. But no independent group has repeated either trial, so the effect size rests on the sellers' own analysis.
Research funding at a glance
12 sources graded. The counts below are the independence tiers assigned in the scorecard above. They count sources, not funding amounts or evidence quality.
Public funding alone does not prove independence, and commercial funding alone does not prove a claim false. Named funders, countries and relationships are in the scorecard and source notes.
Documented integrity events
We did not find, in the sources we opened, any regulatory sanction, warning letter or fraud finding against either trial. The documented controversy is about interpretation and cost, not data integrity. Both drugs were first refused by the EMA committee and later approved on re-examination in a narrower population (Bobbins 2025).
Related research
For wider, evidence-graded context, see these Pure City Research guides:
- Stroke: emergency signs, types, treatment and recovery
- Transient ischaemic attack
- Anticoagulants and bleeding safety
- Antiplatelet medicines
- Depression in older adults
- Sleep disorders in older adults
- Creatine monohydrate: cognition and safety
Frequently asked questions
Do lecanemab and donanemab cure Alzheimer's disease?
No. In the pivotal trials people on the drugs still declined, only more slowly. Lecanemab's difference was 0.45 CDR-SB points over 18 months (CLARITY-AD), and donanemab's was 0.7 (TRAILBLAZER-ALZ 2).
Will a person on treatment notice a difference?
Probably not in a way you could point to. Independent reviewers put the smallest noticeable change in mild cognitive impairment at about 1 CDR-SB point, and the average trial effects were lower (Muir 2024). Some individuals may do better than the average, but the trials do not tell us who.
What is ARIA and how serious is it?
ARIA is brain swelling (ARIA-E) or small bleeds (ARIA-H) found on MRI. Most cases cause no symptoms, but both US labels warn it can be serious and fatal. Symptomatic ARIA-E occurred in 2% of one-copy ApoE4 carriers on lecanemab and 9% of two-copy carriers (FDA label).
Why is an ApoE4 gene test needed?
Because two copies raise the risk of ARIA sharply. The MHRA advises testing before treatment, and its advisers did not find the balance favourable for people with two copies (MHRA). The test result also has implications for relatives, which is a conversation for a genetic counsellor.
Is lecanemab available on the NHS?
According to Alzheimer's Research UK, NICE's final decision in June 2025 was that it will not be available on the NHS in England, with the Scottish Medicines Consortium reaching the same conclusion for Scotland (ARUK). We could not open NICE's own pages to confirm, or check for later changes.
Why did the FDA approve it if the benefit is small?
The FDA said the confirmatory trial verified clinical benefit, and its advisory committee voted unanimously to that effect on 9 June 2023 (FDA). Regulators ask whether the effect is statistically real and whether the risks are acceptable. NICE and Cochrane ask whether the effect is large enough to justify the cost and harm, which is a different question.
Has the Cochrane review settled the argument?
Not entirely. Cochrane found trivial effects across seven antibodies, but critics say pooling failed drugs with the two approved ones understates the result (Bernath 2026). Those critics have disclosed ties to Lilly, Eisai and Biogen, and the Cochrane team has public funding. Both sides are working from the same industry trials.
Do these drugs replace other Alzheimer's care or support?
No. They are not a replacement for care, safety planning or other treatments. Our Alzheimer's disease review, which covers tests, other drugs and prevention, is in preparation and not yet live.
Sources and funding notes
What we could not verify: NICE's own documents (blocked), including its position on donanemab and any appeal; the MHRA's own donanemab announcement; primary company price releases; the profit-sharing terms between Eisai, Biogen and BioArctic; the origin of the "four deaths" figure that ARUK mentions; the full-text funding statements of Ebell 2024 and Gao 2026; the content of the European Alzheimer's Disease Consortium letter; and any independent replication of the pivotal trials. The 6 July 2023 and 22 August 2024 dates come from the FDA and MHRA pages we opened.
- van Dyck CH et al. Lecanemab in Early Alzheimer's Disease (CLARITY-AD). N Engl J Med 2023 — pivotal lecanemab trial; funded by Eisai and Biogen.
- Sims JR et al. Donanemab in Early Symptomatic Alzheimer Disease (TRAILBLAZER-ALZ 2). JAMA 2023 — pivotal donanemab trial; Eli Lilly sponsor; authors include Lilly employees.
- Leqembi (lecanemab-irmb) prescribing information, DailyMed (revised July 2026) — US regulator-approved label; Eisai Inc.
- Kisunla (donanemab-azbt) prescribing information, DailyMed — US regulator-approved label; Eli Lilly.
- FDA: traditional approval of Leqembi (6 July 2023) — US government.
- FDA: approval of Kisunla (July 2024) — US government.
- MHRA: lecanemab licensed (22 August 2024) — UK government regulator.
- EMA: Leqembi EPAR page and Kisunla EPAR page — EU regulator.
- Nonino F et al. Amyloid-beta-targeting monoclonal antibodies. Cochrane Database Syst Rev 2026 — funded in part by Emilia-Romagna region and Italian Ministry of Health; all 17 trials industry-funded.
- Ebell MH et al. Clinically important benefits and harms of anti-amyloid antibodies. Ann Fam Med 2024 — US academic; no funding statement seen in the record.
- Muir RT et al. Minimal clinically important difference in Alzheimer's disease: rapid review. Alzheimers Dement 2024 — Canada; no funding section in record; two authors report Lilly, Eisai or Roche ties.
- Lansdall CJ et al. Establishing clinically meaningful change on outcome assessments. J Prev Alzheimers Dis 2023 — funded by Roche; Roche and RTI Health Solutions authors.
- Bobbins A et al. Lecanemab: the pharmacovigilance perspective. Br J Clin Pharmacol 2025 — UK; funding not stated; no conflicts declared.
- Gao L et al. Cost-effectiveness of donanemab in Australia. Med J Aust 2026 — no funding listed; one author reports Lilly honoraria.
- Nguyen TT et al. Cost-effectiveness analyses of monoclonal antibodies for Alzheimer's disease. Alzheimers Dement (N Y) 2026 — UC Davis; funding not stated; no conflicts declared.
- Burn O et al. Long-term outcomes of lecanemab: a modelling study. J Med Econ 2026 — authors affiliated with Eisai; funding not visible in the record.
- Bernath MM et al. Beyond the Cochrane review. Alzheimers Dement 2026 — NIA grant P30 AG072976; authors report ties to Lilly, Eisai, Biogen and others.
- Owen AM. What have trials targeting amyloid told us to date? J Neurol 2026 — Canada; funded by CIHR; author co-founded Creyos Inc.
- Høilund-Carlsen PF et al. Serious side effects of Alzheimer's immunotherapy demand scrutiny. Ageing Res Rev 2026 — international viewpoint; one author reports pharmaceutical consulting.
- Jessen F et al. Comment by EADC investigators on the CHMP negative recommendation on donanemab. J Prev Alzheimers Dis 2025 — European specialists; most report Lilly, Eisai or other industry ties.
- Alzheimer's Research UK: lecanemab and NICE (updated 19 June 2025) — UK charity; funded by donations; secondary source for NICE.
- Wikipedia: Lecanemab, Donanemab and Eisai (company) — secondary sources, used only for developer, price and date details we could not confirm from primary pages.
Educational research reviewed 6 October 2026. This article is not medical advice; talk to a qualified clinician or pharmacist before starting, stopping or combining treatments.
Have a question — or want us to cover something?
Ask about anything on this page, or request the next deep dive: an ingredient, a supplement, or a health concern. We use published research, evidence syntheses, and regulatory guidance, with clear source links.
One daily research roundup
Get the topics, key findings and links from our new articles in one email. At most one digest a day, only when there is something new.