Lecanemab vs Donanemab: How Two Alzheimer’s Drugs Differ

Lecanemab and donanemab are both monoclonal antibodies that clear amyloid from the brain in early Alzheimer’s disease, but they differ in meaningful ways: what form of amyloid they grab onto, how they are dosed, how long treatment lasts, and their safety profiles. Both slowed cognitive decline by roughly a quarter to a third over about 18 months in their pivotal trials, yet the details underneath those headline numbers matter for deciding which drug fits a given patient. The picture is still sharpening as real-world data and longer-term follow-up come in.

Different Targets on the Same Protein

Amyloid-beta, the protein at the center of both therapies, exists in multiple forms in the brain. Lecanemab is engineered to bind protofibrils, which are soluble clumps of amyloid that are thought to be particularly toxic to neurons. Structurally, lecanemab recognizes amyloid fragments that are only slightly trimmed, missing just one or two amino acids from their original sequence, which helps the antibody avoid latching onto the heavily degraded amyloid circulating in the bloodstream or already locked inside dense plaques.¹ Donanemab, by contrast, homes in on a chemically modified version of amyloid found predominantly in established plaques. Its target is a pyroglutamate modification at position 3 of the amyloid peptide, a feature that marks older, deposited plaque material rather than newly forming aggregates.²

This distinction has practical consequences. Because donanemab targets plaque itself, it tends to strip amyloid from the brain faster and more completely, which is why the drug’s trial design included stopping treatment once amyloid levels dropped below a threshold. Lecanemab’s affinity for soluble protofibrils means it is also clearing toxic intermediates in the spaces between cells, which some researchers argue could matter independently of total plaque burden. Whether one mechanism ultimately proves more beneficial than the other remains an open question; the field does not yet have a head-to-head randomized trial between the two drugs.

What the Pivotal Trials Found

Lecanemab’s phase 3 trial, Clarity AD, enrolled roughly 1,800 people with early Alzheimer’s disease. After 18 months, participants receiving lecanemab showed less decline on the primary outcome measure, CDR-SB (a scale that tracks both cognition and daily function), than those on placebo, with an adjusted difference of about 0.45 points favoring the drug. Cognitive subscales and a measure of daily living activities also favored lecanemab, all reaching statistical significance.³

Donanemab’s pivotal trial, TRAILBLAZER-ALZ 2, enrolled about 1,700 participants and ran for 76 weeks. Among people with low or moderate levels of tau pathology, the CDR-SB difference versus placebo was roughly 0.67 points. In the broader combined population that also included people with higher tau, the benefit was similar at about 0.7 points.⁴ Donanemab’s effect thus appeared numerically a little larger than lecanemab’s on CDR-SB, but the trials differed in design, duration, and patient selection, making direct comparison tricky.

A network meta-analysis that pooled data from 33 randomized trials of various anti-amyloid antibodies found that donanemab ranked highest for amyloid clearance on PET scans and for scores on the ADAS-cog cognitive scale, while lecanemab ranked highest for preserving daily function as measured by the ADCS-ADL.⁵ In short, neither drug clearly dominates the other across every clinical measure. Both produced modest slowing of decline over 18 months, and both achieved statistical significance.⁶

Translating Trial Numbers Into Daily Life

A fraction-of-a-point difference on a rating scale is hard to visualize. One analysis tried to convert those CDR-SB changes into something more tangible by estimating how long treatment might preserve a person’s ability to manage day-to-day tasks independently. For someone starting at a mild disease stage, lecanemab was projected to extend independence in daily activities by about 10 months, and donanemab in the low-to-medium tau group by about 13 months, compared with no treatment.⁷ Those estimates carry wide confidence intervals, but they at least anchor the benefit in terms families care about: months of retained independence rather than abstract scale points.

Long-term extension data from the donanemab program adds another angle. At three years, participants who started donanemab early continued to show slower decline on CDR-SB compared with a matched external control group. People who began donanemab later, after the initial placebo period ended, also benefited, but early starters had a lower risk of progressing to the next clinical stage over three years.⁸ Whether the same pattern will hold for lecanemab at three years is not yet known from published extension data.

Safety and Brain Swelling Risk

The most talked-about side effect of both drugs is ARIA, short for amyloid-related imaging abnormalities. ARIA comes in two flavors: brain swelling or fluid buildup (ARIA-E) and tiny brain bleeds or surface iron deposits (ARIA-H). Most ARIA episodes are caught on routine MRI scans and never cause symptoms a patient would notice. But a minority produce headaches, confusion, or dizziness, and in rare cases the consequences can be serious.

A network meta-analysis of phase 2 and 3 randomized trials found that donanemab carried a higher risk of ARIA-E than lecanemab, ranking alongside aducanumab at the top of the risk spectrum.⁹ Indirect comparison studies have put the overall ARIA risk difference between the two drugs at roughly 10 percentage points, with lecanemab showing lower rates of both ARIA-E and ARIA-H. The gap widened further among people who carry the APOE4 gene variant, a known amplifier of ARIA susceptibility.

Infusion reactions are another safety consideration, though these are largely a nuisance rather than a danger. Among more than 2,700 donanemab-treated participants across the program, about 8% reported an infusion-related reaction. The majority were mild or moderate, and most first reactions occurred within the first four infusions. Serious infusion reactions were rare, occurring in about half a percent of patients. Lecanemab also causes infusion reactions, though the rates in Clarity AD were somewhat lower. A subcutaneous formulation, discussed below, may sidestep many infusion-related issues entirely.

Who Benefits Most and Who Should Be Cautious

Tau pathology turns out to be one of the strongest predictors of how much benefit a person gets from amyloid removal. In the donanemab trial, people with low or medium tau on PET scans had a more favorable response, while those with high tau levels saw a smaller benefit. Analysis of both the Clarity AD and TRAILBLAZER-ALZ 2 programs suggests that patients with lower tau burden represent a distinct group in whom plaque clearance is most likely to translate into clinical stabilization or functional gain.¹⁰ The practical problem is that tau PET scans are expensive and not widely available, so most treatment decisions today are still made without this information.

People who carry two copies of the APOE4 gene variant face a different calculus. A Bayesian reanalysis of both phase 3 trials found that APOE4 homozygotes showed no clear clinical benefit from either drug, while facing a substantially higher risk of ARIA and related complications. The authors argued that excluding these patients from treatment is justifiable given the unfavorable risk-benefit balance.¹¹ This does not apply to people with just one APOE4 copy, who can still benefit, though they need closer monitoring for ARIA.

A common concern among clinicians has been whether patients taking blood thinners face extra danger from ARIA-related brain bleeds. A meta-analysis of clinical trial and real-world data found no significant increase in ARIA-E or ARIA-H rates among patients using oral anticoagulants alongside anti-amyloid therapy. This is reassuring for the many older adults who take blood thinners for atrial fibrillation or other conditions, though careful monitoring remains standard practice.

How Treatment Is Given

Lecanemab and donanemab differ sharply in how they are administered, and this matters for the lived experience of patients and caregivers. Lecanemab was initially approved as a biweekly intravenous infusion, each session lasting about an hour. That schedule means roughly 26 infusion visits per year, a considerable commitment for patients who may already be dealing with transportation and scheduling challenges. A subcutaneous formulation has since been developed, deliverable by autoinjector, with pharmacokinetic studies showing that the autoinjector produced drug exposure about 20 to 25 percent higher than a prefilled syringe, and both routes achieved a half-life of about seven days.¹² A self-administered injection at home could dramatically reduce the treatment burden.

Donanemab is given as a monthly intravenous infusion, which already halves the visit frequency compared to lecanemab’s biweekly IV schedule. More distinctively, donanemab is designed as a time-limited treatment. In the TRAILBLAZER-ALZ 2 trial, patients who achieved sufficient amyloid clearance on PET stopped receiving the drug, with some completing treatment as early as one year. Extension data showed that people who finished treatment by 52 weeks maintained the same degree of slowed progression at three years as those treated longer.¹³ That prospect of stopping treatment is appealing not just for convenience but for reducing cumulative exposure to ARIA risk and holding down costs.

The Brain Volume Puzzle

One counterintuitive finding across anti-amyloid trials is that treated patients actually lost more brain volume on MRI scans than those on placebo, despite doing better clinically. This initially alarmed researchers, since brain shrinkage is normally associated with worsening Alzheimer’s disease. A review in The Lancet Neurology argued that this volume loss is best understood as “pseudo-atrophy,” reflecting the physical removal of amyloid plaques and the resolution of associated inflammation rather than true neuronal damage. The excess volume changes occurred only with antibodies that successfully lowered amyloid, and the available evidence did not link them to worse outcomes.¹⁴ Both lecanemab and donanemab trigger this effect. It matters mostly because clinicians and patients should expect to see it on scans and not mistake it for a sign of harm.

Tracking Response With Blood Tests

Knowing whether a drug is actually doing something in a given patient’s brain without requiring frequent PET scans would be a game-changer. Plasma biomarkers are getting closer to filling that role. In the earlier donanemab trial, blood levels of phosphorylated tau 217, a marker of active amyloid and tau pathology, dropped by about 23% from baseline after donanemab treatment, while levels in the placebo group rose by 6% over the same period. The decline in this blood marker correlated with the amount of amyloid clearance seen on PET scans.¹⁵ Donanemab treatment also reduced tau accumulation in cortical brain regions by about a third compared with placebo, particularly in the temporal, parietal, and frontal lobes among patients who achieved full amyloid clearance.¹⁶ These biomarker shifts hint that clearing amyloid may slow the downstream spread of tau, the protein more directly linked to nerve cell death, though establishing a firm causal chain requires more work.

Cost and the Access Problem

Both drugs carry annual list prices in the range of $26,000 to $32,000 per year in the United States, before considering infusion center costs, MRI monitoring (typically required at minimum four to five times in the first year), and amyloid PET or spinal fluid testing to confirm eligibility. A systematic review of cost-effectiveness analyses found that neither lecanemab nor donanemab met the commonly used willingness-to-pay threshold of $100,000 per quality-adjusted life year. Even at a more generous $150,000 threshold, estimated value-based prices for lecanemab and donanemab came in below roughly $16,000 and $34,000 per year, respectively.¹⁷

A separate modeling study looked at lecanemab from a different angle, comparing it against collaborative care programs that coordinate medical, social, and caregiver support. Collaborative care alone produced far more population-level health gains and actually saved money, yielding an estimated 1.5 million quality-adjusted life years and $300 billion in savings when scaled nationally. Adding lecanemab on top of collaborative care generated an additional 180,000 quality-adjusted life years at a cost of about $39.5 billion, working out to roughly $218,000 per QALY from a societal perspective.¹⁸ The implication is uncomfortable but worth stating plainly: for the money spent on one of these drugs, a care coordination program could help more patients, and the drugs work best when layered on top of good care rather than substituting for it.

Donanemab’s time-limited dosing could shift the economics in its favor. If many patients can complete treatment within 12 to 18 months, total drug costs fall substantially compared with an indefinite regimen. However, the monitoring costs and infrastructure demands remain front-loaded regardless of which drug is used.

Early Real-World Experience

Trial populations are carefully selected: participants tend to be younger, healthier, and less diverse than the broader Alzheimer’s population. Real-world data is beginning to appear but is still limited. A single-site clinical practice cohort treated with lecanemab found safety and efficacy patterns broadly consistent with the trial, though the authors acknowledged the limitations of a small, non-randomized dataset.¹⁹ In Europe, one early tertiary center reported starting 32 patients on anti-amyloid therapy, with donanemab prescribed about three times as often as lecanemab, and one patient switching from lecanemab to donanemab.²⁰ The small numbers underscore how early we are in the rollout. Infrastructure for amyloid confirmation, MRI monitoring, and infusion delivery remains concentrated in academic medical centers, creating geographic and socioeconomic barriers to access that will take years to dismantle.

The patients who reach treatment in the real world tend to be the most informed and best-connected, which is a different problem from whether the drugs work. Whether these therapies can be delivered equitably at scale, and whether they produce the same modest benefits outside the controlled trial environment, are questions that will define the next chapter of Alzheimer’s treatment far more than any head-to-head efficacy comparison.