Five Years of Leqembi: The Gap Against Expected Alzheimer's Decline Kept Widening

Editorial Note: This article provides a solution-oriented perspective on this story while maintaining complete factual accuracy. Original reporting available at Eisai and Biogen, BioArctic.
2.05 CDR-SB points less decline at five years than ADNI natural-history data
62% of low-tau patients scored better on the CDR-SB than at the start
36% lower relative risk of reaching the next disease stage than in ADNI
351 patients treated with lecanemab for five continuous years

After five years on lecanemab, 351 people with early Alzheimer's had declined 2.05 points less on the Clinical Dementia Rating-Sum of Boxes (CDR-SB) than the disease's expected course. That gap grew every year. It stood at 1.00 points at year three and 1.78 at year four, measured against natural-history data from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Eisai and Biogen presented the figures at the 19th Clinical Trials on Alzheimer's Disease (CTAD) conference in Boston on November 16, 2026. Against a second reference group, the BioFINDER cohort, the five-year gap reached 2.54 points.

The placebo group ended with the original 18-month trial, so the five-year numbers set treated patients against outside cohorts, not a randomized control. Lecanemab, sold as Leqembi, is an antibody that clears amyloid, the protein that clumps into plaques in the Alzheimer's brain.

What the scale measures

The CDR-SB runs from 0 to 18, and higher scores mean more impairment. Clinicians rate memory, thinking and everyday function after interviewing the patient and someone who knows them well. Patients in Clarity AD started at an average of 3.2 points. Over the 18-month core trial, published in the New England Journal of Medicine in 2023, the lecanemab group worsened 1.21 points and the placebo group 1.66, a difference of 0.45. The European Academy of Neurology summarizes that as 27 percent slower decline. The five-year line extends that same gap more than three years past the point where the placebo comparison stopped.

The patients who improved

The sharpest numbers come from a subgroup. Clarity AD ran an optional tau PET substudy, a brain scan for tau, the protein that forms tangles inside nerve cells. BioArctic, the Swedish company that partners with Eisai on the drug, describes patients with low tau as being at an earlier stage of disease. Among them, 66 percent showed no decline on the CDR-SB after five years, and 62 percent scored better than at baseline.

Other scales moved the same way. On the ADAS-Cog14, a cognition test, 55 percent showed no decline and 52 percent improved. On the ADCS MCI-ADL, a measure of daily activities, 69 percent showed no decline and 62 percent improved. Improvement from baseline makes a stronger claim than slower decline. Neither release states how many patients sat in the low-tau group, so these shares come without a denominator.

Staying in the early stage longer

The progression figures turn scale points into stages. After five years, 61.9 percent of treated patients had moved to the next stage of disease, such as mild cognitive impairment to mild dementia, against 77.5 percent in the ADNI data. The companies put that at a 36 percent lower relative risk. Progression to moderate or severe dementia split wider: 31.0 percent of treated patients against 48.7 percent in ADNI, a 49 percent lower relative risk.

The core trial enrolled 1,795 people with early Alzheimer's and confirmed amyloid. Of those, 898 received lecanemab at 10 mg/kg by infusion every two weeks, and 897 received placebo. After 18 months, 95 percent of the people who finished the core study joined the extension. Continuous treatment means a patient took lecanemab from day one, so the 351 five-year patients come from the original 898, about 39 percent of them. The releases do not say why the others stopped.

Safety held steady

The companies reported no new safety findings across five years. Amyloid-related imaging abnormalities with edema or effusion (ARIA-E), swelling or fluid seen on brain scans, defined the early worries about this class of drug. ARIA-E struck most often in the first six months. It affected 13.5 percent of patients in the first 12 months, then 3.6 percent, 1.1 percent, 1.3 percent and 0.6 percent in the years that followed. The placebo group had a 1.7 percent rate over 18 months. Patients who sign up for years of treatment face most of that risk up front.

Where the drug stands now

The FDA has approved lecanemab since July 2023, and it now holds approval in 53 countries and regions, with six more reviewing it. A subcutaneous form, LEQEMBI IQLIK, won U.S. approval for maintenance dosing in August 2025 and for starting treatment in July 2026. China and Japan approved the subcutaneous starting dose in September 2026. Eli Lilly's donanemab joined the same early-Alzheimer's field with FDA approval in July 2024.

📍 Multiple Perspectives on the Story

🩺 The Clinician Perspective

Two reference groups, one direction

A single comparison cohort could flatter a result by chance. This one ran against two. Measured against BioFINDER, the gap was 1.42 points at year three and 2.24 at year four, both wider than the ADNI figures. For doctors talking with newly diagnosed patients, a benefit that keeps growing turns the conversation from months gained toward years spent at an earlier stage.

🔬 The Early-Detection Perspective

Find it before tau spreads

If treatment works best when tau is low, the bottleneck moves from the drug to the diagnosis. The AHEAD 3-45 trial, running since July 2020, tests lecanemab in clinically normal people who carry intermediate or elevated brain amyloid. Its results will show whether starting before symptoms widens the window further.

🏥 The Access Perspective

Easier schedules, untested over five years

The five-year cohort took infusions every two weeks. Patients now have lighter options: monthly intravenous maintenance is approved in nine countries and regions and filed in eleven more, and the subcutaneous starting dose is under review in four countries. Long-term benefit only reaches people who stay on treatment, though this dataset does not show whether the newer schedules produce the same five-year curve.

👨‍👩‍👧 The Family Caregiver Perspective

Half the scale is life at home

Three of the CDR's six rated areas cover community affairs, home and hobbies, and personal care, according to the Knight Alzheimer Disease Research Center at Washington University. The other three cover memory, orientation, and judgment and problem solving. A family member or close friend supplies part of the picture in the interview, so a slower-rising score partly records what the people at home see.

⚖️ The Skeptic Perspective

Who stayed, and who they are compared with

Patients who remain on an open-label drug for five years may differ from those who left. If people stopped because of side effects or faster decline, the remaining average looks better than the full group would. Neither release explains how the treated patients were matched to the ADNI and BioFINDER volunteers, who were never screened by the trial's entry rules. The companies also report improvement only for the low-tau subgroup. For most patients, the data describe slower decline, not a reversal.

What is known and what is still open

In this cohort, the distance between treated patients and the expected course of early Alzheimer's grew each year through year five, against two separate reference groups, and the safety picture matched the 18-month trial. Three things stay open. The five-year data came from a conference presentation and company releases and have not appeared in a peer-reviewed paper. The program has not reported how patients with higher tau at the start fared over the same five years. And AHEAD 3-45 has yet to show whether treating before symptoms appear moves the line further.

Frequently Asked Questions

What did the five-year lecanemab data show?

Among 351 people treated with lecanemab (Leqembi) for five continuous years, the gap against the expected course of early Alzheimer's grew each year on the CDR-SB scale: 1.00 points at year three, 1.78 at year four and 2.05 at year five, compared with ADNI natural-history data. Eisai and Biogen presented the data at the CTAD conference in Boston on November 16, 2026.

Did any patients get better on lecanemab?

In a subgroup with low levels of tau, a protein marker of later disease, 62 percent scored better on the CDR-SB after five years than at the start, and 66 percent showed no decline. The companies did not say how many people were in that subgroup, and for most patients the drug slows decline rather than reversing it.

Was there a placebo group for the five-year comparison?

No. The placebo comparison ended with the 18-month core trial. The five-year figures compare treated patients with two outside natural-history cohorts, ADNI and BioFINDER, and the data have not yet appeared in a peer-reviewed paper.

Is lecanemab safe to take for years?

The companies reported no new safety findings over five years. Brain swelling seen on scans (ARIA-E) affected 13.5 percent of patients in the first 12 months, then fell each year to 0.6 percent in the fifth year. The placebo group had a 1.7 percent rate over 18 months.