Lecanemab’s Five-Year Follow-Up: From Longer Disease Control to a Durable Treatment Business
BIIB
Executive Summary
Longer follow-up adds evidence for sustained disease control. External comparisons, treatment retention, ARIA monitoring and Biogen’s collaboration accounting determine how far that evidence can support a commercial forecast. BioScore unverified; no editorial investment rating assigned.
An Alzheimer’s treatment creates value when it preserves abilities that patients and families want to keep: following a conversation, managing a routine, or needing less help. That makes duration an important question for lecanemab, the amyloid-directed antibody marketed as LEQEMBI. A short-term difference matters more if it can become a sustained difference in daily life.
The five-year Clarity AD extension update presented at CTAD adds a longer observation window to that discussion. For investors in Biogen, its commercial relevance is a stronger basis for discussing sustained treatment. It does not, by itself, establish how many newly diagnosed patients will start treatment, remain on it, or generate profitable revenue. Those are connected but separate propositions.
Start with the randomized clinical value
The original Clarity AD study randomized 1,795 people with amyloid-confirmed early Alzheimer’s disease: 898 to lecanemab and 897 to placebo. At 18 months, deterioration on CDR-SB, a scale combining cognition and function, averaged 1.21 versus 1.66 points. The adjusted difference was −0.45 points, with a 95% confidence interval of −0.67 to −0.23 and P<0.001. Both groups worsened on average; the treated group worsened less. [2]
That result is a useful anchor. Random assignment supports attributing the difference to treatment within the study’s population and conditions. Statistical significance addresses uncertainty around the contrast; it does not tell a family exactly which ability an individual will retain, or for how long. A group mean is not an individual promise.
What the longer follow-up adds
The company’s US announcement is dated November 16, 2026, while its international dateline is November 17. It reports 351 participants with five years of continuous treatment. At five years, the reported CDR-SB difference from an external ADNI natural-history comparison was −2.05 points. Reported progression to a subsequent disease stage was 61.9% with treatment versus 77.5% in the external comparison, a 15.6-percentage-point difference. These are company-reported extension analyses, not five-year randomized placebo comparisons. [1]
The clinical opportunity is meaningful: slowing progression over a longer period could preserve independence. The company interprets the widening separation as accumulating benefit. Our interpretation is narrower: the update strengthens the case for investigating durable clinical value, while the exact long-term causal effect remains less certain than the original randomized contrast.
An external control can supply a longer view when continued placebo treatment is unavailable. Its credibility depends on how closely patients, disease severity, assessment schedules and subsequent care align. Statistical adjustment can address measured differences; it cannot guarantee that unmeasured differences have disappeared.
The 15.6-point percentage difference therefore should not be inverted into a causal “number needed to treat.” Nor should the −2.05-point external-control estimate be divided by the −0.45-point randomized result to claim that efficacy multiplied by a particular amount. The comparator and the analysis changed alongside the duration.
The denominator matters for the business as well
The 351 long-exposed participants are not a verified commercial persistence rate. Calculating 351 divided by all originally randomized patients would mix eligibility, treatment assignment, time available to reach five years, discontinuation and follow-up. A proper persistence analysis needs the eligible cohort, numbers at risk over time, reasons for leaving, interruptions and handling of missing outcomes.
The peer-reviewed extension publication available for review covers up to 36 months, with a March 31, 2024 cutoff. It describes extension eligibility and its analytic framework, but cannot substitute for the complete five-year patient flow or analysis materials. The five-year headline estimates reviewed here did not provide the confidence intervals and full attrition information needed to independently reconstruct their precision and sensitivity to missing data. [1,4]
This is not a reason to discard extension evidence. It is a reason to identify the decision it can inform. Observations among continuing participants can inform treatment discussions; they provide a weaker basis for forecasting the experience of every new starter.
Longer use still carries a monitoring burden
The current FDA label restricts initiation to mild cognitive impairment or mild dementia stages and requires confirmation of amyloid pathology. It carries a boxed warning for amyloid-related imaging abnormalities, or ARIA, including edema and bleeding-related changes. Serious and fatal events can occur; ApoE ε4 homozygotes face greater risk. Baseline and follow-up MRI requirements are part of the treatment pathway. [3]
In the label’s core-study safety population, ARIA-E occurred in 113/898 lecanemab recipients versus 15/897 placebo recipients; ARIA-H occurred in 152/898 versus 80/897. These are approximately 18-month core-study observations, not five-year cumulative rates. Lower event frequency later among continuing participants would not erase early risk or demonstrate that every future starter will tolerate treatment. [3]
For investors, monitoring is an operating condition rather than a footnote. A durable treatment business requires diagnostic access, MRI capacity, clinicians able to manage adverse findings, and patients and caregivers able to sustain the treatment routine. The clinical value proposition has to be strong enough to justify that burden for an appropriately selected patient.
Translate clinical duration into financial questions carefully
Biogen’s third-quarter 2026 filing reports $68.4 million of Alzheimer’s collaboration revenue for the three months ended September 30 and $191.6 million for the nine-month period. The filing explains that this line reflects its 50% share of LEQEMBI net product revenue and cost of sales, including royalties, because Biogen is not the principal. It is neither worldwide LEQEMBI product sales nor Biogen’s bottom-line profit or free cash flow. Doubling it would not reconstruct worldwide net sales. [5]
The investment hypothesis is that credible long-term benefit could improve the willingness to diagnose early, begin therapy and continue appropriate treatment. That hypothesis needs evidence at each transition. More eligible patients do not automatically become starts; more starts do not automatically become long-duration exposure; longer exposure does not automatically produce attractive incremental margins.
The most useful commercial evidence would connect newly treated cohorts with subsequent continuation, interruptions, safety management, reimbursement and realized economics. Market-level revenue growth alone cannot reveal which part of that chain improved.
Position in the field and the next tests
Lecanemab operates in a field that includes donanemab, another FDA-approved amyloid-directed treatment for early Alzheimer’s disease. The five-year lecanemab update contains no randomized head-to-head comparison that establishes superiority over it. Differences across development programs should guide questions, not a numerical league table. [3,6]
Within the lecanemab program, ongoing studies in earlier disease settings ask different questions from this extension. The company identifies AHEAD 3-45 and DIAN-TU Tau NexGen among continuing investigations. Their existence is not evidence that prevention, broader eligibility or a particular combination strategy has already succeeded. [1]
For the present market, an especially valuable question is the incremental benefit of continued treatment after the initial phase: which patients need which maintenance approach, and what happens when treatment is interrupted or stopped? The extension offers observations, but it is not a randomized continuation-versus-discontinuation answer.
What would change the assessment
Full five-year methods, confidence intervals, patient flow and sensitivity analyses could make the long-term estimate more or less persuasive. Better evidence on daily function and care needs would sharpen its clinical meaning. Real-world cohorts linking initiation to durable use and financial reporting would strengthen the commercial bridge.
The current update adds a longer horizon to an established treatment effect. It supports a serious discussion about preserving time in earlier disease stages. A sound investment judgment keeps that opportunity visible while requiring separate evidence for durable adoption, manageable treatment burden and the economics attributable to Biogen.
Sources and evidence scope
Information reviewed November 17, 2026. The new five-year findings are company conference disclosures. The randomized trial and earlier 36-month extension are peer-reviewed publications. A complete five-year paper, full analysis plan for the new comparisons and reconstructable patient-flow tables were not obtained; unpublished details are not assumed. [1] Eisai/Biogen, CTAD five-year company disclosure; US page dated November 16, 2026, release dateline November 17. Conference disclosure, not a new peer-reviewed five-year report. https://media-us.eisai.com/2026-11-16-New-Five-Year-Lecanemab-Data-from-Clarity-AD-Study-Long-Term-Extension-Presented-at-CTAD-2026
[2] van Dyck et al., randomized Clarity AD trial; online November 29, 2022. DOI: 10.1056/NEJMoa2212948. Registration: NCT03887455. https://mediacenteratypon.nejmgroup-production.org/NEJMoa2212948.pdf
[3] FDA, LEQEMBI US prescribing information, 2026 version, including July 2026 dosing changes and boxed ARIA warning. https://www.accessdata.fda.gov/drugsatfda_docs/label/2026/761375s001lbl.pdf
[4] van Dyck et al., peer-reviewed open-label extension report, 2025; DOI: 10.1002/alz.70905. Up to 36 months; data cutoff March 31, 2024. This is not the five-year dataset. https://alz-journals.onlinelibrary.wiley.com/doi/10.1002/alz.70905
[5] Biogen, Form 10-Q for the quarter ended September 30, 2026; Note 5 and revenue discussion. Financial reporting, not a clinical outcome source. https://investors.biogen.com/node/31091/html
[6] FDA, donanemab approval announcement, July 2, 2024; competitor indication context, not a head-to-head comparison. https://www.fda.gov/drugs/news-events-human-drugs/fda-approves-treatment-adults-alzheimers-disease