Why the new Alzheimer's drugs check APOE4 first
By MyGeneLog™ Team · Updated September 14, 2026 · New research
The two anti-amyloid Alzheimer's drugs approved in recent years, lecanemab (Leqembi) and donanemab (Kisunla), both come with a genuinely serious safety consideration: ARIA, short for amyloid-related imaging abnormalities — swelling (ARIA-E) or small bleeds (ARIA-H) in the brain, visible on MRI, sometimes symptomatic. And the risk of ARIA is not the same for everyone taking these drugs. It depends heavily on how many copies of the APOE ε4 allele someone carries.
The numbers, from the trial itself
The CLARITY AD trial — the randomized, placebo-controlled study behind lecanemab's approval, published in the New England Journal of Medicine in 2023 — enrolled 1,795 people and stratified randomization by APOE carrier status specifically, because the investigators already expected genotype to matter. It did, sharply:
- ARIA-E (swelling): 5.4% in people with no ε4 allele, 10.9% in heterozygotes (one copy), 32.6% in homozygotes (two copies).
- ARIA-H (microhemorrhage): 11.9% in noncarriers, 14.0% in heterozygotes, 39.0% in homozygotes.
That is not a small gradient. Going from zero copies of ε4 to two roughly sextuples ARIA-E risk. It's exactly the kind of gene-dose relationship — more copies, more effect, in a straight line — that shows up clearly in a randomized trial and is why the FDA-approved label for lecanemab recommends testing APOE ε4 status before starting treatment, specifically to inform that risk conversation between a patient and their doctor.
What this site already had, before this was news
Clinical APOE genotyping for exactly this purpose reads two positions: rs429358 and rs7412. Together, the two determine which of the three common APOE forms — ε2, ε3, or ε4 — someone carries on each copy of the gene, and how many ε4 copies that adds up to. Both positions are already on this site's APOE and Alzheimer's disease risk page, which has covered the ε2/ε3/ε4 system and what each form does to disease risk since well before this month's drug-safety news.
That's the whole connection, stated plainly: the genotype question these drugs now make clinically relevant is the same genotype question this site's APOE page was already built to answer. Nothing here is a new finding — it's the same two positions, newly relevant for a different, very current reason.
This is not a substitute for clinical APOE genotyping. A doctor considering lecanemab or donanemab orders a clinical-grade test through a certified lab, interpreted alongside a person's full medical picture — not a consumer genotype file read against a webpage. If you or someone you know is being evaluated for one of these drugs, that conversation belongs with a neurologist, not here.
© 2026 MyGeneLog™. Licensed under CC BY-NC 4.0 — share or adapt this article, including translations, with credit to MyGeneLog™ and a link back to this page. Commercial use is not permitted without our written permission. Full terms: mygenelog.com/terms.
Quoting this page
Licensed CC BY-NC 4.0 — when citing this, name MyGeneLog™ and link to this exact page. Commercial use needs our written permission.
Why the new Alzheimer's drugs check APOE4 first. MyGeneLog™. https://www.mygenelog.com/updates/apoe4-testing-before-alzheimers-drugs
Frequently asked questions
What is ARIA?
ARIA (amyloid-related imaging abnormality) is a side effect of anti-amyloid Alzheimer's drugs like lecanemab and donanemab — either brain swelling (ARIA-E) or small bleeds (ARIA-H), seen on MRI and sometimes symptomatic.
Why does APOE genotype matter for these drugs specifically?
The CLARITY AD trial found ARIA-E risk rising from 5.4% in people with no APOE e4 allele to 32.6% in people with two copies — a roughly sixfold increase. That gene-dose relationship is why testing is recommended before starting treatment.
Are the variants on this site the same ones a clinical APOE test uses?
The same two positions, rs429358 and rs7412, are what clinical APOE genotyping reads to determine e2/e3/e4 status. This site has covered both on its APOE page independently of this drug-safety news.
Does this mean I should check my own genotype before an Alzheimer's drug decision?
No — that decision belongs with a neurologist and a certified clinical test, not a consumer genotype file. This page explains why genotype matters for these drugs; it is not a substitute for that clinical process.
