Reviews/Anti Amyloid Antibodies/APOE4 homozygotes face higher amyloid-related imaging abnormality risk
Safety · AMAB-SAF-02 · v2
Borderline: within 4 points of the Established/Likely boundary, so a re-grade could land either side.
Several independent anti-amyloid drug programmes point the same way: people who carry two copies of the APOE4 gene tend to have the most imaging side effects (ARIA) during treatment. In a lecanemab dataset, a model put the rate of brain-swelling ARIA at 22.5% for these double-carriers against 5.4% for non-carriers [1], and a separate trial report confirmed the same direction [2]. A large donanemab analysis found the risk rose independently as the number of APOE4 copies increased [5], and an A4 study reported a sizeable effect of the double-carrier genotype on bleeding-type ARIA [7]. Because every line of evidence agrees on direction and the reported gaps are large, the panel treated the higher risk as well supported. The main caution is that the single cleanest homozygote percentage comes from a model built on very small observed samples.
Direction is unanimous across distinct antibodies. [1] models ARIA-E at 22.5% in homozygous carriers versus 6.8% heterozygous and 5.4% noncarriers on lecanemab, states incidence appears highest in homozygous subjects, and reports observed rates of 25% (1/4) in the OLE and 50% (5/10) in the Core. [2] reports from the same Study 201 that ARIA-E incidence was higher in ApoE4 homozygous carriers. [5], across 3030 donanemab-trial participants, found an independent association between ARIA-E and increasing APOE ε4 allele number. [7] reports APOE ε4/ε4 OR=4.8 (p<0.001) for accumulation of hemorrhagic lesions in 1097 A4 participants, while [3] and [6] corroborate a broader carrier-versus-noncarrier gradient for bapineuzumab and gantenerumab. The contested point: The division is over the precision of 'substantially higher,' not its direction. The sharpest homozygote-specific percentage sits in a modeling study [1] whose observed homozygote cells are tiny (1/4, 5/10), and the large donanemab dataset [5] reports an allele-count gradient rather than a standalone homozygote incidence.
All 18 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Anti Amyloid Antibodies, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The cleanest homozygote-specific ARIA-E percentage (22.5% vs 5.4%) comes from a modeling study [1] whose observed homozygote samples are tiny (1/4 in OLE, 5/10 in Core), implying wide uncertainty around the magnitude.
[5] supports increased risk with APOE ε4 allele number but reports no homozygote-specific rate in the corpus text, so the exact homozygote magnitude is inferred from an allele-count gradient rather than a reported homozygote incidence.
Where it remains uncertain. The cleanest homozygote-specific rate derives from a model with very small observed samples (1/4, 5/10) [1], and [5] reports an allele-count gradient rather than a homozygote-specific incidence.
Aleatory
[1]'s observed homozygote rates rest on 1/4 (OLE) and 5/10 (Core), implying wide sampling variability around the magnitude.
Epistemic
[5] reports association with increasing APOE ε4 allele number but no homozygote-specific rate in the corpus text.
Inferential
[3] and [6] report carrier-versus-noncarrier comparisons, not homozygote-versus-heterozygote, so they bear on the claim only indirectly.
The load-bearing papers from a corpus of 18, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Provides the homozygote-specific magnitude: ~4x the noncarrier ARIA-E rate, supporting 'substantially higher'.
“The incidence of ARIA‐E in subjects treated with lecanemab appears highest in APOE4 homozygous subjects.”
Panel auditInternal validity: medium · scope of inference: ARIA-E risk by APOE4 genotype in early AD patients treated with lecanemab 10 mg/kg Q2W.Key limitations: model-based prediction · very small homozygote subgroups · single drug · awaits confirmation
Independent confirmation of elevated homozygote ARIA-E risk from the trial data itself.
“ARIA-E was correlated with maximum lecanemab serum concentration and incidence was higher in apolipoprotein E4 (ApoE4) homozygous carriers.”
Panel auditInternal validity: medium · scope of inference: ARIA-E incidence by ApoE4 status in early AD patients treated with lecanemab.Key limitations: no numbers in abstract · single drug · overall ARIA-E incidence low (<10%)
Large donanemab dataset shows allele-number dose-response, so two-allele homozygotes face the highest risk.
“Risk analysis demonstrated independent associations between ARIA-E and 6 baseline variables, including increased risk with APOE ε4 allele number”
Panel auditInternal validity: high · scope of inference: ARIA-E risk factors in early symptomatic AD treated with donanemab.Key limitations: allele-number effect reported, not a direct homozygote-vs-others magnitude · single drug · some post hoc analyses
Corroborates a carrier>noncarrier ARIA-E gradient with a second antibody, consistent with an APOE4 dose effect.
“In carriers, ARIA-E occurred in 21.2% in the 0.5 mg/kg group”
Panel auditInternal validity: medium · scope of inference: ARIA-E by ε4 carrier status in mild-to-moderate AD treated with bapineuzumab.Key limitations: no homozygote breakdown · dose confounded with carrier status · different drug than claim's focus
4 of 18 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
Unanimous direction across lecanemab, donanemab and A4 evidence plus a large reported magnitude support an Established verdict, held conservative because the sharpest homozygote-specific quantification derives from a modeling study with small observed samples.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Directly observed homozygote ARIA-E rate in a large active-treatment arm. The cleanest homozygote percentage is modeled on small samples [1]; a large observed rate would anchor the magnitude. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: Prespecified homozygote-stratified ARIA-E incidence from a large confirmatory anti-amyloid trial (e.g. the CLARITY-AD data [1] names as needed).
Open question: What is the directly observed ARIA-E incidence in APOE4 homozygotes treated in a large anti-amyloid trial? Why it is still open: The corpus's sharpest homozygote rate is modeled on tiny samples (1/4, 5/10) [1], and [1] states larger-trial confirmation is still needed.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19