New · October 2026 · Treatment Watch
Why don’t anti-amyloid drugs cure Alzheimer’s, and what comes next? The first randomized evidence that lowering tau may slow decline arrived in 2026. Read: Beyond Amyloid: Why Tau Is the Next Frontier →
What is new for APOE ε4 carriers since this guide was last revised:
- ARIA risk rises with each copy of ε4, in trials and in routine care. A 2026 meta-analysis of lecanemab trial and real-world data found a relative risk of ARIA of 1.45 for one copy and 3.54 for two, compared with non-carriers.
- Whether anti-amyloid antibodies work less well in carriers is unresolved. Two 2025 syntheses of the trial data reach different conclusions.
- ALZ-801 (valiltramiprosate) did not meet its Phase 3 primary endpoint. A 2026 biomarker follow-up from the same trial is favorable at the MCI stage, but these remain hypothesis-generating findings that need a confirmatory trial.
Read the full update with sources · Evidence Watch, September 2026 · All Evidence Watch editions
Alzheimer’s Treatments and Emerging Therapies: A Guide for APOE ε4 Carriers
A physician-curated, evidence-graded overview of every category of treatment — from FDA-approved medications to clinical trials to investigational therapies — with APOE ε4–specific data wherever it exists.
The treatment landscape for Alzheimer’s disease has changed dramatically in the past three years. There are now FDA-approved drugs that can slow the disease — not just manage symptoms — and a robust pipeline of new agents in late-stage development. At the same time, unproven therapies are widely promoted online, often with misleading claims.
This page organizes everything into three clear tiers based on the strength of evidence. It is written for patients and families — not physicians — so medical terms are explained in plain language wherever they appear. APOE ε4–specific data is highlighted throughout, because your genotype affects both the benefit and the risk of several treatments.
Tier 1 — FDA-Approved Treatments
These medications are available today, supported by large randomized controlled trials, and approved by the US Food and Drug Administration.
There are two categories of FDA-approved treatment for Alzheimer’s disease: disease-modifying therapies, which target the underlying biological process and can slow progression, and symptomatic medications, which help manage cognitive symptoms without affecting the underlying disease. Both have an important role, and they are often used together.
Disease-Modifying Therapies: Anti-Amyloid Antibodies
These are intravenous infusion medications that work by clearing amyloid protein deposits from the brain — the plaques that have long been associated with Alzheimer’s disease. They represent the first treatments that address a root cause of the disease rather than only managing its symptoms. Both currently approved agents require confirmed amyloid pathology (via a PET brain scan or cerebrospinal fluid test) before starting treatment, and both require APOE ε4 genotyping because your genetic status directly affects your risk of the most significant side effect.
Lecanemab (brand name: Leqembi®)
What it is: A monoclonal antibody — a laboratory-made protein that targets and clears a specific toxic form of amyloid from the brain. Given as an intravenous infusion every two weeks.
What the evidence shows: The CLARITY AD Phase 3 clinical trial enrolled 1,795 participants with early Alzheimer’s disease — either mild memory impairment (called MCI) or mild dementia — and followed them for 18 months. Lecanemab slowed the rate of cognitive and functional decline by 27% compared to placebo. It also cleared amyloid from the brain by an average of 59 Centiloids (the unit used to measure amyloid burden on brain scans). (van Dyck et al., N Engl J Med 2022)
The most important side effect of lecanemab is called ARIA — Amyloid-Related Imaging Abnormalities. This refers to brain swelling (ARIA-E) or small brain bleeds (ARIA-H) that can occur as amyloid is cleared from vessel walls. Most ARIA is mild and shows up only on MRI scans, without any symptoms. Rarely, it can cause headache, confusion, or dizziness. Very rarely, it can be serious.
APOE ε4 genotype is the strongest known predictor of ARIA risk. In the CLARITY AD trial plus its open-label extension (1,612 people treated with lecanemab; Honig et al., Alzheimers Res Ther 2024):
| Genotype | ARIA-E Rate | ARIA-H Rate |
|---|---|---|
| ε4/ε4 (two copies) | 34.5% | 39.8% |
| ε3/ε4 (one copy) | 11.6% | 16.1% |
| Non-carrier | 6.5% | 11.9% |
ARIA-E = brain swelling or fluid; ARIA-H = microbleeds or superficial siderosis. In the 18-month placebo-controlled phase alone, ARIA-E rates were 5.4%, 10.9%, and 32.6% for non-carriers, one copy, and two copies (van Dyck et al., N Engl J Med 2022). ARIA-H also occurs without treatment: 21.1% of homozygotes on placebo had it.
Important: APOE genotyping is required before starting lecanemab. Patients on blood thinners (anticoagulants) should not receive lecanemab until more safety data are available, as this combination substantially increases bleeding risk. Regular MRI monitoring is mandatory during treatment. (Cummings et al., J Prev Alzheimers Dis 2023)
van Dyck CH et al. N Engl J Med. 2022 · Honig LS et al. Alzheimers Res Ther. 2024 · Cummings J et al. J Prev Alzheimers Dis. 2023
Donanemab (brand name: Kisunla®)
What it is: A monoclonal antibody that targets a slightly different form of amyloid — a chemically modified version called pyroglutamated amyloid — that accumulates selectively in plaques. Given as a monthly intravenous infusion. Notably, treatment may be discontinued once amyloid is fully cleared from the brain, which occurred in over half of participants within 12 months — making this the first “treat-to-clear” Alzheimer’s therapy.
What the evidence shows: The TRAILBLAZER-ALZ 2 Phase 3 trial (1,736 participants) used baseline tau PET scans to stratify participants. Among those with lower tau burden — meaning less advanced disease — donanemab slowed cognitive and functional decline by 35%. Among all participants combined, slowing was 22%. Crucially, participants treated at the earliest stages experienced the greatest benefit (up to 60% slowing of progression), reinforcing the principle that earlier treatment yields better outcomes. (Sims et al., JAMA 2023)
ARIA rates with donanemab follow the same genotype-dependent pattern seen with lecanemab. In the TRAILBLAZER-ALZ 2 trial, ARIA-E rates with donanemab by APOE ε4 status were as follows (Sims et al., JAMA 2023). The right-hand column shows the pooled odds across Phase 3 anti-amyloid antibody trials (Jeong et al., Neurology 2025):
| Genotype | ARIA-E Rate | Pooled odds ratio vs. non-carrier (all antibodies) |
|---|---|---|
| ε4/ε4 (two copies) | 40.6% | 5.6 (95% CI 3.8–8.2) |
| ε3/ε4 (one copy) | 22.8% | 1.9 (95% CI 1.5–2.4) |
| Non-carrier | 15.7% | Reference |
In the European Union and the United Kingdom, both antibodies are licensed only for people who are not ε4/ε4 homozygotes, because of ARIA risk; regulatory positions elsewhere vary and change over time. The FDA has approved both agents regardless of APOE genotype. A slower dosing ramp-up lowered ARIA-E at 24 weeks from 23.7% to 13.7% overall, and in the small group of ε4/ε4 homozygotes (about 21 per arm) from 57.1% to 19.0%, while amyloid removal was similar (TRAILBLAZER-ALZ 6; Wang et al., Alzheimers Dement 2025).
Sims JR et al. JAMA. 2023 · Jeong SY et al. Neurology. 2025 · Wang H et al. Alzheimers Dement. 2025
Symptomatic Medications
These medications do not slow or stop the underlying disease process. They work by supporting the brain’s chemical signaling systems to improve memory, thinking, and daily function. They are widely used, often in combination, and remain the first treatment prescribed after an Alzheimer’s diagnosis. Their benefit in APOE ε4 carriers is addressed directly below.
Cholinesterase Inhibitors: Donepezil, Rivastigmine, Galantamine
What they are: This class of medications works by preserving acetylcholine — a chemical messenger in the brain that is critical for memory and learning. In Alzheimer’s disease, the neurons that produce acetylcholine are among the first to be damaged. Cholinesterase inhibitors slow the breakdown of whatever acetylcholine remains, temporarily shoring up the signal. Donepezil (Aricept) is the most commonly prescribed. Rivastigmine is also available as a skin patch, which can be convenient and reduces stomach side effects.
What the evidence shows: A systematic review and meta-analysis of 30 studies found that all three agents produce modest but consistent improvements in cognition compared to placebo. Effect sizes are real but moderate — they tend to stabilize function for a period rather than restore lost ability. (Chen et al., Medicine 2024)
This is a nuanced and clinically important question. The short answer: APOE ε4 carrier status does not change whether you should take these medications. Both carriers and non-carriers benefit, and your genotype is not a reason to prescribe or avoid them differently.
The largest meta-analysis on this question (Cheng et al., 2018, 30 studies) found no statistically significant difference in cognitive response to cholinesterase inhibitors between APOE ε4 carriers and non-carriers. There is one notable exception worth knowing about: in the ADCS MCI trial (Petersen et al., NEJM 2005), a secondary analysis found that ε4 carriers on donepezil had a significantly lower rate of progression from mild memory impairment to Alzheimer’s dementia over three years. Non-carriers showed no such sustained benefit. However, this was a secondary analysis in a trial that was negative on its primary outcome — meaning it was not specifically designed to test this question — and it has not been independently replicated.
The FDA labels for these medications do not differentiate dosing or indication by APOE genotype. For memantine (see below), no APOE ε4–stratified Phase 3 trial data exist at all. (Cheng et al., Dement Geriatr Cogn Disord 2018; Petersen et al., NEJM 2005; Ritchie et al., Neurol Clin Pract 2024)
Cheng YC et al. Dement Geriatr Cogn Disord. 2018 · Petersen RC et al. N Engl J Med. 2005 · Chen Y et al. Medicine. 2024
Memantine (brand name: Namenda®)
What it is: Memantine works through a different mechanism than the cholinesterase inhibitors. It regulates glutamate — a different brain chemical that, when overactive, contributes to the death of nerve cells. It is approved for moderate to severe Alzheimer’s disease and is frequently combined with donepezil.
What the evidence shows: Memantine produces modest improvements in cognition and function in moderate-to-severe disease. Effect sizes are comparable to the cholinesterase inhibitors. It is generally well tolerated, with the most common side effect being dizziness.
APOE ε4 note: No Phase 3 genotype-stratified data for memantine exist in the published literature. Its use in ε4 carriers follows the same standard of care as the general Alzheimer’s population.
Chen Y et al. Medicine. 2024 · Ritchie M et al. Neurol Clin Pract. 2024
Tier 2 — Pipeline Agents in Active Development
These treatments are not yet available outside of clinical trials. They are in Phase 2 or Phase 3 testing, meaning they are being evaluated in large numbers of human participants right now.
The pipeline for Alzheimer’s therapies is more active than at any point in history. Three agents are highlighted here because they either have direct relevance to APOE ε4 carriers, have recently produced significant clinical trial data, or represent a new mechanism of action with strong scientific rationale.
ALZ-801 / Valiltramiprosate (Alzheon)
What it is: ALZ-801 is a daily oral pill — not an infusion — that works by preventing amyloid proteins from clumping together into the toxic clusters (called oligomers) that are thought to damage synapses and nerve cells. This is meaningfully different from the anti-amyloid antibodies, which clear plaques that have already formed. ALZ-801 is being developed specifically and exclusively for APOE ε4 carriers.
What the evidence shows: ALZ-801 has now completed its Phase 3 trial, APOLLOE4 — the only Phase 3 in Alzheimer’s disease ever run exclusively in a genotype-defined population. It enrolled 325 APOE ε4/ε4 homozygotes with early Alzheimer’s disease and ran for 78 weeks. It did not meet its primary endpoint. Cognitive decline on the ADAS-Cog13 was slowed by 11%, which was not statistically significant (p = 0.607). The drug did significantly slow hippocampal atrophy (18% slowing, p = 0.017) and lower plasma p-tau217, confirming that it reaches and engages its target. In a prespecified subgroup at the MCI stage — roughly 62 patients per treatment arm — cognitive and functional measures were nominally favorable, but those p-values are not corrected for multiple comparisons, and the subgroup’s own key secondary measure (CDR-SB) was not significant (p = 0.053). The earlier two-year study in 84 ε4 carriers, which reported stabilized memory scores and falling amyloid biomarkers, was single-arm and open-label — it had no placebo group, so it cannot support a controlled estimate of effect. (Abushakra et al., Drugs 2025; Hey et al., Drugs 2024)
Unlike lecanemab and donanemab — which carry substantially elevated ARIA risk in ε4/ε4 homozygotes — no cases of ARIA were observed in the ALZ-801 Phase 2 trial. This is mechanistically expected: the drug does not clear plaques from blood vessel walls (the process that causes ARIA), but instead prevents oligomer formation upstream.
The absence of ARIA remains a real advantage for ε4/ε4 homozygotes, who are excluded from anti-amyloid antibody treatment in several countries because of ARIA risk — and APOLLOE4 confirmed no increased ARIA-E or ARIA-H at Phase 3 scale. But a favorable safety profile is not evidence of benefit, and this trial did not demonstrate benefit. Establishing that would require a new trial confined to the MCI stage. Drug exposure was not affected by APOE genotype, age, or sex. (Hey et al., Clin Pharmacokinet 2025)
Abushakra S et al. Drugs. 2025;85(11):1455–1472 · Hey JA et al. Drugs. 2024 · Hey JA et al. Clin Pharmacokinet. 2025
Remternetug (Eli Lilly — LY3372993)
What it is: Remternetug is the next-generation anti-amyloid antibody from Eli Lilly, the maker of donanemab. It targets the same form of amyloid as donanemab, but is being developed in both intravenous and subcutaneous (under-the-skin injection) formulations — a practical advantage that could allow patients to self-administer treatment at home rather than attending an infusion center. It is not yet approved by the FDA.
Where it stands: Three trials are currently active. The first Phase 3 trial (TRAILRUNNER-ALZ 1, NCT05463731), whose primary outcome is amyloid plaque clearance, was expected to report in 2026. As of September 2026 we have not found peer-reviewed results; the only Phase 3 data presented so far (CTAD, December 2025) addressed the feasibility of self-injection, not efficacy or ARIA. A larger Phase 3 trial (NCT06653153) is actively enrolling participants through 2030. A third trial through Washington University is evaluating remternetug in people with preclinical (pre-symptom) Alzheimer’s and in those with the rare inherited forms of the disease — extending potential use into secondary prevention. (Kelliny et al., J Neurosci Res 2025; Nonino et al., Cochrane 2026)
APOE ε4–specific ARIA rates for remternetug have not yet been published. Based on class-wide data from comparable anti-amyloid antibodies, the same genotype-dependent ARIA risk pattern is expected. APOE genotyping will almost certainly be required before treatment — as it is for lecanemab and donanemab. We will update this section when peer-reviewed Phase 3 results are published.
Kelliny S et al. J Neurosci Res. 2025 · Panza F et al. Expert Opin Biol Ther. 2025 · Nonino F et al. Cochrane Database. 2026
GLP-1 Receptor Agonists (Semaglutide, Liraglutide)
What they are: GLP-1 receptor agonists are a class of medications originally developed for type 2 diabetes and weight loss — the same class as semaglutide (Ozempic®, Wegovy®) and liraglutide (Victoza®). Interest in their potential for Alzheimer’s disease arose from laboratory and animal studies suggesting they reduce neuroinflammation, improve brain energy metabolism, and protect synapses from amyloid-related damage.
What the Phase 3 trials showed: The EVOKE and EVOKE+ trials — together the largest Alzheimer’s drug trial program ever conducted, enrolling 3,808 participants across 40 countries — tested oral semaglutide 14 mg versus placebo in patients with early Alzheimer’s disease over 104 weeks. Both trials were negative: semaglutide did not slow cognitive decline compared to placebo. CDR-SB scores (a standard measure of cognitive and functional ability) were essentially identical between the treatment and placebo groups in both trials. (Cummings et al., Lancet 2026)
The negative EVOKE results do not necessarily mean GLP-1 receptor agonists have no role in Alzheimer’s prevention. Several researchers have noted that oral semaglutide penetrates the brain poorly — the drug may not reach neurons in sufficient concentration to produce the neuroprotective effects seen in animal models. A separate smaller Phase 2b trial of injectable liraglutide (ELAD trial, published in Nature Medicine 2025) also showed no significant effect on brain metabolism, though one cognitive measure showed a trend toward benefit.
The current scientific hypothesis is that GLP-1 agents may be more effective as preventive treatments — targeting metabolic and inflammatory drivers of disease before significant neurodegeneration has occurred — rather than as treatments for established symptomatic Alzheimer’s. The ISAP trial is testing this directly in people with preclinical (pre-symptom) amyloid-positive Alzheimer’s using tau PET as the primary endpoint. (Koychev et al., BMJ Open 2024)
Bottom line: Taking semaglutide or a similar drug specifically to prevent or treat Alzheimer’s disease is not currently supported by Phase 3 clinical trial evidence. If you take these medications for diabetes or weight management, continue doing so under your physician’s guidance — metabolic health is itself protective against dementia. But they should not be sought out specifically for Alzheimer’s prevention based on current evidence. (Alhowail et al., Front Aging Neurosci 2026)
Cummings JL et al. Lancet. 2026 · Edison P et al. Nature Medicine. 2025 · Koychev I et al. BMJ Open. 2024 · Alhowail AH et al. Front Aging Neurosci. 2026
Tier 3 — Investigational and Fringe Interventions
These interventions are widely discussed online but lack the clinical trial evidence required for a medical recommendation. They are presented here because patients ask about them — not because they are recommended.
A physician-authored platform grounded in peer-reviewed evidence has an obligation to address these therapies honestly. The purpose of this section is not to dismiss patient curiosity, but to give you the tools to evaluate claims critically — and to protect you from interventions that are expensive, unproven, and in some cases potentially harmful. For each agent, the current state of the published evidence is presented without embellishment.
Mesenchymal Stem Cell Therapy (Laromestrocel / Lomecel-B)
What it is: Mesenchymal stem cells are a type of cell found in bone marrow with anti-inflammatory properties. In these trials, cells from a healthy donor are processed and given by intravenous infusion to patients with Alzheimer’s disease. The proposed mechanism is reduction of neuroinflammation rather than direct replacement of damaged neurons.
What the evidence shows: This is the most credible entry in Tier 3 because it has actual published RCT data. A Phase 2a trial of laromestrocel published in Nature Medicine (Rash et al., 2025) enrolled 49 participants with mild Alzheimer’s disease. Treatment met its primary safety endpoint — no serious infusion reactions and no ARIA were observed at any dose. At 39 weeks, treated participants showed 48% less brain volume loss and 62% less hippocampal (memory center) shrinkage compared to placebo. Notably, secondary cognitive measures also trended toward benefit. A Phase 1 trial of the same cellular platform (Lomecel-B) similarly demonstrated safety and early biomarker signals. (Brody et al., Alzheimers Dement 2022)
Important context: These trials enrolled 49 and 33 participants respectively — far too small to draw conclusions about efficacy. Brain volume changes are a surrogate marker, not a direct measure of cognitive benefit. Phase 3 trials with larger populations and longer follow-up are required before any clinical recommendation can be made. This is not a treatment available outside of research. It is, however, a legitimately promising signal that warrants follow-up investigation — which is a different standard than claiming it works.
Rash BG et al. Nature Medicine. 2025 · Brody M et al. Alzheimers Dement. 2022
Rapamycin (mTOR Inhibitor)
What it is: Rapamycin (also known as sirolimus) is an existing medication used in transplant medicine to prevent organ rejection. It works by inhibiting a protein called mTOR, which regulates cell growth, metabolism, and autophagy — the process by which cells clear out damaged components. In Alzheimer’s disease, mTOR is overactive, and this overactivity is thought to impair the brain’s ability to clear amyloid and tau. Rapamycin has attracted attention as a potential longevity and neuroprotective drug.
Preclinical evidence: The laboratory rationale is compelling. A systematic review of 13 animal studies found rapamycin improved learning and memory in transgenic Alzheimer’s mouse models, with evidence of reduced amyloid, reduced tau phosphorylation, improved blood vessel function, and enhanced cellular cleanup processes. In mice expressing human APOE4 specifically, rapamycin normalized brain blood flow and reduced amyloid retention — effects not seen in APOE3 mice, suggesting possible genotype-specific relevance. (Cai et al., J Alzheimers Dis 2024; Lin et al., Neurobiol Dis 2020)
Human evidence — a significant caution: The first completed human clinical trial of rapamycin in Alzheimer’s patients (10 participants, Phase 1, published May 2025) produced concerning findings. At 1 mg/day for 8 weeks, rapamycin was undetectable in the cerebrospinal fluid — meaning it did not reach the brain in measurable concentrations at this dose. Biomarkers of neurodegeneration (phosphorylated tau, GFAP, neurofilament light) moved in an unfavorable direction during treatment. No cognitive benefit was observed. (Gonzales et al., Commun Med 2025)
What this means: The PEARL trial — sometimes cited online as evidence that rapamycin is safe for aging — was a trial in healthy older adults without Alzheimer’s disease. It was not an Alzheimer’s trial and does not speak to efficacy in AD. Its results showed no effect on the primary endpoint (body fat) but some positive secondary signals in women at higher doses.
A Swedish Phase 2a trial (ERAP) is currently testing higher weekly doses with brain PET imaging as the primary endpoint, which may shed light on whether adequate CNS drug levels can be achieved. Until those results are available, self-administering rapamycin for Alzheimer’s prevention or treatment is not supported by clinical evidence and carries real risks, including immunosuppression. (Svensson et al., BMC Neurol 2024)
Gonzales MM et al. Commun Med. 2025 · Cai J et al. J Alzheimers Dis. 2024 · Moel M et al. Aging. 2025 · Lin AL et al. Neurobiol Dis. 2020
Cerebrolysin
What it is: Cerebrolysin is a mixture of peptides (small proteins) and amino acids derived from pig brain tissue, given by intravenous or intramuscular injection. It is widely promoted online — particularly in longevity and biohacking communities — as a cognitive enhancer. It is approved in Austria, China, Germany, Russia, and South Korea, and used in over 50 countries, but it has never received FDA or EMA (European Medicines Agency) central approval for Alzheimer’s disease.
What the evidence shows: Approximately six randomized controlled trials have been published in mild-to-moderate Alzheimer’s disease, most lasting only four to twelve weeks. A 2015 meta-analysis found statistically significant improvements in cognition at four weeks, but this effect was not sustained at six months. No large Phase 3 pivotal trial meeting FDA or EMA registration standards has ever been conducted. (Gauthier et al., Dement Geriatr Cogn Disord 2015)
Significant concerns with the evidence base: No Cochrane review has evaluated cerebrolysin in Alzheimer’s disease. The closest Cochrane assessment, in vascular dementia (6 randomized trials, 597 participants), graded the evidence for cognition and global function as very low quality because of heterogeneity and high risk of bias; every trial that reported its funding was industry-supported, and the reviewers cautioned that any benefit may be too small to be clinically meaningful (Cui et al., Cochrane Database Syst Rev 2019). The Alzheimer’s trials share similar limitations: they are small and short, and the benefit on cognition in the 2015 meta-analysis was significant at four weeks but not at six months (SMD −0.37; 95% CI −0.90 to 0.16). Most trials predate biomarker-confirmed diagnosis, meaning some enrolled participants may not have had AD at all.
There is one APOE ε4–relevant data point: a biomarker substudy found that BDNF (a brain growth factor) increased more in ε4 carriers treated with cerebrolysin, and this correlated with better cognitive performance. This is hypothesis-generating, not evidence of clinical efficacy. No RCT has been designed or powered to assess cerebrolysin efficacy specifically by APOE genotype. Two of its authors were affiliated with the manufacturer, and the strongest signal came from cerebrolysin combined with donepezil. (Alvarez et al., Int J Neuropsychopharmacol 2016)
Gauthier S et al. Dement Geriatr Cogn Disord. 2015 · Cui S et al. Cochrane Database Syst Rev. 2019 (vascular dementia) · Alvarez XA et al. Int J Neuropsychopharmacol. 2016
A Note on Decision-Making as an APOE ε4 Carrier
The treatment landscape for Alzheimer’s disease has changed more in the past three years than in the previous three decades. Two disease-modifying therapies are now available. An oral agent designed specifically for APOE ε4 carriers, ALZ-801, has completed Phase 3; it did not meet its primary endpoint, and its signals at the MCI stage await a confirmatory trial. And the field is moving rapidly.
For APOE ε4 carriers specifically, several principles should guide treatment conversations with your physician: your genotype affects both the benefit and the risk of anti-amyloid therapies; earlier treatment consistently yields better outcomes across all available data; and APOE genotyping is now a standard clinical step before initiating any anti-amyloid therapy.
The symptomatic medications — cholinesterase inhibitors and memantine — remain the standard of care for cognitive symptom management regardless of disease stage. They are not in competition with the newer disease-modifying therapies; they are complementary to them.
- van Dyck CH et al. Lecanemab in Early Alzheimer’s Disease (CLARITY AD). N Engl J Med. 2022
- Honig LS et al. Updated safety results from phase 3 lecanemab study in early Alzheimer’s disease. Alzheimers Res Ther. 2024;16:105. doi:10.1186/s13195-024-01441-8
- Cummings J et al. Lecanemab Appropriate Use Recommendations. J Prev Alzheimers Dis. 2023
- Sims JR et al. Donanemab in early symptomatic Alzheimer disease: the TRAILBLAZER-ALZ 2 randomized clinical trial. JAMA. 2023;330:512–527. doi:10.1001/jama.2023.13239
- Jeong SY et al. Incidence of amyloid-related imaging abnormalities in phase III clinical trials of anti-amyloid-β immunotherapy: an updated meta-analysis. Neurology. 2025;104:e213483. doi:10.1212/WNL.0000000000213483
- Wang H et al. Modified titration of donanemab reduces ARIA risk and maintains amyloid reduction (TRAILBLAZER-ALZ 6). Alzheimers Dement. 2025;21:e70062. doi:10.1002/alz.70062
- Fox NC et al. Treatment for Alzheimer’s Disease. Lancet. 2025
- Chen Y et al. Efficacy and safety of AD drugs — meta-analysis. Medicine. 2024
- Cheng YC et al. APOE ε4 and cholinesterase inhibitor response — meta-analysis. Dement Geriatr Cogn Disord. 2018
- Petersen RC et al. Donepezil for MCI — ADCS Trial. N Engl J Med. 2005
- Ritchie M et al. APOE genetic testing in clinical practice. Neurol Clin Pract. 2024
- Hey JA et al. ALZ-801 Phase 2 in APOE4 carriers. Drugs. 2024
- Hey JA et al. ALZ-801 pharmacokinetics in APOE4 carriers. Clin Pharmacokinet. 2025
- Kelliny S et al. AD and FTD — remternetug pipeline review. J Neurosci Res. 2025
- Nonino F et al. Amyloid-targeting antibodies — Cochrane review. Cochrane Database. 2026
- Cummings JL et al. EVOKE and EVOKE+ trials of semaglutide. Lancet. 2026
- Edison P et al. ELAD trial of liraglutide. Nature Medicine. 2025
- Koychev I et al. ISAP trial protocol — semaglutide in preclinical AD. BMJ Open. 2024
- Rash BG et al. Laromestrocel mesenchymal stem cell Phase 2a. Nature Medicine. 2025
- Brody M et al. Lomecel-B Phase 1 results. Alzheimers Dement. 2022
- Gonzales MM et al. Rapamycin Phase 1 in AD. Commun Med. 2025
- Cai J et al. Rapamycin in AD animal models — meta-analysis. J Alzheimers Dis. 2024
- Lin AL et al. APOE genotype-dependent rapamycin response. Neurobiol Dis. 2020
- Moel M et al. PEARL trial results — rapamycin in healthy aging. Aging. 2025
- Gauthier S et al. Cerebrolysin in mild-to-moderate Alzheimer’s disease: a meta-analysis of randomized controlled clinical trials. Dement Geriatr Cogn Disord. 2015;39:332–347. doi:10.1159/000377672
- Cui S et al. Cerebrolysin for vascular dementia. Cochrane Database Syst Rev. 2019;(11):CD008900. doi:10.1002/14651858.CD008900.pub3 (vascular dementia, not Alzheimer’s disease)
- Alvarez XA et al. Synergistic increase of serum BDNF in Alzheimer patients treated with cerebrolysin and donepezil: association with cognitive improvement in ApoE4 cases. Int J Neuropsychopharmacol. 2016;19(6):pyw024. doi:10.1093/ijnp/pyw024
This article is written for educational purposes only and does not constitute medical advice. Content is authored by a licensed physician and grounded in peer-reviewed literature. All treatment decisions should be made in consultation with a qualified healthcare provider familiar with your complete clinical history, genotype, and individual circumstances.
What the newest research changes for APOE ε4 carriers
This section summarizes peer-reviewed evidence published since this guide was last revised. Each study is graded and explained in detail in Evidence Watch, September 2026.
1. ARIA risk by genotype: new pooled estimates
Four new syntheses confirm that the risk of ARIA rises with each copy of ε4, for both approved antibodies and in routine clinical care as well as in trials.
| Study | Data pooled | Key APOE ε4 finding |
|---|---|---|
| Qi et al., 2026 | Lecanemab: 2 RCTs + 5 real-world studies (1,576 patients) | Relative risk of ARIA vs. non-carriers: 1.45 (one copy), 3.54 (two copies). Overall ARIA 19%; symptomatic ARIA 3%. |
| Shim et al., 2025 | Lecanemab and donanemab: 6 RCTs (5,633 participants) | Carriers vs. non-carriers: ARIA-E 2.19 times higher; ARIA-H 3.45 times higher. |
| Hsu et al., 2025 | 7 antibody trials, including aducanumab (8,010 participants) | Homozygotes vs. non-carriers: 5.53 times the odds of ARIA-E. |
| Pereira da Silva et al., 2025 | Donanemab: 3 RCTs (2,054 participants) | ARIA-E about 12 times more common than on placebo overall (RR 12.39), highest in homozygotes. |
These pooled estimates complement the trial-specific rates in the tables above. The studies use different comparisons (carriers vs. non-carriers, or drug vs. placebo) and different mixes of trials, so their numbers are not directly interchangeable.
2. Do the antibodies work less well in carriers? Still unresolved
Shim and colleagues reported the largest slowing of decline in non-carriers, but they summarized trial subgroups rather than pooling them statistically. Hsu and colleagues, using a network meta-analysis, found no difference in efficacy by carrier status. They also found that the average benefit of the antibodies did not reach the minimally important difference on either the CDR-SB or the ADAS-Cog. The honest summary for carriers: the added risk is well established, while any difference in benefit is uncertain.
3. ALZ-801 (valiltramiprosate): the 2026 biomarker follow-up
A 2026 analysis of the APOLLOE4 trial reports that in the MCI subgroup, plasma p-tau217 fell 36% on drug and rose 17% on placebo (observed cases). Neurofilament light was also lower on drug at week 78, and these biomarker changes correlated modestly with clinical and MRI outcomes. The trial’s primary endpoint was still not met, both papers were written largely by the sponsor, and a biomarker response in a subgroup is not proof of clinical benefit. A confirmatory trial at the MCI stage is needed.
What this means for you. If you are considering lecanemab or donanemab, know your APOE genotype first and discuss ARIA risk for your specific genotype with your neurologist. If you carry two copies, the risk–benefit conversation deserves particular care. If you have had a p-tau217 blood test, see our explainer Your p-tau217 Blood Test and APOE4. New research suggests the same result carries more near-term risk in carriers.
Related reading: Treatment Options for the APOE ε4 Carrier · Do Leqembi and Kisunla Meaningfully Help? · ARIA-E After Donanemab in an APOE ε4 Carrier · Evidence Watch
Sources for this update
- Qi L, et al. Safety profiles of lecanemab: a systematic review and meta-analysis of randomized controlled trials and real-world evidence. J Prev Alzheimers Dis. 2026;13(3):100473. doi:10.1016/j.tjpad.2025.100473
- Shim GH, et al. Influence of patient characteristics on efficacy and safety of anti-amyloid monoclonal antibodies in Alzheimer’s disease. Ageing Res Rev. 2025;114:102981. doi:10.1016/j.arr.2025.102981
- Hsu CW, et al. The efficacy and safety of anti-amyloid monoclonal antibody versus acetylcholinesterase inhibitor. J Prev Alzheimers Dis. 2025;12(8):100195. doi:10.1016/j.tjpad.2025.100195
- Pereira da Silva AM, et al. Efficacy and ε4-stratified risk of donanemab in Alzheimer’s disease. J Alzheimers Dis. 2025;107(2):477–493. doi:10.1177/13872877251361044
- Hey JA, et al. Plasma biomarker effects of oral valiltramiprosate/ALZ-801 in early Alzheimer’s disease from Phase 3 and Phase 2 trials. Drugs. 2026;86(10):1705–1721. doi:10.1007/s40265-026-02381-3
- Abushakra S, et al. Clinical efficacy, safety and imaging effects of oral valiltramiprosate in APOEε4/ε4 homozygotes with early Alzheimer’s disease (APOLLOE4). Drugs. 2025;85(11):1455–1472. doi:10.1007/s40265-025-02250-5