Obicetrapib and APOE4: The First Oral Drug to Move Amyloid AND Tau

Hi Phoenix friend,

(Full Youtube Video link at the bottom)

A cholesterol drug just did something no Alzheimer's drug has ever done in APOE4/4 carriers. In a pre-specified substudy of 1,535 patients published in The Journal of Prevention of Alzheimer's Disease, p-tau217 — the cleanest blood biomarker we have for Alzheimer's pathology — fell 7.81% on obicetrapib and rose 12.67% on placebo, a 20.48% placebo-adjusted treatment difference in APOE4/4 homozygotes (P = 0.010) [Davidson et al., 2025].

If you are an APOE4 or APOE4/4 carrier reading that paragraph for the first time, I want you to sit with it the same way I did. As you know I carry APOE4/4.

You also know I spend a crazy amount of time reading every drug paper that promised something for people like us, and this is the cleanest oral-drug biomarker signal I have ever seen in our genotype.

This article is a careful, calibrated walkthrough of what obicetrapib is, what the BROADWAY APOE4 substudy actually showed, why the genetics predicted it years ago, why earlier CETP drugs failed and this one didn't, and what APOE4 carriers should realistically do with this information right now. Education, not medical advice.

What Is Obicetrapib?

Obicetrapib is a next-generation oral CETP inhibitor — 10 mg once a day — being developed by NewAmsterdam Pharma [Masson et al., 2025]. CETP stands for cholesteryl ester transfer protein. Here it is in one sentence from a 2024 Current Atherosclerosis Reports review: CETP "tends to result in a net mass transfer of cholesteryl esters from HDL to VLDL and LDL, and a net mass transfer of triglycerides from VLDL to LDL and HDL" [Kastelein et al., 2024].

Translation: CETP is a shuttle that moves cholesterol out of your protective HDL particles and into your atherogenic LDL and VLDL particles. Block it, and two things happen at once: LDL drops and HDL climbs. Obicetrapib also potently raises ApoA-1 and ApoE — the apolipoproteins sitting on the surface of HDL particles [Kastelein et al., 2024]. Hold that ApoE piece. It is the key to why this drug matters specifically for APOE4 carriers.

Mechanistically, obicetrapib works at a completely different spot on the cholesterol pathway than anything currently in the lipid toolkit:

• Statins block the liver from making cholesterol.

• Ezetimibe blocks gut absorption.

• PCSK9 inhibitors (evolocumab, alirocumab, inclisiran) increase hepatic LDL receptor recycling.

• Obicetrapib blocks the transfer step in the bloodstream itself.

Different lever. Different target. Stacks on top of everything else. [link: APOE4 lipid management 101]

The BROADWAY Substudy — Why APOE4 Carriers Should Care

BROADWAY is obicetrapib's flagship Phase 3 cardiovascular trial. 2,530 patients with established cardiovascular disease or familial hypercholesterolemia on maximally tolerated statin therapy, randomized 2:1 to obicetrapib 10 mg or placebo, published in The New England Journal of Medicine in 2025 [Nicholls et al., 2025].

Hidden inside BROADWAY — pre-specified, not fished — was an Alzheimer's-biomarker substudy. 1,535 patients had their APOE genotype recorded and baseline plus 12-month plasma p-tau217 measured. That is one of the largest prospective AD-biomarker cohorts ever collected inside a cardiovascular trial. It was published separately in December 2025 in The Journal of Prevention of Alzheimer's Disease. First author Michael Davidson [Davidson et al., 2025].

The abstract opens with this: "Obicetrapib significantly slowed AD biomarker progression over 12 months in participants with ASCVD, with the greatest effects in ApoE4 carriers" [Davidson et al., 2025].

That phrase — greatest effects in ApoE4 carriers — is unusual. Most brain-targeted drugs (lecanemab, donanemab) work less well in APOE4 carriers and cause more side effects. Obicetrapib goes the other direction.

The Numbers: 7.81% / 12.67% / 20.48%

Among APOE4/E4 homozygotes, obicetrapib produced a 7.81% adjusted mean decrease in p-tau217 versus a 12.67% increase on placebo — a 20.48% placebo-adjusted treatment difference (P = 0.010) [Davidson et al., 2025].

The effect scaled with genetic risk. Non-carriers showed a small effect, APOE3/4 heterozygotes a bigger one, APOE4/4 homozygotes the largest effect of any subgroup in the study. And it wasn't just p-tau217. In APOE4/4 participants, obicetrapib produced "consistent improvements across multiple AD biomarkers compared to placebo treatment, with placebo-adjusted benefits ranging from 13.67% to 22.65%" [Davidson et al., 2025] — across p-tau217, Abeta42:40 ratio, the p-tau217/Abeta ratio, GFAP (a marker of neuroinflammation), and NfL (neurodegeneration). All five moved in the right direction. Specifically on GFAP, obicetrapib produced a "-6.39% vs +8.85%" swing in APOE4/4 patients — a ~15-point placebo-adjusted gap on a neuroinflammation marker in 12 months [Davidson et al., 2025].

Davidson's Most Important Quote

The authors — conservative cardiologists and lipidologists, not neuro-maximalists — write this in their discussion:

If you are APOE4/4 and you have been told for twenty years that genetics is destiny, that sentence is the first time a peer-reviewed Phase 3 substudy has contradicted it at the biomarker level.

Three Important Caveats

I have to hit these hard. Otherwise I am not doing my job.

1. Biomarker is not cognitive outcome. No one has measured MMSE, CDR-SB, or real-world dementia incidence on obicetrapib. P-tau217 is arguably the best surrogate we have for AD pathology, but surrogate is the operative word. [link: p-tau217 blood test guide]

2. Small subgroup. The APOE4/4 cell is a slice of 1,535 patients. Pre-specified, yes — but it needs independent replication.

3. Confirming data is partial. A larger overlapping analysis (Davidson et al., 2025, AAIC abstract, 1,727 patients) showed concordant results: in APOE4 carriers, obicetrapib stabilized p-tau217 (0% increase vs 5.7% with placebo, P = 0.03), and the p-tau217/Abeta42:Abeta40 ratio rose only 2.1% on obicetrapib vs 10.2% on placebo (P = 0.005) [Davidson et al., 2025, AAIC]. That's reassuring internal consistency, but it's the same clinical program — not a different trial.

Honest synthesis: this is the strongest oral-drug AD biomarker signal ever reported in APOE4/4 patients. It is not yet proof of dementia prevention. We want a prospective randomized cognitive-endpoint trial. Until then, treat it as a major hypothesis with unusually strong backing.

Why This Plausibly Works — HDL, ApoE, and Amyloid Clearance

Why would a drug designed to shift blood lipids touch brain pathology? The mechanistic story is coherent and every step has independent support.

APOE is a lipid-carrier protein. You have two copies of the APOE gene. The e4 variant produces a protein that traffics cholesterol less efficiently in the brain. Cerebral cholesterol moves on HDL-like particles that carry ApoE on their surface; those particles help clear beta-amyloid out of the small vessels of the brain (the cerebrovasculature where APOE4 carriers are especially vulnerable to cerebral amyloid angiopathy, or CAA).

When you inhibit CETP — which is what obicetrapib does — you raise HDL particles and the ApoE riding on them [Kastelein et al., 2024]. From the 2025 mechanistic review in Journal of Cardiology and Cardiovascular Sciences: "In mice, genetic and pharmacological studies have shown that HDL levels are highly associated with CAA and that peripheral injection of synthetic HDL particles stimulates clearance of both Abeta42 and Abeta40 from the brain" [Poliakova & Wellington, 2025].

So the chain is: obicetrapib → more HDL particles carrying more ApoE → better cerebrovascular amyloid clearance → slower biomarker progression. Not a hand-wave. A pathway.

The Genetics Predicted It — Mendelian Randomization

Here is the part that should have made us see this coming.

Mendelian randomization (MR) uses genetic variants as a natural experiment. Because variants are randomly distributed at conception, people who inherit a variant that lowers CETP are essentially enrolled in a 60-year randomized trial of CETP inhibition — one they never signed up for. Schmidt and colleagues ran exactly that analysis in 2024 in Alzheimer's Research & Therapy:

LBD is Lewy body dementia (the subtype Robin Williams had). Translation: genetically lower CETP is associated with meaningfully lower LBD risk in APOE4 carriers and only marginally in non-carriers — and the interaction p-value is 5.81 × 10⁻⁴, so the difference between those effect sizes is highly significant. The authors concluded: "inhibition of CETP may be a viable strategy to treat dementia, with a more pronounced effect expected in APOE-epsilon4 carriers" [Schmidt et al., 2024]. The geneticists told us that in 2024. Obicetrapib delivered it in 2025.

This converges with older genetics. In 2006, Nir Barzilai's team studied Ashkenazi centenarians and found that "subjects with MMSE > 25 were twice as likely to have the CETP VV genotype (29% vs 14%, p = 0.02)" — a CETP variant associated with lower CETP activity tracked with preserved cognition into extreme old age [Barzilai et al., 2006]. The 2015 Cache County Study (4,486 participants, up to 12 years of follow-up) extended this: "an average 0.6-point decrease per year in the rate of cognitive decline for each additional valine (p < 0.011)" at the CETP I405V locus [Lythgoe et al., 2015]. Important calibration from the same paper: "CETP I405V is associated with preserved cognition over time but is not associated with LOAD status" [Lythgoe et al., 2015]. This is a cognitive-trajectory signal, not a confirmed Alzheimer's-diagnosis signal.

That convergence is what makes this story different from most "early promising signal" stories.

Why Earlier CETP Drugs Failed — and Why This One Didn't

CETP inhibition has a body count. Three prior drugs died on the vine:

• Torcetrapib (Pfizer) — raised cardiovascular events and death; terminated 2006.

• Dalcetrapib (Roche) — futility.

• Evacetrapib (Lilly) — futility.

• Anacetrapib (Merck) — statistically positive but small effect; shelved after it was found to accumulate in adipose tissue.

Torcetrapib is the cautionary tale that matters. From the 2024 review: torcetrapib "had structure-related off-target effects causing increased blood pressure, as well as increased aldosterone, steroid, and endothelin-1 levels, and electrolyte abnormalities" [Kastelein et al., 2024]. The CETP target wasn't the problem. The molecule was dirty.

Obicetrapib was specifically engineered to avoid those off-target effects. From the BROOKLYN Phase 3 trial: "obicetrapib was observed to be well-tolerated, with safety results comparable to placebo and no increase in blood pressure. The treatment discontinuation rate for the obicetrapib arm was 7.6% versus 14.4% for placebo" [Nissen et al., 2025]. No BP signal. Lower dropouts than placebo. The 2025 meta-analysis of seven RCTs (n=3,381) confirmed: "there were no significant differences in adverse events" [Araujo et al., 2025]. Clean molecule. The target was never cursed — the old drugs were.

The Lipid Story — LDL, Lp(a), and a Diabetes Bonus

The headline LDL effect from BROADWAY: "The least-squares mean percent change from baseline to day 84 in the LDL cholesterol level was -29.9% (95% CI, -32.1 to -27.8) in the obicetrapib group, as compared with 2.7% (95% CI, -0.4 to 5.8) in the placebo group" [Nicholls et al., 2025]. That is a ~32.6 percentage-point placebo-adjusted LDL reduction on top of maximum-dose statins, durable through day 365.

The 2025 meta-analysis pooled all seven obicetrapib RCTs: "obicetrapib significantly reduced mean LDL-C (MD: -37.21%; 95% CI: -41.53 to -32.90; p < 0.01), lipoprotein(a) (MD: -37.16%; 95% CI: -43.63 to -30.70; p < 0.01), apolipoprotein B (MD: -24.65%; 95% CI: -28.71 to -20.59; p < 0.01)" [Araujo et al., 2025]. Every trial. Same direction. Same magnitude. Which in science is the best compliment you can pay.

Lp(a): the under-discussed win for APOE4 carriers

Lipoprotein(a) is a mostly genetically determined, LDL-like particle that raises cardiovascular, stroke, and vascular-dementia risk independently of LDL-C and APOE. Until very recently, no approved drug lowered it meaningfully (pelacarsen, lepodisiran, olpasiran, muvalaplin are in development but not yet FDA-approved).

In BROOKLYN: "Treatment with obicetrapib resulted in placebo-adjusted reductions in apolipoprotein B of -24.4%, non-HDL cholesterol of -34.5% and lipoprotein(a) of -45.9%, as well as a placebo-adjusted increase in high-density lipoprotein cholesterol of +138.7%" [Nissen et al., 2025]. A 46% Lp(a) reduction from an oral drug is not the NEJM-headline story — but for a 58-year-old APOE4/4 carrier whose parent had a stroke on top of dementia, that is potentially huge. [link: Lp(a) and APOE4 cardiovascular risk]

The diabetes bonus

Statins slightly raise new-onset diabetes risk. The meta-analysis found obicetrapib does the opposite: "obicetrapib also reduced the incidence of new-onset diabetes (RR: 0.88; 95% CI: 0.80 to 0.97; p = 0.01)" [Araujo et al., 2025]. A ~12% relative reduction in new diabetes diagnoses is not the story — but if you are APOE4 and already watching glucose, that's a meaningful tailwind.

The MACE signal (exploratory)

BROADWAY was not powered as an outcomes trial, but the exploratory MACE analysis showed a "21% relative reduction in major adverse cardiovascular events (MACE) (hazard ratio [HR], 0.79; 95% CI, 0.54-1.15)" [Nicholls et al., 2025]. Pooled with BROOKLYN (JACC 2025): the coronary composite was "lower with obicetrapib (3.9% vs 5.0%; HR: 0.77; 95% CI: 0.54-1.11; P = 0.16), with a risk reduction in the second 6 months (HR: 0.60; 95% CI: 0.37-0.99; P = 0.04)" [Nicholls et al., JACC 2025]. That late-divergence pattern is exactly what you'd expect from an LDL-lowering drug — plaques take time to register a number change. Until PREVAIL reads out, there is no confirmed cardiovascular mortality benefit. The signal is encouraging. It is not proof.

TANDEM — the single-pill oral combo

TANDEM (The Lancet, 2025) tested obicetrapib 10 mg + ezetimibe 10 mg as a fixed-dose combination: "At day 84, percent differences in LDL cholesterol reduction with the FDC were -48.6% (95% CI -58.3 to -38.9) versus placebo" [Sarraju et al., 2025]. That approaches injectable PCSK9 inhibitor territory — from a once-a-day oral pill.

PREVAIL — The Trial That Could Change Everything

PREVAIL (NCT05202509) is the cardiovascular outcomes trial: 9,541 patients with established cardiovascular disease, ≥30-month median follow-up, hard MACE as primary endpoint. Status: active, not recruiting. Readout expected November 2026.

If PREVAIL demonstrates a hard-outcome CV benefit, three things follow:

1. FDA filing becomes straightforward.

2. Broad clinical availability and insurance coverage within 12-18 months of approval.

3. A dedicated APOE4 AD-prevention trial becomes likely — and that is the trial that could definitively answer the cognitive-outcome question for our genotype.

Until PREVAIL reads out, obicetrapib is pre-approval. You cannot pick it up at Walgreens.

What Should APOE4 Carriers Do Now?

Here is my actual framework as an APOE4/4 carrier, a pharmacist, and the person building Phoenix.

1. Get your baseline now. If you don't have recent labs with an advanced lipid panel (ApoB, Lp(a), fasting insulin), get them. If you haven't had a plasma p-tau217 drawn, know it is now available as a commercial blood test. You want your starting number before any therapy decision. Phoenix's Bloodwork module tracks all of these against APOE4-specific reference ranges — because normal-for-population is not the same as optimal-for-APOE4.

2. If you're already on max statin + ezetimibe + a PCSK9 inhibitor and still above ApoB/LDL goal — know that obicetrapib is the next wave. Talk to your lipidologist now. Ask where you sit relative to current 2024/2025 APOE4-appropriate targets (many of us should be aiming lower than standard guidelines). When the drug is approved, you want to be first in line.

3. If you're APOE4/4 and worried specifically about the AD trajectory — track PREVAIL. November 2026 readout. Positive CV outcomes → FDA filing → likely dedicated APOE4 prevention trial. That future trial will need enrollees. You want to be a candidate.

4. Attack every lever we already know about. Obicetrapib's mechanism tells us HDL-ApoE flux and Lp(a) matter for the APOE4 brain. We can't drop Lp(a) 46% without a drug, but we can raise HDL and protect ApoE function: aerobic exercise raises HDL; Mediterranean / MIND-pattern eating improves ApoE-mediated lipid handling; sleep fidelity protects glymphatic clearance; resistance training protects against vascular dementia. None of these individually matches obicetrapib's effect size. Collectively, started in your 50s, they move the needle. [link: APOE4 lifestyle protocol stack]

A Calibrated Bottom Line

What we have:

• Three positive Phase 3 trials on lipid endpoints [Nicholls et al., 2025; Nissen et al., 2025; Sarraju et al., 2025]

• A pooled MACE signal that looks real and emerges after 6 months [Nicholls et al., JACC 2025]

• A pre-specified APOE4 biomarker substudy with the strongest oral-drug AD biomarker signal ever reported [Davidson et al., 2025]

• An independent confirming analysis [Davidson et al., AAIC 2025]

• Two decades of genetics pointing the same direction [Schmidt et al., 2024; Barzilai et al., 2006; Lythgoe et al., 2015]

• A coherent HDL-ApoE-amyloid mechanism [Kastelein et al., 2024; Poliakova & Wellington, 2025]

• A clean safety profile across 3,381 randomized patients [Araujo et al., 2025]

What we don't have:

• A cognitive outcomes trial on obicetrapib (MMSE, CDR-SB not measured)

• A dedicated APOE4 prevention trial

• Long-term safety beyond ~12 months of published follow-up

• Confirmed cardiovascular mortality benefit (PREVAIL pending)

• FDA approval or pricing

If you asked me what drug I'm watching hardest for APOE4 carriers through 2026-2027, this is the answer. It could still disappoint — science does that. But the evidence pyramid is unusually strong for a pre-approval compound.

Where Phoenix Fits In

I built The Phoenix Community for APOE4 carriers, high-family-risk folks, and clinicians navigating this field in real time. A few pieces of the platform are built for exactly the obicetrapib decision window:

• Bloodwork — track ApoB, Lp(a), lipid particle sizing, and plasma p-tau217 against APOE4-specific reference ranges, not population averages.

• Clinical Trials module — we track active APOE4-relevant trials including the obicetrapib-adjacent pipeline and flag them to members who may qualify.

• Pods — small APOE4 carrier cohorts that run structured protocols together; consistency is easier with people in the same boat.

• Experiments & XP-Packs — structured protocols for the levers you can act on now: sleep, MIND-pattern eating, exercise dosing, supplement stacks, all logged against your labs.

If you want to navigate the obicetrapib window with people paying the same kind of attention you are, the door is open.

Frequently Asked Questions

Is obicetrapib FDA-approved?
Not as of April 2026. Obicetrapib is in late-stage Phase 3 development with NewAmsterdam Pharma. The PREVAIL cardiovascular outcomes trial (n=9,541) is expected to read out in November 2026, and a positive readout is expected to trigger FDA filing. No pricing or availability information exists yet.

Does obicetrapib work better in APOE4 carriers?
At the biomarker level, yes — and that is the unusual part. The BROADWAY APOE4 substudy showed the largest placebo-adjusted p-tau217 benefit in APOE4/4 homozygotes (20.48%, P = 0.010) [Davidson et al., 2025], and Schmidt's 2024 Mendelian randomization predicted an APOE4-specific dementia protection from CETP inhibition [Schmidt et al., 2024]. Calibration: this is a biomarker advantage and a genetics-derived prediction. We do not yet have a head-to-head cognitive-outcome trial comparing obicetrapib's effect in APOE4 carriers vs non-carriers.

What's the difference between obicetrapib and statins, ezetimibe, or PCSK9 inhibitors?
All four lower LDL but hit different targets. Statins block hepatic cholesterol synthesis. Ezetimibe blocks intestinal absorption. PCSK9 inhibitors increase hepatic LDL-receptor recycling. Obicetrapib blocks CETP-mediated cholesterol transfer in the bloodstream itself, which simultaneously lowers LDL ~30%, raises HDL ~139%, and lowers Lp(a) ~37-46% [Nicholls et al., 2025; Nissen et al., 2025]. It stacks on top of the other three mechanistically.

When will obicetrapib be available?
Speculative. If PREVAIL reads out positively in late 2026, FDA filing would follow. Typical FDA review for a compound with positive outcomes data is 10-12 months. A realistic earliest-possible commercial-availability window is late 2027 to 2028. Do not take dates from me — take them from the FDA and from NewAmsterdam Pharma's disclosures.

Can APOE4 carriers access obicetrapib clinical trials now?
The major trials (BROADWAY, BROOKLYN, TANDEM) are completed. PREVAIL is active but not recruiting. The small Phase 2a early-AD pilot (n=13, APOE4 carriers only) is also complete but was not peer-reviewed in a journal; treat its preliminary findings as hypothesis-generating. For active APOE4-relevant trials, check ClinicalTrials.gov and — if you are a Phoenix member — the Clinical Trials module, which surfaces obicetrapib-adjacent and other lipid-targeted AD prevention studies as they open.

—

Kevin

Most Newsletters? One-way street.
How boring…
This is the Phoenix Community. So let's make it a two-way street.
Got a question? Feedback?
Hit reply. I read every single one.