In 29 people with two APOE4 copies inside a heart-drug trial, a year of a once-daily cholesterol pill dropped a blood Alzheimer marker by about 8%, while the same marker rose about 13% on placebo. That 20-point gap is the number that made headlines. [1]
It is a blood-marker finding, not a memory finding. It comes from a 29-person subgroup inside a cardiovascular study. It is still the first randomized signal that an oral lipid drug can move Alzheimer blood markers in the genotype that needs that news most. It does not belong on your to-do list yet.
What this page answers
- What obicetrapib is, and why a heart drug showed up in Alzheimer news
- What changed in p-tau217, including the two-copy subgroup
- What a blood-marker shift is not
What the pill does
Obicetrapib is a once-daily 10 mg pill that blocks cholesteryl ester transfer protein, or CETP. CETP is an enzyme that moves cholesterol from HDL particles onto LDL particles. Block it and LDL cholesterol falls while HDL cholesterol rises.
In the parent BROADWAY trial, 2,530 people with atherosclerotic cardiovascular disease or inherited very high cholesterol, already on the strongest lipid-lowering therapy they could tolerate, were randomized 2:1 to obicetrapib or placebo for 12 months. [2] LDL cholesterol (low-density lipoprotein, whose particles can deposit cholesterol in artery walls) fell about 30% on drug versus a 2.7% rise on placebo. HDL cholesterol (high-density lipoprotein, particles involved in returning cholesterol toward the liver) rose about 136%, and lipoprotein(a) (an inherited LDL-like particle with an extra protein tail, tracked once because diet barely moves it; see Lp(a) and APOE4) fell about 34%. [1] The lipid and safety results were reported separately as a pivotal heart program. [3]
Earlier CETP blockers failed as heart drugs, sometimes with safety problems, so the class carries a long shadow. Obicetrapib is a later, more potent molecule. That history is why a brain-marker headline from this class needs extra skepticism, not extra hope. John Kastelein, a co-author, is also a co-developer of the drug, which is a conflict to keep in view. For the CETP idea as it sounds on a podcast, see our Kastelein summary.
What changed in the Alzheimer blood markers
The Alzheimer analysis was planned in advance on stored plasma. Of 2,530 randomized participants, 1,535 had known APOE status and a p-tau217 reading (a phosphorylated tau-protein fragment in blood that tracks Alzheimer-related amyloid and tau biology; higher is generally worse; background in blood-based biomarkers) above the test’s lowest reliable value at baseline and at 12 months. Median age was 67, 67% were men, and 85% were Caucasian. The genotype split was 1,045 with ε3/ε3, 338 with ε3/ε4, 29 with ε4/ε4, plus smaller ε2 groups. [1]
Baseline p-tau217 already tracked genotype: median 0.39 pg/mL in ε3/ε3, 0.46 in ε3/ε4, and 0.56 in ε4/ε4. That is the biology you would expect, sitting in a cardiology clinic rather than a memory clinic.
Over 12 months, obicetrapib slowed the rise in p-tau217 in the full analysis set: adjusted mean +2.09% versus +4.94% on placebo. In all ε4 carriers, the rise was +1.92% versus +6.91%. In the 29 people with two copies, drug fell 7.81% and placebo rose 12.67%. Two other injury markers in that tiny two-copy group moved the same way. GFAP, a marker of astrocyte stress, fell 6.39% versus an 8.85% rise on placebo. Neurofilament light, a marker of axon injury, fell 10.49% versus a 6.82% rise. [1]
People with higher drug levels in blood at the end of the study also had bigger p-tau217 improvements. That is at least consistent with a dose-linked effect rather than a random wobble.
Why a CETP pill might matter to an ε4 brain
APOE’s day job is moving cholesterol. The ε4 version does that job poorly inside the brain and, in blood, tends to ride along with higher LDL and ApoB. One working hypothesis is that HDL particles carrying more cholesterol, and in particular their main protein ApoA-I, help pull amyloid out of vessel walls. That is still mostly lab-model work, laid out in HDL, ApoA-I, and cerebral amyloid clearance. Blocking CETP is a drug way to load those HDL particles and drop LDL at the same time.
There is also older human genetics: people with naturally low CETP activity have been reported to decline more slowly, with a stronger signal in APOE4 carriers. Genetics is not a drug trial. It is the reason anyone thought to look at p-tau217 inside BROADWAY.
Lipid and Alzheimer-marker shifts at a glance
| What | Figure | Context |
|---|---|---|
| Parent trial | 2,530 randomized, 12 months, 10 mg daily | lipid trial, not a dementia trial |
| Alzheimer analysis | 1,535 with APOE + paired p-tau217 | 890 lacked APOE status |
| Two-copy subgroup | 29 people | the headline 20.5-point gap |
| p-tau217, all participants | +2.09% vs +4.94% placebo | P = 0.025 |
| p-tau217, ε4 carriers | +1.92% vs +6.91% placebo | P = 0.041 |
| p-tau217, two copies | −7.81% vs +12.67% placebo | 20.48-point difference, P = 0.010 |
| LDL-C in BROADWAY | −29.9% vs +2.7% placebo | 32.6-point difference |
| Cognition | not measured | the missing endpoint |
What a blood-marker shift is not
BROADWAY did not test memory, daily function, or dementia. A 12-month shift in p-tau217 can mean the drug touched Alzheimer biology, or it can mean a lipid-linked assay moved without changing the brain’s future. The authors ran several comparisons without a correction for that, and the two-copy result sits on 29 people. About 35% of the trial lacked APOE status. The investigators include people with a financial stake in the molecule.
Prior CETP drugs taught the field that spectacular HDL numbers can fail to help patients. Until a dedicated Alzheimer or prevention trial reports thinking outcomes, treat this as a mechanism check, not a treatment. It also does not compete with the lifestyle package in FINGER and US POINTER. Those trials measured thinking. This one measured a blood protein.
What to watch, and what not to change
- Do not try to get obicetrapib for your brain. It is not approved as an Alzheimer drug. If a clinician ever offers it, that conversation will be about lipids, not about p-tau217.
- Keep the lipid work that is already evidence-based. ApoB, not HDL-C, is still the particle count that tracks artery injury. See ApoB versus LDL and APOE4, cholesterol, and the heart.
- Watch the next trial, not this substudy. The useful milestones are a cognition-endpoint trial in carriers, a repeat of the two-copy blood-marker split, and whether regulators ever treat p-tau217 change as a prevention stand-in. For other experimental bets, see APOE-targeted therapies.
- If prevention research appeals to you, finding a trial is the way a biomarker story becomes a human one.
Common questions
Why did two-copy carriers show the biggest shift? They started with the highest p-tau217, and the analysis found larger drug effects where baseline levels were higher. It could be biology. It could also be a small-subgroup artifact. Both can be true at once.
Should I raise my HDL with anything else? Chasing the HDL number has a poor track record. The ApoA-I deep dive is about particle function and vessel health, not about buying an HDL supplement.
A heart drug moved Alzheimer blood markers in carriers, including a striking two-copy split on 29 people. Watch the next trial. Do not rearrange your care around a biomarker substudy.