Lp(a) as a drug target: what the trials have measured, and what they have not

Your lipid panel comes back and one line on it has not budged — not after the diet changes, not after a year on a statin. It is labelled Lp(a). The practical question follows immediately: can this number be treated, and if it comes down, does anything change for your heart? Since September 2026 there is a first measured answer to the second half of that question, and it is not the one the field expected.

This is an informational article. It is not a diagnosis, not a treatment plan, and it does not replace a consultation with your doctor.

Why doesn't this number respond to diet?

Lipoprotein(a) is an LDL-like particle with an extra protein, apolipoprotein(a), bolted onto it. How much of it circulates in you is set largely by the LPA gene you inherited. The practical consequence is blunt: Lp(a) is largely genetic, and diet does not meaningfully change it.

That is why it became a drug target rather than a lifestyle target. For the number to fall, something has to interfere with the particle's assembly or with the genetic instructions for building it — and all four drugs below do one of those two things.

One wrinkle before you compare your result to anything you read: these trials set their entry thresholds in different units. Pelacarsen's phase 2 required 60 mg/dL or more, printed as 150 nmol/L; its phase 3 used 70 mg/dL, given as roughly 149 nmol/L; the olpasiran trial above 150 nmol/L, the lepodisiran and muvalaplin trials 175 nmol/L or above. The mg/dL-to-nmol/L correspondence in those papers is approximate, so check which unit your laboratory used.

Can the number be brought down?

In the people these trials enrolled it came down, and far down — measured in four separate randomised, double-blind, placebo-controlled phase 2 trials, each a few hundred people selected for high Lp(a), most already on cholesterol treatment.

Pelacarsen is an antisense oligonucleotide aimed at the messenger RNA for apolipoprotein(a). In 286 people with cardiovascular disease and Lp(a) at or above 60 mg/dL, injected under the skin for six to twelve months, mean Lp(a) fell by 35% to 80% depending on the schedule, against 6% on placebo. Those are within-group changes, not placebo-subtracted differences, and the abstract does not attach a single analysis week to them.

Olpasiran is a small interfering RNA against LPA. In 281 people with atherosclerotic cardiovascular disease and Lp(a) above 150 nmol/L, the placebo-adjusted change at week 36 ran from −70.5% at the lowest dose to −101.1% at 225 mg every 12 weeks. A figure past −100% is arithmetic, not biology: placebo drifted up 3.6% while the drug came down. Note who was in that trial: median LDL cholesterol was already 67.5 mg/dL, 88% on a statin, 23% on a PCSK9 inhibitor. This measured what is left after good treatment, not untreated cholesterol.

Lepodisiran is a longer-acting siRNA against the same target. In 320 randomised participants with a median starting Lp(a) of 253.9 nmol/L, the primary endpoint averaged the placebo-adjusted change across days 60 to 180: −93.9 percentage points (95% CI −95.1 to −92.5) for the pooled 400 mg groups. After a second 400 mg dose the level was still 74.2 percentage points below baseline at day 540 — about eighteen months. Serious adverse events occurred in 35 participants, none judged related by the investigators.

Muvalaplin is the odd one out: a daily tablet, a small molecule that blocks the particle from being assembled rather than silencing the gene's message. In 233 adults with Lp(a) at or above 175 nmol/L plus atherosclerotic disease, diabetes or familial hypercholesterolaemia, twelve weeks at 240 mg a day gave a placebo-adjusted reduction of 85.8% (95% CI 83.1 to 88.0) on an assay that counts intact particles.

That last one carries the most useful lesson here. In the same people at the same dose, the conventional apolipoprotein(a) assay read 68.9% — and on that assay the 60 mg dose scored slightly higher than 240 mg, at 70.0%. The headline figure depends on which assay you read it off. Twelve weeks is also a short exposure.

So why did the first outcomes trial not work?

Lowering Lp(a) and preventing a heart attack are two different claims, and until 2026 no trial of the second had reported a result. Lp(a)HORIZON, a phase 3 outcomes trial, reported first. It randomised 8,323 people with established cardiovascular disease — a previous heart attack, an ischaemic stroke, or symptomatic peripheral artery disease — and Lp(a) at or above 70 mg/dL, one to one, to monthly pelacarsen 80 mg or placebo on top of optimised usual care. The design wrote in a minimum of two and a half years of follow-up and ran until 993 adjudicated primary events had accumulated. The primary endpoint: time to cardiovascular death, non-fatal heart attack, non-fatal stroke, or urgent coronary revascularisation requiring hospitalisation — tested in the whole trial population and, as a second primary endpoint, in those at or above 90 mg/dL.

On 4 September 2026 the sponsor announced that this endpoint was not met in the overall population. Lp(a) was lower on pelacarsen. The events did not follow.

Two features of that announcement matter as much as the result. It is a company statement, not a peer-reviewed paper. And it withholds the figures that would let anyone interpret it: no hazard ratio, no confidence interval, no event counts, no number for how much Lp(a) fell on treatment. Full data are promised for a future congress.

What this does not mean

• It does not show that Lp(a) is irrelevant. A trial that misses its endpoint has not demonstrated the effect it was built to detect, in its own conditions and its own population — and that is the whole of what a miss establishes. With no hazard ratio and no confidence interval released, nothing has been ruled out. Whether a deeper reduction, a higher starting level, longer exposure or a different endpoint would also fail is untested; the trial's second primary endpoint, in the population at or above 90 mg/dL, has not been reported.

• It does not transfer to the other three drugs: different molecules, different depths of lowering, outcome trials still running.

• In the other direction — a phase 2 win is a laboratory endpoint. A concentration in a tube is not an event in a person. None of the four is an approved medicine, and the sponsor's own September statement said no approved treatment aimed at Lp(a) exists.

• None of the four phase 2 trials used an active cholesterol drug as comparator. Placebo was added to usual care.

• If your Lp(a) is normal, you were outside every one of these populations. Nothing here was measured in people like you.

Where do the other three stand?

All three are in phase 3 outcome trials with no readout. Olpasiran — OCEAN(a)-Outcomes, 7,297 enrolled, active and no longer recruiting, primary completion estimated 31 March 2028. Lepodisiran — ACCLAIM-Lp(a), active and no longer recruiting, enrolment still listed as an estimate of 17,300 rather than a locked count, estimated March 2029. Muvalaplin — MOVE-Lp(a), recruiting, estimated enrolment 10,450, estimated March 2031. Those are registry estimates, and registry estimates move.

What to do with your number

Treat it as information about a largely inherited trait, not a verdict. How much your own Lp(a) contributes to your own risk is a different question from the one these four trials asked, and they do not answer it. Check the unit before comparing yourself to any threshold you read elsewhere.

The conversation worth having with a cardiologist is about the whole picture: the rest of the lipid panel, blood pressure, family history, what is already being treated. Not about a drug that does not exist as a prescription.

Because Lp(a) is largely genetic and does not respond to diet, it is worth having in writing rather than in memory. If you keep your labs in Lonevi, it sits in your card alongside the rest of your lipid profile.

Sources

• Tsimikas S et al. Lipoprotein(a) Reduction in Persons with Cardiovascular Disease (pelacarsen phase 2). N Engl J Med 2020;382:244–255. PMID 31893580 · doi:10.1056/NEJMoa1905239

• Cho L et al. Design and Rationale of the Lp(a)HORIZON Trial. Am Heart J 2025;287:1–9. PMID 40185318 · doi:10.1016/j.ahj.2025.03.019

• Lp(a)HORIZON (pelacarsen phase 3 outcomes), registry record: 8,323 participants, completed. NCT04023552. The topline of 4 September 2026 comes from the sponsor's own announcement and has not been peer-reviewed or published; no hazard ratio or event counts were released with it.

• O'Donoghue ML et al. Small Interfering RNA to Reduce Lipoprotein(a) in Cardiovascular Disease (olpasiran, OCEAN(a)-DOSE). N Engl J Med 2022;387:1855–1864. PMID 36342163 · doi:10.1056/NEJMoa2211023

• OCEAN(a)-Outcomes (olpasiran phase 3), registry record: 7,297 enrolled, active not recruiting. NCT05581303

• Nissen SE et al. Lepodisiran — A Long-Duration Small Interfering RNA Targeting Lipoprotein(a) (ALPACA). N Engl J Med 2025;392:1673–1683. PMID 40162643 · doi:10.1056/NEJMoa2415818

• ACCLAIM-Lp(a) (lepodisiran phase 3), registry record: estimated enrolment 17,300, active not recruiting. NCT06292013

• Nicholls SJ et al. Oral Muvalaplin for Lowering of Lipoprotein(a): A Randomized Clinical Trial (KRAKEN). JAMA 2025;333:222–231. PMID 39556768 · doi:10.1001/jama.2024.24017

• MOVE-Lp(a) (muvalaplin phase 3 outcomes), registry record: recruiting, estimated enrolment 10,450. NCT07157774