Pelacarsen Cut Lp(a) by 78% and Prevented No Heart Attacks. Four Trials With 46,000 Patients Are Testing the Same Idea
The full Lp(a)HORIZON results, published November 7, show MACE in 12.3% of pelacarsen patients and 12.1% on placebo. ClinicalTrials.gov lists four more Lp(a) outcome trials from Amgen and Eli Lilly, planned to enrol 46,047 people.

Lipoprotein(a), or Lp(a), is a cholesterol-carrying particle whose level is set mostly by genes. High levels go with more heart attacks and strokes, and statins do not lower it. For a decade the bet has been simple: a drug that lowers Lp(a) should prevent cardiovascular events.
On November 7, the first outcome trial of that bet was presented in full at the American Heart Association meeting and published in The New England Journal of Medicine. Novartis's pelacarsen lowered Lp(a) as designed. It did not change how many patients had a heart attack, stroke, urgent revascularization or cardiovascular death.
What did Lp(a)HORIZON show?
The trial randomized 8,323 adults with established cardiovascular disease and Lp(a) of at least 70 mg/dL to monthly pelacarsen or placebo, at 797 sites in 42 countries. Median follow-up was 4.7 years.
Pelacarsen lowered Lp(a) by 78% and left major cardiovascular events unchanged: 12.3% against 12.1% on placebo.
Lp(a)HORIZON, 8,323 patients with established cardiovascular disease. Left: change in Lp(a) from baseline at 12 months. Right: share of patients with a first major adverse cardiovascular event over a median 4.7 years.
Source: Cho et al., NEJM, November 7, 2026, as reported by HCPLive and TCTMD; AHA Scientific Sessions release. Events: 511 of 4,157 patients on pelacarsen and 503 of 4,165 on placebo. Hazard ratio 1.02 (95% CI 0.90 to 1.15), P = 0.76.
View data
| measure | pelacarsen | placebo |
|---|---|---|
| Lp(a) change at 12 months | -78.4 | -3.2 |
| Patients with a major cardiovascular event | 12.3 | 12.1 |
The numbers are as flat as an outcome trial gets. Median Lp(a) started at about 236 nmol/L. At 12 months it was 64.4 nmol/L on pelacarsen and 233.4 nmol/L on placebo. The primary endpoint occurred in 511 of 4,157 pelacarsen patients and 503 of 4,165 on placebo, a hazard ratio of 1.02 with a confidence interval from 0.90 to 1.15.
The patients with the highest levels did no better. In those with Lp(a) of 90 mg/dL or more, about 192 nmol/L, the hazard ratio was 1.00. Senior author Steven Nissen summarized it in the AHA release: "it didn't work."
Did anything else go wrong?
Serious adverse events were similar, 37.3% on pelacarsen and 41.1% on placebo, according to the AHA release. Some reactions were more common with the drug: flu-like symptoms, injection site reactions and allergic reactions, including 19 cases of anaphylaxis on pelacarsen and none on placebo. About 21% of pelacarsen patients stopped the study drug because of adverse reactions, against just over 7% on placebo.
The trial register already shows the consequences. A planned US study of pelacarsen after acute coronary syndrome (NCT07625306) is withdrawn, with the HORIZON result given as the reason. A 5,700-patient open-label extension for HORIZON participants (NCT07517263) is listed as active and no longer recruiting.
Who is still running the experiment?
Three other drugs that lower Lp(a), working in different ways, are in cardiovascular outcome trials. A search of ClinicalTrials.gov for Phase 3 Lp(a) studies returns four of them, all with major cardiovascular events as the primary endpoint.
The four Lp(a) outcome trials still running plan to enrol 46,047 patients, 5.5 times as many as the one that just failed.
Phase 3 cardiovascular outcome trials of Lp(a)-lowering drugs: actual enrolment for completed or closed trials, planned enrolment for those still recruiting. Years are estimated primary completion dates.
Source: ClinicalTrials.gov API v2, queried November 9, 2026 (NCT04023552, NCT05581303, NCT06292013, NCT07157774, NCT07136012). OCEAN(a)-Outcomes enrolment is actual. The other three ongoing trials report estimated enrolment. Completion dates are sponsor estimates and often move.
View data
| trial | patients |
|---|---|
| Lp(a)HORIZON, pelacarsen (Novartis), completed | 8323 |
| OCEAN(a)-Outcomes, olpasiran (Amgen), results due 2028 | 7297 |
| ACCLAIM-Lp(a), lepodisiran (Lilly), results due 2029 | 17300 |
| MOVE-Lp(a), muvalaplin (Lilly), results due 2031 | 10450 |
| OCEAN(a)-PreEvent, olpasiran (Amgen), results due 2031 | 11000 |
The two olpasiran trials and lepodisiran are small interfering RNAs, a different way of silencing the same gene that pelacarsen, an antisense drug, targets. Muvalaplin is a pill that blocks the particle from assembling. Lp(a) cutoffs also differ: OCEAN(a) requires 200 nmol/L or more, and the Lilly trials 175 nmol/L, higher than HORIZON's 70 mg/dL, about 148 nmol/L.
The design that matters most is who is enrolled. HORIZON took only patients who had already had a cardiovascular event. OCEAN(a)-PreEvent excludes anyone with a prior event, and the two Lilly trials accept people at risk of a first event as well as those with one.
Is the Lp(a) hypothesis dead?
Not on this evidence alone. Several explanations were offered at the meeting, and the trial cannot tell them apart. Patients were already well treated, with a median LDL cholesterol of about 65 mg/dL and about 78% on a high-intensity statin. Lp(a) is high from birth, so lowering it for under five years in people who already have heart disease may be too little, too late. And the cutoff of 70 mg/dL may have included many patients whose risk from Lp(a) was modest.
What the trial does rule out is the simple version of the bet: that a 78% drop in the number on the lab report translates into fewer events for people who already have cardiovascular disease.
Limits
We read the NEJM paper through its press coverage and the AHA release; the full article is behind a paywall. Event counts and the hazard ratio come from HCPLive and were checked against TCTMD, which reports the same figures. The serious adverse event and discontinuation numbers are from the AHA release. Our trial table comes from one ClinicalTrials.gov search and may miss outcome trials registered without the phrase lipoprotein(a), including trials registered only outside the US. Planned enrolment and dates are sponsor estimates.
Sources
- American Heart Association, Lp(a) HORIZON trial: Lp(a) fell by 78%, but cardiovascular outcomes unexpectedly neutral, November 7, 2026: 8,323 patients, 797 sites in 42 countries, 14,249 screened, adverse events, discontinuations, 19 cases of anaphylaxis, the Nissen quote.
- Cho L, Nicholls SJ, Frndak S, et al. Pelacarsen and cardiovascular outcomes in patients with high lipoprotein(a). N Engl J Med, published online November 7, 2026 (not read directly; figures via sources 3 and 4).
- HCPLive, Lp(a) HORIZON shows pelacarsen lowers Lp(a) 78% but not cardiovascular events, November 7, 2026: 511 of 4,157 and 503 of 4,165 events, HR 1.02 (0.90 to 1.15), P = 0.76, Lp(a) change of 78.4% and 3.2%, median LDL cholesterol 64.6 mg/dL.
- TCTMD, Full Lp(a)HORIZON results affirm lack of pelacarsen benefit in secondary prevention, November 7, 2026: median Lp(a) 236 nmol/L at baseline, 64.4 and 233.4 nmol/L at 12 months, subgroup HR 1.00.
- ClinicalTrials.gov API v2, queried November 9, 2026: NCT04023552, NCT05581303, NCT06292013, NCT07157774, NCT07136012, NCT07517263, NCT07625306.
- Related: Half of FDA's new drugs already target rare diseases and how to read Regeneron's muscle data for GLP-1 users.
Recommended Reading

FDA Approves Sunharbo, a Factor XIa Inhibitor: Ischemic Stroke Down 26%, Major Bleeding Up 0.2 Points

Five Weeks of RASONQUE Sales Came to 89% of Lumakras’s Best US Year

Treasury Wants to Fence Off Chinese Biotech. Industry Trials With a Site in China Already Equal 68% of the US Count
Discussion (0)
Sign in to join the conversation and share your thoughts
No comments yet. Be the first to start the discussion!