Connect with us

NEWS

Pelacarsen Cut Lp(a) 78%, and Heart Attacks Did Not Fall

Pelacarsen cut lipoprotein(a) 78% in Horizon and did not prevent heart attacks, leaving newly screened patients with a risk number and no proven Lp(a) drug.

Published

on

Pelacarsen cut lipoprotein(a) 78.4% in 8,322 adults with prior heart disease, and it did not cut heart attacks, strokes, or urgent artery procedures.

The Cleveland Clinic-led Lp(a)HORIZON trial did not reduce major heart events after 4.7 years of follow-up. Those events hit 12.3% of people on the monthly shot and 12.1% on placebo. Results went to the American Heart Association meeting in Chicago on Nov. 7, 2026, and into the New England Journal of Medicine the same day.

Pelacarsen Flattened Lp(a) and Left Events Untouched

Pelacarsen is an antisense oligonucleotide, a short strand of genetic material meant to stop the liver from making apolipoprotein(a), the signature piece of the Lp(a) particle. Novartis funded the trial. People injected 80 mg under the skin once a month, or a matching dummy shot.

Investigators randomized 8,350 adults at 797 sites in 42 countries from December 2019 through July 2022 and analyzed 8,322 of them for efficacy, 4,157 on pelacarsen and 4,165 on placebo. Mean age was 59.7 years. Twenty-seven percent were women. About 78% of the group was White. Everyone already had a heart attack, a stroke, or symptomatic peripheral artery disease between 3 months and 10 years before joining, plus an Lp(a) level of at least 70 mg/dL.

Median Lp(a) at the start sat at 236 nmol/L. At 12 months it was 64.4 nmol/L on pelacarsen and 233.4 nmol/L on placebo. That is an observed 78.4% drop versus 3.2% on placebo, and an estimated 77.7 percentage-point treatment difference. Over the whole trial the time-averaged fall was 161.2 nmol/L versus 17.1 nmol/L.

The lab win did not move the clinical end point, a mix of heart-related death, nonfatal heart attack, nonfatal stroke, or urgent coronary revascularization that needed a hospital stay. Steven Nissen, chief academic officer of Cleveland Clinic’s Heart, Vascular and Thoracic Institute and the trial’s study chair, had expected the opposite. “I really thought that this was going to be a slam dunk,” said Dr. Amanda Doran, an assistant professor of cardiovascular medicine at Vanderbilt University Medical Center, one of the nearly 800 sites. Vanderbilt staff had put years into the study. “We were pretty invested, so it obviously hurts.”

Nissen said the result took him by extreme surprise. Dr. Karol Watson, co-director of the UCLA Program in Preventive Cardiology, who was not on the trial, called it as flat a large heart study as she had seen in a long time. Novartis later said it would discontinue the pelacarsen atherosclerotic cardiovascular disease program, including ongoing studies and a managed access program.

The Screening Rule That Outran the Drug

The timing is awkward for anyone who just got this blood test. The 2026 ACC/AHA multisociety dyslipidemia guideline, printed in Circulation, gives a Class I recommendation to measure Lp(a) at least once in all adults for atherosclerotic risk assessment. That is the guideline’s strongest wording: the test should be done.

The American Heart Association now tells the public that about 1 in 5 people worldwide have high Lp(a), usually with no symptoms, and that a regular cholesterol panel does not catch it. Cleveland Clinic estimates elevated levels in 64 million people in the United States and 1.4 billion worldwide. The particle is 80% to 90% set by genes. Diet and workouts barely move it.

High, in the 2026 U.S. document, starts at 125 nmol/L, or 50 mg/dL, a level tied to about a 1.4-fold rise in atherosclerotic risk. At 250 nmol/L, or 100 mg/dL, estimated risk roughly doubles. The same guideline still treats that number as a reason to push LDL cholesterol down harder, not as a license to wait for an Lp(a) shot.

Donald Lloyd-Jones, who helped write that document, said after the Chicago session that Horizon does not pull the Class I test off the books. The point of measuring, he said, is to spot a risk-enhancing factor and decide whether to start or intensify treatments that already have outcome data.

THE YEAR THE TEST BEAT THE SHOT

  1. March 13, 2026: The ACC/AHA dyslipidemia guideline posts ahead of print with a Class I call for one lifetime Lp(a) measurement in every adult.
  2. April 28, 2026: The full text appears in Circulation, retiring the 2018 blood-cholesterol guideline and adding a section on people with high Lp(a).
  3. September 2026: Horizon’s top-line readout says the primary end point was missed.
  4. November 7, 2026: Full results hit the AHA meeting and the New England Journal of Medicine; Novartis later ends the pelacarsen heart-disease program.
  5. November 13, 2026: The European Atherosclerosis Society says high Lp(a) remains an inherited risk factor and still wants the once-in-a-lifetime test.

That sequence is the practical problem sitting in clinic inboxes. Millions of people can now be told they carry a genetic cholesterol their food and gym time will not fix, while the first dedicated drug has already had its outcomes test and lost.

Why a 78% Drop Changed No Heart Attacks

Horizon was not a sloppy study of untreated patients. After a 30- to 90-day stretch to tighten background care, median LDL cholesterol sat near 65 mg/dL. About 78% of people were on a high-intensity statin. Mean systolic pressure was 125 mm Hg. Mean high-sensitivity C-reactive protein was 0.8 mg/L. “These patients were optimally treated with existing therapies,” Nissen said. “Their cardiovascular risk factors were managed as well as in any cardiovascular outcome trial to date.”

THE HORIZON SCORECARD

Measure Pelacarsen Placebo
People in the efficacy analysis 4,157 4,165
Primary heart events 511 (12.3%) 503 (12.1%)
Events if baseline Lp(a) was at least 90 mg/dL 402 of 3,266 (12.3%) 403 of 3,274 (12.3%)
Median Lp(a) at 12 months 64.4 nmol/L 233.4 nmol/L
LDL cholesterol at 12 months 62.3 mg/dL 71.2 mg/dL
Serious adverse events 37.3% 41.1%
Stopped the study drug for side effects 20.6% 7.5%

The 12.3 percent versus 12.1 percent split on the primary end point gave a hazard ratio of 1.02 (0.90 to 1.15), P=0.76. In the prespecified group with Lp(a) of at least 90 mg/dL, 78.6% of each arm, the ratio was 1.00, P=0.99. Heart attack, stroke, urgent revascularization, and heart death, taken one at a time, did not separate. A three-part mix that dropped urgent procedures, a coronary-focused mix, and all-cause death were also null. So were on-treatment analyses, uninterrupted-treatment analyses, total-event counts, and landmark looks by how far Lp(a) had fallen at 3 months.

These results were disappointing and unexpected, particularly given the findings of prior observational and Mendelian randomization studies.

Steven Nissen, M.D., study chair of Lp(a)HORIZON, Cleveland Clinic

Nissen told colleagues the finding was not a near miss. The hazard ratio sat on 1.02, and the secondary cuts landed in the same place. Pelacarsen also came with a tolerability cost. Drug-related adverse events hit 56.8% versus 18.0% on placebo. Flu-like constitutional symptoms ran 31.9% versus 5.8%. Investigator-reported hypersensitivity ran 17.1% versus 5.4%. Any premature stop of the regimen ran 35.1% versus 18.3%. Mean time on the study drug was 3.5 years versus 4.0 years. Serious adverse events were not higher on the drug (37.3% versus 41.1%), which is why the story is a failed benefit, not a safety blow-up.

Two explanations now sit on the table, and they can both be true. Watson noted that Lp(a) is high from conception. Intervening at a mean age of 59.7 for 4.7 years may be late if decades of plaque and clot risk have already accrued. Nissen added that genetic studies describe a lifetime of lower exposure, which is a different experiment from a midlife shot. The other explanation is residual risk. Horizon’s event rate ran low enough that the trial had to run longer than planned to collect the needed events. If LDL, blood pressure, and antiplatelet drugs had already squeezed the risk Lp(a) is supposed to add, a 78.4% lab change would have little left to show.

Pelacarsen also lowered LDL by 13.3 percentage points versus placebo, to 62.3 mg/dL from a placebo value of 71.2 mg/dL at one year. Even that extra LDL nudge did not show up in events. Leslie Cho, the trial’s first author and co-section head of preventive cardiology at Cleveland Clinic, still called Lp(a) a causal risk factor and said Horizon shows how much LDL lowering still matters.

Keep LDL Low, Because That Still Worked

Sotirios Tsimikas, a University of California San Diego cardiologist, an Lp(a) clinic director, and a Horizon investigator, wrote after the embargo lifted that he does not read the trial as a failure for patients who already live with a high number. He sees a workable plan that does not wait on a new molecule: keep LDL cholesterol below 65 mg/dL or lower, keep blood pressure at 120 or lower, treat the rest of the usual list, and take medicines that already have outcome proof. “Lp(a) risk can be managed,” he wrote, aimed at people in his clinic who have carried years of worry about a number food cannot change.

That reading fits the trial’s own baseline. These were not people left on weak statins. They were already at LDL levels many primary-care offices still treat as a win. John Wilkins, a Northwestern Medicine cardiologist and a co-author of the U.S. Lp(a) guidance, said the miss does not erase Lp(a) as a marker that spots people more likely to have an event. It means this drug, in this already-treated secondary-prevention group, did not change that event rate.

WHAT STILL CUTS RISK WITH A HIGH LP(A)

  • LDL cholesterol: Drive it as low as your clinician can with a high-intensity statin and, if needed, ezetimibe or a PCSK9 drug, the path Cho said Horizon still supports.
  • Blood pressure: Horizon’s mean systolic reading was 125 mm Hg; Tsimikas’s clinic target is 120 or lower.
  • The rest of the usual list: Diabetes control, no tobacco, sleep, weight, and antiplatelet drugs when they are already indicated after an event.
  • Family context: A high result is a reason to ask first-degree relatives about their own test, because the trait clusters in families.

Doran put the same idea in plainer clinic language: people with high Lp(a) need to know it so they will chase prevention that already works. Watson asked how much Lp(a)-linked risk was even left once LDL, pressure, and aspirin were in place. Horizon cannot answer that with a neat percentage. It can only show that subtracting most of the circulating particle, for 4.7 years, in people whose other numbers were already tight, did not change the event curve.

Four Remaining Trials Aim at Different Patients

Nissen said no further study of pelacarsen is planned, and that one trial does not close the idea of lowering Lp(a). The next shots are mostly small interfering RNA drugs, which silence the same liver message by a different chemical route and, in earlier studies, drove Lp(a) down further than pelacarsen did. They are also aimed at somewhat different people.

THE TRIALS STILL RUNNING

Drug Maker Event trial How it differs from Horizon
Lepodisiran (siRNA shot) Eli Lilly ACCLAIM-Lp(a), 17,300 people, primary completion estimated March 2029 Includes people with established disease and people at risk for a first event
Olpasiran (siRNA shot) Amgen OCEAN(a)-Outcomes, 7,297 people, estimated around March 2028 Primary mix is coronary death, heart attack, or urgent revascularization, with no stroke component
Olpasiran (siRNA shot) Amgen OCEAN(a)-PreEvent Built to test first-event prevention, not another secondary-prevention rerun
Muvalaplin (daily pill) Eli Lilly MOVE-Lp(a) Oral small molecule, not an injection

Lilly’s 17,300-person ACCLAIM-Lp(a) study is already listed as active and not recruiting, with results estimated in March 2029. Silence Therapeutics’ zerlasiran, another siRNA shot, does not yet have a registered outcomes trial. Early CRISPR work aimed at switching off the LPA gene is further back. Nissen has also pointed to aortic stenosis studies, a valve disease tied to Lp(a) that Horizon was not built to test.

If the “too late, too treated” reading of Horizon is right, the trials that enroll people before a first event, and the ones that demand higher starting Lp(a) than Horizon’s 70 mg/dL, are the ones that can still rescue the drug idea. If the reading is wrong, and midlife Lp(a) lowering simply does not change events once LDL is already low, those trials will come back as flat as Horizon. Either way, none of them reports tomorrow. Patients getting an Lp(a) result this winter will not be waiting on pelacarsen.

After Horizon, Europe Still Wants the Test

On Nov. 13, 2026, the European Atherosclerosis Society posted a formal statement on the post-Horizon period. High Lp(a) remains an inherited risk factor, independent of LDL and the other usual markers, the society said. It still recommends measuring Lp(a) at least once in everyone, using the result to guide prevention, especially LDL lowering, and to find relatives who may share the trait. It also said more trials are needed before anyone can claim that lowering the particle prevents disease.

That is a colder sentence than the “slam dunk” mood that surrounded pelacarsen a year ago, and it is also more useful. The test still sorts people. The first dedicated drug did not change what happened next in people who had already had an event and whose LDL and blood pressure were already handled. Nissen’s own close was that Horizon is not the end of the Lp(a) story, only the end of this molecule in this setting.

Novartis has closed the pelacarsen heart-disease program. The 2026 U.S. guideline still calls for one Lp(a) measurement in every adult. A high result remains a reason to treat LDL, pressure, and the rest of the proven list with more urgency, while the later shots enroll different patients and ask a narrower question than Horizon did.

Frequently Asked Questions

Does a Standard Cholesterol Panel Include Lipoprotein(a)?

No. A usual lipid panel reports total cholesterol, LDL, HDL, and triglycerides. Lp(a) is a separate blood test a clinician has to order, which is why so many people with a “normal cholesterol” result have never seen this number. The American Heart Association says that separate test is the only way to know your level.

How Often Should Adults Repeat an Lp(a) Test?

Once is enough for most people, because the level is largely genetic and stays in a similar range over years. The 2026 U.S. dyslipidemia guideline allows a repeat when an intermediate result coincides with menopause, pregnancy, new chronic kidney disease, or hypothyroidism, states that can move Lp(a). Statins can raise it slightly and PCSK9 inhibitors can lower it, which is another reason an intermediate value sometimes gets a second look.

If My Lp(a) Is High, Should Relatives Be Tested?

The 2026 U.S. guideline gives a Class I recommendation to test first-degree relatives, parents, siblings, and children, when a person has familial hypercholesterolemia, premature atherosclerotic disease, or high Lp(a). Levels also tend to run higher in people of African descent and in South Asian populations, another reason a family conversation around the test is worth having.

Is There an Approved Drug That Lowers Lp(a) to Prevent Heart Attacks?

No. No Lp(a)-specific medicine is approved to prevent heart attacks or strokes. Pelacarsen was the first large outcomes test of that idea. After Horizon, Novartis discontinued the pelacarsen atherosclerotic cardiovascular disease program, including a managed access program. Other shots and one daily pill remain in event trials whose first readouts are estimated in 2028 and 2029.

Disclaimer: This article is news reporting on a completed clinical trial and on published heart-risk guidance. It is for information only and is not a diagnosis, a treatment plan, or personal medical advice. It does not tell anyone to start, stop, or change a medicine, supplement, diet, exercise routine, or screening test. Talk with a licensed physician or cardiologist who knows your history before you act on an Lp(a) result or on any drug discussed here. Figures, trial statuses, and guideline language reflect the sources available on the dates named in the article and can change as later studies finish.

Harry is the editor of REMEDIES HEALTH, an independent health title that he owns and runs, covering fitness, nutrition, food, mental health, public health and home remedies. He has been in journalism for ten years, a reporter before he was an editor, with most of that time on health and science, where the gap between a headline and the study behind it is usually the story. Articles are built from peer-reviewed trials, systematic reviews and meta-analyses, trial registry records, and the guidance published by public health bodies, with each study reported alongside its size, duration, comparator and funding source. Remedies are covered by what the evidence actually shows, including when it shows nothing, and fitness guidance is checked against training research rather than gym folklore. Nutrition numbers are verified against food composition databases before publication. Mistakes are handled under a public corrections policy, and a corrected article carries a note explaining the change. Nothing on the site replaces a clinician; readers with symptoms or on medication should seek proper medical care before changing what they do. Harry answers reader mail at support@remedieshealthfitness.com.

Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending