Drug cut Lp(a) cholesterol by 78% yet heart attacks did not fall

Pelacarsen, a monthly injection designed to lower lipoprotein(a), cut levels of this inherited form of cholesterol by about 78% in the phase 3 Lp(a) HORIZON trial. It made no difference to heart attacks, strokes or deaths from heart disease. After a median of 4.7 years, 12.3% of people on pelacarsen had a major cardiovascular event, against 12.1% on placebo.

The results were presented on 7 November 2026 as a late-breaking trial at the American Heart Association's Scientific Sessions in Chicago and published the same day in the New England Journal of Medicine. A second paper with extended analyses appeared in NEJM Evidence. Novartis, which licensed pelacarsen from Ionis Pharmaceuticals, funded the study. Cleveland Clinic led it.

This was the first outcomes trial of any drug built to lower Lp(a). Doctors and drug companies had been waiting years for it, because Lp(a) affects roughly one in five people and has no approved targeted treatment. Novartis had already said in a September statement that the trial missed its main goal. The full data, released in Chicago, show how complete the miss was.

What is lipoprotein(a)?

Lp(a) is a cholesterol-carrying particle in the blood. It looks a lot like LDL, the "bad" cholesterol, with an extra protein called apolipoprotein(a) attached. Like LDL, it can help build up the fatty plaque that narrows arteries.

Your Lp(a) level is mostly set by your genes. According to the American Heart Association, diet and exercise barely move it, and neither do statins. The AHA puts the risk this way. A level of 125 nmol/L or more is linked with about a 1.4-fold higher long-term risk of heart attack or stroke, and a level of 250 nmol/L with at least double the risk.

Most people have never had it measured. The 2026 multi-society guideline on managing blood fats, cited by the AHA, recommends checking Lp(a) at least once in adulthood. High Lp(a) causes no symptoms of its own.

Large genetic studies, known as Mendelian randomisation studies, found that people born with low Lp(a) have fewer heart attacks over a lifetime. That made Lp(a) look like a direct cause of heart disease, and so a promising target for a drug. Steven Nissen, the trial's study chair at Cleveland Clinic, told Consult QD that the results were "disappointing and unexpected, particularly given the findings of prior observational and Mendelian randomization studies."

How the trial worked

Lp(a) HORIZON was double-blind and placebo-controlled. Neither patients nor their doctors knew who was getting the real drug. Our guide to placebo and blinding explains why that protects results like these from bias.

The trial ran at 797 sites in 42 countries. It enrolled adults who had already had a heart attack, a stroke caused by a clot or narrowed leg arteries, and whose Lp(a) was at least 70 mg/dL (about 148 nmol/L). That threshold puts someone in roughly the top 20% of the world's population.

Before randomisation, people had 30 to 90 days to get their other risk factors to guideline targets if needed. About 75% already met the targets for blood pressure and cholesterol. Patients were then assigned one to one to an 80 mg injection under the skin of pelacarsen or a matching placebo, once a month.

Pelacarsen is an antisense oligonucleotide. It is a short strand of synthetic genetic material that enters liver cells and blocks them from making apolipoprotein(a), so less Lp(a) reaches the blood. The site has covered another drug of this type, eplontersen for ATTR cardiomyopathy, which also came out of Ionis.

These were well-treated patients. Their median LDL was already low and their average systolic blood pressure was about 125 mm Hg, according to Cleveland Clinic's summary. Nissen described their risk factors as managed "as well as in any cardiovascular outcome trial to date."

The main goal was the time to a first major cardiovascular event. The trial defined that as death from cardiovascular causes, a non-fatal heart attack, a non-fatal stroke or an urgent procedure to reopen a heart artery that needed a hospital stay.

What the results showed

The drug did its job on the blood test. At 12 months, median Lp(a) was 64.4 nmol/L on pelacarsen and 233.4 nmol/L on placebo, a treatment difference of 77.7%. The effect lasted, with an average reduction of about 74% over the whole trial, the American College of Cardiology reported. LDL also fell by 13.3 percentage points more than on placebo, according to Cleveland Clinic.

It made no difference to events. A primary event happened to 511 people on pelacarsen (12.3%) and 503 on placebo (12.1%). The hazard ratio was 1.02, with a 95% confidence interval of 0.90 to 1.15. A hazard ratio of 1 means no difference at all.

The trial had a second main test in people with the highest levels, 90 mg/dL (about 192 nmol/L) or more. These made up around 78.6% of each group. The answer was the same: 402 events on pelacarsen and 403 on placebo, 12.3% in both groups, a hazard ratio of 1.00.

Here is the same result in plain numbers. Picture 1,000 people like those in the trial. Over almost five years, about 123 on pelacarsen and about 121 on placebo would have had a heart attack, stroke, urgent artery procedure or heart death. The 2 extra events on the drug sit well inside the play of chance. Our explainer on absolute and relative risk shows how to do this sum for any trial.

"I cannot avoid the disappointing conclusion, it didn't work," Nissen said in the AHA news release. He told Consult QD the result was not "a near miss."

Could the analysis have missed a benefit?

The team looked hard for one. The NEJM Evidence paper sliced the data in several ways, and Healio reported the main hazard ratios.

None of these comes close to a benefit. People whose Lp(a) fell the most did no better than those on placebo. Subgroups split by starting Lp(a), LDL, inflammation (hsCRP) and time since the first heart event also showed no difference. Death from any cause was also similar between groups.

"No matter how we looked at the data, the result was the same, no effect," Nissen told Consult QD.

Side effects and why so many stopped

Serious adverse events were slightly less common on pelacarsen (37.3%) than on placebo (41.1%). Drug-related side effects were far more common, at 56.8% against 18.0%. These included flu-like symptoms, injection site reactions and allergic reactions.

Investigators reported hypersensitivity reactions in 17.1% on pelacarsen and 5.4% on placebo. Nineteen people on pelacarsen had anaphylaxis, the most severe kind of allergic reaction. No one on placebo did, and no deaths came from these reactions, according to the AHA.

Side effects led 20.6% of people on pelacarsen to stop the drug, against 7.5% on placebo. In total, 35.1% on pelacarsen stopped treatment early for any reason, against 18.3% on placebo. Investment analysts at William Blair called this "one major caveat," BioSpace reported. The on-treatment and uninterrupted-treatment analyses in the table above were designed to test it, and they were also neutral.

Why did lowering Lp(a) not help?

Nobody knows yet. The investigators have offered a few explanations, and each one is still a hypothesis.

The first is timing. People with naturally low Lp(a) have had it from birth. "Here we intervened and lowered patients' Lp(a) to a more normal level for a median of 4.7 years, but that's not a lifetime," Nissen told Consult QD. Decades of exposure may do damage that a few years of treatment can't undo.

The second is that the patients were already very well treated. With low LDL and controlled blood pressure, there may have been less left for an Lp(a) drug to fix. The event rate was lower than expected, and the trial had to run longer to collect enough events.

A third is who took part. Everyone had existing heart disease. The trial says nothing about preventing a first heart attack in people with high Lp(a). The AHA also flags that 78% of participants were white and less than 4% had African ancestry, even though average Lp(a) is about twice as high in people of African ancestry.

Leslie Cho, the trial's first author and principal investigator at Cleveland Clinic, said Lp(a) "is a causal risk factor" regardless of these findings, and that the trial shows how important lowering LDL cholesterol still is.

There is also a plainer reading. Lowering Lp(a) in adults who already have heart disease, on top of good standard care, may not cut events at all. A failed trial can still be informative, as our guide on what it means when a clinical trial fails explains.

What happens to the other Lp(a) drugs?

Several Lp(a) drugs are still in large outcome trials. Nissen named the injected small interfering RNA (siRNA) drugs olpasiran, from Amgen, and lepodisiran, from Eli Lilly, and the oral pill muvalaplin, also from Lilly. The siRNA drugs lower Lp(a) further than pelacarsen does. According to Nissen, their outcome trials also include people without a previous heart event.

The olpasiran trial, OCEAN(a)-Outcomes, is registered on ClinicalTrials.gov. BioSpace reports that topline results are expected in early 2028, with first data from Lilly's ACCLAIM-Lp(a) trial of lepodisiran expected in March 2029. Analysts at Leerink Partners cut their estimate of olpasiran's chance of success from 50% to 20% after seeing the full HORIZON data, the same report said.

Nissen told Consult QD that a single negative trial does not settle whether lowering high Lp(a) is useful. The site covered a similar story for a different blood fat, triglycerides, where the injection plozasiran cut pancreatitis attacks.

What this means if you have high Lp(a)

Pelacarsen is not approved anywhere, and no medicine is approved specifically to lower Lp(a). Nothing in this trial changes what doctors can prescribe today.

What still lowers risk is managing the things that can be treated: LDL cholesterol, blood pressure, blood sugar and smoking. The AHA and the trial's leaders at Cleveland Clinic said knowing your Lp(a) level is still worthwhile, because it can change how hard your doctor pushes on those other risk factors.

If you are taking part in a different Lp(a) trial, Nissen's advice in the AHA release was direct: "don't drop out because of these findings." Talk to the trial team if you have questions. Our guide on how to find and join a clinical trial explains how participants can ask about a study's progress.

Questions worth taking to your next appointment:

• Has my Lp(a) ever been measured, and what was the number?
• Given my Lp(a), is my LDL cholesterol low enough?
• Am I eligible for any Lp(a) trial that is still recruiting near me?

This article is for information only and is not medical advice. Do not start, stop or change any medicine without talking to your doctor.