Pelacarsen drove lipoprotein(a) down sharply, but events did not budge
The Lp(a) HORIZON trial delivered a result that will reverberate through lipid clinics for years: pelacarsen lowered lipoprotein(a) by a mean of 78%, yet the reduction did not translate into fewer major cardiovascular events. In a 7-year, phase III, randomized study of more than 8,300 adults with established cardiovascular disease and elevated Lp(a), the time to first nonfatal heart attack, stroke, urgent revascularization, or death from cardiovascular causes remained unchanged between the active treatment and placebo groups.
This was not a small pharmacodynamic signal buried under uncertainty. The drug worked on its intended target. What failed was the clinical endpoint.
A biologically precise target met a clinically neutral outcome
Lp(a) is not ordinary cholesterol. It is a genetically driven lipoprotein particle with a structure that overlaps with LDL cholesterol and a risk profile that has long pointed toward arterial plaque formation, inflammation, and thrombosis. In theory, lowering it should reduce the burden of atherosclerotic disease, especially in people who already carry established cardiovascular injury. Pelacarsen was designed to silence hepatic production of lipoprotein(a) by using a synthetic genetic fragment that enters liver cells and blocks the pathway responsible for its creation.
That mechanism produced a large biomarker reduction. Mean Lp(a) fell by 78% across treatment groups, including the subgroup with baseline levels at or above 90 mg/dL, a range that captures the highest-risk tail of the population. Yet even in that enriched subgroup, cardiovascular outcomes remained flat. The mismatch between biology and bedside outcome was consistent across the prespecified analyses and across extended outcome combinations.
The cohort was large, global, and already aggressively treated
Investigators enrolled 8,323 adults at 797 sites across 42 countries after screening 14,249 people. The population had substantial baseline risk: mean age 60 years, 27% women, and 26% with diabetes, almost all type 2. Most had already lived through major atherosclerotic disease. Eighty-two percent had a prior heart attack, 10% had a previous clot-related stroke, 14% had symptomatic peripheral artery disease, and 18% had multiple cardiovascular events. Early vascular disease was common as well, with heart attack or stroke before age 55 in men or before age 65 in women documented in 37%.
Median Lp(a) at enrollment was 109 mg/dL, and all participants met the trial’s threshold of at least 70 mg/dL. About three-quarters were already at guideline targets for blood pressure and cholesterol and did not require additional treatment at baseline. Nearly all were taking antiplatelet therapy, and 77% to 78% were on high-intensity statins. In other words, this was not a naïve population in which one new therapy could easily reveal a dramatic effect. These were already-treated patients with residual risk, the exact setting where an Lp(a)-lowering intervention would be expected to prove its worth if the pathway were strongly causal for near-term events.
Safety was acceptable overall, but tolerability was not trivial
The adverse-event profile never disappeared into the background. Serious adverse events occurred in 37.3% of pelacarsen-treated participants and 41.1% of those receiving placebo, so the broad serious-event rate was similar. The more relevant signal was tolerability: flu-like symptoms, injection-site reactions, and acute allergic reactions were more common with pelacarsen. Anaphylaxis occurred in 19 patients receiving the drug and in none assigned to placebo, although no deaths resulted from these allergic reactions.
Discontinuation followed that pattern. About 21% of participants in the pelacarsen group stopped study medication because of adverse reactions, compared with more than 7% in the placebo group. For a preventive therapy that would likely be used over years in asymptomatic patients identified through screening, this matters at the bedside. A lipid drug that lowers a laboratory value but generates substantial attrition cannot be judged only by pharmacology; adherence and tolerability become part of its real-world efficacy.
What the neutral result means for the biology of Lp(a)
The result does not erase the importance of Lp(a) as a cardiovascular risk factor. High levels remain associated with greater long-term risk of heart attack and stroke, and the burden is largely inherited rather than behaviorally acquired. Lifestyle modification has little impact on Lp(a), and statins do not meaningfully lower it. That leaves clinicians with a difficult truth: a common, genetically anchored driver of risk is measurable, but not yet fully actionable with an approved Lp(a)-lowering medicine.
The neutral outcome also raises a sharper mechanistic question. Lp(a) may contribute to atherothrombosis in a way that is more complex than simple concentration reduction can capture, or the trial may have selected a group in whom other residual risks dominated. Timing from the qualifying vascular event, background treatment intensity, and the biology of already stabilized plaques may all have blunted the chance to see an event difference over the study horizon.
Screening still has a role, even after this setback
Measuring Lp(a) remains clinically worthwhile. High levels are usually silent, so patients do not volunteer symptoms that would prompt testing. Lp(a), blood pressure, LDL cholesterol, and blood sugar together shape cardiovascular prevention strategy, and the 2026 multi-society lipid guidance recommends measuring Lp(a) at least once in adulthood. Roughly 1 in 5 people globally are thought to have elevated levels, which makes this a common risk modifier rather than an exotic one.
The trial also leaves open a broader point about prevention. Patients with very high Lp(a) may still need intensive control of every modifiable risk factor, because the pathway remains clinically relevant even if this specific drug did not improve outcomes. For people already enrolled in ongoing Lp(a) studies, dropping out now would only weaken the evidence base. Three other cardiovascular outcome trials of different investigational agents are still underway, and those results will help determine whether the disappointment seen here reflects a class-wide limitation, a drug-specific issue, or a mismatch between target suppression and outcome biology.
The gaps are now part of the story
Representation was uneven, with 78% of participants self-identifying as white and fewer than 4% reporting African ancestry, despite the fact that some groups carry substantially higher average Lp(a) levels. The trial also spanned the COVID-19 era, and some cardiovascular events may have been influenced by pandemic-era care disruptions or infection-related cardiovascular injury that could not be separated from the underlying atherosclerotic signal. Those issues do not negate the result, but they do define the boundaries of what this study can and cannot prove.
Pelacarsen did exactly what it was built to do in the bloodstream. It just did not improve the outcomes that matter most to patients and clinicians.