News

Heart health

A monthly shot cut Lp(a) by 78 percent. Heart attacks and strokes did not move

By Matthew Spencer

Published · Updated

A row of sealed blood sample tubes in a laboratory rack
Illustrative stock photo.

The short answer: in the first big outcomes trial of a drug built to lower Lp(a), the drug worked exactly as designed on the blood test, cutting Lp(a) by about 78 percent. Over nearly five years, heart attacks, strokes, urgent artery-opening procedures and cardiovascular deaths happened at the same rate as placebo.

Lp(a), or lipoprotein(a), is a cholesterol particle similar to LDL, the usual bad cholesterol. Levels are largely set by your genes and stay stable through life. About one in five people worldwide has an elevated reading.

That matters because high Lp(a) tracks with more heart attacks and strokes in population studies, and genetic studies point to a causal role. The catch: statins and lifestyle changes barely move it, and no approved treatment lowers it.

The trial that was supposed to prove the point

Lp(a) HORIZON enrolled 8,323 adults at 797 sites in 42 countries. Everyone had established cardiovascular disease and an Lp(a) of at least 70 mg/dL, roughly the highest fifth of the population. The median level was 236 nmol/L.

Participants injected pelacarsen 80 mg, or a matching placebo, under the skin once a month. Pelacarsen is an antisense drug: a short synthetic genetic strand that tells liver cells to stop producing Lp(a).

The primary question was time to a major cardiovascular event: nonfatal heart attack, nonfatal stroke, urgent revascularization, or death from a cardiovascular cause. Median follow-up ran 4.7 years.

The marker fell. The events did not

Lp(a) dropped about 78 percent on treatment. But a primary event occurred in 12.3 percent of the pelacarsen group and 12.1 percent on placebo, a hazard ratio of about 1.02 - statistically indistinguishable from no effect.

The pattern held everywhere the researchers looked. Patients with the very highest Lp(a), 90 mg/dL or more, saw identical 12.3 percent event rates in both arms. Extended analyses of secondary endpoints, subgroups and on-treatment data were equally flat.

Senior author Steven Nissen of Cleveland Clinic did not soften it: the results were completely neutral across multiple analyses, with basically no effect on outcomes, not even a trend.

Safety was not the story, but it was not clean

Serious adverse events were actually less frequent on pelacarsen than placebo, 37.3 versus 41.1 percent. But injection-site reactions, flu-like symptoms and allergic reactions were more common, and 19 patients on pelacarsen had anaphylaxis versus none on placebo. None of those reactions were fatal.

About 21 percent of patients on pelacarsen stopped the drug because of adverse reactions, compared with just over 7 percent on placebo.

Why the result surprised everyone

The bet behind the trial was strong. Decades of observational data linked high Lp(a) to heart disease, and Mendelian randomization studies, which use genetic variants as a natural experiment, pointed the same way. Lowering the particle was expected to lower the risk.

One honest reading is that a surrogate marker is not a guaranteed outcome. Another is timing: Lp(a) may do its damage over decades, and five years of lowering in people who already had heart disease may be too late, or too short, to show a difference.

Who the result applies to

This was one drug, in one population: people who had already had a cardiovascular event. It says nothing about prevention in healthier people, or about other drug classes. Several siRNA drugs that may cut Lp(a) further are in outcomes trials now, including olpasiran, lepodisiran and muvalaplin.

Nissen's own caution cuts both ways: a single neutral trial does not prove lowering Lp(a) is pointless, any more than the genetic data proved it would work. The concept is bruised, not buried.

What this means for you

If you have never had your Lp(a) measured, the trial changes nothing about the standard advice: major guidelines recommend checking it once in adulthood, because it is genetic, common and invisible without a blood test.

A high result still signals higher average risk, and it is still information your other numbers, LDL, blood pressure and blood sugar, do not give you. What this trial removes is the assumption that a pill to fix the number is just around the corner. For now, the levers that reliably move outcomes remain the unglamorous ones.

Sources (5)

This is news reporting and not medical advice.