Pelacarsen: A Decade-Long Bet on a Stubborn Molecule

Lipoprotein a has been taught to generations of medical student as major risk for heart disease. There were families with high Lp(a) and generations of premature cardiac deaths. But medical science must appraise long held truths. As most of you know, I supervise clinical trials on vaccines. Cardiac trials are much more complicated. Read below:

Picture a drug that does exactly what scientists designed it to do. It hits its target. The numbers on the lab report improve. Everyone expects good news.

Then the good news never comes.

That is what happened to pelacarsen, a drug from Novartis built to fight heart disease. On September 4, 2026, the company announced that its huge, years-long study of the drug had failed. Not because the drug didn’t work the way it was supposed to. It failed because doing the one thing it promised, lowering a fatty particle in the blood, didn’t actually protect anyone’s heart.

The particle is called lipoprotein(a), or Lp(a) for short. Doctors pronounce it “L-P-little-a.”

About one in five people on Earth carry high levels of it. It runs in families. You inherit your level the way you inherit eye color, and no amount of salad or jogging will change it. Statins, the cholesterol pills millions of people take every day, barely touch it either.

High Lp(a) raises your risk of heart attack and stroke. Doctors have known this for years. What they didn’t know is whether lowering it would actually help. Nobody had ever tested that question in a large, careful study. Pelacarsen was supposed to give the answer.

The story starts with a company called Ionis Pharmaceuticals, which builds drugs that block the body from making harmful proteins in the first place. Ionis tested an early version of pelacarsen on humans for the first time in 2015.

Novartis liked what it saw. In 2017 the two companies started working together. In February 2019, Novartis paid $150 million up front to take full control of the drug, with the chance to pay Ionis up to $675 million more if things went well.

Things did not go well.

From that 2019 deal to this month’s disappointing results, seven years passed. Count the earlier work at Ionis, and the whole project stretched past a decade. Heart drugs take this long because researchers cannot rush the wait for heart attacks and strokes to happen, or not happen, in thousands of real patients. That waiting is the whole point of the study. There is no shortcut.

The study had a name built for drama: Lp(a)HORIZON. Here is what it looked like on the ground.

Doctors screened more than 14,000 people. They enrolled 8,323 of them, at 797 hospitals and clinics, spread across 42 countries and six continents. Every one of those patients had already survived a heart attack, a stroke, or serious artery disease, and every one had high Lp(a).

Half the patients got a monthly shot of pelacarsen. Half got a placebo shot that did nothing. Neither the patients nor their doctors knew which was which, a safeguard that keeps hope and bias from creeping into the results.

Patients were followed for years. It had lowered Lp(a), just as expected. But when the researchers compared the two groups, patients on the drug were not having fewer heart attacks or strokes than patients on the placebo.

“Although lower Lp(a) levels were observed with pelacarsen, the findings did not demonstrate that this translated into reduced cardiovascular risk in the overall study population,” said Dr. Shreeram Aradhye, Novartis’s chief medical officer, in the company’s announcement.

In plain terms: the drug did its job in the bloodstream. It just didn’t save lives.

Novartis has not published a final price tag for Lp(a)HORIZON. Trials of this size and length usually cost more than a billion dollars. Add the years Ionis spent inventing the drug before Novartis ever got involved.  and the true cost of finding out that pelacarsen doesn’t work is almost ethereal.

It would be easy to read this as a story about a drug that flopped. It is really a story about how much modern medicine still doesn’t know.

This was the first time anyone had ever tested, in a large and rigorous way, whether lowering Lp(a) protects the heart. The assumption seemed reasonable. High Lp(a) tracks with heart disease, so lowering it should help, the same way lowering LDL cholesterol helps. That assumption just took a serious hit.

Novartis plans to present the full results at a major heart conference later this year, likely the American Heart Association’s meeting in November. Researchers want to know whether the drug simply didn’t lower Lp(a) enough, or whether certain patients benefited even if the whole group didn’t, or whether something else entirely explains the result.

Other companies are racing down the same road. Amgen has its own Lp(a) drug in testing. So does Eli Lilly. Both are watching this result closely, and both now face a harder question than they did last week: what if the whole idea, chasing this molecule at all, needs a second look?

Nobody knows yet. That uncertainty, more than the failed trial itself, may be the real headline here. (written with assist from Claude AI Sonnet)