// Contents
78%. That is how far, on average, a monthly shot of pelacarsen pushed down Lp(a), a particle in the blood. Yet heart attacks, strokes and cardiac deaths were just as common as on placebo: 12.3% versus 12.1%. The drug hit its target dead center, and the heart never noticed.
Researchers tested it on more than 8,300 adults in 42 countries. Every one of them had already survived a heart attack, a stroke or narrowed leg arteries, and the team followed them for 4.7 years, counting heart attacks, strokes, emergency heart procedures and cardiovascular deaths. The team presented the results on November 7, 2026, at the American Heart Association meeting in Chicago and published them the same day in the New England Journal of Medicine. Steven Nissen, the Cleveland Clinic cardiologist who led the trial, told NPR: “I was shocked. I had high hopes for success.”
The trial, called Lp(a)HORIZON, was the first large test of a simple idea: lower Lp(a) and you get fewer heart attacks. The answer matters to one person in five on Earth, and most of them have never had their level checked.
An inheritance you can’t diet away
Lp(a), short for lipoprotein(a), is a cousin of LDL, the “bad” cholesterol. It ferries fat through the bloodstream, and in large amounts it helps plaque build up in artery walls. Your genes set the level almost entirely, and it barely moves from birth to old age. High Lp(a) causes no symptoms. None.
According to the American Heart Association, about one person in five worldwide has elevated levels, roughly 1.4 billion people. At 125 nmol/L or above, the long-term risk of heart attack or stroke is about 1.4 times higher; at 250 nmol/L, at least double. Healthy eating and exercise hardly shift the number, and neither do statins. There is still no approved drug that targets Lp(a). People with severe inherited cholesterol disorders sometimes have their blood filtered on a dialysis-like machine, but that option reaches very few.
New US cholesterol guidelines, published in 2026, tell every adult to get an Lp(a) test at least once. Pelacarsen was supposed to be the next step: find a high level, prescribe a drug.

Why almost everyone expected it to work
For decades, observational studies had linked Lp(a) to heart attacks. On its own, a link like that proves little: people with high levels might differ in some other way. The genetic evidence looked far stronger.
Geneticists turned to Mendelian randomization, a kind of experiment that nature has already run. At conception, some people happen to inherit gene variants for low Lp(a) and others get variants for high Lp(a). Each group then lives with its level for life. The low group has fewer heart attacks, so the particle itself is probably to blame, not lifestyle.
Studies like these suggested that Lp(a) does damage by itself. Donald Lloyd-Jones, a cardiologist and past president of the American Heart Association, told a media briefing that Lp(a) meets every test for a cause of atherosclerosis, the plaque buildup in arteries, except one: nobody had yet lowered the particle and shown that people did better. Lp(a)HORIZON was designed to supply that proof.
Ionis discovered the drug, and Novartis licensed it in 2019. Novartis announced the failure early, on September 4, 2026, and the market noticed: both companies’ shares fell that same evening.
What a clean failure looks like
Pelacarsen works on the genetic instructions themselves. The liver builds the particle from a set of instructions, and the drug intercepts them on their way to the assembly line, so the part never gets made. Half the participants injected themselves at home with 80 mg once a month; the other half injected a placebo. Nobody knew who got which. Not the patients, not the doctors.
On blood tests, everything went to plan. After a year, Lp(a) had fallen by an average of 78% in people on the drug, compared with 3% on placebo. LDL cholesterol fell by 13% along the way. Judged on lab results alone, the trial would have been a triumph.
It wasn’t. It was judged on heart attacks. Nissen stressed that this was not a near miss: however you slice the data, the answer is the same: no effect. Among participants with the highest starting levels, 192 nmol/L or more, event rates in the two groups were identical. Deaths from any cause matched too.
// Lab results vs. heart attacks
Karol Watson, a cardiologist at the University of California, Los Angeles, reviewed the trial at the meeting and summed it up: compelling biology, compelling genetics, everything went right, and yet cardiovascular events didn’t change.
Side effects and dropouts
Adverse reactions that doctors linked to the study drug hit 56.8% of the pelacarsen group and 18% of the placebo group. The most common were flu-like symptoms (31.9% vs. 5.8%) and injection-site reactions (25.6% vs. 2.9%). Nineteen people on the drug had anaphylaxis, compared with none on placebo, and nobody died of an allergic reaction. After a suspected case of anaphylaxis in March 2023, injections were given only under medical supervision for six months. Side effects made 20.6% of the drug group quit, versus 7.5% on placebo; counting every reason, 35.1% versus 18.3% stopped early. Serious adverse events, oddly, were less common on the drug: 37.3% vs. 41.1%. The authors add one more caveat: some events happened during the pandemic, and there is no way to tell which ones were linked to COVID-19.
Five theories of why the particle went and the risk stayed
Too late. The favorite theory. Genetically low Lp(a) protects a person from day one, while trial participants started pelacarsen at around 60 and took it for less than five years. “It may be that a person needs to have lifelong Lp(a) reduction to be protected,” Nissen said. The drug turned off the tap after the basement had already flooded. Lloyd-Jones put it as a question: are we too far down the pathway to be able to fully reverse it?

Treated too well. Participants had a typical LDL of about 65 mg/dL and an average systolic blood pressure of 125 mm Hg. Nissen said these were among the lowest numbers of any trial of its kind. When every other risk factor is already pinned down, there is little room left for a new drug to help.
A weak drug. Pelacarsen falls short of the next generation: olpasiran cut Lp(a) by more than 90% in early trials. And a third of the people on pelacarsen quit their injections early. Lloyd-Jones wondered aloud whether side effects ate into the benefit and called it an open question.
The wrong people. David Maron, a cardiologist at Stanford, thinks the trial should have set a higher Lp(a) bar for entry. Fewer than 4% of participants were of African ancestry, a group whose average level is about twice as high. Then again, the participants with the highest levels showed no benefit either.
The wrong target. The most uncomfortable option: Lp(a) may not be a cause at all, just a marker for something else that raises risk. Doctors interviewed by NPR named it among the possibilities, although the genetic evidence points the other way. Only the next drugs can settle it.
Three drugs waiting in line
Novartis has shut down its entire pelacarsen program for atherosclerosis, including ongoing studies and its early-access program. The idea survived. Three other drugs are waiting in line. Amgen’s olpasiran and Lilly’s lepodisiran both lower the particle more sharply than pelacarsen. The third, Lilly’s muvalaplin, is a pill instead of a shot. Some of the new trials also enroll people who have never had a heart attack or stroke; one olpasiran trial is devoted entirely to preventing a first one. That will partly test the “too late” theory.
Nissen is urging volunteers in those trials not to drop out because of his result. “A single trial does not provide definitive evidence of a lack of benefit of lowering lipoprotein(a),” he said. “We have to be humble about what our trial did show, what it didn’t show.” Measuring Lp(a), he says, is still worth it.
Advice for people with high Lp(a) is thin for now. Leslie Cho, a Cleveland Clinic cardiologist and first author of the NEJM paper, points out that Lp(a) is still a causal risk factor and that lowering LDL remains crucial. There is no drug for the particle itself, so for now the test reveals a risk, not a treatment plan.
In the genetic studies, protection went to people whose Lp(a) had been low since birth. If that is really the key, how early would treatment have to start for a shot to do what genes do over a lifetime?







