CHICAGO, IL—The Lp(a)HORIZON investigators combed through the data, looked at multiple endpoints in a myriad of ways, but there was absolutely no benefit to adding the antisense oligonucleotide (ASO) pelacarsen to guideline-directed medical therapy (GDMT) in patients with atherosclerotic cardiovascular disease (ASCVD).
In fact, event curves for the primary and nearly every secondary endpoint were virtually superimposable, reported researchers today at the American Heart Association (AHA) Scientific Sessions.
“I’ve never seen so many overlapping Kaplan-Meier curves in my career,” lead Lp(a)HORIZON investigator Steven Nissen, MD (Cleveland Clinic, OH), said during a press briefing. “It’s a lot of neutral results.”
Donald Lloyd-Jones, MD (Boston University School of Medicine, MA), past-president of the AHA who was not involved in the trial, said these new data show researchers that “there is no benefit from this approach, this strategy, this compound.”
The topline results of Lp(a)HORIZON have been known since September when drug manufacturer Novartis announced the trial missed its primary endpoint. The AHA presentation, along with the simultaneous publication in the New England Journal of Medicine, marks the first detailed look at the outcomes.
“Although these results are disappointing, they provide scientific insights that help inform future research efforts,” said Nissen. “The Lp(a)HORIZON trial is not the end of the story.”
Karol Watson, MD, PhD (University of California Los Angeles), the discussant following the presentation, said that when the topline Lp(a)HORIZON results were announced, “you could almost hear a collective gasp around the world.” The trial ultimately showed there was a disconnect between what could be reduced and what could be prevented.
“Even with compelling biology, compelling genetics, everything went right but cardiovascular events didn’t change,” she said. “We still have so much to learn about Lp(a).”
I’ve never seen so many overlapping Kaplan-Meier curves in my career.
Steven Nissen, MD
Despite the shock and disappointment of Lp(a)HORIZON, there are several ongoing studies testing Lp(a)-lowering therapies, including the secondary-prevention OCEAN(a) Outcomes trial and primary-prevention OCEAN(a) PreEvent study with the small-interfering RNA (siRNA) olpasiran (Amgen). The ACCLAIM-Lp(a) investigators are testing lepodisiran (Eli Lilly), another siRNA, in a primary and secondary prevention population. Finally, the MOVE-Lp(a) researchers are evaluating muvalaplin (Eli Lilly), a small oral molecule that blocks Lp(a) production in the liver, in patients with a prior ASCVD event or those at high risk for a first.
Olpasiran, lepodisiran, and muvalaplin have all demonstrated an ability to significantly slash Lp(a) levels, although the true test will be the ongoing outcome studies.
“When you get a really neutral trial, as we got, it doesn’t raise expectations for the other trials,” said Nissen. “But science, it’s a zigzag pathway. Sometimes we zig before we zag, and there have been pretty good examples where the first drug in a class failed for a variety of reasons…I really think you can’t close the door with a single negative trial.”
Deepak Bhatt, MD (Icahn School of Medicine at Mount Sinai, New York, NY), who wasn’t involved in Lp(a)HORIZON, told Cardiology Now News that the easy conclusion would be to say the Lp(a)-lowering hypothesis is finished, but that would be premature.
“I don’t think we should ever generalize to say a class is dead because one particular drug is,” said Bhatt. “Even though this trial looks pretty negative—I don’t really see much supporting any benefit of this particular compound—but I think it’s too soon to say as a whole that Lp(a)-lowering therapeutics are dead.”
Pelacarsen is a second-generation ASO targeting mRNA transcribed by LPA, the gene that codes for apolipoprotein(a) in the liver. Apolipoprotein(a) is needed to produce Lp(a) so blocking its production leads to reductions in circulating Lp(a) concentrations.
High Lp(a) is a known, genetically determined risk factor for ASCVD and calcific aortic valve disease. Observational and Mendelian randomization studies have consistently shown that high Lp(a) levels are linked with ASCVD risk and many international guidelines, including the most recent American College of Cardiology (ACC)/AHA dyslipidemia guidelines, recommend measuring levels at least once in a person’s lifetime. High Lp(a) levels are an indication for more intensive LDL cholesterol lowering and risk-factor management (class 1 indication).
Lp(a)HORIZON: the Details
With Lp(a)HORIZON, investigators randomized 8,350 patients (mean age 59.7 years; 27.0% female) with preexisting cardiovascular disease and elevated Lp(a) levels, defined as 70 mg/dL or higher (approximately 148 nmol/L), to pelacarsen 80 mg administered once monthly or matching placebo.
The study protocol included a 30- to 90-day period that allowed physicians to optimize patients on GDMT for the secondary prevention of ASCVD. Overall, 50.3% of randomized patients had a premature MI or ischemic stroke (before age 55 in men and 65 years in women) or peripheral artery disease before age 50.
Patients in both arms were well treated on GDMT, with 77.6% of patients taking a high-intensity statin. In all, 56.8% were taking ezetimibe or a bile acid sequestrant, 11.1% were taking a PCSK9 inhibitor, and 81.0% were taking aspirin. The median baseline LDL at randomization was 64.6 mg/dL and mean systolic blood pressure was 125 mmHg. The median Lp(a) level was 236 nmol/L.
Pelacarsen cut Lp(a) levels by 78.4% at 12 months compared with a 3.2% reduction in the placebo-treated patients.
The primary endpoint, a 4-point MACE composite of cardiovascular death, nonfatal MI, nonfatal stroke, or urgent coronary revascularization, occurred in 12.3% of the pelacarsen-treated patients and 12.1% of the placebo-treated group (HR 1.02; 95% CI 0.90-1.15) over a median follow-up of 4.7 years. There was no effect on any of the individual components of the primary endpoint.
I think it’s too soon to say as a whole that Lp(a)-lowering
Deepak Bhatt, MD
therapeutics are dead.
Even in patients with Lp(a) levels 90 mg/dL or higher at baseline (approximately 192 nmol/L), a prespecified subgroup, treatment with pelacarsen did not reduce the risk of the primary endpoint compared with placebo: 12.3% in both treatment arms (HR 1.00; 95% CI 0.87-1.15).
There was no significant difference between the pelacarsen- and placebo-treated patients with respect to the 3-point MACE endpoint (cardiovascular death, nonfatal MI, or nonfatal stroke) nor was there any difference when the 3-point MACE endpoint used coronary heart disease mortality instead. Again, pelacarsen did not reduce either of the 3-point MACE endpoints in those with Lp(a) levels 90 mg/dL or higher at baseline.
In a prespecified on-treatment analysis and uninterrupted-treatment analysis, the latter including patients who completed the trial without stopping treatment for more than 60 days, the results were unchanged compared with the intention-to-treat analysis. Even in patients with the largest reductions in Lp(a) from baseline, there was no favorable effect with treatment.
Adverse events leading to drug discontinuation occurred in 20.6% of the pelacarsen-treated patients and 7.5% of those randomized to placebo. The most common adverse events were constitutional symptoms/influenza-like reactions and investigator-reported hypersensitivity reactions.
What Went Wrong?
During the late-breaking science session, the experts said there is extensive evidence that Lp(a) is a causal risk factor, and not just a marker of risk, for ASCVD. One question that emerged was whether the siRNA therapies, olpasiran and lepodisiran, might be more effective than pelacarsen. The siRNA therapies are better at lowering Lp(a) and have less toxicity.
“The siRNAs that are being tested are more potent and from what we’ve seen of them so far, they tend to be very well tolerated,” said Bhatt.
He also pointed to the past as grounds for some optimism. The GLP-1 receptor agonists are all quite good at lowering glucose levels, but only some agents reduce the risk of cardiovascular events. Omega-3 fatty acids also delivered mixed results in different trials. Icosapent ethyl in REDUCE-IT reduced the risk of ASCVD events in patients with hypertriglyceridemia, and more recently in the PISCES study, but a high-dose fatty acid formulation in STRENGTH had no impact on clinical outcomes.
Another hypothesis is that patients in Lp(a)HORIZON might have been treated so well that ASCVD risk was attenuated. It’s possible, said Watson, that aggressive LDL cholesterol lowering mitigated some of the risk associated with Lp(a). It’s also possible that Lp(a) lowering might be more effective in a primary-prevention setting as opposed to secondary prevention.
To Cardiology Now News, Nissen said Lp(a)HORIZON showed that physicians did an excellent job with secondary prevention strategies. “We did everything we could for these patients,” he said. “We kept telling investigators, ‘Give them the best guideline-recommended therapy you can.’ And guess what? It worked and it’s a positive message.”
While patients were well treated, Lloyd-Jones noted that cardiovascular events continued to accrue over time.
“In no prevention trial, especially secondary prevention, are we getting anywhere close to zero events,” he told Cardiology Now News. “We must get better at implementing what we know to reduce event rates, but we also must innovate and understand new pathways where we can get close to zero.”
Don’t Abandon Lp(a)-Lowering Yet
Lloyd-Jones, who was involved in drafting the ACC/AHA dyslipidemia guidelines, stressed that they stand by their recommendation to measure Lp(a) at least once for ASCVD risk assessment. They did not make the recommendation because Lp(a) was a target for therapy, he said.
One could make an argument that measuring Lp(a) is not a good use of healthcare dollars, said Bhatt. Today, Lp(a) is not a modifiable risk factor and while it can be used to guide therapy, Bhatt said doctors should already be targeting LDL cholesterol and systolic blood pressure as well as pushing weight loss and smoking cessation. Despite the arguments against measuring it, Bhatt said he does check Lp(a) levels, noting it’s not a very expensive test and it can be used to motivate patients.
With the null results, one thing investigators and other experts worry about is whether the bad news will be contagious. Nissen expressed concern that negative headlines will lead patients to bail out from ongoing Lp(a)-lowering trials or for sponsors to get cold feet and yank funding.
“It’s going to be my mission to try to convince the people that pay the bills, which is the sponsors, not to pull the plug because we really need to understand more,” said Nissen. “Who benefits? Who doesn’t? Are there other populations that can be studied? There’s a lot more to be said and we’re going to have to wait and see.”
Source:
Cho L, Nicholls SJ, Frndak S, et al. Pelacarsen and cardiovascular outcomes in patients with high lipoprotein(a). N Engl J Med. Published online November 7, 2026.
Disclosures:
Nissen reports research funding from Madrigal Pharmaceuticals.
Lloyd-Jones and Watson report no conflicts of interest.