Pelacarsen Cut Lp(a) by 80%. Hearts Did Not Care.
The Biomarker That Would Not Bend
Roughly one in five people on Earth carries elevated Lp(a), a waxy particle made of fat and protein. Unlike LDL cholesterol, Lp(a) levels are almost entirely dictated by genetics: diet cannot lower them, statins barely touch them, and they stay elevated from birth to death. For decades, cardiologists watched high-Lp(a) patients suffer heart attacks anyway and built a hypothesis: lowering the particle would prevent those events. Novartis and its partner Ionis Pharmaceuticals spent more than six years testing that hypothesis with an antisense drug called pelacarsen. On September 4, 2026, they announced the answer.
8,323 Patients. Six Years. No Benefit.
The Lp(a)HORIZON Phase 3 trial enrolled 8,323 patients with established cardiovascular disease and Lp(a) of at least 70 mg/dL. All were already on guideline-standard care: statins, blood pressure drugs, antiplatelet therapy. Mean LDL cholesterol across the population was around 66 mg/dL, extremely well controlled. Against that background, pelacarsen lowered Lp(a) levels substantially but did not significantly reduce the composite primary endpoint: cardiovascular death, non-fatal heart attack, non-fatal stroke, or urgent hospitalised revascularisation. The trial had been designed to detect a hazard ratio of approximately 0.80 on that endpoint, at more than 90% statistical power, once 993 events had accumulated. The events accumulated. The drug did not reach the threshold.
Patients enrolled in Lp(a)HORIZON
8,323
Duration of the outcomes trial
over 6 years
Minimum Lp(a) for enrolment
70 mg/dL
Lp(a) reduction achieved by pelacarsen
approximately 80%
Lp(a) reduction in Phase 2 at highest dose
up to 80%
Global population with elevated Lp(a)
roughly 20%
NVS (Novartis ADR) drop in pre-market, Sep 5
12.7% to $139.56
IONS (Ionis) drop in pre-market, Sep 5
13.1% to $50.51
AMGN (Amgen) after-hours drop, Sep 4
approximately 5%
"These are not the results we hoped for, but they provide important evidence that advances scientific understanding.
"Shreeram Aradhye · CMO at Novartis
Why It Failed: Three Theories
The most charitable reading is background noise. With LDL at 66 mg/dL, residual risk may have been too low for a Lp(a) effect to register. A drug reducing events by 10% in a high-risk cohort may produce no signal in one this well-treated. A second theory targets the drug. Pelacarsen cut Lp(a) by around 80%. The competing siRNA drugs achieve above 90 to 95%. If there is a clinical threshold, pelacarsen may have stayed above it. The third theory is that Lp(a) is a bystander. It may travel with cardiovascular risk rather than causing it, making it a reliable marker but a useless target. If true, the 95% reductions olpasiran and lepodisiran achieve will also fail. The full dataset, due at an upcoming cardiology congress, will be the first real evidence for or against.
80% Was Not Enough. Can 95% Do Better?
Amgen's olpasiran is a small interfering RNA (siRNA) that silences the same gene and achieves reductions above 95% at the highest dose. Its Phase 3 outcomes trial, OCEAN(a)-OUTCOMES, enrolled 7,297 patients and is designed to read out around March 2028. Dosing is every 12 weeks. Eli Lilly's lepodisiran is also an siRNA. In Phase 2, a single 400 mg dose reduced Lp(a) by 93.9% at six months, with that reduction persisting above 90% a full year after a single injection. Its Phase 3 trial, ACCLAIM-Lp(a), plans to enrol 16,700 patients (double the pelacarsen cohort) with completion expected in 2029. Silence Therapeutics is running a fourth candidate, zerlasiran, through Phase 2. Amgen and Lilly have both said they will not alter their trial designs on the basis of the pelacarsen topline results. They are waiting for the full dataset.
A Third Big Pharma Failure in Eight Weeks
Novartis is the third large pharmaceutical company to report a major cardiovascular outcomes trial failure in two months, following similar setbacks at Novo Nordisk and AstraZeneca. For Ionis, it is the second cardiovascular Phase 3 failure of 2026: eplontersen missed its endpoint in July. Ionis shares lost approximately $877 million in market value in a single session and were trading near a 52-week low. The broader question is now the most debated in cardiovascular medicine: is Lp(a) a genuine driver of heart events, or merely a marker? LDL cholesterol was once similarly contested. The statins proved the hypothesis by delivering hard outcomes. Lp(a) still awaits its statin moment. Whether olpasiran or lepodisiran can provide it, from a deeper reduction baseline than pelacarsen managed, will be known by 2028 at the earliest.
