Novartis' Pelacarsen Misses Primary Endpoint in Lp(a)HORIZON Cardiovascular Outcomes Trial

Novartis (NYSE: NVS) said Friday that pelacarsen, its antisense therapy for elevated lipoprotein(a), missed the primary endpoint in the Phase 3 Lp(a)HORIZON cardiovascular outcomes trial. The drug did not significantly cut the risk of major cardiovascular events versus placebo, despite earlier data showing it lowers Lp(a) by up to 80%.

The numbers

Lp(a)HORIZON randomized 8,323 patients with elevated Lp(a) and established cardiovascular disease to pelacarsen or placebo:

• Primary endpoint (4-point MACE): missed, no significant reduction in cardiovascular death, non-fatal myocardial infarction, non-fatal stroke, or urgent coronary revascularization requiring hospitalization
• Prior biomarker data: up to 80% Lp(a) reduction versus placebo in a mid-stage study
• Specific data: not disclosed; Novartis said it will present full results at a future medical congress

Why it matters

Lp(a)HORIZON was the first Phase 3 cardiovascular outcomes trial to test whether lowering Lp(a) actually reduces cardiovascular risk in patients who are already on optimal care. That question has driven billions of dollars in R&D, and the answer from pelacarsen is no benefit on hard events.

The read-through is what makes this a sector story, not just a Novartis story. Amgen’s olpasiran (Phase 3 OCEAN(a), topline expected 2028) and Eli Lilly’s lepodisiran (Phase 3 ACCLAIM-Lp(a), data expected 2029) are both siRNA therapies pursuing the same Lp(a)-lowering hypothesis, and both have shown even deeper reductions than pelacarsen, with olpasiran topping 97% and lepodisiran nearly 94%. Merck also paid nearly $2 billion in biobucks last year for an oral Lp(a) inhibitor from Hengrui.

Analysts framed the miss as a setback, not a death blow, for the target. BMO called it “a meaningful setback for dedicated Lp(a)-lowering therapies” and said it “raises the burden of proof” for lepodisiran and olpasiran. Truist noted olpasiran’s deeper Lp(a) reduction and OCEAN(a)‘s enrollment of a higher-risk population as reasons the Amgen program could still succeed.

Pelacarsen is being advanced with Ionis Pharmaceuticals, which developed the antisense candidate. For background on the antisense versus RNAi platform debate, see our RNA therapeutics primer.

What to watch

Novartis will present the full Lp(a)HORIZON data at a future congress, and the Lp(a) thesis now rests on whether olpasiran and lepodisiran can show a cardiovascular benefit that pelacarsen could not. This was Novartis’ first of two Phase 3 failures this week; see our coverage of the del-desiran HARBOR miss. The topline details are on ClinicalTrials.gov, and Novartis’ statement is on its media releases page.

What It Means for the Trade

The first-order point: biomarker reduction is not clinical benefit, and Lp(a)HORIZON is now the canonical proof. An 80% Lp(a) reduction that fails to move MACE forces the field (and investors) to confront three possible explanations, each with different read-throughs: (1) Lp(a) is a marker, not a mechanism — causally less important than the Mendelian-randomization genetics suggested; (2) the reduction wasn’t deep or fast enough — 80% mean lowering still leaves substantial absolute Lp(a) in the highest-risk patients, whereas olpasiran’s 97%+ and lepodisiran’s ~94% approach near-complete suppression; (3) the population was too treated — established-CVD patients on maximal statin/PCSK9/antiplatelet therapy have a residual-event pool driven by other pathways. The siRNA programs’ sponsors will argue (2) and (3); the target’s skeptics will argue (1). Until OCEAN(a) reads out in 2028, neither side can be proven wrong.

For Novartis specifically, the loss is a pipeline-pillar event. Cardiovascular outcomes was one of the two big growth legs (with nuclear and radioligand therapy) of the post-Entresto story, and pelacarsen (with Ionis) was the lead asset in the cardiometabolic extension. Add the same-week del-desiran HAMBOR miss in myotonic dystrophy and Novartis has now burned two Phase 3 programs in a week — a pattern that shifts the 2026-2027 narrative onto remaining catalysts (remixes of Entresto, the RLT franchise, Kisqali’s expansion) and invites questions about pipeline prioritization discipline.

For the sector, mark the calendar effects: Amgen’s OCEAN(a) (2028 topline) and Lilly’s ACCLAIM-Lp(a) (2029) are now the field’s entire clinical verdict. Merck’s ~$2 billion oral-Lp(a) bet (Hengrui-derived) is even further out and suddenly looks like the most exposed of the follow-ons — it’s paying for a hypothesis that just got harder to prove.

What the Full Dataset Will Decide

The congress presentation matters more than most, because the failure’s texture determines the siRNA programs’ odds. Watch four things: the hazard ratio’s direction and confidence intervals (a numerically-favorable-but-nonsignificant HR near 0.9 keeps the “not deep enough” argument alive; an HR at or above 1.0 argues the target itself is weak); the baseline and achieved Lp(a) distribution by event status (if events clustered in patients whose Lp(a) didn’t fall far, dose-depth is the story); subgroup performance by baseline Lp(a) (genetic-predisposition arguments predict benefit concentrated in the very-highest baseline strata); and whether Lp(a)HORIZON’s monthly subcutaneous antisense dosing produced the adherence gaps that plague long inflection-free trials.

One technical note worth knowing: pelacarsen is an antisense oligonucleotide from Ionis, dosed monthly, with kidney-tissue exposure that required monitoring; the siRNA rivals are two-to-four-times-yearly injections with deeper and more durable knockdown. If depth-and-durability is the differentiator, the RNAi platform (Alnylam-licensed chemistry inside both Amgen’s and Lilly’s candidates) inherits the better hand — a subtle but real read-through to the antisense-versus-RNAi platform debate we track in our RNA therapeutics primer.

Risks and What Could Go Wrong

For the target-class bulls: if OCEAN(a) misses in 2028, the entire Lp(a)-lowering thesis — billions in R&D, three Phase 3 outcomes programs, an oral small-molecule race — collapses at once, and the biomarker-investment framework (“lower the causal marker, events follow”) takes its biggest hit since CETP inhibitors. For the Lp(a)-agnostics: even a positive OCEAN(a) would need a large enough HR to justify two-to-four-times-yearly injectable pricing against an events-averse but cost-conscious cardiologist base. For Novartis holders: the near-term risk is narrative — a second consecutive year of headline Phase 3 misses compresses the multiple even where the base business (Kisqali, Leqvio, RLT) performs. The trade, as ever with outcomes-trial misses, is patience: the next real information arrives at the congress presentation, and the real verdict in 2028. Anything traded on the topline adjectives between now and then is positioning, not investing.