Does lowering Lp(a) actually cut cardiovascular events? The outcome trials — HORIZON, OCEAN(a) and ACCLAIM — and why the first readout is still not in

Evidence-first notes on bioscience and deep tech, at the edge of the lab and the market. Information only — not investment advice, and not medical advice.

The 30-second version

• What. Earlier parts of this series established that Lp(a) is causal (Mendelian randomization), genetically fixed, and now pharmacologically lowerable by 80–90%+ with siRNA/ASO/oral agents. But biomarker lowering is not the clinical question. The clinical question — does lowering Lp(a) reduce hard cardiovascular events? — is answered only by large outcome trials, and as of this writing not one has reported a hard result.
• So what. The first readout, Lp(a)-HORIZON (pelacarsen, Novartis/Ionis, n = 8,323, secondary prevention), had an estimated primary-completion date of 2026-06-30 that has now passed, yet the registry status is still ACTIVE_NOT_RECRUITING and no topline press release or congress presentation is publicly confirmed. The actual result (hazard ratio, confidence interval, p-value) is [unverified] and is not reproduced or guessed anywhere in this piece. Whether lowering translates into fewer events remains undetermined.
• Now what. HORIZON is an unusually single-trial-dependent inflection point. Confirmation, if it comes, is years out: OCEAN(a)-OUTCOMES (olpasiran, Amgen, n = 7,297, registry readout 2028-03-31) and ACCLAIM-Lp(a) (lepodisiran, Eli Lilly, the only trial including primary prevention, ~2029). A new primary-prevention trial, OCEAN(a)-PreEvent (olpasiran, ~2031), has also appeared. The endpoints differ (HORIZON 4-point expanded MACE vs OCEAN(a) 3-point, stroke excluded), so cross-trial comparison needs adjustment. (All trials are manufacturer-sponsored.)

The five-minute read

Biomarker is solved; the hard-outcome translation is not

From a cardio-kidney-metabolic (CKM) perspective, Lp(a) is the genetic axis of residual risk that survives statins, PCSK9 inhibitors, SGLT2 inhibitors and GLP-1 agents. The prior parts of this series settled three things: causality (Mendelian randomization), genetic fixedness (the KIV-2 copy-number repeat), and 80–90%+ biomarker lowering by the new agents. None of those answers the final clinical question — whether lowering the number lowers the event rate. That translation, surrogate to hard outcome, is decided only in large event-driven randomized controlled trials. And on 2026-07-12, none of the three outcome trials has released a hard result. This is a chapter that waits for the answer rather than reporting it — and stating that honestly is the backbone of this part.

The first readout is overdue, and its result is isolated as unverified

Lp(a)-HORIZON (NCT04023552) is the direction-setter because it is the first Lp(a) cardiovascular outcome trial to reach completion. It enrolled 8,323 patients with established cardiovascular disease and screening Lp(a) ≥ 70 mg/dL, randomizing pelacarsen 80 mg monthly subcutaneously against placebo on top of standard care, across 797 sites in 42 countries. Its co-primary endpoints are time to first expanded MACE (cardiovascular death, non-fatal myocardial infarction, non-fatal stroke, urgent coronary revascularization) — one in the full ≥ 70 mg/dL population, and one in the ≥ 90 mg/dL subgroup, a design that bakes in the “lower more, gain more” absolute-reduction logic.

Company guidance pointed to Phase 3 data in the first half of 2026 (with regulatory filing in the second half if positive), after a slip from 2025 driven by the pace of blinded event accrual. The registry’s estimated primary-completion date, 2026-06-30, has now passed — yet status remains ACTIVE_NOT_RECRUITING and no confirmed topline is public. This is the textbook window just before a catalyst: overdue but not reported. The result is therefore isolated as [unverified], and no HR / CI / p-value is invented here.

The expected effect size — anchored to genetics, but only an upper bound

The Mendelian-randomization anchor is that a genetically 10 mg/dL lower Lp(a) corresponds to roughly a 5.8% lower coronary risk, so matching LDL’s 20% relative risk reduction would require lowering Lp(a) by about 101.5 mg/dL — meaning the absolute milligram reduction, not the percentage, is the primary determinant of expected benefit. In the ≥ 90 mg/dL co-primary subgroup, an ~80% pelacarsen reduction implies a large absolute drop, and the naive linear conversion lands in the ~20%+ relative-reduction region. But this is an upper-bound estimate: Mendelian randomization reflects lifelong exposure, whereas a trial delivers a few years of intervention against plaque that has already formed, so the realistic expected effect should be read conservatively.

Deep dive

1. Background

Lp(a) is a low-density-lipoprotein-like particle carrying apolipoprotein(a), whose plasma concentration is roughly 80–90% genetically determined by the LPA locus and largely unmodified by lifestyle or statins. Across the prior parts, the case for causality, for genetic fixedness, and for potent pharmacological lowering was assembled. The remaining and decisive question is clinical: does lowering the biomarker lower the event rate? That is what the outcome trials exist to answer, and this part dissects their design while being explicit that the treatment benefit itself is not yet demonstrated.

2. What these trials are designed to establish

• Lp(a)-HORIZON (direction-setter): pelacarsen 80 mg monthly, n = 8,323, established cardiovascular disease, Lp(a) ≥ 70 mg/dL, with a ≥ 90 mg/dL co-primary subgroup, expanded 4-point MACE. It is the first Lp(a) cardiovascular outcome trial to reach completion — which is exactly why the class hinges on it. If HORIZON is significant on the primary MACE endpoint, the “lowering reduces events” hypothesis is confirmed for the first time and the whole class (olpasiran, lepodisiran) is re-rated upward; if the confidence interval crosses 1, a surrogate-disappointment narrative (see the CETP-inhibitor analogy in Section 6) reverberates across the field.
• OCEAN(a)-OUTCOMES (second read): olpasiran, n = 7,297, ASCVD secondary prevention, with a 3-point endpoint (CHD death, MI, urgent revascularization) that excludes stroke — a real definitional difference from HORIZON’s 4-point endpoint. As a 2028 readout it functions as a confirmation/replication trial, provided the endpoint difference is accounted for in any cross-trial comparison. Note the correction to earlier estimates: N is 7,297 (not ~6,000) and the registry readout is 2028-03 (not 2026–2027).
• OCEAN(a)-PreEvent (new, primary prevention): a separate olpasiran primary-prevention outcome trial, n ≈ 11,000, currently recruiting, estimated completion 2031-10-20. Amgen has thus adopted a two-track strategy — secondary prevention (2028) plus primary prevention (2031) — a development not present in earlier framing of this series.
• ACCLAIM-Lp(a) (latest but broadest): lepodisiran dosed once every six months, registry N = 17,300 (estimate; press reports have cited ~12,500, so the figure is flagged unverified pending protocol reconciliation). It is the only trial to include both secondary and primary prevention in one study, with an event-driven MACE-4 composite that runs until ≥ 1,276 primary events accrue over roughly 4.75 years, completion estimated 2029-03. Including primary prevention widens the potential label the most, but a primary-prevention population has a lower baseline event rate, so the absolute benefit is likely shallower — which is why the trial needs a large N and long duration.

3. Methodological strengths and limits

• Strengths: All are large, event-driven, blinded-to-accrual outcome RCTs with registry-verified designs, endpoints and sample sizes. The class spreads its bets across secondary and (now) primary prevention, and across three independent drug modalities.
• Limit 1 — surrogate risk (top): Lp(a) lowering is a surrogate. History (CETP inhibitors, niacin) shows that a good-looking lipid biomarker can fail to translate into outcomes; 90% lowering does not automatically mean fewer events. HORIZON is the first direct test of that proposition, and until it reports the treatment benefit is unestablished.
• Limit 2 — single-trial dependence: the class effectively rests on one readout (HORIZON), with OCEAN(a) two years behind and ACCLAIM three. If HORIZON is ambiguous, confirmation faces a multi-year gap.
• Limit 3 — Mendelian randomization is not trial magnitude: lifelong genetic exposure overstates what a few years of drug can achieve; the naive milligram-to-RRR conversion is an upper bound.
• Limit 4 — endpoint heterogeneity: HORIZON’s 4-point expanded MACE (stroke included) versus OCEAN(a)’s 3-point endpoint (stroke excluded) makes direct cross-trial effect-size comparison fragile.
• Limit 5 — sample-size version mismatch: ACCLAIM registry 17,300 versus press ~12,500 suggests a possible protocol amendment; flagged unverified.

4. Neighbouring domains

There is a methodological cross-domain hook worth naming: these are large event-driven trials that stay blinded until a fixed cumulative event count (≥ 1,276 in ACCLAIM) accrues, which makes the exact readout date structurally uncertain — the same class of problem that AI-based event prediction and adaptive trial design aim to reduce. On the CKM side, Lp(a) sits alongside SGLT2 inhibitors, GLP-1 agents, finerenone and statins/PCSK9 inhibitors as the one axis of residual cardiovascular risk that is genetic and, until now, pharmacologically untouchable — which is why the translation question carries weight beyond a single drug class. This is a cross-trial, cross-mechanism reading, not a head-to-head claim.

5. Commercialization and market context (TRL, companies)

• TRL: the class holds at TRL 7 — large Phase 3 outcome trials in progress but hard outcome not yet demonstrated. A positive, significant HORIZON topline could move the class to a conditional TRL 8; a failure carries class-wide re-rating risk downward.
• Listed companies: pelacarsen (Novartis / Ionis), olpasiran (Amgen, AMGN), lepodisiran (Eli Lilly, LLY). All four outcome trials are manufacturer-sponsored, and every design figure is attributed to the registry and the sponsors.
• Share-price sensitivity: because HORIZON is the first and single direction-setting readout, its result — in either direction — would re-rate not only Novartis/Ionis but the whole class. This piece deliberately does not state or imply that result. The actual HORIZON topline, its disclosure date and venue are all unconfirmed and must not be inferred.

6. The skeptic’s counterpoint

Mandatory caveats inherited from the original deep-dive’s skeptic gate (§6, proceed-with-caveats, near hold):

• Treatment benefit is not demonstrated. Design, N, endpoints and timelines are registry-verified (verified-clean), but the therapeutic benefit on MACE itself is unverified, so any conclusive implication is at hold level until HORIZON reports. Re-adjudicate immediately on topline release.
• Surrogate risk is the top concern. Lp(a) lowering is a surrogate; the CETP-inhibitor and niacin precedents show a strong biomarker can fail to translate. HORIZON is the first direct test.
• Single-trial dependence. The class rests on one readout, with confirmation two to three years out.
• Mendelian randomization ≠ RCT magnitude. Lifelong exposure overstates a few years of drug; the milligram-to-RRR conversion is an upper bound.
• Endpoint heterogeneity and sample-size mismatch. HORIZON 4-point vs OCEAN(a) 3-point (stroke excluded); ACCLAIM N 17,300 (registry) vs ~12,500 (press), flagged unverified.
• Sponsorship attribution. All trials are manufacturer-sponsored; all quantitative claims are attributed to the registry, sponsors and their sources.
• Falsifiable predictions. (1) HORIZON’s ≥ 90 mg/dL co-primary subgroup will show a larger relative reduction than the full population (absolute-lowering logic); if not, the “milligrams drive benefit” frame weakens. (2) OCEAN(a)’s 3-point (stroke-excluded) effect size will differ systematically from HORIZON’s 4-point by endpoint composition. (3) Primary-prevention trials (ACCLAIM, OCEAN(a)-PreEvent) will show a smaller absolute benefit than secondary prevention, requiring larger N and longer duration.

7. What to watch

1. The HORIZON topline itself — direction, primary-endpoint significance, and specifically whether the ≥ 90 mg/dL co-primary subgroup outperforms the full population. Until disclosed, treated as [unverified].
2. Whether OCEAN(a)-OUTCOMES (2028) replicates HORIZON once the 3-point vs 4-point endpoint difference is reconciled.
3. Primary-prevention signal size in ACCLAIM and OCEAN(a)-PreEvent — the absolute-benefit test.
4. Reconciliation of the ACCLAIM sample size (registry 17,300 vs press ~12,500).

References

• Lp(a)-HORIZON — pelacarsen cardiovascular outcomes study (NCT04023552). ClinicalTrials.gov (status ACTIVE_NOT_RECRUITING; estimated primary completion 2026-06-30). https://clinicaltrials.gov/study/NCT04023552
• OCEAN(a)-OUTCOMES / TIMI 75 — olpasiran (NCT05581303). ClinicalTrials.gov. https://clinicaltrials.gov/study/NCT05581303 (TIMI Study Group: https://timi.org/oceana-outcomes-timi-75/)
• ACCLAIM-Lp(a) — lepodisiran (NCT06292013). ClinicalTrials.gov. https://clinicaltrials.gov/study/NCT06292013
• OCEAN(a)-PreEvent — olpasiran primary prevention (NCT07136012). ClinicalTrials.gov (status RECRUITING; estimated completion 2031-10-20). https://clinicaltrials.gov/study/NCT07136012
• Lp(a)-HORIZON trial design. American Heart Journal (2025). https://www.sciencedirect.com/science/article/pii/S0002870325001012
• Ionis announces enrollment completion of the Phase 3 Lp(a)-HORIZON study of pelacarsen. PR Newswire. https://www.prnewswire.com/news-releases/ionis-announces-enrollment-completion-of-phase-3-lpa-horizon-cardiovascular-outcomes-study-of-pelacarsen-301589746.html
• 2026 cardiovascular catalysts: Lp(a) on the horizon. BioCentury (2026). https://www.biocentury.com/article/658135/2026-cardiovascular-catalysts-lp-a-on-the-horizon
• Novartis extends timeline for pelacarsen Phase 3 cardiovascular trial to 2026. Medpath (2025-01). https://trial.medpath.com/news/5789d0b880688613/novartis-extends-timeline-for-pelacarsen-phase-3-cardiovascular-trial-to-2026

Note: registry fields, sample sizes, endpoints and timelines were triangulated from clinicaltrials.gov plus company and press sources and should be reconfirmed. The HORIZON topline result is unconfirmed as of 2026-07-12 and is not reproduced here. Copyrighted full text is not redistributed; only summary, commentary and quotations ≤ 150 characters are included.

Disclosure

This post is for information only. It is not investment advice, and it is not medical advice — treatment decisions must be made in consultation with a qualified healthcare professional. The author holds no position in the named securities (Novartis, IONS, AMGN, LLY).

COI note: the outcome trials cited here (Lp(a)-HORIZON, OCEAN(a)-OUTCOMES, OCEAN(a)-PreEvent, ACCLAIM-Lp(a)) are each sponsored by the respective drug’s manufacturer (Novartis/Ionis, Amgen, Eli Lilly). All design figures — sample size, endpoints, timing, status — are attributed to the trial registry and the sponsors. This piece does not assert that any manufacturer’s drug reduces cardiovascular events: that treatment benefit is unverified pending the first hard-outcome readout, and no efficacy result (HR, CI, p-value) is stated or implied.