Two recent major cardiovascular trials have now delivered an uncomfortable but important lesson: successfully changing a biological marker does not necessarily mean that patients will have fewer heart attacks, strokes, or cardiovascular deaths.

The Phase III ZEUS trial of ziltivekimab and the Phase III Lp(a) HORIZON trial of pelacarsen approached cardiovascular disease from very different directions. ZEUS targeted inflammation, specifically interleukin-6 (IL-6). Lp(a) HORIZON targeted lipoprotein(a), or Lp(a), an inherited lipid particle strongly associated with cardiovascular disease.

Both drugs did what they were designed to do biologically. Yet neither trial demonstrated the anticipated reduction in cardiovascular events. That convergence makes the two studies particularly important to consider together.

ZEUS: Lowering Inflammation Was Not Enough

The rationale behind ZEUS was compelling. Atherosclerosis is not simply a disorder of cholesterol accumulation. It is also an inflammatory disease. Previous research, particularly the landmark CANTOS trial , established that selectively reducing inflammation could lower cardiovascular event rates without lowering LDL cholesterol.

Ziltivekimab was designed to take that concept a step further by directly inhibiting IL-6, a major signaling protein involved in vascular inflammation.

Earlier evidence looked remarkably promising. In the Phase II RESCUE trial , ziltivekimab produced profound reductions in high-sensitivity C-reactive protein, or hsCRP, a commonly measured marker of inflammation. Depending on dose, hsCRP fell by roughly 77% to 92%. Other markers related to inflammation and clotting also declined.

ZEUS was designed to answer the question that biomarker studies could not: would suppressing this pathway actually prevent cardiovascular events?

The trial enrolled 6,376 patients with a particularly high-risk combination of established atherosclerotic cardiovascular disease, chronic kidney disease, and systemic inflammation. Participants received either ziltivekimab 15 mg once monthly or placebo in addition to contemporary cardiovascular therapy.

The answer was unexpectedly clear. Although ziltivekimab successfully inhibited the IL-6 pathway and lowered the inflammatory marker, high sensitivity C-reactive protein (hsCRP), it did not reduce the composite risk of cardiovascular death, nonfatal heart attack, or nonfatal stroke. The hazard ratio was 0.99, essentially indicating no difference from placebo.

Overall adverse events and serious adverse events were similar between groups, although serious infections occurred more frequently among patients receiving ziltivekimab.

ZEUS therefore separates two concepts that are frequently blurred together: biological activity and clinical efficacy are not the same thing.

It does not mean that inflammation is irrelevant to atherosclerosis. Rather, it suggests that blocking this particular inflammatory pathway, at this stage of disease, in this particular patient population, was insufficient to alter cardiovascular outcomes.

Lp(a) HORIZON: Another Powerful Biomarker Reduction Meets An Outcomes Test

On September 4, 2026, cardiovascular medicine received a second major negative trial result.

Novartis announced that Lp(a) HORIZON did not meet its primary endpoint. Pelacarsen lowered Lp(a), but that reduction did not translate into a statistically demonstrated reduction in cardiovascular events in the overall study population.

The result is particularly consequential because the biological and genetic case against LP(a) is strong. LP(a) resembles an LDL particle attached to a protein called apolipoprotein(a). Its concentration is predominantly inherited and resistant to diet and exercise. Elevated LP(a) has also been linked to atherosclerosis, heart attacks (myocardial infarction), stroke, peripheral vascular disease, and calcific aortic valve disease. Roughly one in five people worldwide has elevated levels of Lp(a).

Pelacarsen attacks the problem near its source. It is an antisense oligonucleotide that interferes with production of apolipoprotein(a) in the liver and can markedly reduce circulating Lp(a).

Lp(a)HORIZON enrolled 8,323 patients who already had cardiovascular disease and an Lp(a) concentration of at least 70 mg/dL. Participants received monthly pelacarsen or placebo on top of standard preventive treatment and medications.

The primary outcome combined cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, and urgent coronary revascularization requiring hospitalization. The endpoint was evaluated both across the overall population and in patients with particularly high Lp(a) levels.

According to the September 4th topline announcement , pelacarsen lowered Lp(a), but the trial failed to show that this translated into fewer primary cardiovascular events.

The detailed data remain critical. Novartis has not yet released the full event rates, achieved Lp(a) reductions, subgroup findings, component endpoints, or other analyses that will be necessary to understand precisely why the trial was neutral.

“The Lp(a)HORIZON trial teaches us yet again that the relationship between elevated Lp(a) levels and increased risk of cardiovascular events is more complex than we thought. Previously, we had seen niacin in the AIM-HIGH trial lower Lp(a) levels but this also did not translate in a reduction in cardiovascular events. We have other medications such as PCSK9 inhibitors that lower Lp(a) levels by nearly 30 percent but have not been studied for this indication. Similar drugs to pelacarsen are still in trials and it will be interesting to see what the results will be,” said Dr. Abeer Berry , a board-certified cardiologist and Director of Noninvasive Cardiology at DMC Huron Valley-Sinai Hospital in Commerce Township, Michigan, who was not involved in either study.

Why The Two Trials Matter Together

ZEUS and Lp(a) HORIZON were testing different hypotheses, but they collided with the same principle. Modern cardiovascular medicine increasingly identifies “residual risk”--the risk that remains even when LDL cholesterol, blood pressure, diabetes, smoking, and other conventional factors are aggressively managed.

The ZEUS trial targeted residual inflammatory risk, and Lp(a)HORIZON targeted residual genetically mediated lipoprotein risk. Neither intervention reduced its respective biological target and automatically produced fewer clinical events.

That should temper a growing tendency in precision medicine to equate a compelling biomarker with a validated therapeutic target.

That said, these findings should not be interpreted as proof that either IL-6 or LP(a) is biologically unimportant. A biomarker can be associated with disease--and may even contribute causally to disease--without guaranteeing that modifying it later in life will reverse decades of accumulated vascular injury. Timing may matter and duration of treatment may matter more importantly. The degree of target reduction may also be crucial. And the selected population and background therapy may matter. And different methods of modifying the same biological pathway may produce different outcomes.

These distinctions will be especially important for Lp(a). Other agents, including olpasiran and lepodisiran, suppress Lp(a) through different RNA-based technologies and are being evaluated in large cardiovascular outcomes programs. The neutral Lp(a)HORIZON result therefore raises the evidentiary bar for those programs rather than automatically determining their outcomes.

The same reasoning applies to inflammation. ZEUS does not erase CANTOS or decades of vascular and immunology focused research. It demonstrates that direct IL-6 ligand inhibition with ziltivekimab did not reduce events in patients with established cardiovascular disease, chronic kidney disease, and persistent inflammation. Whether different inflammatory targets, different patient phenotypes, or treatment during acute heart attacks (myocardial infarction) or heart failure produces another result remains an open question.

Berry also emphasized that inflammatory risk remains an open therapeutic question: “It’s common for there to be residual inflammation that persists outside of treatment of elevated cholesterol, sugar and blood pressure in patients. Although the ZEUS trial proved that reducing inflammation did not reduce the risk of cardiovascular events, there is still a question as to whether using other inflammatory marker inhibitors outside of IL-6 may be successful as we saw somewhat with colchicine, for example.”

The most important contribution of ZEUS and Lp(a) HORIZON may ultimately be methodological rather than pharmacological.

Cardiovascular medicine has become extraordinarily adept at identifying molecules associated with risk. The harder question is determining which of those molecules represent therapeutically modifiable causes of disease.

Both trials crossed that final bridge from biology to outcomes--and failed to demonstrate clinical benefit despite successful target engagement. That is not scientific failure. It is precisely why large, randomized controlled trials are necessary.

As Berry put it, “The findings of these studies are not indications of the failure of science, but actually demonstrate how science works. Biology allows us to form hypotheses and scientific trials teach us if they are true or false.”

A laboratory measurement can tell us whether a drug reached its target. Only a properly conducted clinical outcomes trial can tell us whether changing that target actually helps patients live longer or avoid heart attacks and strokes.

Taken together, ZEUS and Lp(a) HORIZON offer a sobering reminder for the next era of cardiovascular precision medicine: the ultimate endpoint is not a lower biomarker--it’s a healthier patient.