When the First Trial Fools Us

Why early enthusiasm can lead us to misread evidence

Note to readers: Starting this Wednesday, and for the rest of July, my posts will be directed more toward a broader audience of patients, families, and anyone curious about their health. I will continue to post clinical and research-focused content, but for the next month or so, the mix will shift a bit.

My first review paper and book chapter were about Lp(a) excess and its historical treatments, including niacin, estrogen, and neomycin (1,2). So I am very excited that later this summer, we will get our first look at the cardiovascular disease (CVD) outcome data for pelacarsen, an Lp(a)-lowering therapy. The Lp(a) HORIZON trial is testing pelacarsen, compared with placebo, in people with established CVD and elevated Lp(a) (3). This study has been eagerly awaited because Lp(a) is a genetically determined, largely unmodifiable CVD risk factor, and we still do not know whether lowering it reduces cardiovascular events (2,4), despite compelling Mendelian randomization evidence (5) and strong biological plausibility. The field is primed to believe, which is precisely the moment to be careful.

In my last post, I outlined how I read a randomized clinical trial before I decide whether it should change my practice. When the Lp(a) HORIZON study is reported, I will apply every one of those steps, but a few will deserve particular attention.

First, what was the background lipid-lowering therapy and was it optimized? According to the design paper, patients were treated aggressively by the standards at the time the study was designed, with ~77% on high-intensity statins, but only ~56% on ezetimibe and ~11% on a PCSK9 inhibitor. The median baseline LDL-C was 65.7 mg/dL. But by the standards of the recent 2026 ACC/AHA Dyslipidemia guidelines, many, if not most, patients with established atherosclerotic CVD now would be treated toward an LDL-C goal of <55 mg/dL (6). Thus, if patients in Lp(a) HORIZON were not already on high-intensity statins, ezetimibe, and a PCSK9 inhibitor when indicated, then any apparent benefit from pelacarsen may partly reflect residual undertreated apoB-related risk, rather than Lp(a)-lowering specifically. Indeed, if apoB fell meaningfully in the treatment arm, isolating the Lp(a) effect from a more general lipoprotein-lowering effect will be difficult, because even modest reductions in LDL-C after correcting for Lp(a)-cholesterol or in apoB could affect event rates (7). A key question will be whether achieved apoB and LDL-C levels are similar between groups, or whether differences in apoB explain part of the observed effect.

Second, what are the absolute event rates in each arm, how large are the between-arm differences, and which event types are reduced? In Lp(a) HORIZON, the primary endpoint is a composite of cardiovascular death, nonfatal MI, nonfatal stroke, or urgent coronary revascularization requiring hospitalization (3). These endpoints sit on a hierarchy: cardiovascular death and nonfatal stroke matter much more to patients than nonfatal MI, which matters more than revascularization. Also, all-cause mortality is a secondary endpoint worth examining carefully, but it is a high bar to clear in 3 to 6 years. I hope for a markedly positive trial, but suspect that if the study is positive, the benefits will be driven primarily by reductions in nonfatal events. Then we will debate in what circumstances we should use the novel Lp(a)-lowering agents rather than further apoB reduction with existing agents, as we await trials like OCEAN(a)-Outcomes (TIMI 75) with olpasiran, among other studies.

I will share a full analysis when the data are available. But the reason to raise these questions now, before the trial reports, is that the moment of maximum enthusiasm is the moment of maximum vulnerability to misreading evidence. We have been here before.

The statin + niacin story

The HDL-Atherosclerosis Treatment Study (HATS) enrolled 160 patients with coronary artery disease and low HDL-C. In the key comparison, simvastatin plus niacin was associated with a 90% relative reduction in the composite of death, myocardial infarction, stroke, or revascularization (8). The authors were cautious about the small sample, but the cardiology community embraced the results and niacin’s reputation as a lipid therapy with hard CVD outcome data was established.

It wasn’t until a decade later that the assumption of benefit for adding niacin to a statin was challenged, when the AIM-HIGH study of >3,400 patients was stopped early for futility (9). And two years later, HPS2-THRIVE (10) settled the question in >25,000 patients: niacin added to statins provided no reduction in major vascular events and was associated with an increase in serious adverse events. There were some baseline differences between participants in these three studies, but the question about niacin’s usefulness was settled, and the answer was the opposite of what HATS had suggested.

The phenomenon of effect-size inflation (the “winner’s curse”) operates most powerfully when event counts are small, because trials that randomly deviate toward larger-than-true effects are more likely to cross significance thresholds and get published. HATS was a useful hypothesis-generating study that got mistaken for confirmatory evidence. The lesson is not that small trials are useless: it is that implausibly large effects in small samples deserve proportionally large skepticism.

What to watch for in August

The anticipation around Lp(a) HORIZON is legitimate, and I share it: I would love another treatment in my CVD risk-reducing arsenal, especially for people with high Lp(a). But I am cautious. Obviously, Lp(a) HORIZON is much larger than HATS, but high excitement and plausible biology are precisely when overinterpretation is most likely.

When the data are published, I will read the methods, the tables, and the supplements before I read the conclusions. I will look at the absolute event rates in each arm. I will ask which event types were reduced and by how much. I will ask what happened to apoB. I will ask whether the follow-up was long enough for the biology to express itself in hard endpoints. And I will remember that the trials that hold up do not need spin, because the data tell their story.

The trials that hold up do not need spin. The data tell their story.

References:

1. Stein JH, Rosenson RS. Lipoprotein Lp(a) excess and coronary heart disease. Arch Intern Med 1997;157(11):1170-6.

2. Rosenson RS, Stein JH. Lipoprotein(a). UpToDate. Wolters Kluwer.

3. Cho L, Nicholls SJ, Nordestgaard BG, et al. Design and Rationale of Lp(a)HORIZON Trial: Assessing the Effect of Lipoprotein(a) Lowering With Pelacarsen on Major Cardiovascular Events in Patients With CVD and Elevated Lp(a). Am Heart J 2025;287:1-9. doi: 10.1016/j.ahj.2025.03.019.

4. Tsimikas S. A Test in Context: Lipoprotein(a): Diagnosis, Prognosis, Controversies, and Emerging Therapies. J Am Coll Cardiol 2017;69(6):692-711. doi: 10.1016/j.jacc.2016.11.042.

5. Burgess S, Ference BA, Staley JR, et al. European Prospective Investigation Into Cancer and Nutrition–Cardiovascular Disease (EPIC-CVD) Consortium. Association of LPA Variants With Risk of Coronary Disease and the Implications for Lipoprotein(a)-Lowering Therapies: A Mendelian Randomization Analysis. JAMA Cardiol 2018;3(7):619-627. doi: 10.1001/jamacardio.2018.1470. PMID: 29926099.

6. Blumenthal RS, Morris PB, Gaudino M, et al. 2026 ACC/AHA/AACVPR/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Dyslipidemia. J Am Coll Cardiol. Published online March 13, 2026. doi:10.1016/j.jacc.2025.11.016.

7. Yeang C, Karwatowska-Prokopczuk E, Su F, et al. Effect of Pelacarsen on Lipoprotein(a) Cholesterol and Corrected Low-Density Lipoprotein Cholesterol. J Am Coll Cardiol 2022;79(11):1035-1046. doi: 10.1016/j.jacc.2021.12.032.

8. Brown BG, Zhao XQ, Chait A, et al. Simvastatin and niacin, antioxidant vitamins, or the combination for the prevention of coronary disease. N Engl J Med 2001; 345(22):1583-92. doi: 10.1056/NEJMoa011090.

9. AIM-HIGH Investigators; Boden WE, Probstfield JL, Anderson T, et al. Niacin in patients with low HDL cholesterol levels receiving intensive statin therapy. N Engl J Med 2011; 365(24):2255-67. doi: 10.1056/NEJMoa1107579.

10. HPS2-THRIVE Collaborative Group; Landray MJ, Haynes R, Hopewell JC, et al. Effects of extended-release niacin with laropiprant in high-risk patients. N Engl J Med 2014;371(3):203-12. doi: 10.1056/NEJMoa1300955.