The Lipoprotein(a) Paradox: From Vulnerable Plaque to a Missed Endpoint

On 4 September 2026, the Lp(a)HORIZON trial reported that pelacarsen did not reduce its primary composite endpoint in 8,323 patients with established atherosclerotic cardiovascular disease and lipoprotein(a) of at least 70 mg/dL. The full dataset has not been presented. This commentary argues that the topline result should neither be dismissed nor allowed to erase the genetic and epidemiologic evidence preceding it, because the two bodies of evidence answer different questions. Mendelian randomization describes lifelong exposure and cannot be translated directly into the effect of a finite-duration drug. Imaging associates Lp(a) with plaque burden and, in one post hoc analysis, independently with focal vulnerable plaque; neither is a clinical outcome. Intensive LDL-C lowering itself modifies plaque vulnerability, so no exclusive division of labor between the two lipoproteins is supportable. The most parsimonious reading of a negative trial remains that the intervention does not work, and that possibility is held open here. Five parameters are proposed — achieved absolute concentration, the prespecified ≥90 mg/dL population, coronary endpoint components, follow-up distribution with event-time curves, and background lipid-lowering therapy and adherence — that determine what the trial actually tested, whatever the direction of the result.