Novartis and Novo Nordisk Cardiovascular Trials Show No Benefit Despite Genetic Targets

The Novartis Lp(a) HORIZON trial randomized over 8,000 patients with elevated Lp(a) to pelacarsen or placebo. Lp(a) levels fell substantially, yet the composite endpoint of cardiovascular death, myocardial infarction, and stroke showed no difference at 4 years. Absolute event rates remained 12.4 percent versus 12.1 percent. A parallel Novo Nordisk study of ziltivekimab in 6,000 post-MI patients lowered hs-CRP by 65 percent but produced identical rates of major adverse events.

These results follow repeated disappointments with genetically validated targets. Earlier PCSK9 and ANGPTL3 programs succeeded only after large outcomes trials; many others, including CETP inhibitors, did not. The pattern indicates that genetic associations identify risk markers more reliably than causal pathways amenable to single-drug intervention.

AI platforms that mine genomic and proteomic data to prioritize assets inherit the same limitation. They optimize for target engagement rather than clinical translation, where redundant pathways and long latency dilute surrogate effects. The AHA presentations underscore that faster candidate generation does not address the core attrition driver of human biology.

Future programs will require adaptive enrichment designs and longer follow-up to detect modest benefits. Regulatory agencies may also demand demonstration that AI-selected targets improve net clinical outcomes, not merely biomarkers, before granting accelerated approval.