Recent heart drug clinical trial failures highlight limitations of artificial intelligence in research
Recent clinical trial setbacks have renewed skepticism over claims that artificial intelligence and advanced research technologies can rapidly transform drug discovery, particularly in cardiovascular medicine. While developers have long anticipated that computational tools and modern genetics would accelerate novel therapies, heart attacks and strokes—the leading killer in the developed world—continue to expose persistent limitations in translating theoretical predictions into effective clinical treatments.
The gap between high-tech expectations and clinical reality was highlighted by two major studies presented on Saturday at the American Heart Association's annual meeting. In one trial, Novartis tested pelacarsen, a medication designed around genetic findings suggesting that reducing a blood substance called lp(a) would curb cardiovascular risk. Although pelacarsen successfully lowered lp(a) levels, it produced no effect on the rate of heart attacks and strokes. In another large study, Novo Nordisk evaluated an anti-inflammatory therapy targeting the protein IL-6, which likewise failed to demonstrate any patient benefit. Together, the trial failures underscore that biological complexity still presents formidable barriers that algorithmic and genetic models cannot easily overcome.
Continue in the app — vote & join in ➔