Novartis, Novo Nordisk Heart Drug Failures Should Temper AI Hype
Two large trials of heart drugs from Novartis and Novo Nordisk found no benefit against heart attacks and strokes, results presented at the American Heart Association meeting. A medicine writer says the setbacks are a caution for claims that new technology will cure disease.
Two large trials of heart drugs from Novartis and Novo Nordisk showed no benefit in preventing heart attacks and strokes, according to results presented at the American Heart Association's annual meeting in November 2026. The heart drug failures should make investors and the public more cautious about promises that new technology will quickly deliver cures, argues Matthew Herper, a senior medicine writer who covered the meeting.
What the two trials found
Novartis tested a drug called pelacarsen. It lowers a substance in the blood called Lp(a), which genetic research had indicated should reduce cardiovascular risk. The trial found no effect on the rate of heart attacks and strokes.
Novo Nordisk ran a separate large study of an anti-inflammatory drug that targets a protein called IL-6. That study also showed no benefit.
Both drugs were aimed at targets that looked promising on paper. Neither delivered in patients.
Why the results matter
Heart attacks and strokes together are the largest killer in the developed world, Herper noted. Drugmakers have spent years searching for new ways to prevent them.
Both trials were built on newer ideas about how the disease works. Herper says the genetics behind these targets was supposed to improve the odds of success. In his words, it "did anything but."
The AI comparison
Herper draws a wider lesson for anyone pitching artificial intelligence, or any other new technology, as a way to cure every disease. He points to the repeated record in heart disease, where pharmaceutical researchers have learned that technologies meant to speed new medicines to patients have fallen short.
His point is not about one drug. It is that a stronger scientific rationale, and faster tools for finding targets, have not guaranteed that a medicine works in people.
For companies, a failed late-stage trial is costly. These studies involved large numbers of patients and tested whether a drug cut real events, such as heart attacks and strokes, rather than a number on a lab test. Lowering Lp(a) did exactly what the drug was designed to do, yet it did not change the outcomes that matter to patients.
Neither Novartis nor Novo Nordisk had said what it would do next with the programs in the material reviewed for this report. Herper's argument is that the field's record should temper expectations the next time a new technology is promised as a cure.
