Novo Nordisk stunned the cardiology world last week when it announced that a large study it was running on an experimental drug for heart disease had failed. The drug, widely expected to be a blockbuster, was no better than a placebo.

The drug, ziltivekimab, reduces inflammation, which is generally thought to be a cause of atherosclerosis. Stopping that inflammation was hoped to be a way to save lives.

Now the drug’s failure is raising a troubling question. Is inflammation really a cause or is it just a marker of atherosclerosis, the way gray hair is a marker of aging?

Cardiologists had mixed reactions to the news. Dr. Sanjay Kaul, a cardiologist at Cedars Sinai, quoted Thomas Huxley on X: “The great tragedy of science — the slaying of a beautiful hypothesis by an ugly fact.”

Dr. David Maron of Stanford, who is president of the American Society for Preventive Cardiology, said the result “comes as a surprise and a big disappointment.”

But he added, something could have been wrong with the study: Perhaps it just wasn’t the right drug to target heart disease inflammation, or maybe patients in the trial were too sick to be helped by anti-inflammatory medications. There is too much evidence, he said, from pathology studies of plaques in arteries, animal studies, epidemiology and other clinical trials, to toss out the inflammation hypothesis.

“Inflammation is more than a marker of disease, it is a cause of the disease,” Dr. Maron said.

The Novo Nordisk study, called Zeus, involved more than 6,300 patients with heart disease or advanced kidney disease. They also had high levels in their blood of a marker of inflammation known as C reactive protein, or CRP. And they had high levels of interleukin-6, or il-6, a chemical that causes inflammation.

The drug did what it was supposed to do. It made il-6 levels plunge. CRP levels fell in concert.

With those signals indicating that inflammation was controlled, it seemed only logical that heart disease complications and cardiovascular death rates would drop, too. But that didn’t happen.

“I think we’re all in a state of shock,” said Dr. Martha Gulati of Houston Methodist, past president of the American Society for Preventive Cardiology.

Novo Nordisk said in a statement that it was continuing to test ziltivekimab in two other studies of patients at risk of heart attacks and cardiovascular complications and deaths. Those studies will conclude in 2027.

The idea that inflammation causes atherosclerosis dates back more than a quarter-century when Dr. Paul Ridker of Brigham and Women’s Hospital published a paper positing the link.

Dr. Sanjit Jolly, a medical student at the time, vividly remembers a 2002 Time magazine cover story “Beyond Cholesterol,” based on Dr. Ridker’s work. It stated that inflammation, measured by CRP in the blood, was linked to heart attacks. “I was really excited by this idea,” he said.

Suggestive evidence began to accumulate.

Two studies tested colchicine, an anti-inflammatory drug used to treat gout, in heart patients. It reduced heart disease complications but had inconsistent effects on inflammation.

An experimental anti-inflammatory drug made by Novartis had a small beneficial effect in heart patients but increased the risk of fatal infections. Novartis did not market the drug for heart disease, but it seemed additional proof of concept for the inflammation hypothesis.

As a professor at McMaster University in Canada, Dr. Jolly saw a chance in 2018 to do what he hoped would be a definitive test of inflammation. It was larger than the previous studies, involving more than 7,000 heart disease patients who were given colchicine or a placebo. Levels of CRP fell in those taking the drug, but they fared no better than those taking a placebo.

“People said, ‘Something has to be wrong with your study,’” Dr. Jolly said. “I said: ‘I am a scientist. I have to wait for trials like Zeus to come out.’”

Now, with Zeus results that are nearly identical to the results from Dr. Jolly’s similarly sized study, he is left wondering what will happen next, in the two ongoing Novo Nordisk trials of ziltivekimab.

Others are also starting to wonder if perhaps they have been pinning their hopes on a hypothesis that is wrong.

Dr. C. Michael Gibson, a Harvard cardiologist, said he had not given up on the inflammation hypothesis, but was starting to question it. “I’ve been following the inflammation hypothesis for a couple of decades,” he said. “The data was very compelling. There was at least a good correlation between inflammation and outcomes.” The hypothesis, he added, “seemed to make good sense.”

But, he said, the Novo Nordisk study gives him pause.

Dr. Gulati too has not given up on the hypothesis. But, she said, she is tempering some of her enthusiasm. She had hoped, she said, that by adding an inflammation treatment to treatments to lower LDL, the bad type of cholesterol, “we could stomp out heart disease.”

“I actually thought this would be the new frontier,” she said.

Now she is asking a new question: “Does treating inflammation even matter by the time you have atherosclerosis? Or is it too late?”