Stents Show No Benefit at 25 Years in COURAGE

Opening a blocked artery with a stent doesn't lower the long-term risk of death or heart attack for people with stable coronary artery disease. That's the finding of an extended follow-up of the COURAGE trial, published in JAMA and presented on November 7 at the 2026 American Heart Association Scientific Sessions in Chicago. The analysis, led by Neel M. Butala, MD, of the University of Colorado Anschutz Medical Campus, tracked patients for up to 25 years, the longest follow-up the question has had.

The numbers are close to identical. The estimated 25-year incidence of death or heart attack was 93.3 percent among patients assigned to percutaneous coronary intervention plus optimal medical therapy, and 91 percent among those assigned to medical therapy alone. The hazard ratio was 1.01, with a 95 percent confidence interval of 0.88 to 1.15 and a P value of .89. Death alone ran 83.9 percent against 82.2 percent. Heart attack ran 54.6 percent against 54.0 percent.

The follow-up covered 959 veterans treated at 15 Veterans Affairs hospitals, 42 percent of the 2,287 patients the original trial randomized. Of those, 485 had been assigned to the procedure plus medical therapy and 474 to medical therapy alone. Their mean age was 63, and 98 percent were men. Researchers followed them for a median of 15 years through linked VA and Medicare records.

This isn't a new randomized trial, and the study says so plainly. It's a post hoc observational follow-up of the original trial's participants, with outcomes drawn from administrative and laboratory data rather than trial adjudication. The method is how a 25-year look is possible at all. It also means heart attacks can be missed or misclassified in ways a running trial would catch, a limit the researchers name themselves.

The result extends a line that hasn't moved in two decades. COURAGE first reported in 2007, after a median follow-up of 4.6 years, that adding the procedure to medical therapy didn't reduce death or heart attack. A 2015 extension, at a median of 11.9 years, found no survival difference either. Each longer look has tested whether a benefit was hiding further out. At 25 years, it still isn't there.

The limits matter for how far the finding travels. The patients were overwhelmingly male veterans at high baseline cardiovascular risk. Their procedures used the technology of their era, including bare-metal stents, not the drug-eluting stents and imaging guidance in use now. Current guidelines still recommend the procedure for chest pain that persists despite medical therapy. It relieves symptoms. What this quarter-century of data doesn't support is doing it up front, in stable patients, to prevent a heart attack or a death.

One more measure closes the loop. About half of each group ended up having an artery-opening procedure during the follow-up anyway: 53.5 percent of the medical-therapy group had a first revascularization, and 50.4 percent of the procedure group had a repeat one. The difference isn't significant. Starting with medicine didn't strand patients without procedures. It reserved the procedure for the patients whose symptoms called for it, and the long-term outcomes landed in the same place.