Clinical trial · Research · 7 Nov 2026
In 5,179 people with moderate or severe ischemia, an invasive strategy did not lower 11-year mortality
Source: JAMA
What the research found
JAMA published an extended follow-up of a randomized trial on 7 November 2026. Hochman, Anthopolos, and colleagues report ISCHEMIA-EXTEND. The original trial enrolled 5,179 people with moderate or severe ischemia, a median age of 64 years (IQR 58 to 70), 23 percent women, 42 percent with diabetes, and a median ejection fraction of 60 percent. Entry required an ejection fraction of at least 35 percent, and excluded left main coronary disease, unacceptable angina, and a recent acute coronary syndrome. People were randomized at 320 sites in 37 countries to an invasive strategy (guideline-directed medical therapy plus angiography and revascularization when indicated) or a conservative strategy (guideline-directed medical therapy, with an invasive procedure only if that therapy failed). Follow-up ran through December 2025. Median follow-up was 9 years. There were 1,071 deaths (782 of them in the extension): 672 cardiovascular, 359 not cardiovascular, and 40 unclassified. At 11 years, all-cause mortality was 25.8 percent with the invasive strategy and 26.9 percent with the conservative strategy, a difference of -1.0 percentage points (95 percent CI -4.2 to 2.1). The adjusted hazard ratio was 0.97 (0.86 to 1.09). A Bayesian estimate put the probability of more than a 1 percentage point survival benefit at 18 percent. Cardiovascular mortality was 15.8 percent versus 17.1 percent, difference -1.4 percentage points (-4.0 to 1.3), adjusted hazard ratio 0.88 (0.75 to 1.02). Mortality that was not cardiovascular was 9.1 percent versus 8.1 percent, difference +1.0 percentage points (-0.9 to 2.9), adjusted hazard ratio 1.18 (0.96 to 1.45). The restricted mean survival difference was 14 days (-41 to 70). Registration NCT04894877. DOI 10.1001/jama.2026.20375. Not medical advice.
Evidence level: Clinical trial. This is among the stronger kinds of evidence, but one study rarely settles a question.
More research news
- In 14,665 adults, an AI electrocardiogram flagged slightly more atrial findings in 90 days, and one swapped event would erase the result
- After adjustment, depression or anxiety lined up with a small rise in late atrial fibrillation in 44,338 people who had cardiac surgery, mainly in men
- In genetic data from hundreds of thousands of women, a lower Lp(a) signal was not tied to worse pregnancy or fertility outcomes
- In 9,476 adults in Brazil with type 2 diabetes and high cardiovascular risk, a systolic target under 120 mm Hg did not reduce major events versus under 140
- In notes on 28,113 people, most of those above their LDL goal had no escalation of lipid-lowering therapy