JAMA November 07, 2026

The ISCHEMIA Trial a Decade Later—What Have We Learned?

Cutlip DE, Yeh RW

Bottom Line

Extended 9-year follow-up of the ISCHEMIA trial confirms that a routine initial invasive strategy for stable coronary disease with moderate-to-severe ischemia offers no long-term overall survival advantage compared to an initial conservative strategy focused on optimal medical therapy.

Key Findings

1. At a median follow-up of 9.0 years (and a maximum of 13.4 years), there was no significant difference in all-cause mortality between the initial invasive and initial conservative strategies (approximately 26% vs. 27%, respectively).

2. Among the original 5,179 randomized participants, a total of 1,071 deaths were recorded, of which 782 occurred during the extended follow-up (ISCHEMIA-EXTEND) phase.

3. While earlier interim 5.7-year analyses noted a potential reduction in cardiovascular mortality (HR ~0.78) offset by an increase in non-cardiovascular mortality, the nearly 10-year follow-up demonstrates strict parity in overall long-term survival.

4. The data solidifies the conclusion that while upfront invasive intervention effectively relieves angina symptoms and improves quality of life, it does not prolong life in patients with stable coronary artery disease.

Study Design

Design

RCT (Extended Follow-Up)

Open-Label

Sample

5,179

Patients

Duration

9.0 yr

Median

Setting

Multicenter, 37 countries

Population Patients with stable chronic coronary disease, preserved left ventricular ejection fraction (≥35%), and moderate-to-severe ischemia on stress testing. Patients with unprotected left main coronary artery disease (≥50% stenosis) or unacceptable angina were excluded.

Intervention Initial invasive strategy: Optimal guideline-directed medical therapy plus routine cardiac catheterization and subsequent revascularization (PCI or CABG) when anatomically feasible.

Comparator Initial conservative strategy: Optimal guideline-directed medical therapy alone, with angiography and revascularization reserved strictly for patients who fail medical therapy (refractory angina, acute coronary syndromes, or heart failure).

Outcome All-cause mortality.

Study Limitations

• The ISCHEMIA-EXTEND phase utilized a pragmatic, highly streamlined observational design focused primarily on all-cause and cause-specific mortality; thus, non-fatal events (such as non-fatal myocardial infarction, heart failure, and repeat revascularizations) and long-term angina status were not systematically collected.

• The study was open-label, which could introduce bias regarding subsequent downstream crossover to revascularization and long-term adherence to guideline-directed medical therapy.

• Ascertaining the definitive cause of death (cardiovascular vs. non-cardiovascular) in a long-term, pragmatic extension phase can be challenging and subject to misclassification.

• Findings cannot be extrapolated to cohorts deliberately excluded from the trial, such as patients with significant left main coronary artery disease (>50% stenosis), severe heart failure (LVEF <35%), recent acute coronary syndromes, or highly symptomatic angina unacceptable to the patient.

Clinical Significance

This 10-year follow-up cements optimal guideline-directed medical therapy as the cornerstone and primary standard of care for stable chronic coronary disease, even in the presence of objectively moderate-to-severe ischemia. It definitively shifts the clinical paradigm: routine upfront coronary angiography and revascularization should be explicitly reserved for symptom relief (angina control) in patients refractory to medical therapy, rather than being deployed as a blanket strategy to prevent death or prolong life.

Historical Context

For decades, the prevailing 'oculostenotic reflex' mandated aggressive revascularization for stable coronary disease under the biological assumption that opening narrowed arteries would invariably prevent myocardial infarction and prolong survival. Landmark trials like COURAGE (2007) and BARI 2D (2009) challenged this view, demonstrating no survival benefit for PCI over medical therapy. However, critics argued these trials failed to select patients with sufficient ischemia or lacked modern imaging and physiologic assessment. The original ISCHEMIA trial (2020) resolved these critiques by enrolling patients with proven moderate-to-severe ischemia while using CCTA to exclude left main disease. Its initial 3.2-year follow-up showed no reduction in major adverse cardiovascular events but confirmed better angina relief. Because the event curves appeared to diverge slightly late in the initial follow-up, the ISCHEMIA-EXTEND study was heavily anticipated to see if a delayed mortality benefit would emerge. The 2026 nearly decade-long results conclusively settle the debate, demonstrating that the initial neutral mortality findings are highly durable.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

How does the pathophysiology of stable coronary artery disease and myocardial ischemia explain why bypassing or stenting a focal stenosis might not improve overall survival compared to optimal medical therapy?

Key Response

In stable CAD, acute myocardial infarctions typically arise from the rupture of vulnerable, non-flow-limiting plaques rather than the gradual progression of stable, flow-limiting plaques. Optimal medical therapy (like statins and antiplatelets) stabilizes all plaques systemically, whereas an invasive strategy only treats a specific, already-stable focal stenosis, explaining the lack of a survival advantage.

Resident

A 65-year-old patient with stable angina and a stress test showing severe ischemia asks if they need a stent to prevent a heart attack. Based on the 9-year ISCHEMIA follow-up, how should you counsel this patient regarding the roles of stents versus medical therapy?

Key Response

The ISCHEMIA trial demonstrated that an initial invasive strategy does not improve overall survival or reduce the risk of all-cause mortality compared to optimal medical therapy in stable CAD. Residents must learn to counsel patients that stenting is primarily for symptom relief (angina reduction) rather than life prolongation or preventing future myocardial infarctions in this specific patient population.

Fellow

Although the overall ISCHEMIA trial follow-up showed no survival benefit for the invasive strategy, how should we interpret the early increase in periprocedural MIs versus the late reduction in spontaneous MIs in the invasive arm, and how does this affect decision-making for patients with high symptom burden?

Key Response

The ISCHEMIA trial showed crossing hazard curves for MI, with periprocedural MIs higher early in the invasive arm, but spontaneous MIs lower over the long term. Fellows need to weigh these competing risks. For a patient with a high symptom burden refractory to OMT, the late reduction in spontaneous MI may provide additional reassurance for an invasive approach, provided the patient understands the upfront procedural risks and the lack of an overall mortality benefit.

Attending

Given the 9-year follow-up data confirming no survival benefit for routine revascularization in stable CAD, how can we systematically restructure our clinical workflows and shared decision-making models to prevent the oculostenotic reflex among practicing interventionalists and referring providers?

Key Response

Changing ingrained clinical habits is difficult. Attendings must lead systemic changes, such as implementing strict appropriateness criteria for cath lab activation, emphasizing OMT optimization first, and utilizing decision aids that clearly delineate that PCI in stable CAD is for symptom control, thereby overcoming the long-standing bias that fixing a blockage saves a life.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The ISCHEMIA trial excluded patients with unacceptable angina, left main disease, and severe heart failure. How do these exclusion criteria introduce selection bias into the survival analysis at the 9-year mark, and what advanced statistical methods could be used to estimate the effect size if these populations were included?

Key Response

The exclusion of high-risk or highly symptomatic patients limits generalizability. Assessing the impact of this selection bias using propensity score matching with observational registries or inverse probability weighting could help researchers model the theoretical treatment effect across the broader spectrum of CAD, refining our understanding of heterogeneous treatment effects.

Journal Editor

In reviewing the 9-year follow-up manuscript of the ISCHEMIA trial, how should editors scrutinize the impact of crossover from the conservative to the invasive arm over nearly a decade, and what sensitivity analyses must be mandated to ensure the intention-to-treat conclusions remain robust?

Key Response

Over a 9-year follow-up, significant crossover is inevitable as patients in the conservative arm develop worsening symptoms or acute events. A rigorous peer review must demand per-protocol and as-treated analyses, alongside time-dependent covariate adjustments, to ensure that the lack of survival difference in the primary intention-to-treat analysis is not merely an artifact of delayed revascularization in the conservative cohort.

Guideline Committee

Current ACC/AHA guidelines emphasize optimal medical therapy as first-line for stable ischemic heart disease (Class 1) but leave room for revascularization to improve survival in certain high-risk subgroups. How does the definitive 9-year null mortality finding from ISCHEMIA prompt a revision of the Class of Recommendation or Level of Evidence for initial revascularization in asymptomatic or mildly symptomatic severe ischemia?

Key Response

The extended follow-up provides Level of Evidence A that routine invasive strategies do not confer a survival advantage in stable moderate-to-severe ischemia (excluding left main disease). Guideline committees must downgrade or remove residual recommendations suggesting a survival benefit for revascularization in these specific stable phenotypes, solidifying OMT as the undisputed cornerstone of management.

Clinical Landscape

Noteworthy Related Trials

2007

COURAGE Trial

n = 2,287 · NEJM

Tested

Percutaneous coronary intervention (PCI) plus optimal medical therapy

Population

Patients with stable coronary artery disease

Comparator

Optimal medical therapy alone

Endpoint

Death from any cause and nonfatal myocardial infarction

Key result: PCI did not reduce the risk of death, myocardial infarction, or other major cardiovascular events compared to optimal medical therapy alone.

2009

BARI 2D Trial

n = 2,368 · NEJM

Tested

Prompt revascularization (PCI or CABG) plus intensive medical therapy

Population

Patients with type 2 diabetes and stable ischemic heart disease

Comparator

Intensive medical therapy alone

Endpoint

All-cause mortality

Key result: There was no significant difference in survival or major cardiovascular events between prompt revascularization and medical therapy alone.

2017

ORBITA Trial

n = 200 · Lancet

Tested

PCI with optimal medical therapy

Population

Patients with severe single-vessel stable coronary artery disease

Comparator

Sham procedure (placebo) with optimal medical therapy

Endpoint

Difference in exercise time increment

Key result: PCI did not significantly increase exercise time or improve angina frequency compared to a sham procedure.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis