ISCHEMIA Trial: Was It Worth the Wait?

Rasha Al-Lamee, MBBS, MA, PhD; Alice K. Jacobs, MD


Circulation. 2020;142(5):517-519. 

In This Article

Design, Conduct, and Results of the Ischemia Trial

The largest clinical trial in SIDH was designed to answer the following question: Is there a benefit of cardiac catheterization and, if feasible, revascularization in addition to guideline-directed medical therapy (GDMT) in stable patients with at least moderate ischemia on a stress test? Notable potential limitations of previous trials were considered. Selection bias was reduced by randomization before knowledge of coronary anatomy. Most patients underwent blinded computed tomographic coronary angiography adjudicated by a core laboratory that excluded patients with significant left main stem disease or unobstructed coronary arteries (<50% stenosis in patients undergoing stress imaging, <70% stenosis in patients undergoing exercise treadmill testing). Meticulous measures of adherence were introduced to improve attainment of optimal medical therapy. Patients were excluded if they had unacceptable angina despite GDMT, were dissatisfied with medical management of angina, or had a history of medication nonadherence to minimize crossover from the conservative to the invasive arm (notably the trial was analyzed by intention to treat).

This truly international trial was conducted in 38 countries and recruited >5000 patients. The investigators prepublished the trial design, explanations for protocol changes, baseline clinical characteristics, and an interim report on medication adherence and treatment goal attainment ahead of the primary result presentation. The openness and inclusivity of this approach set a new standard for trial reporting.

ISCHEMIA showed that, at a median follow-up of 3.2 years, in patients with SIHD and moderate to severe ischemia, an initial invasive strategy did not reduce the risk of the primary end point of cardiovascular death, MI, hospitalization for unstable angina, heart failure, or resuscitated cardiac arrest compared with a conservative strategy. There was also no difference in the major secondary end point of cardiovascular mortality or MI, although there were fewer spontaneous MIs in the invasive arm. A secondary analysis showed that the incidence of the primary outcome was sensitive to the definition used for MI. When analyzed using the secondary definition of MI, the procedural MI rate increased; the clinical importance of this remains unknown. In addition, albeit in an unblinded trial, an invasive strategy led to a greater improvement in symptom control and quality of life.

As is often necessary in strategy trials, several changes to design were made during the enrollment phase to ensure adequate sample size and end-point events. These included making minimal changes in the inclusion criteria for the definition of at least moderate ischemia, allowing exercise tolerance testing as an entry test, balancing a reduction in sample size from 8000 to 5000 with prolongation of the follow-up phase to allow adequate event accrual, and expanding the ideal 2-component primary end point of cardiovascular mortality and MI to the wider prespecified 5-component primary end point. Debates will continue as to how much these changes influenced the results. It is noteworthy that the change in the primary end point had no effect on the main conclusions because rates of hospitalization for unstable angina, heart failure, and resuscitated cardiac arrest were low and group comparisons showed similar results for the primary and major secondary end points.