Are We At Risk of Depriving Patients Lifesaving Cardiac Surgery?

Implications of the ISCHEMIA Trial for Coronary Artery Bypass Graft Surgery

John H. Alexander, MD, MHS; Peter K. Smith, MD


Circulation. 2020;142(19):1797-1798. 

The ISCHEMIA trial (International Study of Comparative Health Effectiveness with Medical and Invasive Approaches) is one of the most anticipated, important, and complicated clinical trials ever conducted.[1] Its objective was to answer whether, in stable patients with at least moderate ischemia on stress testing, there is a benefit to cardiac catheterization and, if feasible, revascularization on top of optimal medical therapy. Before randomization, patients underwent computed tomography (CT) angiography to rule out significant left main coronary artery disease. In patients who were randomly assigned to cardiac catheterization, the choice of revascularization, with percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) surgery, was left to the treating physician. Of those who were randomly assigned to catheterization, 96% underwent angiography and 79% underwent revascularization, 74% with PCI and 26% with CABG surgery.[1] The primary outcome of ISCHEMIA was a composite of cardiovascular death, myocardial infarction, hospitalization for unstable angina or heart failure, or resuscitated cardiac arrest. Important secondary outcomes included cardiovascular death or myocardial infarction and quality of life, including the extent of angina. With an average of 3.2 years of follow-up, the trial demonstrated no benefit of catheterization (adjusted hazard ratio 0.93 [95% CI, 0.80–1.08]) on the primary outcome beyond optimal medical therapy alone.[1] Catheterization did reduce angina and there was a numeric (1.8% absolute) early increase in mostly periprocedural myocardial infarction and a later numeric (1.6% absolute) decrease in spontaneous myocardial infarction with catheterization in comparison with optimal medical therapy.[1]

What do the results of ISCHEMIA mean for the appropriate evaluation of patients for CABG surgery? In ISCHEMIA, more than two-thirds of patients had at least monthly angina, 45% had diabetes, and most had near-normal left ventricular function. Although patients with significant left main coronary disease on CT angiogram were excluded, 45% had >50% stenosis of 3 coronary vessels and 45% had >50% stenosis of the proximal left anterior descending artery. We know from clinical trials in the 1980s that CABG surgery is superior to the medical therapy of that time in patients with multivessel coronary artery disease. From more recent trials, we know that CABG surgery is superior to PCI in patients with complex multivessel coronary disease (as assessed by SYNTAX [Synergy Between Percutaneous Coronary Intervention With Taxus and Cardiac Surgery] score), those with concomitant diabetes, and in patients with left ventricular systolic dysfunction.[2–5]

The ISCHEMIA trial failed to demonstrate a benefit of catheterization on clinical events; however, ISCHEMIA evaluated a diagnostic strategy, and then combined 2 entirely different therapeutic interventions (PCI and CABG) and compared them with optimal medical therapy alone. Any effect (benefit or harm) of PCI or CABG might be obscured by the effects of the other procedure.

Subgroup analyses of the ISCHEMIA trial suggest that patients with more extensive coronary disease and with proximal left anterior descending artery disease tended (although not statistically significant) to do better with an invasive approach.[1] It will be important to understand the clinical and angiographic characteristics of the patients in ISCHEMIA who did and did not undergo CABG surgery. It will also be tempting to look at the outcomes of patients who underwent CABG; however, the ISCHEMIA trial was not designed to evaluate CABG versus optimal medical therapy. There is no appropriate randomized control group for those who underwent CABG surgery. The only possible analysis would be a post hoc, underpowered, nonrandomized, propensity-matched, postbaseline subgroup analysis with probably too short a period of follow-up to detect any benefit of CABG surgery.

The results of ISCHEMIA must be incorporated into the context of what we already know regarding the benefits of PCI and CABG surgery in patients with ischemic heart disease. To implement the results of ISCHEMIA in practice, patients will need to have a CT angiogram to exclude significant left main coronary artery disease. Beyond left main disease, this CT angiogram should also be used to exclude surgical coronary artery disease. We should be offering CABG surgery to appropriately selected patients with complex 3-vessel coronary artery disease or 3-vessel coronary artery disease with concomitant diabetes or left ventricular dysfunction.