Interpreting Myocardial Infarction Analyses in ISCHEMIA

Separating Facts From Fallacy

Raffaele De Caterina; David L. Brown


Eur Heart J. 2021;42(31):2986-2989. 

In This Article

Abstract and Introduction


Graphical Abstract: Hard endpoints in the ISCHEMIA trial. (A) Myocardial infarction according to the primary definition in the trial. (B) Prognostically relevant myocardial infarctions according to the criteria of the third and fourth Universal Definition of Myocardial Infarction. (C) All-cause deaths. CON, conservative arm; INV, invasive arm. From refs.[1,15]


After a decade of planning and execution, the International Study of Comparative Health Effectiveness with Medical and Invasive Approaches (ISCHEMIA) trial was published in April 2020.[1] This landmark study found that in patients with chronic coronary syndromes (CCS) both the composite primary endpoint [cardiovascular death, myocardial infarction (MI), hospitalization for unstable angina or heart failure, or resuscitated cardiac arrest] and secondary endpoint (cardiovascular death and MI) did not differ between invasive and conservative strategies, refuting the long-held belief in a prognostic benefit from revascularization by percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) surgery among patients with CCS who had significant inducible myocardial ischaemia at baseline.[1] In the invasive group of ISCHEMIA, 79% of patients underwent revascularization (PCI in 74%; CABG in 26%), whereas 21% did not. Periprocedural (Types 4a and 5) and late procedure-related MIs (Types 4b and 4c) were more frequent in the invasive arm, whereas late, spontaneous MIs were more common in the conservative arm. Although periprocedural MI was not associated with increased all-cause or cardiovascular mortality, late procedure-related MI was associated with a nearly four- and seven-fold increased risk in death and cardiovascular death, respectively. Type 1 MI was associated with a 2.4- and 3.3-fold increase in all-cause and cardiovascular death. On this basis, some have concluded that early revascularization should be offered to patients with CCS to prevent 'prognostically important' infarctions, i.e. spontaneous MIs while ignoring periprocedural and late procedure-related MIs (Graphical abstract). This commentary seeks to provide a rebuttal to this speculative conclusion by considering both biological plausibility and alternative evidence-based explanations, namely that (i) 'procedural' MIs are not innocuous and (ii) the higher mortality associated with revascularization-related (Types 4a, 4b, 4c, and 5) MI completely counterbalances the alleged mortality reduction attributed to a lower rate of Type 1 MI, resulting in net neutrality of total death rates.