Periprocedural myocardial injury in the EXCEL trial

This commentary refers to ‘Impact of large periprocedural myocardial infarction on mortality after percutaneous coronary intervention and coronary artery bypass grafting for left main disease: an analysis from the EXCEL trial’, by O. Ben-Yehuda et al., 2019;40:1930–1941. The association between periprocedural myocardial infarction (PMI) and mortality after percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) for left main disease was presented in the European Heart Journal recently.1 In this study, the authors have analysed data from the EXCEL trial and concluded that PMI was strongly associated with increased 3-year mortality.2 Creatine kinase-MB levels were used to quantify myocardial injury and are presented in five strata normalized to the upper reference level (URL), the upper being ≥10× URL. The mortality-plots in Figure 4 demonstrate the risk of death following PCI or CABG in four strata of creatine kinase-MB. Logically, myocardial injury leading to early post-operative death would be associated with higher creatine kinase-MB levels than 10× URL. The distribution of creatine kinase-MB levels is not reported, and we are unable to find shared data. The combined endpoint of death, stroke, or myocardial infarction (MI) (both PMI and subsequent spontaneous or procedural MI) was in the EXCEL trial not significantly different at 5 years.2 All-cause mortality at 5 years, however, was higher in the PCI group. Periprocedural myocardial infarction was higher in the CABG group and contributes as a component of the combined endpoint to the conflicting results between all-cause mortality and the combined endpoint. Hence, the definition and implications of PMI are important for the interpretation of the EXCEL trial. In the EXCEL trial, the authors discuss that the definition of MI used is at odds with both the prior and current universal definitions of MI and the authors state that PMI ‘was independently predictive of late death’. Although this seems to be true for PCI (Figure 4b), for CABG this association is due to early mortality (Figure 4c). To better understand the implications and long-term prognostic value of PMI following CABG, we encourage the authors to share data, in particular the distribution of creatine kinase-MB levels and the values associated with CABG, PCI, early and late mortality, respectively. Funded by Oslo University Hospital (employer). Conflict of interest: none declared.