Has the frequency of peri-procedural myocardial infarction reduced using third universal definition?

Abstract

Background: Peri-procedural myocardial infarction (MI) in the 2012 universal definition requires, in addition to cardiac biomarker elevations, ischaemic chest pain and/or abnormal findings on ECGs or imaging. As MI, including post percutaneous coronary intervention (PCI) is an important endpoint in clinical trials and as post-PCI Troponin T (TnT) elevations are prognostic in patients with stable coronary heart disease (CHD), we determined whether the frequency of post-PCI MI was reduced, and whether it was associated with increased rates of late death/MI. Methods and results: To compare 2007 and 2012 peri-procedural MI definitions, we evaluated TnT levels within 24 hours post-PCI and assessed other criteria in747 patients undergoing PCI, for stable-CHD (18%), for non-ST elevation MI (43%) and for unstable angina (40%). The patients' mean age was 64±12years, 75% were male, 28% had diabetes and 60% had ACC/AHA class B2/C lesions. Pre-PCI TnT levels were normal ( 5xURL, 89 (18%) had levels >3xURL. Patients with elevated pre-PCI TnT levels, 88 (35%) had >20% elevation in TnT levels post-PCI. Among patients with an increase in post-PCI TnT levels to >5xURL or had increase by > 20% above elevated pre-PCI TnT levels (using third universal MI definition), the frequencies of both ischaemic chest pain and ECG changes was (only)7%. Peri-procedural MI occurred in 177 (24%) patients had using 2007 definition whereas 142 (19%) had post-PCI TnT levels either >5xURL or > 20% increased above elevated pre-PCI levels (p<0.0001). At median follow-up of 37 [IQR: 17-52] months, rates of death/MI in patients with stable-CHD were 28% and 10% in those with, and without, post-PCI TnT levels >5xURL, respectively (p=0.05), whereas these event rates in patients with ACS who did, and did not, have post-PCI TnT >5xURL or >20% above elevated pre-PCI TnT levels were 26% and 24% (p=0.58). Conclusion: The diagnosis of peri-procedural MI was less frequent when the 2012 universal definition was used compared to the 2007 definition. This has implications for clinical trial endpoint frequency and thus trial costs. Peri-procedural MI using the 2012 definition was prognostic in patients with stable-CHD.