Background: Creatine kinase-MB (CK-MB) elevation after percutaneous coronary intervention (PCI) has been associated with increased risk for mortality. Although most studies have defined periprocedural myocardial infarction (pMI) as an elevation in CK-MB >3× upper limit of normal (ULN), use of different CK-MB assays and variation in site-specific definitions of the ULN may limit the value of such relative thresholds.
Methods and results: We used data from the multicenter Evaluation of Drug-Eluting Stents and Ischemic Events (EVENT) registry to examine the impact of variations in site-specific thresholds for CK-MB elevation on the incidence of pMI as well as the relationship between absolute peak levels of CK-MB after PCI and 1-year mortality. The study cohort consisted of 6347 patients who underwent nonemergent PCI and had normal CK-MB at baseline. Across the 59 study centers, the ULN for CK-MB ranged from 2.6 to 10.4 ng/mL (median, 5.0 ng/mL), and there was an inverse relationship between the site-specific ULN and the incidence of pMI (defined as CK-MB elevation >3× ULN). Although any postprocedure elevation of CK-MB was associated with an adverse prognosis, in categorical analyses, only CK-MB ≥50 ng/mL was independently associated with increased 1-year mortality (hazard ratio, 4.71; 95% confidence interval, 2.42 to 9.13; P<0.001). Spline analysis using peak CK-MB as a continuous variable suggested a graded, nonlinear relationship with 1-year mortality, with an inflection point at ≈30 ng/mL.
Conclusions: Among unselected patients undergoing PCI, there is a graded relationship between CK-MB elevation after PCI and 1-year mortality that is particularly strong for large CK-MB elevations (>30 to 50 ng/mL). Future studies that include pMI as a clinical end point should consider using a core laboratory to assess CK-MB (to ensure consistency) and raising the threshold for defining pMI above current levels (to enhance clinical relevance).