Redefinition of myocardial infarction after cardiac surgery: a two-center clinical evaluation of the proposed european association for cardio-thoracic surgery algorithm for the diagnosis of perioperative myocardial injury and infarction
Maaike S Y Thio, Peter G Noordzij, Henk J T Ruven, Ted Reniers, Jurriën M Ten Berg, Rohit M Oemrawsingh, Marloes L P Langelaan, Thijs C D RettigAbstract
Objectives
This study evaluated the clinical applicability and relevance of the European Association for Cardio-thoracic surgery (EACTS) algorithm for the diagnosis of perioperative myocardial injury and myocardial infarction (PMI) after cardiac surgery, compared to the Fourth Universal Definition of Myocardial Infarction (4UD).
Methods
This was a prospective two-center cohort study in adults undergoing a range of elective and urgent cardiac surgeries. PMI and myocardial injury were diagnosed using both the EACTS algorithm and the 4UD. Primary outcome was to compare PMI and myocardial injury incidences between definitions. One-year mortality hazard ratios (HR) adjusted for EuroSCORE II were calculated for perioperative biomarker elevation, myocardial injury and PMI versus no myocardial injury.
Results
1,142 patients were included. Most common procedures were coronary artery bypass grafting (53.0%) and isolated valve surgery (24.4%). The EACTS algorithm diagnosed 2.5% (n = 28) of patients as PMI (versus 2.6%, n = 30 with 4UD), 29.9% (n = 341) as myocardial injury (versus 87.2%, n = 996 with 4UD), 37.6% (n = 429) as perioperative biomarker elevation and 30.1% (n = 344) as no myocardial injury (versus 10.2%, n = 116 with 4UD). Adjusted HRs for one-year mortality with the EACTS algorithm were 12.3 (95% CI 2.0–74.7) for PMI (versus 13.7 (95% CI 1.5–124.6) with 4UD), 3.6 (95% CI 0.8–16.6) for myocardial injury (versus 2.3 (95% CI 0.3–17.0) with 4UD) and 2.7 (95% CI 0.6–12.6) for perioperative biomarker elevation.
Conclusions
Implementation of the EACTS algorithm led to a similar incidence of PMI but substantially reduced myocardial injury incidence compared to the 4UD, while maintaining its clinically relevant association with mortality.