Optimal cut-off value for cardiac troponin I in ruling out Type 5 myocardial infarction

Abstract

The clinical classification of myocardial infarction (MI) into five types was introduced in 2007 as a component of the universal definition. A Type 5 MI was defined as a MI related to coronary artery bypass surgery. In a setting of patients undergoing elective coronary artery bypass grafting, we set out (i) to describe the pattern of multiple serial cardiac troponin I (cTnI) measurements within 72 h postoperatively and (ii) to determine the optimal cardiac troponin I cut-off value in ruling in or ruling out a Type 5 MI. In 2011-2012, patients with two- and three-vessel disease scheduled for elective on-pump coronary artery bypass grafting were considered. Samples for cTnI were drawn before and 0, 2, 4, 6, 12, 24, 48 and 72 h after surgery. Analysis for cardiac troponin I was performed by use of the Abbott Architect c16000 system with an upper reference limit (URL) of 30 ng/l. The diagnosis of a Type 5 MI was prospectively made by a consultant cardiologist and was based on clinical, electrocardiographic and imaging data together with routine sampling and measurements of cTnI, but without knowledge of the results of serial study cTnI measurements. Of the 141 eligible patients, 99 (70%) qualified for final enrollment. In 8 patients (8%), the clinical diagnosis of a Type 5 MI was made. Patients without Type 5 MI (n = 91) had a median cTnI peak value of 7675 ng/l compared with 20 500 ng/l in Type 5 MI patients (P = 0.01). By use of receiver operating characteristic curves, optimal cut-off values for identifying Type 5 MI were defined as 7970 ng/l (corresponding to 266 times the URL) 12 h postoperatively and 9950 ng/l (corresponding to 331 times the URL) 24 h postoperatively. These cut-off values resulted in negative predictive values of 0.99 (12 h) and 0.99 (24 h). Positive predictive values were 0.23 (12 h) and 0.35 (24 h). In clinically stable patients undergoing elective coronary artery bypass grafting, measurements of cTnI are useful in ruling out a Type 5 MI.