Validation of acute myocardial infarction (AMI) cut-off point for troponin I on the ADVIA Centaur™ assay

The preferred biomarker for myocardial damage is cardiac troponin. Any increased value exceeding the 99^th percentile of a reference control group (by an assay with CV < 10%) in the appropriate clinical setting is associated with acute myocardial infarction. We evaluated an automated platform, ADVIA Centaur™. An essential step in the evaluation of troponin I level is to establish a clinically relevant cut-off point for AMI diagnosis. Four hundred sixty-nine heparinized plasma samples were drawn from 161 patients presenting with symptoms suggestive of myocardial infarction. Part 1: In 80 patients, troponin I was measured on two different analytical systems (Abbott AxSYM™ and Bayer ADVIA Centaur™). Clinical diagnosis of AMI was used as the 'gold standard' for the AMI. Results were compared using Receiver Operator Characteristic (ROC) curve analysis. Part 2: To determine the clinical sensitivity and specificity within different time intervals after presentation, an additional 81 patients, each with a series of at least 3 samples, were analyzed by ADVIA Centaur™ Assay. Twenty-six patients with undetermined clinical diagnosis were excluded from the analysis (17 in part 1 and 9 in part 2). Part 1: Of 63 subjects, 16 (25.4%) had AMI, 47 (74.6%) did not. Based on our data, 0.77 ng/ mL (plasma) cut-off point in the Centaur assay optimizes both sensitivity and specificity (100% & 98% respectively). Our previously established cut-off in use by the AxSYM™ is 2.5 ng/mL (sensitivity and specificity of 94% & 81%). Part 2: Of 72 subjects, 53 (73.6%) had AMI, 19 (26.4%) did not. ROC analysis of cTnI at time intervals after presentation (T1=0, T2=>0 to <3, T3=3 to <6, T4=6 to <12, T5=>12 hours) showed maximum efficiency at cut-off points of 0.20, 0.51, 0.61 and 0.93 ng/mL respectively. We conclude that in plasma, 0.77 ng/mL is the best cut-off for the diagnosis of AMI based on the established clinical diagnosis. The precision is adequate for our clinical application in the range of the cut-off.