Diagnosis of acute myocardial infarction: super-sensitivity vs. high-sensitivity cardiac troponin I assay

Abstract

Purpose: The favorite biomarker in diagnosing acute myocardial infarction (AMI) is cardiac troponin. Allowing troponin detection in >50% of the presumably healthy population high-sensitivity cardiac troponin assays have been established to improve accuracy of AMI diagnosis. Recently, detecting troponin in >95% of the general population, super-sensitivity cardiac troponin assays have been developed. The clinical utility of these troponin assays for the diagnosis of AMI is unclear. Aim of the study was to directly compare diagnostic performance of a super-sensitivity cardiac troponin assay with a high-sensitivity cardiac troponin assay in patients with suspected AMI. Methods: 1389 patients with new-onset chest pain were included in the study. 934 patients were diagnosed with non-coronary chest-pain, 184 with unstable angina pectoris, 271 with AMI. Cardiac troponin I (cTnI) concentration was measured on admission (0h) and 3 hours later (3h) by a high-sensitivity (h-s) assay (ARCHITECT STAT highly sensitive Troponin I immunoassay, Abbott Diagnostics, Abbott Park, IL, USA) and a super-sensitivity (s-s) assay (Erenna® cTnI immunoassay, Singulex, Inc., Alameda, CA, USA). As diagnostic cutoff concentration (P99) assay- and gender specific 99th percentiles, derived from the same reference population, were used. The diagnostic algorithm incorporated cTnI elevation (>P99) at either time point and predefined relative 3-h changes in cTnI concentration (ΔcTnI). Accordingly, for rule-in AMI, patients were divided in two subgroups: those with cTnI (0h) >P99 + ΔcTnI and that with cTnI (0h) ≤P99 and cTnI (3h) >P99 + ΔcTnI. Results: Rule-out AMI was achieved by a cTnI (3h) ≤P99 with negative predictive values of 98.6% (s-s assay) and 99.5% (h-s assay; P=0.008) and with specificities of 95.1% (s-s assay) and 88.3% (h-s assay; P P99, rule-in AMI via ΔcTnI ≥100% resulted in sensitivities of 53.8% (s-s assay) and 50.2% (h-s assay; P=0.25) and a positive predictive value (PPV) of 95% for both assays. In patients with cTnI (0h) ≤P99, cTnI (3h) >P99 and assay-specific ΔcTnI (≥100% s-s assay, ≥250% h-s assay) yielded sensitivities of 85.9% (s-s assay) and 90.9% (h-s assay; P=0.24) and PPVs of 83.9% (s-s assay) and 83.3% (h-s assay). Conclusions: Using assay-specific relative changes of cTnI concentration to rule-in AMI in patients with new-onset chest pain, the diagnostic performance of both assays were found to be equivalent. The direct comparison of super- and high-sensitivity troponin assay shows only minor differences in ruling-out AMI.