High sensitive cardiac troponin T: Testing the test

Abstract

Background High sensitive cardiac troponin T (hs-TnT) found its way into everyday clinical routine to diagnose acute myocardial infarction (AMI). However, its levels vary considerably based on the underlying pathophysiology of the patients. Hence we sought to test the applicability of the currently only available hs-TnT assay (Roche Diagnostics, Switzerland) to diagnose acute myocardial infarction. Methods and patients Retrospectively, we analyzed the hs-TnT results of 1573 patients admitted to a level A university hospital emergency department. Overall 323 patients had an acute cardiac event defined as Non-ST Elevated Myocardial Infarction (NSTEMI) and 286 patients had a ST-Elevated Myocardial Infarction (STEMI). 964 patients served as controls, consisting of patients with other cardiac and non-cardiac morbidity. Results The sensitivity of hs-TnT for detecting an acute cardiac event was more than 92% overall. The specificity varied around 35% depending on the respective patient cohort. ROC curve analysis of the initial hs-TnT results showed that the AUC in total cardiac events (STEMI and NSTEMI) was 0.81. Detailed analysis resulted in an AUC of 0.79 in NSTEMI and 0.84 in STEMI patients detected via the initial hs-TnT. We further tested the ESC algorithm for detecting NSTEMI and obtained a sensitivity of about 83%, while 43% of all non-NSTEMIs are classified as NSTEMIs. Conclusion We show that the specificity of hs-TnT for AMI is very low and conclude that the current assay including its delta values represents myocardial damage of any origin. This damage alone does not substantiate an AMI diagnosis even when international algorithms are applied. © 2016