Diagnostic value of circulating microRNAs compared to high-sensitivity troponin T for the detection of non-ST-segment elevation myocardial infarction

Aims To assess the diagnostic value of microRNAs (miRNAs) for the detection of non-ST-segment elevation myocardial infarction (NSTEMI). Methods and results A total of 1042 patients presenting between August 2014 and April 2017 to the emergency department with the suspected acute coronary syndrome were included. Non-ST-segment elevation myocardial infarction was diagnosed per criteria of the fourth Universal definition of myocardial infarction (UDMI) using high-sensitivity troponin T (hs-cTnT). Expression levels of eleven microRNAs (miR-21, miR-22, miR-29a, miR-92a, miR-122, miR-126, miR-132, miR-133, miR-134, miR-191, and miR-423) were determined using RT-qPCR. Discrimination of NSTEMI was assessed for individual and a panel of miRNAs compared to the hs-cTnT reference using C-statistics and reclassification analysis. NSTEMI was diagnosed in 137 (13.1%) patients. The area under the curve (AUC) of the hs-cTnT based reference was 0.937. In a multivariate model, three miRNAs (miR-122, miR-133, and miR-134) were found to be associated with NSTEMI with AUCs between 0.506 and 0.656. A panel consisting of these miRNAs revealed an AUC of 0.662 for the diagnosis of NSTEMI. The AUC of the combination of the miRNA panel and troponin reference was significantly lower than the reference standard (AUC: 0.897 vs. 0.937, P=0.006). Despite a significant improvement of NSTEMI reclassification measured by IDI and NRI, miRNAs did not improve the specificity of hs-cTnT kinetic changes for the diagnosis of NSTEMI (Delta AUC: 0.04). Conclusion Although single miRNAs are significantly associated with the diagnosis of NSTEMI a miRNA panel does not add diagnostic accuracy to the hs-cTnT reference considering baseline values and kinetic changes as recommended by the fourth version of UDMI. [GRAPHICS]

Automated summary

This study evaluated the diagnostic value of 11 microRNAs for NSTEMI, finding that a panel of three microRNAs could be associated with NSTEMI diagnosis, but did not improve diagnostic accuracy compared to the hs-cTnT reference standard.