Myeloperoxidase in the diagnosis of acute coronary syndromes: The importance of spectrum

Background: Myeloperoxidase (MPO) is proposed for risk stratification in patients with suspected acute coronary syndromes (ACSs). We determined if MPO has diagnostic value in patients being evaluated for ACS. Method: MIDAS was an 18-center prospective study enrolling suspected ACS emergency department patients who presented <8 hours after symptom onset and in whom serial cardiac markers and objective cardiac perfusion testing were planned. Blinded MPO (Biosite, Inc, San Diego, CA) and troponin I (Triage Cardio 3; Biosite, Inc) were drawn at arrival, and Troponin I (TnI) was measured at 90, 180, and 360 minutes. Final diagnoses were adjudicated by the local investigator blinded to study assay. Results: Of 1,018 patients, 54% were male, 26% black, with a mean age of 58 ± 13 years. Diagnoses were ACS in 288 (23%) and noncardiac chest pain (NCCP) in 788 (77%). Of patients with ACS, 94 (9.2%) had a myocardial infarction (MI) at presentation (69 non-ST-elevation MI, 25 ST-elevation MI), and 136 had unstable angina. Using a cutpoint of 210 ng/mL to provide 90% specificity, MPO had a sensitivity of 0.18; negative predictive value, 0.69; positive predictive value, 0.47; negative likelihood ratio, 0.91; and a positive likelihood ratio of 1.83 to differentiate ACS and NCCP. Because of the large overlap of quartiles, MPO was not clinically useful to predict serial TnI changes. The C statistics ± 95% CI for MPO differentiating ACS from NCCP and for AMI versus NCCP were 0.629 ± 0.04 and 0.666 ± 0.06, respectively. Conclusions: Myeloperoxidase has insufficient accuracy for decision making in patients with suspected ACS.