Role of Cardiac Magnetic Resonance Imaging and TroponinT in Definitive Diagnosis of Myocardial Infarction With Nonobstructive Coronary Arteries (MINOCA).

Abstract

It is unknown whether the degree of high-sensitivity troponin T (hsTropT) elevation in patients with suspected myocardial infarction without obstructive coronary arteries (MINOCA) presentations can help predict the likelihood of an abnormal cardiac magnetic resonance (CMR) scan. In this study we describe the diagnostic utility of CMR in patients with MINOCA and assesses the effect of peak hsTropT levels at presentation on CMR diagnostic yield. Records of consecutive patients (n= 1407) referred for CMR at a tertiary referral hospital between January 2016 and September 2021 were reviewed. A total of 70 patients met the criteria of MINOCA including ischemic chest pain, elevated peak hsTropT, and nonobstructive coronary artery disease (< 50% stenosis). The peak hsTropT levels within 72 hours of admission were identified. CMR images were generated using a 3.0 T Siemens scanner. Predictors of having an abnormal CMR were evaluated. CMR established a diagnosis in 71% (n= 50) of patients, with the most common CMR diagnosis being myopericarditis (n= 27; 39%). Time to CMR was an independent predictor of a normal CMR scan (odds ratio, 0.98; 95% confidence interval, 0.97-0.999). Peak hsTropT had a high diagnostic accuracy for identifying patients with an abnormal CMR scan (area under the receiver operator characteristic curve, 0.81; P < 0.001). The optimal hsTropT cutoff was 166 ng/L, with 72% sensitivity and specificity. A troponin value ≥ 166 ng/L was independently predictive of an abnormal CMR scan (odds ratio, 4.76; 95% confidence interval, 1.32-17.11). HsTropT and early CMR imaging are independently predictive of an abnormal CMR scan in patients with MINOCA. Additionally, the use of a hsTropT cutoff provides incremental predictive value to clinical parameters and time to CMR scanning in determining an abnormal scan.