Cardiovascular magnetic resonance imaging after normal echocardiography in myocardial infarction with nonobstructed coronary arteries

Background In patients with myocardial infarction with nonobstructive coronary arteries (MINOCA), cardiovascular magnetic resonance imaging (CMR) can often establish a causal diagnosis. It is unknown if CMR is warranted in patients with normal echocardiographic findings, or if measurement of high-sensitive troponin T (hs-TnT) and NT-pro-BNP can be of value in selecting patients for further imaging. Objectives To assess the diagnostic accuracy of echocardiography and hs-TnT and NT-pro-BNP for identifying patients receiving any diagnosis using CMR. Methods We included patients (n = 123) from the SMINC2 study who underwent same day CMR and echocardiography, at a median of 3 days after hospital admission for MINOCA. Normal echocardiography was defined as left ventricular ejection fraction ≥ 55%, absolute global longitudinal strain ≥ 17%, E/e′ ≤ 14, and no regional wall motion abnormalities. Logistic regression models were fitted to assess the probability of CMR diagnosis at increasing levels of hs-TnT and NT-pro-BNP. Results Of patients with a normal echocardiographic examination, 23/33 (70%) received a diagnosis using CMR. Pathological echocardiography identified patients with a diagnosis using CMR with a sensitivity of 77%, specificity 38%, positive predictive value 82%, and negative predictive value 30%, respectively. There was no level of hs-TnT or NT-pro-BNP below which a CMR diagnosis could be reliably excluded. Conclusions The majority of patients with MINOCA and a normal echocardiogram will receive a diagnosis by CMR. A CMR diagnosis was common even among patients with low levels of biomarkers. CMR should be recommended regardless of echocardiographic findings and hs-TnT or NT-pro-BNP levels in patients with MINOCA.