Cardiovascular Magnetic Resonance in Nonischemic Myocardial In fl ammation Expert Recommendations

This JACC Scienti fi c Expert Panel provides consensus recommendations for an update of the cardiovascular magnetic resonance (CMR) diagnostic criteria for myocardial in fl ammation in patients with suspected acute or active myocardial in fl ammation (Lake Louise Criteria) that include options to use parametric mapping techniques. While each parameter may indicate myocardial in fl ammation, the authors propose that CMR provides strong evidence for myocardial in fl ammation, with increasing speci fi city, if the CMR scan demonstrates the combination of myocardial edema with other CMR markers of in fl ammatory myocardial injury. This is based on at least one T2-based criterion (global or regional increase of myocardial T2 relaxation time or an increased signal intensity in T2-weighted CMR images), with at least one T1-based criterion (increased myocardial T1, extracellular volume, or late gadolinium enhancement). While having both a positive T2-based marker and a T1-based marker will increase speci fi city for diagnosing acute myocardial in fl ammation, having only one (i.e., T2-based OR T1-based) marker may still support a diagnosis of acute myocardial in fl ammation in an appropriate clinical scenario, albeit with less speci fi city. The update is expected to improve the diagnostic accuracy of CMR further in detecting myocardial in fl ammation. (J Am Coll Cardiol 2018;72:3158 – 76) © 2018 Published by Elsevier on behalf of the American College of Cardiology Foundation. The original Lake Louise Criteria I (left-most any out are supported by the largest but heterogeneous body of evidence. The evidence on other cardiac magnetic resonance (CMR) criteria combinations is also heterogeneous and much sparser. Nevertheless, the data suggest that novel mapping techniques may offer at least theoretical advantage over the original Lake Louise Criteria, pending further validation. The estimated area-under-the-curve (AUC*) is calculated as the average of the sensitivity and speci city reported for each in published studies. The bubbles tend to rise toward higher values of estimated area-under-the-curve for combinations that include mapping for at least study sample