T1 mapping cardiovascular magnetic resonance imaging to detect myocarditis-Impact of slice orientation on the diagnostic performance

Background: T1 mapping is a promising diagnostic tool to improve the diagnostic accuracy of cardiovascular magnetic resonance (CMR) in patients with suspected myocarditis. However, there are currently no data on the potential influence of slice orientation on the diagnostic performance of CMR. Thus, we compared the diagnostic performance of global myocardial T1 and extracellular volume (ECV) values to differentiate patients with myocarditis from healthy individuals between different slice orientations. Methods: This study included 48 patients with clinically defined myocarditis and 13 healthy controls who underwent CMR at 1.5 T. A modified Look-Locker inversion-recovery (MOLLI) sequence was used for T1 mapping before and 15 min after administration of 0.075 mmol/kg Gadolinium-BOPTA. T1 mapping was performed on three short and on three long axes slices, respectively. Native Tl, post-contrast T1 and extracellular volume (ECV)-BOPTA maps were calculated using a dedicated plug-in written for the OsiriX software and compared between the mean value of three short-axes slices (3SAX), the central short-axis (1SAX), the mean value of three long-axes slices (3LAX), the four-chamber view (4CH), the three-chamber view (3CH) and the two-chamber view (2CH). Results: There were significantly lower native T1 values on 3LAX (1081 ms (1037-1131 ms)) compared to 3SAX (1107 ms (1069-1143 ms), p = 0.0022) in patients with myocarditis, but not in controls (1026 ms (1009-1059 ms) vs. 1039 ms (1023-1055 ms), p = 0.2719). The areas under the curve (AUC) to discriminate between myocarditis and healthy controls by native myocardial T1 were 0.85 (p < 0.0001) on 3SAX, 0.85 (p < 0.0001) on 1SAX, 0.76 (p = 0.0002) on 3LAX, 0.70 (p = 0.0075) on 4CH, 0.72 (p = 0.0020) on 3CH and 0.75 (p = 0.0003) on 2CH. The AUCs for ECV-BOPTA were 0.83 (p <0.0001) on 3 SAX, 0.82 (p <0.0001) on 1SAX, 0.77 (p = 0.0005) on 3LAX, 0.71 (p = 0.0079) on 4CH, 0.69 (p = 0.0371) on 3CH and 0.75 (p = 0.0006) on 2CH. Conclusion: Native T1 and ECV-BOPTA on short axes slices provide a better diagnostic performance in myocarditis than long axes slices since long axes slices seem to underestimate native myocardial T1 in myocarditis. T1 mapping in suspected myocarditis can be restricted to a single mid-ventricular short-axis slice without a significant loss in diagnostic performance. (C) 2016 Elsevier Ireland Ltd. All rights reserved.