Non-invasive differentiation of idiopathic inflammatory myopathy with cardiac involvement from acute viral myocarditis using cardiovascular magnetic resonance imaging T1 and T2 mapping

Fig. 1. Mapping parameters estimation in the skeletal muscles. As a first step, visible skeletal muscles were delineated on the first TI image after setting the adequate gray levels. Second, each muscle region of interest (ROI) was propagated on all the following TI images. Third, a zoom was performed around each propagated ROIs and gray levels were adjusted to allow their reshaping and displacement in order to avoid fat and vessels as well as to adapt to muscle deformation when necessary. Finally, for each ROI, gray levels were averaged for successive TI images and the resulting signal is fitted with an exponential model to calculate native T1 in the above example. Of note, this process is performed first on native T1 dataset and the resulting ROIs are copied on post-contrast T1 and T2 datasets and the correction process is repeated

Fig. 2. Myocardial T1 and T2 mapping and late gadolinium enhancement (LGE) images. Examples of myocardial T1 mapping in a healthy control subject (female, 33 years old), a patient with AVM (acute viral myocarditis, female, 51 years old) and a patient with IIM (idiopathic inflammatory myositis with cardiac involvement, male, 51 years old). On the first row, native T1 maps are shown (949 ms in the healthy control, 1019 ms in the AVM patient and 1001 ms in the IIM patient. On the second row, post contrast T1 times were measured 332 ms, 277 ms and 306 ms. On the third row, T2 relaxation times were measured 45 ms, 52 ms and 50 ms. Corresponding late gadolinium slices are shown on the last row with slight inferior and infero-lateral sub-epicardial enhancement in the AVM patient, while a minimal infero-lateral subepicardial enhancement can be discussed in the IIM patient. Post contrast T1 in the pectoralis muscle was 469 ms, 499 ms and 399 ms, thus differentiating the IIM patient from the AVM patient and the healthy control

Fig. 3. ROC curves illustrating the ability of thoracic skeletal muscles and myocardium mapping parameters to differentiate acute viral myocarditis (AVM) from idiopathic inflammatory myopathy (IIM) myocarditis patients. Area under the ROC curve (AUC) for each mapping parameter are indicated, as well as optimized cutoff values with corresponding sensitivities, specificities and accuracies. ECV = extracellular volume fraction, %

Fig. 4. Comparison of different CMR mapping parameters for the measured muscle groups in healthy controls. Error bars indicate 95%-confidence-interval. P-value of Kruskal-Wallis test between muscle groups is indicated for each mapping parameter. ECV = extracellular volume fraction, %

Fig. 5. Schematic workflow in patients undergoing CMR for suspected myocarditis. The cutoffs in this study for myocardial inflammation were > 988 ms for native T1 and > 50 ms for T2 in the myocardium, while in the thoracic skeletal muscles, inflammation was best identified with a cutoff of < 431 ms for post contrast T1 and an ECV > 12%. Please note that cutoffs for mapping parameters might significantly vary between centers with different vendors, field strengths and acquisition parameters, so identification of normal values in healthy volunteers is recommended for each CMR magnet [29]. CMR = cardiovascular magnetic resonance; ECV = extracellular volume fraction; IIM = idiopathic inflammatory myopathy