Diagnostic implications of magnetic resonance feature tracking derived myocardial strain parameters in acute myocarditis

Abstract

Purpose: The present study aims to evaluate the diagnostic value of cardiac magnetic resonance (CMR) feature tracking (FT) derived strain-analysis of both ventricles in patients with acute myocarditis (ACM) in order to improve its currently still challenging non-invasive diagnosis.

Methods: CMR cine data of 31 patients with clinically suspected ACM and confirmation of diagnosis by CMR according to the Lake Louise criteria as well as 14 patients with clinically diagnosed ACM but inconspicuous CMR were retrospectively analyzed. 20 healthy volunteers (HV) served as a control. Analysis of global longitudinal, circumferential and radial strain and strain rate of both ventricles was performed in one long-axis and three short-axis slices using a dedicated FT-software (TomTec Imaging Systems).

Results: Patients with ACM showed significantly reduced LV longitudinal strain (-12.7 ± 6.5 vs. -16.8 ± 5.9%, p=0.021) and LV circumferential strain (LVCirStrain; -22.9 ± 5.7 vs. -27.8 ± 4.4 %, p<0.001) compared to HV. Conversely, they showed improved basal RV circumferential strain rate (BasalRVCirSR; -0.70 ± 0.23 vs. -0.47 ± 0.31s(-1), p=0.009). In ACM patients with preserved EF, BasalRVCirSR was significantly increased compared to HV while LV strain was not significantly different between both groups. In multinominal logistic regression analysis, LVCirStrain and BasalRVCirSR proved to be the best independent predictors of ACM with preserved EF. A combined cut-off of -0.53s(-1) for BasalRVCirSR and of -29.0% for LVCirStrain allowed a classification of ACM patients with preserved EF with a sensitivity of 89% and a specificity of 80%. Also patients with clinical ACM but inconspicuous CMR showed a significantly improved BasalRVCirSR and a cut-off of -0.77s(-1) allowed a classification of ACM patients with a sensitivity of 70% and a specificity of 90%, while all other CMR parameters were normal.

Conclusions: The defined cut-offs for LVCirStrain and BasalRVCirSR allow a prediction of ACM with high sensitivity and specificity, even in patients with preserved EF and in patients with otherwise completely inconspicuous CMR. Our results point to a discriminative power especially of RV strain analysis in the CMR-based diagnosis of ACM.

Keywords: Acute myocarditis; Cardiovascular magnetic resonance; Feature tracking; Lake Louise criteria; Myocardial strain.