0486 : Predictive value of late gadolinium enhancement quantification with cardiac magnetic resonance imaging in acute myocarditis

Abstract

Prognostic factors after acute myocarditis are lacking, yielding heterogeneity in follow-up practices. We hypothesized that quantifying the late gadolinium enhancement (LGE) during cardiac magnetic resonance imaging (CMRI) could be an appealing tool to predict the risk of cardiovascular event. This pilot study aimed at evaluating the variation in LGE Semi- Quantitative Score (SQS) between CMRI at admission and a second CMRI three months later, as a predictive tool of adverse cardiac events following an acute viral myocarditis. 49 patients with acute viral myocarditis as confirmed by CMRI Lake Louise criteria were included. Patients had a second CMRI at 3 months and were followed clinically during one year. No cardiovascular event occurred between discharge and the second CMRI at three months. LGE was quantified at M0 and M3, using SQS already validated in ischemic cardiomyopathy. SQS predictive value was assessed in the acute phase and at one year using a combination of death from cardiovascular cause, heart failure or severe arrhythmias as main outcome. In most cases (n=39, 79,5%), LGE decreased over the time after an acute myocarditis and rarely disappeared (n=5, 10%). At admission, we did not identify any biomarker (troponin, CPK, or CRP) or imaging parameter (SQS) capable of predicting future cardiac events. Of note SQS was not correlated to cardiovascular (troponin: r 2 =0.22, p=0.001; CPK: r 2 =0.11, p=0.03) or inflammation (CRP:ρ=0,35, p=0.02) biomarkers. However, variation of SQS between admission and 3 months exhibited promising predictive value for adverse cardiac events with an AUC of 97,5% (sensitivity =90, specificity =74%). This analysis showed that a reduction of SQS lower than 42% was a risk factor for cardiac event. The follow up of LGE decrease with SQS at three months is a useful tool to predict cardiovascular events in the year following an acute viral myocarditis. The author hereby declares no conflict of interest