CMR in patients with severe myocarditis: diagnostic value of quantitative tissue markers including extracellular volume imaging.

objectivesThis study evaluated the accuracy of T2, T1, and extracellular volume (ECV) quantification as novel quantitative tissue markers in comparison with standard "Lake-Louise" cardiac magnetic resonance (CMR) criteria to diagnose myocarditis.

backgroundNovel approaches using T2 and T1 mapping may overcome the limitations of signal intensity-based parameters, which would potentially result in a better diagnostic accuracy compared with standard CMR techniques in suspected myocarditis.

methodsCMR was performed in 104 patients with myocarditis and 21 control subjects at 1.5-T. Patients with myocarditis underwent CMR 2 weeks (interquartile range: 1 to 7 weeks) after presentation with new-onset heart failure (n = 66) or acute chest pain (n = 38). T2 and T1 mapping were implemented into a standard protocol including T2-weighted (T2w), early gadolinium enhancement (EGE) CMR, and late gadolinium enhancement (LGE) CMR. T2 quantification was performed using a free-breathing, navigator-gated multiecho sequence. T1 quantification was performed using the modified Look-Locker inversion recovery sequence before and after administration of 0.075 mmol/kg gadobenate dimeglumine. T2, T1, and ECV maps were generated using a plug-in for the OsiriX software (Pixmeo, Bernex, Switzerland) to calculate mean global myocardial T2, T1, and ECV values.

resultsThe diagnostic accuracies of conventional CMR were 70% (95% confidence interval [CI]: 61% to 77%) for T2w CMR, 59% (95% CI: 56% to 73%) for EGE, and 67% (95% CI: 59% to 75%) for LGE. The diagnostic accuracies of mapping techniques were 63% (95% CI: 53% to 73%) for myocardial T2, 69% (95% CI: 60% to 76%) for native myocardial T1, and 76% (95% CI: 68% to 82%) for global myocardial ECV. The diagnostic accuracy of CMR was significantly improved to 90% (95% CI: 84% to 95%) by a stepwise approach, using the presence of LGE and myocardial ECV ≥27% as diagnostic criteria, compared with 79% (95% CI: 71% to 85%; p = 0.0043) for the Lake-Louise criteria.

conclusionsIn patients with clinical evidence for subacute, severe myocarditis, ECV quantification with LGE imaging significantly improved the diagnostic accuracy of CMR compared with standard Lake-Louise criteria.