Abstract
BACKGROUND Cardiac magnetic resonance imaging (CMRI) is considered to be useful for the diagnosis of myocarditis, and the Lake Louise Criteria (LLC) has been proved to be of significance as the standard of diagnosis. However, the diagnostic performance of LLC-based CMRI for myocarditis compared with endomyocardial biopsy (EMB) has not been quantitatively evaluated in a meta-analysis. MATERIAL AND METHODS The databases PubMed, Cochrane's Library, and EMBASE were searched to identify studies on LLC and its individual components for the diagnosis of myocarditis. EMB was the control reference. The sensitivity, specificity, and positive and negative diagnostic likelihood ratios were calculated with a random-effects model. The area under the receiver operating characteristic curve (AUC) was estimated to show overall effectiveness. RESULTS We included 9 cohorts (614 patients) of patients with suspected MC. The combined sensitivities, specificities, and AUCs for T1-weighed global relative enhancement were 0.66, 0.73, and 0.71; for T2-weighed edema ratio they were 0.52, 0.73, and 0.72; for the late gadolinium enhancement, they were 0.70, 0.57, and 0.67; and for LLC-based CMRI they were 0.70, 0.56, and 0.70, respectively. Subgroup analysis indicated that the sensitivities, specificities, and diagnostic accuracies of LLC and its individual component-based CMRI seemed to be similar in patients with acute or chronic myocarditis. Results of the Deeks' funnel plot asymmetry test showed no significant publication bias among the studies. CONCLUSIONS CMRI based on LLC or its individual components seems to have moderate accuracy in diagnosis of acute or chronic myocarditis.