Cardiac magnetic resonance for ventricular arrhythmias: a systematic review and meta-analysis

Background Cardiac magnetic resonance (CMR) allows comprehensive myocardial tissue characterisation, revealing areas of myocardial inflammation or fibrosis that may predispose to ventricular arrhythmias (VAs). With this study, we aimed to estimate the prevalence of structural heart disease (SHD) and decipher the prognostic implications of CMR in selected patients presenting with significant VAs. Methods Electronic databases were searched for studies enrolling adult patients that underwent CMR for diagnostic or prognostic purposes in the setting of significant VAs. A random effects model meta-analysis of proportions was performed to estimate the prevalence of SHD. HRs were pooled together in order to evaluate the prognostic value of CMR. Results The prevalence of SHD was reported in 18 studies. In all-comers with significant VAs, the pooled rate of SHD post-CMR evaluation was 39% (24% in the subgroup of premature ventricular contractions and/or non-sustained ventricular tachycardia vs 63% in the subgroup of more complex VAs). A change in diagnosis after use of CMR ranged from 21% to 66% with a pooled average of 35% (29%–41%). A non-ischaemic cardiomyopathy was the most frequently identified SHD (56%), followed by ischaemic heart disease (21%) and hypertrophic cardiomyopathy (5%). After pooling together data from six studies, we found that the presence of late gadolinium enhancement was associated with increased risk of major adverse outcomes in patients with significant VAs (pooled HR: 1.79; 95% CI 1.33 to 2.42). Conclusion CMR is a valuable tool in the diagnostic and prognostic evaluation of patients with VAs. CMR should be considered early after initial evaluation in the diagnostic algorithm for VAs of unclear aetiology as this strategy may also define prognosis and improve risk stratification.