Published in:
03-07-2021 | Sudden Cardiac Death
The prognostic role of late gadolinium enhancement on cardiac magnetic resonance in patients with nonischemic cardiomyopathy and reduced ejection fraction, implanted with cardioverter defibrillators for primary prevention. A systematic review and meta-analysis
Published in: Journal of Interventional Cardiac Electrophysiology | Issue 3/2022
Login to get accessAbstract
Background
Previous studies suggest that late gadolinium enhancement (LGE) on cardiac magnetic resonance (CMR) is associated with arrhythmic events in patients with nonischemic cardiomyopathy (NICM), while others have questioned the role of left ventricular ejection fraction (LVEF) as a sole predictor of future events.
Objectives
To evaluate the role of LGE on CMR in identifying patients with NICM and reduced LVEF for whom a benefit from defibrillator implantation for primary prevention is not anticipated, thus they are mainly exposed to potential risks.
Methods
Major electronic databases were searched for studies reporting the incidence of appropriate device therapy (ADT), sudden cardiac death (SCD), and cardiac death based on the presence of LGE on CMR, among patients with NICM and reduced LVEF, implanted with a cardioverter defibrillator for primary prevention.
Results
Eleven studies (1652 patients, 947 with LGE) were included in the final analysis. LGE presence was strongly associated with ADT (logOR: 1.95, 95%CI: 1.21–2.69) and cardiac death (logOR: 0.91, 95%CI: 0.14–1.68), but not with SCD (logOR: 0.26, 95%CI: −1.09–1.6). Diagnostic accuracy analysis demonstrated that contrast enhancement is a sensitive marker of future ADT and cardiac death (93%, 95%CI: 85.8–96.7%; 82.9%, 95%CI: 70.6–90.7%; respectively), with moderate specificity ( 44%, 95%CI: 27.2–62.6%; 37.7%, 95%CI: 23.4–54.6%; respectively).
Conclusion
LGE is a highly sensitive predictor of ADT and cardiac death in NICM patients implanted with a defibrillator for primary prevention. However, due to moderate specificity, derivation of a cutoff with adequate predictive values and probably a multifactorial approach are needed to improve discrimination of patients who will not benefit from ICDs.