Using the 3D architecture of scar to predict life-threatening ventricular arrhythmias, still a long way to go

Abstract Background Late gadolinium enhancement (LGE) has been proposed as an independent predictor of ventricular arrhythmias. Purpose The purpose of this study was to assess if myocardial scar characterization could enhance the risk stratification for life-threatening arrhythmias and sudden cardiac death (SCD). Methods We included patients with an indication for ICD or CRT-D implantation who underwent cardiac magnetic resonance for clinical proposes since February/2018 and in whom a 3D-LGE dataset was obtained. Patients with channelopathies (n=2) or inappropriate imaging quality (imaging artifacts; n=7) were excluded. Scar characterization using ADAS software was performed in 3D-LGE datasets in all but 16 patients, where 2D datasets were used. The primary endpoint was the composite of appropriate ICD therapy (classified as ATP or shock), sustained ventricular tachycardia or SCD. Results A total of 116 patients were analysed (mean age 66 ± 14 years; 81% male; mean LVEF 34 ± 14%; 74 patients with ischemic and 42 with non-ischemic cardiomyopathy; 40 patients received a device in the setting of secondary prevention). During a median follow-up of 2.3 years (IQR 1.3 – 3.3 years) there were 23 events (18 appropriate ICD therapies [9 shocks and 9 ATP], 3 episodes of VT under the threshold for ICD therapy and 2 SCD). No statistically significant differences were found between patients with or without events in terms of scar mass, border zone (BZ) mass, BZ channels (BZC), BZC mass, number of channels detected, and scar heterogeneity (BZ mass / scar mass ratio) - all p values > 0.2 – Figure. Restricting the analysis to only primary prevention cases yielded similar results. Overall, 26 patients did not show any channel. Four of these (all with non-ischemic LGE) experienced an arrhythmic event, yielding a negative predictive value of 83% (95% CI 64-93%) for the absence of channels. Conclusion In this cohort with still relatively limited follow-up duration, no single parameter reflecting scar tissue characterization was able to predict appropriate device therapies or sudden cardiac death.