P1106Fundamental differences and predictors of scar arrhythmogenicity in ischemic patients using cardiac magnetic resonance: a propensity score-matching study

Abstract

Abstract Introduction Late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) permits to identify the arrhythmogenic substrate (AS) in ischemic patients. However, it is unknown why the majority of them never develop ventricular tachycardias (VT), irrespectively of their left ventricular ejection fraction (LVEF). Purpose To characterize the fundamental differences and potential predictors of scar arrhythmogenicity in post-myocardial infarction (MI) patients with and without VT. Methods 36 consecutive ischemic patients with no arrhythmia evidence underwent a LGE-CMR study 4 years after the MI (controls). Scar data were compared with those obtained from 49 ischemic patients referred for VT substrate ablation (cases). Propensity score matching (PSM) was performed to adjust for age, LVEF, scar mass, and time from MI. The myocardium was segmented in 10 layers (endo- to epicardium), characterizing the core, border zone (BZ) and BZ channels (BZCs) using a dedicated post-processing software. Results Compared to controls, cases were significantly older (67.3 ± 9.1 vs. 56.5 ± 11), had lower LVEF (33.1 ± 10.1 vs. 51 ± 9.4), greater scar mass (33.9 ± 17.2 vs. 14.2 ± 11.6 g), BZ mass (21.1 ± 9.9 vs. 9.6 ± 7.6 g), core mas (12.6 ± 8.8 vs. ± g), number of BZC (2.9 ± 1.4 vs. 1.1 ± 1.1) and BZC mass (10.5 ± 4.2 vs. 2.3 ± 2.4 g) (p < 0.001 in all cases). After PSM (2:1) adjustment[BJG1] , cases had more BZCs (2.9 ± 1.4 vs. 2.2 ± 0.6; p = 0.01) and a greater BZC mass (10.5 ± 4.2 vs. 4.6 ± 2.6 g; p < 0.001). In the multivariable logistic regression analysis, the BZC mass was the only independent predictor of being a case [OR 2.3 (1.5–3.4); p < 0.001]. Receiver operating characteristic curve analysis identified a cut-off point of BZC mass >4.28 g (AUC 0.98; p < 0.001), showing 100% sensitivity and 91% specificity for cases’ discrimination. Conclusions Compared with the cases, an otherwise similar control group (PS-matched for age, LVEF, scar mass, and time from MI) showed fewer BZC and a reduced BZC mass. BZC mass was the only independent predictor of being a case. A BZC mass cut-off point of > 4.28 g showed a 100% sensitivity and 91% specificity for the identification of ischemic patients with documented VT. Abstract Figure. Mean BZC mass and ROC curve analysis