RELATIONSHIP OF ELECTRICAL MARKERS OF SUDDEN CARDIAC ARREST RISK AND CHARACTERISTICS OF MYOCARDIAL SCAR

Abstract

Left ventricular ejection fraction (LVEF) is the standard method used to identify patients at risk for sudden cardiac arrest (SCA) after myocardial infarction (MI). Yet, electrical markers and myocardial scar appear to provide more accurate information on SCA risk. We assessed the relationships between two established electrical markers, Holter-based T-wave alternans (TWA) and Heart Rate Turbulence (HRT) with scar, as assessed using late gadolinium-enhanced cardiac magnetic resonance imaging (LGE-CMR), in post-MI patients with a range of LVEF values to assess the correlation between these measures. Consecutive patients in sinus rhythm and no contraindication to CMR-LGE were enrolled 3 to 15 months after MI. HRT slope and maximal TWA were measured from 24-hour high-resolution ECG recordings (SEER12) and commercial software (MARS, GE Healthcare). CMR-LGE imaging was used to assess the extent, location and characteristics of myocardial scar (LV percent at 5-SD) using a 1.5 Tesla scanner and CVI42 software (Circle Cardiovascular Imaging). Electrical and scar data were independently assessed and compared using variance least squares regression, t-tests and the non-parametric test of trend. The 99 patients enrolled included 83 participants in the Risk Estimation Following Infarction Noninvasive Evaluation-ICD efficacy (REFINE-ICD) study (LVEF 36 to 50%) and others with an LVEF < 35% (n = 6) or > 50% (n = 10). Median LVEF was 42% (inter-quartile range 38% to 45%). Most were male (87%), their mean age was 61 years, 76% had recent documented ST-Elevation MI and 25% had a history of multiple MI events. A linear relationship between TWA and the extent of myocardial scar was observed (p = 0.002). Yet, maximal mean TWA values only tended to be higher in the 32 patients with anterior transmural scar (55 μV) vs. the 68 others (51 μV; p = 0.09). Further, maximal TWA was not related to the number of regions with scar (p = 0.2). No significant linear relationship between HRT slope and scar burden was observed (p = 0.6). Further, no significant relationship between HRT slope and scar was found in terms of transmurality (p = 0.6), location (p = 0.3) or the number of regions affected (p = 0.4). These data indicate that TWA is weakly associated, and HRT is not associated with myocardial scar extent and location among patients with prior myocardial infarction. Accordingly, electrical versus CMR-based markers of risk appear poorly correlated and offer independent insight into arrhythmogeneisis among this patient population.