A comparison of cardiac magnetic resonance imaging peri-infarct border zone quantification strategies for the prediction of ventricular tachyarrhythmia inducibility

Abstract

Background: Peri-infarct border zone (BZ) as quantified by late gadolinium enhancement (LGE) on cardiac magnetic resonance imaging (MRI) has been proposed as a risk stratification tool, and is associated with increased mortality. BZ has been measured by various methods in the literature. We assessed which BZ analysis best predicts inducible arrhythmia during electrophysiological study (EPS).
Methods: LGE was performed in 47 patients with coronary artery disease referred for EPS to assess for ventricular tachycardia (VT). LGE data was analyzed for BZ quantification by 3 previously published methods. Method I (BZ-I) used pixels 2–3 standard deviations over the mean of normal tissue, expressed as % of left ventricular mass, Method II (BZ-II, as described by Yan) and Method III (BZ-III, as described by Schmidt). EPS results were classified as negative (non-inducible) or positive (monomorphic VT — MVT).
Results: There were 47 subjects-age 61.7 years, 72% male. During EPS, 20 patients were non-inducible and 18 had induced MVT. Ejection fraction was not significantly different between non-inducible patients and those with MVT (34.1% vs. 28.5%, p = 0.13). BZ-I was significantly different (1.4% vs. 2.6%, p = 0.001), but not BZ-II (7.9% vs. 6.9%, p = 0.68) or BZ-III (2.7 g vs. 2.1 g, p = 0.88). Multivariate analysis demonstrated that only BZ-I was an independent predictor of EPS outcome after controling for infarct size (OR 1.97 per % change, 95% CI 1.04–3.73, p = 0.04).
Conclusions: This study demonstrates significant variability between the published methods for measuring BZ. Also, BZ-I is a stronger predictor of inducible MVT during EPS than ejection fraction and infarct size. BZ may be another LGE marker of elevated risk of arrhythmia.