Focus Session
SCMR 22nd Annual Scientific Sessions
Jihye Jang, MSc
Research Fellow
Beth Israel Deaconess Medical Center and Harvard Medical School
John Whitaker, MD
Clinical Research Fellow
King's College London
Eran Leshem, MD
Research Fellow
Beth Israel Deaconess Medical Center
Long Ngo, PhD
Associate professor
Beth Israel Deaconess Medical Center
Shiro Nakamori, MD
Research fellow
Beth Israel Deaconess Medical Center
Warren Manning, MD
Professor
Beth Israel Deaconess Medical Center
Elad Anter, MD
Assistant Professor
Beth Israel Deaconess Medical Center
Reza Nezafat, PhD
Associate Professor of Medicine
Department of Medicine (Cardiovascular Division) Beth Israel Deaconess Medical Center
Background: Slow conduction may contribute to the formation of an arrhythmogenic substrate in the region of ventricular scar [1]. Conduction velocity is reduced in the presence of myocardial fibrosis or heterogeneous connectivity in tissue microarchitecture [2]. With the ability of LGE to visualize myocardial fibrosis [3], it may offer a unique ability to predict areas of slow conduction velocity non-invasively. We sought to investigate associations between conduction velocity (CV) measured using catheter mapping and LGE, myocardial wall thickness, and their heterogeneity.
Methods: Six swine with healed anterior wall infarction [4] were studied and the processing pipeline is shown in Figure 1. In-vivo LGE imaging was performed using a dark-blood LGE sequence [5], and electrophysiology study was performed using the CARTO electroanatomic mapping (EAM) system (Biosense Webster, CA). CV was estimated using a triangulation technique [6] from a local activation time acquired during EAM. LGE and wall thickness heterogeneity were defined as standard deviation of 20 nearest neighboring points of LGE signal intensity and LV wall thickness that were projected on the endocardial surface mesh. CMR and EAM data were spatially aligned using landmarks registration followed by hammer mapping. A generalized linear modeling was performed to investigate the mean level difference of CV in LGE enhanced/normal and thinned/normal wall thickness regions, and the association of CV with heterogeneity in LGE and wall thickness for each individual level and an aggregated model.
Results: In spatially aligned EAM and CMR data, regions of LGE scar and thinned LV wall shows slowed CV (Figure 2). In the mixed model of all animals, LGE enhanced region has a mean CV that is significantly lower than the CV mean from LGE normal (p<0.001; Table 1A). For the LV wall thickness, CV is significantly higher for non-thinned wall regions (p<0.001; Table 1B). LGE heterogeneity showed a negative association with conduction velocity when LGE normalized signal was used as a confounding variable (slope: -0.009, p=0.005; Table 1C). Wall thickness heterogeneity showed a negative association with conduction velocity when wall thickness was used as a confounding variable (slope: -0.011, p=0.023; Table 1D).
Conclusion: The areas with low conduction velocity, as measured by EAM, is located at hyperenhanced region on LGE, and thinned myocardium. Heterogeneous regions of LGE and wall thickness shows lower conduction velocity.