3255Characterization of the arrhythmogenic substrate with multimodality imaging in ischemic patients undergoing VT ablation: relationship between cardiac computed tomography and magnetic resonance

Abstract

Abstract Background Cardiac magnetic resonance (CMR)-aided ventricular tachycardia (VT) substrate ablation has shown to improve VT recurrence-free survival, through a better identification of the arrhythmogenic substrate. However, the access to CMR may be limited in certain centers or sometimes Its use can be contraindicated in patients with cardiac implantable electronic device. Cardiac computed tomography (CT) has shown to improve the results of substrate ablation, correlating with low-voltage areas and local abnormal ventricular activity, and identifying ridges of myocardial tissue (CT-channels) that may be appropriate target sites for ablation. Purpose To evaluate the correlation between CT and CMR imaging in identifying anatomical heterogeneous tissue channels (CMR-channels) or CT-channels in ischemic patients undergoing VT substrate ablation. Methods The study included 30 post-myocardial infarction (MI) patients (mean age 69±10; 94% male, left ventricular ejection fraction 35±10%), who underwent both CMR and cardiac CT before VT substrate ablation. Using a dedicated post-processing software, the myocardium was segmented in 10 layers from endocardium to epicardium both for the CMR and CT, characterizing the presence of CMR-channels and CT-channels, respectively, by two blinded operators, assigned either to CMR or CT analysis. CMR-channels were classified as endocardial (CMR-channels in layer <50%), epicardial (CMR-channels in layers ≥50%) or transmural (in both endo and epicardial layers). Presence and location of CT and CMR-channels were compared. Results In 26/30 patients (86.7%) 91 CT-channels (mean 3.0±1.9 per patient) were identified while 30/30 (100%) showed CMR-channels (n=76; mean 2.4±1.2 per patient). We found 190 CT-channel entrances (mean 6.3±4.1 per patient), and 275 CMR-channel entrances (mean 8.9±4.9 per patient) on cardiac CT and CMR, respectively. There were 47/91 (51.6%) true positive CT-channels. On the contrary, 44/91 (48.4%) CT-channels were considered false positives [19/91 (20.9%) identified out of CMR scar], and 29/76 (38.2%) CMR-channels could not be identified on CT. Thirty-six out of 76 (47.4%) CMR-channels were considered as non-endocardial (epi- or transmural). Twenty-nine out of 36 (80.5%) non-endocardial CMR-channels were coincident with CT-channels. CT and CMR Channels Conclusion CT shows a modest sensitivity in identifying CMR-channels and fails in ascertain their complexity, underestimating the number of entrances; however, channels location at CT fit well with CMR for those classified as transmural or epicardial.