LAT velocity vectors: a novel tool for ventricular functional substrate characterization

Abstract Background Catheter ablation is a cornerstone treatment for scar-related ventricular arrhythmias arising in patients affected by structural heart disease. LAT Velocity Vectors (LVV) is a novel feature for wave velocity visualization over ventricular substrate maps during sinus or paced rhythm. Purpose To evaluate conduction velocity characteristics of LVV according to local anatomical and functional substrate features. Methods Fifteen consecutive patients prospectively enrolled in the ongoing VOYAGE clinical trial for cardiac magnetic resonance (CMR) aided VT ablation at our center were retrospectively analyzed. Multidetector computed tomography (MDCT), and late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) imaging was performed before procedure. LGE-CMR was processed with ADAS3D for scar characterization and heterogeneous tissue channel (HTC) identification. All patients underwent high density endocardial, epicardial or endo-epicardial mapping with a multipolar catheter during sinus or ventricular paced rhythm. LVV were generated with CARTO V8 electroanatomical mapping suite, compression was set to maximum value for discrete LVV identification and conduction velocity (CV) was calculated for each LVV. LVV were compared in terms of CV according to colocalization (< 5 mm) with scar and border zone (BZ) (< 1,5 mV bipolar) areas, deceleration zones (DZ) and HTCs. Results A total of 185 LVVs were identified in 15 patients (12,3 LVVs per patient, mean age 61,9±9 years, mean LVEF 37,3±8%, LVEDV 171,1±70 mL), 61 (33%) LVVs were epicardial. Overall mean LVV CV was 0,32±18 mm/ms. In scar/BZ areas, mean CV was significantly lower than healthy tissue (0,30±0,16 vs 0,42±0,21 mm/ms, P<0,01). LVVs corresponding to DZs also showed significantly lower CV than LVV outside DZs (0,28±0,16 vs 0,41±0,20 mm/ms, P<0,01). LVVs colocalized to ADAS3D HTCs also showed lower CV than LVV outside HTC area (0,25±0,14 vs 0,39±0,20 mm/ms, P<0,01). When excluding LVVs outside the scar, LVV CV was still significantly lower both in case of DZ (0,26±0,14 vs 0,39±0,17 mm/ms, P<0,01) and HTC (0,25±0,13 vs 0,37±0,18 mm/ms, P<0,01) colocalization. ROC analysis for HTC prediction using LVV (scar/BZ only) found best cutoff < 0,26 mm/ms (68% sensitivity, 64% specificity) with 0,72 AUC. Conclusion LAT velocity vectors showed significantly slower conduction when colocalized to scar areas, DZs and HTCs, and hold potential as a novel tool for easy functional substrate characterization and corridor prediction. Further studies with larger samples are necessary to confirm these results.