Scar Tissue Characterization to Predict Arrhythmia Recurrence in Patients Undergoing Ventricular Tachycardia Ablation: Cardiac Computed Tomography Compared with Cardiac Magnetic Resonance

Abstract Background The role of imaging in evaluating patients undergoing ventricular tachycardia (VT) ablation is crucial for planning, procedural success, and clinical outcomes. In daily practice, cardiac computed tomography (CCT) assessing wall thinning and late iodine uptake (LI) and cardiac magnetic resonance (CMR) assessing late-gadolinium enhancement (LGE) are most commonly implemented. However, data comparing the two imaging modalities for scar tissue characterization is scarce. Aims To compare the performance of both imaging modalities for scar tissue characterization in ischemic (ICM) and non-ischemic (NICM) patients undergoing VT ablation and predict procedural outcomes. Methods In a retrospective analysis, consecutive patients undergoing both CCT and LGE-CMR before scheduled VT ablation were included. The presence and extent of scar was assessed by means of each imaging modality using dedicated softwares (MUSIC and ADAS). To compare scar distribution and agreement between the two modalities, a scar classification scheme was used for all 17 AHA segments: 0 points (no scar present), 1 point (scar <50% of total segment area), 2 points (scar representing 50-99% of total segment area) or 3 points (scar present in 100% of the segment area).<scar<100%)><scar<100%> Results 36 patients (67±10 years; 97% male; LVEF 39±10%; 72% ischemic) undergoing CCT and LGE-CMR before scheduled VT ablation were included. In the ICM group, mean detected scar burden was higher than in the NICM group on LGE-CMR (total volume: 62.3±32.9g vs. 32.8±9.5g, p<0.001), using wall thinning model (WT)(mean scar area 69.0±45.6 cm2 vs. 23.4±23.6cm2, p=0.005) as well as LI (dense scar area: 39.2±28.1 cm2 vs. 11.0±6.3 cm2 , p=0.005). The absolute score difference per segment and patient between CCT and LGE-CMR was calculated and summed for all patients divided by the number of patients. Differences were more pronounced for NICM compared to ICM and LGE-CMR showed better concordance to LI than WT in both NICMP and ICM (Figure). Total procedure time was 205±68 min (ablation time 28±16 min; fluoroscopy time 15 ± 12 min). During a follow-up period of 19±7 months, 87 sustained VT episodes in 12 patients (33%) were documented. Three patients (8%) underwent a second VT ablation and one patient (3%) died during follow-up. In ICM, total scar volume on LGE-CMR and scar area on LI showed numerically higher predictive value for VT recurrence than total scar area by means of WT (AUC: LGE 0.76, LI 0.78 and WT 0.65 respectively), while in NICM scar characterization by means of WT yielded higher performance to predict VT recurrence than LI or LGE-CMR (AUC: LGE 0.67, LI 0.58, WT 0.83 respectively). Conclusions In patients undergoing VT ablation, both LGE-CMR and CCT are valuable for scar tissue characterization. Late iodine uptake on CCT showed a high concordance with LGE-CMR in both NICM and ICM.</scar<100%></scar<100%)>