16-34: Specific CT features indicative of fibrofatty replacement are associated with local conduction delay in ARVC confirmed by endocardial and epicardial electroanatomical mapping

Abstract

Introduction: ARVC is a hereditary disease characterized by fibrofatty replacement. Fibrofatty infiltration interspersed with bundles of viable myocardium may provide the substrate for localized conduction delay and ventricular tachycardia (VT). Cardiac CT allows delineation of fat and myocardium with high spatial resolution. We aimed to evaluate whether specific CT features may be indicative to fibrofatty replacement patterns that are associated with slow conduction using integrated electroanatomical mapping (EAM) as gold standard. Method: ARVC patients who underwent combined endocardial and epicardial EAM for VT ablation with real-time integration of CT-derived fat mesh were enrolled. Intracardiac electrocardiograms (ICEG) were categorized as abnormal (consistent with conduction delay based on ICEG characteristics) or normal (all remaining points independent from voltages). Abnormal ICEG were subdivided in fragmented (amplitude/duration [mV/ms] ratio ≤0.05 and duration > 50ms) and late potential (inscribing after QRS, separated by isoelectric segment > 20 ms). Reversed registration was performed to superimpose EAM points on the contour of the post processed short-axis CT slice. On short axis slices, fat infiltration was quantified, defined as the percentage of image pixels with Hounsfield units from -100 to -50 HU, within the subepicardial 4 mm layer of the RV free wall. For each mapping site the corresponding percentage of fat infiltration was calculated and compared to the ICEG characteristics. Patients without structural heart disease matched for age and body mass index served as controls for CT derived fat within the subepicardial layer. Receiver operating characteristics curve analysis was performed to determine the optimal cut-off value for the percentage of intramural fat, to differentiate between in ARVC patients and controls. Results: Eight ARVC patients with combined endocardial and epicardial EAM and CT image integration (all male, age 47 ± 18 years, BMI 24 ± 3 kg/m2, ARVC associated mutation in 6 [75%]) and 8 healthy controls (all male, age 48 ± 17 years, BMI 24 ± 4 kg/m2) were enrolled. Of the 4209 exported mapping points; 3416 (81%) were normal and 785 (19%) were abnormal (232 fragmented EGM and 553 late potentials). The median fat percentage was significantly higher at sites with local conduction delay compared to normal sites (5% [IQR, 0-18%] vs. 0% [IQR, 0-7%], P= 0.001, panel A). The mean fat infiltration was higher in ARVC patients compared to healthy matched controls (5.7 ± 3.9% vs. 1.3 ± 0.6%, p= 0.008, panel B). The optimal cut off of fat percentage to differentiate between ARVC patients and healthy matched controls was 2% fat (sensitivity 88% and specificity 100%). Conclusion: A subepicardial fat percentage of 2% may discriminate between ARVC patients and normal controls. In ARVC patients, areas with conduction delay based on electrogram characteristics show a higher amount of fat infiltration, compared to areas without delay. A critical amount of fat infiltration may be required to cause conduction delay, which may be detectable by pre-procedural CT. Whether cardiac CT can be used for the early diagnosis of ARVC and for imaging guided substrate ablation requires further investigation in a larger group of patients.