Peak frequency analysis for epicardial contact mapping: How (in)frequent is epicardial adipose tissue?

Combined endo-epicardial catheter ablation is an established treatment for ventricular arrhythmias. Abnormal omnipolar voltage (OV) <1.5mV is a surrogate of underlying myocardial fibrosis/scar. Yet, epicardial adipose tissue (EAT) is a common confounder when interpretating epicardial OV maps and can result in "false" low voltage annotation even in the absence of scar. It is also an important limitation for radiofrequency delivery due to its insulating properties preventing adequate energy delivery to the myocardial tissue. Thus, intraprocedural data about EAT localisation and extent is a crucial information for the operator. The "peak frequency" (PF) metric, recently introduced into EnSiteTM X (Abbott), estimates the highest frequency content (Hz) of the EGM. Its application for endocardial ventricular mapping has been reported.1Payne J.E. Woods C. Elshazly M.B. Matthews A. Kroman A. Feng Z. et al.A novel automated peak frequency annotation algorithm for identifying deceleration zones and ventricular tachycardia ablation sites.Heart Rhythm. 2024; 21: 27-33Google Scholar 2Merino J.L. Yamaguchi T. Kim S. Relan J. Castrejon-Castrejon S. Otsubo T. et al.PO-04-120 Differentiation of Near Field from Far Field Electrograms in the Low voltage Region recorded during Ventricular Tachycardia improves discrimination of sites of arrhythmia ablation.Heart Rhythm. 2023; 20: S530-531Google Scholar On epicardial maps, low PF values may indicate far field signals and could indirectly support identification of sites with EAT. Here, we present an epicardial contact map of a patient with suspected ARVC and RVOT VT with integrated anatomical CT segmentation. The OV map (1) revealed low voltage areas (<1.5mV) along the acute margin of the RV, peri-tricuspid and RVOT as well as patchy areas over the interventricular groove. The novel emphasis map feature (4) was set to highlight EGMs with >250Hz on the OV map to indicate sharp "near field" EGMs. It allowed to distinguish two sites over the epicardial RVOT and peri-tricuspid with low voltage- high frequency EGMs from sites with "false" low voltage, low frequency due to EAT and/or poor catheter contact. Semi-automated imaging of epicardial adipose tissue (EAT) in cardiac CT using Adas3D (2) is shown as a reference to establish the anatomical EAT pattern and demonstrates reasonable agreement with the PF map (3). Supported by complementary VE/VT activation and pace mapping, the area over the RVOT was targeted (white/yellow square) and non-inducibility achieved. In this particular case, the second low voltage-high frequency peri-tricuspid area (dashed lines) was not ablated as no additional substrate modification was pursued following confirmation of none-inducibility. Future work should focus on a systematic assessment of PF maps and CT derived EAT to establish the most accurate thresholds for PF to identify EAT as well as integrate information about electrode-tissue coupling (catheter contact) as another important confounder for "false" low voltage.