Assessing the field of view of multisize electrodes in ischemic cardiomyopathy by validating against ex-vivo high resolution cardiac magnetic resonance

Abstract

Type of funding sources: Public hospital(s). Main funding source(s): Leiden University Medical Centre Substrate identification after myocardial infarction (MI) relies on voltage mapping. Early reperfusion results in non-transmural scar (NTS). Narrow-spaced microelectrodes (ME) are thought to have a limited field of view (FOV). The role of ME for delineation of NTS is unclear. To evaluate mapping with multi-size electrodes for identifying NTS, validated against high-resolution ex-vivo cardiac magnetic resonance imaging (HR-LGE-CMR). Nine swine with early reperfusion MI underwent endocardial electroanatomical voltage mapping (EAVM) with the QDOT catheter which incorporates three ME in the 3.5mm tip electrode. HR-LGE-CMR (0.3mm slices) were obtained and merged with EAVM. At each EAVM point a transmural cylinder (5mm radius) was projected on the CMR and the volume of viable myocardium (VM) in the cylinder quantified (Otsu method). Unipolar (UV) and bipolar (BV) voltages from conventional (c) and microelectrodes (µ) were related to VM. Cut-off values for normal myocardium were based on 5th percentiles of areas without fibrosis. In each swine 220 (IQR 216-260) mapping points were collected (total 2322 points). Cut-off for normal myocardium were 3.18 mV, 0.85 mV, 3.28mV and 1.93 mV for UVc, BVc, UVµ and BVµ, respectively. Wall thickness (WT) was reduced in areas with fibrosis vs. no fibrosis (5.4mm, IQR 3.2- 6.9 vs. 7.4mm, IQR 5.3 - 9.6). All voltages were reduced at sites with fibrosis vs. no fibrosis (UVs 4.4mV vs. 6.8mV; BVc 1.2mV vs. 2.1mV; UVµ 2.5mV vs. 5.5mV, and BVµ 2.9mV vs. 5.4mV, all p<0.001). For areas with any fibrosis, all voltages increased with increasing WT up to the maximal WT of 13mm (fig. 1). Similarly, all voltages increased with an increase in VM volumes from >200mm3 to >600 mm3 (equivalent to a cylinder with h=7.64mm) (fig. 2) with the strongest correlation for UVµ (r=0.47). Below a volume of 200mm3 VM voltages did not correlate significantly with VM. In NTS, UVc, BVc, and, notably, BVµ and UVµ increase with increasing WT and increasing transmural volume of VM, with the biggest role perhaps laid out for UVµ to estimate transmural VM. EAVM cannot accurately delineate areas with the lowest amount of VM, potentially due to insufficient far field cancellation in NTS. Both these findings argue against a limited FOV of ME.