Assessment of scar between atria, rhythms and atrial fibrillation types: does ultra high density mapping offer new insights?

FUNDING ACKNOWLEDGEMENTS: Type of funding sources: Private company. Main funding source(s): Boston Scientific BACKGROUND: 3D electroanatomical mapping systems demonstrate atrial scar by recording local tissue voltages as a surrogate. Increased scar has been shown with both progression of atrial fibrillation, (AF, paroxysmal (PAF) to persistent (PeAF)), and between sinus rhythm (SR) and AF on low density mapping systems. The introduction of ultra high density mapping allows a more detailed comparison. OBJECTIVES: Assessment of atrial scar determined by voltages and surface area between atria, rhythm, and AF type. METHODS: Ultra high density 3D electroanatomical maps of left and right atria (LA, RA) were created. PeAF patients had maps created in AF and SR following DC cardioversion. PAF patients had maps created in SR only. Electrograms (EGMs) on corresponding AF and SR maps were paired using exported co-ordinates. Percentage surface area of scar was and assigned as low voltage myocardium (LVM, ≤ 0.05mV), intermediate, (IVM, 0.05 - 0.5mV), or normal, (NVM, > 0.5mV). RESULTS: 95 maps were produced using 913480 EGMs. Paired SR-AF EGM voltages had a moderate correlation in the LA (Bipolar: Pearson’s R = 0.32, unipolar: R = 0.36, both p <0.0005), and weak correlation in the RA, (Bipolar: R = 0.19, unipolar: R = 0.25, both p < 0.0005). Paired voltages were significantly higher in SR in both atria, LA/RA, bipolar/unipolar all p < 0.0005). There was significantly more LVM/IVM percentage surface area in AF over SR in the LA, (SR to AF - LVM: 6.4 ± 6.9 to 12.2 ± 9.5; IVM 41.0 ± 12.1 to 50.5 ± 10.0; NVM: 51.7 ±17.2 to 37.3 ± 16.1, p < 0.0005), but not in the RA (SR to AF – LVM: 6.5 ± 5.3 to 8.4 ± 8.3; IVM 42.0 ± 10.1 to 44.6 ± 6.9;NVM: 51.5 ± 13.6 to 47.0 ± 13.8, p = 0.62) Figure 1A. Global tissue voltages in SR were significantly higher in PAF over PeAF in both atria, (LA –bipolar: 1.13 vs 0.52mV, unipolar 1.96 vs 1.44mV, RA – bipolar 0.93 vs 0.57mV, unipolar: 1.68 vs 1.33mV all p < 0.0005). There was significantly more LVM/IVM surface area demonstrated in PeAF over PAF in the LA, (PAF to PeAF - LVM: 0.6 ± 0.7 to 6.6 ± 6.8; IVM 29.5 ± 12.3 to 43.1 ±11.4; NVM: 69.8 ± 12.8 to 50.3 ± 16.4, p = 0.01), but not in the RA, (PAF to PeAF - LVM: 4.6 ± 3.6 to 6.5 ± 5.3; IVM 32.7 ± 9.5 to 42.0 ±10.1; NVM: 62.7 ± 12.1 to 51.5 ± 13.6, p = 0.20). Figure 1B Comparing atria, the LA had a significantly lower voltage in SR, but this difference was small (LA/RA = 0.89, p < 0.0005). The percentage surface area of scar was comparable, (LA to RA - LVM: 4.0 ± 5.6 to 5.6 ± 4.5; IVM: 38.4 ± 37.6 to 37.6 ± 10.6; NVM: 57.5 ±18.4 to 56.8 ± 13.8). Figure 1C CONCLUSIONS: Ultra high density mapping confirms atrial voltages are higher in SR over AF, and PAF over PeAF. The RA is comparable to the LA for tissue voltages but does not demonstrate a change in surface area of scar between AF and SR or PAF and PeAF. This may reflect the complexity of AF wavefronts in the LA compared to the RA leading to more bipole orientation related overestimation of scar in the fibrillating LA. [Figure: see text]