Electrophysiological and histological characterization of atrial scarring in a model of isolated atrial myocardial infarction.

Characterization of atrial myocardial infarction is hampered by the frequent concurrence of ventricular infarction. Theoretically, atrial infarct scarring could be recognized by multifrequency tissue impedance, like in ventricular infarction, but this remains to be proven. This study aimed at developing a model of atrial infarction to assess the potential of multifrequency impedance to recognize areas of atrial infarct scar. Methods: Seven anesthetized pigs were submitted to transcatheter occlusion of atrial coronary branches arising from the left coronary circumflex artery. Six weeks later the animals were anesthetized and underwent atrial voltage mapping and multifrequency impedance recordings. The hearts were thereafter extracted for anatomopathological study. Two additional pigs not submitted to atrial branch occlusion were used as controls. Selective occlusion of the atrial branches induced areas of healed infarction in the left atrium in 6 of the 7 cases. Endocardial mapping of the left atrium showed reduced multi-frequency impedance (Phase angle at 307 kHz: from -17.1° ± 5.0° to -8.9° ± 2.6°, < .01) and low-voltage of bipolar electrograms (.2 ± 0.1 mV vs. 1.9 ± 1.5 mV vs., < .01) in areas affected by the infarction. Data variability of the impedance phase angle was lower than that of bipolar voltage (coefficient of variability of phase angle at307 kHz vs. bipolar voltage: .30 vs. .77). Histological analysis excluded the presence of ventricular infarction. Selective occlusion of atrial coronary branches permits to set up a model of selective atrial infarction. Atrial multifrequency impedance mapping allowed recognition of atrial infarct scarring with lesser data variability than local bipolar voltage mapping. Our model may have potential applicability on the study of atrial arrhythmia mechanisms.