Evaluation of a Fully Automatic Measurement of Short-Term Variability of Repolarization on Intracardiac Electrograms in the Chronic Atrioventricular Block Dog

Background: Short-term variability of repolarization (STV) of the monophasic action potential duration (MAPD) or activation recovery interval (ARI) on the intracardiac electrogram (EGM) increases abruptly prior to the occurrence of ventricular arrhythmias in the chronic AV-block (CAVB) dog model. Therefore, this parameter might be suitable for continuous monitoring of imminent arrhythmias using the EGM stored on an implanted device. However, 24/7 monitoring would require automatic STV-ARI measurement by the device. Objective: To evaluate a newly developed automatic measurement of STV-ARI for prediction of dofetilide-induced Torsade de Pointes (TdP) arrhythmias in the CAVB-dog. Methods: Two retrospective analyses were done on data from recently performed dog experiments. 1.) In 7 anesthetised CAVB-dogs, the new automatic STV-ARI method was compared with the gold standard STV-MAPD at baseline and after dofetilide administration (0.025 mg/kg in 5 minutes). 2.) The predictive value of the automatic method was compared to currently used STV-ARI methods, i.e. slope method and fiducial segment averaging (FSA) method, in 11 inducible (≥3 TdP arrhythmias) and 10 non-inducible CAVB-dogs. Results: 1.) The automatic measurement of STV-ARI had good correlation with STV-MAPD (r2 = 0.89; p<0.001). Bland-Altman analysis showed a small bias of 0.06 ms with limits of agreement between -0.63 ms and 0.76 ms. 2.) STV-ARI of all three methods was significantly different between inducible and non-inducible dogs after dofetilide. The automatic method showed the highest predictive performance with an area under the ROC-curve of 0.93, compared to 0.85 and 0.87 of the slope and FSA method, respectively. With a threshold of STV set at 1.69 ms, STVARI measured with the automatic method had a sensitivity of 0.91 and specificity of 0.90 in differentiating inducible from non-inducible subjects. Conclusion: We developed a fully-automatic method for measurement of STV-ARI on the intracardiac EGM that can accurately predict the occurrence of ventricular arrhythmias in the CAVB-dog. Future integration of this method into implantable devices could provide the opportunity for 24/7 monitoring of arrhythmic risk.