Relative timing of near-field and far-field electrograms can determine the tachyarrhythmia site of origin

Background Despite improving algorithms, inappropriate shocks for supraventricular tachycardia (SVT) still occur in a significant number of patients with implantable cardioverter-defibrillators (ICDs). This makes the discovery of novel discriminators that use existing ICD hardware an attractive proposition. Objective We hypothesized that the delay of activation onset from the device-detected, far-field electrogram (EGM) to the near-field, bipole EGM would allow the differentiation of ventricular tachycardias (VTs) from SVTs. Methods Proof of principle was demonstrated by rapid pacing in the right atrium, right ventricle, and left ventricle in healthy patients undergoing atrial fibrillation ablation procedures (n = 17). Using real-life ICD recordings, the equivalent measurements were made in a derivation cohort (n = 26) and cutoff predictive values obtained. Finally, the selected values were validated in a separate group of recordings (n = 82). Results In healthy patients, significant differences in the far-field to near-field EGM activation onsets were observed between right atrial (14.7 ± 2.7 ms), right ventricular (36.3 ± 8 ms), and left ventricular (57.8 ± 10.3 ms; P < .001) pacing. In the derivation ICD cohort, the median far-field to near-field onset delay was significantly shorter in SVT (24.5 ms; interquartile range, 15.3–47.5 ms) than in VT (118.5 ms; interquartile range, 102.5–131.5 ms) (P < .001). Using a cutoff of 100 ms in the validation cohort, SVT was successfully discriminated from VT with a sensitivity and specificity of 88% and a negative predictive value of 94.2%. Conclusion The delay between far-field and near-field EGMs offers a potential new discrimination tool to reduce inappropriate ICD therapies and aid interpretation of single-lead device tracings. Despite improving algorithms, inappropriate shocks for supraventricular tachycardia (SVT) still occur in a significant number of patients with implantable cardioverter-defibrillators (ICDs). This makes the discovery of novel discriminators that use existing ICD hardware an attractive proposition. We hypothesized that the delay of activation onset from the device-detected, far-field electrogram (EGM) to the near-field, bipole EGM would allow the differentiation of ventricular tachycardias (VTs) from SVTs. Proof of principle was demonstrated by rapid pacing in the right atrium, right ventricle, and left ventricle in healthy patients undergoing atrial fibrillation ablation procedures (n = 17). Using real-life ICD recordings, the equivalent measurements were made in a derivation cohort (n = 26) and cutoff predictive values obtained. Finally, the selected values were validated in a separate group of recordings (n = 82). In healthy patients, significant differences in the far-field to near-field EGM activation onsets were observed between right atrial (14.7 ± 2.7 ms), right ventricular (36.3 ± 8 ms), and left ventricular (57.8 ± 10.3 ms; P < .001) pacing. In the derivation ICD cohort, the median far-field to near-field onset delay was significantly shorter in SVT (24.5 ms; interquartile range, 15.3–47.5 ms) than in VT (118.5 ms; interquartile range, 102.5–131.5 ms) (P < .001). Using a cutoff of 100 ms in the validation cohort, SVT was successfully discriminated from VT with a sensitivity and specificity of 88% and a negative predictive value of 94.2%. The delay between far-field and near-field EGMs offers a potential new discrimination tool to reduce inappropriate ICD therapies and aid interpretation of single-lead device tracings.