Diagnostic reliability of AV synchrony self-diagnostics in leadless VDD pacemakers

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Introduction Leadless pacemakers (PMs) capable of atrio-ventricular (AV) synchronous pacing have been introduced recently. These devices provide mechanical atrial sensing by detection of the atrial contraction via an accelerometer. Atrial tracking may be disturbed by external influences such as body motions and higher heart rates in real life. To track the amount of AV synchronous pacing, the device statistic classifies all sensed and paced QRS complexes according to presumed AV synchrony. The reliability of this self-diagnostics to estimate the true degree of AV synchrony, however, is insufficiently studied. Purpose To investigate the informative value of the device statistics offered by leadless VDD PMs regarding true AV snychrony. Methods We prospectively included all patients who received a leadless VDD PM between 07/2020 and 05/2021 at our center in this observational study. During the regular outpatient follow-ups, device interrogation was performed, which included evaluation of the PM statistics. True AV synchrony (defined as a QRS complex preceded by a p-wave within 300ms) was analyzed repeatedly during follow-up using Holter ECGs. Results We analysed 34 Holter ECGs from 20 outpatients (816 hours of ECG in total). Patients had a median age of 80 years (interquartile range 76-86 years), 55% were females. For Holter ECG sequences that showed high degree or complete AV-Block (>80% ventricular pacing), the percentage of paced beats that were presumed to be AV synchronous by the device statistic (ratio "AM-VP"/total VP) correlated well with AV synchrony as assessed using Holter-ECGs (Kendall’s τ=0.312, p<0.001). AV synchrony in the Holter ECG was different in patients with <66.6% presumed AV synchrony than in patients with >66.6% presumed AV synchrony (p<0.001, figure). For Holter ECG sequences with mostly preserved intrinsic AV conduction (<20% ventricular pacing), the ratio "AM-VP"/total VP was not predictive for true AV synchrony (Kendall’s τ=0.07, p=n.s.). In this situation, however, "VS only" (Kendall’s τ=0.110, p=0.005) correlated with true AV synchrony (due to AV conduction mode switch) and "VP only" showed an inverse correlation with AV synchrony (Kendall’s τ=-0.215, p<0.001). Conclusion Leadless PMs provide device statistics that show a correlation with true AV synchrony. The clinical setting as well as the device programming (e.g. AV conduction mode switch) significantly influence the predictive value of the different parameters of the device’s statistics.