An Unusual Case of Pacemaker‐Mediated Tachycardia With Alternating Ventricular Capture. What Is the Mechanism?

A 75-year-old man presented for a routine follow-up at the device clinic. He had received a dual-chamber pacemaker with left bundle branch pacing (Zenex DR St. Jude Medical Abbott, St. Paul, MN, USA; and Medtronic 3830 Selectsecure) 3 months earlier. Prior to pacemaker implantation, the patient presented with syncope due to complete atrioventricular block, which required the implantation of a temporary pacemaker. Transiently, the patient exhibited an escape rhythm at 20 beats per minute with a narrow QRS morphology. CSP was confirmed at the time of implantation by observing transition from left septal pacing to selective left bundle branch pacing (sLBBP) during the ventricular capture test at an output of 0.8 V at 0.4 ms. The device settings were DDD 60–130 bpm, with a sensed atrioventricular interval (sAVI) of 150 ms and paced AVI (pAVI) of 180 ms and postventricular atrioventricular refractory period (PVARP) of 225–275 ms. The ventricular lead output was programmed in a unipolar configuration with the autocapture control activated. At 3-month follow-up visit, the patient reported intermittent dizziness and progressive exertional dyspnea since implantation, as well as resting episodes of 100–110 bpm tachycardia observed on his smartwatch, prompting visits to the emergency department. Device interrogation revealed several pacemaker-mediated tachycardia (PMT) events at 100–115 bpm. A 12-lead electrocardiogram (Figure 1A,B) showed an incessant spike-preceded narrow complex tachycardia at 103 bpm with two alternating left bundle branch area capture patterns. What diagnostic information can be deduced from these tracings? Conduction system pacing (CSP) has emerged as a more physiological alternative to right ventricular pacing 1. Recent evidence suggests its use not only as a biventricular resynchronization alternative but also for bradycardia indications with high expected pacing demands in order to prevent pacing-induced left ventricular dysfunction 2. This report outlines a case of pacemaker-mediated tachycardia in which changes in left bundle area capture patterns are responsible for unique, unexpected findings regarding this entity. To confirm the diagnosis of PMT, atrial sensing was previously analyzed in detail, excluding far-field R-wave sensing as well as the presence of signals during the atrial blanking period. Once PMT was confirmed, we initially attempted to interrupt the tachycardia by prolonging the PVARP. However, this was not feasible without compromising a reasonable maximum tracking rate (MTR), due to long and variable ventriculoatrial (VA) intervals. Therefore, the PMT prevention algorithm was unable to terminate the tachycardia, which exhibited an incessant behavior. Consequently, a concealed, slowly conducting accessory pathway was suspected, and an electrophysiological study was performed. A multipolar catheter was positioned within the coronary sinus, and a 4 mm-ablation non-irrigation catheter at the His bundle or right ventricle (RV). Baseline rhythm under VVI programmed at 30 bpm showed complete anterograde AV block without spontaneous ventricular escape (Figure 2A). Programmed ventricular stimulation showed retrograde VA conduction, with a concentric atrial activation sequence and a decremental VA/VH interval, eventually leading to VA block without a retrograde His signal (Figure 2B). VH timing was measured from the onset of the stimulus artifact to the onset of the His recording in the lead of interest. In the presence of retrograde conduction through the left bundle branch: In the absence of retrograde conduction through the left bundle branch: The long VH interval would be attributable to slow conduction across the septum until engagement of retrograde conduction through the right bundle branch. In both scenarios, the short VH interval observed after left septal capture could be explained by a more rapid incorporation of the impulse into the right bundle branch (Figure 3). Finally, an auriculoventricular node ablation was successfully performed during tachycardia, confirming retrograde VA block by right ventricle stimulation. Tachycardia was no longer inducible after switching to DDD mode. PMT is a reentrant tachycardia traditionally observed in patients with dual-chamber transvenous pacemakers 3. The underlying mechanism involves ventricular pacing triggering retrograde conduction through the atrioventricular (AV) node or other structures, resulting in atrial depolarization. This atrial activity is subsequently sensed and tracked by the pacemaker, leading to sustained ventricular pacing. Persistent retrograde conduction establishes a reentrant circuit, potentially resulting in an “endless loop” tachycardia. The retrograde P wave that initiates PMT can be triggered by 1: premature ventricular contraction 2 loss of atrial capture 3; atrial ectopy 4; the conclusion of automatic ventricular threshold testing performed in VVI mode, which generates a retrograde P wave that is sensed upon switching back to DDD mode 5; inappropriately long AV delay promoted by atrioventricular search algorithms to minimize right ventricular pacing. Another VA-conduction–dependent pacemaker arrhythmia is repetitive nonreentrant ventriculoatrial synchrony (RNRVAS). In contrast to PMT, RNRVAS occurs when retrograde P waves fall within the PVARP, resulting in functional undersensing and leading to atrial pacing without capture, typically at slower rates. In patients with PMT, applying a magnet over the device to shift it to an asynchronous mode (DOO) or switching to non-atrial tracking modes (DDI or DVI) can be used to manually interrupt the tachycardia. Identifying retrograde conduction, assessing the VA conduction interval, and, if feasible, programming a PVARP of adequate duration to prevent the detection of retrograde P waves are essential to preventing the onset of PMT. However, this may limit the device's upper tracking rate (PVARP + AV delay) and consequently affect automatic AV delay adjustments. Extensions of PVARP following a detected ventricular premature contraction, as well as various PMT termination algorithms, have been developed to address this challenge. Although our patient's pacemaker was programmed with a specific PMT termination algorithm, it required detection of eight consecutive AS–VP cycles and ± 16 ms stability of VP–AS intervals 4. If these conditions were met, an AV delay prolongation (50 ms) or a change in the MTR would be deployed to check for subsequent VA variations, confirming PMT if retrograde VA conduction remains unaffected. When PMT is confirmed, the next AS is not followed by a VP, and an AP is delivered 330 ms after the sensed P wave. Successful atrial capture breaks the cycle and prevents retrograde conduction after the next ventricular-paced complex. The PMT prevention algorithm failed to stop our patient's PMT, probably due to prolonged and variable VA conduction. Data on this specific entity in a similar scenario have been reported in the case of “left ventricle-only” stimulation in CRT-P patient 5. In our case, we observed VH prolongation consistently coinciding with selective LBBP, leading to VA prolongation and subsequent CL modifications (Figure 3), which were essential to evade the PMT prevention algorithm. Although transition from ns-LBBP to s-LBBP is mainly observed shortly after lead fixation 6, because of myocardial threshold elevation due to the current of injury (COI), changes in CSP patterns can be observed during follow-up, as we observed during threshold test at the device clinic. Retrograde VA conduction in conventional right ventricular stimulation in apex or septum often shows shorter VA intervals that are more suitable for PTM prevention algorithms. Therefore, our case illustrates that dual-chamber transvenous pacemakers with LBBAP can produce pacemaker-mediated reentrant arrhythmias with long VA conduction times that could be challenging to treat with conventional reprogramming options. Close attention must be taken to different grades of CSP transition patterns in order to understand and manage this unique entity. Josué López-Baizán has been directly involved in patient care and drafted the manuscript. Juan Acosta has been directly involved in patient care, has undertaken and interpreted patient investigations, and has reviewed the manuscript for significant intellectual content. Manuel Frutos-López has been directly involved in patient care and reviewed the manuscript for significant intellectual content. Marcos Baz-González has been directly involved in patient care. Juan Antonio Sánchez-Brotons has been directly involved in patient care. Eduardo Arana-Rueda has been directly involved in patient care. Alonso Pedrote has been directly involved in patient care. The authors have nothing to report. The authors have nothing to report. The data that support the findings of this study are available from the corresponding author upon reasonable request.