Premature ventricular complexes in survivors of idiopathic ventricular fibrillation: short-coupled near major arrhythmic events, but longer-coupled remotely

Idiopathic ventricular fibrillation (VF) and polymorphic ventricular tachycardia (pVT) are often triggered by short-coupled premature ventricular complexes (SC-PVCs), with coupling intervals (CI) <350 ms. However, the timing, burden, and conditions of these PVCs remain largely unknown.

To investigate the CIs and burden of (SC-)PVCs in idiopathic VF/pVT survivors, their relation to the time since the last event (i.e., VF/pVT episodes, appropriate ICD shocks, or antitachypacing), and to compare their characteristics between day and night.

In 48 idiopathic VF/pVT survivors (26 male, 21 on β-blockers, no other antiarrhythmics), we analyzed extended ECG recordings (duration 42 [26-55] minutes, reported as median [Q1-Q3]), and extracted the shortest CI at rest. These were correlated to the time passed since the last event using a linear model, and tested for significance with an F-test. Differences between β-blocker users and non-users were assessed with a Mann-Whitney U test or chi-squared test.

From 15 of these patients, 37 Holter recordings (24 hours each; 2 [1-4] recordings/patient; 13 [10-287] days post-event) were analyzed for day-night differences in CI, hourly PVC burden, and median RR intervals preceding PVCs, using a linear mixed-effects model, and tested for significance with a t-test.

In the extended ECG recordings, a positive correlation was observed between the shortest PVC CI and the logarithmic time since the last event (p=0.03, Fig. 1A). PVC occurrence was more frequent in β-blocker users than in non-users (18/21 vs. 14/27, respectively, p=0.05), but SC-PVC occurrence (2/21 vs. 3/27, p=0.98), CI (410 [379-439] vs. 436 [411-482] ms, p=0.22), or time since the last event (483 [106-2022] vs. 446 [40-3262] days, p = 1.00) were not significantly different (Fig. 1B).

Holter analysis revealed that the shortest CI was shorter during daytime than at night (341 [325-373] vs. 392 [353-458] ms, p=0.0001), and the same was true for median RR-intervals preceding PVCs (762 [709-802] vs. 864 [812-976] ms, p<0.0001) (Fig. 1C). No significant differences were found in the hourly PVC burden (63 [3-259] vs. 5 [1-163]/hour, p=0.78) or the hourly SC-PVC burden (0 [0.1-0.4] vs 0 [0-0]/hour, p=0.29) (Fig. 1D).

In two patients with longitudinal Holter data, an arrhythmia recurred in between recordings, coinciding with a shorter CI, which later increased again in one patient (Fig. 2).

In survivors of idiopathic VF/pVT, PVC CIs correlated positively with time since the last event, and were shorter during the day than at night. Occurrence of short-coupled PVCs was less frequent in patients further from the last event, and at night. In two patients, arrhythmia recurrence was linked to a shorter CI. These findings suggest that the potential triggers for idiopathic VF/pVT vary over time. In the presence of a (dynamic) substrate, this variability may help set the stage for arrhythmia initiation.Figure 1  Figure 2