Purkinje network and myocardial substrate at the onset of human ventricular fibrillation: implications for catheter ablation

Abstract

Aims: Mapping data of human ventricular fibrillation (VF) are limited. We performed detailed mapping of the activities underlying the onset of VF and targeted ablation in patients with structural cardiac abnormalities.

Methods and results: We evaluated 54 patients (50 ± 16 years) with VF in the setting of ischaemic (n = 15), hypertrophic (n = 8) or dilated cardiomyopathy (n = 12), or Brugada syndrome (n = 19). Ventricular fibrillation was mapped using body-surface mapping to identify driver (reentrant and focal) areas and invasive Purkinje mapping. Purkinje drivers were defined as Purkinje activities faster than the local ventricular rate. Structural substrate was delineated by electrogram criteria and by imaging. Catheter ablation was performed in 41 patients with recurrent VF. Sixty-one episodes of spontaneous (n = 10) or induced (n = 51) VF were mapped. Ventricular fibrillation was organized for the initial 5.0 ± 3.4 s, exhibiting large wavefronts with similar cycle lengths (CLs) across both ventricles (197 ± 23 vs. 196 ± 22 ms, P = 0.9). Most drivers (81%) originated from areas associated with the structural substrate. The Purkinje system was implicated as a trigger or driver in 43% of patients with cardiomyopathy. The transition to disorganized VF was associated with the acceleration of initial reentrant activities (CL shortening from 187 ± 17 to 175 ± 20 ms, P < 0.001), then spatial dissemination of drivers. Purkinje and substrate ablation resulted in the reduction of VF recurrences from a pre-procedural median of seven episodes [interquartile range (IQR) 4-16] to 0 episode (IQR 0-2) (P < 0.001) at 56 ± 30 months.

Conclusions: The onset of human VF is sustained by activities originating from Purkinje and structural substrate, before spreading throughout the ventricles to establish disorganized VF. Targeted ablation results in effective reduction of VF burden.

Key question: The initial phase of human ventricular fibrillation (VF) is critical as it involves the primary activities leading to sustained VF and arrhythmic sudden death. The origin of such activities is unknown.

Key finding: Body-surface mapping shows that most drivers (≈80%) during the initial VF phase originate from electrophysiologically defined structural substrates. Repetitive Purkinje activities can be elicited by programmed stimulation and are implicated as drivers in 37% of cardiomyopathy patients.

Take-home message: The onset of human VF is mostly associated with activities from the Purkinje network and structural substrate, before spreading throughout the ventricles to establish sustained VF. Targeted ablation reduces or eliminates VF recurrence.

Keywords: Ablation; Brugada syndrome; Cardiomyopathy; Purkinje system; Sudden cardiac death; Ventricular fibrillation.

Structured Graphical Abstract

Ventricular fibrillation (VF) onset in humans—Purkinje and structural substrate govern the transition from trigger to disorganized VF. Schematic view of initial VF activities in patients with cardiac structural abnormalities. The upper panel shows an electrocardiogram of spontaneous VF onset in a patient with a prior history of myocardial infarction. The three illustrations show the sequence of trigger, initial organized VF, and disorganized VF. The trigger is shown as a red star close to the structural substrate (mottled white area). Initial VF activities are represented as localized waves generated from the ventricular or Purkinje substrate. Then the acceleration of activities in parallel with previously described changes (reduction in action potential duration, Ca handling…) leads to dissemination of activities and VF disorganization.

Figure 1

Evolution of ventricular fibrillation in the initial seconds—transition to disorganized ventricular fibrillation. (A) Evolution of ventricular fibrillation frequency (top) and organization (bottom) as a function of time. Ventricular fibrillation cycle length is measured every 2 s on body-surface mapping and the average of right ventricle and left ventricle cycle lengths used. The maximal cycle length decrease over all recorded ventricular fibrillation time (P < 0.001*) occurred in the initial 4 s (window 2–4 s compared with 0–2 s). Ventricular fibrillation organization is quantified every second by a non-dipolar component index (NDI) measuring QRS complex dispersion over the body-surface leads. The duration of organized phase was based on an NDI value <3% indicating organized ventricular fibrillation, whereas a value ≥3% indicated disorganized ventricular fibrillation. (B) Transition to disorganized ventricular fibrillation in one patient. Large waveforms on the electrocardiogram and discrete endocardial electrograms are seen at the initial 4 s. Subsequently, the acceleration of the frequency is associated with fragmentation of endocardial electrograms (at 4–5 s) then fine fibrillatory waves on the electrocardiogram. The spatial organization assessed by the non-dipolar component index is indicated for every second.

Figure 2

Rotational activity: serial images and acceleration during the initial seconds of ventricular fibrillation. (A) Patient with dilated cardiomyopathy and rotational activity in the right ventricle (total 12 rotations, the driver area in this patient, is shown in Figure 4B). Rotational activity is shown as full phase progression around a centre-point in colour-coded phases of wave propagation. The series of images show two full rotations (cycle length 150 ms) in colour-coded activation maps around a pivot point (asterisk in first and last image). (B) In another patient with dilated cardiomyopathy, rotational activities occurred in lateral left ventricle at the beginning of ventricular fibrillation (1–1.9 s) and one second later (3.1–4.3 s), in addition to other activities (not shown). The top trace shows a sample unipolar electrogram. Rotational activity is shown as full phase progression around a center-point in colour-coded phases of wave propagation. Coloured points indicate the sites where unipolar electrograms are recorded around the pivot point and show sequential activation (timing shown by red points). The mean local cycle length is 191 ms during the 1–1.9 s ventricular fibrillation interval and has decreased to 172 ms during the 3.1–4.3 s ventricular fibrillation interval.

Figure 3

Main drivers during initial organized ventricular fibrillation. Epicardial views in anterior, inferior, and left lateral projections, showing driver distribution during the first 5 s after ventricular fibrillation initiation. The dotted lines depict the interventricular region (IVR—indicating septal projection) along the left anterior descending and interventricular coronary arteries. Rotational activities are shown in red areas. The numbers indicate the number of rotations at each area. Main drivers (as defined in the text) are indicated by white arrows. Note that drivers are present in interventricular region in all cases. CM, cardiomyopathy.

Figure 4

Driver locations in relation to substrate. In three patients, the locations of drivers are shown in relation to structural substrate defined on electrophysiological criteria or scar imaging. (A) A case of ischaemic cardiomyopathy with drivers (full circle) in the inferior left ventricle associated with myocardial infarction scar on magnetic resonance imaging (middle) and low voltage area on electrogram mapping (right). There is an overlapping of driver area with the substrate (dotted circle) defined on imaging or mapping technique. (B) A case of dilated cardiomyopathy with a right ventricle driver area (full circle) related to low voltage (middle) and fractionated and late electrograms (right). This patient had no scar in the right ventricle on magnetic resonance imaging. Note that the driver is located at the border of abnormal electrogram substrate. (C) A case of Brugada syndrome with drivers in the RV outflow tract and inferior right ventricle (left), overlapping with abnormal epicardial electrograms (right). After ablation at the RV outflow tract, ventricular fibrillation was still inducible by two extrastimuli with persistent drivers in the inferior right ventricle. Ablation at the inferior right ventricle resulted in ventricular fibrillation non-inducibility and no recurrence of ventricular fibrillation for 62 months. Note that inferior right ventricle displays mildly abnormal electrograms not fulfilling all abnormality criteria (electrogram duration is 55 ms instead of being >70 ms). The colours of driver areas are not uniform as different system versions have been used. AP, anteroposterior; LAD, left anterior descending coronary artery.

Figure 5

Inducible repetitive activity in distal Purkinje in hypertrophic cardiomyopathy. The upper panels show a recording in sinus rhythm and ventricular fibrillation induction with programmed stimulation in a 29-year-old man. A decapolar catheter along the distal fascicle indicates that each of the initial ventricular beats is associated with Purkinje activity (red asterisks); the latter has a more rapid cycle length (mean 219 ms) than the local ventricle (230 ms). The lower panels show a recording in sinus rhythm and ventricular fibrillation induction with programmed stimulation in a 63-year-old man. Ventricular fibrillation induction is associated with 1:1 Purkinje activity. The multispline catheter shows Purkinje activities on several bipoles (red asterisks); Purkinje cycle length (mean 237 ms) is shorter than local ventricular cycle lengths (252 ms). Another ventricular fibrillation window occurring 1 s later is also shown. Note that ventricular electrograms are significantly fractionated and disturb Purkinje recognition. This is clearly seen (five small arrows) in the last paced beat inducing ventricular fibrillation. In both patients, ventricular fibrillation was non-inducible after Purkinje ablation. The ablation lesions targeting the Purkinje arborization (red tags) are shown for the upper case. In both patients, the 12-lead electrocardiograms of ventricular fibrillation induction and voltage maps are shown in Supplementary material online, Figure S6.

Figure 6

Number of ventricular fibrillation episodes terminated by implantable cardioverter-defibrillator shocks, before and after ablation—individually and full group. The left panel shows the number of ventricular fibrillation episodes terminated by implantable cardioverter-defibrillator shocks, before and after ablation, in individual patients. The right panel shows actuarial survival curve for ventricular fibrillation recurrence after catheter ablation.