Spatial distribution of electrical reconnection after pulmonary vein isolation in patients with recurrent paroxysmal atrial fibrillation

This study was approved by the Medical Ethics Committee of our institution and all patients gave written informed consent.

Data were collected between November 2011 and June 2014. Eighty-six consecutive patients previously successfully treated with PVI for paroxysmal AF and accepted for a second procedure due to recurrent symptomatic paroxysmal AF were eligible for inclusion in this study. A team consisting of four experienced electrophysiologists performed all PV isolations (both initial and second procedures). Patients who had an unsuccessful first procedure, i. e. no acute and complete isolation of all PVs, or who had more than one previous PVI, were not eligible for this study. In addition, patients with a left common PV, with five PVs or with atrial tachyarrhythmias other than AF were also excluded from this study.

The initial procedure was performed using one of two different techniques, both routinely performed. In one group of patients (n = 38), a single tip, 4 mm irrigated RF ablation catheter (s-RF) was used in combination with a CARTO (Biosense Webster Inc., CA, USA) or EnSite NavX (St. Jude Medical Inc., MN, USA) navigation system. In this population point-by-point wide antral circumferential ablation was performed, with an irrigated single tip ablation catheter without contact force feedback (information about adequacy of tip-to-tissue contact, aimed to optimise energy delivery), since this option was not available at the time of the index procedure. In the other group (n = 48), a multi-electrode, duty-cycled (technology combining uni- and bi-polar ablation) RF catheter (PVAC, Medtronic Inc., MN, USA) was used with an EP navigator (Philips Medical Systems, Eindhoven, the Netherlands) for imaging support. In patients treated with the latter catheter type, a non-steerable transseptal sheath was used. For both groups, disappearance of PV signals and exit block during pacing from the PVs were used to assess procedural acute success after a waiting time of 30 minutes. Adenosine or other additional means to assess dormant reconnections were not used [9].

All patients were treated according to the same routine procedure for recurrence of paroxysmal AF, independent of the technique used in the initial procedure. In this second procedure only X-ray was used for imaging. In patients treated with warfarin, oral anticoagulation was not interrupted, whereas novel oral anticoagulants (NOACs) were stopped 24 hours before the procedure. During the intervention, all patients were heparinised to achieve an activated clotting time >300 seconds, after double transseptal puncture. All PVs were assessed for the presence and the exact location of electrical reconnection using a circular mapping catheter (either a Lasso™ [Biosense Webster Inc., CA, USA] or a Reflection™ [St. Jude Medical Inc., MN, USA] catheter). A reconnection site was defined as the location in the previous ablation line that was associated with the site of earliest activation as assessed by the circular mapping catheter, which was placed just distal from the ablation line. In case of multiple sites of reconnection, the site of earliest activation was always targeted first. The location of reconnection was expressed by projecting a virtual clock around each PV ostium with the left PVs and right PVs in RAO and LAO projections, respectively. The operator determined, based on X‑ray imaging, the location of reconnection by linking it to one of the 12 hours in the virtual clock. In the left PVs, 3 and 9 o’clock corresponded to the centre of the anterior and posterior quadrant, respectively, in contrast to the right PVs where 3 o’clock and 9 o’clock corresponded to the centre of the posterior and anterior quadrant, respectively. See Fig. 1 for schematic representation. Sites of reconnection were ablated using a single tip irrigated RF ablation catheter (CELSIUS Thermo-Cool, Biosense Webster Inc., CA, USA or Quickflex, St. Jude Medical Inc., MN, USA). Each ablation was applied for 60 seconds at a maximum of 30 Watts and was repeated if necessary. After successful local ablation, the PV was checked for additional reconnections and treated in the same way as described above, if necessary until complete PVI was achieved.

All patients had reconnection of at least 1 PV, with 2.9 ± 0.9 and 3.2 ± 0.7 reconnected PVs for patients with previous s‑RF and PVAC procedures, respectively (p = 0.193). In each reconnected PV, there were 2.7 ± 1.6 and 2.9 ± 1.7 gaps of electrical connection found (s-RF and PVAC group, respectively [p = 0.779]) (Tab. 2).

There were predilection sites of electrical reconnection depending on the initial ablation technique. Fig. 2 displays the spatial distribution of electrical reconnection in each of the four PVs expressed in the 12-segment model. The red shaded area indicates the range with statistical difference (p < 0.05) when comparing the two techniques. The figure demonstrates that after previous s‑RF ablation reconnection is found more frequently superior in the right upper PV whereas in patients who had a previous PVAC ablation, reconnection was more often located inferior in the right lower PV. The figure also shows two additional small areas (1 segment each) in the right upper and left lower pulmonary vein in which there was a statistical difference between both techniques.

The duration of RF-energy application per gap needed to achieve electrical re-isolation was 72 ± 36 and 75 ± 44 seconds for patients with previous s‑RF and PVAC ablation, respectively (p = 0.892). In both groups, the duration of RF-energy application for re-isolation was equally distributed between the PVs (Tab. 2) and there were no spatial differences with respect to the 12-segment model.