Prospective observational study of consecutive patients with frequent symptomatic PVCs referred for catheter ablation at our center from January to September 2018. These patients underwent ablation using a pre-specified mapping strategy with the PMF software and the PentaRay catheter. Procedural endpoints and acute and 12-month follow-up results were assessed and compared to a 12-month retrospective cohort of patients who underwent complex PVC ablation with standard activation mapping performed with the use of the Thermocool SmartTouch irrigated tip contact force-sensing ablation catheter (ST - Biosense Webster).

All patients provided written informed consent and the study was approved by the local institutional ethics committee.

Patients from both the study and control groups were eligible for inclusion in the study if they presented at least one criteria from group A plus at least one from group B. Group A criteria included: i) symptomatic drug-refractory PVCs, or ii) a decline in left ventricular ejection fraction (LVEF) due to frequent PVCs. Group B criteria included: i) a previous history of PVC ablation, or ii) a left-sided PVC. All patients underwent evaluation for any underlying structural heart disease, and major coronary artery disease had to be ruled out by cardiac angiography or stress testing in cases of LVEF dysfunction.

Exclusion criteria were as follows: presence of structural heart disease (ischemic heart disease, significant valvular disease or genetic or infiltrative cardiomyopathy), pregnancy and any clinical contraindication for the ablation procedure.

All procedures in both the prospective and retrospective groups were performed by the same primary operator. They were conducted under local anesthesia and all anti-arrhythmic drugs (AAD) stopped at least five half-lives before the procedure. For left-sided procedures, a retrograde approach was first choice, but access via transseptal puncture was also obtained whenever necessary. Intravenous heparin was administered to maintain an activated clotting time ≥300 seconds for left-sided procedures.

Anatomical mapping data was collected using a 3D mapping system (CARTO3®, Biosense Webster) and respiratory noise was excluded (AccuResp Module, Biosense Webster). Activation mapping was performed using a high-density system (Confidense® Module, Biosense Webster) and signal annotation conducted using Wavefront activation software (Biosense Webster), which analyses and integrates bipolar signals and the corresponding negative onset of unipolar signal.5

The use of intracardiac echocardiography was left to the physician's discretion.

Ablation was performed using a ST ablation catheter, in power-controlled mode with temperature limited to 43°C and saline irrigation speed of 17 mL/min. The power setting was 20 W and could be titrated up to 40 W, depending on the region being targeted.

When the ectopic focus was located in the coronary cusps, coronary angiography (CA) was routinely performed. If the focus was at a safe distance (>6 mm)2 from the coronary artery, radiofrequency (RF) was delivered. After ablation, CA was performed again in order to exclude any damage to the coronary arteries. When ablation was performed beneath the coronary cusps, CA was not required given the probable distance from the coronary arteries.

After ablation, patients were monitored for at least 30 minutes in de novo cases and 60 minutes in redo procedures to avoid recurrence.

All AAD were discontinued after successful procedures.

Systematic transthoracic echocardiography was performed before discharge to exclude aortic valve damage or pericardial effusion. A 24-h Holter was also performed before discharge.

This new software was released with version 6 of the CARTO3® system. It allows the acquisition of local activation time (LAT) points based on real time comparison of the 12-lead electrocardiogram (ECG) morphology with a morphology of interest (template). Essentially, this software enables a correlation to be made between the predefined morphology and every real time beat.

Each activation point collected with the mapping catheter must follow some intrinsic steps before validation (Figure 1A-H): (A) definition of the pattern – selection of the template morphology to be mapped and adjustment of the window of interest (WOI); (B) single channel correlation – the fit between the template and the current beat is calculated for each of the 12 leads using Pearson correlation formula; calculation of overall weighted – (C) in this step the absolute voltage is measured for each lead (for the template and for the beat); (D) lead correlation – the total sum of each pair of leads (template and beat) is calculated and then (E) divided by the total sum of all leads weight; (F) lead contribution – the lead correlation is multiplied by the weighted average and the result (G) is the final weighted average score; (H) phase shift of ±40 ms – this step takes into consideration a possible offset between reference annotations in beats with similar morphology. Finally, the Pattern Matching percentage is compared to the threshold – if the matching percentage is equal to or higher than the threshold, the beat is considered to represent the template morphology and is incorporated into the activation map.

Figure 1.

(A-H) Pattern Matching Filter (PMF) algorithm workflow (as described in technical presentation for health care professionals; with permission from Biosense Webster). (I-L) Utilization of the PMF software in a patient with premature ventricular contraction arising from the left coronary cusp. I. Selection of the premature ventricular contraction as the template morphology to be mapped. J. Window of interest (WOI) adjustment in order to include only the QRS complex. K. A 95% minimum threshold was selected and during continuous mapping, each premature ventricular contraction and its respective correlation were visualized. L. Since the vast majority of premature ventricular contraction presented a correlation of between 97 and 99%, WOI was adjusted and the minimum threshold was set at 97%. After all the steps were performed, we started collecting geometry and activation points with the PentaRay catheter.

(0.31MB).

The approach was similar to the aforementioned, but the whole anatomy was constructed with the Thermocool SmartTouch catheter, which has a 3.5-mm electrode tip with interelectrode spacings of 1-6-2 mm. After defining the geometry with FAM, we started collecting LAT points. Points were also acquired automatically (position stability=3 mm, maximum density and a minimum tissue contact of 3 g), however each point required manual confirmation before validation and incorporation into the activation map.

The seven patients from the PMF group are described in Table 1 – one was a 13-year old male undergoing a redo ablation for RVOT PVC, while the other six patients had PVCs arising from the LV (five from the LVOT and one from the anterolateral papillary muscle). All seven patients had a high PVC burden (>32 000 PVC/24 h), with multiple and long episodes of bigeminy in all of them. The mean LVEF was 47±8% and three patients had LVEF<45%.

The mean number of points acquired with the PentaRay catheter was 507±213 before ablation during 15±10 min of mapping. In the first patient, who was undergoing a redo ablation of a PVC arising near the His bundle, mapping was carried out with the PentaRay in both the RVOT and the coronary cusps, particularly the non-coronary cusp (NCC) (Figure 2A). However, since the earliest activation site was located in the RVOT, with an optimal unipolar signal, ablation was performed there. In the other patients, the earliest LAT was located in the left coronary cusp (LCC) (n=3) – one in the LCC (Figure 3) and two beneath (Figure 2B), the papillary muscle (n=1), the aortic mitral continuity (AMC) (n=1) and the right (RCC)-NCC commissure (n=1). Median procedure and RF ablation times were 67±42 min and 294±112 s, respectively. Acute success was achieved in all patients (100%). Coronary angiography was performed in three patients – two cases of LCC PVCs and one with PVCs arising from the RCC-NCC commissure. In all patients, the coronary arteries were at a safe distance from the ablation catheter.

Figure 2.

A. Premature ventricular contraction (PVC) activation map, projected in a posterior view, with the earliest activation point located in the right ventricular outflow tract (RVOT) slightly above the His bundle (5 mm away). This high-density activation mapping, performed with the PMF and the PentaRay mapping catheter, included 816 points (379 in the RVOT), resulting in a very detailed and accurate map. B. This high-density activation mapping included 308 points, with the earliest activation point just beneath the left coronary cusp. The figure is projected in the upper left posterior view. Also, in both maps performed with the PMF software and the PentaRay mapping catheter, the bipolar endocardial electrograms were less fractionated and had a higher voltage amplitude signal compared to the bipolar electrogram acquired with the ablation catheter, and so potentially more accurate. C. PVC activation map, projected in upper posterior view, with the earliest activation point in the aortomitral continuity. This activation map was performed with the standard approach and ST ablation catheter. Only 104 local activation time points were collected, since each point needs manual confirmation before validation and incorporation into the activation map, resulting in a less detailed map.

(0.27MB).

Figure 3.

Premature ventricular contraction activation map, projected in an upper anterior view, with the earliest activation point located in the left coronary cusp. Both maps were performed in 12 minutes. A. This high-density activation map was produced with the PMF and the PentaRay catheter and included 547 points, resulting in a very detailed and accurate map. B. Standard activation map performed without PMF. Points were collected with the ablation catheter and required manual validation, resulting in fewer points being collected in the same amount of time – 42 points. Also, the bipolar endocardial electrogram had a lower voltage amplitude signal with larger electrogram duration compared to the bipolar electrogram acquired with the PentaRay catheter.

(0.21MB).

At 12 months post-procedure, there was no clinical recurrence, with all patients presenting a reduction in PVC burden of at least 80% compared with baseline, despite being off AAD. Of the three patients with LVEF<45%, two had complete normalization of LVEF.

Only four patients matched the inclusion criteria (Table 2) and were included in our study – all of them had PVCs from the LV (from the LCC, the RCC, the AMC (Figure 2C) and the posteromedial papillary muscle, respectively). Similar to patients from the PMF group, all four control patients had a high PVC burden (>23 000 PVC/24 h). Two patients had LVEF<45%.

During 36±24 min of mapping, the mean number of points acquired with the ablation catheter was 90±62. Median procedure and RF ablation times were 130±54 min and 706±613 s, respectively. Acute success was achieved in three of the four patients. The patient with PVCs arising from the RCC had a recurrence during the waiting period (mapping in this patient was also performed in the RVOT, but ablation was unsuccessful). Coronary angiography was performed in two patients – with LCC and RCC PVCs, respectively. In both cases, the coronary arteries were at a safe distance from the ablation catheter.

At 12 months post-procedure, the three patients who had had a successful ablation remained symptom free, with a PVC burden reduction of >80%, despite being off AAD. The two patients with a LVEF<45% had full LVEF recovery.