A local registry of patients who had PVI with the use of vHPSD was prospectively observed and analysed. The high-power short duration (HPSD) and STSF groups were retrospectively collated following compatibility with inclusion criteria. Patients over the age of 18 with first time PVI ablation for symptomatic atrial fibrillation were included. Subjects with any previous catheter or surgical ablation were excluded.

This study was registered as a quality improvement project into the current RF technologies utilised for PVI procedures at the University Hospitals of Leicester.

Procedures were either carried out under general anaesthetic (GA) or sedation with local anaesthesia. Patients listed for sedation and local anaesthesia procedures were given intravenous (IV) diazepam 5 mg, morphine 5 mg, metoclopramide 10 mg and 1 g Paracetamol at the start of the procedure. Further bolus doses of IV Midazolam 1–2 mg and morphine were administered as required during the procedure. Venous access was gained via the right femoral vein under ultrasound guidance. All patients underwent 3D electro-anatomical mapping using the CARTO 3 system (Biosense Webster, Inc., Irvine, CA, USA) with either a Lasso or PentaRay catheter (Biosense Webster, Inc., Irvine, CA, USA) following standard trans-septal puncture.

The general ablation strategy for PVI was wide area circumferential ablation (WACA) with substrate ablation conducted in the posterior wall +/– roof +/– anterior wall when indicated.

An open-irrigated tip catheter (QDOT MICRO, Biosense Webster, Inc., Irvine, CA, USA) was utilised. Visualizable steerable sheaths were used by certain operators when required however were not routinely used. The QDOT MICRO ablation catheter incorporates six thermocouples situated in the catheter tip, facilitated for accurate local temperature measurement. The positioning of the electrodes is optimised to monitor the temperature at both perpendicular and parallel catheter orientations which provides the basis for a susceptible feedback system of catheter-tissue interface temperature and thus, catheter stability during energy application [6, 7]. During the RF application, the vHPSD algorithm continually modulates the power based on the greatest surface temperature detected by the thermocouples. The primary setting utilised for the QMODE+ group was 90 W, 4 s; irrigation at 8 mL/min; recommended contact force setting of 5–25 grams. The temperature setting followed a cut-off of 65 degrees Celsius. This approach promotes resistive heating and reduces conductive heating [5, 8, 9].

The QDOT MICRO Ablation catheter was also utilised for conventional ablation in the HPSD patient cohort. Within this mode, the system adjusts the irrigation flow rate and power based on the recorded temperature to stabilize the catheter tip temperature.

The Thermocool Smart-touch SF (Biosense Webster, Inc., Irvine, CA, USA) open-irrigated tip catheter was used for conventional ablation in the STSF patient cohort. Ablation was conducted in the power-control mode with energy application limited to 40 W at the anterior segments and guided by the ablation index (AI) [10].

An activated clotting time (ACT) was maintained between 300–400 ms and was checked regularly (in 15–30 min interval) as per ACT once access to the left atrium was achieved. Operators aimed for an overlap of anterior wall lesions and thus intralesional distance was $\le$4 mm, while posterior wall lesions was 6 mm. PVI was confirmed with entrance and exit block using pacing manoeuvres and/or adenosine challenge to confirm absence of pulmonary vein (PV) reconnection. If PV conduction was still apparent, additional lesions were performed where indicated with either the vHPSD or standard power setting (40–50 W in QMODE) respectively if the gap was on the anterior (90 W) or posterior (50 W) portion of the left atrium. The durability of PVI lesions was observed with a wait period of 30 minutes, to ensure there was no further acute reconnection. Cavotricuspid isthmus (CTI) ablation was performed with standard RF ablation technique when typical atrial flutter was documented previously or observed during the procedure. Substrate lesions were performed at the discretion of the operator for persistent AF.

The primary outcomes were focused on intraoperative variables corresponding to the following: PVI ablation duration (time between from the first lesion of the WACA to confirmation of PV isolation of the last ablation lesion), duration of energy application for PVI, short-term efficacy of procedure and incidence of procedural adverse events (PAE). The short-term efficacy of the procedure was determined if PVI was confirmed following pacing manoeuvres, adenosine or isoproterenol challenge. The procedural metrics were collected via the CARTO 3 system (Biosense Webster, Inc., Irvine, CA, USA) and LABSYSTEM™ PRO EP Recording system (Boston Scientific, Inc., Marlborough, MA, USA).

The outcome corroborating to safety was derived from the incidence of adverse events which included cardiac tamponade or perforation, significant vascular access complication, stroke, transient ischaemic attack and periprocedural mortality. As per post-procedural management, patients were monitored as inpatients overnight and received a focussed transthoracic echocardiogram to exclude any pericardial effusion immediately after the procedure and prior to hospital discharge.

Secondary outcomes were sedation requirement and fluoroscopy duration. Patient symptoms particularly corresponding to pain and discomfort were routinely reviewed by the procedural team during the intraoperative period. The procedural dosage of IV morphine and IV midazolam was extracted from the procedural drug chart and checked with the catheter laboratory control drug record.

Antiarrhythmic pharmacological management subsequent to the procedure was at the discretion of the operator.

Categorical variables were expressed as frequency and percentage. Median and interquartile range (25th–75th percentiles) was used to describe non-parametric continuous data. Mean and standard deviation was used to describe parametric continuous data. The Shapiro-Wilk test was used to ascertain for normality.

Comparisons between unpaired groups of parametric data was conducted with the Student’s t-test while non-parametric data unpaired comparisons were conducted with the Mann-Whitney U test. Categorical variables were compared with the use of $\chi$${}^2$ test. The one-way analysis of variance (one-way ANOVA) with Bonferroni test for pairwise comparison was used for parametric comparison of three or more groups. The Kruskal-Wallis test with Dunn’s multiple’s comparison test was used for non-parametric comparison of three or more groups. A p 0.05 was considered statistically significant.

Statistical analysis was conducted with the use of GraphPad Prism 8 (GraphPad Software, Inc., San Diego, CA, USA).

Between August 2020 and January 2022, a total of 100 patients underwent procedures prospectively in the vHPSD group according to the study criteria. Thirty and forty-five patients were retrospectively included in the HPSD and STSF groups, respectively.

Baseline characteristics and demographics are summarised in Table 1. The median age (interquartile range, IQR) of the vHPSD, HPSD and STSF groups were 62.5 (56–69), 61 (53.5–68) and 59 (53–65) with the prevalence of the male gender at approximately 2/3 (71%, 70% and 73%). The prevalence of paroxysmal AF in the vHPSD, HPSD and STSF groups were 67%, 53%, and 58%, respectively. Additionally, the prevalence of persistent AF in the vHPSD, HPSD, and STSF groups were 33%, 47%, and 42%. The median CHA₂DS₂-VASc (IQR) of the vHPSD, HPSD and STSF groups were 1 (1–2), 2 (0–2.25) and 1 (0–2). The mean (confidence interval, CI) left atrial volume index (LAVI) across the vHPSD, HPSD and STSF groups were in the mild category at 29 $\pm$ 17.9 mL/m${}^2$, 30.6 $\pm$ 13.1 mL/m${}^2$ and 32.8 $\pm$ 13.6 mL/m${}^2$, respectively.

Of the patients in the vHPSD, HPSD and STSF group, substrate ablation was performed in addition to PVI in 40 (40%), 17 (56.7%) and 20 (44.4%) patients, respectively. The median number (IQR) of RF energy applications required for PVI alone in the vHPSD group, vHPSD and STSF group was 111 (96.75–129.3), 93 (80.25–112.5) and 85 (74.5–99.5) (p 0.0001) (Table 2) (Fig. 1). Of the vHPSD group, the median (IQR) number of vHPSD and conventional RF lesions were 101 (90.5–119) and 3 (0–14). The mean total RF ablation time for the vHPSD, HPSD and STSF groups were 9.87 $\pm$ 4.16 min, 33.9 $\pm$ 7.49 min and 36.0 $\pm$ 10.5 min (p 0.0001), respectively (Fig. 1). The mean total time required for PVI in the vHPSD, HPSD and STSF groups was 67.7 $\pm$ 29.7 min, 92.9 $\pm$ 25.7 min and 93.6 $\pm$ 29.1 min (p 0.0001). The mean fluoroscopy time corresponding to the vHPSD, HPSD and STSF group was 12.9 $\pm$ 8.00 min, 14.1 $\pm$ 12.8 min and 16.1 $\pm$ 10.1 min (p = 0.1981).

Fig. 1.

Procedural outcomes. (a) Number of lesions per patient. (b) Energy application time. (c) Total PVI duration. (d) Analgesia and sedation requirement. RF, radiofrequency ablation; vHPSD, very high power-short duration; WACA, wide area circumferential ablation; STSF, THERMOCOOL SMARTTOUCH™ SF; PVI, pulmonary vein isolation; L, left; R, right.

Within the vHPSD, HPSD and STSF groups, 14, 8 and 2 patients received general anaesthesia. The mean intra-procedural administered dose of intravenous morphine in the vHPSD, HPSD and STSF groups were 10.2 $\pm$ 3.43 mg, 16.1 $\pm$ 4.58 mg and 15.3 $\pm$ 3.94 mg (p 0.0001) in patients who had PVI under conscious sedation (Table 3) (Fig. 1). The mean intra-procedural administered dose of intravenous midazolam in the vHPSD, HPSD and STSF groups were 4.04 $\pm$ 3.24 mg, 8.63 $\pm$ 5.22 mg and 8.56 $\pm$ 4.72 mg (p 0.0001).

PVI was successfully achieved in all patients in the study. Fifty-six patients (56%) in the vHPSD group required supplementary conventional RF lesions to attain PVI. First-pass PVI of all PVs was attained in 55 (55%), 21 (70%) and 29 (64%) patients in the vHPSD, HPSD and STSF group, respectively (p = 0.134). The vHPSD, HPSD and STSF groups demonstrated evidence of acute pulmonary vein reconnection in 45%, 30% and 35.6% of patients (p = 0.134). The most common site of residual conduction in case of non-first pass isolation was at the left pulmonary vein carina.

None of the patients in this study died. Three PAEs were observed in the vHPSD group. One patient experienced an intraprocedural haemorrhagic pericardial effusion with cardiac tamponade, which was immediately treated with pericardiocentesis and initiation of the major haemorrhage protocol, including transfusion of packed red cells and prothrombin complex concentrate. Haemodynamic stability was promptly attained and the patient was transferred to the coronary care unit (CCU) and made a good clinical recovery with no subsequent re-hospitalization noted. One patient experienced a transient episode of junctional rhythm with sinus bradycardia. An exercise tolerance test conducted 48 hours later did not reveal any persistent atrioventricular (AV) block requiring pacemaker insertion. The third event was a 2:1 atrioventricular block that occurred following the ablation of a slow pathway subsequent to PVI for treatment of incessant atrioventricular nodal reentrant tachycardia (AVRNT) and paroxysmal atrial fibrillation (PAF). The PVI component of the procedure was performed with vHPSD while the slow pathway ablation was conducted by conventional ablation setting. The patient developed 2:1 AV block following slow pathway ablation for AVNRT. A temporary pacing wire was inserted followed by a dual chamber permanent pacemaker a few days later. There was no re-admission for procedure-related adverse events.

One PAE was noted in the HPSD group. A cardiac tamponade occurred which required immediate pericardial drain insertion. The patient made a full recovery following a period of monitoring in a high-dependency unit setting.

Four PAEs were observed in the STSF group; one cardiac tamponade, one pericardial effusion not requiring drainage, one access site haematoma and one suspected left anterior circulation stroke. The cardiac tamponade occurred following a transoesophageal echocardiogram and PVI under general anaesthesia, a post-procedural transthoracic echocardiogram (TTE) identified a severe pericardial effusion that required a pericardial drain, following which the patient had an uneventful recovery. The suspected cerebrovascular accident (CVA) occurred following PVI as the patient experienced left-sided weakness and was subsequently transferred to the stroke unit. Following a period of observation, the patient made a full recovery. The access site haematoma resolved with manual compression. One small pericardial effusion was noted on routine post-procedure TTE. A period of monitoring followed by a repeat TTE the next day revealed no evidence of progression of the pericardial effusion or haemodynamic compromise and hence did not require treatment.

The authors acknowledge several limitations in this service evaluation study. Firstly, due to the novelty and recent introduction of this technology, the number of available cases to contribute to the sample size was limited. Patients were prospectively recruited consecutively. The two-year inclusion period was during the COVID-19 pandemic where there was significant disruption to facilitation of PVI procedures. Additionally, other PVI procedures were performed by other ablation technologies such as Cryoablation and some RF ablation procedures were performed with other mapping system. The decision in determining between available ablation options were typically operator dependent. With respect to the control population, the authors felt the sample size was sufficient for comparison. Secondly, due to the service evaluation essence of this study, a formal randomisation process for patient selection and allocation was not employed. Furthermore, with regards to the ablation procedure, multiple different operators contributed to the sample size and thus introduced an inevitable degree of heterogeneity with regard to the technicalities of the intervention. Three operators with significant experience in catheter ablation ranging from 5 to 20 years, participated in the study. Most importantly, as an early service evaluation into real-world feasibility, this study does not provide data for longitudinal outcomes.

Thus, it would be prudent to eventually pursue a study design with a well-defined protocol and appropriate sample size determined from statistical power calculations. More importantly, a longitudinal follow-up period of at least 12 months would be beneficial to ascertaining longer term outcomes corresponding to overall effectiveness in reducing sinus remission.