To assess 1-year efficacy of CA of AF in relation to BR function modification and type of ablation energy used.

In the period between 2016 and 2018, we performed an observational, prospective, single-center study (ClinicalTrials.gov Identifier: NCT03811639) that enrolled patients admitted for CA of paroxysmal AF—named ABLANSAF study (Changes in Cardiac Autonomic Nervous System Following Atrial Fibrillation Ablation). The study was approved by the Local Ethics Committee (No 65/PB/2015) and all participants provided written informed consent.

A total of 78 patients were included in the ABLANSAF study during this period and 69 patients completed a 1-year follow-up. The inclusion criteria were as follows: (1) paroxysmal AF, (2) first CA, (3) sinus rhythm at the time of BR measurements, (4) no sick sinus syndrome, (5) no implanted pacemaker or defibrillator, (6) no PV anatomy favoring any specific ablation technique (for example, the common trunk of the left PV favoring RF), (7) no additional cavo-tricuspid CA or another atrial linear application, and (8) no evidence of severe heart failure defined as left ventricular ejection fraction <35%.

Patients were admitted to the hospital one day before the procedure. We assessed the BR function twice—before CA at 7:00 am on the day of the procedure and 2 days after CA. The BR parameters were measured during a shortened protocol of 25-min tilt test at rest, while supine, and after tilting (70°, no nitroglycerine challenge). The Task Force Monitor (TFM) software (CNSystem, 2007, version 2.2, Austria) was used for offline analysis (10). The producer of the Task Force Monitor CNS System declared under sole responsibility that the products were in compliance with directive 93/42/EEC. The TFM passed the tests for the CE quality mark (CE 0408, TUeV Austria, Vienna). TFM, certified in Europe, provides correct and reliable hemodynamic data (11). By identifying the sequences of at least three consecutive heartbeats with either a progressive increase in systolic blood pressure (SBP) and consequent prolongation of the R-R interval or a progressive decrease in SBP and resultant shortening of the R-R interval, the BR function was measured. The SBP and R-R interval changes must have been at least 1 mmHg over 4 ms. To determine BR sensitivity (BRS, in ms/mmHg), we set the interval between R-R and SBP values at 0 beats and computed the slope of the regression line between the R-R intervals and the SBP values for each sequence. An original example of ECG tracings with R-R intervals and SBP values (plethysmographic method) is shown in Figure 1.

Three BR parameters were calculated non-invasively during supine and tilt positions upon tilt testing. The following three parameters were assessed: 1.

BR event count (BREC) depicting overall BR activity (the number of events per minute),

Medications, such as beta blockers or antiarrhythmic drugs, remained unchanged before or directly after CA.

Procedural details were previously described (9). The RF CA procedure was performed using a single transseptal puncture and a Thermocool SmartTouch catheter (Biosense Webster, USA). The CARTO 3 system, ablation index (550 AI on the anterior wall was 550, and 380 on the posterior wall) for the module, and the CLOSE protocol were used (12). Intracardiac echocardiography (ICE) was used during the performance of all procedures. PVI was confirmed using a diagnostic circular catheter or by pacing from inside the ablation line. CB was performed using a flexible 15F sheath (Flex-Cath Advance, Medtronic, USA) and second-generation CB (Arctic front 2AF281 balloon, Medtronic, USA) with a diagnostic Achieve electrode (Medtronic, USA) to confirm PVI.

Patients were followed up for 1 year after CA. They underwent serial 4- to 7-day Holter ECG recordings (DMS 300-4A, DM Software, Stateline, NV, United States) at 3, 6, and 12 months after CA. Any antiarrhythmic medication was left unchanged up to 3 months after ablation and was stopped if no AF recurrence was observed. Effective ablation was defined as no AF, or atrial tachycardia (AT) episodes lasting longer than 30 s were recorded during Holter ECG or using standard ECG. The QoL was assessed using a dedicated scale (the University of Toronto Atrial Fibrillation Severity Scale—AFSS). The AFSS is a 19-item self-administered questionnaire designed to capture AF-related burden and its symptoms (frequency, duration, and severity of episodes) and healthcare utilization (13). The score of the scale is a summary that averages the frequency, duration, and patient-detected severity of AF episodes (14). The higher the score the greater the AF burden (13). We analyzed global wellbeing, the question about AF severity: higher score = more severe AF and the global number of points that characterize AF burden.

Statistical analyses were conducted using SAS statistical package ver. 9.4 (SAS Institute Inc., Cary, NC, USA). Continuous variables are expressed as mean ± SD (normal distribution) or median and the 25th–75th percentile range for skewed distribution. Normality was verified using the Shapiro–Wilk test. Between-group comparisons were made using Student's t-test or the non-parametric Mann–Whitney test, as appropriate. Within-subject comparisons at two time points were performed using a paired Student's t-test or the non-parametric signed-rank test. Due to the skewness of the distributions, the correction in relation to the baseline measurements was carried out using the robust regression method. The strength of linear relationships between numerical variables was measured by using Spearman's correlation coefficient. The statistical significance of the coefficients was tested. Categorical data are reported as counts and percentages. The differences in proportion between groups were analyzed using Fisher's exact test or Pearson’s χ² test. All statistical tests were treated as two-sided and evaluated at a significance level of 0.05.

As described previously (9), the baseline BR parameters, both while supine and after tilting, were similar in both groups. CA reduced all of the evaluated BR parameters in the whole group while supine and after tilting (9). The decrease in BRS parameters was more marked in the CB group than in the RF group when assessing BREC (9).

The 1-year follow-up was completed by 35 (89.7%) patients from the CB group and by 34 (87.2%) subjects from the RF group. After CA, all of the evaluated BR parameters decreased in both groups (9). A significant negative correlation was found between the difference in baroreceptor parameters (before and after ablation) and baseline values of baroreceptor parameters in both groups in the supine phase and after tilting. Regression coefficients showed significantly greater dependence of the decrease in BRS event count from the baseline values in CB vs. RF ablation. This relationship is statistically significant both for supine and tilt positions in the CB and RF groups [regression coefficient supine: −0.88 ±  0.11 vs. −0.45 ± 0.09; p = 0.003; tilt position: −0.99 ± 0.11 and −0.67 ± 0.07; p = 0.015, respectively). On repeated Holter ECG monitoring, the rate of efficacy of CB and RF was similar at 12 months (31/35 (88.6%) vs. 26/34 (76.5%); p = 0.185, respectively]. After CA, the BR function decreased in the whole group, but the values of the BR parameters did not differ between responders and non-responders, both while supine and after tilting (Table 2). The reduction in the BR values was not related to the efficacy depicted in Holter monitoring. The CA-induced changes (delta) in all BR parameters were similar in responders vs. non-responders, in both supine and tilt positions (Table 3). Although no significant differences were observed between responders and non-responders, in the RF ablation group, the BREC and BEI values in the supine position were approximately twofold lower in the supine position before and after ablation in responders compared with non-responders, which resulted in differences in BREC (after tilting, p = 0.047) and BEI (supine, p = 0.037) in responders (Table 4). There were no significant differences in BR parameters in non-responders and responders after ablation compared with baseline measurements in the supine and tilt positions in both groups (Table 5). However, data related to the above are underpowered because of the relatively small number of patients in both groups.

During tilt provocation after ablation, the values of the BR parameters were significantly reduced, but they did not differ between non-responders and responders.

According to the AFSS, the global score of perceived AF burden was significantly decreased in both groups after 3, 6, and 12 months (Table 6). Patients in the CB group were characterized by a significantly lower score after 3 months in comparison with those in the RF group. Similarly, the global score was lower in the CB group than in the RF group after 6 and 12 months (Table 6). However, after adjustment for baseline values, these differences became non-significant (Table 7).

The strengths of the study are as follows: (1) A well-defined and homogeneous population of patients with paroxysmal AF. (2) Use of BRs to evaluate CA-induced changes in BR function. (3) Use of Holter monitoring as an objective way to assess AF recurrence. (4) Confirmation of BR attenuation following CA of AF. (5) Demonstrating that CB leads to a greater decrease in BR function than RF does, but the long-term effectiveness of both procedures is similar.

First, the number of patients included in the study was relatively small. The number of patients was too low to compare the predictive value of BR reflex changes on the outcome. Second, this study was an unblinded and non-randomized one; however, demographic and clinical characteristics were similar in both patient groups. Third, due to a lack of data, the effectiveness of CA was assessed based on only Holter monitoring and not compared with BRS values because of the fact that baroreflex function was assessed only shortly after ablation and not at a later time point. The fact that baroreflex function was assessed only shortly after ablation and not at a later time point did not allow an evaluation of longitudinal changes. Finally, we did not use other tools to assess BR parameters, such as extracardiac vagal stimulation (ECVS), but at the time of conducting the study, we had no access to the ECVS device.