This prospective cohort study consecutively enrolled patients with paroxysmal or persistent AF referred to our center according to routine clinical practice to undergo PV ablation from June 2015 to February 2017. Patient follow-up was performed for 12 months. All patients signed written informed consent for the study, which had previously been approved by the local research ethics committee.

The exclusion criteria were as follows: a) patient refusal to sign the informed consent; b) an inability to perform clinical follow-up or biological sample collection; c) a personal history of severe renal impairment (eGFR < 30mL/min/1.73 m2); d) patients on hemodialysis; e) acute myocardial infarction or percutaneous coronary intervention in the 6 months before the ablation; and f) AF ablation in the 6 months before the performance of the study.

The analysis included BNP, corin, and galectin-3 measurements at baseline and 1 year of follow-up. Blood samples for analysis of corin and galectin-3 were centrifuged and frozen at −80°C after extraction until their analysis.

The ablation procedure was performed as described in the supplementary data and as illustrated in figure 1 of the supplementary data.

After the ablation, patients continued their antiarrhythmic therapy and oral anticoagulation for the first 3 months, considered the window period. At the 3-month follow-up, Holter monitoring was performed for 4 to 7 days, as well as blood and urine analysis. At 12 months, a 24-hour Holter electrocardiogram was performed, as well as urine and blood biomarker analysis. Any atrial arrhythmia (AF, flutter, or atrial tachycardia) lasting ≥ 30seconds beyond the window period was considered recurrence. Because no event recorder was available, patients who experienced palpitations were directed to go to their health care center or emergency department in an effort to document possible recurrences.

The eGFR was calculated according to the 2012 KDIGO guidelines,12 which recommend its estimation using the CKD-EPI formula. The eGFR was measured at baseline and 3 and 12 months after the ablation. Any increase during follow-up vs baseline was considered an eGFR improvement. In patients who received iodinated contrast during the procedure for rotational angiography, a 24-hour analysis was performed to rule out contrast nephropathy, defined as an absolute increase in serum creatinine ≥ 0.5mg/dL or a relative increase ≥ 25% within 24hours after contrast exposure.13

Numerical variables are expressed as mean ± standard deviation or, for nonnormally distributed variables, as median [interquartile range]. Categorical variables are expressed as absolute and relative frequencies. To compare the possible differences in categorical variables between the groups according to type of AF and eGFR group, the Pearson chi-square test was used or the Fisher exact test, as appropriate. For numerical variables, the t test and ANOVA were used for independent samples and the Mann-Whitney and Kruskal-Wallis for nonparametric variables. Variable normality was assessed with the Shapiro-Wilk test. The repeated measures general linear model was applied to compare the changes over time in the different follow-up parameters, considering time as an individual factor and the comparison group (eg, the different cardiovascular risk factors and treatments) as an interindividual factor. The McNemar test was used to compare qualitative parameters at different times. A mixed-effects linear regression model was performed to analyze the variables influencing the changes in the eGFR. It was adjusted for time, recurrence, and the remaining possible confounding variables, and the patients were considered a random effect in the model. First, the null model was adjusted to consider patient variability alone (model I); then, it was adjusted by the time variable (model II); finally, it was adjusted by the other covariables (model III). All comparisons were considered significant at P <.05. Data were analyzed with SPSS 19.0 software (SPSS Inc; Chicago, Illinois, United States).

Out of 174 patients who underwent PV ablation during the study inclusion period, 124 were enrolled. The reasons for patient exclusion were as follows: 1 due to severe renal impairment (2%; eGFR, 23mL/min); 5 due to ablation within the previous 6 months (10%); 6 due to their participation in another study that was incompatible with ours (12%); 26 due to patient refusal or impossibility of follow-up (52%); and 12 because biomarker sampling could not be performed (24%).

The mean patient age was 55 ± 10 (range, 22-75) years and 86 were men (69.4%). The PV ablation was performed using radiofrequency in 98 patients (79%) and using cryoablation in 26 (21%). In addition, 75 patients had paroxysmal AF (60.5%) and 49 had persistent AF (39.5%). The ablation procedure was the first in 96 patients (77.4%) but a previous ablation procedure had been performed in 28 (22.6%). The baseline eGFR was 90.8 [73.8-90.8] mL/min/1.73 m2 and the baseline creatinine was 0.88 [0.78-1.03] mg/dL. Patients were divided into 3 groups according to their baseline eGFR: group 1 (67 patients, 54%) with a normal eGFR (≥ 90mL/min/1.73 m2); group 2 (48 patients, 38.7%) with a slightly reduced eGFR (60-89mL/min/1.73 m2); and group 3 (9 patients, 7.3%) with a moderately reduced eGFR (30-59mL/min/1.73 m2). The 9 patients with an eGFR <60mL/min/1.73 m2 had no recorded history of renal impairment but all were hypertensive, 2 were diabetic, and 1 had systemic lupus erythematosus as possible causes of the renal impairment deterioration. No patients had an eGFR <30mL/min/1.73 m2 because it was an exclusion criterion. No contrast nephropathy occurred within 24hours after the ablation.

The clinical, echocardiographic, and blood parameter characteristics of the patients according to AF type and baseline eGFR are shown in table 1 and table 2, respectively. Patients with persistent AF were more frequently treated with diuretics, beta-blockers, and anticoagulants than those with paroxysmal AF; they also had a larger left atrial diameter. In contrast, patients with paroxysmal AF had a higher frequency of hypercholesterolemia and a greater use of antiarrhythmic drugs. There were no differences between the 2 groups in baseline eGFR, baseline creatinine, or biomarker levels. When patients were grouped by baseline eGFR, those with worse eGFR (groups 2 and 3) were older, were more likely to have hypertension and structural heart disease, and had a higher body mass index, greater use of angiotensin-converting enzyme inhibitors/angiotensin receptor blockers, diuretics, and anticoagulants, and a higher baseline BNP concentration.

Circumferential PV isolation was performed in all patients. Isolation of all targeted veins was achieved in 121 patients (97.6%). A second ablation procedure was performed in 9 patients (7%) during the first postprocedural year. This procedure was performed with radiofrequency in all patients; these cases were considered recurrences. At 12 months, 84 patients (67.7%) were recurrence free; of these, 57 had paroxysmal AF (76%) and 27 had persistent AF (55%) (P=.013). Of the 40 patients with recurrence, the recurrence was paroxysmal AF in 28 (22.6%), persistent AF in 8 (6.5%), and atypical flutter in 4 (3.2%). Recurrences were documented in 16 patients (12.9%) using electrocardiography performed during emergency department treatment, in 20 (16.1%) using electrocardiography or Holter monitoring performed during clinical follow-up, and in 4 (3.3%) using both methods. Of the 84 patients without 12-month recurrence, 11 (13%) were taking antiarrhythmic drugs at the end of follow-up vs 25 patients with recurrence (62.5%) (P <.0001). No relationship was found between the biomarkers and recurrence; on multivariate analysis, the only variable associated with recurrence was the presence of persistent AF (odds ratio=2.3; 95% confidence interval [95%CI], 0.9-5.7).

At the 1-year follow-up and in the overall patient group, the eGFR increased from 90.8 [77.8-100.0] to 95.6 [84.8-103.3] mL/min/1.73 m2 (P <.0001) and the creatinine level decreased from 0.88 [0.78-1.03] to 0.82 [0.76-0.92] mg/dL (P <.0001). There was no difference in the eGFR improvement according to recurrence during the window period (P=.095). However, there were differences among the baseline, 3-month, and 12-month eGFRs. These differences were due to time, not group, although there was an interaction between time and group (table 3). The eGFR increased until the end of follow-up by 2.9 [–2.7 to 11.8] mL/min/1.73 m2 on average, with significant differences between patients with 12-month recurrence and those without (–1.1 [–6.0 to 8.8] vs 7.1 [–0.6 to 14.2] mL/min/1.73 m2; P=.017). The 12-month improvement in eGFR was inversely proportional to the baseline eGFR. Thus, patients with an eGFR 30-59mL/min/1.73 m2 showed the largest increase in eGFR (group 1, –0.2 [–4.3 to 3.8] mL/min/1.73 m2; group 2, 9.1 [0.9 to 21.5] mL/min/1.73 m2; group 3, 32.9 [19.4 to 47.6] mL/min/1.73 m2; P=.001) (figure 1). Of patients without 12-month recurrence, 5% were in group 3 and 41.2% were in group 2 at baseline, whereas they numbered 0% and 30%, respectively, at 12 months (P <.0001) (figure 2). In contrast, patients with recurrence comprised the same percentage of all patients in each group at baseline and at 12 months (P=.006). The clinical, echocardiographic, and blood parameter characteristics were similar in patients with an eGFR improvement at the end of follow-up and in those without such an improvement, although the group with an eGFR improvement showed a tendency for a higher number of diabetic patients, better left ventricular ejection fraction, greater use of anticoagulants, and lower use of angiotensin-converting enzyme inhibitors/angiotensin receptor blockers. The only variable associated with an eGFR improvement was recurrence presence/absence (table 4).

Figure 1.

Changes in eGFR at 1 year of follow-up based on baseline function according to group. eGFR, estimated glomerular filtration rate.

(0.07MB).

Figure 2.

Changes in the eGFR distribution according to the presence of recurrence at 12 months. eGFR, estimated glomerular filtration rate.

(0.1MB).

Patients without 12-month recurrence had a baseline eGFR similar to that of patients with recurrence (90.8 [75.7-100.0] vs 90.9 [69.8-100.3] mL/min/1.73 m2; P=.708). There were also no significant differences in recurrence according to baseline eGFR group (32.8% in group 1 vs 29.2% in group 2 vs 44.4% in group 3; P=.660).

BNP levels decreased from 47.8 [17.5-85.7] to 32.6 [13.1-69.4] pg/mL at 1 year of follow-up (P=.013), although there were no significant differences between patients with and without recurrence at 1 year (P=.465). No relationships were found either between BNP and any of the other variables studied. Corin levels increased from 7.63 [4.72-10.64] to 10.60 [9.00-12.23] ng/mL at 12 months (P <.001), without differences according to recurrence (P=.461). Corin levels were higher at 12 months in the persistent AF group than in the paroxysmal AF group (P=.011); there were no differences in the other variables studied. Galectin-3 levels increased from 343.5 [266.4-411.7] pg/mL at baseline to 371.7 [310.8-470.5] pg/mL at 12 months (P=.025), but no associations were found with any of the variables studied.

The mixed-effects multivariate model showed a significant increase in the eGFR over time, with an average increase of 2.8 (95%CI, 1.7-4.0) mL/min/1.73 m2 during follow-up. Patients with recurrence had a nonsignificantly lower eGFR (–1.4; 95%CI, –5.4 to 2.5mL/min/1.73 m2). Male sex was directly associated with an eGFR increase of 7.1 (95%CI, 3.0-11.3) mL/min/1.73 m2; the rate improved by 6.3 (95%CI, 1.8-10.9) mL/min/1.73 m2 in patients not taking diuretics, whereas age was inversely associated with an eGFR deterioration of –0.71 (95%CI, –0.9 to –0.5) mL/min/1.73 m2 per year (table 5). None of the biomarkers studied was associated with the eGFR improvement.

Although most factors possibly associated with renal function were studied, there may be other confounding factors that could explain the changes in the eGFR. The detection of recurrences in our study could have been improved through the use of event recorders and it would also have been useful to quantify the AF burden. The eGFR increase in all of the patients with recurrence was small and its relevance in routine clinical practice needs to be determined.