Original article
P-wave indices as predictors of atrial fibrillation recurrence after pulmonary vein isolation in normal left atrial size Giacomo Mugnaia,b, Gian-Battista Chierchiaa, Carlo de Asmundisa, Justo Julia´a, Giulio Contea, Juan Sieira-Moreta, Lucio Capulzinia, Kristel Wautersa, Moises Rodriguez-Man˜eroa, Giuseppe Cicontea, Giannis Baltogiannisa, Giacomo Di Giovannia, Yukio Saitoha and Pedro Brugadaa Aims Prolonged P-wave duration and dispersion are universally accepted noninvasive markers for atrial electrical remodeling. Our aim was to analyze P-wave indices as predictors of atrial fibrillation recurrence after pulmonary vein isolation in patients with normal left atrial size. Methods From January 2008 to December 2011, 426 patients with drug-resistant symptomatic paroxysmal atrial fibrillation underwent pulmonary vein isolation as an index procedure by conventional radiofrequency or cryoballoon ablation in our center. Patients with left atrial dilatation, poor-quality electrocardiograms, atrial pacemaker stimulation, and those undergoing repeat procedures were excluded. A total of 201 patients were analyzed during a mean follow-up of 22 W 16 months. Results Patients with prolonged P-wave duration had higher rates of atrial fibrillation recurrences compared with those without prolonged P-wave duration (49 vs. 14%; P < 0.001). Atrial fibrillation recurrence was significantly associated with prolonged P-wave duration (129 W 13 vs. 119 W 11 ms; P < 0.001) and P-wave dispersion (54 W 12 vs. 42 W 10 ms; P < 0.001) compared with those who remained in sinus rhythm. P-wave duration and dispersion were
Introduction Atrial fibrillation is the most common rhythm disturbance in the western society. Pulmonary vein isolation remains the cornerstone for the treatment of symptomatic drugrefractory paroxysmal atrial fibrillation.1–5 Atrial remodeling is a term to describe alterations in the properties or functions of atrial tissue and can be distinguished in electrical and structural remodeling. Electrical atrial remodeling is progressively perpetuated by rapid rates of atrial fibrillation and contributes itself to atrial fibrillation progression; it is characterized by a decrease in atrial effective refractory periods (ERPs), nonuniform conduction slowing, and increased dispersion in atrial refractoriness.6–8 Prolonged P-wave duration and P-wave dispersion assessed using a surface ECG are noninvasive markers for atrial electrical remodeling.6,9 In some patients undergoing pulmonary vein isolation, the atrial tissue may have already been remodeled and pulmonary 1558-2027 ß 2016 Italian Federation of Cardiology. All rights reserved.
independently associated with atrial fibrillation recurrence (hazard ratio 1.045, 95% confidence interval 1.027–1.063, P < 0.001; and hazard ratio 1.049, 95% confidence interval 1.022–1.078, P < 0.001, respectively), after adjusting for left atrial size and age. Conclusion Prolonged P-wave duration and dispersion were found to be independently associated with higher recurrence rates of atrial fibrillation after pulmonary vein isolation in patients with normal left atrial dimension. Therefore, a prolongation of P-wave indices may help to identify those patients in whom electrical remodeling has already occurred and a more extensive ablation may be indicated. J Cardiovasc Med 2016, 17:194–200 Keywords: atrial fibrillation, P-wave dispersion, P-wave duration, pulmonary vein isolation a
Heart Rhythm Management Centre, UZ Brussel-VUB, Brussels, Belgium and Department of Cardiology, University Hospital of Verona, Verona, Italy
b
Correspondence to Giacomo Mugnai, MD, Heart Rhythm Management Centre, Universitair Ziekenhuis Brussel, Vrije Universiteit Brussel, 101 Laarbeeklaan, 1090 Brussels, Belgium Tel: +32 24773038; fax: +32 24776851; e-mail: [email protected] Received 30 April 2014 Revised 26 August 2014 Accepted 27 August 2014
vein isolation alone might not guarantee the persistence of sinus rhythm on the long term. Previously published studies also showed that P-wave duration and dispersion were good predictors of atrial fibrillation relapse after RF ablation; however, in these studies, prolonged P-wave duration and dispersion were associated with left atrial dilatation.10,11 In the present study, we sought to evaluate P-wave indices as independent predictors for atrial fibrillation recurrence following pulmonary vein isolation in patients with normal left atrial size.
Methods From January 2008 to December 2011, all consecutive patients having undergone pulmonary vein isolation as an index procedure in our department for documented symptomatic paroxysmal atrial fibrillation were considered for our retrospective analysis. In all procedures, DOI:10.2459/JCM.0000000000000220
© 2016 Italian Federation of Cardiology. All rights reserved
P-wave indices in predicting atrial fibrillation Mugnai et al. 195
the acute end point was the achievement of electrical pulmonary vein isolation. Paroxysmal atrial fibrillation was defined as self-terminating atrial fibrillation episodes lasting for less than 7 days as per guidelines.12 Exclusion criteria were repeat procedures for recurrent atrial fibrillation after the index procedure, left atrial dilatation (indexed diameter 24 mm/m2), poor-quality ECGs (2 uninterpretable leads), atrial pacemaker stimulation, acute complications disrupting procedure continuation, and persistent atrial fibrillation. Patients having no sinus rhythm on ECG surface within 6 months prior to the procedure were also excluded. Prior to the procedure, all patients underwent two-dimensional (2D) transthoracic echocardiography to assess left atrial size and function, and left ventricular ejection fraction, and to rule out any structural and/or valvular disease. The left atrial diameter was assessed by 2D transthoracic echocardiography as the left atrial anteroposterior diameter measured at parasternal long-axis M-mode recordings and then indexed to the body surface area. A cardiac computed tomography and a transesophageal echocardiography were performed the day before ablation to analyze left atrial and pulmonary vein anatomy, and to rule out the presence of an intracardiac thrombus. A simultaneous recording of standard 12-lead ECG was performed in all patients during sinus rhythm, the day before the ablation. The 12-lead ECGs, recorded at 25.0 mm/s and calibrated to 1 mV/cm, were directly uploaded on the hospital ECG database at 300 dots per inch. These images were amplified 10, and then intervals were measured using software with semi-automatic callipers (GNU image manipulation program 2.8.2, Berkeley, California, USA). All the ECGs were analyzed by a blinded assessor, with borderline results being reassessed by a blinded electrophysiologist. Patients presenting ECGs with at least two uninterpretable leads were excluded from the analysis. The onset of the P-wave was defined as the point of initial upward or downward deflection from the baseline and the P-wave offset as its returning point to the isoelectric baseline. Maximum P-wave was defined as the longest P-wave duration in any lead and minimum P-wave as the shortest P-wave duration in any lead. Prolonged P-wave duration was defined as maximal P-wave duration of at least 120 ms.13 P-wave dispersion was calculated as the difference between maximum and minimum P-wave duration. In patients with prolonged P-wave duration, the morphology and characteristics of the P-wave were classified in type 1 or 3 according to the interatrial block consensus criteria.13 Evaluation of the P-wave terminal force (PWTF) was also performed by multiplying the P-wave duration in seconds with the P-wave amplitude of the negative terminal portion of the P-wave in V1 in millimeters. Increased PWTF was defined for values of 0.04 mm/s or less (larger than one small square). Finally, P-wave axis was also assessed.
Ablation was performed with either RF or cryoballoon technique. Both techniques were carried out in our center as previously described.14 A 2D transthoracic echocardiogram was repeated the day after the procedure to rule out the presence of pericardial effusion. Therapy with low-molecular-weight heparin was started the same day following ablation. Oral anticoagulation was started the day following the procedure. Patients were discharged on both oral anticoagulation and low-molecular-weight heparin. When a target international normalized ratio of 2–3 was reached, low-molecular-weight heparin was stopped and oral anticoagulation was continued for at least 3 months. Antiarrhythmic therapy was continued for 2 months following the procedure and stopped if the patient was free of atrial fibrillation relapse. Clinical follow-up consisted of physical examinations, ECG, and 24-h Holter recording performed at 1, 3, 6, and 12 months after ablation and every 6 months after the first year. A blanking period of 3 months was considered for the study. All documented episodes of atrial tachyarrhythmias lasting for at least 30 s were considered as a recurrence. Continuous variables were expressed as mean SD and, if appropriate, compared using the Student’s t test. Categorical variables were expressed as numbers and percentages and, if appropriate, compared with the chisquare analysis. The atrial fibrillation-free rates according to P-wave duration less than 120 and at least 120 ms, and P-wave dispersion less than 45 and at least 45 ms were analyzed using Kaplan–Meier analysis with the log-rank test. Pearson’s correlation coefficient, and univariate and multivariate Cox regression using significant variables were also performed. A P value less than 0.05 was deemed statistically significant. A receiver-operating characteristic (ROC) curve was constructed to assess the sensitivity and specificity of various cut-points of P-wave duration and P-wave dispersion in the prediction of atrial fibrillation recurrences. Statistical analysis was performed using SPSS 20.0.0 (IBM Inc., Armonk, New York, USA).
Results A total of 426 consecutive patients having undergone pulmonary vein isolation for paroxysmal atrial fibrillation as an index procedure were studied. Among them, 195 were excluded because of left atrial dilatation (indexed diameter 24 mm/m2), 10 were excluded for impossibility to retrieve information about their follow-up, 13 were excluded because of poor quality of their ECGs, 5 were excluded because of atrial pacing, and 2 because of the occurrence of an acute complication impeding procedure termination. Thus, in all, 201 patients (145 men, 56 11 years) were finally considered in our analysis; of these, 133 patients (66%) underwent pulmonary vein isolation using RF ablation procedure and 68 (34%) using
© 2016 Italian Federation of Cardiology. All rights reserved
196 Journal of Cardiovascular Medicine 2016, Vol 17 No 3
cryoballon ablation procedure. All patients were affected by highly symptomatic, drug-resistant paroxysmal atrial fibrillation. Baseline characteristics are summarized in Table 1. Mean indexed left atrial diameter was 20.4 2.2 mm/m2 (median 20.3, range 15–23.9 mm/m2). Prolonged P-wave duration was most frequently measured in lead II (115/201, 57%) followed by V5 (38/ 201, 19%), and lead I (12/201, 6%). Patients with prolonged P-wave duration had greater P-wave dispersion (51.9 11.9 vs. 39.8 9.1; P < 0.001) and minimum P-wave duration (79.7 10.8 vs. 73.0 9.4; P ¼ 0.007). No differences in age, sex, arterial hypertension, left atrial dimension, left ventricular ejection fraction, and P-wave axis were observed between patients with prolonged and normal P-wave duration (Table 2). P-wave duration and P-wave dispersion presented a good positive correlation (r ¼ 0.61, P < 0.001). P-wave duration was also found moderately correlated with age (r ¼ 0.23, P ¼ 0.01) and weakly correlated with P-wave axis (r ¼ 0.15, P ¼ 0.05). P-wave dispersion was not significantly correlated with age (r ¼ 0.11, P ¼ 0.12) and P-wave axis (r ¼ 0.13, P ¼ 0.07). Mean procedural and fluoroscopy times were, respectively, 147 66 and 34 21 min. Procedural time was significantly shorter in the cryoballoon group compared with the radiofrequency ablation group (112 58 vs. 192 49 min; P < 0.000001); fluoroscopy time was similar in both the groups (31 17 min for cryoballoon ablation vs. 36 14 min for radiofrequency ablation; P ¼ 0.45). In all patients, pulmonary veins were completely isolated at the end of the procedure. During a mean follow-up of 22 16 months (median 18, range 4–68 months), atrial fibrillation recurrences occurred in 63 patients (31.3%); of these, 52 patients (82%) had atrial fibrillation recurrence within 12 months after the ablation and 39 of them (75%) underwent a repeat procedure whilst 13 (25%) did not have a re-do procedure. First recurrences after the first year occurred in 11 patients (18%), of whom 5 (45%) Table 1
underwent a repeat ablation and the remaining 6 (55%) did not experience a re-do ablation. As shown in Table 2, patients with prolonged P-wave duration had greater recurrence rates than those without prolonged P-wave duration (49 vs. 14%; P < 0.001). Similarly, patients with atrial fibrillation recurrence had greater mean P-wave duration (129.5 13.1 vs. 119.2 10.7 ms; P < 0.001) and P-wave dispersion (53.9 12.5 vs. 42.2 10.2; P < 0.001) compared with patients without atrial fibrillation recurrence by univariate analysis (Table 1). After multivariate Cox regression, P-wave duration and P-wave dispersion were found significant predictors of atrial fibrillation recurrence, independently of left atrial dimension and age [hazard ratio 1.045, 95% confidence interval (CI) 1.027–1.063, P < 0.001; and hazard ratio 1.049, 95% CI 1.022–1.078, P < 0.001, respectively]. No other variables were independently associated with atrial fibrillation recurrence. ROC curve analysis was performed for different P-wave duration and P-wave dispersion cut-off points (Fig. 1). The area under the ROC curve for P-wave duration was 0.740 (95% CI 0.668–0.810, P < 0.001). The best combination of sensitivity and specificity was found for P-wave duration to be at least 120 ms (respectively, 77.8 and 62.6%), along with positive predictive value of 49% and negative predictive value of 86% (Fig. 1a). Using a cut-off of P-wave duration of at least 140 ms, the specificity and positive predictive values increased to 97.1 and 73%, respectively, but with a huge decrease in sensitivity, which collapsed to 17.5%. The area under the ROC curve for P-wave dispersion was 0.759 (95% CI 0.685–0.852, P < 0.001). P-wave dispersion of 45 ms showed the best combined sensitivity and specificity (71.4 and 68.8%, respectively), along with positive predictive value of 51% and negative predictive value of 84% (Fig. 1b). Increasing the P-wave dispersion cut-off to at least 55 ms, the specificity and the predictive positive value were enhanced to 87.1 and 63.2%, respectively, along with a decrease in sensitivity (49.2%) and in negative predictive value (78%). Times to atrial
Clinical characteristics of the total population and comparison between patients with and without atrial fibrillation recurrences
Clinical characteristics Number Mean age (years) Men Radiofrequency/cryoablation Arterial hypertension Mean indexed left atrial diameter (mm/cm2) BMI (kg/m2) Coronary artery disease Mean P-wave duration (ms) Mean P-wave dispersion (ms) Mean left ventricular ejection fraction (%) P-wave minimum (ms) P-wave axis (8) 1st degree interatrial block 3rd degree interatrial block P-wave terminal force (mm/s)
Total
Recurrence
Nonrecurrence
P value
201 56.2 11 145 (72.1%) 133/68 (66/34%) 62 (31%) 20.4 2.2 26.9 4.3 17 (8%) 122.4 12.4 45.8 12.2 59.1 5.1 76.3 10.6 49.6 24.2 100 (50%) 4 (2%) 0.035 0.02
63 55.1 10.2 47 (75%) 43/20 (68/32%) 18 (29%) 20.7 2 27.7 4.7 6 (10%) 129.5 13.1 53.9 12.5 59.2 3.5 75.9 11.4 52.1 20.9 49 (78%) 3 (5%) 0.04 0.02
138 56.7 10.9 98 (71%) 90/48 (65/35%) 44 (32%) 20.2 2.3 26.5 4.1 11 (8%) 119.2 10.7 42.2 10.2 59 5.7 76.5 10.3 48.5 25.5 51 (37%) 1 (1%) 0.03 0.03
0.31 0.58 0.41 0.74 0.13 0.07 0.94
P-wave indices as predictors of atrial fibrillation recurrence after pulmonary vein isolation in normal left atrial size Giacomo Mugnaia,b, Gian-Battista Chierchiaa, Carlo de Asmundisa, Justo Julia´a, Giulio Contea, Juan Sieira-Moreta, Lucio Capulzinia, Kristel Wautersa, Moises Rodriguez-Man˜eroa, Giuseppe Cicontea, Giannis Baltogiannisa, Giacomo Di Giovannia, Yukio Saitoha and Pedro Brugadaa Aims Prolonged P-wave duration and dispersion are universally accepted noninvasive markers for atrial electrical remodeling. Our aim was to analyze P-wave indices as predictors of atrial fibrillation recurrence after pulmonary vein isolation in patients with normal left atrial size. Methods From January 2008 to December 2011, 426 patients with drug-resistant symptomatic paroxysmal atrial fibrillation underwent pulmonary vein isolation as an index procedure by conventional radiofrequency or cryoballoon ablation in our center. Patients with left atrial dilatation, poor-quality electrocardiograms, atrial pacemaker stimulation, and those undergoing repeat procedures were excluded. A total of 201 patients were analyzed during a mean follow-up of 22 W 16 months. Results Patients with prolonged P-wave duration had higher rates of atrial fibrillation recurrences compared with those without prolonged P-wave duration (49 vs. 14%; P < 0.001). Atrial fibrillation recurrence was significantly associated with prolonged P-wave duration (129 W 13 vs. 119 W 11 ms; P < 0.001) and P-wave dispersion (54 W 12 vs. 42 W 10 ms; P < 0.001) compared with those who remained in sinus rhythm. P-wave duration and dispersion were
Introduction Atrial fibrillation is the most common rhythm disturbance in the western society. Pulmonary vein isolation remains the cornerstone for the treatment of symptomatic drugrefractory paroxysmal atrial fibrillation.1–5 Atrial remodeling is a term to describe alterations in the properties or functions of atrial tissue and can be distinguished in electrical and structural remodeling. Electrical atrial remodeling is progressively perpetuated by rapid rates of atrial fibrillation and contributes itself to atrial fibrillation progression; it is characterized by a decrease in atrial effective refractory periods (ERPs), nonuniform conduction slowing, and increased dispersion in atrial refractoriness.6–8 Prolonged P-wave duration and P-wave dispersion assessed using a surface ECG are noninvasive markers for atrial electrical remodeling.6,9 In some patients undergoing pulmonary vein isolation, the atrial tissue may have already been remodeled and pulmonary 1558-2027 ß 2016 Italian Federation of Cardiology. All rights reserved.
independently associated with atrial fibrillation recurrence (hazard ratio 1.045, 95% confidence interval 1.027–1.063, P < 0.001; and hazard ratio 1.049, 95% confidence interval 1.022–1.078, P < 0.001, respectively), after adjusting for left atrial size and age. Conclusion Prolonged P-wave duration and dispersion were found to be independently associated with higher recurrence rates of atrial fibrillation after pulmonary vein isolation in patients with normal left atrial dimension. Therefore, a prolongation of P-wave indices may help to identify those patients in whom electrical remodeling has already occurred and a more extensive ablation may be indicated. J Cardiovasc Med 2016, 17:194–200 Keywords: atrial fibrillation, P-wave dispersion, P-wave duration, pulmonary vein isolation a
Heart Rhythm Management Centre, UZ Brussel-VUB, Brussels, Belgium and Department of Cardiology, University Hospital of Verona, Verona, Italy
b
Correspondence to Giacomo Mugnai, MD, Heart Rhythm Management Centre, Universitair Ziekenhuis Brussel, Vrije Universiteit Brussel, 101 Laarbeeklaan, 1090 Brussels, Belgium Tel: +32 24773038; fax: +32 24776851; e-mail: [email protected] Received 30 April 2014 Revised 26 August 2014 Accepted 27 August 2014
vein isolation alone might not guarantee the persistence of sinus rhythm on the long term. Previously published studies also showed that P-wave duration and dispersion were good predictors of atrial fibrillation relapse after RF ablation; however, in these studies, prolonged P-wave duration and dispersion were associated with left atrial dilatation.10,11 In the present study, we sought to evaluate P-wave indices as independent predictors for atrial fibrillation recurrence following pulmonary vein isolation in patients with normal left atrial size.
Methods From January 2008 to December 2011, all consecutive patients having undergone pulmonary vein isolation as an index procedure in our department for documented symptomatic paroxysmal atrial fibrillation were considered for our retrospective analysis. In all procedures, DOI:10.2459/JCM.0000000000000220
© 2016 Italian Federation of Cardiology. All rights reserved
P-wave indices in predicting atrial fibrillation Mugnai et al. 195
the acute end point was the achievement of electrical pulmonary vein isolation. Paroxysmal atrial fibrillation was defined as self-terminating atrial fibrillation episodes lasting for less than 7 days as per guidelines.12 Exclusion criteria were repeat procedures for recurrent atrial fibrillation after the index procedure, left atrial dilatation (indexed diameter 24 mm/m2), poor-quality ECGs (2 uninterpretable leads), atrial pacemaker stimulation, acute complications disrupting procedure continuation, and persistent atrial fibrillation. Patients having no sinus rhythm on ECG surface within 6 months prior to the procedure were also excluded. Prior to the procedure, all patients underwent two-dimensional (2D) transthoracic echocardiography to assess left atrial size and function, and left ventricular ejection fraction, and to rule out any structural and/or valvular disease. The left atrial diameter was assessed by 2D transthoracic echocardiography as the left atrial anteroposterior diameter measured at parasternal long-axis M-mode recordings and then indexed to the body surface area. A cardiac computed tomography and a transesophageal echocardiography were performed the day before ablation to analyze left atrial and pulmonary vein anatomy, and to rule out the presence of an intracardiac thrombus. A simultaneous recording of standard 12-lead ECG was performed in all patients during sinus rhythm, the day before the ablation. The 12-lead ECGs, recorded at 25.0 mm/s and calibrated to 1 mV/cm, were directly uploaded on the hospital ECG database at 300 dots per inch. These images were amplified 10, and then intervals were measured using software with semi-automatic callipers (GNU image manipulation program 2.8.2, Berkeley, California, USA). All the ECGs were analyzed by a blinded assessor, with borderline results being reassessed by a blinded electrophysiologist. Patients presenting ECGs with at least two uninterpretable leads were excluded from the analysis. The onset of the P-wave was defined as the point of initial upward or downward deflection from the baseline and the P-wave offset as its returning point to the isoelectric baseline. Maximum P-wave was defined as the longest P-wave duration in any lead and minimum P-wave as the shortest P-wave duration in any lead. Prolonged P-wave duration was defined as maximal P-wave duration of at least 120 ms.13 P-wave dispersion was calculated as the difference between maximum and minimum P-wave duration. In patients with prolonged P-wave duration, the morphology and characteristics of the P-wave were classified in type 1 or 3 according to the interatrial block consensus criteria.13 Evaluation of the P-wave terminal force (PWTF) was also performed by multiplying the P-wave duration in seconds with the P-wave amplitude of the negative terminal portion of the P-wave in V1 in millimeters. Increased PWTF was defined for values of 0.04 mm/s or less (larger than one small square). Finally, P-wave axis was also assessed.
Ablation was performed with either RF or cryoballoon technique. Both techniques were carried out in our center as previously described.14 A 2D transthoracic echocardiogram was repeated the day after the procedure to rule out the presence of pericardial effusion. Therapy with low-molecular-weight heparin was started the same day following ablation. Oral anticoagulation was started the day following the procedure. Patients were discharged on both oral anticoagulation and low-molecular-weight heparin. When a target international normalized ratio of 2–3 was reached, low-molecular-weight heparin was stopped and oral anticoagulation was continued for at least 3 months. Antiarrhythmic therapy was continued for 2 months following the procedure and stopped if the patient was free of atrial fibrillation relapse. Clinical follow-up consisted of physical examinations, ECG, and 24-h Holter recording performed at 1, 3, 6, and 12 months after ablation and every 6 months after the first year. A blanking period of 3 months was considered for the study. All documented episodes of atrial tachyarrhythmias lasting for at least 30 s were considered as a recurrence. Continuous variables were expressed as mean SD and, if appropriate, compared using the Student’s t test. Categorical variables were expressed as numbers and percentages and, if appropriate, compared with the chisquare analysis. The atrial fibrillation-free rates according to P-wave duration less than 120 and at least 120 ms, and P-wave dispersion less than 45 and at least 45 ms were analyzed using Kaplan–Meier analysis with the log-rank test. Pearson’s correlation coefficient, and univariate and multivariate Cox regression using significant variables were also performed. A P value less than 0.05 was deemed statistically significant. A receiver-operating characteristic (ROC) curve was constructed to assess the sensitivity and specificity of various cut-points of P-wave duration and P-wave dispersion in the prediction of atrial fibrillation recurrences. Statistical analysis was performed using SPSS 20.0.0 (IBM Inc., Armonk, New York, USA).
Results A total of 426 consecutive patients having undergone pulmonary vein isolation for paroxysmal atrial fibrillation as an index procedure were studied. Among them, 195 were excluded because of left atrial dilatation (indexed diameter 24 mm/m2), 10 were excluded for impossibility to retrieve information about their follow-up, 13 were excluded because of poor quality of their ECGs, 5 were excluded because of atrial pacing, and 2 because of the occurrence of an acute complication impeding procedure termination. Thus, in all, 201 patients (145 men, 56 11 years) were finally considered in our analysis; of these, 133 patients (66%) underwent pulmonary vein isolation using RF ablation procedure and 68 (34%) using
© 2016 Italian Federation of Cardiology. All rights reserved
196 Journal of Cardiovascular Medicine 2016, Vol 17 No 3
cryoballon ablation procedure. All patients were affected by highly symptomatic, drug-resistant paroxysmal atrial fibrillation. Baseline characteristics are summarized in Table 1. Mean indexed left atrial diameter was 20.4 2.2 mm/m2 (median 20.3, range 15–23.9 mm/m2). Prolonged P-wave duration was most frequently measured in lead II (115/201, 57%) followed by V5 (38/ 201, 19%), and lead I (12/201, 6%). Patients with prolonged P-wave duration had greater P-wave dispersion (51.9 11.9 vs. 39.8 9.1; P < 0.001) and minimum P-wave duration (79.7 10.8 vs. 73.0 9.4; P ¼ 0.007). No differences in age, sex, arterial hypertension, left atrial dimension, left ventricular ejection fraction, and P-wave axis were observed between patients with prolonged and normal P-wave duration (Table 2). P-wave duration and P-wave dispersion presented a good positive correlation (r ¼ 0.61, P < 0.001). P-wave duration was also found moderately correlated with age (r ¼ 0.23, P ¼ 0.01) and weakly correlated with P-wave axis (r ¼ 0.15, P ¼ 0.05). P-wave dispersion was not significantly correlated with age (r ¼ 0.11, P ¼ 0.12) and P-wave axis (r ¼ 0.13, P ¼ 0.07). Mean procedural and fluoroscopy times were, respectively, 147 66 and 34 21 min. Procedural time was significantly shorter in the cryoballoon group compared with the radiofrequency ablation group (112 58 vs. 192 49 min; P < 0.000001); fluoroscopy time was similar in both the groups (31 17 min for cryoballoon ablation vs. 36 14 min for radiofrequency ablation; P ¼ 0.45). In all patients, pulmonary veins were completely isolated at the end of the procedure. During a mean follow-up of 22 16 months (median 18, range 4–68 months), atrial fibrillation recurrences occurred in 63 patients (31.3%); of these, 52 patients (82%) had atrial fibrillation recurrence within 12 months after the ablation and 39 of them (75%) underwent a repeat procedure whilst 13 (25%) did not have a re-do procedure. First recurrences after the first year occurred in 11 patients (18%), of whom 5 (45%) Table 1
underwent a repeat ablation and the remaining 6 (55%) did not experience a re-do ablation. As shown in Table 2, patients with prolonged P-wave duration had greater recurrence rates than those without prolonged P-wave duration (49 vs. 14%; P < 0.001). Similarly, patients with atrial fibrillation recurrence had greater mean P-wave duration (129.5 13.1 vs. 119.2 10.7 ms; P < 0.001) and P-wave dispersion (53.9 12.5 vs. 42.2 10.2; P < 0.001) compared with patients without atrial fibrillation recurrence by univariate analysis (Table 1). After multivariate Cox regression, P-wave duration and P-wave dispersion were found significant predictors of atrial fibrillation recurrence, independently of left atrial dimension and age [hazard ratio 1.045, 95% confidence interval (CI) 1.027–1.063, P < 0.001; and hazard ratio 1.049, 95% CI 1.022–1.078, P < 0.001, respectively]. No other variables were independently associated with atrial fibrillation recurrence. ROC curve analysis was performed for different P-wave duration and P-wave dispersion cut-off points (Fig. 1). The area under the ROC curve for P-wave duration was 0.740 (95% CI 0.668–0.810, P < 0.001). The best combination of sensitivity and specificity was found for P-wave duration to be at least 120 ms (respectively, 77.8 and 62.6%), along with positive predictive value of 49% and negative predictive value of 86% (Fig. 1a). Using a cut-off of P-wave duration of at least 140 ms, the specificity and positive predictive values increased to 97.1 and 73%, respectively, but with a huge decrease in sensitivity, which collapsed to 17.5%. The area under the ROC curve for P-wave dispersion was 0.759 (95% CI 0.685–0.852, P < 0.001). P-wave dispersion of 45 ms showed the best combined sensitivity and specificity (71.4 and 68.8%, respectively), along with positive predictive value of 51% and negative predictive value of 84% (Fig. 1b). Increasing the P-wave dispersion cut-off to at least 55 ms, the specificity and the predictive positive value were enhanced to 87.1 and 63.2%, respectively, along with a decrease in sensitivity (49.2%) and in negative predictive value (78%). Times to atrial
Clinical characteristics of the total population and comparison between patients with and without atrial fibrillation recurrences
Clinical characteristics Number Mean age (years) Men Radiofrequency/cryoablation Arterial hypertension Mean indexed left atrial diameter (mm/cm2) BMI (kg/m2) Coronary artery disease Mean P-wave duration (ms) Mean P-wave dispersion (ms) Mean left ventricular ejection fraction (%) P-wave minimum (ms) P-wave axis (8) 1st degree interatrial block 3rd degree interatrial block P-wave terminal force (mm/s)
Total
Recurrence
Nonrecurrence
P value
201 56.2 11 145 (72.1%) 133/68 (66/34%) 62 (31%) 20.4 2.2 26.9 4.3 17 (8%) 122.4 12.4 45.8 12.2 59.1 5.1 76.3 10.6 49.6 24.2 100 (50%) 4 (2%) 0.035 0.02
63 55.1 10.2 47 (75%) 43/20 (68/32%) 18 (29%) 20.7 2 27.7 4.7 6 (10%) 129.5 13.1 53.9 12.5 59.2 3.5 75.9 11.4 52.1 20.9 49 (78%) 3 (5%) 0.04 0.02
138 56.7 10.9 98 (71%) 90/48 (65/35%) 44 (32%) 20.2 2.3 26.5 4.1 11 (8%) 119.2 10.7 42.2 10.2 59 5.7 76.5 10.3 48.5 25.5 51 (37%) 1 (1%) 0.03 0.03
0.31 0.58 0.41 0.74 0.13 0.07 0.94
