Changes in P-wave area and P-wave duration after circumferential pulmonary vein isolation

Transcription

1 Europace (2010) 12, doi: /europace/eup410 CLINICAL RESEARCH Ablation for Atrial Fibrillation Changes in P-wave area and P-wave duration after circumferential pulmonary vein isolation Katarina Van Beeumen 1,2 *, Richard Houben 3, Rene Tavernier 4, Stefan Ketels 1, and Mattias Duytschaever 1,4 1 Department of Cardiology, University Hospital, Ghent, Belgium; 2 Department of Cardiology, St-Lucas Hospital, Ghent, Belgium; 3 Medtronic Bakken Research Center, Maastricht, The Netherlands; and 4 Department of Cardiology, St-Jan Hospital, Bruges, Belgium Received 12 July 2009; accepted after revision 24 November 2009; online publish-ahead-of-print 3 January 2010 Aims The effect of circumferential pulmonary vein isolation (CPVI) on P-wave characteristics is not clear. We used the signal-averaged (SA) electrocardiogram (ECG) and the ECG derived vector cardiogram (dvcg) to study the influence of CPVI on P-wave duration (PWD) and P-wave area (PWA) and studied whether changes were associated with successful outcome after initial CPVI.... Methods Thirty-nine patients ( years, 72% males) underwent CPVI for paroxysmal or persistent atrial fibrillation (AF). and results For each patient, an ECG recording was taken at the start and end of the ablation procedure. dvcg was derived using the inverse Dower transform. PWD was defined by manual annotation of earliest onset and latest offset of the SA-P-wave. PWA was calculated as the area under the SA-ECG curve averaged for the 12 ECG leads (PWA- ECG) and SA-dVCG curve (PWA-dVCG). Successful outcome after CPVI was defined as freedom from symptomatic and asymptomatic AF at the end of follow-up ( months). Average PWD decreased from to ms (P, 0.01). PWA-ECG and PWA-dVCG decreased markedly from to mvms (P, 0.001) and from to mvms (P, 0.001). Parameters of PWA were not different between successes (n ¼ 31) and failures (n ¼ 8). In contrast, PWD after ablation was significantly shorter in patients with successful outcome ( vs ms, P, 0.05).... Conclusion (i) CPVI results in a modest but significant shortening in PWD and a marked decrease in PWA. (ii) PWD was significantly shorter in cases of successful outcome after CPVI Keywords Atrial fibrillation Circumferential pulmonary vein isolation Signal-averaging electrocardiogram Derived vector cardiogram P-wave area P-wave duration Introduction Percutaneous circumferential pulmonary vein isolation (CPVI) for paroxysmal or short standing recurrent atrial fibrillation (AF) results in maintenance of sinus rhythm (SR) without antiarrhythmic drugs in 70 90% of patients. 1 4 Its beneficial effect has been related to trigger elimination, autonomic modulation, and/or elimination of drivers at the posterior wall and pulmonary vein junction. 5,6 The beneficial effect of CPVI might also be mediated by ablation-induced reduction in critical electrical mass, as both intra-atrial conduction delay and increased tissue area are associated with an increased propensity for the development of AF Pappone et al. 6 calculated that during circumferential PV ablation, 28% of the total left atrial area is ablated and demonstrated that a smaller area of LA ablation is a predictor of AF recurrence at follow-up. We hypothesized that this reduction in electrical atrial tissue mass by ablation would result in reduction in P-wave area (PWA) and/or P-wave duration (PWD). To analyse changes in PWA and PWD, the signal-averaged (SA) 12-lead surface electrocardiogram (ECG) and the ECG derived vector cardiogram (dvcg) were used. We also studied whether parameters of PWA and PWD were predictors for successful outcome after CPVI. * Corresponding author. Tel: þ ; fax: þ , katarina.vanbeeumen@azstlucas.be Published on behalf of the European Society of Cardiology. All rights reserved. & The Author For permissions please journals.permissions@oxfordjournals.org.

2 P-wave area and P-wave duration after CPVI 799 Methods Study population The patient population consisted of 39 consecutive patients (mean age years, 72% males) referred for percutaneous catheter ablation of highly symptomatic drug resistant paroxysmal and short standing persistent AF. Patient characteristics are described in Table 1. All patients gave informed consent to participate in the study protocol. Circumferential pulmonary vein isolation Treatment with b-blocking agents and all other anti-arrhythmic drugs except amiodarone was discontinued at least four to five half-life periods before the ablation procedure. Procedure was performed during SR and under general anaesthesia using continuous propofol infusion for sedation, cisatracurium for neuromuscular blockade, and remifentanil for analgesia. Patients that were in AF before the procedure (n ¼ 5, 13%) were first converted to SR with an external synchronized direct current shock. CARTO-and LASSO-guided CPVI was performed as previously described elsewhere. 11 The end-point for PV ablation was complete abolishment of local PV electrograms recorded with the circular mapping catheter. In case PVPs remained after anatomical circumferential lesions were created, ablation was applied at the circumference of the ablation circle (at bridging potentials), using the mapping catheter to identify the point of earliest breakthrough. After ablation, each PV was mapped for permanent PV isolation (PVI) during al least 30 min. At the end of the procedure a dedicated voltage and activation map of the left atrium (+100 points) was performed. Post-ablation follow-up After the procedure, all patients were monitored in a coronary care unit for 2 days with a continuous heparin infusion (target activating clotting time of 180 s) for the first 24 h. A transthoracic echocardiogram was performed in all patients to rule out pericardial effusion. At discharge, anti-arrhythmic drugs (sotalol 80 mg bid or flecainide 100 mg bid) were given for 1 month. Irrespective of risk for stroke, oral vitamin K antagonist therapy was started as well with a target INR of 2 3. Table 1 Baseline patient characteristics (n 5 39) Mean age (years) Males, n 28 (72%) AF type Paroxysmal, n 32 (82%) Short standing persistent, n 7 (18%) Mean duration of AF (years) Arterial hypertension, n 13 (33%) Structural heart disease, n 3 (8%) LA diameter, 2D echo (mm) LA volume, CARTO (ml) Previous AARD, n Amiodarone, n 12 (31%) Others, n 27 (69%) Vitamin K antagonist, n 16 (42%) AF, atrial fibrillation; LA, left atrial; AARD, anti-arrhythmic drugs; 2D, 2-dimensional. ECG recordings and data processing During the ablation procedure, continuous 12-lead ECG, and intracardiac signals were recorded, digitized (sample frequency 977 Hz) and stored for offline analysis (Cardiolab, Prucka Engineering, Houston, TX, USA). P-wave analysis was performed on ECG strips without atrial and/or ventricular ectopic beats (20 s), selected from the beginning and the end of the ablation procedure. In case of prior external shock, ECG strips used for P-wave analysis were taken at least 5 min after electrical cardioversion. From leads V1 V6, I and II, the dvcg was calculated using the inverse Dower transform. 12,13 Carlson et al. 12 demonstrated that this inverse dower transform performed on a subset of leads provides a derived VCG comparable with the VCG of orthogonal leads. From each selected ECG and dvcg strip the R waves were detected after correction of the baseline by band-pass filtering ( Hz) and using a slope base algorithm. 14 This technique greatly reduced noise enabling accurate identification of the start, peak and end of the P-wave. Detected R-waves were used as anchor points for extraction of ECG and dvcg beats (1 s). After removal of possible remaining premature atrial and ventricular premature beats in all ECG and dvcg strips by the software program, the sinus beats (22 + 4) were time aligned, de-trended and SA. P-wave analysis After signal averaging, delineation of the P-wave was obtained by creating a time window preceding each QRS complex containing the P-wave. In case of long PQ-interval and/or long PWD, the width and position of the time window were adapted to ensure full coverage of the P-wave. Two independent observers, blinded to the outcome of the patient, defined SA-PWD by manual annotation of the time difference between the earliest onset of the P-wave (in any lead) and the latest offset of the P-wave (in any lead) with a time resolution of 1 ms (Figure 1A). The Pearson correlation for reproducibility of SA-PWD data between both independent observers was r ¼ 0.67 (P, 0.05). The amplitude of the P-wave was measured with respect to a baseline that was constructed by linear interpolation (dashed line) between the amplitudes at inception and termination of the P-wave. PWA was calculated as the area under the SA ECG signal (PWA-ECG) and SA-dVCG curve (PWA-dVCG) from onset to termination of the P-wave. PWA analysis was performed on the first (I) and second (II) half of the P-wave and on the terminal 40 ms (RMS 40) and 20 ms (RMS 20) of the P-wave. Follow-up evaluation Each patient was scheduled for follow-up examination at 1 and 3 months and every 3 months thereafter. Mean follow-up duration was months (range 4 22 months). At each follow-up examination, 7-day Holter monitoring was performed to verify the absence of AF. After 1 month of follow-up, anti-arrhythmic drugs were discontinued. Anticoagulation was continued if the patient had a CHADS 2 score 2. A successful outcome was defined as freedom from symptomatic and asymptomatic AF without anti-arrhythmic drugs at follow-up with a blanking period of 1 month. Statistical analysis Statistical analysis was performed using SPSS 12.0 (SPSS Inc, Chicago, IL, USA). All data are presented as mean + standard deviation (SD) for continuous variables. Results within one group with normal distribution were compared using paired t-test. Comparison between groups with normal distribution was performed using independent sample t-test. Statistical significance was established at P, 0.05.

3 800 K. Van Beeumen et al. example is given in Figure 2. In this patient, baseline heart rate was 65 bpm while heart rate after ablation increased to 85 bpm. Clearly, except for lead AVL, there was a marked decrease in P-wave amplitude from 0.15 to 0.2 mv before ablation to 0.1 mv after ablation. After a mean follow-up of months (range 4 22 months), 80% (n ¼ 31) of the patients remained free of symptomatic and asymptomatic AF after the first procedure. Mean number of RF applications and mean LA CARTO volume were not statistically different between patients with and without AF recurrence at follow-up ( vs , P ¼ NS and vs ml, P ¼ NS). Figure 1 (A) P-wave analysis. Measurement of P-wave duration and P-wave area. The total duration of the P-wave was 124 ms, and the P-wave area 3.13 mvms. Within the first 62 ms (I), the area under the P-wave was 1.51 mvms whereas for the terminal part (II) the area was 1.62 mvms. (B) Representative example of PWD and PWA (dvcg) before and after CPVI in a 50-year-old patient with paroxysmal atrial fibrillation and successful outcome after CPVI. Before ablation, the duration of the P-wave was 120 ms with a total area (PWA) of 6.22 mvms. Within the first (I) and terminal (II) part of the P-wave, the area under the curve was 3.60 and 2.62 mvms. After PV ablation, the P-wave slightly shortened to 112 ms (DPWD ¼ 8 ms) and PWA was reduced by 59% to 2.54 mvms. Within the trailing and terminal part of the P-wave, the area reduced to 1.82 and 0.72 mvms. Results Circumferential pulmonary vein isolation and follow-up In all patients, complete PVI was obtained with a mean RF time of min and a fluoroscopy time of min. Mean procedure time was h. Mean heart rate before and immediately after ablation was and bpm, respectively (P, 0.001). In all patients, a macroscopic decrease in P-wave amplitude was seen on the surface ECG after ablation. A representative Ablation-induced changes in P-wave duration Overall, we observed a significant shortening of the PWD after CPVI. In Figure 1B, a representative example is given of ablation-induced shortening in PWD (dvcg is shown). This patient (with successful outcome after CPVI) demonstrates marked shortening in PWD from 120 ms before to 112 ms (DPWD 8 ms) after ablation. Mean data are shown in Figure 3A. The average PWD decreased from ms before CPVI (PWD pre) to ms after CPVI (PWD post), resulting in a DPWD of ms (P ¼ 0.005). When comparing PWD data in both patient groups (Table 2), we observed that patients with a successful outcome after CPVI had a significantly shorter PWD post compared with patients with AF recurrence ( vs ms, P ¼ 0.04). Furthermore, as shown in Figure 3A, all patients with successful outcome demonstrated a significant shortening in PWD after ablation ( to ms, DPWD ms, P ¼ 0.003), whereas patients with AF recurrence did not show a significant change in PWD ( to ms, DPWD ms, P ¼ NS). The difference in the change in PWD between successes and failures was not significant, although borderline (P ¼ 0.07). Data on PWD were comparable for patients with amiodarone vs. the total group (P ¼ NS). Ablation-induced changes in P-wave area PWA-ECG and PWA-dVCG decreased significantly after CPVI in all patients and subgroups. A representative example of ablation-induced decrease in PWA-dVCG is shown in Figure 1B. PWA-dVCG decreased from 6.22 mvms before to 2.54 mvms after ablation, resulting in a DPWA-dVCG of 59%. Compared with the first part of the PWA, the second part of the PWA decreased more markedly (72 vs. 49%). Mean results for all patients are shown in Figure 3, lower panel. Mean PWA-ECG decreased significantly after CPVI (from to mvms P, 0.001). Also PWA-dVCG decreased significantly (from to mvms, P, 0.001). Changes were more pronounced in PWA-dVCG than in the PWA-ECG (DPWA-dVCG 240%, DPWA-ECG 221%, P, 0.001). Separate analysis of the first and second parts of the P-wave revealed a more pronounced decrease in the second part of PWA after ablation, both for the 12 ECG leads (DPWA-ECG part I 211%, DPWA-ECG part II 228%,

4 P-wave area and P-wave duration after CPVI 801 Figure 2 Representative 12-lead ECG before and after CPVI showing macroscopic reduction in P-wave amplitude. P, 0.001) as for the VCG (DPWA-dVCG part I 232%, DPWA-dVCG part II 243%, P ¼ 0.01). Changes in RMS40 and RMS20 were comparable. RMS40 PWA-dVCG decreased from to mvms (253%, P, 0.01), RMS20 PWA-dVCG from to mvms (233%, P ¼ 0.01). RMS40 PWA-ECG decreased from to mvms (238%, P, 0.01), RMS20 PWA-ECG from to mvms (266%, P ¼ 0.01). Data comparing PWA before and after ablation in patients with and without successful outcome showed no significant difference (Table 2). Data on PWA-ECG and PWA-dVCG were comparable for patients with amiodarone vs. the total group (P ¼ NS). Intra-cardiac mapping of the left atrium before and after ablation Bipolar electrogram voltage within the circles decreased from mv before ablation to mv after ablation (P,0.01). Ablation also markedly reduced electrogram voltage at the posterior wall of the left atrium from to mv (P,0.01). This reduction was consistent in all patients. In contrast, ablation did not affect left atrial activation. Conduction time from the onset of the P-wave to the latest left atrial activation (i.e. at the mitral annulus region) was similar before and after ablation ( vs ms, P ¼ NS). Also the total left atrial activation time (difference from earliest to the latest activation time) remained unchanged ( vs ms, P ¼ NS). Discussion Main findings (i) Using detailed analysis of the SA P-wave we observed a moderate but significant shortening of the P-wave after catheter ablation of AF. (ii) Circumferential PVI markedly decreases PWA in all patients, especially in the second part of the P-wave. (iii) Duration of the P-wave after catheter ablation of AF is associated with a successful outcome at follow-up. P-wave duration after catheter ablation of atrial fibrillation Our findings are in agreement with two previously reported studies on PWD after pulmonary vein isolation. Ogawa et al. 15 measured SA PWD before and after segmental ostial catheter ablation (n ¼ 27) in patients with paroxysmal and persistent AF. Patients without AF recurrence at follow-up demonstrated a significant reduction in PWD from to ms(DPWD ms, P, ), whereas patients with AF recurrence demonstrated no significant shortening in PWD (DPWD ms, P ¼ NS). A cut-off value of 5 ms in P-wave reduction had a sensitivity of 92% and specificity of 71% to predict good clinical outcome. Similarly, Okumara et al. 16 measured SA PWD before and after both ostial and circumferential PVI (n ¼ 51) in patients with paroxysmal and persistent AF. A baseline PWD of.150 ms was a powerful predictor of recurrence after CPVI and patients with successful outcome demonstrated a statistically significant shortening in PWD ( to ms, P, 0.01 vs to ms, P ¼ 0.2). In contrast, Redfearn et al. 17 demonstrated a prolongation in PWD after wide area circumferential ablation. Most likely, prolongation in PWD was a reflection of gaps in the encircling lines, because electrical isolation was not an end-point in that study. The pathogenesis of shortening of the P-wave after CPVI (and its possible predictive value on outcome) is not clear. It has been suggested that shortening in PWD may be explained by electrical disconnection of the PVs. 15,16 This hypothesis implies that the

5 802 K. Van Beeumen et al. Figure 3 (A) Bar plots showing mean PWD before and after CPVI, for all patients (n ¼ 39) and for successes (n ¼ 31) and failures (n ¼ 8). (B) Bar plots showing mean PWA before and after CPVI, for all patients. ECG, electrocardiogram; dvcg, derived vectorcardiogram. baseline terminal part of the P-wave is represented by late activation of the LA PV region. However, as shown by Date et al. 18, excitation of myocardial sleeves in the PVs mainly plays a role in the formation of the middle part of the P-wave in patients with drug refractory paroxysmal AF, and subsequent ostial electrical isolation did not result in a change in PWD. Also Lemery et al. 19 demonstrated that in patients with a history of AF, different regions of the atria contribute to the terminal portion of the P-wave. Our data are in agreement with these prior studies: we observed (i) that despite invariable electrical PVI in each patient, PWD did not shorten in all cases and (ii) that latest left atrial intra-cardiac activation which was not affected by ablation occurred at the lateral mitral annulus region. A second hypothesis might be that electrical isolation of the PVs only leads to shortening of the P-wave in patients with broad LA PV connections whereas in patients with discrete LA PV connections, PVI might not result in PWD shortening. This latter hypothesis might also explain why shortening of the P-wave was only significant in the success group. Patients with broad LA PV connections are known to have increased PV arrhythmogenesis and are therefore expected to respond to PVI. 20 Vice versa, in patients with discrete LA PV connections, there is a higher likelihood of failure after PVI because of non-pv mediated AF. Thirdly, shortening of PWD after ablation (and its possible predictive value on outcome) might be due to ablation-induced vagal denervation. Ketels et al. 11 observed an immediate vagal denervation during ongoing PVI and also in the present study heart rate had increased at the end of the procedure. Vagal denervation (by atropine) is known to shorten P wave duration. 21 The underlying mechanism of this effect is unclear and even paradoxical given the fact that pure acceleration of heart rate is expected to prolong PWD (ratedependent slowing of atrial conduction). Finally, other factors like changes in fluid status or the use of anaesthetics are less likely to have influenced the results of our study. 22,23

6 P-wave area and P-wave duration after CPVI 803 Table 2 Predictors of successful outcome after CPVI Total (n 5 39) Successes (n 5 31) Failures (n 5 8) P-value... PWD pre ms ms ms 0.42 PWD post ms ms ms 0.04* PWA-ECG pre mvms mvms mvms 0.54 PWA-ECG I pre mvms mvms mvms 0.74 PWA-ECG II pre mvms mvms mvms 0.46 PWA-ECG post mvms mvms mvms 0.42 PWA-ECG I post mvms mvms mvms 0.59 PWA-ECG II post mvms mvms mvms 0.31 PWA-dVCG pre mvms mvms mvms 0.80 PWA-dVCG I pre mvms mvms mvms 0.68 PWA-dVCG II pre mvms mvms mvms 0.98 PWA-dVCG post mvms mvms mvms 0.70 PWA-dVCG I post mvms mvms mvms 0.60 PWA-dVCG II post mvms mvms mvms 0.86 *Data on PWD and PWA before and after ablation. Only PWD post was statistically different between patients with and without successful outcome after CPVI. P-wave area after catheter ablation of atrial fibrillation Although changes in P-wave amplitude can be easily observed on the surface ECG, to our knowledge, no previous study has quantified changes in PWA after catheter ablation of AF. Using both the SA 12-lead surface ECG and the ECG dvcg, we observed a marked and significant decrease in the PWA of 20 40%. PWA was even more reduced in the second part and terminal parts (RMS40 and RMS20) of the P-wave. Ablation-induced reduction in PWA is most likely related to a reduction in electrical mass of the left atrium after encircling of the PVs. Indeed, the degree of ablation-induced reduction in PWA (20 40%) correlates well with previous reports showing that PV encircling results in an area of low voltage bipolar electrograms of, respectively, 30% (voltage,0.1 mv) and 40% (voltage,0.5 mv). 6,24 Likewise, in the present study, reduction of PWA was paralleled by marked reduction in bipolar electrogram voltage not only within the encircled PV area (+28% of LA surface) but also at the posterior wall of LA in between the circles. Ablation-induced reduction in PWA was more marked in the ECG dvcg compared with the 12-lead surface ECG. This may not be surprising, as the amplitude of the P-wave vector (PWA-dVCG) itself is not influenced by variations in electrical axis, whereas the amplitude of the P-wave of an ECG lead is influenced by both length and axis of the P-wave vector. As a consequence, a reduction in PWA-ECG after ablation due to a decrease in atrial electrical mass may result in a decrease in P-wave amplitude, but may be partially compensated due to changes in the axis of the vector. As our aim was to study changes in magnitude of electrical activity (vector amplitude) and not changes in spatial orientation of electrical activity (vector axis), PWA-dVCG might be more accurate than PWA-ECG to describe changes in magnitude of atrial electrical mass. In the present study, ablation-induced reduction in PWA was no predictor for clinical outcome after catheter ablation of AF. Previous studies on electrical mass reduction show varying results. Pappone et al. 6 demonstrated that a large low-voltage area after ablation was a predictor of successful outcome after PV ablation in 297 patients with paroxysmal AF. On the other hand, in the study by Verma et al. 24 the low voltage area was identical between successes and failures. Clinical implications The present study elucidates the underlying mechanism of the macroscopic decrease in PWA, easily seen on the surface ECG. However, besides the cognitive value of this manuscript, our findings might also have clinical implications. Our study indicates that PWD measured immediately before and after circumferential PVI can be used as a predictor of clinical outcome at follow-up. Further studies are needed to determine whether a possible cut-off value for PWD post can be used as an on-line tool to guide the optimal ablation strategy during the procedure itself. In patients with lack of P-wave shortening immediately after electrical isolation of the pulmonary veins, additional modification of the substrate might be required. Before we can implement such a strategy however, further research is needed to investigate the exact cause of P wave changes and whether this betrays a modifiable substrate that may be amenable to ablation. Only then on-line evaluation of PWA or PDW (e.g. via automated signal-averaging algorithms) could be used to guide catheter ablation. Study limitations In the present study, only a limited number of patients were included. Therefore, changes in PWA could not (yet) be identified as a predictor of success. Calculation of the PWD was performed after manual annotation of earliest onset and latest offset of the P-wave on the SA-ECG. In the present study, we did not perform signal-averaging of intra-cardiac electrograms although it allows for better registration of small potentials and has previously been described. 25 However, in the present study, unstable position

7 804 K. Van Beeumen et al. of the mapping catheter would limit detection and beat-to-beat averaging of the signals within a short period of time and would also make comparison of the averaged electrograms before and after ablation more difficult. In the present study, we used a standard 12-lead ECG position and not orthogonal lead placement for calculation of the VCG. Carlson et al. 12 demonstrated that the inverse dower transform (performed on a subset of leads), provides a derived VCG comparable with the VCG of orthogonal leads. Conflict of interest: none declared. References 1. Pappone C, Rosanio S, Oreto G, Tocchi M, Gugliotta F, Vicedomini G et al. Circumferential radiofrequency ablation of pulmonary vein ostia. A new anatomic approach for curing atrial fibrillation. Circulation 2000;102: Liu X, Long D, Dong J, Hu F, Yu R, Tang R et al. Is circumferential pulmonary vein isolation preferable to stepwise segmental pulmonary vein isolation for patients with paroxysmal atrial fibrillation? A randomised study. Circ J 2006;70: Ouyang F, Bänsch D, Ernst S, Schaumann A, Hachiya H, Chen M et al. Complete isolation of left atrium surrounding the pulmonary veins. New insights from the double-lasso technique in paroxysmal atrial fibrillation. Circulation 2004;110: Ouyang F, Antz M, Ernst S, Hachiya H, Mavrakis H, Deger FT et al. Recovered pulmonary vein conduction as a dominant factor for recurrent atrial tachyarrhythmias after complete circular isolation of the pulmonary veins: lessons from double Lasso technique. Circulation 2005;111: Oral H. Mechanisms of atrial fibrillation: lessons from studies in patients. Prog Cardiovasc Dis 2005;48: Pappone C, Santinelli V, Manguso F, Vicedomini G, Gugliotta F, Augello G et al. Pulmonary vein denervation enhances long-term benefit after circumferential ablation for paroxysmal atrial fibrillation. Circulation 2004;109: Xia Y, Hertervig E, Kongstad O, Ljungström E, Platonov P, Holm M et al. Deterioration of interatrial conduction in patients with paroxysmal atrial fibrillation: Electroanatomic mapping of the right atrium and coronary sinus. Heart Rhythm 2004;1: Budeus M, Felix O, Hennersdorf M, Wieneke H, Erbel R, Sack S. Prediction of conversion from paroxysmal to permanent atrial fibrillation. PACE 2007;30: Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA. Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation 1995;92: Byrd GD, Prasad SM, Ripplinger CM, Cassily TR, Schuessler RB, Boineau JP et al. Importance of geometry and refractory period in sustaining atrial fibrillation. Testing the critical mass hypothesis. Circulation 2005;112:I7 I Ketels S, Houben R, Van Beeumen K, Tavernier R, Duytschaever M. Incidence, timing and characteristics of acute changes in heart rate during ongoing circumferential pulmonary vein isolation. Europace 2008;10: Carlson J, Havmoller R, Herreros A, Platonov P, Johansson R, Olsson B. Can orthogonal lead indicators of propensity to atrial fibrillation be accurately assessed from the 12-lead ECG? Europace 2005;7:S39 S Guillem MS, Sahakian AV, Swiryn S. Derivation of orthogonal leads from the 12-lead electrocardiogram. Performance of an atrial-based transform for the derivation of P loops. J electrocardiol 2008;41: Pan J, Tompkins WJ. A real-time QRS detection algorithm. IEEE Trans Biomed Eng 1985;32: Ogawa M, Kumagai K, Vakulenko M, Yasuda T, Siegerman C, Garfinkel A et al. Reduction of P-wave duration and successful pulmonary vein isolation in patients with atrial fibrillation. J Cardiovasc Electrophysiol 2007;18: Okumura Y, Watanabe I, Ohkubo K, Ashino S, Kofune M, Hashimoto K et al. Prediction of the efficacy of pulmonary vein isolation for the treatment of atrial fibrillation by the signal-averaged p-wave duration. Pacing Clin Electrophysiol 2007;30: Redfearn D, Skanes A, Gula J, Griffith M, Marchall H, Stafford P et al. Noninvasive assessment of atrial substrate change after wide area circumferential ablation: a comparison with segmental pulmonary vein isolation. Ann Noninvasive Electrocardiol 2007;12: Date T, Yamane T, Inada K, Matsuo S, Kanzaki Y, Miyanaga S et al. The effects of pulmonary vein isolation on the morphology of p waves: the contribution of pulmonary vein muscle excitation to the formation of p waves. Pacing Clin Electrophysiol 2007;30: Lemery R, Birnie D, Tang ASL, Green M, Gollob M, Hendry M et al. Normal atrial activation and voltage during sinus rhythm in the human heart: an endocardial and epicardial mapping study in patients with a history of atrial fibrillation. J Cardiovasc Electrophysiol 2007;18: Nakagawa H, Aoyama H, Beckman KJ, Po SS, Wu R, Lockwood D et al. Relation between pulmonary vein firing and extent of left atrial-pulmonary vein connection in patients with atrial fibrillation. Circulation 2004;109: Cheema A, Ahmed M, Kadish A, Goldberger J. Effects of autonomic stimulation and blockade on signal-averaged P wave duration. J Am Coll Cardiol 1995;26: Rizzo V, Di Maio F, Campbell SV, Tallarico D, Petretto F, Lorido A et al. Left ventricular function, cardiac dysrhythmias, atrial activation, and volumes in nondipper hypertensive individuals with left ventricular hypertrophy. Am Heart J 2000;139: Owczuk R, Wujtewicz MA, Sawicka W, Polak-Krzeminska A, Suszynska-Mosiewicz A, Raczynska K et al. Effect of anaesthetic agents on p-wave dispersion on the electrocardiogram: comparison of propofol and desflurane. Clin Exp Pharmacol Physiol 2008;35: Verma A, Kilicaslan F, Pisano E, Marrouche NF, Fanelli R, Brachmann J et al. Response of atrial fibrillation to pulmonary vein antrum isolation is directly related to resumption and delay of pulmonary vein conduction. Circulation 2005;112: Berger RD, Nsah E, Calkins H. Signal-averaged intracardiac electrograms: a new method to detect kent potentials. J Cardiovasc Electrophysiol 1997;8: