Atrial electrical remodeling has been shown after termination of atrial flutter (AFL); however, whether abnormalities persist beyond an arrhythmic episode is not known. We aimed to characterize the atrial substrate, remote from arrhythmia, in patients with typical AFL. We compared 20 patients, studied remote from episodes of typical AFL and without a history of atrial fibrillation, to 20 reference patients. Multipolar catheters placed at the lateral right atrium (RA), coronary sinus, crista terminalis, and septal RA measured the effective refractory period at 5 sites; conduction characteristics at the crista terminalis; and the conduction time along the lateral RA and coronary sinus. Electroanatomic right atrial maps were created to determine regional differences in voltage and conduction. Patients with AFL demonstrated the following compared to the reference patients: a larger right atrial volume (121 ± 30 vs 83 ± 24 ml, p = 0.005); a prolonged P-wave duration (122 ± 18 vs 102 ± 11 ms, p = 0.007); a longer right atrial activation time (107 ± 23 vs 85 ± 14 ms, p = 0.02); a prolonged conduction time along the lateral RA (67 ± 4 vs 47 ± 3 ms, p <0.001); a slower mean conduction velocity (1.2 ± 0.2 vs 2.1 ± 0.6 mm/ms, p <0.001); a greater proportion of fractionated electrographic findings (16 ± 4% vs 10 ± 6%, p = 0.006); more frequent abnormal electrographic findings at the crista terminalis (4.1 ± 2.6 vs 1.0 ± 1.1, p = 0.001); a prolonged corrected sinus node recovery time (318 ± 71 vs 203 ± 94 ms, p = 0.02); a trend toward greater effective refractory period (232 ± 29 vs 213 ± 12 ms, p = 0.06); and a lower voltage (2.1 ± 0.5 vs 3.0 ± 0.5 mV, p <0.001). In conclusion, studied remote from arrhythmia, patients with AFL demonstrated significant and diffuse atrial abnormalities characterized by structural changes, conduction abnormalities, and sinus node dysfunction. These persisting abnormalities characterize the substrate underlying typical AFL and may account for the subsequent development of atrial fibrillation.
The circuit of typical right atrial flutter (AFL) has been well characterized, with the crista terminalis and tricuspid annulus defined as the posterior and anterior conduction barriers, respectively. Several studies have observed atrial electrical remodeling on termination of arrhythmia that reverses within weeks of arrhythmia termination. However, these changes that follow tachycardia do not explain why patients develop AFL. In addition, although ablation of the cavotricuspid isthmus is an effective treatment of AFL, the reason many patients subsequently develop atrial fibrillation is not known. We hypothesized that patients with AFL would demonstrate a diffusely abnormal atrial substrate that would persist remote from arrhythmia to account for the occurrence of AFL and the subsequent propensity toward atrial fibrillation. To evaluate the substrate without electrical remodeling effects of the arrhythmia itself, we performed high-density electrophysiologic and electroanatomic atrial mapping remote in time from the arrhythmia.
Methods
The present study included 20 patients without antecedent atrial fibrillation undergoing ablation for typical AFL and a reference group of 20 patients undergoing ablation for atrioventricular reentry tachycardia or atrioventricular nodal reentry tachycardia. Patients with AFL were selected on the basis of having had ≥2 previous episodes of typical AFL confirmed on a 12-lead electrocardiogram. These patients underwent continuous monitoring for 7 days immediately before the study. Patients were excluded if they had previous electrocardiographic evidence of atrial fibrillation, symptoms of arrhythmia in the previous month, or any atrial arrhythmia >30 seconds in the week before ablation. All antiarrhythmic medication was ceased ≥5 half-lives before the study. All patients provided written informed consent to the study protocol, which had been approved by the Human Research Ethics Committee of the Royal Adelaide Hospital.
The electrophysiologic study was performed in the postabsorptive state with sedation using midazolam and fentanyl. Subsequently, the patients with AFL underwent cavotricuspid isthmus ablation using an externally irrigated ablation catheter (Biosense-Webster, Diamond Bar, California) with a delivered power of 35 W and irrigation rates of 30 to 60 ml/min. The end point of ablation was bidirectional conduction block across the isthmus using established criteria. The following catheters were positioned for the study protocol as previously described : a 10-pole catheter (2-5-2 mm interelectrode spacing, Daig Electrophysiology, Minnetonka, Minnesota) within the coronary sinus with the proximal bipole at the coronary sinus ostium as determined in the best septal left anterior oblique position; a 20-pole “crista” catheter (1-3-1 mm interelectrode spacing, Biosense-Webster) placed along the crista terminalis with the distal tip superiorly such that the second bipole lay at the junction of the superior vena cava and right atrium (RA), stabilized by a long sheath (CSTA, Daig Electrophysiology) to ensure close apposition to the posterolateral RA; and a 20-pole catheter (2-5-2 mm interelectrode spacing, Daig Electrophysiology) placed initially along the lateral RA and then moved to the high right atrial septum. The atrial effective refractory period (ERP) was evaluated at twice the diastolic threshold at cycle lengths of 600 and 450 ms using an 8-beat drive followed by an extra stimulus, starting with a coupling interval of 150 ms and increasing in 10 ms increments. ERP was defined as the longest coupling interval failing to propagate to the atrium. At each site, the ERP was measured 3 times at each cycle length and averaged. If the ERP varied by >10 ms, an additional 2 measurements were made, and the total number was averaged. The ERP was measured at the following sites: distal coronary sinus; proximal coronary sinus; low-lateral RA; high-lateral RA; and high-septal RA. Atrial fibrillation induced by ERP testing lasting >5 minutes was considered sustained; when this occurred, no additional data were acquired. The atrial conduction time was assessed along linearly placed catheters by pacing the distal bipole (1,2) and determining the conduction time to a proximal bipole (9,10) at the coronary sinus and lateral RA. The conduction time at each site was averaged for 10 beats during stable capture at 600 and 450 ms cycle lengths. The P-wave duration was determined by the mean of 10 beats on the electrocardiographic lead II. The number of bipoles on the crista terminalis catheter with discrete double potentials separated by an isoelectric interval or complex fractionated activity of ≥50 ms in duration and the maximum electrogram duration were determined during sinus rhythm, constant pacing, and for the shortest-coupled captured extra stimulus from the proximal coronary sinus and low-lateral RA. The corrected sinus node recovery time was determined after a 30 second drive train at cycle lengths of 600 and 450 ms, correcting for the baseline cycle length. At each cycle length, the mean corrected sinus node recovery time was determined 3 times and averaged.
Electroanatomic maps were created of the RA during sinus rhythm using the CARTO mapping system (Biosense-Webster). The electroanatomic mapping system has been previously described in detail; the accuracy of the sensor position has been previously validated and was 0.8 mm and 5°. In brief, the system records the surface electrocardiogram and bipolar electrograms filtered at 30 to 400 Hz from the mapping and reference catheters. Points were acquired in the auto-freeze mode if the stability criteria in space (≤6 mm) and local activation time (≤5 ms) were met. Mapping was performed with an equal distribution of points using a fill-threshold of 15 mm. The editing of points was performed off-line. The local activation time was manually annotated to the peak of the largest amplitude deflection on the bipolar electrograms. In the presence of double potentials, this was annotated at the largest potential. If the bipolar electrogram displayed equivalent maximum positive and negative deflections, the maximum negative deflection on the simultaneously acquired unipolar electrogram was used to annotate the local activation time. Points not conforming to the surface electrocardiogram P-wave morphology or <75% of the maximum voltage of the preceding electrogram were excluded. The regional atrial bipolar voltage and conduction velocity were analyzed as previously described and are detailed below. For the purposes of evaluating the regional voltage differences, each atrium was segmented using previously validated off-line software. The RA was segmented as the high- and low-lateral RA, high- and low-posterior RA, high- and low-septal RA, and anterior RA. At each of these regions, an average voltage of 10 points was determined. The voltage of points identified by region was exported for analysis. Low voltage points were defined as points with a bipolar voltage of ≤0.5 mV and electrically silent points as the absence of recordable activity or a bipolar voltage amplitude of ≤0.05 mV (the noise level of the system). Isochronal activation maps (5 ms intervals) of the atria were created, and the regional conduction velocity was determined in the direction of the wave front propagation (least isochronal crowding). An approximation of the conduction velocity was determined by expressing the distance between 2 points as a function of the difference in the local activation time. The mean conduction velocity for each region was determined by averaging the conduction velocity between 5 pairs of points, as previously described. For the purposes of evaluating the regional conduction differences, each atrium was segmented as above. The proportion of points demonstrating complex electrograms was determined using the following definitions: fractionated signals—complex activity of ≥50 ms in duration; and double potentials—potentials separated by an isoelectric interval with an electrogram duration of ≥50 ms.
Continuous variables are reported as the mean ± SD or the median and interquartile range, as appropriate. Categorical variables are reported as the number and percentage. Proportions were compared using Fisher’s exact test. Comparisons between means were analyzed using analysis of variance and paired or unpaired t tests, as appropriate. Comparisons with adjustment for multiple sampling within patients were performed using a mixed linear model for continuous data or a logistic generalized estimating equation for categorical data. Statistical tests were performed using the Statistical Package for Social Sciences, version 15 (SPSS, Chicago, Illinois), and statistical significance was set at p <0.05.
Results
The baseline patient characteristics are summarized in Table 1 . Patients with AFL had a median AFL history of 14 months (interquartile range 2 to 17), with no history of arrhythmia for ≥1 month (range 1 to 8) before the study. The groups were well matched for age, left ventricular dimensions, and hypertrophy; however, patients with AFL had larger left atria.
| Variable | AFL Group (n = 20) | Reference Group (n = 20) | p Value |
|---|---|---|---|
| Age (years) | 59 ± 10 | 59 ± 13 | 0.9 |
| Hypertension | 4 (20%) | 2 (10%) | 0.7 |
| Diabetes | 1 (5%) | 2 (10%) | 1 |
| Ischemic heart disease | 1 (5%) | 2 (10%) | 1 |
| Structural heart disease ⁎ | 4 (20%) | 2 (10%) | 0.7 |
| Left atrial parasternal size (mm) | 41 ± 6 | 36 ± 6 | 0.03 |
| Left atrial area (cm 2 ) | 23 ± 5 | 18 ± 3 | 0.02 |
| Right atrial area (cm 2 ) | 19 ± 4 | 15 ± 5 | 0.08 |
| Left ventricular end-diastolic dimension (mm) | 50 ± 4 | 48 ± 6 | 0.2 |
| Left ventricular end-systolic dimension (mm) | 30 ± 6 | 32 ± 6 | 0.5 |
| Interventricular septum (mm) | 11 ± 4 | 11 ± 2 | 0.9 |
⁎ Patients with atrial flutter had mild left ventricular impairment (n = 2), aortic valve replacement (n = 1), and hypertrophic cardiomyopathy (n = 1); reference patients had mild left ventricular impairment (n = 2).
For patients with AFL, the mean ERP across all sites demonstrated a trend to be greater than that of reference patients (at 600 ms, 244 ± 25 vs 227 ± 15 ms, p = 0.07; at 450 ms, 232 ± 29 vs 213 ± 12 ms, p = 0.06). At each site, patients with AFL demonstrated equal or greater ERPs compared to the reference group. However, this difference was only statistically significant at one site by post hoc testing—the high-lateral RA ( Figure 1 ). The ERP was greater after the 600 ms drive trains than after the 450 ms drive trains for both patients with AFL (244 ± 25 vs 235 ± 30, p = 0.02) and reference patients (227 ± 15 vs 213 ± 12, p <0.001).
The conduction time along the linear catheters was longer in patients with AFL than in reference patients (p = 0.04), and this demonstrated site-dependence (p <0.001 for interaction). The post hoc analysis performed within the mixed linear model revealed slower conduction along the lateral RA in patients with AFL than in reference group (67 ± 4 vs 47 ± 3 ms, p <0.001) and no difference in conduction time at the coronary sinus (45 ± 4 vs 46 ± 3 ms, p = 0.8). The P-wave duration was prolonged in patients with AFL compared to that in reference patients (122 ± 18 vs 102 ± 11 ms, p = 0.007).
Site-specific conduction abnormalities at the crista terminalis during sinus rhythm were more apparent in patients with AFL than in reference patients. During sinus rhythm, the patients with AFL had a greater number of bipoles demonstrating double potentials or fractionated signals on the crista terminalis catheter than reference patients (4.1 ± 2.6 vs 1.0 ± 1.1, respectively, p = 0.001), and the maximum electrogram duration was longer (80 ± 26 vs 53 ± 7 ms, respectively, p = 0.004). These differences were more marked during pacing and increased further with extra stimulus ( Figure 2 ). These data highlight the functional nature of conduction delay at the crista terminalis, as evidenced by the variation in the extent of conduction abnormalities depending on the rate and site of stimulation. Patients with AFL had a longer corrected sinus node recovery time at 450 ms (318 ± 71 vs 203 ± 94 ms, p = 0.02); however, at 600 ms this lengthening did not reach significance (266 ± 113 vs 194 ± 74 ms, p = 0.1).
A total of 97 ± 19 points per patient were analyzed in the RA using electroanatomic mapping. The mean right atrial volume by electroanatomic mapping was greater in patients with AFL than in reference patients (121 ± 30 vs 83 ± 24 ml, p = 0.005). The mean right atrial bipolar voltage was reduced in patients with AFL compared to that in reference patients (2.1 ± 0.5 vs 3.0 ± 0.5 mV, p <0.001). Regional differences in bipolar voltage are illustrated in Figure 3 . Significant differences between groups were seen in the high-lateral, high-posterior, and anterior regions. Additionally, points at the high-posterior region were more likely to be low voltage (≤0.5 mV) in patients with AFL (odds ratio 3.7, 95% confidence interval 2.1 to 6.7). No areas of electrical silence (≤0.05 mV) were observed. Illustrative examples of electroanatomic maps in representative patients are shown in Figure 4 .
