Characterising the difference in electrophysiological substrate and outcomes between heart failure and non-heart failure patients with persistent atrial fibrillation

Characterising the difference in electrophysiological substrate and outcomes between heart... Abstract Aims Characterizing the differences in substrate and clinical outcome between heart failure (HF) and non-heart failure (non-HF) patients undergoing persistent atrial fibrillation (AF) ablation. Methods and results Using complex fractionated electrograms (CFE) as a surrogate marker of substrate complexity, we compared the bi-atrial substrate in patients with persistent AF with and without HF, at baseline and after ablation, to determine its impact on clinical outcome. In this retrospective analysis of two prospective studies, 60 patients underwent de-novo step-wise left atrial (LA) ablation, 30 with normal left ventricular ejection fraction (LVEF) ≥ 50% (non-HF group) and 30 with LVEF ≤ 35% (HF group). Multiple high-density bi-atrial CFE maps were acquired along with AF cycle length (AFCL) at each procedural stage. Change in bi-atrial CFE areas, AFCL and outcome data were then compared. In the non-HF group, higher CFE-areas were found at baseline and at each step of the procedure in the LA. In both LA and the right atrium (RA), baseline and final CFE area were also higher in the non-HF group. Single procedure, arrhythmia-free survival at 1 year was higher in the HF group compared with the non-HF group (72% vs. 43%, log rank P = 0.04). Final total bi-atrial CFE area was an independent predictor of arrhythmia recurrence. Conclusions CFE represents an important surrogate marker of atrial substrate complexity. The atrial substrate in persistent AF differs between HF and non-HF with the latter representing a more complex ‘primary’ bi-atrial myopathy. LA focussed ablation results in more extensive substrate modification in HF and better clinical outcomes as compared with non-HF. Catheter ablation, Atrial Fibrillation, Heart Failure, Persistent What’s new? There are important differences in electrophysiological substrate in persistent/LSPAF patients with and without HF. Persistent/LSPAF without HF (or other structural heart disease) can represent a more advanced ‘primary’ atrial myopathy with higher baseline CFE burden in both LA and RA, as compared with AF associated with HF. An LA focussed ablation strategy resulted in better clinical outcomes in the HF group. Residual bi-atrial CFE burden post-LA stepwise ablation was an independent predictor of clinical success. CFE may be more useful as a surrogate marker of atrial substrate complexity than a primary target of ablation. Introduction Catheter ablation for persistent and long-standing persistent AF (LSPAF) remains one of the most challenging areas of clinical electrophysiology. Clinical outcomes remain suboptimal, despite significant technical and technological advances and the use of adjunct ablation strategies to pulmonary vein isolation (PVI).1,2 This challenge reflects the complex pathological milieu that underlies persistent AF substrates which may be further complicated by different disease states. Complex fractioned atrial electrograms (CFEs) are thought to reflect areas of slow conduction, localized re-entry or pivot points for change in wavelet orientation. Although they may not represent target sites for AF ablation, they are likely to represent surrogate markers of substrate complexity that may differ in different underlying disease state that maintain AF.3,4 This study was designed to evaluate the hypothesis that the underlying atrial substrate and response to conventional stepwise LA ablation is different in persistent and LSPAF patients with and without HF. Methods Study population This is a retrospective analysis of data collected from two prospective studies where patients received the same ablation lesion sets using the same mapping and ablation technologies. Sixty patients in total underwent de novo CA for symptomatic persistent/LSPAF. Thirty patients with normal LVEF ≥ 50% (non-HF) recruited from a prospective clinical trial (NCT01385358) and compared with 30 HF patients with an LVEF ≤ 35% (HF) from a previous prospective study (NCT00878384).5 All patients gave written informed consent for the study, approved by the local ethics committee in accordance with the Declaration of Helsinki. Ablation and mapping protocol Under general anaesthesia, extensive bi-atrial CFE mapping was performed using an electroanatomical mapping system (EnSite VelocityTM, St Jude Medical, MN, USA) and a multi-electrode high density (AFocus IITM, St Jude Medical) mapping catheter. The step-wise ablation protocol has been previously described and consists of PVI, linear lesions (LL) (roof and mitral isthmus) and LA CFE ablation using non-contact force 3.5 mm irrigated-tip D-F SF ablation catheters (Thermocool Navistar, Biosense Webster).5 Bi-atrial CFE maps were acquired at baseline and post CFE ablation, with additional LA only CFE maps after PVI and LL (Figure 1). At each stage, the right (RAA) and left atrial appendage (LAA) cycle length (CL) was recorded. Figure 1 View largeDownload slide Procedural protocols for non-HF and HF groups—sequence of treatment and mapping. Protocol summary is shown in the diagram. The number of maps at each stage is shown in grey boxes. The ablation and DC cardioversion stages are shown in white boxes. Number of AF terminations is shown at each stage of procedure. Technical software failure resulted in the loss of one baseline, two final RA maps, and three final LA maps in the non-HF group. RA, right atrium; LA, left atrium; PVI, pulmonary vein isolation; CFE, complex fractionated electrograms; LL, linear lesions. Figure 1 View largeDownload slide Procedural protocols for non-HF and HF groups—sequence of treatment and mapping. Protocol summary is shown in the diagram. The number of maps at each stage is shown in grey boxes. The ablation and DC cardioversion stages are shown in white boxes. Number of AF terminations is shown at each stage of procedure. Technical software failure resulted in the loss of one baseline, two final RA maps, and three final LA maps in the non-HF group. RA, right atrium; LA, left atrium; PVI, pulmonary vein isolation; CFE, complex fractionated electrograms; LL, linear lesions. All CFE sites marked in the post LL maps were ablated. If AF persisted after LA CFE ablation, direct current cardioversion was used to restore sinus rhythm (SR). PVI and LL bi-directional block were then assessed using conventional techniques with incomplete lines re-ablated to achieve bi-directional block. Cavo-tricuspid isthmus ablation was only performed if there was documented typical atrial flutter. If AF terminated to atrial tachycardia (AT), the protocol was terminated and the AT mapped and ablated. CFE mapping Using the roving high-density mapping catheter, signals were recorded from all 19 bipoles over 5 s per acquisition until the whole endocardial surface was covered. The mitral/tricuspid annuli were defined by electrogram characteristics and the contained area excluded. The NavX CFE mean tool was used to identify high-frequency atrial signals with multiple components with the following settings: electrogram width <10 ms (to remove far field signals), refractory period 30 ms (values below this being regarded as nonphysiological for local reactivation), and interpolation and surface projection <10 mm based on previous studies. Voltage detection threshold was adjusted to exclude background noise and avoid false-positive CFE annotation and fixed for subsequent maps. Points >10 mm from the surface, and those displaying electric interference, were deleted. Scar was defined as <0.05 mV. CFE mean was defined as the mean time between consecutive deflections during a 5-s recording period. CFE in this and the previous study was defined as sites with CFE mean ≤120 ms.5 Analysis of CFE data As previously described, each atrium was segmented into three regions to categorise distribution of CFE.5 The LA regions were anterior, posterior and appendage. The RA regions were lateral, septal and appendage. All contiguous zones of CFE within each segment of LA and RA were delineated and the area enclosed defined as CFE-area. The LA CFE-area, however, excluded the PVI encircling lesions, linear lesions (within 5 mm), and the mitral valve annulus. Follow-up The primary study endpoint was freedom from recurrent AF or atrial tachyarrhythmias (AT) ≥30 s after a single procedure and off AADs after the 3-month blanking period. Class I and III antiarrhythmic drugs (AADs) were discontinued before or at ablation. Patients were followed up for a minimum of 48 h–7 days ambulatory monitoring at 3, 6, and 12 months. Reported symptoms outside these time points were assessed with 12-lead ECG and further ambulatory ECG as indicated.4 Statistical analysis Continuous data are presented as mean (SD) or median (IQR), and categorical data as number and percentage. Paired t-tests (within group) and unpaired t-tests (between groups) were used to analyse change in CFE-area (absolute and percentage coverage of atrial surface area) and AFCL with Bonferroni correction to account for multiple comparisons. Arrhythmia-free survival was analysed by Kaplan–Meier survival curves with log-rank comparisons using the Mantel–Cox test. Cox regression was used to assess predictors of arrhythmia-free survival and only variables that had P < 0.10 were included in the multivariable model. P values <0.05 were considered statistically significant. Data were analysed using GraphPad Prism version 6.00 for Mac (GraphPad Software, San Diego, CA, USA) and R statistical software (version 3.1.2). Results Baseline characteristics are shown in Table 1. Mean duration of continuous AF was greater than 12 months in both groups. In the HF cohort, 40% were ischaemic in aetiology. The LA diameter on transthoracic echocardiography (TTE) was smaller (although similar surface area on electroanatomical map) and hypertension was more prevalent in the non-HF group. Table 1 Baseline characteristics (N = 60)   Non-HF  HF  P-value  Age (y), mean ± SD  66 ± 9  63 ± 9  0.09  Sex (male), n (%)  23 (48)  25 (52)  0.51  Duration of continuous AF (months), mean ± SD  19 ± 4  24 ± 8  0.28  LA size, mean ± SD  44 (5)  50 (7)  <0.001  LVEF (%), mean ± SD  60 ± 8  25 ± 11  <0.001  Hypertension, n (%)  21 (72)  10 (33)  0.004  Coronary artery disease, n (%)  9 (31)  12 (40)  0.59  Diabetes Mellitus, n (%)  1 (3)  7 (23)  0.05  Beta-blocker, n (%)  18 (72)  28 (93)  0.06  CHA2DS2-VASC, mean ± SD  1.9 ± 1.1  2.7 ± 1.5  0.03    Non-HF  HF  P-value  Age (y), mean ± SD  66 ± 9  63 ± 9  0.09  Sex (male), n (%)  23 (48)  25 (52)  0.51  Duration of continuous AF (months), mean ± SD  19 ± 4  24 ± 8  0.28  LA size, mean ± SD  44 (5)  50 (7)  <0.001  LVEF (%), mean ± SD  60 ± 8  25 ± 11  <0.001  Hypertension, n (%)  21 (72)  10 (33)  0.004  Coronary artery disease, n (%)  9 (31)  12 (40)  0.59  Diabetes Mellitus, n (%)  1 (3)  7 (23)  0.05  Beta-blocker, n (%)  18 (72)  28 (93)  0.06  CHA2DS2-VASC, mean ± SD  1.9 ± 1.1  2.7 ± 1.5  0.03  n = 30. AF, atrial fibrillation; LA, left atrial; LVEF, left ventricular ejection fraction. Procedural and mapping data Total procedural time (271 ± 55 vs. 331 ± 55 mins; P < 0.001), fluoroscopy time (52 ± 16 vs. 79 ± 18 min; P < 0.001), total RF time (58 ± 15 vs. 82 ± 19 min; P < 0.001 and PVI RF time (28 ± 10 vs. 46 ± 17 min; P < 0.001) were all shorter in the non-HF group (Table 2). Number of points per map in the LA (475 ± 191 vs. 479 ± 99; P = 0.92) and RA (410 ± 161 vs. 373 ± 96; P = 0.14) did not differ between the non-HF and HF groups, respectively. Total LA surface area (210 ± 38 vs. 213 ± 43; P = 0.75) and LA CFE area analysis zone (114 ± 30 vs. 117 ± 24; P = 0.62) were similar but the total RA surface area (178 ± 41 vs. 136 ± 33; P < 0.001) was higher in the non-HF group. Termination of AF during ablation (termination to SR without DC cardioversion) was observed in 11 cases, 5 (2 during roof line and 3 via AT during CFE ablation) in the non-HF group and 6 (4 via AT and 2 direct to SR during CFE ablation) in the HF group (Figure 2) . Table 2 Comparison of procedural and mapping data between non-HF and HF groups   Non-HF  HF  P value  Total procedural time (min)  271 (±55)  331 (±55)  0.001  Fluoroscopy time (min)  52 (±16)  79 (±18)  <0.001  Total RF time (min)  58 (±15)  82 (±19)  <0.001  PVI RF time (min)  28 (±10)  46 (±17)  <0.001  Roof line RF time (min)  3.4 (±1.4)  3.3 (±1.7)  0.97  Mitral isthmus line RF time (min)  4.9 (±2.4)  4.0 (±2.4)  0.16  CFE RF time (min)  9.4 (±7.0)  12.1 (±7.7)  0.17  LA points per map  475 (±191)  479 (±99)  0.92  RA points per map  410 (±161)  373 (±96)  0.14  LA CFE area analysis zone (cm2)  114 (±30)  117 (±24)  0.62  Total LA surface area (cm2)  210 (±38)  213 (±43)  0.75  Total RA surface area (cm2)  178 (±41)  136 (±33)  <0.001  Unblocked linear lesions  0/30  3/30  0.07    Non-HF  HF  P value  Total procedural time (min)  271 (±55)  331 (±55)  0.001  Fluoroscopy time (min)  52 (±16)  79 (±18)  <0.001  Total RF time (min)  58 (±15)  82 (±19)  <0.001  PVI RF time (min)  28 (±10)  46 (±17)  <0.001  Roof line RF time (min)  3.4 (±1.4)  3.3 (±1.7)  0.97  Mitral isthmus line RF time (min)  4.9 (±2.4)  4.0 (±2.4)  0.16  CFE RF time (min)  9.4 (±7.0)  12.1 (±7.7)  0.17  LA points per map  475 (±191)  479 (±99)  0.92  RA points per map  410 (±161)  373 (±96)  0.14  LA CFE area analysis zone (cm2)  114 (±30)  117 (±24)  0.62  Total LA surface area (cm2)  210 (±38)  213 (±43)  0.75  Total RA surface area (cm2)  178 (±41)  136 (±33)  <0.001  Unblocked linear lesions  0/30  3/30  0.07  Figure 2 View largeDownload slide CFE-mean (CFE mean ≤120 ms) map with surface markings of CFE distribution. Posterior aspect of LA geometry with CFE-mean map. The map shown is after PVI ablation; however, offline surface annotation was used to annotate CFE areas with colour coded annotation schema at each stage of the procedure: baseline (red), post-PVI (amber), post-linear lesions (green), final post CFE ablation (blue). The majority of these annotations are left projected onto the map surface for illustrative purposes. Figure 2 View largeDownload slide CFE-mean (CFE mean ≤120 ms) map with surface markings of CFE distribution. Posterior aspect of LA geometry with CFE-mean map. The map shown is after PVI ablation; however, offline surface annotation was used to annotate CFE areas with colour coded annotation schema at each stage of the procedure: baseline (red), post-PVI (amber), post-linear lesions (green), final post CFE ablation (blue). The majority of these annotations are left projected onto the map surface for illustrative purposes. Impact of ablation on remote LA CFE area (intragroup analysis) The impact of ablation upon the CFE area for each group are described below. Results are also shown as percentage of analysed LA surface (Figure 3). More detailed data for the constituent segments of each atrium are shown in Supplementary material online, Table 1. In the non-HF group, a sequential reduction of the LA CFE area was observed from baseline 27.6 ± 11.8 cm2 (24.4 ± 10.2% of analysed LA surface) to post-PVI 21.6 ± 12.5 cm2 (18.6 ± 10.4%, P = 0.002 vs. baseline), and after the addition of LL to 16.5 ± 9.8 cm2 (14.8 ± 9.5%, P = 0.008 vs. post-PVI). Direct CFE ablation further reduced final LA CFE-area, compared with post-LL analysis, to 7.8 ± 8.7 cm2 (6.7 ± 6.6%, P = 0.001 vs. post-LL). Figure 3 View largeDownload slide Impact of stepwise LA ablation on LA CFE and RA CFE area in non-HF and HF groups. Panel A—graphs show the percentage (mean ± SD) coverage of CFE of the atrial surface (after exclusion of PVI and LL), within the segmented LA at baseline and after each stage of ablation. Panel B—graphs show the percentage (mean ± SD) coverage of CFE of the atrial surface within the segmented RA at baseline (pre) and after the completion of all the steps of LA ablation (post) The first graph is of the Non-HF group and the second the HF group allowing for visual comparison of CFE quantification between the two groups. P values are shown for comparisons between steps of ablation and denoted * <0.05, ** <0.01, and *** <0.001. Figure 3 View largeDownload slide Impact of stepwise LA ablation on LA CFE and RA CFE area in non-HF and HF groups. Panel A—graphs show the percentage (mean ± SD) coverage of CFE of the atrial surface (after exclusion of PVI and LL), within the segmented LA at baseline and after each stage of ablation. Panel B—graphs show the percentage (mean ± SD) coverage of CFE of the atrial surface within the segmented RA at baseline (pre) and after the completion of all the steps of LA ablation (post) The first graph is of the Non-HF group and the second the HF group allowing for visual comparison of CFE quantification between the two groups. P values are shown for comparisons between steps of ablation and denoted * <0.05, ** <0.01, and *** <0.001. In the HF group, a similar trend was seen with sequential reduction of the LA CFE area from baseline 18.3 ± 12.0 cm2 (16.2 ± 10.6% of the analyzed LA surface) to post-PVI 10.2 ± 7.1 cm2 (9.0 ± 6.6%; P < 0.001 vs. baseline) and after the addition of LL and compared with post-PVI analysis, total LA CFE area reduced to 7.7 ± 6.5 cm2 (6.9 ± 5.9%; P = 0.006). Direct CFE ablation further reduced final LA CFE area, compared with post-LL ablation analysis, to 3.1 ± 3.5 cm2 (2.8 ± 3.0%; P = 0.002). Impact of ablation on remote RA CFE area (intragroup analysis) In the non-HF group, LA ablation also resulted in reduction of the RA CFE area, from baseline 39.8 ± 19.3 cm2 (22.1 ± 9.2% of the RA surface) to 24.1 ± 12.9 cm2 (13.3 ± 8.1%, P = 0.0001 vs. baseline) in the final RA CFE map (see Supplementary material online, Table 1). Similarly, in the HF group, baseline RA CFE area was reduced from 25.9 ± 14.1 cm2 (19.2 ± 10.3% of the total RA surface) to 12.9 ± 11.8 cm2 (9.9 ± 7.8%; P < 0.001) in the final RA CFE map. Results are also shown as the percentage of analysed RA surface (Figure 3). Comparing the impact of ablation on remote CFE-area between groups In the LA, CFE areas in the non-HF patients were all higher than in the HF group at each step of the procedure (see Supplementary material online, Table 1): baseline total CFE area (27.6 ± 11.8 vs. 18.3 ± 12.0 cm2; P = 0.004); post-PVI total CFE area (21.6 ± 12.5 vs. 10.2 ± 7.1 cm2; P = 0.001), post-LL total CFE area (16.5 ± 9.8 vs. 7.7 ± 6.5 cm2; P < 0.001), and final total CFE area (7.8 ± 8.7 vs. 3.1 ± 3.5 cm2; P = 0.018). Similarly, in the RA, the baseline total RA CFE area (39.8 ± 19.3 vs. 25.9 ± 14.1 cm2; P = 0.002) and final total RA CFE area (24.1 ± 12.9 vs. 12.9 ± 11.8 cm2; P = 0.003) were also higher in the non-HF group. Impact of ablation upon AFCL The impact of LA ablation on bi-atrial AFCL exhibited two different patterns of change (Table 3). In the HF group, both the LA (161 ± 27 to 180 ± 42; P = 0.004) and RA (167 ± 33 to 178 ± 40; P < 0.001) AFCL prolonged in parallel, whereas in the non-HF group, only the LA (153 ± 18 to 173 ± 29; P < 0.0001) AFCL prolonged resulting in a divergent AFCL pattern. Table 3 Overall change in AFCL pre- and post-LA ablation with intragroup analysis for both non-HF and HF groups   NON-HF   HF     AFCL (ms)  P value  AFCL (ms)  P value  RA baseline  163±18  0.26  167±32  0.004  RA final  168±24  178±40  LA baseline  153±18  <0.0001  161±28  <0.0001  LA final  173±29  180±42    NON-HF   HF     AFCL (ms)  P value  AFCL (ms)  P value  RA baseline  163±18  0.26  167±32  0.004  RA final  168±24  178±40  LA baseline  153±18  <0.0001  161±28  <0.0001  LA final  173±29  180±42  Clinical outcome—single procedure arrhythmia-free success Single procedure arrhythmia-free survival at 1 year, off AADs, was 72% for the HF group vs. 43% for the non-HF group (log rank, P = 0.04, Figure 4). The final total bi-atrial CFE area was the only independent predictor of arrhythmia recurrence in multivariable analysis (Table 4). In the non-HF group, one patient had a primary intra-cerebral event in the context of a supra-therapeutic INR level within the blanking period. In the HF group, two patients died from progressive HF at 1 and 11 months. Table 4 Cox regression analysis model for arrhythmia recurrence after a single ablation procedure n = 60   Univariable analysis   Multivariable analysis     HR (95% CI)  P value Valuevalue  HR (95% CI)  P value  Male  0.50 (0.22–1.14)  0.10      Age/yr  0.99 (0.95–1.03)  0.64      LA size/5mm  0.96 (0.81–1.15)  0.67      LV ejection fraction/5%  1.41 (0.96–2.09)  0.08  1.01 (0.61–1.69)  0.96  AF, duration/months  0.99 (0.97–1.02)  0.76      Baseline LA CFE area/cm2  1.02 (0.99–1.05)  0.25      Baseline RA CFE area/cm2  1.01 (0.99–1.04)  0.22      Baseline total bi-atrial CFE area/cm2  1.01 (1.00–1.02)  0.20      Final total bi-atrial CFE area/cm2  1.03 (1.01–1.06)  0.001  1.03 (1.01–1.06)  0.01  Reduction in total bi-atrial CFE area/cm2  0.99 (0.98–1.01)  0.43      Total RF, duration/10 min  0.82 (0.68–1.00)  0.05  0.87 (0.69–1.09)  0.23    Univariable analysis   Multivariable analysis     HR (95% CI)  P value Valuevalue  HR (95% CI)  P value  Male  0.50 (0.22–1.14)  0.10      Age/yr  0.99 (0.95–1.03)  0.64      LA size/5mm  0.96 (0.81–1.15)  0.67      LV ejection fraction/5%  1.41 (0.96–2.09)  0.08  1.01 (0.61–1.69)  0.96  AF, duration/months  0.99 (0.97–1.02)  0.76      Baseline LA CFE area/cm2  1.02 (0.99–1.05)  0.25      Baseline RA CFE area/cm2  1.01 (0.99–1.04)  0.22      Baseline total bi-atrial CFE area/cm2  1.01 (1.00–1.02)  0.20      Final total bi-atrial CFE area/cm2  1.03 (1.01–1.06)  0.001  1.03 (1.01–1.06)  0.01  Reduction in total bi-atrial CFE area/cm2  0.99 (0.98–1.01)  0.43      Total RF, duration/10 min  0.82 (0.68–1.00)  0.05  0.87 (0.69–1.09)  0.23  HR, hazard ratio; CI, confidence interval; AF, atrial fibrillation; CFE, complex fractionated electrogram; LA, left atrial; LV, left ventricular; RA, right atrial; RF, radiofrequency ablation. Figure 4 View largeDownload slide Freedom from atrial arrhythmias (single procedure) at 12 months. Non-HF vs. HF groups, single procedure Kaplan–Meier curves of atrial arrhythmia-free survival off AAD (censored at 365 days). Success is defined as freedom from atrial arrhythmia after a 3-month blanking period. The two survival curves were compared using the log-rank (Mantel–Cox) test. Figure 4 View largeDownload slide Freedom from atrial arrhythmias (single procedure) at 12 months. Non-HF vs. HF groups, single procedure Kaplan–Meier curves of atrial arrhythmia-free survival off AAD (censored at 365 days). Success is defined as freedom from atrial arrhythmia after a 3-month blanking period. The two survival curves were compared using the log-rank (Mantel–Cox) test. Discussion The principle findings in this study are that in persistent/LSPAF patients: The baseline atrial electrophysiological substrate is different between the HF and non-HF group with the latter exhibiting higher total CFE areas in both the LA and RA. LA step-wise ablation results in more extensive substrate modification in the HF group with lower final CFE areas in both LA and RA which can be related to significantly higher single-procedure arrhythmia-free survival at 12 months (72% vs. 43%, log rank P=0.04) in the HF group as compared with the non-HF group. The residual bi-atrial CFE area post LA focussed stepwise ablation was an independent predictor of arrhythmia recurrence. CFE burden In addition to the above findings, our data show the sequential reduction in remote CFE areas bi-atrially after PVI and LL ablation in both groups, except of the LAA in the non-HF group (Figure 3). This is in line with previous data and substantiates previous claims that many CFE represent incidental or passive activation rather than source activity.5,6 What is new is the fact that the residual bi-atrial CFE area was shown to be important in predicting clinical success. This residual CFE area in our study comprises not only CFE in the LAA and the RA, which are areas that have not been ablated, but also ‘new’ CFE in the LA despite targeted ablation of all CFE sites post-PVI and LL lesions. Our protocol did not include a further stage to repeatedly target the final CFE areas which is the technique used by some centres when pursuing the endpoint of AF termination but it does raise the question as to whether these residual CFE are a marker for arrhythmia continuation without being an ablation target. Given the recent negative randomised clinical data showing no incremental benefits in ablating CFE,7–9 the remote effects on CFE post PVI and LL,5,6 and the more extensive distribution of CFE in persistent AF as compared to paroxysmal AF,10,11 one can conclude that CFE may actually be a dynamic surrogate marker of the atrial electrophysiological substrate rather than represent an anatomical target for ablation. Atrial substrates in different disease states Intuitively AF substrates would be expected to differ in different pathological states.12 In patients with paroxysmal AF and heart failure, recent data showed improved outcomes when non-PV trigger ablation was undertaken in addition to PVI.13 Catheter ablation using extensive substrate modification strategies have also shown good outcomes, in persistent AF and heart failure, with 70-81% freedom from arrhythmia rates which is consistent with our HF group success rates.14–18 In a human study of chronic atrial stretch and volume overload secondary to ASD, Morton et al. showed greater electrical remodelling in these patients when compared with normal.19 In left ventricular dysfunction, Roberts-Thomson et al.20 demonstrated greater conduction delay, heterogeneity, and anisotropy, compared with normal. Our data goes rather than go a step further to show the electrophysiological manifestation of these remodelling changes relevant to catheter ablation but interestingly shows that persistent/LSPAF patients harbour more abnormal and complex electrical substrates in the absence of heart failure. These findings at first seem to be at odds with the widely accepted notion that structural heart disease confers additional complexity to the atrial substrate. However, the important distinction to be made is whether the arrhythmia is the primary disturbance or secondary to another cause. James Cox in his field of AF surgery previously proposed a different classification for AF based on underlying cause.21 ‘Primary’ AF was described as those without another cardiac co-morbidity serious enough to warrant concomitant surgical intervention, and ‘secondary’ AF as due to another left heart condition (e.g. ischaemia, heart failure and valvular heart disease). This was based on the finding that ‘primary’ non-paroxysmal AF harboured a very abnormal bi-atrial substrate, which mandated bi-atrial surgical lesions to achieve similar efficacy to that of secondary AF treatable by surgical procedures confined to the LA.21 This concept is strongly supported by our data and suggests that more importance should be given to the bi-atrial aspect of the substrate in patients with persistent/LSPAF in the absence of structural heart disease, whose ‘primary’ atrial myopathy is likely to be more aggressive than those with a ‘secondary’ atrial myopathy and hence less reversible. This is further supported by previous LA focused ablation strategies in persistent AF achieving similarly poor success rates (∼40%) to our data,1,22 although there are also data with higher outcomes reported.23 Role of the RA The exact role of the RA in the pathogenesis and maintenance of persistent/LSPAF remains controversial. We found that CFE areas in the RA were not only higher pre- and post- remote LA ablation in the non-HF group but that this contributed to the residual bi-atrial CFE predicting outcome in cox regression analysis. Furthermore, the RA surface area as calculated from the geometry of the 3D electroanatomical map was also larger in this group. These findings provoke the hypothesis that the RA is at least of equal importance to the LA in patients with a ‘primary’ atrial myopathy. This is further supported by the two different patterns of AFCL prolongation seen in the RA in response to remote LA ablation. In the HF group, the parallel increase in AFCL in both atria implies that LA ablation had a significant impact first and foremost on the LA substrate, and then in turn on the RA, reflecting its likely bystander role. In contrast, in the non-HF group, the AFCL only prolonged within the LA resulting in a RA to LA frequency gradient. This divergent pattern of AFCL change suggests that the RA may well have harboured driver sources in a significant proportion of these patients. This agrees with data from Hocini et al.,9 who demonstrated that where LA ablation failed to terminate AF but resulted in an RA to LA AFCL frequency gradient, RA ablation terminated AF in 50%. Rostock et al. also found that RA ablation terminated AF in 26% of chronic AF patients and more recently, Narayan et al. used novel computational mapping techniques to demonstrate that 24% of drivers in a predominantly persistent AF population were located in the RA and that ablation of these sources resulted in excellent freedom from arrhythmia rates at 9 months.24,25 Ravelli et al.26 used another novel technique combining analysis of AFCL and fibrillatory wave analysis to show that 23% of potential AF rotors in persistent AF were found in the RA. Despite the increasing evidence that the RA is important to outcomes in a significant proportion of patients with non-paroxysmal AF, its electrical role in the HF population is not known. It is conceivable, however, that in left ventricular dysfunction, the pressure and volume overload affects the LA more than the RA, such that the LA becomes the dominant arrhythmogenic chamber of AF. Our findings from this study suggest that persistent/LSPAF without structural heart disease represents a more advanced ‘primary’ atrial myopathy. These are more complex bi-atrial substrates consisting of more extensive CFE distribution both before and after LA focussed ablation, compared with the ‘secondary’ atrial myopathy found in HF. Furthermore, the outcome data suggest that LA only ablation—albeit with a similarly extensive lesion set in both non-HF and HF groups—is not as effective in these advanced ‘primary’ atrial myopathies. By showing that associated structural heart disease does not universally imply more advanced substrates our study highlights the heterogeneous nature of AF substrates and the need to adopt more tailored, patient specific ablation strategies.25,27 Therefore, thought should be given to whether the AF is a ‘primary’ or ‘secondary’ process when counselling patients pre-procedure and to help guide the ablation strategy. Looking to the future, we need to accurately and prospectively identify which patients will benefit from bi-atrial ablation, and to incorporate this into individualized strategies for ablation to improve outcomes, particularly in non-paroxysmal AF. Limitations The main limitation of this study is the relatively small sample size, and there remains a small possibility that the outcomes could be due to chance. Another limitation is the fact that RA CFE maps were not taken in parallel with the LA CFE maps at all time points. In doing so the stepwise effect of LA ablation on RA CFE cannot be ascertained, although the overall effect of LA ablation is examined. Additionally, these two groups were studied sequentially rather than contemporaneously and this may incorporate unmeasured differences to account for the differences in clinical outcome. Conclusions Persistent/LSPAF in the absence of HF represents a more complex ‘primary’ bi-atrial myopathy than that when AF occurs in association with HF. LA step-wise ablation results in more extensive substrate modification in HF compared to non-HF and is related to better clinical outcomes. Post ablation bi-atrial total CFE burden was an independent predictor of clinical success highlighting the role of CFE as an important surrogate marker of underlying atrial substrate. Clinical Trial Registration—URL: http://www.clinicaltrials.gov. Unique identifiers: NCT01385358 and NCT00878384. Supplementary material Supplementary material is available at Europace online. Funding S.H. and D.J. received education grant support from St Jude Medical. S.H and T.W and this project were supported by National Institute of Health Research grant EME 12/127/127. Conflict of interest: none declared. References 1 Rostock T, Salukhe TV, Steven D, Drewitz I, Hoffman B, Bock K et al.   Long-term single-and multiple-procedure outcome and predictors of success after catheter ablation for persistent atrial fibrillation. Heart Rhythm  2011; 8: 1391– 7. Google Scholar CrossRef Search ADS PubMed  2 Tilz RR, Chun KR, Schmidt B, Fuernkranz A, Wissner E, Koester I et al.   Catheter ablation of long-standing persistent atrial fibrillation: a lesson from circumferential pulmonary vein isolation. J Cardiovasc Electrophysiol  2010; 21: 1085– 93. Google Scholar CrossRef Search ADS PubMed  3 Konings KT, Kirchhof CJ, Smeets JR, Wellens HJ, Penn OC, Allessie MA. High-density mapping of electrically induced atrial fibrillation in humans. Circulation  1994; 89: 1665– 80. Google Scholar CrossRef Search ADS PubMed  4 Calkins H, Kuck KH, Cappato R, Brugada J, Camm AJ, Shih-Ann C et al.   2012 HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation. Europace  2012; 14: 528– 606. Google Scholar CrossRef Search ADS PubMed  5 Jones DG, Haldar S, Jarman J, Johar S, Hussain W, Markides V et al.   Impact of stepwise ablation on the bi-atrial substrate in patients with persistent atrial fibrillation and heart failure. Circ Arrhythm Electrophysiol  2013; 6: 761– 8. Google Scholar CrossRef Search ADS PubMed  6 Matsuo S, Yamane T, Date T, Tokutake K, Hioki M, Narul R et al.   Substrate modification by pulmonary vein isolation and left atrial linear ablation in patients with persistent atrial fibrillation: *its impact on complex fractionated atrial electrograms. J Cardiovasc Electrophysiol  2012; 23: 962– 70. Google Scholar CrossRef Search ADS PubMed  7 Oral H, Chugh A, Good E, Crawford T, Sarrazin J, Kuhne M et al.   Randomized evaluation of right atrial ablation after left atrial ablation of complex fractionated atrial electrograms for long-lasting persistent atrial fibrillation. Circ Arrhythm Electrophysiol  2008; 1: 6– 13. Google Scholar CrossRef Search ADS PubMed  8 Wynn GJ, Das M, Bonnett LJ, Panikker S, Wong T, Gupta D. Efficacy of catheter ablation for persistent atrial fibrillation: a systematic review and meta-analysis of evidence from randomized and nonrandom- ized controlled trials. Circ Arrhythm Electrophysiol  2014; 7: 841– 52. doi: 10.1161/CIRCEP.114.001759. Google Scholar CrossRef Search ADS PubMed  9 Hocini M, Nault I, Wright M, Veenhuyzen G, Narayan S, Jais P et al.   Disparate evolution of right and left atrial rate during ablation of long-lasting persistent atrial fibrillation. J Am Coll Cardiol  2010; 55: 1007– 16. Google Scholar CrossRef Search ADS PubMed  10 Scherr D, Dalal D, Cheema A, Cheng A, Henrikson C, Spragg D et al.   Automated detection and characterization of complex fractionated atrial electrograms in human left atrium during atrial fibrillation. Heart Rhythm  2007; 4: 1013– 20. Google Scholar CrossRef Search ADS PubMed  11 Ciaccio EJ, Biviano AB, Whang W, Vest JA, Gambhir A, Einstein AJ et al.   Differences in repeating patterns of complex fractionated left atrial electrograms in longstanding persistent as compared with paroxysmal atrial fibrillation. Circ Arrhythm Electrophysiol  2011; 4: 470– 7. Google Scholar CrossRef Search ADS PubMed  12 Goette A, Kalman JM, Aguinaga L, Akar J, Cabrera JA, Chen SA et al.   EHRA/HRS/APHRS/SOLAECE expert consensus on atrial cardiomyopathies: definition, characterization, and clinical implication. Europace  2016; 18: 1455– 90. Google Scholar CrossRef Search ADS PubMed  13 Zhao Y, Di Biase L, Trivedi C, Mohanty S, Bai R, Mohanty P, Gianni C et al.   Importance of non-pulmonary vein triggers ablation to achieve long-term freedom from paroxysmal atrial fibrillation in patients with low ejection fraction. Heart Rhythm  2016; 13: 141– 9. Google Scholar CrossRef Search ADS PubMed  14 Di Biase L, Mohanty P, Mohanty S, Santangeli P, Trivedi C, Lakkireddy D et al.   Ablation versus amiodarone for treatment of persistent atrial fibrillation in patients with congestive heart failure and an implanted device: results from the AATAC multicenter randomized trial. Circulation  2016; 133: 1637– 44. Google Scholar CrossRef Search ADS PubMed  15 Hunter RJ, Berriman TJ, Diab I, Kamdar R, Richmond L, Baker V et al.   A randomized controlled trial of catheter ablation versus medical treatment of atrial fibrillation in heart failure (the CAMTAF trial). Circ Arrhythm Electrophysiol  2014; 7: 31– 8. doi: 10.1161/CIRCEP.113.000806. Google Scholar CrossRef Search ADS PubMed  16 Jones DG, Haldar S, Hussain W, Sharma R, Francis DP, Rahman-Haley S et al.   A randomized trial to assess catheter ablation versus rate control in the management of persistent atrial fibrillation in heart failure (ARC-HF). J Am Coll Cardiol  2013; 61: 1894– 903. Google Scholar CrossRef Search ADS PubMed  17 Anselmino M, Matta M, Castagno D, Giustetto C, Gaita F. Catheter ablation of atrial fibrillation in chronic heart failure: state-of-the-art and future perspectives. Europace  2016; 18: 638– 47. Google Scholar CrossRef Search ADS PubMed  18 Ullah W, Ling LH, Prabhu S, Lee G, Kistler P, Finlay MC et al.   Catheter ablation of atrial fibrillation in patients with heart failure: impact of maintaining sinus rhythm on heart failure status and long-term rates of stroke and death. Europace  2016; 18: 679– 86. Google Scholar CrossRef Search ADS PubMed  19 Morton JB, Sanders P, Vohra JK, Sparks P, Morgan J, Spence S et al.   The effect of chronic atrial stretch on atrial electrical remodeling in patients with an atrial septal defect. Circulation  2003; 107: 1775– 82. Google Scholar CrossRef Search ADS PubMed  20 Roberts-Thomson KC, Stevenson I, Kistler PM, Haqqani HM, Spence SJ, Goldblatt JC et al.   The role of chronic atrial stretch and atrial fibrillation on posterior left atrial wall conduction. Heart Rhythm  2009; 6: 1109– 17. Google Scholar CrossRef Search ADS PubMed  21 Cox JL. The longstanding, persistent confusion surrounding surgery for atrial fibrillation. J Thorac Cardiovasc Surg  2010; 139: 1374– 86. Google Scholar CrossRef Search ADS PubMed  22 McCready JW, Smedley T, Lambiase PD, Ahsan SY, Segal OR, Rowland E et al.   Predictors of recurrence following radiofrequency ablation for persistent atrial fibrillation. Europace  2011; 13: 355– 61. Google Scholar CrossRef Search ADS PubMed  23 Elayi CS, Di Biase L, Barrett CChing CK, al Aly M, Lucciola M, Bai R et al.   Atrial fibrillation termination as a procedural endpoint during ablation in long-standing persistent atrial fibrillation. Heart Rhythm  2010; 7: 1216– 23. Google Scholar CrossRef Search ADS PubMed  24 Rostock T, Steven D, Hoffmann B, Servatius H, Drewitz I, Sydow K et al.   Chronic atrial fibrillation is a biatrial arrhythmia: data from catheter ablation of chronic atrial fibrillation aiming arrhythmia termination using a sequential ablation approach. Circ Arrhythm Electrophysiol  2008; 1: 344– 53. Google Scholar CrossRef Search ADS PubMed  25 Narayan S, Krummen D, Shivkumar K, Clopton P, Wouter-Jan R, Miller J. Treatment of atrial fibrillation by the ablation of localized sources: CONFIRM trial. J Am Coll Cardiol  2012; 60: 628– 36. Google Scholar CrossRef Search ADS PubMed  26 Ravelli F, Masè M, Cristoforetti A, Del Greco M, Centonze M, Marini M et al.   Anatomic localization of rapid repetitive sources in persistent atrial fibrillation: fusion of biatrial CT images with wave similarity/cycle length maps. JACC Cardiovasc Imaging  2012; 5: 1211– 20. Google Scholar CrossRef Search ADS PubMed  27 Haissaguerre M, Hocini M, Denis A, Shah A, Komatsu Y, Yamashita S et al.   Driver domains in persistent atrial fibrillation. Circulation  2014; 130: 530– 8. Google Scholar CrossRef Search ADS PubMed  Published on behalf of the European Society of Cardiology. All rights reserved. © The Author 2017. For permissions, please email: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Europace Oxford University Press

Characterising the difference in electrophysiological substrate and outcomes between heart failure and non-heart failure patients with persistent atrial fibrillation

Loading next page...
 
/lp/ou_press/characterising-the-difference-in-electrophysiological-substrate-and-Asx1CEChTQ
Publisher
Oxford University Press
Copyright
Published on behalf of the European Society of Cardiology. All rights reserved. © The Author 2017. For permissions, please email: journals.permissions@oup.com.
ISSN
1099-5129
eISSN
1532-2092
D.O.I.
10.1093/europace/euw380
Publisher site
See Article on Publisher Site

Abstract

Abstract Aims Characterizing the differences in substrate and clinical outcome between heart failure (HF) and non-heart failure (non-HF) patients undergoing persistent atrial fibrillation (AF) ablation. Methods and results Using complex fractionated electrograms (CFE) as a surrogate marker of substrate complexity, we compared the bi-atrial substrate in patients with persistent AF with and without HF, at baseline and after ablation, to determine its impact on clinical outcome. In this retrospective analysis of two prospective studies, 60 patients underwent de-novo step-wise left atrial (LA) ablation, 30 with normal left ventricular ejection fraction (LVEF) ≥ 50% (non-HF group) and 30 with LVEF ≤ 35% (HF group). Multiple high-density bi-atrial CFE maps were acquired along with AF cycle length (AFCL) at each procedural stage. Change in bi-atrial CFE areas, AFCL and outcome data were then compared. In the non-HF group, higher CFE-areas were found at baseline and at each step of the procedure in the LA. In both LA and the right atrium (RA), baseline and final CFE area were also higher in the non-HF group. Single procedure, arrhythmia-free survival at 1 year was higher in the HF group compared with the non-HF group (72% vs. 43%, log rank P = 0.04). Final total bi-atrial CFE area was an independent predictor of arrhythmia recurrence. Conclusions CFE represents an important surrogate marker of atrial substrate complexity. The atrial substrate in persistent AF differs between HF and non-HF with the latter representing a more complex ‘primary’ bi-atrial myopathy. LA focussed ablation results in more extensive substrate modification in HF and better clinical outcomes as compared with non-HF. Catheter ablation, Atrial Fibrillation, Heart Failure, Persistent What’s new? There are important differences in electrophysiological substrate in persistent/LSPAF patients with and without HF. Persistent/LSPAF without HF (or other structural heart disease) can represent a more advanced ‘primary’ atrial myopathy with higher baseline CFE burden in both LA and RA, as compared with AF associated with HF. An LA focussed ablation strategy resulted in better clinical outcomes in the HF group. Residual bi-atrial CFE burden post-LA stepwise ablation was an independent predictor of clinical success. CFE may be more useful as a surrogate marker of atrial substrate complexity than a primary target of ablation. Introduction Catheter ablation for persistent and long-standing persistent AF (LSPAF) remains one of the most challenging areas of clinical electrophysiology. Clinical outcomes remain suboptimal, despite significant technical and technological advances and the use of adjunct ablation strategies to pulmonary vein isolation (PVI).1,2 This challenge reflects the complex pathological milieu that underlies persistent AF substrates which may be further complicated by different disease states. Complex fractioned atrial electrograms (CFEs) are thought to reflect areas of slow conduction, localized re-entry or pivot points for change in wavelet orientation. Although they may not represent target sites for AF ablation, they are likely to represent surrogate markers of substrate complexity that may differ in different underlying disease state that maintain AF.3,4 This study was designed to evaluate the hypothesis that the underlying atrial substrate and response to conventional stepwise LA ablation is different in persistent and LSPAF patients with and without HF. Methods Study population This is a retrospective analysis of data collected from two prospective studies where patients received the same ablation lesion sets using the same mapping and ablation technologies. Sixty patients in total underwent de novo CA for symptomatic persistent/LSPAF. Thirty patients with normal LVEF ≥ 50% (non-HF) recruited from a prospective clinical trial (NCT01385358) and compared with 30 HF patients with an LVEF ≤ 35% (HF) from a previous prospective study (NCT00878384).5 All patients gave written informed consent for the study, approved by the local ethics committee in accordance with the Declaration of Helsinki. Ablation and mapping protocol Under general anaesthesia, extensive bi-atrial CFE mapping was performed using an electroanatomical mapping system (EnSite VelocityTM, St Jude Medical, MN, USA) and a multi-electrode high density (AFocus IITM, St Jude Medical) mapping catheter. The step-wise ablation protocol has been previously described and consists of PVI, linear lesions (LL) (roof and mitral isthmus) and LA CFE ablation using non-contact force 3.5 mm irrigated-tip D-F SF ablation catheters (Thermocool Navistar, Biosense Webster).5 Bi-atrial CFE maps were acquired at baseline and post CFE ablation, with additional LA only CFE maps after PVI and LL (Figure 1). At each stage, the right (RAA) and left atrial appendage (LAA) cycle length (CL) was recorded. Figure 1 View largeDownload slide Procedural protocols for non-HF and HF groups—sequence of treatment and mapping. Protocol summary is shown in the diagram. The number of maps at each stage is shown in grey boxes. The ablation and DC cardioversion stages are shown in white boxes. Number of AF terminations is shown at each stage of procedure. Technical software failure resulted in the loss of one baseline, two final RA maps, and three final LA maps in the non-HF group. RA, right atrium; LA, left atrium; PVI, pulmonary vein isolation; CFE, complex fractionated electrograms; LL, linear lesions. Figure 1 View largeDownload slide Procedural protocols for non-HF and HF groups—sequence of treatment and mapping. Protocol summary is shown in the diagram. The number of maps at each stage is shown in grey boxes. The ablation and DC cardioversion stages are shown in white boxes. Number of AF terminations is shown at each stage of procedure. Technical software failure resulted in the loss of one baseline, two final RA maps, and three final LA maps in the non-HF group. RA, right atrium; LA, left atrium; PVI, pulmonary vein isolation; CFE, complex fractionated electrograms; LL, linear lesions. All CFE sites marked in the post LL maps were ablated. If AF persisted after LA CFE ablation, direct current cardioversion was used to restore sinus rhythm (SR). PVI and LL bi-directional block were then assessed using conventional techniques with incomplete lines re-ablated to achieve bi-directional block. Cavo-tricuspid isthmus ablation was only performed if there was documented typical atrial flutter. If AF terminated to atrial tachycardia (AT), the protocol was terminated and the AT mapped and ablated. CFE mapping Using the roving high-density mapping catheter, signals were recorded from all 19 bipoles over 5 s per acquisition until the whole endocardial surface was covered. The mitral/tricuspid annuli were defined by electrogram characteristics and the contained area excluded. The NavX CFE mean tool was used to identify high-frequency atrial signals with multiple components with the following settings: electrogram width <10 ms (to remove far field signals), refractory period 30 ms (values below this being regarded as nonphysiological for local reactivation), and interpolation and surface projection <10 mm based on previous studies. Voltage detection threshold was adjusted to exclude background noise and avoid false-positive CFE annotation and fixed for subsequent maps. Points >10 mm from the surface, and those displaying electric interference, were deleted. Scar was defined as <0.05 mV. CFE mean was defined as the mean time between consecutive deflections during a 5-s recording period. CFE in this and the previous study was defined as sites with CFE mean ≤120 ms.5 Analysis of CFE data As previously described, each atrium was segmented into three regions to categorise distribution of CFE.5 The LA regions were anterior, posterior and appendage. The RA regions were lateral, septal and appendage. All contiguous zones of CFE within each segment of LA and RA were delineated and the area enclosed defined as CFE-area. The LA CFE-area, however, excluded the PVI encircling lesions, linear lesions (within 5 mm), and the mitral valve annulus. Follow-up The primary study endpoint was freedom from recurrent AF or atrial tachyarrhythmias (AT) ≥30 s after a single procedure and off AADs after the 3-month blanking period. Class I and III antiarrhythmic drugs (AADs) were discontinued before or at ablation. Patients were followed up for a minimum of 48 h–7 days ambulatory monitoring at 3, 6, and 12 months. Reported symptoms outside these time points were assessed with 12-lead ECG and further ambulatory ECG as indicated.4 Statistical analysis Continuous data are presented as mean (SD) or median (IQR), and categorical data as number and percentage. Paired t-tests (within group) and unpaired t-tests (between groups) were used to analyse change in CFE-area (absolute and percentage coverage of atrial surface area) and AFCL with Bonferroni correction to account for multiple comparisons. Arrhythmia-free survival was analysed by Kaplan–Meier survival curves with log-rank comparisons using the Mantel–Cox test. Cox regression was used to assess predictors of arrhythmia-free survival and only variables that had P < 0.10 were included in the multivariable model. P values <0.05 were considered statistically significant. Data were analysed using GraphPad Prism version 6.00 for Mac (GraphPad Software, San Diego, CA, USA) and R statistical software (version 3.1.2). Results Baseline characteristics are shown in Table 1. Mean duration of continuous AF was greater than 12 months in both groups. In the HF cohort, 40% were ischaemic in aetiology. The LA diameter on transthoracic echocardiography (TTE) was smaller (although similar surface area on electroanatomical map) and hypertension was more prevalent in the non-HF group. Table 1 Baseline characteristics (N = 60)   Non-HF  HF  P-value  Age (y), mean ± SD  66 ± 9  63 ± 9  0.09  Sex (male), n (%)  23 (48)  25 (52)  0.51  Duration of continuous AF (months), mean ± SD  19 ± 4  24 ± 8  0.28  LA size, mean ± SD  44 (5)  50 (7)  <0.001  LVEF (%), mean ± SD  60 ± 8  25 ± 11  <0.001  Hypertension, n (%)  21 (72)  10 (33)  0.004  Coronary artery disease, n (%)  9 (31)  12 (40)  0.59  Diabetes Mellitus, n (%)  1 (3)  7 (23)  0.05  Beta-blocker, n (%)  18 (72)  28 (93)  0.06  CHA2DS2-VASC, mean ± SD  1.9 ± 1.1  2.7 ± 1.5  0.03    Non-HF  HF  P-value  Age (y), mean ± SD  66 ± 9  63 ± 9  0.09  Sex (male), n (%)  23 (48)  25 (52)  0.51  Duration of continuous AF (months), mean ± SD  19 ± 4  24 ± 8  0.28  LA size, mean ± SD  44 (5)  50 (7)  <0.001  LVEF (%), mean ± SD  60 ± 8  25 ± 11  <0.001  Hypertension, n (%)  21 (72)  10 (33)  0.004  Coronary artery disease, n (%)  9 (31)  12 (40)  0.59  Diabetes Mellitus, n (%)  1 (3)  7 (23)  0.05  Beta-blocker, n (%)  18 (72)  28 (93)  0.06  CHA2DS2-VASC, mean ± SD  1.9 ± 1.1  2.7 ± 1.5  0.03  n = 30. AF, atrial fibrillation; LA, left atrial; LVEF, left ventricular ejection fraction. Procedural and mapping data Total procedural time (271 ± 55 vs. 331 ± 55 mins; P < 0.001), fluoroscopy time (52 ± 16 vs. 79 ± 18 min; P < 0.001), total RF time (58 ± 15 vs. 82 ± 19 min; P < 0.001 and PVI RF time (28 ± 10 vs. 46 ± 17 min; P < 0.001) were all shorter in the non-HF group (Table 2). Number of points per map in the LA (475 ± 191 vs. 479 ± 99; P = 0.92) and RA (410 ± 161 vs. 373 ± 96; P = 0.14) did not differ between the non-HF and HF groups, respectively. Total LA surface area (210 ± 38 vs. 213 ± 43; P = 0.75) and LA CFE area analysis zone (114 ± 30 vs. 117 ± 24; P = 0.62) were similar but the total RA surface area (178 ± 41 vs. 136 ± 33; P < 0.001) was higher in the non-HF group. Termination of AF during ablation (termination to SR without DC cardioversion) was observed in 11 cases, 5 (2 during roof line and 3 via AT during CFE ablation) in the non-HF group and 6 (4 via AT and 2 direct to SR during CFE ablation) in the HF group (Figure 2) . Table 2 Comparison of procedural and mapping data between non-HF and HF groups   Non-HF  HF  P value  Total procedural time (min)  271 (±55)  331 (±55)  0.001  Fluoroscopy time (min)  52 (±16)  79 (±18)  <0.001  Total RF time (min)  58 (±15)  82 (±19)  <0.001  PVI RF time (min)  28 (±10)  46 (±17)  <0.001  Roof line RF time (min)  3.4 (±1.4)  3.3 (±1.7)  0.97  Mitral isthmus line RF time (min)  4.9 (±2.4)  4.0 (±2.4)  0.16  CFE RF time (min)  9.4 (±7.0)  12.1 (±7.7)  0.17  LA points per map  475 (±191)  479 (±99)  0.92  RA points per map  410 (±161)  373 (±96)  0.14  LA CFE area analysis zone (cm2)  114 (±30)  117 (±24)  0.62  Total LA surface area (cm2)  210 (±38)  213 (±43)  0.75  Total RA surface area (cm2)  178 (±41)  136 (±33)  <0.001  Unblocked linear lesions  0/30  3/30  0.07    Non-HF  HF  P value  Total procedural time (min)  271 (±55)  331 (±55)  0.001  Fluoroscopy time (min)  52 (±16)  79 (±18)  <0.001  Total RF time (min)  58 (±15)  82 (±19)  <0.001  PVI RF time (min)  28 (±10)  46 (±17)  <0.001  Roof line RF time (min)  3.4 (±1.4)  3.3 (±1.7)  0.97  Mitral isthmus line RF time (min)  4.9 (±2.4)  4.0 (±2.4)  0.16  CFE RF time (min)  9.4 (±7.0)  12.1 (±7.7)  0.17  LA points per map  475 (±191)  479 (±99)  0.92  RA points per map  410 (±161)  373 (±96)  0.14  LA CFE area analysis zone (cm2)  114 (±30)  117 (±24)  0.62  Total LA surface area (cm2)  210 (±38)  213 (±43)  0.75  Total RA surface area (cm2)  178 (±41)  136 (±33)  <0.001  Unblocked linear lesions  0/30  3/30  0.07  Figure 2 View largeDownload slide CFE-mean (CFE mean ≤120 ms) map with surface markings of CFE distribution. Posterior aspect of LA geometry with CFE-mean map. The map shown is after PVI ablation; however, offline surface annotation was used to annotate CFE areas with colour coded annotation schema at each stage of the procedure: baseline (red), post-PVI (amber), post-linear lesions (green), final post CFE ablation (blue). The majority of these annotations are left projected onto the map surface for illustrative purposes. Figure 2 View largeDownload slide CFE-mean (CFE mean ≤120 ms) map with surface markings of CFE distribution. Posterior aspect of LA geometry with CFE-mean map. The map shown is after PVI ablation; however, offline surface annotation was used to annotate CFE areas with colour coded annotation schema at each stage of the procedure: baseline (red), post-PVI (amber), post-linear lesions (green), final post CFE ablation (blue). The majority of these annotations are left projected onto the map surface for illustrative purposes. Impact of ablation on remote LA CFE area (intragroup analysis) The impact of ablation upon the CFE area for each group are described below. Results are also shown as percentage of analysed LA surface (Figure 3). More detailed data for the constituent segments of each atrium are shown in Supplementary material online, Table 1. In the non-HF group, a sequential reduction of the LA CFE area was observed from baseline 27.6 ± 11.8 cm2 (24.4 ± 10.2% of analysed LA surface) to post-PVI 21.6 ± 12.5 cm2 (18.6 ± 10.4%, P = 0.002 vs. baseline), and after the addition of LL to 16.5 ± 9.8 cm2 (14.8 ± 9.5%, P = 0.008 vs. post-PVI). Direct CFE ablation further reduced final LA CFE-area, compared with post-LL analysis, to 7.8 ± 8.7 cm2 (6.7 ± 6.6%, P = 0.001 vs. post-LL). Figure 3 View largeDownload slide Impact of stepwise LA ablation on LA CFE and RA CFE area in non-HF and HF groups. Panel A—graphs show the percentage (mean ± SD) coverage of CFE of the atrial surface (after exclusion of PVI and LL), within the segmented LA at baseline and after each stage of ablation. Panel B—graphs show the percentage (mean ± SD) coverage of CFE of the atrial surface within the segmented RA at baseline (pre) and after the completion of all the steps of LA ablation (post) The first graph is of the Non-HF group and the second the HF group allowing for visual comparison of CFE quantification between the two groups. P values are shown for comparisons between steps of ablation and denoted * <0.05, ** <0.01, and *** <0.001. Figure 3 View largeDownload slide Impact of stepwise LA ablation on LA CFE and RA CFE area in non-HF and HF groups. Panel A—graphs show the percentage (mean ± SD) coverage of CFE of the atrial surface (after exclusion of PVI and LL), within the segmented LA at baseline and after each stage of ablation. Panel B—graphs show the percentage (mean ± SD) coverage of CFE of the atrial surface within the segmented RA at baseline (pre) and after the completion of all the steps of LA ablation (post) The first graph is of the Non-HF group and the second the HF group allowing for visual comparison of CFE quantification between the two groups. P values are shown for comparisons between steps of ablation and denoted * <0.05, ** <0.01, and *** <0.001. In the HF group, a similar trend was seen with sequential reduction of the LA CFE area from baseline 18.3 ± 12.0 cm2 (16.2 ± 10.6% of the analyzed LA surface) to post-PVI 10.2 ± 7.1 cm2 (9.0 ± 6.6%; P < 0.001 vs. baseline) and after the addition of LL and compared with post-PVI analysis, total LA CFE area reduced to 7.7 ± 6.5 cm2 (6.9 ± 5.9%; P = 0.006). Direct CFE ablation further reduced final LA CFE area, compared with post-LL ablation analysis, to 3.1 ± 3.5 cm2 (2.8 ± 3.0%; P = 0.002). Impact of ablation on remote RA CFE area (intragroup analysis) In the non-HF group, LA ablation also resulted in reduction of the RA CFE area, from baseline 39.8 ± 19.3 cm2 (22.1 ± 9.2% of the RA surface) to 24.1 ± 12.9 cm2 (13.3 ± 8.1%, P = 0.0001 vs. baseline) in the final RA CFE map (see Supplementary material online, Table 1). Similarly, in the HF group, baseline RA CFE area was reduced from 25.9 ± 14.1 cm2 (19.2 ± 10.3% of the total RA surface) to 12.9 ± 11.8 cm2 (9.9 ± 7.8%; P < 0.001) in the final RA CFE map. Results are also shown as the percentage of analysed RA surface (Figure 3). Comparing the impact of ablation on remote CFE-area between groups In the LA, CFE areas in the non-HF patients were all higher than in the HF group at each step of the procedure (see Supplementary material online, Table 1): baseline total CFE area (27.6 ± 11.8 vs. 18.3 ± 12.0 cm2; P = 0.004); post-PVI total CFE area (21.6 ± 12.5 vs. 10.2 ± 7.1 cm2; P = 0.001), post-LL total CFE area (16.5 ± 9.8 vs. 7.7 ± 6.5 cm2; P < 0.001), and final total CFE area (7.8 ± 8.7 vs. 3.1 ± 3.5 cm2; P = 0.018). Similarly, in the RA, the baseline total RA CFE area (39.8 ± 19.3 vs. 25.9 ± 14.1 cm2; P = 0.002) and final total RA CFE area (24.1 ± 12.9 vs. 12.9 ± 11.8 cm2; P = 0.003) were also higher in the non-HF group. Impact of ablation upon AFCL The impact of LA ablation on bi-atrial AFCL exhibited two different patterns of change (Table 3). In the HF group, both the LA (161 ± 27 to 180 ± 42; P = 0.004) and RA (167 ± 33 to 178 ± 40; P < 0.001) AFCL prolonged in parallel, whereas in the non-HF group, only the LA (153 ± 18 to 173 ± 29; P < 0.0001) AFCL prolonged resulting in a divergent AFCL pattern. Table 3 Overall change in AFCL pre- and post-LA ablation with intragroup analysis for both non-HF and HF groups   NON-HF   HF     AFCL (ms)  P value  AFCL (ms)  P value  RA baseline  163±18  0.26  167±32  0.004  RA final  168±24  178±40  LA baseline  153±18  <0.0001  161±28  <0.0001  LA final  173±29  180±42    NON-HF   HF     AFCL (ms)  P value  AFCL (ms)  P value  RA baseline  163±18  0.26  167±32  0.004  RA final  168±24  178±40  LA baseline  153±18  <0.0001  161±28  <0.0001  LA final  173±29  180±42  Clinical outcome—single procedure arrhythmia-free success Single procedure arrhythmia-free survival at 1 year, off AADs, was 72% for the HF group vs. 43% for the non-HF group (log rank, P = 0.04, Figure 4). The final total bi-atrial CFE area was the only independent predictor of arrhythmia recurrence in multivariable analysis (Table 4). In the non-HF group, one patient had a primary intra-cerebral event in the context of a supra-therapeutic INR level within the blanking period. In the HF group, two patients died from progressive HF at 1 and 11 months. Table 4 Cox regression analysis model for arrhythmia recurrence after a single ablation procedure n = 60   Univariable analysis   Multivariable analysis     HR (95% CI)  P value Valuevalue  HR (95% CI)  P value  Male  0.50 (0.22–1.14)  0.10      Age/yr  0.99 (0.95–1.03)  0.64      LA size/5mm  0.96 (0.81–1.15)  0.67      LV ejection fraction/5%  1.41 (0.96–2.09)  0.08  1.01 (0.61–1.69)  0.96  AF, duration/months  0.99 (0.97–1.02)  0.76      Baseline LA CFE area/cm2  1.02 (0.99–1.05)  0.25      Baseline RA CFE area/cm2  1.01 (0.99–1.04)  0.22      Baseline total bi-atrial CFE area/cm2  1.01 (1.00–1.02)  0.20      Final total bi-atrial CFE area/cm2  1.03 (1.01–1.06)  0.001  1.03 (1.01–1.06)  0.01  Reduction in total bi-atrial CFE area/cm2  0.99 (0.98–1.01)  0.43      Total RF, duration/10 min  0.82 (0.68–1.00)  0.05  0.87 (0.69–1.09)  0.23    Univariable analysis   Multivariable analysis     HR (95% CI)  P value Valuevalue  HR (95% CI)  P value  Male  0.50 (0.22–1.14)  0.10      Age/yr  0.99 (0.95–1.03)  0.64      LA size/5mm  0.96 (0.81–1.15)  0.67      LV ejection fraction/5%  1.41 (0.96–2.09)  0.08  1.01 (0.61–1.69)  0.96  AF, duration/months  0.99 (0.97–1.02)  0.76      Baseline LA CFE area/cm2  1.02 (0.99–1.05)  0.25      Baseline RA CFE area/cm2  1.01 (0.99–1.04)  0.22      Baseline total bi-atrial CFE area/cm2  1.01 (1.00–1.02)  0.20      Final total bi-atrial CFE area/cm2  1.03 (1.01–1.06)  0.001  1.03 (1.01–1.06)  0.01  Reduction in total bi-atrial CFE area/cm2  0.99 (0.98–1.01)  0.43      Total RF, duration/10 min  0.82 (0.68–1.00)  0.05  0.87 (0.69–1.09)  0.23  HR, hazard ratio; CI, confidence interval; AF, atrial fibrillation; CFE, complex fractionated electrogram; LA, left atrial; LV, left ventricular; RA, right atrial; RF, radiofrequency ablation. Figure 4 View largeDownload slide Freedom from atrial arrhythmias (single procedure) at 12 months. Non-HF vs. HF groups, single procedure Kaplan–Meier curves of atrial arrhythmia-free survival off AAD (censored at 365 days). Success is defined as freedom from atrial arrhythmia after a 3-month blanking period. The two survival curves were compared using the log-rank (Mantel–Cox) test. Figure 4 View largeDownload slide Freedom from atrial arrhythmias (single procedure) at 12 months. Non-HF vs. HF groups, single procedure Kaplan–Meier curves of atrial arrhythmia-free survival off AAD (censored at 365 days). Success is defined as freedom from atrial arrhythmia after a 3-month blanking period. The two survival curves were compared using the log-rank (Mantel–Cox) test. Discussion The principle findings in this study are that in persistent/LSPAF patients: The baseline atrial electrophysiological substrate is different between the HF and non-HF group with the latter exhibiting higher total CFE areas in both the LA and RA. LA step-wise ablation results in more extensive substrate modification in the HF group with lower final CFE areas in both LA and RA which can be related to significantly higher single-procedure arrhythmia-free survival at 12 months (72% vs. 43%, log rank P=0.04) in the HF group as compared with the non-HF group. The residual bi-atrial CFE area post LA focussed stepwise ablation was an independent predictor of arrhythmia recurrence. CFE burden In addition to the above findings, our data show the sequential reduction in remote CFE areas bi-atrially after PVI and LL ablation in both groups, except of the LAA in the non-HF group (Figure 3). This is in line with previous data and substantiates previous claims that many CFE represent incidental or passive activation rather than source activity.5,6 What is new is the fact that the residual bi-atrial CFE area was shown to be important in predicting clinical success. This residual CFE area in our study comprises not only CFE in the LAA and the RA, which are areas that have not been ablated, but also ‘new’ CFE in the LA despite targeted ablation of all CFE sites post-PVI and LL lesions. Our protocol did not include a further stage to repeatedly target the final CFE areas which is the technique used by some centres when pursuing the endpoint of AF termination but it does raise the question as to whether these residual CFE are a marker for arrhythmia continuation without being an ablation target. Given the recent negative randomised clinical data showing no incremental benefits in ablating CFE,7–9 the remote effects on CFE post PVI and LL,5,6 and the more extensive distribution of CFE in persistent AF as compared to paroxysmal AF,10,11 one can conclude that CFE may actually be a dynamic surrogate marker of the atrial electrophysiological substrate rather than represent an anatomical target for ablation. Atrial substrates in different disease states Intuitively AF substrates would be expected to differ in different pathological states.12 In patients with paroxysmal AF and heart failure, recent data showed improved outcomes when non-PV trigger ablation was undertaken in addition to PVI.13 Catheter ablation using extensive substrate modification strategies have also shown good outcomes, in persistent AF and heart failure, with 70-81% freedom from arrhythmia rates which is consistent with our HF group success rates.14–18 In a human study of chronic atrial stretch and volume overload secondary to ASD, Morton et al. showed greater electrical remodelling in these patients when compared with normal.19 In left ventricular dysfunction, Roberts-Thomson et al.20 demonstrated greater conduction delay, heterogeneity, and anisotropy, compared with normal. Our data goes rather than go a step further to show the electrophysiological manifestation of these remodelling changes relevant to catheter ablation but interestingly shows that persistent/LSPAF patients harbour more abnormal and complex electrical substrates in the absence of heart failure. These findings at first seem to be at odds with the widely accepted notion that structural heart disease confers additional complexity to the atrial substrate. However, the important distinction to be made is whether the arrhythmia is the primary disturbance or secondary to another cause. James Cox in his field of AF surgery previously proposed a different classification for AF based on underlying cause.21 ‘Primary’ AF was described as those without another cardiac co-morbidity serious enough to warrant concomitant surgical intervention, and ‘secondary’ AF as due to another left heart condition (e.g. ischaemia, heart failure and valvular heart disease). This was based on the finding that ‘primary’ non-paroxysmal AF harboured a very abnormal bi-atrial substrate, which mandated bi-atrial surgical lesions to achieve similar efficacy to that of secondary AF treatable by surgical procedures confined to the LA.21 This concept is strongly supported by our data and suggests that more importance should be given to the bi-atrial aspect of the substrate in patients with persistent/LSPAF in the absence of structural heart disease, whose ‘primary’ atrial myopathy is likely to be more aggressive than those with a ‘secondary’ atrial myopathy and hence less reversible. This is further supported by previous LA focused ablation strategies in persistent AF achieving similarly poor success rates (∼40%) to our data,1,22 although there are also data with higher outcomes reported.23 Role of the RA The exact role of the RA in the pathogenesis and maintenance of persistent/LSPAF remains controversial. We found that CFE areas in the RA were not only higher pre- and post- remote LA ablation in the non-HF group but that this contributed to the residual bi-atrial CFE predicting outcome in cox regression analysis. Furthermore, the RA surface area as calculated from the geometry of the 3D electroanatomical map was also larger in this group. These findings provoke the hypothesis that the RA is at least of equal importance to the LA in patients with a ‘primary’ atrial myopathy. This is further supported by the two different patterns of AFCL prolongation seen in the RA in response to remote LA ablation. In the HF group, the parallel increase in AFCL in both atria implies that LA ablation had a significant impact first and foremost on the LA substrate, and then in turn on the RA, reflecting its likely bystander role. In contrast, in the non-HF group, the AFCL only prolonged within the LA resulting in a RA to LA frequency gradient. This divergent pattern of AFCL change suggests that the RA may well have harboured driver sources in a significant proportion of these patients. This agrees with data from Hocini et al.,9 who demonstrated that where LA ablation failed to terminate AF but resulted in an RA to LA AFCL frequency gradient, RA ablation terminated AF in 50%. Rostock et al. also found that RA ablation terminated AF in 26% of chronic AF patients and more recently, Narayan et al. used novel computational mapping techniques to demonstrate that 24% of drivers in a predominantly persistent AF population were located in the RA and that ablation of these sources resulted in excellent freedom from arrhythmia rates at 9 months.24,25 Ravelli et al.26 used another novel technique combining analysis of AFCL and fibrillatory wave analysis to show that 23% of potential AF rotors in persistent AF were found in the RA. Despite the increasing evidence that the RA is important to outcomes in a significant proportion of patients with non-paroxysmal AF, its electrical role in the HF population is not known. It is conceivable, however, that in left ventricular dysfunction, the pressure and volume overload affects the LA more than the RA, such that the LA becomes the dominant arrhythmogenic chamber of AF. Our findings from this study suggest that persistent/LSPAF without structural heart disease represents a more advanced ‘primary’ atrial myopathy. These are more complex bi-atrial substrates consisting of more extensive CFE distribution both before and after LA focussed ablation, compared with the ‘secondary’ atrial myopathy found in HF. Furthermore, the outcome data suggest that LA only ablation—albeit with a similarly extensive lesion set in both non-HF and HF groups—is not as effective in these advanced ‘primary’ atrial myopathies. By showing that associated structural heart disease does not universally imply more advanced substrates our study highlights the heterogeneous nature of AF substrates and the need to adopt more tailored, patient specific ablation strategies.25,27 Therefore, thought should be given to whether the AF is a ‘primary’ or ‘secondary’ process when counselling patients pre-procedure and to help guide the ablation strategy. Looking to the future, we need to accurately and prospectively identify which patients will benefit from bi-atrial ablation, and to incorporate this into individualized strategies for ablation to improve outcomes, particularly in non-paroxysmal AF. Limitations The main limitation of this study is the relatively small sample size, and there remains a small possibility that the outcomes could be due to chance. Another limitation is the fact that RA CFE maps were not taken in parallel with the LA CFE maps at all time points. In doing so the stepwise effect of LA ablation on RA CFE cannot be ascertained, although the overall effect of LA ablation is examined. Additionally, these two groups were studied sequentially rather than contemporaneously and this may incorporate unmeasured differences to account for the differences in clinical outcome. Conclusions Persistent/LSPAF in the absence of HF represents a more complex ‘primary’ bi-atrial myopathy than that when AF occurs in association with HF. LA step-wise ablation results in more extensive substrate modification in HF compared to non-HF and is related to better clinical outcomes. Post ablation bi-atrial total CFE burden was an independent predictor of clinical success highlighting the role of CFE as an important surrogate marker of underlying atrial substrate. Clinical Trial Registration—URL: http://www.clinicaltrials.gov. Unique identifiers: NCT01385358 and NCT00878384. Supplementary material Supplementary material is available at Europace online. Funding S.H. and D.J. received education grant support from St Jude Medical. S.H and T.W and this project were supported by National Institute of Health Research grant EME 12/127/127. Conflict of interest: none declared. References 1 Rostock T, Salukhe TV, Steven D, Drewitz I, Hoffman B, Bock K et al.   Long-term single-and multiple-procedure outcome and predictors of success after catheter ablation for persistent atrial fibrillation. Heart Rhythm  2011; 8: 1391– 7. Google Scholar CrossRef Search ADS PubMed  2 Tilz RR, Chun KR, Schmidt B, Fuernkranz A, Wissner E, Koester I et al.   Catheter ablation of long-standing persistent atrial fibrillation: a lesson from circumferential pulmonary vein isolation. J Cardiovasc Electrophysiol  2010; 21: 1085– 93. Google Scholar CrossRef Search ADS PubMed  3 Konings KT, Kirchhof CJ, Smeets JR, Wellens HJ, Penn OC, Allessie MA. High-density mapping of electrically induced atrial fibrillation in humans. Circulation  1994; 89: 1665– 80. Google Scholar CrossRef Search ADS PubMed  4 Calkins H, Kuck KH, Cappato R, Brugada J, Camm AJ, Shih-Ann C et al.   2012 HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation. Europace  2012; 14: 528– 606. Google Scholar CrossRef Search ADS PubMed  5 Jones DG, Haldar S, Jarman J, Johar S, Hussain W, Markides V et al.   Impact of stepwise ablation on the bi-atrial substrate in patients with persistent atrial fibrillation and heart failure. Circ Arrhythm Electrophysiol  2013; 6: 761– 8. Google Scholar CrossRef Search ADS PubMed  6 Matsuo S, Yamane T, Date T, Tokutake K, Hioki M, Narul R et al.   Substrate modification by pulmonary vein isolation and left atrial linear ablation in patients with persistent atrial fibrillation: *its impact on complex fractionated atrial electrograms. J Cardiovasc Electrophysiol  2012; 23: 962– 70. Google Scholar CrossRef Search ADS PubMed  7 Oral H, Chugh A, Good E, Crawford T, Sarrazin J, Kuhne M et al.   Randomized evaluation of right atrial ablation after left atrial ablation of complex fractionated atrial electrograms for long-lasting persistent atrial fibrillation. Circ Arrhythm Electrophysiol  2008; 1: 6– 13. Google Scholar CrossRef Search ADS PubMed  8 Wynn GJ, Das M, Bonnett LJ, Panikker S, Wong T, Gupta D. Efficacy of catheter ablation for persistent atrial fibrillation: a systematic review and meta-analysis of evidence from randomized and nonrandom- ized controlled trials. Circ Arrhythm Electrophysiol  2014; 7: 841– 52. doi: 10.1161/CIRCEP.114.001759. Google Scholar CrossRef Search ADS PubMed  9 Hocini M, Nault I, Wright M, Veenhuyzen G, Narayan S, Jais P et al.   Disparate evolution of right and left atrial rate during ablation of long-lasting persistent atrial fibrillation. J Am Coll Cardiol  2010; 55: 1007– 16. Google Scholar CrossRef Search ADS PubMed  10 Scherr D, Dalal D, Cheema A, Cheng A, Henrikson C, Spragg D et al.   Automated detection and characterization of complex fractionated atrial electrograms in human left atrium during atrial fibrillation. Heart Rhythm  2007; 4: 1013– 20. Google Scholar CrossRef Search ADS PubMed  11 Ciaccio EJ, Biviano AB, Whang W, Vest JA, Gambhir A, Einstein AJ et al.   Differences in repeating patterns of complex fractionated left atrial electrograms in longstanding persistent as compared with paroxysmal atrial fibrillation. Circ Arrhythm Electrophysiol  2011; 4: 470– 7. Google Scholar CrossRef Search ADS PubMed  12 Goette A, Kalman JM, Aguinaga L, Akar J, Cabrera JA, Chen SA et al.   EHRA/HRS/APHRS/SOLAECE expert consensus on atrial cardiomyopathies: definition, characterization, and clinical implication. Europace  2016; 18: 1455– 90. Google Scholar CrossRef Search ADS PubMed  13 Zhao Y, Di Biase L, Trivedi C, Mohanty S, Bai R, Mohanty P, Gianni C et al.   Importance of non-pulmonary vein triggers ablation to achieve long-term freedom from paroxysmal atrial fibrillation in patients with low ejection fraction. Heart Rhythm  2016; 13: 141– 9. Google Scholar CrossRef Search ADS PubMed  14 Di Biase L, Mohanty P, Mohanty S, Santangeli P, Trivedi C, Lakkireddy D et al.   Ablation versus amiodarone for treatment of persistent atrial fibrillation in patients with congestive heart failure and an implanted device: results from the AATAC multicenter randomized trial. Circulation  2016; 133: 1637– 44. Google Scholar CrossRef Search ADS PubMed  15 Hunter RJ, Berriman TJ, Diab I, Kamdar R, Richmond L, Baker V et al.   A randomized controlled trial of catheter ablation versus medical treatment of atrial fibrillation in heart failure (the CAMTAF trial). Circ Arrhythm Electrophysiol  2014; 7: 31– 8. doi: 10.1161/CIRCEP.113.000806. Google Scholar CrossRef Search ADS PubMed  16 Jones DG, Haldar S, Hussain W, Sharma R, Francis DP, Rahman-Haley S et al.   A randomized trial to assess catheter ablation versus rate control in the management of persistent atrial fibrillation in heart failure (ARC-HF). J Am Coll Cardiol  2013; 61: 1894– 903. Google Scholar CrossRef Search ADS PubMed  17 Anselmino M, Matta M, Castagno D, Giustetto C, Gaita F. Catheter ablation of atrial fibrillation in chronic heart failure: state-of-the-art and future perspectives. Europace  2016; 18: 638– 47. Google Scholar CrossRef Search ADS PubMed  18 Ullah W, Ling LH, Prabhu S, Lee G, Kistler P, Finlay MC et al.   Catheter ablation of atrial fibrillation in patients with heart failure: impact of maintaining sinus rhythm on heart failure status and long-term rates of stroke and death. Europace  2016; 18: 679– 86. Google Scholar CrossRef Search ADS PubMed  19 Morton JB, Sanders P, Vohra JK, Sparks P, Morgan J, Spence S et al.   The effect of chronic atrial stretch on atrial electrical remodeling in patients with an atrial septal defect. Circulation  2003; 107: 1775– 82. Google Scholar CrossRef Search ADS PubMed  20 Roberts-Thomson KC, Stevenson I, Kistler PM, Haqqani HM, Spence SJ, Goldblatt JC et al.   The role of chronic atrial stretch and atrial fibrillation on posterior left atrial wall conduction. Heart Rhythm  2009; 6: 1109– 17. Google Scholar CrossRef Search ADS PubMed  21 Cox JL. The longstanding, persistent confusion surrounding surgery for atrial fibrillation. J Thorac Cardiovasc Surg  2010; 139: 1374– 86. Google Scholar CrossRef Search ADS PubMed  22 McCready JW, Smedley T, Lambiase PD, Ahsan SY, Segal OR, Rowland E et al.   Predictors of recurrence following radiofrequency ablation for persistent atrial fibrillation. Europace  2011; 13: 355– 61. Google Scholar CrossRef Search ADS PubMed  23 Elayi CS, Di Biase L, Barrett CChing CK, al Aly M, Lucciola M, Bai R et al.   Atrial fibrillation termination as a procedural endpoint during ablation in long-standing persistent atrial fibrillation. Heart Rhythm  2010; 7: 1216– 23. Google Scholar CrossRef Search ADS PubMed  24 Rostock T, Steven D, Hoffmann B, Servatius H, Drewitz I, Sydow K et al.   Chronic atrial fibrillation is a biatrial arrhythmia: data from catheter ablation of chronic atrial fibrillation aiming arrhythmia termination using a sequential ablation approach. Circ Arrhythm Electrophysiol  2008; 1: 344– 53. Google Scholar CrossRef Search ADS PubMed  25 Narayan S, Krummen D, Shivkumar K, Clopton P, Wouter-Jan R, Miller J. Treatment of atrial fibrillation by the ablation of localized sources: CONFIRM trial. J Am Coll Cardiol  2012; 60: 628– 36. Google Scholar CrossRef Search ADS PubMed  26 Ravelli F, Masè M, Cristoforetti A, Del Greco M, Centonze M, Marini M et al.   Anatomic localization of rapid repetitive sources in persistent atrial fibrillation: fusion of biatrial CT images with wave similarity/cycle length maps. JACC Cardiovasc Imaging  2012; 5: 1211– 20. Google Scholar CrossRef Search ADS PubMed  27 Haissaguerre M, Hocini M, Denis A, Shah A, Komatsu Y, Yamashita S et al.   Driver domains in persistent atrial fibrillation. Circulation  2014; 130: 530– 8. Google Scholar CrossRef Search ADS PubMed  Published on behalf of the European Society of Cardiology. All rights reserved. © The Author 2017. For permissions, please email: journals.permissions@oup.com.

Journal

EuropaceOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off