- Email: [email protected]
Relationship among complex signals, short cycle length activity, and dominant frequency in patients with long-lasting persistent AF: A high-density epicardial mapping study in humans
Relationship among complex signals, short cycle length activity, and dominant frequency in patients with long-lasting persistent AF: A high-density epicardial mapping study in humans Geoffrey Lee, MBChB,* Kurt Roberts-Thomson, MBBS, PhD,† Andrew Madry, BSc, PhD,* Steven Spence, ACCT,* Andrew Teh, MBBS,* Patrick M. Heck, BM, BCh, DM,* Saurabh Kumar, BSc, MBBS,* Peter M. Kistler, MBBS, PhD,*‡ Joseph B. Morton, MBBS, PhD,* Prashanthan Sanders, MBBS, PhD,† Jonathan M. Kalman MBBS, PhD* From the *Department of Cardiology, Royal Melbourne Hospital and Department of Medicine, University of Melbourne, Melbourne, Australia, †Centre for Heart Rhythm Disorders, University of Adelaide and the Royal Adelaide Hospital, Adelaide, Australia, and ‡Department of Cardiology, The Alfred Hospital, and the Baker IDI Diabetes and Heart Research Institute, Melbourne, Australia. BACKGROUND Complex fractionated atrial electrograms (CFAEs) and regions of high dominant frequency (DF) both may identify sites critical to the maintenance of atrial fibrillation (AF). CFAEs may be defined by either (1) complex multicomponent electrograms (EGMs) and/or continuous electrical activity (multicomponent/continuous EGM) or (2) discrete high-frequency EGMs. OBJECTIVE The purpose of this study was to test if the 2 definitions of CFAE identify the same arrhythmia substrate and determine the relationship of CFAE to areas of high DF. METHODS High-density epicardial mapping of the posterior left atrium was performed in 10 patients with long-lasting persistent AF. Point-by-point analysis was performed to determine the spatial distribution and correlation of CFAE defined as either (1) multicomponent/continuous-EGMs or (2) AF cycle length ⬍120 ms. Additionally, spatial analysis was performed to determine the relationship of high DF sites to CFAE sites defined by each of the 2 definitions. RESULTS The percentage of sites deemed CFAE varied markedly between patients and was different depending on the definition of CFAE adopted. There was a poor correlation between CFAE defined by multicomponent/continuous EGMs and AF cycle length ⬍120
Introduction 1
Nademanee et al first described complex fractionated atrial electrograms (CFAEs) as target sites for substrate-based ablation in patients with atrial fibrillation (AF) and reported good clinical outcomes. However, subsequent studies of CFAE ablation used alone or in addition to pulmonary vein isolation have yielded contradictory results.2– 4 These varied outcomes potentially may be explained by the different Address reprint requests and correspondence: Dr. Jonathan M. Kalman, Department of Cardiology, Royal Melbourne Hospital, Melbourne, Australia 3050. E-mail address: [email protected]. (Received April 11, 2011; accepted May 24, 2011.)
ms (r ⫽ 0.18). High DF sites were arranged in clusters evenly distributed throughout the posterior left atrium, with 4.2 ⫾ 1.0 high DF clusters per patient. Although there was poor point-bypoint correlation between multicomponent/continuous EGMs and high DF sites (r ⫽ 0.107), spatial analysis revealed that 96% of multicomponent/continuous EGMs were found adjacent to and partially surrounding (ⱕ5 mm) high DF sites. CONCLUSION There is poor anatomic overlap between CFAE defined by multicomponent/continuous EGMs and CFAE defined by AF cycle length ⬍120 ms. Multicomponent/continuous EGMs are found adjacent to and surrounding sites of high DF. Further studies are needed to determine the mechanisms responsible for these different signals. KEYWORDS Atrial fibrillation; Complex fractionated atrial electrogram; Dominant frequency ABBREVIATIONS AF ⫽ atrial fibrillation; CFAE ⫽ complex fractionated atrial electrogram; CL ⫽ cycle length; DF ⫽ dominant frequency; EGM ⫽ electrogram; IQR ⫽ interquartile range; LA ⫽ left atrium; MAP ⫽ monophasic action potential (Heart Rhythm 2011;8:1714 –1719) © 2011 Heart Rhythm Society. All rights reserved.
definitions of CFAE in clinical use. CFAEs have been defined by the presence of either multicomponent or continuous electrograms (multicomponent/continuous EGM) or discrete atrial EGMs with short cycle length (CL) activity (⬍120 ms). Although the 2 definitions of CFAE both have been used to identify potential AF sources, it remains unclear whether they signify the same underlying electrophysiologic mechanism or whether they identify the same anatomic sites. In order to standardize CFAE definition, automated computer algorithms that define CFAE based on AFCL criteria (⬍120 ms) have been developed. Dominant frequency (DF) analysis has also been used in an attempt to define with improved specificity driver sites within the atria
1547-5271/$ -see front matter © 2011 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2011.05.021
Lee et al
Relationship Among Complex Electrograms, Short Cycle Length Activity, and High Dominant Frequency
that may be targets for ablation. Although these varied definitions have been used to identify ablation sites, their relationship remains unclear. Using simultaneous high-density epicardial mapping in patients with long-standing persistent AF, the aims of the current study were to determine the prevalence and spatial correlation of CFAE using 2 definitions—(1) multicomponent/continuous EGMs and (2) AFCL ⬍120 ms—and to determine the relationship of these sites to areas of high DF.
Study population This study consisted of 10 patients with long-lasting persistent AF undergoing elective cardiac surgery for mitral regurgitation (n ⫽ 7), aortic stenosis (n ⫽ 2), and ischemic heart disease (n ⫽ 1). The mean and median durations of long-lasting persistent AF were 9.1 ⫾ 11.8 years and 4.5 years, respectively. In accordance with the HRS Expert Consensus Statement,5 persistent AF was defined as AF that sustained for more than 7 days or that lasted less than 7 days but necessitated pharmacologic or electrical cardioversion. Included within the category of persistent AF was “longstanding persistent AF,” which is defined as continuous AF of greater than 1-year’s duration. Baseline characteristics of all 10 patients are shown in Table 1. All antiarrhythmic medications were stopped more than 5 half-lives prior to the surgery. All patients gave written and informed consent prior to the surgery, and the study protocol was approved by the research and ethics committee of Melbourne Health.
Study protocol Following median sternotomy and prior to the institution of cardiopulmonary bypass, high-density atrial epicardial mapping was performed. A custom-made high-density triangular epicardial plaque comprising 128 silver-plated copper electrodes with an interelectrode distance of 2.5 mm (effective mapping area 6.75 cm2) was placed in the oblique sinus on the posterior left atrium (LA) between the pulmonary veins, as previously described.6 Once plaque position was stable, continuous 10-second bipolar atrial EGMs were recorded using a computerized mapping system (Unemap, Uniservices, Auckland, New Zealand) for offline analysis. Table 1
1715
Ten-second EGMs were sampled at 1000 Hz with a bandpass filter ⬍400 Hz used.
High-density spatial analysis All EGMs were acquired during AF. Using the 128-point epicardial electrode matrix as a frame of reference, we performed simultaneous point-by-point analysis at each location for the presence of 3 specific characteristics: (1) multicomponent/continuous EGM; (2) AFCL ⬍120 ms; and (3) DF. The following spatial relationships were analyzed: (1) multicomponent/continuous EGM to AFCL ⬍120 ms; (2) multicomponent/continuous EGM to high DF; and (3) AFCL ⬍120 ms to high DF.
Assessment of visual fractionation All visual EGM analyses were performed at 100 mm/s sweep speed. Multicomponent/continuous EGM was defined as atrial EGMs with ⱖ3 deflections of ⬎50-ms duration without an isoelectric baseline and/or continuous electrical activity, as previously described.1,7 For each 10-second recording of AF, the signal had to fulfill the above criteria for ⬎75% of the recording in order for that site to be deemed fractionated. In order to determine the degree of intra- and interobserver variability in EGM classification, the principal investigator and a second observer (PH) were asked to classify a blinded sample of 100 atrial EGMs from the data as either multicomponent/continuous EGM or non–multicomponent/ continuous EGM using the above study criteria. For the principal investigator, there was a 98% intraobserver agreement in the classification of multicomponent/continuous EGM sites (kappa ⫽ 0.91, P ⬍.001). For the second observer (PH), there was 94% interobserver agreement with the principal investigator in the classification of multicomponent/continuous EGM sites (kappa ⫽ 0.75, P ⬍.001).
Assessment of AF cycle length Unemap signals were imported into customized computer software (Cardiac ElectroPhysiology Analysis System [CEPAS]; Cuoretech Pty Ltd, Sydney, Australia) that enabled automated AFCL calculation and DF analysis. Similar
Patient characteristics
Patient
Age (y)
1 2 3 4 5 6 7 8 9 10
35 75 77 66 64 82 79 60 65 82
Structural heart disease MR MR MR MR MR CAGS MR AS AS MR
Ejection fraction (%)
Left atrial diameter (cm)
Left ventricular end-diastolic diameter (cm)
New york heart association failure grade
Years in continuous atrial fibrillation
⬎59% 40%–59% ⬎59% ⬎59% ⬎59% ⬎59% ⬎59% ⬎59% ⬎59% 40%–59%
5.7 6.4 5.4 5.8 5.9 5.8 6.0 4.9 4.3 4.6
6.5 6.7 4.9 5.5 6.6 5.0 5.1 5.1 5.0 5.5
2 2 1 2 3 2 2 2 2 2
3 5 17 3 10 4 2 40 5 2
AS ⫽ aortic stenosis; MR ⫽ mitral regurgitation; CAGS ⫽ coronary artery graft surgery.
1716
Heart Rhythm, Vol 8, No 11, November 2011
to commercially available software, CEPAS has specific user-defined characteristics to identify EGM activations. These include (1) a baseline noise threshold, (2) EGM width criterion to avoid detection of broad far-field activations, (3) EGM slope, and (4) EGM “refractory” periods to avoid multiple detections within the same activation. The baseline noise threshold for CFAE determination was adjusted according to the baseline noise of the electrical recordings, and EGM detection parameters were standardized as previously suggested by Aizer et al.8 A noise threshold of 0.1 mV, width criterion 10 ms and refractory period of 49 ms was used for our analysis. All automated EGM annotations were visually verified to confirm correct annotation of activations. Mean AFCL of each point was calculated for the 10-second recording period. AFCL ⬍120 ms was specifically used to define CFAE because this has been the cutoff level used in previous studies.1,3,4
ables are reported as number and percentages. Proportions were compared using Chi-square test and Fisher exact test as appropriate. DF and AF cycle length were not normally distributed. As such, comparisons of these variables were analyzed using the Mann-Whitney U test. Point-by-point correlation of sites of visual fractionation, AFCL ⬍100 ms, and AFCL ⬍120 ms was performed using the McNemar test and Pearson r statistics. Receiver operating characteristic curves were constructed to determine the accuracy of AFCL to detect visual fractionation using an area under the curve calculation. Area under the curve cutoff values of 0.9 –1.0, 0.8 – 0.9, 0.7– 0.8, 0.6 – 0.7, and ⬍0.6 were used to classify the accuracy of the test as excellent, good, fair, poor, or not accurate, respectively.10 P ⬍ .05 was considered significant.
Assessment of DF
Patient characteristics
Within the same 10-second recording window, spectral analysis for determination of DF was analyzed. Exported signals were rectified, filtered using a Butterworth filter, and edge-tapered with a Hanning window. DF was determined by fast Fourier transform using zero padding with spectral resolution of 0.1 Hz. The DF of each individual electrode recording site was defined as the frequency demonstrating the highest power within the 3-to 15-Hz frequency domain. A high DF site within the recording plaque was identified if the local electrode DF was 20% greater than the DF of adjacent electrodes.9
Baseline characteristics of the 10 patients are given in Table 1. All patients had a history of long-lasting persistent AF of greater than 1 year’s duration.
Statistical analysis Normality of all quantitative data variables was checked using the Kolmogorov-Smirnov test. Continuous variables are reported as mean ⫾ SD and median and interquartile range (IQR), as appropriate. Categorical vari-
Results
CFAE by definition Figure 1 shows the percentage of plaque fractionation per patient using the 2 different definitions of CFAE. The percentage of sites deemed CFAE varied markedly between patients (76% to 14% by AFCL ⬍120 ms definition) and was significantly different using the 2 definitions. CFAE defined by multicomponent/continuous EGM was less prevalent than CFAE defined purely by AFCL criterion. Overall, CFAE defined by multicomponent/continuous EGM was present in 22% ⫾ 9% of all posterior wall sites. CFAE defined by AFCL ⬍120 ms was present in 40% ⫾ 20% of posterior wall sites.
Figure 1 Percentage of posterior left atrial wall fractionation is shown for each patient using the 2 definitions of complex fractionated atrial electrogram (CFAE). The difference in percentage CFAE defined by multicomponent/continuous electrogram (EGM) vs atrial fibrillation cycle length (AFCL) ⬍120 ms was statistically significant in all patients (P ⬍.05).
Lee et al
Relationship Among Complex Electrograms, Short Cycle Length Activity, and High Dominant Frequency
1717
Relationship between AFCL and multicomponent/continuous EGM
Spatial relationship of multicomponent/continuous EGMs and high DF
Median AFCL of the posterior LA wall was 115 ms (IQR 93–157). We assessed the ability of an AFCL ⬍120 ms criterion to detect multicomponent/continuous EGM and found that it was a poor predictor of multicomponent/ continuous EGMs. The percentage of points with AFCL ⬍120 ms that showed multicomponent/continuous EGMs was 33% ⫾ 18%. Conversely, the percentage of points with multicomponent/continuous EGMs that showed AFCL ⬍120 ms was 62% ⫾ 30%. Using AFCL ⬍120 ms, the sensitivity, specificity, positive predictive value, and negative predictive value was 61%, 57%, 32%, and 82%, respectively. Receiver operating characteristic curves confirmed that AFCL ⬍120 ms is a poor tool for predicting sites of multicomponent/continuous EGMs with area under the curve value of 0.59.
Using the epicardial plaque as a frame of reference, the spatial relationship between sites of multicomponent/continuous EGM and high DF was assessed for each patient (Figure 2). There was a poor direct spatial correlation between sites of multicomponent/continuous EGM and sites of high DF (r ⫽ 0.107), with only 23.1% of multicomponent/continuous EGM sites occurring at the same location as a site of high DF. However, spatial analysis confirmed that the vast majority (84%) of the multicomponent/continuous EGM sites occurred directly adjacent (ⱕ2.5 mm) to a site of high DF. Of the remaining multicomponent/continuous EGM sites, 12% were 2.5–5 mm and 4% were 5–10 mm away from a high DF site. No multicomponent/continuous EGM was seen ⬎10 mm away from a high DF site.
DF analysis
Spatial relationship of AFCL and high DF sites
Median DF for the posterior wall was 5.5 Hz (IQR 4.8 – 6.7). Mean number of individual high DF sites identified per patient was 17 ⫾ 5. High DF sites were arranged in clusters, with 4.2 ⫾ 1.0 high DF clusters per patients. Median DF of the high DF sites was 8.7 Hz (IQR 8.7–10.4) compared with 5.4 Hz (IQR 4.5–5.9) for non– high DF sites (P ⫽ .001). The high DF clusters were distributed equally throughout each mapping area, and no characteristic patterns were identified. In particular, high DF clusters did not localize to the lateral edges of the plaque (ie, adjacent to the pulmonary veins). There was no difference in the DF of multicomponent/continuous EGM sites compared to non–multicomponent/continuous EGM sites: 5.6 Hz (IQR 4.6 –7.6) compared to 5.5Hz (IQR 4.9 – 6.6) (P ⫽ .56).
Using the same methodology as above, we analyzed the spatial relationship between high DF sites and AFCL ⬍120 ms sites (Figure 2). There was a poor direct spatial correlation between sites of AFCL ⬍120 ms and high DF (r ⫽ 0.05, P ⫽ .99). Only 17.6% of short CL sites (AFCL ⬍120 ms) occurred at sites of high DF (P ⬍.001), and only 54.7% of these sites were directly adjacent (ⱕ2.5 mm) to a high DF site (P ⫽ .6).
Figure 2 Spatial analysis for patient 6. Left: Spatial relationship between sites of multicomponent/continuous electrogram (EGM) (yellow boxes) and high dominant frequency (DF) (red boxes). Right: Spatial relationship between atrial fibrillation cycle length (AFCL) ⬍120 ms (gray boxes) and high DF. Red– yellow boxes represent sites of anatomic overlap between high DF and multicomponent/ continuous EGM. Red–gray boxes represent sites of anatomic overlap between high DF and AFCL ⬍120 ms. Left: Although there is little direct anatomic overlap between sites of multicomponent/continuous EGM and high DF, the majority of multicomponent/continuous EGM sites are found clustered around areas of high DF. Right: Poor anatomic correlation exists between high DF and short AFCL sites. Representative electrograms of sites of multicomponent/continuous EGM, AFCL⬎120 ms, and a site with AFCL ⬍120 ms are shown. A representative example of the spectral analysis from a high DF site is shown with DF peak of 10.2 Hz.
Discussion The main findings of this study are as follows. 1. The prevalence of CFAE defined by multicomponent/ continuous EGM differs significantly from the prevalence of CFAE defined by short CL activity.
1718 2. Spatial analysis confirms there is poor anatomic overlap between the definitions of CFAE. Multicomponent/continuous EGMs and sites of short CL activity (AFCL ⬍120 ms) identify different atrial regions. 3. Simultaneous high-density epicardial mapping reveals that sites of multicomponent/continuous EGM occur adjacent to and surrounding areas of high-frequency activation identified by DF analysis. This is consistent with the hypothesis that CFAEs represent areas of wavebreak around a high-frequency focus as previously described in an animal study.11 4. There is a poor anatomic correlation between sites of short AFCL (⬍120 ms) and sites of high DF.
CFAE: Definition, prevalence, and anatomic distribution Since the work of Nademanee et al,1 ablation of complex fractionated signals (CFAEs) has become a standard component of procedures for cure of persistent AF. These signals potentially identify the anatomic sites critical to AF maintenance, in the form of focal drivers or small rotors. However, this putative link between the recorded signal and the presence of an AF source remains unconfirmed. Reports of the efficacy of CFAE ablation have differed widely. In part, this may be related to the different definitions of CFAE that are in clinical use: visual fractionation or short CL activity. In the current study using high-density simultaneous epicardial mapping, we demonstrate that different CFAE definitions result in a widely varying CFAE prevalence and a poor correlation in anatomic location according to CFAE definition. Previous studies have shown that although most regions of the LA show some degree of multicomponent/continuous EGMs,12,13 the anatomic locations of highest prevalence are at the septum, LA roof, posterior wall, and within the coronary sinus.1 The high prevalence of CFAEs at sites of anatomic overlap (eg, coronary sinus and LA septum) suggest that, in some instances, complex signals may result from the summation of far- and near-field signals from adjacent but differentially activated atrial structures.14,15 When CFAE has been defined by the presence of short CL activity (AFCL ⬍120 ms) using endocardial mapping systems, it may be very widespread, with reported prevalence of up to 86% of all atrial signals.16 –18 Using high-density endocardial contact mapping, Rostock et al19 found uniform shortening of AFCL preceded the appearance of multicomponent/continuous EGMs and observed an inverse relationship between the duration of multicomponent/continuous EGMs and preceding AFCL. The stereotypical distribution of multicomponent/continuous EGMs to anatomic areas of complex muscle fiber overlap and its relationship to preceding AFCL suggest that multicomponent/continuous EGMs may be determined by rate and local tissue conduction properties rather than necessarily reflecting the site of a focal source.20,21 Recently, Narayan et al15 used monophasic action potential (MAP) recordings to further classify fractionated EGMs into distinct functional subtypes. In this
Heart Rhythm, Vol 8, No 11, November 2011 study, 67% of all MAP recordings analyzed from CFAE sites were characterized by the presence of dissociated signals superimposed on local MAP recordings, suggesting that the majority of CFAEs were the result of superimposed far-field atrial activations from overlying atrial structures. In contrast, only a small proportion (8%) of CFAE exhibited rapid, discrete, organized MAP recording activity consistent with an AF driver.
DF vs visual fractionation Few studies have looked at the spatial relationship between multicomponent/continuous EGMs and high DF. In an acute ovine model of acetylcholine-induced AF, Kalifa et al11 tested the hypothesis that areas of fast, highly periodic activity near the posterior LA wall would give rise to the most fractionated EGMs. They observed that fractionation was lowest at the center of the high DF area and highest at surrounding low-frequency border zones where wavebreak and beat-to-beat variability in wavefront direction and velocity occurred, suggesting a passive phenomenon. Using complex computer modeling to analyze optical and electrical recordings obtained from isolated Langendorff-perfused sheep hearts, Zlochiver et al22 found that waves emanating from a moving rotor consistently undergo beat-to-beat changes in activation directionality and AFCL and hence provide a direct mechanism for EGM fractionation. This study is the first to use simultaneous high-density epicardial mapping during human AF to characterize the spatial relationship between multicomponent/continuous EGMs and sites of high DF identified by spectral analysis. In this study, the vast majority of multicomponent/continuous EGMs occurred at sites directly adjacent (ⱕ2.5 mm) to a high DF site. If high DF identifies a focal source, then visual fractionation may represent wavefront breakup at the periphery. In humans, this may be exacerbated by the presence of regions of atrial remodeling resulting in slowed conduction.6
AFCL and DF In a Langendorff-perfused ovine model of cholinergic AF, Skanes et al23 showed a strong correlation between the AFCL of periodic activations and DF, suggesting that a significant proportion of the mapped atria was activated in a 1:1 pattern. In contrast, high-density endocardial mapping studies performed in humans have found a poor correlation between high frequency sites detected by AFCL and DF analysis.24,25 Stiles et al24 looked at the spatial relationship of CFAE identified by the NavX algorithm (CFE mean ⬍120 ms) to the 20 highest DF sites obtained from a point-by-point endocardial map during AF. Using points taken from a 20-pole PentaRay catheter (4-mm interelectrode spacing; Biosense-Webster, Diamond Bar, CA) they observed a poor point-by-point correlation between CFAE (NavX algorithm) and DF. CFAE areas were located adjacent to areas of high DF, with 80% ⱕ10 mm from a high DF area. A more detailed analysis was limited by the size of the
Lee et al
Relationship Among Complex Electrograms, Short Cycle Length Activity, and High Dominant Frequency
recording bipole and the sequential nature of data acquisition. Human studies exploring the relationship between AFCL and DF using epicardial26 or endocardial mapping24,25 have yielded divergent results. A number of hypotheses have been advanced to explain the poor correlation seen in some studies. Among them, the presence of advanced remodeling with loss of 1:1 conduction,27,28 the presence of “endocardial– epicardial dissociation,”29 and the difficulty of determining local atrial activation rates when signals have complex morphology30 all may be important.
Study limitations Mapping of CFAE, AFCL, and DF was limited to the posterior LA, and whether the same relationships exist in other locations is unknown. The duration of persistent AF in this study is significantly longer compared to previous CFAE ablation series. It is unclear if progressive electroanatomic remodeling affects the spatial relationships described in this study. The temporal stability of CFAE and DF has been questioned recently,31 and it is possible that a longer recording time may be required to fully understand the relationship between these parameters. However, longduration recordings to guide ablation in clinical procedures may lack feasibility.
Conclusion The prevalence of CFAEs in the posterior LA varies markedly among patients and is highly dependent on the definition of CFAE adopted. There is poor anatomic overlap between CFAE defined by multicomponent/continuous EGM and CFAE defined by short CL activity. Areas of multicomponent/continuous EGMs are recorded adjacent to regions of high DF. Further studies are needed to determine the mechanisms responsible for these different signals.
References 1. Nademanee K, McKenzie J, Kosar E, et al. A new approach for catheter ablation of atrial fibrillation: mapping of the electrophysiologic substrate. J Am Coll Cardiol 2004;43:2044 –2053. 2. Oral H, Chugh A, Yoshida K, et al. A randomized assessment of the incremental role of ablation of complex fractionated atrial electrograms after antral pulmonary vein isolation for long-lasting persistent atrial fibrillation. J Am Coll Cardiol 2009;53:782–789. 3. Elayi CS, Verma A, Di Biase L, et al. Ablation for longstanding permanent atrial fibrillation: results from a randomized study comparing three different strategies. Heart Rhythm 2008;5:1658 –1664. 4. Verma A, Novak P, Macle L, et al. A prospective, multicenter evaluation of ablating complex fractionated electrograms (CFEs) during atrial fibrillation (AF) identified by an automated mapping algorithm: acute effects on AF and efficacy as an adjuvant strategy. Heart Rhythm 2008;5:198 –205. 5. Calkins H, Brugada J, Packer DL, et al. HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for personnel, policy, procedures and follow-up. A report of the Heart Rhythm Society (HRS) Task Force on catheter and surgical ablation of atrial fibrillation. Heart Rhythm 2007;4:816 – 861. 6. Roberts-Thomson KC, Stevenson IH, Kistler PM, et al. Anatomically determined functional conduction delay in the posterior left atrium relationship to structural heart disease. J Am Coll Cardiol 2008;51:856 – 862. 7. Oral H, Chugh A, Good E, et al. Radiofrequency catheter ablation of chronic atrial fibrillation guided by complex electrograms. Circulation 2007;115:2606 – 2612.
1719
8. Aizer A, Holmes DS, Garlitski AC, et al. Standardization and validation of an automated algorithm to identify fractionation as a guide for atrial fibrillation ablation. Heart Rhythm 2008;5:1134 –1141. 9. Sanders P, Berenfeld O, Hocini M, et al. Spectral analysis identifies sites of high-frequency activity maintaining atrial fibrillation in humans. Circulation 2005;112:789 –797. 10. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29 –36. 11. Kalifa J, Tanaka K, Zaitsev AV, et al. Mechanisms of wave fractionation at boundaries of high-frequency excitation in the posterior left atrium of the isolated sheep heart during atrial fibrillation. Circulation 2006;113:626 – 633. 12. Jais P, Haissaguerre M, Shah DC, Chouairi S, Clementy J. Regional disparities of endocardial atrial activation in paroxysmal atrial fibrillation. Pacing Clin Electrophysiol 1996;19:1998 –2003. 13. Tada H, Yoshida K, Chugh A, et al. Prevalence and characteristics of continuous electrical activity in patients with paroxysmal and persistent atrial fibrillation. J Cardiovasc Electrophysiol 2008;19:606 – 612. 14. Teh AW, Kalman JM, Kistler PM, et al. Prevalence of fractionated electrograms in the coronary sinus: comparison between patients with persistent or paroxysmal atrial fibrillation and a control population. Heart Rhythm 2010;7:1200 –1204. 15. Narayan SM, Wright M, Derval N, et al. Classifying fractionated electrograms in human atrial fibrillation using monophasic action potentials and activation mapping: evidence for localized drivers, rate acceleration, and nonlocal signal etiologies. Heart Rhythm 2011;8:244 –253. 16. Scherr D, Dalal D, Cheema A, et al. Automated detection and characterization of complex fractionated atrial electrograms in human left atrium during atrial fibrillation. Heart Rhythm 2007;4:1013–1020. 17. Wu J, Estner H, Luik A, et al. Automatic 3D mapping of complex fractionated atrial electrograms (CFAE) in patients with paroxysmal and persistent atrial fibrillation. J Cardiovasc Electrophysiol 2008;19:897–903. 18. Roux JF, Gojraty S, Bala R, et al. Complex fractionated electrogram distribution and temporal stability in patients undergoing atrial fibrillation ablation. J Cardiovasc Electrophysiol 2008;19:815– 820. 19. Rostock T, Rotter M, Sanders P, et al. High-density activation mapping of fractionated electrograms in the atria of patients with paroxysmal atrial fibrillation. Heart Rhythm 2006;3:27–34. 20. Spach MS, Dolber PC. Relating extracellular potentials and their derivatives to anisotropic propagation at a microscopic level in human cardiac muscle. Evidence for electrical uncoupling of side-to-side fiber connections with increasing age. Circ Res 1986;58:356 –371. 21. Konings KT, Kirchhof CJ, Smeets JR, Wellens HJ, Penn OC, Allessie MA. High-density mapping of electrically induced atrial fibrillation in humans. Circulation 1994;89:1665–1680. 22. Zlochiver S, Yamazaki M, Kalifa J, Berenfeld O. Rotor meandering contributes to irregularity in electrograms during atrial fibrillation. Heart Rhythm 2008;5: 846 – 854. 23. Skanes AC, Mandapati R, Berenfeld O, Davidenko JM, Jalife J. Spatiotemporal periodicity during atrial fibrillation in the isolated sheep heart. Circulation 1998;98:1236 –1248. 24. Stiles MK, Brooks AG, Kuklik P, et al. High-density mapping of atrial fibrillation in humans: relationship between high-frequency activation and electrogram fractionation. J Cardiovasc Electrophysiol 2008;19:1245–1253. 25. Elvan A, Linnenbank AC, van Bemmel MW, et al. Dominant frequency of atrial fibrillation correlates poorly with atrial fibrillation cycle length. Circ Arrhythm Electrophysiol 2009;2:634 – 644. 26. Schuessler RB, Kay MW, Melby SJ, Branham BH, Boineau JP, Damiano RJJ. Spatial and temporal stability of the dominant frequency of activation in human atrial fibrillation. J Electrocardiol 2006;39:S7–S12. 27. Sih HJ, Zipes DP, Berbari EJ, Adams DE, Olgin JE. Differences in organization between acute and chronic atrial fibrillation in dogs. J Am Coll Cardiol 2000; 36:924 –931. 28. Ryu K, Sahadevan J, Khrestian CM, Stambler BS, Waldo AL. Use of fast Fourier transform analysis of atrial electrograms for rapid characterization of atrial activation-implications for delineating possible mechanisms of atrial tachyarrhythmias. J Cardiovasc Electrophysiol 2006;17:198 –206. 29. de Groot NM, Houben RP, Smeets JL, et al. Electropathological substrate of longstanding persistent atrial fibrillation in patients with structural heart disease: epicardial breakthrough. Circulation. 2010;122:1674 –1682. 30. Ng J, Kadish AH, Goldberger JJ. Effect of electrogram characteristics on the relationship of dominant frequency to atrial activation rate in atrial fibrillation. Heart Rhythm 2006;3:1295–1305. 31. Habel N, Znojkiewicz P, Thompson N, et al. The temporal variability of dominant frequency and complex fractionated atrial electrograms constrains the validity of sequential mapping in human atrial fibrillation. Heart Rhythm 2010;7:586–593.
Introduction 1
Nademanee et al first described complex fractionated atrial electrograms (CFAEs) as target sites for substrate-based ablation in patients with atrial fibrillation (AF) and reported good clinical outcomes. However, subsequent studies of CFAE ablation used alone or in addition to pulmonary vein isolation have yielded contradictory results.2– 4 These varied outcomes potentially may be explained by the different Address reprint requests and correspondence: Dr. Jonathan M. Kalman, Department of Cardiology, Royal Melbourne Hospital, Melbourne, Australia 3050. E-mail address: [email protected]. (Received April 11, 2011; accepted May 24, 2011.)
ms (r ⫽ 0.18). High DF sites were arranged in clusters evenly distributed throughout the posterior left atrium, with 4.2 ⫾ 1.0 high DF clusters per patient. Although there was poor point-bypoint correlation between multicomponent/continuous EGMs and high DF sites (r ⫽ 0.107), spatial analysis revealed that 96% of multicomponent/continuous EGMs were found adjacent to and partially surrounding (ⱕ5 mm) high DF sites. CONCLUSION There is poor anatomic overlap between CFAE defined by multicomponent/continuous EGMs and CFAE defined by AF cycle length ⬍120 ms. Multicomponent/continuous EGMs are found adjacent to and surrounding sites of high DF. Further studies are needed to determine the mechanisms responsible for these different signals. KEYWORDS Atrial fibrillation; Complex fractionated atrial electrogram; Dominant frequency ABBREVIATIONS AF ⫽ atrial fibrillation; CFAE ⫽ complex fractionated atrial electrogram; CL ⫽ cycle length; DF ⫽ dominant frequency; EGM ⫽ electrogram; IQR ⫽ interquartile range; LA ⫽ left atrium; MAP ⫽ monophasic action potential (Heart Rhythm 2011;8:1714 –1719) © 2011 Heart Rhythm Society. All rights reserved.
definitions of CFAE in clinical use. CFAEs have been defined by the presence of either multicomponent or continuous electrograms (multicomponent/continuous EGM) or discrete atrial EGMs with short cycle length (CL) activity (⬍120 ms). Although the 2 definitions of CFAE both have been used to identify potential AF sources, it remains unclear whether they signify the same underlying electrophysiologic mechanism or whether they identify the same anatomic sites. In order to standardize CFAE definition, automated computer algorithms that define CFAE based on AFCL criteria (⬍120 ms) have been developed. Dominant frequency (DF) analysis has also been used in an attempt to define with improved specificity driver sites within the atria
1547-5271/$ -see front matter © 2011 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2011.05.021
Lee et al
Relationship Among Complex Electrograms, Short Cycle Length Activity, and High Dominant Frequency
that may be targets for ablation. Although these varied definitions have been used to identify ablation sites, their relationship remains unclear. Using simultaneous high-density epicardial mapping in patients with long-standing persistent AF, the aims of the current study were to determine the prevalence and spatial correlation of CFAE using 2 definitions—(1) multicomponent/continuous EGMs and (2) AFCL ⬍120 ms—and to determine the relationship of these sites to areas of high DF.
Study population This study consisted of 10 patients with long-lasting persistent AF undergoing elective cardiac surgery for mitral regurgitation (n ⫽ 7), aortic stenosis (n ⫽ 2), and ischemic heart disease (n ⫽ 1). The mean and median durations of long-lasting persistent AF were 9.1 ⫾ 11.8 years and 4.5 years, respectively. In accordance with the HRS Expert Consensus Statement,5 persistent AF was defined as AF that sustained for more than 7 days or that lasted less than 7 days but necessitated pharmacologic or electrical cardioversion. Included within the category of persistent AF was “longstanding persistent AF,” which is defined as continuous AF of greater than 1-year’s duration. Baseline characteristics of all 10 patients are shown in Table 1. All antiarrhythmic medications were stopped more than 5 half-lives prior to the surgery. All patients gave written and informed consent prior to the surgery, and the study protocol was approved by the research and ethics committee of Melbourne Health.
Study protocol Following median sternotomy and prior to the institution of cardiopulmonary bypass, high-density atrial epicardial mapping was performed. A custom-made high-density triangular epicardial plaque comprising 128 silver-plated copper electrodes with an interelectrode distance of 2.5 mm (effective mapping area 6.75 cm2) was placed in the oblique sinus on the posterior left atrium (LA) between the pulmonary veins, as previously described.6 Once plaque position was stable, continuous 10-second bipolar atrial EGMs were recorded using a computerized mapping system (Unemap, Uniservices, Auckland, New Zealand) for offline analysis. Table 1
1715
Ten-second EGMs were sampled at 1000 Hz with a bandpass filter ⬍400 Hz used.
High-density spatial analysis All EGMs were acquired during AF. Using the 128-point epicardial electrode matrix as a frame of reference, we performed simultaneous point-by-point analysis at each location for the presence of 3 specific characteristics: (1) multicomponent/continuous EGM; (2) AFCL ⬍120 ms; and (3) DF. The following spatial relationships were analyzed: (1) multicomponent/continuous EGM to AFCL ⬍120 ms; (2) multicomponent/continuous EGM to high DF; and (3) AFCL ⬍120 ms to high DF.
Assessment of visual fractionation All visual EGM analyses were performed at 100 mm/s sweep speed. Multicomponent/continuous EGM was defined as atrial EGMs with ⱖ3 deflections of ⬎50-ms duration without an isoelectric baseline and/or continuous electrical activity, as previously described.1,7 For each 10-second recording of AF, the signal had to fulfill the above criteria for ⬎75% of the recording in order for that site to be deemed fractionated. In order to determine the degree of intra- and interobserver variability in EGM classification, the principal investigator and a second observer (PH) were asked to classify a blinded sample of 100 atrial EGMs from the data as either multicomponent/continuous EGM or non–multicomponent/ continuous EGM using the above study criteria. For the principal investigator, there was a 98% intraobserver agreement in the classification of multicomponent/continuous EGM sites (kappa ⫽ 0.91, P ⬍.001). For the second observer (PH), there was 94% interobserver agreement with the principal investigator in the classification of multicomponent/continuous EGM sites (kappa ⫽ 0.75, P ⬍.001).
Assessment of AF cycle length Unemap signals were imported into customized computer software (Cardiac ElectroPhysiology Analysis System [CEPAS]; Cuoretech Pty Ltd, Sydney, Australia) that enabled automated AFCL calculation and DF analysis. Similar
Patient characteristics
Patient
Age (y)
1 2 3 4 5 6 7 8 9 10
35 75 77 66 64 82 79 60 65 82
Structural heart disease MR MR MR MR MR CAGS MR AS AS MR
Ejection fraction (%)
Left atrial diameter (cm)
Left ventricular end-diastolic diameter (cm)
New york heart association failure grade
Years in continuous atrial fibrillation
⬎59% 40%–59% ⬎59% ⬎59% ⬎59% ⬎59% ⬎59% ⬎59% ⬎59% 40%–59%
5.7 6.4 5.4 5.8 5.9 5.8 6.0 4.9 4.3 4.6
6.5 6.7 4.9 5.5 6.6 5.0 5.1 5.1 5.0 5.5
2 2 1 2 3 2 2 2 2 2
3 5 17 3 10 4 2 40 5 2
AS ⫽ aortic stenosis; MR ⫽ mitral regurgitation; CAGS ⫽ coronary artery graft surgery.
1716
Heart Rhythm, Vol 8, No 11, November 2011
to commercially available software, CEPAS has specific user-defined characteristics to identify EGM activations. These include (1) a baseline noise threshold, (2) EGM width criterion to avoid detection of broad far-field activations, (3) EGM slope, and (4) EGM “refractory” periods to avoid multiple detections within the same activation. The baseline noise threshold for CFAE determination was adjusted according to the baseline noise of the electrical recordings, and EGM detection parameters were standardized as previously suggested by Aizer et al.8 A noise threshold of 0.1 mV, width criterion 10 ms and refractory period of 49 ms was used for our analysis. All automated EGM annotations were visually verified to confirm correct annotation of activations. Mean AFCL of each point was calculated for the 10-second recording period. AFCL ⬍120 ms was specifically used to define CFAE because this has been the cutoff level used in previous studies.1,3,4
ables are reported as number and percentages. Proportions were compared using Chi-square test and Fisher exact test as appropriate. DF and AF cycle length were not normally distributed. As such, comparisons of these variables were analyzed using the Mann-Whitney U test. Point-by-point correlation of sites of visual fractionation, AFCL ⬍100 ms, and AFCL ⬍120 ms was performed using the McNemar test and Pearson r statistics. Receiver operating characteristic curves were constructed to determine the accuracy of AFCL to detect visual fractionation using an area under the curve calculation. Area under the curve cutoff values of 0.9 –1.0, 0.8 – 0.9, 0.7– 0.8, 0.6 – 0.7, and ⬍0.6 were used to classify the accuracy of the test as excellent, good, fair, poor, or not accurate, respectively.10 P ⬍ .05 was considered significant.
Assessment of DF
Patient characteristics
Within the same 10-second recording window, spectral analysis for determination of DF was analyzed. Exported signals were rectified, filtered using a Butterworth filter, and edge-tapered with a Hanning window. DF was determined by fast Fourier transform using zero padding with spectral resolution of 0.1 Hz. The DF of each individual electrode recording site was defined as the frequency demonstrating the highest power within the 3-to 15-Hz frequency domain. A high DF site within the recording plaque was identified if the local electrode DF was 20% greater than the DF of adjacent electrodes.9
Baseline characteristics of the 10 patients are given in Table 1. All patients had a history of long-lasting persistent AF of greater than 1 year’s duration.
Statistical analysis Normality of all quantitative data variables was checked using the Kolmogorov-Smirnov test. Continuous variables are reported as mean ⫾ SD and median and interquartile range (IQR), as appropriate. Categorical vari-
Results
CFAE by definition Figure 1 shows the percentage of plaque fractionation per patient using the 2 different definitions of CFAE. The percentage of sites deemed CFAE varied markedly between patients (76% to 14% by AFCL ⬍120 ms definition) and was significantly different using the 2 definitions. CFAE defined by multicomponent/continuous EGM was less prevalent than CFAE defined purely by AFCL criterion. Overall, CFAE defined by multicomponent/continuous EGM was present in 22% ⫾ 9% of all posterior wall sites. CFAE defined by AFCL ⬍120 ms was present in 40% ⫾ 20% of posterior wall sites.
Figure 1 Percentage of posterior left atrial wall fractionation is shown for each patient using the 2 definitions of complex fractionated atrial electrogram (CFAE). The difference in percentage CFAE defined by multicomponent/continuous electrogram (EGM) vs atrial fibrillation cycle length (AFCL) ⬍120 ms was statistically significant in all patients (P ⬍.05).
Lee et al
Relationship Among Complex Electrograms, Short Cycle Length Activity, and High Dominant Frequency
1717
Relationship between AFCL and multicomponent/continuous EGM
Spatial relationship of multicomponent/continuous EGMs and high DF
Median AFCL of the posterior LA wall was 115 ms (IQR 93–157). We assessed the ability of an AFCL ⬍120 ms criterion to detect multicomponent/continuous EGM and found that it was a poor predictor of multicomponent/ continuous EGMs. The percentage of points with AFCL ⬍120 ms that showed multicomponent/continuous EGMs was 33% ⫾ 18%. Conversely, the percentage of points with multicomponent/continuous EGMs that showed AFCL ⬍120 ms was 62% ⫾ 30%. Using AFCL ⬍120 ms, the sensitivity, specificity, positive predictive value, and negative predictive value was 61%, 57%, 32%, and 82%, respectively. Receiver operating characteristic curves confirmed that AFCL ⬍120 ms is a poor tool for predicting sites of multicomponent/continuous EGMs with area under the curve value of 0.59.
Using the epicardial plaque as a frame of reference, the spatial relationship between sites of multicomponent/continuous EGM and high DF was assessed for each patient (Figure 2). There was a poor direct spatial correlation between sites of multicomponent/continuous EGM and sites of high DF (r ⫽ 0.107), with only 23.1% of multicomponent/continuous EGM sites occurring at the same location as a site of high DF. However, spatial analysis confirmed that the vast majority (84%) of the multicomponent/continuous EGM sites occurred directly adjacent (ⱕ2.5 mm) to a site of high DF. Of the remaining multicomponent/continuous EGM sites, 12% were 2.5–5 mm and 4% were 5–10 mm away from a high DF site. No multicomponent/continuous EGM was seen ⬎10 mm away from a high DF site.
DF analysis
Spatial relationship of AFCL and high DF sites
Median DF for the posterior wall was 5.5 Hz (IQR 4.8 – 6.7). Mean number of individual high DF sites identified per patient was 17 ⫾ 5. High DF sites were arranged in clusters, with 4.2 ⫾ 1.0 high DF clusters per patients. Median DF of the high DF sites was 8.7 Hz (IQR 8.7–10.4) compared with 5.4 Hz (IQR 4.5–5.9) for non– high DF sites (P ⫽ .001). The high DF clusters were distributed equally throughout each mapping area, and no characteristic patterns were identified. In particular, high DF clusters did not localize to the lateral edges of the plaque (ie, adjacent to the pulmonary veins). There was no difference in the DF of multicomponent/continuous EGM sites compared to non–multicomponent/continuous EGM sites: 5.6 Hz (IQR 4.6 –7.6) compared to 5.5Hz (IQR 4.9 – 6.6) (P ⫽ .56).
Using the same methodology as above, we analyzed the spatial relationship between high DF sites and AFCL ⬍120 ms sites (Figure 2). There was a poor direct spatial correlation between sites of AFCL ⬍120 ms and high DF (r ⫽ 0.05, P ⫽ .99). Only 17.6% of short CL sites (AFCL ⬍120 ms) occurred at sites of high DF (P ⬍.001), and only 54.7% of these sites were directly adjacent (ⱕ2.5 mm) to a high DF site (P ⫽ .6).
Figure 2 Spatial analysis for patient 6. Left: Spatial relationship between sites of multicomponent/continuous electrogram (EGM) (yellow boxes) and high dominant frequency (DF) (red boxes). Right: Spatial relationship between atrial fibrillation cycle length (AFCL) ⬍120 ms (gray boxes) and high DF. Red– yellow boxes represent sites of anatomic overlap between high DF and multicomponent/ continuous EGM. Red–gray boxes represent sites of anatomic overlap between high DF and AFCL ⬍120 ms. Left: Although there is little direct anatomic overlap between sites of multicomponent/continuous EGM and high DF, the majority of multicomponent/continuous EGM sites are found clustered around areas of high DF. Right: Poor anatomic correlation exists between high DF and short AFCL sites. Representative electrograms of sites of multicomponent/continuous EGM, AFCL⬎120 ms, and a site with AFCL ⬍120 ms are shown. A representative example of the spectral analysis from a high DF site is shown with DF peak of 10.2 Hz.
Discussion The main findings of this study are as follows. 1. The prevalence of CFAE defined by multicomponent/ continuous EGM differs significantly from the prevalence of CFAE defined by short CL activity.
1718 2. Spatial analysis confirms there is poor anatomic overlap between the definitions of CFAE. Multicomponent/continuous EGMs and sites of short CL activity (AFCL ⬍120 ms) identify different atrial regions. 3. Simultaneous high-density epicardial mapping reveals that sites of multicomponent/continuous EGM occur adjacent to and surrounding areas of high-frequency activation identified by DF analysis. This is consistent with the hypothesis that CFAEs represent areas of wavebreak around a high-frequency focus as previously described in an animal study.11 4. There is a poor anatomic correlation between sites of short AFCL (⬍120 ms) and sites of high DF.
CFAE: Definition, prevalence, and anatomic distribution Since the work of Nademanee et al,1 ablation of complex fractionated signals (CFAEs) has become a standard component of procedures for cure of persistent AF. These signals potentially identify the anatomic sites critical to AF maintenance, in the form of focal drivers or small rotors. However, this putative link between the recorded signal and the presence of an AF source remains unconfirmed. Reports of the efficacy of CFAE ablation have differed widely. In part, this may be related to the different definitions of CFAE that are in clinical use: visual fractionation or short CL activity. In the current study using high-density simultaneous epicardial mapping, we demonstrate that different CFAE definitions result in a widely varying CFAE prevalence and a poor correlation in anatomic location according to CFAE definition. Previous studies have shown that although most regions of the LA show some degree of multicomponent/continuous EGMs,12,13 the anatomic locations of highest prevalence are at the septum, LA roof, posterior wall, and within the coronary sinus.1 The high prevalence of CFAEs at sites of anatomic overlap (eg, coronary sinus and LA septum) suggest that, in some instances, complex signals may result from the summation of far- and near-field signals from adjacent but differentially activated atrial structures.14,15 When CFAE has been defined by the presence of short CL activity (AFCL ⬍120 ms) using endocardial mapping systems, it may be very widespread, with reported prevalence of up to 86% of all atrial signals.16 –18 Using high-density endocardial contact mapping, Rostock et al19 found uniform shortening of AFCL preceded the appearance of multicomponent/continuous EGMs and observed an inverse relationship between the duration of multicomponent/continuous EGMs and preceding AFCL. The stereotypical distribution of multicomponent/continuous EGMs to anatomic areas of complex muscle fiber overlap and its relationship to preceding AFCL suggest that multicomponent/continuous EGMs may be determined by rate and local tissue conduction properties rather than necessarily reflecting the site of a focal source.20,21 Recently, Narayan et al15 used monophasic action potential (MAP) recordings to further classify fractionated EGMs into distinct functional subtypes. In this
Heart Rhythm, Vol 8, No 11, November 2011 study, 67% of all MAP recordings analyzed from CFAE sites were characterized by the presence of dissociated signals superimposed on local MAP recordings, suggesting that the majority of CFAEs were the result of superimposed far-field atrial activations from overlying atrial structures. In contrast, only a small proportion (8%) of CFAE exhibited rapid, discrete, organized MAP recording activity consistent with an AF driver.
DF vs visual fractionation Few studies have looked at the spatial relationship between multicomponent/continuous EGMs and high DF. In an acute ovine model of acetylcholine-induced AF, Kalifa et al11 tested the hypothesis that areas of fast, highly periodic activity near the posterior LA wall would give rise to the most fractionated EGMs. They observed that fractionation was lowest at the center of the high DF area and highest at surrounding low-frequency border zones where wavebreak and beat-to-beat variability in wavefront direction and velocity occurred, suggesting a passive phenomenon. Using complex computer modeling to analyze optical and electrical recordings obtained from isolated Langendorff-perfused sheep hearts, Zlochiver et al22 found that waves emanating from a moving rotor consistently undergo beat-to-beat changes in activation directionality and AFCL and hence provide a direct mechanism for EGM fractionation. This study is the first to use simultaneous high-density epicardial mapping during human AF to characterize the spatial relationship between multicomponent/continuous EGMs and sites of high DF identified by spectral analysis. In this study, the vast majority of multicomponent/continuous EGMs occurred at sites directly adjacent (ⱕ2.5 mm) to a high DF site. If high DF identifies a focal source, then visual fractionation may represent wavefront breakup at the periphery. In humans, this may be exacerbated by the presence of regions of atrial remodeling resulting in slowed conduction.6
AFCL and DF In a Langendorff-perfused ovine model of cholinergic AF, Skanes et al23 showed a strong correlation between the AFCL of periodic activations and DF, suggesting that a significant proportion of the mapped atria was activated in a 1:1 pattern. In contrast, high-density endocardial mapping studies performed in humans have found a poor correlation between high frequency sites detected by AFCL and DF analysis.24,25 Stiles et al24 looked at the spatial relationship of CFAE identified by the NavX algorithm (CFE mean ⬍120 ms) to the 20 highest DF sites obtained from a point-by-point endocardial map during AF. Using points taken from a 20-pole PentaRay catheter (4-mm interelectrode spacing; Biosense-Webster, Diamond Bar, CA) they observed a poor point-by-point correlation between CFAE (NavX algorithm) and DF. CFAE areas were located adjacent to areas of high DF, with 80% ⱕ10 mm from a high DF area. A more detailed analysis was limited by the size of the
Lee et al
Relationship Among Complex Electrograms, Short Cycle Length Activity, and High Dominant Frequency
recording bipole and the sequential nature of data acquisition. Human studies exploring the relationship between AFCL and DF using epicardial26 or endocardial mapping24,25 have yielded divergent results. A number of hypotheses have been advanced to explain the poor correlation seen in some studies. Among them, the presence of advanced remodeling with loss of 1:1 conduction,27,28 the presence of “endocardial– epicardial dissociation,”29 and the difficulty of determining local atrial activation rates when signals have complex morphology30 all may be important.
Study limitations Mapping of CFAE, AFCL, and DF was limited to the posterior LA, and whether the same relationships exist in other locations is unknown. The duration of persistent AF in this study is significantly longer compared to previous CFAE ablation series. It is unclear if progressive electroanatomic remodeling affects the spatial relationships described in this study. The temporal stability of CFAE and DF has been questioned recently,31 and it is possible that a longer recording time may be required to fully understand the relationship between these parameters. However, longduration recordings to guide ablation in clinical procedures may lack feasibility.
Conclusion The prevalence of CFAEs in the posterior LA varies markedly among patients and is highly dependent on the definition of CFAE adopted. There is poor anatomic overlap between CFAE defined by multicomponent/continuous EGM and CFAE defined by short CL activity. Areas of multicomponent/continuous EGMs are recorded adjacent to regions of high DF. Further studies are needed to determine the mechanisms responsible for these different signals.
References 1. Nademanee K, McKenzie J, Kosar E, et al. A new approach for catheter ablation of atrial fibrillation: mapping of the electrophysiologic substrate. J Am Coll Cardiol 2004;43:2044 –2053. 2. Oral H, Chugh A, Yoshida K, et al. A randomized assessment of the incremental role of ablation of complex fractionated atrial electrograms after antral pulmonary vein isolation for long-lasting persistent atrial fibrillation. J Am Coll Cardiol 2009;53:782–789. 3. Elayi CS, Verma A, Di Biase L, et al. Ablation for longstanding permanent atrial fibrillation: results from a randomized study comparing three different strategies. Heart Rhythm 2008;5:1658 –1664. 4. Verma A, Novak P, Macle L, et al. A prospective, multicenter evaluation of ablating complex fractionated electrograms (CFEs) during atrial fibrillation (AF) identified by an automated mapping algorithm: acute effects on AF and efficacy as an adjuvant strategy. Heart Rhythm 2008;5:198 –205. 5. Calkins H, Brugada J, Packer DL, et al. HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for personnel, policy, procedures and follow-up. A report of the Heart Rhythm Society (HRS) Task Force on catheter and surgical ablation of atrial fibrillation. Heart Rhythm 2007;4:816 – 861. 6. Roberts-Thomson KC, Stevenson IH, Kistler PM, et al. Anatomically determined functional conduction delay in the posterior left atrium relationship to structural heart disease. J Am Coll Cardiol 2008;51:856 – 862. 7. Oral H, Chugh A, Good E, et al. Radiofrequency catheter ablation of chronic atrial fibrillation guided by complex electrograms. Circulation 2007;115:2606 – 2612.
1719
8. Aizer A, Holmes DS, Garlitski AC, et al. Standardization and validation of an automated algorithm to identify fractionation as a guide for atrial fibrillation ablation. Heart Rhythm 2008;5:1134 –1141. 9. Sanders P, Berenfeld O, Hocini M, et al. Spectral analysis identifies sites of high-frequency activity maintaining atrial fibrillation in humans. Circulation 2005;112:789 –797. 10. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29 –36. 11. Kalifa J, Tanaka K, Zaitsev AV, et al. Mechanisms of wave fractionation at boundaries of high-frequency excitation in the posterior left atrium of the isolated sheep heart during atrial fibrillation. Circulation 2006;113:626 – 633. 12. Jais P, Haissaguerre M, Shah DC, Chouairi S, Clementy J. Regional disparities of endocardial atrial activation in paroxysmal atrial fibrillation. Pacing Clin Electrophysiol 1996;19:1998 –2003. 13. Tada H, Yoshida K, Chugh A, et al. Prevalence and characteristics of continuous electrical activity in patients with paroxysmal and persistent atrial fibrillation. J Cardiovasc Electrophysiol 2008;19:606 – 612. 14. Teh AW, Kalman JM, Kistler PM, et al. Prevalence of fractionated electrograms in the coronary sinus: comparison between patients with persistent or paroxysmal atrial fibrillation and a control population. Heart Rhythm 2010;7:1200 –1204. 15. Narayan SM, Wright M, Derval N, et al. Classifying fractionated electrograms in human atrial fibrillation using monophasic action potentials and activation mapping: evidence for localized drivers, rate acceleration, and nonlocal signal etiologies. Heart Rhythm 2011;8:244 –253. 16. Scherr D, Dalal D, Cheema A, et al. Automated detection and characterization of complex fractionated atrial electrograms in human left atrium during atrial fibrillation. Heart Rhythm 2007;4:1013–1020. 17. Wu J, Estner H, Luik A, et al. Automatic 3D mapping of complex fractionated atrial electrograms (CFAE) in patients with paroxysmal and persistent atrial fibrillation. J Cardiovasc Electrophysiol 2008;19:897–903. 18. Roux JF, Gojraty S, Bala R, et al. Complex fractionated electrogram distribution and temporal stability in patients undergoing atrial fibrillation ablation. J Cardiovasc Electrophysiol 2008;19:815– 820. 19. Rostock T, Rotter M, Sanders P, et al. High-density activation mapping of fractionated electrograms in the atria of patients with paroxysmal atrial fibrillation. Heart Rhythm 2006;3:27–34. 20. Spach MS, Dolber PC. Relating extracellular potentials and their derivatives to anisotropic propagation at a microscopic level in human cardiac muscle. Evidence for electrical uncoupling of side-to-side fiber connections with increasing age. Circ Res 1986;58:356 –371. 21. Konings KT, Kirchhof CJ, Smeets JR, Wellens HJ, Penn OC, Allessie MA. High-density mapping of electrically induced atrial fibrillation in humans. Circulation 1994;89:1665–1680. 22. Zlochiver S, Yamazaki M, Kalifa J, Berenfeld O. Rotor meandering contributes to irregularity in electrograms during atrial fibrillation. Heart Rhythm 2008;5: 846 – 854. 23. Skanes AC, Mandapati R, Berenfeld O, Davidenko JM, Jalife J. Spatiotemporal periodicity during atrial fibrillation in the isolated sheep heart. Circulation 1998;98:1236 –1248. 24. Stiles MK, Brooks AG, Kuklik P, et al. High-density mapping of atrial fibrillation in humans: relationship between high-frequency activation and electrogram fractionation. J Cardiovasc Electrophysiol 2008;19:1245–1253. 25. Elvan A, Linnenbank AC, van Bemmel MW, et al. Dominant frequency of atrial fibrillation correlates poorly with atrial fibrillation cycle length. Circ Arrhythm Electrophysiol 2009;2:634 – 644. 26. Schuessler RB, Kay MW, Melby SJ, Branham BH, Boineau JP, Damiano RJJ. Spatial and temporal stability of the dominant frequency of activation in human atrial fibrillation. J Electrocardiol 2006;39:S7–S12. 27. Sih HJ, Zipes DP, Berbari EJ, Adams DE, Olgin JE. Differences in organization between acute and chronic atrial fibrillation in dogs. J Am Coll Cardiol 2000; 36:924 –931. 28. Ryu K, Sahadevan J, Khrestian CM, Stambler BS, Waldo AL. Use of fast Fourier transform analysis of atrial electrograms for rapid characterization of atrial activation-implications for delineating possible mechanisms of atrial tachyarrhythmias. J Cardiovasc Electrophysiol 2006;17:198 –206. 29. de Groot NM, Houben RP, Smeets JL, et al. Electropathological substrate of longstanding persistent atrial fibrillation in patients with structural heart disease: epicardial breakthrough. Circulation. 2010;122:1674 –1682. 30. Ng J, Kadish AH, Goldberger JJ. Effect of electrogram characteristics on the relationship of dominant frequency to atrial activation rate in atrial fibrillation. Heart Rhythm 2006;3:1295–1305. 31. Habel N, Znojkiewicz P, Thompson N, et al. The temporal variability of dominant frequency and complex fractionated atrial electrograms constrains the validity of sequential mapping in human atrial fibrillation. Heart Rhythm 2010;7:586–593.