This website is intended for healthcare professionals only.

Hospital Healthcare Europe
Hospital Healthcare Europe

Treating complex fractionated atrial electrograms

Leonardo Calo
1 January, 2008  

Leonardo Calò
Chief of Unit of Electrophysiology
Ermenegildo de Ruvo  

Luigi Sciarra

Ernesto Lioy
Division of Cardiology
Policlinico Casilino
Rome, Italy

Complex fractionated or fragmented atrial electrograms (CFAEs) are a new target of atrial fibrillation (AF) ablation, but their definition is still unclear. In fact, different types of electrograms are included in the term CFAE.

The term CFAE was used for the first time by Nademanee et al,(1) who defined CFAEs as follows:

  • Atrial electrograms with fractionated electrograms composed of two deflections or more, and/or perturbation of the baseline with continuous deflection of a prolonged activation complex over a 10-second recording period.
  • Atrial electrograms with a very short cycle length of <120 ms averaged over a 10-second recording period. Therefore, these authors included in CFAEs true continuous fragmented atrial electrograms and ­rapid atrial activity. Recently, Haissaguerre and colleagues(2) considered CFAEs to be fractionated ­potentials exhibiting multiple deflections from the isoelectric line (≥3 deflections) and/or potentials with ­continuous electrical activity without an isoelectric line.

Konings et al defined fragmented potentials as electrograms exhibiting more than two negative deflections within 50 ms.(3) The limit of 50 ms was based on the assumption that the atrial refractory period during AF and the interval between two successive fibrillation waves were ≥50 ms.

Rapid atrial activity (long double potentials, according to Konings’ definition) is generally ­characterised by sharp atrial electrograms with a very short cycle length (<120 ms in Nademanee's definition). It may be the expression of collision/overlap of different wavelets entering the same area, or may be determined by focal sources. It is important to underline that Konings et al conducted a unipolar mapping, while most endocardial studies utilised bipolar recordings. In fact, while bipolar electrograms are usually preferred because they are less sensitive to noise, they are the result of the phase difference between the activation of two poles. Thus, in the case of AF, where the direction of propagation changes ­continuously, bipolar recordings show a higher spatiotemporal variation in morphology than unipolar signals.

Little information exists on the spatial distribution of CFAEs in the left and right atrium. Nademanee et al(1) mainly localised CFAEs in the interatrial septum, proximal coronary sinus, pulmonary veins, mitral annulus and left atrial roof. Previous human mapping studies(4–6) have found the septum as the area where CFAEs were mostly present. Oral et al,(7) using Nademanee’s definition of CFAEs, have observed the highest prevalence of CFAEs in the anterior wall of the left atrium and in the pulmonary vein antrum or ostium.

CFAE definition
But what are CFAEs? CFAEs are thought to indicate areas of nonuniform wavelet propagation and slowed conduction, pivot points around areas of functional and anatomical blocks that may be implicated in arrhythmia perpetuation.(3) The fragmentation of atrial electrograms can be determined by:

  • Conduction velocity (eg, area of fibrosis, tissue anisotropy).
  • Spatial dispersion of refractory periods: in the areas with longer refractory periods, the action ­potentials may not be able to be achieved when the activation rate encroaches on the refractory periods – in other words, some areas are able and other areas are not able to maintain 1:1 capture.
  • Presence of specific anatomical characteristics (areas of wide muscle connections favouring ­wavefront collision, or presence of anatomical barriers between atrial muscle bundles).

Furthermore, CFAE can be determined by high-frequency sources (mother rotors). In fact, mapping of AF in sheep models has demonstrated the presence of CFAEs around areas of high-frequency excitation.(8) The fast centrifugal propagation of the wavefront from the core of a rotor may determine wave break and fibrillatory conduction. Therefore, it could be hypothesised that CFAEs are adjacent to the AF driver and can favour the maintenance of AF by multiple re-entry.

This hypothesis is also supported by the observation that the occurrence of fragmented electrograms is significantly associated with a preceding shortening of FF interval, and the duration of ­complex atrial fragmented electrograms is inversely correlated with the preceding FF interval (shorter FF interval, greater duration of CFAEs).(2) Finally, it has been proposed that CFAEs are areas of presence of ­ganglionated plexi, where the local vagal stimulation shortens the local effective refractory periods and causes the formation of CFAEs.(9,10)

CFAE mapping
The importance of CFAE mapping is determined by the seminal observation by Nademanee and ­colleagues that selective CFAE ablation may terminate AF and prevent its recurrence. In the course of ­ablation, the rhythm often converts from AF to an organised macro- or micro-reentrant atrial ­tachycardias (ATs) or atrial flutters (AFLs) in three-quarters of patients. One-third of nonparoxysmal AF needs to repeat ablation because of AT and AFL recurrencies. Also, in a stepwise approach conducted in patients with a long-lasting persistent AF,(11) conversion to sinus rhythm occurred during ablation in 87% of patients, but ablation of CFAEs was only a part of a sequence that included empirical PV isolation and a variety of left atrial linear lesions. Conversely, in a study conducted by Oral et al in 100 patients with chronic AF, only selective ablation of CFAE determined conversion to sinus rhythm in only 12% of patients. These findings could douse any enthusiasm for using CFAEs as target sites for AF substrates, even if the aforementioned studies presented relevant methodological differences in the mapping ­procedure, ablation settings and procedural endpoints.

Thus, the possibility to use an advanced mapping system not only to represent the atrial anatomy but also to efficiently map the chambers during AF and rapidly identify CFAE sites has become one promising approach. In the past we tagged fragmented areas during electroanatomic mapping; and, when possible, the lesions were performed both in left and right atrium along these sites.(12) Currently, we are able to identify CFAEs using the CFAE tool algorithm (CARTO XP, Biosense Wester). It is a ­powerful tool that can be customised to identify different atrial patterns of activation during AF. By modifying preloaded settings it is possible to recognise continuous fragmented electrograms and rapid atrial activity.

CFAE ablation is a suitable and appealing way to treat persistent and permanent AF. While pulmonary veins are the most important target for paroxysmal AF ablation, substrate ablation may be a new option to treat long-lasting AF. Pulmonary veins often remain one of the key areas, after the ­septum, the coronary sinus ostium with its proximal part and the left atrial roof. All these areas need to be ablated to achieve conversion to sinus rhythm. In fact, according to this hypothesis, if such areas are selectively eliminated by catheter ablation, wavelet re-entry would stop, thereby preventing perpetuation of AF. Therefore, in each patient a tailored radiofrequency ablation of specific areas where CFAEs are present would be desirable. Procedural times may be longer than pulmonary vein techniques because of the need for an extensive biatrial mapping and atrial tachycardia recurrences, but new tools have been developed to shorten the length of AF ablation mapping. One of the difficulties in trying to reproduce these results is the relative subjectivity inherent in defining whether a particular electrogram is “complex” enough. In an effort to standardise the definition of a CFAE site, there has been a significant amount of work in developing signal-processing software to analyse atrial electrograms during AF.


  1. Nademanee K, et al. J Am Coll Cardiol 2004;43:2044-53.
  2. Rostock T, et al. Heart Rhythm 2006;3:27-34.
  3. Konings KTS, et al. Circulation 1997;95:1231-41.
  4. Jais P, et al. Regional disparities of endocardial atrial activation in paroxysmal AF. PACE. 1996;19 (Pt II): 1998-2003.
  5. Gaita F, et al. Circulation 1998;97:2136-45.
  6. Gaita F, et al. J Am Coll Cardiol 2001;37:534-41.
  7. Oral H, et al. Circulation 2007;115:2606-12.
  8. Kalifa J, et al. Circulation 2006;113:626-33.
  9. Scherlag BJ, et al. J Am Coll Cardiol 2005;45:1878-86.
  10. Hou Y, et al. J Am Coll Cardiol 2007;50:61-68.
  11. Haissaguerre M, et al. J Cardiovasc Electrophysiol
  12. Calò L, et al. J Am Coll Cardiol 2006;20;47(12):2504-12.2005;16:1125-37.