Advances in Atrial Flutter Mapping:

Several investigators have identified critical components of the atrial flutter reentrant circuit.1–8 Careful analysis of atrial activation patterns, specific interventions (pacing and ablation), and confirmatory data advanced knowledge of atrial flutter even further.9–13 Now, based on detailed, sophisticated, and effective three-dimensional mapping technologies, understanding of activation patterns during atrial flutter is even more refined. Our insights have expanded.14–17 Present tools identify the involved pathways, required boundaries, conduction properties, and refractory periods, similar and different, from patient to patient, in atrial flutter. Three-dimensional noncontact mapping may advance our knowledge further, but do we need to know more? Does this mapping approach help us to understand atrial flutter, and is it of practical value or importance? Typical atrial flutter involves (1) anatomic reentry around the tricuspid annulus with an excitable gap and an area of slow conduction4,15,18; (2) activation of smooth and roughened portions of the right atrium11,19; (3) functional or fixed conduction delay or block in the crista terminalis, sinus venosa region, and/or eustachian ridge20–23; and (4) a quiescent left atrium. The entire right atrium is not required for flutter propagation. Variability in conduction is possible. The mysterious relationship between fibrillation and flutter may be understood better by delineating the structures involved in reentry. The study by Tai et al.14 in this issue of the Journal involves 26 patients with “typical” atrial flutter who undergo noncontact three-dimensional mapping in preparation for ablation of isthmus-dependent (20 clockwise and 6 counterclockwise) atrial flutter. Three activation patterns were identified: (1) conduction block in the crista terminalis; (2) conduction around and through the crista terminalis as a single incomplete line of block (the majority); and (3) activation around double lines of block, including the sinus venosa region. Because this study was small, even more than three activation patterns are conceivable. Double loop reentry, including reentry through the crista terminalis (possibly passive, and not a driver of the circuit), occurred in 12 of 26 patients. These widely different activation patterns are somewhat surprising, as atrial flutter on the surface ECG and based on years of standard intracardiac mapping techniques appears monotonously similar. After cavotricuspid isthmus ablation, upper loop reentry was induced, ablated, and “cured” by radiofrequency applications to the crista terminalis through which conduction occurred in half the patients. The data suggest that upper loop reentry using tissue in the crista terminalis is possible and common.14 One of six patients with a complete line of conduction block in the crista terminalis had history of atrial fibrillation compared with four of seventeen with conduction through the crista terminalis. Two of three patients with conduction block in the crista terminalis and sinus venosa region had atrial fibrillation. The data are intriguing because the report provides new insight into right atrial activation during atrial flutter. The data also are perplexing and of concern. Several questions arise. (1) Do the data apply to patients with typical atrial flutter? If not, then for which patients do they apply? (2) Should the approach to atrial flutter ablation change based on this? (3) Are the data valid? (4) Is the crista or sinus venosa involved in atrial flutter propagation? (5) Are atrial flutter and fibrillation related by conduction patterns in the crista? (6) Are conduction properties of the crista terminalis unique to those who have atrial flutter? Answers to these questions are not yet apparent. Ablation of the cavotricuspid isthmus, one of the many success stories of ablation, is effective and safe to stop, and prevent typical atrial flutter in most patients. Recurrent flutter, although uncommon, is generally due to return of isthmus conduction24 or misinterpretation of the initial data, not due to development of upper loop reentry. Prior data from Tai's own laboratory confirm this to be the case.25 To the contrary, now Tai et al. note that patients with isthmus-dependent flutter commonly require ablation at the crista terminalis as well as the cavotricuspid isthmus based on flutter induction and activation maps. This need is inconsistent with extensive experience. It even is inconsistent with their own data suggesting that 4.2% (not >50% as stated here14) of those with typical atrial flutter have atypical atrial flutter recurrence during even longer follow-up than in the present study (29 ± 17 months vs 8.4 ± 4.1 months).26 Some, not all, will have upper loop reentry.27 Inducible atypical atrial flutter alone may be of little importance; however, the 30% occurrence of atrial fibrillation after atrial flutter ablation is important.28,29 Many studies support the resounding success of isthmus ablation for typical atrial flutter. Bidirectional cavotricuspid isthmus block generally cures the problem. Do we really need to do more? Do we really need noncontact mapping to ensure success? Upper loop reentry can be a real problem, but not for most patients with typical atrial flutter. This report14 does not provide comparative (randomized to crista vs no crista applications) or long-term data supporting the need for prevention of this potential form of reentry. Elimination of the reentrant circuit responsible for upper loop reentry may not be needed. These patients did not have manifest upper loop reentry. Ablation of this arrhythmia is like curing a patient of a disease he or she did not have. Friedman et al.21 assessed right atrial activation in 28 patients with counterclockwise (n = 20), clockwise (n = 3), or clockwise and counterclockwise (n = 5) atrial flutter. In all patients, a line of functional (not fixed) block was observed in the posteromedial, sinus venosa, region of the right atrium, manifested by the presence of double potentials. Conduction block was not present at the crista terminalis. Other data from the laboratory of Tai et al. and in a prior manuscript suggest that functional conduction abnormalities in the crista terminalis may be associated with typical atrial flutter.20 Tai et al. evaluated conduction properties of the crista terminalis in 12 patients with typical atrial flutter compared to 12 patients without clinically apparent but with inducible atrial flutter. During sinus rhythm, no double potentials were recorded along the crista terminalis in either group, but conduction properties varied between groups. Those with clinically apparent atrial flutter had poor transverse conduction in the crista terminalis. This finding has been confirmed by other investigators.30,31 Using a similar noncontact mapping technique, Chen et al.22 assessed activation in the right atrium in 39 patients with typical atrial flutter. A single line of block at the crista terminalis was found in 18 patients. Two lines of block at the crista terminalis and the sinus venosa were found in 20 patients, and one line of block at the lateral wall in 1 patient. Block was observed in the lower portion of the crista terminalis in atrial flutter and during pacing. Chen concluded that lines of block may form the posterior boundaries of the flutter circuit. Block in the lower part of the crista terminalis was fixed while the upper part was functional. There is still much to learn about right atrial activation in typical atrial flutter. Tai et al.14 stated that the “direction of the activation wavefront demonstrated by noncontact mapping was consistent with the activation sequence recorded by the Halo catheter in all patients; ” however, we need to take the investigators at their word. No electrogram data are presented. The investigators had an excellent opportunity to validate results of their noncontact mapping by providing real electrograms and comparing them to virtual electrograms. The data in this report are based on noncontact mapping. This technique may be accurate, correct, and validated in small patient numbers,32 but concerns remain. Depending on the location of the balloon, the extent of atrial mapping, and the methods to interpret the data, results can vary widely. The figures, such as their Figures 5 and 8, are open to interpretation and may not show what is stated. Based on these data, the approach to atrial flutter ablation would require substantial change, including the need for noncontact or other forms of detailed mapping and attempts at induction of atypical forms of atrial flutter. I am not yet convinced. The implication of the double-potential virtual electrograms is somewhat problematic. Assuming the validity of the signals to show activation timing, the use of unipolar virtual signals to indicate a line of block by the presence of double potentials is not substantiated but is extrapolated from recordings of double potentials obtained by other methodologies.3,5 Considering that double-potential recordings by a bipole touching heart muscle is open to query regarding its meaning, can virtual recordings provide indisputable evidence for conduction block? Data from Tai et al.14 are not strong enough to recommend abandonment of well-founded standard approaches to atrial flutter ablation in lieu of potential quixotic adventures that ultimately may cause patient harm. Isthmus ablation of typical atrial flutter is generally successful. The authors agree to that. Their data14 are inconsistent with extensive experience and may lead to utilization of unnecessary and potentially harmful mapping and ablation techniques that could create a new proarrhythmic circuit or damage normal cardiac and extracardiac structures. Data from the study by Tai et al.14 are highly illuminating. They shed new light on what we know and what we do not know about atrial flutter. We need to know more about global atrial conduction, atypical flutter, and the relationship of atrial flutter to atrial fibrillation. Then, we will be positioned better to care for patients with atrial flutter and atrial fibrillation. This study enhances our understanding of atrial flutter. Alone, it is not enough evidence to change a successful, established, ablation approach.