Case Presentation
A 63-year-old male without structural or ischemic heart disease has a history of longstanding paroxysmal palpitations that stop with vagal maneuvers. He underwent spinal fusion surgery for low back pain without complications. On post-operative day 2, atrial flutter (AFL) was recorded with a 2:1 response rate (Figure 1), easily seen after 6 mg of IV adenosine was administered. Initial medical therapy failed to control the episodes and/or the rate. An electrophysiology study was undertaken to study and ablate atrial flutter.
Clockwise isthmus-dependent right atrial flutter was induced with rapid right atrial lateral wall pacing (Figure 2). Counterclockwise atrial flutter was not inducible from the coronary sinus. This may have been due to difficulty capturing the atria with rapid pacing from multiple coronary sinus sites despite maximal pacing output. Induced sustained atrial flutter stopped with radiofrequency lesions delivered via an 8 mm tip catheter placed in the right atrial cavotricuspid isthmus. Subsequently, bidirectional block was not apparent but after delivery of several more isthmus lesions, a new tachycardia was noted.
This tachycardia was stopped and induced with rapid atrial pacing. The new tachycardia (cycle length = 290 ms) had a 1:1 AV relationship and appeared to be atrioventricular nodal reentry tachycardia (AVNRT) (Figure 3). This was proven with pacing maneuvers. Premature atrial extrastimuli dissociated the right and left atria from the tachycardia. Ventricular premature beats delivered when the His bundle was refractory did not reset the tachycardia or preexcite the atria, but 6 mg of adenosine IV stopped the tachycardia (Figure 4).
Figure 2.
Burst pacing from the Halo distal catheter induces clockwise atrial flutter.

An AV node function curve assessment indicated an 80 ms “jump” during delivery of 10 ms decremental atrial extrastimuli at a fixed cycle length. The tachycardia started with a jump in the AV node function curve. The ablation catheter was then placed in the location of the slow pathway near the posterior inferior septum. Delivery of a two-minute 8 mm lesion with adequate power (>60 watts) and temperature (>50˚ C) did not result in junctional beats but did eliminate the jump, echo beats and tachycardia initiation. Additional isthmus lesions delivered after the new tachycardia was rendered noninducible initiated bidirectional isthmus block. After 30 minutes, neither tachycardia was induced with and without isoproterenol, burst pacing or atrial extrastimuli at a fixed pacing cycle length. Bidirectional block persisted for at least 30 minutes.

Discussion
This case demonstrates three unique issues. First, the patient’s clinical flutter, most likely “typical” type I counterclockwise atrial flutter, was not inducible likely due to difficulty in 1:1 left atrial capture. Clockwise isthmus-dependent right atrial flutter was induced and ablated. Second, AVNRT was induced during cavotricuspid-isthmus ablation. Third, no junctional beats occurred during slow pathway ablation.
Figure 3.
Spontaneous AVNRT induced during isthmus ablation.

The most common type of atrial flutter (“typical” type I) is a macroreentry, anatomically-demarcated circuit with an excitable gap. It rotates counterclockwise around the tricuspid annulus and is bound by the crista terminalis, tricuspid valve annulus, the IVC and SVC orifices, the Eustachian ridge and the coronary sinus os. The circuit incorporates an “isthmus” that is critical for reentry to occur. The surface electrocardiogram often shows a “saw tooth” appearance (negative “F” waves in the inferior leads and an upright “F” wave in lead V1). A less common form of type I atrial flutter propagates in the opposite (clockwise) direction around the tricuspid annulus as seen in the left anterior oblique position. Uncommon type I flutter has upright “F” waves inferiorly.
Concealed entrainment, with pacing faster than the rate of the atrial flutter from the isthmus during tachycardia, can assess isthmus dependence. When pacing from the isthmus during tachycardia, the first post-pacing interval is at the tachycardia cycle length, and during pacing, the atrial activation is the same as during tachycardia.
The cavotricuspid isthmus appears to be a mutual electrical zone of slow conduction for both clockwise and counterclockwise AFL. Exit sites from the isthmus for clockwise flutter may differ versus counterclockwise atrial flutter. More than one form may exist. Ching-Tai and colleagues studied 30 patients with clockwise atrial flutter.1 Eighteen patients had both counterclockwise and clockwise flutters; 12 had only clockwise atrial flutter. Both forms of atrial flutter had similar cycle lengths (232 ± 30 versus 226 ± 25 msec, p = 0.526), and all exhibited concealed entrainment when pacing from the isthmus. Some patients with clockwise atrial flutter (n=20) had biphasic P waves with an exit site of slow conduction area located at the low posterolateral right atrium; 10 patients with positive P waves in the inferior leads had the presumed exit site located at the mid-high posterolateral right atrium. Linear radiofrequency lesions between the IVC and the tricuspid annulus eliminated clockwise atrial flutter in all patients.
Figure 4.
Adenosine blocks conduction of the tachycardia in the fast pathway.

However, non-isthmus dependent clockwise flutter can also occur. Shulong et al studied 12 patients with clockwise atrial flutter based on surface ECG tracings2 and proven with intracardiac recordings. Entrainment pacing was performed in 12 patients and electroanatomic mapping was performed in 7 patients, respectively. Clockwise lower loop reentry utilized the lower right atrium involving the IVC in 7 patients, figure-of-8 (double-loop) reentry around both the IVC and tricuspid annulus in 4 patients, and single reentrant loop around the tricuspid annulus in 1 patient. Radiofrequency ablation in the isthmus between the tricuspid annulus and IVC terminated tachycardia in all patients.
Other right atrial flutters (scar-mediated, lower loop reentry, upper loop reentry) are not dependent on the cavotricuspid isthmus.These flutters may passively activate the isthmus, but concealed entrainment from the isthmus is not seen. The isthmus may act as an innocent bystander for these flutters. Right atrial flutters show right to left atrial activation during atrial flutter as seen by coronary sinus atrial activation, but all right to left atrial activation flutters are not right sided.
AVNRT during ablation of flutter was unexpected. While these two distinct tachycardias generally do not share similar anatomic space, some reports suggest they do. AVNRT may utilize perinodal tissue extending beyond the slow and fast pathways. The slow pathway is postero-inferior and the fast pathway is antero-superior in their respective projections in the AV node.
The slow pathway and the isthmus responsible for atrial flutter do share similar or at least adjacent anatomic locations. Some investigators have called this the “shared space”, i.e., the low posterior right atrium near the coronary sinus os. Interian and colleagues evaluated 29 patients with supraventricular tachycardias thought to be AVNRT.3 Of these patients, 52% had AVNRT and atrial flutter both induced, and 47% had an ablation performed at the tricuspid annulus above the coronary sinus. Neither AVNRT nor atrial flutter could be induced after ablation. Repeat electrophysiology testing after successful ablation (mean = 6 days) did not demonstrate inducibility of either AVNRT or atrial flutter with or without isoproterenol. These investigators indicated that there does appear to be a “shared pathway” where both arrhythmias can be ablated at the same site at the same time. These results have been confirmed by other studies.4,5 We are not aware of data indicating the presence of inducible AVNRT after atrial flutter ablation.
Junctional rhythm during slow-pathway ablation has been the most sensitive but a nonspecific marker of successful ablation of AVNRT. Slow pathway ablation without junctional rhythm occurs, but is rare. Hsieh and colleagues evaluated 353 patients with AVNRT who underwent slow pathway ablation.6 In this study, 20/353 patients had a successful ablation site located in the posterior right atrium without junctional rhythm during slow-pathway ablation versus 200/353 patients with junctional rhythm during ablation (p < 0.001). Lack of junctional rhythm during slow pathway modifications was more common with fast-slow rather than slow-fast AVNRT (30% versus 3%; p = 0.001).

Conclusions
This case highlights several key findings regarding atrial flutter and AVNRT ablation. Typical type I atrial flutter is counterclockwise, but clockwise atrial flutter can also be isthmus dependent. Atrial flutter and AVNRT can share similar anatomic space and can coexist. Finally, absence of a junctional rhythm during attempted slow pathway modification is not necessarily evidence for failure, but is less common. The patient presented in this case had an uneventful course after ablation and has had no recurrent tachycardias.