Response to Adenosine Differentiates Focal From Macroreentrant Atrial Tachycardia
Validation Using Three-Dimensional Electroanatomic Mapping
Background— We previously proposed that adenosine has mechanism-specific effects on atrial tachycardia (AT), such that adenosine terminates AT attributable to triggered activity, transiently suppresses automatic rhythms, and has no effect on macroreentrant AT. This, however, remains controversial, because other studies have reported that adenosine terminates reentrant AT. To clarify this issue, we used 3D electroanatomic mapping to delineate the tachycardia circuit and thereby determine whether the response to adenosine differentiates focal from macroreentrant AT.
Methods and Results— We examined the effect of adenosine on 43 ATs in 42 consecutive patients (59±15 years of age; 26 female) who received adenosine during tachycardia and whose mechanism of AT was characterized by pharmacological perturbation, entrainment, 3D electroanatomic mapping, and results of radiofrequency ablation. Eight tachycardias were macroreentrant (noncavotricuspid isthmus-dependent), and 35 ATs were focal (either triggered or automatic). Adenosine administered during AT (at doses sufficient to result in AV block) terminated or transiently suppressed focal AT in 33 of 35 cases, whereas 8 of 8 macroreentrant ATs were adenosine insensitive (P<0.001). Twenty-eight of 35 focal ATs were located along the crista terminalis or tricuspid annulus.
Conclusions— The response of AT to adenosine can immediately differentiate atrial tachycardia arising from a focal source from that attributable to macroreentry. This finding can be exploited to facilitate developing a focused, strategic ablative approach at the onset of a procedure.
Received July 31, 2002; revision received September 13, 2002; accepted September 16, 2002.
Adenosine is an endogenous compound that acts on supraventricular tissue by 2 cellular mechanisms: activating IKACh,Ado to hyperpolarize the membrane potential toward the potassium equilibrium potential (≈−90 mV) and shortening action potential duration by increasing potassium conductance.1 The latter effect decreases the wavelength of activation and therefore would be expected to have a proarrhythmic rather than antiarrhythmic effect on reentrant arrhythmias. Adenosine also has indirect antiadrenergic effects, decreasing intracellular cAMP, resulting in inhibition of the L-type calcium current (ICa(L)) as well as the transient inward current (ITi).2 These effects are consistent with adenosine-mediated termination of atrial tachycardia (AT) attributable to cAMP-dependent triggered activity and transient suppression of AT attributable to enhanced automaticity.
AT can be focal in origin or macroreentrant, involving conduction around anatomic, incisional, or functional barriers. We previously proposed that adenosine has mechanism-specific effects on atrial tachycardia,3,4⇓ such that it terminates or suppresses focal AT (attributable to triggered activity or enhanced automaticity, respectively) and has no effect on macroreentrant AT (with the rare exception of that involving partially depolarized tissue).5 The effect of adenosine on reentrant AT remains controversial, however, because other studies6,7⇓ have reported termination with adenosine in up to 90% of patients.6 Therefore, we sought to clarify the effect of adenosine on AT, using 3D electroanatomic mapping to more precisely delineate the tachycardia circuit.
Forty-two consecutive patients (59±15 years of age; 26 female) who presented for electrophysiological evaluation and radiofrequency ablation of AT and who received adenosine during tachycardia comprise this series, including 8 patients with noncavotricuspid isthmus-dependent macroreentrant AT and 34 with focal AT (1 patient received adenosine during 2 different focal ATs). Their baseline characteristics are listed in the Table. Three-dimensional electroanatomic mapping (CARTO, Biosense-Webster) was performed in all cases as described previously.8
Structural heart disease was present in 23 (55%) patients. Thirty-two (76%) patients were taking antiarrhythmic agents or AV nodal blocking agents at the time of the procedure. These included β-blockers (n=19), calcium channel blockers (n=15), digoxin (n=14), and antiarrhythmic agents (amiodarone=2; sotalol, flecainide, quinidine, n=1 each).
Baseline Electrophysiological Study
This study was approved by our institutional review board. After giving informed written consent, patients underwent electrophysiological testing after an overnight fast. Patients were locally anesthetized with 0.25% bupivacaine and sedated with intravenous midazolam and morphine. Quadripolar 6F catheters (Bard Electrophysiology Division, C.R. Bard, Inc or Daig Corp) were advanced to the His bundle position and right ventricular apex. Right atrial electrogram recordings were obtained with either a quadripolar catheter positioned in the high right atrium (RA) or a 7F duodecapolar Halo catheter (Cordis Webster) positioned along the tricuspid annulus. A 6F decapolar catheter (Daig Corp) was positioned in the coronary sinus to record left atrial activity. Bipolar intracardiac electrograms were filtered at 30 to 500 Hz and recorded on optical disk (Prucka Engineering). If additional left atrial mapping or ablation was required, a transseptal atrial puncture was performed.
The stimulation protocol included rapid atrial and ventricular pacing and introduction of multiple atrial and ventricular extrastimuli at several basic drive cycle lengths. Stimuli were delivered as rectangular pulses of 2-ms duration at 4 times diastolic threshold. To facilitate induction of sustained AT, when necessary, programmed stimulation was repeated after isoproterenol or dobutamine was infused to decrease the sinus cycle length by ≈30%.
Electroanatomic mapping was performed with a reference locator pad (on the patient’s back) for spatial reference and a bipolar intracardiac electrogram (from the coronary sinus) used as temporal reference. A 7F deflectable, 4-mm-tip quadripolar catheter (Biosense-Webster) was used for activation mapping and ablation.
Atrial Tachycardia Diagnosis
AT was distinguished from other supraventricular tachycardias, including atrioventricular (AV) nodal reentry and AV reciprocating tachycardia, by standard electrophysiological criteria, as described in detail previously.4
Focal AT was defined on the basis of centrifugal atrial activation pattern, dissociation of nearly the entire atria from the tachycardia with atrial extrastimuli, early local atrial activation relative to the surface P wave, and inability to entrain the tachycardia. Annular focal AT was identified when the above criteria were met, fluoroscopic and 3D electroanatomic sites were consistent with an annular site, and atrial and ventricular electrograms were present simultaneously, with an atrial to ventricular ratio between 0.2 and 2.0 at the site of successful ablation.
Focal AT consistent with enhanced automaticity had the following characteristics3: spontaneous initiation or termination, failure to initiate with programmed stimulation, and demonstration of warm-up and cool-down phenomena. In contrast, focal AT consistent with triggered activity demonstrated reproducible initiation and termination with programmed stimulation, as well as sensitivity to adenosine and verapamil (when administered).
Macroreentrant AT was defined by demonstration of concealed or manifest entrainment of the tachycardia with atrial pacing, activation mapping using CARTO identifying macroreentry, and continuous activation of the atria during diastole at multiple adjacent sites. Cavotricuspid isthmus-dependent atrial flutter was excluded by atrial activation pattern and by demonstrating that the cavotricuspid isthmus was not a critical part of the tachycardia circuit.
Adenosine (Adenocard; Fujisawa) was administered as a rapid bolus through a central venous catheter, followed by a 10-mL flush of normal saline. The initial dose of adenosine was 3 to 6 mg, with the dose titrated incrementally by 3 to 6 mg until AT terminated or was suppressed or AV block occurred. Response of AT to adenosine was defined as sensitive, insensitive, or nonspecific (induction of atrial fibrillation [AF]). Sensitivity was additionally subdivided into termination (occurring in the absence of an atrial premature complex [APC]) or transient suppression (associated with spontaneous reinitiation within 20 seconds). Patients were excluded if termination was demonstrated solely because of an APC. Verapamil (5 to 20 mg IV) was also infused in 7 patients to determine its effect on AT.
Radiofrequency (RF) energy was applied using a 7F 4-mm-tip ablation catheter with temperature feedback control to a target of 60°C and maximal-allowed power output of 50 W. RF energy was administered for up to 60 seconds per application and was terminated if significant chest discomfort, catheter movement, or impedance rise occurred.
Results are presented as mean±SD where appropriate. Comparison of focal and macroreentrant ATs was performed using the Mann-Whitney test (for adenosine dose), Fisher exact test (for comparing sex distribution, prevalence of heart disease, and likelihood of being on cardiac medications), and two-tailed t test for independent samples (age and tachycardia cycle length). P≤0.05 was considered statistically significant.
In 35 of 43 cases, tachycardia was initiated with programmed stimulation (5 of these patients required concurrent infusion of isoproterenol or dobutamine). Eight patients presented to the electrophysiology laboratory in incessant AT (without termination of AT before ablation) or had recurrent spontaneous episodes of AT.
Focal Atrial Tachycardia
Focal AT was induced with rapid atrial pacing or atrial extrastimuli in 32 of 35 cases, 3 of whom also had AT induced with ventricular pacing or ventricular extrastimuli (in the presence of VA conduction). In 5 patients, concurrent infusion of catecholamine (isoproterenol or dobutamine) was used to facilitate induction of AT. The mean AT cycle length was 435±88 ms (range, 260 to 630 ms). Tachycardia characteristics are included in the Table.
A focal origin of AT was identified in all patients on the basis of centrifugal activation of the atria from early sites, which was confirmed with electroanatomic mapping. Six arose from the crista terminalis, 22 from the tricuspid annulus (including 6 from the triangle of Koch), 3 from the mitral annulus, 2 from the septal RA, and 1 each from the right atrial appendage and posterior RA. Entrainment could not be demonstrated in any patient. Adenosine terminated focal AT in 31 of 35 cases, transiently suppressed AT in 2 patients (for <20 seconds), and resulted in AF in 1 patient. One focal AT arising from the posteroseptal RA was insensitive to 6 mg of adenosine (despite demonstration of AV block). Verapamil terminated AT in 6 of 6 patients, all of whom were adenosine-sensitive.
Fractionated potentials were observed at sites of earliest atrial activation in 4 of 35 ATs (one each from the tricuspid annulus, crista terminalis, interatrial septum, and mitral annulus), all of which were adenosine-sensitive. RF ablation was attempted in 30 cases of focal AT and was successful in 28 of 30 (93%). In 1 patient (No. 11) with a para-Hisian focus, the procedure had to be terminated prematurely because of sedative-induced patient agitation. Figure 1 (patient No. 7) represents a focal, adenosine- and verapamil-sensitive AT arising from the triangle of Koch, contiguous to the His bundle. Figure 2 represents a focal AT from the anterior mitral annulus (patient No. 34).
Macroreentrant Atrial Tachycardia
Macroreentrant AT was induced with atrial extrastimuli in 3 of 8 patients. Five patients presented in incessant tachycardia. The mean tachycardia cycle length was 321±70 ms (range, 240 to 440 ms). Five ATs involved the RA, and 3 involved the left atrium. The tachycardia circuits are delineated in Figure 3; AT characteristics are listed in the Table. Entrainment was confirmed in 7 of 8 patients. Entrainment could not be demonstrated in 1 patient because AT repeatedly terminated during atrial pacing. Electroanatomic activation mapping confirmed macroreentrant circuits (with 90±8% of the tachycardia cycle length mapped). A representative example of adenosine-insensitive AT is shown in Figure 4, in a patient (No. 39) who underwent repair of tetralogy of Fallot and ostium primum atrial septal defect with a figure-of-eight reentrant AT involving a critical isthmus between an atriotomy scar and the inferior vena cava. Concealed entrainment was demonstrated from within this isthmus (yellow dot). A linear lesion across the isthmus terminated AT. Another example of adenosine-insensitive macroreentry is shown in Figure 5, which depicts the RA activation sequence in a patient (No. 38) with Ebstein’s anomaly and atrial septal defect repair who developed AT. The circuit was also a figure-of-eight, with a critical isthmus between the atriotomy scar and tricuspid annulus. A linear RF lesion between the scar and annulus terminated AT.
Adenosine failed to terminate macroreentrant AT in all 8 patients; adenosine induced AF in 1 patient. Verapamil was given to 1 macroreentrant AT and had no effect. RF ablation was successful in 5 of 7 (71%) patients.
Comparison of Groups
Patients with macroreentrant AT had a higher likelihood of structural heart disease than those with focal AT (88% versus 47%; P=0.05). The cycle length of macroreentrant AT was significantly shorter than that of focal AT (321 versus 435 ms; P=0.001). There was no significant difference in age (55±18 versus 60±15 years, respectively; P=NS) or sex (62% versus 63% female) between the 2 groups. The dose of adenosine administered to patients with macroreentrant AT was higher than the dose given to those with focal AT (10.1±2.7 mg versus 6.7±4.1 mg, respectively; P=0.001).
The principal finding in this study is that sensitivity to adenosine predicts whether an AT is focal or macroreentrant. Focal AT terminates or suppresses in response to adenosine, whereas macroreentrant AT is adenosine insensitive. The response to adenosine allows for the rapid determination of whether an AT is focal, facilitating the strategic ablative approach at the onset of a procedure.
Focal Atrial Tachycardia: “Ring of Fire”
In our series, most focal ATs arose from either the crista terminalis or tricuspid annulus. The crista terminalis has long been appreciated as a “ring of fire,” reportedly accounting for most focal right ATs. Kalman et al9 described the presence of an early and fractionated potential at the site of successful ablation in most of the ATs from this region and hypothesized that, because the crista terminalis is an area with marked anisotropic conduction, it may be susceptible to microreentry. Clinically, however, definitive evidence for microreentry in the atria does not exist, and the pharmacological response of these ATs to adenosine is consistent with either calcium-dependent microreentry or cAMP-dependent triggered activity. In addition, we found fractionated potentials at the site of successful ablation in only a minority (4 of 35) of focal ATs.
The tricuspid annulus has recently been described as another source of focal ATs10,11⇓ and accounts for >50% of focal right ATs presenting for evaluation in our laboratory. Periannular cells are histologically similar to atrial cells but resemble nodal cells in their electrophysiological characteristics and by their lack of connexin-43.12 Because adenosine has a constellation of unique effects in periannular atrial tissue, including decreasing the upstroke velocity, amplitude, and duration of the action potential, it could potentially terminate triggered activity or calcium-dependent microreentry through its direct effect on action potential duration (shortening) or indirectly via its antiadrenergic properties, each of which reduces intracellular calcium.
The triangle of Koch has been reported to be the source of a discrete form of atrial tachycardia.13,14⇓ These ATs resemble the fast/slow form of AV node reentry with respect to ECG characteristics and electrophysiological properties. Earliest atrial activation is recorded from the low anteroseptal RA, and these tachycardias are sensitive to adenosine and verapamil. However, the tachycardia persists despite conduction block in the AV node. Figures from the sites of successful ablation, although not commented on in these previous reports, reveal atrial and ventricular electrograms, consistent with an annular site, similar to the 6 patients in our series who had this form of AT. On the basis of these data, it is likely that most triangle of Koch ATs should be regarded as a subset of a much broader group of tachycardias more appropriately classified as tricuspid annular AT.
Macroreentrant Atrial Tachycardia
Incisional or lesional atrial tachycardias have been reported after surgery for congenital heart disease15 and mitral valvular disease.8 Most patients in this series who presented with macroreentrant AT had underlying structural heart disease and had undergone cardiac surgery requiring atriotomy. Surgical scars serve as anatomic conduction barriers, along with native structures (eg, AV valves, inferior and superior vena cavae, and pulmonary veins), contributing to the substrate for stable macroreentry. Conceptually, this is analogous to cavotricuspid isthmus-dependent atrial flutter, in which the tricuspid valve, coronary sinus, fossa ovalis, crista terminalis, and inferior vena cava form barriers to conduction. We have previously demonstrated that atrial flutter is insensitive to adenosine.3 Like atrial flutter, incisional ATs do not terminate with adenosine. The exception to this rule is the rare form of macroreentrant AT involving tissue with decremental conduction properties. These circuits are likely comprised of partially depolarized atrial tissue that, as a consequence, behave like AV nodal tissue.16
It has been previously reported that the response of AT to adenosine (and verapamil) does not distinguish between tachycardia attributable to triggered activity from that attributable to reentrant AT.6,7⇓ However, criteria used to define reentry frequently did not exclude triggered activity as the etiology of arrhythmia. Progressive fusion was not always demonstrated (only constant fusion).6 In addition, inducibility with programmed stimulation is a relatively nonspecific response and only excludes automaticity. Furthermore, an inverse relationship between the pacing cycle length or coupling interval and the interval between the last paced beat and the first beat of AT can also be observed with arrhythmias attributable to triggered activity, possibly because of induction of outward sodium current.17 Thus, some ATs may have been misclassified as reentrant, despite being attributable to triggered activity. Also, if adenosine termination occurred because of an APC (a nonspecific response), this would also have skewed the results.
This series includes only patients who received adenosine during tachycardia, with confirmation of their circuit by electroanatomic mapping. Patients with nonsustained AT were not given adenosine because of difficulty in interpreting the response and were not included in this study. Therefore, no definitive conclusions can be made regarding the relative prevalence of one form of AT compared with another.
It is possible that higher doses of adenosine may have terminated macroreentrant AT. However, this is unlikely, because the mean dose of adenosine given to macroreentrant AT was higher that that for focal AT and there was evidence of adenosine effect on the AV node. Finally, three-dimensional electroanatomic mapping techniques do not presently have the resolution to distinguish between triggered activity and microreentry, both of which can appear as focal ATs.
Adenosine is useful in distinguishing focal from macroreentrant AT. Therefore, the response of AT to adenosine can be helpful in developing a focused, strategic ablative approach at the onset of a procedure, with initial mapping of adenosine-sensitive AT performed in the regions of the crista terminalis and tricuspid annulus (including the triangle of Koch). Insensitivity to adenosine suggests a macroreentrant circuit. Previous conflicting data may have been attributable to the nonspecific effects of adenosine (ie, induction of AF or termination by adenosine-induced premature atrial beats) or to incomplete delineation and demonstration of macroreentry.
This work was supported in part by grants from the National Institutes of Health (RO1 HL-56139), the American Heart Association Grant-in-Aid (New York City Affiliate), the Maurice and Corinne Greenberg Arrhythmia Research Grant, the Raymond & Beverly Sackler Foundation, and the Michael Wolk Foundation.
- ↵Belardinelli L, Isenberg G. Isolated atrial myocytes: adenosine and acetylcholine increase potassium conductance. Am J Physiol. 1983; 224: H734–H737.
- ↵Lerman BB, Belardinelli L. Cardiac electrophysiology of adenosine: basic and clinical concepts. Circulation. 1991; 83: 1499–1509.
- ↵Engelstein ED, Lippman N, Stein KM, et al. Mechanism-specific effects of adenosine on atrial tachycardia. Circulation. 1994; 89: 2645–2654.
- ↵Chen SA, Chiang CE, Yang CJ, et al. Sustained atrial tachycardia in adult patients: electrophysiological characteristics, pharmacological response, possible mechanisms, and effects of radiofrequency ablation. Circulation. 1994; 90: 1262–1278.
- ↵Glatter KA, Cheng J, Dorostkar P, et al. Electrophysiologic effects of adenosine in patients with supraventricular tachycardia. Circulation. 1999; 99: 1034–1040.
- ↵McGuire MA, deBakker JM, Vermeulen JT, et al. Atrioventricular junctional tissue: discrepancy between histological and electrophysiological characteristics. Circulation. 1996; 94: 571–577.
- ↵Roos-Hesselink J, Perlroth MG, McGhie J, et al. Atrial arrhythmias in adults after repair of tetralogy of Fallot: correlations with clinical, exercise, and echocardiographic findings. Circulation. 1995; 91: 2214–2219.
- ↵Jalife J, Delmar M, Davidenko JM, et al. Basic mechanisms of cardiac arrhythmias. In: Basic Cardiac Electrophysiology for the Clinician. Armonk, NY: Futura Publishing; 1999: 212.