Interesting electrophysiological findings in a patient with coexistence of atrial tachycardia originating from coronary sinus and slow-fast atrioventricular nodal reentrant tachycardia

Abstract

Slow-fast atrioventricular nodal tachycardia (AVNRT) has various electrophysiological aspects due to atrioventricular (AV) nodal physiology. In addition, concomitantly another form of arrhythmia with AVNRT, especially atrial tachycardia (AT), was an infrequent arrhythmia. A 38-year-old female with narrow QRS tachycardia underwent electrophysiological study due to frequent faintness. The electrophysiological study disclosed the coexistence of AT originating from coronary sinus (CS) with slow-fast AVNRT. We easily diagnosed AT originating from CS and terminated with several radiofrequency ablations (RFA) around CS. The diagnosis of slow-fast AVNRT, however, was somewhat difficult due to the following findings: (1) small amount of adenosine triphosphate (ATP) could terminate slow-fast AVNRT reproducibly; (2) we could provoke slow-fast AVNRT only by RV pacing with isoproterenol infusion. With other electrophysiological findings, we diagnosed slow-fast AVNRT. Radiofrequency energy was delivered initially in the posteroseptal region, followed by inside CS, and finally in the middle septal region, which completed the slow pathway ablation. After the procedure, we could never provoke these arrhythmias.

<Learning objective: Coexistence of focal AT originating from CS with slow-fast AVNRT is a rare phenomenon. Furthermore, slow-fast AVNRT could show unusual characteristic as following: (1) small amount of ATP terminates slow-fast AVNRT; (2) atrial pacing never provoked slow-fast AVNRT with isoproterenol infusion whereas ventricular pacing did, which depends on the physiological characteristic of the dual AV nodal pathway. Accordingly, we should precisely assess the obtained electrophysiological findings.>

Introduction

Coronary sinus (CS) is an uncommon site of origin for focal atrial tachycardia (AT), which was reported to be 6.7% of focal AT [1]. The success rate of radiofrequency ablation (RFA) was high at the point where the discrete potential was recorded [2]. On the contrary, slow-fast atrioventricular nodal tachycardia (AVNRT) is a well-known cause of paroxysmal supraventricular tachycardia (PSVT). However, in some cases, its diagnosis is difficult due to the electrophysiological characteristics. In our case, anterior slow pathway had high sensitivity to a small dose of adenosine triphosphate (ATP) and, moreover, we could not provoke slow-fast AVNRT by atrial pacing with or without isoproterenol infusion, which nearly misled us. Here, we describe the experience of coexistence of AT originating from CS with slow-fast AVNRT, besides their interesting electrophysiological findings.

Case report

A 38-year-old female was referred to our clinic with a complaint of repetitive faintness and palpitation. Ambulatory electrocardiogram (ECG) documented narrow QRS tachycardia with 1:1 atrioventricular (AV) conduction. There was no 12-lead ECG during the tachycardia. Echocardiography elucidated no structural heart disease. Electrophysiological study and subsequent catheter ablation were performed with the setting of multipolar electrode catheter in the high right atrium, His bundle, CS, and right ventricle (RV). Electroanatomical mapping was performed with 3-D electroanatomical mapping system (Ensite Velocity™, St. Jude Medical. Co., Ltd., St Paul, MN, USA) with a mapping catheter (Ensite Array™, St. Jude Medical. Co., Ltd.). RFA was performed with a 4-mm-tip temperature-controlled non-irrigated catheter. Initially, atrial overdrive pacing provoked a narrow QRS tachycardia (hereafter, SVT1). 12-Lead ECG showed long RP tachycardia with P-wave negative deflection in inferior leads (Fig. 1, upper panel). Tachycardia cycle length (TCL) was 420 ms with early activation site in CS ostium (Fig. 2A). RV overdrive pacing showed 1:1 ventriculoatrial (VA) conduction limited to a maximum of 100 bpm. The atrial activation sequence in electrode catheter during SVT1 was different from during RV pacing. Isopotential map showed a breakout site in the CS. Virtual unipolar electrogram (VUE) showed rS pattern, which represented that earlier atrial site was located inside CS. (Fig. 2C). In addition, return cycle after entrainment pacing from CS ostium was not identical with from ​His bundle area—VA linking was not observed (Fig. 3A and B). These findings were consistent with focal AT originating from the CS. SVT1 was sustained and accelerated with isoproterenol administration—TCL became 310 ms (Fig. 2B). Radiofrequency energy was delivered initially in CS ostium, targeting temperature of 55° and a power of 30 W. About 10 times of RFA around the CS was required to eliminate SVT1 where the discrete potential preceding the P-wave onset was recorded in distal ablation catheter (Fig. 2B), which implied the substrate for SVT1 expanded to inside CS (Fig. 2D).

Fig. 1.

Upper panel: 12-lead electrocardiogram (ECG) during SVT1 showed long RP tachycardia and P-wave deflection was negative in lead II, III, and aVF. Lower panel: 12-lead ECG during SVT2 showed narrow QRS tachycardia with the hidden P wave in QRS complex.

Fig. 2.

(A) Intracardiac electrogram (ICE) during SVT1 with the tachycardia cycle length (TCL) of 420 ms. The ICE showed the earliest activation site in the coronary sinus (CS) ostium where the discrete potential was recorded. (B) After isoproterenol infusion, SVT1 was accelerated to TCL of 310 ms and atrial activation sequence was the same as before infusion. The ablation catheter, which was located in the CS ostium, recorded discrete potential preceding P-wave onset. (C) Isopotential mapping elucidated the breakout of CS. The white arrow indicates the direction of activation from the CS. Surface electrocardiogram and VUE were indicated at the bottom of panel. (D) Radiofrequency energy was delivered in the CS ostium to the inside CS. Targeted ablation site was indicated by circles. SVT1 was finally terminated at the blue circle portion. Abl-d, distal ablation electrode; Abl-p, proximal ablation electrode; VUE, virtual unipolar electrode; CSp, proximal coronary sinus electrode; CSd, distal coronary sinus electrode; Hisp, proximal His bundle electrode; Hisd, distal His bundle electrode; RV, right ventricle electrode.

Fig. 3.

(A, B) When atrial entrainment pacing from coronary sinus ostium during SVT1, first return cycle (VA interval) was 205 ms whereas 280 ms from His bundle area. (C) Intracardial electrogram during RV pacing with isoproterenol infusion was indicated. Atrial early activation site was His bundle area (red arrow head). (D) During SVT2, atrial early activation site was His bundle area (red arrow head) and its activation sequence was identical to when RV pacing. Abl-d, distal ablation electrode; Abl-p, proximal ablation electrode; CSp, proximal coronary sinus electrode; CSd, distal coronary sinus electrode; Hisp, proximal His bundle electrode; Hisd, distal His bundle electrode; RV, right ventricle electrode.

Another narrow QRS tachycardia (SVT2) with TCL of 380 ms was induced and sustained by RV extra stimulation with isoproterenol infusion. 12-Lead ECG did not show the obvious P-wave (Fig. 1, lower panel). The electrophysiological study revealed the following findings: (1) atrial early activation site during tachycardia was the His bundle area, whose atrial activation sequence was identical with when RV burst pacing (Fig. 3C and D); (2) both AV and VA conduction showed decremental property and ventricular stimulation at refractory time of the His bundle could not affect the tachycardia cycle length; (3) atrial overdrive pacing from CS ostium could demonstrate manifest entrainment; (4) during entrainment study with atrial pacing, the VA interval of the return beat was the same as the VA interval of the tachycardia, which was known as VA linking; (5) the initiation of the tachycardia by RV extra stimulation showed V-A-V sequence pattern; (6) SVT2 could be provoked only by ventricular pacing—atrial program stimulation at basic cycle length of 400 ms or 600 ms could never provoke SVT2, besides the jump up in the atrio-His(AH) interval with or without isoproterenol infusion; (7) we could terminate SVT2 using small amount of ATP (2 mg) with AV block. These electrophysiological findings, except for the incidence of (6) and (7), indicated SVT2 as slow-fast AVNRT. However, the incidence of (6) and (7) was unusual for the characteristic of slow-fast AVNRT, which delayed our diagnosis of SVT2. We performed RFA with imaginable two possibilities of the tachycardia: slow-fast AVNRT and adenosine-sensitive AT. Adenosine-sensitive AT, which was characterized by the feature of high sensitivity to ATP, was inconsistent with the incidence of (4) and (5). We finally performed slow pathway ablation initially in the posterior atrial septum, followed by inside CS, and in the middle atrial septum targeting temperature of 55° and a power of 30 W, which finally achieved the slow pathway ablation. After the procedure, SVT1 and SVT2 were never provoked again by all of pacing maneuvers with or without isoproterenol infusion. No arrhythmias have recurred in this patient at least four months after these interventions.

Discussion

This case suggests two interesting issues: (1) there was a coexistence of AT originating from CS with slow-fast AVNRT; (2) SVT2, which was disclosed as slow-fast AVNRT, had high sensitivity to ATP and could be induced only by RV pacing. The CS, which is an uncommon site of origin for focal AT, sometimes becomes a critical part of arrhythmia substrate because of its musculature and the other tissues [1]. In addition, a discrete potential, which was recorded in the CS, was frequently targeted with RFA, which demonstrated that the CS is an integral part of the circuit [3]. Coexistence of slow-fast AVNRT with other forms of arrhythmia has been investigated in some studies. Sticherling et al. demonstrated that AT was induced in 15% of patients who underwent slow pathway ablation [4]. The most common cause of concomitant arrhythmia with slow-fast AVNRT was reported to be atrial fibrillation, followed by AT which accounted for 8% of all cases [5]. However, the exact mechanism of the coexistence of AVNRT with AT, especially the AT originating from Koch triangle, remains indeterminable although there is estimated to be an interaction between AV nodal tissue and perinodal tissue [6]. Focused on the SVT2 mechanism, especially regarding the high sensitivity to ATP or adenosine, there are some reports about a validation of the response to adenosine for AV nodal conduction. The antegrade slow pathway was reported to block with adenosine of 7.2 ± 4.7 mg whereas fast pathway with adenosine of 2.7 ± 3.0 mg [7]. According to other reports, a 6 mg bolus of adenosine is the recommended initial dose to terminate PSVT and has an efficacy rate of 60%. A 12 mg dose has an efficacy rate of approximately 90% [8]. Rankin et al. reported that the average effective dosage of adenosine to terminate PSVT was 3.8 mg and of ATP was 6.6 mg, suggesting molar equipotency [9]. These findings can support the unusualness in the present case; a 2 mg bolus of ATP reproducibly terminated slow-fast AVNRT, which was a lower dosage than applied in previous reports. In terms of the way of initiation, slow-fast AVNRT was usually induced by atrial pacing more easily than ventricular pacing and its induction basically requires dual-AV-nodal-pathway physiology. Moreover, the easiness of induction is dependent on the difference of the effective refractory period (ERP) between the antegrade fast and slow pathway—the shorter the difference is, the more easily the induction was by atrial pacing. Lee et al. reported that slow-fast AVNRT which was induced only by ventricular pacing accounted for 3% and, in addition, the differences of the ERP between the dual AV nodal pathway and the ERP of antegrade fast pathway were shorter in the group of slow-fast AVNRT induced only by ventricular pacing than the others [10]. We could never provoke slow-fast AVNRT by atrial pacing and, moreover, the jump-up of AH interval could not be observed by atrial extra stimulation with or without isoproterenol infusion, which suggests the underlying electrophysiological feature in AV node: similar ERP between the dual AV nodal pathway existed and, besides, the ERP of the dual AV nodal pathway was both extremely short—atrial refractory period exceed the ERP of the AV node.

Slow-fast AVNRT has a diverse electrophysiological feature whose mechanism was not clearly revealed. We electrophysiologists should be aware of the possibility that coexistence of slow-fast AVNRT with AT originating from CS and, in addition, slow-fast AVNRT could present various initiation patterns dependent on the characteristic of the dual AV nodal pathway, including the evidence that the anterior slow pathway could be blocked with a small amount of ATP.

Conflict of interest

All authors have nothing to declare.