A case of paroxysmal atrial fibrillation with a non-pulmonary vein trigger identified by intravenous adenosine triphosphate infusion

Abstract

A 54-year-old woman was referred to our institution with frequent chest discomfort and was diagnosed with drug-refractory paroxysmal atrial fibrillation. Radiofrequency catheter ablation (RFCA) was performed using a three-dimensional electroanatomic mapping system. After completion of left and right circumferential pulmonary vein isolation (CPVI), an intravenous bolus of adenosine triphosphate (ATP, 20 mg) was administered to evaluate the electric reconduction between the pulmonary vein (PV) and left atrium (LA). Although no PV–LA reconduction was observed, atrial fibrillation (AF) was reproducibly induced. As the duration of AF was very short (<20 s), no further RFCA to the LA was performed. One month later, the patient presented with frequent atrial tachyarrhythmias (ATs), and RFCA was repeated. Although no electric reconduction was observed in the left- or right-sided PVs, incessant ATs and AF were induced after an intravenous bolus administration of ATP. The earliest atrial activation site initiating ATs was consistently identified from electrodes positioned in the superior vena cava (SVC), and both ATs and AF were no longer inducible after electric isolation of the SVC. ATP-induced PV/non-PV ectopy may be a marker of increased susceptibility to autonomic triggers of AF and could potentially predict recurrent AF after CPVI.

1. Introduction

Although the pulmonary veins (PVs) represent the predominant source of atrial fibrillation (AF), non-PV triggers play an important role in initiating and maintaining AF in approximately 20% of cases [1–3]. Episodes of atrial tachyarrhythmias (ATs) and AF originating from non-PV triggers are often unpredictable and difficult to identify, owing to their transient duration and diverse locations. Furthermore, the precise location may remain unknown, even with the use of a three-dimensional electroanatomic mapping system (3DEAM).

Several studies have indicated that an intravenous injection of adenosine, in the form of adenosine triphosphate (ATP), can induce the transient reconnection of isolated PVs after electric isolation, consistent with unmasking dormant conduction between the PVs and the left atrium (LA) [4,5]. Furthermore, ATP, when given as an intravenous bolus, can induce AF [6]. In addition, several studies have recently described the usefulness of an ATP injection for inducing and identifying PV and/or non-PV triggers after circumferential pulmonary vein ablation (CPVI) [7,8]. Here, we describe a case of paroxysmal AF originating from a non-PV trigger, which was precisely identified using ATP infusion and successfully treated using radiofrequency catheter ablation (RFCA).

2. Case report

A 54-year-old woman was referred to our institution because of frequent episodes of palpitation and chest discomfort. Although she was taking several antiarrhythmic drugs, a symptomatic 12-lead electrocardiogram (ECG) revealed AF, and RFCA was indicated. After obtaining informed consent from the patient, RFCA was performed under deep sedation/analgesia using propofol and dexmedetomidine. Two circular mapping catheters (EPstar Libero, Japan Lifeline Inc., Tokyo, Japan) were placed in the superior and inferior PVs, respectively, via a transseptal puncture site, and the left and right-sided ipsilateral PVs were circumferentially and extensively ablated, using 3DEAM (CARTO, Biosense Webster, Inc., Diamond Bar, CA, USA) under electrophysiological guidance. After the initial electric isolation of all 4 PVs, a 20-mg ATP bolus was injected to provoke dormant PV conduction during coronary sinus pacing. Although no electric reconduction was observed in the left or right PVs, AF was reproducibly induced (Fig. 1). The duration of AF was very short (<20 s), and no further AF episodes were observed, even during an intravenous drip infusion of isoproterenol (ISP: dosage up to 5 μg/min), so no further mapping or RFCA was performed, with the exception of linear ablation at the cavotricuspid isthmus.

Fig. 1.

Induction of AF via ATP infusion. Two circular catheters were positioned at both the superior and inferior right (a) and left (b) PVs. The asterisk indicates electrically dissociated PV potentials. The earliest atrial activation site initiating AF was obtained from the electrodes (black arrow heads) positioned between the HRA and SVC. RSPV, right superior pulmonary vein; RIPV, right inferior pulmonary vein; LSPV, left superior pulmonary vein; LIPV, left inferior pulmonary vein; AP, atrial pacing; VP, ventricular pacing; AF, atrial fibrillation; ATP, adenosine triphosphate; PV, pulmonary vein; HRA, upper right atrium; and SVC, superior vena cava.

However, the patient was later re-admitted with frequent palpitation symptoms similar to those exhibited prior to the initial RFCA. Despite oral class I antiarrhythmic drug and beta-blocker treatment, ECG revealed incessant ATs and symptomatic AF (Fig. 2), so repeat RFCA was performed. Intravenous bolus infusion of ATP did not produce dormant electrical reconduction in either the left or right PVs, but incessant AF/ATs were induced. The earliest atrial activation site initiating incessant AF/ATs was consistently observed from the electrodes placed between the superior vena cava (SVC) and the upper right atrium, which was the location identical to that observed in the initial procedure. To identify the precise ectopic origin, a circular multielectrode-mapping catheter was positioned superior to the atriocaval junction within the SVC based on venography. During repetitive ATs, the activation recorded from the mapping catheter was consistently the earliest, and AT was diagnosed as being triggered by a non-PV focus within the SVC (Fig. 3). Electric isolation of the SVC was therefore performed (Fig. 4), after which AF/ATs were no longer inducible. Only dissociated potentials from the mapping catheter in the SVC were observed, despite repeated ATP bolus injections with and without concomitant ISP infusion. The patient did not exhibit any AF/ATs during a follow-up period of approximately 6 months.

Fig. 2.

Symptomatic 12-lead and monitoring ECG after initial RFCA. AF/ATs from both 12-lead (upper) and monitoring ECG (lower; black arrows) were incessant during sinus rhythm (SR). ECG, electrocardiogram; RFCA, radiofrequency catheter ablation; AF, atrial fibrillation; and AT, atrial tachyarrhythmias.

Fig. 3.

Initiation of AF and subsequent ATs after ATP infusion. (a) After ATP bolus infusion, AF was initiated via the earliest atrial activation site observed from the electrodes (black arrow heads) located between the HRA and SVC, identical in location to the initial procedure. (b) Incessant ATs (black arrows) were induced via the earliest atrial activation site (black arrow heads), consistent with AF initiation. (c) Note the spiky potentials (black arrow heads), which were observed from the circular mapping catheter (described as “Lasso” in figure) positioned at the SVC, above the atriocaval junction. These potentials preceded the atrial electrograms during incessant ATs. LSPV, left superior pulmonary vein; LIPV, left inferior pulmonary vein; HBE, His bundle electrogram; AP, atrial pacing; SR, sinus rhythm; AF, atrial fibrillation; AT, atrial tachyarrhythmias; ATP, adenosine triphosphate; HRA, upper right atrium; ans SVC, superior vena cava.

Fig. 4.

Intracardiac recordings and catheter position during electric SVC isolation. (a) The spiky potentials (asterisk) from the circular mapping catheter (described as “Lasso” in figure), located above the atriocaval junction guided by SVC venography, disappeared during RF energy application. (b, c) Yellow arrow heads indicate the presumed earliest atrial activation site before circular catheter positioning. Venography was performed via the 2 long sheaths and a circular mapping catheter was inserted through one of these long sheaths, positioned above the presumed earliest atrial activation site. Note the RFCA sites (yellow dots) in the lower part than the circular catheter position guided by 3DEAM (Ensite-NavX). CS, coronary sinus; HBE, His bundle electrogram; HRA, upper right atrium; RAO, right anterior oblique; SVC, superior vena cava; RF, radiofrequency; and RFCA, radiofrequency catheter ablation.

3. Discussion

Adenosine is known to hyperpolarize atria cells via activation of the A1 adenosine receptor-mediated inwardly rectifying potassium currents (I_KAdo), which shorten the action potential duration (APD) and the effective refractory period [9]. While established, the mechanisms behind the association of these ionic interactions, adenosine-induced PV (and/or non-PV) trigger activation, and clinical AF remain unclear.

Several reports have indicated that adenosine (in the form of ATP) can generate autonomic activation of PV triggers. Cheung et al. [10] presumed that adenosine-induced PV triggers were in part elicited by activation of parasympathetic triggers because adenosine and acetylcholine (ACh), both of which are parasympathetic neurotransmitters with cardiac stimulatory actions, act on identical cell signaling pathways to produce significant antiadrenergic effects [11]. Adenosine and ACh share similar G-protein receptor-effector coupling systems (class A18 rhodopsin-like; adenosine interacts with the A1 receptor, and ACh with muscarinic [M2] receptor). Activation of these systems generates the outward potassium currents I_KAdo and M2 receptor-mediated inwardly rectifying potassium currents, respectively, resulting in hyperpolarization of the cardiac cell membrane and shortening of the atrial APD and refractory period. Such ionic interactions would provide a favorable substrate for atrial arrhythmogenic activities such as AF/ATs.

With respect to ACh, several reports have described the relevance between the ACh-mediated cholinergic effect, which stimulates the intrinsic cardiac autonomic nervous system (ICANS), and PV and/or non-PV triggers. In dogs, administration of ACh into the ganglionated plexi (GP) fat pad, a major ICANS situated at the PV–LA junction, induced spontaneous PV triggers and AF [12]. In addition, Lu et al. [13] demonstrated that the hyperactivity of the SVC-aorta-GP axis, another major ICANS, induced by direct injection of ACh could lead to rapid firing from the SVC. Under the existence of both efferent parasympathetic and sympathetic neurons in the GPs, stimulation of these major ICANSs would sympathetically produce a high cytosolic Ca²⁺ enhanced by calcium transients and parasympathetically induce APD shortening. Subsequently, enhanced intracellular Ca²⁺ accumulation occurs at membrane voltages negative to the equilibrium potential for the sodium–calcium exchanger (NCX) due to enhanced repolarization, which acts as its forward mode resulting in an inward current, leading to generation of an early after depolarization (EAD), and re-excitation of the myocardium, causing PV and/or non-PV firing [14].

Considering the fact that both ATP and ACh share G-protein mediated autonomic effects, ATP-induced PV and/or non-PV firing may be partly due to stimulation of the GP, as is observed with ACh administration.

On the other hand, several authors reported that an ATP bolus injection could promote significant sympathomimetic effects. Biaggioni et al. suggested that intravenous ATP administration primarily activates afferent nerves via arterial (or carotid body) chemoreceptors or baroreceptors resulting in sympathetic activation [15]. This increase in sympathetic tone is frequently followed by an abrupt shift toward vagal predominance, which may facilitate the induction of AF from the arrhythmogenic site [16,17].

Therefore, the ATP-induced non-PV trigger originating within the SVC in the present case might have been elicited by both sympathetic and parasympathetic activity. The suggested mechanism is the partial stimulation of the SVC-aorta-GP axis via G-protein signaling that hyperactivates Ca²⁺ transient leading to an inward current via the NCX during APD shortening, which subsequently generates a triggered EAD leading to non-PV firing.

The major concern of ATP administration is whether these triggers could be induced reproducibly and whether they could be induced from the same source. This issue remains controversial and to our knowledge, relevant systematic studies have not been reported. The lack of reproducible arrhythmogenic activity from the same source may limit the accuracy of 3DEAM, making its precise identification difficult. In the present case, incessant AF/ATs originating from the SVC were reproducibly induced by ATP infusion. Given the above-mentioned concerns, however, we performed electric SVC isolation, rather than focal trigger ablation, which is essential for identifying the exact trigger source.

The ability of ISP to provoke PV-/non-PV foci when administered by an intravenous drip infusion with an incremental dosage (1–2 μg/min, increasing up to 20 μg/min) is widely recognized [18]. However, both the onset of action and the decline in activity are delayed. This requires high doses and a continuous infusion for at least 15 min to override the sedation effect and simultaneously produce physiologic catecholamine discharge as previously proposed [7], which limits the clinical usefulness of the drug. In the present case, the maximum ISP dosage during the initial procedure was 5 μg/min, which was discontinued because of hemodynamic instability; therefore, the sedation effect could not be negated, limiting the provocation of the non-PV trigger.

To eradicate ATP- or ISP-mediated concerns, Zhang et al. have recently demonstrated the effectiveness of using ATP in combination with ISP infusion for unmasking non-PV triggers [19]. In their study, among 39 patients in whom ATP reproducibly induced AF, only 5 μg/min of ISP was used before the administration of ATP, which was comparable to the present case. This approach, used to increase the heart rate by ≈20 bpm from baseline, was designed to elicit an obvious change in autonomic tone, leading to higher reproducibility and longer durability of ATP-induced non-PV triggers [5,20]. As a result, none of the patients showed AF originating from a non-PV trigger during ISP infusion alone, with AF only revealed by additional ATP injection. The findings from this case indicate that ATP administration could be useful when the optimal ISP dosage cannot be tolerated. Furthermore, to achieve higher reproducibility and longer durability of ATP-induced non-PV triggers, the combination of ATP and ISP, might be more effective for unmasking than either drug alone, although further investigation is needed to confirm the combined effect by these drugs.

In conclusion, ATP-induced AF after CPVI is strongly associated with spontaneous AF via non-PV triggers, as described in the present case. ATP injection may be useful for identifying fibrillatory arrhythmogenic sites.

Conflicts of interest

None.