Adenosine-Induced Atrioventricular Block in Patients With Unexplained Syncope
The Diagnostic Value of ATP Testing
Background ATP and its related nucleoside, adenosine, are ubiquitous biological compounds with potent depressant activity on the atrioventricular node. We hypothesized that an increased susceptibility of the atrioventricular node to adenosine may, in some cases, play a role in the genesis of syncope.
Methods and Results The study was performed in two parts. In part 1, we evaluated the effects of a bolus injection of 20 mg ATP in a group of 60 patients (57±19 years, 31 men) with syncope of unexplained origin and in 90 control subjects without syncope (55±17 years, 46 men). In control subjects, the upper 95th percentile of the maximum RR interval distribution, during ATP-induced atrioventricular block (AVB), was 6000 ms. In the syncope group, 28% of patients had a maximum RR interval above this limit (P=.000). The distribution of the maximum RR interval below the 95th percentile was similar in the two groups. In part 2, we validated the ATP test in 24 patients who had the fortuitous ECG recording of a spontaneous syncope caused by a transient asystolic pause (AVB in 15 and sinus arrest in 9). The ATP test caused AVB with an asystolic pause of ≥6000 ms in 53% of the patients with documented AVB but in none (0%) of the patients with documented sinus arrest (P=.01). Among the patients with spontaneous AVB, the ATP test was abnormal in 6 of the 7 patients (86%) in whom all conventional investigations for syncope had been negative and in 2 of the 8 patients (25%) who had shown positivity (P=.03).
Conclusions An increased susceptibility to ATP testing is present in patients with SUO and patients with syncope due to paroxysmal AVB. Thus, a logical inference is that ATP testing can be used to identify patients with syncope due to paroxysmal AVB. The results of this study form the necessary background for future prospective studies with an aim to validate this assumption.
Adenosine triphosphate and its related nucleoside, adenosine, are ubiquitous biological compounds that exert a potent depressant activity on the AV node; this can result in transient AVB. ATP and adenosine are released from myocardial cells under physiological and pathological conditions (eg, during myocardial oxygen supply/demand imbalance) and have similar effects. The negative dromotropic action of ATP is due to its rapid catabolism to adenosine and the subsequent action of adenosine at purinoceptor sites.1–4 AVB has sometimes been observed after exogenous ATP or adenosine infusion in patients undergoing electrophysiological studies,4 in patients with paroxysmal supraventricular tachycardia,5 and in patients undergoing adenosine stress testing for the diagnosis of coronary artery disease.6,7 At higher doses, an intravenous bolus of ATP or adenosine has been seen to cause transient AVB in many patients with neurally mediated syncope or sick sinus syndrome and in control subjects; the AVB has sometimes been associated with a prolonged asystolic ventricular pause.8–10
Therefore, because of its powerful negative effect on AV conduction, we hypothesized that an increased susceptibility of the AV node to adenosine may play a role in the genesis of some cases of syncope. The aims of the present study were to evaluate the normal range of responses to an intravenous bolus of ATP (ATP testing) in control subjects without syncope and the diagnostic value of ATP testing in patients with SUO and patients with syncope due to paroxysmal AVB.
The study was made up of two parts. In part 1, we evaluated the ATP test in a group of patients with SUO and control subjects. In part 2, we validated the ATP test in a group of patients who had the fortuitous ECG recording of a spontaneous syncope caused by a transient asystolic pause.
The ATP test was performed in 60 patients affected by SUO and 90 control subjects without syncope.
Patients With SUO
SUO was defined in patients in whom the cause of syncope had remained unexplained despite a standardized basic evaluation that consisted of (1) a complete history, physical examination, and neurological evaluation, (2) baseline laboratory testing, (3) 12-lead ECG, (4) ECG monitoring of ≥24-hour duration, (5) chest radiographic examination, (6) M-mode and two-dimensional echocardiographic evaluation of cardiac function, (7) carotid sinus massage, (8) head-up tilt-testing, (9) electrophysiological study (performed in selected patients with structural heart disease or abnormal ECG, or complex premature beats11), and (10) definite evaluation of any clinical or historic findings suggestive of the cause of the syncope. In accordance with the literature,12 patients with the following characteristics were considered to have a definite or potential cause of syncope and were therefore excluded from the study: history suggestive of vasovagal syncope (if a precipitating event such as fear, severe pain, or instrumentation could be identified), history suggestive of situational syncope (if syncope was clearly correlated with coughing, micturition, defecation, or swallowing), a positive response to carotid sinus massage or head-up tilt-testing as previously described,13–15 postural hypotension, conversion reaction, seizure disorders, transient ischemic attack, subclavian steal syndrome, drug-induced syncope, aortic stenosis, pulmonary hypertension, hypertrophic cardiomyopathy, dysrhythmias (eg, sick sinus syndrome, symptomatic supraventricular tachycardia, second- or third-degree AVB, ventricular tachycardia of >5 beats), or generally accepted abnormalities of the electrophysiological study.11,14
Of 494 patients referred to our units for investigation of syncope between January 1995 and December 1996, the cause of syncope remained unexplained in 84 (Fig 1⇓). To avoid possible confusion in the results of ATP test, we excluded an additional 24 patients who had minor electrical abnormalities of the impulse formation or of the conduction system (eg, sinus bradycardia <50 bpm, first-degree AVB, bundle-branch block) or were taking drugs that could impair AV conduction properties or have potential interactions with ATP (eg, digitalis, β-blockers, calcium antagonists, antiarrhythmics). The remaining 60 patients were included in the study and underwent ATP testing.
The control group consisted of patients without a history of syncope or presyncope and without the above-mentioned cardiac electrical abnormalities of impulse formation or conduction system. They were recruited during the same period as the SUO patients.
The ATP test was performed in 24 patients affected by recurrent syncopes who, during ECG monitoring, had a fortuitous recording of intermittent asystolic pause (which caused syncope) with full return to normal heart rhythm at the end of the episode. Apart from syncope, patients had no other acute clinical disease or any other reason for hospitalization; they were in good clinical condition without any physical restriction. We excluded patients in whom chronic or intermittent second- or third-degree AVB or sick sinus syndrome had previously been documented and patients who had bradycardia during acute myocardial infarction or other acute diseases or due to the adverse effects of medications. In addition to ATP testing, all these patients underwent a full electrophysiological study, carotid sinus massage, and head-up tilt-testing without and with pharmacological challenge with nitroglycerin. We have previously shown that in most cases, these examinations are able to identify the exact mechanism of asystolic syncope.14 The protocols of execution of the carotid sinus massage and head-up tilt-test have been previously described.13–15 The protocol of the electrophysiological study included assessment of AV conduction in the baseline state, during incremental atrial pacing, and after intravenous administration of 1 mg/kg ajmaline and assessment of sinus node function in the baseline state and after autonomic blockade as previously described.13,14
Protocol of ATP Test
ATP (20 mg; Striadyne) was dissolved in 10 mL of saline solution and injected very rapidly (<3 seconds) into a suitable antecubital vein with the patient in the supine position. Continuous recording of ECG tracing and noninvasive beat-to-beat arterial blood pressure by means of the Finapres method16,17 were performed during, and for 2 minutes after, drug administration. For the purpose of this study, we evaluated the longest RR interval and the maximum drop in systolic blood pressure (defined as the difference between the value observed immediately before drug administration and the lowest value observed after drug administration, excluding that of the first 2 beats after the prolonged asystolic pauses).
It is well documented in the literature1,3,4,9,10 that the maximum bradycardic effect after a bolus of ATP usually occurs after 10 to 20 seconds (the latency time necessary for the drug to reach the heart); this persists for up to 20 seconds and is followed by sinus tachycardia for up to 2 minutes. Hypotension occurs during and immediately after the bradycardic phase and is sometimes followed by moderate hypertension. Facial flushing, shortness of breath, and chest pressure are frequent side effects, but due to the rapid deactivation of the drugs, these are transient and well tolerated by the patient. All patients gave informed consent.
Average data are presented as mean±SD. Comparison of patients’ characteristics or proportions was done by means of χ2 or Fisher’s exact test or t statistics, when appropriate. A value of P<.05 was considered significant.
The clinical characteristics of patients with SUO and control subjects are shown in Table 1⇓. The two groups had similar clinical characteristics.
ATP Test in the Control Group
The ATP test caused transient third-degree AVB in 26 subjects (29%). The median RRmax was 1600 ms (range, 480 to 8000 ms). Only one subject had a sinus arrest of 3000 ms associated with AVB; no patients had sinus arrest of >2000 ms alone. Systolic blood pressure dropped by 31±20 mm Hg (range, 0 to 80 mm Hg). The distributions of RRmax and systolic blood pressure drop are shown in Fig 2⇓. The mean RRmax was longer in women than in men (2943±1847 versus 1871±1688 ms, P=.005), in subjects ≥55 years old than in those <55 years old (2743±1787 versus 2015±1840 ms, P=.06), and in subjects with resting systolic blood pressure of ≥150 mm Hg than in those with <150 mm Hg (3030±2206 versus 1543±873 ms, P=.001). RRmax was not influenced by underlying cardiopathy (2835±1841 versus 2186±1815 ms, P=.12) or by ECG abnormalities (2477±1707 versus 2369±1891 ms, P=.81).
ATP Test in Patients With SUO
ATP induced transient third-degree AVB in 29 (48%) of SUO patients (P=.01 compared with control subjects). The median RRmax was 2200 ms (range, 700 to 13 000 ms). Two patients also had a sinus arrest of 2300 and 2500 ms associated with AVB; no patients had sinus arrest of >2000 ms alone. The number of SUO patients who had RRmax above the values corresponding to the 95th and 99th percentiles of control group distribution was significantly higher than the number in the control group (Table 2⇓); those with values below the 95th percentile had a distribution similar to that of control subjects. This suggests that a value of RRmax of ≥6000 ms can be considered abnormal. In SUO patients, there also was a slightly higher drop in systolic blood pressure (38±22 mm Hg drop: P=.047 compared with control subjects).
Thus, an increased susceptibility to ATP (defined as RRmax of ≥6000 ms) was present in 17 SUO patients (28%). There was a prevalence of women (11 women and 6 men), with a mean age of 66±20 years (range, 25 to 87 years). They had had a median of three syncopal episodes (mean 4±3) during the previous 3±4 years; all syncopal episodes had occurred suddenly in 13 patients, whereas they were sometimes preceded by dizziness, diaphoresis, or vomiting in the remaining patients; triggering factors had never been identified. Due to the sudden onset, 8 patients had reported trauma as a consequence of the loss of consciousness. An underlying organic heart disease was present in 7 patients. Compared with the ATP-negative patients, the ATP-positive patients were older (66±20 versus 53±19 years; P=.028), had a female prevalence (65% versus 42%; P=.068); and were more likely to have sudden syncopal episodes (71% versus 35%; P=.004) that caused trauma (47% versus 9%; P=.002).
The reproducibility of the ATP test was evaluated after 1 to 4 days in 14 patients: RRmax of ≥6000 ms was still present in 11 of these patients (79%): RRmax was 8349±3264 ms during the first test and 6927±3232 ms during the second test (17% decrease; P=.26). ATP test results were only slightly influenced by premedication with 0.02 mg/kg atropine IV (8 patients; RRmax, 8892±2021 ms before versus 7900±2008 ms after the drug; 11% decrease; P=.06), whereas they normalized in all patients after premedication with oral theophylline (serum levels, 11.5±2.6 μg/mL) (12 patients; RRmax, 8878±2153 ms before versus 1530±962 ms after the drug; 83% decrease; P=.001. Eleven patients received chronic treatment with 600 mg/d theophilline PO and were followed-up for a mean of 13±10 months. A total of three syncopal recurrences occurred in 2 patients, with a decrease in the recurrence rate from 1.8 per year before treatment to 0.26 per year during treatment.
Complications of the ATP Test
A sensation of lightheadedness was the most frequently encountered symptom in patients with prolonged pauses. Bradycardia-dependent syncope occurred in no control subjects and 7 SUO patients; syncope was of short duration, with a full recovery within a few seconds, and did not require any treatment. Nonsustained atrial tachyarrhythmias occurred in 1 patient at the end of the bradycardic phase.
The data for the 24 patients with a documented episode of syncope caused by a transient asystolic pause are shown in Table 3⇓. There were 15 patients with syncope due to paroxysmal AVB and 9 patients with syncope due to sinus arrest (control group). The ATP test caused AVB with an asystolic pause of ≥6000 ms in 8 (53%) and 0 (0%) patients, respectively (P=.01); moreover, an additional 2 patients in the AVB group showed pauses of 5000 ms. The ATP-induced AVB was very similar to the spontaneous AVB, as evidenced by the cases of patients 2 and 3 shown in Figs 3⇓ and 4⇓. Furthermore, among patients with documented spontaneous AVB, the ATP test showed an abnormal response in 6 of the 7 patients (86%; patients 1 through 7) in whom all conventional investigations were negative, whereas it was abnormal in only 2 of the 8 patients (25%; patients 8 through 17) who had a positive response to carotid sinus massage or head-up tilt-testing or AV conduction abnormalities that suggested another mechanism responsible for AVB (P=.03). Compared with the patients with sinus arrest, the difference was highly significant for those patients with a negative work-up (P=.001) but not for those with a positive work-up (P=.21).
The main results of this study are that the patients with SUO show an increased susceptibility to ATP testing in comparison with those without syncope and ATP testing is able to reproduce AVB (and suggest the mechanism) in patients with spontaneous paroxysmal AVB, especially in those without abnormalities of the AV conduction or autonomic nervous system. If these latter patients had not had the fortuitous documentation of paroxysmal AVB at time of syncope, they would have been categorized as SUO patients. The logical inference is that ATP testing can identify patients with syncope due to transient AVB even when the electrophysiological findings and other conventional tests are unremarkable. However, this remains an interesting hypothesis to be confirmed. The results of the present study form the necessary background for future prospective studies aimed to demonstrate the causal relationship between increased susceptibility to adenosine and syncope due to paroxysmal AVB.
In our institutions, ATP testing suggested such a diagnosis in 28% of SUO patients and 3.4% of all patients referred for study of syncope. In these patients, syncope usually occurred without warning symptoms or triggering factors and was more frequent in female and older patients; it seemed to be independent of the presence of underlying heart disease.
The ATP test was simple, easy to perform, safe, and without complications, and positive responses were sufficiently reproducible.
The clinical usefulness of ATP tests for identifying patients with SUO is independent of the exact pathophysiological mechanism of spontaneous AVB, which remains unclear. In the literature, an adenosine-mediated mechanism has been assumed to be responsible for some clinical cases of AVB that were resistant to atropine and reversed by theophylline, an adenosine antagonist. In those cases, AVB was a complication of myocardial infarction,18,19 cardiac transplant rejection,20 atrial fibrillation with slow ventricular response,21 or dipyridamole infusion,2 which elevates endogenous adenosine levels.
The cause of the hypersensitivity to exogenous adenosine found in our patients is not clear because we carefully excluded all patients with manifest or subtle AV conduction disorders or those taking drugs that block AV conduction. We cannot exclude that unknown AV conduction abnormalities not recognizable by means of standard clinical and electrophysiological evaluation could constitute the substrate responsible for the hypersensitivity to adenosine. Whether a positive response to the ATP test also identified an adenosine-mediated mechanism of the paroxysmal AVB—due to an increased release of endogenous adenosine under physiological or pathological conditions (eg, hypotension or ischemia)—or whether it simply revealed a nonspecific susceptibility of the AV node to different triggers (eg, vagal hyperactivity or intermittent AV node conduction disorders) that could not otherwise be recognized is unknown. Shen et al9 recently suggested that adenosine may be a potential modulator for vasovagal syncope. In their study, a bolus of IV adenosine in a patient in the upright position was able to induce, 15 to 60 seconds after injection, a typical vasovagal response in 22 of 85 patients (26%) with syncope and negative electrophysiological testing. A sympathetic activation by adenosine, either directly (ie, on cardiac excitatory efferent nerves) or indirectly (ie, vasodilation and reflex sympathetic activation), was assumed to be the mechanism responsible for the vasovagal reaction. All except 4 of their adenosine-positive patients had also a positive response to the head-up tilt-test. In that respect, the study of Shen et al differs from ours. Indeed, in the present study, we carefully excluded those patients with a history of neurally mediated syncope, and all patients had a negative response to the head-up tilt-test; moreover, the mechanism of induction of AVB was different from vagal stimulation because it was not eliminated by atropine. Thus, the mechanism of syncope in our patients seems to be different from that found by Shen et al.9 However, the ATP test reproduced the spontaneous AVB in 2 of 8 patients with a neurally mediated susceptibility or intrinsic AV conduction abnormalities (Table 3⇑), thus raising the possibility that common mechanisms or some degree of overlapping exists among different etiologies of spontaneous paroxysmal AVB.
Adenosine has been proposed as a putative mediator of sick sinus syndrome due to its direct negative chronotropic effect on the sick sinus node.1,3,8 The present study shows that when the sinus node is normal, adenosine has little or no effect on sinus node function.
The effect of a bolus of ATP on systolic blood pressure is well known, although it has not been systematically investigated.1,6,9 We found a trend toward a higher hypotensive response in patients with SUO than in control subjects. Hypotension could play an adjunctive role in the genesis of syncope in patients with AVB.
Mechanism of ATP Test
Two conditions are necessary to cause a positive asystolic response to the ATP test: the induction of a complete AVB and, at the same time, the exaggeration of the mechanism of overdrive suppression of the ventricular escape rhythm. Adenosine affects the AV node by stimulating the time-dependent outward potassium current, a current that is also stimulated by acetylcholine. It exerts its primary and most powerful effect on the N cells of the AV node. Adenosine has little or no effect on the distal AV node or His-Purkinje system in the basal state but may significantly slow the AV nodal or His-Purkinje escape rhythm in the presence of sympathetic stimulation.22 Thus, adenosine attenuates the cardiac stimulatory actions of catecholamines and indirectly depresses ventricular automaticity. In these circumstances, the sudden cessation of propagation of conducted impulses to the ventricles may exaggerate the mechanism of overdrive suppression of idioventricular pacemakers.23 Overdrive suppression may also occur in the normal heart, although it is enhanced in conditions of damage to the specialized tissues or anoxia or as a consequence of the depressant effects of drugs.22
Management of Patients With ATP-Induced AVB
The study was not designed to address management of patients with ATP-induced AVB. We treated most patients with theophylline, an adenosine receptor antagonist. A low recurrence rate of syncope was observed; no patients developed stable AVB. The positive clinical response to theophylline suggests that endogenous release of adenosine plays a role in the mechanism of syncope. The potential role of this specific treatment must be evaluated in well-conducted prospective studies that compared it with the natural course of the syndrome and the effect of pacing therapy.
Selected Abbreviations and Acronyms
|RRmax||=||maximum ventricular pause induced by ATP infusion (ATP test)|
|SUO||=||syncope of unexplained origin|
Dr Del Rosso’s current address is Department of Cardiology, Ospedale S. Pietro Igneo, Fucecchio, Italy. Dr Pellinghelli’s current address is Department of Cardiology, Ospedale di Oglio Po, Cremona, Italy.
- Received May 19, 1997.
- Revision received August 11, 1997.
- Accepted August 20, 1997.
- Copyright © 1997 by American Heart Association
Lerman B, Belardinelli L. Cardiac electrophysiology of adenosine: basic and clinical concepts. Circulation. 1991;83:1499–1508.
Pelleg A, Mitsuoka T, Michelson E, Menduke H. Adenosine mediates the negative chronotropic action of adenosine 5′-triphosphate in the canine sinus node. J Pharmacol Exp Ther. 1987;242:791–795.
Belardinelli L, Linden J, Berne RM. The cardiac effects of adenosine. Prog Cardiovasc Dis. 1989;22:73–97.
DiMarco J, Miles W, Akhtar M, Milstein S, Sharma A, Platia E, McGovern B, Scheinman M, Govier W. Adenosine for paroxysmal supraventricular tachycardia: dose ranging and comparison with verapamil. Ann Intern Med. 1990;113:104–110.
Brignole M, Menozzi C, Alboni P, Oddone D, Gianfranchi L, Gaggioli G, Lolli G, Paparella N. The effect of exogenous adenosine in patients with neurally-mediated syncope and sick sinus syndrome. PACE. 1994;17(pt II):2211–2216.
Benedini G, Cuccia C, Bolognesi R, Affatato A, Gallo G, Renaldini E, Visioli O. Value of purinic compounds in assessing sinus node dysfunction in man: a new diagnostic method. Eur Heart J. 1984;5:394–403.
Petersen MEV, Williams TR, Sutton R. A comparison of non-invasive continuous finger blood pressure measurements (Finapres) with intra-arterial pressure during prolonged head-up tilt. Eur Heart J. 1995;16:1647–1654.
Alboni P, Ratto B, Scarfò S, Rossi P, Cappato R, Paparella N. Dromotropic effects of oral theophylline in patients with atrial fibrillation and slow ventricular response. Eur Heart J. 1991;12:630–634.
Lerman B, Wesley R, DiMarco J, Haines D, Belardinelli L. Antiadrenergic effects of adenosine on His-Purkinje automaticity. J Clin Invest. 1988;82:2127–2135.
Vassalle M. On the mechanisms underlying cardiac standstill: factors determining success or failure of escape pacemakers in the heart. J Am Coll Cardiol. 1985;5:35B–42B.