|
|
||||||||||
|
J Am Coll Cardiol, 2006; 47:587-593, doi:10.1016/j.jacc.2005.09.043
(Published online 13 January 2006). © 2006 by the American College of Cardiology Foundation |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

* Department of Cardiology, University Hospital La Timone, Marseille, France
Department of Mathematics and Biostatistics, University Hospital La Timone, Marseille, France
Manuscript received May 18, 2005; revised manuscript received August 29, 2005, accepted September 8, 2005.
* Reprint requests and correspondence: Prof. Jean-Claude Deharo, Hôpital La Timone Adultes, Département de Cardiologie, 264, Rue Saint Pierre, 13 385 Marseille Cx 05, France (Email: jean-claude.deharo{at}ap-hm.fr).
| Abstract |
|---|
|
|
|---|
BACKGROUND: Heart rhythm obtained during provocative condition is often used to guide therapy in VVS. To date there is no conclusive evidence that the heart rhythm observed during a positive HUT can predict heart rhythm during VVS or that the heart rhythm observed during a spontaneous syncope will be identical to the recurrent syncope.
METHODS: Twenty-five consecutive VVS patients (age 60.2 ± 17.1 years; 14 women,) presenting with frequent syncopes (6.9 ± 4.6 episodes/year) and a positive HUT (cardioinhibitory in 8 patients) were implanted with an ILR. Seven of them also had a positive adenosine triphosphate (ATP) test.
RESULTS: Follow-up was 17.0 ± 3.6 months. Thirty VVS were observed in 12 patients. Nine episodes showed bradycardia of <40 beats/min or asystole; progressive sinus bradycardia preceding sinus arrest was the most frequent electrocardiographic finding. Twenty-one syncopes occurred without severe bradycardia. The heart rhythm observed during the first syncope was identical to the recurrence. No correlation was found between slow heart rate at the ILR interrogation and a cardioinhibitory HUT response (p = 1.0) or a positive ATP test (p = 1.0).
CONCLUSIONS: In highly symptomatic patients with VVS, the heart rhythm observed during spontaneous syncope does not correlate with the HUT. The heart rhythm during the first spontaneous syncope is identical to the recurrent syncope.
| |||||
Heart rhythm can be continuously monitored with an insertable loop recorder (ILR) (Reveal 9525 and Reveal Plus 9526; Medtronic, Minneapolis, Minnesota) (13,14). This device, inserted subcutaneously in the left chest region for 14 to 18 months, is capable of monitoring up to 42 min of a single-lead electrocardiogram (ECG). It can be patient-triggered or, with the Reveal Plus, automatically activated.
Highly symptomatic patients with vasovagal syncope were implanted to prospectively analyze the heart rhythm during spontaneous syncope recurrences. Heart rhythms observed during HUT and the adenosine triphosphate (ATP) test were also correlated.
| Patients and methods |
|---|
|
|
|---|
Study protocol. The contents of the study protocol obtained the approval of the hospital ethics committee. All eligible patients received device therapy with an ILR, and follow-up for 18 months was planned.
Because of the possible overlap between vasovagal syncope and adenosine-sensitive syncope (15,16), an ATP test was performed at protocol initiation.
Patients were systematically seen by the investigators for clinical evaluation, then for ILR interrogation one month after implantation, every three months until the end of the study, and after each syncope recurrence. The patients received no drug treatment for vasovagal syncope. They received counseling on the precipitating factors and warning prodromes in the aim of preventing or postponing syncope. In the advent of a pacemaker implant, the ILR was explanted.
During follow-up, patients were asked about syncope recurrences (presyncopes were not considered for analysis). The ILR recordings were retrieved and downloaded for analysis.
Test procedure.
Tilt test
Baseline HUT was performed in our department, as previously described (17), in the morning after an overnight fast. A 10-min supine period was followed by a maximum of 45 min of tilt at a 60° angle. Noninvasive finger heart rate and blood pressure were continuously recorded through a photoplethysmographic device (Finapres; Ohmeda, Englewood, Colorado), and beat-by-beat values were stored on a computer. The ECG was continuously monitored and recorded on Holter tapes (Pathfinder; Del Mar Reynolds Medical, Hertford, United Kingdom) throughout the HUT. The HUT was considered positive when the patient developed syncope or intolerable presyncope associated with significant arterial hypotension. In case of negative baseline HUT, a second phase of 5 min supine and 10 min upright was performed with drug provocation using either 300 µg sublingual nitroglycerin or an incremental infusion of isoproterenol at a rate varying from 1 to 3 µg/min. Nitroglycerin was our first-choice drug, and isoproterenol was used only in patients who had a previously negative HUT with nitroglycerin provocation at another institution. Responses to HUT were classified according to the modified Vasovagal Syncope International Study (VASIS) classification (18).
ATP test
The ATP test was not performed on the same day as the HUT. All patients were in sinus rhythm, free of antiarrhythmic drug therapy, and in the supine position with continuous ECG recording at 25 mm/s. They received a 20-ml intravenous bolus injection of 20-mg ATP (Striadyne; Wyeth Pharmaceuticals, Paris, France) followed by a 20-ml flush of 5% glucose. The test was considered positive when a ventricular pause of
6 s occurred, due to either complete atrioventricular block or sinoatrial block (19).
ILR implant and follow-up. The ILR (Reveal 9525 in the first three patients and Reveal Plus 9526 in all the others) was implanted in the electrophysiology room, subcutaneously in the left chest region.
After implantation, patients and a family member were instructed on the use of the patient activator, a means to trigger the device memory after each syncope. The Reveal model requires an external activation to record a single-lead ECG. The Reveal Plus model has additional recording capabilities based on automatic detection of bradycardia or tachycardia. This device was programmed to store the ECG when a bradycardia of <40 beats/min or a tachycardia of >165 beats/min occurred.
Patients were nevertheless strongly encouraged to activate the device after each syncope. When recordings were automatically stored, only the episodes which correlated with a syncopal attack were considered for analysis. Presyncopes were not considered.
For comparison with the HUT data, rhythm and rate of the ILR tracings were classified as follows:
40 beats/min and without either patient activation or the presence of an automatic recording by the device (in patients equipped with a Reveal Plus model). Statistics. Data are expressed as mean values ± standard deviation, and qualitative variables were compared by use of the McNemar test. A value of p < 0.05 was considered significant.
| Results |
|---|
|
|
|---|
|
Seven patients had a positive ATP test with a maximum RR interval of 7.6 ± 1.6 s.
Follow-up. Figure 1 summarizes the outcome of the patients. Two ILR were explanted after four and six months, one because of breast neoplasia requiring radiotherapy and the other for local infection. All of the 23 remaining patients had complete follow-up of 18 months, and 12 of them presented with 31 syncopes (2.6 ± 1.8 per patient). The first syncope occurred 4.8 ± 4.7 months after inclusion. Twenty-three syncopes were documented by patient-activated recordings. In patients implanted with a Reveal Plus, failure of patient-activation occurred in eight episodes, two documented by automatic activation and six not documented by automatic activation. According to the classification, "slow heart rate" was ruled out in the six latter cases. In the three patients implanted with a Reveal, no activation failure was reported. Electrocardiographic findings during recurrences are reported in Table 2.
|
|
The 30 remaining syncopal attacks were considered to be vasovagal and were analyzed. They represented 2.5 ± 1.7 episodes per patient. Slow heart rate (<40 beats/min) was observed in nine episodes (5 patients). Sinus arrest preceded by sinus bradycardia and leading to an asystole of 10.5 ± 6.2 s (maximum 22.8 s) was observed in six cases (three patients), severe sinus bradycardia was observed in two cases (two patients), and sudden-onset atrioventricular block was observed in one case.
The distribution of the number of recurrences per patient was the following: one in five patients; two in two patients; three in two patients; five in three patients. The seven patients with more than one syncope showed the same type of heart rhythm pattern for each of the recurrences (heart rate
40 beats/min in six patients and <40 beats/min in one patient).
Figure 2 shows the correlation observed between HUT responses, ATP test, and ILR-documented syncopes. There is no statistical correlation between cardioinhibitory HUT (p = 1.0; McNemar test) or positive ATP test (p = 1.0; McNemar test) and slow heart rate during recurrences.
|
| Discussion |
|---|
|
|
|---|
Heart rate pattern at the time of syncope. Patients with severe bradycardia showed a progressive decrease in heart rate leading to asystole due to sinus arrest in the vast majority of the cases (6 of 9 syncope). This pattern has been regarded as highly suggestive of neurocardiogenic syncope (10) and is confirmed in our study. The mean length of sinus arrest in the present study was 10.5 ± 6.2 s, which is shorter than previously reported in similar patients (10,11,21). Moya et al. (10) suggest that long asystolic pauses might be regarded as typical of neurally mediated syncope without warning in patients with advanced age and aspecific clinical presentation. In our series of patients, mean age was lower than in the ISSUE study and the selection of patients was made on the basis of a typical vasovagal history, which might explain the difference observed. Atrioventricular block was observed in one patient in our series of vasovagal syncopes. This heart rhythm has been regarded as very unusual in vasovagal patients except when it is accompanied by sinus bradycardia or arrest (22), but Mendoza et al. (23) could document paroxysmal atrioventricular block in six patients with positive HUT who experienced spontaneous syncope or presyncope during Holter monitoring. In these patients, sinus rate was stable before the onset of paroxysmal atrioventricular block; electrophysiologic studies were negative and the ATP test positive in only one case. These observations are similar to our patient #18, who experienced a third-degree atrioventricular block documented during the first syncopal attack seven months after ILR implant. This 8.8-s atrioventricular block period occurred in the morning and was preceded by normal sinus rhythm at 65 beats/min for 10 minutes. The patient was a 69-year-old woman with a 10-year history of typical vasovagal syncope, often in similar environmental circumstances. Baseline ECG was normal and passive HUT was positive (type 3). The ATP test showed a 9.5-s AV block. Electrophysiologic study showed no conduction disturbance. The recorded syncope was considered to be vasovagal but might also correspond to an adenosine-sensitive syncope (15,22).
Although there are few further data concerning ECG recordings during spontaneous vasovagal syncopal episodes, in the studies available severe bradycardia has been more frequently reported (10,11) than in our series. In the ISSUE study a subgroup of 29 patients with unexplained syncope had a positive HUT. Eight patients had syncope recurrences, with asystole or bradycardia in six (75%); syncope with normal sinus rhythm or sinus tachycardia occurred in only 25% of patients but was the more frequent mechanism during presyncope. The difference observed in our study might be due to the difference in the patients, primarily the younger age in our study subjects than in the tilt-positive patients of the ISSUE study (60.2 ± 17.1 years vs. 64 ± 15 years, respectively).
In patients with either stable sinus rhythm, bradycardia of >40 beats/min, or sinus tachycardia, severe bradycardia can be excluded as the cause of syncope, whereas a vasodepressor vasovagal mechanism can be suspected only because of the absence of blood pressure monitoring. This assumption seems reasonable in case of mild bradycardia where a mixed syncope can be suspected. In case of a stable heart rate, a pure vasodepressor vasovagal syncope is hypothesized because our patient population was highly selected on the basis of clinical data and a positive HUT. Nevertheless, orthostatic hypotension without a compensatory increase in heart rate due to autonomic failure cannot be excluded, particularly in the oldest patients. Psychogenic or neurologic mechanisms are also possible. Moreover, because syncope was not witnessed, we cannot exclude that true syncope did not occur in some cases.
The pattern of progressive heart rate increase before syncope, observed twice in one patient (#15), is also compatible with the mechanism of neurocardiogenic syncope, because it has been described during HUT (24) and spontaneous syncope in ATP- and tilt-positive patients (10,22).
Mechanism of the recurrences. In the seven patients who experienced more than one vasovagal syncope, the heart rhythm observed during the first recurrence (i.e., cardioinhibitory or not) was predictive of the rhythm observed during subsequent recurrences. This result of our study is in accordance with the preliminary data reported in the ISSUE study (10), where three patients had several syncopal recurrences. In these three cases, the recurrence mechanism was predicted by the first episode. We confirm these data in our patients who experienced one to five recurrences. As previously suggested by others, this result might have important implications for the choice of therapy in a similar population of highly symptomatic patients once the first recurrence has been documented through an ILR (10). Interestingly, in a different group of patients with positive ATP and HUT tests, Donateo et al. (22) observed very similar heart rate findings during the second or the third event.
Patient #4 experienced six syncopes during the study, including one post-tachycardia pause and five vasovagal syncopes without severe bradycardia. Two distinct syncope mechanisms were therefore shown by the ILR. Based on the clinical history we considered that the vasovagal mechanism was predominant. As in the other vasovagal patients, the same heart rate (no severe bradycardia) was consistently observed in all the documented vasovagal episodes.
Correlation between spontaneous and induced syncope. In our study, the type of response observed during HUT was not predictive of the mechanism of spontaneous syncope. This data is consistent with the results of the ISSUE study (10), where spontaneous asystolic syncope was observed more frequently than expected on the basis of the HUT results. Our study shows that severe bradycardia during spontaneous syncope can be seen not only in patients with type 2 HUT response but also in patients with type 1 or type 3 responses. On the other hand, spontaneous syncope with no severe bradycardia was observed in patients with a type 2 HUT response (2 of 4). This observation might be part of an explanation for the results of two recently published randomized studies on pacing therapy in vasovagal patients: In both studies, pacing therapy failed to show effectiveness, even in the subgroup of patients presenting with severe bradycardia at HUT (8,9).
The ATP test has been proposed (16) for assessing the role of cardioinhibition in vasovagal patients. In our tilt-positive patients, 28% also had a positive ATP test, which is in agreement with previous studies (22,25). Of these patients, four experienced syncope and one had a cardioinhibitory vasovagal syncope. Therefore, a positive ATP test did not predict severe bradycardia during vasovagal syncope. In a study of patients with positive ATP test, Donateo et al. (22) observed ILR-documented syncopal relapse in eight patients who also presented with a positive response during tilt testing. Bradycardia was the predominant heart rhythm observed, with sinus arrest observed in five cases and sinus bradycardia in two cases. The small number of patients with a positive ATP test in our study might explain the difference between our results and the ones from Donateo et al. (22). Nevertheless, in the present study, a positive ATP test was observed in two patients who had syncope with bradycardia: Patient #18 and Patient #4 (for her first syncope).
Role of ILR in vasovagal syncope. Knowledge of the heart rhythm at the time of vasovagal syncope resulted in the choice of therapy in seven patients, including pacemaker implant or vasoactive drugs. Our study was not designed to evaluate the appropriateness of the ILR-guided therapy, and this question might be addressed in future studies. Despite a seemingly typical vasovagal presentation of syncopes, the ILR documented one syncopal episode that was not related to a vasovagal mechanism (Patient #4, first syncope). Such observations are of clinical importance in highly symptomatic patients and emphasize the role of the ILR in the choice of treatment in this subset of patients, although the presentation seems typically vasovagal.
Study limitations. A limitation of the present study is that the number of patients is small and they were highly selected (only 25 of 497 patients referred for suspected vasovagal syncope). This is due to the strict inclusion criteria, based on a typical vasovagal clinical history and the severity of the symptoms, leading to the implantation of the ILR. However, the relatively long follow-up period allowed documentation of a large number of syncopal attacks.
Owing to this limitation, the results should not be extrapolated to larger and less selected groups of vasovagal patients before being compared with the results of larger studies.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T Pezawas, G Stix, J Kastner, B Schneider, M Wolzt, and H Schmidinger Implantable loop recorder in unexplained syncope: classification, mechanism, transient loss of consciousness and role of major depressive disorder in patients with and without structural heart disease Heart, April 1, 2008; 94(4): e17 - e17. [Abstract] [Full Text] [PDF] |
||||
![]() |
Authors/Task Force Members, P. E. Vardas, A. Auricchio, J.-J. Blanc, J.-C. Daubert, H. Drexler, H. Ector, M. Gasparini, C. Linde, F. B. Morgado, et al. Guidelines for cardiac pacing and cardiac resynchronization therapy: The Task Force for Cardiac Pacing and Cardiac Resynchronization Therapy of the European Society of Cardiology. Developed in Collaboration with the European Heart Rhythm Association Europace, October 1, 2007; 9(10): 959 - 998. [Full Text] [PDF] |
||||
![]() |
Authors/Task Force Members, P. E. Vardas, A. Auricchio, J.-J. Blanc, J.-C. Daubert, H. Drexler, H. Ector, M. Gasparini, C. Linde, F. B. Morgado, et al. Guidelines for cardiac pacing and cardiac resynchronization therapy: The Task Force for Cardiac Pacing and Cardiac Resynchronization Therapy of the European Society of Cardiology. Developed in Collaboration with the European Heart Rhythm Association Eur. Heart J., September 2, 2007; 28(18): 2256 - 2295. [Full Text] [PDF] |
||||
![]() |
R. Maggi, C. Menozzi, M. Brignole, C. Podoleanu, M. Iori, R. Sutton, A. Moya, F. Giada, S. Orazi, and N. Grovale Cardioinhibitory carotid sinus hypersensitivity predicts an asystolic mechanism of spontaneous neurally mediated syncope Europace, August 1, 2007; 9(8): 563 - 567. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Viskin, D. Justo, and A. Halkin Should the 'adenosine-challenge test' be part of the routine work-up for syncope? Europace, August 1, 2007; 9(8): 557 - 558. [Full Text] [PDF] |
||||
![]() |
J. B. P. Stephenson Clinical Diagnosis of Syncopes (Including So-called Breath-Holding Spells) Without Electroencephalography or Ocular Compression J Child Neurol, April 1, 2007; 22(4): 502 - 508. [Abstract] [PDF] |
||||
![]() |
M. Brignole, R. Sutton, C. Menozzi, R. Garcia-Civera, A. Moya, W. Wieling, D. Andresen, D. G. Benditt, N. Grovale, T. De Santo, et al. Lack of correlation between the responses to tilt testing and adenosine triphosphate test and the mechanism of spontaneous neurally mediated syncope Eur. Heart J., September 2, 2006; 27(18): 2232 - 2239. [Abstract] [Full Text] [PDF] |
||||
![]() |
Tilt-Test Results Don't Predict Heart Rhythm During Spontaneous Vasovagal Syncope Journal Watch Neurology, May 4, 2006; 2006(504): 10 - 10. [Full Text] |
||||
![]() |
Tilt-Test Results Don't Predict Heart Rhythm During Spontaneous Vasovagal Syncope Journal Watch Cardiology, March 23, 2006; 2006(323): 2 - 2. [Full Text] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |