The optimal medical therapy of patients with vasovagal syncope (VVS) remains controversial. Fluoxetine is effective against anxiety and panic disorders, while its use has shown promising results for VVS. Anxiety sensitivity is a personality trait observed in a considerable proportion of patients with VVS, associated with predisposition to anxiety and panic disorders. Our aim was to examine whether fluoxetine exerts beneficial effects regarding VVS prevention in the subset of patients with anxiety sensitivity.
We assessed 106 patients with typical history of recurrent VVS, without other comorbidities, and a diagnostic, positive head-up tilt test. A psychiatric examination ruled out clinical psychiatric disease. Their psychological, stress-related profile was assessed by the Anxiety Sensitivity Index (ASI) questionnaire, a 16-item questionnaire, assessing fear of anxiety-related sensations, previously studied in VVS. Patients scoring positive for ASI (n = 60, 57% of the population) were randomized in a 2:1 fashion to receive either 10–40 mg fluoxetine daily (n = 40) or placebo (n = 20), and were followed-up for 1 year. A significant difference was observed between patients receiving fluoxetine and those with placebo, regarding the distribution of syncope-free time during the study (P < 0.05). A significant difference was also observed between the two groups regarding presyncopal events and the total number of patients who experienced syncope or presyncope during follow-up.
Sensitivity to anxiety is a common personality trait in recurrent VVS. Fluoxetine is superior to placebo against syncope in these patients. This drug may be a first-line pharmacological treatment for this difficult-to-treat group.
Anxiety sensitivity can be specifically assessed and is observed in the majority of patients with recurrent vasovagal syncope.
Fluoxetine is superior to placebo in its effect to prevent recurrent episodes in such patients.
Fluoxetine can be a first-line therapy for this population, which is the most difficult-to-treat patient group with syncope.
Introduction
The optimal medical therapy of patients with vasovagal syncope (VVS) is not yet fully defined,1 possibly because this is a syndrome with more than one mechanisms of various significance among patients. Peripheral or central mechanisms2–4 may prevail in the pathogenesis of syncope.
Patients with recurrent episodes frequently display subclinical minor psychiatric disorders, mainly anxiety, somatization- and panic- related problems,5,6 as well as psychological distress;7 both are linked to symptom persistence.6,7 Most of these people do not want and may never need psychiatric support. Anxiety-related psychosocial distress, as a specific personality trait, is observed in a considerable proportion of patients with VVS and unexplained syncope.7
Ιn the population of patients with VVS, the superiority of serotonin reuptake inhibitors (SSRIs) over placebo has not been definitively proven.8,9 As it is known, SSRIs exhibit various clinically important central nervous system actions and might be more effective in the treatment of patients with VVS and psychological distress, although without clinically overt psychiatric disorders.
Therefore, we aimed at examining whether fluoxetine, an SSRI with promising results for VVS treatment, exerts beneficial effects relative to placebo in its ability to prevent VVS in the subset of patients testing positive for anxiety-related psychological distress, already identified in VVS by a specific, simple, clinically relevant questionnaire (Anxiety Sensitivity Index, ASI).10
Methods
In a prospective, two-arm, placebo-controlled, parallel design randomized way, we assessed 106 patients with typical history of recurrent VVS, all with a diagnostic, positive head-up tilt test (HUT) with clomipramine drug challenge, as previously described.3 It was decided to perform tilt testing before protocol initiation in order to confirm the diagnosis of VVS by a second test, in addition to the typical history of recurrent VVS, which is—indeed—the gold standard for VVS diagnosis. Patients were eligible if they had age ≥18 years and ≥2 syncopal spells in the 6 months preceding enrolment. Patients with known non-cardiovascular or cardiovascular diseases, taking systematically medicine for any reason, or unable to give informed consent were not studied. The study complies with the Declaration of Helsinki, the local ethics committee has approved the research protocol, and informed consent has been obtained from all study subjects.
A psychiatric examination ruled out clinically significant psychiatric disease. The psychiatric assessment included history taking by an interview which ruled out overt obsessions (unwanted and distressing thoughts or impulses), compulsions (urges to do irrational or apparently useless acts), delusions (fixed false beliefs), distress being expressed in physical symptoms (e.g. headache, abdominal pain), mental symptoms (e.g. phobic behaviour, depression), or social behaviour (e.g. withdrawal, rebelliousness). The psychiatric assessment also included mental status examination, using observation and questions to evaluate several domains of mental function, including speech, emotional expression, thinking, perception, and cognitive functions. No structured questionnaire has been used.
Patients were also not included if, during a 5-min stand test, had postural tachycardia (heart rate increase >30 beats per minute) or orthostatic hypotension (systolic blood pressure decrease >20 mmHg). Systolic and diastolic blood pressure were assessed every min during the first 5 min of HUT. Following this, blood pressure was assessed every 5 min and when deemed appropriate.
Patients' psychological, stress-related profile was assessed following tilt test by the ASI questionnaire, a 16-item questionnaire10 assessing fear of anxiety-related sensations, previously studied in VVS.11 This is based on beliefs that anxiety-related sensations can be harmful; thus, they can lead to additional anxiety, fear, illness, embarrassment, and loss of control. Anxiety Sensitivity Index is a validated scale that specifically differentiates between anxiety and anxiety sensitivity, which is more of a personality trait and less of a ‘disease.’ However, this trait is associated with predisposition to anxiety and panic disorders.10 It demonstrates good criterion validity, internal consistency (α = 0.82–0.91), as well as good test–retest reliability (r = 0.75). A mean of 33 is considered as positive result.7,10
Patients scoring positive for anxiety sensitivity (n = 60, 36 males/24 females) were randomized in a 2:1 way to receive either 10–40 mg fluoxetine daily (n = 40) or placebo (n = 20), and they were all followed for 1 year. A 2:1 randomization was chosen to allocate fewer patients into the placebo group, with only a modest loss in statistical power. We followed an FFPFPF way of randomizing consecutive patients, where F = Fluoxetine and P = Placebo administration. This was done to ensure that the 2:1 randomization will be evenly distributed among consecutive patients, since cold and hot weather (i.e. winter vs. summer) may affect syncope recurrence. In addition, this way of randomization permits blinding.
All patients received standard of care for VVS, including explanation of the risk and reassurance about prognosis of VVS, motivation to drink a high volume of water, avoidance of precipitating factors, and counselling about isometric leg and arm counterpressure manoeuvres, as indicated by current guidelines for syncope.11 Treatment with fluoxetine was initiated with a dose of 10 mg; after 1 week, the dose was increased to 20 mg. The dose was increased after 8 weeks of treatment in case of syncope or severe presyncope relapse. Patients completed a diary of presyncopal and syncopal spells. They were assessed every 3 months for 12 months and were supplied with study drugs at those times. They were asked to notify the study doctor if they experienced a syncopal spell between clinic visits and were interviewed by the syncope clinic within 1 week of the date of the episode. Only syncopal episodes occurring after the first month of treatment were included in the analysis, since this time period is necessary for the effective down-regulation of central serotonergic activity by fluoxetine.
Statistical analysis
The primary outcome measure was the time to the first recurrence of syncope. Patients who discontinued treatment, dropped out of the study or were withdrawn by their physician for reasons related to the treatment were considered treatment failures at that time. A secondary hypothesis was that fluoxetine will reduce the frequency of recurrent syncopal and presyncopal spells relative to placebo.
Continuous variables are presented as mean ± SE. Log-rank test was used to compare the time to symptom recurrence between the 2 groups. Differences of continuous variables between groups were determined by t-test. Forward stepwise discriminant analysis was used to test whether supine systolic blood pressure, its initial upright values and its changes during tilt test predict syncope recurrence. P value of <0.05 was considered statistically significant. STATISTICA software package, version 10, was used for analysis.
Results
Demographical data and clinical characteristics of study patients are depicted in Table 1. No differences were observed between the two treatment groups regarding demographics or symptom characteristics. A relatively low age was observed for both groups, presumably due to the fact that patients with comorbidities were not included in the study. Thirty-five of the 40 patients of the fluoxetine group received the drug at a dose of 20 mg/day. The remaining five patients had only 10 mg prescribed, due to low systolic blood pressure (≤100 mmHg) after 1 week of low-dose fluoxetine. Of these patients, two experienced syncope during follow-up. Of the five patients who had syncope during fluoxetine treatment, the dosage was increased to 30 mg in 3. No patient received 40 mg daily.
Patient demographics and characteristics
| Fluoxetine (n = 40) | Placebo (n = 20) | P-value | |
|---|---|---|---|
| Age (years) | 39 ± 2.4 | 41 ± 2.5 | NS |
| Sex | 24 men/16 women | 11 men/9 women | NS |
| History of syncope spells (n) | |||
| Total | 6 ± 0.9 | 5 ± 1.0 | NS |
| Last 6 months | 2 ± 0.2 | 2 ± 0.3 | NS |
| History of presyncope spells (n) | |||
| Total | 8 ± 2.3 | 9 ± 2.1 | NS |
| Last 6 months | 4 ± 0.2 | 5 ± 0.7 | NS |
| ‘Stress’-related syncope | 29 (72%) | 14 (70%) | NS |
| Baseline supine blood pressure (mmHg) | |||
| Systolic | 123 ± 1.6 | 122 ± 2.4 | NS |
| Diastolic | 74 ± 0.8 | 74 ± 0.9 | NS |
| Baseline supine heart rate (bpm) | 72 ± 1.1 | 70 ± 1.2 | NS |
| Fluoxetine (n = 40) | Placebo (n = 20) | P-value | |
|---|---|---|---|
| Age (years) | 39 ± 2.4 | 41 ± 2.5 | NS |
| Sex | 24 men/16 women | 11 men/9 women | NS |
| History of syncope spells (n) | |||
| Total | 6 ± 0.9 | 5 ± 1.0 | NS |
| Last 6 months | 2 ± 0.2 | 2 ± 0.3 | NS |
| History of presyncope spells (n) | |||
| Total | 8 ± 2.3 | 9 ± 2.1 | NS |
| Last 6 months | 4 ± 0.2 | 5 ± 0.7 | NS |
| ‘Stress’-related syncope | 29 (72%) | 14 (70%) | NS |
| Baseline supine blood pressure (mmHg) | |||
| Systolic | 123 ± 1.6 | 122 ± 2.4 | NS |
| Diastolic | 74 ± 0.8 | 74 ± 0.9 | NS |
| Baseline supine heart rate (bpm) | 72 ± 1.1 | 70 ± 1.2 | NS |
Patient demographics and characteristics
| Fluoxetine (n = 40) | Placebo (n = 20) | P-value | |
|---|---|---|---|
| Age (years) | 39 ± 2.4 | 41 ± 2.5 | NS |
| Sex | 24 men/16 women | 11 men/9 women | NS |
| History of syncope spells (n) | |||
| Total | 6 ± 0.9 | 5 ± 1.0 | NS |
| Last 6 months | 2 ± 0.2 | 2 ± 0.3 | NS |
| History of presyncope spells (n) | |||
| Total | 8 ± 2.3 | 9 ± 2.1 | NS |
| Last 6 months | 4 ± 0.2 | 5 ± 0.7 | NS |
| ‘Stress’-related syncope | 29 (72%) | 14 (70%) | NS |
| Baseline supine blood pressure (mmHg) | |||
| Systolic | 123 ± 1.6 | 122 ± 2.4 | NS |
| Diastolic | 74 ± 0.8 | 74 ± 0.9 | NS |
| Baseline supine heart rate (bpm) | 72 ± 1.1 | 70 ± 1.2 | NS |
| Fluoxetine (n = 40) | Placebo (n = 20) | P-value | |
|---|---|---|---|
| Age (years) | 39 ± 2.4 | 41 ± 2.5 | NS |
| Sex | 24 men/16 women | 11 men/9 women | NS |
| History of syncope spells (n) | |||
| Total | 6 ± 0.9 | 5 ± 1.0 | NS |
| Last 6 months | 2 ± 0.2 | 2 ± 0.3 | NS |
| History of presyncope spells (n) | |||
| Total | 8 ± 2.3 | 9 ± 2.1 | NS |
| Last 6 months | 4 ± 0.2 | 5 ± 0.7 | NS |
| ‘Stress’-related syncope | 29 (72%) | 14 (70%) | NS |
| Baseline supine blood pressure (mmHg) | |||
| Systolic | 123 ± 1.6 | 122 ± 2.4 | NS |
| Diastolic | 74 ± 0.8 | 74 ± 0.9 | NS |
| Baseline supine heart rate (bpm) | 72 ± 1.1 | 70 ± 1.2 | NS |
Two patients of the fluoxetine arm discontinued treatment due to side-effects (arterial hypotension in one and gastrointenstinal discomfort in the other), while one subject was lost to follow-up. One patient of the placebo arm discontinued treatment due to nausea.
The effect of fluoxetine vs. placebo on syncope recurrence.
The effect of fluoxetine vs. placebo on presyncope recurrence.
Discussion
The main finding of the present study is that fluoxetine treatment is more effective than placebo as a first-line pharmaceutical treatment of patients with recurrent VVS and psychosocial distress of the anxiety sensitivity type. Patients with recurrent episodes often exhibit this personality trait.
Anxiety sensitivity and vasovagal syncope
An important issue in this study regarding VVS is the distinction between anxiety and anxiety sensitivity.
Anxiety in general has long been recognized as a predisposing factor for VVS. However, benzodiazepines (the classical anti-anxiety medications) have not been proved better than placebo12 against syncope.
The ASI test used in this study for the evaluation of the specific personality trait of anxiety sensitivity has previously been assessed in VVS and unexplained syncope and has been found positive in up to 25% of the population.7 This trait is related to the notion that anxiety is linked to severe negative effects.13,14 Accordingly, people who believe that anxiety has few or no negative effects may be able to cope with a relatively high level of exposure to anxiety-provoking stimuli. In contrast, people who believe that anxiety has terrible effects, may tend to anxiety reactions that grow in anticipation of severe consequences. Although it is only a personality trait, anxiety sensitivity implies a tendency to show exaggerated and prolonged reactions to anxiety-provoking stimuli. This is relevant to the pathophysiology of VVS, which can be considered a type of exaggerated response to physiological stimuli.
Superiority of serotonin reuptake inhibitors are known to be effective for clinical situations like panic attacks and other anxiety disorders, which all share with VVS the increased anxiety sensitivity as a common predisposing factor.13,14 Moreover, a relationship between the substrate of minor psychiatric disorders and syncope exists, while psychiatric treatment results in the improvement of syncopal and psychiatric symptoms, implying a common pathophysiological and clinical pathway in these diseases.15
Serotonin reuptake inhibition and vasovagal syncope therapy
Drugs acting on the central nervous system by inhibiting the reuptake of serotonin in the synapse like fluoxetine, sertraline, or paroxetine have previously been used in VVS with encouraging results, especially in patients resistant to or intolerant to other therapies.8,9,16 Fluoxetine was compared with propranolol in a randomized, placebo-controlled study conducted in 94 patients and showed no differences in syncope-free period. A significant improvement in quality of life was observed with fluoxetine.8 Paroxetine, another SSRI, was evaluated in a randomized, double-blind, placebo-controlled study in which 68 consecutive patients with recurrent VVS and positive head-up tilt were randomized to paroxetine 20 mg/day or placebo.9 After 1 month of treatment, the response rates (negative tilt) were 61.8% vs. 38.2% (P < 0.001) in the paroxetine and placebo groups, respectively. Paroxetine significantly improved symptoms compared with placebo (17.6% vs. 52.9, P < 0.0001). However, in another study, paroxetine has failed to prevent vasovagal reactions associated with peripheral stimuli (lower body negative pressure) in healthy volunteers,17 suggesting that paroxetine may be more efficacious when central mechanisms prevail in the pathogenesis of VVS, such as was the case in our specifically selected study patients.
Finding that fluoxetine is very effective in patients with VVS scoring positive for anxiety sensitivity adds to our therapeutic armamentarium, because this drug may be used as an effective first-line treatment in a significant percentage of patients with recurrent VVS, as assessed in our present population. The importance of this result is also related to the fact that very few drugs have been more efficacious than placebo in VVS, namely α-adrenergic agonists and pacemakers in highly selected patients.
The effect of fluoxetine may be a class effect of selective serotonin inhibitors, but may also be drug-specific, since fluoxetine—unlike other selective serotonin uptake inhibitors—also acts on central nervous system norepinephrine and dopamine concentrations.18 As it has recently been reported,19 drugs acting on central norepinephrine kinetics may be another promising therapy for VVS.
Limitation of the study
A standardized education protocol, which significantly reduces traumatic injuries and syncope recurrence in patients with VVS,20 has not been applied in the present study. All patients received standard of care for VVS, as indicated by current guidelines for syncope.11 This has been decided in order to keep the protocol as simple as possible.
Conclusions
In conclusion, fluoxetine is superior to placebo in recurrent VVS with anxiety-related psychosocial distress, which is frequently observed in patients with recurrent events. This drug can constitute an efficient first-line pharmacological treatment in this difficult-to-treat group.
Conflict of interest: none declared.
