Syncope is defined as the sudden loss of consciousness associated with the absence of postural tone, and is usually followed by a complete and rapid recovery. It is a very common clinical problem. The overall incidence of this first report of syncope is that it occurs in 6.2 per 1,000 patients; the incidence rate increases with age, especially after age 70.¹ The estimated 10-year cumulative incidence of syncope is 6%, and up to 22% of patients experience recurrent syncope.

       The differential diagnosis of syncope is extensive (Figure 1).

Figure 1.

Differential diagnosis for syncope.

The management of the patient who presents with syncope begins with the history and physical examination and almost always includes an electrocardiogram (ECG) and echocardiogram. The initial step in diagnosis involves distinguishing cardiac from non-cardiac causes of syncope.

Cardiac Syncope
       Cardiac causes of syncope include disorders of autonomic function, such as neurally-mediated syncope (e.g., vasovagal syncope, carotid sinus hypersensitivity), chronic orthostatic intolerance (i.e., postural orthostatic tachycardia syndrome [POTS]), and orthostatic hypotension (secondary to volume depletion, systemic illness, use of a vasoactive drug or pure autonomic failure/multiple system atrophy), disorders related to obstruction to blood flow and arrhythmias (bradyarrhythmias and tachyarrhythmias). Patients in whom syncope results from obstruction to blood flow are readily identified by echocardiography and rarely pose a diagnostic or therapeutic dilemma. Therefore, the remainder of this update will focus on differentiating patients with syncope due to autonomic dysfunction from those with an arrhythmic etiology. This differentiation has prognostic implications, since the mortality of patients with arrhythmic syncope is double that of patients without a history of syncope or those in whom syncope is vasovagal in etiology.¹

       Disorders of autonomic function. These disorders are usually not associated with underlying structural heart disease. Therefore, in a patient with a normal ECG and no structural heart disease, one of the disorders of autonomic function should be strongly considered (Figure 2).

Figure 2.

An algorithm for the evaluation of patients with suspected cardiac syncope.

       The most common disorder of autonomic function is neurally-mediated syncope, which includes vasovagal and carotid sinus syncope. In both of these conditions, syncope results from bradycardia and/or hypotension. It is important to note that in many patients with carotid sinus hypersensitivity, the hemodynamic alterations may be manifest only in the upright state, and that in nearly a third of patients, the major hemodynamic effect is a pure vasodepressor response.² Therefore, in diagnosing carotid sinus hypersensitivity, patients should be evaluated while supine and upright with the use of one of several commercially available devices capable of recording blood pressure non-invasively on a beat-to-beat basis.
       Chronic orthostatic intolerance, formerly referred to as POTS, is characterized by pronounced orthostatic tachycardia without associated hypotension. The predominant symptoms include lightheadedness, dizziness, palpitations, chest pain and syncope. In some patients, this condition is related to a defect in the norepinephrine transporter, which results in an increase in the norepinephrine concentration.3 In the upright state, norepinephrine levels characteristically exceed 600 pg/ml.^3,4
       Although most commonly a result of volume depletion, orthostatic hypotension may represent a sign of generalized autonomic dysfunction. An identifiable cause is often present, such as a systemic disease (e.g., diabetes mellitus), toxic agents (e.g., alcohol) or use of a vasoactive medication. Often a cause cannot be identified; in these instances, one must consider pure autonomic failure (when other neurologic features are absent) or multiple system atrophy (associated with other neurologic features, Shy-Drager syndrome).⁵
       Tilt table testing is useful in differentiating these syndromes (Figure 3).

Figure 3.

The tilt test response in various disorders of autonomic function. NMS = neurally-mediated syncope; COI = chronic orthostatic intolerance; PAF = pure autonomic failure.

In neurally-mediated syncope, in response to upright tilt, the heart rate and blood pressure initially remain stable. However, there is then an abrupt decline in the heart rate and/or blood pressure, which results in syncope. In contrast, patients with chronic orthostatic intolerance have an immediate increase in heart rate (greater than 30 bpm and to more than 120 bpm) without a change in blood pressure, and patients with orthostatic hypotension due to autonomic dysfunction exhibit a progressive decline in blood pressure (greater than 20 mmHg) without a change in heart rate.⁶
       Although tilt testing is widely used to evaluate patients with suspected neurally-mediated syncope, there is no standardized tilt test protocol.⁷ Currently used protocols usually consist of an initial drug-free tilt phase followed by a phase using pharmacologic provocation, most commonly with isoproterenol, nitroglycerin, or adenosine. At our institution, we abandoned nitroglycerin tilt testing because of concerns over test specificity.^8,9 We now routinely use adenosine to facilitate the induction of neurally-mediated syncope.¹⁰ Adenosine, like isoproterenol, has sympathomimetic effects, which are mediated through baroreflex and chemoreceptor activation. Single-stage adenosine tilt testing has a diagnostic yield comparable to a two-stage test protocol using drug-free and isoproterenol tilt stages.¹⁰ An attractive feature of adenosine tilt testing is that in takes only three minutes to complete.
       The major problem with tilt testing remains its poor sensitivity when using a protocol that maintains high specificity.¹¹ In addition, the sensitivity of tilt testing worsens significantly as the patient population being evaluated becomes older.¹² Furthermore, prospective studies have demonstrated that in patients without structural heart disease who present with syncope, the clinical outcome is similar, irrespective of the tilt test response.¹³ As a result, the clinical role for tilt testing remains controversial.
       Another frustrating area relates to the treatment of patients with vasovagal syncope. For one, the natural history of vasovagal syncope is variable. Multiple events often occur over a relatively short period and are followed by longer, relatively symptom-free periods.¹⁴ Interestingly, the frequency of syncopal episodes decreases substantially following tilt testing, whether patients are treated or not.¹⁵ Therefore, it has been difficult to evaluate the therapeutic effect of any single intervention.
       The need to initiate therapy needs to be individualized; in general, therapy is attempted in patients with recurrent episodes of syncope or in those in whom syncope is associated with trauma. A variety of therapies have been suggested for patients with vasovagal syncope. The most commonly used treatments include beta blockers, selective serotonin reuptake inhibitors, b-agonists, anticholinergic agents, mineralocorticoids, and permanent pacemakers.¹⁴ Unfortunately, most of these therapies have been evaluated in only small, short-term, uncontrolled and unblinded studies.
       To date, only the SSRI agent paroxetine has been shown in a double-blind, randomized, placebo-controlled clinical trial to significantly reduce the incidence of recurrent syncope.¹⁶ However, a confirmatory study is still needed. Although there was great initial enthusiasm for using permanent pacemakers (with enhanced features such as rate-drop programming), the negative results of the North American Vasovagal Pacing Study II effectively eliminated the routine use of pacing for patients with vasovagal syncope.¹⁷

       Arrhythmias. An important arrhythmic cause for syncope is a bradyarrhythmia, such as sinus bradycardia resulting from sinus node dysfunction or atrioventricular (AV) block resulting from AV node or His-Purkinje system dysfunction. Electrophysiologic (EP) studies are most useful when the patient’s baseline ECG demonstrates sinus bradycardia or conduction system disease (prolonged PR interval and/or bundle branch block). However, although EP testing has high specificity, it too lacks sensitivity for diagnosing patients at risk for symptomatic bradycardia. For example, with the use of an implanted loop recorder, it has been possible to demonstrate that in a third of patients with bundle branch block (specifically, right bundle branch and fascicular block or left bundle branch block), recurrent syncope results from paroxysmal AV block, even though the EP study is “normal.”¹⁸
       The disappointing yield of tilt testing and EP testing in patients with syncope in the absence of structural heart disease has necessitated the use of alternative diagnostic strategies. The most promising is the insertable loop recorder, a leadless “black box” (61 x 19 x 8 mm; 8 cc; 17 gm) with two self-contained electrodes that is implanted subcutaneously in a left parasternal location. The unit records a single-lead ECG continuously in 42-minute “loops” (either automatically or when activated by a patient) over its 14-month battery life. In a recent randomized study, a strategy of early loop recorder implantation was shown to be superior to a more conventional strategy of EP and tilt testing in making a definitive diagnosis (which was usually a form of bradycardia) in patients without structural heart disease presenting with syncope.¹⁹
       The yield of EP testing in patients with syncope is greatest in those with underlying structural heart disease, especially in those with an underlying ischemic cardiomyopathy. Sustained monomorphic ventricular tachycardia is inducible in 40% of these patients.²⁰ Since these patients are at high risk for sudden cardiac death, an implantable cardioverter-defibrillator (ICD) is usually inserted. (We have found that induction of ventricular fibrillation, especially when achieved with triple ventricular extrastimuli, is not of prognostic significance.²¹) Within 15 months of ICD implantation, 40% of patients receive an appropriate therapy from the ICD for management of recurrent ventricular tachycardia or fibrillation.²⁰ The risk of recurrent events is greatest in patients with a prolonged QRS duration (>= 120 msec).²² Surprisingly, inducible patients, despite treatment with an ICD, have a higher overall mortality than non-inducible patients.²⁰
       This EP-guided approach is most effective for patients with an ischemic cardiomyopathy. In patients with non-ischemic cardiomyopathy, the negative predictive value of EP testing is poor. Patients with non-ischemic cardiomyopathy whose left ventricular ejection fraction is less than 30% appear to have a particularly high mortality.²³ Therefore, ICD implantation has been advocated in all of these patients who present with syncope, irrespective of the findings at EP testing.²⁴ A similar approach is probably warranted in patients with high-risk conditions such as hypertrophic cardiomyopathy, Brugada syndrome and congenital long QT syndrome who present with syncope.

Non-Cardiac Syncope
       Syncope due to a non-cardiac etiology is associated with a benign prognosis. The most common non-cardiac causes are seizures and psychiatric disorders, such as panic and anxiety attacks. Seizures may be difficult to distinguish from cardiac syncope, since up to 12% of patients with syncope have an associated convulsive reaction (“convulsive syncope”).²⁵ In fact, up to 30% of patients diagnosed with epilepsy may in fact suffer from convulsive syncope. Features suggestive of convulsive syncope include relationship to the upright posture, absence of tonic-clonic movements or automatisms, rapid recovery, absence of post-ictal disorientation, and a normal interictal EEG.²⁶ The entity of convulsive syncope should be considered in all patients with suspected epilepsy, especially when the history is atypical for epilepsy or when there is an inadequate response to anticonvulsant medication.

Conclusions
       Syncope remains a common clinical problem. In most patients, a diagnosis can be readily made on the basis of the initial history and physical examination in conjunction with review of the patient’s ECG and echocardiogram. Patients with structural heart disease who present with syncope are at high risk of death; these patients, in general, should undergo EP testing and be considered for ICD implantation.
       In patients without structural heart disease, syncope is usually the result of bradycardia. Although these patients have traditionally undergone “directed” EP testing (i.e., a sinus node and His-bundle evaluation) and tilt testing, this approach has poor sensitivity. Recent data suggest that loop recorder implantation may be a preferable initial strategy, especially in those patients in whom a definitive diagnosis is essential, such as those in whom syncope is recurrent or associated with trauma. However, in many of these patients, bradycardia is a manifestation of neurally-mediated syncope, a disorder for which effective therapeutic options remain limited.