The cornerstone of the syncope evaluation is a comprehensive history and detailed physical examination (2,11). The history should include any information on current or previous cardiovascular conditions, current medications, comorbid conditions, and a
Young patients (< 35 yr) Neurocardiogenic Psychiatric Situational Epileptic seizures* Long QT Syndrome* Hypertrophic cardiomyopathy* Supraventricular tachycardia* Mid-life (35-65 yr) Neurocardiogenic Cardiac arrhythmias Older patient
Arrhythmic Dysautonomic (orthostatic hypotension) Drug-induced Multifactorial
*Important but less common causes.
thorough family history. The family history should include any history of syncope, seizure disorders, congenital heart disease, premature atherosclerosis, arrhythmias, unexplained death, or Sudden Infant Death Syndrome. Information should be obtained as to the clinical nature of the syncopal episodes. Such information should include the frequency of episodes, relationship to body position and exertion, prodromal symptoms, the presence or absence of any convulsive activity, and the presence or absence of a postictal state. It should be kept in mind, however, that convulsive activity may accompany any condition where cerebral hypoperfusion occurs; true seizure disorders rarely present as syncope. Any observations from witnesses may be useful.
The physician should keep in mind that various psychologic factors may play a significant role in the patient's condition. In all chronic disorders there is often a common and complex interplay between the physical and mental, leading to a "physiology of illness" (7). In some patients, syncope may be purely psychogenic in nature (12).
It is imperative that the physician should explore the potential for drug use as a cause of syncope. Illicit use of recreational drugs, such as cocaine, heroin, methylam-phetamine, and alcohol may result in syncope because of heart rhythm or blood pressure disturbances (2). A number of therapeutic agents may also produce syncope, either by provocation or exacerbation of arrhythmias or by the volume depleting or vascular relaxing effects of antihypertensives.
The physical exam may be extremely helpful in ascertaining an etiology for syncope. Patients must be checked for the presence or absence of orthostasis. Such a maneuver may provide a diagnosis. Carotid sinus massage should be performed if no evidence for carotid occlusive disease is suspected. The remainder of the cardiac examination
Diagnostic Evaluation of Syncope
Further testing or treatment dependent upon data obtained
Stress testing Tilt-table testing Event recorder
Further testing or treatment dependent on data obtained
In "select" patients proceed with EPS, coronary arteriography, or prolonged ambulatory monitoring should focus on identifying significant abnormalities, such as congenital or valvular heart disease, that could potentially be a cause for syncope.
In addition to potentially suggesting a cause for syncope, the history and physical exam may also allow risk stratification to be performed. Although the 1-yr mortality rate for patients with cardiovascular syncope has been identified as approx 30%, this varies depending upon the etiology (10). Patients without primary electrical abnormalities or structural heart disease generally have a better prognosis than patients with structural heart disease or a conduction system abnormality. In addition, the results of the history and physical examination are utilized to direct further evaluation. It is critical to attempt to identify individuals in whom structural heart disease of any sort is present (see Table 3).
A note of special interest shall be made in the evaluation of the patient with a single syncopal episode. It is sometimes incorrectly assumed that a single episode of unexplained syncope is not of significant magnitude to warrant a thorough evaluation. This notion can prove to be very dangerous. Syncope of uncertain etiology can be caused by a wide array of causes, and the prognosis after a single episode of syncope is variable in relation to the underlying cause. A single syncopal episode may be the only warning of a potentially life-threatening disorder.
One common question is whether to hospitalize a patient after an episode of syncope. A patient with known cardiovascular disease, in whom there is a high suspicion of conditions such as ventricular tachycardia (VT) may benefit from hospitalization. In this subpopulation, hospitalization could prevent recurrent symptoms, injury, and even death. However, most patients with syncope do not seem to benefit from hospitalization,
Patients with Syncope Who Should be Considered for Hospitalization
1. The elderly.
2. Patients with suspected or proven structural heart disease.
3. A potential arrhythmic cause of syncope (i.e., Long QT syndrome).
4. Severe orthostatic hypotension.
5. Unexplained syncope that has led to significant injury.
6. Concomitant conditions that require treatment.
7. Syncope associated with new neurologic findings, including weakness or hemiparesis).
and in some the social and economic impact may actually be detrimental. Before hospitalizing a patient with syncope, you must have a clear idea of what diagnostic and therapeutic goals you would like to achieve.
Studies on the effects of hospitalization for syncope have revealed an average length of stay of 6-7 d, usually with no observable benefit (2). In one study of 161 patients with syncope, 78% were hospitalized with an average stay of 6.8 d (13). Of these patients, a cause was identified in only 17%; the cause was cardiac in 7%. The cost of hospitalization can be quite expensive. Mozes et al. reported that the average syncope admission costs up to $38,565 per patient (14). A list of potential syncope patients who might be considered for admission is found in Table 4.
Our opinion is that a baseline 12-lead scalar electrocardiogram (ECG) should be part of the evaluation in all patients with syncope. Although the yield of such testing is not entirely certain, valuable information can be obtained. In younger patients the clinician should evaluate for evidence of pre-excitation, long QT, accelerated AV conduction or conduction block, right ventricular dysplasia, or hypertrophic cardiomy-opathy. In older individuals, evidence for coronary artery disease (CAD) (Q waves), conduction block or bundle-branch block, and hypertensive heart disease are among the abnormalities to look for. Although abnormalities noted may not assign a cause to syncope, they may serve to risk-stratify the patient and direct further evaluation. An example would be that in the evaluation of syncope in a 50-yr-old male patient, the presence of Q waves would direct evaluation towards ischemic heart disease. If the ECG is nonrevealing, further noninvasive evaluation is warranted.
Echocardiography may be very useful as a next step of evaluation. Certainly abnormalities such as HOCM or other structural heart disease would be detected by echocardiography. However, the issue is whether or not routine echocardiography can be recommended.
The majority of data involving screening echocardiography has been drawn from athletes and not persons with prior syncope (15-18). These data suggest a very low yield in identifying significant abnormalities. Murray et al. reported that 10% of the athletes screened had findings that merited medical follow-up (15). However, the majority of these athletes (11 of 13) with abnormalities had only mitral-valve prolapse.
Recchia and Brazilai examined the value of echocardiography in 128 patients with syncope admitted over an 8-mo period (19). Echocardiography was performed in 64% of these patients. In the patients for whom echocardiography was performed, the majority were normal. The authors found, however, that in patients with a normal ECG and no clinical markers for cardiovascular disease, 100% of the echocardiograms were normal. In patients with a clinical or electrocardiographic indicator of cardiovascular disease, 34% had abnormal echocardiograms. The authors concluded that in the absence of a clinical or electrocardiographic predictor of cardiovascular disease, the yield of echo-cardiography was exceptionally low. This conclusion should be somewhat tempered, because 36% of the patients did not have echocardiography; this may have affected the results (19). Also, the study was a retroprospective evaluation. At this point, although the yield of an echocardiogram is likely to be low with a lack of substantial perspective data, there should be sufficient latitude granted to clinical judgment in this area.
Certainly in individuals for whom an ischemic etiology is suspected, exercise testing may be appropriate. However, exercise testing is useful in other subsets of patients as well, and should be considered in individuals for whom syncope occurred during exercise or exertion.
It is now appreciated that in younger patients with structurally normal hearts, exercise-induced neurocardiogenic syncope is a legitimate entity (15,20-24). These patients have the features of neurally mediated syncope present on both exercise and head-upright tilt testing. Sakaguchi et al. found neurocardiogenic syncope to be the leading cause of syncope in patients with syncope associated with exercise (20).
Exercise-related syncope may be caused by other nonarrhythmic, nonischemic etiologies. Both post-exercise asystole (22,25,26) and post-exercise postural hypotension are capable of provoking syncope.
It is also possible to have nonischemic exercise-induced arrhythmia (28-33). In some patients, electrophysiology study may be negative, and exercise may be the only method to provoke arrhythmia.
Ischemic-induced exercise-related syncope may be the result of a permanent substrate or to a transient ischemia itself (33). Although this condition is generally believed to occur in the older population, other age groups may be affected. In a study of sports-related sudden death in individuals with a mean age of 32 yr, 26% were believed to be caused by atherosclerotic heart disease (34). Younger patients without atherosclerotic heart disease may also experience transient ischemia from conditions such as anomalous coronary artery distribution (35,36), and hypoplastic (37) or tunnel coronary arteries (35).
In patients with cardiovascular syncope and no structural heart disease or ischemic heart disease, neurocardiogenic syncope is the most common etiology (38). This has been demonstrated in adults and children (38-40). Much controversy has been generated over the appropriate angle of tilt and duration of tilt. A recent American College of Cardiology (ACC) consensus document recommended an angle of 60-80 degrees with a duration of 30-45 min for the baseline tilt test (41). These parameters appear to provide a reasonable sensitivity. False-positive rates are generally less than 10% (43). Practices regarding pharmacologic provocation vary widely; a number of pharmacologic agents have been used, including isoproterenol, epinephrine, edrophonium, and nitroglycerine.
The use of isoproterenol provocation has also come under attack as being nonspecific; however, this is dependent upon the dosing protocol (43,44). Most experts would agree on a protocol of low-dose isoproterenol infusion. In a recent trial, Natale et al. evaluated 150 control patients in the baseline state, at various isoproterenol doses (47). They found that in the baseline state, a tilt angle of 60-70 degrees provided an excellent specificity. Testing at an 80-degree angle for greater than 10 min was less specific. The addition of a low-dose isoproterenol protocol reduced the specificity minimally (from 92-88%). Higher doses of isoproterenol (3-5 |g/min) substantially affected specificity.
Our current practice is to perform a baseline test at an angle of 70 degrees for 30 min. If necessary, a subsequent isoproterenol test is done in a low-dose protocol. We titrate an isoproterenol dose to increase heart rate by 25-30% over the baseline supine rate and then perform tilt testing at a 70-degree angle for 10 min.
Several investigators have explored the potential of other provocative agents during tilt-table testing, primarily nitroglycerin and adenosine (41). Similar sensitivities have been obtained with these agents, although it may be that a positive test with one of these agents and not another may identify different subgroups of neurocardiogenic syncope (42). Further studies will be necessary to better define the utility of each of these various agents.
The utility of electrophysiology study (EPS) in patients with syncope is largely dependent upon the patient's underlying cardiac substrate. A thorough history, physical exam, and noninvasive testing are likely to identify patients in whom a tachyarrhythmia or bradyarrhythmia is known or suspected to be the causative agent of syncope. Such individuals would include these with ischemic heart disease, other structural heart disease such as cardiomyopathy, or those with known or suspected conduction system abnormalities. Guidelines were established as to the appropriateness of EPS in such patients (43). A class I indication is given to EPS in patients with structural heart disease and syncope that remains unexplained after appropriate initial evaluation. More controversial is the role of EPS in patients with unexplained syncope without structural heart disease.
In general terms, the diagnostic yield of EPS is low in patients with syncope and no structural heart disease (44). In one review (45), the role of electrophysiology testing in syncope of unknown origin was examined in four studies totaling 327 patients. EPS provided a diagnosis for syncope in 71% of patients with structural heart disease and 36% of patients without heart disease (46-49). One study examined nine major series dealing with electrophysiology only in patients with syncope. A diagnostic abnormality was found in 59% of the combined total of 647 patients (50).
Current guidelines provide a class II indication for EPS in patients with recurrent syncope without structural heart disease and with a negative head-up tilt test (43). In patients with no structural heart disease, neurocardiogenic syncope is the more likely etiology (51). In a study evaluating unexplained syncope, patients underwent EPS, and if this was negative, then underwent head-up tilt testing (52). A diagnosis was established in 74% of the patients. In patients with abnormal EPS, structural heart disease was present in 76% of these patients. In contrast, in the group of patients with a negative EPS and a subsequent positive tilt test, only 6% had structural heart disease.
Several reports have described the utility of Holter monitoring for the evaluation of syncope. The predominant use is in substantiating or excluding a dysrhythmia as the etiology of the event(s) (53-63). Many of the studies were summarized in a 1990 review of DiMarco and Philbrick (53). The trials were retrospective studies (54-60), and attempted to correlate symptoms with arrhythmias. Correlation between symptoms and simultaneous ECG findings could only be made on 22% of patients; the remaining 78% had nondiagnostic studies.
Interestingly, between 4% and 30% of patients who were asymptomatic during these recordings had significant arrhythmias. DiMarco and Philbrick concluded that ambulatory Holter monitoring has a relatively low yield unless symptoms are frequent an/or the period of monitoring is prolonged (53).
More recently Pasyk and Srediniawa (61) evaluated 32 patients with syncope utilizing 72-h Holter monitoring. Syncope occurred in 16 of 32 patients. Cardiac arrhythmias were the cause of syncope in 14 of these 16 patients. Therefore, a 72-h monitoring period appeared to provide a reasonable diagnostic yield. Racco et al. (62) prospectively evaluated 134 patients with syncope. Although 102 of 134 patients showed some rhythm disturbance, continuous ECG recording provided a decisive diagnosis in only six patients. They concluded that continuous ECG recording was of limited value in the evaluation of syncope. However, they believed that because of its moderate cost and noninvasive nature, it may prove worthwhile in certain cases.
Beauregard et al. examined the combined use of 24-h electrocardiographic and electroencephalographic monitoring in 45 patients with syncope (63). Isolated cardiac rhythm abnormalities were seen in 21 patients, and none of these were symptomatic. Isolated EEG abnormalities were observed in 11 patients. Simultaneous ECG and EEG abnormalities were seen in four patients. In two of these four patients, an unsuspected etiology for syncope was noted.
In 1986 Braun et al. reported the utility of memory-loop event recording in 100 patients, and 74 of these patients had either presyncope or syncope (64). Sixty-nine patients made recordings, and 56 patients made recordings believed to be diagnostic. We are not aware of prospective direct comparison of loop-event recorders vs Holter monitoring in the evaluation of syncope. However, in a 1996 publication, Kinlay et al. found the cardiac-event recorder superior to 48-h Holter monitoring in patients with palpitations (65). In this study, patients with syncope were excluded.
Our own beliefs regarding Holter monitoring are similar to those of DiMarco and Philbrick (53): unless patients are having frequent episodes and/or the monitoring period is prolonged, the diagnostic yield in syncope is likely to be low. We generally employ loop-event recording more frequently than Holter monitoring.
Another interesting monitoring device is an implantable long-term subcutaneous (SC) monitoring device. Krahn et al. evaluated such a device in 16 patients with syncope for whom ambulatory monitoring, EPS, and tilt-table testing were all nondiagnostic (66). In a monitoring period of 4 ± 4 mo, 15 of 16 patients had syncope and a diagnosis was established for all 15 patients. Nine of these 15 patients had an arrhythmic cause of syncope; 7 of these 9 were bradyarrhythmias, and 2 of 9 had tachyarrhythmias. This device may be of benefit in selected patients with very rare episodes.
Several studies have evaluated the usefulness of signal-average electrocardiography in identifying patients with syncope for whom EPS may be appropriate (67-70). These studies examined the relationship between late potentials and the risk of VT. These studies were summarized by Kjellgren and Gomes in a 1993 review (71). They concluded that late potentials appeared to offer a good sensitivity and specificity in identifying individuals in whom VT was inducible on EPS. However, they cautioned that the utility is variable depending on the underlying substrate. More recently, Morillo et al. evaluated the usefulness of Signal-averaged electrocardiogram (SAECG) head-up tilt-table testing and EPS in 70 patients with recurrent unexplained syncope (72). Overall, a cause for syncope was found in 77% of these patients. A positive SAECG was more likely to be found in patients with a prior myocardial infarction (MI) and lower ejection fraction than patients with a negative SAECG. The patients with a positive ECG were also much more likely to have inducible VT on EPS. They conclude a positive SAECG along with a previous MI and ejection fraction less than 40% were highly predictive of inducible VT.
Was this article helpful?