Syncope is one of common medical problems, defined as a sudden and temporary loss of consciousness with a deficit of postural tone which is spontaneously recovered without electrical or chemical cardioversion^1,2). Syncope is experienced by 3% of adult males and 3–5% of adult females and syncope is the main reason for admission in 1% of inpatients of general hospitals and in 3% of patients visiting an emergency room³⁾.

The cause of syncope is often very difficult to determine. Therefore, syncope remains still unexplained in 40–50% of the cases, in spite of thorough noninvasive and invasive diagnostic investigations, because the mechanism or cause of syncope is various and complex, the episodes are brief and unpredictable, spontaneous remission rate is high and there is no reliable diagnostic method^4–7).

Unexplained syncope or syncope of unknown origin (SUO) has been most commonly reported to be vasovagal syncope and head-up tilt test (HUT) has been utilized as a valuable diagnostic test of vasovagal syncope^2,8–13). The sensitivity of HUT, however, is rather low, ranging from 32 to 70%. Therefore, several protocols of HUT with or without pharmacologic provocation were developed to improve the sensitivity, thereby reaching 90%^2,8–16). This study was performed to determine the value of the HUT-isp in evaluation of patients with SUO and predicting the efficacy of long-term drug therapy using beta-blocker, aminophylline or disopyramide.

Twenty-nine out of 67 patients presenting with SUO had various prodromal symptoms such as palpitation in 12 patients, cold sweating in 10, dizziness in 9, dyspnea in 8, nausea in 5, abdominal pain in 2, general weakness in 2 and visual blurring in 2.

Thirty-six patients who developed syncope during the HUT-isp complained of various discomforts such as dizziness in 24 patients, nausea in 18, blurred vision in 14, diaphoresis in 12, headache in 2 and general weakness in 2, as a prodromal symptom. Fifty-six (83.6%) of 67 patients developed positive vasovagal response to the HUT-isp, which was also observed in 10 (33.3%) of 30 control subjects (Table 1). Type of positive responses in the patient group (n=56) was vasodepressive type in 41 (73.2%), mixed in 11 (19.6%), and cardioinhibitory in 4 (7.1%) (Table 1). The sensitivity of the HUT-isp was 83.6%, specificity 66.7%, accuracy 78.4% and positive predictive value 84.8%.

The positive responses in patients group were developed in the 3rd stage in 43 (76.8%) patients, 2nd stage in 9 (16.1%), and 1st stage in 4 (7.1%) and, in control subjects, 3rd stage in 7 (70%), 2nd stage in 1 (10%), and 1st stage in 2 (20%). There was no significant difference in the degree of fall in BP and HR at the onset of positive responses between patients group and control group (Table 2).

One of 3 drugs (carteolol 10 mg/day, aminophylline 450 mg/day, disopyramide 300 mg/day) was randomly given to the 56 patients with a positive HUT-isp and their short-term effect was evaluated with a repeat HUT-isp in 27 patients after 2–4 weeks on medication (Fig. 1). Carteolol turned out to be effective in 12 (85.7%) of 14 patients, aminophylline in 1 (14.3%) of 7 and disopyramide in 7 (58.3%) of 12 (Fig. 1, 2). None of the 20 patients in whom the repeat HUT-isp became negative after medication developed a recurrent syncope during the follow-up period of 175±212 days (26–623 days).

Syncope is caused by various cardiac and non-cardiac diseases. Aortic stenosis, pulmonary embolism, mitral stenosis, ventricular tachycardia, cardiac conduction disturbances, myocardial infarction and mitral valve prolapse are reported as important causes of cardiac syncope. Non-cardiac causes of syncope are epilepsy, postural hypotension, cerebrovascular disease, carotid sinus hypersensitivity, vasovagal syncope, drug, hyperventilation and hypoglycemia^3,18). However, vasovagal syncope is known as the most common cause of syncope in otherwise healthy people. Recently, HUT has been used as a useful test to evaluate patients with unexplained syncope, although the underlying mechanisms of tilt-induced hypotension in patients with vasovagal syncope are not clearly understood yet¹⁹⁾.

When a normal man is standing, about 300 to 800 ml of blood is pooled in the lower extremities, resulting in decrease of venous return, cardiac output and temporary decrease of the arterial blood pressure in order. This is normally compensated by baroreceptor reflex-mediated increase in the sympathetic outflow that increases cardiac contractility, heart rate and peripheral vascular resistance. However, in a susceptible subject, activation of the left ventricular mechanoreceptors (C-fibers) due to increased cardiac contractility could override the baroreceptor reflex, resulting in peripheral vasodilation and/or bradycardia due to the resultant diminution of sympathetic tone and enhancement of parasympathetic activity^{20,21,34,38,40)}.

Several protocols of HUT which are different in tilt angle, duration of the test and drug used for provocation have been reported. We performed HUT with isoproterenol provocation to stimulate the mechanoreceptors in the heart more vigorously and to thereby reproduce reflex vasodilation and/or bradycardia²²⁾.

In our study, the HUT-isp was positive in 83.6% of SUO patients and in 33.3% of control subjects. Therefore, the sensitivity of our protocol of HUT-isp in diagnosing vasovagal syncope from SUO was 83.6%, specificity 66.7%, accuracy 78.4% and positive predictive value 84.8%. Four (7.1%) of the 56 patients with a positive HUT-isp were positive in the 1st stage of the HUT-isp with no isoproterenol infusion. The positive rate of 7.1% in baseline HUT with no provocation in our study is much lower than the 32–70% which were previously reported by other investigators^{2,8–16,22–27)}. This difference may be explained partly by the fact that this study included patients with only one episode of syncope and the duration of the baseline tilt was shorter than the others (20 min vs 40–60 min). According to the reports, the false positive rate of HUT ranged from 5 to 15%, which largely depends on the duration (30–90 min) and the tilt angle (40–80 degrees)^{2,10–12,23–25)}. In our study, false positive rate was 33.3%, which is somewhat higher than that of the previous report¹²⁾. This relatively high false positive rate in our study was thought to be caused by the higher tilt angle (80 degrees) and higher infusion dose of isoproterenol.

Several studies have reported that beta-adrenergic receptor blockers, aminophylline or disopyramide can be used for long-term management of vasovagal syncope^23,28). The exact mechanisms by which these drugs prevent vasovagal syncope are unclear. It has been, however, suggested that beta-blockers can lower the discharge frequency of mechanoreceptors and by allowing unopposed activation of alpha receptors, also elicit a peripheral vasoconstriction^34,35). Disopyramide is known to have negative inotropic, anticholinergic and peripheral vasoconstricting effects, which may account for its efficacy in preventing vasovagal syncope³⁶⁾. Theophylline is the selective antagonist of adenosine A1 receptor through which adenosine evokes negative inotropic and chronotropic effect on the heart³⁷⁾. Adenosine is an endogenous cardioprotective agent, particularly in the event of myocardial ischemia. Although the exact role of adenosine in vasovagal syncope has not been determined yet, prevention of vasovagal syncope with theophylline has been reported.

In this study, carteolol, a beta-blocker, was most effective in preventing vasovagal syncope. The efficacy of oral disopyramide and aminophylline in preventing vasovagal syncope was reported to be 100% and 82%, respectively, in some uncontrolled trials^28,37,41). In contrast, disopyramide was effective in 58% of the patients and aminophylline in 14% in our series. But this result should be interpreted with great caution because these drugs were administered in a small number of patients (12 and 7 cases, respectively) and more importantly, blood level of the drugs was not monitored. The cardiac pacing is barely used since it is not effective in preventing syncope. It is, however, applied to selected patients whose bradycardia or asystole are very serious in order to prevent sudden death^2,23,29).

It is noteworthy that none of the patients who were put on one of 3 drugs based on the results of a repeat HUT-isp to evaluate the short-term drug effect developed a recurrent syncope. This finding is strong evidence that the HUT-isp can be used to assess and guide the long-term drug therapy in patients with vasovagal syncope.

This study has some limitations. First of all, patients with only one episode of syncope were included, even though the episode was very recent (< 4 weeks). Inclusion of this patient population may increase false negativity or positivity of the test since non-recurrent syncope may be provoked by a transient derangement of autonomic nervous system which will neither persist nor recur after recovering from the syncope. Second, our protocol of the HUT-isp is rather aggressive in terms of tilt angle (80 degrees) and isoproterenol infusion dose (5 μg/min), which may contribute relatively high false positivity of our HUT-isp. Third, drug effect was tested without confirming whether the therapeutic blood level of the drugs was achieved or not. Nevertheless, long-term drug therapy based on the results of serial drug testing with a repeat HUT-isp was very successful.

Data from the present study suggest that the HUT-isp is safe and useful to evaluate patients with syncope of unknown origin and to guide long-term drug therapy in vasovagal syncope. However, further studies are needed to improve both sensitivity and specificity of the test and to standardize the protocol.