Dynamic Changes in T‐Wave Amplitude during Tilt Table Testing: Correlation with Outcomes

Abstract

Background: Changes in autonomic tone may play a role in syncope. Autonomic tone has been shown to affect cardiac repolarization in the ECG. Changes in the T wave can be seen during head‐up tilt table (HUT) testing with unknown significance or relationship to outcomes.

Methods: Twelve‐lead ECGs during HUT testing from 150 patients were reviewed from a prospectively collected registry database. ECGs during supine‐rest, 30‐45‐70° tilt, and 5‐minute supine recovery were reviewed. Changes in the T wave, that is, decreased amplitude with or without becoming negative or flipping from negative to positive, were recorded for each stage. Outcomes of the HUT test include nondiagnostic, postural orthostatic hypotension (POH), postural orthostatic tachycardia syndrome (POTS), and vasovagal response (VVR). Age (Younger: <50 year old; Older: ≥50 year old) and gender were analyzed.

Results: Of 150 patients (108 women; 80 Younger), 135 had T‐wave changes during HUT; changes resolved in 114 patients during supine recovery. Changes mostly occurred in inferior and anterolateral leads. POH occurred in 114 patients, POTS in 67, and VVR in 30. T‐wave changes in V₁ inversely correlated with POH (P = 0.005). T‐wave changes in inferior leads II, III, aVF and in anterolateral leads V₃–V₆ positively correlated with POTS (P < 0.05). Female gender and younger age correlated with POTS independent of the leads (P < 0.05). Concomitant T‐wave changes in V₅ and V₆ correlated with VVR; changes in aVF also correlated with VVR (P < 0.05).

Conclusions: Dynamic T‐wave changes during HUT testing in inferior and anterolateral leads are associated with POTS and VVR independent of age and gender. Changes in autonomic tone may play a role and need further study.

Head‐up tilt table (HUT) testing is commonly performed in evaluating syncope, especially if neurocardiogenic syncope or postural orthostatic hypotension (POH) is suspected and is useful in the diagnosis of postural orthostatic tachycardia syndrome (POTS). Static levels of, and dynamic changes in, autonomic tone are hypothesized to play a major role in the pathogenesis of these conditions, the mechanisms are not fully understood. It is suggested that a faster rise in adrenaline levels occurs in those with vasovagal faints compared with those who do not faint under similar circumstances. ¹ Others found that norepinephrine levels increased to a much smaller extent than did epinephrine levels during HUT testing or with spontaneous syncope, suggesting that a sympathoadrenal imbalance precedes syncope. ²

Autonomic tone has recently been suggested to affect cardiac repolarization. ³ , ⁴ , ⁵ , ⁶ , ⁷ Heterogeneity in cardiac repolarization may be seen on a surface electrocardiogram (ECG); several markers of heterogeneity in repolarization have been studied including ST height, QT or JT dispersion, T peak to T end, and T‐wave complexity among others. Different methods have been used to elucidate the effects of autonomic tone on cardiac repolarization, including heart rate variability analysis of patients with early repolarization syndrome, analysis of the effects of medications such as isoproterenol, atropine, or propranolol that affect overall autonomic tone, and analysis of patients with spinal cord injury (with impaired sympathetic outflow). It has been suggested that HUT testing may be useful in a heart transplant population to uncover increased dispersion in ventricular repolarization, representing heterogenous sympathetic reinnervation. ⁸

Alterations in T‐wave amplitude are frequently seen during HUT testing. The mechanism behind these changes is not fully understood, but they may be due to dynamic changes in autonomic tone. The diagnostic or prognostic significance of these changes are unknown. The current study analyzes the T‐wave changes during HUT testing and their relationships with outcomes of the test. Alterations in T‐wave amplitude are hypothesized to correlate with outcomes of the test.

METHODS

In a retrospective analysis, the results of 12‐lead ECGs during HUT testing from 150 consecutive patients who received a HUT test as part of syncope evaluation were reviewed from a prospectively collected registry database in a study approved by the Institutional Review Board. Exclusion criteria include those not being to complete the HUT test because of reasons other than symptoms associated with syncope (i.e., back pain or unable to obtain IV access). ECGs were recorded during a protocol standard at this institution: supine rest, 30° tilt (2 minutes), 45° tilt (2 minutes), 70° tilt (45 minutes), and 5‐minute supine‐recovery stages. All ECGs were manually reviewed by an experienced cardiologist. Alterations in the T wave throughout tilt and recovery stages were recorded, including decreased amplitude with or without becoming negative or flipping from negative to positive (Fig. 1).

Figure 1.

Lead III during each stage of the HUT test. Note that in this example, T‐wave changes are still initially evident after return to supine recovery with normalization of heart rates (compare the T wave in 45° tilt and supine recovery Minute 1 with similar heart rates).

Standardized outcomes of the HUT test based on blood pressure and heart rate changes were included. POH was defined as either a 20 mmHg fall in systolic pressure or 10 mmHg fall in diastolic pressure, POTS was defined as a sustained heart rate increase of greater than 30 beats/minute or an increase to 120 beats/minute or greater, and vasovagal response (VVR) was defined as a sudden drop in both blood pressure and heart rate. Patients were divided into two age groups: Younger: <50 years old, Older: ≥50 years old. Gender was also included in the analysis. Multivariate analysis of variance was used to determine the relationship between each factor and outcomes of the HUT test, P <0.05 was considered significant.

RESULTS

Of 150 consecutive patients, 108 were female and 80 were Younger. T‐wave alterations occurred in 135 patients at any point during the test; these changes resolved in 114 patients after returning to supine position during recovery. In 21 patients, T‐wave changes lasted through the end of the 5‐minute recovery.

T‐Wave Alterations Per Lead

T‐wave alterations were reviewed for each lead (Fig. 2). T‐wave changes ranged from decreased amplitude, to becoming negative, or flipping from negative to positive (rare). For example, in lead V₅ there was no change in 58 patients, a decreased T‐wave amplitude was seen in 59 patients, the T wave became negative in 32 patients, and the T wave flipped from negative to positive in 1 patient. Among all leads, changes were most frequently seen in inferior leads II (75 patients), III (91 patients), and aVF (75 patients) as well as anterolateral leads V₄ (81 patients), V₅ (92 patients), and V₆ (97 patients).

Figure 2.

Frequency of T‐wave amplitude changes per lead.

Outcome Analysis

POH, POTS, and (actual or impending) VVR were included in the analysis of outcomes. When considering only one main diagnosis per test (if subjects had VVR, then VVR was the main diagnosis, otherwise if subjects had both POTS and POH, then POTS was the main diagnosis), then the distribution of main diagnoses was VVR in 30 patients, POTS in 50, POH in 51, and nondiagnostic in 19. When considering the diagnoses even if present in conjunction with another diagnosis, then POTS was diagnosed in 67 patients, POH was diagnosed in 114, and VVR was diagnosed in 30. Using multivariate analysis of variance, the effects of age, gender, and T‐wave changes in each individual were compared with outcomes of the HUT test (all diagnoses even if present in conjunction with other diagnoses); each individual lead was tested in a model that included age and gender (Table 1). T‐wave changes in V₁ inversely correlated with POH (P = 0.005). T‐wave changes in inferior leads II, III, aVF, and in anterolateral leads V₃–V₆ strongly and positively correlated with POTS (P < 0.05). Female gender and younger age correlated with POTS independent of the effects of ECG leads (P < 0.05). Concomitant T‐wave changes in lateral leads V₅ and V₆ correlated with VVR; T‐wave changes in aVF alone correlated with VVR as well.

Table 1.

Multivariate Analysis of Age, Gender, and Individual Leads Versus Outcomes

	Outcomes
		POH	POTS	VVR
	I	ns	ns	ns
	II	ns	P < 0.001	ns
	III	ns	P = 0.003	ns
	aVR	ns	ns	ns
	aVL	ns	ns	ns
Leads	aVF	ns	P < 0.001	P = 0.0446
	V1	P = 0.005*	ns	ns
	V2	ns	ns	ns
	V3	ns	P = 0.0039	ns
	V4	ns	P < 0.001	ns
	V5	ns	P < 0.001	P = 0.058
	V6	ns	P < 0.001	P = 0.053
	V5+V6	ns	P < 0.001	P = 0.0401
	Age	ns	P = 0.007	ns
	Gender	ns	P < 0.001	ns

ns = nonsignificant; *P value. = inverse correlation (T‐wave changes in V1 inversely correlate with diagnosis of POH).

DISCUSSION

Main Findings

Changes in the T wave occurred throughout the stages of the HUT test, with most changes resolving with return to supine recovery. Since some T‐wave changes persisted even after return to a supine position and a normalization of heart rate (Fig. 1), this argues against a positional or spatial explanation for these generally dramatic changes during tilt; in addition, it argues against these changes being solely due to faster heart rates. Moreover, T‐wave changes occurred mainly in inferior and anterolateral leads and correlated with POTS and VVR independent of the effects of age and gender. In addition, right chest lead V₁ was inversely correlated with POH. Female gender, as well as younger age, was associated with POTS.

ECG Changes

The ECG changes shown during head up tilt table testing are hypothesized to be due to alterations in autonomic tone. However, other mechanisms may affect the ST segment and T wave. Myocardial ischemia can produce similar changes on the ECG; however, subjects in the current study did not complain of chest pain during tilt and many of them were young patients with no obvious risk factors for coronary artery disease. A faster heart rate itself can cause changes in cardiac repolarization. Pacing has been shown to alter ion channel activity; Viswanathan et al. ⁹ showed IKs increased with rapid pacing. Other studies have shown that Ito can have a reactivation time constant of up to 600 ms. ¹⁰ These suggest that during faster heart rates, several ionic channels may be in an inactive state, and this would therefore alter ionic currents; a transmural gradient of ionic currents during faster heart rates may result in alterations in cardiac repolarization. Although this mechanism likely plays a role in altering cardiac repolarization during faster heart rates, it is unlikely to fully account for the generally dramatic ECG changes (i.e., T‐wave inversion) and should also resolve with normalization of heart rates where ion channel reactivation time no longer plays a role. In the present study, the ECG changes in 21 subjects were still apparent even after 5 minutes of supine recovery; ECG changes were still seen during recovery even though heart rates had returned to pretilt levels (Fig. 1, compare the T waves at 45° tilt and supine recovery‐Minute 1 with similar heart rates). Although 135 out of 150 subjects had ECG changes, only 67 of 150 subjects had POTS with heart rate increases of at least 30 bpm. Lastly, a spatial or positional affect on the ECG may play a role. However, ECG changes from this mechanism should have been apparent soon after position change, whereas this was not always the case (Fig. 1, compare 70° Minute 1 and 70° Minute 20) and should resolve after return to supine recovery (Fig. 1); in 21 subjects ECG changes lasted throughout 5 minutes of supine recovery.

Previous Literature

The pathogenesis of syncope is still being explored. Many hypothesize that dynamic changes in autonomic tone play a major role. Benditt et al. ¹ studied 22 patients undergoing HUT testing in syncope evaluation, measured adrenaline and norepinephrine levels throughout the HUT test, and found that a faster rise in adrenaline levels occurs in those with vasovagal faints compared with those who did not faint. Goldstein et al. ² studied 25 patients with chronic orthostatic intolerance and found that norepinephrine levels increased to a much smaller extent than did epinephrine levels during tilt table testing or with supine spontaneous syncope, suggesting that a sympathoadrenal imbalance precedes neurocardiogenic syncope. However, measurement of catecholamines is not an exact measure of autonomic effects during a HUT test or syncope. ¹¹ It can be affected by changes in circulation during tilt testing or syncope, catecholamine kinetics, and does not address the possibility of alteration in the sensitivity of betareceptors.

Recently, it is suggested that a loss of cardiac chaos plays a role in syncope. Suzuki et al. ¹² studied 26 patients with vasovagal syncope and 14 matched controls and measured the Lyapunov exponent calculated from 200 consecutive RR intervals; cardiac chaos is defined as the presence of a positive finite exponent. They found that a decrease in cardiac chaos during a HUT test as measured by the Lyapunov exponent was specific to those with vasovagal syncope regardless of the results of the HUT test.

There has recently been an increased effort to understand factors that affect cardiac repolarization, especially in the study of early repolarization syndrome ¹³ and of Brugada syndrome ¹⁴ with ST‐segment elevations in the precordial leads. It has been recently shown that alterations in autonomic tone can affect cardiac repolarization. Lehmann et al., ⁴ in a study of subjects with spinal cord injury and thus disrupted spinal cord injury, found that the ST‐segment was significantly elevated; the administration of the sympathomimetic agent isoproterenol in these subjects markedly decreased ST‐height. Using heart rate variability analysis to assess autonomic activity, Demir et al. ⁷ studying subjects with early repolarization suggested that an increased vagal tone rather than a decreased sympathetic tone may be responsible for the increased ST height in young men with early repolarization. In showing that subjects with early repolarization had increased aerobic capacity reflecting increased vagal tone, Haydar et al. ⁶ suggested that slower heart rates and higher ST height are of vagal origin in these subjects. In the study of subjects with Brugada syndrome, Miyazaki et al. ⁵ found that ST height was decreased after beta‐adrenoceptor stimulation with isoproterenol.

The current study demonstrates that changes on the ECG relating to ventricular repolarization correlate with outcomes of the HUT test. Since there is a possibility that these changes are influenced by dynamic alterations in autonomic tone, it necessitates further investigation to understand the role of the autonomic nervous system in syncope.

Limitations

One limitation is that though T‐wave changes in general were dramatic and easily visible (i.e., inversion of the T wave), the actual degree of change was not quantified. Likewise, T‐wave complexity or other measurements of repolarization such as ST height or QT dispersion were not analyzed. Another limitation is that the ECG reviewer was not blinded to the results of the HUT test and interpreted the test itself. However, the results of the HUT test (POH, POTS, VVR) are strictly defined by quantitative heart rate and blood pressure changes and are not based on ECG changes.