ECG Repolarization Waves: Their Genesis and Clinical Implications

Abstract

The electrocardiographic (ECG) manifestation of ventricular repolarization includes J (Osborn), T, and U waves. On the basis of biophysical principles of ECG recording, any wave on the body surface ECG represents a coincident voltage gradient generated by cellular electrical activity within the heart. The J wave is a deflection with a dome that appears on the ECG after the QRS complex. A transmural voltage gradient during initial ventricular repolarization, which results from the presence of a prominent action potential notch mediated by the transient outward potassium current (I_to) in epicardium but not endocardium, is responsible for the registration of the J wave on the ECG. Clinical entities that are associated with J waves (the J‐wave syndrome) include the early repolarization syndrome, the Brugada syndrome and idiopathic ventricular fibrillation related to a prominent J wave in the inferior leads. The T wave marks the final phase of ventricular repolarization and is a symbol of transmural dispersion of repolarization (TDR) in the ventricles. An excessively prolonged QT interval with enhanced TDR predisposes people to develop torsade de pointes. The malignant “R‐on‐T” phenomenon, i.e., an extrasystole that originates on the preceding T wave, is due to transmural propagation of phase 2 reentry or phase 2 early afterdepolarization. A pathological “U” wave as seen with hypokalemia is the consequence of electrical interaction among ventricular myocardial layers at action potential phase 3 of which repolarization slows. A physiological U wave is thought to be due to delayed repolarization of the Purkinje system.

INTRODUCTION

The electrocardiogram (ECG) recording is a noninvasive, inexpensive, and useful test widely used in clinical practice. It is used to diagnose heart disorders like cardiac arrhythmias, but it may also be used to detect cardiac effects of noncardiac factors such as electrolyte disturbances and drug toxicity. Although the ECG does not directly record electrical activity from the heart, every single ECG waveform reflects a potential change in the electrical field on the body surface generated solely by the heart that functions as a central electromotive source to produce the change. On the ECG, all waveforms can be categorized into two groups: depolarization and repolarization. Ventricular depolarization is depicted by the QRS complex, and ventricular repolarization is represented by the J (Osborn), T, and U waves. Progress in basic and clinical research of electrical heterogeneity across the ventricular wall in the past two decades has greatly advanced our understanding of the ionic and cellular basis for ventricular repolarization waves and their clinical implications. This article attempts to elucidate some of the elusive ECG syndromes and entities related to ventricular repolarization based on previously demonstrated and published theories.

J (OSBORN) WAVE

The J wave is a deflection with a dome or hump morphology that appears on the ECG after the QRS complex or as a small secondary R wave (R′). Also referred to as the Osborn wave because of Osborn's landmark description in the early 1950s, ¹ the J wave has been commonly observed on the ECG of some species, such as baboons and dogs, under physiological conditions. In humans, the J wave is partially or completely buried in the QRS complex under normal conditions. A prominent J wave on the ECG is often associated with conditions of hypothermia and hypercalcemia. ² , ³ However, a distinct J wave on the ECG can be seen in some normal individuals with the early repolarization syndrome although it is sometimes diagnosed mistakenly as intraventricular block. ⁴ , ⁵

Earlier studies attributed the J wave to a variety of factors including hypoxia, injury current, acidosis, delayed ventricular depolarization, and early ventricular repolarization. ⁶ The fact that the J wave immediately follows the completion of ventricular activation (QRS complex) indicates the presence of a transient transmural voltage gradient during early ventricular repolarization. In the late 1980s, Antzelevitch and his colleagues proposed a difference in repolarization phases 1 and 2 of the action potential between canine ventricular epicardium and endocardium as the basis for the ECG J wave. ⁷ , ⁸ Ventricular epicardium commonly displays action potentials with a prominent notch (spike and dome) largely mediated by a 4‐aminopyridine‐sensitive transient outward current (I_to). The absence of a prominent notch in the endocardium is the consequence of a much smaller I_to. A prominent I_to‐mediated action potential notch in ventricular epicardium but not endocardium may produce a transmural voltage gradient during early ventricular repolarization that could register as a J wave or J point elevation in the ECG. Direct evidence in support of this hypothesis was obtained recently in the arterially perfused canine ventricular wedge preparation. ⁹ As shown in Figure 1, when activation starts with endocardium and spreads transmurally to epicardium, a distinct J wave coincident with the epicardial action potential notch is seen on the ECG. Conversely, if activation starts from the epicardium and travels to the endocardium, the J wave disappears from the ECG because it is buried within the QRS complex. The data obtained from the wedge preparation have also demonstrated a linear correlation between the J wave size and I_to‐mediated action potential notch in epicardium. Therefore, factors that influence I_to kinetics and ventricular activation sequence can modify the J wave on the ECG. The prominent J wave induced by hypothermia is the result of a marked accentuation of the spike and dome morphology of epicardial action potentials probably due to the effect of cold temperatures to slow the activation kinetic of I_to less than that of the L‐type Ca²⁺ current. ⁹ With a normal ventricular activation sequence from endocardium to epicardium, an acceleration of heart rate reduces I_to due to its slow recovery from inactivation, resulting in a decrease in the J wave size. ¹⁰ This property of the J wave is particularly useful in distinguishing the J wave from the terminal part of a notched QRS complex. Since a notched QRS is the consequence of altered ventricular activation, an increase in heart rate may amplify its terminal size. Similarly, some sodium channel blockers that possess additional I_to inhibition, such as quinidine and disopyromide, reduce the J wave size. ¹¹ , ¹² However, they would not be expected to have an effect on the terminal part of a notched QRS.

Figure 1.

Effect of ventricular activation sequence on the J wave on the ECG. (A) When the wedge preparation is stimulated from the endocardial surface with epicardium activated last, a J wave on the ECG is temporally aligned with I_to‐mediated epicardial action potential notch. (B) When the preparation is paced from the epicardial surface with endocardium activated last, the epicardial action potential notch is coincident with the QRS, and a J wave is no longer observed (reprinted from Ref. 9 with permission).

Clinical Syndromes or Diseases that Are Associated with the J Wave (the J Wave Syndrome)

The Brugada Syndrome

The Brugada syndrome is a clinical entity described by the Brugada brothers. ¹³ It manifests clinically as recurrent syncope and sudden cardiac death from ventricular fibrillation (VF). ¹³ , ¹⁴ Although observed worldwide, it is more common in young to middle‐aged Asian males. The ECG of the Brugada patient is characterized by a normal QT interval, apparent incomplete right branch bundle block (RBBB), and ST segment elevation in the right precordial leads (V₁–V₃) or inferior leads (II, III, and aVF) in the absence of myocardial ischemia, electrolyte disturbance, or defined structural heart disease. ¹⁰ , ¹³ , ¹⁴ The apparent incomplete RBBB on the ECG in patients with the Brugada syndrome is actually an accentuated J wave. ¹¹ , ¹⁵ This is supported by the clinical observations that the J wave and associated ST segment elevation in the Brugada patients is significantly influenced by heart rate and autonomic tone. ¹⁰ , ¹⁶

Progress in the research of ventricular repolarization heterogeneity in the past decade has greatly enhanced our understanding of the ionic and cellular basis for ST segment elevation and arrhythmogenesis in the Brugada syndrome. Parallel to the clinical observations by the Brugada brothers in the early 1990s, ¹³ Antzelevitch and his colleagues published several important findings regarding electrical heterogeneity across the canine ventricular wall. ⁸ , ⁹ They found that the I_to‐mediated action potential dome in the epicardium is responsible not only for the J wave on the ECG but also for the development of ST segment elevation and polymorphic ventricular tachycardia (VT) or ventricular fibrillation under certain pathophysiological conditions. An outward shift of currents due either to a decrease in inward currents (I_Na and I_Ca) or an increase in outward currents (I_K–ATP) may lead to two different fates of the I_to‐mediated action potential dome in the epicardium depending on the size of I_to and I_to‐mediated phase 1 magnitude. ¹¹ , ¹⁷ , ¹⁸ When I_to is prominent, as it is in right ventricular epicardium, ¹¹ , ¹⁸ , ¹⁹ an outward shift of currents can accentuate the I_to‐mediated action potential notch in the epicardium to a more negative potential. The L‐type calcium current then fails to activate, and the action potential dome fails to develop, leading to an all‐or‐none repolarization at the end of phase 1. All‐or‐none repolarization, i.e., complete loss of the action potential dome, results in marked abbreviation of action potentials (Fig. 2). ¹¹ , ¹⁹ , ²⁰ In contrast, when I_to is weak as seen in the left ventricular epicardium, an outward shift of currents may only cause a depression (partial loss) of the dome and therefore a minimal change in action potential duration (APD). Loss of the I_to‐mediated action potential dome in epicardium but not endocardium generates a transmural voltage gradient that manifests as ST segment elevation on the ECG. The ST segment elevation due to complete loss of the epicardial action potential dome has a coved morphology as seen in the Brugada syndrome (Fig. 2). ¹¹ , ²¹ , ²² The ST segment elevation due to partial loss of action potential dome in epicardium but not endocardium is usually concave upward or saddleback as seen in the early repolarization syndrome (please see the discussion below). A defect in the sodium channel due to a gene mutation in SCN5A that leads to a decrease in the inward current is linked with the Brugada syndrome. ²³ , ²⁴ It provides the rationale for the use of sodium channel blockers, which increase the outward shift of currents at phases 1 and 2 by inhibition of I_Na, to unmask the concealed form of the Brugada syndrome. ²² , ²⁵

Figure 2.

Loss of I_to‐mediated epicardial action potential dome and ST segment elevation. Upper panel: Simultaneous recordings of a transmural ECG and transmembrane action potentials from one endocardial (Endo) and two epicardial (Epi 1 and Epi 2) sites. Loss of I_to‐mediated epicardial action potential dome induced by pinacidil leads to ST segment elevation. The 4‐aminopyridine (4‐AP), a specific I_to blocker, restores epicardial dome thus normalizing ST segment. Lower panel: Ventricular fibrillation is initiated by a closely coupled R‐on‐T ectopic beat via phase 2 reentry due to complete loss of epicardial dome at Epi 2 but not at Epi 1 (reprinted from Ref.11 with permission).

Complete loss of the epicardial action potential dome may not be uniform; that is, loss of the dome may occur at some sites but not others, probably due to intrinsic differences in I_to in the epicardium and/or regional differences in pathophysiological conditions. Interestingly, when the I_to‐mediated action potential dome is completely lost, the action potential abbreviates by 40–70%. However, when the dome is just about to disappear, the action potential prolongs paradoxically. This leads to a marked repolarization gradient on the epicardium (Fig. 2, lower panel). ¹¹ Propagation of the action potential dome from sites at which it is maintained to sites at which it is lost causes local re‐excitation via phase 2 reentry, leading to the development of a closely coupled “R‐on‐T” extrasystole (Fig. 2). An increase in the transmural dispersion of repolarization (TDR), also present in this condition, may facilitate transmural propagation of the R‐on‐T extrasystole and provide further substrate for the development of polymorphic ventricular tachycardia and fibrillation. ¹¹ It is worth emphasizing that although I_to may not necessarily be the specific ionic target of pathophysiological alterations, like the sodium channel mutation described above, a prominent I_to‐mediated action potential dome is a prerequisite for the development of phase 2 reentry. This is the reason why the Brugada syndrome and idiopathic ventricular fibrillation are predominantly seen in males whose I_to is significantly greater than that seen in females. ²⁶

The Early Repolarization Syndrome

The early repolarization syndrome is characterized by a distinct J wave or J‐point elevation, concave upward ST segment elevation defined predominantly in the left precordial leads V₄–V₆, with a broad and upright T wave. The early repolarization syndrome is a benign ECG phenomenon that is more commonly seen in young healthy men and athletes. ¹⁸ Interest in this well‐known syndrome has recently been rekindled because of its electrocardiographic similarity to the highly arrhythmogenic Brugada syndrome. ¹⁷ , ²⁷

The underlying mechanism for ST segment elevation in the early repolarization syndrome is principally similar to that for ST elevation in the Brugada syndrome. As shown in Figure 3, pinacidil, a potassium channel opener, leads to partial loss of the I_to‐mediated action potential dome in canine left ventricular epicardium. As mentioned above, the ST segment elevation due to partial loss of the action potential dome in epicardium but not endocardium is usually concave upward or saddleback on the ECG. This resembles the early repolarization syndrome observed in humans (Fig. 3B). Experimental data obtained from our laboratory have demonstrated that complete loss of the I_to‐mediated action potential dome in canine left ventricular epicardium rarely occurs due to a much smaller I_to compared to canine right ventricular epicardium. Perhaps, this explains why the early repolarization syndrome is a benign entity.

Figure 3.

Cellular basis for the early repolarization syndrome. (A) Surface ECG (lead V5) recorded from a 17‐year‐old healthy African‐American man. Note the presence of a small J wave and marked ST segment elevation. (B) Simultaneous recording of transmembrane action potentials from epicardial (Epi) and endocardial (Endo) regions and a transmural ECG in an isolated arterially perfused canine left ventricular wedge. A J wave in the transmural ECG is present due to the action potential notch in the epicardium but not the endocardium. Perfusion of the preparation with pinacidil (2 μmol/l), an ATP‐sensitive potassium channel opener, causes partial loss of the action potential dome in the epicardium, resulting in ST segment elevation in the ECG resembling the early repolarization syndrome.

As in the Brugada syndrome, heart rate and autonomic tone modify not only the J wave but also the ST segment elevation in the early repolarization syndrome. ²⁸ However, the ionic basis for the outward shift of currents during action potential phases 1 and 2 that are responsible for partial loss of the left ventricular epicardial action potential dome in the early repolarization syndrome is unknown.

Sudden Cardiac Death Associated with a Prominent J Wave in the Inferior Leads

In the ECG precordial leads V₁–V₃, the J wave and ST segment elevation are the signs that indicate a high risk of sudden cardiac death, i.e., the Brugada syndrome. In contrast, the J wave and ST segment elevation in V₄–V₆ are benign, i.e., the early repolarization syndrome. In other words, pathophysiological alterations of J wave that involve the right ventricle are arrhythmogenic; those that involve the left ventricular anterolateral regions are benign. What would be the consequences of a prominent J wave and ST segment elevation in the inferior leads, i.e., the pathophysiological alterations of the J wave that involve the inferior region of left ventricle? We previously reported a case of recurrent VF in a young Asian male who did not exhibit structural heart disease, but who had a prominent J wave and ST segment elevation in the inferior leads (Fig. 4). ¹⁰ A number of similar cases have been reported since then. ²⁹ , ³⁰ , ³¹ In 1990 and 1993, Aizawa et al. reported several cases of VF in Japanese men that were associated with a “notch” in the late part of the QRS complex in the inferior leads. ³² , ³³ At that time, the authors interpreted this as an intraventricular conduction delay. ³³ We now know that the “notch” was probably a J wave because its properties were consistent with those of the J wave: its size decreased with fast heart rate but increased after a long pause.

Figure 4.

Upper Panel: Twelve‐lead ECG recorded from a 29‐year‐old Asian man who survived cardiac arrest. Prominent J waves and ST segment elevation in inferior leads II, III, and aVF were observed. Lower Panel: The onset of VF in the same patient during his hospital stay (reprinted from Ref.10 with permission).

The above cases may be considered variants of the Brugada syndrome since mutations in the cardiac sodium channel gene SCN5A caused the J wave and ST segment elevation in the right precordial leads as well as in the inferior leads. ³¹ , ³⁴ On the other hand, in the cases of VF in patients with a prominent J wave and ST segment elevation in the inferior leads, early repolarization in the left precordial leads is often present (Fig. 4). ¹⁰ , ³² , ³³ Therefore, all three clinical entities are identical in terms of their ionic and cellular basis and may be referred to an I_to‐mediated J‐wave syndrome. The only difference among these clinical entities is the difference in I_to density and associated J wave size. The fact that patients with hypothermia and a resultant prominent J wave predispose to the development of VF ⁶ indicates that I_to density or kinetics and I_to‐mediated J wave size play a key role in arrhythmogenesis.

T WAVE

According to biophysical principles of the ECG recording, all waves on the body surface ECG are the manifestation of a voltage gradient within the heart. A positive T wave, one whose polarity is the same as the QRS complex, represents such a voltage gradient during ventricular repolarization. Attempts to unlock the cellular basis of the T wave began more than one century ago when the ECG recording technique was in its infancy, ³⁵ , ³⁶ and its genesis had been a matter of debate through the entire 20th century. ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ A concordant polarity of the QRS complex and T wave indicates that ventricular depolarization and repolarization waves travel in opposite directions. Therefore, some regions of ventricle that are depolarized first have to repolarize last in order to generate a concordant QRS and T on the ECG.

In 1976, Cohen et al. demonstrated a difference in action potential duration between the apex and the base of sheep left ventricle, ³⁸ proposing that an apico‐basal voltage gradient was responsible for the positive T wave. ³⁸ , ³⁹ However, the difference in APD between “apex” and “base” observed in this study was actually the difference in repolarization between epicardium and endocardium; they recorded action potentials from the epicardium of the apex and the endocardium of the base. ³⁸ A more recent study has demonstrated that there is no significant apico‐basal voltage gradient on the same myocardial layer of canine left ventricle. ⁴¹

Another possibility for the concordance of polarity of the QRS complex and the T wave is that normal ventricular electrical repolarization may be opposite to ventricular depolarization, a process that can generate a transmural voltage gradient. ⁴⁰ , ⁴² In an elegant study in the early 1980s, Higuchi and Nakaya demonstrated the effect of repolarization sequences across ventricular wall on T wave polarity. ⁴² In their study, cooling of the canine epicardial surface was associated with longer APD in the epicardium and resulted in a negative T wave. In contrast, warming of the epicardium was associated with APD shortening in the epicardium and led to a positive T wave. They concluded that an APD that is at least 40–60 ms longer in the endocardium than in the epicardium is required to generate a positive T wave. ⁴² The contribution of a transmural voltage gradient across the ventricular wall due to the difference in repolarization time between epicardial and endocardial layers is illustrated in Figure 5.

Figure 5.

Effect of ventricular activation on the T wave on the ECG. (A) ECG tracing V4 was recorded from a patient in our EP lab who underwent radiofrequency (RF) ablation for his accessory pathway. In the first three beats, pre‐excitation of the ventricle was associated with a prolonged QRS complex and a small positive T wave. Immediately after the interruption of accessory conduction by RF ablation, the normalization of the QRS was associated with an increase in T wave amplitude and T_p–e interval. (B) Pacing site dependent changes in the QT and T_p–e intervals on the ECG. Pacing from right ventricular endocardium (RVEndoP, conventional pacing) yielded a QT interval of 485 ms. Immediately after switching to left ventricular epicardial pacing (LVEpiP), the QT interval prolonged to 580 ms, and the T_p–e interval also increased. (C) Cellular basis for pacing site dependent changes in the QT and TDR. The QT interval and TDR increased respectively from 284 and 43 ms to 303 ms and 67 ms when switching from the endocardial to epicardial despite no changes in APD. Panels B and C are reprinted with modifications from Ref.77 with permission.

The ionic and cellular basis for the genesis of a transmural gradient was not established until early 1990s when a new subpopulation of cells in canine left ventricular myocardium, i.e., M cell, was discovered by Dr. Antzelevitch and his colleagues. ⁴³ M cells that reside in the deep subendocardium have been well described in canine and human left ventricles. ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ Epicardial, endocardial, and M cells differ from each other primarily by their repolarization properties. The hallmark of M cells is a tendency to have their action potentials prolong disproportionately to epicardium or endocardium in response to slowing of pacing rate or to drugs that prolong the action potential, playing an important role in abnormal ventricular repolarization such as long QT syndrome. The electrical interplay among these three myocardial layers during ventricular repolarization underlies the genesis of various T wave morphologies as demonstrated in the isolated arterially perfused ventricular wedge preparation under both physiologic and pathologic conditions. ⁴¹ , ⁴⁷ , ⁴⁸ , ⁴⁹ With a positive T wave on the ECG, the peak of the T wave (T_peak) is always coincident with the repolarization of epicardium, ¹⁸ , ⁴¹ , ⁴⁸ , ⁴⁹ , ⁵⁰ whereas the end of T wave (T_end) is dictated either by the repolarization of M cells (e.g., in canine ⁴¹ , ⁵⁰ ), or by the repolarization of the subendocardium and endocardium (e.g., in rabbit ¹⁸ , ⁴⁸ , ⁴⁹ ). Regardless, the interval from the T_peak to the T_end (T_p–e) represents the maximal TDR of the ventricle. In fact, the T_p–e interval exhibits a rate dependence in humans ⁵¹ and has served as a useful index in the assessment of arrhythmic risk in patients with the long QT syndrome. ⁵² , ⁵³ , ⁵⁴ , ⁵⁵ It is worthy to emphasize that caution should be taken when extrapolating the data obtained from the isolated ventricular wedge preparation to understanding of the T wave in humans because both nondipolar (local) and dipolar (global) repolarization components may contribute to the genesis of the T wave on the body surface ECG. ⁵⁶ The relative contribution of each of these two repolarization forces to the T wave may vary among different clinical conditions.

Ionic features of M cells that distinguish them from epicardium and endocardium include the presence of a smaller slowly activating delayed rectifier potassium current (I_Ks) but a larger late sodium current (late I_Na). ⁴⁶ However, the rapidly activating delayed rectifier potassium current (I_Kr), the ionic target by most APD prolonging agents, is similar in all three types of cells. Therefore, the difference in the I_Ks:I_Kr density ratio among the three transmural cell types contributes importantly to heterogeneous repolarization across the canine left ventricular wall. ⁴⁶ The mutations of genes that encode I_Ks and I_Kr have been implicated in most types of congenital long QT syndromes. ⁵⁷ It is worth emphasizing that a distinct layer of M cells in the left ventricle may not be seen in smaller species. For example, data obtained from our laboratory have demonstrated that repolarization time increases gradually from left ventricular epicardium to endocardium in this model. ⁴⁸ , ⁴⁹ I_Ks and I_Kr density were similar between subendocardium and endocardium in the rabbit, ⁵⁸ indicating that the endocardium may function like M cells in smaller animals that have thinner left ventricular walls.

Clinical Syndromes and Phenomena that are Associated with the T Wave

Long QT Syndrome (LQTS)

The congenital and acquired LQTS is a cardiac disease characterized by the appearance of a long QT interval on the ECG, notched T waves, U waves, and an atypical polymorphic VT known as Torsade de Pointes (TdP) that can lead to sudden cardiac death. ⁵⁹ , ⁶⁰ , ⁶¹ Since LQTS has been reviewed extensively in the literatures, ⁵⁷ , ⁵⁹ , ⁶⁰ it will only be discussed briefly here.

Genetic analysis has identified mutations of at least six ionic genes located on chromosomes 3, 4, 7, 11, 17, and 21 that are responsible for defects in I_Na (SCN5A, LQT3), I_Kr (HERG and MiRP1, LQT2 and LQT6), I_Ks (KVLQT1 and MinK, LQT1 and LQT5), and I_K1 (Kir2.1, LQT7). ⁵⁷ Acquired LQTS is caused by drugs that either inhibit outward currents of I_Kr and/or I_Ks or enhance the inward current such as I_Na. However, agents that target I_Kr are responsible for most clinical cases of TdP. Preferential prolongation of the cells in the subendocardium and endocardium is thought to cause QT prolongation, the phenotypic appearance of abnormal T waves, pathologic U waves, and the development of TdP. ¹⁸ , ⁴¹ , ⁴⁸ , ⁶² It is now generally accepted that phase 2 early afterdepolarization (EAD) as the initiating mechanism, and functional reentry due to an enhanced TDR as the maintenance mechanism, are responsible for the development of TdP. ⁴¹ , ⁴⁷ , ⁴⁸ , ⁶² , ⁶³

Women exhibit an enhanced susceptibility to develop TdP under conditions of LQTS, ⁶⁴ , ⁶⁵ , ⁶⁶ which may be modulated by age. ⁶⁷ This sexual difference in the TdP risk may be related to the intrinsic differences in ventricular repolarization across the ventricular walls that manifest as the differences in the T wave morphology, e.g., T wave amplitude, slopes of the ascending and descending limbs of the T wave and the T_p–e interval, on the body surface ECG. ⁵¹ , ⁶⁸ , ⁶⁹ However, the physiological significance of a sexual difference in the T wave morphology remains unknown.

T Wave Alternans (TWA)

TWA is characterized by beat‐to‐beat variation of T wave morphology, amplitude, and/or polarity, and it may precede the onset of malignant ventricular arrhythmias. ⁷⁰ As expected, TWA is often associated with a beat‐to‐beat change in QT interval and has been observed under various pathophysiological conditions including ischemia, ventricular hypertrophy, electrolyte disturbances, and the long QT syndrome. ⁷⁰ , ⁷¹

The mechanisms underlying the genesis of TWA are not fully understood. It is certain, however, that TWA is the ECG manifestation of beat‐to‐beat changes in ventricular repolarization. ⁷⁰ There is evidence that TWA under conditions of a long QT interval during relatively fast pacing rates is largely the consequence of alteration of APD in the subendocardium during alternate beats. This is thought to be due to intracellular calcium cycling. ⁷¹ , ⁷² In the setting of ventricular hypertrophy and failure (for example, the rabbit renovascular hypertension model), endocardium but not epicardium exhibits a significant beat‐to‐beat change in APD at slow pacing rates. ⁴⁹ Phase 2 EAD may be generated from endocardium during a beat with longer APD, leading to polymorphic VT under conditions of enhanced TDR. ⁴⁹

“R‐on‐T” Phenomenon

It is well‐known that an R‐on‐T extrasystole is associated with polymorphic VT and VF particularly under conditions of myocardial ischemia, LQTS and Brugada syndrome. The R‐on‐T phenomenon was first described by Smirk in 1949 as “R waves interrupting T Waves.” ⁷³ Theoretically, any transmurally conducted electrical activity during phase 2 or 3 of the action potential can manifest as an “R‐on‐T” extrasystole on the ECG. The terminal limb of the T wave encompasses the vulnerable window representing the interval of time during which some ventricular myocytes have recovered from refractoriness while others have not. ⁴¹ Therefore, an impulse arriving at the terminal limb of the T wave could potentially capture the vulnerable window and initiate polymorphic VT or VF. Two mechanisms have been demonstrated in the genesis of an R‐on‐T extrasystole: transmurally propagated phase 2 reentry and phase 2 EAD. Phase 2 reentry, secondary to the heterogeneous loss of the epicardial action potential dome, can produce a closely coupled R‐on‐T extrasystole with a normal QT interval, a mechanism probably responsible for the initiation of polymorphic VT and VF in the Brugada syndrome and acute transmural myocardial ischemia. ¹¹ , ¹⁷ , ¹⁸ On the other hand, phase 2 EAD is associated with QT prolongation. Phase 2 EAD, which is generated from endocardium or subendocardium in the presence of APD prolonging agents, can propagate transmurally under conditions of enhanced TDR and produce an R‐on‐T extrasystole capable of initiating TdP (Fig. 6). ⁴⁸

Figure 6.

(A) An R‐on‐T ectopic beat (arrow) initiated an episode of TdP in a patient with a markedly prolonged QT interval who had taken cisapride (reprinted from ^Ref.78 with permission). (B) Cisapride preferentially prolonged endocardial APD, resulting in the development of phase 2 EAD in an isolated rabbit left ventricular wedge preparation. Transmural propagation of EAD produced R‐on‐T ectopic beats (arrows) that were able to initiate TdP.

U WAVE

A physiologic U wave in the presence of normal serum potassium concentration is a small wave following the T wave. Its height is generally less than one fourth of the T wave height. The T–U junction is situated at or close to the isoelectric line, but it may be slightly deviated. ⁷⁴ This is significantly different from a pathologic “U” wave seen under conditions such as hypokalemia. It is usually similar to or larger than the so‐called “T” wave with no isoelectric line between the two waves.

Although a physiologic U wave may be observed in normal adults, it is particularly prevalent in patients with the congenital long QT syndrome. The cellular basis for the electrocardiographic U wave has been a matter of debate for many years. Several hypotheses have been proposed to explain this last repolarization component of the ventricles. The likely cellular basis responsible for the physiologic U wave may be Purkinje fiber network or M cells with properties of delayed repolarization. ⁷⁵ , ⁷⁶ As mentioned previously, recent data have shown that M cells, electrically coupled with endocardium and epicardium, play a determining role in the genesis of the T wave. ⁴¹ The hypothesis that the Purkinje fiber network is responsible for the physiologic U wave seems most plausible, but direct evidence is still lacking. ⁷⁵

From a mechanistic viewpoint, the pathologic U wave may be different from the physiologic U wave. In contrast to hyperkalemia, the rate of repolarization of phase 3 of the action potential during hypokalemia is slowed, giving rise to small opposing voltage gradients that crossover producing a low amplitude bifid T wave. ⁴¹ Under these conditions, full repolarization of the epicardium marks the peak of the second component of T wave, whereas repolarization of both endocardium and M cells contributes importantly to the descending limb of the second component of T wave. The apparent “T–U” complex seen during hypokalemia is in fact a T wave whose ascending limb is interrupted. ¹⁸ , ⁴¹ In clinical practice, however, to distinguish between pathologic and physiologic U waves may be difficult.

Advances in modern laboratory technology have permitted in depth research into many mysteries related to the electrophysiological behaviors of the heart. Some ECG waveforms on the electrocardiogram that were described many years ago, and others discovered within the last one or two decades, now have plausible ionic and cellular explanations. Despite this, more interrogation must be pursued in order to bridge the gap between basic electrophysiological research and clinical ECG phenomena.