Sudden Cardiac Death in Athletes or the Young and Healthy

The conduction system has been carefully studied in several young persons who were living a “normal” life and died suddenly. Routine autopsies are often unremarkable. In the majority, the heart is hypertrophied and enlarged to a mild, moderate, or marked extent; however, at the gross level, no significant abnormalities were seen.³ A serial section examination of the conduction system and the surrounding myocardium reveals varying types of abnormalities, either of a congenital or acquired nature. Congenital abnormalities can exist in the sinoatrial (SA) node, atrioventricular (AV) node, or the AV bundle and the bundle branches (Figure 46-3). Abnormal formation of the SA or AV nodes, such as a double SA node, a double AV node, or the abnormal location of the SA and AV nodes in unexpected areas, can be seen pathologically. The AV bundle may be considerably fragmented into several components, abnormally located, or both. In addition, acquired pathologic findings exist in the form of focal myocardial disarray, fat, or fibrosis to varying degrees, disrupting or replacing parts of the specialized conduction fibers and the surrounding myocardium (see Figure 46-3) with arteriolosclerosis (small vessel disease) of the ventricular septum. In general, the findings in the conduction system are accompanied by pathologic findings in the surrounding myocardium. Mononuclear cell infiltration in the approaches to the SA node and in the SA node itself is present.³

FIGURE 46-3 Conduction system and adjoining myocardial disease in a 25-year-old man who collapsed while playing basketball. He was found to be in ventricular fibrillation and could not be resuscitated. Photomicrograph of the left bundle branch shows falling out of cells. Note the endocardial thickening and numerous focal areas of fibrosis of the ventricular septum (Weigert-van Gieson stain ×30). Arrows point to the loss of left bundle branch fibers. E, Endocardium; V, ventricular septum; Fi, fibrosis.

(From Bharati S, Lev M: The cardiac conduction system in unexplained sudden death, Mount Kisco, NY, 1990, Futura.)

Despite such pathologic findings in and around the conduction system, these persons were considered to be asymptomatic, and SCD was the first manifestation of the disorder. Clinically, in almost all, VF was the only common denominator observed by the paramedics at the time of resuscitation. It can be hypothesized that varying types of congenital or acquired pathologic findings in and around the conduction system may remain “silent,” and these individuals are asymptomatic for long periods. Other disorders (e.g., right ventricular cardiomyopathy, hypertrophic cardiomyopathy, idiopathic dilated cardiomyopathy) can lead to VT degenerating to VF, the first evident manifestation of the disease. During an altered physiological or metabolic state, anatomic findings, pathologic findings, or both may trigger an arrhythmia that progresses to VT, VF, and SCD.³

Familial Sudden Cardiac Death

A genetic tendency for the development of an abnormal conduction system, the surrounding myocardium, or both may also lead to VT, VF, and SCD (Figure 46-4).^5,6 Familial occurrence of ventricular arrhythmias may be related to the many genetic mutations associated with congenital long QT syndrome (LQTS) or Brugada syndrome. These disorders of ion channel function may also lead to cellular dysfunction and pathologic changes such as fatty infiltration, fibrosis, and disruption of the conduction system and the adjoining myocardium.^7,8 These disorders are discussed in other chapters in this text and are alluded to in subsequent sections of this chapter.

FIGURE 46-4 Cardiac conduction system histology in a 16-year-old girl with a familial history of alternating bi-directional tachycardia who died suddenly. Fatal ventricular arrhythmias occurred in three generations, including two siblings and the mother. The conduction system revealed degenerative changes. Photomicrograph demonstrates vacuolar degeneration of left bundle branch (hematoxylin-eosin stain ×104). Arrows point to vacuolar degeneration of main left bundle branch.

(From Gault JH, Cantwell J, Lev M, Braunwald E: Fatal familial cardiac arrhythmias, Am J Cardiol 29:548–553, 1972.)

Of note, in many cases of SCD in young patients, no gross anatomic or histologic abnormalities are identified.

Ventricular Fibrillation and Sudden Cardiac Death

The epidemiology of VF intertwines with the available data on SCD (or cardiac arrest) and its documentation by emergency medical system (EMS) personnel.⁹ In 1998, an epidemiologic prospective study from northeast Italy reported an incidence of cardiac arrest of 95 of 100,000 persons per year.¹⁰ In this study, VF accounted for 30.2% of the initially recorded rhythms, asystole for 48.3%, and pulseless electrical activity for 21.5%. A similar study of out-of-hospital cardiac arrest confirmed VT or VF as the initial rhythm in 59 (30%) of 197 patients of a patient cohort from the Los Angeles area.¹¹ Although VT and VF were consistently more likely to be associated with return of spontaneous circulation in these two studies, the time dependency of the initial recorded rhythm is still debated. The Ontario Prehospital Advanced Life Support (OPALS) study quoted an annual incidence of 58 out-of-hospital cardiac arrests per 100,000 persons, and in the subgroup of EMS-witnessed cardiac arrests, VT and VF accounted for 34.2% of cases.¹² However, a recent analysis from Sweden estimated that patients with an electrocardiogram (ECG) taken within the first 10 minutes from a witnessed cardiac arrest have an incidence of VF of 50% to 60%. Linear regression further estimated that ECGs taken within the first 4 minutes should have an incidence of 75% to 80%.¹³

Ventricular Fibrillation and Population Considerations

Despite the debate on the true proportion and incidence of VF in prehospital cardiac arrest, event rates range from 250,000 to more than 450,000 per year in the United States.¹⁴ This absolute number is heavily influenced by population dynamics and the defined subpopulation being discussed.

The median incidence of cardiac arrest in 10 large Canadian and U.S. communities is 52 cases per 10,000 annually.¹⁵ The well-documented high-risk subgroups (ischemic heart disease with low left ventricular ejection fraction [LVEF], complex ventricular ectopy, prior hospital admission for congestive heart failure [CHF], or previous cardiac arrest) contribute a minority of the total number of cardiac arrests, although the subgroup percent incidence per year is the highest (Figure 46-5).

FIGURE 46-5 Risk of sudden cardiac death, with influence of pre-existing heart disease on magnitude of risk in middle-aged and older adults. The prevalent etiologies are a function of age. EF, Ejection fraction; VT/VF, ventricular tachycardia/ventricular fibrillation; MI, myocardial infarction.

(Modified from Myerburg RJ, Kessler K, Castellanos A: Sudden cardiac death: Structure, function and time-dependence of risk, Circulation 85[suppl I]:2–10, 1992.)

A screening tool or preventative intervention would need to be applied to 999 of 1000 persons to influence the 1 of 1000 previously unidentified persons destined for a cardiac arrest. The Framingham study demonstrated a large disparity from the highest to lowest decile (14-fold increase in risk) for SCD measuring the well-known risk factors such as age, family history, gender, tobacco use, hyperlipidemia, and hypertension.¹⁶ Electron beam computed tomography for the detection of coronary artery calcification has been correlated to coronary event risk.¹⁷ A recent publication compared the Framingham risk index and coronary artery calcification in an autopsy series of SCD victims.¹⁸ The two risk assessment methods had modest correlation with each other (63%). Of the cases, 83.5% had a coronary artery calcification score or Framingham risk index above average for age. The remaining 17.5% highlight the need for exploration of new cardiac risk factors (e.g., fibrinogen, homocysteine, infectious agents, or ion channel abnormalities) that could be applied to the population at large.

The high-risk subpopulations have been targeted in recent publications for epidemiologic analysis. The Worcester Heart Attack Study described a relatively fixed incidence rate of 4.7% for VF in hospitalized patients with validated acute myocardial infarction (MI) over a 22-year period (1975 to 1997).¹⁹ The Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico (GISSI-2) database reported an incidence of early-onset VF (<4 hours after onset of acute MI) and late VF (>4 to 48 hours) of 3.1% and 0.6%, respectively. In-hospital prognosis was worse in patients with VF than without VF, but the postdischarge to 6-month mortality rate was similar for both VF subgroups and controls.²⁰ The patient with prior cardiac arrest has a similar “time dependence” of risk of recurrent events, as was presented by Furukawa et al (Figure 46-6).²¹

FIGURE 46-6 Risk of recurrent cardiac arrest among survivors of out-of-hospital cardiac arrest. The highest risk of recurrent cardiac arrest is within the first 12 to 24 months. In the earlier part of that period, ejection fraction (EF) is the most powerful predictor, whereas persistent inducibility of ventricular tachyarrhythmias during electrophysiological studies becomes an important predictor beyond that period. Nonetheless, EF remains a predictor throughout. F/U, Follow-up.

(Modified from Furukawa T, Rozanski JJ, Nogami A, et al: Time-dependent risk of and predictors for cardiac arrest recurrence in survivors of out-of-hospital cardiac arrest with chronic coronary artery disease, Circulation 80:599–608, 1989.)

Low LVEF was a strong predictor of the highest risk of recurrence (11% in the first 6 months). Persistent inducibility at electrophysiological study (EPS) was the better predictor for recurrent events from 6 to 42 months, but the subsequent risk fell to 0.8% over the final 6 months of the study period. In the patients who entered the Antiarrhythmics Versus Implantable Defibrillators (AVID) study with VF, the survival curve shows the same early recurrence (<6 months) in the patients treated with antiarrhythmic drugs as those who did not receive this therapy (Figure 46-7).²²

FIGURE 46-7 Survival free of arrhythmic cardiac death in patients with ventricular fibrillation who qualified for the Antiarrhythmics Versus Implantable Defibrillators (AVID) study. Nonarrhythmic cardiac and noncardiac deaths are censored.

(From the AVID Investigators: Causes of death in the AVID study, J Am Coll Cardiol 34:1552–1559, 1999.)

Clinical trials are assessing the “static” risk of cardiac events outside the setting of acute hospitalization.²³ The Multicenter Automatic Defibrillator Implantation Trial (MADIT-II) trial tested implantable cardioverter-defibrillator (ICD) therapy versus standard medical therapy in patients with ischemic heart disease, an LVEF less than 30%, and previous MI. It showed a 31% relative risk reduction with ICD therapy. The Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) trial compared three arms of therapy in patients with congestive heart failure and LVEF less than 30%: best conventional heart failure therapy with placebo, best conventional heart failure therapy with amiodarone, and best conventional heart failure therapy with ICD.²³ In a preliminary report, there was a 23% relative reduction in all-cause mortality with the ICD compared with placebo (7.2% to 5.8% per year), and no reduction in mortality rate with amiodarone compared with placebo. The most recent developments in assessing risk of cardiac arrest have been in the blossoming technology of molecular genetics. The familial types of LQTS and their respective genetic bases have been well described elsewhere. An increasingly recognized syndrome of right bundle branch block (RBBB), ST segment elevation, and aborted SCD (Brugada syndrome) has also been ascribed to a genetic mutation in the gene encoding the cardiac sodium channel. Its prevalence in a European study by ECG screening was estimated at 0.1% of the general population.²⁴ If the only effective therapy for the syndrome is ICD implantation, the economic impact on the public health budget is obvious.

A trial of the automated external defibrillator (AED) to be used in the home by trained spouses or cohabitants of individuals with prior anterior MI (the Home Automated External Defibrillator Trial [HAT]) was not able to show a reduction in mortality rate, in part because of the unexpectedly low incidence of VF as the initial recorded arrhythmia in these patients.²⁵

The epidemiology of VF and SCD are closely aligned. From a public health perspective, the largest number of cardiac arrest survivors will have the well-described cardiac risk factors from the Framingham study (e.g., hypertension, hyperlipidemia). An intervention for the population as a whole, however, will require treatment of a vast majority to prevent a single cardiac arrest. The high-risk subset of pre–cardiac arrest (although a minority of the total number) is currently undergoing research investigating prophylactic ICD implantation and the cost effectiveness of such a practice. New screening tools that are widely applicable, economical, and cost-effective are needed for future progress in SCD prevention.

Dynamics of Re-entrant Ventricular Fibrillation

VF is generated by self-sustained re-entrant circuits that turn around regions of conduction block. In some circumstances, the zone of conduction block can be anatomic in nature, such as the site of a healed infarct or a region of fibrosis. In such cases, the re-entrant circuit is anchored to the site of block and results in a stable rotor and monomorphic VT. When the zone of conduction block is functional in nature, the region of refractoriness is transient both spatially and temporally. Re-entrant circuits that form around such zones can therefore meander freely or change in size, providing an important mechanism for VF. However, the nature of re-entrant waves and the mechanisms by which they lead to VF are still not entirely clear.

Nature of Fibrillatory Wavefronts

Re-entrant excitation is the fundamental mechanism responsible for VF. The precise nature of these wavefronts, their formation, and sustainability are key to understanding VF. Various theories exist regarding the nature of re-entrant waves. One such theory is the leading circle hypothesis, originally put forward by Allessie and coworkers to explain atrial fibrillation (AF).²⁶ This hypothesis states that a re-entrant circuit is the smallest pathway in which the impulse can continue to circulate and that the core is kept permanently refractory (i.e., there is no excitable gap) (Figure 46-8). However, Janse questioned whether the core is truly refractory during re-entry, since membrane potential is relatively normal in this region and diastolic intervals are relatively long.²⁷ Furthermore, the presence of an excitable gap allows wavefronts (spontaneous or induced) to invade the area and terminate or entrain the re-entrant circuit; Allessie et al were able to terminate tachycardia with premature beats (i.e., anti-tachycardia pacing).²⁶ It now appears that in most cases of VF, an excitable gap does exist; however, there may still be room for leading circle re-entry in a subset of arrhythmic phenomena, particularly with regard to AF.²⁸

FIGURE 46-8 Leading circle re-entry. Activation map (right) and action potential recordings (left) obtained during steady-state tachycardia in an isolated rabbit atrial preparation. Cells in the central area of the re-entrant circuit show double potentials of low amplitude (tracings 3 and 4). A schematic activation pattern is shown on the lower right. Double bars indicate conduction block; the black section is absolutely refractory; between the head and tail of the re-entrant wavefront, relatively refractory tissue is present. Characteristics for leading circle re-entry are (1) the core is kept in a permanent state of refractoriness by centripetal wavelets, and (2) the head of the leading circle bites into its own relative refractory tail.

(From Allessie MA, Bonke FI, Schopman FJ: Circus movement in rabbit atrial muscle as a mechanism of tachycardia. III. The “leading circle” concept: A new model of circus movement in cardiac tissue without the involvement of an anatomical obstacle, Circ Res 41:9–18, 1977.)

An alternative mechanism for VF is spiral wave re-entry. This theory suggests that the wavefront curves, or forms a spiral, because curvature is negatively related to conduction velocity (because of the source-sink relationship). As a result, the wave is highly curved near the core and moves slowly, but at the distal end the wave speed increases, resulting in a spiral shape. Two-dimensional spiral waves may theoretically exist in surviving two-dimensional layers overlying the healed myocardial infarcts, although three-dimensional waves (i.e., scroll waves) are clearly more relevant to cardiac arrhythmias. A scroll wave can be thought of as a stack of spiral waves in which the cores line up to form a “filament” at the center. Simple scroll waves, in which the filament forms a straight line, have been demonstrated during VF.²⁹ A scroll wave with a core that is perpendicular to the recording surface will appear as a two-dimensional spiral wave, whereas a filament that is parallel to the recording surface will appear as a plane wave. It has been suggested that during VF, scroll waves are usually oriented parallel to the surface, which may explain why mapping studies sometimes fail to detect re-entry during VT or VF. A scroll wave with a ring-shaped filament is called a scroll ring. It is interesting to note that the cross-section of such a wave would represent classic figure-of-eight re-entry. Only indirect experimental evidence is available for the existence of scroll rings, probably because of the difficulty of measuring them, because they manifest only transmurally, never on the surface.³⁰ The shape of a scroll as it is initiated strictly depends on the shape of the inexcitable obstacle encountered by the wavefront. Thus, filaments of varying shape can readily develop in the myocardium. Scroll rings ultimately collapse on themselves, but simple scroll waves can be stabilized and maintained.³¹

Numerous hypotheses have been offered regarding the nature of fibrillatory waves that underlie VF. Three of these theories are illustrated in Figure 46-9: (1) the multiple wavelet hypothesis, (2) the meandering spiral wave hypothesis, and (3) the mother-daughter wavelet hypothesis. The multiple wavelet hypothesis was put forward in the early 1960s by Moe and describes fibrillation as consisting of multiple, nonstationary wavefronts that continuously form, fractionate, and reform (Figure 46-9, A).³² Spatial dispersion of repolarization was an important condition for initiating multiple wavelet VF. Others have suggested that VF is caused by spiral re-entrant waves that meander around the myocardium (see Figure 46-9, B).³³ According to this theory, the chaotic appearance of the ECG in VF is attributable to the meandering of a single spiral wave rotor throughout the ventricular myocardium. Jalife suggested that the mechanism of fibrillation may consist of a stable, periodic re-entrant wave that gives off “daughter wavelets” that meander and fractionate (see Figure 46-9, C).³⁴ According to this hypothesis, unstable daughter wavelets form because of local gradients of refractoriness and give rise to the fibrillatory patterns of the ECG.

FIGURE 46-9 Various theories of fibrillatory wavefront behavior. A, Multiple wavelet hypothesis. B, Meandering spiral wave hypothesis. C, Mother-daughter wavelet hypothesis.

Substrates for Ventricular Fibrillation: Heterogeneities of Repolarization and Refractoriness

Various anatomic and functional substrates can lead to the development of the fibrillatory dynamics previously described. Early studies of AF suggested that the heterogeneity of refractoriness was a necessary substrate for fibrillation in the heart.³² Although it was later shown that VF can occur in hearts without large dispersions of refractoriness (i.e., in normal hearts), or in modeling studies using simulations of electrically homogenous tissue, these studies may not be clinically relevant.³⁵ It is more difficult to initiate VF in a normal heart, and the vast majority of fibrillatory episodes occur in diseased hearts, in which the gradients of repolarization kinetics (and therefore refractoriness) are almost certainly abnormal. The important role of repolarization in the development of re-entrant substrates has recently been emphasized by studies demonstrating the inherent heterogeneities in repolarization properties that exist transmurally.³⁶ Antzelevitch and others have demonstrated that trans-mural heterogeneities of repolarization are critical in the development of the substrates for re-entry under various drug-induced or disease-induced conditions.^37,38

Rosenbaum et al recently developed a hypothesis that explains how the critical heterogeneities of repolarization that provide a substrate for re-entry may occur.³⁹ Beat-by-beat alternation of cardiac action potential duration (APD) has been shown to underlie T-wave alternans, an electrocardiographic indicator that correlates well with the risk of SCD.³⁹ We have demonstrated that when alternans occurs in various regions of the myocardium in a “discordant” manner (i.e., some regions are on a long-short-long cycle, and others are on a short-long-short cycle), steep gradients of repolarization develop and reverse the direction on a beat-by-beat basis, which can lead to conduction block, re-entry, and VF.⁴⁰

In related work, Laurita et al demonstrated that premature beats modulate the dispersion of repolarization in a manner that has direct effects on vulnerability to VF.⁴¹ As the S₁-S₂ interval is shortened, gradients of repolarization decrease concomitant to a decrease in vulnerability to VF, but then increase with a concomitant increase in VF vulnerability as S₁-S₂ is further shortened (Figure 46-10). Such biphasic modulation of dispersion relates back to the concept of the “vulnerable period,” which arose from critical studies by Moe, who described the window of time during the ECG T wave when vulnerability to VF is greatest.³² Thus, dispersion of repolarization and refractoriness appears to play a major role in the mechanism of re-entrant VF.

FIGURE 46-10 Modulated dispersion hypothesis and vulnerability to ventricular fibrillation. A, Dependence of mean repolarization (S₂-RT, blue circles) and dispersion of repolarization (S₂-DISP, red circles) on S₁-S₂ coupling interval. B, Dependence of arrhythmia vulnerability (S₂-VFT) on premature coupling interval. Dispersion of repolarization (A, red circles) and vulnerability to fibrillation (B) were modulated in a similar biphasic fashion, with minimum vulnerability (i.e., maximum S₂-VFT) and minimum dispersion occurring at the same S₁-S₂ coupling interval (255 ms, dashed arrow). This inherent mechanism may provide protection against premature stimuli delivered at moderate coupling intervals.

(From Laurita KR, Girouard SD, Akar FG, et al: Modulated dispersion explains changes in arrhythmia vulnerability during premature stimulation of the heart, Circulation 98:2774–2780, 1998.)

Restitution Hypothesis

Restitution is a property of cardiac myocytes that dictates the APDs of a premature action potential after a period of steady-state rhythm. The APD of the extrasystolic beat is determined not only by the S₁-S₂ interval but also by the APD of the pre-extrasystolic beat. Furthermore, gradients of restitution kinetics are known to exist across the epicardial surface such that cells in different regions of the heart having equal baseline APDs could have different extrasystolic APDs following the same S₁-S₂ interval.⁴² In theory, if gradients of restitution kinetics are steep enough, underlying heterogeneities of repolarization would not be necessary to create dispersions of repolarization following a premature beat.⁴² However, cardiac tissue is known to exhibit repolarization heterogeneities across epicardial and trans-mural surfaces, so the role of restitution may be to enhance existing dispersion of repolarization and promote VF.

One possible mechanism to describe the role of restitution in VF is the restitution hypothesis, which states that when the restitution curve (a plot of APD vs. diastolic interval) has a slope greater than 1, VF may be initiated. For example, a wavefront encroaching on the tail end of a previous wave creates a gradient of diastolic interval between the two (assuming the degree of curvature is not identical); in regions of tissue where restitution is steep, this will create large dispersions of APD along the second wavefront. Some areas of the wave may have such a short diastolic interval that they fail to propagate, resulting in wavefront fractionation. This was demonstrated by Garfinkel et al, who also used pharmacologic agents to flatten the restitution curve slope and decrease vulnerability to VF. Koller et al showed that elevating extracellular potassium decreases the portion of the restitution curve where the slope is 1 or more and converts VF to a periodic rhythm.^43,44 Other investigators have shown that the biphasic nature of a restitution curve, demonstrated in a number of cardiac tissue types, is also an important determinant of vulnerability to VF.⁴⁵ More recently, the restitution hypothesis has been applied to restitution curves for both APD and conduction velocity.⁴⁶

Myocardial Ischemia

VF is commonly caused by acute myocardial ischemia. Under experimental conditions, it is often difficult to provoke VF in normal tissue using premature stimuli; however, in ischemic tissue, premature beats exacerbate dispersion of refractoriness and readily lead to tachyarrhythmias.⁴⁷ This is because of the various electrical alterations that occur during ischemia and the consequent formation of various substrates that promote re-entry and VF.

At the cellular level, acute ischemia results in depolarization of the resting membrane potential, decreased maximum rise rate and amplitude of the action potential, and decreased APD, as well as reduced intercellular coupling.^48–53 Depolarization of the membrane is attributed largely to accumulation of extracellular potassium, as is post-repolarization refractoriness.^54,55 Acidosis can produce a small depolarization of resting membrane potential and also decreases gap junction conductance and cell-to-cell coupling.⁵⁶ The net result of the electrical changes that occur during ischemia is the formation of various substrates that can lead to re-entry and wavebreak, such as slowed conduction velocity, increased dispersion of refractoriness, and alternans.

Healed Myocardial Infarcts

The process of tissue healing and scar formation after an episode of acute myocardial ischemia involves necrosis of the infarcted region as well as swelling and hypertrophy of the noninfarcted region as it attempts to compensate for the loss of cardiac muscle.⁵⁷ Heterogeneity of repolarization may result from the uneven prolongation of APD in the hypertrophied post-MI ventricle, changes in expression of gap junction proteins, or both.^58,59 A healed infarct may also provide an anatomic substrate for arrhythmias. Infarcts vary from simple surviving muscle bundles that form accessory pathways to complex subendocardial sheetlike structures, which are linked to the surrounding myocardium by multiple connecting bundles, creating complex matrices of conductive tissue that may promote multiple re-entrant circuits. In particular, the combined effects of slowed conduction and the presence of structural anomalies is significant because in tissue where excitability is low, a wavefront breaking past an obstacle will curl at the inner ends (where the break occurred) to create figure-of-eight re-entry, whereas in a normally excitable medium, the wave tends to reform on the other side of the obstacle.⁶⁰

Autonomic Modulation of Ventricular Fibrillation

It has been well established that increases in sympathetic tone increase the risk for SCD, whereas vagal reflexes have the opposite effect.^61–63 Studies in dogs have shown that left cardiac sympathetic denervation, or left stellectomy, increases survival and decreases arrhythmias during acute MI.⁶⁴ In contrast, unilateral blockade of the right stellate ganglion by cooling increases the number of arrhythmias after coronary occlusion.⁶⁵ Many clinical trials have shown that β-adrenergic receptor blocking drugs have a potential benefit in preventing SCD in patients with MI, whereas anti–α-adrenergic agents have not been shown to be effective.^66–69

It has been suggested that during acute MI, sympathetic nerves that traverse the myocardial wall are damaged, leading to denervation hypersensitivity.⁷⁰ This, in turn, leads to spatially inhomogeneous responses to β-adrenergic stimulation across the ventricle, forming potential substrates for re-entrant excitation. Further investigation is required to elucidate fully the mechanistic role of β-adrenergic stimulation and arrhythmogenesis.

Genetic Substrates for Ventricular Fibrillation

Brugada syndrome is a cardiac disease characterized by an elevated ST segment that is unrelated to ischemia, Q-T interval prolongation, electrolyte abnormalities, or structural heart disease.⁷¹ Patients with Brugada syndrome are at increased risk of SCD from VF. The syndrome is caused by a mutation in the gene encoding the cardiac sodium channel (SCN5A) and is common in Southeastern Asia, where it is the second highest cause of death among young men.⁷² In Europe, gender specificity is less distinct, although affected women tend to be less symptomatic than affected men. Patients with Brugada syndrome often die in their sleep, and a possible relationship to sudden infant death syndrome has been suggested.⁷³

It is believed that the mechanism of VF in patients with Brugada syndrome is related to reduced function of the mutant late inward sodium current, forcing the balance of plateau currents in favor of repolarization. This is why any further impairment of sodium channel function (e.g., by sodium-blocking drugs) can exacerbate the electrocardiographic and arrhythmogenic phenotype in this disorder.⁷⁴ Because epicardial, not endocardial, cells possess the transient outward current (I_to), they are most susceptible to the repolarizing effect of the sodium channel mutation in Brugada syndrome. Marked and selective shortening of epicardial action potentials relative to endocardial action potentials produces a transmural voltage gradient during the plateau that, in turn, is believed to account for the characteristic pattern of ST segment elevation seen in these patients.⁷⁵ Transmural action potential gradients also account for the presumed mechanism of VF (i.e., phase 2 re-entry).

In a recent provocative study, Haisaguerre et al identified a select group of patients with recurrent idiopathic VF that seemed to originate with premature beats emanating from the distal His-Pukinje system; these patients were successfully treated with radiofrequency ablation targeting the originating premature beats, with a low rate of VF recurrence.⁷⁶

In summary, VF is a complex arrhythmia associated with many distinct electrophysiological mechanisms that no doubt highly depend on the disease state in question. Although most often associated with acute ischemia or healed MI, VF can occur in the absence of any structural heart disease, such as heritable disorders (e.g., Brugada syndrome). Greater understanding of these various mechanisms is required to guide the development of novel pharmacologic approaches and targets that can ultimately be used in the treatment and prevention of VF. Until then, clinicians will have to rely on electrical defibrillation, which is highly effective irrespective of the underlying VF mechanism.

Electrocardiographic Features of Patients at High Risk for Ventricular Fibrillation

Electrocardiographic recordings obtained at the onset of VF provide insight into events that precipitate SCD. Documentation of these events has been obtained by continuous monitoring in hospital telemetry units and by ambulatory monitoring. The patients involved in these studies had a high incidence of CAD, and most had frequent or complex ventricular ectopy. The terminal events were associated with sinus arrest, complete heart block, or ventricular asystole in approximately 10% of patients; in 90%, VF was preceded by VT or ventricular flutter of variable duration.^77–80 ST segment and T-wave changes indicative of ischemia related to acute MI or coronary spasm have also been reported to precede terminal ventricular arrhythmias.^77–80 Other studies indicate that bradycardia and electromechanical dissociation are important causes of SCD in patients with advanced heart failure and nonischemic cardiomyopathy.⁸¹ These observations have significant implications for strategies to reduce mortality rates in patients at risk for SCD.

The diagnostic role of the ECG in identifying patients with genetic disorders associated with SCD from ventricular arrhythmias has continued to evolve over the past decade. The electrocardiographic characteristics of LQTS, Brugada syndrome, and arrhythmogenic right ventricular cardiomyopathy (ARVC) have gained widespread attention. This is vital to the prevention of SCD in young patients with these disorders, particularly if these patients participate in athletics. A recent study has suggested a higher incidence of early repolarization changes in the inferolateral leads (J-point elevation and notching in the terminal portion of the QRS complex) among survivors of cardiac arrest caused by idiopathic VF when compared with control subjects.⁸²

The pattern of early repolarization associated with highest risk seems to be global or inferolateral early repolarization with prominent J waves on the resting ECG.⁸³ It is not clear if a single disorder of repolarization causes at least some of the cases previously labeled “idiopathic VF,” or, as seems more likely, a spectrum of disorders is associated with abnormal and heterogeneous repolarization that predisposes otherwise normal individuals to VF. These disorders have been collectively named early repolarization disease.⁸⁴

Although the prevalence of repolarization abnormalities in patients resuscitated from VF with no evident structural heart disease is higher than expected, these abnormalities are also common in young, healthy individuals. The sensitivity, specificity, and predictive accuracy of this finding are not clear. The risk of cardiac arrest in asymptomatic patients with early repolarization is likely very low.

Long QT Syndrome

The electrocardiographic manifestations of LQTS include QT prolongation, abnormalities in T-wave morphology, increases in QT dispersion, T-wave alternans, and a relative degree of bradycardia in children (Figure 46-11). The upper limits of normal for the QTc values are 460 to 470 ms for females and 440 to 460 ms for males.^85,86 Longer QT values may be observed in normal women after puberty.⁸⁷ The degree of QT prolongation does not directly correspond with the risk of syncope, but malignant ventricular arrhythmias are more frequent when the QTc exceeds 600 ms.⁸⁷ A diagnostic dilemma is that the Q-T interval shows temporal variations in patients with this syndrome, and the QTc may fall within the normal range on a random recording.⁸⁸ Garson reported that 6% of LQTS patients had a normal Q-T interval, and data from the International Registry showed that 10% of family members with a QTc less than 440 ms had a cardiac arrest.^89,90 Thus, an ECG with a normal QTc does not exclude the diagnosis if strong suspicion exists that a patient has the syndrome, especially if the QTc is on the border of normal.

FIGURE 46-11 Twelve-lead electrocardiogram of a patient with familial long QT syndrome. Note the prolonged Q-T interval and T-wave abnormalities.

The effect of exercise increases the QTc in patients with LQTS, but this effect is less apparent in patients with LQT3 than in those with other genotypes.⁹¹ Approximately 62% of patients with LQTS exhibit T waves that are biphasic or notched, and a higher incidence of these abnormalities is seen in patients with cardiac events.⁹² The characteristic features are most pronounced in precordial leads V2 to V5. The appearance of notched T waves may be provoked by exercise. The degree of QT dispersion is measured by the difference between the longest and shortest Q-T intervals on the 12-lead ECG and is prolonged in patients with LQTS.⁹³ It is thought to represent increased dispersion of repolarization. Patients who show no change in the degree of QT dispersion when they are treated with β-blockers appear to be at increased risk for cardiac events.^90,94 T-Wave alternans is a beat-to-beat alternation in the amplitude or polarity of the T wave. It appears to be a marker of electrical instability that may precede torsades de pointes.⁹⁴ Children with LQTS often have resting heart rates that are lower than normal and may exhibit a blunted chronotropic response to exercise.^87,95 Torsades de pointes, which is the ventricular arrhythmia associated with LQTS, is characterized by the undulating amplitude of the QRS complex, which gives the appearance of twisting about its axis. The onset is frequently associated with pause-dependent ventricular ectopy that falls on the T wave.^96–98

Brugada Syndrome

The ECG patterns associated with Brugada syndrome are (1) a terminal R′ in lead V1 with complete or incomplete RBBB; (2) convex downward (coved) ST segment elevation equal to 0.1 mV in lead V1 or leads V1 and V2; convex upward (saddle-shaped) ST-segment elevation equal to 0.1 mV; and (3) J-point elevation followed by a downsloping ST segment ending in a negative deflection (triangular shape).⁹⁸ Serial ECGs performed on the same patient may show variation from one pattern of ST-segment elevation to another, normalization, and progressive development of RBBB. Figure 46-12 shows examples of variable Brugada ECG patterns that occurred during an ajmaline test in the same patient.^99,100 The prevalence of the Brugada ECG pattern is reported to be 0.07% to 0.7%, and there is a male predominance that is especially marked in Asians.^101–106 The range in prevalence appears to depend on the criteria used to make the diagnosis. The saddleback ST-segment elevation is more common. The typical coved pattern was found in 0.1% to 0.26% of community-based populations in Japan and Europe.^104–106 In a Japanese study population that underwent ECGs during health examinations, the prevalence of all types of Brugada ECG patterns was 0.7%.¹⁰³ The coved-type ST-segment elevation was found in 38% of subjects with the Brugada pattern, and the saddleback-type ST-segment elevation was seen in 62%. In the same study, the rsR′ pattern in lead V1 was observed in 41%, and the Rsr′ pattern was recorded in 59%. The prevalence of the coved pattern was 0.26%, and the typical Brugada ECG pattern with coved ST-segment elevation and the rsR′ pattern in lead V1 was 0.12%. If only male subjects were considered, the criteria for a Brugada ECG pattern was met in 2.14% of the population, but the typical pattern in males was 0.38%.¹⁰⁴ In a European population studied by Hermida, the prevalence of ST segment elevation was 6.1%; however, only 1 (0.1%) of the 61 subjects who met the study’s criteria for the Brugada pattern had the coved pattern. All of the others had the saddleback pattern.¹⁰³

FIGURE 46-12 Serial changes in a Brugada pattern electrocardiogram recorded in a patient during an ajmaline test. Twelve-lead ECGs are recorded at baseline (A), followed by ajmaline infusion. At 2 minutes, ST-segment elevation that is coving upward is seen in the right precordial leads (B), followed by marked ST-segment elevation in leads V1 to V3 (C).

The prognostic significance of the Brugada ECG pattern is difficult to assess. Brugada reported an 8% incidence of arrhythmic events in an asymptomatic hospital-based population.¹⁰⁶ The degree and type of ST segment elevation require further study for risk stratification of asymptomatic individuals in community-based populations. Matsuo evaluated mortality in patients younger than 50 years who had ECGs recorded during biannual examinations from 1958 through 1999.¹⁰⁷ A total of 32 patients were identified with the Brugada ECG pattern. Seven of these patients died suddenly or died of an unexplained accident. Although total mortality was not increased in patients with the Brugada ECG pattern, the mortality rate from unexpected death was significantly higher. No increase in mortality was observed in studies by Miyasaka, Takenaka, or Priori.^104,105,107 In the Osaka population, one SCD occurred among the 98 subjects with the Brugada ECG pattern during a mean follow-up of 2.6 years.¹⁰⁴ A 3-year follow-up reported by Atarashi of patients with a Brugada ECG pattern found cardiac event–free rates of 67.6% in symptomatic patients and 93.4% in an asymptomatic group.¹⁰⁸ Coved-typed ST segment elevation appeared to be related to cardiac events. The higher incidence in hospital-based studies may reflect referral patterns based on a family history of SCD. Differences in criteria or ECG interpretation affect the diagnosis of this pattern, and the follow-up in most studies is too short to draw definitive conclusions about the long-term prognosis.

Arrhythmogenic Right Ventricular Cardiomyopathy

ECG recordings during sinus rhythm in patients with ARVC have several distinctive features (Figure 46-13).¹⁰⁹ The QRS may be prolonged in the right precordial leads to a greater extent than in leads I or V6. The QRS is often greater than 110 ms in lead V1 (sensitivity, 55%), and a pattern of incomplete RBBB is observed.¹¹⁰ In 30% of cases, the delay in conduction over the right ventricle results in a small potential at the terminal portion of the QRS in lead V1 that has been termed an epsilon (ε) wave, which can be amplified with bipolar recordings over the inferior and superior aspects of the sternum.¹⁰⁹ This can be achieved by repositioning the left arm lead over the xiphoid process, positioning the right arm lead over the manubrium sternum, and applying the left leg lead at the position customarily used for lead V4 or V5.¹¹¹ The other major feature of ECGs recorded from patients with ARVC is inversion of the T waves in the precordial leads, which is observed in 42% to 54% of patients.^111,112 Metzger assessed the value of serial 12-lead ECGs in 20 patients to recognize progression of ARVC over a mean of 71 ± 48 months.¹¹³ Abnormalities were detected in 90% of the patients. The most frequent abnormality was T-wave inversion in the precordial leads. No correlation was demonstrated between the ECG and the extent of disease detected by echocardiography. In the 14 patients who had several ECGs recorded over time, no clear progression of electrocardiographic abnormalities was observed.

FIGURE 46-13 Twelve-lead electrocardiogram recorded from a 19-year-old man with arrhythmogenic right ventricular cardiomyopathy. An ε wave is present in lead V1 at the end of the QRS.

The ventricular arrhythmias associated with ARVC typically show a morphology resembling left bundle branch block in lead V1 with a variable axis. Nava published a clinical profile and long-term follow-up of 37 families with ARVC that demonstrated a correlation between the severity of echocardiographic findings and the severity of ventricular arrhythmias, which were seen in all patients with the severe form of the disease, in 82% with moderate disease, and in 23% in those with mild disease.¹¹² Overall, 60 (45%) of 132 affected living members had ventricular arrhythmias. These included VF in 1, sustained VT in 14, nonsustained VT in 8, ventricular couplets and triplets in 16, and frequent ventricular premature depolarizations in 8. Exercise-induced polymorphic VT was observed in 13 patients with no other documented arrhythmias. Only one patient who was judged to have mild disease had a sustained ventricular arrhythmia. Although the data from this study showed a low incidence of VF, the incidence may be higher. In 19 of the 37 families, the proband died at a young age, and the diagnosis was made at autopsy. One may speculate that some of these subjects had VF. Figure 46-14 shows electrograms recorded from an ICD implanted in a teenager (male) with ARVC and frequent nonsustained ventricular arrhythmias. The recording shows the sudden onset and successful termination of VF, which occurred at night while he was asleep. He had no prior history of syncope or sustained ventricular arrhythmias.

FIGURE 46-14 Electrograms recorded from an implantable cardioverter-defibrillator (ICD) in a patient with arrhythmogenic right ventricular cardiomyopathy. The continuous recordings document the sudden onset of ventricular fibrillation, followed by an ICD shock and conversion to sinus rhythm. The upper channel represents the stored intracardiac electrogram. The lower channel marks sensed events and displays the cycle length.

Evaluation of Transient or Reversible Causes

VF that occurs secondary to a reversible or transient cause may be adequately treated by correction of the reversible cause or by short-term therapy or close observation while awaiting spontaneous resolution of the transient cause.¹¹⁴ The causes of VF with these characteristics are usually readily identified with a few focused investigations. Electrolyte abnormalities are identified with serum electrolyte testing performed as soon as possible after resuscitation. The most common electrolyte abnormalities leading to VF are hypokalemia, hypomagnesemia, or both. When considering the temptation to ascribe an episode of VF wholly to hypokalemia, hypomagnesemia, or both, one must recall that the adrenergic discharge state during resuscitated VF results in redistribution of extracellular potassium and magnesium into the intracellular compartment. Accordingly, relative hypokalemia, hypomagnesemia, or a combination is very common after VF.¹¹⁵ Only marked hypokalemia, hypomagnesemia, or both and confidence that future episodes can be prevented should prompt the belief that the episode of VF had a reversible cause.

The early performance of a 12-lead ECG and serum markers of myocardial necrosis (creatine kinase [muscle or brain type], troponin) will permit identification of the patient whose VF has occurred in the acute phase (first 48 hours) of an MI. Evidence that acute MI produces an environment that constitutes only a transient risk of VF is most convincing for a Q-wave MI.¹¹⁶ Nevertheless, the risk of VF may also be transient in the setting of a non–Q-wave MI.¹¹⁷ However, the practitioner must avoid the temptation to ascribe an episode of VF wholly to an acute MI if the only evidence of acute myocardial necrosis is a marginal elevation in a serum marker of necrosis; such elevations may be secondary to the VF-induced cardiac arrest rather than represent its cause. The diagnosis of MI in the setting of out-of-hospital cardiac arrest from VF is made difficult by the usual absence of a history of chest pain, frequent nondiagnostic ECG abnormalities that may be a consequence of the myocardial ischemia resulting from the global “no-flow or low-flow state” during VF, and the absence of a specific cut-off for the degree of cardiac enzyme elevation that separates MI causing VF from VF causing myocardial necrosis. In general, the presence of a culprit lesion (coronary thrombus or ulcerated plaque) on angiography early after cardiac arrest and well-preserved or normal left ventricular function suggests an ischemic or infarction-related cause.

The other common reversible or transient cause of VF is the use of a proarrhythmic drug: a classic antiarrhythmic drug, other drugs with electrophysiological effects, and recreational drugs, particularly cocaine.¹¹⁸ Use of these agents is determined from the history. However, if the practitioner believes that a report of such drug use will not be forthcoming from a high-risk individual, a toxicology screen is advised. When therapeutic drug use is reported, early determination of a serum concentration of the agent may be important when toxicity related to that agent has a relationship to serum concentrations (i.e., digitalis).

Finally, an episode of VF may be considered reversible if it accompanies a state of extreme physiological derangement that is not expected to be recurrent. Such states include that seen in the immediate postoperative period, with sepsis, and with hemodynamic instability, especially when the treatment of the physiological derangement required administration of sympathomimetic agents.

Identification of Structural Heart Disease

In most of the world, the most common forms of structural heart disease that precipitate VF are atherosclerotic CAD and either congestive or hypertrophic cardiomyopathy. Accordingly, most patients who have been resuscitated from VF require an echocardiographic examination, an exercise test (with or without myocardial perfusion imaging), and cardiac catheterization with coronary angiography.

The echocardiogram primarily serves as an adjunct to the physical examination for identification of myocardial or valvular structural heart disease. It is particularly suited to the documentation of hypertrophic cardiomyopathy and for the detection of other forms of structural heart disease that have escaped clinical detection. Although the echocardiogram is also useful to quantitate left ventricular systolic function, radionuclide ventriculography, cardiac computed tomography, cardiac magnetic resonance imaging, and contrast ventriculography may provide more accurate determinations of this important prognostic variable. Of importance, the echocardiogram is not very sensitive for the detection of early arrhythmogenic right ventricular dysplasia.

Exercise testing of the patient who has been resuscitated from an episode of VF provides information about the inducibility of both exercise-related myocardial ischemia and exercise-related arrhythmias.¹¹⁹ The sensitivity, specificity, and spatial localization of reversible myocardial ischemia can be enhanced by coupling the exercise test with radionuclide or echocardiographic imaging techniques.

Despite the availability of these noninvasive diagnostic procedures, the critical importance of accurate and complete anatomic and functional diagnosis of structural heart disease in a patient who has been resuscitated from VF usually necessitates cardiac catheterization and coronary angiography. Occasionally, the combination of information available from the history of events surrounding the spontaneous episode of VF (especially when preceded by physical or emotional stress accompanied by angina), evidence of reversible myocardial ischemia on exercise testing (especially when associated with evidence of ventricular electrical instability), and documentation of CAD (especially when not associated with severe left ventricular dysfunction) will suggest the possibility that the episode of VF was caused by reversible myocardial ischemia that can be corrected. Although coronary revascularization in this setting has been reported to prevent further episodes of VF in some patients, the effect is not necessarily predictable.¹²⁰ Accordingly, it has become customary to either assume that the revascularization procedure was insufficient for the prevention of VF recurrences or document that VT or VF is not inducible by programmed stimulation performed during a transvenous catheter EPS study after the revascularization procedure—preferably after a preoperative catheter EPS documented the inducibility of sustained VT or VF.

On occasion, specialized procedures are required to identify underlying structural heart disease in selected patients for whom a high index of suspicion exists. These procedures include cardiac magnetic resonance imaging, catheter endomyocardial biopsy, or both for the diagnosis of arrhythmogenic right ventricular dysplasia (ARVD), cardiac magnetic resonance imaging, catheter endomyocardial biopsy, or both for the diagnosis of myocarditis; signal averaged electrocardiography for the diagnosis of ARVD; infusion of a class I antiarrhythmic agent for the diagnosis of Brugada syndrome; and infusion of epinephrine for the diagnosis of LQTS. To date, genetic testing has not yet reached the maturity at which it can be recommended for diagnostic purposes in patients with VF. However, the future promise of genetic testing in this regard is high.

Documentation of the Mechanism of Ventricular Fibrillation

The debate on the advisability of offering a baseline transvenous catheter EPS to all patients who have had an episode of VF in the absence of a reversible or transient cause is still ongoing.¹²¹ The major potential advantage of an EPS for patients with VF is the possibility of demonstrating that the VF was caused by another arrhythmia that would be treated in another way. Of course, VF may result from the degeneration of other tachyarrhythmias best treated by trans-catheter ablation procedures such as supraventricular tachyarrhythmias (including AF in the setting of ventricular pre-excitation) and certain VTs (including bundle branch re-entrant VT). EPS may also help distinguish patients with a permanent VT substrate (typically those with a myocardial scar who have inducible VT) from those without a permanent VT substrate (typically those without a myocardial scar who do not have inducible monomorphic VT). This distinction may assist the practitioner in the selection of treatment modalities. For example, revascularization of CAD is rarely, if ever, useful in the former circumstance. Finally, EPS may also provide information regarding optimal programming of a subsequently implanted ICD. Nevertheless, each of these potential advantages is more likely to be had by the patient who presents with VT.

Identification of Other Disease States

Identification of other disease states that interact with antiarrhythmic therapies to be prescribed for the treatment of VF is an important goal of the investigations performed in this patient population. Screening biochemical investigations for renal or hepatic dysfunction are indicated before the prescription of antiarrhythmic drugs that depend on the renal or hepatic metabolism. Similarly, prescription of those therapies with adverse effect profiles that include interaction with other organ systems should be preceded by screening of the integrity of that organ system (e.g., thyroid function testing and pulmonary function testing in preparation for amiodarone therapy). These screening examinations are then be repeated as necessary during follow-up and can be compared with the baseline evaluations to substantiate change.

Assessment and Monitoring of Selected Antiarrhythmic Therapy

The assessment and monitoring of some forms of antiarrhythmic therapy require other selected investigations. Although now infrequently used, the selection of antiarrhythmic drug therapy by suppressing ventricular premature beats requires a baseline EPS as well as antiarrhythmic drug–free, and drug assessment 24-hour ambulatory ECG examinations along with exercise tolerance tests.^119,122 Similarly, the selection of antiarrhythmic drug therapy using the approach of suppression of ventricular tachyarrhythmias induced by programmed stimulation requires a baseline EPS as well as antiarrhythmic drug–free and drug assessment EPSs.^123,124 Of course, the follow-up of patients with a treated propensity to VF usually requires long-term surveillance—most commonly with repeated 24-hour ambulatory ECG examinations. However, no direct evidence suggests that such surveillance is of value to the patient with VF.

Induction of Ventricular Tachycardia Versus Ventricular Fibrillation

After defining the underlying heart disease, EPS is important for risk stratification for primary VF. It is believed that VF is often preceded by VT, which degenerates into VF, so the induction of VT in the electrophysiological laboratory is indicative of the clinical rhythm (Figure 46-15). Therefore, the principal goal at EPS is the induction of sustained monomorphic VT. In the setting of a healed MI, VT can be induced at EPS in 20% to 45% of patients. The anatomic and electrophysiological substrate of VT is well described. Areas of healed MI, regions of slow conduction, and inhomogeneities of refractoriness all contribute to the specific responses to programmed stimulation and the induction of VT. As the induction of VT at EPS is already well described (see Chapters 25, 26, and 47), the focus of this section is induction of VF.

FIGURE 46-15 A, Induction of rapid monomorphic ventricular tachycardia with triple ventricular extrastimuli in a patient with cardiac arrest. B, A subsequent induction attempt with triple extrastimuli resulted in induction of ventricular fibrillation.

The correlation between the induction of VF in the electrophysiology laboratory and the incidence of VF as the presenting arrhythmia is limited, perhaps because of transient factors or prior VT in either situation.¹²⁷ Among patients with previous cardiac arrest, 20% to 50% do not have VF induced at EPS. VF is inducible in up to 25% of cardiac arrest survivors compared with 3% of patients who presented with monomorphic VT. This suggests that induction of VF in patients who present clinically with VF may be specifically predictive of spontaneous VF episodes, whereas induction of VF in patients who presented with VT may represent a nonspecific response to programmed electrical stimulation.¹²⁸

In patients with asymptomatic nonsustained VT, CAD, and an ejection fraction less than 40%, VF or polymorphic VT is induced in up to 6% of cases.¹²⁹ Clinical variables often fail to distinguish patients with inducible arrhythmias from those without inducible arrhythmias. Ejection fraction is not significantly different in patients with no inducible arrhythmia, inducible nonsustained VT, inducible sustained monomorphic VT, or inducible VF (Figure 46-16). Of the 6% of patients in whom VF or polymorphic VT was induced, 17% died suddenly during follow-up.¹²⁶ The presence of inducible sustained ventricular arrhythmias and the persistence of inducible sustained ventricular arrhythmias on therapy were significant univariate predictors of SCD. However, only the persistence of inducible sustained arrhythmias on therapy was an independent predictor of SCD (Figure 46-17).

FIGURE 46-16 Scatterplot of left ventricular ejection fraction stratified by baseline-induced ventricular arrhythmia. Vertical bars are the mean ± SD of each group. Circled dots represent patients in whom inducible arrhythmias were suppressed. P = not significant among groups. NI, Noninducible; NSVT, nonsustained ventricular tachycardia; SMVT, sustained monomorphic ventricular tachycardia; VF, ventricular fibrillation.

(From Wilber DJ, Olshansky B, Moran JF, Scanlon PJ: Electrophysiological testing and nonsustained ventricular tachycardia: Use and limitations in patients with coronary artery disease and impaired ventricular function, 82[2]:350–358, 1990.)

FIGURE 46-17 Actuarial incidence of sudden cardiac death or cardiac arrest in 97 patients, stratified by treatment subgroup. Three patients with cardiac arrest during serial drug testing were not included in analysis. CA, Cardiac arrest; SD, sudden cardiac death.

(From Wilber DJ, Olshansky B, Moran JF, Scanlon PJ: Electrophysiological testing and nonsustained ventricular tachycardia: Use and limitations in patients with coronary artery disease and impaired ventricular function, 82[2]:350–358.)

Electrophysiological Characteristics Associated with Induction of Ventricular Fibrillation

It is not well understood why VF is induced in some patients in the clinical electrophysiology laboratory and not in others. Because VF can be induced in perfectly normal hearts by using either multiple closely coupled premature stimuli or by shocks applied on the T wave, induction of VF by catheter stimulation is not necessarily associated with a poor prognosis. Several potential reasons exist for the nonspecific response of induction of VF at EPS, including local graded response in normal muscle or decremental conduction block caused by short coupling intervals. This appears to have less to do with dispersion of refractoriness or propagation and more to do with local graded responses in normal muscle. For this reason, premature coupling intervals are usually not shortened below 180 ms to avoid inducing VF that has no diagnostic value.¹²⁹ Other causes of nonspecific responses to programmed stimulation and induction of VF include increased conduction latency and prolongation of local activation time in proximity of the stimulus electrode. Induction of VF is preceded by increased latency (Figure 46-18).¹³⁰

FIGURE 46-18 Tracings of electrocardiogram leads I, II, and V1 and ventricular electrograms. Top, Tracing of the last two ventricular stimuli from the drive (S₁) and the triple extrastimuli that result in the induction of sustained monomorphic ventricular tachycardia. Middle, Induction of ventricular fibrillation. Bottom, No induction of ventricular tachycardia or ventricular fibrillation. VF, Ventricular fibrillation, paper speed 100 ms/cm; VT, ventricular tachycardia.

(From Avitall B, McKinnie J, Jazayeri M, et al: Induction of ventricular fibrillation versus monomorphic ventricular tachycardia during programmed stimulation: Role of premature beat conduction delay, Circulation 85:1271–1278, 1992.)

Another hypothesis suggesting why VF is induced with closely coupled premature beats during programmed stimulation is delayed conduction. Conduction slowing by itself causes dispersion of activation time between sites near the stimulating electrode and distant sites. Also, conduction slowing of a premature beat allows distant sites to have a longer coupling interval of the premature beat compared with that from the pacing site. These factors result in dispersion of refractoriness, with the refractory periods following the premature beat being longer at distant sites than at the pacing site. This, in turn, allows additional premature beats to induce VF. Therefore, the measurement technique itself alters what we are attempting to measure (Table 46-1).¹³¹ In addition, indirect evidence suggests that the mechanism of VF in the setting of a healed MI may be related to areas of slow conduction and stable re-entrant circuits. Occasionally, polymorphic VT or VF can transform to monomorphic VT by type I antiarrhythmic medications.¹³² These tachycardias have been mapped in the operating room and have been cured by endocardial resection. These findings have not been observed in patients with a normal heart and inducible polymorphic VT or VF. Therefore, use of a type I antiarrhythmic agent may be helpful in differentiating between a nonspecific finding from a significant one in the electrophysiology laboratory.

Prognosis and Clinical Relevance

The prognostic value of electrophysiological evaluation has been extensively studied in survivors of cardiac arrest. The reproducibility of inducible VF is strongly predictive of recurrent cardiac arrest. However, the failure to induce arrhythmias with programmed stimulation may not necessarily be associated with a benign prognosis. In a specific subset of patients with CAD, LVEF less than 40%, asymptomatic nonsustained VT, and no inducible arrhythmias, the relative risk of death was 1.8.¹²⁵

In addition, several secondary prevention trials have supported the efficacy of defibrillator implantation in out-of-hospital VF survivors irrespective of the outcome of EPS. Results of the Cardiac Arrest Study Hamburg (CASH), AVID, and the Canadian Implant Defibrillator Study (CIDS) support defibrillator placement in cardiac arrest survivors who have no reversible cause of VF.^133,134

In summary, the usefulness of EPS in patients with VF is evolving. If a clearly reversible cause of secondary VF is identified, correction of the underlying problem is needed. With primary VF, it may be reasonable to consider EPS. The specificity of induction of VF by programmed stimulation varies with the patient’s clinical presentation. The induction of VF may be related to areas of slow conduction or healed MI, local graded responses, decremental conduction block caused by short coupling intervals, and increased latency and prolongation of activation time. Defibrillator insertion may be indicated despite a negative EPS result in these patients.

Rescue of the Patient with Ventricular Fibrillation

Optimal resuscitation from VF requires restoring a perfusing cardiac rhythm as early as possible as well as providing the maximum possible cerebral and coronary blood flow during the resuscitation process. Mechanical aids, which improve cardiac output during CPR, have consistently shown improved outcomes in experimental settings and have shown, in limited human clinical trials, improved early resuscitation and improved survival to hospital admission.¹³⁵ However, definitive clinical trials with respect to improving survival to discharge from hospital are lacking although such interventions may be able to improve long-term outcomes in patients with VF, particularly if the interventions are begun relatively early (e.g., at 5 to 10 minutes) after the onset of cardiac arrest from VF.

Mechanical interventions of this type include the active compression-decompression CPR device, interposed abdominal and chest compression devices, the impedance threshold valve, mechanical or pneumatically driven chest compression devices, or devices inserted into the chest cavity to assist in direct cardiac compression.¹³⁶ All these devices increase cardiac output by increasing intrathoracic pressure during the compression phase of CPR or by increasing venous return during the decompression phase, either by positive abdominal pressure or by negative intrathoracic pressure during the decompression phase of CPR. The recent Prehospital Resuscitation Using an Impedance Valve and Early vs. Delayed Analysis Impedance Threshold Device (ROC PRIMED ITD) study was stopped at the recommendation of the Data and Safety Monitoring Board; the ITD did not improve survival to hospital discharge (see www.nih.gov/news/health/nov2009/nhlbi-06.htm).

Drug Therapy in Acute Management of Ventricular Fibrillation

The benefit of pharmacologic therapy as an adjunct to defibrillation in out-of-hospital VF is incompletely established. Despite many decades of use and ample laboratory experimental evidence of benefit, controlled clinical trials have not provided any clear evidence that either low-dose or high-dose epinephrine is beneficial in the treatment of patients with out-of-hospital VF.¹³⁷ Vasopressin, which appears to be superior to epinephrine in experimental models of cardiac arrest in improving survival from experimental VF, appears to be superior to epinephrine in some studies but not in others.^138–140 However, vasopressin has also not been proven superior to placebo in randomized controlled trials.

Similarly, the evidence base regarding antiarrhythmic therapy is incomplete. The standard therapy used to assist electrical defibrillation has been lidocaine; no evidence from controlled clinical trials is available to suggest that lidocaine is superior to placebo or any other agent in improving survival to hospital admission or survival to hospital discharge.¹⁴¹ In a meta-analysis of prophylactic lidocaine used in the post-infarction period, lidocaine may reduce the incidence of primary VF but appears to increase mortality rate.¹⁴² Lidocaine likely increases defibrillation thresholds and may increase the incidence of asystole. Magnesium is not superior to placebo in in-hospital cardiac arrest caused by VF.^143–145 Bretylium, although studied in VF, is no longer available.

Intravenous (IV) amiodarone for VF resistant to three defibrillation shocks was compared with placebo in a blinded, randomized clinical trial by Kudenchuk et al.¹⁴⁶ The 246 patients with shock-resistant, out-of-hospital VF were randomized to 300 mg amiodarone by IV bolus versus placebo; the primary study endpoint was survival to hospital admission, achieved in 44% of amiodarone-treated patients versus 34% of placebo-treated patients. In a related study, IV amiodarone was compared with IV lidocaine in a blinded, randomized trial of patients with VF persisting after three shocks, IV epinephrine, and a further defibrillation shock. Survival to hospital admission was achieved in 23% of amiodarone-treated patients versus 12% of lidocaine-treated patients (P = .009).¹⁴⁷ Although neither of these studies demonstrated statistically significant improvements in survival to hospital discharge, if any antiarrhythmic drug is to be used in out-of-hospital VF, on the basis of these studies it seems reasonable to consider amiodarone as the drug of choice.

It is important to note that resuscitation from VF is a dynamic clinical situation, and many patients will have multiple recurrences of VF in the seconds to minutes after initial defibrillation and also undergo multiple transitions among VF, asystole, pulseless electrical activity, and a perfusing organized rhythm during the course of a protracted cardiac arrest. The use of adjunctive drug therapy can therefore be considered a means to prevent recurrences of VF as much as to assist in defibrillation. In particular, patients with frequent recurrences of VF may benefit from intensive antiadrenergic therapy, IV amiodarone therapy, and occasionally from anti-ischemic therapy, revascularization therapy, or therapy with an intra-aortic balloon pump.^148,149

“Non-antiarrhythmic” Drugs that Prevent Sudden Cardiac Death

In the setting of heart failure, a cascade of neurohumoral activation initially supports perfusion.¹⁵⁰ Over time, however, the biologically active molecules released in the neurohumoral activation, including the sympathetic nervous system and the renin-angiotensin-aldosterone system, result in progressive left ventricular dysfunction. Pharmacologic antagonism of these two systems has proven to be an effective treatment paradigm.

Angiotensin-Converting Enzyme Inhibitors

The survival advantage conferred by angiotensin-converting enzyme (ACE) inhibitors in patients with heart failure has been demonstrated in a number of studies.^156–159 However, only in the Tandolapril Cardiac Evaluation Trial (TRACE) did the reduction in SCD reach statistical significance.¹⁶⁰ A recently published meta-analysis has now clearly demonstrated that ACE inhibitors reduce SCD.¹⁶⁰ In this study, 15,104 patients from 15 trials were studied. The relative risk of SCD in ACE inhibitor–treated patients was 0.80 (95% CI, 0.70 to 0.92).

Aldosterone Antagonism

The effect of spironolactone on survival was studied in the Randomized Aldactone Evaluation Study (RALES).¹⁶¹ A total of 822 patients on an ACE inhibitor with an LVEF of 35% or less and severe heart failure symptoms were randomly assigned to treatment with spironolactone or placebo. After a mean follow-up of 24 months, the mortality rate was 46% in the placebo group and 35% in the spironolactone group (P < .001). The relative risk of SCD in spironolactone-treated patients was 0.71 (CI, 0.54 to 0.95).

Recent studies have confirmed these findings and extended the benefit to patients with mild heart failure, in whom eplerenone, a selective aldosterone inhibitor, reduces all-cause mortality. A total of 12.5% of patients receiving eplerenone and 15.5% of those receiving placebo died (hazard ratio [HR], 0.76; 95% CI, 0.62 to 0.93; P = .008).¹⁶² Taken together, the data confirm that antagonizing the various components of the neuroendocrine activation that accompanies heart failure results in a reduction in SCD.

Anti-thrombotic and Anticoagulant Therapy

Accumulated data have suggested that ischemia caused by thrombus formation in stenotic coronary arteries may result in SCD. On the basis of pathologic examinations, greater than 80% of SCD cases in patients with ischemic heart disease may be associated with thrombus formation or plaque fissuring.^163,164 In the Second International Study of Infarct Survival (ISIS-2), aspirin use was associated with a reduction in total mortality and SCD.¹⁶⁵ A multivariate analysis of the 6797 participants in the Studies of Left Ventricular Dysfunction (SOLVD) demonstrated that both anti-platelet and anticoagulant therapies were associated with a reduction in the risk of SCD.¹⁶⁶ This provides a rationale for anti-thrombotic therapy, anti-platelet therapy, or a combination of both in patients with left ventricular dysfunction.

Antiarrhythmic Drugs

The importance of ventricular ectopy as a risk factor for SCD in patients with coronary disease is well established. It was therefore reasonable to hypothesize that suppression of premature ventricular complexes (PVCs) using standard antiarrhythmic agents that block sodium or potassium channels might reduce the risk of SCD.

Nonpharmacologic Therapy

Primary Prevention of Sudden Cardiac Death

At this point, no doubt exists about the capacity of the ICD to prevent arrhythmic death. This means that any patient destined to have a potentially fatal ventricular arrhythmia (in spite of otherwise appropriate medical therapy) could potentially benefit from having an ICD in place. A central issue for future investigation is the identification of patients at sufficiently high risk of SCD to justify the morbidity and cost of placing a device. Some subpopulations of patients described earlier may be effectively treated with amiodarone. For instance, a post hoc analysis of the AVID database suggests that ICDs have no benefit over amiodarone for patients with LVEF greater than 35%.¹⁹⁰ A randomized trial in such patients would be needed to make definitive treatment recommendations.

Certain populations of patients are known to be at risk for SCD, but because of competing risks of death, it is not clear whether the ICD will confer a benefit to the population as a whole. The Defibrillator in Acute Myocardial Infarction Trial (DINAMIT) study used a risk factor representing autonomic dysfunction as a risk stratifier to allow defibrillation therapy to more effectively focus on the highest-risk patients. A reduction in heart rate variability (HRV) has been shown to portend a worse prognosis in patients with CAD.^191–193 Patients in DINAMIT had low HRV and an ejection fraction of 35% or less. They were within 40 days of an acute MI. No mortality benefit was seen in patients randomized to the ICD (n = 332) compared with control patients (n = 342), with an HR for all-cause death of 1.08 (95% CI, 0.76 to 1.55) in the ICD versus control groups, over 30 ± 13 months of follow-up.¹⁹⁴

In the Defibrillators In Non-Ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) study, 458 patients with nonischemic cardiomyopathy, an LVEF of 35% or less, and more than 10 PVCs per hour or any episode of unsustained VT were randomized to standard medical therapy versus standard therapy plus ICD. The HR for all-cause mortality after 29 ± 14.4 months of follow-up in the ICD versus control groups was 0.65 (95% CI, 0.4 to 1.06; P = .08).¹⁹⁵

The entire discussion to this point has focused on populations known to be at high risk for SCD. Myerberg and others have emphasized the fact that the majority of cardiac arrests occur in the low-risk, but very large, population of patients whose increased risk has not come to clinical recognition.¹⁹⁶ To make major inroads into the prevention of SCD, learning how to screen the asymptomatic population inexpensively, but safely and effectively, will be necessary. Identifying high-risk patients in the low-risk, asymptomatic population and identifying members of known high-risk populations who are not at high enough risk to justify ICD placement is the next frontier in SCD prevention.