Our understanding of the mechanisms of cardiac arrhythmias has advanced with the introduction of new technologies to study the electrical activity of the heart. These studies show that most arrhythmias are generated by the same ion channels that sustain normal activity. However, alterations in ion channel number and the intracellular and extracellular concentrations of ions in disease states provide the setting for arrhythmias. Structure is also very important, from the changes in protein structure in the long QT and Brugada syndromes to the remolding of the atria and ventricle in ischemia, cardiac hypertrophy and other forms of heart muscle disease. Reentry is the likely mechanism of most sustained arrhythmias in chronic heart disease. Abnormal impulse formation, as evidenced by early and delayed afterdepolarization, is the second major cause of arrhythmias. In some circumstances, such as focal atrial fibrillation, the arrhythmia may be initiated by abnormal impulse formation and sustained by reentry. The clinical setting—family history, drug use, electrolyte imbalance, and electrocardiographic (ECG) findings—may suggest a mechanism. Antiarrhythmic drugs act on membrane ion channels. Specificity, selectivity, and therapeutic index are low. As a result, physical strategies such as ablation and cardioverter defibrillator implantation are playing an increasingly important role in arrhythmia management.

This chapter reviews normal impulse generation and propagation in the heart and the ways that disturbances of either mechanism result in cardiac arrhythmias. Emphasis is placed on observations made at the cellular and subcellular levels. Analysis of the mechanisms of arrhythmias at the higher levels of tissues and in the in situ heart is the focus of Chapters 60 and 62.

The earliest direct recording of electrical activity of the heart used sensitive string galvanometers to record unipolar and bipolar electrogram to study regional spread of excitation. Lewis and colleagues used accurate extracellular recordings to relate propagation velocity to structure. The millimeter resolution of such recording has been replaced by voltage-sensitive dyes with a resolution of microns. These strategies have permitted the mapping of excitation spread in the basal state and during arrhythmias. The basis for normal and abnormal excitation had to await techniques that permitted transmembrane potential recordings. Ling-Gerrard capillary microelectrodes permitted transmembrane action potential recordings from groups of cells with little apparent damage to the cells. Those studies documented the relationship between the initial QRS complex and the T wave on the ECG to periods of rapid depolarization (phase 0) and repolarization (phase 3) of the action potential. Analysis of the ionic basis of the action potential proved more formidable. Membrane potential (V) was varying with time (dV/dT greater than 0) as excitation propagated over the surface of the heart (dV/dX greater than 0). Marmont and Cole recognized that the analysis of excitation required that one of the independent variables (time t, or distance x) had to be controlled (zero) as the experimenter varied V (2,3). These investigators developed the voltage clamp technique, which permitted measurement of the underlying current as a function of time as membrane voltage V was varied. This required that V should not vary with distance (i.e., dV/dX = 0, space clamping). The complexity of naturally occurring cardiac preparation
severely limited space clamping. However, the preparation of single calcium-tolerant cardiac myocytes by Powell and Twist enabled a quantitative voltage clamp of the major currents that contribute to the action potential (4).

Neher and Sakmann introduced the patch clamp technique that permitted quantitative voltage clamp of even the smallest cardiac cells such as those of the sinoatrial (SA) and atrioventricular (AV) nodes (5). The technique also permitted the resolution of current through single ion channels. All the major ion channels that contribute to the generation of the action potential have now been cloned and functionally expressed. Subgroups of arrhythmias occurring in the structurally normal heart have been shown to be monogenetic, arising from mutations of ion channel genes (6,7). The more common arrhythmias, such as ischemic-induced ventricular tachycardia and atrial fibrillation, are of multifactorial origin, with susceptibility genes playing an important role. Analysis of these and other arrhythmias starts with the premise that aberration of the normal mechanisms of impulse initiation and propagation are central to the genesis of arrhythmias. Therefore, normal impulse initiation and propagation are the starting points for the analysis of arrhythmia mechanisms.

The five phases—0, 1, 2, 3, and 4 — of the action potential recorded from a ventricular cell are illustrated in Figure 57.1. Phase 4 describes the membrane potential during diastole. Working myocardial cells maintain a steady diastolic level of approximately –85 mV. Conduction system myocytes in the SA and AV nodes, excitable cells of the mitral and tricuspid valve, the pulmonary veins, and the coronary sinus exhibit slow diastolic depolarization during phase 4. Excitation of resting cells results in a rapid transition from phase 4 to phase 0. This phase represents a period of rapid depolarization resulting in reversal of the transmembrane potential, and lasts less than 1 millisecond. Phase 1 is a period of rapid repolarization that returns the membrane to the level of the plateau. This phase (1) shows the greatest variability among various regions of the heart and among various species. It is prominent in Purkinje and epicardial cells, but it is diminutive or absent in endocardial cells. Phase 2 is the plateau phase of the action potential and lasts up to several hundred milliseconds. It is the phase that most clearly distinguishes the cardiac action potential from the brief action potentials of skeletal muscle and nerve. Phase 3 marks the period of terminal repolarization when the membrane potential is restored to its resting level.

An uneven (nonequilibrium) distribution of ions across the cell membrane is an essential basis for the action potential and is summarized in Table 57.1. Under physiologic conditions, the sodium/potassium (Na⁺/K⁺) pump moves three Na⁺ ions out of the cell in exchange for two extracellular K⁺ ions; one molecule of adenosine triphosphate (ATP) is consumed in the process. Because the stoichiometry of ion movement is not 1:1, the Na⁺/K⁺ pump generates outward current. The overdrive suppression of pacemaker activity following rapid stimulation results from the hyperpolarizing effect of the Na⁺/K⁺ pump current. Arrhythmias resulting from digitalis intoxication are secondary consequences of pump inhibition. The Na⁺/calcium (Na⁺/Ca²⁺) exchanger uses the Na⁺ gradient established by the Na⁺/K⁺ pump to transport Ca⁺ across the cell membrane. The Na⁺/Ca²⁺ exchanger transports Ca²⁺ into the cell at plateau levels of membrane potential and Ca²⁺ out of the cell during diastole. Because the stoichiometry of the exchanger is approximately 3:1, it is also electrogenic. The Na⁺/H⁺ and bicarbonate/chloride (HCO₃^–/Cl^–) exchangers also contribute to the uneven distribution of ions across the cell membrane.

During the action potential, selectively permeable ion channels open or close, and the direction of ion movement (and the resulting current) is determined by the ratio of the intracellular and extracellular concentration of each ion and the membrane potential. Ion movement is passive. We now review the ion channels that are crucial to each phase of the action potential in more detail.

Certain clinical states such as myocardial ischemia or reperfusion, cardiac failure, and myocardial hypertrophy are associated with an increased incidence of cardiac arrhythmias. Because the currents that generate normal action potential play a central role in most arrhythmias, these clinical states must exert some effect on the membrane currents. Membrane ion currents
can be modified in relatively few ways. Pathologic states may modify channel function by changing the transmembrane gradients for the permeant ions; they may alter channel gating by changing intracellular Ca²⁺, intracellular and extracellular pH, intracellular [ATP], and autonomic tone; and they may alter the level of expression of ion channels.

Myocardial ischemia is associated with an early rise in [K⁺]_o, a fall in intracellular and extracellular pH and [ATP], and an increase in [Ca²⁺]_i and [Na⁺]_i. Lysophospholipids also accumulate in the ischemic tissues. Elevation of [K⁺]_o has two major effects: membrane depolarization and an increase in the inward rectifier K⁺ current. The fall in intracellular [ATP] and the associated rise in ADP activates the ATP-sensitive K⁺ channel. In general, most ionic currents are reduced by a fall in intracellular and extracellular pH. The threshold for this effect is variable between ion channels. The lysophospholipids, such as lysophosphatidylcholine, block ion channels. They may also change channel gating (e.g., inactivation of the Na⁺ current is slowed) (18). In experimental models of acute ischemia, reperfusion is associated with a high incidence of ventricular tachycardia and ventricular fibrillation. Indeed, the prevalence of arrhythmias during reperfusion may exceed that during the proceeding ischemic period. In the clinical setting, vasospastic angina and percutaneous transluminal coronary angioplasty are associated with reperfusion after a variable period of ischemia. Accelerated idioventricular rhythm is the most common arrhythmia in clinical reperfusion; ventricular fibrillation has a much lower prevalence than in the experimental setting. Reperfusion is associated with a restoration of [K⁺]_o, pH toward normal, and generation of oxygen-derived free radicals. [Ca²⁺]_i may rise further and reduce the conductance of junctional channels.

Cardiac hypertrophy and failure are associated with a prolongation of the action potential. The most consistent finding is a reduction of the transient outward current. Available data suggest that the decrease in I_to is the result of decreased expression of the channel. The mediators of this suppression are not known. Cytokines may play a role.