Juan Tamargo and Eva Delpón

Chapter Outline

Antiarrhythmic Drugs: Introduction

Cardiac arrhythmias represent a problem in current clinical practice; they are a major cause of morbidity and mortality and are an increasing economic burden for health care systems.¹ Treatment of cardiac arrhythmias with antiarrhythmic drugs (AADs) has two main objectives: relieve symptoms and complications (improve quality of life) and reduce mortality directly related to the arrhythmia. A basic principle in pharmacology is that the best treatment is targeted specifically to disease mechanisms. However, in many patients the ultimate underlying mechanisms of the arrhythmia remain incompletely understood. Thus, the choice of a given AAD is empiric and based on the characteristics of the arrhythmia, the pharmacologic properties of the AAD, and, above all, its safety profile. Moreover, triggers and arrhythmogenic substrates can vary among patients with the same arrhythmia depending on the underlying structural heart disease (i.e., coronary artery disease [CAD], heart failure [HF], left ventricular [LV] hypertrophy, hypertension). This variation could explain why AADs produce widely divergent effects on different patients ranging from termination of the arrhythmia to inefficacy, to exacerbation of the treated arrhythmias or generation of entirely new ones (proarrhythmia).

The selection of pharmacologic or nonpharmacologic therapeutic strategies differs with respect to the origin of the arrhythmia. AADs are used for ventricular tachycardias (VTs) in competition with implantable cardioverter defibrillators (ICDs) and ablation techniques as additional therapy to reduce the number of necessary shocks. Atrial tachyarrhythmias, particularly atrial fibrillation (AF), which is the most common arrhythmia, are the domain for drug treatment, although AF is particularly difficult to treat because of its tendency to become persistent in the course of the disease.

Until recently, arrhythmias were primarily considered a purely electrophysiological problem, and AADs mainly target cardiac Na⁺, Ca²⁺, and K⁺ ion channels (Figure 54-1). They bind to specific receptor sites within the channel drug affinity being strongly modulated by the channel state in a time- and voltage-dependent manner.² In addition, some AADs modulate the autonomic tone, primarily by antagonizing β1-adrenoceptors (β-blockers) or muscarinic receptors (atropine, disopyramide) or by stimulating adenosine A1 receptors (adenosine).

Conventional AADs are most commonly grouped according to the Vaughan-Williams classification in four different groups (Table 54-1): Na⁺ channel-blocking drugs (class I), β-blockers (class II), K⁺ channel blockers (class III) that prolong action potential (AP) duration (APD) and refractoriness, and L-type Ca²⁺ channel blockers (class IV). However, most AADs are “dirty drugs” that in a narrow range of concentrations simultaneously block different types of channels and modulate adrenergic and/or muscarinic receptors (e.g., quinidine, disopyramide, propafenone, amiodarone, dronedarone). Moreover, some old (adenosine, digoxin) and new AADs (ivabradine) cannot be listed in any of the original four classes. Table 54-2 shows drug selection based on the mechanisms underlying cardiac arrhythmias, and Figure 54-2 shows the main cellular targets for AADs.

Na⁺ Channel Blockers: Class I AADs

Na⁺ channel blockers bind and unbind in a strongly time- and voltage-dependent manner to a receptor site within the pore-forming subunit (Nav1.5) of the channel when channels are in the activated or inactivated state.² Na⁺ channel blockers also slow Na⁺ channel reactivation (transition from the inactivated to the resting state) upon repolarization, which prolongs refractoriness independently from changes in APD—that is, they increase the effective ratio of refractory period (ERP) to APD (postrepolarization refractoriness). As a consequence, the effects of class I AAD would be expected to increase: (1) at faster rates of stimulation (use-dependent block), because Na⁺ channels spend more time in activated and inactivated states, and the diastolic time for recovery from drug-induced block is shortened; and (2) in depolarized cardiac tissues (voltage-dependent block) because membrane depolarization inactivates Na⁺ channels and slows recovery from block.

Clinical and experimental data suggest that block of atrial Na⁺ channels can terminate AF. Indeed, propafenone and flecainide are first-choice drugs for AF cardioversion, but only in patients without structural heart disease. Conversely, ventricular Na⁺ channel block is associated with proarrhythmic effects. Class I AADs decrease excitability and slow ventricular conduction and can increase all-cause mortality and, therefore, they have no role (or are contraindicated) in the primary prevention of sudden cardiac death (SCD) because of life-threatening ventricular tachyarrhythmias (VT or ventricular fibrillation [VF]) in high-risk patients with post–myocardial infarction.³ In secondary prevention, the use of class I AADs is limited because of their low efficacy and serious adverse effects that may offset the therapeutic benefit, including proarrhythmia, multiorgan toxicity, reduced cardiac function, or worsening coexisting diseases.^4–6

Class I drugs are subdivided into drugs with intermediate (IA), fast (IB), and slow (IC) onset/offset kinetics of Na⁺ channel block (see Table 54-1). Recently, it has been demonstrated experimentally that drugs with fast onset/offset kinetics of Na⁺ channel block, such as vernakalant and ranolazine, could selectively target atrial versus ventricular Na⁺ channels.^7,8 This targeting would allow them to exhibit a disease-specific component of atrial-selectivity because of the high atrial rate in AF, which enhances block of atrial over ventricular Na⁺ channels. Indeed, it has been proposed that the faster the AADs dissociation kinetics the more selective the atrial effect during AF and the fewer the ventricular proarrhythmic effects, because the fast dissociation kinetics allows recovery from block at slower frequencies (e.g., those of the ventricles either at sinus rhythm [SR] or during AF). Moreover, atrial cells exhibit a slightly more depolarized resting membrane potential (RMP) and a more gradual phase 3 of the AP, which at rapid atrial rates results in a less negative take-off potential.^7,8 In addition, atrial cells exhibit a more negative potential for half-maximum inactivation of Na⁺ channels than ventricular myocytes. Because reactivation depends on membrane potential, fewer Na⁺ channels recover from the inactivated state during diastole in atria than in ventricles. Therefore, Na⁺ channel blockers that preferentially bind to the inactivated state of the channel and have a fast dissociation rate will exhibit atrial selectivity, because steady-state drug binding and consequently channel block will be larger in atria than in ventricles.⁸ However, block of K⁺ channels, particularly those preferentially or exclusively present in the atria, seems to be also required to meet the “class I atrial selective” profile. Otherwise lidocaine, which is highly selective for the inactivated state and exhibits a fast onset and offset kinetics, would be also atrial selective.

Indeed, class IA and IC, like class III AADs, also block several K⁺ channels and prolong ventricular (quinidine) or atrial (propafenone and flecainide) APD, depending on the type of K⁺ channel they preferentially block (discussed later). Furthermore, amiodarone, dronedarone, ranolazine, and vernakalant, which produce an atrial-selective I_Na block and in animal models suppress AF with minor ventricular effects, also inhibit the rapid component of the delayed rectifier current (I_Kr) and other K⁺ currents prolonging atrial APD. This effect that enhances the use-dependent I_Na block by further reducing the diastolic interval at fast rates.8,9

β-BLOCKERS: CLASS II AADs

β-Blockers include a heterogeneous group of drugs that inhibit sympathetic effects. Antiarrhythmic effects of β-blockers are the result of their electrophysiological effects (β-adrenoceptor stimulation modulates several ion currents, including hyperpolarization-activated inward current [I_f], I_Na, I_CaL, I_K1, I_to1, I_Kr, and I_Ks), inhibition of neurohumoral activation (and perhaps of sympathetic hyperinnervation and sprouting), and their antiischemic (reduce myocardial oxygen demands, increase subendocardial perfusion, and improve cardiac metabolism), antihypertensive, and antiproliferative effects.17,18 Indeed, β-blockers exert a beneficial effect on LV remodeling in patients with myocardial infarction (MI) or HF. Furthermore, in the ischemic myocardium, β-blockers decrease dispersion of ventricular repolarization and increase the VF threshold.

Sympathetic hyperactivity is the predominant change in the autonomic tone preceding malignant ventricular arrhythmias and SCD, whereas reduced vagal activity is associated with increased mortality in HF. β-Blockers remain among the few AADs that are effective for both primary and secondary prevention of SCD in different clinical conditions, including acute and chronic myocardial ischemia, congestive HF or LV dysfunction, and hypertrophic cardiomyopathy.3,17,18 Thus, most patients who have a propensity to develop VT/VF should receive β-blockers because they are a rational, mechanism-based therapy for the treatment of these arrhythmias. Moreover, β-blockers decrease the incidence of postoperative AF, which is explained by the role of an increased adrenergic tone in the genesis of this arrhythmia.

By inhibiting I_f and I_CaL, β-blockers decrease spontaneous activity of the sinoatrial node (SAN), inhibit the ectopic activity of the His-Purkinje system (Purkinje fibers are particularly prone to develop abnormal automaticity during increased sympathetic activation), and suppress abnormal automaticity generated in cardiac cells depolarized to potentials between −60 and −40 mV owing to the activation of I_CaL.¹⁸ In the atrioventricular node (AVN), they decrease conduction velocity and prolong refractoriness. β-Blockers are widely used to treat inappropriate sinus tachycardia and exercise-induced and supraventricular and ventricular arrhythmias where an increase in sympathetic tone plays a role¹⁷; they are the most effective drugs for ventricular rate control in AF and are first-choice drugs in the treatment of supraventricular reentrant tachycardias where the AVN forms part of the circuit.^1,17 Finally, β-blockers are effective in controlling the proarrhythmia induced by class I AADs, probably because of their bradycardic effect, which decreases the rate-dependent conduction slowing induced by Na⁺ channel blockers.

β-Blockers decrease all the proarrhythmic actions secondary to the increase of extracellular Ca²⁺ influx through L-type Ca²⁺ channels, sarcoplasmic reticulum Ca²⁺ release, and intracellular Ca²⁺ overload.17,18 Thus, they suppress EADs and DADs and are first-choice drugs in drug-induced torsades de pointes (TdP) and in patients with LQTS types 1 and 2 or catecholaminergic polymorphic VT (CPVT).^19,20 Surprisingly, and even when there are marked pharmacologic differences among β-blockers, there are few direct comparator data regarding their efficacy.

K⁺ Channel Blockers: Class III Antiarrhythmic Effects

These drugs prolong cardiac APD and refractoriness effects that, under theoretical bases, can suppress reentry. Some drugs (e.g., dofetilide) are considered pure class III AADs because they selectively block the channels that generate the I_Kr (named Kv11.1 or HERG), thus prolonging both atrial and ventricular APD and refractoriness in the absence of effects in conduction velocity.9,21 Experimental data demonstrated that selective I_Kr inhibition produces a greater prolongation of APD and refractoriness in atria versus ventricles, at least at normal SR frequencies. Unfortunately, at fast driving frequencies, the relative role of I_Kr in atrial and ventricular repolarization diminishes so much that I_Kr inhibition is not able to prolong APD during the arrhythmia episodes. This property, known as reverse use-dependence, limits the clinical efficacy of these drugs to stop fast arrhythmias, particularly AF. Moreover, marked I_Kr inhibition produces excessive and inhomogeneous ventricular APD prolongation, results in TDR, and predisposes to EADs, which in turn can trigger TdP. Inhibition of I_Kr, and hence proarrhythmic risk, increases in the presence of bradycardia, hypokalemia, HF, or other I_Kr-inhibiting drugs. Indeed, many other drugs, including antihistamines, antipsychotics, antimicrobials, and diuretics, are able to markedly block Kr channels, thus prolonging the QT interval (drug-induced LQTS) and increasing the incidence of TdP and SCD. It has been proposed that prolongation of APD produced by pure I_Kr inhibitors is characterized by a set of four disturbances of repolarization including triangulation of the AP, reverse use-dependence, instability, and dispersion of APD whose magnitude would help to predict drug-induced proarrhythmia risk.²²

HERG channel blockers gain access, from the intracellular side of the membrane, to a binding site located within the central cavity of the channel. The inner vestibule of HERG channels is larger than that of other K⁺ channels and presents two aromatic residues on each of the four subunits assembled to compose the conducting pore.²³ These topological characteristics explain the marked pharmacologic promiscuity of HERG because they allow drugs with rather different chemical structures to establish interactions with the aromatic moieties by π electron stacking and block K⁺ efflux through the pore. Finally, it is worth mentioning that besides the “pure Kr blockers” many AADs, including amiodarone, quinidine, propafenone, flecainide, and ranolazine, also block HERG channels at therapeutic concentrations, an effect that could contribute to their antiarrhythmic properties particularly at the atrial level (i.e., amiodarone, ranolazine). However, unfortunately in some cases, HERG blockade can increase their ventricular proarrhythmic effects (quinidine).

In healthy human atrial and ventricular myocytes driven at SR frequencies, I_Ks seems to be small, and its pharmacologic inhibition does not alter APD.²⁴ Importantly, the role of I_Ks in determining APD rises in prominence at increasing beating frequencies or under β-adrenergic stimulation, which causes channel accumulation in closed states near the open state. In addition, I_Ks helps to terminate the AP when the repolarization reserve is compromised (i.e., I_Kr decrease).²⁵ Moreover, recent data demonstrated that chronic AF increases by 100% I_Ks density in myocytes from both the right and left atria.²⁶ Mathematical models suggest that this chronic AF-induced remodeling of I_Ks density contributes to the atrial APD and refractoriness shortening produced by the arrhythmia, which in turn further enhances reentry, thus favoring AF maintenance. Moreover, there are data indicating that stable fast reentry sources (rotors) occur with significantly higher rotation frequencies, lower conduction velocities, and shorter AP in cells with prominent I_Ks.²⁷ In addition, the frequency-dependent accumulation of I_Ks promotes post-repolarization refractoriness and fibrillatory conduction of waves emanating from rotors. Therefore, it seems reasonable to propose that I_Ks could be an interesting target for the treatment of AF. In fact, some AADs, such as propafenone, quinidine, and amiodarone, inhibit I_Ks—an effect that probably contributes to their antiarrhythmic effects.²¹ Unfortunately, selective inhibition of I_Ks would be difficult to achieve, because molecular determinants for channel block are similar to those required for HERG blockade and would not be devoid of ventricular proarrhythmic effects. Indeed, patients carrying mutations in KCNQ1 or KCNE1 genes, which encoded the α (Kv7.1) and β (minK) subunits that form the channels, present LQTS types 1 and 5, respectively. This finding suggests that excessive inhibition of I_Ks could also lead to a marked prolongation of ventricular APD, particularly under conditions of increased sympathetic tone that promote EADs and TdP.

I_to1 (4-aminopyridine sensitive component of the transient outward current) determines the amplitude of the phase 1 of early repolarization, and thus the height of the AP plateau, which in turn influences the activation of other currents involved in repolarization (mainly I_CaL, I_Kr, and I_Ks).²⁸ Variations of I_to1 strongly influence the shape of cardiac AP, intracellular Ca²⁺ transients, and cardiac contractility. Relative I_to1 density is much higher in atrial, Purkinje, epicardial, and midmyocardial cells than in endocardial cells. I_to1 densities are also reported to be higher in right than in left ventricular myocytes. Prominent I_to1 leads to the “spike and dome” morphology typical of epicardial (particularly right) and some atrial AP (I_to1 is also differentially expressed among atrial cells). HF and chronic AF reduce I_to1 density, this effect being more marked in the left than in the right atria. In regions of the myocardium exhibiting a prominent I_to1, marked accentuation of the AP notch results in a transmural voltage gradient, leading to coved ST-segment elevation diagnostic of Brugada syndrome (BrS).²⁹ Because accentuation of the AP notch and loss of the dome could be secondary to either a decrease in inward currents (I_Na and I_CaL) or an increase in outward potassium currents (mainly I_to1), sodium channel blockers such as procainamide, propafenone, flecainide, ajmaline, and disopyramide can be used to induce or unmask ST-segment elevation in patients with concealed J wave syndromes (BrS and early repolarizing syndromes).²⁹ Conversely, quinidine, which is a potent I_to1 inhibitor, reduces the magnitude of the J wave and normalizes ST-segment elevation. The efficacy of quinidine in the prevention of BrS-associated arrhythmias is currently under evaluation.

I_Kur is carried by Kv1.5 channels and is detected only in the human atria, but not in the ventricle. Therefore, I_Kur inhibitors would be expected to prolong atrial, but not ventricular APD and refractoriness.³⁰ In myocytes from patients with SR, selective I_Kur inhibition shifts the plateau phase to more positive potentials at which the I_Kr is rapidly activated, so that mid to late repolarization is accelerated and APD is slightly shortened.³⁰ This could explain why selective I_Kur inhibitors do not suppress or even induce AF in experimental models.⁸ Indeed, KCNA5 (the gene encoding Kv1.5) loss-of-function mutations can cause AF. Chronic AF reduces I_CaL and markedly shortens atrial APD, so that I_Kur blockers increase the plateau potential over a range (below 0 mV) where the activation of I_Kr and I_Ks is less pronounced, and prolong atrial APD.³⁰ Furthermore, mathematical models predict that I_Kur or I_to1 inhibition can induce rotor termination.³¹ However, chronic AF also reduces I_Kur, particularly in the right atria; this decrease suppresses the endogenous right-to-left gradient of the current.²⁶ As a result, the relative contribution of I_Kur to repolarization in chronic AF can be small, decreasing the atrial sensitivity to I_Kur inhibitors.³⁰ These findings raise serious doubts on the usefulness of “pure” I_Kur blockers, and in fact none of those currently under development have been successfully launched for therapeutics.³² Conversely, vernakalant, which combined blockade of I_Kur, I_to1, I_Kr, and I_KACh, together with fast blockade of inactivated Na⁺ channels produces a high conversion rate to SR in patients with recent-onset AF.

I_K1 contributes to the final phase of repolarization and stabilizes the RMP of atrial and ventricular myocytes.³³ I_K1 increase hyperpolarizes RMP, shortens APD, reduces spontaneous activity, and accelerates and stabilizes arrhythmia-maintaining rotors, playing a role in the genesis of AF and VF.^31,33,34 I_K1 exhibits differential biophysical properties in atrial and ventricular tissues. Moreover, experimental data suggest that the Kir2.1 channel is the major isoform underlying ventricular I_K1, whereas relative contribution of Kir2.2 and 2.3 to atrial I_K1 seems to be greater. I_K1 increases in left and right atrial myocytes from patients with chronic AF, an augmentation that is greater in the left than in the right atria in patients with paroxysmal AF. Left-to-right ventricular differences in I_K1 have been postulated to represent a mechanism for the preferential stabilization of high-frequency rotors in the LV.³⁴ Moreover, patients carrying gain-of-function Kir2.1 mutations exhibit type 3 short QT syndrome characterized by a high incidence of VF and SCD. Furthermore, flecainide and propafenone at therapeutically relevant concentrations selectively increase I_K1 current generated by Kir2.1 homotetramers, an effect that could contribute to their ventricular proarrhythmic effects. Overall, it seems that I_K1 inhibitors could be an efficacious antifibrillatory strategy. Unfortunately, I_K1 inhibition also results in proarrhythmia. In fact, patients carrying loss-of-function Kir2.1 mutations exhibit congenital LQTS type 7 (Andersen-Tawil syndrome) characterized by APD prolongation, diastolic depolarization, and increased propensity for EAD- and DAD-triggered arrhythmias.³⁵ Moreover, the development of Kir2.1-based antiarrhythmics is hampered because Kir2.1 channels are expressed in many excitable tissues, which could lead to organ toxicity. More promising could be selective blockade of Kir2.3 over Kir2.1 channels for the treatment of AF. Kir2.3 channels are indeed preferentially inhibited by several drugs taking advantage of the low affinity of Kir2.3 channels for phosphatidylinositol 4,5-bisphosphate (PIP2), which critically determines the channel function.

Acetylcholine released from vagal nerve terminals activates I_KACh. Because I_KACh is exclusively present in the atria, it has been proposed as a target for development of atrial selective AADs. I_KACh activation slows SAN pacemaker activity, produces a heterogeneous APD and ERP shortening, and hyperpolarizes RMP, increasing Na⁺ channel availability. These effects create a substrate for reentry, accelerate and stabilize high frequency rotors, and contribute to initiation and maintenance of AF.³⁴ Some AADs inhibit I_KACh via the blockade of muscarinic receptors (disopyramide) and may be useful in vagally mediated AF. Others inhibit I_KACh via GTP binding proteins or the KACh channel itself (propafenone, dofetilide, and amiodarone). The new multichannel blocker dronedarone is 100-fold more potent in blocking I_KACh than amiodarone, an effect that could contribute to its antiarrhythmic properties. Importantly, it has been demonstrated that chronic AF also reduces I_KACh, reducing its relative role in atrial repolarization while increasing an I_KACh component (I_KAChC) with constitutive activity (active in the absence of muscarinic ligands).³⁶ Ideally, an AAD should selectively block I_KAChC (an atrial- and disease-specific therapeutic target) without affecting the I_KACh that modulates SAN and AVN. I_KAChC is inhibited by flecainide, AVE0118, and investigational AADs, but highly selective I_KAChC blockers are not currently available.

In up to 80% of cases, VT/VF occurs in the setting of CAD and HF. Attempts to prevent arrhythmic events and SCD with class I and III AADs in the setting of myocardial ischemia have failed. Activation of I_KATP during acute myocardial ischemia results in shortening of the APD, accumulation of extracellular K⁺, membrane depolarization, and slow conduction velocity. These effects render the ischemic heart vulnerable to reentrant arrhythmias. It has been hypothesized that specific I_KATP inhibitors prolong ventricular refractoriness in ischemic tissues with little or no effect on normal tissue, being truly ischemia-selective AADs.³⁷ Conversely, the KATP opener nicorandil shortens the APD, reduces TDR, and suppresses EADs in patients with LQT1.³⁸

Calcium Channel Blockers: Class IV AADs

Nondihydropyridine Ca²⁺-channel blockers (e.g., diltiazem, verapamil) inhibit the I_CaL and exhibit antiischemic and antihypertensive properties. These blockers stabilize the channel in its inactivated state and prolong its reactivation, so that their effects increase at fast rates (use-dependent block) and at depolarized membrane potentials (voltage-dependent block).¹

In cardiac tissues with RMPs positive to −60 mV (i.e., SAN, AVN, and ischemic tissues), I_CaL is responsible for AP depolarization (phase 0) and conduction velocity. Class IV AADs prolong the refractory period and slow conduction through the AVN. The rate-dependent effect on the AVNs is the basis for their use to terminate or prevent reentrant supraventricular tachycardias whose circuit involves the AVN, being an alternative to adenosine, and to control the ventricular rate in patients with AF or flutter. At therapeutic concentrations, diltiazem and verapamil decrease heart rate slightly, because their direct negative chronotropic effect is partly counteracted by their peripheral vasodilator effects that produce a sympathetic-mediated reflex response. However, in patients treated with β-blockers, reflex sympathetic stimulation is inhibited, and diltiazem and verapamil slow heart rate and AVN conduction.

Inhibition of I_CaL suppresses automaticity generated in depolarized ventricular cells (abnormal automatism) and EADs occurring at the plateau phase of AP because of I_CaL reactivation (as marked delay in repolarization allows Ca²⁺ channels to recover from inactivation). Thus, class IV AADs and magnesium sulfate are effective for the treatment EAD-triggered arrhythmias. On the other hand, abnormal Ca_i²⁺ overload activates forward mode of the NCX (Ca²⁺ extrusion) generating a transient inward current (I_ti) that underlies phase 3 EADs and DADs. Class IV AADs and AADs that block Ca²⁺ channels at therapeutic concentrations (e.g., amiodarone, dronedarone, propafenone) suppress DADs by preventing Ca_i²⁺ overload.

Verapamil is a drug of choice in idiopathic outflow tract and fascicular VT and drug-induced TdP; it can be an alternative to β-blockers in CPVT and prevents VTs related to ischemia. However, class IV drugs have no effect on all-cause mortality in patients with prior MI and are contraindicated in congestive HF.³

Gap-Junction Coupling Enhancers

Connexins (Cx) are membrane proteins that assemble into hexameric hemichannels, also known as connexons. At the gap junctions, connexons allow electrical flow between cardiomyocytes and thus regulate intercellular coupling. Changes in Cx expression, location (lateralization), and function occur in many forms of heart disease (i.e., ischemia, hypertrophy, HF, AF) and contribute to arrhythmogenesis as they slow conduction velocity, cause heterogeneities in repolarization, and modulate automaticity. Cx40 has a critical role in mediating atrial conduction, whereas Cx43 seems to be the major ventricular form. Cellular uncoupling owing to dephosphorylation and redistribution of Cx43, together with increased fibrosis⁴⁰ and decreased expression of Na⁺ channels, are implicated in conduction slowing during ischemia, increasing the risk of fatal ventricular arrhythmias. Rotigaptide increases gap junction conductance and prevents acute ischemic arrhythmias. However, it partially reverses the loss of Cx43 but does not restore normal conduction or prevent arrhythmias in the healing infarct border zone, such as after a prolonged period of gap junction remodeling.⁴¹ Thus, it seems that interventions that are effective in restoring cell-to-cell coupling under conditions of acute ischemia might not be effective in a substrate that has already undergone a prolonged period of gap junction remodeling. Gap junction–enhancing drugs can also reduce AF vulnerability in some models of AF (chronic mitral regurgitation and acute ischemia), but not in others (HF or atrial tachypacing).⁴¹ These findings suggest that the role of Cx in AF is disease specific and that gap junction enhancers may be effective only when conduction slowing is due to alterations in gap junctions.

Stretch-Activated Ion Channels

Chronic stretch causes heterogeneous conduction, prolongs refractoriness, and induces structural remodeling (dilatation, hypertrophy, fibrosis), which contributes to both atrial and ventricular arrhythmias. Recent evidence suggests that these electrophysiological effects can be explained via the activation of the nonselective cationic transient receptor potential TRPC6 channels.⁴² In experimental models, nonselective stretch-activated channel blockers (gadolinium and the tarantula peptide GsMtx-4) suppress AF inducibility.¹⁴ Nowadays, modulation of stretch-activated channels as an antiarrhythmic strategy is limited by the lack of effective drugs.

Modulation of Ion Channel Trafficking

Structural heart diseases and many mutations associated with channelopathies affect biogenesis, forward trafficking to the surface membrane subdomains, and degradation of cardiac ion channels.⁴³ Recent evidence demonstrates that AADs inhibit ion conduction and modulate ion channel trafficking or surface density and that channel blockade and trafficking can be targeted independently.^43,44 Some HERG channel blockers (cisapride, E4031, quinidine) can rescue defective HERG surface trafficking, whereas others (ketoconazole and fluoxetine) can decrease HERG surface density. Furthermore, there are drugs that do not block HERG channel conductance, but either decrease (pentamidine, probucol) or increase (fexofenadine) HERG surface density.⁴⁵ Moreover, type 2 LQTS-associated mutant channels can be rescued to the plasma membrane with drugs (E4031, terfenadine, fexofenadine) that likely stabilize misfolded protein in the endoplasmic reticulum, although the drug effectiveness depends on the particular mutation present.⁴⁵

Quinidine induces a stereospecific dose-, time- and subunit-dependent internalization of Kv1.5, concomitant with the classical pore block.⁴³ The specific expression of Kv1.5 channels in human atria raises the possibility of designing drugs to specifically modulate Kv1.5 channel trafficking.⁴³ Such drugs, in theory, can cause a rapid and reversible decrease in I_Kur and a prolongation of atrial APD terminating AF, with minimal risk of ventricular proarrhythmia. However, there is a need for a better understanding of trafficking mechanisms and how chronic decrease of channel expression contributes to arrhythmia-induced remodeling.

Some drugs such as pentamidine and flecainide decrease and increase Kir2.1 channel expression in the membrane, respectively. Flecainide is also able to promote trafficking of some Kir2.1 loss-of-function mutants increasing the amplitude of the outward K⁺ current generated by the channels, if any. This effect could contribute to the empirically observed antiarrhythmic actions produced by flecainide in some patients with Andersen-Tawil syndrome.⁴⁴

Targeting Intracellular Calcium Handling

Abnormal intracellular Ca²⁺ handling can promote both atrial and ventricular arrhythmias; it also has a key role in cardiac electrical and structural remodeling in patients with AF, HF, LV hypertrophy, and CAD.14,46–48 Moreover, CPVT associated with mutations in the sarcoplasmic reticulum Ca²⁺ release channel/ryanodine receptor (RyR2) and calsequestrin (CASQ2) genes destabilize the RyR2 channel complex and increase spontaneous Ca²⁺ release, which under certain conditions (e.g., during exercise or β-adrenergic stimulation) facilitates the development of DADs and triggered arrhythmias.^20,46 Thus, intracellular Ca²⁺-handling proteins (Ca²⁺-ATPase [SERCA2a], phospholamban, RyR2, and its accessory protein calstabin 2, NCX, and CSQ), provide targets for developing new AADs.

Dysregulation of RyR2, characterized by improper SR Ca²⁺ release and increased diastolic Ca²⁺ leak is related to: (1) RyR2 hyperphosphorylation by protein kinase A (PKA) and Ca²⁺/calmodulin-dependent kinase II (CaMKII), which transiently dissociates calstabin 2 from the RyR2 complex increasing open channel probability; (2) increased sensitivity of RyR2 to activation by luminal Ca²⁺; and (3) reactive oxygen species generated in the diseased heart, which make the channel leaky.14,46–48 β-Adrenergic stimulation and renin-angiotensin-aldosterone system (RAAS) can promote triggered arrhythmias via PKA- and CaMKII-induced RyR2 hyperphosphorylation.^46,47 β-Blockers prevent PKA- and CaMKII-mediated hyperphosphorylation of RyR2, depletion of calstabin 2, and SR Ca²⁺ leak being the only effective drugs in improving survival in CPVT.²⁰ β-Blockers and verapamil act synergistically to prevent stress-induced increase in sarcoplasmic reticulum Ca²⁺ content by reducing heart rate, Ca²⁺ influx into the cell, and Ca²⁺ uptake into the sarcoplasmic reticulum.²⁰ However, in patients with CPVT, cardiac events remain considerable even in this combination, so that alternative therapies are needed.

Carvedilol, a nonselective β- and α1-adrenergic antagonist, blocks several ion channels (I_Kur, I_Kr, I_CaL, I_to1, and I_Ks) and exerts antioxidant and antiproliferative effects. It is the only β-blocker that suppresses sarcoplasmic reticulum Ca²⁺ release by directly reducing RyR2 mean open time, independently of its β- or α1-antagonism or antioxidant activities, which likely contributes to its antiarrhythmic effects in HF patients.20,46 Carvedilol analogues with minimal β-blocking activity retain the ability to suppress Ca²⁺ release and prevent CPVT without causing bradycardia. Interestingly, the combination of these analogues with selective β-blockers (metoprolol or bisoprolol) is more effective than each agent alone for preventing CPVT.

Stabilizers of the calstabin 2–RyR2 complex (Rycals) represent a new antiarrhythmic approach. Rycals (such as JTV519 and S44121) increase the affinity of calstabin-2 for PKA-hyperphosphorylated RyR2, reduce channel open probability and prevent arrhythmogenic diastolic Ca²⁺ leak without affecting normal RyR2.14,20,46,48 As a consequence, JTV519 prevents APD alternans and triggered ventricular arrhythmias and reduces Ca²⁺ waves in arrhythmogenic Purkinje fibers following MI. It also reduces ectopic activity and DADs in pulmonary vein cardiomyocytes, decreasing AF inducibility.⁴⁷ Its putative efficacy in the treatment of AF and ventricular arrhythmias associated with HF and CPVT is currently being tested.

Flecainide and propafenone block the open state of RyR2, inhibit arrhythmogenic diastolic Ca²⁺ waves, and suppress DADs and triggered arrhythmias in experimental models and in patients with CPVT.20,47 In addition, their Na⁺ channel–blocking properties prevent DADs from reaching threshold potential and from triggering premature beats. Furthermore, β-blocking activity of propafenone can also contribute to its clinical efficacy in CPVT. As a result, class IC AADs can be an alternative to β-blockers for some patients with CPVT.^20,47

Some RyR2 mutations can lead to an abnormally tight domain-domain interaction that results in an erroneous activation of the channel and diastolic Ca²⁺ leak.⁴⁶ In experimental models, dantrolene stabilizes domain interactions within the RyR2 and prevents aberrant Ca²⁺ release and CPVT.²⁰ The antioxidant edaravone also ameliorates the defective interdomain interaction of the RyR2 and prevents Ca²⁺ leak and LV remodeling during the development of HF.⁴⁸

Even when strong evidence suggests that CaMKII control over Ca²⁺ handling might be more potent than that of PKA, CaMKII inhibitors might not represent a feasible target, because the ubiquitous role of CaMKII in cellular physiology can result in deleterious off-target side effects.⁴⁸ In addition, some CaMKII inhibitors (KN-93) cannot discriminate between CaMKII and CaMKIV and inhibit voltage-gated K⁺ and Ca²⁺ channels. Therefore, to achieve an effective and safe cardiac effect, drugs with tissue and isoform-specific action (CaMKIIδ) are probably required. An alternative strategy is small peptides like autocamtide-2-related inhibitory peptide (AIP) that inhibit CaMKII activation by blocking calmodulin binding to CaMKII.

Pharmacologic Treatment of Inherited Cardiac Arrhythmia Syndromes

During the past decade, there was an explosion of information linking mutations in genes encoding ion channel α- and accessory-subunits and cytoskeletal molecules to inherited cardiac arrhythmia syndromes, including short QT syndrome, LQTS, BrS, AF, idiopathic VF, and CPVT (Table 54-3).^{19,29,35,38,45} Identification of disease-associated genes provides important information on the molecular mechanisms of cardiac arrhythmias and opens the possibility for antiarrhythmic therapies based on genotype and clinical presentation. However, randomized trials have not been conducted because of the rarity of these conditions. Nevertheless, therapy alleviating the mutation consequences depends on the mutation class, although most AADs lack the desired selectivity for a given channel.

Target Cardiac Remodeling: Upstream Therapies

Cardiac arrhythmias produce changes (remodeling) in the electrophysiological (ion channel and gap junction function or expression, abnormal Ca²⁺ handling) and structural (hypertrophy, fibrosis, dilatation) properties of the myocardium often associated with stretch, oxidative stress, inflammation, or ischemia, creating a substrate for reentry and focal activity.13,14 Furthermore, aging and structural heart diseases also produce electrophysiological and structural remodeling that increases susceptibility to arrhythmias and makes arrhythmias more drug resistant.^14,49 However, remodeling is a disease-specific process presenting important differences in signaling pathways and pharmacologic responses among comorbidities. Upstream therapies to prevent or delay myocardial remodeling are an attractive approach in the prophylaxis and treatment of arrhythmias associated with HF, hypertension, or MI.^{13,14,48,50,51} Potential drugs in this category include antiinflammatory agents, RAAS inhibitors (angiotensin-converting enzyme inhibitors [ACEIs], angiotensin II receptor antagonists [ARAs], aldosterone antagonists), omega-3 polyunsaturated fatty acids (PUFAs), statins, and antifibrotic agents.

Other AADs

Adenosine acting on cardiac A1 receptors activates an outward K⁺ current (I_KAdo) present in the atria, SAN, and AVN and inhibits I_f and adenylyl cyclase, which indirectly results in a decrease in catecholamine-stimulated I_CaL and I_TI. As a result, adenosine suppresses the pacemaker activity of SAN and depresses AVN conduction, being the first-line agent for terminating paroxysmal paroxysmal supraventricular tachycardias in which the AV node is part of the reentry pathway. In addition, adenosine suppresses catecholamine-stimulated EADs and DADs and catecholamine-mediated atrial and ventricular arrhythmias.

Digoxin inhibits Na⁺-K⁺–ATPase and decreases intracellular Na⁺ concentration, which in turn, increases via the NCX both Ca_i²⁺ and contractile force. In addition, digoxin enhances vagal tone and reduces sympathetic and RAAS tone. At therapeutic concentrations, digoxin decreases the automaticity of the SAN (vagal stimulation inhibits I_f) and slows conduction and prolongs refractoriness in the AVN. This latter effect is the basis for heart rate control in patients with AF and systolic HF, particularly in unstabilized HF patients in whom β-blockers and calcium antagonists are contraindicated. In permanent AF, the combination of digoxin with these drugs provides a satisfactory rate control both at rest and during exercise. Digoxin has no role in ventricular arrhythmias, does not modify cardiac mortality in patients with congestive HF in SR, and can increase the rate of SCD.

Raised resting heart rate is a strong independent risk factor for cardiovascular events. The antianginal drug ivabradine, a selective I_f inhibitor, added on background therapy reduces the risk of cardiovascular death or hospitalization for worsening HF in patients with symptomatic HF and an LV ejection fraction of 35% or less with heart rate 70 beats/minute or greater.⁶⁰ In addition, ivabradine reverses LV remodeling, suggesting that it prevents disease progression in these patients. Preliminary data suggest that ivabradine is a promising option in patients with inappropriate sinus tachycardia.

Unresolved Questions and Future Strategies

There are many unresolved questions concerning the pharmacologic treatment of arrhythmias. For example, it is still unclear whether targeting an individual ion channel is better than targeting multiple ion channels. By blocking both inward and outward currents, multichannel blockers can create steady-state conditions that avoid large variations in AP repolarization, thereby preventing the development of electrical instability, although the risk of proarrhythmia may increase.

As mentioned, VT/VF associated with SCD preferentially occurs in the setting of myocardial ischemia, and it would be optimal to develop ischemia-selective AADs. Because class I and III AADs have failed to prevent ischemia-induced arrhythmias, antiarrhythmic strategy has been shifted to drugs commonly prescribed in patients with MI or HF that prevent plaque rupture and reduce myocardial ischemia, such as antianginal drugs (β-blockers, ivabradine, ranolazine), statins, RAAS inhibitors, and platelet antiaggregants, in an attempt to eliminate potential triggers for ventricular arrhythmias and, in the long-term, prevent the formation of myocardial scars and SCD. RAAS inhibitors and β-blockers might also limit the substrate for ventricular arrhythmias because they prevent LV remodeling after MI.³

Another unresolved question is the duration and timing of AAD therapy. Chronic antiarrhythmic therapy increases the risk of proarrhythmia, and a suitable timing of antiarrhythmic drug therapy might enhance safety without reducing efficacy. It is unlikely that patients with infrequent, brief, and mild symptomatic arrhythmic episodes should receive long-term AAD therapy. As an alternative strategy, a single oral bolus dose of propafenone or flecainide (“pill in the pocket”) can be administered in selected patients to terminate recent-onset AF outside the hospital once treatment has proved safe in the hospital. However, few prospective data are available on the relative safety of this approach, and the decision to initiate therapy outside the hospital should be carefully individualized.

Because AF is the most common arrhythmia, there are strategies for the development of atrial-selective AADs devoid of ventricular proarrhythmic effects based on the identification of atrial-specific targets. There are preliminary experimental data suggesting that several channels represent a possible atrial therapeutic target, including two-pore–domain K⁺ channels, Ca²⁺-activated nonselective cation channels, small-conductance Ca²⁺-activated K⁺ channels (SK1-3 channels are expressed in the atria), and transient receptor potential channels of the canonical (TRPC1/3/6) and melastin-related (TRPM3/4/6/7) subfamilies.9,14 However, because drugs used to target these currents lack specificity, the functional role of these channels in humans remains uncertain. In the same line, abnormal Ca²⁺ handling plays a major role in arrhythmogenesis, but the developed drugs do not have sufficient specificity.

Conclusions

AADs remain the mainstay of therapy for the majority of patients with cardiac arrhythmias, but there is an unmet need for new drugs that achieve effective suppression and prevention of life-threatening arrhythmias without the risk of proarrhythmia. The rational development of these new AADs should be the final result of a better understanding of the molecular and cellular mechanisms involved in the genesis and maintenance of cardiac arrhythmias in different pathologic substrates and the mechanisms of action of AADs.

Acknowledgments

The authors thank Ricardo Caballero for comment on the manuscript and Paloma Vaquero for editorial assistance in manuscript preparation. This work was supported by grants from the Ministerio de Ciencia e Innovación (SAF2011-30088), Fondo de Investigaciones Sanitarias (PI11/01030), and Red Heracles (RD12/0042/0011).

References