Inhibition of late sodium current to reduce electrical and mechanical dysfunction of ischaemic myocardium

Abstract

This commentary on the review by DA Saint in the current issue of the British Journal of Pharmacology focuses on the pathological role of late I_Na in the heart, the evidence supporting inhibition of late I_Na as a therapeutic target in ischaemic heart disease, and the therapeutic applications and challenges for development of new late I_Na inhibitors. Recent reports from a large clinical outcome trial (MERLIN) of ranolazine, a drug known to inhibit late I_Na, indicated that it was safe and reduced recurrent ischaemia and arrhythmic activity. In combination with other results indicating that inhibition of late I_Na reduces ischaemia, myocardial Ca²⁺ overload, and electrical and mechanical dysfunction when late I_Na is increased, the new clinical trial results suggest that reduction of cardiac late I_Na is safe and therapeutically beneficial.

Introduction

The review by David Saint of the properties of the cardiac persistent late sodium current (I_Na) (Saint, 2007), and of the potential therapeutic role(s) of drugs that inhibit this current, recognizes that late I_Na may be increased in ischaemic and failing hearts and thus may be of general pathologic importance in heart disease. The discussion of blockers of late I_Na in the review is timely, given the recent FDA approval of ranolazine for the treatment of chronic angina (the primary symptom of ischaemic heart disease) and the publication of results from a large clinical outcome trial with the drug (Morrow et al., 2007; Scirica et al., 2007).

Why is enhanced late I_Na pathologic?

Enhancement of late I_Na is pathologic because it increases the intracellular concentration of Na⁺, Na⁺–Ca²⁺ exchange (NCX), Ca²⁺ entry into myocytes and duration of the myocardial action potential (AP). The increase of Ca²⁺ entry may be of sufficient magnitude to cause Ca²⁺ overloading with mechanical (e.g., impaired ventricular diastolic relaxation and reduction of diastolic coronary blood flow) and electrical (e.g., early and delayed afterdepolarizations: EADs and DADs) dysfunction (Belardinelli et al., 2006). Enhancement of late I_Na greatly increases Na⁺ entry during phase 2 of the ventricular AP (Makielski and Farley, 2006). Because net outward current is small during phase 2, an enhanced late I_Na reduces repolarization reserve and increases AP duration (APD). When phases 2 and 3 are prolonged, Ca²⁺ and Na⁺ channels may reopen, and the subsequent cation influx may lead to further APD prolongation and EADs. Spatial heterogeneity of prolongation of APD in the ventricle due to enhanced late I_Na may set the stage for triggered and re-entrant arrhythmias.

Enhanced late I_Na disrupts Na⁺ and Ca²⁺ homeostasis. Sodium entry leads to increases of intracellular Na⁺ concentration, reverse mode (3 Na²⁺ out for 1 Ca²⁺ in) NCX and Ca²⁺ entry. The increase of Ca²⁺ entry during phase 2 of the AP plateau may not only support systolic contraction, but also cause Ca²⁺ overload. Acute cellular Ca²⁺ overloading has negative consequences. Accumulation of Ca²⁺ by the sarcoplasmic reticulum is increased, but an elevation of both stored Ca²⁺ and cytosolic Ca²⁺ concentrations facilitates the occurrence of spontaneous sarcoplasmic reticulum Ca²⁺ release during diastole (Bers, 2001). Spontaneous release of Ca²⁺ from the sarcoplasmic reticulum during diastole increases NCX (Bers, 2001). NCX is electrogenic, and its operation in the forward mode (3 Na⁺ in for 1 Ca²⁺ out) creates a transient inward current that can generate a DAD and triggered arrhythmic activity, as well as an aftercontraction (i.e., force production during diastole). Thus, both reverse (phases 1 and 2 of the AP) and forward (phases 2 and 3 and diastole) modes of NCX are enhanced subsequent to increases of late I_Na and intracellular [Na⁺], although at different times during the cardiac cycle. The late I_Na-enhanced oscillatory increases and decreases of the forward and reverse modes of the electrogenic NCX may become electrically destabilizing and cause formation of DADs.

Elevation of the intracellular Ca²⁺ concentration secondary to an enhanced late I_Na in ventricular myocytes impairs diastolic relaxation, that is, slows active relaxation of the ventricle (a negative lusitropic effect). Maintenance of contractile tension during a prolonged period of active ventricular relaxation increases myocardial oxygen consumption and reduces coronary blood flow during left-ventricular (LV) isovolumic relaxation, when flow is normally at its peak. A reduction of diastolic coronary flow in a region of the myocardium with already reduced flow reserve only worsens the tenuous balance between oxygen supply and demand (see Figure 1 and Belardinelli et al., 2006). An increase of late I_Na exacerbates the situation by facilitating AP prolongation, EADs and DADs. These acute electrical events are associated with increases in the cytosolic Ca²⁺ concentration, force production and delayed or reduced diastolic relaxation that further increase oxygen demand while concurrently reducing diastolic blood flow and myocardial oxygen supply. Thus, increases of late I_Na and intracellular Ca²⁺ worsen LV function. Furthermore, although the role of late I_Na in chronic calcium overloading is unknown, late I_Na is reported to be increased in myocytes from failing hearts and from post-infarction remodeled myocardium (Huang et al., 2001; Undrovinas et al., 2006). Heart failure appears to be a condition of chronic Ca²⁺ overloading that is associated with down-regulation of I_Ca,L, increased NCX, prolongation of APD, decreased contractility and a proarrhythmic state (Bers, 2001).

Figure 1.

Block of late I_Na interrupts a deleterious positive feedback cycle during ischaemia by reducing [Na+]i-dependent calcium overload.

Regional differences in the magnitude of late I_Na in ventricular myocardium have been reported, and augmentation of these differences is proarrhythmic (Antzelevitch et al., 1999; Antzelevitch and Belardinelli, 2006). In ischaemic and structurally abnormal hearts, it is doubtful that increases of late I_Na and cellular Ca²⁺ overloading occur simultaneously and equally throughout the ventricle. Consequently, cell-to-cell and regional dispersion of increases of late I_Na and Na⁺-induced Ca²⁺ overloading during ischaemia may lead to focal regions of EAD and DAD formation that increase the dispersion of ventricular repolarization. These events alter the conduction and chronology of ventricular electrical (and mechanical) activity, facilitate triggered and re-entrant arrhythmic activity, prolong the time needed for relaxation of systolic contraction and reduce contractile efficiency and diastolic blood flow (oxygenation) in the myocardium.

Much evidence suggests that late I_Na is enhanced in ischaemia, although quantization of late I_Na in an intact, blood-perfused heart during ischaemia cannot be done. Myocardial-cell Na⁺ content increases during ischaemia (Bers, 2001). Myocardial-cell Ca²⁺ overload and dysfunction caused by ischaemia and reperfusion can be prevented by blockade of membrane Na⁺ channels or cardiac-specific ablation of NCX, thus implicating increased Na⁺ entry into cells as a proximate cause of Ca²⁺ overload (Belardinelli et al., 2006). Hypoxia is known to impair inactivation of Na⁺ channels and increase late I_Na (Saint, 2006). Tissue hypoxia and reperfusion of ischaemic myocardium are reported to generate metabolites (e.g., palmitoyl-L-carnitine and lysophosphatidylcholine) and reactive oxygen/nitrogen species (e.g., hydrogen peroxide and nitric oxide) that increase late I_Na in ventricular myocytes.

The mechanism of action of ranolazine to reduce electrical and mechanical dysfunction of the ischaemic myocardium

Ranolazine, at concentrations within its therapeutic range (⩽10 μM), significantly reduces late I_Na and reverses or prevents the consequences of increase of late I_Na (Belardinelli et al., 2006). Ranolazine reduces late I_Na in mouse myocytes in which late I_Na is enhanced due to expression of the long-QT syndrome 3 mutant sodium channel, ΔKPQ (Fredj et al., 2006), in ventricular myocytes isolated from normal dogs and from dogs with heart failure (Undrovinas et al., 2006), in guinea pig isolated ventricular myocytes exposed to anemone toxin-II or hydrogen peroxide (H₂O₂) (Song et al., 2004, 2006) and in HEK293 cells expressing human Na_V1.5 channels (Rajamani et al., 2007). The range of values of IC₅₀ (potency) is 5.9–15 μM for ranolazine to reduce late I_Na in various preparations. Ranolazine (1–10 μM) reduced APD and temporal dispersion of APD, and abolished EADs of guinea pig isolated ventricular myocytes in the presence of either anemone toxin-II or H₂O₂, both of which increase late I_Na (Song et al., 2004, 2006), thus supporting the concept that reduction of an increased late I_Na is anti-arrhythmic. Ranolazine (5–20 μM) reversibly and significantly shortened the APD of myocytes isolated from dogs with heart failure, in which late I_Na is increased, and reduced variability of APD (Undrovinas et al., 2006). Ranolazine decreased cellular Ca²⁺ overload caused by anemone toxin-II or ischaemia/reperfusion in rat isolated hearts, and it enhanced recovery of post-ischaemic LV function, coronary flow and coronary vascular conductance (Fraser et al., 2006). As expected, ranolazine had no effect on intracellular diastolic Ca²⁺ concentration, Ca²⁺ transient amplitude, LV function, coronary flow or coronary vascular conductance during aerobic conditions (Fraser et al., 2006). Ranolazine-mediated block of late I_Na was significantly reduced by mutation of a single amino-acid residue (F1760A) in the putative local anaesthetic binding site of the cardiac Na⁺ channel Na_V1.5 (Fredj et al., 2006). Thus, a putative site of ‘binding' of ranolazine in the cardiac Na⁺ channel has been identified. In sum, the above findings strongly suggest that ranolazine alters the pathophysiology associated with Na⁺-dependent Ca⁺² overloading in myocardium by inhibiting late I_Na.

The anti-ischaemic effects of ranolazine occur in the absence of haemodynamic changes. Although ranolazine has been reported to antagonize α₁- and β-adrenergic antagonist activity in radioligand binding and in vitro tissue assays (Letienne et al., 2001), at therapeutic concentrations (1–10 μM), it does not cause bradycardia or hypotension in animals (Wang et al., 1999) or in resting and exercising humans (Pepine and Wolff, 1999; Chaitman et al., 2004a, 2004b; Rousseau et al., 2005). In dogs that are awake, ranolazine at steady-state concentrations ⩽18 μM did not cause significant slowing of heart rate or lowering of arterial blood pressure and did not affect coronary blood flow, coronary vascular resistance or LV contractility (CV Therapeutics, 2003). These findings suggest that effects of ranolazine in vivo on haemodynamic parameters that are regulated by α₁- or β₁-adrenergic-mediated activity are not significant in normal dogs with intact autonomic nervous control mechanisms. Furthermore, ranolazine is cardioprotective in ischaemic protocols, wherein hearts are both electrically paced and perfused at constant rates, and in models using cytotoxic ischaemic metabolites (Belardinelli et al., 2006), wherein selective β-adrenergic-receptor-blocking drugs are not cardioprotective. Thus, the data available to date do not indicate any contribution of the sympatholytic action of ranolazine to its anti-ischaemic and anti-anginal effects.

Inhibitions by ranolazine of ion currents other than late I_Na do not appear to be responsible for the anti-ischaemic and anti-anginal effects of the drug. Inhibition of I_Kr is likely responsible for the prolongation of the QT interval by ranolazine, but blockade of I_Kr has not been shown to reduce ischaemic damage or angina. With other drugs, I_Kr block has been associated with proarrhythmic activity, whereas ranolazine reduces both atrial and ventricular arrhythmic activity in humans (Scirica et al. 2007). The significance of a reported inhibition by ranolazine of late I_Ca (Antzelevitch et al., 2004) is unknown. Finally, ranolazine (1–10 μM) does not increase the duration of the QRS interval (i.e., does not slow the conduction of electrical activity), a finding consistent with its low potency to reduce peak I_Na in the ventricle. The potencies (IC₅₀ values) of ranolazine to inhibit peak I_Na in ventricular myocytes isolated from normal dogs and from dogs with heart failure were 294 and 244 μM, respectively (stimulation rate, 0.1 Hz; holding potential, −140 mV) (Undrovinas et al., 2006). In the same study, the IC₅₀ for ranolazine to inhibit late I_Na was 6.5 μM. These findings are consistent with the results of a large clinical outcome trial indicating that ranolazine does not increase mortality and reduces arrhythmic activity (Morrow et al., 2007; Scirica et al., 2007).

Results of a recent ‘proof-of-concept' study of patients with a long-QT syndrome 3 caused by a KPQ deletion in Na_V1.5 revealed that ranolazine (1.8–4 μM) significantly shortened the QTc interval and improved measures of the rate of diastolic relaxation of the left ventricle (data on file at CV Therapeutics Inc., Palo Alto, USA). In a different study of patients with a previous transmural myocardial infarction and with ejection fractions ranging from 13–55%, ranolazine improved diastolic function of the non-infarcted myocardium (Hayashida et al., 1994). These results are consistent with the hypothesis that enhanced late I_Na and its phenotypic manifestations are normalized by therapeutic concentrations of ranolazine in humans, and they provide compelling evidence that ranolazine inhibits late I_Na of the human heart in vivo.

What are the potential therapeutic applications for an inhibitor of late I_Na?

Both preclinical experience with late I_Na inhibitors and the clinical experience with ranolazine indicate that inhibition of late I_Na improves function of the ischaemic myocardium. Cardiac angina, a symptom of myocardial ischaemia, is relieved by ranolazine (Chaitman et al., 2004a, 2004b; Rousseau et al., 2005). The anti-ischaemic effect of ranolazine, demonstrated in humans by an increased time to 1-mm ST-segment depression during an exercise tolerance test (Pepine and Wolff, 1999; Rousseau et al., 2005), and improved diastolic function (Hayashida et al., 1994) are the likely explanations for the action of ranolazine to alleviate angina in patients with ischaemic heart disease. The observation that late I_Na is increased in myocytes from failing hearts of dog suggests that late I_Na may be increased in heart failure, and thus heart failure may yet prove to be another indication for late I_Na inhibitors (Maltsev et al., 2007). Inhibition of late I_Na reduces the prolonged ventricular APD (i.e., the QT interval), dispersion of repolarization, arrhythmic activity, Ca²⁺ loading of myocardial cells and diastolic dysfunction in heart failure (Undrovinas et al., 2006). Diastolic heart function is improved in ΔKPQ long-QT syndrome 3 patients (as noted above), indicating that enhanced late I_Na may contribute to, and reduction of late I_Na may alleviate, diastolic dysfunction in some patients. Other potential therapeutic applications of late I_Na inhibitors are neuronal and muscular disorders in which mutations in various non-cardiac Na⁺-channel isoforms (e.g., Na_V1.7 and Na_V1.4) are known to underlie the pathological phenotype, as in neuropathic pain, epilepsy, sudden infant death syndrome and muscular paralysis. The past 10 years have seen remarkable advancements in the identification of pathological effects of enhanced late I_Na, causal mechanisms of Na⁺ channelopathies and disease states in which late I_Na is enhanced. We speculate that increases of late I_Na will be found to contribute to impairment of the functional status of excitable cells in many other situations. This is not unexpected, given that excessive late I_Na may disrupt the homeostasis of Na⁺- and Ca²⁺-ion concentrations that maintain normal function of myocardial, neural, muscular and secretory cells. Current evidence of a role for reactive oxygen species to increase late I_Na (Song et al., 2006) suggests that enhancement of late I_Na may be a common phenomenon in processes such as ischaemia and aging.

What are the challenges in developing an inhibitor of late I_Na?

Understanding of the roles of late I_Na in physiology and disease is still unfolding, and development of inhibitors of this current poses several challenges. Ion-channel-blocking drugs used for suppression of cardiac arrhythmias are often unselective for their given ion-channel target (e.g., amiodarone). Na⁺-channel antagonists at present in therapeutic use (e.g., class I anti-arrhythmic agents), unlike many drugs (e.g., ligands of G-protein-coupled receptors), typically bind to their ion-channel targets with micromolar rather than nanomolar affinity. The goals of high potency, efficacy and selectivity that normally drive drug development may or may not be appropriate to find useful, safe blockers of late I_Na. In the final analysis, inhibition of ion-channel function should perhaps be rather subtle and readily reversible, and the combination of ion-channel effects of a drug may determine its success or failure.

When inhibiting cardiac late I_Na, it is preferable to avoid inhibition of peak I_Na, in the heart as well as in other excitable tissues. Drugs (e.g., flecainide, encainide) that inhibit peak I_Na in the ventricle were associated with increased mortality in the Cardiac Arrhythmia Suppression Trial (Echt et al., 1991). This finding suggests that late I_Na-blocking drugs must not alter Na⁺ channel opening and conductance during the AP upstroke (phase 0), to avoid slowing of impulse conduction and reduction of the wavelength of a re-entrant circuit. Reduction of peak I_Na by ranolazine increases with increased frequency of stimulation and decreased membrane potential and appears to be tissue (atria vs ventricle; Burashnikov et al., 2007), isoform (i.e., Na_V1.1–1.9) and mutation (e.g., late I_Nas of various Na_V1.5-mutated channels are differentially responsive to ranolazine) dependent.

Inhibition of the HERG current I_Kr is problematic, and many drugs, including Na⁺-channel blockers, reduce this current. Reduction of I_Kr is frequently associated with an increased incidence of torsades de pointes, a potentially fatal ventricular tachycardia. Reduction of late I_Na offsets the reduction of repolarization reserve and the increase of APD and antagonizes the induction of EADs caused by I_Kr blockers. Ranolazine reduces I_Kr and prolongs, although modestly, the APD, but its effect is not rate (use) dependent, and it does not have proarrhythmic activity. The twofold greater potency of ranolazine to inhibit late I_Na than to inhibit I_Kr may contribute significantly to the anti-arrhythmic effects and absence of proarrhythmic effects of the drug. In post-myocardial infarction patients without ST-segment elevation but with a QTc interval greater than 450 ms, ranolazine significantly (P<0.001) reduced the incidence of ventricular tachycardia (⩾8 beats at ⩾100 bpm, lasting less than 30 s) during 7 days of Holter monitoring by 47% compared with placebo (Scirica et al., 2007).

Lastly, how much reduction of cardiac late I_Na is safe? At therapeutic concentrations, ranolazine reduces late I_Na by about 50%. Would more be better? The answer may depend on the role of late I_Na in normal cardiac function. For example, Purkinje fibres have a greater late I_Na than other cells in the heart, and the duration of the Purkinje fibre AP is significantly shortened upon late I_Na block. Excessive shortening of the Purkinje fibre APD may increase the risk of re-entry. Similar concerns with blockade of physiological roles of late I_Na in other tissues counsel caution in the development of potent drugs for incompletely understood targets such as late I_Na. Nevertheless, the observation that ranolazine reduces ischaemia/angina, diastolic dysfunction and arrhythmic activity in a safe manner suggests that reduction of late I_Na is therapeutically beneficial. However, scientific understanding of late I_Na is in its infancy, and we remind ourselves that ‘All scientific work is liable to be upset or modified by advancing knowledge' (Hill, 1965).

Abbreviations

AP

action potential

APD

action potential duration

DAD

delayed afterdepolarization

EAD

early afterdepolarization

I_Na

sodium current

LV

left ventricle

NCX

Na⁺–Ca²⁺ exchange

Conflict of interest

John C Shryock and Luiz Belardinelli are employees of CV Therapeutics Inc., which markets Ranexa (ranolazine).