Effects of Ionic Remodeling on Cardiac Antiarrhythmic Drug Actions

In this issue of the Journal of Cardiovascular Pharmacology, Sato et al. (1) report on the impact of atrial tachycardia-induced remodeling on the atrial electrophysiologic actions of pilsicainide. Pilsicainide is a class IC antiarrhythmic agent that is widely used in Japan in the treatment of atrial fibrillation (2). Tachycardia-induced remodeling significantly attenuated the refractoriness-prolonging effects of the drug. For example, pilsicainide increased atrial effective refractory period (ERP) by over 20% at a cycle length of 200 ms under control conditions, but by less than 10% after 2 weeks of atrial pacing at 400 bpm. In contrast, the conduction-slowing effects of pilsicainide were unaltered. These observations point to a specific interaction between atrial tachycardia-induced remodeling and the ERP-prolonging effects of the drug. It has long been known that as atrial fibrillation becomes chronic, it becomes increasingly resistant to conversion by cardiac antiarrhythmic drugs (3). This observation has been confirmed repeatedly in the clinical literature (4). With the recognition of the importance of atrial tachycardia-induced remodeling (5), it has become possible to evaluate potential underlying mechanisms in clinically relevant experimental models. Although limited, the available information supports the notion that electrophysiologic remodeling significantly modulates the response of atrial fibrillation to antiarrhythmic drugs. Garratt et al. (6) have reported that goats subjected to electrically maintained atrial fibrillation require larger doses of the class I agent cibenzoline to terminate arrhythmia as atrial fibrillation duration is increased from 24 h to longer periods. We have observed marked differences in the effects of the I Kr -blocking class III agent dofetilide in dogs with sustained atrial fibrillation based on an atrial tachycardia–remodeled substrate versus a substrate resulting from congestive heart failure (7). Dofetilide was ineffective in terminating atrial fibrillation in the tachycardia-remodeled dogs, but was highly effective in dogs with atrial fibrillation induced in the setting of congestive heart failure. A primary mechanism for the effects of atrial tachycardia on atrial electrophysiology and vulnerability to atrial fibrillation is ionic remodeling (8). Downregulation of various cardiac ion channels occurs in response to atrial tachycardia (9). Decreases in L-type Ca 2+ current (I Ca ) density appear to be particularly important in remodeling associated with atrial tachycardia, because they reduce atrial action potential duration (APD) and APD adaptation to rate change (9–11). In addition, several outward K + currents are affected. Transient outward current (I to ) is clearly reduced (9,10,12). The sustained outward current corresponding to the ultrarapid delayed rectifier current, I Kur , has been reported to be reduced in patients with atrial fibrillation in one study but not in another (12,13). Inward rectifier current (I Kl ) has been reported to be increased (10,12) or unchanged (9), and acetylcholine-dependent current may be reduced (14). The net result of these ion channel alterations is an action potential that is shorter than normal, with a short plateau and reduced APD rate adaptation. Tachycardia-dependent ionic remodeling can alter electrophysiologic drug actions in a number of ways. If the ionic current against which a drug is targeted is strongly reduced by remodeling, the drug would be expected to have very little effect on the action potential. Conversely, reductions in the target current do not necessarily result in reduced drug action. For example, atrial fibrillation has been reported to reduce I Kur(12). One would therefore expect reduced APD-prolonging effects by I Kur blockers in atrial fibrillation. However, the peak conductance of a current is not the only determinant of its density during the action potential. Furthermore, blocking a current changes the voltage-time trajectory of the action potential, causing changes in the density of other currents in a way that may antagonize the effects of current blockade. Blocking I Kur raises the plateau level. I K , particularly the rapid component I Kr , is quite sensitive to voltage in the plateau range of the atrial action potential. A more positive plateau activates more I K , which can counteract the APD-prolonging effect of I Kur blockade (15). Because the tachycardia-remodeled action potential has a short plateau, little I K is activated during the plateau and there is no significant I K enhancement by I Kur blockade. Thus despite the downregulation of I Kur , I Kur blockers can actually have larger APD-prolonging effects on the remodeled human atrial action potential than on the normal action potential (15). Similar considerations may explain the effects of tachycardia-induced remodeling on the action of I Kr -blocking drugs. Because little I Kr is activated during the short plateau of the tachycardia-remodeled action potential, the APD prolongation produced by I Kr blockade would be reduced. Other pathologies associated with atrial fibrillation cause atrial ionic remodeling different from that caused by atrial tachycardia. Congestive heart failure reduces atrial I Ca to a modest extent (∼30%), while reducing I to (by ∼50%) and the slow component of I K , I Ks (by ∼30%), and increasing the Na + ,Ca 2+ -exchange current (16). In addition, the mechanism of atrial fibrillation appears to be different from that in tachycardia remodeling (7). Whether because reduced outward currents make atrial myocytes more dependent on I Kr for repolarization in congestive heart failure, or because of differences in the arrhythmia mechanism, atrial fibrillation induced in experimental congestive heart failure models is particularly sensitive to termination by dofetilide. In addition to affecting drug action on the atrium, ionic remodeling can also influence the risk of ventricular proarrhythmia associated with drugs used to treat atrial fibrillation. For example, congestive heart failure is an important risk factor for antiarrhythmic drug–induced torsades de pointes (17). The increased risk may be due to ionic remodeling at the ventricular level associated with congestive heart failure. Congestive heart failure decreases I Ks in ventricular myocytes (18,19), making their repolarization potentially more dependent on I Kr and sensitizing them to I Kr blockers. It also causes ionic remodeling in cardiac Purkinje cells, reducing I to and I Kl and slowing I Ca deactivation (20). These changes do not alter APD, perhaps because of a compensating effect of I Kr activated at the more positive plateau voltage, but substantially sensitize congestive heart failure to APD-prolonging effects of I Kr -blocking drugs (20). The study of Sato et al. points out the potentially important effect that ionic remodeling can have on the actions of an antiarrhythmic drug in an atrial fibrillation model. Unfortunately, because of the limitations of their model, they were unable to link their observations to changes in drug efficacy in terminating or preventing atrial fibrillation. Reduced ERP and wavelength prolongation, as seen by Sato et al., would be expected to translate into reduced efficacy for atrial fibrillation termination, but this expectation remains to be verified. It would also be important to know precisely how tachycardia-induced remodeling altered the ERP-prolonging effects of the drug. Drug-induced conduction-slowing effects were unchanged, so potency for Na + -channel blockade is unlikely to have been altered. Changes in drug effects on repolarization could have occurred, but were not measured in this in vivo study. Alternatively, by changing action potential morphology, atrial tachycardia could have reduced the extent of Na + channel blockade during the action potential plateau and have affected drug-induced postrepolarization refractoriness. A detailed understanding of the consequences of ionic remodeling for antiarrhythmic drug action is essential for improving antiarrhythmic drug therapy. Arrhythmias, including atrial fibrillation, most commonly occur in the presence of cardiac disease that causes ionic remodeling. In addition, cardiac arrhythmias can in themselves cause ionic remodeling, with atrial fibrillation being the clearest and best-understood example. At the moment, we know relatively little about the effects of ionic remodeling on antiarrhythmic action. Much of the limited knowledge available has been obtained from studies, like that by Sato et al., in in vivo models, and the precise underlying cellular and ionic mechanisms remain largely unknown. More work clearly needs to be done in this important area in order to optimize the development of new antiarrhythmic drugs and to improve the use of presently available agents.