Antiarrhythmic activity of amiloride: mechanisms.

We have previously shown that amiloride suppresses the induction of sustained ventricular tachycardia both in dogs late following myocardial infarction and in patients. In those studies the only electrophysiologic correlate of amiloride's antiarrhythmic activity observed was prolongation of ventricular effective refractory period at the zone bordering the infarct. The purpose of this study was to assess the pharmacologic effects of amiloride associated with antiarrhythmic efficacy. However, amiloride has multiple pharmacologic effects, including inhibition of the slow inward calcium current (ICa), inhibition of the sodium-calcium and sodium-hydronium ion exchangers, acidification of intracellular pH resulting in partial inhibition of the inwardly rectifying potassium current (IK1), and increase in serum potassium and magnesium. The approach used in this study was to use selective pharmacologic probes to produce the known components of amiloride's pharmacologic effects. The selective agents consisted of verapamil (partial blockade of ICa), 3',4'-dichlorobenzamil (partial inhibition of the Na-Ca exchanger), 5-(N-ethyl-N-isopropyl) amiloride (partial inhibition of the Na-H exchanger), the combination of these congeners, KCl infusions to increase serum potassium, MgSO4 infusions to increase serum magnesium, the combination of KCl and MgSO4 infusions, barium (partial block of IK1), ryanodine (partial blockade of sarcoplasmic reticulum calcium release), and placebo. In this study only barium produced antiarrhythmic and electrophysiologic effects similar to those of amiloride. However, amiloride prolongs border zone refractoriness selectively, whereas barium prolongs refractoriness diffusely throughout the myocardium. Blockade of ICa by verapamil, increases in serum magnesium and potassium alone or in combination, and partial blockade of sarcoplasmic reticulum by ryanodine were not antiarrhythmic in this model. Monotherapies that produced partial blockade of the Na-Ca and Na-H exchangers separately did not produce antiarrhythmic activity. However, the combination of these amiloride congeners reproduced the antiarrhythmic activity of amiloride but did not prolong border zone refractoriness. From these studies we conclude that the antiarrhythmic activity of amiloride relates to (a) selective blockade of IK1 in the border zone and/or (b) combined inhibition of sodium-calcium and sodium-hydronium ion exchangers.