Profiles of Aprindine, Cibenzoline, Pilsicainide and Pirmenol in The

Home , Aprindine, Indecainide, Pilsicainide

REVIEW ARTICLE Jpn Circ J 1999; 63: 1–12

Proﬁles of Aprindine, Cibenzoline, Pilsicainide and Pirmenol in the Framework of the Sicilian Gambit

Itsuo Kodama, MD*; Satoshi Ogawa, MD**; Hiroshi Inoue, MD†; Hiroshi Kasanuki, MD††; Takao Kato, MD‡; Hideo Mitamura, MD**; Masayasu Hiraoka, MD‡‡; Tsuneaki Sugimoto, MD§; The Guideline Committee for Clinical Use of Antiarrhythmic Drugs in Japan (Working Group of Arrhythmias of the Japanese Society of Electrocardiology)

The Vaughan Williams classification has been used widely by clinicians, cardiologists and researchers engaged in antiarrhythmic drug development and testing in many countries throughout the world since its initial proposal in the early 1970s. However, a major criticism of the Vaughan Williams system arose from the extent to which the categorization of drugs into classes I–IV led to oversimplified views of both shared and divergent actions. The Sicilian Gambit proposed a two-dimensional tabular framework for display of drug actions to solve these problems. From April to December 1996, members of the Guideline Committee met to discuss pharmacologic profiles of 4 antiarrhythmic drugs (aprindine, cibenzoline, pilsicainide, and pirmenol) that were not included in the original spreadsheet but are used widely in clinical practice in Japan. The discussion aimed to fit the drug profiles into the Gambit framework based on all the important literature published to date regarding the actions of the 4 drugs. This report is a summary of that deliberation. (Jpn Circ J 1999; 63: 1–12) Key Words: Antiarrhythmic drugs; Cardiac electrophysiology; The Sicilian Gambit; The Vaughan Williams classification

he Vaughan Williams classification has been used approaches to antiarrhythmic therapy. The second and the widely by clinicians, cardiologists and researchers third Gambit meetings were held at Arden House in T engaged in antiarrhythmic drug development and Harriman, New York in 1993 and 1996. The conclusions of testing in many countries throughout the world since its the second meeting, published as a monograph in 1994,3 initial proposal in the early 1970s.1 This classification is were translated into Japanese by Ogawa in 1995.4 Two admirable in its simplicity with respect to the physiologic Japanese investigators, (Drs S. Ogawa and M. Hiraoka), basis of effects, and it has facilitated education and discus- joined the third meeting. Thus, the Gambit classification sion concerning the beneficial and deleterious actions of process is progressing and expanding. In accord with this drugs. During the past 2 decades, however, our knowledge trend, the Guideline Committee for Clinical Use of of the electrophysiologic mechanisms of arrhythmias has Antiarrhythmic Drugs was organized in Japan in 1996 as a expanded tremendously, and the number of antiarrhythmic Working Group of Arrhythmias of the Japanese Society of agents available for clinical practice has greatly increased. Electrocardiology. Accordingly, updated concepts are becoming progressively A major criticism of the Vaughan Williams system arose difficult to incorporate into the general scheme of the from the extent to which the categorization of drugs into Vaughan Williams classification in the process of ratio- classes I–IV led to oversimplified views of both shared and nally determining the optimal pharmacologic treatment of divergent actions.2,3 Briefly, the problems with the older arrhythmias. In 1991, the Sicilian Gambit was proposed by classification areas follows. the Task Force of the Working Group on Arrhythmias of (1) The classification is hybrid, with classes I and IV the European Society of Cardiology,2 as the outcome of the representing blockade of ion channels, class II, blockade first Gambit meeting in Taormina, Sicily (1990) where of receptors, and class III, a change in a specific electro- basic and clinical investigators attempted a break from physiologic variable (action potential duration). traditional approaches to drug classification. The Sicilian (2) Some drugs, such as amiodarone and bepridil, have Gambit aimed to construct a system where current diverse actions representing several such classes. knowledge and information on arrhythmias could be inte- (3) The classification essentially describes antiarrhyth- grated into a framework of universal pathophysiologic mic drugs as acting via blockade of channels and receptors. The alternate possibility that activation of certain (Received October 15, 1998; accepted November 5, 1998) channels or receptors might be antiarrhythmic is not *Research Institute of Environmental Medicine, Nagoya University, considered. Nagoya, **Keio University, School of Medicine, Tokyo, †Toyama (4) The classification is incomplete; several drugs Medical and Pharmaceutical University, Faculty of Medicine, currently available for the treatment of arrhythmias, such Toyama, ††Tokyo Women’s Medical College, Tokyo, ‡Nippon Medical as cholinergic agonists, digitalis, and adenosine, are not ‡‡ School, Tokyo, Tokyo Medical and Dental University, Medical included. Research Institute, Tokyo, and §Kanto Central Hospital, Tokyo, Japan Mailing address: Satoshi Ogawa, MD, Cardiopulmonary Division, The Sicilian Gambit proposed a 2-dimensional tabular Department of Medicine, Keio University, School of Medicine, 35 framework (spreadsheet) for display of drug actions to Shinanomachi, Shinjuku-Ku, Tokyo, Japan solve these problems.2,3 In the second version of the frame-

Japanese Circulation Journal Vol.63, January 1999 2 KODAMA I et al. work, the molecular targets to be affected by antiarrhyth- and in ventricular myocytes isolated from guinea pig hearts mic drugs (ion channels, receptors, and pumps) are listed using single-sucrose gap or whole-cell voltage-clamp tech- on the left side, with the corresponding clinical effects of niques.13,18,19 In ventricular muscles or myocytes treated the drugs (heart rate, electrocardiographic intervals, with aprindine (2–5μmol/L), application of conditioning ventricular performance, and side-effect potential) appear- depolarization to 0 mV from a holding potential (HP) at ing on the right side of the columns.3 The antiarrhythmic resting level (–80 to –90mV) caused a decrease in Vmax of drugs are listed in vertical rows in order of their predomi- a test action potential elicited 100ms after the clamp pulse. nant molecular targets. This decrease became more marked as clamp pulse dura- The spreadsheet (Gambit framework) has the following tion was prolonged.13,18,19 These authors estimated the advantages. extent of sodium channel block by various antiarrhythmic (1) It does not discard useful information about a drug. drugs, including aprindine, during activated and inactivated Most antiarrhythmic drugs act at several sites, sharing a states.18,19 A percentage decrease in Vmax by the shortest spectrum of action with other agents. Such multifaceted clamp pulse (10ms) from the reference level (tonic block- drug actions can be presented appropriately. ade subtracted value) was defined as an activated channel (2) The listing is flexible. The vertical order of the blockade (ACB), while an additional percentage decrease antiarrhythmic drugs may be changed to emphasize in Vmax occurring when clamp pulse duration was pro- different drug clusters. For example, drugs that act atβ- longed from 10 to 500 ms was defined as an inactivated receptors or those that block potassium channels can be channel blockade (ICB). Blockade of sodium channels placed together. If important new molecular targets are during slow inactivation can be neglected in this definition, found in future basic research, they can be added as new because Vmax of control muscles or myocytes decreased columns. As new effective antiarrhythmic drugs are significantly only when the conditioning 0mV clamp was developed, they would be added by expanding rows. maintained for longer than 500 ms.18,19 ACB and ICB (3) The tabular approach is suitable for programming induced by 2–5μmol/L aprindine were estimated to be diverse drug information into computer software for 1.3–5.4% and 14.6–49.6%, respectively, giving rise to a availability from single channel, single cell to bedside. ratio (ICB/ACB) of 7.4–11.2.13,18,19 In single-cell experi- From April to December 1996, members of the Guide- ments, ICB/ACB ratios for aprindine, mexiletine, and lido- line Committee met 4 times to discuss pharmacologic caine were all larger than 5.0, while those for quinidine and profiles of 4 antiarrhythmic drugs (aprindine, cibenzoline, disopyramide were much less (1.5–1.6).18,19 The authors pilsicainide, and pirmenol) that were not included in the therefore concluded that aprindine appears to block the original spreadsheet but used widely in clinical practice in sodium channel mainly during the inactivated state. Japan. The discussion aimed to fit the drug profiles into the In ventricular muscles or cells treated with aprindine, Gambit framework based on all the important literature additional application of lidocaine or mexiletine has been published to date regarding the actions of the 4 drugs in demonstrated to cause an increase in tonic blockade, but a experimental as well as clinical studies. This report is a decrease of use-dependent blockade, often resulting in a net summary of that deliberation. increase in Vmax at certain stimulation rates.10,13,14,20 In the presence of both aprindine and lidocaine (or mexiletine), Vmax recovered from use-dependent blockade according to Aprindine a dual exponential function, where the short- and long-time Action Potentials constants corresponded to the values for single treatments Aprindine causes a decrease in the maximum upstroke with each drug.10,13,14 These findings suggest that aprindine, velocity (Vmax) of action potentials without affecting the lidocaine, and mexiletine may block the sodium channel by resting membrane potential. This action occurs in cardiac binding with different kinetics to a common receptor site, tissues and cells whose excitation depends on activation of leading to competitive displacement of one another at high the fast sodium channel. Verdonck et al5 and Carmeliet and concentrations.13,21 In support of this hypothesis, Lee et al Verdonck6 demonstrated concentration-dependent decreases have demonstrated that additional application of lidocaine in Vmax in cow and dog Purkinje fibers upon treatment with attenuated the frequency-dependent QRS prolongation aprindine (half-maximal inhibitory concentration on IC50, induced by aprindine in ECGs obtained from patients with ~2μmol/L), in association with marked shortening of the implanted pacemakers.22 action potential duration (APD). They also showed a As for effects on cardiac action potentials that depend on concentration-dependent, aprindine-induced decrease in activation of calcium channels, Carmeliet and Verdonck,6 Vmax in atrial and ventricular muscles from guinea pigs, and Elharrar et al23 have reported that aprindine (3–10 cats and cows, but this inhibition was less prominent than μmol/L) did not affect slow action potentials of K+-depo- in Purkinje fibers; APD of atrial and ventricular muscles larized ventricular muscles or Purkinje fibers. These groups were slightly shortened or unaffected.5,6 Qualitatively therefore suggested that aprindine may lack slow channel similar results were obtained by other investigators in blocking properties. However, in experiments using small cardiac tissues from guinea pigs or rabbits.7–11 sinoatrial (SA) node tissue specimens isolated from rabbit Aprindine-induced Vmax inhibition is enhanced in a hearts, Kotake et al found that aprindine (1–4μmol/L) frequency- and use-dependent manner.5,7,9,10,12–16 The onset caused significant concentration-dependent decreases in and offset kinetics of use-dependent Vmax inhibition are Vmax, in the amplitude of action potentials, and in their intermediate.9,10,12–15 The recovery time constants (τR) of spontaneous firing rate.24 Aprindine (1μmol/L) also was Vmax from use-dependent blockade reported in ventricular shown to reduce Vmax and amplitude of action potentials as muscle or cells ranged from 3 to 8s.9,10,13,14,17 Comparable well as spontaneous firing rate in small atrioventricular values were reported for atrial muscle.15 (AV) node tissue specimens isolated from rabbit hearts.25 Kodama et al examined the voltage- and time-dependence of Vmax inhibition by aprindine in papillary muscles

Japanese Circulation Journal Vol.63, January 1999 Aprindine, Cibenzoline, Pilsicainide and Pirmenol in the Sicilian Gambit 3

Ionic Currents premature contractions (VPC). In a randomized study by Sato et al examined the effects of aprindine on the Pouleur et al in which aprindine (80mg/day), disopyramide inward sodium current (iNa) in single guinea-pig ventricular (400mg/day), or a placebo was given to patients with ven- myocytes using whole-cell voltage-clamp techniques.26 tricular arrhythmias, aprindine was significantly more Aprindine (3μmol/L) caused both tonic and phasic (use- effective (p<0.05) than disopyramide or placebo.32 In a dependent) inhibition of iNa. The IC50 of aprindine for tonic double-blind comparison of the effect of aprindine (60 blockade was 37.7μmol/L at an HP of –140mV, and 0.74 mg/day) and disopyramide (300mg/day) on VPC and SVPC, μmol/L at an HP of –100 mV. Aprindine (3μmol/L) aprindine proved to be significantly more effective.33 (2) shifted the steady-state inactivation curve of iNa in the Paroxysmal supraventricular tachycardia (PSVT).34–37 hyperpolarizing direction by 11.4mV. Phasic block of iNa Intravenous administration of aprindine (100mg; with a by aprindine (3μmol/L) in response to 2-Hz trains of depo- maximum dosage, 2mg/kg) was highly effective in termi- larization to –20mV from an HP at –140mV was enhanced nating AV reciprocal tachycardia and AV nodal reentrant moderately as the duration of each depolarization was tachycardia, with a success rate of 70–100%. In contrast, prolonged from 1.5 to 200 ms. TheτR from aprindine- aprindine failed to convert paroxysmal atrial fibrillation to induced phasic blockade was 5.0 s These authors also sinus rhythm in 15 of 22 cases. Oral aprindine prevented showed that aprindine-induced phasic blockade of iNa induction of PSVT by programmed stimulation in only 2 of persisted after pharmacologic removal of rapid inactiva- 21 cases. (3) Mortality reduction. Two double-blind, tion, finally concluding that aprindine blocks iNa at least randomized, placebo-controlled trials were conducted to partially by binding to channels in the activated state. determine whether aprindine influences mortality in high- Discordance of this conclusion with the Vmax measurement risk patients after acute myocardial infarction. According studies described before11,13,19 could result from different to Gottlieb et al, aprindine reduced mortality at 1 month, experimental conditions such as low temperature and low but had no effect on overall mortality or sudden death rates external sodium concentrations in the iNa experiments. at 1 year.38 The Ghen-Rotterdam aprindine study reported Aprindine has been shown to inhibit the slow inward that aprindine tended to be effective in reducing mortal- current (isi) in multicellular tissue preparations of rabbit SA ity.39,40 and AV nodes.24,25 In SA node cells, aprindine (4μmol/L) Clinical Electrophysiology and Effects on Cardiac reduced isi by approximately 20%,24 whereas the IC50 of Function Aprindine prolonged the AH and HV intervals aprindine for isi inhibition in AV node cells was approxi- and the effective refractory periods of the atrium, ventricle mately 10μmol/L.25 Aprindine-induced isi inhibition in AV and AV node,35 and suppressed antegrade and retrograde node cells was enhanced in voltage- and use-dependent conduction in Kent’s bundle while prolonging its effective manners, suggesting a higher affinity of this compound for refractory period.35 On the ECG, aprindine increased PQ, Ca2+ channels in the inactivated than the resting state.25 In QRS, and QT intervals. The PP interval did not change, but whole-cell voltage-clamp experiments on guinea pig atrial occurrence of sinus arrest has been reported in patients cells, Shibasaki et al demonstrated that 5μmol/L aprindine with sick sinus syndrome. reduced inward calcium current (iCa) by 35%.27 Oral administration of aprindine did not decrease In double-microelectrode voltage-clamp experiments on contractility,41 but aprindine given intravenously produced rabbit SA and AV node preparations, aprindine (1–4μmol/ a moderate, dose-related decrease in peak left ventricular L) caused a moderate reduction of the delayed rectifier K+ (LV) dP/dt and Vmax.42 current (iK) and the hyperpolarization-activated inward Clinical Pharmacology and Side-Effects While in the current (the pacemaker current, if).24,25 Aprindine (5μmol/ United States and Europe the appropriate daily dose of L) also reduced iK in single guinea pig atrial cells without aprindine is in the range of 100–150mg with an effective affecting the inward rectifier K+ current (iK1).27 However, serum concentration of 1–2μg/ml,43 in Japan the dose is relatively little is known and much remains to be investi- only 40–60mg and the serum concentration is only 0.13– gated as to the effects of aprindine on these ionic currents. 0.68μg/ml.44 The difference may be related to a racial difference in sensitivity to aprindine. Aprindine is metabo- Receptors and Pumps lized exclusively by the liver, mainly by the enzyme Aprindine does not possess anticholinergic proper- CYP206,45 and thus may cause a drug interaction with dilti- ties.25,28,29 No in vivo or in vitro experimental data suggest azem by competing for CYP206. The pharmacokinetics of direct effects of aprindine onα- andβ-adrenergic receptors, aprindine are nonlinear.44 adenosine (A1) receptors, or sarcolemmal Na/K pumps.25,28,29 The major side-effects of aprindine include agranulocy- Aprindine at high concentrations has been shown to inter- tosis and cholestatic jaundice, which are thought to be idio- fere with calmodulin activation of such cytosolic enzymes syncratic reactions. Agranulocytosis was reported rarely in as phosphodiesterase and Ca2+-ATPase.30,31 The pharmaco- the United States (3 of 400–500 patients). The incidence of logic significance of this calmodulin inhibition is still agranulocytosis is estimated as approximately 2 occur- unsettled. rences per 1000 patient-years based on data from Europe.46 Other reported side-effects included neuralgia and seizures, Clinical Effects which were related to the dose and serum concentration of Since its development in Belgium in 1973, aprindine has aprindine. been used clinically in 7 European countries. However, clinical investigation of the drug was halted in the United Gambit Profile States due to such side-effects as agranulocytosis. In Japan, The pharmacologic profile of aprindine can be summa- lower doses of aprindine have been recognized as effective rized as follows (Table1). First, aprindine has a potent use- and safe for supraventricular and ventricular arrhythmias. dependent inhibitory action on the cardiac sodium channel. Effects on the Various Tachyarrhythmias (1) Supra- From the onset and offset kinetics of use-dependent block- ventricular premature contractions (SVPC) and ventricular ade, aprindine belongs among the intermediate-kinetic

Japanese Circulation Journal Vol.63, January 1999 4 KODAMA I et al.

Table 1 Gambit Profiles of Aprindine, Cibenzoline, Pilsicainide and Pirmenol Channels Receptors Pumps Clinical Effects ECG Effects Drug Na Na/K CaKIf αβM2 A1 LV function Sinus Rate Extra-cardiac PR QRS JT Fast Med Slow ATPase Aprindine I Cilbenzoline A Pilsicainide A Pirmenol A

For the section on channels, receptors, and pumps, the actions of drugs on sodium (Na), calcium (Ca), potassium (K) and If channels are indicated. Sodium channel blockers are subdivided into 3 groups of actions characterized by fast, medium, and slow time constants for recovery from block. Blockade in the inactivated (I) or activated (A) state is indicated. Drug interaction with receptors α( ,β , muscarinic subtype 2 [M2] and adenosine [A1]) and drug effects on the sodium-potassium pump (Na/K ATPase) are indicated. Circles indicate blocking actions. The intensity of the action is indicated ( low, moderate, high). The absence of a symbol indicates lack of effect. *The clinical and ECG arrows indicate direction (increase, decrease or no substantial change). The effects are those that may be seen at therapeutic plasma levels. drugs, and it appears to block the sodium channel mainly rabbits or dogs were moderately prolonged, while the APD during the inactivated state. Second, aprindine may have of dog Purkinje fibers was shortened.47–50 APD in rabbit modest inhibitory actions on Ca2+ and K+ channels and the Purkinje fibers and guinea pig papillary muscles was unaf- if channel, but the effects of aprindine on these ionic chan- fected by cibenzoline.47,48,51 This discrepancy could be a nels have not yet been thoroughly investigated. Finally, result of differences in ionic currents responsible for repo- aprindine has no significant effects on α-, β-, or A1-recep- larization of the action potential in different cardiac tissues tors, or on Na/K pumps. and animal species.52 Only limited information is available regarding the effects of cibenzoline on cardiac tissues whose excitation depends on activation of slow calcium Cibenzoline channels. In SA node preparations isolated from rabbits, Action Potentials cibenzoline (1–30μmol/L) was shown to cause concentra- In atrial and ventricular muscles as well as in Purkinje tion-dependent decreases in Vmax, action potential ampli- fibers from rabbits and dogs, Millar and Vaughan tude, and rate of spontaneous excitation.47,48,53 Williams47,48 and Dangman49 demonstrated that cibenzoline (1–20μmol/L) decreased Vmax without affecting the Ionic Currents resting membrane potential. Similar concentration-depen- An inhibitory effect of cibenzoline on the iNa was dent Vmax inhibition by cibenzoline was observed in guinea demonstrated in frog atrial muscles using the double pig ventricular muscle.50,51 The IC50 of cibenzoline for Vmax sucrose-gap voltage-clamp method,54 and in single guinea inhibition in those studies was 2–10μmol/L. Cibenzoline- pig ventricular myocytes using the whole-cell voltage- induced Vmax inhibition became more pronounced in a clamp method.55 The IC50 of cibenzoline for iNa in guinea frequency- and use-dependent manner.50,51 The onset and pig ventricular myocytes was estimated as 7.8μmol/L.55 offset kinetics of use-dependent Vmax inhibition with ciben- The effects of cibenzoline on iNa kinetics and the use- and zoline are slow.50,51 TheτR of Vmax from use-dependent voltage-dependence of iNa inhibition by cibenzoline have block in guinea pig papillary muscles was assigned a value not been investigated. of 26.2s by Niwa et al.51 In whole-cell voltage-clamp experiments using guinea Niwa et al also examined the voltage- and time-depen- pig ventricular myocytes, cibenzoline (1–30μmol/L) was dence of Vmax inhibition by cibenzoline in single guinea shown to inhibit iCa in a concentration-dependent man- pig ventricular cells using whole-cell clamp techniques.51 In ner.55–57 The IC50 of cibenzoline for iCa as estimated by myocytes treated with cibenzoline (10μmol/L), application Holk and Osterrieder56 and by Sato et al55 was approxi- of the shortest (10ms) single conditioning depolarization to mately 14μmol/L, but a higher value (30μmol/L) was 0mV from an HP at –80mV caused substantial Vmax reduc- reported by Matsuoka et al.57 This iCa inhibition was tion of a test action potential elicited 100ms after a 10-ms enhanced at higher depolarization frequencies and at more depolarization. Vmax inhibition was enhanced modestly by positive HP.56 Inhibition of iCa by cibenzoline also was further prolongation of clamp pulse duration. These authors demonstrated in frog atrial muscles using double sucrose evaluated the approximate extent of sodium channel block gap voltage-clamp methods54 and in rabbit SA node prepa- during activated and inactivated states using the definition rations using a double-microelectrode voltage-clamp of Kodama et al18,19 with average values for ACB and ICB method.53 by 10μmol/L cibenzoline being 11.0% and 16.4%, respec- Cibenzoline inhibits voltage-dependent K+ channel cur- tively; the resulting ICB/ACB ratio of 1.5 is comparable to rents. Sato et al demonstrated concentration-dependent that for quinidine (1.6) and disopyramide (1.3).51 In guinea inhibition of both the iK and the iK1 in guinea pig ventricu- pig ventricular myocytes treated with 3μmol/L cibenzo- lar myocytes with respective IC50 at 23.0μmol/L and 33.7 line, use-dependent Vmax reduction during a train of 0mV μmol/L.55 In rabbit SA node cells, iK also was shown to be depolarization was enhanced only slightly by a prolonga- decreased by about 30% after application of 30μmol/L tion of each depolarization from 10 to 200 ms.51 These cibenzoline.53 A recent study by Wang et al indicated that observations suggest that cibenzoline may block the both a rapidly activating, E-4031-sensitive component of iK sodium channel mainly during the activated state. (iKr) and a slowly activating, E-4031-insensitive compo- Reports of cibenzoline effects at 1–10μmol/L on APD nent of iK (iKs) are inhibited by cibenzoline.58 are conflicting. APD of atrial and ventricular muscles from Cibenzoline inhibits ligand-gated K+ channel currents as

Japanese Circulation Journal Vol.63, January 1999 Aprindine, Cibenzoline, Pilsicainide and Pirmenol in the Sicilian Gambit 5 well. In whole-cell voltage-clamp experiments in single accounted for the increase in the QT interval.64 Cibenzoline guinea pig ventricular cells by Wu et al59 and Sato et al60 has inconsistent effects on sinus rate in patients;63 when the time-independent outward current induced by 2,4-dini- given orally, cibenzoline does not affect sinus rate signifi- trophenol or by diazoxide was inhibited almost completely cantly,65 but this drug increases sinus rate slightly when by 5–10μmol/L cibenzoline. The authors therefore con- given intravenously.66 The direct inhibitory effect of ciben- cluded that cibenzoline caused potent inhibition of the zoline on the SA node possibly counteracts any clinically ATP-sensitive K+ channel current (iK,ATP) at concentrations significant reflex tachycardia in response to the negative close to the therapeutic plasma level (1–5μmol/L). In inotropic effect of this compound, minimizing changes in support of this impression, Sato et al demonstrated that heart rate.63 cibenzoline (10μmol/L) significantly reversed 2,4-dinitro- Cibenzoline exerts a negative inotropic effect following phenol-induced shortening of APD in guinea pig ventricu- single-dose intravenous administration, causing significant lar cells.60 Millar and Vaughan Williams also have reported transient decreases in ejection fraction (EF), cardiac index, that cibenzoline (2.63μmol/L) prevented APD shortening stroke volume index, maximum velocity of contractile induced by hypoxia in isolated rabbit atria.47 In experiments shortening, and rate of change of LV pressure, accompa- by Niwa et al, however, cibenzoline (10μmol/L), unlike nied by an increase in vascular resistance, within the thera- glibenclamide, failed to prevent hypoxia-induced APD peutic range for plasma concentration (250–350μg/L).63 shortening of guinea pig ventricular muscle.51 Therefore, Effects on blood pressure are minimal.63 A comparative the pharmacologic importance of iK,ATP inhibition as an study indicated that the negative inotropic effects of ciben- action of cibenzoline is still unclear. zoline at doses 1.0 and 1.2mg/kg (iv) were similar to those Effects of cibenzoline on the muscarinic receptor-acti- of disopyramide at 2mg/kg in patients with coronary artery vated, acetylcholine (ACh)-sensitive K+ channel current disease.66 Oral cibenzoline (5mg day–1 kg–1) was shown to (iK,ACh) were examined by Wu et al in guinea pig atrial have no significant effects on left ventricular function.63 myocytes using a whole-cell voltage clamp method.61 They Cibenzoline has been relatively well tolerated in clinical showed a concentration-dependent inhibition of iK,ACh by trials.63 As for extracardiac side-effects, gastrointestinal cibenzoline with IC50 at 8μmol/L. This cibenzoline action disturbances, nervousness and tremulousness, and to a is most likely to be mediated by inhibition of the ACh- lesser extent blurred vision, dry mouth, and urinary reten- sensitive K+ channel itself and/or GTP-binding proteins tion have been observed. Some of these adverse effects are coupled to the channel.61 attributable to the anticholinergic properties of cibenzoline. The compound also sporadically induces hypoglycemia, Receptors and Pumps probably due to inhibition of KATP channels in pancreatic Cibenzoline has anticholinergic effects. Cazes et al cells.67–69 compared the anticholinergic effects of cibenzoline with those of disopyramide in vitro and in vivo animal experi- Gambit Profile (Table1) ments.62 Respective Ki values for inhibition of specific First, cibenzoline has a potent use-dependent inhibitory binding of 3H-quinuclidinyl benzylate (3H-QNB) of ciben- action on the cardiac sodium channel. From the onset and zoline and disopyramide were 15.8μmol/L and 12.0μmol/L offset kinetics of use-dependent blockade, cibenzoline is in a rat heart membrane preparation, and 31.6μmol/L and among the slow-kinetic drugs, and it may block the sodium 7.8μmol/L in cerebral cortical membranes.62 In isolated channel mainly during the activated state. Second, cibenzo- guinea pig ileum, disopyramide was about 15 times more line has inhibitory actions on Ca2+ channels (relatively strongly anticholinergic than cibenzoline.62 According to weak) and K+ channels (moderately potent). The K+ chan- antagonism of the negative chronotropic effect of vagal nels affected by cibenzoline include both voltage-depen- stimulation in anesthetized dogs, intravenously adminis- dent channels (iK, and iK1) and ligand-gated channels tered cibenzoline was shown to be several times less potent (iK,ATP and iK,ACh). Third, cibenzoline has anticholinergic than disopyramide.62 effects, but has no effects on other receptors or on pumps. Cibenzoline may not affect cardiac sympatheticβ-recep- Fourth, cibenzoline inhibits LV function of patients at ther- tors, because it does not inhibit the positive chronotropic apeutic doses, but has no consistent effects on sinus rate. effect of isoproterenol on isolated rabbit SA node tissue Extracardiac side-effects of cibenzoline are relatively preparations or dog Purkinje fibers.47,49 No reports to date weak, but in some patients hypoglycemia may occur. have shown significant effects of cibenzoline onα-adrener- Finally, in clinical ECG, cibenzoline prolongs PR and QRS gic receptors, A1 receptors, or sarcolemmal Na/K pumps in intervals without affecting the JT interval. the heart.

Clinical Effects Pilsicainide Several studies have investigated the electrophysiologic Action Potentials effects of cibenzoline in patients with various arrhyth- In isolated atrial and ventricular muscles from guinea mias.63 Touboul et al have reported dose-dependent pigs and Purkinje fibers from dogs, pilsicainide (3–30 increases in QRS, QT, AH, and HV intervals, and in the μmol/L) was shown to cause a concentration-dependent ventricular effective refractory period (ERP) over an intra- decrease in Vmax without affecting resting membrane venous dose range of 1.55–2.6 mg/kg.64 Prolongation of potential.70–73 Vmax of those tissues constantly stimulated at QRS and HV intervals reﬂecting intraventricular conduc- 1.0Hz was reduced by 30–50% at 30μmol/L pilsicainide.70,71 tion delay was observed at relatively low doses of cibenzo- The APD of Purkinje ﬁbers was shortened by pilsicainde line (1.55–1.8 mg/kg), whereas prolongation of the AH (by 40% at 30μmol/L), while the APD of atrial and ventric- interval reflecting conduction delay in the AV node was ular muscle was unaffected.70,71 Vmax inhibition was en- observed at higher doses.64 The JT interval was unaltered, hanced in a frequency- or use-dependent manner.72,73 Onset suggesting that the increased QRS duration alone and offset kinetics of use-dependent Vmax inhibition with

Japanese Circulation Journal Vol.63, January 1999 6 KODAMA I et al. pilsicainde are relatively slow.57,58 As for theτR from use- upon sudden exposure to ACh.81 These data suggest that dependent blockade, Hattori et al reported a value of 27–28 inhibition of iK,ACh by pilsicainide may resemble that by s in guinea pig papillary muscle.72 disopyramide, being mediated by blockade of muscarinic With respect to Ca2+ channel-dependent slow action (M2) receptors.80,81 Pilsicainide had no significant effects on potentials, pilsicainide at concentrations up to 30μmol/L iK,ATP activated by 2,4-dinitrophenol in guinea pig atrial showed no significant effects on the slow responses of myocytes.59 guinea pig papillary muscles induced by high K+ (22 No information is available concerning the effect of mmol/L) plus isoproterenol (0.2μmol/L).70 In small tissue pilsicainide on if. specimens of rabbit SA and AV nodes, however, pilsicainde (10–20μmol/L) caused slight decreases in Vmax and Receptors and Pumps amplitude of spontaneous action potentials.74,75 Pilsicainide has a weak inhibitory action on iK,ACh, probably through a blockade of M2 receptors as described Ionic Currents before. Unlike that of disopyramide, however, the anti- Direct evidence for iNa inhibition by pilsicainide was cholinergic effect of pilsicainide is minimal or negligible at demonstrated in whole-cell voltage-clamp experiments on therapeutic plasma concentrations (2–10μmol/L).80 Pilsicainide single ventricular myocytes isolated from rat and guinea at antiarrhythmic doses has little or no effects on the auto- pig hearts.71,76–78 Pilsicainide reduced iNa elicited by depo- nomic nervous system of animals in vivo.82,83 Pilsicainide, larization pulses (0.1–0.2Hz) in a concentration-dependent even at a high concentration (30μmol/L), has no significant manner with an IC50 of 20–30μmol/L.71,76 Inhibition of iNa effect on ACh- or histamine-induced contractions of by pilsicainide included both tonic (resting) and phasic isolated guinea pig ileum or tracheal smooth muscle.83 No (use-dependent) components with greater use-dependent reports to date have shown significant effects of pilsi- block of higher depolarization pulse frequencies.71,76–78 cainide on α- or β-adrenergic receptors, A1 receptors, or Recovery of iNa in the presence of pilsicainide (10–30 sarcolemmal Na/K pumps. μmol/ L) followed a double exponential function, most likely reflecting rapid recovery of drug-free Na+ channels and Clinical Effects slow recovery of drug-associated channels with time con- Pilsicainide is a novel class I antiarrhythmic agent devel- stants from 15 to 70ms and 65–580ms, respectively.71,76–78 oped in Japan by the Biomedical Institute of Suntory Inomata et al investigated the state-dependence of sodium Limited. Therefore, to date, clinical data regarding this channel blockade in guinea pig atrial myocytes pretreated drug have been limited to investigations in Japan. This with scorpion toxin (5μmol/L), which inhibits inactivation drug is indicated for supraventricular and ventricular tach- of iNa.77Application of pilsicainide (10μmol/L) to these yarrhythmias, and is considered highly effective and safe. myocytes resulted in a use-dependent decrease in the peak Pilsicainide for intravenous administration has been devel- amplitude of the relatively persistent iNa, and this iNa inhibi- oped, and will soon be clinically available. tion was associated with an accelerated phase current Effect on the Various Tachyarrhythmias In a double- decay.77 They suggested from these findings that pilsi- blind placebo-controlled comparative study of pilsicainide cainide may interact preferentially with the sodium channel (150 mg/day) and disopyramide (300 mg/day) on SVPC in the open state.77 and VPC, reduction of the respective arrhythmia by 75% or The state-dependence of sodium channel block by pilsi- more occurred in 28% and 38% in the disopyramide group, cainde also was tested in single guinea pig ventricular and 52% and 50% in the pilsicainide group.84,85 When the myocytes using multiple 0-mV conditioning voltage-clamp pharmacologic effect of pilsicainide on PSVT was evalu- measurements with an interpulse interval of 500ms from ated electrophysiologically, induction of PSVT was an HP at –85mV.79 Both short (10ms) and long (200ms) prevented in 69–86% of cases.86–88 In a double-blind, pulse trains applied to myocytes treated with pilsicainide placebo-controlled study investigating the acute effect of (10μmol/L) caused a marked use-dependent Vmax reduc- pilsicainide on paroxysmal atrial fibrillation, sinus rhythm tion with the trains, but the enhancement of Vmax reduction was restored in 45% of cases after a single oral dose of 150 by prolongation of clamp pulse duration was minimal or mg, compared to 9% in the placebo group.89 negligible.79 This pattern suggests that repetition of the acti- Clinical Electrophysiology and Effect on Cardiac vated state is required for development of use-dependent Function A single dose of pilsicainide of 150–200 mg blockade of iNa by pilsicainide. significantly prolonged PA, AH and HV intervals, sinoa- Pilsicainide at high concentrations causes minimal to trial conduction time (SACT), and right ventriclular ERP, moderate inhibition of ionic currents passing through Ca2+ but had no influence on sinus node recovery time, ERP of and K+ channels.71,75,80,81 Inomata et al compared the the AV node, functional refractory period of the AV node, potency of iNa and iCa inhibition by pilsicainide in guinea or ERP of the right atrium.86 With regard to effects on elec- pig ventricular myocytes, and reported IC30 of 9.1μmol/L trocardiographic parameters, the QRS and PQ intervals and for iNa and 55μmol/L for iCa.71 They therefore, suggested the P wave width were prolonged in a dose-dependent that the iCa-blocking action of pilsicainide would be negli- manner, but no influence was observed on RR or QTc gible at an effective plasma concentration for treatment of intervals.84,85,88 arrhythmias (2–10μmol/L). Adverse Effects and Clinical Pharmacology During In single guinea pig atrial myocytes, iK,ACh was inhibited postmarketing surveillance, adverse effects of pilsicainide by pilsicainide in a concentration-dependent manner (IC50, were observed in 4.8% of 3768 cases; including first degree 25–30μmol/L) when the current was activated by external AV block in 0.72%, bundle branch block in 0.56%, and application of ACh; however, the current activated by gastrointestinal symptoms in 0.94%. Ventricular tachycar- internal loading with GTPgS was unaffected.80 In experi- dia and ventricular fibrillation were observed in 0.19% and ments using concentration-clamp techniques, pilsicainide 0.08% of cases, respectively. was shown to prolong the latent period of iK,ACh activation Pharmacokinetically, the drug is well fitted by a one-

Japanese Circulation Journal Vol.63, January 1999 Aprindine, Cibenzoline, Pilsicainide and Pirmenol in the Sicilian Gambit 7 compartment model, exhibiting linear pharmacokinetics the sodium channel primarily when it is in the open (acti- with a half-life (T1⁄2) showing a proportional correlation vated) state.93 between the dose and both the area under the curve (AUC) In guinea pig papillary muscles treated with pirmenol, a and Cmax. Cmax and T1⁄2 were found to be 1–2 and 4–5h, partial depolarization of the resting membrane potential respectively. Pilsicainide is considered renally excreted; resulting from increasing the extracellular K+ concentration 75–86% of administered drug is excreted through the from 4 to 8 mmol/L resulted in an acceleration of Vmax kidneys.90,91 recovery from use-dependent blockade despite an increase of steady-state amount of blockade.99 Vmax recovery from Gambit Profile (Table1) use-dependent blockade with lidocaine and mexiletine is First, pilsicainide has a potent use-dependent inhibitory known to be accelerated at more negative membrane poten- action on the cardiac sodium channel. From the onset and tials (hyperpolarization).103,104 This has been attributed to a offset kinetics of this use-dependent blockade, pilsicainide higher affinity for the inactivated state of the sodium is a slow-kinetic drug, and it seems to block the sodium channel than the resting state. Under such conditions, channel mainly during the activated state. Second, pilsi- hyperpolarization would enhance drug dissociation from cainide at therapeutic concentrations has no substantial channel receptors by increasing the resting fraction of effects on other ionic channels, receptors, or pumps. drug-associated channels at the expense of the inactivated fraction.103,104 The voltage-dependence of Vmax recovery for pirmenol is opposite to these drugs, and similar to those Pirmenol reported for penticainide,105 disopyramide,106 and dipra- Action Potentials fenone.102 According to the hypothesis of ‘activation trap- In Purkinje fibers of dogs and rabbits constantly stimu- ping’105,106 the drug molecules, which bind to and unbind lated at 0.7–2Hz, pirmenol (≥1μmol/L) caused a concen- from sodium channels primarily during the activated state, tration-dependent decrease in Vmax without affecting are trapped in the channel when it returns to the resting resting membrane potential (IC50 for Vmax inhibition was state. Because the probability of having the activation gate about 10μmol/L).92–95 The APD of Purkinje fibers at the in the open position (flicker openings) is smaller at more late repolarization phase was moderately prolonged (3– negative membrane potentials, dissociation of drugs from 12%) at relatively low concentrations of pirmenol (1–10 channel receptors during electrical diastole would be accel- μmol/L), but shortened at high concentrations (10–30 erated by depolarization. Dissociation (unbinding) of drugs μmol/L).92–95 In rabbit atrial muscles stimulated at 1–2Hz, from the sodium channels during the activation state (acti- Vmax was decreased by 30–40% and APD at late repolariza- vation unblocking) is enhanced by hyperpolarization of the tion was prolonged by 10–40% by 5–6μmol/L pirmenol.94,96 resting membrane potential,107 leading to a decrease of the Pirmenol also was shown to concentration-dependently steady-state amount of blockade during the stimulation reduce Vmax in ventricular muscles of guinea pigs and trains.102,105–107 Nevertheless, further experimental studies rabbits.96–101 IC50 for Vmax inhibition was about 30μmol/L will be required to substantiate this interpretation. in guinea pig papillary muscles stimulated at 1.0 Hz.97,100 Disagreement prevails as to effects of pirmenol on action APD of the ventricular muscle was slightly prolonged potentials at the SA node. Dukes et al96 and Sawanobori et (5–10%) at low concentrations (3–10μmol/L), but shortened al94 have reported that pirmenol (2–25μmol/L) caused a at high concentrations of pirmenol (30–100μmol/L).97,98,100,101 concentration-dependent prolongation of cycle length of Prolongation of APD (13–30%) by pirmenol (5–10μmol/L) spontaneous excitation in rabbit SA node tissue prepara- also was reported in single atrial and ventricular myocytes tions due to prolongation of repolarization time and a obtained from rabbits.94,102 decrease in pacemaker slope, but had no significant effect Vmax inhibition by pirmenol in Purkinje fibers and on Vmax. Kotake et al however, found that pirmenol (1–30 ventricular muscles was enhanced in a frequency- or use- μmol/L) produced a concentration-dependent decrease in dependent manner.93,95,97–100 Onset and offset kinetics of Vmax of action potentials in small tissue specimens from the use-dependent Vmax inhibition are slow.93,95,97–100 TheτR of rabbit SA node, in association with prolongation of sponta- Vmax from use-dependent blockade was reported as 27–34s neous cycle length and APD.108 No other information is in experiments by Nakaya et al using guinea pig papillary available regarding effects of pirmenol on slow Ca2+ muscles.97,99,100 Comparable values (16–18s) were reported channel-dependent action potentials. by other investigators in experiments on ventricular muscle or Purkinje fibers.95,98 Reichardt et al examined voltage- Ionic Currents and time-dependence of Vmax inhibition in rabbit Purkinje No voltage-clamp studies have been performed to clarify fibers using a double-microelectrode voltage-clamp meth- the effects of pirmenol on iNa in cardiac tissues or cells. od.93 Application of a single conditioning depolarization to Pirmenol (10–30μmol/L) was shown to decrease the –45 mV from an HP at –105 mV in the presence of amplitude of the slow inward current (isi) in multicellular pirmenol (50μmol/L) resulted in a significant decrease in tissue preparations of rabbit SA nodes in association with Vmax of the test action potential elicited 100ms after the slowing of isi recovery from inactivation.108 In whole-cell conditioning clamp. The Vmax reduction consisted of 2 voltage-clamp studies on single ventricular and atrial phases; an initial fall with a very short conditioning depo- myocytes from rabbit or guinea pig hearts, pirmenol (5–30 larization (a few milliseconds) was followed by a very slow μmol/L) decreased the peak amplitude of iCa by 20–22%.94,109 decline with prolongation of depolarization for up to 2.4s.93 As to Ca2+ channel inhibition by pirmenol, however, infor- In experiments with trains of depolarization to –45 mV mation remains very limited, and more extensive voltage- from an HP of –105mV, prolongation of each depolariza- clamp studies will be required to assess the pharmacologic tion from 10 to 500 ms had no significant effect on the significance of this possible action. extent of use-dependent Vmax inhibiton.93 From these obser- Pirmenol inhibits K+-channel currents. In whole-cell vations, Reichardt et al concluded that pirmenol may block voltage clamp experiments on single ventricular myocytes

Japanese Circulation Journal Vol.63, January 1999 8 KODAMA I et al. from guinea pigs or rabbits, Sawanobori et al have demon- of cibenzoline; that is, when given orally, pirmenol does strated that the tail current of iK was decreased by about not affect sinus rate significantly,115,116 but the drug 30% after application of 5μmol/L pirmenol.94 A significant increases sinus rate (by 13%), when given intravenously inhibition of delayed rectifier outward currents (iK or ix) by (50mg initially, then 2.5mg/min).117 This modest increase pirmenol (1–30μmol/L) was observed as well in double- in sinus rate with pirmenol may be due to sympathetic microelectrode voltage-clamp studies on multicellular reflexes responding to the negative inotropic effect of the tissue preparations of rabbit SA node and Purkinje drug.117 fibers.93,108 In single rabbit atrial myocytes, pirmenol was Pirmenol does not reduce EF at daily oral doses up to shown to cause a concentration-dependent decrease in the 500mg,118 but it reduces EF after intravenous administra- 4-aminopyridine (4-AP)-sensitive transient outward current tion.117 However, an increase in sinus rate occurrs possibly (ito), with an IC50 of 10μmol/L.109 Nakajima et al have through sympathetic reflexes, keeping the cardiac index suggested that the pirmenol-induced APD prolongation in unchanged while EF is depressed after intravenous injec- atrial cells is mediated primarily by ito inhibition, and in tion.117 Mean arterial pressure increases after intravenous atrial myocytes, iK1 was unaffected by pirmenol.109 injection.117,119 Pirmenol also suppressed both iK,ACh and adenosine (Ado)- Major extracardiac adverse effects include dry mouth induced K+ currents (iK,Ado) in guinea pig atrial cells with and urinary hesitancy, and are largely due to anticholiner- IC50 at about 1μmol/L and 8μmol/L, respectively.109 The gic properties.115 Otherwise, no other serious extracardiac outward current activated by internal loading with GTPgS adverse effects have been reported. also was inhibited by pirmenol, with IC50 being 8μmol/L.109 These data suggest that pirmenol inhibits muscarinic K+ Gambit Profile (Table1) currents in atrial myocytes through dual mechanisms: both First, pirmenol has a potent use-dependent inhibitory through blockade of muscarinic receptors and through action on the cardiac sodium channel. From onset and direct inhibition of the muscarinic K+ channel itself and/or offset kinetics of use-dependent blockade, pirmenol is a the function of GTP-binding protein coupled to the slow-kinetic drug, and it may block the sodium channel channel.109 primarily in the activated state. Second, a Ca2+ channel- inhibitory action of pirmenol remains to be confirmed. Receptors and Pumps Third, pirmenol has K+ channel blocking action of moder- No experimental evidence suggests that pirmenol has ate potency. The potassium channels affected by pirmenol any action blocking α- or β-adrenergic receptors.96,110 include both voltage-dependent and ligand-gated channels Pirmenol has no significant direct effects on vascular (iK, ito, iK,ACh). Finally, pirmenol has significant anticholin- smooth muscle, nor does it alter responses to a variety of ergic effects on the heart, but its anticholinergic effects on agonists.110 smooth muscle and exocrine glands are relatively weak. Pirmenol has anticholinergic effects. The potency of pirmenol to antagonize carbachol-induced contraction of the isolated guinea pig ileum was significantly less than Discussion that of disopyramide (25% antagonism) or quinidine (50% As the group pursued the task of fitting the pharmaco- antagonism).110 In contrast, the potency of pirmenol to logic profiles of the 4 drugs (aprindine, cibenzoline, pilsi- antagonize the negative inotropic effect of carbachol in the cainide, and pirmenol) into the spreadsheet of the Gambit guinea pig left atrium was greater than that of disopyra- framework, we noticed several problems and limitation in mide, with pA2 values being 6.47 and 5.99, respectively.111 completing the table. This could reflect different affinities of pirmenol to differ- First, no quantitative criteria have been set for grading ent subtypes of muscarinic receptors.112 Cardiac muscarinic the potency of drug effects on the molecular targets (‘low’, receptors are quite unique because of their single receptor ‘moderate’, and ‘high’ are used semiquantitatively). Onset subtype population, specifically M2 receptors.113 On the and offset kinetics of sodium channel blockade are known other hand, smooth muscle contains a mixture of M2 and to be influenced by experimental conditions such as M3 receptors, and exocrine glands contain mainly M3 temperature, external Na+ concentrations, drug concentra- receptors.113,114 In binding studies on membrane preparations, protocols of stimulation, and nonlinear relationships tions from guinea pig left atrium, submandibular gland, and between Vmax and iNa.104,120 For drugs showing substantial urinary bladder smooth muscle using [3H]N-methylscopo- activation trapping and unblocking, such as disopyramide lamine as a ligand, Endou et al found pirmenol to show or pirmenol, a change of the resting membrane potential only one-seventh the affinity for glandular-type muscarinic only by several millivolts could result in a marked alter- receptors (M3) than that of cardiac type muscarinic recep- ation of the τR from use-dependent blockade.102,105–107 tors (M2), while disopyramide and pentisomide had compa- These factors often made it difficult to compare data from rable affinities for M3 and M2 receptors.111,112 different experimental studies. No reports to date have shown significant effects of Second, more than 6 different types of K+ channels have pirmenol on A1 receptors or the function of sarcolemmal been identified in cardiac cells.52,121 These include voltage- Na/K pumps. gated channels including the delayed rectifier K+ current (iK), the transient outward current (ito), and the inward Clinical Effects rectifier K+ current (iK1), and ligand-gated channels such as Electrophysiologic effects of pirmenol have been evalu- the muscarinic acetylcholine receptor-operated K+ current ated in patients with ventricular arrhythmias. PR, QRS, and (iK,ACh), the ATP-sensitive K+ current (iK,ATP), and the Na+- QT intervals all are prolonged slightly by oral pirmenol at activated K+ current (iK,Na). Further, iK is subdivided into daily doses of 150–450 mg.115,116 Prolongation of the QT rapidly activating components (iKr and iKur) and a slowly interval was attributed largely to prolongation of the JT activating component (iKs). Each type of K+ channel interval. Effects of pirmenol on sinus rate is similar to that current plays different roles in the repolarization of action

Japanese Circulation Journal Vol.63, January 1999 Aprindine, Cibenzoline, Pilsicainide and Pirmenol in the Sicilian Gambit 9 potentials and in pacemaking activity.52,121 The physiologic extensive discussion of the pharmacologic profiles of the 4 and pathologic significance of these K+ channel currents drugs from different viewpoints and perspectives to reach a also is different in different cardiac tissues and animal balanced conclusion. We have identified essential data and species.52 The presentation of drug actions on the K+ information that still are lacking but are required to outline channel in a single column is therefore likely to be insuffi- the fundamental profiles of the 4 drugs. The continuing cient and misleading. Further clarification of the specificity discussion will direct urgently needed experimental and of each antiarrhythmic agent for these K+- channel sub- clinical studies. types appears necessary in applying the Gambit concept to clinical arrhythmias, especially atrial tachyarrhythmias; for References example, much information has accumulated about the mechanism of atrial fibrillation, and molecular data relating 1. Vaughan Williams EM: A classification of antiarrhythmic actions reassessed after a decade of new drugs. J Clin Pharmacol 1984; 24: to atrial remodeling, including modification of expression 129–147 + 122–125 of various K channel proteins, have been obtained. 2. Task Force of the Working Group on Arrhythmias of the European Specific K+ channels that play a pivotal part in initiating or Society of Cardiology. The Sicilian Gambit: A new approach to the sustaining atrial fibrillation should be a target of antifibril- classification of antiarrhythmic drugs based on their actions on latory therapy. We now know that paroxysmal atrial fibril- arrhythmogenic mechanisms. Circulation 1991; 84: 1831–1851 3. Members of the Sicilian Gambit. Antiarrhythmic therapy: A patho- lation is triggered exclusively by enhanced vagal activity physiological approach. Armonk, NY: Futura Publishing Company, that activates iK,ACh and shortens atrial refractory periods. 1994 Drugs that inhibit this current either directly or indirectly 4. Ogawa S. Antiarrhythmic therapy: A pathophysiological approach through muscarinic receptor blocking, such as cibenzoline (in Japanese). Tokyo: Igakushoin, 1995 5. Verdonck F, Vereecke J, Vleugels A. Electrophysiological effects of and pirmenol, promise to be more beneficial in preventing aprindine on isolated heart preparations. Eur J Pharmacol 1974; 26: vagally induced paroxysmal atrial fibrillation than disopy- 338–347 ramide, which suppresses this current only indirectly. 6. Carmeliet E, Verdonck F: Effects of aprindine and lidocaine on Furthermore, as atrial remodeling with modification of transmembrane potentials and radioactive K efflux in different + 123–125 cardiac tissues. Acta Cardiol 1974; S18: 73–90 expression of target K channels progresses, once- 7. Steinberg MI, Greenspan K: Intracellular electrophysiological alter- effective drugs may become ineffective. ations in canine cardiac conducting tissue induced by aprindine and As shown in Table 1, we found that cibenzoline and lignocaine. Cardiovasc Res 1976; 10: 236–244 pirmenol, both of which have a pharmacologic profile very 8. Dennis PD, Vaughan Williams EM: Effects on rabbit cardiac potentials of aprindine and indecainide, a new antiarrhythmic agent, in similar to that of disopyramide, have different effects on normoxia and hypoxia. Br J Pharmacol 1985; 85: 11–19 ECG parameters, especially the JT interval. This can be 9. Toyama J, Kamiya K, Kodama I, Yamada K: Frequency- and explained in terms of the difference in K+ channel subtype voltage-dependent effects of aprindine on the upstroke velocity of specificity; these 3 drugs may have different potencies of action potential in guinea pig ventricular muscles. J Cardiovasc action at various K+ channels expressed predominantly in Pharmacol 1987; 9: 165–172 10. Kawamura T, Kodama I, Toyama J, Hayashi H, Saito H, Yamada K: the atrium or in the ventricle. Any drugs that selectively Combined application of class I antiarrhythmic drugs causes ‘addi- prolong atrial refratory periods without QT prolongation tive’, ‘reductive’, or ‘synergistic’ sodium channel block in cardiac should be ideal for treatment of atrial fibrillation. Thus, muscles. Cardiovasc Res 1990; 24: 925–931 further investigations are needed to elucidate K+ channel 11. Honerjager P, Loibl E, Stidl I, Schonsteiner G, Ulm K: Negative inotropic effects of tetrodotoxin and seven class 1 antiarrhythmic subtype specificity for all available agents. drugs in relation to sodium channel blockade. Naunyn Schmiedebergs A third problem with the Gambit approach is that clini- Arch Pharmacol 1986; 332: 184–195 cal effects on LV function, heart rate, and electrocardio- 12. Moyer JW, Hondeghem LM: Rate rhythm and voltage dependent graphic intervals (PR, QRS, JT) are indicated by arrows effects of aprindine: a test of a model of the mechanisms of action of antiarrhythmic drugs. Proc West Pharmacol Soc 1978; 21: 57–61 (increase, decrease, or no change), and extracardiac side- 13. Kodama I, Toyama J, Yamada K: Competitive inhibition of cardiac effects are presented by 3 symbols (low, moderate, or sodium channels by aprindine and lidocaine studied using a high). No quantitative criteria on standards for such estima- maximum upstroke velocity of action potential in guinea pig ventric- tion and grading have been adopted, interfering with data ular muscles. J Pharmacol Exp Ther 1987; 241: 1065–1071 comparison involving different clinical studies with differ- 14. Kamiya K, Kodama I, Toyama J: A combination of inactivated sodium channel blockers causes competitive interaction on dV/dtmax ent protocols. Despite these limitations, we noted that of single ventricular myocytes. Cardiovasc Res 1991; 25: 516–522 unlike other Na+ channel blockers with slow recovery 15. Shirayama T, Inoue D, Inoue M, Tasumi T, Yamahara Y, Asayama J, kinetics, pilsicainide has a relatively minimal effect on LV et al: Electrophysiological effects of sodium channel blockers on function, permitting administration to patients with LV guinea pig left atrium. J Pharmacol Exp Ther 1991; 259: 884–893 16. Iida M, Kodama I, Toyama J: Negative dromotropic effects of Class impairment. I antiarrhythmic drugs in anisotropic ventricular muscle. Cardiovasc In addition, the Gambit framework does not discriminate Res 1996; 31: 640–650 between acute and chronic effects of antiarrhythmic drugs, 17. Woosley RL: Antiarrhythmic drugs: Annu Rev Pharmacol Toxicol which could distort or limit profiling of some drugs. For 1991; 31: 427–455 18. Kodama I, Honjo H, Kamiya K, Toyama J: Two types of sodium instance, long-term administration of amiodarone for channel block by Class-I antiarrhythmic drugs studied by using Vmax several weeks is known to cause quite different effects on of action potential in single ventricular myocytes. J Mol Cell Cardiol ion channels, receptors, pumps, electrocardiographic para- 1990; 22: 1–12 meters, and extracardiac organs than those immediately 19. Kodama I, Toyama J, Takanaka C, Yamada K: Block of activated following initial administration of the drug.126 and inactivated sodium channels by Class-I antiarrhythmic drugs studied by using the maximum upstroke velocity (Vmax) of action Despite such problems and limitations, the group’s task potential in guinea-pig cardiac muscles. J Mol Cell Cardiol 1987; was no doubt worthwhile. The Guideline Committee 19: 367–377 members include selected experts on arrhythmias from 20. Nitta J, Sunami A, Furukawa T, Marumo F, Sawanobori T, Hiraoka M: Combined class I antiarrhythmic agents have differential effects Japan having different specialties including molecular and on tonic and use dependent block of maximum rate of depolarization cellular electrophysiology, in vivo experimental cardiol- of action potentials in guinea pig cardiac muscle. Cardiovasc Res ogy, and clinical cardiology. This diversity facilitated 1992; 26: 462–469

Japanese Circulation Journal Vol.63, January 1999 10 KODAMA I et al.

21. Kamiya K, Toyama J, Honjo H, Kodama I: A model analysis for 45. Ebner T, Eichelbaum M: The metabolism of aprindine in relation to competitive binding of mexiletine and aprindine to the cardiac the sparteine/debrisoquine polymorphism. Br J Clin Pharmacol sodium channel. Pharmacology 1992; 44: 260–268 1993; 35: 426–430 22. Lee J-K, Takanaka C, Nonokawa M, Kato H, Yabe S, Kodama I, et 46. Inabez L, Juan J, Perez E, Carne X, Laporte JR: Agranulocytosis al: Complex frequency–dependent interaction of class-I antiarrhyth- associated with aprindine and other antiarrhythmic drugs: an mic drugs as they affect intraventricular conduction. Eur Heart J epidemiological approach. Eur Heart J 1991; 1: 147–156 1995; 16: 832–841 47. Millar JS, Vaughan Williams EM: Effects on rabbit nodal, atrial, 23. Elharrar V, Bailey JC, Lathrop DA, Zipes DP: Effects of aprindine ventricular and Purkinje cell potentials of a new antiarrhythmic drug, HCl on slow channel action potentials and transient depolarization in cibenzoline, which protects against action potential shortening in canine Purkinje fibers. J Pharmacol Exp Ther 1978; 205: 410–417 hypoxia. Br J Pharmacol 1982; 75: 469–478 24. Kotake H, Igawa O, Miyamoto J, Hasegawa J, Furuse T, Mashiba H: 48. Millar JS, Vaughan Williams EM: Pharmacological mapping of Studies on the bradycardia induced by aprindine in rabbit sinoatrial regional effects in the rabbit heart of some new antiarrhythmic drugs. node cells. Pharmacology 1986; 33: 14–20 Br J Pharmacol 1983; 79: 701–709 25. Tanaka H, Nishimura M, Homma N, Habuchi Y, Watanabe Y: 49. Dangman KH: Cardiac effects of cibenzoline. J Cardiovasc Electrophysiologic actions of aprindine in rabbit atrioventricular Pharmacol 1984; 6: 300–311 node. Naunyn Schmiedebergs Arch Pharmacol 1990; 341: 347–356 50. Satoh H, Ishii M, Hashimoto K: Effect of cibenzoline, a Class I 26. Sato R, Hisatome I, Tanaka Y, Sasaki N, Kotake H, Mashiba H, et al: antiarrhythmic drug, on action potential in canine ventricular muscle. Aprindine blocks the sodium current in guinea-pig ventricular Jpn J Pharmacol 1987; 44: 113–119 myocytes. Naunyn Schmiedebergs Arch Pharmacol 1991; 344: 331– 51. Niwa R, Honjo H, Kodama I, Maruyama K, Toyama J: Sodium 336 channel blocking effects of cibenzoline on guinea-pig ventricular 27. Shibasaki T, Kojima S, Ito H, Hirakawa S: Effects of anti-arrhythmic cells. Eur J Pharmacol 1998; 352: 317–327 drugs on action potential repolarization of guinea-pig atrial myocytes 52. Carmeliet E: Mechanisms and control of repolarization. Eur Heart J (in Japanese). Heart 1991; 23: 327–331 1993; 14(Suppl H): 3–13 28. Georges A: Pharmacological evaluation of aprindine (AC1802), a 53. Kotake H, Matsuoka S, Ogino K, Takami T, Hasegawa J, Mashiba new antiarrhythmic agent. Acta Cardiologica 1973; 28: 166–191 H: Electrophysiological study of cibenzoline in voltage-clamped 29. Kamiya J, Yoshiwara K, Maruyama M, Saito S, Ishii M, Hirayama rabbit sinoatrial node preparations. J Pharmacol Exp Ther 1987; M, et al: General pharmacology of antiarrhythmic drug, aprindine (in 241: 982–986 Japanese). Jpn Pharmacol Ther 1985; 13: 2713–2738 54. Masse C, Cazes M, Sassine A: Effects of cibenzoline, a novel antiar- 30. Levine SN, Hollier B: Aprindine inhibits calmodulin-stimulated rhythmic drug, on action potential and transmembrane currents in phosphodiesterase and CaATPase activities. J Cardiovasc Pharmacol frog atrial muscle. Arch Int Pharmacodyn 1984; 269: 219–235 1983; 5: 151–156 55. Sato T, Wu B, Kiyosue T, Arita M: Effects of cibenzoline, a new 31. Mannhold R, Voigt W, Bast A, Timmerman H: Interaction of class I class Ia antiarrhythmic drug, on various membrane ionic currents and compounds with calmodulin. Arch Pharm (Weinheim) 1990; 323: action potentials of guinea-pig ventricular cells. Naunyn 489–491 Schmiedebergs Arch Pharmacol 1994; 350: 167–173 32. Pouleur H, Chaudron JM, Reyns P: Effects of disopyramide and 56. Holck M, Osterrieder W: Inhibition of the myocardial Ca2+ inward aprindine on arrhythmias after acute myocardial infarction. Eur J current by the class I antiarrhythmic agent, cibenzoline. Br J Cardiol 1977; 5: 397–404 Pharmacol 1986; 87: 705–711 33. Kato K, Hayakawa H, Ueda K, Tanabe A, Iinuma H, Tsuya K, et al: 57. Matsuoka S, Nawada T, Hisatome I, Miyamoto J, Hasegawa J, The clinical evaluation of aprindine (MS-5075) capsules on prema- Kotake H, et al: Comparison of Ca2+ channel inhibitory effects of ture contractions-multi-center double blind study (in Japanese). Jpn J cibenzoline with verapamil on guinea-pig heart. Gen Pharmacol Electrocardiol 1985; 5: 814–838 1991; 22: 87–91 34. Gaum WE, Foster PE, Rosen KM, Wu D, Amai-Y-Leon F, Noble 58. Wang DW, Kiyosue T, Sato T, Arita M: Comparison of the effects of RJ: Treatment of refractory supraventricular tachycardia (SVT) with class I anti-arrhythmic drugs, cibenzoline, mexiletine and flecainide, aprindine. Circulation 1976; 53(Suppl 2): 18 on the delayed rectifier K+ current of guinea-pig ventricular 35. Hambert M, Peskoe SM, Wann SL, Noble J, Fasola AF, Elharrar V, myocytes. J Mol Cell Cardiol 1996; 28: 893–904 Zipes DP: Aprindine for treatment of supraventricular tachycardias 59. Wu B, Sato T, Kiyosue T, Arita M: Blockade of 2,4-dinitrophenol with particular application to WPW syndrome. Am J Cardiol 1977; induced ATP sensitive potassium current in guinea pig ventricular 40: 586–596 myocytes by class I antiarrhythmic drugs. Cardiovasc Res 1992; 26: 36. Schaeffer AH, Greene HL, Reid PR: Comparative efficacy of 1095–1192 aprindine in the management of refractory atrial fibrillation and 60. Sato T, Wu B, Nakamura S, Kiyosue T, Arita M: Cibenzoline flutter. Am J Cardiol 1979; 43: 430 inhibits diazoxide- and; 2,4 dinitrophenol-activated ATP-sensitive 37. Hirao K, Okishige K, Suzuki F, Hiejima K: Combination therapy K+ channels in guinea-pig ventricular cells. Br J Pharmacol 1993; with aprindine and verapamil for paroxysmal supraventricular tachy- 108: 549–556 cardia as assessed by transesophageal atrial pacing. Cardiovasc Drug 61. Wu S-N, Nakajima T, Yamashita T, Hamada E, Hazama H, Iwasawa Ther 1991; 5: 819–826 K, et al: Molecular mechanism of cibenzoline-induced anticholiner- 38. Gottlieb SH, Achoff SC, Mellits ED, Gerslenbith SG, Baughman K, gic action in single atrial myocytes: comparison with effect of Becker L, et al: Prophylactic antiarrhythmic therapy of high-risk disopyramide. J Cardiovasc Pharmacol 1994; 23: 618–624 survivors of myocardial infraction: lower mortality at 1 month but 62. Cazes M, Chassaing C, Martinet M, Cloarec A, Provost D, Boucher not 1 year. Circulation 1987; 75: 792–799 M, et al: Comparison of anticholinergic effects of cibenzoline, 39. Hagemeijer F, Glaser B, Van Durme JP, Lubsen J, Boguert M: disopyramide, and atropine. J Cardiovasc Pharmacol 1990; 15: Design of a study to evaluate drug therapy of serious ventricular 308–316 rhythm disturbances after an acute myocardial infarction. Eur J 63. Harron DWG, Brogden RN, Faulds D, Fitton A: Cibenzoline: A Cardiol 1977; 6: 299–310 review of its pharmacological properties and therapeutic potential in 40. Hagemeijer F, Glaser B, Van Durme JP, Lubsen J, Boguert M, arrhythmias. Drugs 1992; 43: 734–759 Hugenholtz PG: The Ghen-Rotterdam aprindine study: Antiarrhyth- 64. Touboul P, Atallah G, Kirkorian G, de Zuloaga C, Dufour A, mic prophylaxis after myocardial infarction. Excerpta Medica 1979; Aymard MF, et al: Electrophysiologic effects of cibenzoline in 491: 964–966 humans related to dose and plasma concentration. Am Heart J 1986; 41. Simada R, Araki H, Orita Y, Nakagaki O, Nakamura M: Cardiac 112: 333–339 effects of aprindine in patients with or without cardiac dysfunction: 65. Klein RC, Horowitz LD, Rushforth N: Efficacy and safety of oral echocardiographic and clinical evaluation. Excerpta Medica 1983; 5: cibenzoline for ventricular arrhythmias. Am J Cardiol 1986; 57: 348–356 592–597 42. Piessens J, Willems J, Kestelcot H, De Geest H: Effects of aprindine 66. Humen DP, Lesoway R, Kostuk WJ: Acute, single, intravenous on left ventricular contractility in man. Acta Cardiol (Brux) Suppl doses of cibenzoline: an evaluation of safety, tolerance, and hemody- 1974; 18: 203–216 namic effects. Clin Pharmacol Ther 1987; 41: 537–545 43. Vlay SC, Kallman C, Reid SP: The utility of aprindine blood levels 67. Hilleman DE, Mohiuddin SM, Ahmed IS, Dahl JM: Cibenzoline- in the management of ventricular arrhythmias. J Am Coll Cardiol induced hypoglycemia. Drug Intell Clin Pharmacol 1987; 21: 38– 1985; 5: 738–743 40 44. Yokota M, Inagaki H, Uematsu H, Enomoto N, Goto J, Sotobata I, et 68. Gachot BA, Bezier M, Cherrier JF, Daubeze J: Cibenzoline and al: Non-linear pharmacokinetics of aprindine hydrochloride in oral hypoglycemia. Lancet 1988; 8605: 280 administration. Arzneimittelforschung/Drug Res 1987; 37: 184–188 69. Takahashi A, Horie M, Tsuura Y, Ishida H, Ai T, Sasayama S: Block

Japanese Circulation Journal Vol.63, January 1999 Aprindine, Cibenzoline, Pilsicainide and Pirmenol in the Sicilian Gambit 11

of pancreatic ATPsensitive K+ channels and insulinotrophic action of pilsicainide (Pilsicainide suppression trial on atrial fibrillation). by the antiarrhythmic agent, cibenzoline. Br J Pharmacol 1996; 117: Am J Cardiol 1996; 78: 694–697 1749–1755 90. Maeno K, Ikeda T, Takada S, Kobayashi K, Bunko H, Hisada K: The 70. Hattori Y, Inomata N, Aisaka K, Ishihara T: Electrophysiological effect of oral SUN-1165 on hemodynamics at rest and during exer- actions of N-(2, 6-dimethyl phenyl)-8-pyrrolizidine-acetamide cise in patients with ischemic heart disease (in Japanese). Jpn hydrochloride hemihydrate (SUN 1165), a new antiarrhythmic agent. Pharmacol Ther 1990; 18: 4051–4057 J Cardiovasc Pharmacol 1986; 8: 998–1002 91. Nakajima M, Takabatake T, Ohta H, Yamamoto Y, Ishida Y, Hara 71. Inomata N, Ishihara T, Akaike N: SUN 1165: a new antiarrhythmic H, et al: Pharmacokinetics of SUN 1165, a new antiarrhythmic agent, Na current blocker in ventricular myocytes of guinea-pig. Comp in renal dysfunction. Eur J Clin. Pharmacol 1991; 40: 411–414 Biochem Physiol 1987; 87C: 237–243 92. Reder R, Danilo P Jr, Rosen MR: Effects of pirmenol HCl on elec- 72. Hattori Y, Hidaka T, Aisaka K, Satoh F, Ishihara T: Effect of SUN trophysiologic properties of cardiac Purkinje fibers. Eur J Pharmacol 1165, a new potent antiarrhythmic agent, on the kinetics of rate- 1980; 61: 321–333 dependent block of Na channels and ventricular conduction of 93. Reichardt B, Konzen G, Hauswirth O: Pirmenol, a new antiarrhyth- extrasystoles. J Cardiovasc Pharmacol 1988; 11: 407–412 mic drug with potassium- and sodium-channel blocking activity; a 73. Toyama J, Kodama I, Kusunoki T, Ishihara T, Hattori Y, Yamada K: voltage-clamp study in rabbit Purkinje fibers. Naunyn Schmiedebergs Use-dependent blocking action of newly developed lidocaine- Arch Pharmacol 1990; 341: 462–471 analogs on maximum rate of rise of action potentials in guinea pig 94. Sawanobori T, Adaniya H, Yamashita K, Kawano S, Hayami H, papillary muscle. Jpn Heart J 1987; 28: 273–285 Kuga K, et al: Electrophysiologic and antiarrhythmic actions of 74. Saeki T, Yajima K, Sakakibara Y, Suzuki K, Fujinami T: Effects of pirmenol on rabbit and guinea pig cardiac preparations. J Cardiovasc class IC antiarrhythmic agents on action potential in rabbit sinoatrial Pharmacol 1990; 16: 975–983 node. Jpn Heart J 1994; 35(S1): 282–282 95. Lee JH, Rosen MR: Electrophysiologic effects of pirmenol, its 75. Suzuki K, Sakakibara Y, Sugima M, Saeki T, Yamada A: metabolite 2, and enatiomers, on cardiac Purkinje fibers. J Electrophysiological effects of the antiarrhythmic agent pilsicainide Cardiovasc Pharmacol 1993; 22: 416–422 on the rabbit atrioventricular node cell (in Japanese). Jpn J 96. Dukes ID, Vaughan Williams EM, Dennis PD: Electrophysiological Electrocardiol 1992; 12: 713–723 and cardiovascular effects of pirmenol, a new class 1 antiarrhythmic 76. Yatani A, Akaike N: Effects of a new antiarrhythmic compound drug. J Cardiovasc Pharmacol 1986; 8: 227–234 SUN 1165 [N-(2,6-dimethyl phenyl)-8-pyrrolizidineacetamide 97. Nakaya H, Tohse N, Kanno M: Frequency- and voltage-dependent hydrochloride] on the sodium currents in isolated single rat ventricu- depression of maximum upstroke velocity of action potentials by lar cells. Naunyn Schmiedebergs Arch Pharmacol 1984; 326: 163– pirmenol in guinea pig ventricular muscles. Jpn J Pharmacol 1988; 168 48: 423–434 77. Inomata N, Ishihara T, Akaike N: Different time courses of the 98. Hasegawa J, Hirai S, Noguchi N, Hisatome I, Kotake H, Mashiba H: blockade of sodium current by lignocaine and SUN 1165 in single Use-dependent effects of pirmenol on Vmax and conduction in myocytes isolated from guinea-pig atrium. Br J Pharmacol 1989; 98: guinea-pig ventricular myocardium. Br J Pharmacol 1990; 99: 815– 149–154 819 78. Hattori Y, Inomata N: Modes of the Na channel blocking action of 99. Nakaya H, Shishido Y, Oyama Y, Kanno M: Voltage-dependent pilsicainide, a new antiarrhythmic agent, in cardiac cells. Jpn J modification of Vmax recovery from use-dependent block by Pharmacol 1992; 58: 365–373 pirmenol in guinea pig papillary muscles: comparison with other 79. Anno T, Toyama J: State-dependent block of SUN 1165 on Vmax of Class I drugs. J Cardiovasc Pharmacol 1992; 19: 140–147 guinea-pig cardiac myocytes (in Japanese). Annu Rep Res Inst 100. Nakaya H, Hattori Y, Endou M, Gandou S, Kanno M: Environ Med, Nagoya Univ 1994; 15: 182–184 Electrophysiologic and anticholinergic effects of pirmenol 80. Inomata N, Ishihara T, Akaike N: Mechanisms of the anticholinergic enatiomers in guinea-pig myocardium. Naunyn Schmiedebergs Arch effect of SUN 1165 in comparison with flecainide, disopyramide and Pharmacol 1992; 346: 555–562 quinidine in single atrial myocytes isolated from guinea-pig. Br J 101. Nawada T, Tanaka Y, Hirai S, Hisatome I, Hasegawa J, Kotake H, Pharmacol 1991; 104: 1007–1011 Mashiba H: Evaluation of negative inotropic and antiarrhythmic 81. Inomata N, Ohno T, Ishihara T, Akaike N: Antiarrhythmic agents act effects of class 1 antiarrhythmic drugs. Int J Clin Pharmacol Ther differently on the activation phase of the ACh-response in guinea-pig 1994; 32: 347–355 atrial myocytes. Br J Pharmacol 1993; 108: 111–115 102. Kodama I, Suzuki R, Honjo H, Toyama J: Electrophysiological 82. Hirotsu I, Kihara T, Nakamura S, Hattori Y, Hatta M, Kitakaze Y, et effects of diprafenone, a dimethyl congener of propafenone on al: General pharmacological studies on N-(2,6-dimethylphenyl)-8- guinea-pig ventricular cells. Br J Pharmacol 1992; 107: 813–820 pyrrolizidine-acetamide hydrochloride hemihydrate. 1st communica- 103. Hondeghem LM, Katzung BG: Antiarrhythmic agents: The modu- tion: effect on the central nervous system. Arzneimittelforschung/ lated receptor mechanism of action of sodium and calcium channel- Drug Res 1988; 38: 1398–1404 blocking drugs. Annu Rev Pharmacol Toxicol 1984; 24: 387–424 83. Hirotsu I, Kihara T, Hattori Y, Hatta M, Hirose N, Ishihara T, et al: 104. Grant AO, Starmer CF, Strauss HC: Antiarrhythmic drug action: General pharmacological studies on N-(2,6-dimethylphenyl)-8- Blockade of the inward sodium current. Circ Res 1984; 55: 427–439 pyrrolizidine-acetamide hydrochloride hemihydrate. 1st communica- 105. Carmeliet E: Activation block and trapping of penticainide, a disopy- tion: effect on the peripheral nervous system and peripheral organs. ramide analogue, in the Na+ channel of rabbit cardiac Purkinje fibers. Arzneimittelforschung/Drug Res 1988; 38: 1410–1417 Circ Res 1988; 63: 50–60 84. Niitani H, Katoh K, Sugimoto T, Iinuma H, Katagiri T, Hiejima K, et 106. Gruber R, Carmeliet E: The activation gate of the sodium channel al: Clinical evaluation of SUN 1165 in patients with ventricular controls blockade and deblockade by disopyramide in rabbit Purkinje premature contractions: A multicenter double-blind comparative fibers. Br J Pharmacol 1989; 97: 41–50 study with disopyramide (in Japanese). Cardioangiology 1989; 26: 107. Anno T, Hondeghem LM: Interaction of flecainide with guinea pig 216–232 cardiac sodium channels. Importance of activation unblocking to the 85. Niitani H, Katoh K, Sugimoto T, Iinuma H, Katagiri T, Hiejima K, et voltage dependence of recovery. Circ Res 1990; 66: 789–803 al: Clinical evaluation of SUN 1165 in patients with supraventricular 108. Kotake H, Hirai S, Kinugawa T, Ito T, Yamasaki J, Hasegawa J, et premature contractions: A multicenter double-blind comparative al: Effects of pirmenol hydrochloride on the spontaneous action study with disopyramide (in Japanese). Cardioangiology 1989; 26: potentials and membrane current systems of rabbit sinoatrial node 201–215 cells. J Electrocardiol 1988; 21: 355–359 86. Ino T, Atarashi H, Saitoh H: Electrophysiologic and hemodynamic 109. Nakajima T, Hazama H, Yamashita T, Iwasawa K, Hamada E, effects of SUN 1165, a new antiarrhythmic agent. Curr Clin Pract Omata M: Effects of pirmenol on action potentials and membrane Series 1989; 56: 219–221 currents in single atrial myocytes. Eur J Pharmacol 1998; 344: 287– 87. Terazawa T, Suzuki M, Goto T, Kato R, Hayashi H, Ito A, et al: 297 Suppressive effect of SUN 1165 on supraventicular tachycardia. Am 110. Kaplan HR, Mertz TE, Steffe TJ: Preclinical pharmacology of Heart J 1991; 121: 1437 pirmenol. Am J Cardiol 1987; 59: 2H–9H 88. Satoh S, Watanabe J, Keitoku M, Kinoshita M, Sekiguchi N, Endo S, 111. Endou M, Gando S, Hattori Y, Kanno M: Binding profiles of class I et al: Effects of N-(2,6-dimethylphenyl)-8-pyrrolizidine acetamide antiarrhythmic agents to cardiac muscarinic receptors: Competitive hydrochloride hemihydrate on the ventriculo-atrial conductivity of and allosteric interactions with the receptors and their pharmacologi- accessory pathways. Arzneimittelforschung/Drug Res 1989; 39: 908 cal significance. J Pharmacol Exp Ther 1991; 258: 992–998 –911 112. Endou M, Hattori Y, Gando S, Kanno M: Selectivity of class I antiar- 89. Atarashi H, Inoue H, Hiejima K, Hayakawa H, for the PSTAF inves- rhythmic agents, disopyramide, pirmenol, and pentisomide for tigators: Conversion of recent-onset fibrillation by a single oral dose peripheral muscarinic M2 and M3 receptors. J Cardiovasc Pharmacol

Japanese Circulation Journal Vol.63, January 1999 12 KODAMA I et al.

1992; 19: 674–681 patients: A comparison with lidocaine. Eur Heart J 1986; 7: 150– 113. Bonner TI: The molecular basis of muscarinic receptor diversity. 157 Trends Neurosci 1989; 12: 148–151 120. Grant AO, Whalley DW, Wendt DJ: Pharmacology of the cardiac 114. Doods HN, Mathy M-J, Davidesko D, van Charldorp KJ, de Jonge sodium channel. In: Zipes DP, Jalife J, editors. Cardiac electrophysi- A, van Zwieten PA: Selectivity of muscarinic antagonists in radioli- ology: From cell to bedside (II). Philadelphia: WB Saunders, 1995: gand and in vivo experiments for the putative M1, M2 and M3 recep- 247–259 tors. J Pharmacol Exp Ther 1987; 242: 257–262 121. Tseng G-Y: Potassium channels: Their modulation by drugs. In: 115. Anderson JL, Lutz JR, Nappi JM: Pirmenol for control of ventricular Zipes DP, Jalife J, editors. Cardiac electrophysiology: From cell to arrhythmias: Oral dose-ranging and short-term maintenance study. bedside (II). Philadelphia: WB Saunders, 1995: 260–269 Am J Cardiol 1984; 53: 522–527 122. Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA: Atrial fibrilla- 116. Toivonen LK, Nieminen MS, Manninen V, Frick MH: Antiarrhythmic tion begets atrial fibrillation: A study in awake chronically instru- efficacy of pirmenol in the treatment of premature ventricular mented goats. Circulation 1995; 92: 1954–1968 complexes. Eur Heart J 1985; 6: 737–744 123. Le Grand B, Hatem SD, Deroubaix E, Couetil JP, Coraboeuf E: 117. Phillips HR, Stack RS, Davis DD, Morris KG, Lee KL, Shand DG, et Depressed transient outward and calcium currents in dilated human al: Hemodynamic effects of the antiarrhythmic drug pirmenol. Clin atria. Cardiovasc Res 1994; 28: 548–556 Pharmacol Ther 1982; 32: 235–239 124. Goette A, Honeycutt C, Langberg JJ: Electrical remodeling in atrial 118. Salerno DM, Fifield J, Farmer C, Hodges M: Pirmenol. An antiar- fibrillation. Circulation 1996; 94: 2968–2974 rhythmic drug with unique electrocardiographic features - Double- 125. Van Wagoner DR, Pond AL, McCarthy PM, Trimmer JS, Nerbonne blind placebo-controlled comparison with quinidine. Clin Cardiol JM: Outward K+ current densities and Kv1.5 expression are reduced 1991; 14: 25–32 in chronically human atrial fibrillation. Circ Res 1997; 80: 772–781 119. Nieminen MS, Toivonen LK, Menninen V, Frick MH: Acute haemo- 126. Kodama I, Kamiya K, Toyama J: Cellular electropharmacology of dynamic effects of the antiarrhythmic agent pirmenol in cardiac amiodarone. Cardiovasc Res 1997; 35: 13–29

Japanese Circulation Journal Vol.63, January 1999