The anaesthetized rabbit with acute atrioventricular block provides a new model for detecting drug-induced Torsade de Pointes

Transcription

1 British Journal of Pharmacology RESEARCH PAPER British Journal of Pharmacology (2017) The anaesthetized rabbit with acute atrioventricular block provides a new model for detecting drug-induced Torsade de Pointes Correspondence Akira Takahara, Department of Pharmacology and Therapeutics, Faculty of Pharmaceutical Sciences, Toho University, Miyama, Funabashi, Chiba , Japan. akirat@phar.toho-u.ac.jp Received 18 October 2016; Revised 3 May 2017; Accepted 4 May 2017 Mihoko Hagiwara, Seiji Shibuta, Kazuhiro Takada, Ryuichi Kambayashi, Misako Nakajo, Megumi Aimoto, Yoshinobu Nagasawa and Akira Takahara Department of Pharmacology and Therapeutics, Faculty of Pharmaceutical Sciences, Toho University, Chiba, Japan BACKGROUND AND PURPOSE Several rabbit proarrhythmia models have been developed using genetic or pharmacological methods to suppress the slow component of delayed rectifier K + currents in the ventricle, leading to reduction of the repolarization reserve. Here we have characterized a novel rabbit in vivo proarrhythmia model with severe bradycardia caused by acute atrioventricular block (AVB). EXPERIMENTAL APPROACH Bradycardia was induced in isoflurane-anaesthetized rabbits by inducing AVB with catheter ablation, and the ventricle was electrically driven at 60 beats min 1 throughout the experiment except when extrasystoles appeared. We assessed the effects of two antiarrhythmics, two quinolone antibiotics and one antipsychotic drug, which were chosen as positive drugs (dofetilide, sparfloxacin and haloperidol) and negative drugs (amiodarone and moxifloxacin) for induction of Torsades de Pointes (TdP). KEY RESULTS In our model, TdP arrhythmias appeared with high reproducibility after i.v. dofetilide ( μg kg 1 )infive out of six rabbits, sparfloxacin (30 mg kg 1 ) in three out of six rabbits and haloperidol (0.3 3 mg kg 1 ) in two out of six rabbits. The lethal arrhythmias repeatedly appeared and were accompanied with prolongation of the QT interval and early afterdepolarization-like phenomena. Neither amiodarone ( mg kg 1, n =6)normoxifloxacin (3 30 mg kg 1, n = 6) induced such arrhythmias, even when QT intervals were prolonged. CONCLUSIONS AND IMPLICATIONS These results suggest that our model of the unremodelled and bradycardic heart of the anaesthetized rabbit is a useful test system for the detection of drug-induced TdP arrhythmias. Abbreviations AVB, atrioventricular block; BP, blood pressure; ECG, electrocardiogram; I Kr, rapid component of delayed rectifier K + currents; I Ks, slow component of delayed rectifier K + currents; LTV, long-term variability; MAP, monophasic action potential; MAP 90, duration of MAP signal at 90% repolarization level; STV, short-term variability; TdP, Torsades de Pointes 2017 The British Pharmacological Society DOI: /bph.13870

2 M Hagiwara et al. Introduction Drug-induced prolongation of the cardiac QT interval is often associated with the onset of Torsade de Pointes (TdP), resulting in a life-threatening ventricular arrhythmia (Frommeyer and Eckardt, 2016). Most of the drugs that induce TdP have been shown to potently inhibit the rapid component of delayed rectifier K + currents (I Kr ), carried by K v 11.1 channels. This inhibition leads to marked prolongation of the ventricular repolarization (Frommeyer and Eckardt, 2016). It has been suggested that early afterdepolarizations (EADs) and triggered activity occur due to excessive QT-interval prolongation, which is followed by EAD-related TdP arrhythmias and deterioration into ventricular fibrillation. Recently, in vivo proarrhythmia models in animals have been developed to detect the arrhythmogenic potential of compounds during pre-clinical assessments (Thomsen et al., 2006a; Sugiyama, 2008), because the degree of repolarization prolongation does not necessarily correlate with subsequent ventricular arrhythmia development. Among them, the rabbit proarrhythmia model was first established by Carlsson et al. in 1990 for investigating the development of TdP in vivo. Inthismodel,theα 1 -adrenoceptor agonist methoxamine is given i.v. to increase intracellular Ca 2+ concentration, leading to increased triggered activity and increased susceptibility to ventricular arrhythmias (Carlsson et al., 1990; Carlsson, 2008). Recently, other rabbit proarrhythmia models have been developed using genetic or pharmacological methods to suppress the slow component of the delayed rectifier K + currents (I Ks ),leadingto reduction of the repolarization reserve, due to the redundant nature of myocardial repolarization capacity (Roden, 1998; Husti et al., 2015; Major et al., 2016). In the normal rabbit heart, whose sinoatrial rate is usually beats min 1 in vivo, contribution of the I Ks to the ventricular repolarization may be higher than that in human, because the slow deactivation kinetics of I Ks allows it to accumulate at higher frequencies (Stengl et al., 2003). The present study was designed to create and characterize a newly developed proarrhythmia rabbit model with severe bradycardia, which is expected to exhibit reduced repolarization reserve due to smaller contribution of I Ks (Biliczki et al., 2002). Bradycardia was induced by atrioventricular block (AVB) using catheter ablation, and experiments were performed within 5 h after inducing the AVB (acute AVB). The procedure is essentially different from the well-established experimental system, using a canine proarrhythmia model of chronic AVB, which requires 4 weeks of remodelling after the AVB (Vos et al., 1998; Sugiyama et al., 2002). Methods All animal care and experimental procedures were approved by Toho University Animal Care and User Committee (approval number: ), and all efforts were made to minimize suffering. Animal studies are reported in compliance with the ARRIVE guidelines (Kilkenny et al., 2010; McGrath and Lilley, 2015). A total of 36 male New Zealand White rabbits were obtained from Sankyo Labo Service (Tokyo, Japan), and six rabbits were used in each of six experimental groups. The exact group size (n) foreach experimental group is provided in the figure legends, where n refers to independent values, not replicates. Rabbits were randomly assigned across experimental groups. The experimental animals were young adults (16.1 ± 0.3 weeks old, approximately 3 kg). Rabbits were individually housed in stainless steel rabbit cages on a 12 h light (8:00 20:00 h) dark (20:00 8:00 h) cycle and were kept at standard temperature (23 ± 1 C) and relative humidity (50 60%). Food and water were provided ad libitum. New Zealand White rabbits are widely used for pharmacological studies; moreover, the species and strain have previously been used in cardiac safety pharmacology studies of QT interval-prolonging drugs. General procedures Male New Zealand White rabbits were initially anaesthetized with ketamine hydrochloride (35 mg kg 1, i.m.) and xylazine hydrochloride (5 mg kg 1, i.m.). After tracheal intubation, 1.5% isoflurane vaporized with 100% oxygen was inhaled with a ventilator (SN-480-5; Shinano, Tokyo, Japan). The tidalvolumeandrespiratoryrateweresetat6ml kg 1 and 40 strokes min 1 respectively. The right femoral artery was cannulated with a heparinized catheter (100 U ml 1 ) for measurement of the blood pressure (BP). The surface lead II electrocardiogram (ECG) was continuously monitored using a polygraph system (RMP-6008; Nihon Kohden, Tokyo, Japan). Production of complete AVB For production of complete AVB, catheter ablation was carried out as described previously (Takahara et al., 2006; Hagiwara et al., 2016). A quad-polar electrode catheter with a tip of 1 mm (6 Fr, D6DR252RT; Biosense Webster, Diamond Bar, CA, USA) was inserted through the left femoral vein and positioned at the tricuspid valve by watching the bipolar electrograms from the distal electrode pair. The optimal site for the atrioventricular node ablation, namely, the compact atrioventricular node, was determined on the basis of the intracardiac electrogram, in which a very small His deflection was recorded. The power source for atrioventricular node ablation was an electrosurgical generator (SL-1PR; Semco, Tokyo, Japan) delivering continuous unmodulated radiofrequency energy at a frequency of 500 khz. After proper positioning, the radiofrequency energy of 20 W was delivered for 5 10 s from the tip electrode to the return electrode positioned under the animal s back. The endpoint of this procedure was the development of the complete AVB with an onset of stable idioventricular escaped rhythm. The ventricle was electrically driven with a stimulator (SEC- 4103; Nihon Kohden) at 60 beats min 1 throughout the experiment except when extrasystole appeared. The stimulation pulses were rectangular in shape, of 2 V (approximately twice the threshold voltage) and 3 ms duration. After the production of AVB, stabilization periods of h were needed to recover from hypotension or faster idioventricular rhythm (>60 beats min 1 ). A monophasic action potential (MAP) recording/pacing combination catheter (5 Fr, interelectrode distance 1 mm; Physio-Tech, Tokyo, Japan) was positioned at the right 2592 British Journal of Pharmacology (2017)

3 Drug-induced TdP in acute AVB rabbit BJP ventricle through the right jugular vein. The signal was amplified with a differential amplifier (DAM 50; World Precision Instruments, Sarasota, FL, USA). The duration of MAP signals was measured as an interval (ms) at the 90% repolarization level, which was defined as MAP 90. Electrophysiological and cardiovascular parameters were continuously monitored with a polygraph system (RMP-6008; Nihon Kohden) and analysed with a fully automatic analysis system (ecgauto; Emka Technologies, Paris, France). Experimental protocol Electrophysiological and cardiovascular parameters were continuously recorded under the ventricular pacing at 60 beats min 1. After the basal control assessment (C), a low dose of 1 μg kg 1 of dofetilide was infused over 10 min via the left femoral vein using an infusion pump (PHD 2000 Infusion; Instech Laboratories, Plymouth Meeting, PA, USA), and changes in the BP, ECG and MAP 90 were continuously monitored until 30 min after the start of the infusion. Next, a middle dose of 10 μg kg 1 was additionally infused over 10 min, and each parameter was similarly monitored. Finally, a high dose of 100 μg kg 1 was additionally infused over 10 min, and the haemodynamic and electrophysiological changes were monitored until 60 min after the start of the infusion. Similarly, effects of amiodarone (0.3, 3 and 10 mg kg 1 ), sparfloxacin (3 and 30 mg kg 1 ), moxifloxacin (3 and 30 mg kg 1 ) or haloperidol (0.03, 0.3 and 3 mg kg 1 ) were assessed in another series of animals. Each drug was infused over 10 min, except for amiodarone, which was administered over 30 s, as previously reported (Sugiyama et al., 2001a). Each data of the BP, ECG and MAP 90 was obtained from the mean of three recordings, which was assessed 5, 10, 15, 20, 25 and 30 min after the low- and middle-dose administration and 5, 10, 15, 20, 25, 30, 40, 50 and 60 min after the high-dose administration. After termination of the last recordings, an overdose of sodium pentobarbital was administered i.v. to the anaesthetised animal. TdP was defined as a polymorphic ventricular tachycardia, where the QRS complex twisted around the baseline for five or more consecutive beats. Ventricular tachycardia was defined as lasting greater than or equal to five consecutive beats. Premature ventricular contraction with R on T phenomenon was defined by prematurity index (Adgey, 1983), which was calculated by dividing the coupling interval of the premature ventricular contraction (RR 0 ) by the QT interval of the preceding normally conducted beat (=RR 0 /QT). A premature ventricular contraction with a prematurity index of <1 was considered to represent the R on T phenomenon. Premature ventricular contractions with R on T phenomenon, associated with EAD-induced triggered activity, have been knowntobeabletoinitiatethedevelopmentoftdpwhen transmural dispersion of repolarization is exaggerated (Yan et al., 2001). Beat-to-beat analysis For assessment of instability of the ventricular repolarization, the MAP 90 of 31 consecutive beats was measured before and immediately after complete infusion of a drug (or before induction of the first TdP) (Takahara et al., 2006; Hagiwara et al., 2016). Poincaré plots of MAP 90 (n) versus MAP 90 (n + 1) were prepared for each analysis time point. The mean orthogonal distance from the diagonal to the points of the Poincaré plot was determined as short-term variability (STV) (=Σ MAP 90 (n + 1) MAP 90 (n) /[30 2]). Conversely, the mean distance to the mean of the parameter parallel to the diagonal of the Poincaré plot was determined as long-term variability (LTV) (=Σ MAP 90 (n + 1) + MAP 90 (n) 2MAP 90 (mean) /[30 2]). Anaesthetized methoxamine-sensitized rabbit model The anaesthetized methoxamine-sensitized rabbit model (Carlsson et al., 1990) was prepared, with minor modifications. Male New Zealand White rabbits were initially anaesthetized with ketamine hydrochloride (35 mg kg 1, i.m.) and xylazine hydrochloride (5 mg kg 1, i.m.), and 1.5% isoflurane vaporized with 100% oxygen was inhaled with a ventilator (SN-480-5; Shinano). After preparation and instrumentation, baseline recordings were obtained during the 10 min stabilization period. Subsequently, a continuous infusion of methoxamine (15 μg kg 1 min 1 ) was started. After 10 min, dofetilide (1, 10 and 100 μg kg 1 ) was infused concomitantly using the same administration protocol for the acute AVB rabbit. Corrected QT interval was calculated using a formula developed for rabbits [QTc = QT (RR 300)] (Carlsson et al., 1993; Varkevisser et al., 2015). Data and statistical analysis The data and statistical analysis comply with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2015). The raw data were assessed independently by two co-authors to ensure the correctness of conclusions. Blinding was not used since all the measurements were not subjective but strictly quantitative data. Data are presented as the means ± SEM. Statistical data analysis was carried out using SuperANOVA (ver. 1.11; Abacus Concepts, Inc., Berkeley, CA, USA) or GraphPad Prism (ver. 6.01; GraphPad Software, Inc., La Jolla, CA, USA). The statistical significances within a parameter were evaluated by one-way repeated-measures ANOVA followed by contrasts for mean value comparison to compare each of a number of treatments with a pre-drug control. The statistical significances of parameters between models were evaluated by two-way ANOVA followed by post hoc Bonferroni s test. A P-value <0.05 was considered significant. Materials Dofetilide (MW= ), amiodarone hydrochloride (MW = ), sparfloxacin (MW = ), haloperidol (MW = ) and methoxamine hydrochloride (MW = ) were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA); moxifloxacin hydrochloride (MW = ) was purchased from Ark Pharm, Inc. (Libertyville, IL, USA). Dofetilide was dissolved in 10 mm British Journal of Pharmacology (2017)

4 M Hagiwara et al. HCl, which was i.v. infused at a rate of 0.1 ml kg 1 min 1. Amiodarone was dissolved in distilled water with a heating device, which was i.v. administered at a volume of 0.4 ml kg 1. Sparfloxacin, moxifloxacin or haloperidol was dissolved in 0.5% lactate solution, which was i.v. infused at a rate of 0.2 ml kg 1 per min. Heparin sodium, ketamine hydrochloride, xylazine hydrochloride and isoflurane were purchased from AY Pharmaceuticals (Tokyo, Japan), Daiichi Sankyo (Tokyo, Japan), Bayer Yakuhin (Tokyo, Japan) and Mylan Seiyaku (Osaka, Japan), respectively. Nomenclature of targets and ligands Key protein targets and ligands in this article are hyperlinked to corresponding entries in guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Southan et al., 2016), and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 (Alexander et al., 2015a,b,c). Results Comparison of torsadogenic action of antiarrhythmic drugs Figure 1 shows typical tracings of dofetilide-induced TdP arrhythmias in the acute AVB rabbit, obtained after the infusion of 10 μg kg 1. The arrhythmias followed a premature ventricular contraction that occurred during the preceding T wave, known as R on T phenomenon. The time courses of onset of arrhythmias after the administration of dofetilide and amiodarone as well as their haemodynamic and electrophysiological effects are summarized in Figure 2. In the dofetilide group (n = 6), premature ventricular contractions with R on T phenomenon were observed in three out of six animals at the low dose, five out of six animals at the middle dose and all animals at the high dose. Episodes of TdP were detected in four out of six animals at the middle dose Figure 1 Typical tracings of ECG showing dofetilide-induced TdP in the acute AVB rabbit. (A) Compressed tracing of the surface lead II ECG after the administration of dofetilide at 10 μg kg 1. Black inverted triangles represent the onset of TdP. In this rabbit receiving dofetilide, seven episodes of short duration (<10 s) of TdP and one episode of longer duration ( 10 s) of the arrhythmia were observed after the middle dose, and eight episodes of short duration of TdP and nine episodes of longer duration of the arrhythmia were observed after the high dose. (B) A magnified figure of ECG, MAP and BP signals. Dofetilide-induced TdP was observed at 3 min 19 s in this rabbit British Journal of Pharmacology (2017)

5 Drug-induced TdP in acute AVB rabbit BJP Figure 2 Proarrhythmic and electrophysiological effects of dofetilide (n = 6) and amiodarone (n = 6) in the acute AVB rabbit. (A) Summary of proarrhythmic effects of dofetilide (left panel) and amiodarone (right panel). Each coloured bar indicates the responses of each rabbit to drugs, according to the severity of arrhythmias developed in each minute. Red bars indicate the occurrence of ventricular tachycardia (VT) including TdP. Orange bars indicate the occurrence of premature ventricular contraction (PVC) with the R on T phenomenon. (B) Time course of the effects of dofetilide and amiodarone on the QT interval, MAP duration at 90% repolarization level (MAP 90 ), mean BP (MBP) and atrial rate. In the dofetilide group (left panels), the pre-drug control value (C) of the QT interval, MAP 90, MBP and atrial rate was 285±14,240±12ms,45±2mmHgand249±12beatsmin 1, respectively, whereas that in the amiodarone group (right panels) was 316±20,259±16ms,37±3mmHgand220±9beatsmin 1, respectively. In the dofetilide group, the QT interval and MAP 90 were significantly prolonged at the dose of 10 and 100 μg kg 1. The maximal changes in the MAP 90 at 1, 10 and 100 μg kg 1 from pre-drug controlvalue(c)were+48±6,+214±46and+214±51ms,respectively.themaximalchangesintheexperimentalgroupreceiving amiodarone at 0.3, 3 and 10 mg kg 1 were +34 ± 5, +52 ± 5 and +63 ± 12 ms, respectively. Data are presented as mean ± SEM. The filled symbols represent significant differences (P < 0.05) from each pre-drug control value (C). and five out of six animals at the high dose. In the amiodarone group (n = 6), neither premature ventricular contractions with R on T phenomenon nor episode of TdP was observed during the observation period. As shown in Figure 2B, the QT interval and MAP 90 were prolonged by dofetilide and amiodarone. The extent of prolongation of the QT interval and MAP 90 by dofetilide were greater than those by amiodarone. Dofetilide and amiodarone decreased the atrial rate. Comparison of torsadogenic action of antimicrobial drugs Figure 3 shows typical tracings of TdP arrhythmias in our model, obtained after the administration of sparfloxacin (30 mg kg 1 ). The arrhythmias followed a premature ventricular contraction that occurred during the preceding T wave. The time courses of onset of arrhythmias after the administration of sparfloxacin and moxifloxacin as well as their haemodynamic and electrophysiological effects are British Journal of Pharmacology (2017)

6 M Hagiwara et al. Figure 3 Typical tracings of ECG showing sparfloxacin-induced TdP in the acute AVB rabbit. (A) Compressed tracing of the surface lead II ECG after the administration of sparfloxacin at 30 mg kg 1. Black inverted triangles represent the onset of TdP. In this rabbit receiving sparfloxacin at 30 mg kg 1, one episode of short duration (<10 s) of TdP and one episode of longer duration ( 10 s) of the arrhythmia were observed. (B) A magnified figure of ECG, MAP and BP signals. Sparfloxacin-induced TdP was observed at 37 min 0 s in this rabbit. summarizedinfigure4.inthesparfloxacin group (n =6), premature ventricular contractions with R on T phenomenon were observed in three out of six animals at the low dose and four out of six animals at the high dose. Episodes of TdP were detected in three out of six animals at the high dose, and one animal died after the latest TdP that degenerated into ventricular fibrillation. In the moxifloxacin group (n =6), one premature ventricular contraction with R on T phenomenon was observed in one out of six animals at the high dose. As shown in Figure 4B, the QT interval and MAP 90 were prolonged by sparfloxacin and moxifloxacin in a dose-dependent manner. The prolongation of the QT interval and MAP 90 by sparfloxacin was greater than those induced by moxifloxacin. Sparfloxacin and moxifloxacin decreased the atrial rate. Effects of antiarrhythmic and antimicrobial drugs on beat-to-beat variability of ventricular repolarization Typical examples of effects of dofetilide, amiodarone, sparfloxacin and moxifloxacin on the Poincaré plots of the MAP 90 are shown in Figure 5; the beat-to-beat variability of the MAP 90 before and after the drug administration was summarized in Table 1. No significant change in the STV was detected after the administration of dofetilide or amiodarone. However, the STV after the administration of dofetilide at the dose of 100 μg kg 1 was greater than its control values in some animals, although not signficantly overall. Sparfloxacin and moxifloxacin significantly increased the STV at the high dose, which were accompanied with significant prolongation of the MAP 90. Torsadogenic activity of haloperidol Figure 6A shows typical tracings of haloperidol-induced TdP arrhythmias in the acute AVB rabbit, which was observed after the administration of 3 mg kg 1. The arrhythmias followed a premature ventricular contraction that occurred duringtheprecedingtwave.thetimecoursesofonsetof arrhythmias after the administration of haloperidol as well as its haemodynamic and electrophysiological effects (n =6) are summarized in Figure 6B. Premature ventricular contractions with R on T phenomenon were observed in oneoutofsixanimalsatthemiddledoseandfouroutofsix animals at the high dose. Episode of TdP were observed in one out of six animals at the middle or high dose. The QT interval and MAP 90 were prolonged by haloperidol in a 2596 British Journal of Pharmacology (2017)

7 Drug-induced TdP in acute AVB rabbit BJP Figure 4 Proarrhythmic and electrophysiological effects of sparfloxacin (n = 6) and moxifloxacin (n = 6) in the acute AVB rabbit. (A) Summary of proarrhythmic effects of sparfloxacin (left panel) andmoxifloxacin (right panel). Each coloured bar indicates the responses of each rabbit to drugs, according to the severity of arrhythmias developed in each minute. Red bars indicate the occurrence of ventricular tachycardia (VT) including TdP. Orange bars indicate the occurrence of premature ventricular contraction (PVC) with the R on T phenomenon. The black bar indicates ventricular fibrillation (VF) and death. (B) Time course of the effects of sparfloxacin and moxifloxacin on the QT interval, MAP duration at 90% repolarization level (MAP 90 ), mean BP (MBP) and atrial rate. In the sparfloxacin group (left panels), the pre-drug control value (C) of the QT interval, MAP 90,MBP andatrialratewas303±18,252±23ms,45±3mmhgand249±10beatsmin 1, respectively, whereas that in the moxifloxacin group (right panels)was297±12,239±18ms,35±3mmhgand236±8beatsmin 1, respectively. In the sparfloxacin group, the QT interval and MAP 90 were significantly prolonged at the dose of 3 and 30 mg kg 1. The maximal changes in the MAP 90 at 3 and 30 mg kg 1 from pre-drug control value(c)were+83±7and+232±21ms,respectively.themaximalchangesintheexperimentalgroupreceivingmoxifloxacin at 3 and 30 mg kg 1 were +30 ± 9 and +143 ± 8 ms, respectively. Data are presented as mean ± SEM. The filled symbols represent significant differences (P < 0.05) from each pre-drug control value (C). dose-dependent manner. Haloperidol decreased the mean BP and atrial rate. Typical results of the effects of haloperidol on the Poincaré plots of the MAP 90 are shown in Figure 6C, whereas the beat-to-beat variability of the MAP 90 before and after the drug administration was summarized in Table 2. Haloperidol atthehighdosesignificantly increased the STV, which were accompanied with significant prolongation of the MAP 90. The extent of increase of the STV by haloperidol was greater than those observed in the amiodarone or moxifloxacin group. Relationship between induction of TdP and effects on the beat-to-beat variability of ventricular repolarization Changes in the beat-to-beat variability of the MAP 90 induced by the drugs tested (dofetilide, amiodarone, sparfloxacin, British Journal of Pharmacology (2017)

8 M Hagiwara et al. Figure 5 Typical examples of effects of dofetilide, amiodarone, sparfloxacin and moxifloxacin on the Poincaré plots of the MAP 90 in the acute AVB rabbit. Thirty-one beats were plotted for each of the four analysis time points, before and after the administration of dofetilide (1, 10 and 100 μg kg 1 )or amiodarone (0.3, 3 and 10 mg kg 1 ), and each of the three analysis time points, before and after the administration of sparfloxacin (3 and 30- mg kg 1 )ormoxifloxacin (3 and 30 mg kg 1 ). A schematic diagram is shown in the lower right panel. The mean orthogonal distance from the diagonal to the points of the Poincaré plot is determined as STV. Conversely, the mean distance to the mean of the parameter parallel to the diagonal of the Poincaré plot is determined as LTV. TdP(+), TdP arrhythmias developed in the animal; TdP( ), TdP arrhythmia was not developed in the animal. moxifloxacin and haloperidol) in TdP-inducible rabbits (n = 10) were compared with those in TdP non-inducible rabbits (n = 20) (Table 3). The STV at the first dose (prolongation of MAP 90 by %) significantly increased only in the TdP inducible rabbit, whereas significant increment of STV was detected at the second dose (appearance of initial TdP) in the TdP non-inducible rabbit. Anaesthetized methoxamine-sensitized rabbit model The time courses of onset of arrhythmias after the administration of dofetilide as well as their haemodynamic and electrophysiological effects in the anaesthetized methoxamine-sensitized rabbits (n = 6) are summarized in Figure 7A, B. No premature ventricular contractions with R on T phenomenon were observed at the low dose, whereas they appeared in four out of six animals at the middle and high dose. Episodes of TdP were detected in three out of six animals at the middle dose, and one animal died after the latest TdP that degenerated into ventricular fibrillation. At the high dose, episodes of TdP were detected in two out of six animals. The QT interval and MAP 90 were prolonged by dofetilide in a dose-dependent manner. The second-degree AVB with a 2:1 ratio was observed in all animals at the high dose due to excess prolongation of the QT interval. Frequency of premature ventricular contractions with R on T phenomenon and episodes of TdP during the observation period in the methoxamine-sensitized rabbit is summarized in Table 4, which was significantly less than that in the acute AVB rabbit receiving the same doses of dofetilide, as assessed in this study (Figure 2). Typical results of the effects of dofetilide on the Poincaré plots of the MAP 90 in the anaesthetized methoxaminesensitized rabbit are shown in Figure 7C; the beat-to-beat variability of the MAP 90 before and after the drug administration is summarized in Table 5. Dofetilide 2598 British Journal of Pharmacology (2017)

9 Drug-induced TdP in acute AVB rabbit BJP Table 1 Effects of antiarrhythmic and antimicrobial drugs on the beat-to-beat variability of the ventricular repolarization in the acute AVB rabbit Dofetilide (μg kg 1 ) Control STV(ms) 1.9± ± ± ±3.9 LTV(ms) 1.6± ± ±0.8* 5.3 ± 1.3* MAP 90 (ms) 242 ± ± ± ± 73* Amiodarone (mg kg 1 ) Control STV (ms) 2.0 ± ± ± ± 1.0 LTV (ms) 2.0 ± ± ± ± 1.7* MAP 90 (ms) 262 ± ± 16* 316 ± 21* 321 ± 20* Sparfloxacin (mg kg 1 ) Control 3 30 STV(ms) 2.4± ± ±2.1* LTV(ms) 2.1± ±0.4* 5.3 ± 0.4* MAP 90 (ms) 257 ± ± 27* 421 ± 60* Moxifloxacin (mg kg 1 ) Control 3 30 STV(ms) 2.1± ± ±0.3* LTV(ms) 1.7± ± ±0.2* MAP 90 (ms) 240 ± ± 24* 382 ± 20* Data are presented as mean ± SEM (n =6). *P < 0.05, significantly different from corresponding pre-drug control. significantly increased the STV at the high dose accompanied with significant prolongation of the MAP 90. Discussion We assessed the proarrhythmic effects of dofetilide, amiodarone, sparfloxacin, moxifloxacin and haloperidol in the acute AVB rabbit. Because their clinically relevant doses have been reported to be 1 10 μg kg 1,3mg kg 1,3mg kg 1, 3mg kg 1 and mg kg 1, as an i.v. administration, respectively (Satoh et al., 1999, 2000; Sugiyama et al., 2001a,b; Matsuo et al., 2013), the current experiments were carried out at therapeutic to supra-therapeutic dose ranges. In this study, dofetilide and sparfloxacin could induce TdP arrhythmias with high reproducibility, whereas neither amiodarone nor moxifloxacin induced such arrhythmias even in the condition of prolonged QT interval. As shown in Figures 1 and 3, the lethal arrhythmias appeared repeatedly and were accompanied with the QT interval prolongation and EAD-like phenomena. In our recent studies using acute AVB rabbits, drug-induced TdP could be similarly observed after the administration of a class III antiarrhythmic drug, nifekalant, and atypical antipsychotic drugs, risperidone and paliperidone (Hagiwara et al., 2015, 2016). These results suggest that the acute AVB rabbit model may be useful in detecting drugs with a torsadogenic action. As shown in Figures 2 and 4, the QT interval of the acute AVB rabbit at pre-drug control was around 300 ms during ventricular pacing at 60 beats min 1,whichwaslongerthan action potential duration (APD) of the isolated papillary muscle from rabbits that was electrically driven at 1 Hz (APD 90 = 150 ms; Nakaya et al., 1993). Also, the QT interval of the acute AVB rabbit was longer than that of normal rabbits ( ms), those receiving α 1 -adrenoceptor agonist ( ms), transgenic LQT1 rabbits (241 ms), LQT2 rabbits (264 ms) or LQT5 rabbits (140 ms) during sinus rhythm ( beats min 1 )withoutavb(carlssonet al., 1992; Lu et al., 2001a; Brunner et al., 2008; Hagiwara et al., 2015; Major et al., 2016). The sensitivity of the heart to QTinterval prolonging drugs is modified by the contribution of I Ks in the repolarization phase (Jost et al., 2005;Soet al., 2006), whereas the density of I Ks in the isolated ventricular myocytes from rabbits is lower than those from other species like guinea pigs (Lu et al., 2001b). In addition, I Ks in the ventricle is accumulated in a frequency-dependent manner in the presence of β 1 -adrenoceptor stimulation (Stengl et al., 2003; Nissen et al., 2012). Moreover, anaesthetic concentrations of isoflurane have been reported to inhibit I Ks in guinea pig ventricular myocytes (Suzuki et al., 2003). Thus, the greater bradycardia caused by acute AVB in vivo might reduce contribution of I Ks to ventricular repolarization, leading to a prolonged QT interval, in conjunction with reduction of repolarization reserve (Roden, 1998) in this animal model. To confirm the utility of the acute AVB rabbit model for the detection of drug-induced arrhythmias, we assessed the effects of two antiarrhythmics and two quinolone antibiotics, which were chosen as positive drugs (dofetilide and sparfloxacin) and negative drugs (amiodarone and moxifloxacin) for induction of TdP, based on previous investigations of effects of dofetilide (25 μg kg 1, i.v.), amiodarone (5 mg kg 1, i.v.), sparfloxacin (60 mg kg 1, p.o.) or moxifloxacin (2 and 8 mg kg 1, i.v.) (Verduyn et al., 1999; Chiba et al., 2000; Thomsen et al., 2006b) in the canine proarrhythmia model with chronic AVB (Thomsen et al., British Journal of Pharmacology (2017)

10 M Hagiwara et al. Figure 6 Proarrhythmic and electrophysiological effects of haloperidol (n = 6) in the acute AVB rabbit. (A) Typical compressed tracings of the surface lead II ECG showing haloperidol-induced TdP in the acute AVB rabbit. In the rabbit #3 receiving haloperidol at 3 mg kg 1, one episode of short duration (<10 s) of TdP and one episode of longer duration ( 10 s) of the arrhythmia were observed. (B) Summary of proarrhythmic effects of haloperidol (upper panel). Each coloured bar indicates the responses of each rabbit to drugs, according to the severity of arrhythmias developed in each minute. Red bars indicate the occurrence of ventricular tachycardia (VT) including TdP. Orange bars indicate the occurrence of premature ventricular contraction (PVC) with the R on T phenomenon. Time course of the effects of haloperidol on the QT interval, MAP duration at 90% repolarization level (MAP 90 ), mean BP (MBP) and atrial rate (three lower panels). The pre-drug control value (C) of QT interval, MAP 90,MBP andatrialratewas310±23,258±20ms,42±3mmhgand225±6beatsmin 1, respectively. The QT interval and MAP 90 were significantly prolonged at the dose of 0.3 and 3 mg kg 1. The maximal changes in the MAP 90 at 0.03, 0.3 and 3 mg kg 1 from pre-drug control value (C) were +21 ± 6, +93 ± 19, and +252 ± 34 ms, respectively. Data are presented as mean ± SEM. The filled symbols represent significant differences (P < 0.05) from each pre-drug control value (C). (C) Typical examples of effects of haloperidol on the Poincaré plots of the MAP 90 in the acute AVB rabbit. Thirty-one beats were plotted for each of the four analysis time points, before and after the administration of haloperidol (0.03, 0.3 and 3 mg kg 1 ). TdP(+), TdP arrhythmias developed in the animal. 2006a; Sugiyama, 2008). As shown in Figures 1 and 2, the positive reference compound dofetilide markedly increased the QT interval and number of premature ventricular contractions with R on T phenomenon and induced TdP at 10 to 100 μg kg 1. Similar results were obtained with another positive compound sparfloxacin at 30 mg kg British Journal of Pharmacology (2017)

11 Drug-induced TdP in acute AVB rabbit BJP Table 2 Effects of haloperidol on the beat-to-beat variability of the ventricular repolarization in the acute AVB rabbit Haloperidol (mg kg 1 ) Control STV (ms) 2.2 ± ± ± ± 3.1* LTV (ms) 1.9 ± ± ± ± 2.2* MAP 90 (ms) 258 ± ± ± 32* 481 ± 42* Data are presented as mean ± SEM (n =6). *P < 0.05, significantly different from corresponding pre-drug control. Table 3 Relationship between induction of TdP arrhythmias and effects on the beat-to-beat variability of the ventricular repolarization in the acute AVB rabbit TdP inducible rabbits TdP non-inducible rabbits Control 1st dose 2nd dose Control 1st dose 2nd dose STV (ms) 2.2 ± ± 0.6* 3.4 ± ± ± ± 0.7* LTV (ms) 1.9 ± ± 0.6* 5.0 ± 0.6* 1.8 ± ± 0.2* 3.5 ± 0.4* MAP 90 (ms) 248 ± ± 13* 332 ± 25* 253 ± ± 13* 373 ± 21* Data are obtained from TdP inducible (n = 10) and non-inducible (n = 20) acute AVB rabbits in all experimental groups of dofetilide, amiodarone, sparfloxacin, moxifloxacin and haloperidol. The first dose exerted prolongation of MAP 90 by %, and the second dose showed initial induction of TdP. Data are presented as mean ± SEM. *P < 0.05, significantly different from corresponding pre-drug control. (Figures 3 and 4). In contrast, neither amiodarone (0.3 to 10- mg kg 1 )normoxifloxacin (3 and 30 mg kg 1 ) induced TdP in this study (Figures 2 and 4). These results suggest that the sensitivity of acute AVB rabbit model for detection of druginduced arrhythmias may be comparable with that of the canine chronic AVB model. The ideal model to identify unsafe (TdP-inducing) drugs is one that actually produces TdP arrhythmias at clinically relevant doses and does so with reproducibility and with high sensitivity and specificity. Thus, our rabbit model may have potential for understanding the relationships between clinical proarrhythmic events and observations in pre-clinical models. In an earlier experiment using acute AVB dogs, no proarrhythmic action was detected after the administration of dofetilide (25 μg kg 1, i.v.) (Thomsen et al., 2007). More interestingly, the present study indicates that electrophysiological and anatomical remodelling following persistent bradycardia was not necessary for the onset of TdP in the rabbit heart, which contrasts with the canine heart (Vos et al., 1998; Sugiyama et al., 2002). This observation is partly supported by previous in vitro studies that prolonging effects of a class III antiarrhythmic drug nifekalant on the APD in the isolated heart from normal rabbits were more marked than those in the isolated heart from chronic AVB block dogs (Nakaya et al., 1993; Takahara et al., 2007). Electrical remodelling in the canine chronic AVB heart is closely associated with down-regulation of I Kr and I Ks, an up-regulation of the Na + -Ca 2+ exchange current and increased Ca 2+ release from the sarcoplasmic reticulum (Volders et al., 1999;Sipidoet al., 2000). Similar mechanisms may be involved in the rabbit heart except for effects on the Na + Ca 2+ exchanger, as this exchanger did not contribute to dofetilide-induced TdP in the isolated rabbit heart (Lu et al., 2001b; Farkas et al., 2009). Importantly, the volatile anaesthetic, isoflurane, used in this study has been reported to enhance Ca 2+ -induced Ca 2+ release (Akata et al., 2001), which may itself increase intracellular Ca 2+ and thus prove a limitation in this model. Higher doses of isoflurane could be administered to assess whether or not the anaesthesia is affecting the results. As shown in Figure 5 and Table 1, STV of ventricular repolarization period, a surrogate marker of TdP arrhythmias, markedly increased in rabbits receiving dofetilide and sparfloxacin. In a recent electrophysiological study using canine hypertrophied myocytes, changes in this marker were strongly dependent on Ca 2+ release from the sarcoplasmic reticulum through modulation of the L-type Ca 2+ current in a subsarcolemmal microdomain (Antoons et al., 2015). Similar mechanisms might be involved in the acute AVB rabbit. As summarized in Table 3, the STV may be suggested to be a good predictor of initial TdP at a non-torsadogenic dose, based on recalculated data obtained from TdP inducible and non-inducible rabbits. On the other hand, although the high dose of moxifloxacin did not induce TdP, the STV was British Journal of Pharmacology (2017)

12 M Hagiwara et al. Figure 7 Proarrhythmic and electrophysiological effects of dofetilide in the anaesthetized methoxamine-sensitized rabbits (n = 6). (A) Summary of proarrhythmic effects of dofetilide. Each coloured bar indicates the responses of each rabbit to drugs, according to the severity of arrhythmias developed in each minute. Red bars indicate the occurrence of ventricular tachycardia (VT) including TdP. Orange bars indicate the occurrence of premature ventricular contraction (PVC) with the R on T phenomenon. The black column indicates ventricular fibrillation (VF), and death. (B) Time course of the effects of dofetilide on the QT interval, QTc interval, MAP duration at 90% repolarization level (MAP 90 ), mean BP (MBP), atrial rate and ventricular rate. The pre-methoxamine value (Pre) of the QT interval, QTc interval, MAP 90, MBP and heart rate was 163 ± 8, 168±7,132±7ms,52±2mmHgand222±8beatsmin 1, respectively, whereas 10 min after the start of administration of methoxamine, that is, the pre-dofetilide control value (C), were 175 ± 9, 162 ± 10, 146 ± 10 ms, 86 ± 5 mmhg and 169 ± 16 beats min 1,respectively.TheQT interval, QTc interval and MAP 90 were significantly prolonged at the dose of 10 and 100 μg kg 1. The maximal changes in the MAP 90 at 1, 10 and 100 μg kg 1 from pre-dofetilide control value (C) were +31 ± 5, +151 ± 34 and +294 ± 31 ms, respectively. Data are presented as mean ± SEM. The filled symbols represent significant differences (P < 0.05) from each pre-dofetilide control value (C). (C) Typical examples of effects of dofetilide on the Poincaré plots of the MAP 90 in the anaesthetized methoxamine-sensitized rabbit. Thirty-one beats were plotted for each of the five analysis time points; pre, control and after the administration of dofetilide (1, 10 and 100 μg kg 1 ); TdP(+), TdP arrhythmias developed in the animal. slightly but significantly increased. The MAP 90 prolongation by the high dose of moxifloxacin(240±19to382±20ms; +59.1%) was marked and was comparable to that induced by the high dose of dofetilide (+45.9%). As the STV of ventricular repolarization is calculated as an average of difference between the values of adjacent MAP 90 [MAP 90 (n) and MAP 90 (n + 1)], it is possible that the values of STV might be influenced by potent prolonging effects on the MAP 90. Thus, further investigation into the relationship between changes in the STV and onset of TdP will be needed to assure the utility of the surrogate marker in the acute AVB rabbit model British Journal of Pharmacology (2017)

13 Drug-induced TdP in acute AVB rabbit BJP Table 4 Comparison of occurrence of ventricular arrhythmias in the acute AVB rabbit with that in the methoxamine-sensitized rabbit Number of PVCs (R on T type) Number of episodes of TdP Dofetilide (μg kg 1 ) Acute AVB rabbits 3 ± 2 93 ± ± 343 0±0 4±2 24±14 Methoxamine-sensitized rabbits 0 ± 0 27 ± ± 74 0 ± 0 1 ± 1 0 ± 0 Severity of arrhythmia is expressed as numbers of the PVCs with R on T phenomenon and episodes of TdP during the observation period after the administration of dofetilide at 1, 10 and 100 μg kg 1.Significant differences were detected at the high dose. Data are presented as mean ± SEM from acute AVB rabbits (n = 6) and methoxamine-sensitized rabbits (n =6). PVC, premature ventricular contraction. *P < 0.05, significantly different between the models. Table 5 Effects of dofetilide on the beat-to-beat variability of the ventricular repolarization in the anaesthetized methoxamine-sensitized rabbit Dofetilide (μg kg 1 ) Pre Control STV(ms) 0.5± ± ± ± ±1.3* LTV(ms) 0.5± ± ± ±0.6* 5.5 ± 0.5* MAP 90 (ms) 132±7 146±10 175±12 289±43* 428 ± 35* Data are presented as mean ± SEM (n =6). Pre, before administration of methoxamine. *P < 0.05, significantly different from corresponding control. Although the first established rabbit proarrhythmia model (methoxamine-sensitized rabbit) was widely used for screening of a range of drugs (Carlsson et al., 1992; Carlsson, 2008), there is a limitation in assessing proarrhythmic effects of drugs having pharmacological interaction with α 1 - adrenoceptors. In this study, we assessed the effects of a typical antipsychotic drug haloperidol, which is known to inhibit α 1 -adrenoceptors (ED 50 = 0.41 mg kg 1 ) and dopamine D 2 receptors (ED 50 = 0.14 mg kg 1 ) (Schotte et al., 1996).Thesearedoseswhichareclosetothemiddle dose we have used here (0.3mg kg 1 ). As summarized in Figure 6 and Table 2, haloperidol induced TdP accompanied with marked QT interval prolongation and increased STV, which was essentially in accordance with clinical experience (Hunt and Stern, 1995). These results suggest that the acute AVB rabbit could be used for detection of TdP arrhythmias induced by drugs even with an affinity for α 1 -adrenoceptors, in contrast to the methoxamine-sensitized rabbit model. We further assessed the effects of dofetilide on the methoxamine-sensitized rabbit model to better understand the sensitivity of the acute AVB rabbit to drug-induced arrhythmia. In a recent study, the TdP incidence in the methoxamine-sensitized rabbit was dramatically influenced by anaesthesia (Varkevisser et al., 2015). Based on this information, we used isoflurane as the anaesthetic for the methoxamine-sensitized rabbit model and compared the sensitivity of the rabbit heart to torsadogenic action of dofetilide between the two models. In this study, dofetilide induced TdP in four out of six animals with marked prolongation of MAP 90 in the methoxamine-sensitized rabbit model. However, the number of premature ventricular contractions with R on T phenomenon and TdP arrhythmias by dofetilide was less than those in the acute AVB rabbit (Table 4), suggesting that the sensitivity to drug-induced TdP in the acute AVB rabbit is higher than that in the methoxamine-sensitized rabbit. Therefore, the AVB is a better procedure than the methoxamine administration to create a torsadogenic condition in the rabbit heart. The novelty of this model derives from its ability to reproduce proarrhythmic drug effects and its quantitative comparison with the existing rabbit model. In this study, we confirmed that pharmacological inhibition of I Kr channels plays a key role in inducing R on T phenomenon and TdP arrhythmias in the in vivo rabbit bradycardic heart. However, further assessment of the acute AVB rabbit is needed to better understand the mechanistic insight into each of the positive and negative drugs tested in this study and the potential implications for pharmacological screening and clinical practice. In conclusion, the acute AVB rabbit model as sensitive in detecting drug-induced arrhythmias, as the well-established canine chronic AVB model. Therefore, the unremodelled and bradycardic heart of the anaesthetized rabbit could be used to detect drug-induced TdP arrhythmias. Acknowledgements This study was supported in part by JSPS KAKEN-HI (#15K08598). British Journal of Pharmacology (2017)

14 M Hagiwara et al. Author contributions M.H. performed the research, analysed the data, organized the study and contributed to the writing of the manuscript. S.S. performed the research and analysed the data. K.T. performedtheresearchandanalysedthedata.r.k.performed the research and analysed the data. M.N. performed the research and analysed the data. M.A. contributed to the writing of the manuscript. Y.N. performed the research, analysed the data and contributed to the writing of the manuscript. A.T. designed the research plan, organized the study and contributed to the writing of the manuscript. Conflict of interest The authors declare no conflicts of interest. Declaration of transparency and scientific rigour This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research recommended by funding agencies, publishers and other organisations engaged with supporting research. References Adgey AAJ (1983). Warning arrhythmias in ischemic heart disease. Int J Cardiol 4: Akata T, Nakashima M, Izumi K (2001). Comparison of volatile anesthetic actions on intracellular calcium stores of vascular smooth muscle: investigation in isolated systemic resistance arteries. Anesthesiology 94: Alexander SPH, Catterall WA, Kelly E, Marrion N, Peters JA, Benson HE et al. (2015a). The concise guide to PHARMACOLOGY 2015/16: Voltage-gated ion channels. Br J Pharmacol 172: Alexander SPH, Davenport AP, Kelly E, Marrion N, Peters JA, Benson HE et al. (2015b). The concise guide to PHARMACOLOGY 2015/16: G protein-coupled receptors. Br J Pharmacol 172: AlexanderSPH,KellyE,MarrionN,PetersJA,BensonHE,FaccendaE et al. (2015c). The concise guide to PHARMACOLOGY 2015/16: Transporters. Br J Pharmacol 172: Antoons G, Johnson DM, Dries E, Santiago DJ, Ozdemir S, Lenaerts I et al. (2015). Calcium release near L-type calcium channels promotes beat-to-beat variability in ventricular myocytes from the chronic AV block dog. J Mol Cell Cardiol 89: BiliczkiP,VirágL,IostN,PappJG,VarróA(2002).Interactionof different potassium channels in cardiac repolarization in dog ventricular preparations: role of repolarization reserve. Br J Pharmacol 137: BrunnerM,PengX,LiuGX,RenXQ,ZivO,ChoiBRet al. (2008). Mechanisms of cardiac arrhythmias and sudden death in transgenic rabbits with long QT syndrome. J Clin Invest 118: Carlsson L, Almgren O, Duker G (1990). QTU-prolongation and torsades de pointes induced by putative class III antiarrhythmic agents in the rabbit: etiology and interventions. J Cardiovasc Pharmacol 16: Carlsson L (2008). The anaesthetised methoxamine-sensitised rabbit model of torsades de pointes. Pharmacol Ther 119: Carlsson L, Abrahamsson C, Andersson B, Duker G, Schiller-Linhardt G (1993). Proarrhythmic effects of the class III agent almokalant: importance of infusion rate, QT dispersion, and early afterdepolarisations. Cardiovasc Res 27: Carlsson L, Abrahamsson C, Drews L, Duker G (1992). Antiarrhythmic effects of potassium channel openers in rhythm abnormalities related to delayed repolarization. Circulation 85: Chiba K, Sugiyama A, Satoh Y, Shiina H, Hashimoto K (2000). Proarrhythmic effects of fluoroquinolone antibacterial agents: in vivo effects as physiologic substrate for torsades. Toxicol Appl Pharmacol 169: Curtis MJ, Bond RA, Spina D, Ahluwalia A, Alexander SP, Giembycz MA et al. (2015). Experimental design and analysis and their reporting: new guidance for publication in BJP. Br J Pharmacol 172: Farkas AS, Makra P, Csík N, Orosz S, Shattock MJ, Fülöp F et al. (2009). TheroleoftheNa + /Ca 2+ exchanger, I Na and I CaL in the genesis of dofetilide-induced torsades de pointes in isolated, AV-blocked rabbit hearts. Br J Pharmacol 156: Frommeyer G, Eckardt L (2016). Drug-induced proarrhythmia: risk factors and electrophysiological mechanisms. Nat Rev Cardiol 13: Hagiwara M, Kambayashi R, Aimoto M, Nagasawa Y, Takahara A (2016). In vivo analysis of torsadogenic potential of an antipsychotic drug paliperidone using the acute atrioventricular block rabbit as a proarrhythmia model. J Pharmacol Sci 132: Hagiwara M, Kondo N, Chiba T, Takahara A (2015). Proarrhythmic properties of the atrioventricular block heart of the rabbit. Jpn J Electrocardiology 35: Hunt N, Stern TA (1995). The association between intravenous haloperidol and torsades de pointes. Three cases and a literature review. Psychosomatics 36: Husti Z, Tábori K, Juhász V, Hornyik T, Varró A, Baczkó I (2015). Combined inhibition of key potassium currents has different effects on cardiac repolarization reserve and arrhythmia susceptibility in dogs and rabbits. Can J Physiol Pharmacol 93: Jost N, Virág L, Bitay M, Takács J, Lengyel C, Biliczki P et al. (2005). Restricting excessive cardiac action potential and QT prolongation: a vital role for I Ks in human ventricular muscle. Circulation 112: Kilkenny C, Browne W, Cuthill IC, Emerson M, Altman DG (2010). Animal research: reporting in vivo experiments: the ARRIVE guidelines. Br J Pharmacol 160: Lu HR, Remeysen P, Somers K, Saels A, De Clerck F (2001a). Female gender is a risk factor for drug-induced long QT and cardiac arrhythmias in an in vivo rabbit model. J Cardiovasc Electrophysiol 12: Lu Z, Kamiya K, Opthof T, Yasui K, Kodama I (2001b). Density and kinetics of I Kr and I Ks in guinea pig and rabbit ventricular myocytes explain different efficacy of I Ks blockade at high heart rate in guinea pig and rabbit: implications for arrhythmogenesis in humans. Circulation 104: MajorP,BaczkóI,HiripiL,OdeningKE,JuhászV,KohajdaZet al. (2016). A novel transgenic rabbit model with reduced repolarization 2604 British Journal of Pharmacology (2017)