Plays an Important Role in Drug-Induced Cardiac Arrhythmias: Beyond QT-Prolongation and Torsades De Pointes (Tdps)

Journal of Pharmacological and Toxicological Methods 68 (2013) 250–259

Contents lists available at ScienceDirect

Journal of Pharmacological and Toxicological Methods

journal homepage: www.elsevier.com/locate/jpharmtox

Original article A new biomarker – index of Cardiac Electrophysiological Balance (iCEB) – plays an important role in drug-induced cardiac arrhythmias: beyond QT-prolongation and Torsades de Pointes (TdPs)

Hua Rong Lu a,⁎, Gan-Xin Yan b, David J. Gallacher a a Janssen Research and Development, Janssen Pharmaceutica NV, Belgium b Main Line Health Heart Center and Lankenau Institute for Medical Research, Wynnewood, PA, USA article info abstract

Article history: Introduction: In the present study, we investigated whether a new biomarker – index of cardiac electro- Received 9 November 2012 physiological balance (iCEB=QT/QRS) – could predict drug-induced cardiac arrhythmias (CAs), including ven- Accepted 5 January 2013 tricular tachycardia/ventricular fibrillation (VT/VF) and Torsades de Pointes (TdPs). Methods: The rabbit left ventricular arterially-perfused-wedge was used to investigate whether the simple iCEB measured from the Keywords: ECG is reflective of the more difficult measurement of λ (effective refractory period×conduction velocity) iCEB (index of Cardiac Electrophysiological for predicting CAs induced by a number of drugs. Results: Dofetilide concentration-dependently increased Balance) iCEB and λ, predicting potential risk of drug-induced incidence of early afterdepolarizations (EADs) starting Drug-induced arrhythmias μ μ μ μ QRS at 0.01 M. Digoxin (1 and 5 M), encainide (5 and 20 M) and propoxyphene (10 and 100 M) markedly re- QT duced both iCEB and λ, predicting their ability to induce non-TdP-like VT/VF. At 10 μM, both NS1643 and VT/VF: ventricular tachycardia/fibrillation levcromakalim significantly decreased λ and iCEB, which was preceded with presence of non-TdP-like VT/VF. TdPs: Torsades de Pointes Isoprenaline (0.05 to 0.5 μM) significantly reduced both λ and iCEB, which was associated with a high incidence of non-TdP-like VT/VF in most preparations. Other biomarkers (i.e. transmural dispersion of T-wave and instability of the QT interval) predicted only dofetilide-induced long QT and EADs, but did not predict drug-induced risk of non-TdP-like VT/VF. Discussion: Our data from 7 reference drugs of known pro-arrhythmic effects suggests that 1) this non-invasive iCEB predicts potential risk of drug-induced CAs beyond long QT and TdP; 2) iCEB is more useful than the current biomarkers (i.e. transmural dispersion and instability) in predicting potential risks for drug-induced non-TdP-like VT/VF. © 2013 Elsevier Inc. All rights reserved.

1. Introduction use of a broad range of cardiovascular and non-cardiovascular drugs (Cubeddu, 2003; Pugsley, Authier, & Curtis, 2008; Shah, 2004). Cur- Drug-induced QT prolongation and the appearance of Torsade de rent and past established regulatory guidelines (CPMP/986/96 points Pointes (TdP) are recognized as a potential risk associated with the to consider document, 1997 and ICH S7B, 2005) recommend the con- duct of preclinical studies both in vivo and in vitro to detect drug- induced QT-interval prolongation and arrhythmogenic potential. The preclinical guidance was followed more recently by the ICH E14 rec- Abbreviations: BCL, Basic cycle lengths; CAST, Cardiac Arrhythmia Suppression ommendation for conducting thorough QT (TQT) investigations in Trial; CPMP, committee for proprietary medicinal products; CV, conduction velocity; early clinical trials (2005). Drug-induced long QT and its risk for TdP DMSO, dimethyl sulphoxide; EAD, early afterdepolarization; ECG, electrocardiogram; are now well known, because in recent years, extensive progress has λ, lambda; ERP, effective refractory period; E14, ICH Harmonized Tripartite Guideline E14: The Clinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Poten- been made in studying the congenital long QT syndromes (Roden, tial for Non-Antiarrhythmic Drugs; hERG, human ether-à-go-go-related gene; ICH, The 2004), and also different screening strategies for detecting drug- International Conference on Harmonisation of Technical Requirements for Registration induced TdP have been evaluated following the release of CPMP/ of Pharmaceuticals for Human Use; iCEB, index of Cardiac Electrophysiological Balance; 986/96 in 1997. I , the rapidly activating delayed rectifier potassium current; I , the slowly activating Kr Ks However, cardiovascular safety in general is still the major cause of component of delayed rectifier potassium current; INa, sodium current; MAP, mono- phasic action potential; TdPs, Torsades de Pointes; TDR, maximal transmural disper- drug attrition in early drug development and of drug withdrawal from sion of repolarization; Tp-Te, T wave peak-Twave end (measurement of transmural the market (Laverty et al., 2011). In some cases, the identification of dispersion); VF, ventricular fibrillation; VT, ventricular tachycardia. cardiovascular risks is neither detected in earlier clinical trials nor ⁎ Corresponding author at: Safety Pharmacology Research, Translational Sciences, deemed to be biologically or clinically significant until the new medi- Janssen Research and Development, Janssen Pharmaceutica NV, Belgium, B-2340 Beerse,Belgium.Tel.:+3214602215;fax:+3214605839. cines are given to large patient populations for a long period on the E-mail address: [email protected] (H.R. Lu). market (Psaty & Furberg, 2007; Laverty et al., 2011; Redfern et al.,

1056-8719/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.vascn.2013.01.003 H.R. Lu et al. / Journal of Pharmacological and Toxicological Methods 68 (2013) 250–259 251

2010; Redfern & Valentin, 2011). These cardiovascular risks can lead In the present study, we used seven reference drugs, which are to limitations in the use of the medicine, or to drug withdrawal, and known to induce different types of cardiac effects and arrhythmias, can be caused not only by drug-induced long QT and TdP, but also by in order to investigate whether: 1). iCEB (ratio of QT/QRS) generally other factors (Heist & Ruskin, 2010; Shah, 2010). With the exception reﬂects the classic λ (ERPxCV); 2). iCEB can be used as a simple, of patients with genetic hERG defects, little is known about the impli- cost-effective, new and easily measured non-invasive biomarker to cations of QT-shortening caused by drugs (Lu, Hermans, & Gallacher, predict drug-induced cardiac arrhythmias beyond TdP, and 3). this 2012; Shah, 2010) or slowing of cardiac conduction time (widening new biomarker (iCEB) is better than the currently-used biomarkers of QRS) (Gintant, Gallacher, & Pugsley, 2011; Harmer, Valentin, & [i.e. transmural dispersion of T wave (Anttonen, Vaananen, Junttila, Pollard, 2011; Lu et al., 2010; Madias, 2008). The Cardiac Arrhythmia Huikuri, & Viitasalo, 2008; Antzelevitch et al., 2007; Liu et al., 2006; Suppression Trial (CAST, 1989)showedthatﬂecainide and encainide Sicouri, Moro, Litovsky, Elizari, & Antzelevitch, 1997; Yan &

[Class Ic antiarrhythmic, sodium current (INa) blocking agents] were asso- Antzelevitch, 1999) and instability or variability of the QT interval ciated with an increased incidence of sudden cardiac death in (Jacobson, Carlsson, & Duke, 2011; Thomsen et al., 2004; Van der post-infarction patients (Sellers & DiMacro, 1984; Echt et al., 1991). Linde et al., 2005)] in the isolated arterially perfused rabbit ventric- Some published papers show that there are large numbers of drugs ular wedge preparation. which block INa current and could have the propensity to induce cardiac arrhythmias (Gintant et al., 2011; Harmer et al., 2011; Lu et al., 2010). Moreover, some drugs that shorten the QT-interval could also have the 2. Methods potential to induce non-TdP-like VT/VF (VT: ventricular tachycardia; VF: ventricular fibrillation) (Lu et al., 2008). Since it is difficult to capture QT 2.1. Arterially perfused rabbit left ventricular wedge preparations shortening in the setting of the associated fatal arrhythmia (VF), due to lack of 24 h holter monitoring data, and to capture a slowing down of The use of animals in this study was approved by the Institutional conduction associated with non-TdP-like VT/VF, details concerning Animal Care and Use Committee (IACUC) at the Lankenau Institute for drug-induced short QT interval and slowing of conduction are limited. Medical Research. Indeed, the present ICH S7B guideline (Food and Drug Administration, The methods used for surgical preparation, isolation, perfusion, HHS, 2005) does not specifically address the potential for drug- and recording of transmembrane activity from the arterially perfused induced slowing of conduction (QRS widening), acquired short QT, ventricular wedge preparation from male and female rabbits (New and their potential for the development of fatal arrhythmias. Therefore, Zealand white weighing 2.5 to 3 kg), as well as the viability and elec- the currently-applied biomarkers (e.g. instability of QT and transmural trical stability of the preparation, have been described in previous dispersion of repolarization-T wave: TDR) may not be adequate to studies (Yan et al., 2001, Yan, Shimizu, & Antzelevitch, 1998; Liu et detect all types of drug-induced cardiac arrhythmias. al., 2006). Briefly, rabbits were anti-coagulated with heparin and Therefore, in the present study, we would like to introduce a new and anesthetized by intramuscular injection of xylazine (5 mg/kg) and in- non-invasive biomarker: “the index of Cardiac-Electrophysiological Bal- travenous administration of ketamine·HCL (30 to 35 mg/kg). Via a ance” (iCEB), which is the ratio of QT/QRS of the ECG. We speculate left thoracotomy the heart was excised and placed in a cardioplegic that 1) increases/decreases in QT or JT-interval are proportional to solution consisting of cold (4 °C) normal Tyrode's solution. The left those in effective refractory period (ERP), and 2) changes in QRS circumflex branch of the coronary artery was cannulated and per- are similar to those in conduction velocity (CV). Therefore, the iCEB fused with the cardioplegic solution. Unperfused areas of the left ven- (the ratio of QT/QRS) derived from the ECG, should be the equivalent tricle were removed. The preparation was then placed in a small of the classic λ (ERPxCV), and significant changes in iCEB may reflect tissue bath and arterially perfused with Tyrode's solution containing + an imbalance in cardiac electrophysiology, and therefore predict car- 4mMK and buffered with 95% O2 and 5% CO2 (temperature: diac arrhythmias. Therefore, iCEB could be used as a new biomarker 35.7±0.1 °C; mean perfusion pressure at 35 to 40 mm Hg). The ven- to detect drug-induced broad cardiac arrhythmias including TdP and tricular wedge was allowed to equilibrate in the tissue bath for 1 h non-TdP-like VT/VF (Fig. 1). prior to electrical recordings.

Fig. 1. Balance and imbalance of the depolarization (QRS duration) and repolarization (QT interval) of the cardiac electrophysiology. Schematic changes in the index of Cardiac-Electrophysiological Balance (iCEB): slight increases/decrease (↑ or ↓) in iCEB by increase/decrease in QT interval or QRS duration could be no-arrhythmic or even antiarrhythmic (green zone) (balance between the depolarization and repolarization of cardiac electrophysiology) while large changes (↑↑ or ↓↓) in iCEB may be proarrhythmic (imbalance of cardiac electrophysiology) (dark magenta color). TdPs: Torsades de Pointes, VT: ventricular tachycardia, VF: ventricular ﬁbrillation. 252 H.R. Lu et al. / Journal of Pharmacological and Toxicological Methods 68 (2013) 250–259

2.2. Electrophysiological recordings and ERP measurement from the 2.4. Compounds and their selection rabbit ventricular wedge preparation Encainide, propoxyphene, NS1643, isoprenaline, digoxin and A transmural pseudo-ECG signal was recorded using extracellular levcromakalim were obtained from Sigma-Aldrich® (Belgium). silver/silver chloride electrodes placed in the Tyrode's solution bath- Dofetilide was synthesized by TOSLab Ltd. (Ekaterinburg, Russia). ing the preparation 1.0 to 1.5 cm from the epicardial and endocardial Dimethyl sulfoxide (DMSO, ﬁnal maximal bath concentration 0.1%) surface of the left ventricle (not whole the heart). was used as solvent control. The concentrations of the compounds,

From the ECG, we measured transmural dispersion of repolariza- selected for these studies, were based on published IC50 values for tion which is approximately equal to the interval between the peak their respective target blockade or activation and on their reported and the end of the T wave (|Tpeak −Tend|), and rTp-Te (=|Tp−Te|/ effects in proarrhythmic in vitro models (Bentzen et al., 2010; QT∗100) (Liu et al., 2006; Yan & Antzelevitch, 1999). The intervals Harmer et al., 2011; Kennedy, Seifen, Akera, & Brody, 1986; Lenz & of the ECG are similar to those in the intact heart but without P Hilleman, 2000; Lu et al., 2008; Rhodes et al., 2012; Ulens, Daenens, wave. The QT interval was defined as the time from the onset of the & Tygat, 1999; Wu, Carlsson, Liu, Kowey, & Yan, 2005). QRS to the point at which the final downslope of the T wave crossed the isoelectric line. The JT interval is calculated as the QT interval 2.5. Data analysis minus QRS duration (QT interval−QRS duration). The conduction velocity CV (cm/s) was calculated by dividing the longest distance from All values are expressed as mean and standard error of the mean the edge of the preparation to the pacing electrode by the QRS dura- (S.E.M.). Statistically significant differences between solvent and com- tion. The programmed stimulation at 2-fold the diastolic threshold pound were calculated based on their changes from baseline with the from the endocardium was used to determine the effective refractory Wilcoxon–Mann–Whitney test. Two-tailed probabilities of less than period (ERP): 8 regular stimulations (S1) were followed by an extra- 0.05 were considered to indicate statistically significant differences. stimulus (S2). The coupling interval of S1 to S2 was changed dec- rementally by 5 ms until the stimulation failed to initiate excitation 3. Results of the preparation. This S1–S2 interval was then considered as the

Effective Refractory Period (ERP). The cardiac wavelength (lambda: λ) 3.1. Dofetilide (an IKr blocker, mimicking drug-induced Long QT2) was calculated by ERP (ms)×conduction velocity (CV: cm/s) (Girouard &Rosenbaum,2001). All parameters were recorded at a BCL of Dofetilide (n=6) concentration-dependently significantly increased 500 ms and of 2000 ms. iCEB was calculated by both the ratio of both iCEB (ratio of QT/QRS or JT/QRS) (+79% to 168%) and cardiac QT/QRS or JT/QRS, measured from the ECG recordings. The QRS du- wavelength (λ=ERPxCV) (+96 to 194% from baseline at 0.01 μM, ration, QT, JT and Tp–Te intervals were measured manually from 5 0.1 μM, 0.3 μMand1μM, respectively; pb0.05 versus solvent, n=6) consecutive beats within the last minute of the recording period (Table 1,andFig. 2). and the values were then averaged. In order to estimate conduction Dofetilide elicited early afterdepolarizations (EADs) in 2, 6, 6 and 6 velocity, the size of each preparation was measured. out of the 6 preparations at 0.01, 0.1, 0.3 and 1 μM, respectively (ver- The beat-to-beat instability (variability) of the QT interval was sus 0 out of the 6 preparations with solvent; pb0.05 at 0.1, 0.3 and defined according to the method used in our recent studies (Lu, 1 μM, respectively) (see Fig. 2 for an example of dofetilide-induced Vlaminckx, Van de Water, & Gallacher, 2006; Van der Linde et al., EADs at 0.0.1 μM in an isolated arterially-perfused rabbit left ventric- 2005), and similar to the beat-to-beat variability of repolarization in ular wedge). dogs (Thomsen et al., 2004) and in rabbits (Jacobson et al., 2011). In Dofetilide also concentration-dependently increased transmural brief, data are taken from 20 to 30 consecutive QT intervals and plot- dispersion (expressed as rTp-Te) by 28% to 53% from baseline (versus ted in a Poincaré plot. The distance to the x coordinate represents 0% to +1% from baseline with solvent; pb0.05) and concentration- the length of the plot and reflects the long-term instability, and the dependently increased the instability of the QT interval starting at distance to the y coordinate is the width of the plot and reflects the 0.01 μM(pb0.05 only at 1 μM) (Table 2). As expected, dofetilide short-term instability. After calculation of these distances for all data also concentration-dependently prolonged the QT interval (+79% to points, the median value is calculated for the instability parameters +146% from baseline versus 0% to +4% from baseline with solvent; of the total tracing of consecutive 20 to 30 beats. In the present pb0.05), and the JT interval (+90% to 169% from baseline versus study, we reported the short-term beat-to-beat instability as beat- 0 to +4% from baseline with solvent; pb0.05) without significantly to-beat instability in the present study (Van der Linde et al., 2005). changing QRS duration (Table 4). Furthermore, dofetilide significantly TdP score was also calculated in each group (Liu et al., 2006). increased the TdP score starting at 0.01 μM (7 versus 0 with solvent; Ventricular tachycardia (VT) was defined as four or more consec- pb0.05). The TdP score reached a maximum of 10 at 0.1, 0.3 and utive ventricular premature beats, ventricular fibrillation (VF) was 1 μM, respectively (7 to 10 of baseline versus 0 with solvent; pb0.05). defined as a ventricular tachyarrhythmia without identifiable ECG, and in-excitability was defined as the phenomenon where the prepa- 3.2. Digoxin (an activator of Na+/K+ ATPase pump) ration could not follow the electrical stimulation. Digoxin at 0.01 and 0.1 μM did not significantly or physiologically 2.3. Study protocol relevantly change either λ or iCEB. However, at 1 and 5 μM, digoxin significantly decreased both iCEB and λ (to the same degree as λ Stock solutions of the reference drugs were prepared in DMSO. (−7% to −65% of baseline versus +1% to +3% of baseline with sol- Each compound was tested at 4 different concentrations in 6 wedge vent; pb0.05)) (Table 1 and Fig. 3: upper part). preparations. After one-hour equilibration, during which the prepara- Digoxin elicited non-TdP-like ventricular tachycardia (VT) in 4 out tion achieved electrical stability (Yan et al., 1998, 2001; Liu et al., of the 6 preparations at 5 μM (versus 0 out of the 6 preparations; 2006), one of the following 8 treatments was applied for 30 min per pb0.05), and ventricular fibrillation (VF) in 1 and 6 out of the 6 prep- concentration (n=6 per group). arations at 1 and 5 μM, respectively (versus 0 out of the 6 prepara- The preparation was paced at three basic cycle lengths (BCLs) of tions with solvent; pb0.05 at 5 μM) (see an example in Fig. 3). 500 ms, 1000 ms and 2000 ms. A brief period (30 to 60 s) of faster pac- Digoxin at 0.01, 0.1 and 1 μM did not significantly produce rele- ing at a BCL of 500 ms or less was introduced between BCLs of 1000 ms vant change rTp-Te (Table 2). At 5 μM, rTp-Te could not be measured and 2000 ms using bipolar silver electrodes. due to extra beats and changes in the morphology of the T wave. H.R. Lu et al. / Journal of Pharmacological and Toxicological Methods 68 (2013) 250–259 253

Table 1 Table 2 Effects of 7 reference compounds on the cardiac wavelength (λ) and index of Cardiac Elec- Comparison the effects of 7 reference compounds on iCEB with those on the transmurale trophysiological Balance (iCEB) in the rabbit left ventricular arterially-perfused-wedge. dispersion (rTp-Te: ratio) of T wave and instability of QT-interval (in ms) in the rabbit left ventricular arterially-perfused wedge. Baseline Dose 1 Dose 2 Dose 3 Dose 4 (actual unit) Baseline Dose 1 Dose 2 Dose 3 Dose 4 (% change of baseline) iCEB (QT/QRS) λ (ERPxCV) Solvent 7.8±0.5 7.8±0.5 7.9±0.6 8.0±0.5 8.1±0.5 Solvent 10.7±1 0±1 2±1 3±1 3±1 ⁎ ⁎ ⁎ ⁎ Dofetilide 8.4±0.7 14.9±1.9 18.1±2 20.2±2 20±2 Dofetilde 10±1 96±21* 160±32* 194±29* 182±24* ⁎ ⁎ Digoxin 8.8±0.4 9.0±0.5 9.1±0.5 7.8±0.7 4.8±0.4 Digoxin 10.3±1 2±1 3±2 −7±2* −65±2* ⁎ ⁎ Encainide 8.1±0.5 8.4±0.6 8.4±0.6 7.3±0.1 4.8±0.3 Encainide 11±1 1±2 1±2 −12±2* −40±5* ⁎ NS1643 8.7±0.4 8.8±0.4 8.9±0.4 8.6±0.4 7.0±0.5 NS1643 10.8±0.3 3±1 4±1 0±1 −18±3* ⁎ ⁎ Propoxyphene 8.3±0.5 8.4±0.6 8.1±0.6 7.1±0.4 3.8±0.3 Propoxyphene 10.4±0.2 4±1 1±4 −8±3 −45±3* ⁎ ⁎ Levcromakalim 8.7±0.4 8.9±0.4 8.7±0.4 8.8±0.5 3.4±0.2 Levcromakalim 9.8±1 1±1 −1±1 −2±1 −69±4* Isoprenaline 8.1±0.6 7.3±0.4 6.4±0.5⁎ 6.2±0.5 5.4±0.4⁎ Isoprenaline 10±1 −9±4 −18±3* −17±2* −22±9* Transmural dispersion (rTp-Te) iCEB (QT/QRS) Solvent 18±1 18±1 17±1 18±1 18±1 Solvent 7.8±1 0±1 1±1 2±1 3±1 ⁎ ⁎ ⁎ ⁎ Dofetilde 17±1 22.1±1 28±2 26±2 27±3 Dofetilide 8.4±1 79±19* 119±22* 146±21* 145±23* ⁎ Digoxin 18±1 18±1 18±1 16±1 NA Digoxin 8.8±0.4 3±2 4±2 −11±3* −45±5* ⁎ Encainide 18±1 18±1 18±1 15±1 13±1 Encainide 8.2±1 3±1 3±2 −11±3* −41±4* ⁎ NS1643 17±1 17±1 17±1 17±1 13±2 NS1643 8.7±1 2±1 3±1 −1±1 −20±4* ⁎ ⁎ ⁎ Propoxyphene 17±1 17±1 15±1 13±1 12±1 Propoxyphene 8.3±0.5 2±1 −1±4 −14±2* −54±3* ⁎ Levcromakalim 19±1 18±1 18±1 18±1 23±3 Levcromakalim 8.7±0.4 2±1 1±1 2±4 −61±2* Isoprenaline 17±1 16±1 15±1 14±1 14±2 Isoprenaline 8.1±1 −9±4 −21±3* −19±9 −31±4* Instability of QT (ms) iCEB (JT/QRS) Solvent 0.8±0.2 1.0±0.2 1.1±0.2 1.2±0.1 0.8±0.2 Solvent 6.8±1 0±1 2±1 3±1 4±1 Dofetilde 1.1±0.1a 1.7±0.6a 2.3±0.6a 3.3±1.1a 3.8±1.0⁎a Dofetilide 7.4±1 91±22* 137±25* 168±24* 166±27* Digoxin 0.9±0.2 0.7±0 0.8±0.1 0.7±0.04a 0.9±0.1a Digoxin 7.8±0.4 3±2 4±2 −13±4* −51±6* Encainide 0.9±0.1 0.8±0.1 0.7±0.1 0.8±0.1 0.9±0.1 Encainide 7.2±1 3±1 3±2 −13±3* −47±5* NS1643 0.6±0.1 0.5±0.1 0.8±0.2 0.8±0.2 1.0±0.2 NS1643 7.7±0.4 2±1 3±1 0±1 −22±4* Propoxyphene 0.6±0.1 0.6±0.1 0.7±0.2 0.6±0.1 0.7±0.1 Propoxyphene 7.3±0.5 2±1 −2±4 −16±3* −61±3* Levcromakalim 1.0±0.4a 1.1±0.2a 1.1±0.1a 1.1±0.1a 1.1±0.2a Levcromakalim 7.7±0.4 3±1 1±2 2±4 −69±2* Isoprenaline 1.0±0.1 1.0±0.1 1.4±0.2 1.1±0.3 1.1±0.2 Isoprenaline 7.1±1 −10±4 −24±3* −21±10 −36±4* Data are Mean±S.E.M. of n=6 per group. a: values could not be measured due to Data are Mean±S.E.M. of n=6 per group. ERP: effective refractory period; CV: conduction arrhythmias in some of 6 preparations. NA: not available due to severe arrhythmias. velocity. λ was calculated by ERPxCV (in cm); iCEB was calculated by QT/QRS or JT/QRS iCEB: index of Cardiac Electrophysiological Balance and calculated from QT/QRS (in ratio.*: Pb0.05 versus solvent group in bold. Doses of the reference compounds are ratio). listed in the respective ﬁgure. ⁎ Pb0.05 versus solvent group and indicated in bold. Doses of the reference compounds are listed in the respective ﬁgure.

Digoxin did not significantly change the instability of the QT interval at the lower doses tested. Instability could not be measured at the two highest doses in some experiments (in 2 and 3 out of 6 preparations 3.3. Isoprenaline (a ß1-and ß2-adrenoreceptor agonist) at 1 and 5 μM, respectively), because of frequent premature beats. Digoxin at 1 and 5 μMsignificantly shortened the QT interval (−11% Isoprenaline at 0.05 to 0.5 μMsignificantly reduced λ (−18% to and −43% from baseline versus +2% and +4% from baseline with −22% from baseline versus +2% to +3% from baseline with solvent; solvent; pb0.05) and the JT interval (−12% and −48% from baseline pb0.05) and iCEB to a similar degree (−21% to −31% from baseline ver- versus +2% and +4% from baseline with solvent; pb0.05), without sig- sus +1% to +3% from baseline with solvent; pb0.05 at 0.05 and 5 μM). nificantly changing QRS duration (Table 4). Isoprenaline elicited non-TdP-like VT in 4, 6 and 6 out of 6 preparations

Fig. 2. Concentration-dependent effects of dofetilide on cardiac λ (ERPxCV) and iCEB (QT/QRS ratio) in the isolated rabbit left ventricular arterially-perfused-wedge (left side of the ﬁgure). Values were expressed as mean±S.E.M., and as % changes from baseline, *: pb0.05 versus solvent. On the right side of the ﬁgure: an example of dofetilide-induced long QT and EAD (early afterdepolarization) at 0.01 μM. Original endocardial (Endo) action potential and ECG tracings recorded from a rabbit left ventricular wedge preparation at control period (Control) and in the presence of dofetilide at 0.01 μM with EAD. The results indicate that iCEB was equal to λ in predicting dofetilide-induced early afterdepolarization (EAD). 254 H.R. Lu et al. / Journal of Pharmacological and Toxicological Methods 68 (2013) 250–259

Table 3 Summary of the effects of 7 reference compounds on iCEB, QRS-duration, ERP, JT-interval, instability of QT, transmural dispersion (rTp-Te), and on occurrence of arrhythmias in the rabbit left ventricular arterially-perfused wedge.

iCEB QRS-duration ERP JT-interval instability rTp-Te Incidence

Dofetilide ↑ ± ↑ ↑↑↑ EAD Digoxin ↓ ± ↓↓± ± VT/VF Encainide ↓↑ ↑±±↓ VT/VF NS1643 ↓↑ ↓↓± ↓ VT/VF Propoxyphene ↓↑ ↑±±↓ VT/VF Levcromakalim ↓ ± ↓↓± ± VT/VF Isoprenaline ↓ ± ↓↓± ± VT/VF iCEB; index of Cardiac Electrophysiological Balance, ERP: effective refractory period; EAD: early afterdepolarization; VT; ventricular tachycardia; VF: ventricular fibrillation. ↑: significant increase; ↓: significant decrease; ±: no significant change. Data are only taken at the dose with incidence or the highest dose tested.

at 0.05 μM, 0.1 μMand0.5μM, respectively (versus 0 out of the 6 prep- 3.4. Encainide (a class Ic antiarrhythmic agent, an INA blocker) arations with solvent; pb0.05) (Fig. 3:lowerpart). Isoprenaline did not significantly change the instability of the QT Encainide at 0.5 and 1 μM did not significantly change both λ and interval at doses up to 0.5 μM, and did not increase, but tended to de- iCEB. However, at 5 and 20 μM, encainide significantly reduced λ crease rTp-Te (Table 2). In the presence of isoprenaline at 0.1 and (ERPxCV: −12% and −40% from baseline). Encainide significantly re- 0.5 μM, the instability of the QT interval could not be determined in duced iCEB (expressed as ratio of QT/QRS or JT/QRS) in a similar man- about 50% of the preparations due to extra beats, and also Tpeak ner (−11% to −47% from baseline) (Table 1 and Fig. 5). At 20 μM, and the TdP score could not be measured correctly. encainide elicited non-TdP-like VT in 4 out of the 6 preparations (ver- Isoprenaline shortened the QT interval (−9%, −21%, −20% and sus 0 out of the 6 preparations with solvent; pb0.05) (Fig. 4: lower −31% from baseline at 0.01 μM, 0.05 μM, 0.1 μM and 0.5 μM, respec- part). tively, versus≤+4% from baseline with solvent; pb0.05 except for Encainide slightly but significantly prolonged the QT interval (+5%, 0.1 μM), and the JT interval (−10%, −24%, −22% and −36% from +8%, +14% and +12% from baseline at 0.5 μM, 1 μM, 5 μMand baseline at 0.01 μM, 0.05 μM, 0.1 μM and 0.5 μM, respectively, ver- 20 μM, respectively, versus ≤+4% from baseline with solvent; sus≤+4% from baseline with solvent; pb0.05, except for 0.1 μM). pb0.05 except for at 20 μM) and JT interval (+6%,+8%, +12% and Isoprenaline did not significantly change QRS duration (Table 4) and 0% from baseline at 0.5 μM, 1 μM, 5 μMand20μM, respectively, ver- QRS-rate dependency (data not shown). sus≤+4%frombaselinewithsolvent; pb0.05, except for at 20 μM). Encainide dose-dependently increased QRS duration (+3%, +5%, +29% and +96% from baseline, at 0.5 μM, 1 μM, 5 μMand20μM, respectively, versus≤+1% from baseline with solvent; pb0.05 except Table 4 for at 0.5 μM) (Table 4), and QRS-rate-dependency (+5%, +18% Effects of 7 reference compounds on iCEB, QRS-duration, ERP in the rabbit left ventric- and +61% from baseline at 1 μM, 5 μMand20μM, respectively, ver- ular arterially-perfused-wedge. sus 0% of baseline with solvent; pb0.05 except for at 0.5 μM), with- Baseline Dose 1 Dose 2 Dose 3 Dose 4 out significantly changing TdP score (−1to+1ofbaseline). iCEB (QT/QRS) Furthermore, encainide at 0.5 μM, 1 μMand5μMdidnotsignifi- Solvent 7.8±0.5 7.8±0.5 7.9±0.6 8.0±0.5 8.1±0.5 cantly change rTp-Te. At 20 μM, encainide did not increase, but signifi- Dofetilide 8.4±0.7 14.9±1.9* 18.1±2* 20.2±2* 20±2* cantly decreased rTp-Te (−33% from baseline versus 0% from baseline Digoxin 8.8±0.4 9.0±0.5 9.1±0.5 7.8±0.7* 4.8±0.4* with solvent pb0.05) (Table 2). Encainide did not significantly change Encainide 8.1±0.5 8.4±0.6 8.4±0.6 7.3±0.1* 4.8±0.3* NS1643 8.7±0.4 8.8±0.4 8.9±0.4 8.6±0.4 7.0±0.5* the instability of the QT interval (Table 2). Propoxyphene 8.3±0.5 8.4±0.6 8.1±0.6 7.1±0.4* 3.8±0.3* Levcromakalim 8.7±0.4 8.9±0.4* 8.7±0.4 8.8±0.5 3.4±0.2* 3.5. Propoxyphene (a weak opioid agonist that presents with Isoprenaline 8.1±0.6 7.3±0.4 6.4±0.5* 6.2±0.5 5.4±0.4* non-TdP-like arrhythmias associated with overdose in man)

QRS-duration μ fi λ Solvent 37±2 37±2 37±2 37±2 37±2 Propoxyphene at 0.1 and 1 M did not signi cantly change and Dofetilde 36±2 36±2 36±2 36±2 36±2 iCEB. However, at 10 and 100 μM, propoxyphene significantly reduced Digoxin 35±1 35±1 35±1 35±1 36±1 λ (ERPxCV: −8% and −45% from baseline) and iCEB (expressed as a Encainide 37±2 38±2 39±2* 48±4* 73±9* ratio of QT/QRS or JT/QRS) to a similar extent (−14% to -61% from NS1643 35±2 35±2 35±2 36±2 39±3* μ Propoxyphene 37±2 37±2 40±2 47±3* 95±8* baseline) (Table 1 and Fig. 4: upper part). At 100 M, propoxyphene Levcromakalim 34±1 34±1 34±1 34±1 34±1 elicited non-TdP-like VT in 4 out of the 6 preparations (versus 0 out of Isoprenaline 38±1 38±1 37±1 37±1 37±1 the 6 preparations with solvent; pb0.05) (Fig. 4) and in-excitability (the preparation had no response to the electrical stimulation) in 3 ERP out of the 6 preparations (versus 0 out of 6 preparations with solvent; Solvent 248±13 248±14 252±14 254±13 257±13 b Dofetilde 243±17 466±41* 619±66* 695±48* 669±41* p 0.05). Digoxin 248±16 252±19 253±22 230±20* 89±9* Propoxyphene did not increase, but significantly reduced rTp-Te Encainide 261±12 271±14 278±16* 297±15* 302±15* (−11%, −25% and −30% from baseline at 1 μM, 10 μM and 100 μM, NS1643 250±7 255±9 258±10 258±10 225±10* respectively, versus 0% from baseline with solvent) (Table 2). Pro- Propoxyphene 247±6 255±6* 261±7* 283±6* 343±5* fi Levcromakalim 232±8 232±9 231±9 230±8* 71±6* poxyphene at the doses tested, did not signi cantly change the in- Isoprenaline 238±9 216±12 194±9 193±9 182±12 stability of the QT interval (Table 2). Propoxyphene at 100 μM significantly prolonged the QT interval (+17% from baseline versus Data are Mean±S.E.M. of n=6 per group. iCEB: index of Cardiac Electrophysiological b fi Balance, ERP: effective refractory period.*: Pb0.05 versus solvent group in bold. +4% from baseline with solvent; p 0.05), but did not signi cantly Doses of the reference compounds are listed in the respective figure. change the duration of the JT interval. Propoxyphene significantly H.R. Lu et al. / Journal of Pharmacological and Toxicological Methods 68 (2013) 250–259 255

Fig. 3. Effects of digoxin and isoprenaline on cardiac λ (ERPxCV) and iCEB (QT/QRS ratio) in the isolated rabbit left ventricular arterially-perfused-wedge (left side of the figure). Values were expressed as mean±S.E.M., and as % changes from baseline, *: pb0.05 versus solvent. On the right side of the figure: an example of original ECG tracings recorded from a rabbit left ventricular wedge preparation at control period (Control) and in the presence of digoxin at 5 μM or isoprenaline at 0.05 μM. The results indicate that iCEB was equal to λ in predicting digoxin- and isoprenaline-induced non-TdP-like ventricular tachycardia (VT) and ventricular fibrillation (VF). increased QRS duration (+25% and +153% from baseline at 10 μM baseline at 1 μM, 10 μM and 100 μM, respectively, versus 0% of base- and 100 μM versus≤+1% from baseline with solvent; pb0.05) line with solvent; pb0.05) without notable changes in the TdP score (Table 4) and QRS-rate-dependency (+3%, +20% and +84% from (0 to −1 of baseline).

Fig. 4. Effects of propoxyphene and encainide on cardiac λ (ERPxCV) and iCEB (QT/QRS ratio) in the isolated rabbit left ventricular arterially-perfused-wedge (left side of the ﬁgure). Values were expressed as mean±S.E.M., and as % changes from baseline, *: pb0.05 versus solvent. On the right side of the ﬁgure: examples of original ECG tracings recorded from a rabbit left ventricular wedge preparation at control period (Control) and in the presence of propoxyphene at 100 μM and encainide at 20 μM with VT. The results indicate that iCEB was equal to λ in predicting encainide- and propoxyphene-induced non-TdP-like ventricular tachycardia (VT). 256 H.R. Lu et al. / Journal of Pharmacological and Toxicological Methods 68 (2013) 250–259

3.6. Levcromakalim (an IKATP channel opener, mimicking short QTs) pb0.05) without significantly changing QRS duration except for a slight increase at the highest dose (Table 4) and QRS-rate dependency, and Levcromakalim at 10 μM largely reduced both λ (expressed as with a slight reduction in TdP score at 10 μM(−1 of baseline versus ERPxCV: −69% from baseline versus +3% from baseline with solvent; 0withsolvent;pb0.05). NS1463 at doses tested up to 10 μM did not pb0.05). Similarly, levcromakalim at 10 μM also markedly reduced significantly change the instability of the QT interval (Table 2). iCEB (expressed as QT/QRS or JT/QRS: −61% or −69% from baseline, re- In summary, significant increases or reductions in iCEB (QT/QRS ratio) spectively, versus +3% or +4% from baseline with solvent; pb0.05) are associated with its potential risk of induction of long QT/EADs or (Table 1 and Fig. 5). Levcromakalim elicited non-TdP-like VF in 4 out non-TdP types of VT/VF (Table 3). of the 6 preparations (versus 0 out of the 6 preparations with solvent; pb0.05) (see an example in Fig. 5). 4. Discussion Levcromakalim tended to increase rTp-Te at 10 μM (+27% from baseline versus 0% from baseline with solvent, p>0.05), but did not Our data with 7 cardio-active reference compounds tested in the iso- significantly change instability of the QT interval at the doses tested lated rabbit ventricular wedge preparation shows that the new non- (Table 2). As expected, levcromakalim at 10 μM largely shortened invasive biomarker, index of Cardiac-Electrophysiological Balance the QT interval (−60% from baseline versus +4% from baseline with (iCEB: the ratio of QT/QRS), is preceded with not only drug-induced solvent; pb0.05) and the JT interval (−68% from baseline versus long QT and TdP, but also to drug-induced slowing of conduction, QT +4% from baseline with solvent; pb0.05) without significantly chang- shortening and their related non-TdP-like VT/VF. iCEB, measured from ing QRS-duration and QRS-rate dependency. The TdP score was slight- the electrocardiogram (ECG), could serve as an important, new and ly reduced at 10 μM(−1 of baseline versus 0 of baseline with solvent; non-invasive biomarker for potential risks of drug-induced cardiac pb0.05). arrhythmias beyond LQTs and TdP, since currently used biomarkers i.e. transmural dispersion (rTp-Te) and instability of the QT interval 3.7. NS1643 (an Ikr activator mimicking genetic short QT syndrome) are limited to the prediction of drug-induced long QT and TdP.

NS1643 at 10 μM significantly reduced λ (expressed as ERPxCV: 4.1. Can the index of Cardiac-Electrophysiological Balance (iCEB) be −18% from baseline versus +3% from baseline with solvent; used as a more easily measured substitute to the measurement of λ for pb0.05). Similarly, NS1643 at 10 μM also significantly reduced iCEB detecting potential risks of drug-induced cardiac arrhythmias (expressed as QT/QRS or JT/QRS: −20% or −22% from baseline, re- beyond TdP? spectively, versus +3% or +4% from baseline with solvent; pb0.05) (Table 1 and Fig. 5). NS1643 elicited non-TdP-like VT/VF in 1 out of In the present study, our data from 7 reference compounds in- the 6 preparations (versus 0 out of the 6 preparations with solvent; dicate that iCEB (ratio of QT/QRS or JT/QRS, taken directly from ECG p>0.05) (see an example in Fig. 5). recordings) changed parallel with the classic cardiac λ (ERPxCV) NS1643 at 10 μM did not increase but decreased rTp-Te (−20% from (Girouard & Rosenbaum, 2001)(Table 1). Drugs that only slightly in- baseline versus 0% from baseline with solvent, pb0.05). As expected, crease or decrease iCEB may be safe, while drugs markedly increasing NS1643 at 10 μM significantly shortened the QT interval (−11% from or decreasing this parameter could potentially be pro-arrhythmic baseline versus +4% from baseline with solvent; pb0.05) and the JT in- (Fig. 1). λ has been used as a parameter to study ventricular reentrant terval (−14% from baseline versus +4% from baseline with solvent; tachycardia and atrial fibrillation (Aidonidis et al., 2009; Robert et al.,

Fig. 5. Effects of levcromakalim and NS1643 on cardiac λ (ERPxCV) and iCEB (QT/QRS ratio) in the isolated rabbit left ventricular arterially-perfused-wedge (left side of the figure). Values were expressed as mean±S.E.M., and as % changes from baseline, *: pb0.05 versus solvent. On the right side of the figure: Examples of original ECG tracings recorded from a rabbit left ventricular wedge preparation at control period (Control) and in the presence of levcromakalim at 10 μM and NS 1643 at 10 μM with ventricular fibrillation (VF). The results indicate that iCEB was equal to λ in predicting NS 1643- and levcromakalim-induced non-TdP-like VT/VF. H.R. Lu et al. / Journal of Pharmacological and Toxicological Methods 68 (2013) 250–259 257

1996). However, the classical cardiac λ is calculated based on the in- of the action potential significantly change iCEB and may result in car- vasive measurement of ERP by using extra electrical stimulation (s) to diac arrhythmic conditions. the heart in a cardiac electrophysiological pacing laboratory in man. Indeed, in the present study, dofetilide (a potent IKr blocker) in- Although λ is also used as a biomarker in animal models, the inva- duced long QT and high incidences of EADs (risk of TdP) (Lenz & siveness of the ERP measurement may limit the use of this biomarker Hilleman, 2000). This was associated with significant concentration- as a screening tool/biomarker. In addition, as summarized in Table 1: dependent increases in iCEB. These results indicate that iCEB predict- significant drug-induced changes in λ (ERPxCV) are similar to those ed potential risks of drug-induced long QT and TdP risk. On the other in iCEB, and both biomarkers are useful to predict potential risks for hand, the induction of non-TdP-types of VT/VF was associated with drug-induced cardiac arrhythmias. However, iCEB is a simple param- significant and marked decreases in iCEB with drugs slowing conduc- eter, directly measured from ECG recordings without invasive ERP tion (propoxyphene and encainide) and shortening QT (NS1643, di- measurement, and may have advantages over the classical λ mea- goxin, isoprenaline and levcromakalim). The large decreases in iCEB surement, which is based on the invasive measurement of ERP. may therefore be a prerequisite substrate for the initiation of non- However, if a drug (i.e. with a multiple ion-current blocking ef- TdP-types of VT/VF. fects) significantly changes both the duration of the QT interval and Propoxyphene is a synthetic opioid pain reliever and the FDA QRS duration to a similar degree, but it does not significantly change recommended its withdrawal from the market due to its risk of caus- the iCEB, whether this kind of drugs resulting in proarrhythmic or ing cardiac arrhythmias (not LQT and TdP related), heart failure and anti-arrhythmic are unknown in the present study and are beyond hypotension in 2010. Toxic blood concentrations range between 3 the scope of the present manuscript. and 180 μmol/l (Cravey, Shaw, & Nakmura, 1974; Ulens et al., 1999). Encainide was also shown to have a proarrhythmic potential in clinical trials [Cardiac Arrhythmia Suppression Trial (CAST)] (The 4.2. The new non-invasive biomarker-iCEB plays an important role in CAST Investigators, 1989), and in experimental models (Nattel, 1998). drug-induced cardiac arrhythmias beyond TdP In the CAST, INa-current blockers such as encainide (Class Ic) were associated with an increased incidence of sudden cardiac death in The ventricular waveforms on the body ECG are generated by both post-myocardial infarct patients (The CAST Investigators, 1989). Our depolarization and repolarization of the ventricular myocytes' action results indicate that both propoxyphene (100 μM) and encainide potentials, and present themselves as the QRS complex (depolariza- (20 μM) markedly slowed conduction and thereby decreased iCEB, tion) followed by the T wave (repolarization) (Yan, Lankipalli, Burke, which was associated with the induction of non-TdP types of VT/VT Musco, & Kowey, 2003). In physiological conditions there is a dynamic (Fig. 5). balance between depolarization and repolarization. Slight deviations High/toxic doses of digoxin (an activator of Na+/K+ ATPase pump) in the balance i.e. within depolarization and repolarization reserve, (Kennedy et al., 1986) NS1643 (an IKr activator, mimicking are harmless or even can be anti-arrhythmic. For example, lidocaine genetic short QT syndrome) (Lu et al., 2010), isoprenaline (a ß1-and which is a safe Ia antiarrhythmic drug in men, only slightly increases ß2-adrenoreceptor agonist, mimicking catecholaminergic polymor- + QRS duration. Nicorandil (an IK ATP channel agonist) which slightly phic VT) (Balasubramaniam, Grace, Saumarez, Vandenberg, & Huang, shortens the QT interval is also considered as a safe drug. Raloxifene 2003) and of levcromakalim (an IKATPase opener), all largely reduced is an oral selective estrogen receptor modulator (SERM) that has es- iCEB. Significant reductions in iCEB may be therefore associated with trogenic actions on bone and anti-estrogenic actions on the uterus the risk for non-TdP types of VT/VF. The mechanisms by which digitalis and breast. Raloxifene is clinically used without drug-induced long over-dose or toxicity promotes the development of cardiac arrhythmias QT/TdPs. However, raloxifene has potent HERG and IKs blocking activ- including fatal non-TdP types of VT/VF are still not completely clear. In ities in vitro (Liu et al., 2010). All these 3 compounds: lidocaine (n=6 the present study, decreases in iCEB preceded the induction of non-TdP up to 100 μM), raloxifene (up to 10 μM; n=6) and nicorandil (n=6 types of VT/VF. It is already known that isoprenaline at high doses, in- up to 10 μM) have little effects on iCEB (changesb8%) and do not elicit duces cardiac arrhythmias (including delayed afterdepolarizations, trig- cardiac arrhythmias in this model (our unpublished data). We suggest gered activities and non-TdP types of VT/VF) via ß-adrenergic overly 2+ that further studies are needed to determine the threshold for a clini- stimulation of ICa (L) which increases the concentration of Ca .The cally relevant change in iCEB. Moreover, additional studies are also Ca2+-overload maybe caused by Ca2+ release from the sarcoplasmatic needed to establish the role for iCEB in safety risk assessment, either reticulum (SR) via ryanodine 2 receptors (R2R) during diastole, as a stand-alone parameter or in combination with other biomarkers, and therefore provokes non-TdP-like VT/VF (Nam, Burashnikov, & especially for drugs with multiple ion-current blocking activities Antzelevitch, 2005; Watanabe & Knollmann, 2011; Song, Wu, Shryock, such as azimilide and qunidine. In addition, similar to significant & Belardinelli, 2002). However, there are no clear biomarkers to predict changes of all other biomarkers such as instability, transmurale dis- digitalis and isoprenaline-induced arrhythmias. Our data show that persion, QT-interval or QRS duration etc., significant changes in iCEB iCEB can serve as a new biomarker to predict their potential risk of are not always directly linked to the incidence of arrhythmias for non-TdP like VT/VF. each individual experiment. Further studies with large sample size Slowing of conduction (increasing QRS duration) was associated are wanted to further investigate if iCEB can predict whether certain with sudden cardiac death in both man and in experimental models subjects will response with cardiac arrhythmias or not. (Antzelevitch, 2007; Yan & Antzelevitch 1999; Lu et al., 2010; Lu, On the other hand, significant imbalances (either increases or de- Hermans, & Gallacher, 2012; Nattel, 1998). Drugs which slow conduc- creases) between the depolarization and repolarization of the heart tion, thereby decreasing iCEB, and associated with non-TdP-like VT/VF e.g. in pathological conditions (heart failure, inherited LQTs or could be interpreted as “bad” INa blocking drugs like encainide (Class Brugada syndrome etc.) or induced by different types of cardiac and Ic) and propoxyphene (Lu et al., 2010, Lu, Hermans & Gallacher non-cardiac drugs as we have seen in the present study, may be asso- 2012). Indeed, changes in conduction time, ERP or both, promote re- ciated with arrhythmic conditions (Fig. 1). Patients with heart failure entrant arrhythmias (Aidonidis et al., 2009; Robert et al., 1996, 1999; and Brugada syndrome prolonged QRS-duration thereby decreasing Lu et al., 2010; Lu, Hermans & Gallacher 2012). Drugs that increase λ iCEB (QT/QRS) and were associated with increased incidence of tend to increase the risk for TdPs, whereas agents that decrease the sudden cardiac death (Kashani & Barold, 2005; Ohkubo et al., 2011). λ tend to increase the risk for non-TdP VT/VF (Aidonidis et al., 2009; Patients with inherited LQTs significantly prolonged QT-interval Hondeghem, Dujardin, Hoffmann, Dumotier, & De clerk, 2011; thereby significantly increasing iCEB and are associated with inci- Robert et al., 1996, 1999). In 1913, Mines first suggested the role of λ dence for TdPs. Large changes in the depolarization or repolarization in the mechanism of reentry: reentrant excitation was only possible 258 H.R. Lu et al. / Journal of Pharmacological and Toxicological Methods 68 (2013) 250–259 if the λ of the propagating impulse was shorter than the reentrant beats or VT/VF, and it was therefore impossible to analyze 20 to 30 con- path length, i.e. that there is an excitable gap within the reentrant secutive QT intervals. circuit (Mines, 1913). Later on, other studies also suggested that de- λ creases in by shortening the QT interval are related to the initiation 5. Conclusion and maintenance of reentry, which lead to reentrant VT or fibrillation (Robert et al., 1996, 1999; Aidonidis et al., 2009; Girouard et al., 1996). This non-invasive, simple and new biomarker iCEB represents the Robert et al. reported that levcromakalim-induced reentry VT in iso- balance between the depolarization (changes in QRS duration) and λ lated rabbit hearts was associated with a shortening in due to a di- repolarization (changes in the QT interval) of the cardiac action po- rect decrease in ERP without changes in CV (Robert et al., 1999). tential, while transmural dispersion of the T wave and instability of Aidonidis et al. (2009) reported that bimakalim (another IKATP chan- the QT interval only cover alterations in repolarization (changes in λ nel opener) shortened ERP thereby decreased and induced reentrant the QT interval) of the action potential. Therefore, the new biomarker non-TdP-like VT/VF in chronic infarct anesthetized pigs. In the clinic, iCEB may have advantages over the current biomarkers in identifying genetic short QT syndromes, characterized by abnormally short QT in- the potential risk for cardiac arrhythmias, especially for its potential b fi tervals ( 300 ms), are linked to atrial brillation, syncope and sudden to discriminate between long-QT related arrhythmias and non-TdP- death due to non-TdP-like VF (Gaita et al., 2003; Giustetto et al., 2011). like VT/VF. Therefore, a large decrease in iCEB is associated with drug-induced The present study in rabbit left ventricular arterially-perfused non-TdP-like VT/VF, while a marked increase in iCEB is associated wedge suggests that the iCEB may constitute a potentially important with drug-induced TdPs. and non-invasive biomarker in predicting a higher predisposition to different types of drug-induced cardiac arrhythmias beyond LQT and TdP. However, this left ventricular arterially-perfused wedge 4.3. Does iCEB have advantages over established biomarkers such as preparation is an in vitro model under a condition of pacing. The im- transmural dispersion (Tp-Te, rTp-Te) and instability of the QT interval? portance of this biomarker in predicting cardiac arrhythmias in other models both in vitro and in vivo still needs to be established. Further- QT interval prolongation is associated with TdP, but QT prolongation more, the value of iCEB as a tool to predict potential arrhythmias in per se is a poor biomarker of the risk of TdP actually occurring. There- humans, especially in diseased conditions, needs to be evaluated in fore, more predictable bio-markers for TdP, including transmural dis- future studies. persion of the T wave (Tp-Te, rTp-Te) and instability/variability of the In conclusion, our data from 7 compounds with known cardiac ef- QT interval, have been developed in recent years. Increases in trans- fects suggest that: 1) this non-invasive iCEB is qualitatively similar to mural dispersion (rTp-Te) of the T wave and instability of the QT inter- λ, gives an indication towards a higher likelihood for drug-induced val have been well established as associated with drug-induced long QT CAs beyond long QT and TdP; 2) iCEB may be better than the current syndrome, and these two biomarkers are associated with drug-induced biomarkers (i.e. transmural dispersion and instability) in identifying TdP (Jacobson et al., 2011; Liu et al., 2006; Lu et al., 2006; Thomsen et al., the risk for drug-induced potential risks for non-TdP-like VT/VF. 2004; Van der Linde et al., 2005; Yan & Antzelevitch, 1999). Indeed, in the present study, our results with dofetilide (an I blocker) showed Kr Acknowledgments that all 3 biomarkers including iCEB, rTp-Te and instability of the QT-interval predict the compound's ability to cause QT-prolongation Part of this work was supported by an IWT feasibility grant # 110350. and are associated with a risk of producing TdP. However, in the present The authors wish to thank Dr. R. Towart (Janssen Pharmaceutica NV), study, drug-induced non-TdP-like VT/VF is generally not associated Dr. B. Loenders (Janssen Pharmaceutica NV) for their scientificcom- with increases in transmural dispersion or in instability of the QT- ments, Mr. E. Vlaminckx and Mr. K. Van Ammel (Janssen Pharmaceutica interval. Only iCEB appeared to be a good biomarker for this type of ar- NV) for analysis and checking of the data, and K15 (Belgium) for gener- rhythmia, as a decrease in iCEB is indicative of a risk for drug-induced ating the appropriate figures. The experiments used in this manuscript non-TdP-like VT/VF (Tables 2 and 3). were performed by HRS (Heart Rhythm Solutions, USA). With respect to the concept that increases in transmural dispersion may be a general parameter for arrhythmogenesis of LQT/TdP or of short QT (Anttonen et al., 2008; Extramiana & Antzelevitch, 2004; References Antzelevitch, 2007; Antzelevitch et al., 2007; Yan & Antzelevitch, Aidonidis, I., Poyatzi, A., Stamatiou, G., Lymberi, M., Stamatoyannis, N., & Molyvdas, P. A. 1999), in the present study in the rabbit wedge preparation, we (2009). Dose-related shortening of ventricular tachycardia cycle length after adminis- could not exclude its potential role in drug-induced non-TdP-like tration of the KATP channel opener bimakalim in a 4-day-old chronic anesthetized pig – μ μ model. Journal of Cardiovascular Pharmacology and Therapeutics, 14,2222 2230. VT/VF. Encainide (20 M), NS1643 (10 M) and propoxyphene (at Anttonen, O., Vaananen, H., Junttila, J., Huikuri, H. V., & Viitasalo, M. (2008). Electrocar- the highest 3 doses) significantly reduced transmural dispersion of diographic transmural dispersion of repolarizarion in the patients with inherited T wave, in contrast to increases as reported in literature (Anttonen short QT syndrome. Annals of Noninvasive Electrocardiology, 13, 295–300. Antzelevitch, C. (2007). Role of spatial dispersion of repolarization in inherited and ac- et al., 2008; Extramiana & Antzelevitch, 2004; Antzelevitch, 2007; quired sudden cardiac death syndromes. AJP Heart, 293, H2024–H2038. Antzelevitch et al., 2007; Yan & Antzelevitch, 1999). The reason for Antzelevitch, C., Viskin, S., Shimizu, A., Yan, G. -X., Kowey, P., Zhang, L., et al. (2007). this discrepancy is unknown. In cases of drug-induced changes in Does Tpeak–Tend provide an index of transmural dispersion of repolarization? Heart Rhythm, 4, 1114–1119. + chan- morphology of the T wave (which occurs in the presence of Na Balasubramaniam, R., Grace, A. A., Saumarez, R. C., Vandenberg, J. I., & Huang, C. L. nel blockers or QT-shortening drugs), it is difficult to accurately mea- (2003). Electrogram prolongation and nifedipine-suppressible ventricular arrhyth- sure the dispersion (Tp-Te of QT) from ECG recordings. Furthermore, it mias in mice following targeted disruption of KCNE1. Journal de Physiologie, 552, – is also difficult to continuously measure the QT interval when there 535 546. Bentzen, B. H., Bahrke, S., Wu, K., Larsen, A. P., Biermann, J., Koern, G., et al. (2010). The are extra beats in the rabbit wedge model, hampering a correct calcu- hERG channel opener NS1643 shortens the QT in transgenic rabbits with long lation of instability. QT-syndrome 1 in vivo and shortens the APD90 ex vivo, but is associated with in- Indeed, in isolated wedge preparations, compounds like encainide creased risk of arrhythmias. Acta Physiologica, 198(Suppl. 677) O-MON-2-6. The CAST Investigators (1989). Preliminary report: Effect of encainide and flecainide and propoxyphene markedly widened the QRS complex, and altered on mortality in a randomized trial of arrhythmia suppression after myocardial in- the morphology of the T wave (flattened, negative or bi-phasic) — farction. The New England Journal of Medicine, 321, 406–412. Tp-Te can't be accurately measured in these circumstances. In case of Cravey, R. H., Shaw, R. F., & Nakmura, G. R. (1974). Incidence of propoxyphene poising: A report of fetal cases. Journal of Forensic Sciences, 19,72–80. administration of a high dose of digoxin or isoprenaline, instability Cubeddu, L. X. (2003). QT prolongation and fatal arrhythmias: A review of clinical im- was not always measurable, because there were incidences of extra plications and effects of drugs. American Journal of Therapeutics, 10, 452–457. H.R. Lu et al. / Journal of Pharmacological and Toxicological Methods 68 (2013) 250–259 259

Echt, D. S., Liebson, P. R., Mitchell, L. B., Peters, R. W., Obias-Manno, D., Barker, A. H., Nattel, S. (1998). Experimental evidence for proarrhythmic mechanisms of antiar- et al. (1991). Mortality and morbidity in patients receiving encainide, flecainide, rhythmic drugs. Cardiovascular Research, 37, 567–577. or placebo. The Cardiac Arrhythmias Suppression Trial. The New England Journal Nam, G. B., Burashnikov, A., & Antzelevitch, C. (2005). Cellular mechanisms underlying the of Medicine, 324, 781–788. development of catacholaminergic ventricular tachycardia. Circulation, 111, Extramiana, F., & Antzelevitch, C. (2004). Amplified transmurale dispersion of repolar- 2727–2733. ization as the basis for arrhythmogenesis in a canine ventricular-wedge model of Psaty, B. M., & Furberg, C. D. (2007). Rosiglitazone and cardiovascular risk. The New short-QT syndrome. Circulation, 110, 3661–3666. England Journal of Medicine, 356, 2522–2524. Food and Drug Administration, HHS (HHS, 2005). International Conference on Pugsley, M. K., Authier, S., & Curtis, M. J. (2008). Principles of safety pharmacology. Harmonisation; Guidance on S7B nonclinical evaluation of the potential for de- British Journal of Pharmacology, 154, 1382–1399. layed ventricular repolarization (QT interval prolongation) by human pharmaceu- Redfern, W. S., Ewart, L., Hammond, T. G., Bialecki, R., Kinter, L., Lindgren, S., et al. ticals; availability. Federal Register, 70, 61133–61134. (2010). Impact and frequency of different toxicities throughout the pharmaceuti- Gaita, F., Giustetto, C., Bianchi, F., Wolpert, C., Schimpf, R., Riccardi, R., et al. (2003). cal life cycle. The Toxicologist, 114, 1081. Short QT syndrome. A familial cause of sudden death. Circulation, 108, 965–970. Redfern, W. S., & Valentin, J. -P. (2011). Trends in safety pharmacology: Posters pre- Gintant, G. A., Gallacher, D. J., & Pugsley, M. K. (2011). The “overly-sensitive” heart so- sents at the annual meetings of the safety pharmacology society 2001–2010. Jour- dium channel block and QRS interval prolongation. British Journal of Pharmacology, nal of Pharmacological and Toxicological Methods, 64, 102–110. 164, 254–259. Rhodes, S. S., Camara, A. K., Heisner, J. S., Riess, M. L., Aldakkak, M., & Stowe, D. F. Girouard, S. D., Pastore, J. M., Laurita, K. R., Gregory, K. W., & Rosenbaum, D. S. (1996). Op- (2012). Reduced mitochondrial Ca2+ loading and improved functional recovery tical mapping in new guinea-pig model of ventricular tachycardia reveals mecha- after ischemia reperfusion injury in old vs young guineapig hearts. American Jour- nisms for multiple wavelengths in a single reentrant circuit. Circulation, 93,603–613. nal of Physiology - Heart and Circulatory Physiology, 302, H855–H863. Girouard, S. D., & Rosenbaum, D. S. (2001). Role of wavelength adaptation in the initi- Robert, E., Bruelle, P., de la Coussaye, J. E., Juan, J. M., Brugada, J., Peray, P., et al. (1996). ation, maintenance, and pharmacologic suppression of reentry. Journal of Cardio- Electrophysiologic and proarrhythmogenic effects of therapeutic and toxic doses of vascular Electrophysiology, 12, 697–707. imipramine; A study with high resolution ventricular epicardial mapping in rabbit Giustetto, C., Schimpf, R., Mazzanti, A., Scrocco, C., Maury, P., Anttonen, O., et al. (2011). hearts. Journal of Pharmacology and Experimental Therapeutics, 278, 170–178. Long-term follow-up of patients with short QT syndrome. Journal of the American Robert, E., Aya, A. G., de la Coussaye, J. E., Peray, P., Juan, J. M., Brugada, J., et al. (1999). College of Cardiology, 58, 587–595. Dispersion-based reentry: Mechanism of initiation of ventricular tachycardia in Harmer, A. R., Valentin, J. P., & Pollard, C. E. (2011). On the relationship between block isolated rabbit hearts. American Journal of Physiology, 276, H413–H423. of the cardiac Na+ channel and drug-induced prolongation of the QRS complex. Roden, D. M. (2004). Drug-induced prolongation of the QT interval. The New England British Journal of Pharmacology, 164, 260–273. Journal of Medicine, 350, 1013–1022. Heist, E. K., & Ruskin, J. N. (2010). Contemporary reviews in cardiovascular medicine. Sellers, T. D., & DiMacro, J. P. (1984). Sinusoidal ventricular tachycardia associated with Drug-induced arrhythmia. Circulation, 122, 1426–1435. flecainide acetate. Chest, 85, 647–649. Hondeghem, L. M., Dujardin, K., Hoffmann, P., Dumotier, B., & De clerk, F. (2011). Sicouri, S., Moro, S., Litovsky, S., Elizari, M. V., & Antzelevitch, C. (1997). Chronic Drug-induced QTc prolongation dangerously underestimates proarrhythmic poten- amiodarone reduces transmural dispersion of repolarization in the canine heart. tial: Lessons from terfenadine. Journal of Cardiovascular Pharmacology, 57,589–597. Journal of Cardiovascular Electrophysiology, 8, 1269–1279. ICH Hamonized Tripartite Guideline (E14) (2005). The clinical evaluation of QT/QTc interval Shah, R. R. (2004). Pharmacogenetic aspects of drug-induced torsade de pointes: Po- prolongation and proarrhythmic potential for non-antiarrhythmic drugs. Available tential tool for improving clinical drug development and prescribing. Drug Safety, from: http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/ 27, 145–172. Efficacy/E14/Step4/E14_Guideline.pdf. Shah, R. R. (2010). Drug-induced QT interval shortening: potential harbinger of Jacobson, I., Carlsson, L., & Duke, G. (2011). Beat by Beat QT interval variability, but not proarrhythmias and regulatory perspective. British Journal of Pharmacology, 159, QT prolongation per se, predicts drug-induced torsades de pointes in the 58–69. anaesthetized methoxamine-sensitized rabbit. Journal of Pharmacological and Tox- Song, Y., Wu, L., Shryock, J. C., & Belardinelli, L. (2002). Selective attenuation of icological Methods, 63,40–46. isoprenaline-stimulated arrhythmias activity by a partial agonist of adenosine A1 Kashani, A., & Barold, S. S. (2005). Significance of QRS complex duration in patients receptor. Circulation, 105, 118–123. with heart failure. Journal of the American College of Cardiology, 46, 2183–2192. Thomsen, M. B., Verduyn, S. C., Stengl, M. S., Beekman, J. D. M., de Pater, G., van Opstal, Kennedy, R. H., Seifen, E., Akera, T., & Brody, T. M. (1986). Effects of bay K-8644 on inotropic J., et al. (2004). Increased short-term variability of repolarization predicts and arrhythmogenic actions of digoxin. European Journal of Pharmacology, 124,11–20. d-sotalol-induced torsades de pointes in dogs. Circulation, 110, 2453–2459. Laverty, H., Benson, C., Cartwright, E., Cross, M., Garland, C., Hammond, T., et al. (2011). Watanabe,H.,&Knollmann,B.C.(2011).Mechanismunderlyingcatecholaminergic How can we improve our understanding of cardiovascular safety liabilities to de- polymorphic ventricular tachycardia and approaches to therapy. Journal of velop safer medicines? British Journal of Pharmacology, 163, 675–693. Electrocardiology, 44,650–656. Lenz, T. L., & Hilleman, D. E. (2000). Dofetilide: A new antiarrhythmic agent approved Yan, G. X., Shimizu, W., & Antzelevitch, C. (1998). Characteristics and distribution of M for conversion and/or maintenance of atrial fibrillation/atrial flutter. Drugs of Cells in arterially perfused canine left ventricular wedge preparations. Circulation, Today, 36, 759–771. 98, 1921–1927. Liu, T. G., Brown, B. S., Wu, Y., Antzelevitch, C., Kowey, P. R., & Yan, G. X. (2006). Blinded Yan, G. X., & Antzelevitch, C. (1999). Cellular basis for the Brugada syndrome and other validation of the isolated arterially perfused rabbit ventricular wedge in preclinical mechanisms of arrhythmogenesis associated with ST-segment elevation. Circula- assessment of drug-induced proarrhythmias. Heart Rhythm, 3, 948–956. tion, 100, 1660–1666. Liu, H., Yang, L., Jin, M. W., Sun, H. Y., Huang, Y., & Li, G. R. (2010). The selective estrogen Yan, G. X., Rials, S. J., Wu, Y., Liu, T., Xu, X., Marinchak, R. A., et al. (2001). Ventricular receptor modulator raloxifene inhibits cardiac delayed rectifier potassium currents hypertrophy amplifies transmural repolarization dispersion and induces early and voltage-gated sodium current without QTc interval prolongation. Pharmaco- afterdepolarization. American Journal of Physiology — Heart and Circulatory Physiol- logical Research, 62, 384–390. ogy, 281, H1968–H1975. Lu, H. R., Vlaminckx, E., Van de Water, A., & Gallacher, D. J. (2006). Calmodulin antag- Yan, G. X., Lankipalli, R. S., Burke, J. F., Musco, S., & Kowey, P. R. (2003). Ventricular re- onist W-7 prevents sparfloxacin-induced early afterdepolarizations (EADs) in iso- polarization components on the electrocardiogram: Cellular basis and clinical sig- lated rabbit Purkinje fibers: Importance of beat-to-beat instability of the nificance. Journal of the American College of Cardiology, 42, 401–409. repolarization. Journal of Cardiovascular Electrophysiology, 17, 415–422. Ulens, C., Daenens, P., & Tygat, J. (1999). Norpropoxyphene-induced cardiotoxicity is Lu, H. R., Vlaminckx, E., Hermans, A., Rohrbacher, J., Van Ammel, K., Towart, R., et al. associated with changes in ion-selectivity and gating of HERG currents. Cardiovas- (2008). Predicting drug-induced changes in Qt interval and arrhythmias: cular Research, 44, 568–578. QT-shortening drugs point to gaps in the ICHS7B guideline. British Journal of Phar- Van der Linde, H., Van de Water, A., Loots, W., Van Deuren, B., Lu, H. R., Van Ammel, K., macology, 154, 1427–1438. et al. (2005). A new method to calculate the beat-to-beat instability of QT duration Lu, H. R., Rohrbacher, J., Vlaminckx, E., Van Ammel, K., Yan, G. -X., & Gallacher, D. J. (2010). in drug-induced long QT in anaesthetized dogs. Journal of Pharmacological and Tox- Predicting drug-induced lowing of conduction and proarrhythmias: Identifying the icological Methods, 52, 168–177. “bad” sodium channel blockers. British Journal of Pharmacology, 160,60–76. Ohkubo, K., Watenabe, I., Okumura, Y., Ashino, S., Kofune, M., Nagashima, K., et al. Lu, H. R., Hermans, A. N., & Gallacher, D. J. (2012). Does terfenadine-induced ventricular (2011). Prolonged QRS duration in lead V2 and risk of life-threatening ventricular tachycardia/fibrillation directly relate to its QT prolongation and Torsades de arrhythmias in patients with Brugada syndrome. International Heart Journal, 52, pointes? British Journal of Pharmacology, 166, 1490–1502. 98–102. Madias, J. E. (2008). Drug-induced QRS morphology and duration changes. Cardiology Wu, Y., Carlsson, L., Liu, T., Kowey, P. R., & Yan, G. -X. (2005). Assessment of the Journal, 15, 505–509. proarrhythmic potential of the novel antiarrhythmic agent AZD7009 and dofetilide Mines, G. R. (1913). On dynamic equilibrium in the heart. Journal of Physiology (London), in experimental models of torsades de pointes. Journal of Cardiovascular Electro- 46, 349–383. physiology, 16, 898–904. 本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP 图书馆。图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具