Robust anti-arrhythmic efficacy of verapamil and flunarizine against dofetilide-induced TdP arrhythmias is based upon a shared and a different mode of action

Abstract

BACKGROUND AND PURPOSE

The high predisposition to Torsade de Pointes (TdP) in dogs with chronic AV-block (CAVB) is well documented. The anti-arrhythmic efficacy and mode of action of Ca²⁺ channel antagonists, flunarizine and verapamil against TdP were investigated.

EXPERIMENTAL APPROACH

Mongrel dogs with CAVB were selected based on the inducibility of TdP with dofetilide. The effects of flunarizine and verapamil were assessed after TdP and in different experiments to prevent dofetilide-induced TdP. Electrocardiogram and ventricular monophasic action potentials were recorded. Electrophysiological parameters and short-term variability of repolarization (STV) were determined. In vitro, flunarizine and verapamil were added to determine their effect on (i) dofetilide-induced early after depolarizations (EADs) in canine ventricular myocytes (VM); (ii) diastolic Ca²⁺ sparks in RyR2^R4496+/+ mouse myocytes; and (iii) peak and late I_Na in SCN5A-HEK 293 cells.

KEY RESULTS

Dofetilide increased STV prior to TdP and in VM prior to EADs. Both flunarizine and verapamil completely suppressed TdP and reversed STV to baseline values. Complete prevention of TdP was achieved with both drugs, accompanied by the prevention of an increase in STV. Suppression of EADs was confirmed after flunarizine. Only flunarizine blocked late I_Na. Ca²⁺ sparks were reduced with verapamil.

CONCLUSIONS AND IMPLICATIONS

Robust anti-arrhythmic efficacy was seen with both Ca²⁺ channel antagonists. Their divergent electrophysiological actions may be related to different additional effects of the two drugs.

Introduction

In numerous pro-arrhythmic circumstances, calcium [Ca²⁺] overload of the sarcoplasmic reticulum (SR) is the cause of Ca²⁺ leak through ryanodine receptors (RyR). The ensuing Ca²⁺ sparks increase cytosolic [Ca²⁺]_i which in turn may activate the sodium-calcium exchanger (NCX) to generate delayed afterdepolarizations (DADs) and trigger ventricular arrhythmias (de Groot et al., 2000; Pogwizd et al., 2001). The ‘Ca²⁺ antagonists’, flunarizine and verapamil, suppress DADs or DAD-dependent ventricular tachycardias (VTs) either induced by ouabain (Rosen and Danilo, 1980; Jonkman et al., 1986; Vos et al., 1990; Park et al., 1992) or by cathecholamines (Park et al., 1992). In addition, both drugs have also been shown to be effective against Torsade de Pointes (TdP) arrhythmias, whether seen in congenital (Shimizu et al., 1995) or in acquired long QT syndromes (January et al., 1988; Cosio et al., 1991; Verduyn et al., 1995; Carlsson et al., 1996; Aiba et al., 2005; Gallacher et al., 2007). The latter, however, are more likely initiated by early after depolarizations (EADs). The mechanisms underlying these EADs can be reactivation of I_Ca,L, a persistent I_Na or the NCX-mediated inward current (January et al., 1988; Volders et al., 1997; Sipido et al., 2000; Antoons et al., 2007). Although both flunarizine and verapamil belong to the category of Ca²⁺ antagonists, they have additional actions (Zhang et al., 1999; Trepakova et al., 2006), such as blocking the delayed rectifier current (I_Kr) that may negatively affect their anti-arrhythmic efficacy against repolarization-dependent VTs.

The canine model of chronic AV-block (CAVB) has been used to initiate both DAD- and EAD-dependent VTs (Vos et al., 1990; Verduyn et al., 1995; Volders et al., 1997; de Groot et al., 2000; Sipido et al., 2000). The enhanced susceptibility to triggered arrhythmias in this model has been related to SR Ca²⁺ overload and to a diminished repolarization reserve (Vos et al., 1990; de Groot et al., 2000; Sipido et al., 2000; Oros et al., 2008). In this setting, quantification of beat-to-beat variability of repolarization duration (BVR) by short-term variability (STV) has been shown to be a better parameter to predict proarrhythmic predisposition than the QT-time (Thomsen et al., 2006; Oros et al., 2008).

The objectives of this study were to determine if: (i) flunarizine and verapamil prevented and/or suppressed dofetilide-induced TdP in dogs with CAVB; (ii) flunarizine and verapamil improved repolarization reserve, as quantified by STV; (iii) flunarizine was effective against dofetilide-induced increases in STV and EADs in ventricular myocytes isolated from dogs with CAVB; and (iv) their mode of action on Ca²⁺ sparks and late I_Nain vitro differed.

Methods

General

All animal care and experimental handling was in accordance with the ‘European Directive for the Protection of Vertebrate Animals used for Experimental and Scientific Purpose, European Community Directive 86/609/CEE’ and under the regulations of The Committee for Experiments on Animals of Utrecht University, the Netherlands.

A total of 26 adult mongrel dogs (Marshall, USA; 23 ± 3 kg, 16 females) were included. Four weeks after induction of complete AV-block, 22 animals were given a dofetilide test (i.v. infusion of 0.025 mg·kg⁻¹ for 5 min). In this group, five dogs were excluded because they had TdP at baseline (n= 3) or they were non-inducible (n= 2). Repeatability and reproducibility of arrhythmias have been well studied in this model (Oros et al., 2008). In a second group of four animals with CAVB, verapamil and lidocaine were administered in combination to explore their effect on electrophysiological parameters.

All experiments were performed under complete anesthesia, induced with barbiturates and maintained during the experiments with isoflurane (1.5 %). The detailed description of experimental procedures, AV-node ablation, electrocardiogram and monophasic action potential (MAP) recordings (with MAP duration, MAPD at 90 % repolarization), definitions and data analysis, including determination of STV_LV from the left ventricle (LV) MAPD, have been already published (Thomsen et al., 2004; Oros et al., 2006). Early ectopic activity was defined as ectopic beats (EB) initiating before the end of the preceding T wave. Distinction between single (SEBs) and multiple ectopic beats (MEBs) was made as the latter are considered more proarrhythmic (Belardinelli et al., 2003).

Anti-arrhythmic protocols in vivo

Experimental protocols were performed, separated by at least 2 weeks intervals:

Suppression protocol

Approximately 10 min after the start of dofetilide when TdP was reproducibly seen, flunarizine (n= 10; 2 mg·kg⁻¹ for 2 min, i.v.) or verapamil (n= 7; 0.4 mg·kg⁻¹ for 3 min, i.v) were administered to suppress pro-arrhythmic activity.

Validation that TdP remained present in the subsequent 10 min period after dofetilide was recently provided (Oros et al., 2008). To allow comparison of verapamil and flunarizine, the dose of verapamil chosen, had a similar negative inotropic effect in dogs (Vos et al., 1992; Bril et al., 1996) as seen with the anti-arrhythmic dose of flunarizine. Moreover, these equipotent negative haemodynamic effects were confirmed in six sinus rhythm dogs using a 7F catheter (Sentron, Roden, the Netherlands): LV end-systolic pressure decreased by 20% with both drugs: flunarizine from 94 ± 9 to 75 ± 9 mmHg (P < 0.05) and verapamil from 87 ± 13 to 67 ± 9 mmHg (P < 0.05).

Prevention protocol

Whether vulnerability to dofetilide-induced TdP could be prevented by pretreatment with flunarizine (n= 8) or verapamil (n= 6) was investigated in this set of experiments. The dose of dofetilide used was the same as in the previous experiment. Three animals were tested serially with both verapamil and flunarizine.

Combination of drugs

To test if the electrophysiological effects seen with flunarizine were due to a combined block of I_Ca,L and Iate I_Na, verapamil (0.2 mg·kg⁻¹ for 1.5 min) was followed 5 min later by lidocaine (1.5 mg·kg⁻¹ for 1 min), a preferential late I_Na blocker (Fredj et al., 2006).

In vitro experiments

The following concentrations of drugs were used: 1 µM dofetilide, 1 µM and 10 µM flunarizine or verapamil.

Effects of flunarizine on cellular STV

Single myocytes from CAVB dogs were enzymically isolated (Volders et al., 1998). Action potentials were triggered in whole-cell current clamp mode with 2 ms current injections at a cycle length of 2000 ms and recorded with PClamp9 software (Molecular Devices, Sunnyvale, CA, USA). Action potential duration (APD) was measured at 90% repolarization and cellular STV was calculated from 20 successive APDs (STV_APD) similar to the in vivo quantification(Volders et al., 1998; Thomsen et al., 2004). Experiments were performed in Tyrode solution containing (in mmol·L⁻¹): 137 NaCl, 5.4 KCl, 0.5 MgCl₂, 1.8 CaCl₂, 11.8 HEPES and 10 glucose, pH 7.4. Pipettes had a resistance of 2–3 MΩ when filled with pipette solution, containing (in mmol·L⁻¹): 130 KCl, 10 NaCl, 10 HEPES, 5 MgATP and 0.5 MgCl₂, pH 7.2. As with the in vivo experiments, two experimental protocols were used:

Protocol 1: effects of flunarizine on baseline cellular APD and STV in eight myocytes isolated from the LV of four dogs.

Protocol 2: effects of flunarizine on dofetilide-induced EADs. If 1 µM dofetilide induced EADs, flunarizine was added to the Tyrode solution to test its suppressive effect on EADs and dofetilide-increased APD and STV. For these experiments another eight cells [n= 4 right ventricle (RV) and n= 4 LV] from five dogs were used.

Effects of flunarizine and verapamil on I_Na

For recording cardiac peak and late I_Na, SCN5A-HEK 293 cells (expressing Na_v1.5) were superfused with bath solution containing (in mmol·L⁻¹): 140 NaCl, 4.0 KCl, 1.8 CaCl₂, 0.75 MgCl₂ and 5 HEPES (pH adjusted to 7.4 with NaOH). The pipette solution contained (in mmol·L⁻¹): 20 CsCl, 120 CsF, 2 EGTA and 5 HEPES (pH adjusted to 7.4 with CsOH). All experiments were performed at 21 ± 1°C. Whole-cell membrane current was recorded as previously described (Hamill et al., 1981). Computer software (pCLAMP 10.0, Molecular Devices, Sunnyvale, CA) was used to generate voltage-clamp protocols. Patch-clamp amplifier (Multiclamp 700B, Molecular Devices) data were sampled at 5 kHz. Whole-cell capacitance was compensated using the internal voltage-clamp circuitry and about 75–80% of series resistance was compensated. Membrane potentials were not corrected for junction potentials that arise between the pipette and bath solution. Cells were held at −140 mV and dialysed for 5 min before I_Na recording. Data analysis of all measured currents was performed using pCLAMP 10.0 and Origin 7.0 (MicroCal, Northampton, MA) software. To measure the extent of tonic block (first-pulse) by flunarizine or verapamil on peak I_Na, 24-ms depolarizing steps to −20 mV from a holding potential of −140 mV were applied to cells at a rate of 0.1 Hz. The magnitude of peak I_Na in the presence of drug was normalized to the respective control value. To measure the effect of flunarizine or verapamil on late I_Na, the normally small late I_Na was augmented by exposure of cells to Anemone Toxin-II (ATX-II; 3 nM; Song et al., 2008), and the effect of drug to reduce the ATX-II-induced late I_Na was determined. Late I_Na was defined as the magnitude of I_Na between 200 and 220 ms after application of a 220-ms depolarizing step to −20 mV from a holding potential of −140 mV applied at a rate of 0.1 Hz.

Effects of flunarizine and verapamil on Ca²⁺ sparks

Abnormally high spontaneous Ca²⁺ release in diastole (Ca²⁺ sparks) were recorded in intact quiescent myocytes enzymically isolated from homozygous mice carrying the mutation R4496C of the cardiac ryanodine receptor (RyR2^R4496C+/+), which underlies catecholamine polymorphic ventricular tachycardia (CPVT) (Priori et al., 2001; Fernandez-Velasco et al., 2009). To measure the effects of verapamil and flunarizine on spontaneous Ca²⁺ spark activity, cells were loaded with fluorescent Ca²⁺ indicator (Fluo-3 AM) as previously described (Fernandez-Velasco et al., 2009). Cells were recorded under continuous perfusion with Tyrode solution (in mmol·L⁻¹: 140 NaCl, 4 KCl, 1.1 MgCl₂, 10 HEPES, 10 glucose, 1.8 CaCl₂; pH = 7.4, with NaOH) before and following the addition of flunarizine or verapamil for 10 min.

Images were obtained by confocal microscopy (Meta Zeiss LSM 510, objective w.i. 63x, n.a. 1.2) in the line scan mode as previously explained. Image analyses were performed by homemade routines using IDL software (Research System Inc.). Images were corrected for the background fluorescence.

Statistical analysis

Pooled data are expressed as mean ± standard deviation except the result on I_Na where results are presented as mean ± SEM. For the effects of drugs over time, comparisons were performed using a one-way repeated-measures anova followed by a Bonferroni correction. For non-parametric comparisons, the Kruskal–Wallis test was used.

Materials

Flunarizine was supplied by Janssen Pharmaceutica N.V.; verapamil (Isoptin) and isoflurane was from Abbott, Europe; lidocaine from Braun Melsungen AG, Germany. ATX-II and all other chemicals were from Sigma. Channel, receptor and drug nomenclature follows Alexander et al. (2009).

Results

Antiarrhythmic effects of flunarizine

Flunarizine suppression

Dofetilide induced TdP with a median duration of 6.8 s, after 3.1 ± 1 min. After adding flunarizine, all arrhythmias disappeared with the exception of some SEBs in one dog (Figure 1 and Table 1). These anti-arrhythmic effects remained present for more than 10 min. Thereafter, some MEBs returned, albeit less severe. Electrophysiologically, dofetilide increased all repolarization parameters (QT, QT_C, LVMAPD and RVMAPD) and STV_LV before the first EB (2.5 ± 0.5 min. after start dofetilide). Flunarizine decreased the dofetilide-augmented STV_LV and all the other repolarization parameters, to a level similar to baseline (Table 1 upper part and Figure 1).

Figure 1.

Upper panel: anti-arrhythmic effects of flunarizine (suppression) against dofetilide-induced TdP (left) and ectopic activity (right; as single ectopic beats, SEB and multiple ectopic beats, MEB) is shown with an individual example (middle part) of lead II electrocardiogram (ECG) and left ventricular monophasic action potential (LV MAP) recordings (printed at 10 mm·s⁻¹ speed and calibrated at 1 mV per cm for ECG and 20 mV for the MAP recording) on scale paper in baseline (left), with TdP (middle) and after flunarizine. Lower panel illustrates continuous short-term variability (STV_LV) quantification for this experiment. *P < 0.05 versus baseline. TdP, Torsade de Pointes.

Table 1.

Summary of electrophysiological parameters and arrhythmic events in flunarizine suppression and prevention experiments

	Baseline 1	Dofetilide	Flunarizine
RR	1181 ± 87	1291 ± 140	1219 ± 251
QT	436 ± 44	566 ± 29*	435 ± 36$
QTC	421 ± 49	553 ± 40*	425 ± 38$
LV MAPD	355 ± 35	492 ± 53*	367 ± 42$
RV MAPD	310 ± 32	395 ± 68*	333 ± 30$
ΔMAPD	51 ± 28	97 ± 56*	48 ± 32$
STVLV	1.8 ± 0.5	4.5 ± 1.5*	1.5 ± 0.6$
TdP	0 ± 0	11 ± 8*	0 ± 0$
MEB	0 ± 0	13 ± 14*	0 ± 0$
SEB	1 ± 2	48 ± 58*	1 ± 3$

	Baseline 2	Flunarizine	Dofetilide
RR	1239 ± 329	1291 ± 390	1410 ± 462
QT	422 ± 51	380 ± 50*	494 ± 92*#
QTC	413 ± 51	369 ± 41*	476 ± 77*#
LV MAPD	299 ± 44	277 ± 36	380 ± 65*#
RV MAPD	286 ± 39	275 ± 44	348 ± 73*#
ΔMAPD	42 ± 27	22 ± 16	38 ± 42
STVLV	1.5 ± 0.6	1.0 ± 0.5*	1.4 ± 0.5
TdP	0 ± 0	0 ± 0	0 ± 0
MEB	0 ± 1	0 ± 0	0 ± 0
SEB	3 ± 6	3 ± 5	6 ± 10

Maximal effects of flunarizine (5 min) in suppression experiments (upper part) and at the end of the infusion (2 min) in prevention experiments are shown (lower part). Arrhythmias are quantified as average number of events (TdP, MEB, SEB) per 10 min, except for the pretreatment with flunarizine (lower part) where after 5 min dofetilide was added. All electrophysiological parameters are expressed in ms and arrhythmias as average number per time interval.

^*P < 0.05 versus baseline;

^$P < 0.05 versus dofetilide;

^#P < 0.05 versus flunarizine pretreatment.

LV, left ventricle; MAPD, duration of the monophasic action potential; MEB, multiple ectopic beat; RV, right ventricle; SEB, single ectopic beat; STV_LV, short term variability of repolarization, computed from LV MAPD; TdP, Torsade de Pointes.

Flunarizine prevention

Pretreating the same animals with flunarizine resulted in complete prevention of TdP (Figure 2, upper part). During a 10 min. period, dofetilide could only induce few single EBs (6 ± 10 beats/10 min). Flunarizine significantly decreased baseline STV_LV and QTc. After adding dofetilide, STV_LV remained at a level similar to baseline, whereas an increase in QTc could not be prevented by this drug completely (Figure 2 and Table 1, lower part).

Figure 2.

TdP prevention (upper panel) with flunarizine is presented in two serial experiments, first dofetilide alone (left) and with flunarizine pretreatment (right). The effects on QT/QTc (middle part) and short-term variability (STV_LV) in individuals as well as average (lower part) are plotted. TdP, Torsade de Pointes.

Effect of flunarizine on baseline cellular BVR and on dofetilide-induced EADs

In untreated isolated myocytes from dogs with CAVB, flunarizine shortened (at 10 min) both APD (from 418 ± 116 ms in baseline to 312 ± 74 ms, P < 0.05) and cellular STV_APD (baseline 20 ± 10 ms to 11 ± 4 ms, P < 0.05). The time course of changes in APD and STV during an experiment is shown in Figure 3A.

Figure 3.

Anti-arrhythmic effects of flunarizine in isolated ventricular myocytes of the chronic AV-block (CAVB) dog are depicted. (A) 20 superimposed consecutive action potentials (APs) in baseline (left) and after flunarizine (right) as well as the time course of APD (dots) and short-term variability (STV_APD, continuous red line), baseline and with flunarizine perfusion are shown. (B) Similar, 20 superimposed APs in baseline (left), with dofetilide-induced EADs (arrow in middle panel) and after EADs suppression with flunarizine (right) and the temporal behaviour of APD and STV_APD are shown for this experiment. EADs, early after depolarizations.

In dofetilide-treated cells, APD increased from 337 ± 119 to 507 ± 153 ms (P < 0.05) and STV from 14 ± 14 to 65 ± 34 ms (P < 0.05). EADs occurred in eight from a total of nine cells. Addition of flunarizine suppressed all dofetilide-induced EADs (from 8/8 to 0/8, P < 0.05) and reversed APD (289 ± 60 ms) and cellular STV (11 ± 5 ms) to baseline values (P > 0.5 vs. baseline). A representative example is shown in Figure 3B.

Antiarrhythmic effects of verapamil

Verapamil suppression

Similar arrhythmia was seen with dofetilide in this group: TdP induction after 3.7 ± 1 min with a median duration of 7.1 s. All TdP and MEBs were suppressed, while some SEBs remained in three dogs (Table 2, upper part). Verapamil did not affect the dofetilide-prolonged repolarization duration (QT, QTc, LVMAPD and RVMAPD) but reduced the variability of repolarization STV_LV to a level similar to control (Table 2, upper part).

Table 2.

Summary of electrophysiological parameters and arrhythmic events in verapamil suppression and prevention experiments

	Baseline 1	Dofetilide	Verapamil
RR	1361 ± 190	1520 ± 210*	1476 ± 166
QT	456 ± 67	611 ± 92*	557 ± 97*
QTC	424 ± 62	566 ± 87*	516 ± 90*
LV MAPD	349 ± 88	505 ± 110*	466 ± 95*
RV MAPD	305 ± 63	446 ± 135*	392 ± 108*
ΔMAPD	44 ± 39	72 ± 36	92 ± 92
STVLV	1.7 ± 0.4	3.2 ± 1.1*	1.5 ± 0.7$
TdP	0 ± 0	9 ± 5*	0 ± 0
MEB	0 ± 0	9 ± 4*	0 ± 0
SEB	1 ± 1	50 ± 31*	9 ± 15

	Baseline 2	Verapamil	Dofetilide
RR	1285 ± 202	1212 ± 228	1464 ± 240#
QT	442 ± 71	436 ± 57	651 ± 47*#
QTC	417 ± 58	417 ± 41	611 ± 34*#
LV MAPD	332 ± 68	328 ± 34	554 ± 77*#
RV MAPD	324 ± 43	318 ± 37	545 ± 53*#
ΔMAPD	32 ± 29	33 ± 16	30 ± 16
STVLV	1.3 ± 0.4	1.4 ± 0.6	2.3 ± 1.4
TdP	0 ± 0	0 ± 0	0.2 ± 0.4
MEB	0 ± 1	0.2 ± 0.4	3 ± 6
SEB	2 ± 4	1 ± 2	28 ± 44

Maximal effects of verapamil in suppression experiments (at 10 min, upper part) and at the end of the infusion (3 min) in prevention experiments are shown (lower part). All electrophysiological parameters are expressed in ms and arrhythmias (TdP, MEB, SEB) as average number per time interval.

^*P < 0.05 versus baseline;

^$P < 0.05 versus dofetilide;

^#P < 0.05 versus verapamil pretreatment.

LV, left ventricle; MAPD, duration of the monophasic action potential; MEB, multiple ectopic beat; RV, right ventricle; SEB, single ectopic beat; STV_LV, short term variability of repolarization, computed from LV MAPD; TdP, Torsade de Pointes.

Verapamil prevention

Verapamil pretreatment also prevented TdP induction remarkably, only one self-terminating TdP was seen. However, dofetilide was still able to generate numerous SEBs and few MEBs (Table 2, lower part). Verapamil did not change baseline electrophysiological parameters, or STV_LV. The duration of repolarization parameters (QT, QTc, LVMAPD and RVMAPD) was prolonged after adding dofetilide despite verapamil pretreatment. However, the variability of repolarization (STV_LV) was not significantly increased after dofetilide (Table 2, lower part).

Analysis of the mode of action

Effects of flunarizine and verapamil on I_Na

Figure 4 shows the effect of flunarizine (Figure 4A, left) and verapamil (Figure 4B, right) on late I_Na induced by ATX-II (Figure 4). Flunarizine (1 µM, Figure 4A) inhibited late I_Na by 94.4 ± 2.3%, n= 5 cells, P < 0.05). However, at a higher concentration (10 µM), flunarizine had no effect on peak I_Na (tonic block; 4.3 ± 3.0%, n= 4 cells, P > 0.05). In contrast to flunarizine, verapamil (Figure 4B, 10 µM) had no effect on either late I_Na (n= 5 cells, P > 0.05) or peak I_Na (tonic block; 10 µM, n= 6 cells (P > 0.05 and 30 µM, n= 4 cells, P > 0.05), as compared with control.

Figure 4.

Effects of flunarizine (left) and verapamil (right) on late I_Na: representative recordings of late I_Na from a single cell in the absence of drug (black line), during superfusion with 3 nM ATX-II (ATX, grey line) and during superfusion with 1 µM flunarizine (left; red line) or 10 µM verapamil (right; blue line). Insets: expanded traces (last 50 ms following depolarizing pulse) of late I_Na in the absence (black line) and presence of flunarizine (red line) or verapamil (blue line) respectively.

Ca²⁺ sparks study

Acute application of these drugs on the frequency of spontaneous Ca²⁺ sparks in cardiac myocytes expressing a gain-of-function mutation in the RyR2 was examined. Flunarizine (1 µM) did not change the frequency of spontaneous Ca²⁺ sparks in RyR^R4496C cells (Figure 5A, P > 0.05). In contrast, verapamil (10 µM) significantly reduced spontaneous Ca²⁺ spark activity by ≈35% (Figure 5B, P < 0.005).

Figure 5.

(A) Left, representative line-scan images of spontaneous Ca²⁺ sparks recorded in a RyR2^R4496C+/+ ventricular myocyte in the absence (top) or presence (bottom) of 1 µM flunarizine. Right, Ca²⁺ spark occurrence before (control) and during flunarizine (n= 8 cells). (B) Similar, images of spontaneous Ca²⁺ sparks in the absence (top) or presence (bottom) of 10 µM verapamil. Right panel shows the average data in control and with verapamil (n= 11 cells). *P < 0.05 versus control. RyR, ryanodine receptor.

In vivo effects of verapamil in combination with lidocaine

To confirm that STV_LV reduction in baseline by flunarizine was in part due to inhibition of late I_Na, the effects of verapamil combined with lidocaine were explored. The combination of these drugs shortened the duration of repolarization (QTc from 353 ± 35 to 306 ± 21 ms, P < 0.05) and STV_LV was significantly reduced, effects similar to those seen with flunarizine alone (Figure 6).

Figure 6.

Effects of flunarizine (left), verapamil (middle) and the combination of verapamil and lidocaine on baseline short-term variability (STV_LV). Effects of these drugs on baseline STV_LV is shown for individual animals (thin lines) and as a mean value for each group (thick line).

Discussion

The most important findings of this study can be summarized as follows: (i) both Ca²⁺ antagonists flunarizine and verapamil were equally and markedly effective in suppressing and preventing dofetilide-induced TdP; (ii) this anti-arrhythmic effect was reflected in STV_LV, but not in QT or LV MAPD; (iii) flunarizine but not verapamil decreased BVR in baseline, which could be ascribed to its additional late I_Na blocking effect; and (iv) verapamil modestly reduced Ca²⁺ sparks, an effect not seen with flunarizine.

Repolarization-dependent ventricular arrhythmias and EADs

The common Ca²⁺ antagonism of verapamil and flunarizine was used to investigate whether they could improve repolarization reserve, as reflected in protection against dofetilide-induced TdP. These repolarization-dependent arrhythmias normally occur under conditions in which this reserve is ‘challenged beyond capacity’, as in the long QT syndromes, or in congestive heart failure (Tomaselli et al., 1994). Lately, it has been suggested that this reserve can be estimated by STV (Thomsen et al., 2004; 2006; 2007; Hinterseer et al., 2008; Oros et al., 2008). EADs and EAD-dependent triggered activity have been considered as possible initiation mechanisms for TdP in long QT syndromes (el-Sherif et al., 1996; Belardinelli et al., 2003). To develop new anti-arrhythmic drugs, it is important to understand possible targets that are key components of the generation of EADs, such as L-type Ca²⁺ channels, Na⁺ channels (with peak and late I_Na), RyR and its regulating unit calstabin2 (FKBP12.6), and the NCX. EADs have at least two different ways in which they may be generated: (i) window currents, either through the I_Ca,L or late I_Na; and (ii) NCX-mediated inward currents due to an abnormal Ca²⁺ release from the SR. The I_Ca,L has been studied extensively as block of I_Ca,L by verapamil or nitrendipine prevented EADs from developing (Marban et al., 1986; January et al., 1988) and as regional differences in the expression levels of L-type Ca²⁺ channels have implications for the origin of EADs (Sims et al., 2008). This effect may also explain why I_Kr blockade by verapamil (Zhang et al., 1999) and flunarizine are not pro-arrhythmic because an additional block of I_Ca,L protects the heart from developing EADs, despite the QT lengthening induced by I_Kr block (Bril et al., 1996). This balance between I_Ca,L recovery and ventricular repolarization serves also as a physiological stabilizer (Guo et al., 2008).

The second theory, the involvement of abnormal SR Ca²⁺ release in generating EADs, is more controversial (Volders et al., 1997; Antoons et al., 2007). Nevertheless there is evidence that DADs and EADs may occur in the same preparation (Marban et al., 1986; Priori and Corr, 1990; Volders et al., 1997; Song et al., 2008), suggesting a similar etiology. Moreover, EADs and calcium transients have been related (Hamlin and Kijtawornrat, 2008). Indirectly, activation of calcium/calmodulin-dependent protein kinase II (CaMKII) due to an increase in [Ca²⁺]_i might facilitate both I_Ca,L and I_Na, thus inducing EADs and DADs (Anderson et al., 1998). Because of the number of actions of the drugs, the relevance of this alternative (for the conditional phase) was difficult to assess. However, measuring diastolic Ca²⁺ sparks known to underlie DAD generation is an interesting approach to address the question of whether flunarizine and verapamil inhibit the disturbed SR Ca²⁺ release events.

Calcium antagonists: flunarizine and verapamil

Flunarizine does not belong to the cardiovascular categories of Ca²⁺ antagonists. Clinically, the drug has been used to treat neurological disorders, such as migraine and has been termed a calcium overload blocker (van Zwieten, 1986). The latter implies that flunarizine may have an intracellular target. However, until now, only sarcolemmal effects have been described. Besides blocking three types of Ca²⁺ channels (Tytgat et al., 1991; 1996;), I_Ca,L (IC₅₀= 4.6–10 µM), I_Ca,N (0.8 µM) and I_Ca,T (3.3–11 µM), flunarizine is also a potent I_Kr (5.7 nM) and I_Ks (0.7 µM) blocker (Trepakova et al., 2006).

Verapamil is known to block I_Ca,L (0.6–15.5 µM) (Hosey and Lazdunski, 1988), I_Kr (0.1 µM) (Zhang et al., 1999) and I_Ks (5.7–6.3 µM) (Aiba et al., 2005). According to the supplier, our dose of flunarizine will reach a total plasma concentration around 1.7 µM (828 ng·mL⁻¹, MW 477.4) while, for verapamil, this value is around 0.5 µM (Fossa et al., 2002).

In susceptible dogs with CAVB, both drugs were very effective (100% efficacy) in preventing and terminating dofetilide-induced TdP. They were much more effective stronger than other drugs such as the late I_Na blockers, ranolazine and lidocaine, which were effective in approximately 60% of the animals (Antoons et al., 2010), whereas the I_K,ATP agonist levcromakalim was slightly more effective (70%, unpublished data). This confirms that inhibition of I_Ca,L is a very effective way to treat dofetilide-induced TdP, assuming that no other actions are involved (see below).

Additional modes of action

In order to gain more insight in the mode of action of flunarizine and verapamil, we undertook cellular experiments to determine their possible action against late I_Na and Ca²⁺ sparks. Flunarizine but not verapamil blocked late I_Na, which could explain why flunarizine, but not verapamil, was effective against veratridine-induced contractures (Patmore et al., 1989).

Albeit modestly, verapamil (at high concentrations) but not flunarizine could reduce the frequency of diastolic Ca²⁺ sparks in a cell model expressing a gain-of-function mutation in the RyR2 responsible for abnormal Ca²⁺ leakage. These results show that the antiarrhythmic molecular mechanism of verapamil and flunarizine could involve different targets. Regarding flunarizine it is possible to discard an action of flunarizine on RyR2 activity. As mentioned, there is controversy concerning the proposition that SR Ca²⁺ leak may indirectly provide inward currents that contribute to induction of EADs (Fauconnier et al., 2005; Hamlin and Kijtawornrat, 2008). One way to study this is by application of drugs that specifically block this release, like ryanodine or K201. Unfortunately, both drugs do not have a high degree of specificity.

When evaluating the literature concerning ryanodine and its action on DADs or EADs, it becomes apparent that the results are not consistent. Ryanodine is known to be anti-arrhythmic against DADs and DAD-dependent VT (Marban et al., 1986; Priori and Corr, 1990). In dogs with CAVB, ryanodine (10 mg) was effective against drug-induced TdP, whereas ryanodine was not effective against cesium or ATX-II-induced EADs (Marban et al., 1986; Park et al., 1992; Song et al., 2008), but anti-arrhythmic against catecholamine-induced EADs (Priori and Corr, 1990). Ryanodine and flunarizine were both effective against acceleration-induced EADs (Burashnikov and Antzelevitch, 1998). Until there is a specific blocker for unconditional Ca²⁺ leak, it will be difficult to prove SR leakage to be part of the generation of EADs. It is evident however, that adding blocking properties against either late I_Na or Ca²⁺ sparks could generate more anti-arrhythmic ‘strength’. Future studies are necessary to evaluate which of the two additional actions is the most attractive.

Anti-arrhythmic action and STV

The anti-arrhythmic potential of flunarizine and verapamil was clearly reflected by the changes in STV_LV. Its suppressive actions were associated with a reduction in STV_LV, whereas the preventive effects could be seen in keeping STV_LV at low(er) levels. Anti-arrhythmic properties of flunarizine were confirmed in vitro on drug-induced EADs and STV_APD. Thus, STV is an indicator of the ability of the heart to withstand a pro-arrhythmic challenge. However, the action on the other electrophysiological parameters differed. Flunarizine showed a pronounced action on repolarization parameters such as QTc and LV MAPD and cellular APD, whereas the effect of verapamil on repolarization time was much smaller or even absent.

Second, flunarizine decreased baseline STV_LV, suggesting that this drug may increase repolarization reserve. This interpretation is consistent with the greater effect of verapamil combined with lidocaine on STV_LV (Figure 6) and QT LV MAPD than verapamil alone.

The fact that flunarizine decreased APD/QTc while verapamil did not, could in part contribute to the mechanism by which flunarizine reduced STV_LV or STV_APD. However, the contribution of APD to STV_LV was not seen in the suppression experiments where verapamil shortened STV without a significant reduction in APD or QTc.

Study limitations

The results obtained in the SCN5A-HEK 293 cells and/or in myocytes isolated from the transgenic mouse model RyR2^R4496C+/+ may differ from the results that could be obtained from myocytes isolated from CAVB dog hearts.

The blocking effect of verapamil on Ca²⁺ sparks was not very robust and attained at relatively high dosages. Whether this finding has therapeutic consequences is therefore unclear.

In conclusion, a robust anti-arrhythmic efficacy was seen with flunarizine and verapamil. This suppressive and preventive action of the drugs was reflected in STV_LV or cellular STV_APD. Their different electrophysiological response may be related to different additional effects of the two drugs: flunarizine blocks late I_Na, whereas verapamil reduces Ca²⁺ sparks.

Glossary

Abbreviations

APD

action potential duration

BVR

beat-to-beat variability of repolarization duration

[Ca²⁺]_i

intracellular calcium concentration

CAVB

chronic AV-block

DADs

delayed afterdepolarizations

EADs

early after depolarizations

I_Ca,L

L-type calcium current

I_Kr

rapid component of the delayed rectifier current

I_Ks

slow component of the delayed rectifier current

late I_Na

persistent sodium current

LV

left ventricle

MAP

monophasic action potential

MAPD

duration of the monophasic action potential

MEBs

multiple ectopic beats

NCX

sodium-calcium exchanger

RV

right ventricle

RyR

ryanodine receptor

SEBs

single ectopic beats

SR

sarcoplasmic reticulum

STV_APD

short-term variability of repolarization, computed from cellular APD

STV_LV

short-term variability of repolarization, computed from LV MAPD

TdP

Torsade de Pointes

VTs

ventricular tachycardias

Conflicts of interest

SR and LB are employees of Gilead Sciences Inc.