“Early” Class III Drugs for the Treatment of Atrial Fibrillation
Efficacy and Atrial Selectivity of AVE0118 in Remodeled Atria of the Goat
Background— Currently available antiarrhythmic drugs are only moderately effective against atrial fibrillation (AF) and may cause ventricular proarrhythmia. AVE0118 is a blocker of atrium-specific early K+ currents (IKur/Ito).
Methods and Results— Effects of intravenous AVE0118 and dofetilide on atrial effective refractory period (AERP) and inducibility of AF were measured before and after 48-hours of AF-induced electrical remodeling in the goat. During persistent AF (53±19 days), the cardioversion efficacy and effects on atrial wavelength of AVE0118, dofetilide, and ibutilide were evaluated. QT durations were measured during atrial pacing and persistent AF. After 48 hours of AF, the effect of dofetilide on AERP was reduced, and induction of AF was not prevented. In contrast, the class III action of AVE0118 was enhanced, and AF inducibility decreased from 100% to 32% (P<0.001). At 1, 3, and 10 mg · kg−1 · h−1, AVE0118 terminated persistent AF in 1 of 8, 3 of 8, and 5 of 8 goats, respectively. Dofetilide and ibutilide terminated AF in 1 of 5 and 2 of 7 goats. AVE0118 0.5, 1.5, and 5 mg/kg prolonged the AERP during AF and increased the fibrillation wavelength from 6.7±0.6 to 8.5±0.5, 9.7±0.5, and 11.2±0.9 cm (P<0.01). Whereas dofetilide and ibutilide prolonged QT duration, AVE0118 had no appreciable effect.
Conclusions— AVE0118 markedly prolongs the AERP during AF without affecting QT duration. Cardioversion of AF was due to an ≈2-fold increase in fibrillation wavelength. Atrium-selective class III drugs like AVE0118 may be a promising new option for safe and effective cardioversion of AF.
Received January 15, 2004; de novo received March 8, 2004; revision received April 29, 2004; accepted April 30, 2004.
Antiarrhythmic drugs have been used widely for the prevention and termination of atrial fibrillation (AF). Unfortunately, their clinical efficacy has been disappointing, and the risk of proarrhythmia is a significant concern.1,2 The antifibrillatory action of commonly used class III drugs is based on prolongation of the action potential by blockade of the delayed rectifier K+ current (IKr). Clinical studies have shown that these class III drugs can terminate persistent AF in ≈30% of cases.3,4 Their therapeutic range is limited because they prolong the QT duration and induce torsade de pointes arrhythmias.2 Therefore, the development of atrium-specific drugs would be a major step forward.
Wang and colleagues5 showed that an early ultrarapid component of the delayed rectifier (IKur) contributes significantly to repolarization of the human atrial action potential. Because IKur is present in atrial but not ventricular myocytes, they postulated, “It is a potentially promising target for the development of drugs that prevent atrial reentrant arrhythmias without a risk of ventricular pro-arrhythmia.”6
Several experimental IKur blockers are under investigation.7,8 One of them (AVE0118), which, in addition to IKur, blocks the Ito current, is evaluated in the present study. We hypothesized that in electrically remodeled atria, blockade of the early repolarizing currents restores the plateau of the action potential, thereby exerting an antifibrillatory effect. This hypothesis was tested in a goat model of AF in which the electrophysiological and antifibrillatory effects of AVE0118 were compared with the action of conventional class III drugs (dofetilide and ibutilide).
Goat Model of Atrial Fibrillation
Thirteen female goats weighing 52±2 kg were used. The animals were handled according to the European Directive for Animal Research, and the study protocol was approved by the local Animal Investigation Committee. Goats were instrumented as described previously.9 Briefly, during general anesthesia, Teflon-felt plaques with multiple electrodes were sutured onto the free wall of each atrium, Bachmann’s bundle, and the left ventricle. All leads were tunneled subcutaneously to the neck and exteriorized by four 30-pole connectors. Experiments were started 3 to 4 weeks after surgery. AF was induced by a fibrillation pacemaker as described previously.10
The atria were paced with biphasic stimuli of 2-ms duration and 4× threshold. The atrial effective refractory period (AERP) was measured at the free wall of the right and left atria during regular pacing (interval, 400 to 200 ms). Single interpolated stimuli were applied after 8 basic stimuli, starting within the refractory period. The longest interval that failed to capture the atria (2-ms increments) was taken as the AERP. Atrial conduction velocity was measured along Bachmann’s bundle during right atrial pacing. The distance over which conduction velocity was measured ranged from 3.5 to 5 cm.
The length of the fibrillation waves was determined at the right atrial free wall by measuring the refractory period (RPAF) and conduction velocity (CVAF) during AF. RPAF was measured by slow, fixed-rate pacing (1 Hz), resulting in a series of single, randomly coupled, premature stimuli.9 Local capture of AF was evidenced by radial spread of activation from the pacing site and a short delay between stimulus and response. For each coupling interval, the percentage of capture was determined. The shortest interval capturing the atrium ≥50% was taken as the RPAF. CVAF was determined with a mapping electrode containing 5×6 electrodes (interelectrode distance, 4 mm) from the local conduction vectors within areas of 3×3 electrodes. At least 50 AF cycles were used to determine the CVAF.
Inducibility of AF was measured at the right and left atria by single premature stimuli applied during regular pacing (400 ms). In case a premature beat induced a rapid irregular rhythm lasting >1 second, AF was considered inducible. The AF cycle length (AFCL) was measured automatically by an algorithm detecting the negative intrinsic deflection of the fibrillation electrogram. A median value of 300 consecutive intervals was calculated. QT duration was measured during atrial pacing and persistent AF from either an epicardial electrogram or a precordial ECG. Because Bazett’s formula cannot be applied during AF, we used another approach to correct QT duration. In each goat, the relationship between the RR interval and QT duration was determined during 20 seconds of AF. The RR-QT relationship after drug administration was compared with the normal RR-QT relation. In some animals, the effects of class III drugs during persistent AF was evaluated by recording monophasic action potentials (MAPs) under general anesthesia from the endocardium of the right atrium.
The electrophysiological effects of AVE0118 (n=10 goats) and dofetilide (n=6) were measured before and after 48 hours (1 to 4 days) of AF. In 5 goats, both drugs were evaluated. Drug effects were studied on separate days, with at least 10 plasma half-lives between the experiments. AVE0118 and dofetilide were administered intravenously over 1 hour at 3 mg · kg−1 · h−1 and 20 μg · kg−1 · h−1, respectively. AERP, CV, QT duration, and inducibility of AF were measured after 30 minutes of infusion.
In a total of 8 goats, the effects of AVE0118 (n=8), dofetilide (n=5), and ibutilide (n=7) were evaluated during persistent AF (53±19 days). In 6 of these animals, the baseline studies had been performed. AVE0118 was infused in 5 dosages between 0.1 and 10 mg · kg−1 · h−1. Dofetilide and ibutilide were given in dosages of 20 μg · kg−1 · h−1 and 4 mg/h, respectively. Drugs were infused for 1 hour, during which time the AFCL was monitored. After 30 minutes of infusion, RPAF, CVAF, median RR interval, and QT duration were measured. Successful cardioversion was defined as termination of AF within ≤1 hour of drug administration. Plasma concentrations of AVE0118 were determined by Aventis.
Differences between groups were evaluated by paired Student’s t test or by 2-way repeated ANOVA with post hoc Bonferroni’s t test. McNemar’s test was used to compare AF inducibility. Changes in corrected QT duration were evaluated by the 1-sample t test (test value, 0). Differences were considered statistically significant at P<0.05. Results are presented as mean±SEM.
Atrial Refractoriness and Conduction Velocity
Figure 1 shows the effects of dofetilide and AVE0118 before and after 48 hours of AF. In nonremodeled atria, dofetilide and AVE0118 prolonged AERP at all pacing rates. Both drugs showed a smaller class III effect at shorter cycle lengths (reverse frequency dependence). After AF-induced electrical remodeling, the effect of dofetilide was reduced, whereas the action of AVE0118 was enhanced. In Figure 1 (right), the prolongation of atrial refractoriness (Δ-AERP) is plotted. The normal reverse frequency dependence of dofetilide was changed into a frequency-dependent effect after 48 hours of AF because of a loss of class III action at slow rates. The reverse frequency dependence of AVE0118 was reduced. At short cycle lengths (200 ms), the class III effect of AVE0118 was almost tripled (2.8), whereas at 400 ms, the prolongation of atrial refractoriness was enhanced by a factor of 1.4.
Before drug infusion, a physiological rate-dependent slowing of conduction along Bachmann’s bundle was observed from 125±4 to 114±4 cm/s (pacing interval, 400 and 200 ms). After 48 hours of AF, conduction velocity was not changed. Dofetilide did not affect conduction velocity either before or after electrical remodeling (data not shown). At slow pacing rates (400 ms), AVE0118 had no effect on CV in either normal (126±4 versus 123±5 cm/s; P=NS) or remodeled (128±5 versus 126±5 cm/s; P=NS) atria. During rapid pacing (200 ms), AVE0118 slightly slowed conduction from 116±4 to 105±5 cm/s in nonremodeled atria and from 120±5 to 113±6 cm/s in remodeled atria (both P<0.01).
Prevention of AF by Dofetilide and AVE0118
Figure 2 shows the effects of dofetilide and AVE0118 on the inducibility of AF. In nonremodeled atria, single premature stimuli initiated AF in 15% of cases (10 goats, 3 of 20 sites). After 48 hours of AF, the AERP had shortened from 160±6 to 96±5 ms, and the inducibility of AF was 100%. After electrical remodeling, dofetilide prolonged the AERP only by 7±2 ms and did not reduce atrial vulnerability (inducibility, 100%; 6 goats, 12 of 12 sites). In addition, the median duration of AF paroxysms remained the same (3.5 versus 2 seconds; P=0.1). In contrast, AVE0118 prolonged AERP by 65±5 ms and exerted a clear preventive effect on AF inducibility (32%; 10 goats, 6 of 19 sites; P<0.001). The duration of induced AF episodes ranged between 2 and 1260 seconds (median, 6 seconds) before and 1 to 95 seconds (median, 2 seconds) after administration of AVE0118.
Cardioversion of Persistent AF by Dofetilide, Ibutilide, and AVE0118
After 53±19 days of persistent AF, attempts were made to cardiovert AF pharmacologically. Infusion of saline never terminated AF. Administration of dofetilide (20 μg · kg−1 · h−1) or ibutilide (4 mg/h) restored sinus rhythm in 1 of 5 and 2 of 7 goats, respectively. The lowest dosage of AVE0118 (0.1 mg · kg−1 · h−1) restored sinus rhythm in 1 of 6 animals. At higher dosages (1, 3, and 10 mg · kg−1 · h−1), the cardioversion efficacy was increased and AF was terminated in 1, 3, and 5 of 8 goats, respectively.
The effects of dofetilide, ibutilide, and AVE0118 on AFCL are shown in Figure 3. Dofetilide and ibutilide had a moderate effect on AFCL, which was prolonged from 105±5 and 100±4 ms (t=0) to 121±5 and 115±5 ms (t=60; P<0.01). AVE0118 exerted a much stronger effect on AFCL. After 60 minutes of infusion at 0.1, 0.3, 1, 3, and 10 mg · kg−1 · h−1, the AFCL was prolonged by 5±1, 16±2, 30±4, 52±4, and 60±5 ms, respectively (all P<0.01).
In 4 goats with persistent AF, MAPs were recorded from the right atrium during administration of dofetilide (n=1), ibutilide (n=1), and AVE0118 (n=2; 3 mg · kg−1 · h−1). During control, the fibrillation MAPs were characterized by a short duration and an absent plateau. The degree of repolarization showed beat-to-beat variation, and fractionation occasionally occurred. Apart from the prolongation in AF cycle length, dofetilide and ibutilide exerted no clear effects on the duration and shape of the action potentials. AVE0118 caused a significant prolongation of AFCL that was associated with restoration of the plateau phase and an increase in duration of the action potential. This finding demonstrates that the effects of AVE0118 on AFCL are due primarily to prolongation of the atrial action potential.
Effects on Refractory Period and Conduction Velocity During AF
During cardioversion of persistent AF, RPAF and CVAF were measured after 30 minutes of dofetilide infusion (20 μg · kg−1 · h−1; n=4), ibutilide (4 mg/h; n=4), and AVE0118 (1, 3, 10 mg · kg−1 · h−1; n=5). Figure 4 shows a representative example of the effects of ibutilide and AVE0118 on RPAF. During control, the shortest coupling interval that captured AF in 50% of the cases was 84 ms. The 2 tracings show a stimulus with a coupling interval of 84 ms that in 1 case captured and in 1 case did not capture the fibrillating atria. Ibutilide slightly prolonged the RPAF to 95 ms. AVE0118 exerted a much stronger effect, and RPAF was prolonged to 119 and 150 ms at dosages of 0.5 and 5 mg/kg. Figure 5 shows the results of all goats. Shown is the percentage of capture of AF at different coupling intervals before and after drug administration. The S-shaped curves indicate that the RPAF is probabilistic rather than deterministic. Both dofetilide and ibutilide shift the curves slightly to the right, indicating that they prolonged the RPAF only slightly. In contrast, AVE0118 caused a marked rightward shift of the curve because the RPAF was significantly prolonged.
The Table gives results for all goats. Dofetilide and ibutilide prolonged the RPAF from 88±3 and 81±6 ms to 95±6 and 95±8 ms, respectively (dofetilide, P=0.09; ibutilide, P<0.05). AVE0118 increased the RPAF to 144±6 ms at 5 mg/kg (P<0.001). The conduction velocity during AF was measured at the free wall of the right atrium. Figure 6 shows the histograms of the local conduction velocities during control and after AVE0118 (1.5 mg/kg). No effect on atrial conduction was observed. In addition, at higher dosages, AVE0118 caused no changes in CVAF. The atrial wavelength during AF (WLAF=RPAF×CVAF) was increased only slightly by dofetilide or ibutilide. AVE0118 markedly prolonged the wavelength during AF. At dosages of 0.5, 1.5, and 5 mg/kg, the WLAF increased by 28±7%, 53±10%, and 60±7%, respectively (P<0.01). The corresponding plasma levels of AVE0118 were 0.6±0.1, 1.7±0.4, and 4.9±0.8 μg/mL.
Atrial Selectivity of AVE0118
During atrial pacing (400 ms), dofetilide significantly prolonged the QT duration before and after 48 hours of AF (from 252±4 to 266±3 ms and from 239±7 to 259±3 ms, respectively; P<0.05). AVE0118 (1.5 mg/kg) did not cause any changes in QT duration (259±7 versus 261±8 ms and 239±7 versus 246±5 ms; P>0.2). The class III action on the ventricles was also evaluated during persistent AF. Figure 7 shows a representative example of the effects of ibutilide and AVE0118 (1.5 and 5 mg/kg) on QT duration during AF. Because the ventricular rhythm was irregular, QT duration also showed significant beat-to-beat variations. In general, shorter RR intervals were followed by a shorter QT duration and longer RR intervals by a longer QT duration. Infusion of dofetilide prolonged QT duration more during long RR intervals (reverse frequency dependence). At a dosage of 1.5 mg/kg, AVE0118 had no effect on the median QT duration and RR interval. In this example, 5 mg/kg AVE0118 clearly prolonged the median RR interval from 268 to 357 ms. The QT duration prolonged according to its normal rate adaptation. This finding shows that AVE0118 did not cause an independent prolongation of QT duration. In Figure 7 (lower right), effects of the different drugs on the corrected QT duration during AF are given for all goats. Dofetilide and ibutilide prolonged the corrected QT duration by 17±2 and 18±2 ms (P<0.01), whereas AVE0118 did not prolong the corrected QT duration (P>0.25).
In the present study, a novel IKur/Ito blocker (AVE0118) was evaluated in the goat model of AF. It was found that AVE0118 markedly prolonged atrial refractoriness without affecting corrected QT duration. Whereas the class III effect of IKr blockade (dofetilide) was reduced after 48 hours of AF, the action of AVE0118 was enhanced after atrial electrical remodeling. AVE0118 suppressed the inducibility of AF, whereas dofetilide failed to prevent induction of AF paroxysms. During persistent AF, AVE0118 prolonged the atrial wavelength in a dose-dependent way. In 5 of 8 goats, persistent AF was cardioverted by intravenous administration of AVE0118.
Blockade of Ion Channels by AVE0118
The effects of AVE0118 on IKur and Ito have been investigated in CHO cells expressing hKv1.5 (IKur) and hKv4.3 plus KChIP2S (Ito) and in atrial myocytes of the pig.8 In CHO cells, half-maximal inhibition of IKur and Ito occurred in the micromolar range (0.9 to 3.5 μmol/L). Voltage clamp experiments in pig myocytes revealed a 50% reduction in IKur and Ito at comparable concentrations. The IKach current was blocked at an IC50 of 2.6 μmol/L.8 Additionally, in human atrial myocytes, IKur and Ito were blocked by low concentrations of AVE0118.11 Although these data are preliminary, they indicate that AVE0118 is a blocker of IKur, Ito, and IKach.
Effects of Electrical Remodeling on “Early” and “Late” Class III Drugs
It has become clear recently that AF leads to downregulation of several ionic currents, including ICa,L, IKur, and Ito.12–14 Ionic remodeling has important implications for the action of antifibrillatory drugs. It was recently demonstrated that the action of the IKr blocker dl-sotalol was reduced in patients with long-lasting AF.15 In the present study, a reduced action of dofetilide was found. This suggests that in electrically remodeled atria, the late activated IKr current contributes less to atrial repolarization. In a mathematical model of the human atrial action potential, the effect of IKr blockade on the duration of the action potential was reduced in remodeled atrial cells.16 During a short action potential, less IKr was activated; therefore, in this setting, the net prolongation by IKr blockade was minimal. The opposite was true for IKur blockade, which exerted an enhanced effect in remodeled atria. This finding was rather surprising because this current is downregulated in remodeled atria.13,14 A possible explanation for this paradoxical effect is that the ICa,L is reduced to a greater extent than IKur, resulting in a larger contribution of IKur to atrial repolarization. Wirth et al17 showed that AVE0118 prolonged atrial refractoriness more at sites with shorter refractory periods. Nagasawa et al7 reported that the IKur blocker NIP-142 became more effective when the action potential was shortened by vagal stimulation.
Antifibrillatory Action and Atrial Selectivity of AVE0118
During slow pacing, we measured a prolongation of the atrial refractory period of 70% in remodeled atria. Also during AF, the refractory period was prolonged, whereas no changes in conduction velocity were observed. As a result, the wavelength of the fibrillating waves increased from 7 to 11 cm. This prolongation in atrial refractory period and wavelength can adequately explain the antifibrillatory effects of AVE0118. The increase in size of the intra-atrial reentrant pathways and a consequent reduction in number of fibrillation waves increase the likelihood that the fibrillation waves will die out simultaneously and that fibrillation terminates.
Most class III drugs used for the treatment of AF are IKr inhibitors. Unfortunately, blockade of the IKr current also prolongs the ventricular action potentials and QT duration. This increases the chance of torsade de pointes arrhythmias.2 Because IKur is absent in the ventricles, IKur blockers may have a selective action on the atria.6 As shown in the dog by Nattel et al,18 4-aminopyridine (a blocker of IKur and Ito) increased atrial refractoriness without affecting the ventricular refractory period. In the goat, AVE0118 also did not prolong ventricular repolarization while exerting a strong class III effect on the atria. Thus, compared with the IKr blocker dofetilide, AVE0118 was highly atrial selective. In pigs, no prolongation of ventricular refractoriness and QT duration was found after administration of AVE0118.17
Clinical Implications and Limitations
Despite the development of various ablation techniques for the treatment of AF, the development of a safe and effective drug for prevention and termination of AF would have major impact on the management of AF. Our present study suggests that blockers of early K+ currents (IKur and Ito) may have a high clinical potential. The strong class III effect in remodeled atria, together with the absence of an effect on QT duration, offers hope that these drugs can be used effectively without the risk of ventricular arrhythmias. However, caution is warranted in extrapolating our results to humans. One important difference is that the atrial action potential is shorter and the degree of electrical remodeling is higher in the goat than in humans.19 Therefore, it is hard to predict whether the beneficial effects of IKur/Ito blockade, as observed in the goat, will also occur in humans. Another important caveat is that the goat model lacks the pathological changes seen in atria of patients with AF. Hypertension, valvular and coronary artery disease, and heart failure all cause structural changes in the atria such as interstitial fibrosis, apoptosis, and dilatation. Therefore, it remains to be seen whether early class III drugs like AVE0118 will be safe and effective in humans.
This study was supported by grants 902-16-097 and 920-033-122 from the Netherlands Organization for Scientific Research and by a research grant from Aventis Pharma Germany, GmbH.
Dr Gögelein is an employee of and Dr Allessie is a consultant for Aventis Pharma Germany, GmbH.
Wang Z, Fermini B, Nattel S. Sustained depolarization-induced outward current in human atrial myocytes: evidence for a novel delayed rectifier K+ current similar to Kv1.5 cloned channel currents. Circ Res. 1993; 73: 1061–1076.
Li GR, Feng J, Yue L, et al. Evidence for two components of delayed rectifier K+ current in human ventricular myocytes. Circ Res. 1996; 78: 689–696.
Goegelein H. Effects of the novel inhibitor of atrial K+ channels AVE0118 in anesthetized pigs and on ion channels. Pacing Clin Electrophysiol. 2003; 26: 1045. Abstract.
Duytschaever M, Mast F, Killian M, et al. Methods for determining the refractory period and excitable gap during persistent atrial fibrillation in the goat. Circulation. 2001; 104: 957–962.
Wijffels MC, Kirchhof CJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation: a study in awake chronically instrumented goats. Circulation. 1995; 92: 1954–1968.
Bosch RF, Lazlo R, Schneck AC, et al. AVE 0118, an antiarrhythmic drug with novel mechanism of action-block of Ikur and Ito potassium currents in human atrial myocytes. Circulation. 2002; 106 (suppl II): II-22. Abstract.
Yue L, Feng J, Gaspo R, et al. Ionic remodeling underlying action potential changes in a canine model of atrial fibrillation. Circ Res. 1997; 81: 512–525.
Van Wagoner DR, Pond AL, McCarthy PM, et al. Outward K+ current densities and Kv1.5 expression are reduced in chronic human atrial fibrillation. Circ Res. 1997; 80: 772–781.
Bosch RF, Nattel S. Cellular electrophysiology of atrial fibrillation. Cardiovasc Res. 2002; 54: 259–269.
Courtemanche M, Ramirez RJ, Nattel S. Ionic targets for drug therapy and atrial fibrillation-induced electrical remodeling: insights from a mathematical model. Cardiovasc Res. 1999; 42: 477–489.
Wirth KJ, Paehler T, Rosenstein B, et al. Atrial effects of the novel K(+)-channel-blocker AVE0118 in anesthetized pigs. Cardiovasc Res. 2003; 60: 298–306.
Nattel S, Matthews C, De Blasio E, et al. Dose-dependence of 4-aminopyridine plasma concentrations and electrophysiological effects in dogs: potential relevance to ionic mechanisms in vivo. Circulation. 2000; 101: 1179–1184.