Introduction

Top Abstract Introduction Methods Results Discussion References

The blockade of cardiac K⁺ currents to delay repolarization and prolong action potential duration and refractoriness, i.e., Class III electrophysiologic activity, continues to be explored as an antiarrhythmic strategy (Nair and Grant, 1997). The initial clinical assessment of the antiarrhythmic potential of Class III agents has focused on the prevention of malignant ventricular arrhythmias and sudden death (Mason, 1993; Echt et al., 1995; Waldo et al., 1996). However, there is also significant interest in the use of Class III agents in the treatment of supraventricular arrhythmia (Sung et al., 1995; Sedgwick et al., 1995b; Stambler et al., 1996).

The treatment of atrial arrhythmias by currently available Class I Na⁺ channel blocking antiarrhythmic agents is limited by ventricular proarrhythmic effects (Flaker et al., 1992). One potential strategy to improve the safety and efficacy of pharmacologic therapy for atrial arrhythmia is the identification of interventions with atrial-selective activity. Significant differences in action potential configuration and ionic currents in atrial versus ventricular tissue have been reported in animal species and in humans (Hume and Uehara, 1985; Giles and Imaizumi, 1988; Koumi et al., 1995; Amos et al., 1996), suggesting the potential for differential effects on these two tissues with appropriate ion channel blockade. In the present investigation, the relative atrial versus ventricular effects of a variety of reported Class III K⁺ channel blocking agents, which differ by selectivity of cardiac K⁺ channels modulated, were characterized through the determination of effects on refractoriness in vitro. The objective of this study was to determine whether different profiles of atrial versus ventricular activities would be displayed, and in particular whether atrial-selective activity could be achieved, by agents with differing K⁺ channel blocking profiles.

	Methods

Top Abstract Introduction Methods Results Discussion References

Ferret isolated paired right and left atrial preparation. Male ferrets (1.0-1.2 kg) were anesthetized with a mixture of ketamine + xylazine (90 mg ketamine + 2 mg xylazine i.m./ferret). Right and left atrial pairs were excised intact and mounted in 50-ml tissue baths containing Krebs-Henseleit solution (pH = 7.2-7.4) of the following composition (mM): NaCl, 118.0; KCl, 4.0; CaCl₂·2H₂O, 2.0; MgSO₄·7H₂O, 1.2; NaH₂PO₄·H₂O, 1.2; dextrose, 11.1; NaHCO₂, 25.0. Timolol (100 nM) was added to the solution to prevent catecholamine release during electrical stimulation. The bathing solution was aerated with 95%O₂/5%CO₂ and maintained at 32°C. The right atrium of each paired preparation was secured on glass tissue holders in direct contact with platinum pacing electrodes, and the left atrium of each paired preparation was attached with 4-0 suture to an isometric force transducer coupled to a physiologic recorder for monitoring isometric tension. A resting tension of 1.0 g was applied, and the tissues were equilibrated for 120 min with the bathing solution replaced every 20 min. The paired atrial preparations were allowed to equilibrate at their own spontaneous right atrial rate during the equilibration period to promote viability and stability of the preparation, with external pacing applied only for the determination of atrial ERP. Spontaneous rate in the isolated paired atrial preparations at the end of the 2-hr equilibration period was approximately 90 to 100 bpm; spontaneous heart rate in conscious ferrets in our laboratory was estimated to be 150 to 200 bpm. During experimental protocols (described below), atrial ERP was determined at the end of 2-min periods of atrial pacing at 30 bpm above spontaneous atrial rate, approximating a paced rate of 2 Hz, or at 4 Hz as designated by experimental protocol. During periods of pacing for ERP determination, atrial pairs were paced with square wave pulses of 4-msec pulse duration at 2.0 times the excitation threshold. ERP was determined by standard extrastimulus technique in which the extrastimulus was delivered at 2.0 times the excitation threshold at varying coupling intervals relative to basic pacing. The shortest coupling interval resulting in a propagated response was defined as the ERP. Between these brief periods of pacing for ERP determination, the paired atrial preparations were allowed to remain at spontaneous right atrial rate, again to promote viability of the preparation. The paired right and left atrial preparation was used in the present studies because it was found to be a more reliable, durable preparation in pilot studies, i.e., providing more stable and consistent ERP values across the time course and range of pacing frequencies used than either isolated paced right or left atria alone. Additionally, the present studies with the isolated paired atrial preparation were performed at 32°C because pilot studies demonstrated that this preparation degraded, i.e., ERP determinations became inconsistent over time, with pacing at the higher 37°C temperature used in previous studies with papillary muscle preparations.

Ferret isolated right ventricular papillary muscle preparation. The ferret right ventricular papillary muscle preparation was described in detail previously (Baskin et al., 1991). Male ferrets (1.0-1.2 kg) were anesthetized as described above, and right ventricular papillary muscles were excised and mounted in tissue baths containing Krebs-Henseleit solution of pH and composition identical with that described above. Timolol (100 nM) was added to the solution to prevent catecholamine release during electrical stimulation. The bathing solution was aerated with 95%O₂/5%CO₂ and maintained at 32°C or, in specific experimental protocols assessing the effect of temperature on ion channel blocker activity (described below), at 37°C. Each papillary muscle was secured on glass tissue holders in direct contact with platinum pacing electrodes, and the tendinous end of each muscle was attached with 4-0 suture to an isometric force transducer coupled to a physiologic recorder for monitoring isometric tension. A resting tension of 0.5 g was applied, and the tissues were equilibrated for 120 min at a pacing rate of 2 Hz with the bathing solution replaced every 20 min. Papillary muscles were paced with square wave pulses of 4-msec pulse duration at 2.0 times the excitation threshold. During experimental protocols (described below), ventricular ERP was determined by standard extrastimulus technique in which the extrastimulus was delivered at 2.0 times the excitation threshold at varying coupling intervals relative to basic pacing at rates of 1 to 4 Hz, as designated by experimental protocol. The shortest coupling interval resulting in a propagated response was defined as the ERP.

Concentration-ERP response relationship in atrial versus ventricular tissue. The initial characterization of the atrial versus ventricular effects of K⁺ channel blockers was the generation of concentration-ERP response relationships in paired atrial versus papillary muscle preparations. For concentration-response relationships, ERPs were determined in paired atrial preparations paced at 30 bpm above spontaneous right atrial rate, thereby approximating a paced rate of 2 Hz, and in papillary muscles paced at a rate of 2 Hz. Concentration-response studies for both tissues were conducted at 32°C. ERPs were determined at baseline (i.e., after 120 min of equilibration) and after cumulative addition of test agents to the tissue bath with each concentration of test agent incubated for 30 min before ERP determination. Concentration ranges of test agents were as follows: dofetilide, 10 nM to 10 µM; E-4031, 10 nM to 10 µM; d-sotalol, 1 µM to 1 mM; ibutilide, 10 nM to 10 µM; RP58866, 10 nM to 30 µM; azimilide, 100 nM to 100 µM; tedisamil, 100 nM to 100 µM; and indapamide, 1 µM to 1 mM. In this and succeeding studies, stock solutions of test agents (usually 10 mM) were made with distilled water or dimethyl sulfoxide. Serial dilutions to achieve appropriate test agent concentrations for addition to the tissue baths were made with distilled water.

Frequency dependence of atrial versus ventricular ERP response. In separate studies, the effects of varying pacing frequency on K⁺ channel blocker activities were compared in atrial versus ventricular tissue. ERP responses to one individualized concentration of each test agent, determined based on results in the preceding concentration-ERP response relationship study, were measured in paired atrial preparations paced at 30 bpm above spontaneous right atrial rate, thereby approximating a paced rate of 2 Hz, and during a 2-min period of pacing at a rate of 4 Hz. Likewise, ERP responses to the same concentration of test agent were determined in papillary muscles paced at 2 Hz and during a 2-min period of pacing at 4 Hz. In both tissues, the frequency dependence of ERP responses was assessed at 32°C, and ERPs were determined after a 30-min incubation of the test concentration of each agent in the tissue bath. The individualized concentration of each test agent chosen for this frequency-dependence study in both paired atrial and papillary muscle preparations was defined as that concentration of test agent eliciting an approximately 60% increase in atrial ERP in the preceding concentration-response study. In cases where a 60% increase in atrial ERP was not achievable (e.g., as with azimilide and indapamide), near-maximal testable concentrations were chosen for study. Test concentrations in this study were: dofetilide, 100 nM; E-4031, 300 nM; d-sotalol, 300 µM; ibutilide, 300 nM; RP58866, 1 µM; azimilide, 30 µM; tedisamil, 3 µM; and indapamide, 1 mM.

Temperature dependence of ventricular ERP response. Comparisons of K⁺ channel blocker activities in atrial versus ventricular tissue described in the present studies above were conducted in both tissues at 32°C to standardize temperature conditions. As noted previously, pilot studies demonstrated that higher temperatures compromised the stability of the atrial preparation. However, previous characterizations of the ventricular ERP effects of K⁺ channel blockers have been conducted in the ferret right ventricular papillary muscle preparation at a temperature of 37°C. To assess the potential influence of temperature on the activities of representative K⁺ channel blockers, studies assessing effects on ERP in the papillary muscle preparation paced at frequencies of 1 to 4 Hz were conducted at 32°C vs. 37°C. In separate treatment groups equilibrated and maintained at 32°C or 37°C, ERP responses were determined during 5-min periods of pacing at 1, 2, 3 and 4 Hz before and after a 30-min incubation of the test concentration of each agent in the tissue bath. During the 30-min incubation period, papillary muscles were paced at 2 Hz. Test concentrations in this study were identical with those used in the preceding assessment of the frequency dependence of atrial versus ventricular ERP response: dofetilide, 100 nM; azimilide, 30 µM; tedisamil, 3 µM; and indapamide, 1 mM.

Statistical analysis. Data are expressed as mean ± S.E.M. Changes in ERP in atrial and ventricular preparations are expressed as percentage changes from pretreatment baseline to normalize for differences in baseline ERPs between preparations. Comparisons of absolute baseline ERPs between tissues or at two different temperatures were conducted by a two-tailed unpaired Student's t test. Comparisons of absolute baseline ERPs within tissue at two different pacing frequencies were conducted by a two-tailed paired Student's t test.

	Results

Top Abstract Introduction Methods Results Discussion References

Concentration-ERP response relationship in atrial versus ventricular tissue. Concentration-ERP response relationships for the various K⁺ channel blocking agents were generated in paired atrial versus right ventricular papillary muscle preparations at 32°C at a pacing rate of 2 Hz. Under these conditions, baseline ERPs were significantly lower in the atrial versus ventricular preparations: total cohort base-line atrial ERP = 111 ± 2 msec versus ventricular ERP = 165 ± 2 msec (P < .01; n = 64-74). Concentration-ERP response relationships generated in atrial versus papillary muscle preparations are grouped by general profile and depicted in figures 1 to 3. Concentration-ERP response relationships for dofetilide, E-4031, ibutilide and d-sotalol (fig. 1) were characterized by markedly greater efficacies in increasing atrial ERP (approximately 90-110% increases for all four agents) versus ventricular ERP (approximately 10-20% increases for all four agents). For both RP58866 and tedisamil (fig. 2), concentration-ERP response relationships were characterized by similar efficacies for increasing atrial and ventricular ERPs (approximately 60-80% increases for both tissues for RP58866; approximately 150-160% increases for both tissues for tedisamil), but with approximately 10-fold greater potencies for increasing atrial versus ventricular ERP for both agents. Concentration-ERP response relationships for both azimilide and indapamide (fig. 3) were characterized by both similar potencies and efficacies for increasing atrial versus ventricular ERPs (approximately 40-60% increases for both tissues for both agents).

View larger version (32K): [in this window] [in a new window]   Fig. 1.   Concentration-ERP (% of baseline) response relationships of dofetilide, E-4031, d-sotalol and ibutilide in ferret paired atrial and right ventricular papillary muscle preparations at 32°C at a pacing frequency of 2 Hz. Data are mean ± S.E.M. with n = 7-8.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Concentration-ERP (% of baseline) response relationships of RP58866 and tedisamil in ferret paired atrial and right ventricular papillary muscle preparations at 32°C at a pacing frequency of 2 Hz. Data are mean ± S.E.M. with n = 7-12.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   Concentration-ERP (% of baseline) response relationships of azimilide and indapamide in ferret paired atrial and right ventricular papillary muscle preparations at 32°C at a pacing frequency of 2 Hz. Data are mean ± S.E.M. with n = 8-16.

Frequency dependence of atrial versus ventricular ERP response. Frequency-ERP responses for select concentrations of the K⁺ channel blocker test agents (see "Methods" for test concentration criterion) were assessed at pacing frequencies of 2 vs. 4 Hz in paired atrial versus right ventricular papillary muscle preparations at 32°C. At both the 2 and 4 Hz pacing frequencies, baseline ERPs were significantly lower in the atrial versus ventricular preparations: 2 Hz total cohort baseline atrial ERP = 111 ± 2 msec versus ventricular ERP = 159 ± 2 msec (P < .01); 4 Hz total cohort baseline atrial ERP = 101 ± 2 msec versus ventricular ERP = 128 ± 2 msec (P < .01; n = 52-56). Within each tissue, increasing pacing frequency from 2 to 4 Hz resulted in a significant decrease in baseline ERP (P < .01 for both tissues, 2 vs. 4 Hz), i.e., inherent reverse frequency dependence. However, the magnitude of decrease in baseline ERP with increasing pacing frequency was much greater in the ventricular muscle preparation (159 ± 2 msec at 2 Hz decreasing to 128 ± 2 msec at 4 Hz) than in the atrial preparation (111 ± 2 msec at 2 Hz decreasing to 101 ± 2 msec at 4 Hz). Frequency (2 vs. 4 Hz)-ERP responses in atrial versus papillary muscle preparations at 32°C for the K⁺ channel blockers are depicted in figures 4 and 5. For all compounds tested, magnitudes of effect on ERP in atrial versus ventricular tissue at a pacing rate of 2 Hz at the select concentrations were consistent with results obtained in the 2-Hz concentration-ERP response studies described above. Dofetilide, E-4031, ibutilide and d-sotalol (fig. 4, A-D) at the concentrations tested all displayed reverse-frequency-dependent effects on atrial refractoriness, i.e., greater increases in atrial ERP observed at the lower 2-Hz vs. higher 4-Hz pacing rate. In contrast, under the conditions tested, there was no apparent frequency dependence of effect on ventricular ERP for these four agents (fig. 4, A-D). RP58866 (fig. 5A) displayed a frequency-dependence profile similar to those of the preceding four compounds, i.e., marked reverse-frequency-dependent effects on atrial ERP, but a slight forward frequency-dependent effect on ERP in ventricular tissue. Tedisamil (fig. 5B) displayed a profile dissimilar to the preceding five agents, with a forward frequency-dependent effect on ventricular ERP, i.e., lesser increases in ventricular ERP observed at the lower 2-Hz vs. higher 4-Hz pacing rate, but a lack of frequency dependence of effect on atrial ERP. Azimilide (fig. 5C) under these conditions displayed modest forward frequency-dependent effects on ERP in atrial and ventricular tissue, and indapamide (fig. 5D) displayed a forward frequency-dependent effect on ventricular ERP, but frequency-independent increases in atrial ERP.

View larger version (32K): [in this window] [in a new window]   Fig. 4.   Effect on ERP (% of baseline) of dofetilide (100 nM; A), E-4031 (300 nM; B), d-sotalol (300 µM; 4C) and ibutilide (300 nM; D) in ferret paired atrial and right ventricular papillary muscle preparations at 32°C at pacing frequencies of 2 vs. 4 Hz. Data are mean ± S.E.M. with n = 5-11. See "Methods" for test concentration criteria.

View larger version (31K): [in this window] [in a new window]   Fig. 5.   Effect on ERP (% of baseline) of RP58866 (1 µM; A), tedisamil (3 µM; B), azimilide (30 µM; C) and indapamide 1 mM; D) in ferret paired atrial and right ventricular papillary muscle preparations at 32°C at pacing frequencies of 2 vs. 4 Hz. Data are mean ± S.E.M. with n = 5-8. See "Methods" for test concentration criteria.

Temperature dependence of ventricular ERP response. Previous characterizations of the ventricular activities of K⁺ channel blockers, including frequency dependence of ERP response, have been conducted in the ferret right ventricular papillary muscle preparation at 37°C. The present studies comparing K⁺ channel blocker activities in atrial versus ventricular tissue described above were conducted at 32°C to standardize temperature conditions for both tissues. To assess the potential influence of temperature on K⁺ channel blocker activities, studies assessing effects on ERP in the papillary muscle preparation paced at frequencies of 1 to 4 Hz were conducted at 32°C vs. 37°C. Baseline ERPs at both temperatures decreased progressively with increasing pacing frequency, i.e., inherent reverse frequency-dependence: 32°C total cohort baseline ventricular ERP, 1 Hz = 199 ± 3 msec, 2 Hz = 149 ± 3 msec, 3 Hz = 133 ± 3 msec, 4 Hz = 125 ± 2 msec; 37°C total cohort baseline ventricular ERP, 1 Hz = 140 ± 2 msec, 2 Hz = 117 ± 2 msec, 3 Hz = 97 ± 2 msec, 4 Hz = 88 ± 3 msec (n = 16). At each individual pacing frequency in the 1 to 4 Hz range, baseline ventricular ERP always was significantly lower at the 37°C vs. 32°C temperature (P < .01 for all frequencies). Frequency (1-4 Hz)-ERP responses in right ventricular papillary muscle preparations at 32°C vs. 37°C for select concentrations of K⁺ channel blockers (see "Methods" for test concentration criteria) are summarized in figures 6 and 7. For dofetilide (fig. 6A), magnitudes of increase in ERP tended to be greater, particularly at the 1- and 2-Hz pacing rates, at the higher 37°C vs. lower 32°C temperature. Reverse frequency-dependent increases in ventricular refractoriness (i.e., greater increases in ventricular ERP at lower versus higher pacing rates) were clearly apparent with dofetilide at the higher 37°C temperature. However, at the lower 32°C temperature, with the exception of a decrement in ERP effect from 1 Hz to the range of 2 to 4 Hz, there was a virtual lack of frequency dependence of effect of dofetilide. For tedisamil (fig. 6B), azimilide (fig. 7A) and indapamide (fig. 7B), magnitudes of increase in ERP were markedly greater, particularly in the 2- to 4-Hz pacing rate range, at the higher 37°C vs. lower 32°C temperature. For tedisamil (fig. 6B), there was no clear frequency dependence of effect on ventricular refractoriness at 37°C, whereas at 32°C there tended to be a modest forward frequency-dependent effect on refractoriness. Azimilide (fig. 7A) and indapamide (fig. 7B) both displayed clear forward frequency-dependent increases in ventricular refractoriness at 37°C, with much more modest and more variable trends for forward frequency-dependent increases in ventricular refractoriness at the lower 32°C.

View larger version (30K): [in this window] [in a new window]   Fig. 6.   Effect on ERP (% of baseline) of dofetilide (100 nM; A) and tedisamil (3 µM; B) in ferret right ventricular papillary muscle preparations at 32°C vs. 37°C at pacing frequencies of 1 to 4 Hz. Data are mean ± S.E.M. with n = 4. See "Methods" for test concentration criteria.

View larger version (33K): [in this window] [in a new window]   Fig. 7.   Effect on ERP (% of baseline) of azimilide (30 µM; A) and indapamide (1 mM; B) in ferret right ventricular papillary muscle preparations at 32°C vs. 37°C at pacing frequencies of 1 to 4 Hz. Data are mean ± S.E.M. with n = 4. See "Methods" for test concentration criteria.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Significant differences in action potential configuration in atrium versus ventricle have been reported in guinea pig and rabbit, and have been attributed primarily to differences in the magnitudes and kinetics of repolarizing K⁺ currents (Hume and Uehara, 1985; Giles and Imaizumi, 1988). Action potential configuration also differs significantly in human atrium versus ventricle (Trautwein et al., 1962). Recent studies have demonstrated the presence of inward rectifier I_K1, transient outward I_to, and both rapid I_Kr and slow I_Ks components of delayed rectifier currents in both human atrium and ventricle (Escande et al., 1987; Thuringer et al., 1992; Beuckelmann et al., 1993; Nabauer et al., 1993; Wang et al., 1994; Li et al., 1996). Two directed comparisons of K⁺ current subtypes in human atrium versus ventricle have reported differences in I_K1 kinetics and density (Koumi et al., 1995) and I_to kinetics and sensitivity to 4-aminopyridine (Amos et al., 1996) in the two tissues. Additionally, a sustained outward K⁺ current, I_sus or I_so, has been recorded in human atrium but not ventricle (Wang et al., 1993; Amos et al., 1996). The demonstration of significant differences in action potential morphology, K⁺ current subtype density, kinetics and sensitivity to blockers in mammalian atrium versus ventricle suggests the potential to achieve tissue-selective pharmacologic modulation via appropriate ion channel blockade.

Accordingly, the purpose of the present studies was to characterize the relative atrial versus ventricular effects of K⁺ channel blockers varying in selectivities of K⁺ current subtypes modulated. Test agents used in the present studies included the selective I_Kr blockers dofetilide, E-4031 and d-sotalol (Sanguinetti and Jurkiewicz, 1990; Carmeliet 1992), as well as ibutilide, reported to be an activator of a slow inward Na⁺ current but also demonstrated to be a potent blocker of I_Kr (Lee, 1992; Yang et al., 1995). Also included in the present studies were RP58866, a blocker of I_K1 and I_Kr (Escande et al., 1992; Jurkiewicz et al., 1996), and tedisamil, a blocker of I_to and I_Kr as well as sodium (I_Na) and calcium (I_Ca) currents at high concentrations (Dukes and Morad 1989; Dukes et al., 1990; Ohler et al., 1994). Finally, azimilide, a blocker of I_Ks and I_Kr as well as I_Ca (Busch et al., 1994; Fermini et al., 1995), and indapamide, a blocker of I_Ks (Turgeon et al., 1994), also were assessed in these studies. K⁺ channel blockade was quantitated by measurement of ERP, a functional measure of prolongation of action potential duration (i.e., Class III activity), in isolated ferret atrial and right ventricular papillary muscle preparations. Previous studies demonstrated that the ferret right ventricular papillary muscle preparation was sensitive to a wide variety of K⁺ channel blockers including dofetilide, E-4031, d-sotalol, tedisamil and azimilide (Baskin et al., 1991; Wallace et al., 1995; Fermini et al., 1995).

In all protocols in the present studies, baseline ERPs were lower in ferret atrial tissue than ventricular tissue. This observation is consistent with previous reports of shorter action potential duration in guinea pig, rabbit and human atrial vs. ventricular preparations (Trautwein et al., 1962; Hume and Uehara, 1985; Giles and Imaizumi, 1988), suggesting significant differences in the composition of repolarizing currents in the two tissues. Further, different profiles of frequency dependence for both baseline ERP as well as for effects of test agents on ERP were observed in ferret atrial versus ventricular tissues, again suggesting significant differences in repolarizing currents in the two tissues.

The principal result of the present investigation was the demonstration of markedly different profiles for modulation of atrial versus ventricular refractoriness with K⁺ channel blockers of varying subtype selectivity. Selective I_Kr blockade, represented by dofetilide, E-4031 and d-sotalol, displayed marked but not absolute atrial-selective efficacy for increasing refractoriness. Ibutilide, also a potent blocker of I_Kr, displayed atrial selectivity similar to the I_Kr blockers. Azimilide and indapamide, both of which block I_Ks as part of their spectrums of action, displayed essentially "balanced" activities, i.e., both agents elicited increases in atrial and ventricular refractoriness with equivalent potencies and efficacies. RP58866, a blocker of I_K1 and I_Kr, and tedisamil, primarily a blocker of I_to and I_Kr, displayed similar profiles for increasing atrial refractoriness with greater potencies, but with similar efficacies, compared to ventricular refractoriness. Hence, differences in the spectrum of K⁺ channel subtypes modulated resulted in differential effects on atrial versus ventricular refractoriness.

Differences in K⁺ channel selectivity also resulted in different frequency-dependence profiles in atrium versus ventricle. In the atrial preparation at 32°C, dofetilide, E-4031, d-sotalol, ibutilide and RP58866, all of which block I_Kr either selectively or as part of their spectrums of action, displayed reverse frequency-dependent activity. Also in atrium, tedisamil and indapamide displayed frequency-independent activity whereas azimilide displayed modest forward frequency-dependent activity at the concentrations tested. In the ventricular preparation at 32°C, dofetilide, E-4031, d-sotalol and ibutilide displayed essentially frequency-independent effects on ERP, RP58866 and azimilide displayed modest forward frequency-dependent effects on ERP and tedisamil and indapamide displayed clear forward frequency-dependent effects on ERP.

There was no absolute correspondence between profiles for efficacy and potency in concentration-ERP response studies at 2 Hz at 32°C and profiles for frequency dependence at 2 vs. 4 Hz at 32°C. That is, particularly for those test agents with mixed K⁺ channel blocking activities, a given profile for atrial versus ventricular selectivity in concentration-response studies was not predictive for a given frequency-dependence profile. On one hand, dofetilide, E-4031, d-sotalol and ibutilide did display similar profiles for both atrial versus ventricular selectivities and frequency dependence. The similarities in pharmacodynamic profiles among the selective I_Kr blockers and ibutilide in the various protocols in these studies suggests that I_Kr blockade is the predominant mechanism of action of ibutilide in this experimental model system. Conversely, for the other agents with mixed ionic mechanisms of action, there was no predictive relationship between profiles for atrial selectivity in concentration-response studies and frequency dependence. For example, whereas azimilide and indapamide displayed similar atrial versus ventricular selectivity profiles in concentration-response studies, they displayed dissimilar frequency-dependence profiles. The latter observation suggests that differences in K⁺ channel subtype selectivity may differentially and uniquely modulate atrial versus ventricular selectivity and frequency dependence.

Temperature also appears to be an important factor in the assessment of the cardiac electrophysiologic effects of K⁺ channel blockers. The temperature sensitivity of cardiac repolarizing current has been reported previously, with the magnitude of outward K⁺ current reduced at lower temperatures because of reduced activities of regulatory enzymes (Walsh and Kass, 1988). Consistent with this observation, baseline ERPs in the ferret ventricular papillary muscle preparation in a previous (Baskin et al., 1991) as well as in the present study were greater at the lower 32°C vs. higher 37°C temperature tested, indicative of reduced repolarizing K⁺ current at the lower temperature. Consequently, increases in ERP observed with the K⁺ channel blockers used in the present studies generally were greater at the higher 37°C, presumably because of the presence of more blockable K⁺ repolarizing current at higher temperatures. Changes in temperature resulted in alterations not only in absolute magnitude but also in the frequency dependence of change in ERP, which suggests differential effects of temperature on different K⁺ channel subtypes in the papillary muscle preparation. For example, consistent with previous studies with this preparation at higher temperature (Baskin et al., 1991), selective I_Kr blockers displayed a characteristic reverse frequency-dependent activity profile across a wide range of pacing frequency in the ferret right ventricular papillary muscle preparation at the higher 37°C temperature, but not at 32°C. Instability of the ferret atrial preparation at temperatures greater than 32°C precluded a similar assessment in this tissue.

Many caveats must be considered in extrapolating the findings of the present preclinical pharmacologic assessment in isolated tissues to the clinical use of K⁺ channel blockers as antiarrhythmic agents. Most important is the issue of species differences. It is improbable that any animal species will appropriately mimic humans with regard to exact type, mix and densities of K⁺ current subtypes in atrium and ventricle. In this regard, although no clinical study has systematically compared atrial versus ventricular selectivity with a variety of K⁺ channel blockers, clinical studies with racemic sotalol (Echt et al., 1982) and more recently with dofetilide (Sedgwick et al., 1995a) have reported relatively greater increases in atrial versus ventricular ERP, although not to the degree of atrial selectivity displayed by the I_Kr blockers in the ferret tissues in the present study. Clinical electrocardiographic studies also have reported that dofetilide, E-4031 and d-sotalol display reverse frequency-dependent effects on QT interval (Okada et al., 1996), consistent with the activities of these I_Kr blockers on ferret papillary muscle ERP at 37°C. As noted above, temperature is an important factor in the characterization of effects on cardiac refractoriness, with K⁺ channel subtypes potentially modulated to different degrees by changes in temperature. Finally, because of practical considerations, assessments of frequency dependence in atrial versus ventricular tissue and temperature dependence in ventricular tissue were performed with one individualized concentration of each test agent. For K⁺ channel blockers with mixed activities, atrial versus ventricular selectivity and frequency-dependence profiles may well be concentration-dependent. For example, a previous assessment of the mixed I_Kr and I_Ks blocker azimilide, with much lower 0.3 to 3.0 µM test concentrations at which I_Kr blocking activity predominates, reported a reverse frequency-dependent profile in the ferret right ventricular papillary muscle preparation at 37°C (Fermini et al., 1995). In the present studies, with use of a higher 30.0 µM test concentration at which multiple channel subtypes may be modulated, forward frequency-dependent profiles were observed with this agent in the papillary muscle preparation. With these caveats in mind, however, the present studies in ferret atrial and ventricular tissue do provide direct experimental evidence supporting the potential to differentially modulate atrial versus ventricular refractoriness and frequency dependence with ion channel blockers differing in K⁺ channel subtype selectivity. Ultimately, the identification and targeting of an appropriate K⁺ channel subtype or mix of subtypes may result in the achievement of optimal atrial-selective activity for the treatment of supraventricular arrhythmias.