Pinacidil-Induced Electrical Heterogeneity and Extrasystolic Activity in

1177 Pinacidil-Induced Electrical Heterogeneity and Extrasystolic Activity in Canine Ventricular Tissues Does Activation of ATP-Regulated Potassium Current Promote Phase 2 Reentry? Jose M. Di Diego, MD, and Charles Antzelevitch, PhD Background. Pinacidil is known to augment a time-independent outward current in cardiac tissues by activating the ATP-regulated potassium channels. Activation of this current, IK.ATP, is thought to be responsible for increased potassium permeability in ischemia. The contribution of IK.ATP activation to arrhythmogenesis and the role of activation of this current in suppression of arrhythmias are areas of great interest and debate. Because electrical depression attending myocardial ischemia is more accentuated in ventricular epicardium than in endocardium, we endeavored to contrast the effects of pinacidilinduced IK.ATP activation on the electrophysiology of canine ventricular epicardium and endocardium. Methods and Results. Standard microelectrode techniques were used. Pinacidil (1 to 5 umol/l) produced a marked dispersion of repolarization and refractoriness in isolated canine ventricular epicardium as well as between epicardium and endocardium. In endocardium, pinacidil abbreviated action potential duration (APD,) and refractoriness by 8.±2.3%. In epicardium, the effects of pinacidil were nonhomogeneous. At some sites, pinacidil induced an all-or-none repolarization at the end of phase 1 of the action potential, resulting in 55.5±8.7% abbreviation of APD, and refractoriness. Adjacent to these were sites at which the dome was maintained with only minor changes in APD and refractoriness. Extrasystolic activity displaying features of reentry was observed in isolated sheets of epicardium (63.2%) after exposure to pinacidil (1 to 5,umol/L) but never in its absence. Dispersion of repolarization and ectopic activity was most readily induced in epicardium by a slowing of the stimulation rate in the presence of pinacidil. Electrical homogeneity was restored and arrhythmias abolished after washout of pinacidil or addition of either a transient outward current blocker, 4-aminopyridine, or a blocker of the ATP-regulated potassium channels, glybenclamide. Conclusions. Our data suggest that the activation of 'K.ATP can produce a marked dispersion of repolarization and refractoriness in epicardium as well as between epicardium and endocardium, leading to the development of extrasystolic activity via a mechanism that we have called phase 2 reentry. The available data also suggest that blockade of the transient outward current and/or the ATP-regulated potassium channels may be useful antiarrhythmic interventions under ischemic or "ATP depleted" conditions. (Circulation. 1993;88:1177-1189.) Iay WORDs * myocardium * endocardium * electrophysiology * pharmacology * arrhythmias * reentry * potassium * pinacidil * epicardium P inacidil is known to augment ATP-regulated potassium current (IK-ATp) in cardiac tissues.'-3 Recent studies have suggested that activation of ATP-regulated potassium channels (KATp) by pinacidil and several other related agents may represent a novel pharmacological approach in the treatment of arrhythmias caused by repolarization abnormalities4-7 (see References 8 and 9 for reviews). However, other studies have clearly demonstrated arrhythmogenic and proarrhythmic actions of drugs acting to augment IK-ATP.3'81 The activation or opening of these channels during myocardial ischemia is thought to be the cause of the net Received October 1, 1992; revision accepted May 17, 1993. From the Masonic Medical Research Laboratory, Utica, NY. Correspondence to Dr. Charles Antzelevitch, Masonic Medical Research Laboratory, 215 Bleecker Street, Utica, NY 1354. outward movement of potassium that attends metabolic inhibition and may be in large part responsible for the marked shortening of action potential duration (APD) under these conditions.11,2 Also contributing to the abbreviation of APD is a decrease in calcium current.'3,14 Several experimental models have shown that electrical depression attending myocardial ischemia is more accentuated in ventricular epicardium than in endocardium.15-2' Recent work from our laboratory has shown that the differential sensitivity of canine ventricular epicardium and endocardium to ischemia is caused in part by the presence of a prominent transient outward current (Ito) in epicardium but not endocardium.22-25 Marked dispersion of APD and refractoriness leading to extrasystolic activity was found to occur in epicardial preparations exposed to simulated ischemia (KCl 6

1178 Circulation Vol 88, No 3 September 1993 mmol/l, ph 6.8, hypoxia). The dispersion is a result of ischemia-induced loss of the action potential dome (plateau) and marked abbreviation of the epicardial response at some sites but not others. A greater activation of IK-ATP and depression of calcium current in feline epicardial vs endocardial cells after CN-induced metabolic inhibition have also been reported and suggested to contribute to the greater vulnerability of ventricular epicardium during ischemia.26'27 The present study was designed to contrast the effects of pinacidil-induced IK-ATP activation on the electrophysiology of canine ventricular epicardium and endocardium to assess the possible contribution of this component of ischemia to arrhythmogenesis. Preliminary data have been reported previously.28'29 Methods Right ventricular papillary muscles (endocardium: length,.6 to 1. cm; base diameter,.2 to.3 cm) and epicardial strips (approximately 2. x 1.5 x.1 cm) were isolated from hearts removed from anesthetized (sodium pentobarbital, 3 mg/kg) mongrel male dogs. The epicardial preparations consisted of dermatome shavings (Davol Simon Dermatome Power Handle 3293 with cutting head 3295) from the base of the right ventricular free wall. The preparations were allowed to equilibrate for 3 to 4 hours while superfused with oxygenated (95% 2/5% C2) Tyrode's solution maintained at 37±.5C. The composition of the Tyrode's solution was (in mmol/ L): NaCl 129, KCl 4, NaH2PO4.9, NaHCO3 2, CaCl2 1.8, MgSO4.5, and D-glucose 5.5 (ph, 7.35). The tissues were stimulated at basic cycle lengths (BCL) ranging from 5 to 2 milliseconds with rectangular stimuli (2-millisecond duration, 2 to 2.5 times diastolic threshold intensity) delivered through thin silver bipolar electrodes insulated except at the tips. Transmembrane action potentials were recorded from 1 to 3 sites (2 to 3 mm apart) with glass microelectrodes filled with 2.7 mol/l KCl (1 to 3 M k DC resistance) connected to a high-input-impedance-amplification system (World Precision Instruments, New Haven, Conn). Amplified signals were displayed on Tektronix oscilloscopes and either photographed on a 35 mm kymographic camera (Grass) or amplified (model 193-A programmable amplifiers, Cambridge Electronic Designs, Cambridge, England), digitized (model 141 AD/DA system, Cambridge; sampling rate, 7 to 2 khz), analyzed (Everex 386 computer; Spike 2 acquisition and analysis module, Cambridge), and stored on magnetic media (2MB hard disk and digital tape). Recordings were obtained under steady-state stimulation conditions (> 1.5 minutes after change of rate) as well as during rate transitions. Restitution of action potential parameters was determined by use of single test pulses (S2) delivered once after every tenth basic Results Electrophysiological Effects of Pinacidil in Isolated Canine Ventricular Endocardium and Epicardium Recent studies have delineated important distinctions between epicardium and endocardium with respect to action potential configuration and responsiveness to drugs24-283,32-39 (see Reference 22 for review). The presence of a prominent ItO-mediated spike and dome morphology in epicardium but not endocardium has been shown to be largely responsible for the different, sometimes opposite, effects of pharmacological agents in the two tissue types. Fig 1 illustrates that the differential pharmacological responsiveness of endocardium and epicardium, previously demonstrated for calcium and sodium channel blockers and cholinergic and adrenergic agonists and antagonists, applies also to agents that augment potassium current. Fig 1 shows the response of canine ventricular epicardium and endocardium to pinacidil, a KATP channel activator. Shown are transmembrane action potentials recorded at a BCL of 2 milliseconds under control conditions and after 45 minutes of exposure to 1,umol/L pinacidil. Pinacidil produced only a slight abbreviation of APD9 in endocardium (Fig 1, left) but much more pronounced and nonhomogeneous effects in epicardium. The two epicardial action potentials (Fig 1, right), recorded simultaneously from the same preparation, showed an accentuation of phase 1 after exposure to pinacidil. In the upper trace, this effect of pinacidil slowed the emergence of the second upstroke; peak plateau was achieved later and APDgo was prolonged (6 milliseconds; 2.8%). The effect of pinacidil to slightly prolong the APD was seen in 4 of 19 experiments (4 of 48 impalement sites) after 45 minutes of exposure to the drug. In the lower trace, phase 1 (after pinacidil) terminated at a potential negative to -3 mv, and the action potential dome failed to develop; an all-or-none repolarization occurred at the end of phase 1, resulting in a 9millisecond or 42% abbreviation of the epicardial action potential. Our working hypothesis to explain this phenomenon is that the pinacidil-induced IK-ATP and the I,. active at the end of phase 1 together overwhelm the calcium current, resulting in an all-or-none repolarization and abolition of the dome of the epicardial response. There are, no doubt, many factors that contribute to the heterogeneity of response of adjacent epicardial sites to pinacidil, but chief among these is the fact that very small differences in the balance of current at the end of phase 1 can determine whether or not the action poten- beat (S,). The S1-S2 coupling interval was progressively increased from the end of the effective refractory period (ERP) until the next basic beat. APD was measured at 9% repolarization (APD%). The absence of arrhythmic activity or significant change in action potential morphology during a period of 6 to 9 hours after equilibration of the tissues has been demonstrated in previous studies in which similar tissue preparations were used'25.28'332 We conducted two additional control experiments in the present study. Drugs Pinacidil (Leo Pharmaceutical Products, Ballerup, Denmark) was dissolved in ethanol (2% vol/vol) to yield a stock solution of 1 mmol/l. Aliquots were added to Tyrode's solution to achieve final concentrations of 1 to 5,umol/L. 4-Aminopyridine (4-AP) (Sigma Chemical Co, St Louis, Mo) was dissolved in distilled water (stock solution,.5 mol/l) and glybenclamide (Sigma) in ethanol (1% vol/vol) (stock solution, 5 mmol/l). Statistics Statistical analysis was performed using the Student's t test for paired and unpaired data or ANOVA coupled with Scheffe's test, as indicated.

Di Diego and Antzelevitch Pinacidil-Induced Reentry 1179 Endo Epi Pinacidil (1 pm) 'a J f5.i. o a 1 msec FIG 1. Tracings showing electrophysiological effects ofpinacidil (1 umolil) in canine ventricular endocardium (Endo) and epicardium (Epi). Transmembrane action potentials recorded from Endo (left) and Epi (right; two impalements from the same preparation) before (Control) and after 45 minutes of exposure to 1 limolil pinacidil. Basic cycle length, 2 milliseconds. a tial dome develops. Differences in the balance of inward and outward currents may be caused by differences in the intrinsic ionic currents of the cells as well as differences in the responsiveness of the cells to pinacidil. Other factors that can contribute to this heterogeneity include differences in tissue geometry, extracellular space considerations, and differences in passive electrical properties of cells at one epicardial site vs another. Data showing the time course of APD changes during exposure of three epicardial preparations to pinacidil (3 /.mol/l) are graphically presented in Fig 2. Each symbol (El,, and *) depicts the results of a different preparation (three simultaneous impalements in each preparation). This figure illustrates the initial effect of pinacidil to prolong the APD at some epicardial sites as well as the pinacidil-induced great dispersion of repolarization seen with longer exposure. The preparation depicted by shows pinacidil-induced marked abbreviation of APD at two of the three recording sites; the experiment represented by * shows marked abbreviation at one of three recording sites; and that depicted by shows marked abbreviation at all three sites. The Table summarizes the electrophysiological effects of pinacidil in canine ventricular epicardium and endocardium. The results presented are from 19 epicardial preparations in which two or three simultaneous impalements were maintained throughout the experiments (48 observations) and four papillary muscle preparations in which a single impalement was maintained throughout the experimental protocol. Each epicardial preparation was exposed to increasing levels of pinacidil up to 5 pamol/l (range, 1. to 5 pumol/l) in an attempt to induce electrical heterogeneity in the preparation. A similar range of pinacidil concentrations was used to study endocardium. The epicardial data are divided into two groups: those observations in which pinacidil caused loss of the action potential dome and those in which it failed to do so. The average final concentration of pinacidil was similar in the two groups (3.±1.5 and 3.2±1.2 umol/l, respectively). The average concentration used to collect the data in endocardium was comparable (3.±1.4 ttmol/l). The amplitude of phase 1, as well as the membrane voltage at '1 12 ) ~ 12 15 3 45 6 Time (min) FIG 2. Graph showing percent change in action potential duration measured at 9% repolarization (APD%) as a function of time of exposure to pinacidil (3 Zmol/L). Each symbol (o, o, and ) represents the results of a single epicardial preparation from which three simultaneous transmembrane recordings were obtained.

118 Circulation Vol 88, No 3 September 1993 Effect of Pinacidil on Action Potential Parameters in Canine Ventricular Epicardium and Endocardium Epicardium Endocardium n Control P (- dome) n Control P (+ dome) n Control P Amplitude (mv) Phase 31 94.3±5.2 95.4±5.2 17 93.8±6.8 93.2±7.1 4 114.2±1.711 113.3±1.5 Phase 1 31 61.6±4. 58.4±3.8t 17 65.4±6.6* 6.6±5.7t 4 16.±.811 12.±.8* V/Ph1 (mv) 31-22.2±4.1-25.4±4.1t 17-18.4±6.3* -23.1±5.7t 4 23.2±2.511 18.7±1.8* APD9o (ms) 31 189.2±2.5 99.7±18.7t 17 183.4±19.3 168.2±23.9* 4 26.8±14.3 191.5±9.* RMP (mv) 31-83.8±1. -84.1±1.3 17-83.8±1.7-83.7±1.9 4-82.8±2.2-83.3±1.7 Vm. (V/s) 4 182.±13.9 181.8±14.1 4 18.5±15.7 181.2±1.3 2 186.5±13.4 186.±5.6 P (- dome), preparation in which pinacidil abolished the dome. Average pinacidil concentration, 3.± 1.5 ttmol/l (range, 1 to 5 gmol/l). P (+ dome), preparation in which pinacidil did not abolish the dome. Average pinacidil concentration, 3.21.2 gmol/l (range, 2 to 5 umol/l). P (Endocardium), average pinacidil concentration, 3.±1.4 1 mol/l (range, 2 to 5 gmol/l). All values are mean±sd. n, number of individual impalements (23 different experiments). Basic cycle length, 2 milliseconds. All preparations were from the right ventricle. V/Phl, voltage at the end of phase 1; APDu, action potential duration measured at 9% of repolarization; RMP, resting membrane potential. *P<.5, tp<.1 vs control. *P<.5 vs other epicardial control. P<.1 vs epicardial-pinacidil. IIP<.1 vs epicardial controls. the end of phase 1 and APDgo, were significantly reduced in both endocardium and epicardium after exposure to pinacidil. It is noteworthy that the amplitude of phase 1 was smaller and the membrane voltage at the end of phase 1 more negative under control conditions in preparations in which pinacidil caused abolition of the dome. Effect of Pinacidil in Isolated Canine Ventricular Tissues Pretreated With 4-AP Fig 3 illustrates transmembrane activity recorded from canine ventricular endocardium (A) and epicardium (B) under control conditions, during blockade of I,. with 1 mmol/l 4-AP, and after the addition of pinacidil (3,gmol/L) in the continued presence of 4-AP. 4-AP shortened APD9 in epicardium, prolonged this A Endocardium B parameter in endocardium, and greatly attenuated the spike and dome configuration of the epicardial action potential. In the presence of 4-AP, pinacidil produced a comparable abbreviation of the action potential in the two types of tissue (n=3). These data suggest that the presence of a prominent I,.-mediated spike and dome morphology in epicardium but not in endocardium is in large part responsible for the differential responsiveness of the two tissue types to pinacidil. Time and Rate Dependence of the Epicardial APD in the Presence of Pinacidil The time dependence of pinacidil-induced changes in APD of epicardium is illustrated in Fig 4. APD9 is plotted as a function of the S1-S2 interval of premature Epicardium Control 4-AP (1mM) - 4-AP (1mM) Pinacidil (3pM) 1 msec FIG 3. Tracings showing effects ofpinacidil (3,umolIL) in endocardium (A) and epicardium (B) in the presence and absence of the transient outward current blocker 4-aminopyridine (4-AP, 1 mmolil). Basic cycle length, 2 milliseconds. Forty-five minutes of exposure to 4-AP increased action potential duration (APD9,) from 17 to 197 milliseconds in endocardium and decreased APD9, from 184 to 173 milliseconds in epicardium. The addition ofpinacidil (3,umolIL; 45 minutes) caused similar abbreviation ofapd9, in the two tissue types (endocardium, 197 to 171 milliseconds; epicardium, 173 to 133 milliseconds).

Di Diego and Antzelevitch Pinacidil-Induced Reentry 1181 24 EPICARDIUM 2 gc 16 p-12 8 71Control 'o% Pinacidil FIG 4. Graph showing restitution of action potential duration (APD9,) in epicardium in the absence and presence ofpinacidil (2.5,molIL). APD9 is plotted as a function of the S1-S2 interval ofpremature responses elicited once after every 1th basic beat. Basic cycle length, 2 milliseconds. Pinacidil acts to prolong the APD9, of early premature beats and to markedly abbreviate those arriving later in diastole, including the basic beats. 4 4 8 12 S1 -S2 (msec) 16 beats elicited once after every 1 basic beats (BCL of 2 milliseconds). Under control conditions, APD9 increases as a function of the S1-S2 interval. The spike and dome configuration also becomes progressively more prominent in later beats, a reflection of the progressive reactivation of the I. After exposure to pinacidil, the action potential dome of the basic beat is lost (all-or-none repolarization at the end of phase 1), resulting in a marked abbreviation of APD9. Restoration of the dome is seen in responses evoked early in the cycle. Beats introduced later in the cycle once again abbreviate because of loss of the dome. Fig 5 illustrates the dynamic changes in action potential morphology observed in epicardium after an abrupt slowing of the stimulation rate in the presence of pinacidil. The figure shows sequential beat-to-beat changes in action potential morphology after an abrupt slowing of the stimulation rate. The dashed line is the action potential recorded at a BCL of 8 milliseconds. The other traces (solid lines) represent responses recorded immediately after deceleration to a BCL of 2 milliseconds. APD9, although initially prolonged, was markedly abbreviated by the seventh beat. All-or-none repolarization is greatly facilitated at the slower rates by the accentuation of phase 1, which is known to augment with deceleration because of a greater availability of In,. The apparent early afterdepolarizations in beats 5 and 6 are electrotonic manifestations of activity at neighboring sites at which loss of the dome was not synchronous. Pinacidil-Induced Extrasystolic Activity in Epicardium The electrical heterogeneity that develops as a consequence of pinacidil-induced marked abbreviation of the action potential at some epicardial sites but not Pinacidil (3 AM, 3 min) - - - BCL= 8 msec BCL = 2 msec ) 1 msec FIG 5. Tracings showing rate dependence of the epicardial action potential in the presence ofpinacidil. Shown are action potentials recorded after an abrupt deceleration of the stimulation rate from a basic cycle length (BCL) of 8 milliseconds (broken line) to 2 milliseconds (solid lines). The numbers denote the beat number after the change in rate. The apparent early afterdepolarizations in beats 5 and 6 are electrotonic manifestations of activity at neighboring sites where loss of the action potential dome was not synchronous.

1182 Circulation Vol 88, No 3 September 1993 Pinacidil (3pM) 5 mv S1 S2 *2 msec FIG 6. Tracings showing pinacidil-induced reentry in isolated ventricular epicardium. Programmed stimulation. The three traces represent transmembrane activity recorded simultaneously from three different sites of a canine ventricular epicardialpreparationpretreated with pinacidil (3,mol/L; 4 minutes). The first response in each trace is the last of a train of 1 basic beats elicited by stimulation at a basic cycle length of2 milliseconds (S,). The second response is a premature beat evoked by a premature stimulus (S2) introduced at an SrS2 interval of 9 milliseconds. The third beat in the two lower traces represents reentrant reexcitation of these two proximal sites. Conduction of the basic beat was prompt, but that ofthepremature beat was markedly delayed in its attempt to propagate into the area oflonger refractoriness (distal site). The conduction delays were sufficient to permit recurrent reexcitation of the two proximal sites (lower trace). others sets the stage for reentrant arrhythmias. Fig 6 shows an example of extrasystolic activity developing after the introduction of a premature beat. The three traces depict transmembrane activity recorded simultaneously from three sites along a canine ventricular epicardial preparation exposed to 3,umol/L pinacidil for 4 minutes. The first response in each trace is the last of a train of 1 basic beats at a BCL of 2 milliseconds (S,). The second response is a premature beat evoked by a stimulus (S2) introduced at an S1-S2 interval of 9 milliseconds. The action potential at the distal recording (top) shows a dome, whereas those recorded from sites more proximal to the stimulating electrodes are devoid of a dome. Conduction of the premature beat was markedly delayed in its attempt to propagate into the area of longer refractoriness (distal site). The delay was sufficient to permit what can best be characterized as reentrant reexcitation of the two proximal sites (two lower traces). Late premature beats were also observed to give rise to extrasystolic activity, although through a somewhat different mechanism. The example illustrated in Fig 7 was recorded from a canine ventricular epicardial preparation after exposure to 5,umol/L pinacidil. The first propagated response in each panel is the last of a train of 1 beats elicited at a BCL of 1 milliseconds. The second is a response to a premature stimulus introduced progressively later in the cycle. The bottom and top traces were recorded at distances of 3 and 5 mm from the stimulating electrodes. All basic beats were devoid of a dome. Premature beats introduced in early diastole displayed a fully restored dome that propagated normally throughout the preparation, producing uniform electrical depolarization and repolarization (Fig 7A; S1-S2, 3 milliseconds). Premature stimuli introduced very late in diastole also produced uniform electrical activation and repolarization of the epicardial preparation, but in this case the premature response was devoid of a dome (Fig 7D; S1-S2, 8 milliseconds). At intermediate intervals, although activation remained homogeneous, repolarization became heterogeneous, resulting in extrasystolic activity. At an S1-S2 of 445 milliseconds, the dome failed to develop at the distal recording site but did develop in the vicinity of the proximal recording (Fig 7B). The apparent "early afterdepolarization" observed in the terminal part of phase 3 of the distal action potential is not an early afterdepolarization but rather the electrotonic image of the dome of the proximal response, which failed in its attempt to propagate to the distal site. At a slightly longer S1-S2 interval (45 milliseconds), successful propagation of the dome from the proximal to the distal site gave rise to an extrasystole (coupling interval, 15 milliseconds). Note the electrotonic image of the extrasystolic event in the proximal recording, marking its attempt to propagate back to the proximal site. As we will discuss later, this is a novel mechanism of reentry in which a single impulse can split into two independent wave fronts. Extrasystolic activity was also observed under steadystate conditions. Fig 8 illustrates an example of pinacidil-induced extrasystolic activity in an epicardial preparation during basic drive at a BCL of 1 milliseconds. The three traces were recorded simultaneously from three sites along an epicardial preparation exposed to 3 limol/l pinacidil. The top trace was recorded from a site close to a region at which the action potential dome was maintained. This site is also closest to the stimulating electrode (P in Fig 8). The other traces (middle and distal) were recorded at sites at which the normal action potential dome failed to develop. Electrotonic current would be expected to flow from the site displaying a second upstroke to those displaying a short APD and is probably responsible for the generation of an ectopic response at the site depicted by the middle recording. The ectopic impulse propagates normally to the distal site but fails to reexcite the proximal site. In some cases, the extrasystole was observed to reexcite the proximal tissue as well (see Fig lob). Similar arrhythmic activity was observed in 12 of 19 preparations studied (63.2%; pinacidil concentration, 1 to 5,umol/L). Extrasystolic activity occurred over a wide

Di Diego and Antzelevitch Pinacidil-Induced Reentry 1183 A - 51-S2= p 3 c Si-S2-45 D~~~~~~~~~~~~~~~~~~~~~~5 B n old S12 445 U _ S1 -S2 8 l - S S2 Si S2 4 msec FIG 7. Tracings showing pinacidilinduced extrasystolic activity secondary to "late" premature stimulation. The two traces in each panel represent transmembrane activity recorded simultaneously from sites proximal (P) and distal (D) to the stimulating electrodes in a canine ventricular epicardial preparation pretreated with pinacidil (S pumolil). The first response in each panel is the last of a train of 1 basic beats elicited by stimulation at a basic cycle length of 1 milliseconds (S,). The second responses are beats evoked bypremature stimuli (S2) introduced at varying SS2 intervals. Basic beats show loss of the action potential dome. A, Early premature beats show recovery of the dome at all sites, resulting in fairly uniform electrical activity throughout the preparation. B, At intermediate intervals, repolarization became heterogeneous because of recovery of the dome at some sites but not others. The apparent "early afterdepolarization" observed in the distal recording is not an early afterdepolarization but rather the electrotonic image of the dome of the proximal response, which failed in its attempt to propagate to the distal site. C, At a slightly longer S,-S2 interval, successful propagation of the dome from the proximal to the distal site gave rise to an extrasystole (coupling interval, 15 milliseconds); the proximal response is split into two independent wave fronts. D, Premature stimuli introduced very late in diastole also produced uniform electrical activity, in this case caused by loss of the dome at all sites. range of pacing cycle lengths (8 to 2 milliseconds) and showed relatively fixed coupling of the extrasystole to the paced beat. Arrhythmic activity was most readily induced by an abrupt deceleration of the stimulation rate in most preparations. Washout of pinacidil promptly restored electrical homogeneity and aborted all arrhythmic activity (four of four preparations). Blockade of I,, or IK-ATP Restores Electrical Homogeneity and Suppresses Pinacidil-Induced Reentrant Activity in Epicardium Fig 9 shows transmembrane action potentials recorded simultaneously from three sites along a canine ventricular epicardial preparation. Under control conditions, electrical activity throughout the preparation was fairly homogeneous (Fig 9A). Exposure of the preparation to 5 1umol/L pinacidil results in the development of electrical inhomogeneity with loss of the dome at the two proximal sites but not at the distal site. The marked difference in repolarization times results in reentrant excitation at the proximal and middle sites. The activity observed is suggestive of circus movement reentry following a pathway similar to that indicated by the arrows. Fig 9C shows the antiarrhythmic effects of 4-AP (.5 mmol/l, 1 minutes) with continued exposure to pinacidil (5 ttmol/l). 4-AP-induced inhibition of I,. restored electrical homogeneity to the preparation and abolished all arrhythmic activity. Washout of 4-AP was attended by reappearance of electrical inhomogeneity and reentrant activity (Fig loa). The presumed circus movement pathway is once again indicated by the arrows. Fig lob illustrates the antiarrhythmic effects of glybenclamide, an inhibitor of IK-ATP, in the continued presence of pinacidil. The effects of pinacidil were completely reversed after 3 minutes of exposure to glybenclamide (1 limol/l), electrical homogeneity was restored, and all reentrant activity was abolished. 4-AP and glybenclamide were observed to exert similar antiarrhythmic effects in three other preparations in which pinacidil induced arrhythmias.

1184 Circulation Vol 88, No 3 September 1993 PMD I M P O- T 4 4 I 2 msec FIG 8. Tracings showing pinacidil-induced reentry in isolated canine ventricular epicardium during steady-state stimulation; basic cycle length, 1 milliseconds. The action potential traces were recorded simultaneously from three sites along an epicardial preparation pretreated with pinacidil (3,umol/L; 45 minutes). Pinacidil-induced dispersion of repolarization (spike and dome configuration at the proximal but not at the middle and distal sites) gives rise to two independent impulses; the first is caused by the normal propagation of phase, and the second is caused by the propagation ofphase 2 in the same direction. P, proximal; M, middle; D, distal. APD measurements attending 4-AP and glybenclamide reversal of pinacidil-induced dispersion of repolarization are graphically presented in Figs 11 and 12. Shown are data obtained under control conditions, after pinacidil (3 to 5,umol/L, 3 to 6 minutes), and during exposure to 4-AP (.5 to 1 mmol/l, 15 to 3 minutes) or glybenclamide (1 to 2,umol/L, 2 to 45 minutes) in the continued presence of pinacidil. 4-AP prolonged the A Control Distal Proximal B Pinacidil (5 pm) C Pinacidil + 4-AP (5 pm) (.5 mm) action potential at all sites at which pinacidil had caused marked abbreviation but abbreviated or caused little change in APD at sites where the effect of pinacidil on APD was small (Fig 11). Glybenclamide readily reversed the effects of pinacidil at all sites (Fig 12). Discussion KATP channels are potassium channels whose activity is normally inhibited by physiological levels of intracellular ATP (see Reference 4 for review). First identified in cardiac muscle,41 they have been described in pancreatic 13-cells,42 skeletal43 and smooth muscle,44 and neurons.45 Potassium channel activators or openers represent several distinct chemical classes of drugs (eg, cyanoguanidines, benzopyrans, and nicotinamide)46 that act by increasing membrane conductance to potassium, principally through augmentation of the KATP channel current.l4.47 In clinical trials to test their efficacy as antihypertensive agents, potassium channel activators were observed to produce ECG T-wave changes (flat, inverted, or biphasic) in some patients,48 suggesting significant changes in repolarization of ventricular tissues. These findings, coupled with those of other studies demonstrating direct effects of the potassium channel activators to facilitate repolarization of the cardiac action potential in vitro,2,3 prompted a series of investigations into the potential antiarrhythmic as well as arrhythmogenic actions of these drugs. Since electrical changes attending ischemia are believed to be caused in part by activation of KATP channels,4 potassium channel activators such as pinacidil serve to mimic this component of ischemia. Our study was designed to assess the electrophysiological effects of the KATP channel activator pinacidil in isolated canine ventricular epicardium and endocardium and to evaluate the potential arrhythmogenicity of 2 msec FIG 9. Tracings showing the effects of 4-aminopyridine, a transient outward current blocker, on pinacidil-induced extrasystolic activity. Shown are transmembrane action potentials recorded simultaneously from three different sites along a canine ventricular epicardial preparation. A represents the control situation at a basic cycle length of 2 milliseconds. B, recorded after exposure of the preparation to 5 pmolil pinacidil shows the development of electrical inhomogeneity with loss of the dome at the two proximal sites but not at the distal site. The marked difference in repolarization times results in reentrant excitation at theprotimal and middle sites. C shows the effects of 4-aminopyridine (4-AP), a transient outward current blocker (.5 mmolil; 1 minutes), in the continuedpresence ofpinacidil Inhibition of the transient outward current restores electrical homogeneity to the preparation, thus abolishing all arrhythmic activity.

Di Diego and Antzelevitch Pinacidil-Induced Reentry 1185 A Distal, Proximal Pinacidil - 4-AP (5 pm) (.5 mm) B Pinacidil+Glybenclamide (5 pm) (1 pm) 1 2 msec FIG 1. Tracings showing the effects of glybenclamide on pinacidil-induced extrasystolic activity. Washout of 4-aminopyridine (4-AP) (A), in the same preparation as in Fig 8, is attended by reappearance of electrical inhomogeneity and reentrant activity. The presumed circus movement pathway is indicated by the arrows. B illustrates the effects of glybenclamide, an inhibitor of the ATP-sensitive potassium current, in the continuedpresence of pinacidil. The effects of pinacidil were completely reversed after 15 minutes of exposure to 1 pamol/l glybenclamide, electrical homogeneity was restored, and all reentrant activity was abolished. the drug. Our results show that pinacidil can produce major changes in the epicardial action potential at concentrations that exert little effect on endocardium. The differences in the response of the two tissue types to this agent can be explained on the basis of differences in the balance of inward and outward currents present at the end of phase 1 of the action potential. Unlike endocardium, epicardium exhibits a prominent I,mediated phase 1. Inactivation of I,. and activation of the calcium current gives rise to phase 2 and the dome of the epicardial action potential. The imposition of an additional outward current (IK-ATP) during the early phases of the action potential would be expected to accentuate phase 1. Thus, pinacidil would be expected 24-22 - to cause phase 1 to terminate at more negative potentials, as we observed in our epicardial preparations. At these potentials, the inward calcium current (and slowly inactivating sodium current) may be overwhelmed by the outward currents. This, we believe, is the basis for the all-or-none repolarization at the end of phase 1 and marked abbreviation of the epicardial response in the presence of pinacidil. When the balance of current after pinacidil becomes inward at the end of phase 1, the dome is maintained but its emergence may be slowed or delayed (Fig 1). When the dome is maintained, the accentuation of the notch tends to prolong the action potential, whereas the presence of an additional outward current during phase 3 tends to abbreviate it. Both 22-2 - 2 - o 18 E 16 O 18 -,- 1o - E CD 1 14 12 1 - < 14 12 an 5 CONTROL PINMDL 4-P FIG 11. Graph showing 4-aminopyridine (4-AP) reversal of pinacidil-induced dispersion of repolarization. Each symbol (A, o, and *) represents data from individual experiments in which three simultaneous transmembrane recordings were obtained. Action potential duration (APD9) data were obtained under control conditions, after pinacidil (3 to 5,gmol/L; 3 to 6 minutes), and during exposure to 4-AP (.5 to 1 mmol/l; 15 to 3 minutes) in the continued presence of pinacidil. 8-54i a CONTROL PW~DL GLYDENCUMDE FIG 12. Graph showing glybenclamide reversal ofpinacidilinduced dispersion of repolarization. Each symbol represents data from individual experiments in which two or three simultaneous transmembrane recordings were obtained. Action potential duration (APD9,) data were collected under control conditions, after pinacidil (2 to 3 limol/l; 3 to 6 minutes), and during exposure to glybenclamide (1 to 2,gmolIL; 2 to 45 minutes) in the continued presence of pinacidil.

1186 Circulation Vol 88, No 3 September 1993 were observed with low concentrations of pinacidil, especially during the initial 2 to 3 minutes of drug exposure. The absence of a prominent I,. in endocardium is probably responsible for the lack of more dramatic pinacidil-induced changes in action potential morphology (Fig 1). The principal effect of low doses of pinacidil in endocardium is a modest abbreviation of the action potential, as would be expected with relatively small activation of IK-ATP. Our results suggest that the presence of a prominent I,.-mediated spike and dome morphology in epicardium but not endocardium is, in large part, responsible for the differential responsiveness of these two tissue types to pinacidil. This hypothesis is supported by the observation that pinacidil produces similar effects in epicardial and endocardial preparations pretreated with 4-AP, an I,. blocker (Fig 3). Further support for the hypothesis derives from the observation that in epicardium, premature beats, in which I, is greatly diminished (because of slow reactivation kinetics), show only a slight abbreviation in response to pinacidil (Fig 4). Our data suggest that a relatively small increase in the intensity of KATP, as may occur during myocardial ischemia, can contribute importantly to the development of electrical inhomogeneity in the ventricle. Other factors believed to contribute to a differential response of epicardium and endocardium to ischemia include the greater sensitivity of IK-ATP in epicardium to intracellular ATP levels, as demonstrated by Furukawa and coworkers,27 and a greater depression of calcium current in epicardium to metabolic inhibition, as recently described in feline ventricular tissues by Kimura et al.26 Pinacidil-induced loss of the action potential dome causes dispersion of repolarization (and refractoriness) to develop between epicardium and endocardium as well as among different epicardial sites (Figs 1, 2, and 6 through 12). This heterogeneous response to pinacidil creates a substrate favoring the development of arrhythmias. The pinacidil-induced extrasystolic activity observed in our epicardial preparations appears most consistent with a reentrant mechanism in that the propagating impulse fails to die out after normal activation of the tissue but persists to reexcite certain sites after expiration of their refractory period.49 The electrotonic currents, which continue to flow from the site of long APD to the site of short APD after repolarization of the latter, represent the "impulse that fails to die" and that serves to reexcite the tissue. Although the precise mechanism responsible for pinacidil-induced reentry cannot be discerned from the limited data available from this study, our results suggest that heterogeneous repolarization at neighboring epicardial sites gives rise to electrotonic currents that cause reentrant excitation via a mechanism akin to reflection (pseudoreflection) or via a circus movement mechanism. In typical reflection, delayed conduction of the impulse permits local circuit (electrotonic) current generated by activation of distal tissues to feed back to the proximal site after expiration of its refractory period, thus effecting reexcitation. It is the appearance of a prominent action potential dome at site A but not site B that generates local circuit current that reexcites site B after expiration of its brief refractory period. In other words, it is the propagation of the action potential plateau across the same pathway (reflection) or alternate pathway (circus movement) as that used for propagation of phase that is responsible for reentry. In some cases, when the dome is maintained at the proximal site but not the distal site, propagation of the dome leading to reexcitation of the distal site occurs in the same direction as phase. Thus, a single proximal impulse is split, giving rise to two independent impulses at the distal site that may or may not reenter the proximal tissue (Figs 8 and 9B). These are novel mechanisms that we have collectively called "phase 2 reentry." They are conceptually similar to the mechanism that Brugada and Wellens5 called prolonged repolarization-dependent reexcitation and others have called focal reexcitation.51-53 The splitting of a single wave into two is also akin to pseudoreflection.54-57 We have observed and mapped a similar phenomenon in canine ventricular epicardial sheets exposed to simulated ischemia and flecainide.22,32'58'59 A more definitive assessment as to whether and to what extent reflection or circus movement mechanisms are involved must await high-resolution mapping under these conditions. The participation of other mechanisms, namely enhanced and abnormal impulse formation, are discounted on the basis of our failure to observe activity consistent with early or delayed afterdepolarizationinduced triggered activity or any form of pacemaker activity in the experiments performed in the course of this study. Although humps or deflections suggestive of early afterdepolarizations were frequently recorded, in all cases these could be ascribed to the electrotonic interactions among cells at different levels of repolarization. The known effectiveness of pinacidil in suppressing, rather than promoting, arrhythmic activity caused by triggered activity or abnormal automaticity is also relevant to this issue.7 Our results are consistent with those of Chi and coworkers1 showing a profibrillatory effect of pinacidil in the conscious dog exposed to acute ischemia superimposed on a previous infarction. These authors suggested that pinacidil may promote reentrant activity by contributing to the development of a dispersion of refractoriness in the ventricle. The results of Padrini and coauthors6 showing that pinacidil (1 to 1,mol/L) can induce ventricular fibrillation in isolated perfused guinea pig hearts are also consistent with our findings. A preliminary report by Fagbemi et a161 describes the effect of pinacidil to increase the incidence of ventricular fibrillation during hypoxia and reoxygenation of Langendorff-perfused rabbit hearts ([K']., 2.5 mmol/l). The profibrillatory effect of pinacidil was mimicked by phorbol 12,13-dibutyrate. In their study, as in ours, the proarrhythmic effects of both agents were antagonized by glybenclamide. These results were interpreted as suggesting that the potassium channel openers may be influencing the potassium channel via a protein kinase C-associated mechanism. Pinacidil and cromakalim, a benzopyran potassium channel activator, have been shown to exert proarrhythmic effects in isolated rat hearts subjected to low-flow ischemia62 and ischemia/reperfusion.63'64 Potassium channel activators have also been found to produce proarrhythmic effects in isolated Purkinje fibers under some conditions. High concentrations of pinacidil have

Di Diego and Antzelevitch Pinacidil-Induced Reentry 1187 also been shown to accelerate automaticity in bariumtreated Purkinje fibers.3 Contrasting results have been reported in other studies. Carlsson and coworkers4'5 recently reported that pinacidil and two analogues are effective in suppressing clofilium-induced polymorphous ventricular tachyarrhythmias in vivo and early afterdepolarizations and triggered activity in vitro in a rabbit model. The antiarrhythmic potential of potassium channel activators such as pinacidil, cromakalim, and nicorandil has also been explored in several other recent studies. Using isolated canine ventricular myocytes, Spinelli and coworkers7 showed that pinacidil can suppress or diminish early afterdepolarizations induced by Bay K 8644, ketanserin, or applied current, as well as delayed afterdepolarizations induced by ouabain and abnormal automaticity induced by barium. Pinacidil and cromakalim have also been shown to abolish early afterdepolarizations and triggered activity in canine Purkinje fibers exposed to quinidine, cesium, or sematilide.6 Cromakalim65 and nicorandil66 are also effective in decreasing automaticity65 in Purkinje fiber preparations. In vivo data are more limited; in addition to the study by Carlsson and coworkers,4 a study by Fish and coworkers6 demonstrates an effect of pinacidil to blunt arrhythmias induced by cesium in a rabbit model, and a study by Kerr and coworkers67 demonstrates an antiarrhythmic effect of pinacidil in the subacute phase of myocardial infarction in a canine model. Grover et a163 reported that intracoronary cromakalim reduced the incidence of ventricular fibrillation after ischemia/reperfusion in the anesthetized dog. Despite its effects to diminish digitalisinduced delayed afterdepolarizations in isolated ventrcular myocytes,7 pinacidil proved ineffective in suppressing ouabain-induced arrhythmias in a canine model.67 Pinacidil was also without effect in altering the responsiveness of postinfarction canine hearts with respect to electrical induction of tachyarrhythmias.1 In conclusion, our data suggest that a relatively small increase in the intensity of IKATP can contribute importantly to the development of cardiac arrhythmias. The available data suggest that blockade of I,, and/or IK-ATP may be useful antiarrhythmic interventions under ischemic or "ATP-depleted" conditions. The applicability of these findings to humans must be approached with some caution. The clinical relevance of the pinacidil-induced changes observed in canine hearts presumes the presence of similar electrophysiological distinctions between epicardium and endocardium in human ventricles or at least the presence of a similar Itmediated spike and dome in epicardium. Although direct evidence for electrophysiological distinctions between human ventricular endocardium and epicardium is lacking, indirect evidence for the presence of an I,.-mediated (4-AP-sensitive) spike and dome action potential morphology in human ventricular epicardium has recently been described in a preliminary report by Chiamvimonvat and coworkers.68 Although the potential arrhythmogenicity of potassium channel activators remains to be fully defined in the clinic,48 the available data suggest that because of the many variables involved, this issue may be as elusive in humans as it has proved to be in animal models. Marked electrophysiological heterogeneity between epicardium and endocardium as well as among different epicardial sites has been observed in response to a wide variety of agents, including acetylcholine, sodium and calcium channel blockers, and conditions such as hypoxia and ischemia.22 It seems reasonable to speculate that these agents and conditions may act synergistically with pinacidil and other potassium channel activators to produce dispersion of repolarization and proarrhythmia. These findings suggest that in the presence of high vagal tone and/or agents that inhibit calcium current (eg, adenosine), a relatively small increase in IK-ATP, as may occur during mild ischemia, could more readily induce electrical heterogeneity and extrasystolic activity. These other factors are also likely to shift the rate dependence of arrhythmic activity to faster heart rates. A similar shift in rate dependence might also be expected with conditions that give rise to progressively greater activation of IK-ATP, within limits. Although not addressed in this study, data obtained from parallel studies suggest major differences in the spike and dome morphology of right vs left ventricular epicardium in the canine heart. Preliminary studies indicate that pinacidil-induced loss of the dome and arrhythmogenesis are more difficult to achieve in left ventricular epicardium because of a smaller notch in the action potential of this tissue. Loss of the dome could be achieved in the left ventricle with higher concentrations of pinacidil (J.M. Di Diego and C. Antzelevitch, unpublished observation). Like most cardioactive agents, potassium channel activators appear to be able to exert proarrhythmic as well as antiarrhythmic effects, depending on preexisting conditions.7'8 The available data suggest that cardioselective potassium channel activators may prove valuable in combating arrhythmias caused by repolarization abnormalities but may be proarrhythmic in the face of :eentrant or other arrhythmias. It should be noted that ischemia-induced loss of the dome, when homogeneous, may be an important protective mechanism. Suppression of the action potential dome in epicardium after modest activation of IK-ATP could protect the myocardium by inhibiting contractile function in the affected tissue, thus greatly diminishing oxygen demand. The protective role of KATP channel activation in limiting infarct size,6364 preventing contracture,69 and mediating myocardial preconditioning7o suggests that there may be a fine line between the conditions under which potassium channel activators exert antiarrhythmic vs arrhythmogenic actions during ischemia and reperfusion. Acknowledgments This study was supported by grant HL-37396 from the National Institutes of Health, a Fellowship grant from the American Heart Association, New York State Affiliate, and grants from the Charles L. Keith and Clara Miller Foundation and Josephine Lawrence Hopkins Foundation. We wish to thank Judy Hefferon and Robert Goodrow for their skilled technical assistance. Pinacidil was generously provided by Leo Pharmaceutical Products. References 1. Tseng GN, Hoffman BF. Actions of pinacidil on membrane currents in canine ventricular myocytes and their modulation by intracellular ATP and camp. Pflugers Arch. 199;415:414-424. 2. Arena JP, Kass RS. Enhancement of potassium-sensitive current in heart cells by pinacidil: evidence for modulation of the ATPsensitive potassium channel. Circ Res. 1989;65:436-445.

1188 Circulation Vol 88, No 3 September 1993 3. Steinberg MI, Ertel P, Smallwood JK, Wyss V, Zimmerman K. The relation between vascular relaxant and cardiac electrophysiological effects of pinacidil. J Cardiovasc Pharmacol. 1988; 12(suppl 2):S3S4. 4. CarIsson L, Abrahamsson C, Drews C, Duker G. Antiarrhythmic effects of potassium channel openers in rhythm abnormalities related to delayed repolarization in the rabbit. Circulation. 1992; 85:1491-15. 5. CarIsson L. Suppression of clofilium-induced polymorphous ventricular tachyarrhythmias by pinacidil and related pyridylcyanoguanidines in the anesthetized rabbit. Circulation. 199;82(suppl III):III-529. Abstract. 6. Fish FA, Prakash C, Roden DM. Suppression of repolarizationrelated arrhythmias in vitro and in vivo by low-dose potassium channel activator. Circulation. 199;82:1362-1369. 7. Spinelli W, Sorota S, Siegel M, Hoffman BF. Antiarrhythmic actions of the ATP-regulated K' current activated by pinacidil. Circ Res. 1991;68:1127-1137. 8. Antzelevitch C, Di Diego JM. Role of K' channel activators in cardiac electrophysiology and arrhythmias. Circulation. 1992;85: 1627-1629. 9. Lynch JJ, Sanguinetti MC, Kimura S, Bassett AL. Therapeutic potential of modulating potassium currents in the diseased myocardium. FASEB J. 1992;6:2951-296. 1. Chi L, Uprichard ACG, Lucchesi BR. Profibrillatory actions of pinacidil in a conscious canine model of sudden coronary death. J Cardiovasc Pharmacol. 199;15:452-464. 11. Vleugels A, Vereecke J, Carmeliet E. Ionic currents during hypoxia in voltage-clamped cat ventricular muscle. Circ Res. 198; 47:51-58. 12. Isenberg G, Vereecke J, Van der Heyden G, Carmeliet E. The shortening of the action potential by DNP in guinea pig ventricular myocytes is mediated by an increase of the time-independent K conductance. Pflugers Arch. 1983;397:251-259. 13. McDonald TF, MacLeod DP. Metabolism and the electrical activity of anoxic ventricular muscle. J Physiol. (Lond) 1973;229: 559-582. 14. Payet MD, Schanne OF, Ruiz-Ceretti E, Demmers JM. Slow inward currents and outward currents of rat ventricular fibers under anoxia. J Physiol (Paris). 1978;74:31-35. 15. Gilmour RF, Zipes DP. Different electrophysiological responses of canine endocardium and epicardium to combined hyperkalemia, hypoxia, and acidosis. Circ Res. 198;46:814-825. 16. Kimura S, Bassett AL, Kohya T, Kozlovskis PL, Myerburg RJ. Simultaneous recording of action potentials from endocardium and epicardium during ischemia in the isolated cat ventricle: relation of temporal electrophysiologic heterogeneities to arrhythmias. Circulation. 1986;74:41-49. 17. Boineau JP, Cox JL. Slow ventricular activation in acute myocardial infarction: a source of reentrant premature ventricular contractions. Circulation. 1973;48:72-713. 18. Scherlag BJ, El-Sherif N, Hope R, Lazzara R. Characterization and localization of ventricular arrhythmias resulting from myocardial ischemia and infarction. Circ Res. 1974;35:372-383. 19. Williams DO, Scherlag BJ, Hope R, El-Sherif N, Lazzara R. The pathophysiology of malignant ventricular arrhythmias during acute myocardial infarction. Circulation. 1974;5:1163-1172. 2. Elharrar V, Gaum WE, Zipes DP. Effects of drugs on conduction delay and the incidence of ventricular arrhythmias induced by acute coronary occlusion in dogs. Am J Cardiol. 1977;39:544-549. 21. Ruffy R, Lovelace ED, Knoebel SB, Mueller TM, Zipes DP. Relationship between changes in left ventricular bipolar electrograms and regional myocardial blood flow during acute coronary occlusion in the dog. Circ Res. 1979;45:764-77. 22. Antzelevitch C, Sicouri S, Litovsky SH, Lukas A, Krishnan SC, Di Diego JM, Gintant GA, Liu DW. Heterogeneity within the ventricular wall: electrophysiology and pharmacology of epicardial, endocardial and M cells. Circ Res. 1991;69:1427-1449. 23. Lukas A, Antzelevitch C. Reentry induced by "ischemia" in isolated canine subepicardium: role of the transient outward current. PACE. 199;13:5. Abstract. 24. Antzelevitch C, Litovsky SH, Lukas A. Ventricular epicardium vs. endocardium: electrophysiology and pharmacology. In: Zipes D, Jalife J, eds. Cardiac Electrophysiology, From Cell to Bedside. New York, NY: WB Saunders; 199:386-395. 25. Lukas A, Antzelevitch C. The transient outward current contributes to selective depression of electrical activity in canine epicardium under ischemic conditions. PACE. 1989;12:639. Abstract. 26. Kimura S. Bassett AL, Furukawa T, Furukawa N, Myerburg RJ. Differences in the effect of metabolic inhibition on action potentials and calcium currents in endocardial and epicardial cells. Circulation. 1991;84:768-777. 27. Furukawa T, Kimura S, Cuevas J, Furukawa N, Bassett AL, Myerburg RJ. Role of cardiac ATP-regulated potassium channels in differential responses of endocardial and epicardial cells to ischemia. Circ Res. 1991;68:1693-172. 28. Di Diego JM, Antzelevitch C. Pinacidil-induced reentrant arrhythmias in isolated canine ventricular epicardium. Circulation. 199;82(suppl III):III-528. Abstract. 29. Di Diego JM, Antzelevitch C. Antiarrhythmic effects of 4-aminopyridine (4-AP) and glyburide in isolated ventricular myocardium. PACE. 1991;14:649. Abstract. 3. Litovsky SH, Antzelevitch C. Differences in the electrophysiological response of canine ventricular subendocardium and subepicardium to acetylcholine and isoproterenol: a direct effect of acetylcholine in ventricular myocardium. Circ Res. 199;67:615-627. 31. Di Diego JM, Antzelevitch C. The electrophysiologic effects of amiloride in canine ventricular myocardial tissues. FASEB J. 199; 4:A-562. Abstract. 32. Krishnan SC, Antzelevitch C. Sodium channel blockade produces opposite electrophysiologic effects in canine ventricular epicardium and endocardium. Circ Res. 1991;69:277-291. 33. Litovsky SH, Antzelevitch C. Transient outward current prominent in canine ventricular epicardium but not endocardium. Circ Res. 1988;62:116-126. 34. Litovsky SH, Antzelevitch C. Rate dependence of action potential duration and refractoriness in canine ventricular endocardium differs from that of epicardium: the role of the transient outward current. JAm Coll Cardiol. 1989;14:153-166. 35. Litovsky SH, Sicouri S, Antzelevitch C. Electrophysiologic effects of amiodarone on canine ventricular epicardium and endocardium differ. Circulation. 199;82(suppl III):III-529. Abstract. 36. Litovsky SH, Antzelevitch C. Differences in the electrophysiology of ventricular epicardium and endocardium as the basis for the Osborne wave. Circulation. 1989;8(suppl II):II-129. Abstract. 37. Furukawa T, Myerburg RJ, Furukawa N, Bassett AL, Kimura S. Differences in transient outward currents of feline endocardial and epicardial myocytes. Circ Res. 199;67:1287-1291. 38. Kimura S, Bassett AL, Furukawa T, Cuevas J, Myerburg RJ. Electrophysiological properties and responses to simulated ischemia in cat ventricular myocytes of endocardial and epicardial origin. Circ Res. 199;66:469-477. 39. Fedida D, Giles WR. Regional variations in action potentials and transient outward current in myocytes isolated from rabbit left ventricle. J Physiol (Lond). 1991;442:191-29. 4. Nichols CG, Lederer WJ. Adenosine triphosphate-sensitive potassium channels in the cardiovascular system. Am J Physiol. 1991;261:H1675-H1686. 41. Noma A. ATP-regulated K' channels in cardiac muscle. Nature. 1983;135:147-148. 42. Ashcroft FM, Harrison DE, Ashcroft SJH. Glucose induces closure of single potassium channels in isolated rat pancreatic beta-cells. Nature. 1984;312:446-448. 43. Spruce AE, Standen NB, Stanfield PR. Voltage-dependent ATPsensitive potassium channels of skeletal muscle membrane. Nature. 1985;316:736-738. 44. Standen NB, Quayle JM, Davies NW, Brayden JE, Huang Y, Nelson MT. Hyperpolarizing vasodilators activate ATP-sensitive K+ channels in arterial smooth muscle. Science. 1989;245:177-18. 45. Ashford MU, Sturgess NC, Trout NJ, Gardner NJ, Hales CN. Adenosine-5'-triphosphate-sensitive ion channels in neonatal rat cultured central neurones. Pflugers Arch. 1988;412:297-34. 46. Edwards G, Weston AH. Potassium channel openers and vascular relaxation. Pharmacol Ther. 199;48:237-258. 47. Hermsmeyer RK. Pinacidil actions on ion channels in vascular muscle. J Cardiovasc Pharmacol. 1988;12(suppl 2):S17-S22. 48. Goldberg MR. Clinical pharmacology of pinacidil: a prototype for drugs that affect potassium channels. J Cardiovasc Pharmacol. 1988;12(suppl 2):S41-S47. 49. Moe GK. Evidence for reentry as a mechanism for cardiac arrhythmias. Rev Physiol Biochem Pharmacol. 1975;72:55-81. 5. Brugada P, Wellens HJJ. Early afterdepolarizations: role in conduction block, "prolonged repolarization-dependent reexcitation," and tachyarrhythmias in the human heart. PACE. 1985;8:889-896. 51. Hoffman BF. The genesis of cardiac arrhythmias. Prog Cardiovasc Dis. 1966;8:319-329. 52. Han J, Moe GK. Nonuniform recovery of excitability in ventricular muscle. Circ Res. 1964;14:44-6. 53. Janse MJ, Van Capelle FJL, Morsink H, Kleber AG, Wilms- Schopman F, Cardinal R, Naumann D'Alnoncourt C, Durrer D.

Di Diego and Antzelevitch Pinacidil-Induced Reentry 1189 Flow of "injury" current and patterns of excitation during early ventricular arrhythmias in acute regional myocardial ischemia in isolated porcine and canine hearts: evidence for two different arrhythmogenic mechanisms. Circ Res. 198;47:151-167. 54. Scherf D, Cohen J. The Atrioventricular Node and Selected Cardiac Arrhythmias. New York, NY: Grune and Stratton; 1964:23259. 55. Cranefield PF. The Conduction of the Cardiac Impulse. Mount Kisco, NY: Futura; 1975:153-197. 56. Antzelevitch C, Lukas A. Reflection and circus movement reentry in isolated atrial and ventricular tissues. In: Dangman KH, Miura DS, eds. Electrophysiology and Pharmacology of the Heart. A Clinical Guide. New York, NY: Marcel Dekker; 1991:251-275. 57. Janse MJ. Reentry rhythms. In: Fozzard HA, Haber E, Jennings RB, Katz AM, Morgan HE, eds. The Heart and Cardiovascular System. New York, NY: Raven Press; 1986:123-1238. 58. Lukas A, Antzelevitch C. Antiarrhythmic effects of 4-aminopyridine, a transient outward current blocker, on "ischemia"-induced reentry in isolated canine ventricular subepicardium. Circulation. 1991; 84(suppl II):II-56. Abstract. 59. Krishnan SC, Antzelevitch C. Flecainide-induced arrhythmia in canine ventricular epicardium: phase 2 reentry? Circulation. 1993; 87:562-572. 6. Padrini R, Bova S, Cargnelli G, Piovan D, Ferrari M. Effects of pinacidil on guinea-pig isolated perfused heart with particular reference to the proarrhythmic effect. Br J Pharmacol. 1992;15: 715-719. 61. Fagbemi SO, Black SC, Chi L, Cunningham MR, Lucchesi BR. Pinacidil- and phorbol 12,13-dibutyrate-induced ventricular fibrillation: antagonism by glybenclamide. Circulation. 1991;84(suppl II):II-52. Abstract. 62. Wolleben CD, Sanguinetti MC, Siegel PKS. Influence of IK(ATP)- sensitive channel modulators on ischemia-induced fibrillation in isolated rat hearts. J Mol Cell Cardiol. 1992;21:783-788. 63. Grover GJ, Sleph PG, Dzwonczyk S. Pharmacologic profile of cromakalim in the treatment of myocardial ischemia in isolated rat hearts and anesthetized dogs. J Cardiovasc Pharmacol. 199;16: 853-864. 64. Grover GJ, Dzwonczyk S, Parham CS, Sleph PG. The protective effects of cromakalim and pinacidil on reperfusion function and infarct size in isolated perfused rat hearts and anesthetized dogs. Cardiovasc Drugs Ther. 199;4:465-474. 65. Bril A, Man RY. The role of potassium and sodium-calcium exchange currents in the action potential durations of normal Purkinje fibres and Purkinje fibres surviving infarction. Cardiovasc Res. 1989;23:41416. 66. Imanishi S, Arita M, Aomine M, Kiyosue T. Antiarrhythmic effects of nicorandil on canine Purkinje fibers. J Cardiovasc Pharmacol. 1984;6:772-779. 67. Kerr MJ, Wilson R, Shanks RG. Suppression of ventricular arrhythmias after coronary artery ligation by pinacidil, a vasodilator drug. J Cardiovasc Pharmacol. 1985;7:875-883. 68. Chiamvimonvat N, Wang L, Kieser TM, Maitland A, Duff HJ. Evidence for a transient outward current in human ventricular epicardium. Circulation. 1991;84(suppl II):II-14. Abstract. 69. Cole WC, McPherson CD, Sontag D. ATP-regulated K' channels protect the myocardium against ischemia/reperfusion damage. Circ Res. 1991;69:571-581. 7. Gross GJ, Auchampach JA. Blockade of ATP-sensitive potassium channels prevents myocardial preconditioning in dogs. Circ Res. 1992;7:223-233.