Background: Despite a long history of use in the treatment of paroxysmal atrial fibrillation (AF), the efficacy of cardiac glycosides has not been established. If such drugs are beneficial in this condition, the general view is that the benefit must be related to their inotropic actions.
Methods and results: To assess the effects of the rapid-acting cardiac glycoside, acetylstrophanthidin (AS), on AF and AF-induced right atrial (RA) "stunning," RA wall motion (with ultrasonic crystals), RA pressure, and peak first derivative of pressure (dp/dt) (with microtip transducers) were measured before and after 5 min of high-intensity rapid atrial stimulation (10 Hz; 10 mA; 1 ms) and after the cessation of poststimulation AF. Measurements were made in neurally intact and autonomically blockaded dogs both before and after the administration of AS (0.01 mg/kg IV bolus and 0.015 mg/kg/h IV infusion). AS prevented the post-AF reduction in RA peak dp/dt under neurally intact and autonomically blockaded conditions, and it prevented the post-AF increase in the RA end-systolic dimension and the decrease in the percentage of RA systolic shortening with autonomic blockade. AS was beneficial whether or not baseline inotropy was enhanced by AS. The duration of AF following atrial stimulation was the same before and after AS, but when compared to controls, AS treatment appeared to prolong AF.
Conclusions: Cardiac glycosides exert a favorable effect on AF-induced RA stunning, but this action is unrelated to its effects on the duration of AF. (CHEST 2000; 118:1116-1126)
Key words: atrium; cardiac glycosides; contractility; fibrillation
Abbreviations: AF = atrial fibrillation; ANOVA = analysis of variance; AS = acetylstrophanthidin; dp/dt = first derivative of pressure; LV = left ventricle, ventricular; RA = right atrium, atrial; RAEDD = right atrial end-diastolic dimension; RAESD = right atrial end-systolic dimension; RAMD = right atrial maximal dimension; RAP = right atrial pressure; SS = systolic shortening; SSI = systolic shortening index
A better understanding of atrial fibrillation (AF) has emerged from experimental investigations performed during the past several years. Studies using isochronal mapping of the atrium in experimental animals[1,2] and in humans[3,4] have confirmed the "circus movement" theory of AF. These studies have demonstrated that the arrhythmia is a manifestation of multiple, changing circuits[1-4] and conditions that favor the formation of multiple disparate loops (ie, the slowing of atrial conduction and the shortening of atrial refractoriness by reducing the wavelength of atrial impulse) foster the arrhythmia. Moreover, other investigations have demonstrated that AF itself creates conditions that favor the perpetuation of the arrhythmia. AF results in persistent shortening of atrial refractoriness, referred to as "electrical remodeling,"[7-14] and abnormalities of atrial contractile function, so-called "stunning.[14-20] Even just a few minutes of AF can produce these changes,[14,15,20] and the longer that AF lasts, the worse and more persistent the findings.[7,10,14,15,17]
This new information should provide a more rational approach to the therapeutics of AF. Despite a long history of use in AF, the benefits of cardiac glycosides in the treatment of this arrhythmia remain controversial.[21-24] Cardiac glycosides are now used in AF mainly for ventricular rate control, an effect most evident at rest. This dromotropic action of cardiac glycosides is mediated primarily by their neuroexcitatory effects on the atria and atrioventricular node, actions that would also favor the persistence of AF. Because cardiac glycosides have inotropic effects, it would be of interest to know whether these drags might have favorable effects on atrial stunning, and whether such effects relate to the duration of AF.
To assess the effects of cardiac glycosides on post-AF atrial stunning, a model of AF, rapid atrial electrical stimulation for 5 min, which has been shown to produce transient atrial contractile dysfunction,[15,20] was used, and aeetylstrophanthidin (AS), a cardiac glycoside with a time of action overlapping with the development of these abnormalities, was administered. Experiments were performed under neurally intact and autonomically blockaded conditions, and the effects of the glycoside were determined both with baseline conditions already influenced by AS and under conditions without use of the drug. The effects of AS on the duration of AF also were determined.
MATERIALS AND METHODS
Twenty-one conditioned mongrel dogs weighing between 23 and 25 kg were anesthetized with [Alpha]-chloralose, 80 mg/kg IV, after premeditation with morphine, 3 mg/kg IV. If required, supplemental [Alpha]-chloralose, in 10 mg/kg IV doses, was administered. Dogs were intubated and ventilated with a Harvard respirator, tidal volume and respiratory rate were adjusted, lungs were periodically expanded, and supplemental oxygen was administered to keep oxyhemoglobin saturation at [is greater than] 90%. Ringer's lactate (500 to 1,000 mL) supplemented with 20 mEq potassium chloride and 8 mEq magnesium sulfate were administered during preparatory surgery to mitigate the changes in hemodynamics and electrolytes. Body temperature was maintained by heating pads and was monitored by intravaseular thermistors. A right lateral thoracotomy was performed, and the heart was suspended in a pericardial cradle. Intravascular sheaths were positioned in the right femoral artery, right femoral vein, and right external jugular vein; they were used for blood sampling, drug infusions, and vascular accesses for catheters. In experiments that required autonomic blockade, cervical vagi were isolated in the neck.
Pressure and length gauge traces, electrocardiograms, and epicardial electrograms, at a frequency response of 30 to 250 Hz, were displayed and recorded on a multichannel oscillograph at paper speeds of 100 to 150 mm/s. A dual-channel, custom-designed, programmable stimulator (Bloom Associates; Denver, CO) was used to deliver rectangular pulses of 1-ms duration at 10-mA strengths (oscilloscopically calibrated) through an isolation transformer.
A quadripolar catheter was attached to the right atrial (RA) appendage with two fine sutures; the distal pair was used for electrical stimulation, and the proximal pair was used to record local electrical activity. Catheters had 1-mm long electrodes with a 1-mm separation. To maintain catheter patency, 2,000 U sodium heparin were administered IV at the beginning of an experiment.
Hemodynamics: Fluid-filled catheters were positioned in the left ventricle (LV) and RA, and pressures were measured with the manometer in the midchest taken as the zero reference. LV pressure, RA pressure (RAP), and the first derivative of these pressures (dp/dt) were obtained with catheter-tipped micromanometers (model PC-350; Millar Instruments; Houston, TX), which were calibrated in vivo using the fluid-filled catheters. Figure 1 shows a recording of these variables.
[Figure 1 ILLUSTRATION OMITTED]
Atrial Wall Motion: RA diameter was measured with a pair of ultrasonic external transducers (Sonometrics Corp; London, Ontario, Canada) using a sonomicrometer (model 1001 digital sonomicrometer; Sonometrics Corp). One transducer was sutured to the atrial surface on the groove between the aorta and the RA, and the other was sutured to the anterolateral wall. Figure 1 displays the RA maximal dimension (RAMD), RA end-diastolic dimension (RAEDD) (the maximal dimension just before the onset of atrial systolic shortening [SS]), and RA end-systolic dimension (RAESD) (the minimal dimension after atrial SS). Atrial SS is the difference between RAEDD and RAESD. Length measurements are expressed in millimeters. An atrial SS index (SSI) is defined as the SS percentage of RAEDD, which adjusts SS to an estimate of preload.
Hemodynamic and atrial length measurements were made before and after 5 min of rapid atrial stimulation (10 Hz; 10 mA; 1 ms) and after the conversion of induced AF. The duration of AF was noted, and recordings were obtained at 10, 150, and 300 s postconversion. In 11 dogs, following two inductions (separated by, and the second followed by, at least a 10-min recovery period subsequent to the cessation of AF), AS (Sigma; St. Louis, MO), 0.01 mg/kg IV bolus and 0.015 mg/kg/h IV infusion, were administered and rapid atrial stimulation was repeated. After 5 min of atrial stimulation and conversion of AF, the duration of AF was noted and recordings were repeated at the same intervals as those obtained before the administration of the glycoside. Ten minutes after cessation of AF (5 min after the 300-s measurement), a baseline measurement was made (the time at which the maximal effects of AS were evident), and 5 min of rapid atrial stimulation was again initiated. After the cessation of the pacing and conversion of AF, recordings were made at the previously noted intervals. In four dogs, the controls, the same protocol of four experiments was followed, but AS was not administered.
In six dogs, the protocol was replicated exactly, but experiments were performed in the presence of autonomic blockade. [Beta]-Blockade was produced by metoprolol, 0.20 mg/kg IV bolus and 0.40 mg/kg/h IV infusion. Parasympathetic blockade was produced by bilateral cervical vagotomy and atropine, 0.05 mg/kg IV. Experiments were begun 5 to 10 min after the autonomic blockade had been produced, at a time when heart rate and BP had been found to be stable for several minutes.
All data were entered and analyzed using computer software (SPSS/PC+; SPSS, Inc; Chicago, IL). The distribution of each variable was determined by the Kolmogorov-Smirnov test. Parametric variables were analyzed by analysis of variance (ANOVA) and the Student-Newman-Keuls test. Nonparametric variables were analyzed by the Mann-Whitney U test. Values are presented as mean [+ or -] SEM. Experiments, which had been done in pairs, were analyzed by ANOVA to ensure that findings were replicable; the average values of both experiments are presented except where otherwise noted. The null hypothesis was rejected at p = 0.05.
Neurally Intact Conditions
Following 5 min of atrial stimulation and an additional 102 [+ or -] 70 s of persistent AF, a brief period of RA hypercontractility ensued followed by a period of RA contractile dysfunction. The hypercontractile state was characterized by RA peak dp/dt exceeding the baseline value. The highest RA peak dp/dt occurred in the first or second beat after the cessation of AF and averaged 2.7 [+ or -] 0.2 times the baseline value. RA peak dp/dt then gradually declined, falling below baseline after 8.9 [+ or -] 1.2 beats, or 3.8 [+ or -] 0.7 s, and reaching its nadir, generally, after 10 s. Figure 2 illustrates this bimodal contractile response following 5 min of atrial stimulation with a persistent short episode of AF. Although RAMD and RAEDD dimensions tended to increase following AF (RAESD dimensions and SS were even more variable), the changes in RA peak dp/dt (both the initial increase and subsequent decrease) occurred independently of those changes.
[Figure 2 ILLUSTRATION OMITTED]
Table 1 and Figure 3 summarize the atrial findings, and Figure 4 provides a representative recording obtained at baseline and at 10 (lowest values), 150, and 300 s following AF. As shown, 5 min of rapid atrial stimulation producing AF resulted in a significant decrease in RA peak dp/dt, which was no longer different from baseline 2.5 min following the cessation of AF. RAP, RAMD, RAESD, SS, and SSI, were not, however, changed significantly by this perturbation. Heart rate, LV systolic pressure, LV end-diastolic pressure, and LV peak dp/dt also were not affected.
[Figures 3-4 ILLUSTRATION OMITTED]
Table 1--RA Dimensions and Function in Neurally Intact Dogs(*)
(*) Values given as mean [+ or -] SEM. RAA = RA A-wave pressure.
([dagger])p [is less than] 0.05 vs baseline value.
By contrast, as shown in Table 1 and Figure 3, when AS was administered, there was no decrease in RA peak dp/dt following AF. The absence of contractile dysfunction was observed both when baseline values had been obtained immediately before the administration of AS and when they were measured after this glycoside had already been administered and its maximal effects were evident. Figure 5 contrasts the effects of AF on atrial peak dp/dt when AS administration was begun 15 min before atrial stimulation was initiated with the findings in controls (experiments in which the same number of episodes of AF had preceded atrial stimulation but the cardiac glycoside was not given).
[Figure 5 ILLUSTRATION OMITTED]
Autonomic blockade reduced baseline values of RA peak dp/dt (34 [+ or -] 2.3 to 21.2 [+ or -] 0.7 mm Hg/s; p [is less than] 0.001) and LV peak dp/dt (1,535 [+ or -] 94 to 1,185 [+ or -] 64 mm Hg/s; p = 0.024) and increased baseline values of RAMD (13.9 [+ or -] 0.6 to 17.4 [+ or -] 0.8 mm; p [is less than] 0.001) and RAESD (9.8 [+ or -] 0.5 to 11.8 [+ or -] 0.4 mm; p = 0.005). Despite these changes, the contractile response to 5 min of atrial stimulation and an additional 3 [+ or -] 1 s of persistent AF was similar to that observed under neurally intact conditions: a brief period of RA hypercontractility followed by one of RA contractile dysfunction, albeit affecting more variables. After the cessation of AF, RA peak dp/dt was initially 2.8 [+ or -] 0.4 times the baseline value, but gradually declined, falling below baseline after 10.3 [+ or -] 1.8 beats, or 4.3 [+ or -] 0.5 s, and reaching its nadir, generally, after 10 s.
Table 2 and Figure 6 summarize the atrial findings at baseline and at 10, 150, and 300 s after AF. Even though baseline values of variables reflecting RA function were already depressed when compared to the neurally intact values, RA peak dp/dt (ANOVA, p [is less than] 0.013) and SSI (p [is less than] 0.042) decreased and RAESD (p [is less than] 0.015) increased following AF. RAP, RAMD, SS, heart rate, LV peak dp/dt, LVS, and LVED, however, did not change.
[Figure 6 ILLUSTRATION OMITTED]
Table 2--RA Dimensions and Function in Neurally Blockaded Dogs(*)
(*) Values given as mean [+ or -] SD. See Table 1 for abbreviations not in text.
([dagger])p [is less than] 0.05 vs baseline values.
([double dagger])p [is less than] 0.05 vs 10-s value.
As shown in Table 2 and Figure 6, in contrast to the findings without the glycoside, when AS was administered, there was no decrease in RA peak dp/dt following AF. Moreover, AS attenuated the increase in RAESD and the decrease in SSI, so that the changes following AF were no longer significant. The benefits of AS were evident whether or not baseline values had been augmented by the glycoside (ANOVA p [is greater than] 0.4 for RA peak dp/dt and SSI comparing the response to AF with baseline obtained without glycoside vs with baseline demonstrating maximal effect of glycoside).
As shown in Figure 7, the duration of AF induced at baseline was not different from that after the administration of AS. Autonomic blockade sharply reduced the duration of AF; there was, however, no additional change after the administration AS (Fig 7). Although the baseline duration (first two measurements) of AF in control dogs was the same as that found in the neurally intact dogs before the administration of the glycoside (102 [+ or -] 70 vs 106 [+ or -] 72 s), AF duration in the control dogs at a time comparable to that at which the glycoside had been administered was sharply lower (106 [+ or -] 72 vs 7 [+ or -] 3 s; p = 0.12 [Mann-Whitney U test]). Thus, the lack of a difference in AF duration following AS might actually indicate an enhancement of the arrhythmia by the glycoside.
[Figure 7 ILLUSTRATION OMITTED]
The present study demonstrates that the cardiac glycoside AS attenuates the atrial contractile dysfunction, the so-called stunning, that follows short-term AF. This favorable action of AS was observed in neurally intact and autonomically blockaded conditions, and it was found both when baseline atrial function was determined before the administration of the cardiac glycoside and when baseline function had been maximally enhanced by the drug. This effect of AS on atrial contractile function, however, was not associated with any favorable influence on AF itself; in fact, the findings suggest that the cardiac glycoside may actually promote AF even as it exerts a beneficial effect on atrial contractile function.
The model of AF that was used is similar to that of Leistad et al However, in addition to species differences (dogs vs pigs), there were other dissimilarities: higher intensity, albeit fewer, electrical stimuli were used to induce AF in our study. This was done to ensure that AF would be elicited with each experiment. However, higher intensity stimuli produce a greater release of neurohormones that promote AF. With repeated stimulations, the duration of AF tended to shorten, probably related to progressively less electrorelease of these hormones. The neural actions of AS appeared to offset this effect. Although the general findings of the models were similar, a brief hypercontractile period followed by one of reversible hypocontractility, the effects on various variables differed. We failed to observe a significant decrease in atrial SS with AF under neurally intact conditions, a paramount finding in the swine model, but we did find that this occurred with autonomic blockade. The difference was probably caused by a more intense sympathetic response in our study related to the use of [Alpha]-chloralose, an anesthetic that tends to maintain autonomic effects, and the preservation of atrioventricular conduction, which permitted greater systemic effects of AF.
AF is characterized by multiple changing reentrant circuits.[1-4] The arrhythmia is triggered by atrial extrasystoles and atrial tachycardias[30-32] and is sustained by a milieu of shortened and heterogeneous atrial refractoriness.[5,6] AF creates conditions that serve to perpetuate itself, so-called electrical remodeling.[7-13] The fast and irregular stimulations of AF produce marked[7,13] and disparate decreases in atrial refractoriness, the latter enhanced by the inherent differences of the refractory periods at various atrial sites. As our study and others have demonstrated, AF also produces contractile abnormalities, so-called stunning.[15-20] The longer that AF lasts, the longer the changes produced by the arrhythmia will persist.[7,11,14,19] As we have found in the anesthetized dog and as Daoud et al have observed in humans, following several minutes of AF, the hypocontractility that ensues resolves after a few minutes of sinus rhythm. Similarly, the shortening of atrial refractory periods that follows the cessation of AF or tachycardia lasting 10 min resolves also after only a few minutes. By contrast, following the cessation of AF that has been sustained by continuous rapid atrial pacing or following cardioversion of chronic AF in humans, the electrical remodeling and stunning of the atrium may take many days or even weeks to be reversed.
The changes in excitation and contraction that are produced by AF have been related to alterations at the cellular and molecular levels.[17,34] AF has been shown to produce a shortening of the duration of the atrial action potential and a depression of the atrial myocyte function. These changes have been related to a reduction of L-type calcium current and to reduced calcium systolic transients. By contrast, the L-type calcium channel blocker verapamil also has been found to attenuate post-AF electrical remodeling[1,10,12,14] and stunning.
In this context, the actions of AS in AF, and perhaps cardiac glycosides in general, can be examined. The acute neuroexcitatory action of AS, particularly its vagal effects, would explain a lack of benefit of the drag on paroxysmal AF. The vagus nerve shortens atrial refractoriness, increases the heterogeneity of atrial refractory periods, and sustains AF. Previously, we have found a lack of benefit of AS on the duration of AF induced by short bursts of rapid atrial stimulation. Those investigations also demonstrated a prolongation of AF at toxic dosages of AS, effects that can be explained by the intense neurohumoral actions including sympathetic effects, and by the delayed afterdepolarizations (oscillations) that occur with excessive amounts of cardiac glycosides.
Because atrial hypocontractility following AF is associated with reduced calcium transients, the direct cellular action of cardiac glycosides, which increases cytosolic calcium, could explain the favorable action of AS on atrial stunning. The post-AF abnormalities of atrial excitation and contractility, however, have been attributed to calcium overload engendered by frequent depolarizations. This explanation is based, at least in part, on the observation that verapamil antagonizes these changes.[8,10,12,14,16] In addition to its direct actions on L-type calcium channels, IV verapamil also increases plasma catecholamines.[39,40] Although the attenuation of post-AF electrical remodeling by verapamil has been observed in both neurally intact and autonomically blockaded conditions,[8,10,12,14] the beneficial effects of verapamil on post-AF contractile dysfunction have been reported either without blockade or with severe atrial contractile dysfunction. Thus, the actions of verapamil on atrial stunning may be related, at least in part, to the sympathomimetic effects evoked by the drug. Nevertheless, even if the detrimental effects of AF are due to calcium overloading, AS might still improve the contractile function of the stunned atrium in a fashion analogous to the effects of epinephrine and postextrasystolic potentiation on ischemic reperfusion injury, a condition also believed to be caused by calcium overload.
Recent investigations may provide insights into the, "calcium paradox" of AF-induced atrial stunning and into how cardiac glycosides might exert favorable effects on this phenomenon. An increased frequency of depolarizations produces an enhanced entry, of calcium into the cytosol, causing more calcium to be released by the sarcoplasm reticulum, increased calcium transients, and more forceful contractions.[42-44] Concomitantly, increased cytosolic calcium inhibits the inward L-type calcium channel. Following 1 min of rapid atrial stimulation, only hypercontractility is evident, as the contractile force gradually dissipates with the cessation of stimulation but never falls significantly below baseline. With 5 to 15 min of stimulation, the effects of increased cytosolic calcium will exert a favorable effect on contractile proteins yet still will inhibit calcium entry by way of the L-type calcium channel, producing a brief period of hypercontractility that is followed by one of hypocontractility. After longer periods of stimulation, a new steady state might be reached in which cytosolic calcium may even be normal because of a substantial inhibition of calcium entry through the L-type calcium channel. However, when stimulation is stopped, there would be less calcium entry, at least transiently, resulting in the attenuation of calcium release by the sarcoplasmic reticulum and reduced calcium transients, with only atrial hypocontractility evident. Because the enhanced entry of calcium by cardiac glycosides is not dependent on the L-type calcium channel (but is dependent on sodium-potassium adenosine triphosphatase inhibition and on the reverse sodium-calcium exchanger these drugs may ensure that adequate calcium is available to the contractile proteins, even though the L-type calcium channel has been inhibited. AS has, in fact, been found to have a favorable effect on the atrial interval-strength relationship, its actions being more intense when atrial inotropy is depressed. AF also may have adverse effects on atrial blood flow and energetics, which might contribute to atrial stunning but are not likely to be favorably influenced by cardiac glycosides.
Cardiac Glycosides in Paroxysmal AF
The findings in our study are consistent with previously reported clinical investigations that have assessed the effects of cardiac glycosides on paroxysmal AF.[20-24] The view that cardiac glycosides might be beneficial in the treatment of this condition has been based largely on uncontrolled observations, suggesting a high frequency of cardioversion following administration of IV digoxin. The administration of cardiac glycosides given for congestive heart failure also is believed to result in an increased conversion rate of AF to sinus rhythm. Randomized controlled studies, however, have demonstrated little benefit of these drugs for alleviation AF.[21,22,24] When compared to controls, neither IV nor oral cardiac glycosides have been found to increase the number of cardioversions or to shorten the time it takes for the cessation of AF. Moreover, on 24-h ambulatory ECG recordings of patients being monitored for or having demonstrated AF, those patients receiving digoxin had a comparable number of episodes of paroxysmal AF,[23,24] albeit more prolonged attacks, than those not receiving the drug. By contrast, in a randomized, double-blind, placebo-controlled, crossover designed study in which episodes of AF were reported with an event recorder during periods of treatment with digoxin, the interval between episodes of AF lengthened and the average ventricular rate of those events was 15 beats/min slower. Thus, the administration of IV or oral cardiac glycosides does not appear to be effective in converting or shortening episodes of AF,[21,22] findings that are consistent with the effects of AS on experimental AF observed in this study. When AF occurs in patients receiving digoxin, the duration of AF or the ventricular rate at onset also does not appear be improved.[23,24] However, in such patients the average ventricular rate of AF may be slowed (as a dromotropic effect of cardiac glycosides) and the interval between episodes of AF may be lengthened (perhaps in some way related to its antistunning actions).
Limitations and Clinical Implications
The relevance of the findings of this study to clinical AF in humans must consider the acute conditions of the model, species differences, and the manner of induction of AF, in contrast to spontaneous AF, as possible limitations. Also, adaptive changes to more prolonged AF and to cardiac glycosides administered over longer periods also might influence the findings. Our study does not address the possible beneficial effects of cardiac glycosides in reducing the triggers of AF, atrial extrasystoles and tachycardias, and it does not examine the benefits of cardiac glycosides on congestive heart failure that might prevent AF by reducing atrial size and improving neurohumoral conditions. However, to the extent that cardiac glycosides attenuate atrial stunning following the conversion of AF to sinus rhythm, they would enhance ventricular performance by improving atrial "booster" actions and possibly also would reduce the risk of stroke attributable to stunning. Nevertheless, only clinical investigations can determine whether the benefits of cardiac glycosides suggested by our study have importance in humans at risk for paroxysmal AF.
 Allessie MA, Lamers WJEP, Smeets JRLM, et al. Total mapping of atrial excitation during acetylcholine-induced atrial flutter and fibrillation in the isolated canine heart. In: Kulbertus HE, Olsson SB, Schlepper M, eds. Atrial fibrillation. Molndal, Sweden: Lindgren and Soner, 1982; 44-62
 Allessie MA, Lamers WJEP, Bonke FIM, et al. Experimental evaluation of Moe's multiple wavelet hypothesis of AF. In: Zipes DP, Jalife J, eds. Cardiac electrophysiology and arrhythmias. New York, NY: Grune and Stratton, 1985; 265-275
 Cox JL, Canavan TE, Schuessler RB, et al. The surgical treatment of atrial fibrillation: II. Intraoperative electrophysiologic mapping and description of the electrophysiologic basis of atrial flutter and atrial fibrillation. J Thorac Cardiovasc Surg 1991; 101:406-426
 Konings KTS, Kirchhof CJHJ, Smeets JRLM, et al. Hight density mapping of electrically induced atrial fibrillation in man. Circulation 1994; 89:1665-1680
 Moe GK. On the multiple wavelet hypothesis of atrial fibrillation. Arch Int Pharmacodyn Ther 1962; 140:183-188
 Allessie MA, Rensma PL, Lammers WJEP, et al. The role of refractoriness, conduction velocity, and wavelength in initiation of atrial fibrillation in normal conscious dogs. In: Attuel P, Coumel P, Janse M, eds. The atrium in health and disease. Mount Kisco, NY: Futura Publishing, 1989; 27-41
 Wijffels MCEF, Kirchof CJHJ, Dorland R, et al. Atrial fibrillation begets atrial fibrillation: a study in awake chronically instrumented goats. Circulation 1995; 92:1954-1968
 Tieleman RG, De Langen CDJ, Van Gelder IC, et al. Verapamil reduces tachycardia-induced remodeling of the atria. Circulation 1997; 95:1945-1953
 Wijffels MCEG, Kirchhof CJHJ, Dorland R, et al. Electrical remodeling clue to atrial fibrillation in chronically instrumented conscious goats: roles of neurohumoral changes, ischemia, atrial stretch, and high rate of electrical activation. Circulation 1997; 96:3810-3720
 Goette A, Honeycutt C, Langberg JJ. Electrical remodeling in atrial fibrillation: time course and mechanisms. Circulation 1996; 94:2968-2974
 Elvan A, Wylie K, Zipes DP. Pacing-induced chronic atrial fibrillation impairs sinus node function in dogs: electrophysiological remodeling. Circulation 1996; 94:2953-2960
 Daoud EM, Knight BP, Weiss R, et al. Effect of verapamil and procainamide on atrial fibrillation-induced electrical remodeling in humans. Circulation 1997; 96:1542-1550
 Gaspo R, Bosch RF, Talajic M, et al. Functional mechanisms underlying tachycardia-induced sustained atrial fibrillation in a chronic dog model. Circulation 1997; 96:4027-4035
 Yu W-C, Chen S-A, Lee S-H, et al. Tachycardia-induced change of atrial refractory period in humans: rate dependency and effects of antiarrhythmic drugs. Circulation 1998; 97: 2331-2337
 Leistad E, Christensen G, Ilebekk A. Atrial contractile performance after cessation of atrial fibrillation. Am J Physiol 1993; 264(suppl):H104-H109
 Leistad E, Aksnes G, Verburg E, et al. Atrial contractile dysfunction after short-term atrial fibrillation is reduced by verapamil but increased by BAY K8644. Circulation 1996; 93:1747-1754
 Sun H, Gaspo R, Leblanc N, et al. Cellular mechanisms of atrial contractile dysfunction caused by sustained atrial tachycardia. Circulation 1998; 98:719-727
 Shapiro EP, Effron MB, Lima S, et al. Transient atrial dysfunction after conversion of chronic atrial fibrillation to sinus rhythm. Am J Cardiol 1988; 62:1202-1207
 Manning WJ, Leeman DE, Gotch PJ, et al. Pulsed Doppler evaluation of atrial mechanical function after electrical cardioversion of atrial fibrillation. J Am Coll Cardiol 1989; 13:617-623
 Daoud EG, Marcovitz P, Knight BP, et al. Short-term effect of atrial fibrillation on atrial contractile function in humans. Circulation 1999; 99:3024-3027
 Falk RH, Knowlton AA, Bernard SA, et al. Digoxin for converting recent-onset atrial fibrillation to sinus rhythm: a randomized, double-blind trial. Ann Intern Med 1987; 106: 503-506
 Jordaens L, Trouerbach J, Calle P, et al. Conversion of atrial fibrillation to sinus rhythm and rate control by digoxin in comparison to placebo. Eur Heart J 1997; 18:643-648
 Rawles JM, Metcalfe MJ, Jennings K. Time of occurrence, duration and ventricular rate of paroxysmal atrial fibrillation: the effect of digoxin. Br Heart J 1990; 63:225-227
 Murgatroyd FD, Gibson SM, Baiyan X,. et al, for the Controlled Randomized Atrial Fibrillation Trial (CRAFT) Investigators. Double-blind placebo-controlled trial of digoxin in symptomatic paroxysmal atrial fibrillation. Circulation 1999; 99:2765-2770
 Beasley R, Smith DA, McHaffie DJ. Exercise heart rates at different serum digoxin concentrations in patients with atrial fibrillation. BMJ 1985; 290:9-11
 Watanabe AM. Digitalis and the autonomic nervous system. J Am Coll Cardiol 1985; 5(suppl):35A-42A
 Mokraoui AM, Friedman HS, Melniker LA, et al. Effects of acetylstrophanthidin on duration of atrial fibrillation in the neurally-intact and blockaded dog. Cardiovasc Drugs Ther 1988; 2:569-577
 Friedman HS, Wattanasuwan N, Sharafkhaneh A, et al. The comparative effects of drive and test stimulus intensity on myocardial excitability and vulnerability. Pacing Clin Electrophysiol 2000; 23:84-95
 Mendez R, Mendez C. The action of cardiac glycosides on the refractory periods of heart tissues. J Pharmacol Exp Ther 1953; 107:24-36
 Killip T, Gault JH. Mode of onset of atrial fibrillation in man. Am Heart J 1965; 70:172-179
 Hnatkova K, Waktare JEP, Murgatroyd FD, et al. Analysis of the cardiac rhythm preceding episodes of paroxysmal atrial fibrillation. Am Heart J 1998; 135:1010-1019
 Haissaguerre M, Jais P, Shah DC, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 1998; 339:659-666
 Feng J, Yue L, Wang Z, et al. Ionic mechanisms of regional action potential heterogeneity in the canine right atrium. Circ Res 1998; 83:541-551
 Yue L, Feng J, Gaspo R, et al. Ionic remodeling underlying action potential changes in a canine model of atrial fibrillation. Circ Res 1997; 81:512-525
 Allesie R, Nusynowitz M, Abildskov JA, et al. Nonuniform distribution of vagal effects on the atrial refractory period. Am J Physiol 1958; 194:406-410
 Farah A, Loomis TA. The action of cardiac glycosides on experimental flutter. Circulation 1950; 2:742-748
 Rosen MR. Cellular electrophysiology of digitalis toxicity. J Am Coll Cardiol 1985; 5(suppl):22A-34A
 Fozzard HA, Sheets MF. Cellular mechanism of action of cardiac glycosides. J Am Coll Cardiol 1985; 5(suppl):10A-15A
 Imai K, Higashidate S, Prados PR, et al. Relation of blood pressure and catecholamine concentration after administration of calcium antagonists to rats. Biol Pharm Bull 1994; 17:907-910
 Friedman HS, Rodney E, Sinha B, et al. Verapamil prolongs atrial fibrillation by evoking an intense sympathetic neurohumoral effect. J Investig Med 1999; 99:47:293-303
 Becker LC, Levine JH, DiPaula AF, et al. Reversal of dysfunction in postischemic stunned myocardium by epinephrine and postextrasystolic potentiation. J Am Coll Cardiol 1986; 7:580-589
 Schouten VJA, Morad M. Regulation of [Ca.sup.2+] current in frog ventricular myocytes by the holding potential, c-AMP and frequency. Pflugers Arch 1989; 415:1-11
 Lee H-C, Clusin WT. Cytosolic calcium staircase in myocardial cells. Circ Res 1987; 61:934-939
 Wier WG, Yue DT. Intracellular calcium transients underlying the short-term force-interval relationship in ferret ventricular myocardium. J Physiol 1986; 376:507-530
 Sun H, Leblanc N, Nattel S. Mechanisms of inactivation of L-type calcium channels in human atrial myocytes. Am J Physiol 1997; 41:1625-1635
 Katz AM. Effects of digitalis on cell biochemistry: sodium pump inhibition. J Am Coll Cardiol 1985; 5(suppl):16A-21A
 Koch-Weser J, Blinks JR. Analysis of the relation of the positive inotropic action of cardiac glycosides to the frequency of contraction of heart muscle. J Pharmacol 1962; 136:109-122
 White CW, Holida MD, Marcus ML. The effects of acute atrial fibrillation on the vasodilator reserve of the canine atrium. Cardiovasc Res 1986; 20:683-689
 Kitakaze M, Weisman HF, Marban E. Contractile dysfunction and ATP depletion after transient calcium overload in perfused ferret hearts. Circulation 1988; 77:689-695
 Weiner P, Bassan MM, Jarchovsky J, et al. Clinical course of acute atrial fibrillation treated with rapid digitalization. Am Heart J 1983; 105:223-227
 Jennings PB, Makous N, Vander Veer JB. Reversion of atrial fibrillation to sinus rhythm with digitalis therapy. Am J Med Sci 1958; 235:702-706
 Burchell HB. A clinical appraisal of atrial transport function. Lancet 1964; 1:775-779
 Fatkin D, Kuchar DL, Thorburn CW, et al. Transesophageal echocardiography before and during current cardioversion of atrial fibrillation: evidence for "atrial stunning" as a mechanism of thromboembolic complications. J Am Coll Cardiol 1994; 23:307-316
(*) From the Department of Medicine, Long Island College Hospital and SUNY Health Science Center, Brooklyn, NY.
Manuscript received November 11, 1999; revision accepted March 22, 2000.
Correspondence to: Howard S. Friedman, MD, FCCP, 650 First Ave, 3rd Floor, New York, NY 10016; e-mail: HSF2727@ POL.NET
COPYRIGHT 2000 American College of Chest Physicians
COPYRIGHT 2001 Gale Group