Author + information
- Received January 15, 1999
- Revision received August 11, 1999
- Accepted September 10, 1999
- Published online January 1, 2000.
- Kaori Shinagawa, MDa,
- Hideo Mitamura, MDa,* (, )
- Akiko Takeshita, MDa,
- Toshiaki Sato, MDa,
- Hideaki Kanki, MDa,
- Seiji Takatsuki, MDa and
- Satoshi Ogawa, MDa
- ↵*Reprint requests and correspondence: Dr. Hideo Mitamura, Cardiopulmonary Division, Department of Medicine, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan
The purposes of this study were to measure the atrial refractory period and the conduction velocity (CV) during atrial fibrillation (AF) and to explore the antiarrhythmic mechanism of a sodium channel blocker, pilsicainide, during AF.
Sodium channel blockers not only decrease the CV, but also prolong the atrial refractory period, particularly during rapid excitation. Because these effects on the wavelength are counteractive and rate dependent, it is critical to measure these parameters during AF.
In eight dogs, after AF was induced under vagal stimulation, a single extra-stimulus was repeatedly introduced from the left atrium and its capture was statistically determined for each coupling interval. The local CV was also measured during constant capture of the fibrillating atrium by rapid pacing. The same procedure was repeated after pilsicainide administration.
Pilsicainide significantly increased the mode of AF intervals from 81 ± 10 to 107 ± 16 ms (p < 0.01). While the CV was decreased from 0.9 ± 0.1 to 0.7 ± 0.1 m/s (p < 0.02), the effective refractory period during AF was increased from 69 ± 11 ms to 99 ± 17 ms (p < 0.01). As a result, the wavelength was significantly increased by pilsicainide from 6.6 ± 0.9 to 7.6 ± 1.2 cm (p < 0.05).
During AF, whereas the sodium channel blocker pilsicainide decreases CV, it lengthens the wavelength by increasing the refractory period, an action that is likely to contribute to the drug’s ability to terminate the arrhythmia. The direct measurement of refractoriness and CV during AF may provide new insights into the determinations of the arrhythmia and antiarrhythmic drug action.
Although atrial fibrillation (AF) is one of the most common arrhythmias encountered in clinical practice (1), the underlying mechanisms of AF and antiarrhythmic drug actions remain incompletely understood. Based on animal experiments and computer simulations, AF is considered to be maintained by multiple independent wavelets irregularly activating atrial cells at very high rates (2,3). It was found that the maintenance of AF depends on the presence of a number of simultaneous reentering waves. The minimum size of a reentrant wave is related to the wavelength, defined as the product of the local refractory period and the local conduction velocity (CV) (i.e., the distance the wave front travels in one refractory period) (4). Modification of the wavelength is supposed to affect the inducibility and perpetuation of AF. Drugs that prolong the refractory period are expected to prolong the wavelength, whereas those that decrease CV are expected to exert opposite effects.
While sodium channel blocking agents with slow recovery kinetics have been shown to be effective in terminating AF (5), their mechanisms of action are poorly understood. Sodium channel blockers characteristically slow the CV. Given the classically understood determinants of reentry, slowed conduction would decrease the size of the reentrant circuit and increase the number of circuits, increasing thereby the likelihood of AF (6). On the other hand, some class Ic antiarrhythmic drugs have been shown to produce a tachycardia-dependent increase in refractoriness (7,8). Using pilsicainide, a class Ic antiarrhythmic drug (9,10), we have previously shown in a dog model that refractory periods can be prolonged in a use-dependent fashion (8). Therefore, the effects of sodium channel blockers on the wavelength could be counteractive, thus creating a dilemma in the treatment of AF.
Because the effects of class Ic drugs on the CV and the refractory period are rate dependent, the product of these two indexes, that is, the wavelength, should also be rate dependent. However, in previous studies, CV and refractory period were not measured directly during AF. To evaluate the net effect, it is critical to measure these variables during AF. In this experimental study, to explore the mechanism of the antifibrillatory effect of a sodium channel blocker, we attempted to measure the refractory period, CV and wavelength during AF before and after the administration of pilsicainide in a canine vagotonic AF model.
The study protocol was approved by the institutional scientific review committee and conformed to the “Position of the American Heart Association on Research Animal Use” adopted by the Association in November 1984.
Eight adult dogs of either gender weighing between 13.5 and 22.4 kg were anesthetized with ketamine chloride (15 mg/kg) and pentobarbital sodium (20 mg/kg IV), and ventilated with room air. Respiratory parameters were adjusted to maintain physiological levels of arterial blood gases (SaO2>90%). Catheters were inserted into the right femoral artery and the left femoral vein. Through a left intercostal thoracotomy, the pericardium was opened, and the heart was exposed. A rectangular thin plastic sheet containing 96 bipolar electrodes with 1-mm interpolar and 2-mm interelectrode distances was placed on the lateral wall of the left atrium for simultaneous recording of multiple atrial electrograms. A pair of electrodes with 1-mm interpolar distance in the center of the mapping electrode served for pacing using stimulators (Cardiac Stimulator BC-02A; Fukuda Denshi, Tokyo, Japan; Pulse generator 88-1346; Nihon Koden, Tokyo, Japan; and Isolator ss-202J; Nihon Koden), which delivered symmetric biphasic rectangular pulses of 2-ms duration at four times threshold current. Nihon Koden Window Slicer was used to sense bipolar electrograms of 0.75 to 2.5 mV recorded by a bipolar electrode adjacent to the stimulation electrodes. An additional electrode placed next to these electrodes was used for recording unipolar atrial electrograms filtered at 1 to 600 Hz, from which local activation intervals were measured. The arrangement of these electrodes is shown in Figure 1.
A transducer (TP-400-T; Nihon Koden), a direct current coupled isolated preamplifier (frequency range of 0.04 to 500 Hz) and a thermal array recorder (Polygraph system RM6000; Nihon Koden) were used to monitor and record standard surface electrocardiogram (ECG), arterial blood pressure, atrial electrograms and stimulus signals. Activation mapping signals were digitized with 12-bit resolution and a 1-kHz sampling rate, and transmitted to a microcomputer (HPM-7100, time constant 3.3 ms; Fukuda Denshi) for construction of isochrone maps.
The cervical vagal trunks were isolated, and bipolar Teflon-coated stainless steel electrodes were inserted into the center of each nerve via a 22-gauge needle. Bilateral vagal stimulation was delivered by a pulse generator (88-1346; Nihon Koden) with 4-ms pulses of 4 to 6 V and a stimulation frequency of 400/min (6.7 Hz). The strength of vagal stimulation was adjusted in each dog to maintain the sinus rate at about 60 per minute.
Measurement of the atrial effective refractory period by the standard extra-stimulus method
The atrial effective refractory period in the left free wall was assessed at basic cycle lengths of 600, 400, 300, 200 and 150 ms under vagal stimulation. The heart was paced with a train of eight basic (S1) stimuli followed by a premature (S2) stimulus with the coupling interval decreased in 2-ms steps until capture no longer occurred. All basic and premature stimulation was performed using square impulses 2 ms in duration and an intensity of four times the excitability threshold. The effective refractory period was defined as the longest S1-S2 interval failing to produce a propagated response.
Induction of AF
After obtaining baseline electrophysiologic data, AF was induced by burst pacing under bilateral vagal stimulation. Vagotonic AF was induced according to the method described by Wang et al. (9). Under vagal stimulation, a short burst (3 s) of atrial pacing using pulses 2 ms in duration and an intensity four times the excitability threshold at 60-ms interpulse intervals reproducibly induced AF, which persisted for more than 30 min and was terminated within seconds of discontinuing vagal stimulation. Atrial fibrillation was defined as a rapid (>500/min under baseline conditions), irregular atrial rhythm with a varying atrial electrogram morphology.
Measurements of refractory periods during AF
During AF, a single extra-stimulus was introduced from the left atrium at a predetermined coupling interval (CI) after the index local activation recorded and sensed by the adjacent bipolar electrode. A single extra-stimulus was introduced after every 6 to 10 signals of atrial activation using the Pulse Train Generator (Nihon Koden). After 100 single stimuli were applied, the coupling interval of the extra-stimulus was decreased by 5 ms. The atrial activation during AF used for sensing was defined as F0, from which a CI of single extra-stimulus was measured. The atrial activation after F0 was defined as F1. The intervals between F0 and F1 were manually measured 100 times on the unipolar atrial electrograms recorded 1 mm apart from the stimulation electrodes, using a caliper. An example of the recordings used for measurement of intervals during AF is shown in Figure 2. A histogram of 100 F0F1 intervals for each coupling interval was constructed, and was compared with a control histogram consisting of 100 FF intervals of baseline AF (Fig. 3). Capture of the atrium by a stimulus resulted in loss of long F0F1 intervals. When the distribution of a histogram was statistically different from that of the control histogram, capture was considered to be present at a predetermined coupling interval (Fig. 4). The shortest coupling interval at which the atrium was captured during AF was taken as the functional refractory period during AF (FRPAF). The longest coupling interval at which the atrium was not captured during AF was taken as the effective refractory period during AF (ERPAF). Average (Avg FF) and mode (Mode FF) of 100 FF intervals were also obtained during AF. The capture window was then calculated by subtracting FRPAF from Mode FF.
CV and wavelength
Because wave fronts during AF rarely propagate parallel to rows of mapping electrodes, it is difficult to obtain CV during AF. We considered that the CV during constant capture of the fibrillating atrium might represent a better practical approximation than that obtained during sinus rhythm. Hence, the local CV was measured based on the activation mapping of 96 atrial sites during constant capture of the fibrillating atrium by rapid pacing in seven dogs (Fig. 5). Rapid pacing in the center of the mapping electrode at intervals slightly longer than ERPAF resulted in continuous regional captures of AF. Due to technical difficulties, the local CV was not measured in one dog (data excluded from Table 1). The wavelength was calculated as the product of FRPAF and the CV.
After obtaining baseline electrophysiologic data, a loading dose (0.3 mg/kg) of pilsicainide was administered intravenously over 3 min, followed by a maintenance dose (0.03 mg/kg/min) administered by infusion. Five minutes after starting the infusion of pilsicainide, the baseline measurements were repeated. Plasma drug concentrations were measured at the beginning and at the end of the measurements. Then, vagal stimulation was discontinued and AF was allowed to terminate within seconds. After AF was terminated, the atrial effective refractory period during sinus rhythm was measured at various basic cycle lengths using the standard extra-stimulus method under vagal stimulation at the same strength as before.
Values are presented as the mean ± SD. Using Mann-Whitney Utest, the histogram of each corresponding coupling interval was compared with a control histogram. Comparisons between two means were made by Student ttest. Comparisons of effective refractory periods at multiple cycle lengths and ERPAF were performed by analysis of variance followed by the Bonferroni method. A p value of <0.05 was considered statistically significant.
Measurements during baseline AF
The refractory periods during AF could be measured in all eight dogs. During baseline AF, the Avg FF, Mode FF, FRPAF, ERPAF and capture window were 82 ± 10, 81 ± 10, 74 ± 11, 69 ± 11 and 7 ± 2 ms, respectively (Table 1). The capture window was thus shown to exist and be measurable in vagotonic AF in these eight dogs. In seven dogs, the mean CV during AF was 0.9 ± 0.1 m/s, and the calculated wavelength was 6.6 ± 0.9 cm.
Effects of pilsicainide on the electrophysiologic parameters
The effects of pilsicainide on electrophysiologic parameters are also shown in Table 1. Pilsicainide significantly increased Avg FF by 26 ms (p < 0.01), and Mode FF by 26 ms (p < 0.01). Changes in FRPAF and capture window induced by pilsicainide in each dog are shown in Figure 6. The increase of FRPAF was greater than that of Mode FF. As a result, the capture window during pilsicainide administration was shortened significantly to 3 ± 3 ms from 7 ± 2 ms at baseline (p < 0.02). Although the CV decreased by 0.2 ± 0.1 m/s (p < 0.02) during pilsicainide administration, because ERPAF significantly increased by 30 ms from 69 ± 11 ms at baseline to 99 ± 17 ms during drug infusion (p < 0.01), the wavelength was actually increased by pilsicainide by 1.0 cm from 6.6 ± 0.9 cm to 7.6 ± 1.2 cm (p < 0.05).
The values of the effective refractory periods during sinus rhythm determined by the standard extra-stimulus method at multiple cycle lengths and the ERPAF under baseline conditions were compared with their values during pilsicainide administration (Fig. 7). Under baseline conditions, the effective refractory periods during pacing at basic cycle lengths of 150, 200, 300, 400 and 600 ms were 76.3 ± 12.8, 83.8 ± 14.9, 88.3 ± 16.2, 95.3 ± 13.9 and 98.8 ± 13.5 ms, respectively. During baseline AF, ERPAF was 69 ± 11 ms, which was even shorter than the shortest effective refractory period determined by the standard extrastimulus method before induction of AF (basic cycle length 150 ms) (p < 0.05). During pilsicainide administration, the effective refractory periods during pacing at basic cycle lengths of 150, 200, 300, 400 and 600 ms and ERPAF were prolonged to 98.3 ± 5.5, 101.5 ± 4.2, 104.3 ± 9.3, 105.0 ± 8.7, 103.5 ± 9.3 and 99 ± 17 ms, respectively. It is important to note that during pilsicainide administration, the effective refractory period increased particularly at shorter cycle lengths, and the increase was most prominent during AF.
The mean durations required for these measurements under baseline conditions and after pilsicainide administration were 17.3 ± 3.6 and 13.4 ± 2.4 min, respectively. The mean plasma concentrations of pilsicainide at the beginning and at the end of the measurements were 1.90 ± 0.67 and 2.16 ± 0.79 μg/ml, respectively (mean ± SD).
This study has demonstrated that the effective refractory period during AF could be measured by repeated single extra-stimulus method. By combining this measurement of the effective refractory period and that of CV during constant atrial capture, for the first time the wavelength during AF was obtained. Using this method, more detailed assessment of the antifibrillatory effect of drugs became possible. Although a sodium channel blocker, pilsicainide, decreased the CV during vagotonic AF, it lengthened the wavelength by increasing the effective refractory period. It was suggested that the ability of pilsicainide to increase wavelength during AF likely contributes to the drug’s ability to terminate AF.
Atrial capture during AF
The presence of atrial capture during AF has been demonstrated in experimental AF as well as paroxysmal and chronic AF in humans (11–15). Our present study is the first to demonstrate atrial capture during vagotonic AF in dogs.
For the genesis of atrial capture during AF, three mechanisms have been recognized: a gross anatomic obstacle in the reentrant pathway, random reentry and the anisotropic conduction property of atrial myocardium (16). At present, it is unknown to what extent these three possible mechanisms contribute to the creation of an atrial capture during AF. Atrial fibrillation might be perpetuated by a mixture of leading circle reentry, random reentry and macro-anatomical reentry (17,18). If AF arises from different causes (chronic AF of organic heart disease, paroxysmal AF without structural heart disease, vagotonic AF, etc), each mechanism of reentry may, in different ways, contribute to the maintenance of AF. Characteristically, vagal stimulation facilitates atrial reentry by shortening the effective refractory period and creating inhomogeneities in refractoriness without affecting CV (19,20). In our vagotonic AF model, AF is considered to be perpetuated mainly by functionally determined reentry. Therefore, atrial capture demonstrated might be created by a mechanism based on random reentry with the cells not immediately activated by one of the wavelets after recovery of excitability. In this case, the local cycle length will be determined by the local refractory period plus the time the cells have to wait until they are excited by the next fibrillation wave.
Direct measurement of the effective refractory period during AF
Quantitative analysis of atrial capture during AF gives the effective refractory period during AF, which is a vulnerable parameter for the maintenance and termination of AF. To investigate the antifibrillatory mechanisms of antiarrhythmic agents, it is important to measure the effective refractory period directly during AF before and after the drug administration. However, in all the previous studies on the antifibrillatory mechanisms of drugs, the effective refractory period was only obtained by the standard extrastimulus method in the absence of AF. It is reasonable to suspect that the effective refractory period during AF could be different from that obtained during sinus rhythm by the standard extra-stimulus method. Because the action potential duration can be affected by the frequency of activation as well as by the autonomic tone, the effective refractory period could be much shorter during AF when atrial activation is very rapid and the sympathetic tone is heightened. In addition, as discussed later, because sodium channel blockers have use-dependent blocking property, the effect of these drugs on electrophysiological parameters may differ when the excitation is extremely rapid, such as during AF.
Another approach to obtain the effective refractory period during AF uses the entrainment technique by applying rapid atrial pacing with a constant capture. By shortening pacing cycle lengths, one may get a critical cycle length failing a constant capture, which is supposed to represent the effective refractory period. However, rapid pacing during AF may cause further rate-dependent shortening of the refractory period (14). Kirchhof et al. (21)demonstrated that very rapid continuous pacing could cause acceleration of AF by inducing local intraatrial reentry circuits with a revolution time shorter than the pacing interval. Therefore, the effective refractory period obtained using this technique may not necessarily be equal to the effective refractory period inherently present during AF.
To obtain the effective refractory period during AF in a more rational way, we adopted the single extra-stimulus method first described by Mast et al. (22). Considerable beat-to-beat variation in the local fibrillation interval is well recognized during AF and may potentially be due to temporal variations in the duration of local refractoriness (16). The duration of the refractory period depends on multiple factors such as the direction of impulse propagation (anisotropy), the length of the previous interval and electrotonic modulation of the action potential by neighboring wavelets or local intraatrial conduction block (16). Because AF is composed of multiple random reentries, capture or uncapture of a single extra-stimulus would not give the accurate refractory period during AF. Therefore, the introduction of the extra-stimulus at the same coupling interval was repeated 100 times, and the presence or absence of atrial capture was statistically determined.
Comparison of refractory periods between measurements during AF and during sinus rhythm
In many previous studies, the effective refractory period during AF was considered to be close to the effective refractory period determined by the standard extra-stimulus method at a short basic cycle length. In our study, the effective refractory period during AF at baseline was shorter than the effective refractory period determined by the standard extrastimulus method (basic cycle length 150 ms), reflecting the fact that during AF atrial fibers are activated at cycle lengths shorter than 150 ms. This observation supports the importance of measuring refractory period during AF.
Measurement of the CV during AF
Because wave fronts during AF rarely propagate parallel to rows of mapping electrodes, it is difficult to obtain CV during AF. Although it is desirable, it is practically difficult to obtain CV based on captured atrial activation by a single extra-stimulus during AF. Therefore, we decided to measure CV by entraining the atrium during AF. We assume that CV during constant capture of the fibrillating atrium would represent a reasonable practical approximation of that during spontaneous AF.
Application of the new method to evaluating anti-AF effects of sodium channel blockers
One of the most useful applications of this method would be the determination of the wavelength during AF after the administration of sodium channel blockers. Several studies have demonstrated that sodium channel blockers effectively convert clinical AF (5). The antifibrillatory mechanisms of these agents have not been elucidated but appear to be related to prolongation of the wavelength for reentry at rapid rates, despite the fact that these agents are traditionally known to slow conduction (7,23–27). However, almost all previous workers assumed that the wavelength obtained in a nonfibrillatory state at a short pacing cycle length would be close to the wavelength during AF, but did not directly measure the CV nor the effective refractory period during AF. Recently, Wijffels et al. (28)measured the effects of sodium and potassium channel blockers on these parameters during AF in a goat model of chronic AF. Contrary to our observation, they found that the excitable gap was widened with these agents. There are several potential reasons to account for this disagreement. Not only are there species differences, but the model used is also different, such that they used a relatively chronic AF model with advanced electrical as well as structural remodeling, while we used an acute vagotonic AF model. Drugs are different, and the doses may not be comparable, which may also explain the discordance.
Pilsicainide is a pure sodium channel blocker with slow recovery kinetics (27.3 to 28.2 s) (9,10). Although at extremely high doses pilsicainide has been reported to show a vagolytic effect in vitro, the plasma concentration of pilsicainide in our dogs is not considered to exert any vagolytic effects (8). In a clinical trial called PSTAF, the efficacy of a single oral dose of pilsicainide was evaluated (29). This trial demonstrated that conversion of AF to sinus rhythm was achieved within 90 min in 45% of 40 patients receiving pilsicainide, and in 8.6% of 35 patients on placebo (p < 0.01).
Our group has previously demonstrated that pilsicainide successfully terminated vagotonic AF in seven of eight dogs (8). Using monophasic action potential recording in the atrium, we also showed that although pilsicainide did not affect monophasic action potential duration, it caused significant use-dependent prolongation of the atrial effective refractory period, that is, atrial refractoriness extending beyond monophasic action potential duration (postrepolarization refractoriness) (30).
In the present experimental study, using a lower dose of pilsicainide than the dose used in the previous study to avoid premature termination of AF, we demonstrated that the effective refractory period was increased to a greater degree during AF than during rapid pacing (basic cycle length 150 ms), reflecting the use-dependent blocking nature of this drug. Of particular importance in this study was that although pilsicainide slowed atrial conduction during AF, it caused even greater increases in refractoriness. As a result, the wavelength increased during pilsicainide administration compared with their baseline values. Because AF contains multiple reentrant wavelets without a fixed anatomical path, wavelength obtained from this study does not necessarily represent a part of the exact fibrillation path. Nevertheless, the results of this study using a new method to measure electrophysiological parameters during AF suggest that the anti-AF effect of sodium channel blockers may be closely related to the prolongation of refractoriness, rather than to slowing of the impulse conduction.
One of limitations of the study is a relatively low resolution of the current technique in measuring the refractory period. Because repetition of introduction of extra-stimuli 100 times for each coupling interval takes a considerable amount of time, we judged that it would be impractical to measure refractory period during AF by shortening the coupling interval by a smaller than 5-ms step. Similarly, the method of measuring CV during entrainment of the fibrillating atrium may have a limitation, because this is a CV of paced stimuli rather than that of spontaneous atrial activation. Nevertheless, we believe these techniques would provide the best practical approximation of the parameters during AF.
Another limitation of the study is that only the electrograms from a localized part of the left atrium were analyzed. Therefore, the response of atrial tissue in other areas of the left atrium and the right atrium to pilsicainide is unknown. It has been agreed that the atrium is far from a homogeneous milieu and shows a considerable degree of spatial dispersion in the refractory period, excitability and CV (21). Recent experiments suggest an important role of heterogeneity of refractoriness in determining AF (31). Antiarrhythmic drugs may terminate AF not only by prolonging atrial refractoriness and wavelength, but also by making refractory properties more homogeneous across the atria. Although we have previously demonstrated a beneficial effect of pilsicainide in this regard using FF measurements at multiple atrial sites (32), direct determination of the effective refractory period and CV at multiple sites, although technically difficult, would be required to clarify the more general effects of pilsicainide.
Another important limitation of the present study is related to the model we used. Because vagotonic AF could have an electrophysiological milieu different from spontaneously occurring arrhythmia, the findings of this study are specific to vagotonic AF and may not directly apply to patients exhibiting spontaneous chronic or paroxysmal AF. In our canine model, the atrial muscle is considered to be essentially normal, so the effect of pilsicainide shown in this study cannot be extrapolated to pathological atria in patients with chronic remodeled AF with a dilated atrium, where conduction is disturbed in the baseline state and can potentially be slowed to a greater degree by sodium channel blockade. Further observations in different experimental AF models, along with studies in clinical AF, would be of interest.
Measurement of the refractory period during AF by repeated introduction of a single extra-stimulus was feasible in a canine vagal stimulation model. This study demonstrates for the first time that during AF, although a sodium channel blocker decreases CV, it can lengthen the wavelength by increasing the refractory period with abbreviating capture window of the atrium, which may contribute to the termination of AF.
This study was presented, in part, at the 19th Annual Scientific Sessions of the North American Society of Pacing and Electrophysiology, San Diego, May 1998.
- atrial fibrillation
- Avg FF
- average of 100 atrial activation (FF) intervals
- coupling interval
- conduction velocity
- effective refractory period during atrial fibrillation
- functional refractory period during atrial fibrillation
- Mode FF
- mode of 100 FF intervals
- Received January 15, 1999.
- Revision received August 11, 1999.
- Accepted September 10, 1999.
- American College of Cardiology
- Kannel W.E,
- Abbott R.D,
- Savage D.D,
- McNamara P.M
- Moe G.K
- Allessie M.A,
- Lammers W.J.E.P,
- Bonke F.I.M,
- Hollen J
- Wang Z,
- Page P,
- Nattel S
- Inomata N,
- Ishihara T,
- Akaike T
- Allessie M.A,
- Kirchhof C,
- Scheffer G.J,
- et al.
- Capucci A,
- Biffi M,
- Boriani G,
- et al.
- Daoud E.G,
- Pariseau B,
- Niebauer M,
- et al.
- Pandozi C,
- Biancomi L,
- Villani M,
- et al.
- Konings K.T.S,
- Kirchhof C.J.H.J,
- Smeets J.R.L.M,
- et al.
- Zipes D.P,
- Mihalick M.J,
- Robbins G.T
- Schuessler R.B,
- Rosenshtraukh L.V,
- Boineau J.P,
- et al.
- Kirchhof C,
- Chorro F,
- Scheffer G.J,
- et al.
- Mast F,
- Killian M.J,
- Wijffels M.C,
- et al.
- Nattel S,
- Wang Z,
- Pelletier L.C,
- et al.
- Wang J,
- Bourne G.W,
- Wang Z,
- et al.
- Wijffels M,
- Dorland R,
- Killian M,
- et al.
- Whalley D.W,
- Grant A.O
- Liu L,
- Nattel S
- Shinagawa K,
- Mitamura H,
- Sato T,
- et al.