Reproducible Induction of Early Afterdepolarizations and Torsade de Pointes Arrhythmias by d-Sotalol and Pacing in Dogs With Chronic Atrioventricular Block
Abstract It has been well established that antiarrhythmic drugs can also have proarrhythmic effects such as torsade de pointes (TdP) arrhythmias. It was the purpose of this study to create an animal model with a high incidence of reproducible TdP that occurs under clinically relevant circumstances. Experiments were performed in anesthetized dogs that had been in chronic atrioventricular block for 9±6 weeks. TdP inducibility was attempted using different pacing modes before and after the administration of 2 mg/kg d-sotalol. In some experiments, endocardial monophasic action potentials were recorded. d-Sotalol increased the cycle length of the idioventricular rhythm (1475±460 to 1730±570 ms, P<.01) and the QT time (390±65 to 480±85 ms, P<.01). In 10% of the experiments, spontaneous TdP occurred after d-sotalol. The incidence of pacing-dependent TdP was 52% (P<.01). In the inducible group, the cycle length of idioventricular rhythm and QT time were significantly longer despite equal percentage increases in these parameters after d-sotalol in both groups. The pacing modes consisting of more than one frequency change had a higher TdP induction rate (P<.05). Reproducibility of TdP induction (three times or more using the same pacing train) remained present for approximately 60 minutes after d-sotalol and was greater than 90% within the same animal over weeks. TdP induction was related to the presence of early afterdepolarizations on the monophasic action potential recordings: five of six in the inducible group versus two of six in the nonresponders. Inducibility could be further increased to 89% when a second bolus of d-sotalol was administered to noninducible dogs. On the other hand, decreasing QT time by faster basic pacing or administration of isoproterenol, or MgSO4 prevented TdP induction. This effect of MgSO4 coincided with the disappearance of early afterdepolarizations. Our animal model shows a high incidence of reproducible acquired TdP arrhythmias, allowing study of the mechanism and treatment of TdP. TdP induction was related to the combination of a slow ventricular rate, the prolongation of QT time, a sudden induced rate change that often required two or more cycle length changes, and the presence of early afterdepolarizations.
Torsade de pointes (TdP) arrhythmias can be congenital or acquired.1 Acquired TdP often occurs in the presence of drugs that prolong ventricular repolarization, such as class Ia and III drugs. By definition, one of the characteristics of drug-induced TdP is the prolongation of the QT(U) duration. Prolongation of the refractory period of cardiac muscle is at the same time the desired antiarrhythmic goal for the prevention of reentrant tachycardias.
Several canine models exist2 3 4 5 6 7 8 9 10 that enable the study of the morphological or mechanistic characteristics of TdP arrhythmias. TdP initiation is often associated with early afterdepolarizations (EADs).2 3 4 5 Still, we felt that the addition of a model that mimics the clinical circumstances of TdP is important for the comparison of the proarrhythmic potential of drugs to initiate TdP and the study of the mechanisms and treatment of TdP. The clinical conditions known to favor acquired TdP are bradycardia, therapeutic drug doses that prolong QT duration, sequences of short/long/short (SLS) intervals, and hypokalemia and/or hypomagnesemia.1
In the dog we developed a highly reproducible TdP model by mimicking the first three conditions by chronic atrioventricular block, administration of 2 mg/kg d-sotalol, and specific pacing modes. To investigate the model further, we increased basic heart rate by pacing, we administered MgSO4 and isoproterenol, and we recorded monophasic action potentials (MAPs) to demonstrate the involvement of EADs. Finally, we used different pacing modes to clarify the relevance of the number of cycle length changes (eg, the SLS sequence) needed to initiate TdP.
The study protocol was approved by the Committee for Experiments on Animals of the University of Limburg, Maastricht, the Netherlands, and conducted in accordance with the guidelines of the American Physiological Society.
Experiments were performed on 18 adult mongrel dogs of either sex having a body weight between 20 and 35 kg (26±5 kg). In preliminary surgery, a right thoracotomy was performed to induce a permanent complete atrioventricular block by injection of 37% formaldehyde into the atrioventricular junction.11 During the same session, pacing electrodes (Bakken Research Center, Medtronic) were inserted epicardially into the basal portion of the right ventricle and the apex of the left ventricle. The wires were exteriorized through the back of the neck.
Six surface ECG leads and one local ECG were continuously registered and stored on an optical disk. All drugs were administered through a cannula in a cephalic vein.
Anesthesia was induced by intramuscular premedication (1 mL/5 kg: 10 mg oxycodon, 1 mg acepromazine, 0.5 mg atropine) and sodium pentobarbital (20 mg/kg IV). Dogs were artificially ventilated through a cuffed endotracheal tube using a mixture of oxygen, nitrous oxide, and halothane (vapor concentration, 0.5%) by a respirator. Ventilation was controlled by continuous reading of the carbon dioxide concentration in the expired air. A thermal mattress was used to maintain adequate body temperature.
Proper care was taken postoperatively, including antibiotics (1000 mg ampicillin) and analgesics (0.1 mg IM buprenorphine).
Stimulation was done preferably from the right ventricular lead with a programmable stimulator having a synchronizing circuit.12 Unipolar stimuli were given using a pulse width of 2 ms and a stimulus strength of twice the diastolic threshold. As indifferent electrode, we used a needle placed through the skin.
Pacing consisted of four different pacing modes, which represent, respectively, one, two, and three cycle length changes. The pacing modes for the one cycle length change consisted of three stimuli (NVS=3) and continuous pacing for 30 seconds. Both were performed with equal interstimulus intervals and shortened from 1200 to 400 ms in 100-ms steps for NVS=3, whereas continuous pacing was started just below the cycle length of the idioventricular rhythm (CL IVR) and reduced to 1200, 900, 600, and 400 ms. The pacing modes with more than one frequency change were a basic train of eight stimuli followed by an extrastimulus (8+1) and an SLS sequence. The extrastimulus was shortened from 350 ms in 10-ms steps until the effective refractory period (VERP) was reached. VERP was defined as the longest stimulus interval that was not followed by a ventricular complex. VERP was determined during IVR and at paced rates similar to the continuous pacing protocol. SLS sequence consisted of (1) four to eight paced beats with an interval of 600 ms followed by a beat with an interval of 1200 ms and finally an extrastimulus (4 to 8×600/1200/extra), (2) 400/800/extra, and/or (3) 2×300/900/extra. The pacing protocol needed 30 to 40 minutes for completion, and the different pacing modes were applied in a random manner.
TdP incidence was related to a specific pacing train and mode. Moreover, because TdP frequently started before the pacing train was completed, induction was also classified to the specific change in cycle length that seemed to be responsible for TdP initiation. When TdP was induced during the first eight paced beats of either continuous pacing or 8+1, it was classified as being induced after one frequency change. A dog was considered inducible when a TdP arrhythmia could be reproducibly induced by the same pacing train at least three times.
Definition of TdP
A TdP arrhythmia was defined as a polymorphic ventricular tachycardia consisting of five beats or more twisting around the baseline having a rate of more than 200 beats per minute that occur in the setting of a prolonged QT(U) duration.1 These ventricular tachycardias either stop spontaneously or degenerate into ventricular fibrillation. TdP was terminated using cardioversion (60 to 70 J) when it lasted longer than 10 seconds. Defibrillation was performed maximally six times per experiment. TdP was differentiated from ventricular fibrillation using the following characteristics: (1) twisting QRS behavior using all six ECG leads, (2) the change in amplitude of the QRS complex in a single ECG lead, and (3) the frequency of the arrhythmia.
At least 2 weeks after the creation of complete atrioventricular block, dogs were anesthetized (see above). A blood sample was taken to measure plasma levels of calcium, sodium, potassium, and magnesium and to establish kidney function by measuring urea and creatinine. Two defibrillation patches connected to a defibrillator were attached at both sides of the chest. Most dogs were tested at different weeks to investigate the reproducibility of their response.
After the baseline pacing protocol, a bolus of d-sotalol (2 mg/kg per 5 minutes) was administered. Thereafter, 5 minutes elapsed before pacing was resumed. The experiments were divided into two groups: inducible and noninducible. In the inducibile group, we distinguished between TdP that occurred spontaneously after d-sotalol (spontaneous TdP) and TdP that was pacing dependent (pacing-dependent TdP). Because the number of cardioversions per experiment was limited, it was often not possible to repeat the complete pacing protocol in the inducible group.
In most of the experiments belonging to the noninducible group, we added another bolus of d-sotalol (2 mg/kg per 5 minutes) and repeated the pacing protocol.
When TdP was inducible, we performed three procedures. First, the inducibility and reproducibility of TdP arrhythmias were tested further using different pacing modes. Also, the length of the period (minutes) was determined during which a single bolus of d-sotalol in combination with pacing was able to induce TdP. Second, the basic ventricular rate was accelerated by pacing or administration of isoproterenol (10 μg/5 minutes IV). Thereafter, the specific pacing train responsible for TdP induction was repeated. Third, MgSO4 (100 mg/kg per 2 minutes) was administered to assess its suppressive effect against spontaneous TdP and to study its preventive effect against pacing-dependent TdP.
Monophasic Action Potentials
A single endocardial MAP recording was made before and after d-sotalol to detect EADs in 12 experiments. Through either the jugularis vein (n=11) or the carotid artery (n=1), MAP catheters (Franz combination catheter, EPT No. 1650) were placed randomly under fluoroscopy.13 MAP was amplified by a DC-coupled differential amplifier, which is provided with a 20-mV calibration pulse. MAP signals were sampled at a rate of 1 KHz. The recording sites were chosen based on stability, quality of the signal, and an amplitude that had to exceed 15 mV.13 The presence of an EAD on the MAP recording was defined as a retardation in repolarization.2 3 4 5
The following parameters were measured in each experiment by two independent observers: CL IVR, QRS duration, and QT(U) duration. QT(U)c duration was calculated according to Bazett’s formula.14
For determination of statistical difference (P<.05), we used ANOVA for comparisons between more than two groups, followed by Bonferroni’s t test when the F value permitted. Paired Student’s t test was applied to compare data between two groups, and χ2 testing was used when the data were presented as percentages. All data are presented as mean±SD or when indicated as mean±SEM.
In 18 dogs, we performed 42 experiments (2.5±1.2 experiments per dog). The mean time interval between the first experiment and atrioventricular block was 9±6 weeks, and the time between experiments was 14±9 days. All dogs had normal electrolytes except for a slightly decreased potassium level of 3.4 mmol/L in one dog receiving diuretics because of heart failure.
Pacing Under Control Conditions
We were not able to induce TdP by pacing under baseline conditions. However in three dogs, we noticed transient episodes of repolarization abnormalities (occurrence of U waves) in the first spontaneous beats directly after pacing.
Electrophysiological Effects of d-Sotalol
Administration of d-sotalol resulted in a marked increase in the CL IVR from 1475±460 to 1730±570 ms (P<.01) and QT duration from 390±65 to 480±85 ms (P<.01), often occurring in the presence of U waves.
In a subset of dogs (n=8), we investigated the duration of these electrophysiological changes (Fig 1⇓). The increases in CL IVR and QT duration diminished in time but remained present during the investigational period of 50 minutes. Also, QTc showed a similar behavior. No changes were observed in QRS duration.
VERP during spontaneous IVR increased from 265±60 to 325±65 ms (P<.05). The effect of d-sotalol on VERP was reverse use-dependent: the relative increase of VERP was greater at slower heart rates (+23%) than at faster ones (+2%, P≤.01).
Spontaneous TdP Arrhythmias After d-Sotalol
After d-sotalol administration, we noticed four spontaneous appearances of a TdP arrhythmia (4 of 42=10%; Fig 2⇓, panel 3). These episodes were repetitive and often needed interventions, such as cardioversion and backup pacing at faster rates to suppress the arrhythmia. More frequently (50%), ectopic beats with different configurations often arising from the end of the T(U) wave were observed (Fig 2⇓, panel 2).
Inducibility of TdP Arrhythmias After d-Sotalol
After the first bolus of d-sotalol, pacing resulted in reproducible TdP in 22 of 42 experiments (52%, P<.01 versus spontaneous), which appeared in 12 of the 18 dogs. An example is presented in Fig 3⇓. In all dogs with spontaneous TdP, pacing also resulted in TdP induction. Inducibility remained present with the use of similar pacing modes for at least 60 minutes in the 3 dogs in which this was tested.
Inducibility was related to CL IVR and QT(U) duration (Table 1⇓). In the noninducible group (n=20), CL IVR and QT time were significantly shorter, although the relative increase induced by d-sotalol was comparable for both parameters in both groups. Correction for QRS duration (JT interval) revealed no differences in the quantification of the parameter of repolarization.
In only two dogs, one inducible and one noninducible dog, the response in a subsequent experiment changed. All other dogs showed similar responses, either induction or lack of induction. Therefore, the reproducibility is quite high (38 of 42=90%).
Registration of EADs by MAP Recordings
EADs were registered in 7 of 12 experiments (Figs 4 through 6⇓⇓⇓). In half of the MAP experiments, TdP could be induced. The occurrence of ectopic beats was related to the presence of EADs in 4 of 5 experiments (Fig 4⇓). EADs were seen in 5 of 6 inducible and in 2 of 6 noninducible experiments. Fig 5⇓ presents an example of a TdP induction by pacing that is initiated by EAD-dependent triggered beats. Often, EAD amplitude was pronounced during the pacing train.
Effect of Pacing Mode on Inducibility of TdP Arrhythmias
In a subgroup of subsequent, inducible experiments (n=11), we analyzed the induction of 55 episodes of TdP in relation to the pacing mode. Defibrillation was performed in 21 cases, whereas spontaneous termination occurred in 34 TdPs. Induction occurred either during the pacing train or directly thereafter (Figs 3⇑ and 5⇑). In the case of continuous pacing, induction was always seen in the beginning of the pacing train. The incidence of TdP in relation to pacing mode is presented in Table 2⇓. It can be seen that the 8+1 and SLS modes are equally effective and possess a greater ability to induce TdP than the other two modes (P<.05), whereas NVS=3 is more successful than continuous pacing (P<.05).
Induction was related to the number of cycle length changes: Half of the inductions occurred after one change in cycle length, ie, starting during or directly after protocols 1 or 2 or during the first frequency change of protocols 3 and 4 (Table 2⇑, presented as first change). However, inducibility was only possible in the other half of the experiments by at least two frequency changes, as illustrated in Figs 3⇑ and 5⇑.
Inducibility in the Noninducible Group After a Second Bolus of d-Sotalol
After administration of a second bolus of d-sotalol in 14 of the 20 noninducible experiments, inducibility increased to a total of 32 of 36 (89%, P<.01 versus first bolus of d-sotalol). This was accompanied by increases in CL IVR from 1340±190 to 1690±500 ms (P≤.05) and QT(U) duration from 450±60 to 490±55 ms (P≤.05).
Prevention of TdP
Increasing Heart Rate
The basic heart rate was increased by continuous pacing to ±65% of spontaneous CL IVR (n=3). Inducibility was completely prevented but returned when continuous pacing was stopped. Administration of isoproterenol (n=2) decreased CL IVR to ±70% and resulted in complete prevention of inducibility. Thirty minutes after isoproterenol, CL IVR and QT duration returned to their original values, and TdP arrhythmias became inducible again.
Administration of MgSO4
MgSO4 was administered to six pacing-dependent TdP experiments. Two of these dogs also showed spontaneous TdP. In two experiments, a MAP signal was registered. At the time of MgSO4 administration, spontaneous ectopic beats were seen in two experiments (Fig 6⇑, panel 1). MgSO4 suppressed the ectopic beats (panels 2 and 3) and prevented the induction of TdP in all dogs tested (panel 4). This was accompanied by the disappearance of the subthreshold EADs on the MAP recordings. Also, the pacing protocol was completed without the occurrence of ectopic beats. Electrophysiologically, MgSO4 shortened the QT time (Table 3⇓). The QTc duration decreased even more because of the fact that MgSO4 increased the CL IVR.
In two experiments, reinduction of TdP was seen after the effect of MgSO4 had disappeared.
When we summarize the results in this animal model, it becomes obvious that to reproducibly induce TdP arrhythmias, the CL IVR and QT time have to be sufficiently long at the start of the experiment. Shortening of the QT duration by either an increase in heart rate by pacing or administration of isoproterenol or MgSO4 will prevent induction, whereas a second bolus of d-sotalol, which further lengthens QT time and CL IVR, will tend to make the nonsusceptible animals inducible. Both the spontaneous and pacing-dependent TdP rely for their induction on the presence of EADs. Finally, a sudden rate change, often consisting of two or more cycle length changes, is needed to induce TdP under the above conditions.
In recent years, the interest in animal models of TdP arrhythmias has increased considerably.1 2 3 4 5 6 7 8 9 10 Attention has focused on (1) the ionic mechanisms underlying the induction and perpetuation of TdP arrhythmias, (2) screening for parameters and/or circumstances that could predict possible proarrhythmic effects of antiarrhythmic drugs, especially new class III drugs, and (3) TdP prevention by specific interventions, including changes of the activity of the autonomic nervous system and specific antiarrhythmic drugs, including new drugs such as Ik activators.2 3 4 5 6 7 8 9 10 15 16 17 18 19
The intervention most widely used to induce TdP arrhythmias in animals is cesium administration.3 5 7 9 18 In anesthetized dogs with chronic atrioventricular block, we developed a TdP model that shows a high incidence of TdP inducibility together with a high reproducibility. Depending on whether one or two d-sotalol doses are used, the TdP incidence is 52% or 89%, respectively. Reproducibility of TdP induction was present during an experiment for approximately 60 minutes and in repeated experiments over weeks in 90% of the attempts. In most experiments, inducibility depended on abrupt changes in the cycle length applied by pacing.
During the course of our experiments, an article was published that reported a similar approach,10 although important differences exist between the two methodologies. Weissenburger et al10 studied spontaneous induction of TdP arrhythmias in conscious dogs with chronic atrioventricular block. To achieve that, they needed higher doses of the antiarrhythmic drugs in the presence of hypokalemia, β-blockade, or both.
The electrophysiological effects of d-sotalol are in agreement with a therapeutic dose of this drug, and they do represent clinically achieved values.20 The observed reverse use-dependency of VERP has also been described as a property of d-sotalol.21 22 Next to the increases in QT time and VERP, d-sotalol slowed the heart rate (increase in CL IVR). Similar findings for d-sotalol but also other class III agents have recently been reported for dogs in sinus rhythm.23 24 This negative chronotropic effect of class III agents has been attributed almost entirely to prolonged repolarization.
Relevance of the Number of Cycle Length Changes
Spontaneous, drug-induced TdP arrhythmias often show a specific initiating sequence related to the occurrence of spontaneous ectopic beats: SLS intervals that represent three abrupt cycle length changes in a short period of time.1 6 10 15 19 Using a similar pacing mode, we tried to imitate this sequence. Because there is controversy in the literature about the specificity of this sequence,1 25 26 27 we also used other pacing modes to study their possible relevance. Two pacing modes were clearly superior in initiating TdP arrhythmias in our model: SLS and 8+1. The equal incidence of the 8+1 pacing mode is especially of interest because this is worldwide the most often used pacing mode during electrophysiological studies.
Using the MAP catheter, we saw the occurrence of triggered EADs directly after the start of pacing (Fig 5⇑) when the rate was relatively fast and the action potential duration relatively short compared with the IVR. This observation indicates that inadequate adaptation of the duration of the action potential to abrupt changes in cycle length is important for TdP induction. This has also been suggested by others.28 Support for this line of thinking comes from the results of the two pacing modes with fixed interstimulus intervals but differences in duration (NVS=3 versus continuous pacing). The higher inducibility of NVS=3 underscores the importance of an abrupt frequency change rather than the duration of pacing. Additional confirmation comes from the observations that (1) in the case of continuous pacing, TdP always started within the first 10 beats and (2) one cycle length change already induced TdP in half of the experiments. However, when a second or a third interval change is added, inducibility is doubled.
The fact that prolonged pacing is often used to prevent induction of spontaneous drug-induced TdP is another argument for stressing the importance of the abruptness of the cycle length change.
These data should be interpreted with some caution. Because we allowed ourselves only a small number of cardioversions per experiment, we could not always study all pacing modes in the same experiment. We tried to correct this by using the other pacing modes in a subsequent experiment with the same animal. Still, we cannot answer questions concerning cycle length dependence (interstimulus interval) and/or possible interference of different pacing modes on TdP inducibility.
The most effective treatments for clinical acquired TdP are rapid ventricular pacing or isoproterenol or magnesium administration.1 These treatments were all tested and considered effective in preventing TdP arrhythmias.
Several studies have addressed the involvement of triggered activity resulting from EADs in the initiation of TdP.2 3 4 5 15 18 19 Although the contribution of EADs is still not completely resolved, evidence is becoming more convincing that EADs play an important role in the induction of acquired TdP. As seen in the present study, the presence of EADs seems to be related to spontaneous as well as pacing-dependent TdP. However, it is difficult to demonstrate the causal relation between the presence of EADs, triggered beats elicited by EADs, and the actual TdP. The continuous shift in the site of origin of EADs and the elicited beats together with the speed of the arrhythmias complicate the MAP registrations. This technique itself also has some technical difficulties, such as stability and the possibility of registering artifacts. We relied on some good fortune to be at the site of origin of the EAD (Figs 4 through 6⇑⇑⇑). Still, we are convinced that these recordings in a few experiments are representative for the model. We also believe that the experiments with MgSO4 demonstrate that the EADs are not artifacts. Therefore, we think that TdP initiation is EAD dependent. Whether the same mechanism is also responsible for the perpetuation of the arrhythmia has to be established.
Relevance of This Animal Model
A relatively low incidence of spontaneous TdP arrhythmias occurring after d-sotalol administration was observed in these experiments. We defined TdP as five beats or more. More frequently (±50%), ectopic beats occurred arising from the TU wave that seemed to be dependent on EADs. These beats occurred as singles, doubles, or triplets. Reproducible initiation of TdP arrhythmias required cycle length changes that were induced by pacing. Still, we believe that this pacing-dependent TdP can be used to study the mechanism and treatment of spontaneous acquired TdP because (1) in dogs showing spontaneous TdP, pacing was also able to induce TdP reproducibly, (2) both forms of TdP seem to be dependent for their initiation on the presence of EADs (Figs 4⇑ and 5⇑), (3) pacing-dependent TdP showed the same specific ECG characteristics as spontaneous TdP and often terminated spontaneously, and (4) both forms of TdP could be suppressed by MgSO4 (Fig 6⇑), levcromakalim,17 or flunarizine29 and by increasing heart rate.
In this model, we tested the dogs at a relatively long period after application of atrioventricular block (9±6 weeks). The bradycardia-related volume overload and subsequent biventricular hypertrophy (personal observations) could be of importance for the TdP inducibility. Currently, we are documenting these changes over time, both echocardiographically and electrophysiologically, and we are testing the relevance of these changes for TdP inducibility. The observation by Strauss et al21 that application of acute atrioventricular block together with a similar dose of d-sotalol did not result in TdP arrhythmias indicates that additional changes have to be present.
We are currently testing other drugs that prolong action potential duration. Our results indicate that this animal model can be used to compare the proarrhythmic potential of different drugs. Whether this model is suitable for predicting the occurrence of TdP in patients treated with these drugs has to be determined.
In conclusion, this animal model shows a high incidence of reproducible, acquired TdP arrhythmias, allowing the study of the mechanism and treatment of these triggered tachycardias.
This study was supported by a grant from the Dutch Heart Foundation (No. 91.104). The technical assistance of H.D.M. Leunissen and J. van der Zande in performing the experiments and B. van der Steld, MS, for the development of the hardware and software allowing data collection and analysis is greatly appreciated. Also the authors would like to acknowledge the help of the Bakken Research Center (Medtronic), Maastricht, the Netherlands, in providing electrodes.
- Received May 10, 1994.
- Accepted August 31, 1994.
- Copyright © 1995 by American Heart Association
El-Sherif N, Zeiler RH, Craelius W, Gough WB, Henkin R. QTU prolongation and polymorphic ventricular tachyarrhythmias due to bradycardia dependent early after-depolarizations. Circ Res. 1988;63:286-305.
Ben David J, Zipes DP. Differential response to right and left ansae subclaviae stimulation of early afterdepolarizations and ventricular tachycardia induced by cesium in dogs. Circulation. 1988;78:1241-1250.
Hanich RF, Levine JH, Spear JF, Moore EN. Autonomic modulation of ventricular arrhythmia in cesium chloride induced long QT syndrome. Circulation. 1988;77:1149-1161.
Brachmann J, Scherlag BJ, Rosenshtraukh LV, Lazzara R. Bradycardia-dependent triggered activity: relevance to drug-induced multiform ventricular tachycardia. Circulation. 1983;68:846-856.
Bardy GH, Ungerleider RM, Smith WM, Ideker RE. A mechanism of torsade de pointes in a canine model. Circulation. 1983;67:52-59.
Leichter D, Danilo P Jr, Boyden P, Rosen TS, Rosen MR. A canine model of torsade de pointes. PACE. 1988;11:2235-2245.
Steiner CH, Kovalik ATHW. A simple technique for production of chronic complete heart block in dogs. J Appl Physiol. 1968;25:631-632.
Steld AVD, Dassen W, Gorgels APM, Beekman HDM, Wellens HJJ. Flexible multiprocessor system to support electrophysiological investigation in animals. Comput Cardiol. 1984;11:525-528.
Bazett HC. An analysis of the time relations of electrocardiograms. Heart. 1920;7:353-368.
Carlsson L, Abrahamsson C, Drews L, Duker G. Anti-arrhythmic effects of potassium channel openers in rhythm abnormalities related to delayed repolarization. Circulation. 1992;85:1491-1500.
Fish FA, Prahesh C, Roden DM. Suppression of repolarization-related arrhythmias in vitro and in vivo by low dose potassium channel activators. Circulation. 1990;82:1362-1369.
Bardy DS, Inoue H, Kaseda S, Ben-David J, Zipes DP. Magnesium suppression of early afterdepolarizations and ventricular tachyarrhythmias induced by cesium in dogs. Circulation. 1988;77:1395-1402.
Buchanan LV, Kabell G, Brunden MN, Gibson JK. Comparative assessment of ibutilide, d-sotalol, clofilium, E-4031 and UK-68,798 in a rabbit model of pro-arrhythmia. J Cardiovasc Pharmacol. 1993;220:540-549.
Antonaccio MJ, Gomoll AW. Sotalol: pharmacological and anti-arrhythmic effects. Cardiovasc Drug Rev. 1988;6:239-263.
Strauss HC, Bigger JT, Hoffman BF. Electrophysiological and beta receptor blocking effects of MJ 1999 on dog and rabbit cardiac tissue. Circ Res. 1970;24:661-671.
Hondeghem LM, Snijders DJ. Class III antiarrhythmic agents have a lot of potential but a long way to go. Circulation. 1990;81:686-690.
Horowitz L, Greenspan AM, Spielman SR, Josephson ME. Torsade de pointes: electrophysiologic studies in patients without transient pharmacologic or metabolic abnormalities. Circulation. 1981;63:1120-1128.
Vos MA, Verduyn SC, Lipcsei GE, van der Zande J, Leunissen HDM, Gorgels APM. Induction of acquired bradycardia dependent torsade de pointes arrhythmias is prevented by flunarizine: a comparison with MgSO4. Circulation. 1992;86(suppl I):I-560. Abstract.